INFORMATION TO USERS
This was produced from a cop y o f a docum ent sent to us for microfilming. While the m ost advanced technological means to photograph and reproduce this docum ent have been used, the quality is heavily dependent upon the quality o f the material submitted. The following explanation o f techniques is provided to help you understand markings or notations which may appear on this reproduction. 1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)” . If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you o f com plete continuity. 2. When an image on the film ir obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because o f m ovem ent during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should n ot have been filmed, you will find a good image o f the page in the adjacent frame. 3. When a map, drawing or chart, etc., is part o f the material being photo graphed the photographer has follow ed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand corner o f a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until com plete. 4. For any illustrations that cannot .be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department. 5. Some pages in any docum ent may have indistinct print. In all cases we have film ed the best available copy.
University Microfilms Internationa! 300 N. ZEEB ROAD. ANN ARBOR, Ml 48106 18 BEDFORD ROW, LONDON WC1R 4EJ, ENGLAND
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8012253
C u r r e y , Br u c e
MAPPING AREAS LIABLE TO FAMINE IN BANGLADESH
University o f Hawaii
University Microfilms International
PH.D.
300 N. Zeeb Road, Ann Arbor, M I 48106
1979
18 Bedford Row, London WC1R 4EJ, England
Copyright 1980 by Currey, Bruce All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PLEASE NOTE: In a l l cases th is m aterial has been film ed in the best possible way from the a v a ila b le copy. Problems encountered w ith th is document have been id e n t if ie d here w ith a check mark .
1.
Glossy photographs
2.
Colored illu s t r a tio n s
3.
Photographs w ith dark background ________
4.
I llu s t r a tio n s are poor copy
5.
Dr i n t shows through as there is te x t on both sides o f page _________
^
W
6.
In d is t in c t , broken or small p r in t on several p ag es____________ throughout
T ig h tly bound copy w ith p r in t lo s t in spine _________ Computer p rin to u t pages w ith in d is tin c t p r in t _ _ _ _ _
9.
Page(s) lacking when m aterial received , and not a v a ila b le from school or author ________
10.
Page(s) ________ seem to be missing in numbering only as te x t follow s ______
_
7. 8.
11.
Poor carbon copy ________
12.
Not o r ig in a l copy, several pages w ith b lurred
13.
Appendix pages are poor copy _________
14.
O rig in al copy w ith l ig h t type _________
type __
15.
Curling and w rinkled pages
16.
O th e r _____________________________________________________________________
v /
University M ooriim s Internationa} 200 i\ ZEES RD.. ANN AR 30P Ml ^81 06'3 13) 761-4700
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MAPPING AREAS LIABLE TO FAMINE IN BANGLADESH
A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN GEOGRAPHY DECEMBER 1979
By Bruce Currey
Dissertation Committee: Gary A. Fuller, Chairman R. NTarwick Armstrong Brian J. Murton John M. Street Everett A. Wingert Emmanuel Voulgaropoulos
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
We certify that we have read this dissertation and that in our opinion it is satisfactory in scope and quality as a dissertation for the degree of Doctor of Philosophy in Geography.
DISSERTATION COMMITTEE
''Chairman
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
This dissertation is the academic anchor of my project for the Ministry of Relief and Rehabilitation, Government of Bangladesh.
The
project was supported by a USAID, Dacca Project Studies Grant 388-0008. It was implemented through the Johns Hopkins University International Center for Medical Research.
The ICMR supported me as a Research
Associate during most of my stay in Bangladesh 1974-1976. It has been said that copying one person's ideas is plagiarism and copying two persons' ideas is research.
If that were indeed the
scale, then this must be excellent research for I have built it upon the ideas of many people from three continents. This project has been predominantly a Bengali project.
The 180
people who have made the greatest contribution are fully acknowledged on the Map of Areas Liable to Famine in Bangladesh, submitted with this report.
To them I wish to add an additional thank you for bearing with
the extensive interviews and group discussion sessions. Although some seminal ideas for this project were sown in Scotland by A. J. Crosbie, D. R. Macgregor, J. G. Mitchell and W. J. Watson, they flourished in Hawaii under a Ph.D. committee that gave them the chance to bloom.
My chairman, Dr. G. A. Fuller, and committee members R. W.
Armstrong, B. J. Murton, J. M. Street, E. A. Wingert, and E. Voulgaropoulos have throughout formed the fulchrum balancing my Scot's stubbornness and the requirements of an academic discipline.
Many faculty
not on the committee have supported me here in Hawaii, particulary J. H. Chang, M. Chapman, D. W. Fryer, R. J. Fuchs, D. Kornhauser, F. R. Pitts,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv and F. Tilton.
Perhaps as great an emphasis in moulding a dissertation
comes from arguments with one's fellow Ph.D. students M. Ali, A. J. Hampson, and P. Wright. Before implementing the project in Bangladesh I was greatly helped by S. Allison, F. B. Bang, J. Coron, W. U. Drewes, W. H. Mosley, D. W. Parrack, W. Reinke, R. Revelle, the United
D. Sinton, R. Tabors and C. Taylor in
States; S. Kilde, M. F.Lechat, R. Pierpont and F. Soliman
in Europe and R. E. Carlaw in Nepal. The study prospered in Bangladesh with the help of K.M.A. Aziz, L. Chen, A. Chowdhury, R. Gilman, P. 0. Hansen, K. A. Huque, N. Islam, R. Langsten, T. Oliver, J. O'Rorke, M. M. Rahman, S. d'Souza, S. Street, A.
Wahed and G. Warner.
All the hard work in Bangladesh would not
have been possible without K. Islam, Md. Shafique, Md. Shahidullah and M. A. Wahed. Since and typing
Bangladesh the labor has been carried on through the editing stage by F. Jackson, B. Downs, C. Tainsh, and K. Takeshita
and her wonderful staff. by F. Hellinger.
The final typing was designed and implemented
Three people have asked to remain anonymous but they
know they have kept me going when others would have had me stop. Finally, if I have been inspired by any one particular person it is Hugh Brammer.
The failings of this dissertation are the results of
my not having listened to his suggestions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
This dissertation is concerned with mapping areas liable to famine in Bangladesh as a means to changing famine remedial policies from their emphasis on relief to prevention. present famine relief policies.
It discusses the problems of the It argues the need for a better underÂ
standing of the causal mechanisms underlying famine vulnerability in the world and particularly in Bangladesh.
Because of the lack of reliable
data during times of famine in Bangladesh, the dissertation uses a method called the 'Delbecq-Delphi Process’ to gain a consensus view on the areas most liable to famine from village elders, senior retired government officers, aid distributors and technical experts.
The Delbecq-
Delphi Process map is evaluated against a series of mental maps, news paper reports, the places of origin of urban bastee dwellers, and against a famine field survey carried out during the 1974-75 famine.
The final
map of areas liable to famine in Bangladesh illustrates those areas very liable to famine and suggests the structural factors underlying their susceptibility.
The map also includes remedial policies tailored to the
particular components of vulnerability in each of the areas liable to famine.
The policy makers themselves selected and weighted the components
of famine vulnerability on the map and they also suggested the remedial policies for inclusion in the m ap.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
ACKNOWLEDGMENTS ................................................... ABSTRACT
iii
.........................................................
v
LIST OF T A B L E S ...................................................
ix
LIST OF F I G U R E S ...................................................
x
CHAPTER I
1
THE DEFINITION OF F A M I N E ............................. The Famine S y n d r o m e ................................. The Manifestation of Famine in Urban A r e a s .......... The Recent Consolidation of the Symptomatic D e f i n i t i o n ....................................... The Problems of Famine Amelioration by Food Relief . . Putting Relief in a Development Context ............ The Geographical Contribution to Mapping Areas Liable to F a m i n e .................................
3 12
WORLD EXAMPLES OF FAMINE VULNERABILITY...............
43
Famine Vulnerability ................................. The Characteristics of the Community................ The Scale of E n q u i r y ................................. The Evolution of a F a m i n e .......................... Examples of Famine Vulnerability in Different Countries .......................... Ireland and the Great Famine of 1846-51 China and the Shantung Famine of 1925 Ethiopia and the Famine of 1973-74 .................. India and the Bihar Famine of 1967
44 45 49 50
CHAPTER III THE FIELD AREA IN BANGLADESH.........................
85
CHAPTER II
18 21 32 35
51 52 60 67 77
The Unstable Landscape .............................. 85 The Meshing of the S e a s o n s.......................... 89 P o v e r t y ............................................. 94 A Population at R i s k ................................. 98 Faith, Hope and C h a r i t y ............................... 106 Natural Disasters ................................... 113 The Essence of Bangladesh's Famine Vulnerability . . . 123 Examples of Famine Vulnerability within Bangladesh . . 126 Areas Liable to Famine in Bangladesh . ...............130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii CHAPTER IV
THE LACK OF COMPLETE D A T A ............................. 137 Statistical Statistical Statistical Data During
CHAPTER V
METHODOLOGY:
Data in British India .................. 13/ Data on B e n g a l ............................ 139 Data in Bangladesh ........................ 140 Famine Conditions . . '.....................144 THE DELBECQ-DELPHI PROCESS ............
153
The Delbecq Technique .............................. 156 The Delphi Technique ................................ 157 The Delbecq-Delphi Process .......................... 163 Introduction ..................................... 163 Elaboration......................................... 166 Stage 1. Open-ended Questionnaire ................ 171 Mental M a p s ......................................... 174 Stage 2. Editing by Technical E x p e r t s ..............183 Stage 3. Delbecq S e s sions .......................... 185 Stage 4. Delphi S e s s i o n s ........................... 189 Compiling the Map of Areas Liable to F a m i n e .......... 194 Combining the Components of Famine Liability ........ 198 CHAPTER VI
AREAS LIABLE TO FAMINE IN BANGLADESH ................
201
Comparison of the Delbecq-Delphi Map with the Three Mental M a p s ................................... 204 Consistency......................................... 204 Specificity......................................... 204 Disagreement ..................................... 205 Comparison of the Delbecq-Delphi Map with the Newspaper Reports of the 1974-75 Famine .......... 207 Comparison of the Delbecq-Delphi Map with the Map of Origin of Bastee Dwellers of Dacca City . . . . 209 Comparison of the Delbecq-Delphi Map with the 211 Famine Survey of Ten Thanas, 1975 ................ Gathering Data from a Famine Survey . . . . . . . . 211 Differentiating Thanas Liable to Famine .......... 228 Comparing the Indices of Famine Symptoms ........ 242 Final Comparison of the Areal Intensity of Famine Symptoms with the Delbecq-Delphi M a p ...............247 Conclusion............................................. 247 CHAPTER VII
FAMINE REMEDIAL POLICIES IN BANGLADESH ..............
249
National Remedial Policies .......................... 252 252 Revolution of Consciousness ...................... Price Stabilization................................. 252 253 Reorientation of Agricultural Education .......... Improved Coordination of Famine Prevention and Preparedness Planning ................ . . . . 254
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
viii Regional Remedial Policies .......................... 254 F l o o d i n g ........................................... 255 D r o u g h t ............................................. 256 257 Population Pressure .............................. Food D e f i c i t ....................................... 257 Lack of Alternative Employment .................. 258 Low Crop Y i e l d s ..................................... 259 Poor Land Transport ................................. 259 River E r o s i o n ....................................... 260 Cyclones , . . . . 261 Maldistribution of Agricultural Inputs .......... 262 CHAPTER VIII REFLECTIONS, SUMMARY AND C L A I R V O Y A N C E .................265 APPENDICES A. B. C. D. E. F. G. H. BIBLIOGRAPHY
.
Murubis' Questionnaire ........................... 278 Retired Government Officers Questionnaire . . . . 283 Aid Donors' Q u e s t i o n n a i r e ......................... 293 Technical Experts' Questionnaire ................. 303 Example of Delphi Session C a r d ..................... 313 Delbecq-Delphi Session Invitation ............... 314 Famine Field Survey Questionnaire ............... 315 Map of Areas Liable to F a m i n e ..................... 343 ...........
320
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table 1
Page Causes and Effects of the Famine Syndrome Suggested by Four G r o u p s .........................................
4
2
The Famine Syndrome in Bangladesh
.....................
6
3
Perceived Causes of the Bihar Famine ...................
79
4
Estimates of Flood Damage
.............................
143
5
Differences Between the Record and Reality of a Relief Operation .......................................
149
6
A Comparison of the Lists of Factors Considered to Predispose Certain Areas of Bangladesh to Famine at Different Stages in the Delecq-Delphi Process ........
181
7
Background to the 1974-1975 Famine in Eastern Rangpur. .
217
8
Percentage of Households Selling Land Over Famine P e r i o d .................................................
233
9
Famine Death Rates by Occupation of Household Head . . .
241
10
Mortality Rates by Amount of Land Ownership Before Famine of Household H e a d ...............................
242
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Figure
Page A Famine Landscape, by Zainul Abedin ..................
xii
1
Examples of the^ Demographic Effects of the 1974-75 Famine on a Bangladesh Village ........................
2
Outmigration from a Bangladesh Village .................
11
3
Geographical and Temporal Distribution of Street Deaths in Dacca C i t y
14
4
Durvickha (Shortage of Alms) Famine
...................
16
5
Sequences of Delay in the Supply of Foreign Food R e l i e f .................................................
31
6
Famine Vulnerability ...................................
46
7
Regional Demography of the Irish Famine 1846-1851
...
56
8
China— Landscape of Famine .............................
68
9
The Impact of Drought on an Ethiopian F a r m e r ...........
73
. . .
8
10
Bangladesh Reference Map
.............................
86
11
Bangladesh: The Delicate Balance of Man, Land, Season and F o o d .......................................
91
12
Trends in Wholesale Price of Coarse Rice and Real Wages of Agricultural L a b o r e r s .........................
97
13
Bangladesh— Areas of High Population Density, 1974 . . .
101
14
Some Indications of a Population at Risk 1968-1975 . . .
104
15
The Mahasthan T a b l e t ...................................
112
16
Bangladesh: The Historical Record of Natural Calamities, Scarcities and Famines, 17 Districts Between 1757 and 1970
116
17
Bengal Famine C o d e .....................................
132
18
The Delbecq-Delphi Process .............................
155
19
Graphic Representation of the Delphi Rationale
159
.. ..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi Figure
Page
20
Bangladesh: Areas Liable to Famine— Retired Government Officers' Mental Map ........................
176
21
Bangladesh: Areas Liable to Famine— Aid Distributors' Mental Map ..............................
177
22
Bangladesh: Areas Liable to Famine— Technical Experts' Mental M a p .....................................
178
23
Determination of Group S i z e .............................
188
24
Analysis of 3 Delphi S e s s i o n s ...........................
193
25
Evaluation of the Delbecq-Delphi M a p ...................
202
26
Normal Agricultural Calendar (top) and The Brahmaputra River Hydrographs 1955, 1973, 1974
213
27
Photomap of River Erosion ...............................
215
28
Comparison of 1974 Monthly Rainfall with Average Monthly Rainfall .......................................
216
29
Market Price of Rice by M o n t h ...........................
218
30
Field M e t h o d s ...........................................
221
31
Seasonal Sales of Land Over 2 Years in the Bangladesh Thanas of Sunderganj and Rowmari ......................
232
32
Migration of Households from Settlements during Floods in Bangladesh ...................................
239
33
Indices of Famine Symptoms for the 10 Thana Survey
243
. . .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A Famine Landscape, by Zainul A b e d i n (courtesy of the Abedin Family, Dacca)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I THE DEFINITION OF FAMINE
"Gacher gora kete mathai pani dhala" We are watering the tree, But cutting the roots. (Bengali proverb)
It is doubtful that even with a sufficient budget such activities could have taken place in the absence of the trauma created by the famine. (Alan Berg, 1973)
What is a famine? for its amelioration.
Its definition will affect the choice of policies It is a syndrome, a community crisis with many
symptoms and many causes.
The set of famine symptoms is remarkably
similar from one stricken area to another.
Moreover, they are extremely
obvious whenever they occur, and give rise to strong emotions on the part of the urban dweller, the humanist and the media.
The causes of
famine on the other hand, vary considerably from one area to another and often remain hidden to all but the most careful observer.
Recent research
has concentrated on quantifying the more obvious symptoms, such as death by starvation, rather than on examining the more protracted, but less tangible, causes of famine.
Paradoxically it is the results of research
on famine symptoms that are being recognized by decision makers at a time when symptomatic cures in the form of famine relief are being increasingly criticized.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 Research into the multiple causes of famine still remains dis足 jointed and poorly quantified, with few studies at the regional scale. Some decision makers have argued that funds raised to relieve suffering should be channelled towards famine prevention by treating the causes rather than the symptoms.
Such a change of policy has been hampered
by the lack of emotional support for diverting funds away from relief projects and by the lack of substantive information on famine causes. If these causes of famine can be isolated and identified, and at the same time it can be suggested how relief measures can lessen the causes of famine, then these emotional obstacles should disappear.
It
is the goal of this dissertation to demonstrate that in Bangladesh it is possible to map certain areas that are more liable to famine than other areas.
Once it is possible to delimit certain areas as being more
liable to famine then long-term prevention measures can be prepared in advance of a famine crisis occurring in those areas. The relief of immediate suffering and the prevention of possible future famines need not be incompatible.
Emotionally stimulated funding
and the opportunity for change during the community crisis may act as catalysts for famine prevention and for rural development.
The funding
may also provide immediate aid to relief suffering, if appropriate contingency plans are prepared prior to the outbreak of famine symptoms. This dissertation argues that despite the limited research and un足 resolved difficulties in defining famine vulnerability (Chapters II and III), and despite the inadequate data for famine prediction in Bangladesh (Chapter IV), it is nevertheless possible to delimit the areas liable to famine (Chapter V ) .
The accuracy of these findings
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 can be evaluated (Chapter VI) and to the extent they prove accurate it is then possible to compile preparedness plans to mesh with the par足 ticular environmental constraints which predispose certain areas to famine (Chapter VII).
The Famine Syndrome In 1947 A. E. Taylor highlighted the difficulty of giving an elegant definition for famine:
"Famine is like insanity, hard to define,
but glaring enough when recognised." be viewed as:
In this dissertation famine will
a sudden reduction in the capacity of the community to
sustain its poorer members. This syndrome in the community is a very different phenomenon from starvation which is a physiological change affecting individual people (Cahill, 1978). the community.
Famine causes such a rapid change in the whole life of There are not only physiological changes but also social,
behavioral, and psychological changes.
Thus when talking of relation足
ships between famine and disease or famine and fertility we are talking about much broader relationships than when we talk about starvation and disease or starvation and fertility. Some of the many causes and effects of the famine syndrome were suggested by a poll of four groups, each of 45 members, with experience of past famines in Bangladesh.
These same four groups, composed of
village elders, retired government officers, aid distributors and technical experts, all suggested the critical points at which a scarcity (Akal) became a famine (Durvickha) (Table 1). By sifting through these responses and by integrating them both with written accounts of past famines in Bengal and with my own
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1.
Causes and Effects of the Famine Syndrome suggested by Four Groups3
Village Elders
Aid Distributors
Technical Experts
2. Migration to urban 3. After a natural calamity superim足 posed on a scarcity . Thoughts are irrational and only on food
4. People eat leaves, grass, seeds and bonn kochu
5. Increase in the number and timing of beggars
6. People eat leaves, grass, seeds and bonn kochu
6. Increase in begging
o work for anyone 8. A political decisic
10. People eat leaves, grass, seeds and _____bonn kochu______
5 experience during the 1974-75 famine in Bangladesh, I have suggested a conceptual model of the famine syndrome in Bangladesh (Table 2). This syndrome is initiated by a series of assaults which lower a community’s ability to withstand further assault.
Many famine symptoms are suddenly
brought into focus by the triggering effect of a natural or man-made disaster or a series of disasters.
The disruption of agricultural proÂ
duction and storage may be real or speculated. of both. labor.
It may be a combination
There is a rapid decline in the demand for paid agricultural A sudden divergence occurs between the purchasing power of the
community and the price of food.
The latter may rise, but famine
symptoms can occur without any increase in the market price of food (Abed, 1976).
Also a rapid increase in the market price of food need
not imply a famine:
as long ago as 1875 Sir William Wilson Hunter
pointed out that The mere doubling and trebling of prices of rice is not indicative of famine: merely of profiteering by the misfortunes of other districts. (Hunter, 1875) Members of the community who cannot afford to purchase their food supply may either increase their purchasing power by borrowing, begging, stealing, or else by selling or mortgaging their own possessions in a buyer's market.
They may opt to change their diet.
The new diet may
provide less nutrition which, when combined with disease, may increase mortality.
Some community members may opt to leave the community in the
expectation, perhaps unjustified, that employment and relief are avail able in the larger towns and cities.
There, the abnormal concentrations
of often malnourished people, crowded together in unsanitary slums
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
IS I!
iiil!
11fIf11Jj II:!1 _
M
i -i i!i ifii i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
(bustees) , offers the potential for epidemics and thus may further increase mortality. As an example of this syndrome, Figure 1 describes the occurrence of famine in part of Bangladesh in 1974-75.
In a country where rice is
the staple food, the official average price of coarse-grained rice had been rising steadily from a figure of 70 takas per maund (U.S. $158.00 per ton) in 1972. upswing.
In August 1974 the average price of rice took a sudden
Although official figures and aggregated prices undoubtedly
dampened actual bazaar prices, the national average price of rice tripled in taka terms between October 1972 and October 1974, to 210 takas per maund (U.S. $780.00 per ton). After a lag of two months, this sudden upturn in the rice price that had begun in August 1974 was followed by an abrupt increase in the number of unclaimed dead bodies that were collected from the streets of Dacca, the capital city, for burial by the Anjuman Mufidul Islam, a charitable Muslim burial society (see Figures 1 and 3).
There was no expansion of
the burial society's services, but their records indicated a fourteen足 fold increase in the number of street deaths:
from 50 bodies per month
in September to over 700 bodies per month in December 1974.^ Although there were no denominators to the figures of the burial society, the shape of the curve has strongly suggested a cause outside
^The existence of a dysentery epidemic in Dacca during the months preceding and following this sharp rise might lead one to suggest a causal relationship. However, the epidemic admission records at Dacca cholera hospital clearly indicate that the timing of the epidemic did not coincide with the previously noted increase in mortality. Moreover, the curve of street deaths rises much more steeply than the epidemic curve, further suggesting that the two events were not causally related.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 the city.
Some phenomenon or phenomena within the capital's catchment
area had suddenly increased the population at risk to mortality in Dacca.^ In Purundapur, a village of about 1,000 people in a cholera field trial area less than 30 miles from Dacca (see reference map, Figure 10), the demographic symptoms recorded at the same time in 1974-75 (see Figure 1) were reminiscent of those noted in the eighteenth and nineteenth century parish and district records of the now developed areas of e.g., England (Wrigley, 1968), Finland (Jutikala, 1955), France (Ruwet, 1954), Japan (Office of General Affairs, Chumbo Province, 1898) and Sweden (Thomas, 1941), i.e., high death rates, lowered birth rates and out足 migration. Between August 1, 1974 and July 31, 1975, the number of deaths recorded by the village midwife (dai) in Purundapur increased from an annual average for the preceding years of 19 per thousand to 53 per
%ote:
a)
In 1976 the Anjuman Mufidul Islam's records in Dacca again fell to below 100 street deaths per month.
b)
The decline in street deaths in April, May and June 1975 may have been due to the unofficial removal of the bustee dwellers (bastuharas) to planned camps outside the city limits in March 1975.
c)
Similarly the second increase in July and August 1975 may have been due to the unofficial return of the bastuharas and to the influx of unemployed landless laborers during the wet season, after the harvesting of the summer rice (aus) crop.
d)
Although the Anjuman Mufidul Islam served very few of the minority Hindu population, the Anjuman officials claimed that the records of the Satkar Samiti burial society followed the same general trend.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10 thousand.
Eleven of those 53 deaths were noted as being due specifically
to starvation.
After the normal nine month gestation period the number
of births dropped noticeably from an average of 55 per thousand per year during the preceding six years to only 20 per thousand during the twelve months after 1st May 1975. The most sensitive indicator was "out-migration" (see Figure 1) which leapt from an annual average of 62 per thousand per year during the previous six years to 303 in the twelve months immediately after 1st August 1974 without any corresponding "in-migration" being recorded. This out-migration clearly reflected an extreme or crisis situation when compared with the previous rural exodi, (a) as people returned to the cities after the urban political unrest over the Agartala Conspiracy at the end of Ayub Khan's Pakistan regime in late 1968 and early 1969, and (b) as the migration from the village back to Dacca city began in early 1972 at the end of the liberation war (see Figure 1).
In the twelve
months after 1st August 1974, out-migration was not recorded as being in search of food.
Seventy-eight percent of the outmigrants during that
twelve month period were recorded as having left for a better livelihood, thus suggesting poverty-induced migration and the inability of the community to support its rural poor.
The migration recorded as being
because of river erosion suggests only one of the many causes of the famine syndrome in Bangladesh.
3
The magnitude of this "out-migration" is further stressed in Figure 2 where the total breadth of the arrows on both maps emphasizes
^The causes of famine in Purundapur are discussed more fully in Chapter III.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
Figure 2,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12 that out-migration during one famine year (1st August 1974 to 31st July 1975) was almost equivalent to seven years of normal out-migration (1st June 1968 - 30th April 1976. excluding the Famine Period).
Out-migrants
during the famine period reportedly headed for destinations^ similar to those of normal times, including places like Karachi, over 2,000 miles away, thus probably suggesting they were making links with already established contacts.
The subtle change differentiating the famine from
non-famine conditions, as reflected in the maps, is the decrease of rural destinations during the famine period.
Out-migration became
channelled towards the urban areas, as the ability of other rural communities to support additional members was reduced.
The proportion of
the out-migrants moving to Chandpur during the famine period increased. Chandpur is the road, railway and riverboat terminal for the cities of Dacca and Chittagong.
The Manifestation of Famine in Urban Areas It is in the cities, particularly in Dacca, and not in the rural villages like Purundapur, that famine first becomes politically potent. Destitutes from the rural areas concentrate in the cities, and being unable to gain access to the government’s urban rationing system, they irritate the middle-class populations by their overt begging.
In nonÂ
famine times, begging is normally discrete, softened by the presentable
^The destinations are only those reported to the daiis by the migrant's family, relatives or neighbors in the same village. In many cases they may have been guestimates or only the first stopping place.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13 garments of
the beggars and the cover of evening.
During times of
famine, begging becomes a day-long march of many near-naked and skeletal poor. The prime locations for begging also become the foci for street deaths (see Figure 3).
At Azimpur Gate, the entrance to the cemetery,
the beggars could tap the traditional custom of people who, after paying their respects at the grave, gave alms on behalf of their dead relatives and friends.
Similarly, Kamlapur railway station, Sadar Ghat ferry
terminal, and Gulistan bus station were also places where the middleclass travellers were likely to have loose change and where they could be cornered while they were waiting for transport.
The concentration of
street deaths in the old town, near Gandaria, was partly due to its proximity to the headquarters of the burial society.
It may also have
occurred because the congested housing of the old town allowed the beggar to interact with the donor easily, a situation made impossible by the locked gate and compound wall of the houses in the new town. The alms-giving associated with famine, initially felt in the
urban
areas may appear paradoxical, as the Bengali word for famine is Durvickha, "a shortage of alms" (Figure 4).
However, although it was the alms足
giving in the urban areas that drew attention to the famine symptoms, the causes of the famine were in the rural areas where there was indeed a "shortage of alms."
Impoverished rural families reportedly would not
shelter even a close relative for more than one night.
Also in rural
baris or homestead groups, some households were observed to be very near starvation, while their agnatically related neighbors obviously still lived well.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G E O G R A PH IC A L A N D T EM PO R A L D IS T R IB U T IO N OF STR EE T D EA T H S IN D A C C A CITY
D eath s in D acca City by Grid Cell O c to b e r 1 9 7 4 to D ec em b e r 19 7 5
N u m b er o f S tre e t D eath s by M o nth in D acca C ity July 1 9 7 2 to D ec em b e r 1 9 7 5
Figure 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15 The destitutes who migrated to the cities during the famine settled spontaneously on unused urban land, often land that was owned by the government.
The resulting hastily assembled settlements of squatters, the
bustees, irritated the sensibilities of the urban artist (e.g., Figure 4, Abedin, 1943), and of a dignity-conscious government.
They also became
volatile areas of political unrest close to the center of power (Quadir, 1975).
The large-scale out-migration from rural areas made the officers
of the many international aid and charitable organizations aware of the plight of these refugees.
In 1974 the telephone directory for Dacca
listed over fifty such agencies, which with their international mandate, were more able to publicize the disaster than was the economically and politically constrained government. Foreign administrators and reporters have used the concentrated symptoms of famine in urban areas to raise awareness and political support both internally and internationally.
Wherever and whenever these
symptoms occur, they are cruelly apparent and strike the observer with an emotional force not soon forgotten: Still fresh in memory's eye the scene I view, The shrivelled limbs, sunk eyes, and lifeless hue; Still hear the mother's shrieks and infant's moans, Cries of dispair and agonising groans. In wild confusion dead and dying lie;— Hark to the jackals yell and vulture's cry, The dog's fell howl, as midst the glare of day They riot unmolested on their prey! Dire scenes of horror, which no pen can trace, Nor rolling years from memory's page eface. (John Shore, Lord Teignmouth, 1770) Lord Teignmouth's recollections of the Calcutta famine of 1770 bear a desperate similarity to Michael Hornsby's reports of scenes in Dacca in the London Times on November 1st, 1974:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
(.snortage ox du;a; ^ > S k e t c h by Z ai n u l A b e d i n (courtesy of the A b e d i n Family, Dacca)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17 Lavya Khatun is 20 but looks much older. Wrapped in a ragged green sari, she crouches on the ground dazedly waving away the flies which cluster round the scabies sores on the head of her 16 month old daughter. Sx^addled in a filthy rag, the child is a tiny bundle of bones draped with the loose hanging wrinkled skin of a very old man. Occasionally the small body is seized by a dysenteric fit and ejects a thin stream of yellow liquid. (Michael Hornsby, 1974) The melancholy symptoms of death, starvation, migration and disease are common not only to Bengal, but also to the farmers of North East Brazil in the nineteenth century, viz. Thousands of living peasants became but living corpses of flesh and bone. In this weakened state epidemics and diseases killed unchecked, typhoid, smallpox, beriberi, influenza and tuberculosis. The horror of smallpox be足 came very real in the city of Fortaleza during the drought of 1877, where the epidemic spread among the crowds of refugees. For the estimated 80,000 cases in the capital and its suburbs, there were only ten doctors to treat them. In the peak month of December, 1878, 14,491 died of smallpox alone. The day of terror came on 10th December when 1,004 persons succumbed. (Reuben H. Brooks, 1971) and to recent conditions in Ethiopia: "We found the family in the desert about 15 miles north of here," said Arnulf Tjugen, a Norwegian official on the League of Red Cross Societies, who has set up feeding points and a hospital here. "They were digging in the sand with their hands to make a grave for the oldest son, who was 18, and who had just passed away." Dehydration, measles and colds compounded by weakness from the lack of food and water were among the major causes of deaths among the Taureg refugees. (Anonymous, 1973) The universality of these reported melancholic symptoms has re足 cently been emphasized in developed countries by the visual images of the newspaper photograph and the television documentary.
It is true
that the actual groups affected may vary from culture to culture, e.g., in the Anatolian famine of the winter of 1947-48, excess deaths were highest among children less than a year old (Makal, 1954), whereas in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IS Bangladesh prolonged breast-feeding displaced the excess death rate upwards to the 1-4 year old age group (Curlin, Chen and Hussain, 1975), and in Ethiopia the most affected age group were the 5-10 year old males who were away from home herding cattle (Hay, 1976).
Nevertheless the.
pattern of misery, malnutrition and death has been relatively constant and has become increasingly synonymous with the image of famine.
The Recent Consolidation of the Symptomatic Definition During the past decade famines have become virtual field labor atories for nutritionists.
Their studies have concentrated on levels of
nutritional status and death rates, e.g., Ramalingaswami, Deo, Gularia et al. (1971) regarding the Bihar famine of 1966-67; Aall (1970), Davis (1970) and Miller (1970) regarding the Biafran war in Nigeria; Sprague and Foster (1972), and Lowenstein and Sommer (1972) regarding the post liberation war period in Bangladesh; and the survey of the Center for Disease Control (1973) in West Africa, and Jeannee and DeVille de Goyet's studies in Niger in 1976.
In 1974, I too, heedless of a Bengali medical
nutritionist’s remark that "We need a survey of surveys, rather than another survey" (Rahman, 1974), fell into the same mould by mounting a survey to assess the height, weight and arm circumference of several hundred children in an overtly famine stricken area of Northern Bangladesh (Koster and Currey, 1975). These famine surveys were an outcome of the nutrition surveys of Schaeffer (1961) and the ICNND (International Committee on Nutrition for National Defence), which had been primarily designed to examine the chronic problem of malnutrition.
The application of techniques designed
for a chronic situation to a crisis situation has produced the paradox
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19 of widespread deaths occurring while nutritionists discuss the niceties of nutritional measurement ranging from the complexities of serum protein calibration (Ramalingaswami et al., 1971), to the oversimplicity of Arnhold's (1969) QUAC5 stick technique.
Famine survey, from being "an
almost untouched field" (Lechat, 1975), has fast become a field where too much has been done but too little achieved.
This is because famine
survey has become equated with nutritional survey.
Nutritional survey
normally falls within the bailiwick of the Ministry of Health, but famine prevention and relief often fall under the aegis of the Ministry of Relief and Rehabilitation or the Ministry of Agriculture.
Nutrition
surveys are highly sophisticated both in terms of sampling methods, standardization procedures and measurement techniques. tion and implementation takes a great deal of time.
Their organiza
Because of their
association with the health ministry they then have to overcome the difficulties of inter-ministerial permission to proceed. may even be on the wane by the time they are implemented.
The famine They normally
attempt to measure the impact of the famine in the nutritional and mortality symptoms.
All too often they produce self-evident results
such as "documenting an extremely serious undernutrition problem through out the surveyed areas" (Kogan et al., 1977).
Seldom do they document
the available skills and resources for rehabilitation and development in the area.
Rather, the nutritional perspective has achieved the quantifi
cation of the degree of malnutrition and its relationship to the probable
'’The QUAC stick was developed by a QUAker group in order to quickly determine the ratio of Arm £ircumference to body height as a rapid assessment of nutritional status.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20 death rates (Lowenstein and Sommer, 1972).
By diligently quantifying
the terminal symptoms of famine they have reinforced their own definition of famine as being "a widespread food shortage leading to a significant rise in the regional death rate" (Blix et al., 1971; Mayer, 1975). Paradoxically then, at the same time as food relief in response to the terminal symptoms of famine is being increasingly questioned as a solution to the problem, so research on the terminal symptoms is in creasing.
This research, I suggest, is shifting the definition of
famine even further towards one that necessitates symptomatic cure.
By
referring back to Table 1 it can be seen that the three internationally oriented groups, retired government officers, technical experts and aid distributions, ranked the increase in starvation mortality highest in defining a famine, while the village elders ranked it sixth.
The
two groups presently involved in decisions on relief activities„ the technical experts and the aid distributors, both consistently mentioned "high malnutrition levels" in their definition of famine.
Indeed it
ranked third for those groups, but was not even mentioned in the rankings by the village elders or those by the retired government officers.
I
question whether this really reflects an increased understanding of famine by the technical experts and aid distributors or merely an in creased precision in their misunderstanding famine because of better access to nutritional documentation in the available literature. Nutritionists have given international food aid its raison d'dtre: to import food for improving the nutritional status, and then to wait until called upon the next time nutritional levels fall.
They have
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21 justified the politicians’ pet euphemism in which they refer to famine as "a severe food shortage."
The Problems of Famine Amelioration by Food Relief Food relief as a response to famine conditions has been increasingly criticized both by recipients and donors.
Individuals have pointed out
the problems resulting from the resentment and feeling of dependency on the part of the recipient, as well as the possible disincentive effects of food aid on the recipients’ agricultural production.
The donors too are
conscious of the wastage and inappropriateness of relief aid.
With a
commodity like food, which is dependent on weather conditions, scarcities can also occur among donor countries (Sanderson, 1975).
They can and
may, in such a case, refuse to donate to those in greater need.
More
over the food aid that is given from distant international sources often arrives late, and when it does arrive its effectiveness is not certain. Some of these divergent attitudes are discussed below: 1.
The dysfunctionality of relief.
Hyman (1966) coined this phrase
to describe the resentment felt by countries receiving relief aid when they are assigned "beggar-nation" status and even designated as "basket cases."
Gouldner (1960) has suggested
that the resultant "biting of the hand that feeds" stems from a neglect of a "primordial imperative," the need to recipro cate.
Unlike trade, food aid is a one way transfer which may
result in either resentment by governments who are sensitive about their newly obtained national sovereignty, or in a "relief mentality" combined with a loss of independence on the part of the recipient nation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22 2. Relief dependency.
Individuals, when automatically provided
with a free service, may grow to depend upon that service. At times they may be influenced by the provider of that ser vice.
Similarly, governments may begin to rely on politically-
tied external support which may afford a crutch to their own ineptitude.
If the present dependence on relief food aid
continues, changing climatic patterns may greatly increase both the amount of food relief support required by the develop ing countries, as well as the proportion of that aid which is provided with political implications (C.I.A., 1974; Bellagio Conference, 1970).
3.
The disincentive effects of food relief.
The disincentive
effects of food relief have been noted in India by Isenman and Singer, 1976, viz., "The increase in food supplies pro vided by food aid depresses prices received by farmers and causes or supports inadequate agricultural policies by recipients, which together lead to decreases in food pro duction."
Timms (1976) on the basis of empirical observa
tion has noted a similar effect in Bangladesh where the effect might be expected to be greater because of the local high ratio of relief food aid to indigenous production.
Recently Brammer
(1976) has noted, however, an actual increase in indigenous wheat production after the massive influx of relief wheat. The introduction of lower priced wheat to a diet previously restricted to rice may have acted as an incentive for local innovation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23 4.
The wastage of relief.
Burley (1973) estimated the distribution
costs on relief aid to be 40 percent compared to 15 percent for development aid.
Although this comparison may be crude, 1 myself
witnessed the sudden abundance of "soft money" for ad hoc groups to initiate projects without proper monitoring of either costs or effectiveness during the famine in late 1974 and early 1975 in Bangladesh.
The actual distribution of inter足
national relief food during a crisis period is particularly wasteful.
National entry points, whether ports or airports,
are blocked by the sudden abnormal surfeit imposed on scarcity conditions.
Food stocks may rot in heaps, open both to the
elements and to attack by pests.
Pilfering may be widespread.
The connections between the entry points and the area in need are often atrophied or broken. rolling stock and trucks.
There is often a shortage of
Roads and bridges may have been
washed out by recent floods and local officials supervising the transfer of relief foods may be underpaid or unpaid, thus increasing the possibility of pilfering and the willingness to turn a blind eye to the wholesale smuggling of grain to areas where it can be exchanged for cash.
Forcing food
relief through the indigenous distribution system during a famine is often like pushing a camel through the eye of a needle.
The circumvention of this local system by international
organizations or local military personnel may be more efficient and effective, but it often also involves extra manpower and equipment costs, e.g., military guards, power launches,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24 jeeps and helicopters.
It may disrupt development activities by
co-opting key international and national administrative per足 sonnel from permanent programs to serve in ad hoc relief positions.
Superimposed food distribution channels may merely
transfer the location of the bottleneck from the port to a point inland, as when the C-134 transport planes unloaded food supplies on an inaccessible airfield which was nearer the area of need, but which had no entree into the local distribution system.^ There is still potential wastage even if the food relief reaches the local area of need.
In 1974, I witnessed small
children walking distances of up to six miles, under a hot sun and often wading through rivers, to a food distribution center to bring back a packet of five biscuits and a tin of diluted powdered milk.
Recall that Passmore pointed out that
in 1951 that it takes at least 500 calories to walk 5 miles (Passmore, 1951).
5.
The inappropriateness of relief food.
In Bangladesh I per足
sonally saw no evidence to support the rumors of tinned ham being shipped to Muslim countries and slimming pills being included with supplies to famine areas.
I did witness, however,
old men who were visibly starving refusing to eat wheat in
6In 1974 a U.S. C-134 from Clark Air Force Base in the Philippines flew relief aid from Chittagong to Lalmanirhat airport in North Bangladesh, near the famine area. Within 10 days the goods were on sale in Gulshan market, in the foreign residential area of Dacca city, almost 200 miles to the southeast of Lalmanirhat.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25 the form of attar ruti (unleavened bread made from unrefined wheat flour).
They may have been physiologically unable to
digest the wheat because of their bodies' inability to produce the appropriate enzymes.
They may also have been psychologically
unwilling to digest wheat after a lifetime of eating rice. I have also noted essential relief trucks burning scarce fuel supplies transporting custard pies with minimal nutritional value from Chittagong to Dacca.
The pies had been found
unsaleable in the United States and were,thus being donated for relief aid in Bangladesh in order to secure a U.S. tax rebate.
Also in 1975 I ate stale relief biscuits culled from
a tin marked "1961" which had to be smashed open for lack of an accompanying can opener.
6.
Famine relief may be cut off by the donor in times of widespread scarcity.
The occurrence of famines in one or several
developing countries may be coincident with low points in the grain reserves of the developed nations.
At these times, in
alarm and realization of the wastage involved in repeated relief activities, a donor country like the United States might initiate calls for a termination of aid to those in greatest need.
Paddock and Paddock, during the Indian crisis in 1967,
proposed the concept of triage, whereby scarce food resources would be allocated only to those countries deemed "savable" by the donor.
The famines in the Sahel, Ethiopia and Bengal
coincided with the drop in world grain reserves in 1973-75,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26 and there was a resurgence of the calls for "triage" (Green, 1975)
and "lifeboat ethics" (Hardin, 1974), or the saving of
food stocks only for those already on board the lifeboat. Nutritionists like Berg (1975), have failed to grapple with the underlying problem of dependency and have countered the "lifeboat" and "triage" analysis with another nutritional answer by arguing that there is indeed enough food available to supply everyone. A strong resurgence of these calls for "triage" might diminish both famine relief and development aid to the most needy nations, but failure to ameliorate famines in the vulnerable nations will not stabilize their population-resource balance by eliminating the weaker and poorer members.
It will
rather make them increasingly vulnerable to further famines. Mahalanobis (1944), Mukherjee (1944) and Mukherjee (1965), have all documented the massive rural impoverishment that occurred during the Bengal famine of 1943.
Mahalanobis (1943) using the
"number of families transferred from occupations at a higher economic level to occupations at a lower level," found that 3.8 million rural Bengali families suffered a lowering of economic status and decrease of earnings between January 1943 and May 1944.
Mahalanobis also calculated that the war and famine
increased the number of destitutes (those dependent on charity) in Bengal by 480,000 persons.
One of the higher estimates of
mortality during the 1943 Bengal famine was 1.5 million (Greenough, 1977).
It probably claimed a higher proportion of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27 the weaker and poorer members of the community, but it also probably impoverished two and a half times more people than the number that died, making them susceptible to further shortages. Actually, the long-term demographic effects of famines are little researched.
The relatively small proportion of the
population involved in famines in Ethiopia, the Sahel, Bihar sud Bangladesh in the latter half of this century suggest that their long-term demographic effect is minimal.
Faulkingham
and Thorbahn (1975), studying a single village of 1,500 Hausen in the Western Sudan found that five years of drought had had very little effect on the community's demographic structure. Caldwell (1975) has calculated a similar lack of demographic impact for the whole of the Sahel region.
Although the previous
study of Purundapur village in Bangladesh (see Figure 2), in足 dicated considerable short-term local change, Baldwin (1977) has estimated that even if the demographic catastrophies of 1970, 1971 and 1974 (cyclone, war and famine) were to continue to take an equivalent toll every five years until A.D. 2000, the total long-term impact, when calculated against the high growth rate, would lower the total population by 6.4 million from an estimated 142.6 million to an estimated 136.2 million. Baldwin pointed out, however, that the total population would be only slightly lowered by catastrophies.
Its ability to grow
would remain and the crude birth rate would probably not change. Although the demographic results do remain equivocal, there is little historical (see Figure 16) or present evidence to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28 suggest that famine is a long-term correcting mechanism for populations with a population-resource imbalance.
Thus, a
system that would encourage the termination of food or other aid to countries like Bangladesh, through a "triage" policy, is unlikely to solve the problem of famine.
7.
The effectiveness of famine relief.
Although relief agencies
are eager to publicize the successes of their relief programs there are no controlled intervention studies quantifying the effectiveness of a food relief program in a famine area.
Such
a study is not only difficult to implement amidst the emphasis on action rather than research during a crisis, but there is a serious ethical problem with the control group, since they cannot be given a placebo rather than relief.
Evaluation of the
effectiveness of a food relief program therefore normally has to be made on a single population at different time periods. Improved nutritional status is often recorded, but it may be due to the deaths of the most severely malnourished individuals or out-migration of the poorer families who may have carried malnourished members with them.
The improvement may be due,
not to the relief food, but to new food resources that have become available locally with changing crop seasons or new social links with nearby food supplies.
The normal abatement
of a severe epidemic, e.g., dysentery, may similarly improve nutritional status irrespective of food relief inputs. As Lechat (1974) has suggested, the crucial test of meaning足 ful relief effectiveness should come after the relief program is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29 over.
The pattern of nutritional status before and after the
end of relief distribution has not to the best of my knowledge ever been documented, but Dodge (1976) observed a relapse in nutritional status in North Bangladesh in 1975, a few months after feeding centers had ceased functioning.
8.
The tardy arrival of relief food.
The persons requiring famine
relief are not responsible for getting that relief for them足 selves.
There are many links in the network which first trans足
mit the details of the needs to the donor, next corroborates and reinforces the magnitude of the needs, next expedites the relief process, and then finally distributes the relief to the person in need. Needs may be first transmitted to donors by relatives of donor agency employees visiting the city and describing difficulties in the village home (bari).
These may soon be
corroborated through the local government hierarchy.
The
village constable (choukidar) or union council members may report to the Circle Officer (Development), and after "pro足 cessing," the report is then passed on via the Subdivisional Officer (Relief), the Assistant Deputy Commissioner (Relief), and the Divisional Officer, to Dacca. Inertia may have existed at different levels because of the burden of other commitments (e.g., national smallpox searches) during the strained conditions.
Also, officials may have been
reluctant to admit to problems in their own areas (e.g., in 1974 civil servants had to proceed warily because of the threat
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30 of physical harm from politicians and their henchmen).
Those
local civil servants who do admit to the situation may initiate locally-based measures quickly but the local resources often prove insufficient.
The actual transmission of the need for
relief to foreign donors might be made by a lone foreign motor足 cyclist moving through a remote area and reporting crisis con足 ditions to his Dacca headquarters.
This might initially
stimulate a cable abroad, and another reconnaissance party to the area, but it might require several surveys and political meetings before any aid is even allocated to the particular area.
After the allocation decision is made, the actual dis足
tribution of food aid depends upon such factors as:
docking
facilities at the port; the availability of sufficient rolling stock, diesel fuel, which has not been contaminated with water, for the trucks, and spare wheel parts for bullock carts; and the tolerance for overloading of the country boats.
Food
relief distribution is far from the flashing siren and heli足 copter image of relief in the developed world. Documentation of the tardy arrival of relief during two recent famines in Bangladesh and Ethiopia (Figure 5) illustrates the time that elapses between early warnings of famine and the arrival of international relief in the area of need.
In both
instances the international relief arrived only after the time when the local crop production had recovered.
The evidence of the inefficiency and inefficacy of food relief systems for famines does not necessarily negate the concept of a "genesis
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
5.
Sequences
of Delay
in
the
Supply
of
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32 strategy" (Schneider, 1974 and 1976) whereby multilateral buffer stocks are established in preparedness for times of famine (Fitzpatrick, 1977). Although in an extreme famine situation, buffer stocks are likely to have evaporated, the very presence of buffer stocks of grains may minimize wild speculation and unnecessary price increases at the inter足 national level.
Nevertheless, the difficulties previously noted that
are associated with effecting the transfer of relief food grains from regional stroage areas to the stomachs of rural poor do justify the consideration of an alternative strategy for ameliorating famines.
Putting Relief in a Development Context In 1963 Masefield suggested:
"It might be taken as axiomatic that
prevention of famine is even better than relieving it once it has started."
Rational though this may seem, ten years later, on January 1,
1973, the U.N. Security Council report indicated that $1,800 million had been committed to Bangladesh for the U.N. Relief Operation.
That was
four and a half times as much as the United NationsT Development Program global budget of $225 million in 1972. During the six years before the liberation of Bangladesh in 1971 and September 1977, United States economic and humanitarian assistance had provided $48 million to Bangladesh in order to increase agricultural production.
That may seem a large figure, but it is less than one-fifth
of the amount ($253 million) given for disaster relief, and less than one-tenth that loaned for food imports ($518 million) (USAID., 1977). Development economists like Morris have argued that "intense public reactions lead to much more being spent on relief than seems to be called for by the economics of the situation" (Morris, 1975).
He even quoted
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33 the Report of the need
for
Sixth Finance Commission (India) that stressed the
a "link between short term famine
policy and long run develop足
ment policy," but then he argued that relief money should be invested in non-famine prone areas on the grounds of maximum efficiency and in terms of encouraging people to move out of famine prone areas.
This suggestion
apart from implying a form of geographical "triage" (see above) appears politically unsound and in the Bangladesh situation it may not even be feasible because of
the already high population density and the paucity
of areas that might receive such large numbers of out-migrants without themselves becoming famine-prone. Morris's policy also fails to take advantage of the development opportunities that the famine itself provides in the stricken areas: 1.
The blaze of emotional publicity highlights the tragedy of the poor
in an area.
the poor
gain a political potency that they do not otherwise
The results of this publicity are that
possess.
2.
In the glare of the international and national spotlight donors and central government officials need to appear to be doing something in the famine areas for the famine victims.
3.
Graphically demonstrating the failure of past policies, the famine makes the officials and the people of the area more receptive to the knowledge and skills brought in by the influx of external expertise.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34 4.
There is a sudden short-term, but often unprecedented transfer of resources into areas traditionally shunned by investment as having little to offer.
5.
Humanitarian aid to famine areas is usually flexible and seldom tied to particular projects.
6.
Crisis and the need for survival create the possibility that traditional structures and inhibitions in famine stricken areas may be discarded, thus leaving a potential hearth for innovation and change.
As Stepanek (1975) has suggested, to use this amalgam of opportunity "to put relief into a development context" in Bangladesh, it will be necessary to "target relief to the vulnerable groups in the areas affected by famine" (Stepanek, 1975) and to target it through programs that will lead to the long-term development of those areas.
The drought-
induced Bihar famine in 1967 set a precedent of "riding the wave that the urgency and drama of famine create" (Harvey, 1969), B. G. Verghese wrote: and set new standards.
After that famine,
"The famine has released the springs of initiative It has forced the pace of agricultural develop足
ment and resulted in the creation of valuable permanent assets in terms of irrigation works and water supply projects.
The kinship of adversity
has tended to break down ancient prejudices like caste taboos" (Verghese, 1967).
The new political kudos that could be gained from using famine
for development was reinforced by Indira Gandhi, then Prime Minister of India, by her quick public acknowledgment of "the importance of using emergencies to attain broader developmental objectives" (Berg, 1973).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35 However, the Bihar famine was induced by a drought which developed over a two year period.
In Bangladesh, the local famines are often induced
by sudden flooding, and there is thus an even greater need to anticipate the occurrence of famine if Stepanek’s "geographic targeting" is not to find that the target no longer exists.
This dissertation, by mapping the
areas liable to famine in Bangladesh, establishes a geographical target for famine prevention and development programs in order that they may be ready for implementation as soon as relief funds are released.
The Geographical Contribution to Mapping Areas Liable to Famine This delimitation of geographical targets for famine prevention and development programs is a study in applied geography.
The spirit in which
it has been undertaken is one of policy guidance (Hare and Jackson, 1972; Beaujeu-Garnier, 1975; Coppock and Sewell, 1976). "the
It is concerned with
application of a body of geographic knowledge and expertise to
problems that affect the lives of people" (Dawson and Doornkamp, 1973). In the process of providing the policy guidance, it evaluates a method, the ’Delbecq-Delphi Process,' taken from outside the discipline which may have a potential role in other geographical research. The dissertation approaches an old geographical problem of how to delimit areas liable to famine.
It follows ideas from Natural Hazards
Research (White, 1973) which suggest the possibilities of mapping the total vulnerability of an area.
Recent research on territorial social
indicators is then used to suggest solutions to the problems of amal gamating the various components of famine vulnerability on a common index.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
The analysis of areas liable to famine has not formed a major theme in geography comparable with studies of areas liable to floods (e.g., Thorpe and Tweedie, 1956; White, 1961; Macdonald, 1967) or areas liable to disease (e.g., May, 1950; Armstrong, 1965; McGlashan, 1972; Learmonth, 1976).
Many of the pioneering studies have been done out of necessity
by administrators with a geographical perspective.
Colonel R. Baird
Smith’s report on the famine of 1860-61 in the northwest provinces of India was probably the first study to map out the effects of famine and the associated areal variations in the rice prices (Smith, 1861).
W. W.
Hunter’s seminal work in Bengal in determining which tracts were liable to famine (Hunter, 1868) was extended in Annam when the Resident Superieure in 1923 called for: . . . une enqudte gdndrale destind A montrer dans chaque province 'les points ou aires de famine’ A indiquer la frequence des disettes pendant la periode 1901-1921, enfin A determiner leurs causes, causes naturelles (nature du sol, rdgimes des eaux, sdcheresses, inondation, insectes epizooties, etc.) causes exceptionelles (typhons, raz de maree, etc.), causes sociales (absence de petite propridtd, deplacement de la main d'oeuvre, accarpement des grains, etc.). Une telle enqudte mende simultandment par tous les chefs de province permettra l'dstablissement d'une carte generale des rdgions affectdes par la famine en Annam soit d’une faqon periodique, soit d ’une fa^on accidentelle.7 (Chassigneux, 1923) More recently Geoffrey Last (1974), in the Ethiopian Ministry of Educa tion, has been instrumental in setting up a famine warning system for
” . . . a broad investigation to delimit the famine points and retions in each province, to indicate the frequency of scarcities between 1901 and 1921, and also to determine their causes, whether natural (soil conditions, water levels, drought, floods, insects, animal diseases), calamitous (typhoons, tidal waves), or social (lack of small land owners, eviction of labourers, grain procurement etc.). Such an enquiry, led by all the provincial governors at the same time, would allow the compila tion of a general map of the regions affected by recurring and inter mittent famine in Annam."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37 famine vulnerable areas.
In Bangladesh, Hugh Brammer, a land use
advisor with the Food and Agricultural Organisation, has compiled disaster contingency plans for agricultural recovery in each of the disaster prone areas of the country (Brammer, 1975). Academic geographers, in the absence of actual famine field experi ence, have relied on the availability of historical records to map famine areas.
Walton (1952) has used material from the old statistical accounts
written by parish ministers to describe the famine symptoms resulting from the rigorous climatic conditions in pre-nineteenth century north east Scotland.
Dando (1976a) was able to delimit four famine zones in
Russia on the basis of reported famine frequency in chronicles written between A.D. 971 and 1970.
The study unfortunately masks the complexity
of famine vulnerability by first classifying all the famines as man-made and then further grouping them according to a single cause, thus classi fying all Kursk famines as "overpopulation famines."
This oversimplifica
tion is continued in another of Dando’s studies attempting to define past, present and future famine zones at the global scale (Dando, 1976b), but the latter study is further marred sources.
by a lack of specified data
Cousen (1960a and 1960b) and Geddes (1970) have used the
historical records of regional demography to study the causes of famine in different regions.
Cousen’s papers analyzed the geographical patterns
of mortality and emigration in the Irish famine of 1846-51 (Figure 7 ) in an attempt to determine the underlying causes of the famine in different areas.
Brooke (1967) has collated the entries of food shortages in
extensive missionary records of Tanzania, to indicate the areas of recurrent food shortage.
Geddes graphically compared the patterns of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38 population growth at the district level in British India by using the decennial censuses between 1881 and 1931.
Those districts with a
serrated population growth pattern were classed as areas of "recurrent crisis" or "famine tracts."
Although stimulating, the study is limited
by the aggregate nature of the decennial time-scale and the district level area scale which mask both intra-district differences and intercensal migration patterns.
One of the earliest studies by an academic
geographer of areas liable to famine was George T. Renner's study of the former Sudan area (Renner, 1926).
In the study he subjectively amalgamated
the problems of agriculture, nomadism, culture, and climate in order to predict a famine prone area congruent with that affected during the Sahelian famine of the early 1970s. Apart from this intuitive
amalgamation by Renner, past geograph足
ical studies of famine have avoided the problem of fitting the famine liability factors together so that the composite is usable by the policy implementers for "targeting" famine amelioration programs,
Natural
hazards research (white, 1973) has continued the initial work of Montandon (1924) and la Societe de Geographie de Geneve on "la Geographie des Calamites."
Because of the research's new emphasis on the
cultural responses to natural disasters, it has developed ideas of com足 posite vulnerability or total risk in terms of the total effects of the hazard rather than from the total probability of the causes of the hazard.
Initially natural hazards research studied individual hazards,
e.g., White's studies of flooding (1961) and later Ritter and Dupre's "Maps Showing Areas of Potential Inundation of Tsunamis in the San Francisco Bay Region of California" (1972).
At the local scale, Burton
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39 and Moon (1971) and Hewitt and Burton (1971) examined the total "hazard ousness of a place" for London, Ontario.
They compared the perceptions
of the people at risk with the historical record of hazards in the area, but they did not map out the areal dimensions of the total risk.
In the
national inventories of natural hazards, some countries only listed the hazards individually, e.g., Canada and the United States (Visvader and Burton, 1974) and Japan (Nakano, 1974).
The U.S.S.R., with its tradition
of studying the total environment of risk, compiled 29 regions of com posite hazard vulnerability, but the regions were classified according to damage potential (Gerasimov and Zvonkova, 1974).
At the international
scale, both Sheehan and Hewitt (1969) and later Dworkin (1974) have attempted to map the composite effect of "catastrophic events" by plotting the results of a contents analysis of international news chronicles on 10° grid cells on an equal area projection.
Both these
studies exhibit a disregard for the problem of incomplete and regionally non-uniform information available, especially from developing countries, on the effects of disasters.
The studies do evolve a number of
posite scales, but again they
measure effects such as "catastrophic
com
impacts" or "deaths" rather than a composite of the factors predisposing areas of the world to those catastrophic impacts. A physical planner, Lewis, has studied vulnerability in a policy context for disaster preparedness planning.
He has proposed the use of
the "McHarg Method" (McHarg, 1969) to overlay the areas affected by the various hazards in Mexico to produce an amalgamated areal inventory of vulnerability for the country
(Lewis, 1973).
Lewis’s study, as
recent areal studies of "environmental perturbations"
by Polish
withthe and German
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40 researchers on natural hazards (Leszczycki, 1972a, 1972b, 1976; Waksmundski, 1972; Neef, 1972) leaves two major questions unanswered. First, what is the relative importance of each of the individual hazard areas in contributing to the total risk of vulnerability?
Second, once
these vulnerability maps are compiled, what should be done with them, and by whom? The first problem has been partially answered by recent research on ’territorial social indicators for planning' in the United States (Smith, 1973) and in the developing world (Cant, 1976).
The research
has focused on constructing indices of welfare for administrative units with the aim of helping to plan for greater 'spatial justice' by high lighting spatial disparities in welfare.
It has drawn on methods
developed in earlier studies including those of multivariate regional ization (Ginsburg, 1961), those of the social indicators movement (Baster, 1972) and those of factorial ecology (Berry, 1971), to give weightings to particular welfare indicators and then to map the resulting index of "human well-being" for each planning unit (Smith, 1975).
These
weightings have been derived mostly in the form of regression coeffi cients, measuring the relative importance of an independent variable or suspected cause (e.g., low income) on an already measured dependent variable or suspected effect (e.g., number of homicides).
Some weight
ings have been derived from factor loadings which measure the relative similarity of each of a group of suspected variables with some unmeasured common factor which may, or may not, be the concept (e.g., vulnerability) which it is hoped to measure.
Regression coefficients
presume a very
specific measurable "effect" and factor loadings involve human judgments
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41 or often the researcher's individual judgment with its attendant eccen足 tricities in deciding what the concept really is.
Attempting to overcome
these limitations, Adelman and Morris (1971) discussed the use of a panel of experts to rate variables which might describe a nation's economic development.
Schwarzberg (1962) has similarly used a group of
experts to rank regions of India by their level of economic development. Three difficulties of this method were acknowledged by Adelman and Morris: 1.
Concepts may be interpreted differently by different experts.
2.
Opinions about the facts may vary.
3.
Opinions may be biased,
The Delbecq-Delphi Process tested in this dissertation is an attempt to resolve some of these difficulties which arise in mapping areas liable to famine. The second problem which also occurred with previous composite maps of natural hazards, is how to integrate the map of areas liable to famine into the process of policy making.
The problem cannot be com足
pletely solved unless the map compiler is also the policy implementer, but in this study it is partially solved because: 1.
The policy makers requested the map.
2.
The policy implementers were involved in the map compilation
3.
The map is not merely a tool for analyzing famine liability.
process.
It also includes particular policy instruments of minimizing famine liability in particular areas.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42 This dissertation is the outcome of a request by policy implementers to: • . . provide a map of areas liable to famine in Bangladesh which will: 1.
Identify areas to be given priority in food for work programmes.
2.
Identify areas in which relatively high food stock levels should be maintained.
3.
Identify priority areas for disaster preparedness measures. (Ministry of Relief and Rehabilitation, Government of Bangladesh, 1976)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER II WORLD EXAMPLES OF FAMINE VULNERABILITY
Different localities exhibit very different powers of resistance to the strain of famine. W. W. Hunter, 1874
There are certain areas which year after year are visited by floods, and each year there is the call for food supplies. . . . we should not be caught with our pants down again. A. K. M. Ahsan, 1975, Bangladesh Planning Commission
Famine vulnerability means the likelihood of being wounded by famine and not the state of having been wounded.
The concept is uni
versal, but the components vary from time to time and from place to place.
Bangladesh's famine vulnerability is different from that of
other countries, and even within the country, the causes of vulner ability vary.
A complete study of famine vulnerability must entail the
consideration of processes operating on several different levels, from that of the household to the international one; however, such an under taking is beyond the scope of a dissertation.
Therefore this thesis is
limited to examining the physical and socio-economic considerations found in those geographical regions that are persistently liable to famine in Bangladesh.
It is not possible to examine the complex processes under
lying famine vulnerability for all the villages of Bangladesh.
Therefore,
this thesis examines only those areas with certain environmental and socio economic conditions which coincide with the regions that are persistently
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44 liable to famine in Bangladesh.
However a general understanding of the
complex concept of famine vulnerability is first necessary if preventive planning is to recognize the variables in the problems among different countries and even
within the same country.
The term "vulnerable groups" (World Food Programme, 1976) has been applied to people in langerkhanas (gruel kitchens), the camp of destitutes in or near the urban areas of Bangladesh.
The use of this term for this
group is inappropriate because the people in these camps have "already been wounded" by famine and have been forced because of famine to leave their villages.
To attempt to ameliorate famine by giving food relief
to this type of "vulnerable group" is only treating the visible part of the problem.
The bulk of the problem remains untouched.
In fact al足
locating relief only to such "vulnerable groups" may merely encourage more people to come into the city.
Famine Vulnerability The term, "famine vulnerability," is most appropriately applied to the rural areas where famines are generated.
In those areas it is the
product of a complex array of factors which lessen the community's ability to sustain the rural poor.
The concept, however, remains far from
quantification, just as do many definitions of general "disaster vulner足 ability," e.g., The degree to which a community is at risk from the occurrence of extreme physical or natural phenomena where risk refers to the pejorative probability of occurrence, and the degree to which socio-economic and socio-political factors affect the community's capability to absorb and recover from extreme phenomena. (Bradford Disaster Researuch Unit, 1976)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45 Famine vulnerability can be understood in terms of three dimensions: 1.
the characteristics of the community,
2.
the scale of enquiry,
3.
the stage in the evolution of the famine
(Figure 6).
The extent to which each of the community characteristics contributes to famine vulnerability depends on the scale of enquiry.
A factor that may
seem critical when the scale of enquiry is small may in turn be less important on a national or international scale.
Similarly the state to
which famine conditions have evolved also influences the importance of the community characteristics.
Although the subtle complexities of
famine vulnerability defy neat boxing, the figure included here may help to introduce the concept.
The Characteristics of the Community This dimension, shown as the vertical dimension in Figure 6, may be further broken down into four classes: 1.
the production of the community being studied,
2.
the community’s efficiency in using that production,
3.
the storage and savings capacity available to the community,
4.
the degree of isolation of the community.
Production.
Production is a function of the characteristics of the
land, including the amount of land available, its fertility, and the availability of irrigation and drainage; and the level of agricultural technology that is used by the community.
It may be measured by the
annual yields resulting from all stages of the cropping cycle.
The
yield? xtom any territory depend partly upon the crops' abilities to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47 withstand or recover from environmental assaults.
While the tolerance
of different crops to such assaults varies, the effect of environmental assaults on production is a function of their length, severity, timing, and rapidity of onset. Internal production may be augmented from external sources:
pasture
may be grazed in different areas; sea, river, or lake fishing can greatly enlarge the land area of production; and labor can migrate seasonally to other areas for planting and harvesting, thus bringing back payment in cash or kind.
In some areas, families may have links with relations
who work in the merchant marine, or who work in the urban areas of the developed world, or in the oil states. strive to send home income.
These families living abroad may
Whole families may move out of an area during
a disaster period to seek income elsewhere.
This may also be viewed as a
form of increasing or altering territorial production (Waddell, 1974 and 1975).
However, disasters may also assault these extra-territorial sources
of production, such as when a fishing fleet is destroyed by a cyclone.
Efficiency in Using Production. famine vulnerability.
Culture is a major component of
The size, composition, and health of the population
of any territory determine the need for, and utilization of, the produc足 tion available to the community.
The members of the community may consume
and absorb their own produce or use the wages and income from their produce to buy other consumables from outside.
Social status, purchasing
power, political position and disease status all affect the community's ability to utilize its production even in normal times.
For example, the
poor and those of low social rank not only have little to eat, but even what they do eat often passes through underutilized.
The food intake is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48 lost to internal parasites, fevers, and bouts of diarrhea.
In hard times,
the social, economic and political status of the community affects its ability to draw on both emergency territorial supplies such as woodland fruits or normally unacceptable foods from the common land as well as its ability to receive relief from outside the territory.
Famine conditions
usually, but not always, worsen established discrepancies between the "haves" and the "have nots" (Wisner, Westgate and O'Keefe, 1976).
Those
who "have" survive, often at the expense of those who "have not."
Storage and Savings Capacity.
What have people to fall back on?
In times of stress the community's absorptive capacity, its ability to withstand assault without disintegrating, is associated with its own storage capacity and stocks.
It also depends upon the external storage
and savings capacity of the system within which it trades in normal times and from which it seeks assistance in hard times.
Storage, whether in the
household, the village, the open market, the government warehouses, or the international system, includes several factors:
the capacity of the store,
the present stock, the wastage and pest losses, the ease and lagtime for stock replenishment; and the capacity of transport for distribution from central storage to local storage and from local storage to user. Savings broadly covers those forms of famine insurance (McAlpin, 1974) which are exchangeable, but are not specifically stored or ware足 housed.
They may be in the form of currency, landbanks, animal herds,
household articles, or personal belongings, all of which are saleable in an effort to raise purchasing power in hard times.
Unfortunately, those
who need to sell at such times invariably sell at a low price while the prices of the necessities they then buy are extremely high.
Underemployed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49 skills or potential prostitution may be considered as a form of savings. Even credit worthiness and
administrative capability may be viewed as a
resource in seeking aid from donors.
The
greater a community's or nation's
indebtedness, or lack of savings, the more difficult it is to tide over a famine.
Degree of Isolation. Physical isolation whether
Where does the community get help from? from the donor country, the port of entry, the
central or local supply depot, or from local relief works or camps, may be considered in terms of road, rail, sea or river accessibility.
This may
vary seasonally as roads are flooded and rivers become transport arteries during tie wet season.
Isolated communities may be tardy in communica足
ting their need and in justifying their predicament to the outside world. The political and social isolation of certain groups because of caste, religion, political party, or even the absence of a strong political leader to advocate their case, may lower the probability of their receiving outside aid.
Such isolation may also be associated with poor
orientation and an unwillingness to petition for, and receive, both long足 term development aid and short-term relief.
The Scale of Enquiry The factors predisposing a community to famine vary, depending upon the "scale of enquiry" (see the oblique
dimension, Figure 6).
For
example, factors that affect famine vulnerability at a national scale, such as inflation or the lack of agricultural statistics for famine fore足 casting, may not operate at the regional scale, where droughts or floods may be more directly important.
The converse will also hold.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50 At any given scale of enquiry, famine vulnerability may be affected by factors operating at other scales:
a village dependent upon diesel-
driven low lift pumps for accessing irrigation water from a nearby river might be vulnerable to famine because of factors at the international scale such as a worldwide increase in petroleum prices or flooding initiated by widespread deforestation upstream in a neighboring estate. This thesis is argued at the regional scale (Figure 6), but even within a region, famine vulnerability may vary.
One river bank may be
vulnerable because of river erosion, the other because of burial under sterile sand deposits.
Individual villages within the same region may be
more susceptible than others because the settlers are new and not adapted to the area’s problems or because they belong to an underprivileged group. Even within a village, certain households from a particular caste, or households who lost their land as dowries for the marriages of several daughters, may have a particular proclivity towards famine.
The Evolution of a Famine Famines seldom occur as sudden acts of God.
People may become aware
of famine symptoms suddenly, but the "evolution of a famine" (see the horizontal dimension, Figure 6), usually occurs over a number of years. Some factors which are inherent in an area may persistently predispose it to famine.
Under such normal conditions may be included such
relatively static factors as poverty, drought, or a poor transport net work.
These factors, although classed as "normal," may improve or worsen
through time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51 A process or group of processes may begin to operate or to increase their intensity so that they exacerbate or strain the normal conditions: e.g., a significant change in atmospheric circulation patterns may increase an area’s propensity to drought, a particular government policy may in crease the proportion of rural poor who have no land, or a nation’s administration system may decay because of prolonged failure to pay civil servants’ salaries. Eventually an event or succession of discrete events triggers or precipitates an awareness of famine conditions:
a flood occurs or civil
disturbances break out and the resulting disruption brings the famine conditions into focus.
Examples of Famine Vulnerability in Different Countries The concept of famine vulnerability can perhaps best be explained by examining the evaluation of famine in different countries of the xrorld at different times.
The examples illustrate the basic unity of the con
cept, but they also stress the variations that have occurred in different countries at different times.
They emphasize the necessity of gaining
a synthetic understanding of famine vulnerability in any country such as Bangladesh before initiating famine remedial planning. The Irish famine of 1846-51 was chosen to illustrate a relatively recent, but nevertheless historical famine in what is now generally con sidered a non-famine-prone area of the world.
The Shantung famine of
1925, although now also a historical event, bears many similarities to the Bangladesh famine of 1974-75, thus implying that many of the preven tive measures used in the aftermath of the Chinese famine of 1925 might be applicable to the Bangladesh situation today.
Ethiopia and the famine
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52 of 1973-74 illustrates a case contemporary with that of Bangladesh, but one set in a different environmental and cultural context.
Finally, India,
once united with Bangladesh under British rule, is considered.
Bihar,
formerly part of Bengal, and the famine of 1967 are examined to exemplify a case of famine amelioration which was judged to be successful (Berg, 1973; Verghese, 1967).
Ireland and the Great Famine of 1846-51 Production.
Ireland’s agricultural production after the end of the
Napoleonic wars was affected by pressures from three sides:
the land
became increasingly subdivided, the ruling British government based in Whitehall continued to view the country as a peripheral irrelevance from which rents could be collected, and the postwar depression in the grain market meant a decline in grain acreages and the increase of a vulnerable monocrop, the potato. Early marriage and the tradition of subdivision among sons at the time of marriage had increased the percentage of agriculturalists holding less than 5 acres of land to 45 percent by the time of the 1841 census. The figure was only 37 percent in the east of Ireland in counties like Leinster, but in the western counties like Connaught it was over 64 percent (Edwards and Williams, 1956).
These small holdings, often of two
acres or less, in the western counties of Mayo, Galway, Sligo, and Ros common, were among the smallest acreages in Europe.
Lord Devon's Royal
Commission enquiring into the condition of the rural poor quoted details of one estate in order to illustrate the extreme congestion. It consisted of 10,966 acres and supported 10,129 persons. These were divided into two classes: tenants of whom there
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53 were 1,064 with an average holding of almost 10 acres and cottiers of whom 886 shared 536 acres and 256 had no land at all. The density of population was over 600 per square mile in an area dependent only on farming and the return to the owner in rent was £14,750 of which the tenants paid £12,165, an average of £1 2s 2d per acre and the cottiers £2,585 an average of £4 16s 6d which was discharged mainly in labor. (Poor Inquiry, 1835) These small plots increasingly lost their fertility, as the potato, with its large and quick return from a small acreage, was grown on the same land year after year.
The little straw available was used for thatch,
thus causing a dearth of straw for composting.
The main source of
fertilizer was the betail process which involved burning the sod in May and turning the ashes into lazy beds (6? wide planting beds flanked by trenches to carry away excess run-off) for planting the first crop. Areas growing a second crop on the same land in July still depended upon the betail of May. The inland areas of the South and West were different from the oats and cattle mixed farming of the Scots-Irish in northwest Ireland, different from the wheat lands of the drier and British market oriented East, and different from the areas with coastal fisheries,
The South and West had
developed a highly vulnerable mono-crop economy centered on the potato. A few pigs, calves and hens were also raised but they were only used to pay the rent and did not form part of the subsistence diet. tradition suggests that:
Irish oral
"meat was only eaten on Christmas Day and at
New Year," "thumbnails were kept at a certain length to peel potatoes," "those children who went to school would carry cold potatoes for lunch," and "the fishermen had specially made woollen stocking bags to keep sup plies of mashed potatoes warm" (Edwards and Williams, 1956).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54 Serious blights had attacked this highly susceptible monoculture several times in the early 1800s, most notably in 1836, 1837, and 1839 (Woodham-Smith, 1963), but none was so persistent or so widespread as the outbreak of phytophera infestans, a microscopic organism which struck the Irish potato crop repetitively in October 1845, 1846 and again, after a year's respite, in 1848.
The Marquis of Lansdowne, reporting to the
House of Commons in January 1847, estimated the loss of 9-10 million tons of potatoes, the equivalent of 1-1/2 million acres (Edwards and Williams, 1956). Local supplementation of agricultural production was limited. and mineral resources were meager and industry was faltering. development was confined to the northeast.
Coal
Industrial
Local inshore fishing was
restricted by the unseaworthiness of the Irish coracle.
In the summer
months some men would leave to crew the English fishing boats lying off the coast.
Others would stay on the farm but tickle for eels and trout
in the local rivers and loughs. The atmosphere of misery and lack of opportunity at home induced seasonal migration in search of employment elsewhere.
Police counts at
the ports enumerated 57,651 such seasonal migrants in 1841 (HMSO, 1843), heading for the ports of Glasgow, Liverpool, Bristol and London, and thence to areas beyond as they followed the calendar for harvesting labor.
People also migrated permanently to Britain, the United States,
Canada, and Australia, and thus set up bridgeheads to help the estimated two million who left Ireland during the famine period. After the famine broke out, internal migration in Ireland was restricted by the governments, thus confining the distribution of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55 statutory relief to permanent residents of the locality.
There was little
overseas emigration from the intense famine areas of Galway and Clare where mortality was highest (Figure 7) because of the great cost (e.g., 60 shillings to North America).
Galway and Clare, the two counties with
the highest mortality rates, were also the two where at least 50 percent of the population were paupers (recipients of poor law relief) during at least one of the famine years (Cousens, 1960a). emigrate because of poverty.
They were unable to
In contrast the rates of emigration from
north central Ireland (Figure 7), on the periphery of the worst affected area, were high because the peasants, being less poor, were better able to move in response to the famine conditions, and because there was a stronger policy to evict small tenants in those counties (Cousens, 1960b). The impact of this outmigration on vulnerability is unknown because there is no record of the death rates of those who migrated.
If these death
rates of the emigrants were as high or higher than the levels of over 15 percent found in Galway and Clare, then emigration merely transferred rather than lessened famine vulnerability.
Efficiency in Using Production.
The population of Ireland in足
creased by 100 percent between 1770 and 1841 to an estimated 8.5 million at the outbreak of the famine (Connell, 1950).
Early marriages and the
large families supported by the proliferation of the potato crop were primarily responsible for this increase (Wright and Davies, 1974).
On
the eve of the famine, Ireland was more heavily populated than ever before or ever since.
The population was rural, poor, exploited by
prejudicial laws, and susceptible to disease.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
(Standandardised agair
Figure 7.
R e g i o n a l D e m o g r a p h y of the Irish Fa m i n e 1 8 4 6 - 1 8 5 1 (Adapted f r o m C o u s e n s (1960a & 1960b))
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57 The 1841 census registered 80 percent of Ireland's population as living outside of towns or villages. cabins with only a single room.
The poor lived in windowless, mud
Lord Devon's Royal Commission found
"more than 20 percent of the people had very little work and lived in a state of abject poverty" (Poor Inquiry, 1835).
It described a "slough of
unemployment" and pointed out that even "full employment was only eight pence a day for a labourer."
This same poverty is highlighted by
Thackeray in his 'Irish Sketches' describing the "swarms of beggars" and "the number of idling servants" (Thackeray, 1878), The Penal Laws which differentially prohibited Roman Catholics from inheriting property, purchasing land, voting in elections and holding government positions, had been officially abolished by 1838, but the spirit of the laws continued (Beckett, 1966).
Although direct tithes
levied on the Catholics by the Protestant Church of Ireland were abolished, the substitute tithe imposed on the landowners was merely passed on to the Catholics in the form of increased rents.
Beckett neatly clarifies
this anti-Catholic taxation practice by suggesting that "Heresy changed from first hand to second hand" (Beckett, 1966).
Although the essential
purpose of the Penal laws was not to destroy Roman Catholicism, they made sure that Catholics were kept in a position of social, economic, and political inferiority (Curtis, 1964),
As a result, almost all the
landlord class were Protestant by the beginning of the nineteenth century, despite the Protestants being only a 10 percent minority. Increasing rural discontent, as well as 'agrarian outrages' in response to tenant evictions vainly attempted to halt the force which "in the famine years ordained that the small farmer should starve while
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58 in his garden the stock of oats bore the landlord's mark and awaited removal for export" (Edvrards and Williams, 1956).
The landlords, often
alien and absentee, increased rents through middlemen while the landless laborers and insecure cottiers, dependent on their half acre or one acre plot and the landlord's whim, "yearned for ownership or sure tenure" (Freeman, 1957). The potato appears to have provided the necessary calories for the poor in normal times, but as the famine evolved, associated diseases flourished on bodies weakened by the monotonous and decreasing diet of starch.
Scurvy was rife and ruithfola (the 'bloody flux' of bacillary
dysentery) further weakened the poor (Edwards and Williams, 1956). Crowds gathering for relief payments and destitutes huddled in the cabin suburbs (dwellings clustered along the roadsides, analogous to the bustees of Dacca) were the breeding grounds of famine fever (typhus) and relapsing fever.
The 1851 census attributes 20,402 of the famine deaths
to 'starvation', 22,384 to 'dropsy', and 318,805 to the cover-all of 'fever, dysentery, and diarrhea'.
Storage and Savings.
Even in the non-famine years, the Irish Poor
Commission of 1835 estimated that about 2.4 million went hungry in the two months of July and August before the new potato crop was ready.
The
blight that precipitated the famine not only attacked repeatedly over wide areas for three, and in some places four years, but also affected the potatoes stored in pits, barns, and houses, as well as those in the field.
It thus attacked not only the food supply, but also the seed
supply and there are tales of men using goose quills to pick out the eyes of the few unblighted potatoes to eke out the number of seedlings.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59 The poor had but little savings to draw upon.
Even if they had had
land, they would have found it virtually unsaleable as even the larger farmers were selling off and leaving because tenants were unable to pay their rents.
All that the poor had to sell x?ere their bedding, a few
stools, and the vessel in which they normally boiled their potatoes. Attempts to draw on other storage ranged from robbery to seeking alternative foodstuffs from the neighborhood.
Edwards and Williams
(1956) give details from Irish oral history of "sheep stealing, local attacks on meat depots and food wagons as well as night raids on potato and turnip crops."
The Irish had to build bothains or watch-houses
reminiscent of the machas of Bengal) from which they guarded their crops at harvest time.
With the dearth of potatoes, cabbage and turnips were
increasingly used as substitute foods.
The folk calendar incorporated
the term blian na dturnaipi (year of the turnip). acceptable food sources were also tapped.
Other even less
Many of the poor turned to
dogs, horses, asses, dried snails, frogs and hedgehogs as sources of protein and to plants such as fern roots or dandelion leaves, as well as tree bark and nettles to eke out their meager diet. Perhaps one of the most basic although less tangible storage mechanisms which broke down was the communal spirit of the rural people. Hardship and hunger meant a preoccupation with the struggle to survive. Charity and sharing appear to have broken down almost completely as the famine intensified (Edwards and Williams, 1956).
Degree of Isolation.
Physical isolation comparable with the con足
temporary Ethiopian and Sahelian contexts was not a cause of the Irish
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
famine.
The famine persisted for almost seven years, but no part of
Ireland was more than 200 miles from Dublin port and there was already a good system of roads to every village.
Although most movement was by
cart as there were only a dozen miles of railway in Ireland in 1841, it would seem more likely that the good road transport network actually exacerbated the Irish famine.
Certainly the southern and western counties
where the famine was most severe, were most isolated from Britain, but even in 1847, schooners laden with 200 tons of wheat were still leaving Westport in County Mayo (Brody, 1973). Despite rioting to stop this export, the carts continued to carry away the grain of absentee landlords.
Unknowing politicians in Whitehall,
briefed by a Protestant relief commission with only one Catholic member, eloquently continued to support a "laissez-faire" system.
They believed
that grain would flow in response to demand, but in Ireland there were no merchants to promote such a flow, and there was no money among the famine stricken to create such a demand.
Southwest Ireland’s predicament of
social and economic isolation from nineteenth century Britain was articulated by Sir Randolph Routh, the Commissary General, in the state ment quoted by Aykroyd (1975), "You cannot answer the cry of want by a quotation from political economy."
Similarly, cries of want are seldom
understood from afar.
China and the Shantung Famine of 1925 Production.
The northern boundary of Shantung Province
abuts on
the Gulf of Chihli over 230 miles north-north-west of where the southern boundary reaches the Yellow Sea south of the Shantung Peninsula.
This
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
whole province is criss-crossed by the former channels left of the HwangHo.
Aside from over 22 documented smaller changes of course since 2297
B.C., the Hwang-Ho underwent four major course changes within historical records; in 1324, 1852, 1887 and 1889 (Mallory, 1926). resulted.
Each time famine
Floods and droughts were equivalent to famine in Shantung.
The
historical compilations of Alexander Hosie (1878), Yao Shan-Yu (1943) and Co-ching Chu (1926) included only the records of catastrophes like flood and drought.
Famines were understood to be their inevitable consequence.
Of China's 18 provinces, Shan-Yu found Shantung to be the sixth most flood affected; recording 310 floods during the period A.D. 206-1911. Similarly, he found it was the eighth most drought affected province with 519 droughts during the same period.
In the Manchu dynasty from
1644 to 1847 and from 1861 to 1900, the area was recorded as having suf足 fered an average of 27.7 floods per century (Chu, 1926).
The province is
fortunately north of the typhoons and tidal waves so commonly suffered in the coastal districts south of Shanghai.
It is slightly, but not
safely, east of the Lanchow-Kunming fault zone, the main earthquake area in China; however, the active Tancheng-Lukiang fault and other minor faults criss-cross the province from north-north-east to south-south-west (Wilson, 1972).
Shantung lies east of the main locust prone areas, but
large swarms were recorded in A.D. 569, 953, and 991, and in 1937 Buck still noted them as a serious problem, especially for the millet crops (Buck, 1937). Superficially, the 1925 Shantung famine might be diagnosed as yet another flood-induced famine similar to that caused by the Hwang-Ho flood of 1920 in which several hundred thousands died (Mallory, 1926).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62 Certainly during.June, July, and August 1925, the summer monsoon run-off from the 88,000 square mile catchment area of Hwang-Ho, flooded 800 square miles of the south Chihli plain and damaged crops valued at over $20 million.
Burial of agricultural land by sand, and the slow evaporation
of the water from the poorly drained lowlands (Freeman, 1922), not only destroyed the millet and sorghum for the September and October harvest足 ing, but also prevented the planting of the winter wheat. To indict river flooding as the cause of the 1925 famine, however, disregards the negligence which allowed the flood.
In the long term,
Lowdermilk (1924) has stressed the problems of deforestation in North China and the resulting increase in sedimentation downstream.
The
Hwang-Ho flowed "on" rather than "through" the plains, with its bed often raised 9 feet above the surrounding plain between natural and artificial levees (Turner, 1926).
In the short term, the particular flood of 1925
was due to the breakdown of civil administration.
Under the old regime
of the Manchu dynasty, the river conservancy was basin-wide and unified under the control of one man.
After Sun Yat Sen's revolution of 1912,
the water resources of each province were placed under the charge of the provincial governor.
Rather than act as steward of his territory, he
used it as a resource to fund his wars on the neighboring governors. Dike maintenance was given low priority.
Twenty million dollars of crop
damage resulted from the 1925 Shantung flood, but the flooding resulted from a breach in the southern dike of the Hwang-Ho which could have been prevented by $20,000 of dike maintenance the previous year (Mallory, 1926).
In 1925 the river was not even at an unusually high level when
the disaster occurred.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
Efficiency in Using Production.
The effect of flooding in Shantung
was especially great because of the dense population of the North China Plain.
Although Roxby (1925) cited a figure of 500 persons per square
mile for Shantung Province as a whole, dot distribution maps published by the Continuation Committee (1922) suggested densities three times higher in the flood plain area.
Birth rates were high, encouraged by
early marriage and by the need for male descendants to perform the rituals of ancestor worship and to serve as an annuity for the parents’ old age.
The China International Famine Relief Commission (CIFRC) survey
of 240 villages in east and north China in the early 1920s revealed an average family size of 5.7, similar to that of Bangladesh today. Attempts to alleviate population pressure by encouraging migration to the frontier lands of
Manchuria had stubbornly been resisted.
Chinese peasants' retort
was, "Whowould there be to look after the
graves of our ancestors?"
The
Keith Buchanan ascribed this North Chinese
immobility under stress to . . . cultural and social factors: the attachment to family and village, the veneration of the ancestors and the difficulties of language. Partly stability has been imposed by poverty for the peasant lived close to the margins of existence: his savings were meagre and easily wiped out by natural or man-made calamities— and an im poverished family was scarcely likely to be tempted out into the unknown to break in land of doubtful agricultural value. Famine might force a family out, but except when driven by iron necessity the peasant clung to his patch of land, to the known landscape he and his ancestors had created. (Buchanan, 1970) Even in times of normal production, it was difficult for such high densities to support themselves on their meager diet of grain and fresh vegetables in the summer and salt cabbage in the winter.
Mallory
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64 (1926) notes the age-old struggle for existence suggested by the Chinese equivalent of "How do you do?" lema "Have you eaten?")
It is
vj^ ?
ij.
(Jyr lefann-
The disease mix of pre-Maoist China both
exacerbated and reflected the malnutrition and poverty:
cholera and
parasitic worms lowered the effective diet still further; and plague, typhus and kalaazar resulted from the fleas, lice and sand flies con足 comitant with the poor living conditions (Sorre, 1943).
Ernest P.
Bickwell, reporting for the American Red Cross Commission to China (1929), stated, "With the lack of economic margin it is inevitable that large numbers should fall below the existence margin."
The CIFRC survey
reported 50 percent of the interviewed families were receiving a total income from all sources of less than $25 per year.
That would mean that
there was less than $5 per year per person for fuel, clothing, shelter and food. The warlords levied crushing taxation on the provinces.
Even when
they were not at war, which was associated with the burning of crops and looting of stores, the countryside still had to support the private armies at an estimated cost of $200 per soldier per
year (Mallory, 1926).
According to the China Yearbook of 1924, there were
approximately 1,404,000
men in the various armies of China.
As these aimies were vanquished or
disbanded, man}' ex-soldiers, as well as the landless peasants, became bandits in what Buck (1937) termed "a fellowship of the dispossessed." Mallory estimated the number of bandits to he In the tens of thousands.
Savings and Storage.
Even in 'normal' times there was little or
no margin of
safety infood grain availability.
that in 1925
there were few reserve stocks to fall back on.
It is not surprising The chronic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65 situation prior to the acute famine of 1925 was expressed by Edward Thomas Williams (1923) when he stated, "A careful and experienced observer estimates that three millions of people die annually in China for lack of proper sustenance."
In Manchu times the government had
managed to maintain reserves of food for famine years, but they had been built up by forced rural procurement and kept only in the walled cities. During the 1912 revolution the contents of these urban public granaries had been sold by the officials and had not been restocked. The ability to draw on credit had also contracted. after 1912 had progressively weakened foreign aid links.
Warring factions Failure to
repay loans and sometimes even interest on China's railways, built between 1878 and 1920 predominantly with foreign aid, had further lowered China's credit worthiness.
Within China itself, rural credit was not
only inadequate, but interest rates increased in proportion to the degree of poverty.
China's banks were available only to people living
in, or near, the towns.
Even in the towns, normal bank interest rates
were 24-36 percent and they increased to over 100 percent during famines. There was no coordination of reserves with neighboring surplus regions. For the rural farmer, the local grain shop, again with no outside funds, would provide a loan against grain in the field.
If there was
no grain in the field, however, the farmer was left with nothing to mortgage.
The village pawnbroker would take extra clothing, furniture
or curios, but in famine years even this credit would dry up in response to the large-scale pawning of such items.
Moneylenders, who in normal
times lent at exorbitant rates to the poorest against their ability to work, simply moved out in famine years.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
There was no community cooperative society to draw on in times of need or to carry out preventive diking and irrigation during the long winter 'slack season'.
The primary organizational group was the family.
Each family stood by itself with no feeling of responsibility towards either the community or the state.
Degree of Isolation.
In 1926, five of China's twenty-six provinces
were still without railways.
There were only a few thousand miles of
railroad within China compared with a quarter of a million miles in the U.S.A.
Shantung was one of the provinces that did have a railway, but
the few lines that were there were in the hands of the provincial governors.
Exorbitant likin (transit taxes) were imposed at frequent
intervals along the main routes.
Rolling stock were hoarded for use in
military emergencies, but not for times of famine.
Aykroyd (1975)
described the inadequately stored piles of grain that used to lie rotting for want of transport at the railway stations.
The destruction
of bridges by bandits had reduced the railroads to a fraction of their total capacity.
The banditry problem was highlighted by a relief
worker's quotation in the Appendix to the Report of the American Red Cross Commission in China (1929): To have all grain accompanied on rail and trail by sufficient men to guard against theft would mean an unbearable expense even if otherwise practicable. And I know of no way of recruiting a body of men who would be more than a small percent more trust足 worthy than the teamsters themselves. Although the alternative, water transport, was augmented by the Chinese canal system, such transport was slow. suitable for power driven vessels.
There were few waterways
Most of the relief grain imports came
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67 from abroad or Manchuria which meant that junks had to ply upstream against the current. Whether from railway station or port, the feeder lines were only tracks, because China was virtually without roads.
Transport capacity
was limited by the capacity of the carts, oxen, donkeys, wheelbarrows, or human carriers.
Even these small scale supply lines were prey to
brigands (Mallory, 1926 Before the 1925 famine in Shantung, the landscape (Figure 8) and the sequence of events, a.
the run-down socio-economic conditions in a densely settled deltaic area recently disrupted by civil strife,
b.
the flooding possibly affected by deforestation upstream as well as poor conservation measures within the area,
c.
the disruption of agriculture and the dislocation of people by the slow drainage of flood waters and the dumping of sediments, and
d.
the difficulties of access away from the main transportation routes,
were all similar to the conditions in Bangladesh prior to the 1974 Bangladesh famine over half a century later.
Ethiopia and the Famine of 1973-74 Production.
Sixty-five percent of Ethiopia’s 1.2 million square
kilometers of land is classed as agricultural, giving a deceptively high figure of seven acres of agricultural land per capita (USAID, 1974), The predominantly pastoral farming economy is estimated to have supported as many as 26 million cattle, 12
million sheep, 11 million goats, 1.4
million horses, and a similar number of mules prior to the severe
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-jif
1ir
? CHINA - LANDSCAPE OF FAMINE’ ' SSfr
Shantung 1925
ttMHHI
■ iMnMi ii M
Flooding and the loss of possessionsj
Erosion and the breaching of em bankm ents V/r-P
•
k-
'4 *
Sand deposits and the disruption of agriculture
Tem porary m igration and em bankl^ ^ 1^ ^ j J
(After M&llory,1926)
Bruea Currey.
Figure 8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69 droughts of the early 1970s and the famine of 1973-74,
There were
probably about 3.8 million donkeys and a million camels as well.
Only
7-9 percent of the land was actually cultivated for one and sometimes two crop seasons.
They grew teff (Poa abysinnica), wheat, and barley
in the cool highland zone on the 500,000 square kilometer dissected central plateau.
At lower attitudes corn, sorghum and millet formed the
subsistence staple, and coffee and ensete were grown for cash.
Beneath
1,000 meters in the dry peripheral lowlands which receive less than 300 millimeters of precipitation per year, desert scrub predominated except for the tropical plants of the river beds and oases. History describes a prosperous agrarian life amidst a land of rich cedar forests before the advent of the nineteenth century (Hiwet, 1975). Alvarez, in a narrative from the Portuguese embassy in Abyssinia in 1881, depicts a prosperous agricultural economy: Honored guest, do not be amazed, because in the years that we harvest little we gather enough for three years plenty in the country; and if it were not for the multitude of locusts and hail which sometimes do great damage we should not sow the half of what we sow because so much remains that it cannot be believed; so it is sowing wheat, or barley, lentils, pulse or any other seed. We sow so much with the hope that even if the said plagues should come, some would be spoiled the year before, our produce is in such manner abundant that we have no scarcity. (Alvarez, 1881) Unless Alvarez had led an extremely sheltered life in the diplomatic enclave, food supplies appear to have been abundant only seven years prior to the great Ethiopian Famine of 1888-92 (Pankhurst, 1966). then, however, surpluses have declined.
Since
Increased grazing, charcoal
burning and fuel requirements have left only 4 percent of the land forested (Eckholm, 1976).
Winds and flash floods, particularly in the
relatively fertile reddish brown clays and clay loams in the south west
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
highlands, erode the nation's potentially productive areas.
The popula
tion growth rate of 2.6 percent per year coupled with this degradation of existing farmland was forcing people to graze their stock and cultivate land in increasingly marginal areas in the late 1960s.
Simultaneously,
Ethiopia was achieving one of the highest gross domestic product growth rates of the 25 least developed countries (I. D. A., 1974), as the govern ment pursued a policy of cash crop plantation promotion at the expense of dry season grazing land. With its wide gamut of latitudinal (4-18° north) and altitudinal (0-4,000+ meters) differences, Ethiopia has not only a wide range of production opportunities, but also a wide range of possible environmental insults.
The country has a long history of natural disasters:
floods,
droughts, frosts, erosion, earthquakes, army worms and locusts (Pankhurst, 1961).
In the mid-1960s, drought became a recurring problem because the
circumpolar vortex expanded and displaced the latitudinal climatic belts towards the equator (Lamb, 1973).
These disturbances particularly
affected the onset times, regimes, and total amounts of rainfall during the belg season (January to May), when frontal rainfall occurs in association with the low pressure air mass over the north Atlantic. They form the "short” or minor rains in the bimodal regime of the central highlands, but they are the main rains in the monomodal regime of the south and eastern lowlands.
In the areas where the belg rains are
dominant, three successive years of late onset times in the early 1970s critically affected both germination and crops for the subsistence cultivators, as well as abnormally extending the dry season for the pastoralists.
Even in the higher and more westerly areas, where the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71 belg rains are minor compared with the earlier kremt season rains (June to September), a failure of the belg rains nevertheless reduced the meker (harvest resulting from the kremt rains).
A failure of the belg rains
meant that not only was no seed grain available to sow for the kremt crop, but also no food was available to feed the draught animals necessary for tilling the kremt crop.
If lack of rain prevented the land being tilled
during the belg season, the soil became too indurated to till later at the beginning of the kremt season (FNSSS, 1975).
The severe droughts of
1972 and 1973 were exacerbated in some areas, e.g., Koreb, by swarms of Cyrtacanthacrin tataricha, a grasshopper which severely damaged the leaves of the chickpea and pea seedlings (Wood, 1976). Over 4.2 million people, mostly tenants, left the northern provinces of Wollo and Tigre, where the drought hit first, to seek food, land, and work elsewhere.
Unfortunately, these populations moved south as the
effects of the drought also moved south.
They moved into areas like
Eritrea and thus merely aggravated an existing deficit area.
They moved
in search of relief towards the urban areas, and by 1974 it was estimated that there were 30,000 destitutes in relief shelters, but a bungled relief operation provided little aid, and what there was came too late (USAID, 1974).
Efficiency in Using Production.
Of the 25.2 million people who
live on Ethiopia's subsistence and pastoral economy, 50 percent are under 14 years of age and 4 percent are over 55.
This leaves only 11.2 million
in the working population, and 10 million of them have never been to school (USAID, 1974).
The rate of population growth is high (2.6 per足
cent per annum) and might be still higher were it not for the high infant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72 mortality rate caused by successive epidemics of measles, whooping cough, and respiratory disease.
Although smallpox is on the wane, trypanosomia足
sis still saps the strength of the population in the potentially produc足 tive and wetter provinces of the southwest.
High rates of diarrheal
disease hinder food absorption, and in the population concentrations in the relief shelters (e.g., 6,000 in 12 camps in Wollo) even cholera was suspected (Belete, Gebre-Medhin, Hailemariam, Maafi, Vahlquist, and Wolde-Gebriel, 1976).
Disease affects the animal population too, so
that about two million sheep and goats succumb annually.
About 1.2
million of the productive work oxen are also decimated by disease each year. Of the human population who are fit to work, 95 percent are sub足 sistence farmers. farmed.
One-half of Ethiopia's agricultural land is tenant
The tenants pay 33.3 percent to 50 percent of their production
as rent to landlords who are often absentee.
Wood's (1974) conceptual
model (Figure 9) illustrates the inherent vulnerability of tenant farmers as compared with owner cultivators in Ethiopia.
Assuming a normal yield
of 100 percent and a persistent decline from 100 percent to zero over 20 years because of continuously worsening drought, he points out that the tenant has to draw from his savings after 2.5 years to pay his taxes and maintain his consumption level.
The owner cultivator, on the other hand,
does not have to pay rent which is always taken as 50 percent of the annual yield.
Wood proposes that the owner cultivator has to draw on his
savings only after 11 years.
The letter's position is even more impreg足
nable than Wood suggests, however, because in the 'normal year' prior to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE IMPACT OF DROUGHT ON AN ETHIOPIAN FARMER The contrast between tenants and ow ner cultivators (After Wood. 1976)
OWNER CULTIVATORS
TENANTS 100.
100 93 80
H
m 50. 46 33
0.
Figure 9.
74 outbreak of drought, when the yield was 100 percent, he had almost 60 percent surplus compared with the meager 10 percent surplus of the tenant counterpart.
The surplus is presumably above "normal" in some years and
thus provides a large investment against the years with poor yields for the owner cultivator.
The owner cultivator has the further flexibility
of having the option of taking on a tenant who will guarantee him 50 percent of whatever the yield as rent. The repressive land tenure system gives no security of tenure and there is no reimbursement for capital improvements. come only through allegiance to the feudal hierarchy.
Any reimbursements Any decision could
only be ratified from the top of the feudal system, and that system it足 self was clogged by a top-heavy and centralized bureaucracy until 1975. Even in the early 1970s, with the worsening food situation in some areas of the country, and the decline of world food reserves (Walters, 1975), the government was still promoting the export of livestock products to raise money for imperial projects such as defense (George, 1976).
Storage and Savings.
The succession of droughts, coupled with
increasing civil disturbances in the early 1970s, reduced what internal buffer stocks there were in 1971.
Buffer stocks were always few, and
the country was usually prone to a two-month "hungry season."
In
addition, there were the inherent problems of harvest, pest, and storage losses (Last, 1975), as well as a wave of hoarding and profiteering (USAID, 1974).
Ethiopia's low stocks coincided with low grain stocks
in the international system as a result of the siphoning effect of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75 United States-USSR grain deal in response to the latter's drought in 1972 (Sanderson, 1975). The main savings of the rural population were their animal herds. Although not usually used for barter, but rather for draught animals in the highlands and for milk in the lowlands, the animals were sold off in a market glut, and hence at a low price.
The nomads had already lost 80
percent of their cattle, which were particularly prone to the drought conditions, even before this final sale to raise money.
The somewhat
drought resistant sheep were also reduced by 50 percent and the more hardy camel and goat herds were reduced by 30 percent.
Although animal herds
are a buffer against scarcity, once they are gone, they are lost for several years.
They cannot be replenished the next season like crops.
The big cattle holders who stood to suffer the greatest loss were probably the rural rich, but in the feudal system, the poor traditionally suffer first in an attempt to keep the cattle of the rich alive (Wisner et al., 1976).
Degree of Isolation.
Ethiopia exemplifies many of the problems of
physical, social and political isolation associated with famine vulner足 ability.
The ruling Amhara minority constitute only 25 percent of the
total population.
The Amhara are a group of Coptic Christians who
staffed the highly centralized bureaucracy in Addis Abaha. to show indifference to signs of approaching famine.
They appeared
They were unaware
of the situation in several of the politically isolated provinces such as Eritrea.
In other areas, the lack of provincial administrative
capability hindered an efficient warning system.
Even when destitutes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76 camped in the grounds of the Imperial Palace, and after His Imperial Majesty had visited the affected rural areas, his comment to the outside world w as: We have said those who don't work starve . . . Each individual is responsible for his own misfortunes, his fate. (Fallaci, 1973) This politically isolationist stance by an emperor bears an uncanny similarity to the first chapter of the second article of the constitution of the USSR, viz, "He who does not work, neither shall he eat."
Apparent足
ly both imperialists and socialists have turned to their advantage St. Paul's words, "If any man will not work, neither let him eat" (Thessalonians. Chapter 3, Verse 10). In addition to the slow response by the Ethiopian Government, the international community also reacted slowly.
Dalby and Harrison-Church
(1973) suggest that their attention was diverted elsewhere to the simul足 taneous famine conditions in the West African Sahel.
An international
survey to evaluate the gravity of the situation in Ethiopia necessitated the use of three helicopters for six weeks to sample only part of one province (Seaman, Holt, and Rivers, 1974).
When international relief
did begin to arrive, it ran into the problems of regional isolation. Suez Canal had remained closed since the 1967 Arab Israeli war.
The
Grain
ships, when they finally did arrive, could unload at only three Red Sea ports:
Assab and Massawa in Ethiopia, and Djibouti in the French
territory of Affars and Issas. grain by May 1974.
Both Assab and Djibouti were full of
Only three to four freight cars per day were trans足
porting grain inland from Djibouti and 10-15 per day were needed.
Such
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
a large quantity of imported grain suddenly requiring transport inland would have needed about 250 trucks, but only 65 were operational.
The
private truckers were unwilling to operate to drought stricken areas because there was seldom any load for the return journey (USAID, 1974). Many of the affected areas (e.g., Wollo), lacked even a rudimentary road network, and supplies could be imported only by mule, which entailed a 5-7 day ride from the nearest road (IDA, 1974).
Even the United Nations
agencies and the aid missions of the various embassies have been chal足 lenged for their lack of response (Shepherd, 1975).
India and the Bihar Famine of 1967 Production.
India's traditional area of recurrent famines is the
predominantly sorghum and millet area on the Deccan lava plateau in the states of Gujarat and Rajasthan (Bhatia, 1967; Srivastava, 1968; Loveday, 1914; Geddes, 1970), Bihar's 67,196 square miles straddle not the millet, but the rice dominant areas of the Chota Nagpur Plateau and the Gangetic Plain.
India's agricultural production has been increasing erratically
since the 1950s, but the rate of growth had averaged 2.8 percent per year compared with China's much publicized 2.0 percent annual growth rate (Mellor, 1976).
The rate had been even higher in the 1960s in regions like
the Punjab, Gujarat and Andhra where the so-called 'green revolution' was initiated by the introduction of new varieties of high yielding wheats made possible by large fertilizer and irrigation inputs.
The high yield足
ing varieties in the rice areas, however, were at least two years behind those in the wheat areas.
Bihar's development was hindered by its agri足
cultural system being dependent on rice, by its lack of modern irrigation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78 development, by the disrepair of the old zamindarsT (landlords') irrigation schemes, and by its obsolescent land tenure customs.
Even in
normal times, its annual production of 7.5 million tons of food grains left a deficit of 1.3 million tons.
Its dependency on imports meant
that when India's total production fell by 30 million tons over the two drought years of 1965 and 1966 (Berg, 1973), Bihar with its monsoon dependent rice suffered even more. Drought is the predominant natural cause of famine in India (Passmore, 1951).
The 1965-66 drought severely affected seven of
India's twenty-two states including Bihar.
In 1966-67 the upland
bhadoi (September) rice, dependent on the heavy monsoon, was severely affected by drought on the South Bihar Plateau.
Conversely, snow melt
combined with the monsoon flooding of the Kosi, Gandak and Sone rivers, damaged crops in the North Gangetic Plain. resulting trikala (three fold shortage):
Singh (1975) described the food, fodder and water.
A
choukala (four fold shortage) would have been more appropriate because the floods, droughts, and crop damages almost certainly exacerbated the shortage of employment as well.
Even in 1958 Bhattacharjee (1961) had
found that 56 percent of Bihar's farm laborers were underemployed.
The
seasonal underemployment alone amounted to 25 percent, and crop failures like 1965 and 1966 merely extended the underemployment season.
That the
failings and shortages leading to the famine came not only from the drought is corroborated by the results of an unstructured questionnaire inquiring about the causes of the 1967 famine, administered to 84 key government and agency personnel, mostly in South Bihar (see Table 3; CIRTPC, 1969).
The table of responses shows that 81 percent of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79 Table 3 . _-C 4-U_ T C\~lC. DJ1___T7___ -•__ ^
_________________ Causes Mentioned__________ 1. Failure of the Failure of Failure of Failure of
Percentage of Respondents who mentioned______
rains (unspecified) the monsoon the Hathia rains theWinter rains
81 33 9 3
2. Neglect of agriculture (unspecified) Lack of adequate irrigational facilities Paucity of seeds/fertilizers Fragmentation of holdings Iniquitous land distribution
83 79 7 7 5
3.
20 13 5 5 3 3
Defective food policy (unspecified) Hoarding High prices of foodgrains Faulty distribution of foodgrains Weak procurement policy Zonal system of foodgrains
4.
Failure of the bureaucracy
31
5.
Lack of initiative and lethargy in people
31
6.
Population and growth
13
7.
Lack of purchasing power
11
8.
Lack of responsible leadership
9. Caste system
7 6
10.
Non-exploitation of mineral and forest reserves
6
11.
Exploitation by outsiders
6
12.
Exploitation of the rich by the poor
5
13.
Alcoholism
3
14.
Wasteful customs_______________________________________3_____
aFrom an unstructured questionnaire administered to 84 key government and agency personnel. Source:
CIRTPC, 1969.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80 respondents mentioned "failure of the rains" as a cause of the famine, but the percentage of the respondents mentioning the general "neglect of agriculture" as a cause of the famine, was also over eighty.
Efficiency in Using Production.
In 1966 there were 53 million
people in Bihar, and 91.57 percent of them were rural (Ramalingaswami, 1971).
The overall population density was 694 per square mile, a figure
double that of the Indian average, hut less than half the Bangladesh average.
In parts of north Bihar, the population density rose to over
900 persons per square mile. 70,000
Three-fourths of the cultivators in Bihar's
villages owned less than one percent of the land, and 30 percent
or more of them were landless laborers. The cultivators of Bihar were poor and insecure.
India's develop
ment plans in the late 1950s and 1960s had emphasized the growth of heavy indistry and agriculture had been underemphasized.
Whether intentionally
or not, the lack of growth in the agricultural sector meant that there was less drain on food supply because the poor were not given additional purchasing power which would have been immediately spent on food (Hellor, 1976).
In Bihar in particular, the failure of the Bihar Land Reforms
Act of 1950 as well as the amendments of 1954 and 1959 left the share croppers and landless in even greater insecurity than they had been in during the days of the zamindars (landlords) (Januzzi, 1977).
The land
less laborers had expected some occupancy rights to the land they cul tivated, but they merely received a drop in real wages.
The bataiders
(sharecroppers) found not the expected security of tenure for their land, but evictions and unemployment.
Concurrent with the increased polariza
tion of rural wealth (McNicoll, 1975) came agrarian discontent with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81 scions of the old zamindar class vying for power with the elected political panchayat (local council).
Storage and Savings.
With a $27 annual per capita income, the
lowest in India (Verghese, 1967), Bihar was stagnant and indebted even before the onset of the 1965-67 drought.
Successive insults of drought
lowered the calorie intake from 2,200 per capita in 1965 to 1,200 per capita in 1967 (Ramalingaswami, Deo, Guleria, Malhotra, Sood, Prakash, and Sinha, 1971).
Twenty percent of the population had calorie intakes
of less than 1,000 per day.
Berg (1973) describes the situation
poignantly, "The cupboard was bare." In efforts to raise money for food, land was alienated to the money lenders, loans from private agencies carried exorbitant interest rates, and the cattle which had not succumbed were sold off for less than 10 rupees each.
Women and children were seen sweeping in the dust to gather
fallen grain at railway sidings and grain go-downs (storage warehouses). Wild leaves were increasingly consumed, as were roots and tubers and the flower of the wild Mahua tree (Bassia latifolia) (Singh, 1975).
The 38
percent of the state’s population who were tribals and scheduled castes and hence generally landless, were especially vulnerable to the famine as many of their traditional forest lands had been deforested. What was there to draw upon?
The relatively fit were able to
migrate to the industrial areas in search of employment in the Jamshedpur iron and steel industry, and the bauxite, manganese, chromite and copper industries of southeast Bihar.
Ramalingaswami et al. (1971) recorded as
many as 40,000 people leaving one district.
Berg (1973) and Yerghese
(1967) have detailed how they were able to draw on the large underground
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82 aquifers which allowed the sinking of 500,000 hand dug, kutcha (unlined) wells and the brick lining of another 10,000 to make pucca (lined) wells. They were also able to call on a fleet of army water tanks and bullock carts. One of India’s greatest assets was its history of foreign trade relations.
India, unlike China, had maintained its links with developed
nations and food aid channels were open.
Despite the decline of U.S.
aid from its maximum of 25 percent of India's gross investment in 1965 (Mellow, 1976), the U.S. still contributed 20 percent of its 1965-66 wheat crop as aid during the Indian drought.
Some of the wheat which
was late for the 1965-66 drought arrived in time for the 1966-67 drought. Bihar, with a history of famine, had a famine and flood relief code revised in 1957 which established contingency plans for such scarcity and famine conditions.
After the drought of 1965, there was an estabÂ
lished local government relief machinery. had already gained experience.
The imported voluntary agencies
The transportation system had been
streamlined (Berg, 1973) and the food supply did not end at the railway station:
a distributive organization had been prepared which utilized
the existing primary school system.
Degree of Isolation.
After the previous year’s drought in 1965-66,
Bihar was partially sensitized to its rural problems.
In 1966-67 a
master control room and an information system, designed by management systems specialists, collated data from all districts on food stock levels, disease rates, deaths, water levels, crime rates, and food prices.
Radios and telephones were commandeered for use in broadcasting
news of the emergency, not only to the district, but also to the outside
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83 world.
Food flow in the area was relatively smooth.
The Indian govern
ment loaded and moved an average of seven trains a day, with fifty cars per train, and an average of 550 miles per journey.
Docking at the
rate of three a day, 600 ships deposited an average of two billion pounds of food per month (Berg, 1973). Although mortality in Bihar was kept apparently low (seven official starvation deaths in Bihar, 1966), famine symptoms as defined in this thesis undoubtedly existed.
There may have been hundreds, or by high
estimates, several thousand deaths caused by the famine (Berg, 1973). Undoubtedly a more severe famine was averted, but the containment cost was about one billion U.S. dollars, of which about $440 million was from international sources (Singh, 1975).
Without an evaluation of the long
term development effects of such massive relief aid, such a program should not be transferred wholesale to the Bangladesh situation. vades a region relatively slowly.
Drought per
It creates a famine vulnerability very
different from that of the rapid onset flooding found in many parts of Bangladesh. The application of the conceptual model of famine vulnerability to these four examples from Ireland, China, Ethiopia and India has illust r a t e d the increase in famine vulnerability, as a succession of factors have lessened the communities' "degrees of freedom" (Fuller, 1971) or opportunities for further development,
The four examples have specifi
cally illustrated: 1.
that a dichotomous classification of famine causes into "man-made" and "natural" is unilluminating;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.
that while "natural" hazards may precipitate famines, the hazards themselves are often man-induced; and
3.
that the areas of occurrence of natural disasters provide clues to the areas of occurrence of famines.
The progressive weakening in the "degrees of freedom" and the geographical association between natural hazards and famine areas is also found in Bangladesh, the national area studied in this dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER III THE FIELD AREA IN BANGLADESH
The Unstable Landscape And there shall arise after them seven years of famine, And all the plenty shall be forgotten in the land of Egypt; And the famine shall consume the land. (The Bible, King James Version, Genesis: Chapter 41 Verse 30.)
Just as Joseph was able to interpret the pharoah's allegorical dream to anticipate famine in Egypt, so he might have predicted famine for Bangladesh by interpreting the image taken from a heaven 9.5 kilo足 meters high by the Earth Resources and Technology Satellite (Figure 10). Bangladesh is a small country of 55,126 square miles, couched in a downwarp between the Himalyan piedmont to the north and the Bay of Bengal to the south.
The green vegetation hues of the Arakan Yomas mark
the north and eastern boundaries respectively, except for the two dis足 tricts of Chittagong and Chittagong Hill Tracts which intrude into the Eng (Dipterocarpus spp.) forests of Burma and Tripura State (India). The western boundary, delimited by the Radcliffe Award in 1947, bisects the old province of Bengal.
It is not so discernible because it cuts
south, dissecting natural regions with administrative callousness from the piedmont alluvial plains in the north and then across the Barind Tract.
It makes a foray eastwards to within 10 miles below the con足
tentious Farakka Barrage, which has affected surface water and salinity patterns downstream in Bangladesh.
Finally, the western boundary bi足
sects the high plain of the moribund delta before losing itself in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCE M AP
BANGLADESH Figure 10.
Bangla d e s h R e f erence M a p
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright ow ner Further reproduction prohibited without permission.
87 Sunderban Forest on the Bay of Bengal.
That encircling land boundary
from the Naf River in the southeast to the Raimangal River in the southwest is traced through a distance of 2,483 miles.
When added to
the intricate and ever-changing sea boundary of 445 miles, this provides almost 3,000 miles of border for potential smuggling and currency trafficking. Within the boundary, even from on high, the river courses stand out in the reverse color infra-red image, their white sand-flected braidings testifying to the dynamism of the river system.
The three main rivers,
Padma (Ganges), Jamuna (Brahmaputra) and Meghna, gather 1,073 million acre feet of snow melt and rain water, and an estimated 2.4 million tons of sediment load per year from their 600,000 square mile combined catchment area (IBRD, 1971).
The catchment includes the areas of rapid
deforestation (Eckholm, 1976) in the Himalayan foreland, northern Deccan, Himalayan-Tibetan Plateau and the northwest Arakan Range.
The rivers
lose their velocity because of the low gradients x^ithin the country (on the average:0.3-0.6 feet per mile; IBRD, 1971), and thus they drop their sediment load as riverine and deltaic chars (newly emerged lands), and as levees during the monsoon floods. By the time the satellite passes over again, 18 days later, the pattern of braiding visibly changes (Wingert, 1974).
The rivers have
been known to erode land over a mile wide on the river bank in one year (Coleman, 1969).
Despite its apparent aimlessness, the Jamuna has been
moving westwards from its first documented course, when it flanked the western end of the Garo-Khasi escarpment before turning south through Sylhet.
Later it followed the course of what is now called the old
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Brahmaputra channel, through Mymensingh, but between 1787 and 1830 it was diverted to its present steeper Jamuna channel during an abnormal flood.
Perversely, the Padma has countered the Jamuna's general west足
ward movement through history by moving its estuary eastwards.
It has
left behind a trail of decaying and moribund deltaic features, such as bils (marshes), khals (narrow channels) and oxbow lakes.
The Meghna is
not such a volatile river; in its floodplain course, it has the lowest gradient of all (less than 0.06 feet per mile downstream of Bhairab Bazaar).
The extremely low gradient permits the backwater effect of the
tides to penetrate as far north as Sylhet City.
It also forms a holding
area and bottleneck for flash floods from the Garo-Khasi escarpment to the north and the Arakans to the east. The isostatic depression occupied by the Sylhet haors (water filled depressions) holds flood overspill throughout the year (e.g., the satellite images, Figure 10, taken in February still shows large areas of standing water).
During the wet season, the Sylhet Basin becomes an
inland sea eighty miles long.
Other extensive water retaining depres足
sions are also evident in the Chalan Bil (swampy lake), in Rajshahi and Gopalganj Bil in Faridpur. The two outliers of higher land, the pleistocene-sediment-covered plateaux of the Barind and Madhupur tracts, are picked out by the lighter vegetation hues resulting from their good drainage characteristics.
The
intervening graben occupied by the alluvial Jamuna flood plain may have been formed by
tectonic activity associated with the forcing of the Indo-
Australian Plate below the Himalayan zone of the Eurasian Plate (Toksoz, 1975; Molnar and Taponnier, 1977).
This may be continuing, and causes
the ever-present, but poorly perceived, earthquake risk in Bangladesh today (Morgan and Mclntire, 1959; Wilson, 1972; Brammer, 1975).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The Bangladesh landscape presents to the naive newcomer a timeless and somnolent agrarian scene.
In fact, appearances are deceptive.
Bangladesh is characterized by change, including long-term tectonic move足 ments (Berry, 1976), medium- and short-term riverine shifts, seasonal vicissitudes of the monsoonal climate, as well as sudden environmental extremes or hazards such as floods or cyclones.
The Meshing of the Seasons In Bangladesh it is not the quantity of precipitation, but its dis足 tribution throughout the growing seasons, that is the more important criterion in determining famine vulnerability.
The country is well
endowed with precipitation and exotic run-off, but shifts from the average seasonal distribution of precipitation disrupt the agricultural, social and economic calendars (Figure 11). The annual isohyets curve eastwards around a low of 60" on the western border in southwest Rajshaai and Kushtia.
There, the precipita足
tion is only barely adequate for reliable wet season rice cultivation. The isohyets rise northeastwards and southeastwards to high rainfall areas of over 200" per annum in the hill areas of north Sylhet and southern Chittagong district.
About 80 percent of these annual totals
fall during the five monsoon months between the last week in May and the end of October.
The annual precipitation totals are relatively con足
stant (Conrad, 1941) as reflected by the low coefficient of variability1
Coefficient of Variability
standard deviation mean
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
for monsoon precipitation of 9, the lowest in the subcontinent (Das, 1972).
Although the variability in the total precipitation is low,
the critical "onset times" vary.
Awshomoi Brishti (untimely rainfall),
especially that associated with the southwest monsoon's onset which may vary from May 27th to June 23rd, affects crop regimes adversely, espe足 cially those of the staple paddy rice. Crop damage during the kharif (rainy) season from March to October, usually results from precipitation falling outside the region in the river catchment areas upstream.
Run-off from these different catchments
peaks at different times; the Meghna normally peaks first in May and June, the Brahmaputra in July, and the Ganges in August (IBRD, 1971). Changes in the rainfall regime in the upper catchments can alter the river hydrographs downstream.
The synchronization of the Brahmaputra and the
Ganges flood peaks, for example, caused the severe flooding around Dacca in 1974 (Ralph, 1975). The actual amount of rainfall, whether it be too much in the kharif (rainy) season, or too little in the boro (dry) season, is not the only environmental factor affecting crop yields.
The yields are
affected by many variables including the timing and intensity of the rainfall, the infiltration rate and holding capacity of the soil, the evaporation rate, and the stage of development of the crop (e.g., ger足 minating or flowering).
Agricultural drought may be initiated not only
by a shortage in precipitation but also by shortages in the supply of irrigation water which in turn may be caused by a shortage in fuel supplies for the irrigation pumps.
The aus^ (summer rice) normally
2 About 33 percent of the total paddy acreage.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92 sown between March and May and harvested between mid-July and midSeptember is generally an upland rice, and susceptible to the lack of chotta barsak (early rains).
A delay in these chotta barsak postpones
the sowing of the aus or hinders its germination.
This delay in turn
delays harvesting and increases the aus crop's vulnerability to unusually high or early floods in June or July.
Aus rice normally
reaches its maximum height of about 2.5 feet before June.
Any early flood
water rising above this height and keeping the crop submerged for more than five days, will lessen yields by 40-50 percent (IBRD, 1972). During the months of March-May, the aus crop and the relatively flood resistant jute crop are regularly damaged by the hail, high winds and heavy rain that are associated with the nor'wester line-squalls (Kal Baisakhi, or calamities in the mouth of Baisakh) as well as by violent cyclones. 3 Thirty percent of the aman aus in March and April.
(winter rice) crop is broadcast with the
This broadcast aman is thus susceptible to the
same hazards as the aus, but many of the long stemmed varieties, often 23 feet high and sometimes growing at 12 inches per day, can maintain sufficient "freeboard" to stay above all but the most rapidly rising floods.
Sixty-five percent of the aman is not broadcast, but is trans足
planted from nurseries between July and September after sowing in June and July.
This transplanted aman is very susceptible to submergence by
flood water; severe damages occur after only three days of submergence
^About 60 percent of the total paddy acreage.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
(IBRD, 1972).
Both aman crops are vulnerable to the recessions of pre足
cipitation during the monsoon.
The second cyclone season between
October and December may also lodge and rot the aman immediately prior to harvesting between November and January. The boro (dry season) rice crop is planted out between November and February and harvested between March and May.
It depends upon receding
flood water, residual soil moisture, or irrigation. crop is
This dry season
vulnerable both to failure of irrigation systems and to early
flash floods which inundate the low-lying boro habitats first. Just as cropping patterns are closely adjusted and dependent upon the monsoon climate, so seasonality of employment, mobility, food prices, and nutritional levels are dependent on the cropping regime (see the hypothetical multiple calendar, Figure 11).
If sowing or harvesting
is cancelled, delayed or reduced, it means increased underemployment or unemployment for the cultivators, especially the landless laborers. means less wages, whether in cash or kind.
It
Sowing and harvesting
necessitates large-scale seasonal movements of laborers to labor-deficient areas like Sylhet and Noakhali.
This means high short-term concentrations
of people in areas to which they are not accustomed, and hence, increased vulnerability in the event of a disaster.
Many food-deficient areas like
Faridpur depend on the "rice salaries" brought back from Sylhet by the harvesting labor. seasons.
Food prices and nutritional levels fluctuate with the
Even in good years, there is a lean period before the aman
harvesting:
foodstocks are at their lowest level and the prevalence of
kwashiorkor and marasmus in children under five increases by 250 percent between July and November (ICNND, 1966).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Awshomoi brishti (untimely rainfall) and ill-timed flooding dis locate the delicately balanced agricultural calendar, and thus lengthen the "lean season" if they do not precipitate a famine crisis.
Although
such a lack of synchronization among several components of the food system is difficult to document, they are emphasized in Bengali folk lore and proverbs such as those in Khonar Bauchon ("The Words of Khona," Bidyaratna, 1971).
ŠCSPT W3VJ5 V<3T
If the month of Jaistha is very dry And Ashar is rainy Then crops will be plentiful W
^
'3 '^ '
^ tS I '
^
<4^
iKlff c w Khona is saying, listen mankind, "If it does not rain in the months of Sravan and Bhadra And only rains during the day Keeping the skies full of stars at night Then your sorrows will begin."
Vbi?
iSigf w
Storms in the month of Chaitra And flood in Bhadra Will bring human skulls on the ground.
Poverty Bangladesh is a poor country, it is one of the forty poorest countries in the "fourth world" (Hansen, 1976) with a real income per capita of only $80 per annum in 1973 (World Bank; Group, 1975).
It is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95 among the poorest of the poor countries.
For example, while the poor
cultivators in West Bengal live in red brick-tiled and brick-walled houses, their counterparts in Bangladesh must make do with leaking thatch and cane.
Indeed, that real income per capita of $80 masks the dispari足
ties between the gold laden saris worn at functions at the Hotel Inter足 continental in Dacca, and reports of Bengalis themselves of the sharing by three village women of a single threadbare sari as their only modest covering in the extreme duspryha (difficult times) of 1974.
This level
of subsistence is described statistically by Alamgir (1974), who estimated that in 1968-69 there were already 50 million Bengalis living below the "poverty line" (equated with the earning equivalent of a minimum con足 sumption basket of 2,100 calories and 45 grams of protein per person per day).
More recently, Alamgir (1977) estimated that in 1973-74 there
were as many as 64 million people in rural areas of Bangladesh surviving below a minimum living standard. What little wealth there is in Bangladesh is found in its land and people.
If there were a utopian equity of land distribution and equal
productivity throughout the land, there would be less than 1.5 acres of land per household, which would mean less than 0.3 acres per person. Those households fortunate enough to own the average 1.5 acres may rent an additional acre from a more wealthy neighbor, but the total holdings are fragmented often into ten or more plots. any land.
Many Bengalis do not own
In the early 1970s the proportion of farmers classified as
landless was estimated at 40 percent (Abdullah, Hossain, and Nations, 1974), but recent unofficial reports put the figure as high as 60 per足 cent in some areas (Januzzi, 1977).
The landless laborers, when working
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
at the peak seasons of harvesting or sowing, earn about 1 seer (approxi mately 1 kilogram) of rice per day or two full meals instead, plus an amount of money equivalent to another seer of rice to buy his family’s food (Rahman, 1976).
In other seasons they work for food alone, without
the extra seer of rice.
During those slack seasons, less than one in
four will have employment (IBRD, 1972).
Furthermore, these low incomes
have been dropping in real terms since the early 1970s.
Although the
money wage of agricultural labor has increased between 1970 and 1975 from about Taka 2.38 per day to Taka 8.72, the "real wage" of unskilled agri cultural laborers dropped by about 30 percent over the same period (Alamgir, 1977).
The fall in real wages was caused by the 100 percent
increase in the rice price over the same period (Chen and Chaudhury, 1975)
(see Figure 12). The increased polarization of wealth (McNicoll, 1975) and increasing
pauperization leaves rural economic conditions in a state where . . . even a slight impact of an unfavorable exogenous force may seriously jeopardise the chances of survival (of the poor). They are caught in what may be called the "Below Poverty Level Equilibrium Trap." The trap exists between the poverty line and the famine line . . . The famine line is the level of consumption below which the probability of death is unity. (Alamgir, 1977) Although it is the poor who fall below this "famine line," the group most sensitive to famine (i.e., those where the effect may be most easily measured) are not the destitute, but those with some possessions to lose or sell off in the market glut (Mahalanobis, 1946). Against this background of poverty, moral actions often seem to clash with survival economics.
Babies were occasionally sold during
the famines (Hunter, 1868; Ghosh, 1944), and the guardian-less women
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
<u FM
Real wages of agricultural laborers
Figure 12.
Trends in the wholesale price of coar: of agricultural laborer s (1966 = 100)
real wages
resorted to prostitution in efforts to support surviving family members. Many of the lowest paid government workers were not paid for periods of six months in 1975.
The salary of a Circle Officer for Development (the
senior administrator for a thana) was insufficient to support his family of five with rice alone.
Was it corruption when he failed to distribute
a Persian relief blanket?
A Population at Risk International estimates of the 1977 population of Bangladesh suggest a figure of 83.5 million (IBRD, 1974) with an estimated rate of growth between 2.7 percent (Population Reference Bureau, 1977) and 3.01 percent (IBRD, 1974).
The rapid population growth, resulting from the sudden
decline in mortality after the stabilization of agriculture and the food marketing system in the 1930s and 1950s (Mosley, 1975), certainly provides Bangladesh with a potential resource (Harvey, 1974), but at present it constrains development and places increasing numbers of people at risk to famine.
The population is now a highly dependent one, both
on its own proportionately small workforce and on outside food supplies. Standard dependency ratios, indicating the percentage of the popula足 tion under 20 years of age and over 65, relative to the population aged between 20 and 65 (Bogue, 1969), have little meaning in a country where a woman may have borne six children by the age of 20, children below 10 are among the highest paid silk weavers, and male cultivators do not retire unless disabled.
With 50 percent of the population aged below 15,
only 34.73 percent of the population was classified as part of the civilian labor force aged 10 and over in 1968 (IBRD, 1974).
By 1973, the
labor force in Bangladesh was 26.3 million (Dorfman, Alamgir and Tabors,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99 1972).
By the year 2000, this potential labor force will have reached
64-70 million.
Present unemployment, even excluding underemployment, is
already estimated at 30 percent, and by the year 2000 it would reach 35 percent even if all the presently envisaged employment opportunities were provided (Thomas, 1972). Although the 75 percent of the labor force involved in agriculture have raised foodgrain production from about 11 million tons per annum in the mid-1960s to about 14 million metric tons in the mid-1970s, and annual imports have increased over the same period from about 0.5 to 2.0
million metric tons, still the per capita availability of foodgrains
waivers around 450 grams per day (Chen and Chaudhury, 1976).
This
average foodgrain availability per capita figure does not illustrate the real consumption patterns which are skewed by the purchasing power of the consumers (Levinson, 1974).
The national nutrition survey in 1962,
a relatively prosperous year, revealed that 46 percent of the population were classified as consuming inadequate quantities of calories (ICCND, 1965).
Foodgrain availability figures also fail to account for
nutritional losses within the body because of intestinal parasites, recurrent diarrheal illness, and fevers (Mosley, 1975). Pressure on the traditionally intensive agricultural areas is forcing people to live in marginal and vulnerable areas (Brammer, 1975). This redistribution may be "cushioning the impact of the deteriorating food situation" (Chen and Chaudhury, 1976) but the "push" (Haq, 1975) or "forced extrusion" (McNicoll, 1975), from the densely settled areas, may be only a short-term mortgaging of the future. they move are also vulnerable to disruption.
The areas whither
Rural population is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
migrating out of the densely populated districts of Chittagong, Noakhali, Comilla and Dacca (see Figure 13); some move to the urban centers, either temporarily or permanently (the population of urban centers grew at 11 percent per annum between 1961 and 1974), often to squatter conditions (Qadir, 1975), and others move to less populated rural districts where there are "larger farm sizes, lower man-cropping area ratios, and higher levels of per capita rice production" (Chen and Chaudhury, 1975). The rapidly growing districts are the peripheral, sparsely settled districts on the older part of the Bengal delta, particularly Kushtia, Dinajpur,
Jessore, Rajshahi, Rangpur, Bogra and Pabna (Elahi, 1976).
These are the districts with low cropping intensities:
while the central
districts of Dacca and Comilla had crop rotations of mixed aus, deepwater aman, and rabi, the recently growing districts are characterized by an aus, transplanted aman, and fallow rotation.
The districts of Kushtia,
Jessore, Dinajpur and Rangpur, as well as being relatively remote from Dacca, are also relatively prone to drought. Bangladesh's riverine districts with a proclivity for floods, embankment bursts, sediment dumping, and bank erosion now carry high population densities despite their low cropping intensities caused by the long period of
potential inundation.
These high riverine population
densities are little recognized because the National Census Commission (Bangladesh Census, 1976), the World Bank (IBRD, 1974), and the Dacca Geography Department (1975) fail to compensate for Bangladesh's riverine character when calculating population densities and mapping the resulting patterns (Figure 13).
They compute thana population
densities by dividing the thana population by the total area of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
BANGLADESH AREAS OF HIGH POPULATION DENSITY 1974 W
Thanas with over 2 ,0 0 0 persons per square mila
Thanas with over 2,00 0/s q . mile (population -f total administrative area) Thanas with over 2 ,0 0 0 /s q . mile (population approx. dry season land area)
Bruce Currey
Figure 13.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
thana.
Recalculation of the thana population densities, after sub足
tracting planimetered measurements of the approximate dry season river areas from the denominator for each thana, reveals a consistent pattern of high population density along the vulnerable Ganges and Brahmaputra riverine areas. During the annual floods, which on the average cover one-third of Bangladesh (Ralph, 1976), the areal denominator for calculating popula足 tion density becomes even less in the flooded area.
The baris (house足
holds) are left on island mounds clustered linearly among trees on the recent and relic levees of the delta.
Even in 1936, Geddes (1937)
estimated levee population densities at 30,000 persons per square mile. The levee densities have increased since then and I surmise that those families building anew at lower elevations nearer the rivers and towards the
interfluvial bils (lake swamps) are generally the poorer, and hence
more vulnerable groups. These depressing, possible interrelationships between poverty, vulnerability and elevation were exemplified by the location of two of the rehabilitation camps outside Dacca City in early 1975.
The
bastuharas, or slum dwellers were driven out of Dacca by the paramilitary Rakhi Bahini in a cosmetic cleanup operation.
They were resettled in
rehabilitation camps at points where the daily return bus fare to Dacca for work would be prohibitive for them.
The camps for these destitutes
were erected during the dry season, but they were sited below normal monsoon flood levels. Figure 14 shows that the precarious balance between food and population in the early 1970s responds inelastically to major disruptions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103 of the food supply system (Chen and Chaudhury, 1976).
Although both the
"ecological fallacy" and the "atomistic" or "contextual fallacy" (Susser, 1973) may be invoked in considering the juxtaposition of national and "case study" (black profiles) (white profiles) trends (Figure 14), there is a consistency of correspondence between the signs of stress and the disruption of the food supply during the 1971 war of liberation and the flooding followed by famine in 1974.
Although not
all confounding variables are removed, there is a strong association between major social and environmental disruptions and decreased growth rates.
The black profiles of the national statistics in Figure 14
show that although the food supply system was disrupted in 1971 and 1974, the amount
of imported
foodgrain and the per capita foodgrain avail足
ability were relatively high in Bangladesh in 1974. The white profiles (Figure 14) constructed from census data on 112,000
people in the old cholera field trial area in Matlab thana,
Comilla District, show that the crude death rates in 1971 and 1974 rose 50 percent above the rates for the previous years (Chowdhury and Curlin, 1976).
Infant mortality rates are shown to have risen con足
sistently with peak rates of 146.6 per thousand, and 167.2 per thousand in 1971 and 1974, respectively. the crude birth rate dropped. after the
In the wake of the two crisis years, It dropped slightly to 41.8 per thousand
war of liberation (see also Curlin, Chen and Hussain, 1976),
and then plunged to 27.6 per thousand in 1975 after the flood and famine in 1974 and early 1975.
The intrinsic growth rates declined from
3 percent in the late 1960s to 2.5 percent in the early 1970s, and then to the notably low figure of 0.9 percent in 1975.
Chowdhury and Curlin
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105 (1976) have further calculated that 25 percent of the drop in the rate of natural increase in 1974 and 1975 was due to increased mortality and 75 percent to decreased fertility. McKay (1974), in a case study of the mortality effects of the liberation war in the Sylhet district, found in comparing
1972 death
rates with those from the period 1966-70 that deaths due to bacillary dysentery had increased 24 times, anemia 5.8 times, tuberculosis 3.8 times, pneumonia 2.9 times and malnutrition 3.1 times. nutritionally related diseases.
All are
Sommer and Loewenstein (1975), after
an initial anthropometric nutrition study of 8,292 children aged between 1 and 9 in the Matlab area in December 1970, later followed the death rates in the different nutritional classes after 18 months.
The 9
percent of the children who were classified as severely malnourished were later found to have death rates 3.4 times higher than the 50 per足 cent who had been classified as normally nourished. The reductions in fertility stem partially from an increasing median age at marriage (which has risen in Matlab from 15.9 years in 1968 to 17.4 in 1973; Aziz, Chowdhury and Mosely, 1970; Chowdhury, Huffman and Curlin, 1976), but predominantly from a drop in the marital fertility rate.
That malnutrition in women could lead to lower birth
rates was hypothesized by Austin and Levinson (1974).
Frisch (1972)
stressed the necessity for critical body fat levels before menstruation could occur.
In Bangladesh, Chowdhury et al. (1976) have demonstrated a
recent increase in the age at which menarche is reached.
Of the 1956
female cohort who had reached age 13 in 1969 before the nutritional and psychological stresses of the war, 21.1 percent were menstruating by the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106 age of 13.
Of the 1962 cohort who reached the age of 13 in 1975, after
the war and after the famine, only 7.45 percent were menstruating by the age of 13.
Considering the few indicators available, together with the
lack of any contrary evidence, it seems that the Bangladesh population is particularly at risk to any major disruptions which impair food production, food supply, food consumption, or food absorption.
Faith, Hope and Charity Traditional village and kinship sharing, whether religious or secular, appears to operate ineffectively during times of famine, and the vacuum has not been filled by external relief systems. less compatible with famine than with other disasters.
Fatalism is
Similarly, the
extended family sheds its economic supportive links during the pervasive onset of famine.
Even the sneho, or love support, of the nuclear family
breaks down during the crisis.
The community caste systems of support
during scarcity have probably been less strong in Muslim Bangladesh than in many other parts of South Asia. probably removed
The abolition of large landlords has
one of the few remaining famine insurance mechanisms
at the village level.
Exogenous support systems in the form of relief,
although long established, are at present poorly organized to cope with famines which occur infrequently and irregularly.
In 1974 the relief
system was particularly disorganized because of temporary corruption within the system, and temporarily impaired confidence between the donor countries
and Bangladesh.
"Allah sends it, so only Allah can prevent it."
This fatalistic
attitude has been strongly associated with the Bengali perception of "rapid onset" disasters such as the 1970 cyclone (Islam, 1974).
The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
areas affected by that cyclone were the coastal areas of Noakhali and Barisal, both traditionally conservative Muslim areas.
The former,
Noakhali, where Aminul Islam carried out his cyclone research, is the home district of many of the imams (priests) of the Dacca mosques.
At a
national scale, it is less likely that the succession of events which leads slowly to a famine produces so much turning to the sacred for aid. Famine is more probably attributed to such real or "scape-goat" profane causes as "bad government" or "unfavorable price structures." Likewise, family units appear to respond differently to famine con足 ditions as compared with intermittent disasters like tornadoes.
I,
myself, have witnessed villagers from an extended family bringing bhater handi (pots of cooked rice) from several miles distant to help those devastated by the 1976 Faridpur tornado.
These Samaritan responses
clash with my own observations of households starving immediately adjacent to apparently affluent families during the 1974 famine.
The economic
sharing among the extended family contracts; a destitute will seldom be given shelter during a famine for more than one night in a household of his
own extended family.
This survival ethic (compare Turnbull, 1974)
even permeates the nuclear family during famine.
Greenough (1976)
provides evidence of the disintegration of the paribar (family) during the 1943 Bengal famine.
He details the separation of husbands and
wives in extreme situations and also the selling of children, documented elsewhere by Hunter (1968) and Ghosh (1944).
In 1974 Aziz (1976) wit足
nessed suckling children, with mothers absent, trying to suckle the grandmother. At the community level, McAlpin (1975) has proposed that the jajmani system acts as an effective famine insurance mechanism in other areas of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108 South Asia.
An outgrowth of the hindu caste system, the jajmani system
is one in which the jajmans (land controlling castes) are normally responsible for the protection and maintenance of the kamins (dependent castes) (Beidelman, 1959).
In Bangladesh where the caste system has
been dispersed under Muslim dominance, the jajmani system is less of a fixed or dhara relationship with a long-term continuance through inheritance (Greenough, 1976; Zaidi, 1970).
Particularly with the high
proportion of landless laborers and fishermen, the .ja.jman-khamin links tend to be nagat relationships which are casual and require immediate reciprocation, or hat relationships a cash-goods market transfer.
where the tie is only in the form of
Even the dhara relationships have been
shown to collapse under famine conditions in other areas of India (Jodha, 1975; Epstein, 1967).
The ja.jman patrons suspend or delay (often by
moving or absenteeism) their commitments during the crisis.
In Bangladesh
where the nagat and hat relationships are even less binding, the relation足 ship may break down almost as soon as conditions begin to deteriorate. The traditional fall back position in times of need in Bengal was the zamindari system (the landlords of Moghul times were replaced by nonagricultural banias (middlemen) who were then given increased and undefined revenue collecting powers by Lord Cornwallis' Permanent Settlement of 1793 (Mukharjee, 1975).)
This feudal system did act as a
form of aid in hard times for at least one season of crop failure (Brammer, 1975) as the landlords were able to draw on external merchant stocks to relieve their own hari (vassals) in return for later service.
The land足
lords were traditionally Hindu and in 1950, shortly after partition, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109 zamindari system was abolished^ by the East Bengal State Acquisition and Tenancy Act (Government of Pakistan, 1966). The only traditional organizations left in the village today are the sirdars, mattabars, morrols, pramaniks, or murubbis (elders) of the sama.j or former panchayat (council), and the moulvis and mollahs (Muslim priests and religious scholars).
Both the elders and the priests are
held in respect as counsellors in marriage and land disputes rather than as organizers.
The samaj still remains stronger in North Bengal where
their leaders have absolute power.
Although they used to feel responsible
for the welfare of the village, their authority has been eroded recently by factionalism (Zaidi, 1970).
The traditional panchayat was partly
formalized and included in the government structure by the ChowkidariPanchayat Act of 1871 to become the union panchayat or union parishad, the lowest level of local government responsible for policing and taxing the union.
The union parishad acts as the relief distributor for the
villages of the union, particularly to the "D Category" inhabitants, those with little or no land and a large number of dependents, and usually those not on the union parishad tax roll (Rahman, 1976).
The
union parishad, however, has only 9 elected representatives, but although there are a minimum of 9-10 gram (villages) per union, there may be as many as 30 gram per union in other areas.
Thus many gram are without
direct representation in gaining access to relief.
A few gram may have
several members on the union parishad and this may result in a political,
^The Famine Enquiry Commission of 1945 had recommended the abolition of the zamindari system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
rather than a humanitarian distribution of resources.
In a minority
of Bangladesh villages, the Integrated Rural Development Program (IRDP) has encouraged
the formation of self-help organizations, the Krishi
Samabaya Samiti (KSS), based on the Comilla model of rural development cooperatives (Raper, 1970).
These cooperatives, if active (Emmert, 1977),
encourage village cooperation and organization to raise loans through the Thana Central Cooperative Association (TCCA) to stimulate village develop足 ment by the implementation of new irrigation, agricultural, and occasion足 ally educational projects.
Although informally linked by virtue often
of joint membership, the KSS is completely separate from the relief distributing union parishad.
There is presently no government administra足
tive structure to channel relief funding into self-help schemes for development.
The restricted opportunity for development immediately
prior to the outbreak of famine is not conducive to self-help schemes without strong local organization, or without the stimulus of some out足 side agency.
The villagers are caught in McNicoll's (1974) "prisoners'
dilemma" where even though each villager knows he would benefit by cooperative action to improve embankments, construct irrigation bunds, or plant bamboo groves, he does not cooperate because he also knows that if he individually initiates a community project without being matched by the
other village members, then he will be worse off. Relief is not new to Bangladesh, but the scale, source, and methods
of implementation have changed recently.
In Bangladesh relief is not a
western technological discovery imposed on a pristine culture which lies in perfect harmony with nature's cyclical vicissitudes (compare Waddell's findings on Australian relief to the Enga tribespeople of New Guinea
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill (Waddell, 1974 and 1975)).
The Mahasthan tablet (Figure 15), found in
North Bengal and dated around 350 B.C., details . . . a Mauryan Ruler's orders to relieve the famine distress of the Samvamgiyas . . . settled in and about the town. . . . by cash and rice loans . . . to tide over the distress. (Bhandarkar, D.R. 1931) In "The History of Indian Famines," Srivastava (1968) documented similar relief actions throughout Bengal in Moghul and British times. people
Today
in Bangladesh talk about the "growth of a relief mentality."
History, however, suggests that this mentality has been growing through the centuries.
It is the donors who have changed.
The dependence is no
longer upon local rulers but upon distant aid donors as with the recent large scale relief operations of the United Nations and the United States (UNDRO, 1973; USAID, 1974). The "growth of a relief mentality" in response to these recent bilateral and multilateral aid schemes may result from a number of important causes:
from the magnitude of the schemes and the ensuing
fears of their long-term impact on the rural economy;-’ from discontent at the presence of highly paid expatriate monitors who sometimes have limited effectiveness (Rohde and Gardner, 1973); from the untimeliness
■’Cash for work projects:
the infusion of cash may inflate market prices for those who are not fortunate enough to be employed on the project;
Food for work projects:
the infusion of foreign grains may, by depressing local market prices, act as a disincentive to local pro ducers .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
THE M A H A S T H A N TABLET Evidence of relief adm inistration in the 4 th century B.C.
Transcript
nena Sa(m)va(m)giy (a)nam (Galaclanasa) I Dunadina-(maha) mate I sulakhite Pudanagalate I e(ta)w (ni)vahipayisati I Samva(m)giyanam (cha di)ne (tatha) (dha)niyam I nivahsati I da (m)g(a) tiyay (i)k(e) d(eva)(tiya)(yi)kasi I su-atiyayika(si) pi I gamda(kehi) (dhani)(yi)kehi esa kothagale kosam (bhara)(niye)
Translation "To Galadana(Galardana)
of
the
S a m v a m g i y a s . . . (was g r a n t e d )
b y order. The M ah am a t r a f rom the h ig hl y a us picious Pudran ag ar a w i l l c a u s e i t t o b e c a r r i e d o u t . (And l i k e w i s e ) p a d d y h a s b e e n granted to the town during
Samvamgiyas.
(this)
The outbr ea k
ou tb ur s t of
superhuman
(of d i s t r e s s ) agency
in the
s h a l l be
t i d e d over. W h e n t h e r e is an e x c e s s o f p l e n t y , thi s g r a n a r y a n d t h e t r e a s u r y (may b e r e p l e n i s h e d ) w i t h p a d d y a n d t h e
gamdaka c o i n s ."
This inscription containing 7 lines in Brahmi characters of the Mauryan Period on a piece of hard limestone measuring 3.3" by 2.3" by 0.9" was discovered at Mahastangarh village near the Karatoya River in Bogra District, North Bengal on the 30th November 1931. It was found near a high mound which may have been a royal granary (Bhandarkar,1931). Figure 15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
of relief arrivals from distant sources, e.g., the U.S.A., Canada, Australia, and the EEC countries compared with former imports from India and Burma; from the import of inappropriate relief items; from attempts to circumvent local relief distribution systems in the interests of short term expediency; and in limited cases from an understandable jealousy at the success of an expatriate system. Just as the Bangladeshis are skeptical about the donation of this new international relief, so too the donors are not yet fully convinced of the Bangladeshi administration's capability to manage these massive injections of relief.
In 1974 the Dacca delegates of the donor countries
certainly had a lack of faith in key relief recipient personnel.
The
embezzlement of relief funds and the smuggling of foodgrain supplies involved personnel in the government (Washington Post, 1974) and the Bangladesh Red Cross Society Chairman, Gazi Gholam Moustaffa (Bangladesh Times, 1975).
Rumors of government attempts to "cry wolf" regarding
premature reports of massive flooding in order to raise aid donations in early 1974, retarded the relief response when the floods really did occur later.
This two-way skepticism about the handling of the new international
relief causes delays in the implementation of relief programs (e.g., McHenry and Bird, 1977).
It thus increases Bangladesh's vulnerability to
famine.
Natural Disasters Bangladesh is prey to natural disasters.
Global assessments of
disaster risk from 1947 by Sheehan and Hewitt (1969) and more recently by Dworkin (1974), each of whom divided the globe into 10° grid cells, rate the southern Ganges basin cell highest in terms of total loss of life
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
resulting from disasters.
There are, and always have been, many extreme
environmental events which may occur in Bangladesh.
The already described
decrease in the "degrees of freedom" recently has both temporarily lowered the country's capacity to absorb such extreme events. History indicates that natural hazards have long been part and parcel of
the Bangladesh landscape (e.g., Islam 1974).
The "Ain-i-Akbari"
by Abul Fazl-i-Allami (Jarrett, 1949), which chronicles events during the reign of the great Moghul emperor Akbar (1542-1605), refers specifically to "a terrible inundation . . . which swept over the whole of Bakla Sarkar (Bakergunj and Khulna) in 1584 . . . amid thunder and hurricane of wind . . . [causing] nearly two hundred thousand living creatures to perish in the flood in Bengal Subah (division)."
Since the beginning of
the British period in 1757, the gazetteers and statistical accounts of Bengal compiled by the district administrators have documented cases of natural calamities, famines and scarcities.
The gazetteer records are
fraught with inconsistency, bias and incompleteness (see Chapter IV). As an example, the 1943 Bengal famine is not mentioned in Noakhali District, but Noakhali is known to have been one of the most severely affected districts (Mukherjee, 1944 and Mahalanobis et al., 1946).
Some
of the apparent anomalies are, however, explained in the text of the gazetteers, e.g., for Comilla the Eastern Bengal District Gazetteer states: In 1866 when Orissa and Western Bengal were visited by the worst famine of the country, the price of rice in Trippera (Comilla) rose to Rs.5 per maund, but this was paid by outsiders, and most of the people continued to eat their rice at the price it cost them to produce it, while they received a handsome sum for the surplus stock. (Webster, 1910)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115 Similarly in 1896-97, The early cessation of the rains injured the harvest, but there was nothing approaching famine, hardly indeed scarcity. (Webster, 1910) Cycles of famine which may be initially apparent from these historical records must be treated with skepticism.
The cycles may
reflect the history of the record rather than the actual history of the occurrences of famine.
Thus it is tempting to see from Figure 16 a one
hundred year cycle with famines being prevalent around 1770, 1870, and 1970.
These three dates, however, are also times when one xrould
expect recording to be most intensive:
around 1770 the British had just
imposed their rule throughout Bengal and both the new rulers and subjects were probably quick to record any shortages or famines; around 1870 was just before the publication date of several of the gazetteers, so one would expect them to have had either memories of any recent shortages or famines; and similarly around 1970 is very near to the present day and also just before the publication of the most recent series of gazetteers. The record does confirm that famine is Bangladesh.
not a new phenomenon in
Particular years stand out as having had widespread famines:
1769-70, 1787-88, 1866, 1874, 1897, 1907 and 1943.
There is no evidence
of an increase in either the frequency or scale of famine in Bangladesh. Although the numbers of people affected by the natural calamities and famines may be increasing, there is no evidence that famines are affecting an increased percentage of the total population.
During the Bengal famine
of 1769, there were estimates that one-third of the population perished (Hunter, 1868).
A century later, the famine of 1866 is estimated to have
claimed 135,000 lives, or 2-3 percent of the estimated population (Bhatia,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116 THE HISTORICAL RECORD OF NATURAL CALAMITIES, SCARCITIES AND FAMINES 17 DISTRICTS BETWEEN 1757 AND 1970
Figure 16.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
1967).
Estimates of mortality of the Bengal and Bangladesh famines
of 1943 and 1947 were as low as 3 percent of the population (The Famine Inquiry Commission, 1945) and 0.1 percent of the population (Keesing's Contemporary Archives 1975) in the respective years. The repeated juxtaposition of calamities and famines throughout the historical period strengthens the evidence of an association of famine with natural hazards.
The latter appeared to "trigger" the former
despite the agricultural abundance and food surpluses of the province as a whole, until the 1930s (Ahmed 1968). ministrators
It might be assumed their ad足
in Government service were predisposed to associate famines
and scarcity with natural calamities rather than admitting that their own administrative incompetence was one of the causes of famine.
Nevertheless
over 70 percent of the recorded famines and scarcities occurred in the same year, or in the year immediately following, the recording of a calamity.
Famines certainly did occur without any natural calamity being
recorded.
Natural hazards also occurred without "triggering" famines;
thus Sylhet appeared to have suffered from frequent flooding, but there was little correlation between flooding and famine.
This may have been
because of poor recording, or because the flooding occurred in a traditionally surplus area.
It may also have been that frequent flooding
in a food surplus area led to a response mechanism in which larger storage facilities were built to tide the population over all except the periods of most extreme scarcity (Brammer, 1976). appear to "trigger" famines more than others.
Certain disasters
Floods and droughts were
strongly associated with famines, unlike earthquakes which had little immediate effect, either on the standing crops or on the agricultural
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118 employment levels.
The earthquakes may, however, have had a long-term
effect by changing sediment patterns and river gradients, which in turn caused flooding elsewhere in the catchment area up to twenty years after the original earthquake (Berry, 1976).
A famine might have been recorded
in several districts, but the actual natural disaster "triggering" it might have been different in different districts, e.g., the 1866 famine was apparently "triggered" by drought in Rangpur, but in the adjacent district of Bogra, the "triggers" included flood, drought, seasonally maldistributed rainfall and a cyclone.
This does not imply, however, that
natural calamities were not one of the intrinsic causes of famine, but merely that there were real variations in the many causes of famine. Famines in Bangladesh also appear sometimes to have been associated with a sequence of disasters, e.g., in describing the Rangpur famine of
On the 28th May (1787) the Zamindars of Kuchwara . . . presented . . . [to . . . the collector] . . . a petition . . . which set forth that three months incessant rain entirely destroyed the spring harvest . . . The rain con足 tinued unceasingly and the cultivators were in great distress and were abandoning their fields in large numbers, while the cultivation of the winter rice crop was hardly possible, owing to the overflow of the rivers . . .The lowlands were entirely under water, [and] the incessant rain prevented the seeds from germinating, and that even the higher lands, which they had attempted to bring under cultivation, could not be properly attended or weeded, and that in the consequence of the growth of weeds or jungle, which had choked the rice, all such lands had been turned into pasture for the cattle . . . On the 29th July the Collector wrote . . . to the Revenue Board that the unseasonable rains which had commenced on the 26th March, which had [destroyed] the rabi crops on the lowlands, had entirely ceased for the past ten or twelve days . . . The fair weather was of short duration. The rains set in again with renewed violence on the 1st August, just as the cultivators were transplanting the young rice, the rivers again overflowed their banks. But Tista
th e wo r s t was yet that
in all
times
to
co m e .
The m a i n b r a n c h of
is an e r r a t i c r i v e r
the
. . . suddenly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119 swept down from the hills such masses of sand as to form a bar in its course, and bursting its bank it forced its way into the Ghaghar. The channel of the latter stream was utterly in足 adequate to carry off this vast accession of its waters, and the Tista accordingly spread itself over the whole district . . . on the 27th August. The waters at last subsided, leaving the winter rice crop . . . considerably injured. Six weeks of fine weather and the most careful attention to the young crop raised the expectation that the harvest might yet be a fair one. But the calamities of the season were not yet over, and a cyclone next swept over the stricken country early on the morning of the 2nd November, just as the (aman) rice was getting into ear. (Vas, 1911) Despite the incomplete information the gazetteer records did emphasize the relative vulnerability of certain areas of Bangladesh. Rangpur, Dihajpur, Dacca and Faridpur districts stood out as having higher frequencies of famines and scarcities. relatively free: Hill Tracts.
Other districts were
notably Mymensingh, Pabna, Chittagong and Chittagong
The accompanying text in the gazetteer records corrobo足
rates this scanty quantitative evidence.
Thus the Bengal Gazetteer for
Mymensingh points out that, [In Mymensingh] the people can never be seriously afraid of either flood or drought in sufficient quantity to threaten famine. (Sache, 1917) Similarly, O'Malley (1908) describes the low famine vulnerability of Chittagong District in the Eastern Bengal District Gazetteer: The position of Chittagong, between sea and mountain, ensures a heavy and regular rainfall and guarantees it against those famines to which less favored tracts are subject. Shut in as it is on one side by the Bay of Bengal and on the other by hills and many hundreds of miles of forest country, there is a con足 stant inrush of moist winds from the sea, which though subject to fluctuation, never fail entirely. There is a further safe足 guard against famine in the great variety of soils, ranging from high sandy tracts to low swampy marshes; for in a dry season the failure of the crops in the highland is counterbalanced by the fertility of the lowland. By terracing the fields as they slope downwards from the low hill ranges and by damming up the small streams which form after a few showers of rain, the cultivator can generally secure a sufficient supply of water for the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
irrigation of his fields. Consequently any failure of the harvests owing to drought can only be small and partial, and an extensive failure of the all important rice crop is un足 known. A season of high prices is, indeed, not altogether unwelcome to the people generally, as most of them are cultivators keeping stocks of rice for their own consumption, and they are benefited by the high price they obtain for their surplus stock. Unfortunately, the historical record was not only incomplete, but it was also confined to the district scale.
References to sub-district
units were limited to short comments within the narrative, e.g., the Bengal Gazetteer of Bogra specifically mentioned that, the portion of the district west of the Karatoya river comprising thanas of Bogra, Sherpur, Shibgunj, Khetlal and Adamdighi and also partly in thana Panchbibi is most likely to suffer from failure of the rice crop. (Gupta, 1910) Each district had its own "set" or "mix" of disasters.
Floods
were ubiquitous, but droughts were seldom mentioned either in the coastal districts of Noakhali and Bakergunj or in the north east in Mymensingh and Sylhet. country.
Earthquakes were recorded in the north and east of the
Cyclones were certainly most frequent in the coastal districts
of Khulna, Bakergunj and Noakhali, but they were also found away from the coast.
They were recorded occasionally as far north as Rangpur,
Dinajpur, Pabna and Mymensingh, but this may have been a misnomer resulting from local confusion between a cyclone and a severe linesquall. In 1975, Brammer listed twelve major natural hazards which still affect agricultural production in Bangladesh:
1. 2. 3. 4. 5. 6. 7.
Floods Droughts Cyclones Storm surges River erosion Burial by alluvium Linesqualls (Nor'westers)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
8. 9. 10. 11. 12.
Hail Saline incursions Capillary salinity and alkalinity Crop pests and diseases Earthquakes
Diseases and epidemics are excluded because their effect on agricultural production is indirect.
From the historical evidence available it is
not possible to say whether the frequency of these hazards, especially flooding, is increasing, despite Eckholmâ&#x20AC;&#x2122;s inferences suggesting such an increase (Eckholm, 1976).
It is not yet proven even that sediment loads
in Bangladesh rivers have increased significantly in the long term as a result of Himalayan deforestation.
It certainly is not yet possible to
predict any increase or decrease in the often complex successions of natural disasters which concatenate to weaken the rural economy. The impact in terms of the
immediate mortality caused by the
natural hazards is certainly high, e.g., an estimated 225,000 people died in the 1970 coastal cyclone of November 12th and 13th (Islam, 1974), and it is probably increasing.
Although the dollar impact per capita is
low, e.g., $3.00 in the 1970 cyclone (Burton, Kates and White, 1977), because of
the large population denominator, the resulting disruption
to the social system and the development process is probably high; however, it remains uncalculated and perhaps unmeasurable.
There are,
as previously stated, more people now living in the vulnerable areas. Some of these people are voluntary settlers and others live in planned resettlement schemes such as the people living in the Char Elahi Resettlement Scheme (see below, page 127).
The people living in these
traditionally unsettled areas may be seasonal laborers or new permanent settlers.
Whether seasonal or permanent these settlers are often new to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
the vulnerability characteristics of the area, e.g., many of the new settlers on the Rangpur chars are from the Mymensingh district.
Also
they may have to cope with the additional problem of litigation over land rights in the changing river environment;
often being penalized by
the law (Rakshit, 1956) or harassed by the hired latials (guards) of a more powerful party (Adnan, 1976).
The intensification of agriculture
by introducing new crop rotations, increasing yields and improving infrastructure in these areas may increase the potential damage.
Certain
modes of development may also increase the effect of the natural event itself, e.g., a thatch roof shatters in a tornado, but a corrugated-iron roof becomes a lethal weapon.
Some development measures may have
decreased the impact of natural events of intermediate intensity, but they may also have unwittingly increased the impact of the occasional extreme event which exceeds the absorptive capacity of the preventive measure. Islam (1974) has documented the tragic consequences of storm surges breaching coastal embankments originally built to combat salinity inundation.
Similarly one can speculate on the effect of fuel supply
failures for the new deep tube-wells in North Bengal.
Ralph (1975)
interprets various official figures over the short time period, 19621970, to show that "severe damaging floods are occurring with greater frequency."
She ascribes this not only to the increase in floodplain
population, but also to better damage reporting. The threshold between "coping" and "crisis" has probably been decreasing in Bangladesh.
For example, before the 1974 flooding the
people were poorer than in the early 1960s, with less monetary and food reserves, they lived at lower elevations in the flood vulnerable areas,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
and they had been affected by a series of disasters:
the 1970 cyclone,
the 1971 war of liberation, the 1972 drought and the 1973 flood.
These
disasters may have sensitized the international communityâ&#x20AC;&#x2122;s response to the country, but they equally may have calloused local responses.
The
succession of disasters probably increased the number of vulnerable poor in the country (Mahalanobis, 1944).
Rahman (1976) quoted his
grandfather's warning to "keep enough surplus to tide the household over three crop failures."
Since his grandfather's time, there are more in
the household, there is less surplus to store, there are proportionately more living in the vulnerable areas who are dependent upon the surpluses in the non-vulnerable areas in times of distress.
Although the marketing
system has probably improved the ease with which foodgrains may be removed from an area in return for cash, I would speculate that the con verse ability of the relief systems to return foodgrains to the area in times of need has not increased at the same rate, because it is not encouraged by a cash exchange.
Grandfather Rahman's aphorism may have
to be rethought in considering contemporary famine vulnerability.
The Essence of Bangladesh's Famine Vulnerability What then are the main factors predisposing Bangladesh to famine? How do the components of vulnerability in Bangladesh differ from those in other countries?
The conceptual model (Figure 6) outlined earlier
is used to structure such a summary by assessing Bangladesh's vulner ability in terms of its production, the storage and savings capacity available to Bangladesh and the external and internal isolation of the country.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Production. rice.
Bangladesh's foodgrain production is dominated by
Its ability to produce its food is not falling because of soil
erosion as is the case in many of the African states vulnerable to famine.
In theory the tropical monsoon climate could sustain three
cropping seasons in many areas of the country, except for occasional weeks of
low
temperatures which might prevent the rice from flowering
during the dry season. dry spring season.
Irrigation is necessary in many areas during the
Present rice yields of only about nine cwts. per
acre are low compared with other south and east Asian states. yields are increasing slowly. by natural disasters.
The
Extensive damage is caused intermittently
In Bangladesh the increase in crop yields is
restrained by the same lack of incentives to the cultivator and the lack of knowledge about new techniques that were noted in Bihar, the case in Chapter II.
Lack of water at a national scale is not a problem.
Lack of
precipitation is only a relatively minor problem confined to the west and north of the country.
The major problem is a seasonal surfeit of
water with floods affecting 40 percent of the country on average each year and creating a potential pre-famine landscape remarkably similar to those
of the Chihli Plains in pre-Maoist China.
unlike water deficit may be viewed as a resource.
Water surplus, however, That resource still
remains largely untapped as a source of energy, as an opportunity for seasonal transport routes, as potential irrigation water and perhaps most important, as aquaculture and pisciculture areas.
Efficiency in using production.
The majority of Bangladesh's
large agricultural population lives in settled villages in areas carrying extremely high population densities, a situation very different from that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125 found in the nomadic areas of the Sahel.
The population was outgrowing
food supply in the late 1960s and early 1970s.
The new dependence on
outside food aid is primarily a function of population growth, although food imports are more than equalled by post-harvest losses caused by pests and
poor storage.
These internal losses have been occasionally
increased by smuggling and individual consumption losses due to parasites and other diseases.
Internal production might more than match population
growth if the organization of the food system could be improved.
Such
improved organization would have to operate at and between all levels of the present food production and distribution system.
It would have to
include the laborers, the landholders, the merchants, the bureaucracy and the policy makers.
Similarly, improved organization might have the
potential to make the growing population a productive asset rather than an increasing liability in terms of unemployment, poor health, and illiteracy.
Storage and savings.
In economic terms Bangladesh is poor, with few
reserves with which to take the risks
of opportunity.
Since indepen足
dence, neither the individual cultivator nor the nation has had large assets.
Population growth itself has eaten into any reserves.
There
are still harvests providing surpluses but the ability to organize the storage and distribution of such surpluses appears lacking at both national and individual levels.
The increased dependence on inter足
national food supplies makes it more susceptible to declines of stocks and increased prices in that external system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126 Degree of isolation.
Its relatively minor strategic importance,
although lessening the possibilities of massive "mutual security" aid that was available to many of the more strategically located, southeast Asian states, nevertheless, allows the country the benefits of nonalignment and multilateral aid links.
History links Bangladesh with
Europe, the United States and the Commonwealth. the country to Pakistan and Development Bank. with the
Islamic culture ties
the benefits of the oil-backed Islamic
Increasingly, common problems and environments tie it
Third World and especially the People’s Republic of China.
Perhaps the most crucial links are those between Bangladesh and its neighbor, India.
Strong Indo-Bangladesh relations will allow the inter
nationalization of water management, the improved control of border transfers, and the rapid transfer city.
of appropriate aid in times of scar
For the present these aid links are crucial, especially in times
of disaster.
To maintain all the links effectively at critical times
depends upon the credibility of the government in power.
Within such a
small country, few areas are traditionally isolated in the Ethiopian or Chinese sense, but in times of distress, railway distances from the two ports of Chittagong and Khulna and political distances from the cen tralized bureaucracy in Dacca become paramount.
The limited arterial
roads and rail systems are vulnerable to floods and sabotage, and the waterways, although highly developed by the market sector, seem relatively little used for relief.
Examples of Famine Vulnerability Within Bangladesh The above characteristics of famine vulnerability in Bangladesh have referred to the national scale.
At the village level, famine
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
vulnerability has different components in different parts of Bangladesh. The following vignettes exemplify vulnerability in three villages in different areas of the country.
They illustrate how different factors
operated at different scales and different stages before the outbreak of famine symptoms in 1974. The three sites chosen here as examples are Char Elahi, a newly emerged island in the Bay of Bengal; Purundapur Village on the banks of the Meghna river two miles north of Chandpur town; and Rowmari Thana on the border marches of eastern Rangpur (see Figure 10).
Char Elahi.
In 1967 the government of what was then East Pakistan
resettled Muslim refugees from the areas of Agartola and Tripura in India on this recently emerged char island in the of Bengal.
Bay
Having lost all their land and
property in India, each cultivator's household was allocated three acres of land which could only support one crop per year.
All allocations
were equal regardless of the number of mouths to be fed in each household.
In an attempt to settle
this remote area with no road access, they were given the requirements for cultivation, their house, and 25 takas (about 2.3 U.S. dollars) for household utensils. These basic necessities, together with any interim savings, were lost in the November 12th-13th cyclone in 1970.
Relief arrived bringing food,
clothing and some shelter, but the relief also ceased, leaving them once again without savings and capital.
They still had their three acres
of land surrounded by a ten feet high coastal embankment.
In the early
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128 lS70s the embankment fell into disrepair because of administrative break足 down and two years of unusually high water in 1973 and 1974 left a salt crust over much of the char.
The saline deposits damaged the single
crops of aus which the cultivators depend on for agricultural employment in the area.
Purundapur Village.
In the late 1960s the meanderings of the River
Meghna forced a community of Muslim fishermen to quit their char island and resettle on a levee in Purundapur Village on the eastern bank of the Meghna.
They had no
cultivable land like the established villagers, but they had their houses, their savings and some had their fishing boats.
The resettlement site was
near the river port and subdivisional town of Chandpur so there were urban job opportunities especially in the grain trade.
Their savings and
storage levels dropped suddenly in the early 1970s after increasing tension between the newly arrived char people and the settled people of Purundapur literally reached flash point and the by fire.
char people's houses and belongings were razed to the ground
In1974,
Purundapur was reputed to be the only village in the
thana that was "famine stricken" (see Figure 1).
The richer villagers
who had been settled there for a longer time, lost their aus crop and aman seedlings during the
unusually high flooding and river erosion which
removed over a half mile of land.
This was especially severe since it
occurred simultaneously with the contraction of credit.
The char dwellers,
who had already lost their capital in the fire, were even poorer in 1974
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
as fishing in the Meghna had been declining since the early 1970s.
The
irrational craving for rice during the crisis in 1974 reduced the market for fish and the fishermen sold off 50 percent of their boats.
Those
working in Chandpur brought back fewer wages as the grain trade con足 tracted in response to the smaller rural demand.
Rowmari Thana.
This thana in the extreme east of Rangpur District
in North Bengalis cut off from
the rest of the district by the braided
10 mile wide Brahmaputra River.
It lies between
the Brahmaputra to the west and the Garo-Khasi escarpment to the east in India.
Only a river
strewn neck of land connects the southern side of the thana to Mymensingh District
to the south.
There are no roads for vehicular
traffic except
in the dry season when a precocious jeep can precariously balance its way along the remnants of embankments.
Politically and physically, it
would be the least likely thana in Bangladesh to receive capital in足 vestment for development.
It was a bridgehead during the liberation
war in 1971 for the Mukti Bahini to lead the Indian-backed liberation movement and disturbances were widespread.
After liberation, during
the government of the late Sheikh Mujibur Rahman, it became a bridgehead again, this time for the smuggling of rice into India both across the bridge and along the dacoit (bandit) held Brahmaputra river.
Some
villages reported losing as much as 80 percent of their production over the border.
Although this was undoubtedly an exaggeration when viewed
at the aggregate thana level, smuggling undoubtedly exacerbated the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 30
thana1s difficulties after liberation. 1973 and 1974.
Flood levels were high in both
In 1974 there were six flood peaks which destroyed much
of the aus harvest and aman seedlings.
River erosion in Rowmari was
minimal, except on the char islands because of its east bank situation. Burial by sand, however, prevented the planting of flood recovery crops and reduced the available grazing for the cattle on the chars. No planting and no grazing meant less available employment. In all three of these examples, famine vulnerability was associated with natural hazards. the salinity.
In Char Elahi, there was the cyclone and later
The new villagers came to Purundapur because of nearby
river erosion and in 1974, river erosion and flooding were the "final straws that broke the back" of the impoverished community.
Similarly in
Rowmari, the unusually long lasting flooding and burial by sediment further reduced agricultural employment in a year of economic stagnation. In areas liable to natural hazards the likelihood of famine is con足 siderably increased.
The mapping of natural hazards by themselves will
not allow prediction of where famine will occur, but it will allow the dismissal of certain nonhazardous areas as being unlikely to be vulnerable to famine.
Areas Liable to Famine in Bangladesh This dissertation does not propose
to analyze the entire complex
and dynamic concept of famine vulnerability for each village in Bangladesh.
It restricts itself to analyzing the comparatively static
components (e.g., inaccessibility or flood proneness) which coincide with the actual areas liable to famine.
Although superficially this
appears to omit the less tangible process factors (e.g., inflation or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
changes in political ideology), the factors which are associated with the areal pattern of famine vulnerability at the regional scale in Bangladesh have in fact been relatively tangible and static, both in history (Figure 16) and at present (Brammer, 1975).
This assumption is corrobor足
ated in the footnoting of the Bengal Famine Code (Figure 17) which implies that a map of areas liable to famine, once prepared, is relatively permanent unless specific changes occur, such as perhaps the building of an embankment or the change of a river course. The areas liable to famine in Bangladesh are those areas with an inherent susceptibility to outbreaks of famine.
They are the rural
pockets in which localized famines are most likely to occur and where the effects of nationwide famines are likely to be most severe.
They
are the areas where the famine symptoms outlined in Chapter I are most likely to occur.
Such areas are not necessarily the areas where chronic
malnutrition is most severe.
Neither are they equivalent to the "food
deficit" areas where the population's food demand is greater than the local agricultural production. To say an area is liable to famine is of famine.
not to predict the occurrence
Almost all parts of Bangladesh could, in particular and
extreme circumstances, suffer from famine.
[So could, and have almost
all parts of the world (Walford, 1878 and 1879).]
As an example of such
an unpredictable area for a famine outbreak, the border insurgency problems in the Haluaghat area of north Mymensingh District in early 1976 reportedly precipitated localized famine conditions (Jenkins, 1976). It is not yet possible to forecast that famine will break out in a particular area at a certain time.
The eccentricities of nature and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
BENGAL FAMINE CODE
C H A P T E R 1. STANDING PREPARATIONS. 1.—System of Intelligence 2.—Programmes and Estimates. fi. (1)
Statements of areas liable to famine, and programmes of relief works, shall be periodically submitted to Government. (2) District Officers shall submit to the Commissioner Statements or 0f t),e Division, on or before the 1st April pro- of each year, statements in Form A—II, showing gramma of relief the areas in their respective districts which are TsUy^iub'mined'*!o considered liable to famine and the number of perGoremment. sons likelv to require relief in the event of serious famine. These statements shall be accompanied by maps,f showing the areas liable to famine (tracts more liable heing exhibited in shades of dark blue and 'hose less liable in shades of light blue). The district statements and maps need not be submitted by the Commissioner to the Local Government, unless they are especially called for. In considering the liability of a tract to famine, the following are some of the principal points which require attention:— (a) The previous history of the tract in regard to its having been visited by famines or not. (b) The density and economic condition of the population. (c) The nature of the soil and the general capability of the tract to maintain its population. (d) How far cultivation depends on rainfall, and whether the normal rainfall is regular or otherwise. (e) How far the tract is irrigated by rivers, canals, wells, & c. (f) The accessibility of the tract as regards importation of food-grains. F.stimates of the number of persons likely to require employ ment on relief works and gratuitous relief should be made with refer ence (a) to the experience of previous famines and (it) to the general condition of the various sections of the population. In revising the statements year by year, due regard should be had to the increased protection, if any, afforded during the past year by new irrigation works and improved communications, or otherwise. (3) It is impossible, under the urgent stress of actually present famine, either to select those works upon which money can be most usefully spent, or to ensure that those works which are selected are carried out upon the most economical and efficient plan. It is abso lutely necessary, therefore, not only in order that money may not be wasted, but also that advantages are not lost which more careful preparation might have secured to the country, that projects should be prepared and sanctioned in advance which will be sufficient for all practical purposes. T Alter a map has once been prepared, the preparation of a fresh one is, unnecessary, unless there has been some change in the areas liable to famine.
Figure 17.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
politics are still unpredictable.
This inability to predict unfore足
seeable occurrences, such as border insurgency or a transport strike, does not, however, mean that famine is not more likely to occur in some areas than in others.
Rather like life insurance, although we cannot yet
predict the "who and the when," we can at least estimate the "where and the
how many." The concept of areas liable to famine is not new to Bangladesh.
It has evolved as
an integral part of the pre-plan analysis required
before implementing the changing goals of famine relief policy since the British period.
The evolution of this famine policy may be grouped into
four stages and is presently in its fifth stage, viz.: 1.
Experimentation (1858-1879)
2.
Formulation of the Famine Codes (1880-1900)
3.
Preventive Policy (1901-1918) (Strivastava, 1968)
4.
Stagnation (1919-1974)
5.
Using famine as a stimulus to development (1974 onwards)
After the Bengal famine of 1866, the Scottish administrator, Sir William Wilson Hunter, sent out a questionnaire to each district officer in Bengal enquiring about the causes, severity, and potential relief works for famine in their district.
The published report of this survey
entitled "Famine Aspects of Bengal Districts" (Hunter, 1868), stressed that "different localities exhibit very different powers of resistance to the strain of famine." In very superficial and summary form, these results were later included in the Imperial Gazetteer of India (1907) which stated under the section entitled "Famine:
Tracts Liable Thereto" that:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
In an agricultural country like Bengal the failure of the crops must always cause considerable distress, the degree of which varies with the nature and extent of the failure, the material condition of the people and their character, and lastly the accessibility or otherwise of the tract affected. The great cause of deficient harvests is insufficient or badly distributed rainfall. Sometimes much damage is done by floods, and sometimes, though more rarely, by blights or locusts; but in such cases the area affected is generally limited. The crop which is most sensitive to a short or badly dis tributed rainfall is the winter rice, which requires copious showers in May and a punctual commencement of the monsoon, but it is especially dependent on the continuance of the mon soon through September and the early days of October; it is this crop which is most liable to failure in adverse con ditions. It follows that, if the rain is uncertain, the tracts most liable to famine are those in which the winter rice is the staple crop; but there, a serious failure of the winter rains is unknown and the subsoil water level is so high that, in years when the rainfall is moderately efficient, the ground retains sufficient moisture to prevent anything approaching a total loss of the crops. The whole of Dacca and Chittagong Divisions are therefore excluded from the list of tracts liable to famine. Here the only danger of disaster arises from cyclonic storm waves which at intervals burst over the country and in their wide train spread ruin and desolation. In other parts of Bengal proper, where also the winter rice is as a rule the principal crop, the immunity from famine is less complete; but the rainfall is usually ample, and the areas liable to famine are less extensive than in the other subprovinces. From time to time the submontane tracts have been swept by disastrous floods; and, when the embankments on the left of the Bhagirathi give way, floods occasionally break across Murshidabad. and Nadia Districts. (Imperial Gazetteer of India, 1907) By 1913 these ideas had been incorporated in the standing orders of the Bengal Famine Code*’ (Government of Bengal, 1913) (see Figure 17). For want of any updating or replacement, this document remains the statute for famine contingency planning in Bangladesh today.
The
The Famine Codes were compiled for the provinces of British India to outline detailed administrative procedures, e.g., famine declaration, relief works organization, etc., for implementing contingency plans in the event of famine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135 Planning Commission and the Ministry of Relief and Rehabilitation recognized
the need for disaster preparedness planning if development is
to progress unhindered (Planning Commission, BDG, 1976).
They also
realized that there are certain areas which persistently have to request food relief and rehabilitation aid as a result of natural calamities (Ahsan, 1976).
The ability to anticipate areas of need and preparedness
for the timely "targetting" of appropriate relief inputs to those areas so that the catastrophe stimulates rather than impedes development is as yet unrealized.
Although there is evidence of change, there is still
a strong tendency for relief to be allocated only as a corrective rather than also as a preventive or development measure (Dodge and Wiebe, 1976). This dissertation demonstrates a method which makes possible the mapping of areas persistently liable to famine in Bangladesh.
By
enhancing the ability to anticipate famines it improves disaster pre足 paredness.
It thus makes possible the drafting of long term preventive
plans which are meshed with the regional famine vulnerability character足 istics before potential famines occur where the risk is greatest.
The
method builds upon the six principal points which the Bengal Famine Code (Figure 17) advises should be given attention when compiling maps of tracts liable to famine, viz.: 1.
The previous history of the tract in regard to its having been visited by famines or not.
2. The density and economic condition of the people. 3. The nature of the soil and the general capability of the tract to maintain its population. 4. How far cultivation depends on rainfall, and whether the normal rainfall is regular or otherwise.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136 5.
How far the tract is irrigated by rivers, canals, wells, etc.
6.
The accessibility of the tract as regards the importation of foodgrains.
Few such maps of
areas liable to famine were ever compiled.^
The
poor response has probably been mostly because of administrative difficulties, but the problems of compiling such a map from so nebulous and non-operational a series of indicators probably encouraged the inaction.
The six principal points suggest rather than specify actual
areas liable to famine.
They avoid grappling with the idea of total
liability and the relative importance of each of the six points.
In
considering projects to provide relief, the famine code made no specific attempt to tailor such relief measures towards lessening the long term liability of the
particular areas affected.
These limitations in
mapping the areas liable to famine are the problems this dissertation tries to overcome.
^Examination of the Famine Proceedings reveals that these maps were seldom submitted (Sims, 1977). Careful archival investigation at the India Office Library and Records, London and the West Bengal State Archives, Calcutta have traced only one District report (Dinajpur, 1899) which was accompanied by such a map. This map is no longer avail足 able in the West Bengal State Archives (Mulcherjee, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER IV THE LACK OF COMPLETE DATA
On the threshold of my inquiry, I was confronted with what has been held to be a great want to an administrator in Bengal, the want of agricultural statistics. A. P. MacDonnell (1876). Report on the Food Grain Supply and Statistical Review of the Relief Operations in the Distressed Areas of Behar and Bengal during the Famine of 1873-74.
We always exaggerate the intensity of famine con足 ditions at least twofold, in order to get support, and to get it quickly. Member of the Salvation Army Delegation, Dacca, Bangladesh, 1976.
Statistical Data in British India Compared with the geographic data banks (e.g., Hagerstrand and Kuklinski, 1971) of developed countries, Bangladesh, like other areas of the developing world (e.g., Baker, 1977), but unlike most of the rest of the Indian sub-continent, has a poorly developed statistical information system.
Historically, British India developed a more complete
statistical system than other colonized areas of the developing world, e.g., the Dutch Indies (Furnivall, 1948) or French Africa (Hailey, 1938; Caldwell, 1975).
The relatively settled agricultural condition of
India's people, compared with the situation of the nomadic Taureg in West Africa or that of
the Muri pastoralists of South West Ethiopia,
greatly facilitated the gathering of social and agricultural statistics. Also the relatively greater education of the British colonial
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138 administrators compared with the predominantly lower middle-class French and Dutch foreign service officers, aided the replication of the home government's statistical system in British India.
Although the system
of gathering statistics in British India was probably more advanced than in other developing countries it was still far from complete, of dubious reliability, and prone to the biased colonial perspective of those collecting and compiling the data (Bansil, 1974 and Baker, 1975).
For
several reasons, the statistical system of British India was even less complete in Bengal. In the
late eighteenth century, the East India Company first carried
out surveys to acquire social, economic and environmental information. Those early surveys, according to Davis (1951), were "fragmentary, un足 systematic and lacking in uniformity."
More information became avail足
able after 1869 when the Statistical Survey of India was carried out in all fifteen of the British provinces.
The situation further improved
when the results of the survey were condensed into the volumes of the Imperial Gazetteer of India and published in 1881.
Then in 1885 and
1901 similar gazetteers were published which also included serial statistics such as the results of the 1881 Census of India and the 1861 Statistical Abstract. Davis (1951) has argued that these statistics, although inaccurate and lacking much of the elemental ancillary data, are sufficient to estimate trends, but not to document individual events, although such major events as those associated with reform or scandal were detailed by post facto commissions, e.g., The Royal Commission on Agriculture in India (1927) or the Commission on Famine in Bengal and Orissa (1866).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139 Even today the relative difficulties of measuring events in the sub足 continent are epitomized by the lack of an effective registration system for vital information on the population of the rural areas.
Statistical Data on Bengal In comparison with other areas of India, Bengal has had an even poorer statistical system throughout its history.
In pre-Moghul times
(pre-1201) the Aryan administration did not extend as far east as Bengal, where the local Buddhist kingdoms had an "immensely inferior administra足 tive system" (Floud, 1940).
This may have been because the land was
individually owned by each cultivator in Bengal. land ownership.
There was no community
Thus there was less need to maintain community records
of the land owned by each community member. Later, in Moghul times, the Emperor Akbar's detailed revenue assess足 ments implemented throughout India by Raja Todar Mai, the revenue minister, were not applied to Bengal because it had not yet been annexed.
The
Ain-i-Akbari (Akbar's chronicles) (1877, 1927) suggested that after the annexation of Bengal, Akbar merely retained the existing "nasaq" or lump revenue assessment system in Bengal.
In contrast with the other regions
of Akbar's India, no statistical figures were collected to show actual land areas for Orissa, Bihar or Bengal (Floud, 1940). In British times (1757-1947), Lord Cornwallis' permanent revenue settlement (1793), according to which rents were fixed in perpetuity, actually exacerbated Bengal's inferior statistical system in a "Catch22" situation.
Since there were neither maps nor reliable records of
the areas of the holdings
of individual raiyats (cultivators) in
Bengal (Floud, 1940), Cornwallis taxed the zamindars (landlords) in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140 Bengal.
This zamindari system contrasted with the raiyatwari (cultivator
brokerage) and mahalwari (community brokerage) systems in other parts of India.
In turn, this meant that the government officers, being in close
contact only with the landlords, were deprived of firsthand knowledge of rural conditions in Bengal, and
thus were unable to collect accurate
statistics to reflect them.
Statistical Data in Bangladesh At the end of the Pakistan regime (1947-1971), the country "lacked statistical information on perhaps one-half of the economy" (Rabbani, 1968).
Attempts to improve the statistical system by an advisory ser足
vice from the U.S. Department of Commerce, Bureau of Census, under an USAID agreement, were suspended in 1969 because of the decreased develop足 ment funds allocated to the eastern wing of Pakistan and the "low priority accorded statistical services by the controlling officials of the East Pakistan Government" (Elkinton, 1976).
The International Bank
for Rural Development (1974) summarized the statistical situation at the beginning of the Bangladesh period, thus: (a)
The size of the total statistical output is quite large for a country at that stage of development.
(b)
The relevance to needs of the statistics being turned out is good, with one major exception. Relatively little attention and effort is paid to the production of national accounts estimates . . .
(c)
The quality of statistics is mostly only fair, and in some fields poor . . .
(d)
The timeliness of statistics is poor with delays of as much as two to three years for many series and even more in the case of printed reports . . .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141 (e)
The impartiality of statistics is weak, with many figures reported being adjusted, usually by small amounts, on the insistence of other agencies. However, given the poor quality of many statistics, it is often difficult to judge whether statistics were closer to or further away from the truth after adjustment. IBRD, 1974. p.i.
These limitations in the national statistics are magnified at lower levels of
analysis.
Comparatively few statistical series are available
at an administrative level below that of the District.
Notable exceptions
are the National Census which records elementary data, i.e., total popu足 lation by sex, number of literates, number of households, for each of the country's 4,266 Unions (Bangladesh Bureau of Statistics, 1976), and the Bangladesh Soil Survey, which has carefully mapped and inventoried soil environment conditions for each soil association throughout the country (U N D P, 1971). Agricultural statistics still remain among the nation's most serious data deficiencies.
Rice production accounts for one-third of the gross
domestic product, but it is still "guestimated" by the combination of subjective assessments based on a 1944-45 benchmark "norm" and objective crop estimates from sample harvest areas (Elkinton, 1976).
Over-
exuberant plans for satellite crop assessments (Bangladesh Atomic Energy Commission, 1974) still remain unrealized.
The state of agricultural
statistics is perhaps best understood in terms of an official remark made by an experienced foreign agricultural advisor:
"They would be
better off if they didn't have the false security of having them." Statistics on specific events, or interruptions in production figures, even when the event is less emotional than famine (see beloxtf), are at least as incomplete as those in the standard statistical series,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142 if not more.
Some of the results of incomplete reporting of events,
in this case floods, are revealed in Table 4 (IBRD, 1972).^ cases data are nonexistent.
In some
They may never have been collected, either
intentionally or unintentionally.
They may have been collected and then
lost through negligence, perhaps even to termites.
The data may have
been collected and then destroyed through civil disturbance or natural calamity.
They may perhaps have been officially destroyed as were many
"Class C" government documents after independence in 1971.
Taccavi
loan payments distributed as a form of disaster recovery grant since Moghul times would have provided a basis for mapping disaster prone areas, but because they were
only "Class C" documents they too were
destroyed in 1971. Data on events may also be unreliable, either because of actual errors, or because of bias in reporting due to pressures from either the victims of the events, or from those who must deal with the event. Furthermore, information for some events may not be accessible, either because it cannot be located because there is a (perhaps monetary) human problem in securing the release of the data, or because the data are classified.
The last problem increases if the information needed is
for a localized situation, thus requiring detailed information at a large scale, e.g., air photographs.
(As regional problems normally require
only the less-detailed satellite images, the data are easier to acquire.) Non-availability of information also increases if the problem being
â&#x2013; ^The statistics available to IBRD are not necessarily available to a researcher in Bangladesh, although the opposite also holds.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
Estimates of Flood Damage (Millions of Takasâ&#x20AC;&#x201D; Current Prices) 1962 estimates of Aman crop damage by district: District Mymensingh Sylhet Rangpur Bogra Pabna Dacca Raj shahi Faridpur Total
W A P D A 55.52 36.75 3.25 19.58 63.69 60.42 4.89 10.61
76.808 88.666 25.525 43.419 43.283 47.797
254.71
391.227
1962-70 estimates of losses to crops and property: Relief and Rehabilitation Dept.
Year
1962 1963 1964 1965 1966 1967 1968 1969 1970
Agricultural Directorate
554 ? 183 ? 577 ? 1164 82 1380
â&#x20AC;&#x201D;
65.720
Agricultural Directorate 707 60 158 36 480 84 1186 307
1
(IBRD, 1972)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144 studied encroaches upon a sensitive topic, or if it involves sensitive geographical areas such as those near international boundaries. Data may be available only for certain periods of time.
Some
statistics, e.g., the census, are only published every ten years.
Others
are not available for periods of disruption, e.g., a civil war, either because of a failure to record during the disruption or an unwillingness to release data for that period at the time.
Some may be made available
either deliberately or by default, only after the time when they are crucially needed. Finally, some statistics on events are available only for certain areas.
Particular districts or units may not have sent in reports.
The
data may have been collected only for a certain project in a certain area, e.g., tube-well sinking in parts of North Bengal; and even then, limited data may be available only at the aggregate level, such as the district.
Data During Famine Conditions In the previous chapters, some of the symptoms and causes of famines were reviewed.
These characteristics make the gathering of complete
data in famine situations particularly difficult. As the famine evolves, there is a gradual erosion of the social and administrative framework, including a deterioration in the national statistical system prior to the actual outbreak of crisis conditions. The resulting backlog of record keeping and collation inexorably entices civil servants to fudge information.
They may merely replicate past
records so that any change is not recorded.
Later, when it is acknowl足
edged that a change has occurred, they may overcompensate to ensure that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145 the change is visible in the record to their superiors.
Those officials
involved in record keeping may be temporarily involved in, or assigned to, other ad hoc duties as social and economic conditions deteriorate.
Lower
level record-keeping officials, if not devoting their energies to ensuring the survival of
their family and relatives, may have little
incentive to complete records and forms, particularly if the central administration has failed to pay their wages during the same period. The
poorly developed areas, where the prevalence of famines is
greatest and the effects most severe, are often remote from centers of communication as well.
They are thus the areas most likely to be cut
off from the information collation and evaluation center.
They are also
the areas where past calamities may have already disrupted the statis足 tical system. Famine is an abnormal event.
Different phenomena occur which require
new statistical systems if they are to be measured.
During normal times,
however, it is not feasible to maintain a continuous record of such phenomena.
Thus, begging is not perceived as a problem in normal
times, and it
little behooves a developing country with scarce
resources, to set up a continuous surveillance system to monitor begging patterns.
Unusual data sources have to be tapped for information
during a crisis.
The records of the Anjuman Mufidul Islam, the
charitable Muslim burial society, were utilized to examine street death trends in Dacca, but neither those who designed the records, nor those who kept them as suitable meals for the termites that had eaten through the protein rich ink of earlier records, could anticipate the importance of such data for
documenting the incidence of famine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
Crisis, by nature, demands action, not words. even spurn thought.
It may, in fact,
Careful record keeping is synonymous with words.
Masefield (1963) has pointed out the scarcity of documented relief studies.
Death and starvation engender relief action rather than experi足
mentation from the humanitarian, whether he is a Bengali or a foreign national.
The classical "control experiment" of the scientist is
unethical in such situations.
It is immoral to give a "placebo" for
food relief to a control group in a famine area.
Without this control
on results, the interpretation of data gathered in a famine situation may be ambiguous. Disruption normally occurs prior to the outbreak of a famine. There is often an unusual displacement of areas.
people from the rural
Before a famine, large scale river erosion or flooding may
destroy some homes and force other households to move.
Thus, households
which might be the unit for measuring the intensity of famine symptoms, are
no longer there.
Existing maps, and even recent aerial photographic
coverage, become outdated and occasionally useless, as an aid to data collection.
Efforts to raise purchasing power by the sales of
possessions may include the transfer
of a door section, which may have
on it the malaria eradication number, to an owner several villages distant.
The
loss of the household malaria eradication number effective足
ly means the loss of that household from any survey of households for purposes of monitoring the effects of famine (e.g., Mahalanobis, 1946). During famines, the groups with
the largest "non-response" in
surveys may be the most affected group: who have migrated.
the poor, the dead, and those
Traditional famine surveys, like those of Mahalanobis,
may have omitted many of the poor whose houses were not included on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147 union parishad (local council) tax roll, or who, indeed, may have had no household.
The famine victims may have left their homes.
As
destitutes, they may have followed daily begging patterns which brought the household together around a cooking fire, not at their former house hold location, but near a town bazaar, often only after nine o â&#x20AC;&#x2122;clock at night. The emotion associated with famine exaggerates any mistakes.
The
opposing political party in a crisis may exploit statistics that condemn the ruling party (e.g., Hariharma, 1976).
Jodha (1975) has described
how the involvement of politicians in Rajasthan's famine-reporting system has made reports from local officials irrelevant, reducing their function to that of "furnishing statistics for decisions already taken." The Bengal Famine Code (Government of Bengal, 1913) did describe a famine warning and monitoring
system whereby local government officials were
responsible for regularly reporting famine symptoms to the provincial capital.
However, even if such a system had been maintained, it is
doubtful whether officials could have contributed accurate information to such a system while under the strong political pressure existing in some areas in late 1974.
As an example of such difficulty, there were
three assassinations of local government officials in Kurigram sub division, which were reputed to have been at the behest of the local member of parliament. The local authorities may wish to exaggerate figures on the degree of hardship in their area in order to secure increased aid from the central government.
The latter may, however, play down the magnitude
of a local problem until such time as they are pressed into declaring
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148 a crisis, or until evidence appears to substantiate a request for inter足 national aid.
A central government policy may indirectly bias statistics
emanating from a famine area.
Thus Alan Berg, in an article with the
optimistic title, "Famine Contained," relating to the Bihar experience in 1966-67, inserts a small footnote which states: Local officials were personally held responsible for starvations in their jurisdictions and thus, any report from them of a death, would reflect their own inability to prevent it. As a result, deaths were sometimes attributed to other causes. (Berg, 1971) Once the central government has declared a famine, however, the statis足 tics detailing the distress may be exaggerated "in order to get support and get it quickly" (Member of Salvation Army Delegation, Dacca, 1976). Whether
exaggerated or underestimated, the statistics provided in times
of famine can rarely be trusted. Relief statistics which are gathered in times of famine are often gathered to justify or deny the need for distribution of relief, and other attempts at famine amelioration. Because relief strategies are short-term, they seldom contain anti足 corruption measures.
Thus, they are open to exploitation.
Jodha (1975)
has quantified the differences between recorded figures and actual figures (Table 5) for a relief operation in Rajasthan.
Even a figure
traditionally used in the declaration of famine,^ the record of the number of days that relief was said to have operated, was exaggerated from an actual figure of 73 to a recorded figure of 95.
Table 5 also
2 In the Bengal Famine Code, 1913, the District Officer had to submit a report with a view to famine being declared when (a) test works began to attract workers in considerable numbers or (b) gratuitous relief appears likely to be required on a considerable scale, i.e., when 1/2% of any thana or larger area in the district has been on relief for two months continuously.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149 Table 5 Differences Between the Record and Reality of a Relief Operation (Based on details of 113 laborers at one location in Rajasthan, India)
Item Number
of days relief work operated
Average number of workers employed per day (for 73 days) Road Construction per worker per day (feet)
As Recorded
In Practice
95
73
935
462
1.2
2.5
Total cash wage payment per week (Rupees)
4907
2569
Total grain (wage) payment per week (quintiles)
36.7
16.2
115
35
45.6
37.3
Expenses on water and other facilities per week (Rupees) Average earning (cash + kind) per worker per month (Rupees)
(After Jodha, 1975)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150 shows that there was a threefold exaggeration of the "weekly expenses for water and other necessities," i.e., items that are more difficult to monitor than "days that relief operated."
The difficulties of any
individual official attempting to correct such discrepancies was high足 lighted in a vignette about a revenue collector in Barmer, who visited relief works in the district of western Rajasthan in India in 1970: He found that in most areas the number of workers as per the muster-roll exceeded the total population of the areas by 50 to 70 percent. Disgusted by this, he ordered complete closure of all relief works in certain tehsils and initiated the supply of relief (part loan, part dole) on the basis of ration cards. He found this method of relief very economical and effective. He was transferred to a district of South Eastern Rajasthan where droughts are very uncommon. (Jodha, 1975) A final characteristic of a famine situation which may affect the completeness of the data is the influx of foreign nationals.
These
outsiders from information-conscious countries may seek to improve or even circumvent the local information system, but because of their naivete in the novel cultural situation, and their reliance on indigenous sources of
information, they may not only be provided with spurious
information, but they may also misinterpret the data that are provided (Ratcliff, 1977). Neither should the cloak of sanctity associated with relief activities hide the opportunities for deliberately biased information associated with foreign intervention programs.
Nor should it be assumed
that the activities of the foreign nations involved in the crisis situation are incorruptible.
As Cuny et al. have pointed out (Davis
1977), the outside organizations in a disaster
situation generally
have no accountability to the famine victims and little, if any,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151 accountability to anyone.
They do, however, have self interest which
may be buttressed by "bending" the data.
The United Nations
organiza足
tions need to "show how valuable they are to ensure their future growth and funding" (Davis, 1977).
Overseas governments often have to "fulfill
their foreign policy aims by rewarding faithful allies, and at the same time provide opportunities for businesses at home."
Overseas charities,
too, have to "show their supporters and rival agencies how valuable they are to ensure their future growth and
funding," and the "foreign
experts often have a need to impress rivals with their superior knowledge or skills" (Davis, 1977). With famine characteristics "blurring" the collection of statistics, it is not
surprising that the record of famines is incomplete.
A
glance at the historical record of famines in the various gazetteers and statistical accounts reveals the lack of consistency among records (Figure 16).
The records were certainly published in different periods,
ranging from 1874-1976, but it is nevertheless remarkable that there are only two instances where all five sources agree that a famine or scarcity occurred in that district, in that year.
Even this apparent
consistency may only result from occasionally consistent plagiarism from the first record to the subsequent records. In the absence of a complete historical record of famines in Bangladesh, it was surprising to find that 75 percent of 135 decision makers associated with famine problems in Bangladesh (see Figure 18), replied
in the affirmative to the interview question "Do you think it
is possible to predict when local famines will occur in Bangladesh?" (See Appendices B, C, D.)
Indeed, 96 percent of one group of decision
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152 makers, all Bengali retired government
officers, from the total of
135, agreed that it was possible to predict famine.
However, closer
inspection of the responses revealed that less than 1 percent of
the
101 initially responding in the affirmative, did so without qualification. The other 99 percent, all qualified their affirmative reply by indicating that prediction was only possible after a disaster or series of disasters had occurred.
The majority also added the qualification that
additional monitoring of information on purchasing capacity, rice prices, storage levels, migration patterns, and local reports would have to be carried out if the famine were to be predicted.
Even such a
short term prediction is thus predicated upon the organization of an improved nationwide data base. Any attempt to map areas liable to famine in Bangladesh is thus faced
with a lack of complete data:
the historical record of statistics,
although better than many areas of the developing world, is nevertheless incomplete because of the particular revenue system used in Bengal, and the present day statistical system is mediocre because of the recent political, administrative, and environmental disruptions.
Data relating
to famines, both historically and at present, are particularly prone to bias and incompleteness.
Because of the lack of complete data, it is
not yet possible to forecast famines in Bangladesh.
This dissertation
argues, however, that rather than waiting for several decades for the availability of data like those in the developed world, it is possible to use experienced judgments to carefully select sufficient information from the presently available data to map the areas liable to famine in Bangladesh.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER V METHODOLOGY:
THE D EL BE CQ -D EL PH I
PROCESS
But we cannot wait under the present conditions of the world . . . Until we get all the exact and detailed information. Lord
Boyd Orr
Quebec Speech, 1945
The proper place for the scientist— once in a while at least— is in the midst of the unknown, the chaotic, the dimly seen, the unmanageable, the mysterious, the not yet well phrased. This is where a problem oriented science xrould have him be as often as necessary. And this is just where he is discouraged from going by a means stressing approach to science. A. H. Maslow, 1946
In Bangladesh there is a dearth of data on famine problems, but there is a group of knowledgeable people in the country who have experi enced repeated famines and other environmental disruptions.
Because
there were virtually no objective data available on either the areal distribution of famines or their causes in Bangladesh,^ I decided to use an intuitive method.
As a foreign researcher new to the country
and to the problem, I assumed that the judgment and insight of those with experience with the problem of famine in Bangladesh could provide me with opinions that would be more likely to be correct than my own naive speculation.
I therefore proposed to determine the consensus
■*■566 Chapter IV.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
Probability of
Speculation
Opinion
Knowledge
of knowledgeable opinion by integrating two social science techniques, the Delbecq method and the Delphi method, into a technique that I am calling the 'Delbecq-Delphi Processâ&#x20AC;&#x2122; (Figure 18).
I assumed that the
integration of these group techniques would best allow me to reach a consensus of views while minimizing the potential effects of my own lack of training in group psychology. Both the Delbecq and Delphi methods involve attempts to elicit a consensus judgment from a nominal group, i.e., a group where verbal exchange and argument are excluded.
The Delbecq method or process has
been principally used to derive a list of items pertinent to a par ticular problem and then to evaluate each item against a rank or ordinal scale.
The Delphi method, on the other hand, has been used to evaluate
a list of items against an interval scale of measurement.
Derived from
planning programs involving citizen participation, Delbecq has only recently been used in the social sciences.
It remains untested.
Delphi
is an older method which has been used by social scientists in laboratory situations, but it too remains untested in a field situation.
Both
techniques are designed to minimize bias in committee decision making.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE DELBECQ-DELPHI PROCESS
2. Editing by Technical Experts:
3. Delbecq Sessions:
F .gure 18.
The Delbecq-Delphi Process
155
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1. Open-Ended Questionnaires:
156 The Delbecq Technique The Delbecq technique, initially called the Nominal Group Technique was developed by Andre L. Delbecq and Andrew H. Van de Ven in 1968.
As
described in the literature (Delbecq and Van de Ven, 1971 and 1972; Delbecq, Van de Ven, and Gustafson, 1975), the technique has four stages: 1.
Members of the group silently write down their ideas.
2.
The essence of each individual's ideas is presented to the group as a whole.
3.
Group members discuss each recorded idea for clarification and evaluation.
4.
Individual group members vote silently on priority ideas, with the group decision being mathematically derived through rank ordering or rating. (Delbecq et al., 1975)
Although stages 1, 2 and 3 are adapted and integrated into "The DelbecqDelphi Process" in this research design, I have termed the rank ordering strategy of stage 4 as the "Delbecq Session." This rank ordering strategy requires that the members of the nominal group, upon request from a group mediator or non-voting chairman, rank a specified number of items from a given list by writing a value (N), where N corresponds to the total number of items, on a card next to the item judged to be the most important.
The value (N-l) is then written
on another card against the item judged to be the second most important and so on until all the items have been given values.
Then the total
'judgment scores' for each item are tallied for the entire group, and the items are ranked by total scores.
That is, the item with the highest
aggregate score is ranked first, the item with the next highest aggre足 gate score is ranked second, and so on.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157 To my knowledge, this simple technique has never before been used in either a geographical study or in any study of a problem in the developing world.
The Delphi Technique The Delphi Technique's first documented use was to improve the prediction of
winners at horse races in 1948 (see Quade, 1967).
The
technique received the seal of approval in social science, however, when Dalkey and Helmer used it in 1953 to evaluate the potential "effects of strategic bombing on individual targets in the U.S." (Dalkey and Helmer, 1963).
During the 1950s and 1960s, studies by other proponents of Delphi
at the RAND Corporation provided rather indirect support for the tech nique's predictive powers.
For example, student groups in several
instances were asked to "guestimate" facts verifiable from almanac sources, and their answers proved generally accurate (Brown, Cochrane, and Dalkey, 1969; see also Helmer and Rescher, 1959).
However, none of
these studies dealt with "real-world" problems. The Delphi technique has evolved through this research to become a method of refining and quantifying group judgments on variables about which exact knowledge is not available.
There have been three basic
features of Delphi: 1.
Anonymous Response by each individual in the group, e.g., by questionnaire, to minimize bias from dominant members.
2.
Iteration and Controlled Feedback allowing each individual in a group to re-evaluate his or her individual judgment in a second round after he or she knows the groupâ&#x20AC;&#x2122;s initial consensus judgment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158 3.
Statistical Definition of
the Group Response by use of the
median response to ensure that the judgment of each individual member of the group is represented in the group response and yet to minimize the effects of a few extreme judgments.
Thus
the need for group uniformity is circumvented. The rationale for the technique, based on the adage that two (or n) heads are better than one, is illustrated in Figure 19, which suggests that in a second round of evaluations the nominal group's average response on any item will converge towards the median response of the group on the first round.
Also the average response is likely to move towards the
most plausible value, because respondents who were unsure in their initial response will change more than others in the second round. counter argument that it would not be the
There is a
knowledgeable group members,
but rather those with stubborn or dogmatic personalities who would have a propensity not to change their weighting in the second round, but this has been partly dispelled by the Mulgrave and Ducanis (1975) study, which compared Delphi participants' ratings on Rokeach's Dogmatism Scale (Rokeach, 1960) with their willingness to change their views between rounds one and two.
Those with "dogmatic" personalities were found to
be more likely to change, perhaps because in the absence of any clearly defined authority to follow, they saw the first round median as the "authority." By the 1970s over 1,000 Delphi studies had been completed (Linstone and Turoff, 1975).
Although the technique "had catapulted into inter足
national prominence" (Sackman, 1974), with studies assuming its effective足 ness appearing in the United States, United Kingdom, Sweden, U.S.S.R. and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
- Wide
deviations
of
group.
the
- Deviations
WORST T he m e d i a n of
the
GROUP or
group
depend
upon
responses
group
size
â&#x20AC;&#x2122;S w i n g e r s '
-
'Hold-outs'
th e g r o u p
the
Thus
those u nsure of
-
those
sure show
deviation narrows
of
to
judgment a value
the most
less and
50% of
or (M)
the
the m edian
JUDGMENT
average is
who
are
far
the
of
the m ost
than more
individual
nearer
response
nearer
(T)
likely value
inclination
members
expertise.
GROUP
likely v alue
their
their judgment
1 median
group
The m e d i a n
response
is n e a r e r
reassess
th e r o u n d
and
individual
NORMAL
st p r o b a b l y t r u e v a l u e (T) t h a n 50 % o f t h e i n d i v i d u a l r e s p o n s e s .
median,
from
JUDGMENT
average (M)
in t h e
than
responses.
from the
group
round
or
those
already near
to r e a s s e s s
their
judgment.
approaches
the mo st
1
response.
probably
true value.
F i g u r e 19.
Graphic
Representation
of
the D e l p h i R a t i o n a l e
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
Japan; literature critical to Delphi was sparse by comparison.
Those
attacks on the technique that have appeared have cited (1) the limitations of other forecasting techniques in dealing with surprise events, inadequate data and unpredictable interactions (Quinn, 1971); (2) the unproven assumption that "experts” are more knowledgeable than the general popula tion (Bedford, 1972); (3) the loss of data because of the lack of inter action by participants in Delphi sessions (Milkovitch, Annoni, and Mahoney, 1972);
(4) the inadequate attention paid to the psychological
values of the participants (Weaver, 1970);
(5) the failure to capitalize
on the extensive mathematical literature on subjective probability theory (Morris, 1971); and (6) the uncritical acceptance of the median value as the best estimate of the expert opinion without removing possible con founding factors (Derian and Morize, 1973). The most thoroughly disparaging review of Delphi was that by Sackman (1974) who evaluated the technique against the American Psychological Association's standards for social experimentation and opinion questionnaires.
He concluded that conventional Delphi "neglects
virtually every area of professional standards for questionnaire design, administration, application and validation." Sackman particularly criticized most Delphi studies because of: 1.
their "lack of adequate formalization and definition of the target domain";
2.
the lack of precision in "operationally defining the characteristics of the panelist sample," their relevant professional experience qualifications and the directions under which they made their judgments;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161 3.
the lack of "a systematic relationship between the selected panelist sample and the objectively defined population";
4.
the lack of documentation of the "drop-out group" from the sample selected;
5.
the lack of careful documentation of how the comprehensive
6.
the lack of documented procedures for pruning down the initial
7.
the "lack of careful definition of the items evaluated";
8.
the lack of "statistical documentation of variance in responses";
items are elicited from panelists;
list of items to the final set used for the study;
9.
the lack of documentation of the extent of validity between independent judgments and dependent judgments; and
10.
the lack of careful documentation of the "accountability for the interpretation of the results."
Reviews by proponents of Delphi (e.g., Pill, 1971; Welty, 1972 and Linstone and Turroff, 1975), have responded to these criticisms by arguing that "The Delphi technique should be used at high levels of uncertainty, and that
one must accept the difficulty of gauging its usefulness."
It has elsewhere been suggested that "its eventual usefulness will be judged by its performance, rather than by any abstract of its worth" (Pill, 1971). Despite these criticisms, Delphi has been applied to a number of uses in different fields, including management (e.g., Steiner, 1969), futures studies (e.g., Bedford, 1972), business (e.g., Milkovich et al., 1972), clinical medicine (Inglefinger, 1974), and food policy planning
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162 (Agarwal et al., 1976).
In addition, its techniques are also slowly being
adapted by geographers.
Paul and Landini (1975) undertook a Delphistudy
to "determine what were the most important land use data items to be interpreted from remotely sensed imagery for the Santa Cruz mountains," but the first explicit use of Delphi by a geographer was the study in which Smil (1975) investigated "Future Developments in Energy Resources and their Environmental Impacts."
Later, one of Smil's students applied
the techniques to forecasting urban growth in Winnipeg (Thraves, 1976). The effectiveness of Delphi in these geographic studies was neither con足 firmed nor in any other way evaluated, however.
Moreover, the studies
included minimal descriptions of the methodology.
Both studies appear
to have credulously accepted Delphi with no reservations.
This same
blind faith is apparently being transferred to research on environmental perception.
Geographers editing the report on the Man and Biosphere
Project 13:
Perception of Environmental Quality, have without restraint,
encouraged the use of the Delphi technique in policy formulation (Burton, 1973).
Thus the general lack of rigor discussed by Sackman (see above)
is being carried over into Delphi studies The focus of this dissertation is to
in geography. provide a solution
to the
real problem of delimiting areas liable to famine in Bangladesh. not primarily designed as a test of a methodology.
It is
However, one of the
essential by-products of the research reported here is an evaluation of the results of the subjective Delphi ratings by comparing them with real-world phenomena.
Moreover, to insure that the technique itself is
evaluated, rather than merely the implementation of the technique, I have attempted to compensate for Sackman's criticisms (see above) by using and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163 documenting a significantly more rigorous methodology than those used in the earlier applications of Delphi to geography.
Specifically, I have
countered Sackman's criticisms with the following steps: 1.
Careful definition of the area of inquiry.
2.
Careful documentation of
3.
Careful selection of
the characteristics of panelists.
panelists on the basis of professional
training and scaled experience levels, and documentation of the selection procedures used. 4. 5.
Careful documentation of the 'drop out' group in this study. Careful documentation of
the elicitation of comprehensive
items from the panelists. 6.
Careful documentation of the selection of specific items for evaluation from among the elicited list of items.
7.
Careful
definition of the items evaluated.
8.
Careful
documentation of variation of scores between groups.
9.
Careful
documentation of the extent of agreement between
independent judgments (1st round) and dependent judgments (2nd round). 10.
Careful
documentation of the limitations of the test
scores.
The Delbecq-Delphi Process
Introduction The Delbecq-Delphi Process used in this study (Figure 18) consisted of four stages through which it was possible to elicit, refine, evaluate and weight the importance of factors that might reasonably correlate with the areas liable to famine in Bangladesh.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164 Stage 1 consisted of administering an open-ended questionnaire to four groups, each of 45 people, who are or have been involved with famine problems of Bangladesh:
village elders, senior government officers, aid
distributors, and technical experts.
2
These open-ended questionnaires
(see Appendices) attempted to derive a profile of the training and experience
of
the respondents, elicit their definitions of famine,
find out whether they thought famine prediction was possible, and deter mine whether they thought certain areas in Bangladesh were more liable to famine than other areas.
If they thought there were areas more liable to
famine than others, they were asked to list the factors they considered led to this greater liability.
Next, each respondent was asked to
identify on a map those specific areas he or she believed to be especially susceptible to famine and to state for each such area the factors they associated with the famine liability.
Finally, the respondents were asked
to suggest possible preventive plans that either the villagers themselves or the government could implement in those areas. Stage 2.
The 60 factors elicited in Stage 1 were split into nine
general areas of concern (e.g., agricultural concerns, economic con cerns) .
Nine technical experts— one expert for each area of concern—
were then contacted individually and asked to edit the factors subsumed under
their areas of expertise.
The editing consisted of two stages.
First the experts eliminated from future consideration those factors for
2
These persons are all individually named in the acknowledgments on the "Map of Areas Liable to Famine in Bangladesh" (see Appendix H).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165 which reliable information on each area of Bangladesh was lacking. Next each expert determined for each remaining factor a â&#x20AC;&#x2122;critical value' such that
if the critical value was exceeded in a geographical area the
factor was likely to contribute to famine liability. Stage 3 .
A list of the
thirty remaining factors together with their
critical values were then distributed to three separate groups of nine persons each:
3 nine
retired senior government officers, nine aid
distributors, and the same nine technical experts who participated in Stage 2.
Unfortunately, it was not logistically feasible to involve the
village elders in this stage of the process.
The three groups were then
invited to three nominal group meetings at separate times.
In this
Delbecq session they each ranked the five factors they thought were the most important in determining the areas liable to famine in Bangladesh. Stage 4 .
After amalgamating their rankings a list of the ten most
highly ranked factors together with their critical values were then evaluated in greater detail by the same groups of nine in a Delphi session.
Each member of the group was asked to score on a card (see
Appendix E) the list of ten factors on a scale 1-100, just as they might do in scoring an essay or examination.
If a group member judged
the areas liable to famine to be exactly equivalent to the areas for which the critical value of a given factor was exceeded, then that factor was assigned a value of 100.
Factors judged to match less well
with the areas liable to famine were assigned lower values.
After each
of the ten factors had been assigned a value, the mediator, in this case
3
These 27 persons are also individually named in the group acknowledgments on the Map of Areas Liable to Famine in Bangladesh (see Appendix H ) .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
the researcher, selected the median value which had been previously selected by the nine panelists for each factor.
The group w a s ’ then
given the median scores of the values they had assigned to weight the ten factors again.
and were asked
Some members changed their judgments
while others remained the same on some or all of the factors.
The
median values assigned in this second round by each group were used as the group’s weighting of the importance of the factors. After standardizing these weightings for each group and then com bining the weightings for all three groups, it was then possible to com pile a composite map by amalgamating the areas defined by the critical values with their assigned weightings. index of
The map thus derived was an areal
famine liability, where higher numerical values indicated
judgment of greater famine liability.
This map was then compared with
the combined individual mental maps of areas liable to famine that had been generated
in Stage 1.
Elaboration As noted in the outline above, four groups of people involved in decision-making related to famine problems in Bangladesh were included in the initial open-ended questionnaire survey:
village elders, retired
senior government officers, aid distributors, and technical experts.
The
four groups were chosen because of their combined experience with a broad range of famine problems.
Although the group of village elders
was' chosen to represent experience at a local scale, the other three groups were chosen because of their nationwide experience. retired government officers had
The senior
long experience in different areas of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167 Bangladesh and had dealt with famine problems before.
The aid distrib足
utors were the contemporary decision makers involved with relief and rehabilitation programs during and after the 1974-75 famine in Bangladesh. The technical experts were from those groups who presumably provide the background information for contemporary decision making.
Unfortunately,
the study omitted a fifth group, the landless laborers who are most severely affected by famines. "elitist."
The method may therefore be viewed as
This is a failing, but it is also a realistic decision in
that famines may reflect the inability of landless laborers to make development decisions under the status quo.
Also the landless laborers'
knowledge of conditions at a national scale may be even more limited than that of the village elders. The Village Elders (Murrubbis).
The group of Bengali-speaking
village elders was selected from nine villages along an east-west line through the center of a reputedly famine-prone thana, Bhedarganj, in Madaripur subdivision of Faridpur District.
For each of four villages
on the adi (stable land) and five villages on the char (unstable riverine land), five murrubbis were nominated by the primary school master, who attested that the men nominated were indeed relatively established and respected village elders rather than matbars, the members of power groups who have political ties to the urban centers.
None of the
murrubbis chosen claimed to be less than 30 years old, and the average age claimed was 53.
Sixty-seven percent had been born in their villages,
and all but 3 of the 15 who were not, had lived within Bhedarganj Thana since before the great Bengal cyclone of 1919.
All the Murrubbis could
could recall the time when Bangladesh was part of British India:
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
168 percent could recall the 1919 cyclone.
Although 24 percent of them
could not read or write and seven percent were only educated through the teachings of the Quran, 29 percent had completed primary education, 31 percent had attended a high school or a madrasa (Islamic High School), and 4 nine percent had
matriculated.
None, however, had gone to college.
Eighty-seven percent of them were owner-cultivators and only seven per cent were tenant sharecroppers.
One was a businessman and one a teacher,
but none were in government service. Senior Retired Government Officers.
This group of 45 were
identified
through 9 initial nominations by the Fresident of the Retired Government Officers' Club of Bangladesh, and then by seeking additional nominations from the initial nine. over 60 years of age.
All the retired senior government officers were They all had excellent command of English with a
B.A. degree, usually from Calcutta University (66 percent) or from Dacca University (22 percent), but a few had been at Aligarh University. Sixty percent had an M.A. or L.L.B. as well.
Most had been disciplined
in history, economics, political science, English literature, or law, but eleven percent had been trained in the sciences, before entering the â&#x20AC;˘ service as district magistrates.
Most of them had joined the Bengal or
Assam Civil Service or the Indian Police Service in the 1930s.
They had
been in service for an average of forty years and each had served in an average of ten posts, in both urban and district stations.
Ninety perÂ
cent had at least visited every one of Bangladesh's nineteen former
To matriculate in Bangladesh means to pass school final examination usually at age 13. It does not have the traditional English meaning of to enroll or to attend a school.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169 districts, and many of them had visited all of the fifty-nine earlier subdivisions.
Only 10 percent had been posted in fewer than 14 districts
and only 9 percent had visited fewer than 15 districts.
Many of them
(50 percent) had also served in Pakistan (former West Pakistan) and 40 percent had taken work-study tours to such places as the U.S.A.
Fifty
percent of them had reached the rank of Secretary before retirement in the early 1970s.
Twenty-two percent of them, moreover, had recently
been recalled from official retirement to serve in senior administrative or consultative capacities to 1976.
the Martial Law Government established in
Ninety percent had been involved in relief administration during
their service career, and 45 percent had worked in famine relief during the
1943 famine.
Seventy-eight percent of them had read through the
Bengal Famine Code. Aid Distributors.
This group of 45 were selected from a list drawn
up by the staff of the Ministry of Relief and Rehabilitation, the United States Agency for International Development (USAID) Dacca staff, and by the Director Agricultural Development Agencies in Bangladesh (ADAB), the coordinating body for voluntary agencies in Bangladesh.
Roughly
half of this group were Bangladeshis and half were foreign nationals from the United States, the United Kingdom, Sweden, Germany, The Netherlands, Australia, and Sri Lanka.
This 23:22 ratio is, I think,
representative of contemporary relief and rehabilitation decision-making at a national scale in Bangladesh.
They were all administrators in aid
distributing organizations, with 11 percent from the Ministry of Relief and Rehabilitation, 18 percent from private Bengali aid agencies, 16 percent from technical aid sections in the various foreign embassies, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170 the remainder from foreign-funded aid agencies in Bangladesh.
The
foreign nationals included persons on the traditional two year assignment abroad.
Most had either had "on the job training" with a relief type of
development mission, or with the civil service, or else had been educated in either international management or administration.
Some 11 percent
of the group had also gained broader experience (and probably some biases as well) from other famine areas such as the Sahel or Nigeria.
During
their time in Bangladesh, 20 percent of the foreign nationals had never been outside Dacca, and only 14 percent of them had worked outside a district town.
On the average, however, the aid distributors had been to
all but four or five districts of Bangladesh.
Eighty-one percent of
them were professionally involved in relief or development programs operating in one or more districts. relief
Sixty-six percent were supervising
or development programs in areas they themselves later declared
to be "areas liable to famine." Technical Experts. expertise.
This group of 45 was selected from 9 fields of
These fields were initially identified from factors sug足
gested in the Bengal Famine Code (Government of Bengal, 1913) as being problems pertinent to the delimiation of famine areas in Bengal. the help of the
With
World Food Program Representative in Dacca, I modified
these suggested factors to fit contemporary expertise. The nine fields were: 1. 2. 3. 4. 5. 6. 7. 8. 9.
Agriculture Agricultural Economics Climatology Food Storage Population Problems Public Health Nutrition Social Science Transport Logistics Water Resources
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Experts from within each field of expertise were nominated by both the Planning Commission of the Bangladesh Government and by the staff of USAID, Dacca, each nominating approximately 50 percent of the group. Again, about half were Bangladeshis and half were foreign nationals from the United States (19 percent), the United Kingdom (11 percent), The Netherlands (3 percent), Norway (5 percent), South Africa (3 percent), Australia (3 percent), and Haiti (3 percent). were women.
Only 5 percent of them
The technical experts served as staff members in departments
of the Bangladesh Government (31 percent), other Bengali institutions (23 percent), and as overseas technical consultants with the United Nations (20 percent), bilateral missions (14 percent), or other agencies (11 percent).
Fifty percent of the technical experts had visited all
but two of Bangladesh's 19 districts, but only 30 percent of them had worked in the rural areas outside of a district town for more than three months.
Sixty percent of the foreign Technical Experts had been working
in Bangladesh for over five years.
Thus the average technical expert
had worked in Bangladesh between seven and fifteen years, substantially longer than the average aid distributors.
Only 22 percent of the tech足
nical experts claimed that they had read the Bengal Famine Code however. Stage 1.
Open-Ended Questionnaires
A basic open-ended questionnaire was designed.
It included 18
general questions about famine, famine prediction, areas liable to famine, famine prevention and the backgrounds of the respondents.
The basic
questionnaire was then adapted for administering to four different groups of people:
village elders, senior retired government officers,
aid distributors, and technical experts (Appendices A, B, C, and D).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172 The questionnaires were then pretested with a class of Bengali university students and a small number of representatives from each of the four groups.
After modification the questionnaires were then administered to
45 people from each of the four groups. The questionnaire for the village elders was translated into Bengali, and questions relating to areas outside the village were removed.
The questionnaires were administered in Bengali to the village
elders by Dacca University students with homes in that region, after the researcher had visited the area and chosen the sample of murrubbis and monitored the pretesting of the questionnaire.
Sixty-six percent of
the questionnaires to the other three groups were administered by the researcher himself.
The remaining 34 percent were administered by his
principal research assistant who had accompanied the researcher on the initial interviews. The interviews were conducted in relaxed conditions.
Those with the
village elders were usually administered with the interviewer and respondent both sitting on piris (3-4 inch high bench seat for squatting on) and surrounded by family and village members, some of whom might add supportive or corrective information in between puffing contentedly on their hookah (water-pipe).
The interviews with the retired government
officers were normally conducted over nashta (refreshments) either in the early morning or just before dusk or at night.
During Ramzan (the month
of fasting), the possible hours were limited by the five wagt (times) of namaz (prayers), which were rigidly adhered to by the majority of the retired officers.
The average times for interviews with these first
two groups were 60 minutes and 75 minutes, respectively.
Interviews with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173 the aid distributors and technical experts were substantially shorter, lasting for an average of only 35 minutes and 50 minutes respectively. The majority of these interviews were conducted in the offices or at the place of work in less relaxed conditions, but attempts were made to interview as many of them as possible when they were at home.
The inter足
viewing of these four groups continued over a period of eight months; no one refused to be interviewed.^
In each of the 180 interviews, the
respondents were first asked for their definition of both scarcity and famine and they were next asked to describe the critical point at which scarcity became famine (see Chapter I).
They were then invited to
comment on the possibilities of being able to predict famines in Bangladesh (see Chapter IV ).
After this the respondents were given a
crude, but operational definition of famine, i.e., "a situation in which government gruel kitchens have to operate," and were asked if there were certain factors which made some areas of Bangladesh persistently more liable to localized famine.
(The murrubbis were asked if there were
certain factors that made their particular village more liable to famine than other villages.)
Apart from three respondents who said they did
not know, the remainder of the 180 respondents said "Yes."
Those who
had answered "Yes" were then encouraged to list what the factors were. The respondents provided a list of six responses on the average.
These
were then included in an overall list of 60 factors (Table 6).
^The letter authorizing the study from the Ministry of Relief and Rehabilitation, Government of Bangladesh, helped initially in carrying out the interviews.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174 Mental Maps The respondents from three of the groups, i.e., senior retired government officers, aid distributors and technical experts, were then asked to map those areas that they thought were persistently more liable to localized famine, where government gruel kitchens had to operate.^ This they did by ringing the areas with a felt pen on a tracing paper overlay registered on a large 40" x 50" standard reference map of Bangladesh (1:500,000) which was positioned on a large sloping easel. The reactions to this request were then recorded for each member of each of the three groups. complete the task.
Three of the 135 persons said they were unable to Five others were observed to be having difficulty
in reading the reference map. task within ten minutes.
The remaining 127 people completed the
Four distinctive characteristics in the
application of this method were noted. 1.
Respondents ringed broad and unspecific areas rather than attempting to delimit particular boundaries.
This may reflect
both a lack of specificity in the real areas liable to famine, as well as the respondents' understandable generalization because of inadequate knowledge.
Also it is probably the
normal reaction to such an exercise. 2.
Respondents were found to avoid outlining up to and over the international boundary, and over all the deltaic islands.
^This free mapping response format for mapping environmental hazards has recently been demonstrated and documented in the developed world by Smith(1976), but was initiated by the researcher and his colleagues in mapping environmental hazards in Hilo, Hawaii in 1973 (Currey et al., 1973).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175 Thus at a later point high values in the center of the islands, e.g., Bhola, or areas within 5 kms. of international boundaries, e.g., Rowmari, were â&#x20AC;&#x2122;extended' to the coastline or to the boundary line. 3.
Some respondents ringed "places," e.g., Faridpur, while others ringed "areas" over which certain phenomena, e.g., flooding, occurred.
4.
Some respondents appeared to associate famine with the name of a place, e.g., Faridpur, without knowing its areal dimensions.
They might thus ring an area after locating
Faridpur Town rather than locating the
whole district.
After completion, the map overlays were collated on a base map of 5 kilometer by 5 kilometer grid cells.
If 50 percent or more of a grid
cell fell over an area outlined by a respondent, that grid cell was allocated a weight of 1 point.
If it fell over an area outlined by two
respondents it was allocated two points and so on.
The points for each
of the 3125 cells on each of these maps were then totalled and the upper and second quartiles were then chosen as areas of "very strong" and "strong" agreement for each of the three groups. These composite "mental maps" (Figures 20, 21, and 22) of areas liable to famine exhibit a pattern which is consistent for all three of the groups.
The retired government officers, aid distributors and
technical experts all agree that the Jamuna channel, Faridpur and the coastal and island areas of both Barisal and Noakhali on either side of the Meghna Delta, are areas liable to famine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
176
' Strong Agreement (upper quartile):
ÂŁ3'
Figure 20,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Very Strong Agreement (upper quartile):
B
Figure 21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
Figure 22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
There are slight inconsistencies between groups. 1.
The retired government officers are not in "very strong agreement" on the famine liable areas of eastern Rangpur. This may be because their longer time perspective supports a greater emphasis on Noakhali, Bhola and Faridpur.
It may
be, however, that having mostly retired in the early 1970s they were less involved in the 1974 famine in eastern Rangpur than were the other groups.
Unlike the other groups, the
retired government officers were
however, in "strong agreement"
that the boro-growing haor area of Sylhet was an area liable to famine.
This may be because recent improvements in the area,
by the introduction of
high yield variety rices and drainage
had not been implemented during their period of service.
Their
"strong agreement" on the Sylhet area may also be because they associated flooding with famine, and assumed a direct relation足 ship in all areas including the extensive flooded basin areas of Sylhet.
This latter suggestion may also partly explain
their "strong agreement" on extending the famine prone area of Faridpur south-westwards to include the river and the bil (swamp) areas of Gopalganj. 2.
The Aid Distributors unlike the technical experts and the retired government officers, were only in "very strong agree足 ment" about one area: 75 famine.
Eastern Rangpur, the area of the
Other areas were given lesser emphasis
1974-
This
probably reflected their short average period of involvement
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180 with famine problems in Bangladesh.
It also probably
resulted from their recent involvement with the eastern Rangpur famines. 3.
The Technical Experts tended to detail more areas than the other two groups.
They were in "very strong" agreement on
three main areas:
the Jamuna Channel, Faridpur and the coastal
areas near the mouth of the Meghna.
It is noteworthy that their
areas of "very strong agreement," unlike those of the other groups did extend downstream to the southern part of the Jamuna Channel and included a much larger area of the cyclone-prone coastal area. island.
Their area of strong agreement included Sandwip
Surprisingly, they did not register "very strong
agreement" in the case of eastern Faridpur, e.g., Bhedarganj. Their areas of "strong agreement" did include the old flood water channels and drought prone interfluves of North Bengal, two small outliers in south-western Sylhet, and a small outlier on the Chittagong coastal plain. After completing his or her mental map, the respondents from those same three groups were then asked to explain why each of the areas they had ringed
were liable to famine.
These responses were then integrated into
Table 6 as an attempt to augment or reinforce the initial question of why certain areas of Bangladesh were persistently liable to famine. The respondents from the three groups were then asked first, what the
t villages
themselves could do, and then what the central government could
do to prevent famine returning to those areas which they had ringed. The village elders were similarly asked what they themselves could do,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
£ •5
S
1 S
i
■g
!
I %
i s
1| *
1
i
!i
i i
i|
1 I
""-^-■-•"-sasssssssssaaRas
Jssssss
ssaassssss
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183 and then what the central government could do to prevent famine return ing to thei^ village.
The responses to these questions on famine preÂ
vention are described later in Chapter VI.
The open-ended questionnaires
were then completed by a series of questions about the respondentâ&#x20AC;&#x2122;s training and experience in Bangladesh, particularly in respect to work in famine areas or on famine-related problems.
It was hoped to use these
"profiles of experience" to select the "most experienced" respondents to attend the later stages in the Delbecq-Delphi Process.
Stage 2:
Editing by Technical Experts
The list of 60 factors thought to make certain areas of Bangladesh liable to famine (Table 6) was then divided up into smaller lists, according to the most pertinent field of expertise and then given to the most appropriate of the selected technical experts^ from each of the nine fields for editing. For each factor in their own short list, the experts were asked to evaluate if there were nationwide quantitative data available which were sufficiently comprehensive, reliable, and for "sufficiently small areas" to aid in mapping thanas liable to famine in Bangladesh.
If not,
they were asked to suggest a surrogate for which information might be available.
As an example, one of the initial respondents had suggested
that areas where there were "excessive social activities" were areas in Bangladesh that were more liable uo famine.
The technical expert chosen
^The experts selected for editing were the same as those selected for the Delbecq and Delphi sessions. The process by which they were selected is therefore described in Stage 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184 to represent the "social science" field suggested that although no data were available to show areal variations in "excessive social activities," the amount of money spent on buying beef cattle for sacrificing at the feast of Eid-ul-Azha might serve as an indicator of "excessive social activities."
Thus he proposed that it might be possible to map varia足
tions in the price of beef on the eve of Eid-ul-Azha at the principal markets in Bangladesh.
Often the articulation of a concept had to be
crude if the reliability of the data was to be maintained.
Thus with no
reliable acreages for cropped areas it was necessary to define a concept like "population pressure" in terms of gross population density, rather than refining it to include arable cropping intensity as the denominator. If information either for the original factor or for any surrogates suggested were unavailable, that factor was removed from the list (see Table 6).
It might be argued that these omissions were equivalent to
"throwing out the baby with (or even instead of) the bathwater," and that it was the more important correlates of areas liable to famine that were mistakenly omitted for want of available information,
However, the
factors left on the list (Table 6) are more likely to be "the baby" than "the bathwater," because the list of the factors remaining after editing is significantly more similar to the list provided by the respondents after they had ringed the areas they thought were liable to famine (Table 6).
Thus once they had focused their attention upon the famine
liability of particular areas, respondents omitted many "bathwater"-like factors such as "graft and corruption" and "poor administrative struc足 ture," which although possibly contributing significantly to famine liability, need not necessarily have aided in the differentiation of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185 particular areas liable to famine.
Nevertheless, certain pertinent
factors, e.g., "high percentage of landless cultivators," were unfortunately omitted for want of sufficiently reliable data.^
More
over, the limitations of available data limit all forms of research and the problem is unusually acute for nationwide areal data in a developing country. Each of the nine experts was then asked to calibrate each of the short list of factors by deciding on certain pivotal or "go-no-go" values. The area delimited by mapping those values of a factor which were above the pivotal value was judged to be equivalent to the area in which the factor was correlated with the areas liable to
famine.^
For some factors
the pivotal value was stated as some amount of variation above or below the national mean for the value being measured, e.g., population pressure — areas with a population density one standard deviation greater than the mean; for other factors, the areas so designated by a reputable survey group, e.g., Soil Survey of Bangladesh, were taken as being equivalent to the area above the pivotal value, e.g., areas liable to river erosion.
Stage 3— Delbecq Sessions The list of the 30 edited factors, each with its critical value, was then circulated to the members of three separate groups of nine selected
^Since this time, Dr. T. Januzzi has carried out a national survey of the landless population of Bangladesh for the U.S. Agency for Inter national Development, Dacca. This survey was, however, done on a sample survey and does not include thana by thana information. ^The assumption that technical experts can indeed perceive such thresholds and foresee the national pattern is suspect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
186 retired government officers, aid distributors and technical experts. Unfortunately, it was not possible to include the village elders in the latter stages of the Delbecq-Delphi Process.
There were both logistic
and language problems'*'速 which would have prevented the village elders from being involved in the Delbecq-Delphi sessions.
Also it is unlikely
that the village elders would have had a sufficiently nationwide per足 spective to have helped in the final stages of mapping the areas liable to famine. It was mentioned earlier that I had planned to select for each of the three groups the nine persons from the original groups of 45 who had the greatest experience with famine situations in the rural areas of Bangladesh.
Those initial selections had to be changed, however, due to
the realities of the situation.
Many of the originally-selected panel
members were ranking officials with numerous responsibilities that would have made it impossible for them to attend all the sessions. had to cancel because of deaths in their families.
Others
Changes in the
scheduling of Delbecq-Delphi meeting nights because of events such as the death of the national poet, Kazi Nazrul Islam, or the death of Mao Tse Tung also made it impossible to maintain the initial selections. Thus, although I initially attempted to maintain the validity of "the careful selection of panelists on the basis of professional training and scaled experience levels" (Sackman, 1975), this was not possible.
As
an observer, however, I judged that the eventual group members, although
'*'速Note: Andrew L. Delbecq has, however, used the Delbecq Technique for a comprehensive health planning exercise in Micronesia, using trans足 lators to facilitate communication between group members speaking different languages (Delbecq, Van de Ven and Gustafson, 1975).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187 not all the most experienced in famine problems and rural areas, were at least equivalent to or above the average experience levels of the initial three groups of 45 members whence the three sample groups of nine members were drawn. The three groups of nine members met nominally on three separate occasions.
In terms of creativity, it has been argued that it would have
been more productive to have had the three groups interacting (Delbecq, Van de Ven, and Gustafson, 1975), but there are counter arguments (Van de Ven and Delbecq, 1971).
I decided that such discussion between the
groups with my lack of training in group discussion methods, might have led to "maximum feasible misunderstanding" (Moynihan, 1969).
Thus, just
as I had decided to use the nominal group technique to minimize inter足 action, so I also opted to keep all three groups apart, although they were each evaluating the list of factors derived from all the four initial groups. I chose to have nine members in each of the groups for three reasons: 1.
The same groups were to be used for both the Delbecq and the Delphi sessions.
For Delphi studies, a group of nine maximized
the effect of large group sizes; it traded off average group error in judgments (Dalkey, 1969), against decreasing "manageability," with increasing group size (Bouchard and Hare, 1970) 2.
(see Figure 23).
Nine was also an aliquot part of 45, the number in the original groups.
3.
Being an odd number, the median judgment of the group on any item could rapidly be derived by ranking the values
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188 and determining the value of the middle value, i.e., the 5th.
Average Group
Manageability of the Group
Error
Bouchard and Hare (1970)
Dalkey (1969)
No . in Group
Figure 23.
At each of the three Delbecq sessions^ the nine members initially met informally to discuss the list of thirty factors and the critical values that had been distributed previously.
Where possible, the group moder足
ator, in this case the researcher, answered any questions on the meaning or the terminology used in the list.
On one occasion, x^hen the nine
The Delbecq and Delphi sessions were held on three separate evenings at the Dacca residence of the researcher. The system was pre足 tested with a group of University students from the Department of Geography, Dacca University. Official invitations were sent out to each group member (Appendix ). Transport was provided where necessary. Both the Delbecq and Delphi sessions took place around a circular table at one end of the dining room. The circular table meant that no person was per足 ceived as being the head of the table or the dominant discussant because of their position at the table. The Delbecq session was conducted over an appetizer before the main meal. The Delphi session was completed during the coffee hour.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189 technical experts met, the session was continued without the group being in general agreement on the criterion used for the delimitation of drought. After the initial discussion, the nine members then became a nominal group and were asked to rank in terms of their importance in differenti足 ating the areas liable to famine, the five most important factors from the list of 30 they had previously considered.
Each group member
placed on a 3" x 5" card marked "First Choice," the number of the factor he thought was most important and handed the card to the moderator. Each member then wrote the number of the second most important factor on the card marked "Second Choice" and so on, until they had handed in all five cards in less than 10 minutes.
The rankings for all the group
members were then aggregated by assigning any factor given a "first choice" ranking a score of 5, any given with a "second choice" ranking a score of 4, etc.
The scores for each factor were then added and the
10 highest ranked factors for each of the groups were selected for later analysis in the Delphi session (Appendix. E).
Stage 4:
Delphi Sessions
The ten highest ranked factors for each of the groups were then typed on nine 10" x 8" cards (Example, Appendix E) and distributed to each mem足 ber of the nominal group.
Each card contained a number which identified
the particular group member
.^
Cards did not include a column for "self
12 Each member sat by a prominently displayed identification number on the circular table. These numbers aided in the rapid matching of respondents of Delphi cards during the session.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190 rating" (Dalkey, 1969), in which group members could indicate a sub足 jective assessment of their own competence to evaluate any particular factor.
Although Dalkey found such ratings could be used to further
analyze the weightings, I omitted them after the pilot study-^ because the additional column appeared to confuse respondents.
Also each card
contained at the top the list of factors and a pilot test factor, typed in italics, to demonstrate the rationale of Delphi to the participants. The group was asked to "guestimate" the population of Dacca City as defined and documented in the 1974 census.
Answers varied between .5
million and 7 millon on the first round for the senior retired government officers.
The first round median was 1.5 million.
After being informed
of the first round median, some members of the group reassessed their initial judgments and the second median was found to be nearer the documented value of 1.73 million.
This example was also used with the
other two groups, the aid distributors and the technical experts. recorded similar ranges of deviation in the first round.
Both
Both, however,
recorded medians of 1.6 million for both the first and second round medians.
This
unwillingness to change shown by the latter two groups
as compared with the retired government officers was continued into the later weightings of
the 10 factors.
After this initial example, each member of the nominal group was asked to score each indicator or factor as defined by its critical value, on a scale from 0 to 100, just as if they were marking a school
13 The system was pretested in Dacca with a group of university students from the Department of Geography, Dacca University.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191 examination.
An indicator judged to be exactly equivalent to the areas
liable to famine in Bangladesh was to be assigned a score of 100. indicators judged to be less equivalent lower scores.
Those
were to be assigned progressively
Members of the nominal group then handed their Delphi
cards, face downwards, to the mediator who was able to quickly select the median value for the group, i.e., the fifth ranked value, and thus the middle card, for each indicator.
These 10 first round medians were
then quickly written on each of the nine Delphi cards, and then each member's card was returned to him.
The mediator then invited group
members, if they wished, to change their scores after seeing and hearing the "average" (the median) of the group for each indicator from the first round.
After any changes were made the cards were collected and
the second round medians w7ere then calculated. initiated.
No further rounds were
It had been decided, prior to initiating the session, only
to have two rounds, thus "maximising the possibilities of lessening group error after iteration" (Dalkey, 1969), and yet minimising lack of interest during successive iterations. All Delphi sessions were completed in a convivial atmosphere.
The
retired government officers stated they had enjoyed playing the game. Their Delphi session was very much a reunion night for former colleagues. When any of the retired government officers asked a question on how to complete his Delphi card, one or other of his colleagues would carefully explain the method to him.
Several of the aid distributors appeared to
have difficulty in comprehending the different significance of factors which were associated with variations in famine liability within Bangladesh, as opposed to factors that made Bangladesh as a whole
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192 vulnerable
to famine.
A few of the technical experts appeared to be
uneasy in trying to weight the indicators because they disagreed with, or were unable to assess, the areal implications of the critical values assigned. The broad results of the three Delphi sessions are analyzed graphically in Figure 24, which shows the patterns of the medians and interquartile ranges for rounds one and two on all 10 indicators for each of the three groups.^
There was certainly general agreement in
the ordinal scoring of the 10 indicators by the three graphs.
This is
shown by the general leftward displacement of the medians in all three groups between indicator 1 and 10.
Although there were some changes in
weightings between rounds one and two and deviations did decrease in some instances, as exemplified by the second round black polygon being shorter horizontally than the white polygon, there were two instances in which there was actually more deviation in the second round than in the first round.
The median actually moved in only three instances, each time on
a different indicator.
Visually interpreting the polygons, I surmise
that the groups had difficulty understanding the concepts within the indicators that they were being asked to weight.
This is specially
reflected by the broad polygons on the initial round seen in the scores of the retired senior government officers.
It was very different for a
panel member, such as a retired senior government officer, to score a
â&#x2013; ^The omission of the upper and lower quartiles certainly masks the extremes of variability that occurred in the assessments of some indica tors, but it probably aids the interpretation of the general principles.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
24.
193
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194 specific almanac-like question, such as the population of Dacca City in 1974, compared with a nebulous concept, like "population pressure," even though a statistical definition was included. On the basis of the overall ordinal consistency among the three groups, I assumed that an amalgamation of the groups' interval scores would be the best estimate I could determine to weight each indicator for inclusion in an index which would allow the mapping of areas liable to famine.
Each group's median scores for each of the ten indicators were
then standardized against a possible total of 100, thus assigning equal weight to each of the group's judgments.
The average for each of the
three groups' 10 median values was then calculated, and these in turn were then standardized against a possible total of 100.
Compiling the Map of Areas Liable to Famine It was then possible to construct a composite map of the areas liable to famine (see Appendix H), by "overlaying" (McHarg, 1969) 10 component maps of an area delimited by an assigned critical value and each given a Delphi weighting.
The "overlaying" of the 10 components
was done by simple linear addition.
In describing the process I tried
to highlight some of the difficulties inherent in such an addition. These difficulties have been more thoroughly discussed by Macdougall (1975) in his discussion of map overlays. Areal data were collected for the critical values which had been suggested by the technical experts for each of the 10 indicators chosen by the three groups.
I then reviewed the areal data actually available.
In some cases there were less than the technical experts had suggested. Then I re-examined the critical values proposed by the technical experts.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195 These reviews led to a reassessment of some of the definitions and critical values previously assigned to the indicators. The ten component indicators and their weightings as a percentage (in brackets) for the composite map were finally as follows: A. (16.4%).
FLOODING AREAS: All areas suffering extensive flood damages during years of severe flooding included in 1962-67 period (BWAPDA/IBRD, 1972).
The areal delimitation of this indicator
is based on a 5 year period in which BWAPDA (Bangladesh Water and Power Development Authority) flood damage records were considered by IBRD (International Bank for Reconstruction and Development) to be relatively consistent.
B.(15.4%).
DROUGHT AREAS: of receiving
The area with less than 80 percent probability
2inches
of rainfall in any
2week
period during
the critical Aman seedling period June 13th to September 15th. (IBRD 1972).
This area was interpolated from eight control
points using criteria suggested by the local agricultural expert who stressed the importance of the main rice crop to Bangladesh and the need for surface water availability at the critical rice seedling period, a factor not taken into account in traditional water balance drought definitions.
C.(13.7%).
POPULATION PRESSURE:
All areas with a 1974 population density
which was greater than one standard deviation above the national mean.
(Bangladesh Census, 1975).
Without reliable
small crop production statistics, population pressure was measured simply comparing the ratio of the thana population to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196 the dry season land area.
The latter were the values re足
computed by planimeter as described in Chapter III.
D.(13.4%)
FOOD DEFICIT AREAS:
Thanas classified as non-surplus areas
(Food Directorate, Bangladesh Government, 1977).
The areas
defined by this indicator are those not included in the government list of Food Surplus Thanas, based on crop production and population. IV,
As described in Chapters III and
such figures are contestable because of variations in crop
production estimates and because of internal wage labor migration bringing surplus rice back to rice deficit areas. Although the indicator is overtly collinear with the "popu足 lation pressure" and "low crop yield" areas, all three groups unanimously argued strongly for its inclusion on the basis of "one of the best available indicators."
E.(12.6%).
LACK OF ALTERNATIVE EMPLOYMENT AREAS:
All areas with the
percentage of the civilian labor force employed in agriculture in 1974 being greater than one standard deviation above the mean.
(Bangladesh Census, 1975).
only down to the subdivision level.
These data were available They nevertheless sug足
gested areas with a strong dependency on agriculture and little local secondary or tertiary employment.
Apart from
areas near Dacca, it is unlikely to mask a large percent of the population "in service" in a large city.
F.(11.8%).
LOW CROP YIELD AREAS:
All non-forested one crop areas
(IBRD Land Use Associations Map, 1972).
These data were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197 portrayed by Land Development Units (physiographic units) with similar soil and hydrographic conditions in terms of agricultural development.
Areas where only one crop is
grown may sometimes be high yielding areas, e.g., the boro growing areas of Sylhet, but without adequate storage facilities and without a system ensuring that surpluses are stored, they are vulnerable to famine in years where a natural disaster destroys the one and only crop.
G.(6.9%).
POOR LAND TRANSPORT AREAS: All areas further than 10 kilometers from an all weather road.
At present relief
supplies from Dacca and Chittagong port are predominantly transported overland using the road or rail network.
The
rail net is even more limited than the all weather road net defined in the Bangladesh Transport study (UNDP, 1973) and
10kilometers
was approximately equivalent to a bullock
cart journey through difficult monsoon landscape conditions in one day.
H.(5.6%).
RIVER EROSION AREAS:
Land Development Units where river
erosion occurs (Soil Survey Development, 1975).
Little
detailed quantification of river erosion has been completed, and bank line erosion rates vary greatly from year to year. Land Development Units are physiographic units.
The Land
Development Units constituting the recently averaged flood plains of the major rivers were taken to give a crude indication of the areas liable to river erosion.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198 I.(1.8%).
CYCLONES:
Approximate inland limit of frequent cyclone
risk (Brammer, 1975).
Although there are historical records
of cyclones and cyclone tracks over several centuries, a quantified areal study of the intensity and frequency of cyclones is not yet available.
Brammer's area of frequent
cyclone risk is based on his own experience of cyclones during the past
20years
and his own piecing together of the
historical records.
J.(1.8%).
MALDISTRIBUTION OF AGRICULTURAL INPUTS: All areas of North Bengal, i.e., west and north of the Jamuna and Ganges Rivers. Although there are many undocumented social and political as well as distance factors which affect the distribution pattern of agricultural inputs, the agricultural technical experts agreed that the broad area of North Bengal was con足 sistently and ubiquitously poorly supplied with agricultural inputs in comparison with the rest of the country.
The areas defined by each of the critical values on these ten com足 ponents were then explicitly inserted as inset maps for comparison with the following map of thecombined components
Combining
(see Appendix H).
the Components of Famine Liability
The areas defined by each of the critical values for the ten com足 ponents each had an "importance weighting" or a multiplier factor derived by combining and standardizing the results of the three Delphi sessions. These "importance weightings" were on an interval scale, and thus it was mathematically possible to combine them by addition into an areal index.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
199
The rationale for such a linear combination was marred by the observable interdependence among the components, particularly between areas of population pressure and areas of food deficit, between the latter areas and
low crop yield areas and between both flood and cyclone areas and
low crop yield areas.
Durfee (1972) has suggested that factor analysis
might be used to derive a set of independent factors in such a situation, but it seemed irrational to apply such a sophisticated technique to such a limited data base and to obfuscate weightings derived by the explicit Delphi sessions.
This decision was partially supported by Hopkinsâ&#x20AC;&#x2122; (1977)
recent findings in evaluating methods for generating composite maps. According to Hopkins, "limited experimental experience suggests that, because of the data requirements and difficulty of interpretation, using factor analysis
to identify independent factors is not worthwhile."
Acknowledging the inherent weakness caused by the interdependence (see also Macdougall, 1975), the 10 area components of famine liability were combined on a 5 km. square grid system covering the entire country. If a component was found to occur in more than 50 percent of the area of any grid cell, the "importance weighting" was then entered in that cell. After such entries had been made for each of the ten components, the
1100) of
entries were then added to get an index value ( ability for each cell.
famine vulnerÂ
A frequency graph was constructed showing these
index values against the number of cells with such values. graph, Appendix H.) for each
10 units
(See inset
A running mean describing the average number of cells
of the famine liability index was then computed and
drawn to minimise the artificial breaks in the distribution caused by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200 the limited number of possible combinations of the components' "impor足 tance weightings." The "areas very liable to famine" were then classified as all those thanas having a majority of grid cells with a famine liability index value greater than 65, i.e., greater than the highest significant break in the pattern of the 10 unit running mean. inset graph in Map in Appendix K).
(See solid black line on
Similarly thanas classed as being
"liable to famine" were those with index values between 65 and 45, i.e., between the highest significant break and the next highest significant break in the running mean pattern.
It was then possible to map the
thanas that were "very liable to famine" (in red on Map, Appendix H), and the thanas that were "liable to famine" (in yellow on Map, Appendix H).
The class intervals were thus selected at the most natural
breaks in the distribution.
This does not mean that there is a really
significant difference in policy terms between a thana with an index of 65 (e.g., Mohammapur in Jessore District) at the lower end of the "red" (areas very liable to famine) and a thana with an index of 64
(e.g.,
Gopalgarj in Faridpur District) at the high end of the "yellow" (areas liable to famine).
Similarly a thana with an index of only 45 at the
lower end of the yellow classification is not significantly different from one with 44 which would not be classed as liable to famine at all. Each thana was also assigned a pattern coding and a number to describe its actual famine index value and which of the ten components had contributed to the index value in that thana.
(See main legend,
Map, Appendix H .)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VI AREAS LIABLE TO FAMINE IN BANGLADESH
The most obvious empirical test of the predictive accuracy of the map produced by the Delbecq-Delphi method would be to compare it with the areal patterns of a significant number of future famines in Bangla足 desh.
One expected effect of the famine vulnerability map presented
here, however, is the initiation of preventive plans which would in turn hopefully lower the famine liability of certain areas. a test could not give reliable results.
Thus, such
Moreover, the lack of accurate
information on the areal extent of past famines (Chapter IV) also pre足 cludes any possibility of testing the map against past geographical patterns.
Thus a more subtle and inferential evaluation must be used.
Specifically, it seems possible (see Figure 25) to evaluate the map produced by the Delbecq-Delphi process by comparing the resulting map's areal consistency with:
1.
the combined mental maps of the three groups who constructed
2.
newspaper reports of the extent of the 1974-75 famine;
3.
a map showing the place of origin-*- of the bustee dwellers
4.
a famine survey which recorded significant differences in
the Delbecq-Delphi map;
(squatters) in Dacca City 1974 (Oadir, 1975); and
the famine symptoms within a ten thana area of north Bangladesh during the 1974-75 famine.
^The importance of rural out-migration to the cities during famine was demonstrated in Chapter I of the dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
EVALUATION OF THE DELBECQ-DELPHI MAP
Figure 25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203 We shall see from the first of these comparisons that the broad regional patterns portrayed by the mental maps are also repeated in the Delbecq-Delphi map. it
The method of constructing the latter probably gives
an inherently greater specifity than the former.
The few areas of
overt disagreement between the mental maps and the Delbecq-Delphi maps are probably caused by failures in both the methods of map compilation. From the second of the comparisons we shall see that the areal pattern of famine symptoms reported in a national newspaper is very con足 sistent with the pattern of famine liability portrayed on the DelbecqDelphi map.
The few discrepancies are not easily explainable, but they
may be due to the
problem of area scale, i.e., comparing symptoms at
one scale with causes on another. The third of the comparisons will show that the pattern of places of origin of the bustee dwellers (squatters) in Dacca city correlates strongly with the pattern of famine vulnerability within the probable catchment area for migration to the capital.
However, the unavailability
of data for other cities with different catchment regions in Bangladesh prevents a national scale comparison with the pattern of the DelbecqDelphi map. However, from the fourth of these comparisons we shall see that there
is almost no agreement at a local scale when the composite
intensity of famine symptoms collected in a cross-sectional famine survey for ten thanas
is compared with the Delbecq-Delphi index of famine
liability for each of those same ten thanas.
This will confirm the
inability of the Delbecq-Delphi method, organized on the national scale to predict famine symptoms for any given year at the local level.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
204 Comparison of the Delbecq-Delphi Map with the three Mental Maps
Consistency Accepting the differences in emphasis between the groups detailed in Chapter
V, there is a consistent pattern of famine vulnerability in
all three mental maps along the Jamuna, Padma and Lower Meghna channels and into the deltaic chars on the Bay of Bengal. clearly repeated in the Delbecq-Delphi Map.
This broad pattern is
Similarly, the "very strong
agreement" on the Rangpur area of North Bengal, the riverine area of Faridpur and the coastal areas of Noakhali, is generally consistent with areas determined to be very liable to famine on the Delbecq-Delphi Map.
Specificity The Delbecq-Delphi Map is constructed from specifically defined criteria.
It is compiled from components with carefully delimited areas.
Intrinsically it is more likely to be a more specific areal pattern than will the aggregation of areas subjectively ringed by the large groups of mental map respondents.
The tendency of the respondents to ring
"places" associated with famine vulnerability, rather than to delimit particular areas of vulnerability, was mentioned in Chapter
V.
This
may explain the concentration around Faridpur town in Faridpur District on the mental maps, compared with the specific concentration on Madaripur subdivision around Naria, Bhedarganj, Janjira and Gosairhat, and Char Badrasen in the Delphi Map.
Similarly, this greater specificity probably
explains the highlighting on the Delbecq-Delphi Map of Shudharam thana in Noakhali, Barisal Kotwali, Swaruphati, Jhalahati, and Rajapur in Barisal and Bhurungamari, Lalmanirhat, Kurigram, Ulipur, Chilmari, Rowmari and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
205 Fulchari in Eastern Rangpur.
The lower Jamuna area agreed upon by the
technical experts is probably equivalent to the thanas of Kazipur and Sirajgunj, Belkuchi and Chankali on the west bank and Sarishbati, Gopalpur and Tangail on the east bank.
Greater specificity may also explain the
Delbecq-Delphi map's delimitation of the haor area of northern Sylhet as opposed to the more central area ringed by the retired government officers.
To suggest this would, however, contradict the argument pro足
posed in Chapter V
that this area was an anachronistic famine area.
It may indeed be that this area is given a high liability index value because of the definition which equates one crop areas with low crop-yield areas.
Certainly the boro areas in these haors are liable to early
flash flooding, but larger storage and strong national and international savings links may counteract this area's susceptibility.
As a corollary
of the more specific delimitation of certain thanas by the Delbecq-Delphi map, thanas are also omitted from the broad areas of the mental maps. Sandwip Island, within easy boat access from Chittagong, is calculated not to be famine prone.
The southernmost areas of Noakhali and Barisal
on the Bay of Bengal are also emphasized less on the Delbecq-Delphi map. This
may be because of their lower
population densities and the
potential protection offered by coastal embankment project.
The central
area of Faridpur District (like Maksudpur Thana) is likewise emphasized on the Delbecq-Delphi map.
Disagreement There are also some overt disagreements between the areas shown as being liable to famine on the mental maps and those so delimited on the Delbecq-Delphi map.
These disagreements are most probably due to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206 inherent drawbacks of the method of mental mapping.
These drawbacks
may be reflected in the mental mapping of areas liable to famine by the representatives of false historical images, false peripheral images and false associative images. Thus Chittagong coastal plain has historically been considered resilient to famine (Chapter III and Figure 16).
The mental maps reflect
this historical image and none of the three mental maps show the Chittagong coastal plain as being liable to famine.
This historical image is per足
haps reinforced by the association of the coastal plain of Chittagong with the stock piles of grain imports to Chittagong port itself.
How足
ever, as recognized in the 1845-49 famine in Ireland, large quantities of grain may pass by the very door of the poor and they may still experience famines.
Famine liability of the Chittagong plain, because of the lack
of alternative employment, the vulnerability of the single crop and the possibilities of drought, flood and cyclones, highlighted by the DelbecqDelphi map is substantiated both by the newspaper reports of famine con足 ditions in the area in 1974-75 (see Figure 25) and by my own personal observation of famine symptoms in the area in 1974-75. False peripheral images may occur when an area has a marginal location far from the respondent's area of everyday life.
Thus remote
border areas far from Dacca City may be forgotten by Dacca based respondents completing maps in
Dacca city, although they
know of fami
susceptibility in such areas. Thus thanas like Tetulia, Thakurgaon and Ranisankail in Dinajpur District; Gomastapur, Shibganj and Nawabganj thanas in Rajshahi District and some of the border thanas, e.g., Tala,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
207 Debhata, Assasuni, Kaliganj and Shymnagar in Khulna District are in足 cluded on the Delbecq-Delphi map, but are not mentioned in any of the mental maps.
The suggestion that these disagreements are caused by false
peripheral images is unsubstantiated, and needs further study. Finally, false associative images may occur when an area is associated with a particular attribute, but localized pockets within the area are associated with different attributes.
Images of these areas may be
generalized in a form of ecological fallacy.
Thus Comilla district as a
whole is seen generally as bountiful, while its famine prone pockets, e.g., Matlab, Hajigunj, Chandpur, Faridpur and Ramganj, and the thanas along the Gumti River are omitted from the mental maps at the national scale, but are identified in the Delbecq-Delphi map.
In this case the
pattern portrayed on the Delbecq-Delphi map is corroborated both by the pattern of newspaper reports of famine and by the origin of bastee dwellers in Dacca in 1974 (see Figure 25).
The newspaper reports and the
Purundapur village data (Figure 1 ) both indicate that famine occurred in Matlab thana in 1974 and the origin of Dacca bastee dwellers appear to be concentrated in the thanas along the Gumti River. In the cases of overt disagreement between the patterns on the Delbecq-Delphi map and those on the mental maps, the Delbecq-Delphi patterns appear to be more rational than those of the mental maps.
Comparison of the Delbecq-Delphi Map with the Newspaper Reports of the 1974-1975 Famine The contents of the Bangladesh Observer, one of two national newspapers in English, were analyzed in each daily edition between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
208 August 1, 1974 and July 31, 1975.
This made it possible to map locations
reported as having "famine conditions," "severe food shortage," or "government gruel kitchens operating."
A more complete record would
probably have been available in Bengali-language newspapers Ittefaq and Dainik Bangla, which have extensive rural networks of correspondents and serve a rural readership.
Unfortunately neither the two Bengali-
language newspapers nor the alternate English-language paper The Bangla足 desh Times were available for analysis of the 1974-75 period in the Federal depository in Hawaii. When mapping the famine areas from the reports in the Bangladesh Observer (Figure 25), reports of natural disasters that did not mention resulting food problems were not included. ditions were not always located by thana. were omitted.
The reports of famine con足 Those located only by district
If only a subdivision was mentioned, all thanas within
the subdivision were included.
If a village or Union within a thana
were mentioned, all thanas within the subdivision were included.
If a
village or Union within a thana were mentioned, then the whole thana was included.
An area only needed one citation to be mapped.
The areal pattern of reported famine symptoms is consistent with the pattern of famine vulnerability portrayed on the Delbecq-Delphi map and the mental maps.
Thus, the general pattern seen along the Jamuna
channel from Kurigram subdivision in the north through Sirajgunj and Gopalpur thanas south into Manikgunj subdivision in the west of Dacca district, and the patterns of famine cited in the papers for Madaripur thana in Faridpur, Munshiganj in south-east Dacca, Chandpur thana in Comilla and Bhola and Mehendiganj thanas in northern Barisal district
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
209 all conform to the pattern predicted on the Delbecq-Delphi map and on the mental maps. The newspaper reports also provide support for the Delbecq-Delphi map where it is in conflict with the mental maps.
The reports mention
famine conditions during 1974-75 in Satkira subdivision and in both Bagerhat and Kachua thana in Khulna, as well as in many of the coastal thanas of Chittagong district and in Nawabganj subdivision of Rajshahi district.
The newspaper reports also confirm Kotwali and Kishorganj
thanas inMymensingh District to have experienced famine during 1974-75. There are, in fact, only three discrepancies between the newspaper reports of famine and the pattern of potential famine vulnerability shown on the Delbecq-Delphi map.
The newspaper reported famine conditions on
Sandwip Island in Chittagong District, in Chuadanga Subdivision of Kushtia district and in Joypurhat thana in Bogra district while none of these three areas is specified as being liable to famine on the DelbecqDelphi map.
Clearly, these three areas should be given further study.
The discrepancy in Chuadanga Subdivision is probably because two whole thanas were classified as suffering from famine on the basis of newspaper reports which only mentioned two Unions within those two thanas.
Comparison of the Delbecq-Delphi Map with the Map of Origin of Bastee Dwellers of Dacca City Between 15th March and 15th May 1974, Sayeda Rowshan Qadir and the staff of the local government Institute conducted a sample survey of 200 bastuhara or homeless households living in three of the bastee or squatter areas of Dacca City (Qadir, 1974).
The selected bastees,
Babupura, Nayapaltan and Lalmatia, were chosen to represent "large,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
210 medium and small sized bastees in diferent localities of the city." Qadir describes the findings from the sample survey of the origin of the bastee people in Dacca City thus: Although people came from as many as 10 thanas of Faridpur, 11 thanas of Barisal, 12 thanas of Dacca, 10 thanas of Comilla and 5 thanas of Noakhali, major influx occurred from only a few thanas. These were Munshiganj, Tungibari and Lauhaganj in Dacca, accounting for 61%; Naria, Janjira, Bhedarganj, Palong, Shibchar and Goshairhat in Faridpur accounting for 87%; Hijla and Borhanuddin in Barisal ac足 %; Daudkandi and Muradnagar in Comilla counting for accounting for 73% of the respective district's total migration to the three bastees. Almost all these thanas are found to be riverine which are subject to breaches and flood. (Qadir, 1974; see Figure 25)
68
Different bastees in Dacca City drew from different catchment areas, and thest. results only show the origin of bastuharas from within the Dacca catchment area.
Data on the origin of bastee dwellers in Chittagong and
Khulna, although published, were not available.
Neither does the dis足
tribution plotted by Qadir's group directly show the migration generated by the 1974-75 famine, because it was taken prior to the famine crisis period (see Chapter I, Figure 1 ).
Assuming however that famine-induced
migration is merely a sudden acceleration of normal out-migration patterns from vulnerable areas, Qadir's map of the origins of bastuharas within the Dacca catchment area certainly corroborates the pattern of the Delbecq-Delphi map, particularly for the concentration of dots on the river bank of Madaripur subdivision, Faridpur, and Hizla thanas in northern Barisal.
The aggregate pattern in Kasba, Muradnagar, Homna,
Daudkandi and Debiduar, along the Gumti river in northern Comilla con足 firms the Delbecq-Delphi map pattern.
(The mental maps did not demar足
cate any areas in northern Comilla.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
211 Comparison of the Delbecq-Delphi Map with the Famine Survey of Ten Thanas, 1975
Gathering Data from a Famine Survey Unfortunately, it was not possible to conduct a control study to differentiate the severity of famine symptoms in areas very liable to famine and areas not
classified as being liable to famine.
It was,
however, possible to compare differences among thanas in an area which was classified on the Delbecq-Delphi map as being liable to famine, which was recorded as exhibiting severe symptoms of famine in the national newspapers,^
and which was declared to be a famine area by the Government
of Bangladesh. The Government of Bangladesh, Ministry of Health gave the researcher permission to carry out a survey of famine symptoms in a 1300 square mile area of northern Bangladesh, in May and June 1975, after the famine symptoms had begun to decline.
As we shall see later after describing
the results of this survey of famine symptoms in different thanas, it was not
possible to correlate the intensity of famine symptoms with the
Delbecq-Delphi Index of Famine Liability in the same local areas. famine survey covered
10 thanas:
The
nine thanas of eastern Rangpur
(Nageswari, Kurigram, Ulipur, Chilmari, Rowmari, Sunderganj, Gaibandha, Fulchari and Shaghatta) and Dewanganj thana in north-west Mymensingh. These 10 thanas lay athwart the Brahmaputra River for the first 60 miles of its flow through Bangladesh, after entering from Assam in India.
^Ittefaq (1974) Famine in Rangpur, The Daily Ittefaq, Thursday, October 24, page 1, cols. 2, 3 and 4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
212 Although Akbar (1974) has described the famine in that area as being primarily induced by such "man-made" causes, the political unrest in the area following liberation and the magnitude of grain and jute smuggling that occurred across the international border in 1974, the evidence overleaf suggests rather that the predominant causes of the severity of symptoms in this particular area were untimely flooding and river erosion. The 1974 flooding of the Brahmaptra was more damaging than the famous flood of 1955.
Although in 1974 the flood peak was 0.5 feet
lower than in 1955, Figure 26 illustrates the greater destructive power of the flood in 1974 in the Eastern Rangpur area.
It was above danger
level^ for 44 days compared with only 26 days in 1955:
there were
also six steeply rising flood peaks in 1974, not one slow rise as in 1955.
Meshing the hydrographs with the local agricultural calendar
(Figure 26, top), it can be seen that the first peak in 1974 rose early and very rapidly in late June to interfere with the aus (summer rice) harvest and the last flood peak came as late as September when the aman (winter rice) is normally past both the seedling and transplant stage, and into the growth part of the cycle.
Similarly the hydrograph from the
Tista, the major tributary from the Himalayan foreland which joins the Brahmaputra in the area at Chilmari, indicates that the flood waters never crossed the danger level in 1955, but in 1974 they rose above the danger level for 29 days. 3 The concept of "danger level" (bankful level, plus six inches) is ambiguous in Bangladesh where left and right banks vary greatly in ele足 vation. At any recording station, however, the given danger level is taken to be relatively stable over time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
NORMAL AGRICULTURAL CALENDAR GROWTH HARVEST
1
AUS - » : TRAN SPLANT
AMAN THE BRAHMAPUTRA RiVER AT BAHADURABAD GHAT DANGER LEVEL
65 63.5
§ Lj 1 9 7 4 (4 4 ) 1973(21)
55
1955(26) (4 4 ) INDICATES TOTAL NUMBER OF DAYS ABOVE DANGER LEVEL
50 JUNE
JULY
AUGUST
SEPTEMBER
MONTHS Figure 26.
Normal Agricultural Calendar (top) and The Brahmaputra River at Bahadurabad Ghat
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
214 With these extensive high flood levels, bank line migration averaging
1000 ft.
westwards on the eroding right bank in one flood
season, may be inferred from recent work on channel processes of the Brahmaputra (Coleman, 1969).
This massive river erosion is seen in the
photomap (Figure 27) and was substantiated by reports from local officials.
Such an abnormal erosion rate along the 60 miles of the
Brahmaputra in the sample area alone, where riverine population den sities reach
2000per
square mile, would have suddenly imposed over
24,000 people without homes or lands on the already run-down rural infra structure between June and October 1974. Concurrent with this extensive erosion of the right bank occurred the deposition of
sterile sand on the left bank, abnormally high rain
fall within the area (Figure 28) and ponding of the late flood water by water hyacinth (Pontederia crassipes). Apart from this circumstantial evidence, my contention that environ mental factors rather than smuggling predisposed this particular area to famine in 1974-75 is reinforced by an article by Amrita Rangaswami (1975), which documents congruent famine symptoms in the same riverine environ ment, but further upstream, around Gauhati on the Indian side of the international border.
Certainly many of the factors associated with the
famine in Eastern Rangpur in 1974-75 were man-made (Table 7), but the reasons that Eastern Rangpur area was worst affected were predominantly environmental.
It should also be noted that in this area where famine
symptoms were recorded, there was always rice in the subdivisional markets (Figure 29), and in the thana later found to be most severely affected, Rowmari, it was reported that the rice price did not even rise (Abed, 1976).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
215
riaQ?SW3 RIVER
CHANGES
I N THE
SURVEY AREA U SI N G OF LANDSAT
TEN
TONAL
THANA ISOLATIONS
IMAGERY OF THE
BRAHMAPUTRA R I V E R AND 1 9 7 6 .
B lack
=
1974
W hite
=
19 76 R i v e r
R ive r
Scale
=
1 :50 0,000
DRY SEASO
COURSES
IN
197^
Course Course
Ulipur One
In c h
= 8 mi 1e s
S u nderg anj
I
Gsibsndhd
No Sequentiaf3*^
’
~T~ —
I magery A v a i l a b
Fuichari
D ew an g an j
S h a g h a tt a
F ig u re
27.
Photo map
of
R ive r
E ro s io n .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Comparison of 1974 Monthly Rainfall with Average Monthly Rainfall]
1974 Monthly Rainfall
□
Mean Monthly Rainfall (1934-1969)
V777?\
Rainfall (inches)
T Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Note:- Rainfall in July 1974 : 30". This has a probability of 0.05, i.e. once i
Dec 20 years. Figure
Figure 28.
8
Comparison of 1974 Monthly Rainfall with Average Monthly Rainfall
216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
STATION RANGPUR: (25° 45' N., 89° IS' E.) Elevation 112ft.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
:e>(el
29.
Market
K u rig ra m
Price
of Rice
by Month
(as reported
by
the
subdivisional
marketing
department).
218
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The area of Eastern Rangpur affected by the famine had a total population of 311,257 households (1,836,415 individuals) in 10 thanas. The famine survey was designed to include 800 households (4,720 individ足 uals) .
Although the total sample of 800 households was large, the sample
size for each thana was small, because funds were limited and also because it was necessary to complete the survey of all ten thanas before the break of the monsoon in mid June.
The 800 households were selected
equally from each of the ten thanas providing 80 households per thana. The 80 households were then selected from four villages, using a "probability proportional to size" sampling framework to ensure that each household had an equal chance of being chosen, whether it was in a large or small village.
All the households and the villages in the ten
thanas, or more specifically the mouzas (revenue villages), were stratified in order to include a representative proportion of households in villages that were recorded by the Subdivisional Officer (Relief), as being severely affected by the 1974 flooding.
Within each mouza, house足
hold numbers were systematically selected from the list and map of house足 hold malaria numbers, while maintaining a modulus of N/n (N was the total number of households in the village and n was the number of households to be sampled), in order to ensure a total coverage of all households in the village.
If any household had moved, che whither, why and when of
the move was noted, and the next malaria household number was chosen in lieu of the missing household.
Although there were no household refusals,
loss, wasted forms and illegibility meant that only 790 of the house足 hold questionnaires were available for analysis.
In no case did the
number of households analyzed fall below 78 in a thana.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
220 Each of the four survey teams consisted of two members, one of whom was from the Rangpur area.
It was arranged for the teams to be intro足
duced to the ovi bhabok (household head), by the village malaria officer. They were introduced as a "survey team from UNICEF."
The mention of
UNICEF undoubtedly helped the rapport between the members of the survey team and those surveyed, because UNICEF's tubewell program for improving drinking water was fully acknowledged by the population.
However,
knowledge that the team was associated with an aid giving organization may have biased respondents' replies so that symptoms of distress were exaggerated.
In 65 percent of the cases the household head was available
and answered the questions in a relaxed atmosphere, sitting around on piris (low stools) or chairs outside the bari (house) compound, offering the interviewers koler pani and am (cool water from tubewells and mangoes), and themselves often drawing on the hookah in between responses. If an answer was slow in coming, onlookers would volunteer.
For 35
percent of the households where the household head was not available, another family member or neighbor would automatically present himself or herself.
Fifty-seven percent of the respondents were women, and the
purdah system normally required that they be interviewed with the help of a male intermediary repeating the questions through the side of the hut to an invisible respondent. The interviewers administered a questionnaire (see Appendix G and Figure 30) and recorded anthropometic measurements on all available children in the household aged between 2 and 10 years.^
^Age being difficult to ascertain beyond the age of four, 130 cms and 25 kg were taken as upper limits equivalent to age 10. Age was determined by first standing all the children in line in ascending height. A local historical calendar was then used to calculate ages, proceeding from the shortest to the tallest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22:
FIELD METHODS
|THE T E A M
WEIGHING
HEIGHTING
INTERVIEWING
Figure 30.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
222 This survey of famine symptoms was implemented only after careful standardization of the field methods.
The questionnaires were initially
pretested in a village north of Dacca, and then completely revised after a pilot survey in Sankardah village, a char island in the center of the Tista R i v e r . T h e anthropometric measurement scales were standardized against other ongoing nutrition surveys in Bangladesh, e.g., The Bangladesh Nutrition Survey.
For the weighting and heighting, the
weighing scales were standardized against a beam scale in the Cholera Research Laboratory, Dacca, and then standardized daily against each other.
The four persons involved in measuring the weights and heights
were initially trained in Dacca and then later their measurements were checked against each other on 25 children in the field prior to beginning the actual survey.
Throughout this standardization procedure there was
unanimous agreement to within 0.1 kilogram for the weights and to within 0.1 cm for the heights of all 25 children.
There was only one exception:
a child was found by one measure to weigh 0.2 kgs. less than the weight determined by the other three measures.
This was later found to be
caused by an unrecorded urination in the interim.
All children were
measured barefoot wearing only a sari or lungi, and any "knots" of food or other material in the garment were carefully removed. Although the evaluation is primarily concerned with areal variations, a description of the "average household" in the survey may help lead
^By definition survey conditions are differeint in a famine area (see Chapter IV).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in to understanding some of these differences.
The households, identi足
fied by all those on the malaria card at a particular malaria number, were normally an agnatically related family of 5.9 members, who shared the same cooking fire.
Sometimes there might only be one household
member, but households with as many as 18 members were recorded. the total, 719 (91 percent) households were Muslim. households (9 percent) were Hindu. Buddhist or Christian.
Of
The remaining 71
None of the sample households was
The head of the household was normally (93
percent) a male, married (85 percent) and usually claimed to be an owner cultivator (43 percent).
Some (28 percent) were cultivators and a few
(5.3 percent) were businessmen.
Surprisingly, only 2.4 percent of the
household heads were fishermen although the sample area straddled the Brahmaputra River. occupations.
Servants made up another 2.2 percent of the primary
The old, the disabled, the unemployed and the beggars made
up a further 7.7 percent.
Twenty-five percent of the household heads had
a secondary occupation, augmenting their primary income by cultivating their own land (10.1 percent), laboring on someone else's (6.4 percent), doing business (2.7 percent), fishing (0.4 percent) and even begging (0.9 percent). Each household's wealth, or its lack thereof, was bound up in its land.
The average acreage per household was only 1.15 acres at the time
of the survey compared with the Bangladesh average of about 3 acres (Tyers, 1978^).
Only 7.7 percent of the households claimed to have more
than 5 acres and 36 percent had no land at all.
The cattle and ploughs
necessary for tilling the land were similarly scarce.
Although nine of
the households had 8 cattle at the time of the survey, the average number
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
224 of cattle per household was 1.2, and 52 percent of the households had no cattle at all. plough.
Similarly 53 percent of the households did not have a
Goats are
the alternateto cattle for sacrifice in hard times,
but 81 percent of the householdsdid not even have a goat.
The average
number of goats per household was only .33, or one goat between three households.
Corrugated-iron roofs are often a sign of wealth and status,
but only 20 percent of the houses boasted such roofs, the remainder had thatch roofs and in some cases even those had been sold.
With few
possessions, migration was relatively easy, but trade and relief to revitalize the area were limited since only 8 percent of the households had either a cart or a boat. During the nine or ten months between the end of the last floods and the beginning of the survey, households had been selling off their possessions to raise their purchasing power. had sold 0.8 acres of land.
On average each household
In some thanas the land registry office
showed that over 1000 acres had been officially transferred in a single month in early 1975.
This did not include the unofficial sales and
mortgaging that undoubtedly occurred.
In the same period 35 percent of
the households had
sold off someof
holds had lost 0.6
cattle within10 months, i.e., one-third of their
cattle had been sold.
their cattle.
Each ofthe house足
Although only 13 percent of the households had
sold off goats, the average household loss of .26 goats meant that 44 percent of
the goat population had been sold.
Less than 1 percent of
the households sold off their ploughs or their cart or boat, but in the case of the carts and boats this meant a 10 percent loss in overall numbers over the nine month period.
With 5.1 percent of the households
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
225 selling off their corrugated iron roofs, it appears that at least
20
percent of the richer households were also affected by the famine. items had also been sold:
6.1percent
Smaller
of the households had sold off
trees or wood, 9.4 percent had also sold off furniture, old carved wooden chairs, or perhaps their almari (wooden dresser) or shinduk (wooden chest) used to store the household prized possessions, which may have included the malaria card.
The womenâ&#x20AC;&#x2122;s ornaments, often given at the time of
marriage as a form of insurance, had been sold off by 16.2 percent of the households. seemed to be the
The most sensitive indicator of famine induced selling household utensils, e.g., haripatil (cooking pots),
thala (brass plates), chanuch (brass cooking spoons), kashar gelash (brass drinking vessels) and bodna (toilet jugs).
Twenty-six percent
of the households were recorded as having sold such household utensils. Household food habits and nutritional status had also changed.
Even
at the time of the survey in May-June 1975, when the acute famine crisis was declining, 519 or
66percent
of the 790 households were still eating
less than one rice meal per day, which is in sharp contrast to the "normal" three rice meals per day in rural Bangladesh.
During the peak
famine period, from September 1974 to March 1975, the population had resorted to eating alternative "famine foodstuffs" which though ubiquitous, are considered unacceptable in normal times, e.g., thor (the hard external trunk of the plaintain sapling, Musa paradisica) . and bonn kochu or wild arum (Araceae spp.), wheat gruels, khichuris, rice husks, leaves, and even fen (starchy water drained after the rice is cooked).
Of the 790
households, 450 were eating the plaintain trunk during the peak famine period, but by the time of the survey this number had dropped to 50. Similarly the number of sample households eating bonn kochu over the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
226
same period had dropped from 325 to 125.
The results of the short term
decline of the calorific intake were revealed in weight for height data. Anthropometric measurement detected 19 percent of the children aged between
2and 10as
being below 80 percent of the normal weight for height
(Harvard Standards Stuart and Stevenson), i.e., severely malnourished. This was severe compared with Bangladesh "norm" established by the 1966 Nutrition Survey of about 5 percent being severely malnourished.
It
was low, however, compared with two other nutrition surveys in one thana (Rowmari) during the peak of the famine. QUAC stick survey and Harvard Standards.
One of those surveys was a
the other was a weight for height survey using
Both reported that 38 percent of the children at that
time were severely malnourished.
Although the low weight for height
data at the time of the later survey were probably the results of a temporary famine, the weight for age figures showing 33 percent of the children between ages 2 and 10 being below 60 percent of the International Weight for Age Standard suggest long term nutritional deprivation in the area. Households were unable to purchase and consume sufficient food. This dietary deficiency was coupled with the increased prevalence of diarrhoeal disease, particularly shigella shiga observed by smallpox doctors in the area and evidenced by overtly prolapsed rectums on several children.
Together malnutrition and disease in the area increased
the crude death rate to 49 per thousand.
If this high death rate had
indeed lasted for a whole year, it would have meant that the famine had been associated with 55,000 excess deaths within the survey area.
The
assumption is unlikely, however, as the death rate appears to have peaked
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
227 in the four months between September and December, so the computation of a yearly rate from the nine month survey data is invalid.
Just as the
death rate increased, so the crude birth rate showed a response to the physical stress in the area, by falling to 27 per thousand for the year. Some households opted to move, either in toto or in part.
Assuming
that the updating of the village malaria maps in mid-1973 was accurate,
21percent
of households in the area had moved in toto in the two year
period before the mid-1975 famine survey.
Neighbors, who still lived
near to the house sites of households who had moved away, reported that
68percent
of the emigrant households had moved "in search of food and
work" and that 32 percent had moved "to a safer place for settlement and work."
One crude inference might be that
68percent
of the emigrant
households had therefore moved because of the famine and 32 percent had moved because of the preceding flood. During the nine months between the end of the flooding in September 1974 and the survey in mid-1975, only 2 percent of the individuals in the total population had migrated, suggesting that famine-induced migration in the area generally means the movement of the whole household rather than individuals. Relief operations in the survey area had been intensive.
Thirty-
three percent of the households surveyed had had a member working on "test relief."^
Remembering that traditionally famine was officially
^"Test relief" provided public works at depressed wage levels to "test" how many people were willing to work for how long at such a low rate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
228 declared when 0.5 percent of the population was found to be consistentlyworking on the test relief projects (Bengal Famine Code), this figure is indeed large.
Forty percent of the households had had a member attending
gruel kitchens to receive cocked food.
Almost half the households (46
percent) stated that they had received "food aid," but further questioning revealed that they had received only one allocation in most cases.
Differentiating Thanas Liable to Famine Syndromes, with many symptoms, like famine are intrinsically difficult to quantify and hence to compare.
Can we really say that a
certain area, in this case a thana, was more affected than another thana? Which symptoms should be chosen to measure the intensity? symptoms be weighted more than other symptoms?
Should some
Just as it was tempting
to toss all the components of famine vulnerability into a factor analysis program, and then use only the components which the computer returned with a high factor score, as the meaningful components, so with the famine symptoms.
Before abrogating
total responsibility in this case,
however, it may again be more appropriate to select the most appropriate symptoms, using the best judgment available, in this case the researcher's experience throughout the famine survey area.
Each of the areas or
thanas may be ranked for each of the symptoms, portrayed together, graphically and then the total collage interpreted by the research in the light of the known facts.
I consider that in a situation where so little
is known about the causal relations between the symptoms, that this is the best available method. I have chosen eight symptoms, as a manageable number easily por足 trayed graphically, to represent the famine status in each of the ten
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
229 thanas surveyed.
I have had to omit several symptoms of the famine
syndrome (Chapter I) for want of information, i.e., crime, begging, borrowing, unsanitary conditions and disease.
Some of these, e.g.,
crime and begging, might not occur within the areas where the famines are generated.
Also I have omitted certain symptoms, e.g., a decline
in the number of births, or the selling off of trees and wood, for want of a sufficiently large sample size to detect significant differences between thanas.
The eight chosen symptoms are thus:
Percent of households selling household utensils:— Haripatil (commonly household utensils) are not normally sold. the household from generation to generation.
They remain in
In times of crisis, how
ever, the metal utensils offer small easily saleable items which are often replaceable by cheap pottery or wooden ware.
Households can raise
money by selling their thala (large heavy plates) and basun (smaller plates) which may be brass or enamel.
Similarly the chamuch (brass
cooking spoons) and handi (cooking pots) are sold together with the katari (heavy chopping blade) and the chota (griddle).
Even the kolshi
(large water pot) and the lota or badna (toilet jugs) and the kashar gelash (brass drinking vessels) may be dispensed with to raise the house hold purchasing capacity.
Although the different religious groups may
have had different types of utensils (Aziz, 1976), e.g., the Hindus may have had a higher proportion of brassware, nevertheless there was no significant difference (p=0.93) in the proportion of Hindu households selling off utensils, compared with the Muslim households.
Although house
hold utensils are items of relatively little value, they still reflect the suggestion (Chapter I) that famine symptoms may not necessarily be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
230 most visible among the poorest.
Thus the axiom that you have to "have"
before you can "sell" off lessens the resilience of
the poor, while the
rich having greater means of storage and savings are not forced to sell so much.
It was not only among the landless that the percentage of
households selling off their utensils was very high.
Twenty-nine percent
of the households who had been landless before the famine sold off their household utensils, but the percentage was even slightly higher for the households who had between zero and two acres of land before the famine.
The
percent selling utensils dropped to only 18 percent for
households owning between two and five acres of land, and it fell even lower, to 7 percent, for households owning more than five acres of land
000).
before the famine (p=- .
This same principle that those who show greatest loss are the poor, but not the
poorest was reflected when the pattern of utensil sales
was compared against the primary occupation of the household head (p=0.00).
Those who claimed to be beggars, with perhaps only one
aluminium bowl for holding poisha (coins), sold off only 18.2 percent of their utensils.
Households headed by landless laborers, servants and
fishermen, with few other forms of dispensable savings, had higher pro足 portions selling off their utensils: 57.9 percent respectively.
37.4 percent, 57.1 percent and
Those headed by businessmen, although probably
000),
richer, still had significantly less land (p= .
hence fewer savings
options and 31 percent of them had to sell their utensils.
Only 16.7
percent of the households headed by owner cultivators, however, needed to sell their utensils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
231 Percent of average household acreage lost:â&#x20AC;&#x201D; In Bangladesh, a manâ&#x20AC;&#x2122;s land is his wealth.
Real wealth is in paddy land (cf. Mahalanobis Survey,
1944), but in this study, figures were gathered only for the total land area.
No attempt was made to record the quality of land or the number
of crops per year.
The loss of land in an area like eastern Rangpur,
where all but a small minority of the civilian labor force are employed in agriculture, not only means increased dependency on the remaining landholding households for work, but it also means fewer degrees of freedom to rent to others or mortgage or sell that land in later times of need. In some thanas the acreage of officially registered land sales at the thana registry offices showed increases of up to ten times during the famine period from September 1974 to March 1975, compared with the transactions during the twelve preceding months.
Some thanas in the
survey area showed much larger increases than others (e.g., Figure 31). The more credible figures recorded by the sample survey at the household level also showed that large numbers of households had transferred their land. The percentage of households who had sold their land increased as the size of their original land holdings decreased for those households who made small sales of land between one-tenth and two-tenths of an acre.
This negative correlation between the percentage of households
making land sales and
the original size of the landholding was obscured
for sales above two-tenths of an acre.
Sales of land areas over three-
tenths of an acre showed an almost uniform pattern among different classes of original holdings:
34 percent, 32 percent, and 34 percent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
232
1,000 ROWMARI SUNDERGANJ
900
800
Uj 700
Qj 600
kl
500
5
4 00
^
300
200
100 0 1973
1974
1975
YEARS Figure 31.
Seasonal sales of land over two years in the Bangladesh thanas at Sunderganj and Rowmari.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
233 among households formerly having small, medium, and large holdings, respectively.
These larger sales may not have been famine induced, but
rather a response to other market forces, e.g., land prices may have risen despite the glut as happened in the 1943 famine (cf. Mahalanobis, 1946). Following the pattern of household utensil sales, there was no significant difference in the proportions of Muslim or Hindu households selling land.
However, it was interesting to note that all the Hindu
households selling land sold it to buy food and, unlike some of the Muslim households, none sold it to buy better land elsewhere, possibly suggesting either the reluctance or the difficulty for Hindus investing elsewhere in land in Bangladesh.
Table
8
Percentage of Households Selling Land over the Famine Period 0-0.09 acres
0.10-0.19 acres
Cultivator as household head (N=343)
3
18
Laborer as household head (N=222)
9
16
The sale of the small amounts of land, i.e., 0-0.09 and 0.10-0.19 acres followed the same pattern as that of the sales of household utensils, when compared between households headed by cultivators and those headed by laborers.
The percentage of households headed by
laborers, and selling off land was consistently higher than that for households headed by cultivators in both the 0-0.9 acre and 0.10-0.19 acre sales.
This confirms that the laborers are not only the more
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
234 susceptible to famine, having small land plots initially, but that they are also the more severely affected, leaving them in an even more vulnerable position should another famine occur. The percentage of household land acreage lost was selected as an indicator of famine severity because it does indicate rural impoverish ment and had been
used in past famine surveys (Mukherjee, 1974;
Mahalanobis, 1946), but as an indicator it masks many other confounding variables.
Percent increase in the number of landless laborers:â&#x20AC;&#x201D; Land sales alone do not indicate whether those selling have sold all or only part of their land. no land holding.
Landlessness has been used here to mean households with These households are totally dependent on employment,
and hence food supply, from others.
They are thus the first groups to
suffer in times of economic depression, unless they have other income sources, e.g., fishing.
The percentage of landless households in the
survey area increased from 27.7 percent to 35.9 percent between September 1974 and May-June 1975.
This increase in landlessness was almost
entirely (97 percent) from those households that had previously owned less than two acres of land. The occupations of household heads featuring the greatest change towards landlessness were those with skills, e.g., tailors (40 percent), carpenters (25 percent) and teachers (14 percent).
Households headed by
a laborer suffered more than households headed by owner cultivators: 17.1 percent of the laborers became landless.
It is interesting to note
that even households having businessmen and servants as the household
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
235 heads:
i.e., those presumably with ties in the urban areas, suffered as
well, with
11percent
of each category becoming landless.
Percent eating less than one rice meal per day:â&#x20AC;&#x201D; Despite government efforts to introduce a dry season wheat crop, particularly in north Bengal, and the prevalence of other grain crop alternatives, e.g., china kaon or pearl millet (Pennisetum glaucum) in some of the study char (newly emerged riverine land) areas, rice is nevertheless the staple grain.
Old people sometimes appeared psychologically unable to digest
wheat even when dying from starvation, so accustomed were they to rice. The rural Bangladesh norm of three rice meals per day, usually garnished with some protein in the form of beef, goat, chicken, prawn or fish,
with
a vegetable curry is probably a stereotype rather than a norm
today.
Although the famine survey was conducted after the acute famine crisis was over, there were still more than 500 of the 790 households that were eating less than one rice meal per day. The number
of rice meals that households were eating at the time
of the famine survey rose consistently when compared with the acreage of the household after the 1974 flood and before the famine began.
Ninety
percent of the households who had held no land before the famine, were eating less than one rice meal per day in May-June 1975.
The percentage
on less than one rice meal per day fell to 70 percent for those who had owned 0-2 acres, to 44 percent for those who had owned 2-5 acres and to 20 percent for households that had owned over 5 acres of land. The occupation of the household heads in the few households that were still eating three or more rice meals each day was invariably an "owner cultivator."
Only 46 percent of the households headed by "owner
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
236 cultivators" were eating a rice meal less than once a day.
This perÂ
00
centage significantly different (p= . ) from that of the households headed by laborers (95.6 percent) or fishermen (90 percent).
All (100
percent) of the households headed by servants and beggars were eating less than one rice meal per day. This indicator seems to reflect socio-economic status as well, but it may be clouded by differences in the quality of rice intake, the availability of other staples or lack of rice in certain areas, the varying definitions of "a rice meal" and the possibility of relief doles confounding the picture presented by the indicator.
Percent of households eating bonn kochu (wild arum) between 1st Ramzan and the Bengali New Year:â&#x20AC;&#x201D; Foods which were both non-seasonal and unacceptable in the normal times exhibited different consumption patterns between the famine period (September 1974 to March 1975) and the period after the famine (March 1975 to May-June 1975).
Although the intake of
plantain saplings (bananas) appeared to be most sensitive to famine conditions with
460 households using it as a food during the famine
period, compared with only 60 households after the famine period, bonn kochu (wild arum, Araceae spp) was chosen as a famine indicator because it was free from the difficulties of definition associated with the banana tree and also because its intake, was only slightly less sensitive to famine conditions than that of
the plantain sapling.
If someone is
eating any part of the bonn kochu as opposed to the man kochu or the pani kochu (all Araceae spp.), they are eating a famine food, one that lacerates the interior of the mouth.
Although those eating the hard
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
237 trunk of the plantain are also eating a famine food, some people con sume the softer interior parts of the trunk as a delicacy even in normal times.
Percentage of children aged between 2 and 10 less than 80 percent of the weight for height standards (Harvard);— Short term "wasting" as opposed
to long term "stunting" has efficiently been measured by com
paring the
weight against the height of people in a disaster affected
population (Waterlow,
1973).
Children, i.e., persons less than 10 years
of age are normally the most sensitive nutritional group during a famine and children aged
2-10were
chosen because it was assumed that a proportion
of the children less than two were still benefiting from breast feeding. Because of the difficulties, in assessing age above about 4 to 5 years in Bangladesh, children above 10 were equated with those weighing more than 25 kgs. or measuring higher than 120 cms.
The weight-height
ratios of males and females were then compared xtfith the 50th percentile of the Harvard weight for height standards and males and females under 80 percent of their standard were considered to be severely malnourished. Such an anthropometric indicator has the benefit of being a direct measurement as opposed to other indicators which are often only reported perceptions.
In this study the indicator appeared to be consistent from
both males and females because no significant differences were found between the percentages of males and females who were less than 80 per cent of their respective standard weights for their height.
Although the
indicator had these positive characteristics, there were nevertheless several limitations:
some children were away playing in the fields and
were not available for weighing and heighting; some children, perhaps those
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
238 who were most malnourished, refused to come out of their houses for weighing and heighting; and others threw such tantrums that weighing and heighting were not possible.
It is also important to note that the
indicator is certainly objective, but it can only measure the nutritional status of those children who have not yet died of starvation and those who have not yet migrated from the village with their families.
Another
limitation of such an anthropometric indicator is that a community's status may fall because of an epidemic disease which may not be associated with a famine at all.
Finally nutritional status may rebound quickly from
severe malnutrition because feeding programs have been introduced among certain groups, but the other symptoms of famine may still remain in an area.
Percent of households who moved in search of food or work outside the thana in the past two years;â&#x20AC;&#x201D; If households were listed on the malaria lists which had been updated two years prior to the survey, but were no longer found, they were assumed to have moved.
In 96 percent of the
cases this was corroborated by the neighbors.
Often the bhiti (raised
earth floors) of the huts were still visible although grassed over. Banana plants or coconut trees would also mark former household locations. Neighbors were also able to say why, and approximately whither, households had moved. In Figure 32 the size of the circles represents the number of house hold migrations within the past two years.
It can be seen that there are
some thanas such as Ulipur, Chilmari, and Shaghatta where the total number of households migrating has been larger than in other thanas.
However,
closer inspection of the pie-graphs reveals that in a thana like
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
239
K £ Y : %Households who moved mid 1973-1975
iNAGESWARI
KURIGRAM
iULIPUR iROWMARI ICHILMARI SUNDERGONJI
IGAIBANDHA
FULCHARI DEWANGANJ SHAGHATTA
PERCENT HOUSEHOLDS MOVING: P77777, OUTSIDE THANA IN SEARCH OF * 2 ** FOOD AND WORK OUTSIDE THANA TO SAFER PLACE FOR SE TTLE M E N T r r - 7- i WITHIN TH ANA IN SEARCH OF 1 FOOD AND WORK PS*™ WITHIN TH ANA TO SAFER PLACE FOR SE TTLE M E N T § f j ^ TO UNKNOWN DESTINATION AREA OF CIRCLE PROPORTIONAL TO PERCENT HOUSEHOLDS MOVING
Figure 32.
Migration of households from settlements during floods in Bangladesh.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
240 Shaghatta, over 30 percent of the household migration was within the thana in search of a safer place to settle in.
The migration was
probably the result of extensive river erosion which removed up to one mile of river bank in the monsoon of 1974 (see photomap, Figure 27), forcing families to relocate in temporary bastee conditions on flood embankments or on land within the thana owned by relatives.
The
thana was therefore able to absorb this sudden displacement and re location of people.
Households that were reported to have moved to a
location outside the thana "in search of food and work" were more likely to reflect famine symptoms because such a response suggests the inability of the locale to sustain them.
Percent of the population dying from "causes" recorded as starvation or "lack of food":-— If out-migration has not occurred, death is the terminal symptom of famine.
It indicates the community’s in
ability to sustain a community member.
Forty-seven percent of the 130
deaths in the famine survey area between September 1974 and May-June 1975 were reportedly caused by starvation or lack of food.
Sixteen
percent were ascribed to diarrhoea or cholera and 5 percent were thought to have been caused by dysentery. for the remaining 32 percent.
Fever, smallpox and others accounted
Deaths in the area for those nine months
were dominated by "starvation" and "starvation-related" deaths. These starvation-related deaths were significantly higher, both in absolute numbers, and as a percentage of the surveyed population among the young and old. old males. increased
The increased death rate was greater among young and
This normal preponderance of male deaths may have been during the famine as the cultural cosetting of the males was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
241 Table 9 Famine Death Rates by Occupation of Household Head
Occupation of the No. of H.H. Household Head_______________ without Deaths
No. of H.H. with Deaths
% H.H. with Deaths
Cultivator
306
37
10.8
Laborer
179
43
19.4
Businessman
39
3
Fisherman
17
2
10.5
5
29.4
Teacher
12 6
Beggar
14
Servant
Carpenter
4
Housework
15
Old or Disabled
31
Tailor
5
Priest or Imam
3
Self Profession
19
Self-employed
16
1
Unemployed Others
4
%
1 8 0 6 7
0 1 1 2 0 1
TOTAL
671
117
TOTAL
85.2
14.8
7.1
14.28 36.36
0.0 29.0 18.4
0.0 25.0 5.0
11.1 0.0 20.0
Chi Square = 28.64444 with 15 degrees of freedom Significance = 0.0179
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
242 relaxed.
The number of deaths documented in the famine survey that
were ascribed to starvation appeared to be particularly high among the older men aged 50-59 and 60-69. It was the poor households, i.e., those with little or no land, who suffered the most severe mortality between the September 1974 famine and the survey in May-June 1975.
Table 10 Mortality Rates by Amount of Land Ownership Before Famine of Household Head
Amount of land before the famine
HH without Deaths
HH with Row Deaths (%) Total
None
167
52
(24)
Less than 2 acres
306
43
(
2-5 acres
135
14
Over 5 acres
65
Column Total
673
8 117
12) (1 0) (1 1) (15)
(%)
219
( 27.7)
349
( 44.2)
149
( 18.9)
72
(
790
(
9.2)
100.0)
Chi Square = 20.00952 with 4 degrees of freedom.
Thus the relative risk of a household having a death or deaths during the famine period was over twice as great for households that had had no land as for households with 5 acres of land.
Comparing the Indices of Famine Symptoms Similarly Table 9 shows that the relative risk of a household having a death or deaths during the famine was 3-1/2 times greater if the house足 hold was headed by a beggar than if it was headed by an owner cultivator. Households headed by servants showed a relative risk three times as great
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
243
Indices of fa m in e sym ptom s fo r the ten thana survey
0 iiS iS: 52.6 â&#x2013; Nages
Dewanganj
17 9 iu S i& iHHBH 64 1 liffiijjBBI
Famine symptom with an unusually high ranking compared with other symptoms in that thana ipared with other symptoms in that that
Stidderganj 0 .2 8 NOTE: A polygon is drawn ,. = Ranking of that thai
/anganj 0.42
immediately after the peak famir
Bruce Currey
Figure 33.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
244 as those headed by owner cultivators and the households headed by laborers were twice as susceptible to having a death or deaths during the famine as those headed by owner cultivators.
Households headed by
businessmen appeared to have a low relative risk. These eight symptoms chosen to characterize famine in the survey area may have risen, peaked and declined at different times in different areas.
The survey recorded their occurrence at different times or periods:
some through measurement, others through past recall.
All of them have
limitations because the underlying processes which bring about famine are still little understood.
Because of this the causal relationships among
the variables is also poorly understood.
Rather than mask such limita足
tions in the form of factor scores on unknown factors, the eight indica足 tors were ranked for each of the ten thanas and portrayed cartographically in Figure 33.
The polygonal graphs, a concept originally developed by
the Program for International Development at Clark University (1976) show certain thanas to have been more severely affected by many of the famine symptoms than were other thanas. sistently equal radii.
None of the polygons have con足
This portrays the lack of consistency among
the indicators, i.e., segments indented in the shaded polygon suggest that a particular indicator fails to corroborate the severity of that symptom compared with other highly ranked indicators in the thana, e.g., the relatively low percentage of severely malnourished children in Rowmari.
These inconsistencies have also been highlighted by the shading
of deviant rankings for each indicator in each thana on the table accompanying the diagram.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
245 The polygonal diagrams point out that three of the thanas, Rowmari, Fulchari and Chilmari, were very severely affected on most of the symptoms. The concave segment on the Rowmari and Chilmari polygons corresponds to the low percentage of children who were classified as severely malnourished. This is probably because the anthropometric measurements were made in June 1975 after the famine when the most malnourished members had died or migrated from two of the worst affected areas and after feeding pro grams had been concentrated in these areas.
Chilmari also ranked low on
the indicators for both the eating of bonn kochu (wild arum) and starvation-related mortality.
The former might be caused by the avail
ability of the feeding stations as substitutes or by all of the bonn kochu having been eaten by that time.
The low rating on the mortality indicator
may be associated with the location of a railway terminus in the center of the small thana, i.e., relief may come in quickly by rail or individual members of families may have "caught a ride" (often on the train roof) to Kurigram to beg or to look for relief. Some of those thanas, e.g., Nageswari, Kurigram, Sunderganj, Shaghatta and Dewanganj, which are less severely affected in terms of the eight indicators nevertheless show aberrations on a few indicators.
Thus
Nageswari, although bountiful in terms of its man-land ratio, and relatively free from flooding because of its generally higher elevation, nevertheless had a high percentage of household out-migration.
Dis
cussion with local officials suggested that there had indeed been scarcity but not famine in Nageswari.
The relatively high rate of house
hold out-migration may have been due to the proximity of other oppor tunities because access to Kurigram and points west is easy via an army
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
246 road and an excellent bus service from the north-east.
Ulipur thana*s
high percentage of households migrating may also have been due to attrac tions elsewhere because of proximity to Kurigram town from the south. Kurigram thana's high percentage of land loss may have been caused by its urban nature, i.e., townspeople from Kurigram or people visiting the town saw an opportunity to buy up land during the scarcity.
The high
percentage (55 percent) of households eating bonn kochu in Sunderganj thana remains unexplained.
The thana was relatively protected from the
Tista and Brahmaputra flooding by the "right flood embankment" which had remained intact there.
All local officials were in general agreement that
the thana had not been severely affected by the famine and this is sup ported by the generally low ranking on the other indicators.
In Shaghatta
thana the high percentage increase in landlessness (12.5 percent) was probably partially the result of extensive river erosion on the right bank of the Brahmaputra during the 1974 monsoon.
Local officials
reported that "bank line migration" had been as great as one mile from east to west.
Whether Shaghatta's high ranking on the "sales of house
hold utensils" and the "eating of bonn kochu" is connected with the resulting household migration away from the river bank towards flood embankments further within the thana remains unsolved.
Dewanganj thana*s
high ranking on "mortality caused by starvation and food lack" and the "percentage of children aged between 2 and 10 who were less than 80 per cent of the standard weight for height" was probably caused by the in clusion of Hatibanga village, a large village of 26 households consisting of 33 percent of the sample from the thana.
According to local officials
the severity of famine in Hatibanga village was unique in Dewanganj thana
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
247 because of its riverine situation.
Despite severe flooding and river
erosion its remote situation had precluded the distribution of relief to that thana and the 26 households sampled had been severely affected by the terminal symptoms of famine.
Final Comparison
of
the A r e a l
Intensity
of
Famine
Symptoms wit h
th e
Delbecq-Delphi Map
The comparison of these composite indices of famine symptoms during the 19 74-75 famine with the index of famine derived on the Delbecq-Delphi map (see Appendix H) shows no agreement.
Even a comparison of the
dichotomous groupings (i.e., thanas "very liable to famine" and thanas ranked 1-5 on the famine symptom index versus thanas liable to famine and thanas ranked 6-10 on the famine symptom index) does not show any agreement (Figure 25).
As discussed in Chapter II, the Delbecq-Delphi
process, which attempts to delimit liability at a regional scale, is unable to cope with the components of famine vulnerability which are described above at a local scale.
In Chapter II it was also suggested
that the components of famine vulnerability in a small local area such as this would be likely to change from year to year or even season to season, whereas the regional pattern of famine liability is likely to be more persistent.
Conclusion
The findings of this chapter support the proposition that the Delbecq-Delphi map compiled in this study is able to differentiate areas liable to famine at a regional scale.
This is supported by the map's
general consistency with the mental maps that various groups have of such
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
248 areas, by newspaper reports of areas affected by the 1974-75 famine, and by the map of the places of origin of bustee (slum) dwellers who have left their homes within the Dacca catchment area and have then settled in urban slum conditions.
A detailed inter-thana comparison of 10 famine
affected thanas, however, suggests that the Delbecq-Delphi map does not predict the intensity of famine symptoms at the thana level.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VII FAMINE REMEDIAL POLICIES
Comment Faire
Reculer
IN B A N G L A D E S H
la P a u v r e t e Rur a l e
Gilbert Etienne The of
famine
transformed
the p ast
some
of
the
inertia
i nto energy. B.
G.
Verghese
The Delbecq-Delphi map allows those making decisions on famine prevention and development at the national scale to focus on particular regions of Bangladesh that are more liable to famine than others.
The
map focuses attention on 124 thanas which are probably liable to famine and emphasizes an additional 40 thanas which are probably very liable to famine.
The areas classed as being liable to famine include 26 percent
of the land area of Bangladesh and 29 percent of the population.
The
small area which is very liable to famine covers only 9 percent of the area, but an additional 12 percent of the population.
The opportunities
to relocate the 21.4 million people from the area liable to famine and the additional 7.9 million from the area very liable to famine are limited (ChapterIII and Revelle, 1970).
The problem therefore has to be
tackled principally within the areas themselves. The
simple act
of
fo cusing on these
t he p r o b l e m of a n t i c i p a t i n g these
regions
brings
of f a m i n e p r e v e n t i o n , preparations
shape
famines to
showing
areas brings
there.
The very
attention
to b e a r
the h e r e t o f o r e neb u l o u s
nationwide problem
i t as a m a n a g e a b l e
for w h i c h
task
on
act of h i g h l i g h t i n g
specific
ca n be made.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
250 The map has the advantage over a table listing the famine liable thanas because on the map the distinct areal pattern of areas liable to famine can be discerned. coastal.
Most such areas are seen to be riverine or
In fact, the 40 thanas that are classed as being very liable to
famine all abut on a major river or coast.
Also, the proportion of
thanas liable to famine is seen to be particularly high in north Bengal, i.e., north of the Ganges river and west of the Brahmaputra river. The map being
also
indicates
associated with
documentation of change
of
thought
these
t e n of
famine
correlates
from symptom
possible preventive measures; aid
to
increased
indigenous
the mo st
liability
factors The
as
cartographic
areas
liable
to
c a use;
from relief measures
thus h o p e f u l l y
famine
selected
to p o s s i b l e and
of
crucial
in these areas.
encourages
from increased
a to
external
self-help.
Knowing the areas liable to famine and the correlates of famine liability it should be possible to suggest remedial policies appropriate both to the particular area and to the particular cause.
Any such remedial
policies, however, as with any policy, must be sieved through the par足 ticular social and geographical situation where they are to be implemented. Some of the more important aspects of this situation in Bangladesh are the limited and constantly changing area of arable land, the seasonality and variability of the monsoon and its consequent hydrological patterns, the large and increasingly dependent and unemployed population, the con足 tinuing impoverishment of the rural landless population, and the over足 whelming inability to evolve a coherent rural community initiative (see Chapter III).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
251 The need to filter policies through such an indigenous context is often avoided or scrimped on by short term foreign consultants.
The
Delbecq-Delphi process (Figure 18), by allowing the involvement of local expertise and long term experience from the initial stages of the planning process, is assumed to partially provide a built-in contextual filter. Although in this case 25 percent of the initial 180 respondents were nonBengali, the majority were nationals who had been raised in their own country.
Certainly some had served a colonial government, some had
travelled or been educated abroad, and most had read western literature, but they, nevertheless, represent the present Bengali decision makers. The bulk of the remedial policies they suggested for famine prevention in Bangladesh had thus been winnowed through their own experience of the country and through their cultural intuition. All the 180 respondents were initially asked to suggest ways in which famine could be prevented from returning to the areas they had ringed as being liable to famine (see Appendix A: Questions I and J; Appendix B: Questions I and J; Appendix C: Questions I and J; and Appendix D: Questions I and J).
They were asked what the villagers themselves
could do and what the government could do. into two classes:
The responses fell broadly
national policies and local or regional policies.
However, the local or regional policies were related to the possible causes of the famine rather than to a particular famine location.
This
should have been expected in carrying out such a study at the national scale.
However, knowing the areas affected by the possible causes (see
the inserts on Map, Appendix H), it is possible to allocate preventive measures proposed to remedy these possible causes to those same areas (see the blue text on Map, Appendix H).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
252 National Remedial Policies
Suggested policies in response to the questions (see Appendices A, B, C, D) on potential preventive measures at the national scale ranged from the broad to the specific:
from realization of the need for
a "revolution of consciousness" to the more tangible "price stabilization" and "reorientation of agricultural education," to specific suggestions from "improved coordination" of famine prevention and "preparedness planning."
R e v o l u t i o n of
Consciousness
Although Bangladesh has a strong Muslim culture and strong western ties, the Bengalis' suggestions seemed to mirror together the thoughts of both Paulo Freire (1970) and present-day Chinese development strate足 gies.
To the need for a "revolution of consciousness" was added the
need for the "rich to help the poor."
The general call for "social
reform" was backed by the calls for the "sharing of resources" and even a "commune system."
Echoes of Chinese thoughts were further reflected in
the urge to "use manpower not bull-dozers" and to "train students in rural areas."
Along with this strong appeal to "gamble on new ideas" and to
encourage both "self help" and "self ownership of agricultural land," the overriding call was for "a definite rural policy."
There was, however,
either ambiguity or indifference as to whether this should be achieved by "a move towards socialism" or a "return to the old landlord system."
Price Stabilization A strong consensus of respondents suggested that the central govern足 ment could beneficially manipulate prices before and during a famine.
In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
253 the long term it was suggested that the government "maintain a base price for the main crops and publish this price at least one month before sowing."
"Price subsidies" were also suggested to "encourage cereal
production." normal times."
This was supported by calls to "increase the rice price in When scarcity and famine broke out, however, "price
control on essential foods" was suggested despite "the possibility of a mushrooming black market."
It was suggested that Indo-Bangladesh black-
marketeering be reduced during times of famine by appropriate "devaluation of the taka" and "strong military policing of the border and all road, rail and boat traffic."
As a corollary to the rising food prices during
times of famine the slump in prices for sales of possessions might be countered by an "improved government disaster credit system."
The
"availability of government loans rather than moha.jan (moneylender) loans" might "stem the tendency towards polarization of wealth."
Price stabil足
ization during famine might best be maintained by the "provision of unemployment specifically tailored to the skills of the poor."
Reorientation of Agricultural Education Respondents repeatedly stressed the need to "change the basic educational curriculum," to "lessen the emphasis on Political Science and English" and "establish agricultural sciences as a priority."
They
emphasized the need for a "greater appreciation of the value of manual labor."
It was advised that the "agricultural university be reoriented"
and that "agricultural education be regionalized," so that "agricul足 turalists from saline areas could be trained in saline areas" and "crop innovations could be matched with the particular hazard conditions" of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
254 different
areas.
monsoon crops," and
"drought
flood
cycles
Attempts and
crop
should be made
research should
resistant plants" and
to
"reduce
stress
as w e l l as
dependence
"famine
"meshing
o n th e
avoidance
crops"
cropping patterns with
drought patterns."
Improved Coordination of Famine Prevention and Preparedness Planning The respondents stressed the need to "improve the coordination" of "overall disaster and famine prevention planning."
Such coordination
would be international, involving "land use and water agreements with areas upstream in India, China and Nepal."
Nationally there should be
more "integrated planning between ministries" and there was a need articulated to "streamline the Ministry of Relief and Rehabilitation." Famine preparedness planning should include "greater involvement in famine forecasting" including the "organization of a cell with 2-3 officers" with "technical expertise to monitor famine warning indicators," e.g., "flood patterns," "rainfall trends," "sowing times," "satellite imagery analysis," "labor wages," "market prices," "transport avail足 ability," "local official reports" and "migration patterns."
Respondents
specifically mentioned the need to "identify famine prone areas" to "target relief to such areas early," to "identify and train famine remedial manpower among existing staff" (e.g., "Circle Officer Develop足 ment," or the "Thana Agricultural Officer") in such areas and to "have small projects ready at subdivisional and thana level in order to provide employment."
Regional Remedial Policies Although in answer
to
the
regional
the q u e s t i o n s
remedial policies
suggested by
on remedial policies
do d e v e l o p
the re spondents f r o m s o m e of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
255 the national policies, I decided to arrange the responses around the 10 regionally related components of famine liability in particular areas of the country (see Map in Appendix H ) .
Flooding
Destruction from flooding may be lessened by the "selection of crops whose growth patterns mesh with the flood patterns."
This would entail
"less dependence on the kharif and encouragement of rabi crops," e.g., "sweet potato," "potato," "sorghum," "groundnuts," "millet," "wheat" and "boro rice."
Areas prone to early flash floods, e.g., Sylhet, Comilla,
Chittagong, Rangpur and Dinajpur, might experiment with "early maturing and stiff-strawed varieties of boro and aus."
"Post flood recovery
methods" should include "raised seed beds," "pre-germinated seeds," "the forced growth of seedlings on coconut mats," and "experiments with flood recovery crops" which are "appropriate to the area and to the particular time of year." "Control of flooding is unlikely," but there are "possibilities of adjusting the timing and height of floods'' by:
1
( ) "putting the coinciding
flood peaks of the tributaries out of phase to lessen the total flood height where the Brahmaputra and the Meghna meet"; (2) "afforestation on embankments and around tributary areas"; (3) "channelling flood sediments into bil areas"; (4) "local dyking and drainage schemes"; (5) "the demarcation of flood storage areas" to "use as fisheries" or as "areas for concentrating water hyacinth" "for mulch" and for "the pro足 duction of biogas."
"Experiments should be made with raised platforms for
vegetables," e.g., "puishak, in rainflooded areas"; with "emergency
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
256 floating gardens anchored near the baris"; and with "higher yielding varieties of deep water rice." "Storage in flood areas should be in elevated godowns" that are "easily accessible during the floods."
"The annual renewal of seed
banks and cattle food stocks" should be accompanied by an "improved system of agricultural loans."
Drought In traditionally drought-prone areas, "drought-tolerant crops" such as "millets, sorghum, sweet potatoes and groundnuts" should be encouraged. Improved "knowledge of both the seasonal extent and depth of local ground足 water and potential water holding areas near rivers" might suggest irrigation improvements through such "small scale systems as bamboo tubewells."
Further "investigation of such traditional irrigation
systems" as "swing baskets and dons," should be considered particularly "for emergency use after natural disasters."
"Irrigation should be
encouraged to increase rabi crops," but attempts should also be made to "avoid excessive reliance on petroleum fuels.
The possibility of
"employing 'hydraulic rams' or 'windmills' in areas with coastal or river breezes" should be explored.
"A cooperative system for the repair and
maintenance of existing pumps" becomes particularly important in times of drought. If labor is available and there is a sufficiently high water table, experiments should be made to "resuscitate the derelict pukurs (tanks)." If the "technical and tenurial problems" of the tanks can be overcome they might be used not only "for irrigation," but "for fisheries" as well.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
257 "Planting shade trees" and the carefully controlled "introduction of fecal material" might be tried to "increase the productivity of the tanks
Population Pressure Overpopulation at the village level becomes apparent because of the "failure to organize rural communities" and to "develop appropriate skills among their members."
Village needs might be better articulated if "the
landless and the artisans" were given "a voice in the Union Parishad (Union council)."
Aid might be channelled to "encourage landless
villagers" to "attend functional education programs tailored to villager's perceived needs," e.g., "mulching" or "afforestation techniques."
The
"traditional social security system under the zamindars (landlords)," where "the rich tried to help the poor" in all but the worst years, should be "reinstated through the gram panchayat (village judiciary committee)."
Aid might be allocated to panchayats only on the basis of
their proven ability to:
1
( ) "upgrade conditions for the landless";
(2) "organize the use of newly reclaimed land"; and (3) "organize wage labor migration prior to a crisis."
Food Deficit Some of the officially designated food deficit areas, e.g., Faridpur, become surplus areas in reality after "migratory wage laborers return from surplus areas," e.g., Sylhet.
Although "improved rodent-proof
dharma golas (community storage)," especially if located "near the areas vulnerable to natural disaster," might be effective if community organiza足 tion were improved, at present "improved individual bari golas (household
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
258 storage
)11seems
likely to be more effective.
Agricultural "extension
workers might revive traditional storage methods," e.g., "oiling clay jalas and kolshi," "hermetically sealing them with clay" and "inserting gypsum," if available or "charcoal," to "reduce humidity."
If they are
available, the use of "pucca drums" might also "reduce rodent, pest, moth, weevil and mildew attack."
"Improved fumigation" with carbon
bisulphide and "drying techniques" should also be explored for both house足 hold storage and local supply depots. The growth of such "alternative crops" as "dates, bananas or potatoes" on non-cereal land might be encouraged by improvements not only in storage but also in "food preservation techniques," e.g., "drying" or "carbide gassing."
Lack of Alternative Employment Disaster relief aid might provide a "unique opportunity to revitalize rural industry" if basic craft skills are extant, but "long term foreign market links must also be established."
A preliminary "survey of existing
skills," e.g., in "bele mati (clay)," "kansha (brass," "paat shilpa (jute handicrafts)," and jaal bona (net making)," should suggest areas in need of "functional education for the revival of traditional craftsmanship." After a natural disaster the "timely distribution of agricultural inputs for recovery," e.g., "seeds," "fertilizer," "ploughs," "kodals (digging hoes)," and "hand irrigation systems," could minimize the need for additional employment.
"Should immediate recovery be impeded" because
of ponding, salinity or sterile sand deposition, then "previously arranged food for work programs for the unskilled and landless," e.g., "dyking,"
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
259 "afforestation," "tank clearing," and "multipurpose canal excavation," should be quickly "put into operation, before the physical condition of the family wage earners deteriorates."
In many cases the "need for
employment may be anticipated in the 'slack' famine season between the monsoon flooding and the aman harvest." "If community organization can be strengthened, a communal system
natural disaster, certain wage-earning community members might be sent in search of temporary employment" in "fishing" or in "urban areas."
Low Crop Yields Many of these one crop areas actually have a surplus in normal years.
"Cooperative" and "enlarged individual storage systems," like
those in Sylhet, could retain such surplus in the area for times of disaster.
Investment in "horticulture," "poultry," "livestock," and
"bamboo groves" can be viewed as an additional form of storage for lean periods.
The present "excessive reliance on one crop" necessitates close
attention to "possible blight or saline incursion in such areas."
Where
possible, "early maturing varieties of rice" should also be introduced to "minimize the risk from early flash floods."
Poor Land Transport In times of disaster, "remote areas must be kept in better communi足 cation with Dacca," either by allowing increased "use of existing army radio links by public officials," or by providing an "independent disaster radio system."
If these radio links were in contact with a "disaster
information unit in Dacca," an "improved version of the Bengal Famine
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
260 Code" system of "monitoring scarcity conditions and disasters" could be established.
Villager elders might also "listen daily to a specific
frequency at a particular time for disaster warnings." Although "improved road communications are required" in such areas as "the Barind tract," "water transport should be increasingly considered for the distribution of relief" in other areas, particularly "the coastal areas west of Barisal." famine are riverine."
"Most of the thanas which are very liable to Estimates of the likely "seasonality of disaster"
in different areas might help to "determine the need for all-xveather roads to replace dirt tracks."
"Country boat and bullock cart owners
might be paid and prepared to serve as post-disaster transport." The vulnerability of remote areas might be reduced by "decentral足 izing storage facilities from Chittagong," by "improving the condition of godowns in remote areas," by "strong law enforcement in border areas" and by "improved control of border commerce to prevent smuggling."
"Rice
cordoning within the country should be used only with caution." "Strengthening village community organization," although "difficult to institute," could "lessen the loss of grain from remote villages in times of impending famine."
River Erosion With the possible exception of urban areas, "no efforts should be made to control erosion at this time" "unless it occurs in minor tributaries" where it can be combatted with "afforestation."
Possible
adjustments, however, might include the "introduction of quick maturing varieties of crops in erosion-prone areas."
These would then allow
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
261 "early harvesting before forced migration must occur."
If there are
"alternative skills" in the area, e.g., "fishing," "boat building" or "net making," they should be encouraged.
Rapid "recovery on the
depositional bank" might benefit from "increased research on legal aspects of riverine land transfer" and "rigorous enforcement of riverine land laws."
"Increased research in sand land agricultural techniques,"
"greater knowledge of germination times in different soil and moisture conditions" for "sorghum," "millet" and "sweet potatoes," should eventually allow a rapid recovery of employment levels on the depositional banks and chars.
Cyclones Although cyclones occur relatively suddenly and have "disastrous immediate effects," the "surviving population is initially healthy and well-nourished." are available."
They are thus "able to recover if the means of recovery The "development of water communication for relief
distribution" is particularly important in areas that are susceptible to cyclones.
Although "people have been reluctant to use the killa
system of raised mounds for shelter," "the killas should nonetheless be retained for sheltering water, food, seeds, cattle and farm equipment." A "water level embankment system might also be developed for sheltering fishing and relief boats."
"The low level storm surge and salinity
embankments should be carefully maintained and forested if possible." It is important that "sluices be carefully maintained," and where possible, "systems of controlled breaching" should be tried. be encouraged on all non-cereal land."
"Afforestation should
Besides those trees that provide
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
262 shelter, e.g., "coconuts" and "betelnuts," other "trees might also provide an income source," e.g., "bamboo" and "bananas."
"Settlement of
cyclone-prone areas should be minimized" until the "vulnerability of the one-crop areas prone to saline incursion and cyclone" is reduced by the "planting of an additional crop in the rabi season" and until the "development of fishing," "fish farming" and related "employment for the slack period" in the agricultural calendar.
Any "settlement developments
that do occur should be on elevated grounds" and "enclosed by trees" and should "avoid the use of corrugated sheeting," which can become a lethal weapon in high winds.
Maldistribution of Agricultural Inputs Inputs for agricultural production have to "reach the appropriate people at the appropriate place and at the appropriate time."
Although
this is "basically an administrative and organizational problem," "dis足 tribution directly to the cultivators" can be facilitated by the "use of boats during the wet season in the flooded areas."
"Increased perennial
use should be made of water transport."
"Supply centers should be de足
centralized and moved to north Bengal."
"Village cooperatives should
permit the membership of landless laborers." "Priority for aid to these cooperatives" should be determined by their ability to provide quality inputs by critical dates to small cul足 tivators."
"Storage facilities for agricultural inputs" should be
increased and improved.
Experimentation with such "mulches" as "pukur
sludge," "weeds" and "water hyacinth," should be tried in an effort to "lessen the vulnerable dependence on petroleum based fertilizer inputs."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
263 These suggested famine remedial policies,^ both national and regional, are neither mutually exclusive nor immutable.
Many still need
greater refining and further adjustment to local areas.
They merely
offer a broad gamut of alternatives:
possible options if famine can be
anticipated in certain areas of Bangladesh.
They offer substitutes for
the unilinear doles of food relief which have been the recent response to starvation.
2
However, they have to be implemented while the
degrees
of freedom" in terms of individuals' physical strength, and the com足 munities' social integrity still remain. The national policies suggested by the respondents emphasized the need for a "revolution of consciousness" and the regional policies sug足 gested by them persistently allude to the need for "community organiza足 tion" and "cooperation."
These both indirectly highlight the difficulty
in persuading an impoverished Bengali agriculturalist to individually risk taking time off from his own work for the good for the community, e.g., to build a small scale irrigation system, unless he is certain that all the other necessary members of the community will also take the risk and join him on the community project which would bring them all the
^The likelihood that these remedial measures will be included in famine prevention plans is high because the decision makers suggested them. They chose the components of famine liability, they weighted those components, and thus they suggested the areas liable to famine where such remedial measures should be implemented. 2 The famine remedial policies suggested by the respondents and out足 lined in this chapter bear a marked similarity to those suggested after the compound experience of the nineteenth and early twentieth century famine commissions. This may reflect the education and experience of the respondents or rather it may suggest the reinvention of the wheel after a dark age between colonial rule and full national sovereignty.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
264 benefits of development.
Perhaps relief aid, if properly channelled
to anticipated projects through an established village administrative structure like the newly suggested "gram sarkar" (village government) (Chaudhuri, 1978) can provide the needed stimulus to lessen that risk. The time of crisis can become the time for the poor to gain access to resources for a longer term.
If landless workers can organize themselves
they might be able to provide an irrigation canal for several landowners in return for the guarantee of a certain percentage profit from those lands over the next 25 years. If the remedial policies suggested in this chapter are to be successfully implemented amidst the famine crisis and if "relief is to be put into a development context" by the carrying out of these policies, it is essential that there must be some such mechanism for breaking out of the present community-level development dilemma.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VIII REFLECTIONS, SUMMARY AND CLAIRVOYANCE
Like Topsey, this dissertation "just growed" while I was in Bangladesh.
The proposition that it would be "possible to delimit areas
liable to famine" (Chapter I) and thus emphasize famine prevention and development rather than short term relief, evolved from personal experi足 ence in the 1974-75 Bangladesh famine.
The methodology to defend this
argument was also implemented ini vivo in the aftermath of the 1974-75 famine. Despite a warning prior to reaching Bangladesh that any research even tangential to the food situation and food allocation would be severely frowned upon by the government (Taylor, 1974), the implementation of the proposition survived a government coup (August 15, 1975) and a severe reprimand for "unauthorized investigation" by the then Secretary to the Ministry of Health (T. Hossain, 1975) for having carried out a famine survey under the auspices of the Ministry of Relief and Rehabili足 tation.
Even with three separately addressed letters each bearing the
embossed seal of approval of a different government ministry, this geographic and cartographic research on a "hot issue" like famine also had to sidestep the suspicions associated with using maps and aerial photographs, particularly in border areas (see Chapter IV). The need for time to adjust to these political pressures was
con足
tradicted by the continual requests for interim reports by the sponsors. These reports required in quantity,
primarily to satisfy theU.S.
based
auditors of the sponsors, but the content and quality of the reports if
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
266 released in Bangladesh would probably have affected the implementation of the remainder of the project.
As a foreign researcher I was under the
umbrella of an expatriate university medical research group, a group who customarily expect rapidly produced results on small scale human experi足 ments.
They expected a similar time span to complete a regional scale
project in applied geography. The third force in the trihedral situation affecting the proposition was the need to maintain "controls" on the argument.
The proposition was
continually balancing between the need for the battery of statistical significance tests required by traditional scientific research and the need for data to make decisions in a real life tragedy.
As an example, it
was seen in Chapter VI how the famine survey was delayed until after the famine by attempts to evolve a sophisticated sampling strategy and to standardize measurements with other nutrition studies (as well as by the need for government permission). After completing such an applied geographical study of a contemporary issue in the field, there is a tendency to feel somehow dissatisfied and thus to set about producing a list of desiderata to answer the cliche question "What you would do differently if you were to begin the research again"?
Such an ex post facto list, however, is largely irrelevant,
because to conduct an applied geographical study of this nature now, one would perforce begin in a completely different context from the earlier study. The most obvious desideratum would be time.
The Delbecq-Delphi
Process (Chapter V) derives much of its effectiveness from its simplicity, but this cannot be equated with speed.
Time is required to elect a broad
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
267 and representative initial target population who can bring a spectrum of perspectives to an often previously ill-defined problem.
Time is required
to sieve out the representative and available group, with the greatest experience, to join the Delbecq and Delphi sessions.
Much more time than
was allocated in this study is required in the editing stage (Figure 18) where an additional discussion session between the experts from the different fields would probably produce a more coherent list of factors and more meaningful criteria.
Such a meeting might also suggest data
sources that actually were available, rather than those that were thought to be available.
Time is also required to arrange to bring the groups
together at a time convenient to all members.
If there had been time it
might have been possible to display a map showing the actual areal extent of each of the 10 components before the Delphi panelists weighted each of them.
It might also be preferable to have an additional meeting once the
final Delbecq-Delphi map was prepared to include the responses of panel members after viewing the "gestalt" of the finished map which was compiled from the components which they had initially selected and weighted. Although such a multifaceted problem transcends disciplines and often government ministries, the time requirements could be reduced by institutionalizing an improved form of the Delbecq-Delphi process either within a supra-ministerial organization, like the planning commission, or, if the problem were more specific, within a ministry like the Ministry of Relief and Rehabilitation.
Such institutionalization of the technique
should be done by a senior government officer such as the secretary of a ministry moderating a group of lower ranked officials such as the section officers who were already under his jurisdiction.
A geographical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
268 component could be brought to this planning by the secretary of the ministry requesting the deputy commissioners for each district to come to Dacca for some modified form of Delbecq-Delphi session on a particular problem.
This would ensure a comprehensive areal coverage of different
regions. In the attempt to examine the regional components of famine liabil足 ity this study has changed scales with chameleonic ease, at times perhaps committing both ecological and contextual fallacies (Susser, 1973) and once almost certainly invalidating an evaluation technique (see Famine Survey section, Chapter VI).
The need, however, to involve a variety of
scales in comprehending famine vulnerability was introduced early in Chapter II, and both micro-scale case studies and macro-scale national and international contexts have been culled to bring a balance to the meso-scale approach by highlighting its limitations and potentials respectively.
It is tempting to crawl behind the rigor of the micro-scale
study to approach asympote of incontestable truth (see page 154), but already in Bangladesh there has been a deluge of fine-grained case studies, most of which have been foreign funded.
Some of these are
under increasing public attack recently (Chowdhury, 1976; Briscoe, 1978) for they have minimal effect on providing policy information for national programs.
All too often the results of a single micro-scale study are
equated with national norms for reports to Washington, Geneva, or London where they then serve international interests wishing to make international comparisons.
What the Bangladesh government needs, however, is information
to make internal comparisons between regions of Bangladesh.
At the
regional scale, as has been documented in this dissertation (Chapters IV
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
269 and V) , there is an aching void in the information base.
The Delbecq-
Delphi Process, as developed in this dissertation, as well as introducing focused judgment to this area, has also served to highlight many of these yet unmeasured attributes necessary for planning at the regional level (Chapter V, Table
6).
Despite the limited data available in this case, the proposition was upheld: Bangladesh."
it was possible "to delimit areas liable to famine in There are certain regions of Bangladesh, namely the riverine
and coastal areas and north Bengal, that are more susceptible to famine than others. The initial appearance of the map of areas liable to famine in Bangladesh may give a false impression of accuracy.
The broad regional
patterns differentiated by the red, yellow and white classification has been found to be valid (see Chapter VII).
However, the apparently
detailed thana by thana pattern of famine liability still eludes us after using this form of the Delbecq-Delphi Process (see Chapter V).
There
are still several limitations in the process which require further improvement and testing in real world situations: 1.
It is still not proven that the additional Delphi session after the Delbecq session in the Delbecq-Delphi Process really makes a substantial improvement in the level of scaling attained. The apparent interval scale is probably still an ordinal scale in the minds of the panelists.
2.
Delphi panelists were probably unclear about the implications of some of the concepts they were considering and weighting.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.
The Delphi panelists were probably unable to grasp the nationwide significance of some concepts, e.g., flooding.
4.
It is unlikely that the significance of any concept, however carefully defined, is constant throughout a nation even a small one like Bangladesh.
Thus lack of rain for a given period may
have a very different effect in Noakhali than in Rangpur. 5.
Similarly there is no certainty that the weighted concepts or factors can be added and linearity assumed throughout Bangladesh (see Chapter V).
It is even possible that in some
areas one factor might have to be crudely subtracted from another to give combined risk, e.g., flooding and drought.
6.
The Delbecq-Delphi Process still suffers from the lack of interaction both among the different groups and among the individuals of each group.
Groups and individuals were
deliberately isolated from each other in this process because of the researcher's lack of training in group discussion techniques.
With researchers having experience in handling
interactive groups much greater concept definition and more meaningful scaling might have been achieved. In addition to these limitations on the certainty of delimiting the areas liable to famine, four other limitations of the study should be re-emphasized: 1.
Not all the people, nor all the villagers in the thanas liable to famine are necessarily going to be affected by famine.
Those
with savings and storage, or with strong links outside the area are likely to remain resilient to famine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.
The probabilities described in this study are low.
They do not
suggest that areas which are not classed as "very liable" or "liable" to famine will not suffer from famine in the future. They may, but present conditions in those thanas make them less likely to suffer from famine. 3.
This study does not suggest a solution to the problem of famine.
Such a solution can only come from widespread changes
in social responsibility (see Myrdal, 1968 and Hall, 1976) both between rich nations and poor nations and between the rich and poor in Bangladesh, such as the respondents began to suggest in Chapter VII.
This study rather suggests a longer term
holding strategy than does the current emphasis on relief measures. 4.
The development of marginal famine prone areas initiated at times of famine is not a proven method of improving the quality of life of the poor in those areas.
Many others have cogently
argued the traditional efficiency model of industrial investment in the expanding areas such as the cities in Taiwan (Fuchs and Street, 1980) and which will in turn absorb the rural poor. Although such an urban biased (Lipton, 1977) strategy would undoubtedly help urban poor in Bangladesh the scale of the problem is so great (Alamgir, 1977) and the industrial skill and education levels so low, that I suggest that perhaps both strategies may be required.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
272 How long this geographic pattern of famine liability is likely to persist depends upon both physical and economic factors.
I foresee the
broad riverine and coastal pattern persisting unless there are major river changes due to tectonic activity.
The famine liability of north Bengal
is likely to increase if population migration to the area continues at a rate greater than the improvement of dependable irrigation.
Rapid popu足
lation in-migration to Sylhet and the Chittagong Hill Tracts might in足 crease their famine susceptibility.
Also in Sylhet population pressure
might become too great for the traditional storage mechansims to cope with.
The Hill Tracts, although presently sparsely populated, are also
susceptible to development.
The agriculturalists there have lost their
land on the valley floor because of the Kaptai Dam and thus agriculture is now encroaching on the steep forested hillsides which are prone to landslides and which traditionally were the storehouse for alternate foods in times of famine. over small areas.
The pattern of famine liability may improve
Famine liability might be reduced if projects were
implemented in certain thanas, either some of the remedial projects suggested in Chapter VII or other integrated projects which couple the provision of perennial employment and the lessening of risk to natural hazards. If Bangladesh is likely to suffer another famine within the next 30 years, if the past trends of political unrest, administrative break足 down, and successive natural disasters continue, what chance is there that the overall policy of regional targeting of famine prevention policies will be implemented?
Pessimistically, I would foresee the plan
being suddenly resuscitated on the eve of the next famine as administra足 tors frantically search for documentation related to famine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
273 Optimistically, however, there is a chance that parts, at least, of the plan may still be acted upon in the final eddy of the ebb of the 197475 famine.
Backed by the political "clout" of many of the decision
makers themselves, the plan still has the potential to be implemented, although it may be late to "ride to success on the ebb tide."
The
present military-backed government of Bangladesh under President Ziaur Rahman is advised by many of the respondents involved in the DelbecqDelphi Process.
General Zia, as he is popularly known, is himself from
North Bengal and has publicly declared his support for that neglected region.
Having a military background, and hence having been trained in
the use of maps for strategic planning, the general is likely to quickly grasp the potential of the Delbecq-Delphi map for policy planning.
The
present military support and relative internal stability make policing of the major border routes possible.
Bangladesh's relations with India,
and with other Asian States, apart from with Burma to the south-east,'*' are better than ever before should integrated river basin development be contemplated.
General Zia's political backers include the pro-Peking
National Awami Party and relations with Peking have been strengthened. Development ideas are being exchanged and there might be significant Chinese influence in the setting up of the village level councils for self-help. Foreign aid, primarily from western countries, continues to meet 54 percent of the national budget and 80 percent of the public development
"^Recently the influx of over 200,000 Rohingiyas (Muslim refugees from Burma) into Bangladesh has initiated intense discussions concerning their return (The Guardian Weekly, 1978).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ilk expenditures (Barung, 1978).
Politically the Bangladesh government is
unlikely to be unanimous about any suggested targetting of aid.
Far from
viewing the designation of certain areas as being famine liable as a stigma, some politicians are more likely to view it as an asset in terms of soliciting development aid.
Such politicians from areas not desig足
nated as liable to famine are thus unlikely to support the policy, simply because their areas will not be given priority for such aid.
Thus, the
deliberate intent on the part of the aid-giving countries to target development aid is perhaps necessary, however unethical, in order to push such a policy through a government that is still so dependent on external aid. Assuming that the plans and policies are partly or wholly accepted, how should a geographer continue the research on famine begun by this dissertation?
At present I see two principal directions in which further
research might proceed:
one would continue the theme of defining areas
liable to famine both in Bangladesh and elsewhere, and the other would proceed to a further stage by examining the early warning signals of famine in some of the areas now delimited as being liable to famine. Within Bangladesh the accuracy of the areas designated as being liable to famine could and should be improved down to the union level so that specific projects can be prepared in readiness for implementation. It might be possible to use improved historical areal information on past famines, e.g., the data on the symptoms of the 1896-97 famine in Bengal, which were documented at the union level, but which have previously been unavailable to researchers.
Alternatively, it has been suggested (Berry,
1976) that a nationwide survey of village elders, asking them to recall
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
275 past famines and scarcities, might permit further areal differentiation. The problem of regional and local dialectical differences, both in linguistic terms and in the perception of scarcity and famine, however, might lessen the value of such a study.
Rather it would seem more
appropriate to implement the Delbecq-Delphi Process in each of the 62 subdivisions using local officials, village elders and, if possible, landless laborers, as the initial groups of respondents. The application of the Delbecq-Delphi Process is limited neither to Bangladesh nor to famine problems.
It could, with improvement, be
developed as a general planning tool in developing countries where areal data is limited.
I myself would propose its use initially for futher
famine liability sutdies in India, Indonesia, Cambodia, Vietnam, Afghanistan, Ethiopia, Haiti and Kenya.
Not only would they form useful
policy studies, but they would also help to elicit the different com足 ponents of famine liability in these different countries.
The applica足
tion of the technique to the Sahelian countries, e.g., Mauritania, Senegal, Mali, Upper Volta, Niger and Chad, should be encouraged, but modifications will have to be made to accommodate the areal shifts in liability as a result of the large proportions of the population who are nomadic. This dissertation has intentionally avoided becoming involved in short term famine prediction.
In Chapter II it was stated that too little
was understood about the interrelationships among the components of famine vulnerability to predict when famine was going to occur.
However in
seeking to understand the structural components that make certain areas liable to famine, the dissertation has uncovered some initial findings that would encourage further research on short term famine warning systems,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
276 particularly in the areas found to be liable to famine in Bangladesh. Figure 11 in Chapter II, suggested the importance of seasonal cycles in weather, crops, hydrology, employment and nutrition.
Figure 26 in
Chapter V showed how a disruption in the meshing of the flood and crop cycles may have been the "trigger mechansism" for the 1974-75 famine. Figure 1
in Chapter I showed the demographic trends for Purundapur
village during the famine and suggested how the timing of the onset famine symptoms compared with those triggering mechanisms.
Chapter V
also suggested a series of other famine symptoms that could be monitored through time, e.g., land sales, sale of possessions, migration, nutritional levels and the eating of famine food stuffs.
By integrating
the longitudinal monitoring of these trends, seasonal cycles and dis足 ruptions, it might be possible to suggest a short term predictive model for famine.
Once again, however, the early detection of famine symptoms
will require the drawing of information from many scales:
from inter足
national reports of the grain trade and rainfall patterns, and from village level perceptions of change in the local social and economic conditions. Undoubtedly these will vary from one part of the country to another. All these possibilities offer potential research fields for the geographer.
Famine research is intrinsically geographic:
it purports
to examine a man-environment problem; one which is multifaceted and required synthesis to understand, and one which has an overt areal com足 ponent conducive to cartographic analysis. However, just as this research began by suggesting that sending a nutritionist to a famine was like sending a neurosurgeon to a polio epidemic, so it is now necessary to question the appropriateness of an applied geographer's perspective in policy oriented famine research.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
277 At the beginning of this summary chapter, I stressed how this particular study had "just growed" and how it had "growed" in the midst of a trihedral situation, with the stresses of politic behavior, of time constraints, and of the search for acceptable levels of significance.
In
the future, research on famine is likely to become increasingly political. It will always involve "hot issues" and demand quick results.
The temptaÂ
tion to bury oneâ&#x20AC;&#x2122;s head only in the pure sand of theoretical geography must be avoided.
Although this study lacks a high degree of specificity,
it has nevertheless achieved six original contributions to geography as a discipline. 1.
It has successfully brought together the knowledge of groups from disparate backgrounds to examine the problem of famine vulnerability in a geographical context.
2.
It has demonstrated for the first time the possibilities of "spatial targetting" of famine relief and preventive measures.
3.
It has developed and evaluated a geographical model for assessing famine vulnerability which has potential applicability at various scales and in various countries.
4. It has involved policy makers in this compilation stage of a policy planning map. 5. It has integrated remedial policies with an analytical map of famine vulnerability.
6.
Hopefully, it has also made a significant contribution to a change in philosophy from palliative relief aid to preventive and development-oriented aid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A M U R U B I S 1 QUESTIONNAIRE
Perception of Famine in Bhedargan.j Thana by Murrubbis Survey Code: Sample N o . : Name (Respondent):________________________________ Village Name:______________________
J.L. No.:
Bari Name:_________________________
Malaria No. :
Religion (Respondent): 1.
fj Islam
2. / / Hindu 3. / / Other Age (Respondent):
/__/ /
/ /__/
Perceived Income Level of Household: 1. 2.
fj Well off fj Middle
3. / / Poor Date of Interview:___ Name (Interviewer):__ Time Interview Begins Duration
Time Interview Ends:_ Estimate of Response:
of
Interview
mins.
/~7 Very Willing / / Normal / /
Reticent or
Unwilling
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
279 Scrutinized by:______________________________ Coded by:____________________________________ Verified by:_________________________________ Punched by:__________________________________ Verified by:_________________________________ A.
Do you remember within your lifetime? 1. / / Partition 2. 3.
fj 50 fj 26
Mananther Cyclone
4. / / None of the above
5. /~~7 Don't know B.
Were you born in this village? 1. 2.
fj No ÂŁ7 Yes
3. / / Don't know If N o , have you lived in a neighbouring village? 1.
/ / Since
before partition
2.
/ / Since
before the 50 Mananther
3.
/ / Since
before the 26 Cyclone
4 . / / None of the above
5. / / Don't know C.
What is an 'ovab'?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
280
D.
What
E.
At what
F.
is
the difference between
an
point does
become
an
'awarab'
'awarab'
a
and a
^ a h a t '?
'kahafc1?
How many 'awarab* do you recall in this village? 1.
/~7
One
2-
/~7
Two
3. /~7 Three 4. / / Four 5. / V Five
6. /
/ More than five
7. /~7 None
8. / G.
/ Don't know
In your lifetime when vas the worst 'awarab' or 'kahafc' in this village? Bangla Year
H.
fj fj
fj
English Year
fj fj
fj
Why do 'awarabs' visit your village?
_____________________________
1.
2
.________________________________________________
3.__________________________________________________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
281
4 .______________________________________
.________________________________
5
6.
____________________________________________
7_._____________________________
_8 ._________ 9 ._______________________________________________________________________
10 . _____________________________________________ I.
What
can
coming
the peop l e
of
to y o u r v i l l ag e
the village
do to p revent
the
'awarab'
again?
_____________________________
JU
._________________________________________________
2
3.___________________________________________________________
Âą._____________________________ 5 . _______________________
7 ._________________
8. _____________________________________________
9.___________________________________________________________ 10 J.
.
What
_____________________________ can
the government
do
to p r e v e n t
'awarab'
coming
again
to your village?
_._____________________________________
1 2
._____________________________
3_.___________________________________________________________________
4,.__________________________________________________________________
5_.................................... ....... ........... ......
_____________________________ 7 ._________________________________________________________________________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
282
9.__________________________________________________________ 10 ._____________________________________________________ K.
Do y o u k n o w â&#x20AC;&#x2122;a w a r a b 1? 1. f j
anywhere outside Faridpur which
is p r o n e
to
No
2.
f j
3.
/ / D o n 11 k n o w
Yes
Th a n k y o u for all the ans w er s to o u r q u e s t i o n s a b o u t awarab. N o w w e just hav e two final questions about your education and occupation.
L.
What
education
d i d yor. r e c e i v e ?
1.
/~7
2.
/
3.
/ / Primary
4.
/ / High
5.
/ / Matriculation
Cannot
/ Quran
only
School
6.
/ y
H.S.C.
7.
/~7
B .A.
M.
What
read or write
or Madrasa
is y o u r p r i m a r y
occupation?
1. /~7 Retired 2.
/ / Owner/Cultivator
3.
/~7 Tenant/Sharecropper
4.
/ / Businessman
5.
/ /
Imam/Head
of M ad r a s a
6. / ~ Teacher 7.
/ /
Service
8.
/ /
Other
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX B RETIRED
GOVERNMENT
OFFICERS
QUESTIONNAIRE
OPINIONS ON FAMINE BY RETIRED GOVERNMENT OFFICERS Time Begun Date:________
In connection with a project for the Ministry of Relief and Rehabilitation of the Bangladesh Government to "Map the Areas Liable to Famine In Bangladesh", your name has been suggested as a retired Government officer with considerable experience in different districts of Bangladesh, especially in connection with scarcities and famines. Would you be willing to initiate this project by answering a set of questions about your experience of famines and scarcities in Bangladesh. I'd like to start straightway to look at the differences between scarcities and famine by asking for your definition of both: A.
What is a scarcity?
B.
What is a famine?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
284 -2-
C.
At what point can an area in Bangladesh be declared to be a "famine area" as opposed to a "scarcity area"?
D.
Do you think that it is possible to predict when local famines will occur in Bangladesh
O N0 fj
Yes
fj
D O N â&#x20AC;&#x2122;T KNOW If YES, how?
Any qualifications or exceptions mentioned:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-3E.
Do you think there are certain factors which make some areas of Bangladesh persistently more lia,ble to localized famine where government gruel kitchens have to operate.
HI
NO
f~ l
YES
/
7
DON'T KNOW If Y E S , which factors?
1)________________________________ 2 )__________________________________________________________ 3 )________________________________________________________________; ________________
4 )_____________________________________________________________________________ 5>________________________________________________________________________________
6)__________________________________________________________ 7 )________________________________________________________________________________________ 8 )_______________ i________________________________________________________________ 9 )________________________________________________________________________________________
NOTE Register Map. On this map of Bangladesh, could you carefully outline with thin felt tip marker the rural areas persistently liable to famine where government gruel kitchens have to operate. NOTE Any areas about which you are uncertain ring with a dotted line.
Reproduced with permission of the copyright ow ner Further reproduction prohibited without permission.
286
F.
Are there any towns in Bangladesh that you think are persistently liable to famine where government gruel kitchens have to operate? (NOTE: Do not mark the towns on the map.) /
7
NO
/
7
YES
/
/
D O N â&#x20AC;&#x2122;T KNOW
If Y E S , which towns?
1 ) ________________________________________
2 )______________________________ 3)_________________________________________
AL________________ 5 )_________________________________________
6 )______________________________ 7)_____________________________________________ You outlines __________________ areas on the map. Could you rank them by putting a (1) in the area most persistently liable to famine and then a (2) in the area next most liable to famine and so on down to Number _____ . NOTE:
If any 2 areas are considered to rank equally, mark them clearly (1A) and (IB) and then rank the next one (3).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
287
G
Now for each of the areas y o u ’ve outlined, can you suggest reasons why those areas are persistently liable to famine: (If you want to group certain areas together, just tell me which numbers to put together.)
Area No(s)
___
11
Area No(s)
1)
—
Area No(s)
_______________________
_________________
12
II—
Area No(s)_____ —
___
----------------------
Area No(s)
1)__________
Area No( s )
12_________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
288
Do you think that there are any measures that the villagers themselves can take to prevent famine visiting their villages? /
/
NO
/
7
YES
/
7
DON'T KNOW
If Y E S ,for each of the areas you outlined, list possible measures: (If you wish to group certain areas together, just tell me which numbers are to be grouped.) Area No(s)
________________
1)
Area No(s)
Area No(s)_________ _______________
1)________________________
___
Area No(s)
1)________________________
I)________
Area No(s)
Area No(s)
1)
___
1 )________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
289
I.
Do you think that there are any measures that the government can take to prevent famines in those areas?
/
7
NO
/
7
YES
/
7
DON'T KNOW
If Y E S , list the measures for each area. (Again, if you wish to group certain areas together, just tell me which numbers are to be grouped.) Area No(s)
Area No(s)
1)____
1)____
Area No(s)
Area No(s)
1)____
1)____
Area No(s)
Area N o(s)
1)____
1)____
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
J.
In discussing those areas liable to famine, you have drawn on your experience as a former government officer. Could you now run through details of your career with me, noting any degrees you took after high school, the different government posts you held, and for how long, and whether you were ever involved in relief areas in any of the districts. (a)
Education
' ___________________________
(b)
Entered Service:
(c)
Government Service Record:
Year Beginning
District
Period of Service
Post
Involvement in Relief
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
291
K.
Which year did you retire from Government Service?
L.
During your career did you ever read through the Bengal Famine Code?
f~l
NO
LJ
M.
_________
yes
/
/DONâ&#x20AC;&#x2122;T KNOW
/
7
OTHER, SPECIFY:___________________________________________________
Which of Bangladesh's 1? Districts have you never visited? (Include "passing through" a;3 "a visit' 1.
LJ
Bakerganj
11.
LJ
Kushtia
2,
n
Bogra
12.
LJ
Mymensingh
3.
n
Chittagong
13.
LJ
Noakhali
4.
n
Chittagong M. T.
14.
LJ
Pabna
5.
n
Comilla
15.
LJ
Patuakhali
6.
n
Dacca
16.
LJ
Raj shahi
7.
LJ
Dinajpur
17.
LJ
Rangpur
8.
n
Faridpur
18.
n
Sylhet
9.
rj
Jessore
19.
n
Tangail
10.
LJ
Khulna
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
292 N.
Could you give the names and addresses of five other retired government officers, who have served in relief operations in rural areas of Bangladesh and who are now living in Dacca. NAME
ADDRESS
TELEPHONE
2.
3.
4.
0.
Finally, would you yourself be interested and available, if necessary, to participate at a seminar and dinner to explore this problem further on __________ at 7:00 p.m.
LJ
N0
LJ
^es
/
DON'T KNOW
7
THANK YOU FOR ALL YOUR HELP!
TIME: ___ DURATION: DATE: ___
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX C AID DONORS' QUESTIONNAIRE OPINIONS ON FAMINE BY AID DONORS ___________________
Agency Time Begun Date:_______
In connection with a project for the Ministry of Relief and Rehabilitation of the Bangladesh Government to "Map the Areas Liable to Famine in Bangladesh", your name has been suggested as an Aid Donor with expertise i n ___________________________ . Would you be willing to initiate this project by answering a set of questions about your experience of famines and scarcities in Bangladesh? I'd like to start straightway to look famines by asking for your definition A.
What is a scarcity?
B.
What is a famine?
at the differences between scarcities and of both:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C.
Ac what point can an area in Bangladesh be declared to he a "famine area" as opposed to a "scarcity area"?
D.
Do you think that it is possible to predict when local famines wil l occur in Bangladesh
O N0 LJ
yes
[J
DON'T KNOW If YES, how?
Any qualifications or exceptions mentioned:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
E.
Do you think there are certain factors which make some areas of Bangladesh persistently more liable to localized famine where government gruel kitchens have to operate.
n n /
no YES
7
DON'T KNOW If YES, which factors?
1 )________________________________________ : __________________________________
2)_________________________________ : : : ________ i)____________: _________________ A)___________________________________________________________ 5)___________________________________
1
*_
:
___
7 )___________________________________________________
8)________________________________________ 9)_________________________________________________________________________________
NOTE Register Map. On this map of Bangladesh, could you carefully outline with thin felt tip marker the rural areas persistently liable to famine where government gruel kitchens have to operate. NOTE Any areas about which you are uncertain ring with a dotted line.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
296 F.
Are there any towns in Bangladesh that you think are persistently liable to famine where government gruel kitchens have to operate? (NOTE: Do not mark the towns on the map.)
LJ
N0
/
7 YES
/
7 DON'T
KNOW
If Y E S , which towns?
1)___________________________ ;_______________
2)
; ___________________
3 )___________________ __________________ 4 )_____________
_
______________________
5 )__________________________________________
6)____________________________ 7)__ _______________________________ You outlines areas on the map. Could you rank them by -putting a (1) in the area most persistently liable to famine and then a (2) in the area next most liable to famine and so on down to Number ______. NOTE:
If any 2 areas are considered to rank equally, (IB) and then rank the next one (3).
mark them clearly
(1A) and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G.
Now for each of the areas y o u â&#x20AC;&#x2122;ve outlined, can you suggest reasons why those areas are persistently liable to famine. (If you want to group certain areas together, just tell me which numbers to put together.).
Area No(s)
___
12
Area No(s)
Area No(s) â&#x2013;
---
12________ ________________ _
Area No(s)_
___
___
1}__________________________________
Area No(s)
i)________
Area No(s)
12----- --- ---------------- l)_______________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Do you think that there are any measures that the villagers themselves can take to prevent famine visiting their villages? /
7
NO
/
7
YES
/
7
D O N â&#x20AC;&#x2122;T KNOW
If YES,for each of the areas you outlined, list possible measures: (If you wish to group certain areas together, just tell me which numbers are to be grouped.) Area No(s)
________________
1)
Area No(s)
___
Area No(s)
1)________________________
Area No(s)
1)
Area No(s)_________ _______________
1)_________________________
II_______
________________
;
Area No(s)
1)________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Do you think that there are any measures that the government can take to prevent famines in those areas? /
7
NO
/
7
YES
/
/
DON'T KNOW
If Y E S , list the measures for each area. (Again, if you wish to group certain areas together, just tell me which numbers are to be grouped.) Area No(s)
___
Area N o(s)
1)_______________: ___________
Area No(s)
___
â&#x20AC;&#x2DC;
.
___
n____________________________
Area No(s)
1)
1)_______
Area No(s) ___
Area N o(s)
1)
1)______
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
J.
Of those
In which of the areas are you still supporting or have you supported programmes?
In which of the areas have you had, or do you still have, your programmes running?
In which of the areas have you yourself worked for over a week at any one time?
areas outlined:
□ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □
Which of the areas have you
□ □ □ □ □ □ □ □ In which of the areas do you know that government gruel kitchens have operated?
□□□□□□□□
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Include 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
LJ n LJ n LJ LJ LJ n LJ LJ
"passing through" as a
visit")
Bakerganj
11
Bogra
12
LJ n n LJ LJ LJ LJ LJ LJ
Chittagong
13
Chittagong M. T.
14
Comilla
15
Dacca
16
Dinajpur
17
Faridpur
18
Jessore
19
Kushtia Mymensingh Noakhali Pabna Patuakhali Raj shahi Rangpur Sylhet Tangail
Khulna
Have you yourself worked : i rural Bangladesh outside of a district town for over 3 months?
J
N0
7
yes
7 ’ DON'T
KNOW
How long have you been working Bangladesh? here several times.) _/
Less
than 1 year
_/
1 - 2
years
_/
2 - 5
years
_/
5 - 20 years
/
/
(In total, if y o u ’ve been
Greater than 20 years
Have you ever read the Bengal Famine Code?
17 N0 17 YES J
DO N ’T KNOW
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0-
Can you give me a rough description of what your job here entails:
P.
What was your training and background before you began this job:
Q.
Do you think you'd be willing and available to participate at a seminar and dinner to explore this problem further on ____________________________
LJ
NO
/
7
YES
/
7
D O N â&#x20AC;&#x2122;T KNOW
THANK YOU FOR ALL YOUR H E L P ! Time: _________________ Duration of Interview: D a t e :___________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX D TECHNICAL EXPERTS' QUESTIONNAIRE
OPINIONS ON FAMINE BY TECHNICAL EXPERTS N a m e :_________________________________
A gency:_____________________
Nationality:_________________________
Time Begun:_________________ Date:_______________________
In connection with a project for the Ministry of Relief and Rehabilitation of the Bangladesh Government to "Map the Areas Liable to Famine in Bangladesh", your name has been suggested as a technical expert with expertise in ___________ ______________________________ . Would you be willing to initiate this project by answering a set of questi'' *.s about your experience of famines and scarcities in Bangladesh. I â&#x20AC;&#x2122;d like to start straightway to look at the differences between scarcities and famines by asking for your definition of both: A.
What is a scarcity?
B.
What is a famine?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C.
At what point can an area in Bangladesh b e declared to be a "famine area" as opposed to a "scarcity area"?
Do you think that it is possible to predict when local famines w i l l occur in Bangladesh
O
N0
LJ
yes
/~
DON'T K NOW If YES, how?
Any qualifications or exceptions mentioned:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
305
E.
Do you think there are certain factors which make some areas of Bangladesh persistently more liable to localized famine where government gruel kitchens have to operate.
n n I
ho YES 1
D O N â&#x20AC;&#x2122;T KNOW
â&#x20AC;˘
If Y E S , which factors?'
1)__________________________________ : ______________________________ 2)_____________________________________________ 3)
_ _______
_ _ ___________________________________
' 4)_________________________________________________________________
y
5 ________ 6)
:
:
. .
7)________________
8)_________________________________________________________________ 9)_________________________________________________________________ NOTE Register Map. On this map of Bangladesh, could you carefully outline with thin felt tip marker the rural areas persistently liable to famine where government gruel kitchens have to operate. NOTE Any areas about which you are uncertain ring with a dotted line.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
306 -4-
F.
Are there any towns in Bangladesh that you think are persistently liable to famine where government gruel kitchens have to operate? (NOTE: Do not mark the towns on the map.) L J
NO
/
7
YES
/
7
D O N â&#x20AC;&#x2122;T
KNOW
If YES, which towns?
1
)
______________________
2]
; ________
3 )_______________________ ;______________________
4)_________
-___________________
1)__________________________________ 6}__________________________________
2JL___________________________ You outlines areas on the map. Could you rank them by putting a (1) in the area most persistently liable to famine and then a (2) in the area next most liable to famine and so on down to Number _____ . N OTE:
If any 2 areas areconsidered torank (IB) and then rank the next one (3).
equally, mark them clearly
(1A) and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G.
Now for each of the areas you've outlined, can you suggest reasons w hy those areas are persistently liable to famine. (If you want to group certain areas together, just tell me which numbers to put together.)
Area N o ( s )
1)___ 2)______ 3)_______ A)_________ 5 )_________
6)______ Area N o(s)
1)_______
Area N o(s)
11_____
21_ __ D___ ___ A)___ 6)______ Area No(s)
11_____
2)
2)_______
3)_______
3)_______
A)_________ 5 )_________
6)______ Area No ( s )
A)_________
n___ 6)______ Area No ( s )
1)___
11_____
2)______
2)_______
3)_______ A)_________ 5)_________
3)_______ A) '
5)_______
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
308
-6 -
H.
Do you think that there are any measures that the villagers themselves can take to prevent famine visiting their villages?
LJ
NO
/
7
YES
/
7
DON'T KNOW
If YES,for each of the areas you outlined, list possible measures: (If you wish to group certain areas together, just tell me which numbers are to be grouped.) Area N o( s)
________________
1)___________________
Area No(s)
___
Area No(s)_________ _______________
n______________
Area No(s)
1)
11______
Area No(s)
Area No(s)
1)
1)______
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
309
Do you think that there are any measures that the government can take to prevent famines in those areas?
LJ
LJ
N0
YES
LJ
D0N'T k*50â&#x20AC;?
If Y E S , list the measures for each area. (Again, if you wish to group certain areas together, just tell me which numbers are to be grouped.) Area No(s)
___
1)
Area No(s)
Area No(s)
___
1)_____________________
________________
Area No(s)_________ _______________
1)
11__________
Area No(s) ___
Area No(s)
1)
1)_______
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
310
J.
Of those ____________ areas outlined:
In which of the areas are you still connected or have you been connected with aid or research programmes?
In which of the areas have you yourself worked for over a week at any one
Which of the areas have you visited? (Include "passing through" as a "visit")
In which of the areas do you know that government gruel kitchens have operated?
□ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □ □
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
311
(Include 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
L.
M.
/ 7 / 7 / 7 / 7 rr / 7 /~7 / 7 / 7 rr
"passing through" as a " visitâ&#x20AC;? ) Bakerganj
11.
Bogra.
12.
Chittagong
13.
Chittagong M. T.
14.
Faridpur
18.
7 /~7 / 7 /7 rr / 7 / 7
Jessore
19.
/~7
Comilla
15.
Dacca
16.
Dinajpur
17.
/
Kushtia Mymensingh Noakhali Pabna Patuakhali Raj shahi Rangpur Sylhet Tangail
Khulna
Have you yourself worked in rural Bangladesh outside of a district town for over 3 months? /
7
NO
/
7
YES
/
7 ' DON'T KNOW
How long have you been working Bangladesh? here several times.) /___ / Less
than 1 year
/___ / 1 - 2
years
/
years
72 - 5
_/__ /
5-20
(In total, if you've been
years
_/___ / Greater than N.
n
20 years
Have you ever read the Bengal Famine Code?
/ 7
NO
/ 7
YES
/
7 DON'T
KNOW
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
312
0.
Can you give me a rough description of what your job here entails:
P.
What was your training and background before you began this job:
Q.
Do you think you'd be willing and available to participate at a seminar and dinner to explore this problem further on ___________________________
LJ /
7
/ 7
NO YES DON'T KNOW
THANK YOU FOR ALL YOUR HELP! Time: _________________ Duration of Interview: Dat e :__________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX E.
EXAMPLE OF DELPHI SESSION CARD
IN CONFIDENCE Number After writing the score for any factor, turn over the card Âťnd pass it towards your right to the moderator.
FACTOR Example: The population of Dacca, 1974
1. 2. 3. 4.
5. 6. 7. 8. 9. 10.
Round I AVERAGE Roundf 2 AVERAGE
APPENDIX F DELBECQ-DELPHI SESSION INVITATION T o help wit h the r e s e a r c h project
M A P P I N G A R EAS L I AB L E TO F A M I N E IN BA N G L A D E S H
yo u are
invited to attend
A S E M I N A R &â&#x2013; D I N N E R for Aid G i v e r s
On:
Monday 3 0t h August 1976 at 7:4 5 p.m.
At;
IIouse N o * 106 Road N o * 12 Gu l s h a n
N.B. - T r a n s p o r t w i l l be provided, which w i l l coll e ct you a f ter 7:15 p.m. - C o n f i r m a t i o n w i l l be m a d e n e arer t h e tim e that yo u ar e a v a i lable. Programme
(See Over)
M A P P I N G AR E A S L IABLE TO F A M I N E IN BA N G L A D E S H N a sta Shortbread C old Dr i n k s Tea S e l e c t i o n of the five mos t
impo r t a n t factors
Palau F r ied Rui Maa s C h i n g r i Maas & Coco n u t Roast C h i c k e n Mutton Curry Salad Pudd ing Doi W e i g h t i n g of t h e r e l a t i v e
importance of th e d i f f erent facto: Te a Coffee Col d Dri n k s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
i I! I
i i
i
ll APPENDIX
G.
ii
f||j ili1
ill ill Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner
fage 2
2. Since th e b e g in n in g o f la st Ramzan, have any ch ild ren been b o rn alive in y o u r household? 2 = Y es 3 = D on't know I f Y es, how many?
^ ^ Total Male
O /
Female
/â&#x20AC;&#x201D; /
7
3. Since th e b eg in n in g of la s t Ramzan, have an y members of y o u r household le ft th e houce n e v e r to re tu rn ?
If Y es, fill in th e following ta b le:
SI #
Pop u lar Name
Sex 1 = Male 2 = Female
Age (Y ears) (Now)
Relationship
to
Household Head
Date of Move 1 = 1st Rnmzan-Eid U1 Azha 2 = Bid U1 A zha-B engali New Y ear 3 = B engali New Y ea r-P re sen t
O ccupation B efore
Sequel 1 = Contact 2 = No Contact 3 = Died 4 = D on't Know
1. 2. 3. 4. 5. 4.
Since th e b e g in n in g of la st Ramzan, hav e an y members of y o u r family died at home? 1 = No 2 = Yes 3 = D on't know If Y es, fill in th e following tab le on Page 3.
316
Further reproduction prohibited without permission.
Marital S tatu s on D ep artu re 1 = Single 2 = M arried 3 = Widowed 4 = Divorced 5 = S ep arated
Page 3
Cl Cl Q
I II
ii
IS s 1
I
I|5
tl! Ii t
iii!
Sfi's
m
s
lip
!
- 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Possessions
# Sold Since 1st of Last Ramzan
# Possessed a t P re sen t
Reason fo r Sale 1 = To buy food 2 = O th er - Specify
H Possessed a t 1st Ramzan
B uffalo Cow Bull Goat Plough C .I . Roof C art Boat O th e r p o ssessio n s sold to b u y food: 1. / ~7 K itchen u te n sils 2. /__/F u r n i tu r e 3. i / O rnam ents (g o ld /silv er)
8.
4. /
7 Wood
5. /
7 O the rs - Specify: ______________________________
H as any member o f y o u r household e v e r b een on te s t relief sin ce th e beg in n ing of la st Ramzan? 1 = No 2 = Yes 3 = D on't know
rj
If Y es, when? ,
1 2 3 4
= Betw een th e b eginn in g of la st Ramzan and B engali New Y ear = Betw een Bengali New Year and th e P re sen t = Both = D on't know
rj
318
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Page 4 7. D etails of household p o sse ssio n s:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9.
Hns your household ever received food relief since the beginning of last Ramzan?
10. Has any member of your household attended a gruel kitchen for food since the beginning of last Ramzan? 1 = No 2 = Yes 3 = Don't know 11. How many times p e r day have your household usually been having a rice meal since Bengali New Year? 1 = Three or more times 2 = Two times 3 = One time 4 = Never 5 = Other - S p e c ify :_______________________________________________________ 6 = Don't know 12. What was th e usual food for your household members: A. Between beginning of last Ramzan and Bengali New Year? _
B . Between Bengali New Year and the Present?
BIBLIOGRAPHY
Aall, C. (I960). Relief, Nutrition and Health in the Nigerian/Biafran War. Journal of Tropical Pediatrics, Vol. 16, Monograph 9. Abed, F. (1976). Personal Communication from the Director of the Bangladesh Rural Advancement Committee who implemented a rehabilitation cum development program in Rowmari Thana in 1974-75. Abedin, Z. (1943). Personal Communication from the family of the late Zainul Abedin in the midst of the 1943 Bengal famine. Abdullah, A., Hossain, M. and Nations, R. (1974). SIDA/ILO Report on Integrated Rural Development Programme, IRDP, Dacca, pp. 21-23. Adelman, I. and Morris, C. J. (1971). Analysis of Variance Techniques for the Study of Economic Development. Journal of Development , No. 3, pp. 1-20. Studies, Vol.
8
Adrian, S. (1976). Land, Power and Violence in Barisal Villages. Working Paper No. , The Village Study Group, Dacca University, Dacca.
6
Agarwal, S. N., Rohatgi, K., Bowonda, B. and Rohatgi, P. K. (1976). The Food Problem in India: a Delphi Study of Possible Solutions. Food Policy, February, pp. 170-173. Ahmed, N. (1968). An Economic Geography of East Pakistan. University Press, London.
Oxford
Ahsan, A.K.M. (1976). Disaster Preparedness Plan for Bangladesh, Planning Commission, Government of Peoples Republic of Bangladesh, Chairman's Introduction, May 4. Akbar, A. (1975). 1974 Famine in Bangladesh, Monograph (unpublished). Department of Social Welfare, Rajshahi University, Rajshahi. Alamgir, M. (1974). Some Analysis of Distribution of Income, Con足 sumption, Saving and Poverty in Bangladesh. The Bangladesh Economic Review, Vol. 2, No. 4. Alamgir, Mohiuddin (1977). Bangladesh: Acme of Below Poverty Level Equilibrium Trap. Bangladesh Institute of Development Studies, Dacca. Alvarez, F. (1881). Narrative of the Portuguese Embassy to Abyssinia. London, pp. 66-67.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
321 Anonymous (1973). Way of Life Shifting for Tauregs. Friday, September 28, page 10, column 3.
New York Times,
Armstrong, R. W. (1965). Medical Geography: An Emerging Specialty. International Pathology, Vol. , pp. 61-63.
6
Arnhold, R. (1969). The "Quae" Stick: A Field Measure Used by the Quaker Service Team in Nigeria. Journal of Tropical Pediatrics, Vol. 15, pp. 243-247. Aykroyd, W. R. (1975). The Conquest of Famine. E. P. Dutton & Co., Ltd., New York.
Readers Digest Press,
Aziz, K.M.A., Chowdhury, A.K.M.A., and Mosley, W. H. (1970). Patterns of Marriage: a Study in Rural Pakistan, Sixth Annual Conference of Pakistan Sociological Association, Lahore, July 11-1 2 .
Aziz, K.M.A. (1976). Laboratory.
Personal Communication.
Cholera Research
Baldwin, C. S. (1977). Catastrophe in Bangladesh: an Examination of Alternative Population Growth Possibilities 1975-2000. Asian Survey, Vol. 17, No. 4, pp. 345-357. Ball, N. (1976). Understanding the Causes of African Famine. The Journal of Modern African Studies, Vol. 14, Issue 3, No. 3, September. Bangladesh Census (1976). Union Population Statistics, 1974, Bulletin 3, Bangladesh Bureau of Statistics, Dacca. Barang, M. (1978). Latest Theories Tested Here. Review, May 19, pp. 34-36.
Far Eastern Economic
Baster, N. (1972). Development Indicators: an Introduction. of Development Studies, Vol. , No. 3, pp. 1-20.
8
Journal
Beaujeu-Garnier, J. (1975). The Operative Role of Geographers. International Social Science Journal, Vol. XXVII, No. 2, pp. 275-287. Batelle Memorial Institute (1971). Design of an Environmental Evaluation System, Report No. PB 201 743, National Technical Information Service, Springfield, Virginia. Beckett, J. C. (1966). The Making of Modern Ireland, 1605-1925. Faber and Faber, London. Bedford, M. T. (1972). The Future of Communication Services in the Home. Bell Canada Business Planning. Bell Canada, Ottawa.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
322 Beidelman, T. 0. (1959). A Comparative Analysis of the Ja.jroani System. Locust Valley, New York, Monograph VIII, Association for Asian Studies. Belete, S., Gebre-Medâ&#x20AC;&#x2122; n in, M., Hailemariam, B., Maffi, M. , Vahlquist, B., and Wolde-Gebriel, Z. (1976). Famine in Ethiopia: Study of Shelter Population in the Wollo Region. Courier of the Inter national Children's Center, Vol. XXVI, pp. 113-120. Berg, A. (1971). Famine Contained: Notes and Lessons from the Bihar Experience. Famine: Nutrition and Relief Operations in Times of Disaster, Blix, G., Hofvander, Y. and Vahlquist, B., Eds., Almquist and Wiksell, Uppsala, pp. 113-129. Berg, A. (1973). The Nutrition Factor. The Brookings Institution, Washington, Appendix A, pp. 211-221. Berg, A. (1975). June 15.
The Trouble with Triage.
The New York Times Magazine,
Berry, B.J.L., Ed. (1971). Comparative Factorial Ecology. Geography, Vol. 47, No. 2, Supplement, pp. 209-367.
Economic
Berry, L. (1978). Environment: A Critical Aspect of Development and Food Production in Bangladesh. Ecology of Food and Nutrition, Vol. 7, No. 2, pp. 73-86. Bhandavhar, D. R. (1931). Mauryan Brahmi Inscription of Marasthan. Epigraphia Indica, Vol. XXI, No. 14, pp. 83-91. Bhatia, B. M. (1967). Famines in India 1860-1965: A Study in Some Aspects of the Economic History of India, 2nd Edition, Asia Publishing House, Bombay and New York. Bhattacharjee. J. P. (1961). Underemployment Among Indian Farmers: An Analysis of its Nature and Extent Based on Data for Bihar. Artha Vijnana, Vol. 3, No. 3, pp. 246-278. Bidyaratna, H. P. (1971).
Khonar Bauchon.
Raustan Book Depot, Dacca.
Blix, G., Hofvander, Y . , and Vahlquist, B. (1971). Famine: A Symposium Dealing with Nutrition and Relief Operations in Times of Disaster. The Swedish Nutrition Foundation, Almquist and Wiksell, Stockholm. Bradford Disaster Research Unit (1976). Some Definitions. Disaster Research 3 . Disaster Research Unit, University of Bradford, Bradford, p. 3. Brammer, H. (1975). Disaster Preparedness Planning, Precautionary and Rehabilitation Measures for Agriculture. Soil Survey Interpretation Division, Department of Soil Survey, Dacca, p. 15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
323 Brammer, H. (1976). Personal Communication During a Meeting on Food Aid by the Voluntary Agencies at ADAB Headquarters in Dacca. Brody, Hugh (1973).
Inishkillane.
Penguin, London.
Brooke, C. (1967). Types of Food Shortages in Tanzania. Review, Vol. LVII, No. 3.
Geographical
Brooks, R. H. (1971). Human Response to Recurrent Drought in North足 eastern Brazil. Professional Geographer, Vol. XXIII, No. 1, pp. 40-44. Brown, B., Cochrane, S., and Dalkey, N. (1969). RM 5957, PR, Rand, Santa Monica.
The Delphi Method.
Buchanan, R. (1970). The Transformation of the Chinese Earth. Praeger, New York. Buck, J. L. (1937). University. Buck, P. (1937).
Land Utilization in China, Vol. 1 - Text, Nanking
All Men are Brothers.
New York.
Burley, L. A. (1973). Disaster Relief Administration in the Third World. International Development Review, Vol. 1, pp. 8-12. Burton, I., Ed. (1973). Expert Panel on Project 13: Perception of Environmental Quality. M.A.B. Report Series No. 9, Unesco, Paris. Burton, I., Kares, R. W. , and White, G. F. (1977). The Environment as Hazard. Oxford University Press, New York (in press). Burton, I. and Moon, K. (1971). The Perception of the Hazardousness of a Place, A Comparative Study of Five Natural Hazards in London, Ontario. I.G.U. Man and Environment, Unesco Seminar, August 4-9Caudollo. Cahill, G. F. (1978). Physiology of Acute Starvation in Man. of Food and Nutrition, Vol. , No. 4, pp. 221-230.
6
Ecology
Caldwell, J. C. (1975). The Sahelian Drought and Its Demographic Implications, Paper No. , Overseas Liaison Committee, American Council on Education, December.
8
Calkins, Philip B. (1976). Stability and Change in Landholding and Revenue Systems in Bengal, Chapter 2 in Rural Development in Bangladesh and Pakistan. East-West Center Book, Honolulu, pp. 9-28. Cant, R. G. (1976). Territorial Indicators and National Planning. The Use of Socio-Economic Indicators in Development Planning. Unesco Press, Paris.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
324 Carlson, Gary P. (1972). A Discrete Model for Inventory Control and Disaster Planning. IEEETSMC 2:1 (January), pp. 16-23. Center for Disease Control (1974). Nutritional Surveillance in West Africa, 1973. Disaster in the Desert, Hal Sheets and Roger Morris, Special Report, Humanitarian Policy Studies, The Carnegie Endowment for International Peace, Washington, pp. 76-123. Central Intelligence Agency (1975). Potential Implications of Trends in World Population, Pood Production and Climate. Washington. Chassingneux, E. (1923). La lutte contre la famine en Annam. Francaise Mars, pp. 70-72.
L â&#x20AC;&#x2122;Asie
Chen, L. and Chandhury, R. H. (1975). Demographic Change and Food Production in Bangladesh, 1960-74. Population and Development Review, Vol. 1, No. 2, pp. 201-228. China Continuation Committee (1922). China. Shanghai.
The Christian Occupation of
Chinese Government Bureau of Economic Information (1972). Famine Relief in 1926. Chinese Economic Journal, Vol. 1, No. , pp. 762-768.
8
Chinese Government Bureau of Economic Information (1928). Famines in Shantung. Chinese Economic Journal, Vol. II, No. 1, pp. 36-43. Chowdhury, A.K.M.A. and Curlin, G. T. (1976). Recent Trends in Fertility and Mortality in Rural Bangladesh, Preliminary Paper. Cholera Research Laboratory, Dacca. Chowdhury, A., Huffman, S. L., Curlin, G. T. (1976). Malnutrition Menarche and Marriage in Rural Bangladesh. Cholera Research Laboratory, Dacca. Chu, Co Ching (1926). Climatic Pulsations During Historic Times in China. Geographical Review, Vol. 16, pp. 274-79. Chu, Co Ching (1931). Climatic Changes During Historic Times in China. Journal of the North China Branch of the Royal Asiatic Society, Vol. 62, pp. 32-40. CIRTPC (1969). Famine Relief in Bihar. Central Institute of Research and Training in Public Cooperation, New Delhi. Cliffe, L. (1974). Capitalism or Feudalism? the Famine in Ethiopia. Review of African Political Economy, No. 1, p. 34. Coleman, J. M. (1969). The Brahmaputra River: Channel Processes and Sedimentation. Sedimentary Geology, Vol. 3, pp. 129-.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
325 Conrad, V. (1941). The Variability of Precipitation. Review, January, pp. 5-10.
Monthly Weather
Connell, K. H. (1950). Land and Population in Ireland, 1780-1875. Economic History Review, Second Series, Vol. 2, p. 278. Connell, K. H. (1955-56). Marriage in Ireland after the Famine: the Diffusion of the Match. Journal of the Statistical and Social Inquiry Society of Ireland XIX, pp. 82-103. Coppock, J. T. and Sewell, W.R.D. (1976). Policy. Pergamon Press, New York.
Spatial Dimensions of Public
Cousens, S. E. (1960a). Regional Death Rates in Ireland During the Great Famine, from 1846 to 1851. Population Studies, Vol. XIV, No. 1, July, pp. 55-74. Cousens, S. H. (1960b). The Regional Pattern of Emigration During the Great Irish Famine, 1846-1851. Transactions and Papers of the Institute of British Geographers, No. 28, pp. 119-134. Curlin, G. T., Chen, L. C. and Hussein, S. B. (1976). Demographic Crisis: The Impact of the Bangladesh Civil War (1971) on Births and Deaths in a Rural Area of Bangladesh. Population Studies, Vol. 30, No. 1, pp. 87-106. Currey, B., Rockie, J., Ng, S. K., Scudder, R. , Marsh, D. and Sulkey, C. (1973). Hazards in Hilo. Unpublished Monograph, Department of Geography, University of Hawaii, Manoa. Curtis, E. (1964).
A History of Ireland.
Dalby, D. and Harrison Church, Eds. (1973). London.
Methuen, London,
6th
Edition.
Drought in Africa.
Dalkey, N. R. (1969). The Delphi Method: An Experimental Study of Group Opinion. RAND, Santa Monica, R.M. - 5888 - PR. (June). Dalkey, H. and Brown, B. (1971). Comparison of Group Judgement Techniques with Short Range Predictions and Almanac Questions, Report. 678 - ARPA, Advanced Research Projects Agents, RAND, Santa Monica. Dalkey, N., and Helmer, 0. (1963). An Experimental Application of the Delphi Method to the Use of Experts. Management Science, pp. 458467. Dando, W. A. (1976). Man Made Famines: Some Geographical Insights from an Exploratory Study of a Millennium of Russian Famines. Ecology of Food & Nutrition, Vol. 4, pp. 29-234.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
326 Dando, W. A. (1976). Six Millennium of Famine: Proceedings of the AAG, Vol. , pp. 29-37.
8
Das, P. K. (1972).
The Monsoons.
Map & Model.
Edward Arnold, London.
Das, Gupta, A. and Roy, S. G. (1976). Population Estimates for Bangla足 desh: The Use of a Specific Transitional Population Model. Population Studies, Vol. 30, No. 1. Davis, I. (1977). pp. 21-23.
The Intervenors.
New Internationalist, No. 53,
Davis, L. E. (1971). Epidemiology of Famine in the Nigerian Crisis: Rapid Estimation of Malnutrition by Height and Arm Circumference in Large Populations. American Journal of Clinical Nutrition, Vol. 24, pp. 358-364. Dawson, J. A. & Doornkamp, J. C., Eds. (1973). Evaluating the Human Environment: Essays in Applied Geography. Edward Arnold, London. Delbecq, A. L. and Van de Ven, A. H. (1971). A Group Process Model for Problem Identification and Program Planning. Journal of Applied Behavioral Science, Vol. 7, No. 4. Delbecq, A. L. and Van de Ven, A. H. (1972). The Nominal Group as a Research Instrument for Exploratory Health Studies. American Journal of Public Health, pp. 337-342. Delbecq, A. L. , Van de Ven, A. H. and Gustafson, D. H. Group Techniques for Program Planning. Scott, Foresman and Company, Glenview, Illinois. Dikshit, K. H. (1931). The Inscription of Mahasthan. Vol. XXI, No. 14, p. 87.
Epigraphia Indica,
Dodge, C. P. (1976). Personal Communication from the then Director of Oxfam who Financed Part of a Feeding Program in Rowmari Thana. Dodge, C. P. and Wiebe, P. D. (1976). Famine Relief and Development in Bangladesh. Economic and Political Weekly, Vol. XI, No. 22, pp. 809-817. Dorfmen, R. , Alamgir, M. and Tabors, R. (1972). A Framework for Long Term Economic Planning in Bangladesh. Working Paper, Center for Population Studies, Harvard University, Cambridge, Massachusetts. Dutt, A. K. and Ahmed, H. (1974). Population Pressures Focus, Vol. XXV, No. 3, pp. 1-10.
in Bangladesh.
Durfee, R. C. (1972). The Use of Factor and Cluster Analysis in Regional Modeling, ORNL-TM 3720, Oak Ridge, Tennessee, Oak Ridge National Laboratory.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
327 Dworkin, J. (1974). Global Trends in Natural Disasters, 1947-1973. Natural Hazards Research Working Paper, No. 26, University of Colorado. Eckholm, Eric P. (1976). Losing Ground: Environmental Stress and World Food Prospects. W. W. Norton & Co., Inc., New York. Edwards, R. D. and Williams, T. D., Eds. (1956). The Great Famine: Studies in Irish History, 1945-52. Brown and Nolem, Ltd., Dublin. Elahi, K. M. (1976). Spatial Analysis of Population Variations in Bangladesh, 1951-1961, Paper Read at Bangladesh National Geographical Association, Rajshahi University. Elkinson, Charles M. (1976). East Pakistanâ&#x20AC;&#x2122;s Agricultural Planning and Development, 1955-1969: Its Legacy for Bangladesh, Chapter 4, in Rural Development in Bangladesh, E. Pakistan, Ed. Robert D. Stevens, Hemzes Alavi, Peter J. Berrocci, East-West Center Book, Honolulu. Emmert, J. P. (1977). Breakdown in Organisational Ideology: The Expansion of Comilla-Type Cooperatives in Bangladesh. Unpublished Monograph, East-West Center. Epstein, S. (1967). Productive Efficiency and Customary Systems of Rewards in Rural South India, in R. Firth, Ed., Themes in Economic Anthropology. London, Tavistock, pp. 230-233. Etienne, G. (1977). Du Bengale Brittannique Au Bangladesh. Tiers-Monde, Vol. XVIII, No. 72, pp. 709-772. Fallaci, 0. (1973).
Revue
Chicago Tribune, June 24th.
Famine Inquiry Commission (1975). India Press, New Delhi.
Report on Bengal.
Government of
Faulkingham, R. N. and Thorbahn, P. F. (1975). Population Dynamics and Drought: A Village in Niger. Population Studies, Vol. 29, No. 3, pp. 463-477. Fitzpatrick, J. (1977). The Role of Emergency Grain Reserves. Vol. 1, No. 3, pp. 217-221.
Disasters,
FNSSS (1975). The Belg Rainsâ&#x20AC;&#x201D; A Survey of Their Effect on Food Production. Food & Nutrition Surveillance System Secretariate, Addis Ababa. Freeman, T. W. (1957). Pre-Famine Ireland: A Study in Historical Geography. Manchester University Press. Freire, Paulo (1970). Pedagogy of the Oppressed, Translated by Myra Bergman Ramos. Herder and Herder, New York.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
328 Frisch, R. E. (1972). Weight at Menarch: Similarity for Well Nourished and Under Nourished Girls at Differing Ages, and Evidence for Historical Constancy. Paedatrics, Vol. r3, No. 3, pp. 445-450. Fuchs, R. J. and Street (1980). Land Constraints and Development Planning in Taiwan. Journal of Developing Areas, April, 1980 (forthcoming). Fuller, G. A. (1971). The Geography of Prophylaxis: An Example of Intuitive Schemes and Spatial Competition in Latin America. Antipode, Vol. 3, No. 1, pp. 21-30. Furnivall, J. S. (1948). Colonial Policy and Practices: A Comparative Study of Burma and Netherlands India. Cambridge University Press, Cambridge. Geddes, A. (1937). The Population of Bengal, Its Distribution and Changes, A Contribution to Geographical Method. The Geographical Journal, Vol. LXXXIX, No. 4, pp. 344-368. Geddes, A. (1970). The Population Geography of the Indian Sub-Continent during an Era of Change, in Understanding Society, Ed., Social Sciences Foundation Course Team, pp. 620-636. The Open University Press, Macmillan & Co., London. Gerassimov, I. P., and Zvonkova, T. B. (1974). Natural Hazards in the Territory of the USSR: Study Control and Warning, in Natural Hazards: Local, National & Global, Ed., Gilbert L. White. Oxford University Press, London. Ghosh, K. C. (1944). Lahore.
Famines in Bengal:
1770-1943.
Hero Public cions,
Ginsberg, N . , Ed. (1961). Atlas of Economic Development. Illinois, University of Chicago Press.
Chicago,
Gouldner, A. W. (1960). The Norm of Reciprocity: A Preliminary Statement. American Sociological Review, Vol. XXV, No, 2, pp. 161-162. Government of Bengal-Revenue Department (1913). Famine Code. Superintendent of Government Printing, Bengal Government Press, Alipore, Bengal/NIPA Reprint, Dacca. Government of Pakistan (1964). Land: and Publications, Karachi. Green, W. (1975). pp. 9-51.
Triage.
A New Pattern.
Dept, of Films
The New York Times Magazine, January 5,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
329 Greenough, P. R. (1975). Famine Mortality, Destitution and Victimisation, Bengal, 1943-46. Paper, 11th Annua Bengal Studies Conference, University of Iowa, Iowa City, Iowa. Greenough, P. R. (1976). Famine Victimisation in a Traditionally Famine Free Region: Bengal 1943-44. Paper Presented, Panel on Famines in South Asia: Theory,History and Policy. Greenough, P. R. (1977). Mortality dueto the Bengal Famine of 1943-44. The Available Data and their Interpretation. Paper Presented at the Midwestern Meeting of theAssociation for Asian Studies, Historical Demography of Asia Panel, Da Kalb, Illinois. The Guardian Weekly (1978). Refuge of the Desparate, Vol. 118, No. 25, Week Ending June 24th, p. 1. Hailey, Lord (1938).
African Survey.
Hagerstrand, E. A., Kuklinski, R. (1973). Information Systems for Regional Development, A Seminarâ&#x20AC;&#x201D; General Papers, Gleerup, Lund, Sweden, Lund Studies in Geography, Series B, No. 37. Hall, E. T. (1976).
Beyond Culture.
Anchor Press, New York.
Hanley, S. B. and Yamamura, K. Population in Pre-Industrial Japan, in Population and Social Change, Ed., Glass, D. V., and Revelle, R. Edward Arnold, London. Haque, Wahidul et al. (1971). Towards a Theory of Rural Development. Development Dialogue, No. 2. Hansen, R. D. (1976). The U.S. and World Development Agenda for Action, Overseas Development Council, Praeger, New York. Hardin, Garrett (1974). pp. 561-568.
Living on a Lifeboat.
Bio-Science, Vol. 24,
Harvey, D. (1974). Population, Resources, and the Ideology of Science. Economic Geography, Vol. 50, No. 3, pp. 256-277. Harvey, P. D. (1969). Development Potential in Famine Relief: The Bihar Model. International Development Review, Vol. XI, No. 4, pp. 7-9. Hare, F. K. and Jackson, C. J. (1972). Environment: A Geographer's Perspective. Geographical Paper No. 52, Lands Directorate, Ottawa. Hay, R. (1976). Personal Communication from UNICEF Nutrition Surveillance Officer, Addis Ababa after the Ecology of Famine Conference AAAS Annual Meeting in Boston, February 20.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
330 Hewitt, K. and Burton, I. (1971). The Hazardousness of a Place, A Regional Ecology of Damaging Events. Research Publication No. Dept, of Geography, University of Toronto. Hiwed, A. (1975). Paper No. 1.
6,
Ethiopia: from Autocracy to Revolution. Occasional Review of African Political Economy, pp. 17-18.
H.M.S.O. (1843). Report of the Commissioners appointed to rate the Census of Ireland for the Year 1841. HMSO, Dublin. Hogan, R. C., Broske, S. P., Davis, J. P., Eckerson, D., Epler, G. , Guyer, B. J., Kloth, T. J., Kolff, C. A., Ross, R., Rosenborg, R. L., Staehling, N. W., Lane, J. M. (1977). Sahel Nutrition Surveys, 1974-1975. Disasters, Vol. 1, No. 2, pp. 117-124. Hopkins, L. D. (1977). Methods for Generating Land Suitability Maps: A Comparative Evaluation. Journal of the American Institute of Planners, pp. 386-400. Hornsby, M. (1974). Why a Good Harvest May Not Save Thousands from Starvation in West Bengal. Times, 18th October, p. 16. Hosier, A. (1873). Droughts in China, A.D. 620 to 1643. Journal of the North China Branch of the Royal Asiatic Society, Vol. 12, pp. 51-98. Hosie, A. (1878-79). pp. 371-372. Hunter, W. W. (1868).
Floods in China, 630-1630.
Annals of Rural Bengal.
China Review, Vol. 7,
Smith and Elder, London.
Hynam, A. S. (1966). The Dysfunctionally of Unrequited Giving. Human Organisation. Vol. 25, No. 1, pp. 42-45. IBRD (1972). Bangladesh: Land and Water Resources Sector Study, Sector Overview, International Bank for Reconstruction and Development, Washington. IBRD (1974). Bangladesh Development in a Rural Economy, International Bank for Reconstruction and Development, Washington. ICNND (1966). Nutrition Survey of East Pakistan, March 1962-January 1964. International Committee for Nutrition and National Defence, National Institute of Health, Bethesda. Imperial Gazetteer of India (1907). pp. 101-105.
Bengal, Vol. 11, Provincial Series,
International Development Association (1974). Appraisal of Drought Areas Rehabilitation Pro.ject, Washington.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
331 International Development Association (1974). Ethiopian Drought, Addis Abbaba.
Status Report on the
Isenman, P. J. and Singer, H. W. (1977). Food Aid: Disincentive Effects and their Policy Implications, in World Food Problem and U.S. Food Politics and Policies, 1972-76, Ed., R. B. Talbot. Iowa State University Press (Ames), pp. 247-251. Islam, M. A. (1974). Tropical Cyclones: Coastal Bangladesh, in Natural Hazards Local, National & Global, Ed., Gilbert F. White. Oxford University Press, London. Jannuzi, F. T. (1977). Personal Communication to the Author, Texas A & M University. Jannuzi, F. T. (1977). Agrarian Crisis in India: University of Texas Press, Austin.
The Case of Bihar.
Jarrett, H. S. (1949). Translation: Ain-i-Akbari of Abul Fazl-iAllani. Calcutta, Royal Asiatic Society of Bengal, p. 136. Jenkins, A. (1976). Personal Communication, Oxfam, Deputy Field Director, Dacca. Jodha, N. S. (1975). Famine and Famine Policies, Some Evidence. Economic & Political Weekly, Vol. X. Johnson, B.L.C. London.
(1975).
Bangladesh.
Empirical
Heinnemen, Educational Books,
Jutikkala, E. (1977). The Structure of Mortality during Catastrophic Years in a Pre-Industrial Society. Historiallisia Tutkimuksia 105, pp. 202-206. Keesing's Contemporary Archives (1975). Bangladesh: Disasters, 20th-26th January, p. 269.
Flood & Famine
Kennedy, R. E. (1973). The Irish: Emigration, Marriage & Fertility University of California Press, Berkeley. Koster, F. and Currey, B. (1975). 1974, Oxfam, Dacca.
Rowmari Nutrition Survey, December
Lamb, H. H. (1973). Some Comments on Atmospheric Pressure Variations in the Northern Hemisphere, in Drought in Africa, Eds., Dalby, 0. and Harrison Church, R. J., 1973 Symposium Center for African Studies, University of London, London, p. 27. Last, G. C. (1975). Relief and Rehabilitation and the Education System, A Working Paper, Planning Unit, Ministry of Education, Government of Ethiopia.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
332 Learmouth, A.T.A. (1970). The I.G.U. Commission on Medical Geography, 1972-76, Geoforum, Vol. 7, No. 2. Lechat,. M. F. (1974). Personal Communication from Director of the Center for Research on the Epidemiology of Disasters, Louvain, Brussels. Lechat, M. F. (1975). Disaster Epidemiology, Editorial, International Journal of Epidemiology, Vol. 4, No. 1, pp. 5-7. Leszczycki, S. (1972). Maps of Human Perturbations of the Environment, Paper No. 29, I.G.U. Man & Environment Commission, Calgary. Leszczycki, S. (1972). The Geographic Participation in Solving Protection Problems of the Human Environment. Geographica Polonica, Vol. 22, pp. 145-150. Leszczycki, S. (1976). Mapping of Environmental Perturbations, Personal Communication, February 20, 1976. Levinson, F. J. (1974). Morinda: An Economic Analysis of Malnutrition Among Young Children in Rural India. Cornell-MIT. International Nutrition Policy Series, Cambridge, Massachusetts. Lewis, J. (1973). Planning for Disasterâ&#x20AC;&#x201D; Proposals for a Working Method. Disaster Research Unit, University of Bradford, England. Lifschultz, L. (1974). Bangladesh: A State of Seige. Economic Review, August 30th, pp. 47-51.
Far Eastern
Linstone, H. A. (1975). Eight Basic Pitfalls: A checklist in The Delphi Method: Techniques and Applications, Linstone, H. A. and Turoff, M . , Eds., Addison Wesley, Reading, Mass., pp. 573-586. Lipton, M. (1977). Why Poor People Stay Poor: Urban Bias in World Development, Harvard University Press, Cambridge, Massachusetts. Loewenstein, M. and Sommer, A. (1972). Bangladesh Health Nutrition Survey, UNROD Information Paper 13, Dacca. Loveday, A. (1914).
History and Economics of Indian Famines, London.
Lowdermilk, W. C. (1924). Erosion and Floods in the Yellow River Watershed. Journal of Forestry, Vol. 22, pp. 11-18. Lyle, J. and von Wodtke, M. (1974). An Information System for Environmental Planning. Journal of the American Institute of Planners, Vol. 40, No. , pp. 394-413.
6
McAlpin, M. B. (1975). The Demographic Impact of Famines in Bombay Presidency 1871-1931. Some Preliminary Findings, Working Paper, Department of Economics, Tufts University.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
333 McAlpin, M. B. (1974). Railroads Prices and Peasant Rationality India 1860-1900. Journal of Economic History, September. Macdonnell, A. P. (1876). Report on the Food Grain Supply and Statistical Review of the Relief Operations in the Distressed Areas of Behar and Bengal during the Famine of 1873-74. Calcutta, Bengal Secretariat Press. MacDonald, G. T. (1967). A Report on the Hydrological Implications of Flood Mitigation Works on the Flood Plain of McCleay River. University of New England, Research Series in Applied Geography, No. 9. Macdougall, E. B. (1975). The Accuracy of Map Overlays. Planning, Vol. 2, no. 1, pp. 23-30.
Landscape
McGlashan, N. D, (1972). Medical Geography: An Introduction, in Medical Geography: Techniques and Field Studies. London, Methuen, pp. 3-5. McGrath, J. E. and Altmen, I. (1966). Small Group Research: A Synthesis and Critique of the Field. Holt, Rinehart, and Winston, New York. McHarg, I. L. (1969). Design with Nature. Press, Doubleday & Company, New York.
Doubleday/Natural History
McHenry, D. F. and Bird, K. (1977). Food Bungle in Bangladesh. Foreign Policy, No. 27, pp. 72-88. Mackay, D. M. (1974). The Effects of Civil War on the Health of a Rural Community in Bangladesh. Journal of Tropical Medicine and Hygiene, Vol. 77, pp. 120-127. McNicoll, G. (1977). Community-level Population Policy: An Exploration. Population and Development Review, Vol. 1, No. 1, pp. 1-22. Mahalanobis, P. C. (1947). A Sample Survey of the After Effects of the Bengal Famine of 1943. Sankya: The Indian Journal of Statistics, Vol. 7, Part 4, pp. 337-400. Majlis, (1977). A Basketful of Hope. September 23, pp. 22-25.
Far Eastern Economic Review,
Makal, M. (1954).
Valentine Mitchell, London.
A Village in Anatolia.
Mallory, W. H. (1926). China: Society, New York.
Land of Famine.
American Geographical
Mallory, W. H. (1927). China's Famine Problem. Materiaux pour 1'Etude des Calamities, New Eds. (July-Sept.), pp. 101-118.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
334 Malone, C. B. and Taylor, J. B. (1924). The Study of Chinese Rural Economy, Report of Chinese Social and Political Science Review, Peking. Masefield, G. B. (1953). Famine; Its Prevention and Relief. Oxford University Press, London. Maslow, A. H. (1946). Problem-Centering Vs. Means-Centering in Science. Philosophy of Science, Vol. 13, No. 4, October, pp. 326-331. Mason, J. B., Hay, R. ÂĽ., Holt, J. Seaman, J., Bowden, M. R. (1974). Nature, 19th April No. 248, pp. 646-650. May, J. M. (1950). Medical Geography: Its Methods and Objectives. Geographical Review, Vol. 40, pp. 9-41. Mead, D. W. (1915). Floods and Famine in China. Vol. XIV, pp. 261-264.
Journal of Geography
Mellor, J. W. (1976). The Agriculture of India. Vol. 235, No. 3, pp. 154-163.
Scientific American,
Milkovich, G. T., Annoni, A. J. and Mahoney, T. A. (1972). The Use of The Delphi Procedures in Manpower Forecasting, TR-7007 (AD 747-651), Minnesota University, Minneapolis Center for the Study of Organizational Performance and Human Effectiveness. Miller, D. S. and Holt, J.F.J. (1975). The Ethiopian Famine. Proceedings of the Nutrition Society, Vol. 34, No. 167, pp. 167172. Miller, J. P. (1970). Lancet, p. 1330.
Medical Relief in the Nigerian Civil War.
Ministry of Relief and Rehabilitation, BDG (1976).
Contract Statement.
Molnar, P. and Tapponier, P. (1977). The Collision between India and Eurasia. Scientific American, Vol. 236, No. 4, pp. 30-41. Montandon, Raoul (1924). La Geographie des Calamities, Materiaux Pour l'Etudes-des Calamities, Vol. 1, No. 2. Morgan, J. P. and Mclntire (1959). Quaternary Geology of the Bengal Basin, East Pakistan and India. Bulletin of the Geological Society of America, Vol. 70, pp. 310-342. Morris, M. D. (1975). Neededâ&#x20AC;&#x201D; A New Famine Policy. Economic and Political Weekly, Annual Number, February, pp. 283-294. Morris, P. A. (1971). Bayesian Expert Resolution, Doctoral Dissertation, Stanford University.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
335 Mosley, W. H. , Chaudhury, A.K.M.A. and Aziz, K.M.A. (1977). Demographic Characteristics of a Population Laboratory in Rural East Pakistan. Population Research, NICHHO. Mosley, W. (1975). Health, Nutrition and Mortality in Bangladesh. Paper for Publication, Department of Population Dynamics, Johns Hopkins University, School of Public Health, Baltimore, October, 1975. Moynihan, D. P. (1969). Maximum Feasible Misunderstanding; Action in the War on Poverty. Free Press, New York.
Community
Mukerjee, K. (1947). The Famine of 1943 and the Nature of Land Transfer in a Village in Bengal. Modern Review, Vol. (April), pp. 309-312.
8
Mukerjee, K. (1965). Agriculture, Famine and Rehabilitation in South Asia: A Regional Approach. The University Press, Santiniketon. Mukherjee, N. (1975). A Bengal Zamindar: Utterpara and His Times (1808-1888).
Jaykrishna Mukher.jee of Firma K. L. Mukhopadhyay.
Mukerjee, R. K. (1944). Effect of the Food Crisis of 1943 on the Rural Population of Noakhali, Bengal. Science and Culture, Vol. 10, Nos. 5 & , pp. 185-91 and 231-238.
6
Mukherjee, S.T.K. Personal Communication, Director of Archives, Government of West Bengal, May , 1977.
6
Mulgrave, N. W. and Ducanis, A. J. (1975). Propensity to Change Responses in a Delphi Round as Function of Dogmatism, in The Delphi Method: Techniques and Applications, Linstone, H. A. and Turroff, Eds.Addison-Wesley, Reading, Mass., pp. 288-290. Murray, T., Rogers, P., Sinton, D., Steinitz, D . , Toth, R. and Way, 0. (1971). Honey Hill: A Systems Analysis for Planning the Multiple Use of Controlled Water Areas. 2 Vol. Report nos. A.D. 736343 and AD 736-344, Springfield, Va., National Technical Information Service. Myrdal, G. (1968). Asian Drama: An Inquiry into the Poverty of Nations. Pantheon, Random House, New York. Nakano, T. (1974). Natural Hazards: Report from Japan in Natural Hazards: Local, National and Global, Ed., Gilbert L. White, Oxford University Press, London. Neef, E. (1972). Some Notes on Mapping Environmental Disturbances, Presented at IGU Man and Environment Commission Meeting, July 24th-30th, Calgary.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
336
Okayama District Administration (1888). Tempo nendo kikin jokyo torishirabesho (Report of a famine in Tempo Province), Kurashiki Municipal Library, Okayama. Paddock, W. and Paddock Q. (1976). and Company, Boston, Toronto. Pankhurst, R.K.P. (1961). Ethiopia, London.
Time of
Famines.
Little, Brown
Introduction to the Economic History of
Pankhurst, R.K.P. (1966). The Great Ethiopian Famine of 1885-1892, a New Assessment. Journal of the History of Medicine and Allied Sciences, Vol. 21, Part 1, pp. 95-124, Part 2, pp. 271-294. Pankhurst, R.K.P. (1972). The History of Famine and Pestilence in Ethiopia Prior to the Founding of Gondar. Journal of Ethiopian Studies, Vol. 10, pp. 37-66. Passmore, R. (1951). Famine in India: Lancet, August 18, p. 303.
An Historical Survey.
The
Patri, T., Streatfield, D. and Ingmire, T. (1970). The Santa Cruz Mountains Regional Pilot Study: Early Warning System. Department of Landscape Architecture, University of California, Berkeley. Paul, C. and Landini, K. (1975). Results of Phase One of Land Use Information Delphi Study. Land Use Information Systems Project, NASA, Washington, Report 43-22. Pill, J. (1971). The Delphi Method: Substance, Context, A Critique and an Annotated Bibliography. Socio-Economic Planning Science, Vol. 5, pp. 57-71. Poor Inquiry (1535). Report of the Commissioners on the Condition of the Poor Classes in Ireland, Vol. XXXII, Session No. 309. Population Reference Bureau (1977). World Population Data Sheet, Population Reference Bureau, Washington. Program for Instructional Development, Clark University (1976). The Environmental Context of Development: A Program to Analyse Environmental Issues and Trends in Eastern and Southern Africa. Draft Report, Worcester, Massachusetts. Qadir, S. R Dacca.
(1975).
Bastees of Dacca.
Quade, E. S. (1967). Cost Effectiveness: The Rand Corporation, P-3529
Local Government
Institute,
Some Trends in Analysis.
Quinn, J. B. (1971). Technological Forecasting. Harvard Review Forecast Series, No. 21215, pp. 51-53.
Business
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
337 Rabbani, A.K.M. Ghulam (1968). An Approach to the Fourth Plan Formulation in Pakistan. Mimeographed. Dacca, Planning and Development Department, Government of East Pakistan. Rahman, M. M. (1974). Personal Communication from the Acting Director of the Cholera Research Laboratory, Dacca. Rahman, M. M. (1976). Report from Bangladesh. Ecology of Famine Symposium, AAAS, Annual Meeting, Boston, 1976. Rahman, M. M. (1978). The Causes and Effects of Famine in the Rural Population: A Report from Bangladesh. Ecology of Food and Nutrition, Vol. 7, No. 2, pp. 99-102. Rakshit (1956). The Law of Alluvial Lands and Fishery in India, Pakistan, and Burma. Eastern Law House Private Ltd., Calcutta. Ralph, K. A. (1975). Perception and Adjustment to Flood in the Meghna Flood Plain, M.A. Thesis, Dept, of Geography, University of Hawaii. Ramalingaswami, F., Deo, M. G., Guleria, J. S., Malhotra, K. K., Sood, S. K., Prakash, 0. and Sinha, R.V.N. (1971). Studies of the Bihar Famine of 1966-67, in Symposium on Famine: Nutrition and Relief Operations in Times of Disaster, Ed., Blix, G. Hofvander, Y. and Vahlquist, B., pp. 94-112. Almquist and Wiksell,r Uppsala. Rangasami, A. (1975). The Paupers of Kholisabhita Hindupara; Report on a Famine. Economic and Political Weekly, Vol. X, No. 5-7, pp. 266-277. Raper, A. F. (1970). Rural Development in Action. University Press.
Ithaca, Cornell
Ratcliffe, John W. (1970). Analyst Biases in KAP Surveys: A Cross Cultural Comparison. Studies in Family Planning, Vol. 7, No. 11. Revelle, R. and Thomas, H. A. (1970). Population and Food in East Pakistan. Journal of Industrial and Tropical Hygiene, Vol. VII, pp. 27-43. Rokeach, M. (1960).
The Open and Closed Mind.
Bantam Books, New York.
Red Cross, U.S. (1929). The Report of the American Red Cross Commission to China. The American National Red Cross, Washington, D.C. Renner, G. T. (1926). A Famine Zone in Africa: Review, Vol. 16, pp. 583-596.
The Sudan.
Geographical
Ritter, J. R. and Dupre, W. R. (1972). Maps Showing Areas of Potential Inundation of Tsunamis in the San Francisco Bay Region of California, USGS.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
338 Rockefeller Foundation (1976). Climate Change, Food Production and Interstate Conflict, A Bellagio Conference, 1975, Working Paper, February. Rohde, J. E. and Gardner, P. (1973). Refugees in India: Innovative Health Care Programs. Disaster in Bangladesh, Ed., Chen, L., Oxford University Press, London. Roxby, Percy M. (1925). The Distribution of Population in China. Geographical Review XV , pp. 1-24. F.oyal Commission on Agricultureâ&#x20AC;&#x201D; India (1927). Bengal, Vol. 4, London. Ruwet, M. J. (1954). Crises Morales?
Evidence Taken in
Crises Demographiques: Problems Economiques o'u Population, No. 3, pp. 451-476.
Sachman, H. (1974). Delphi Assessment: Expert Opinion, Forecasting and Group Process, R 1283-PR, RAND, Santa Monica. Sanderson, F.H. (1975). The Great Food Fumble. No. 4188, pp. 503-509. Schneider, S.H. (1974). July th, p. 18.
6
Specter at the Feast.
Schneider, S.H. (1976). The Genesis Strategy: Survival. Plenum Press, New York.
Science, Vol.
88,
The National Observer,
Climate and Global
Schuler, E. A. and Zaidi, Hafeez, S. M. (1961). Response to Village Disaster: Tornado and Hailstorm in South Rampur. Journal of the Pakistan Academy, Comilla, Vol. II, Nos. 2 and 3, pp. 1-14, 27-46. Schwartzberg, J. E. (1962). Three Approaches to the Mapping of Economic Development in India. Annals of the Association of American Geographers, Vol. 52, No. 4, pp. 455-468. Seamen, J., Holt, J.F.J. and Rivers, J.P.W. (1974). Kaererghe Under Drought, Relief and Rehabilitation Commission, Ethiopia. Sen, A. (1977). Starvation and Exchange Entitlements: A General Approach and its Application to the Great Bengal Famine. Cambridge Journal of Economics, Vol. 1, pp. 33-59. Shan-Yu, Y. (1943). The Geographical Distribution of Floods and Droughts in Chinese History, 206 B.C. - A.D. 1911. The Far Eastern Quarterly, Vol. II, No. 4, pp. 357-378. Sheehan, L. and Hewitt, K. (1969). A Pilot Survey of Global Natural Disasters of the Past Twenty Years, Working Paper No. 11, Natural Hazards Research, University of Toronto.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
339 Shepard, J. and Walker, G. (1974). Drought, USAID, Addis Ababa.
Status Report on the Ethiopian
Shore, John (1843). Memoirs of the Life and Correspondence of John Lord Teignmouth by his son Lord Teignmouth, 2 vols. Hatchard and Sons, London. Siegel, S. (1956). Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. Sims, J. M. (1977). Personal Communication to Author. Library and Records, London, March 2. Singh, K. (1975). The Indian 1967 Famine: Peoples Publishing House, New Delhi.
India Office
A Study in Crisis and Change.
Smith, D. M. (1973). The Geography of Social Well-Being in the United States. An Introduction to Territorial Social Indicators. McGraw-Hill, New York. Smith, D. M. (1975). Mapping Human Well Being. International Social Science Journal, Vol. XXVII, No. 2, pp. 364-371. Smith, G. C. (1976). Responses of Residents and Policy Makers to Urban Environmental Hazards. Area, Vol. , No. 4, pp. 279-283.
8
Smith, R. B. (1861). Report on the Famine of 1860-61. Government of India, List 2 (863), India Office Library and Records, London. Sommer, A. and Loewenstein, M. S. (1975). Nutritional Status and Mortality: A Prospective Validation of the QUAC Stick. American Journal of Clinical Nutrition, Vol. 28, pp. 287-292. Sorokin, P. A. (1975). Hunger as a Factor in Human Affairs (translated by Elena P. Sorokin). University Presses of Florida, Gainesville. Sorre, M. (1943). Les Fondements Biologiques, Tome 1 - Les Fondaments de la Geographie humaine, Paris, p. 377. Spitz, P. Ed. (1976). UNISRD, Geneva.
Famine Risk in the Modern World, Proposal,
Sprague, J. G. and Foster, S. 0. (1972). Second Bangladesh National Nutritional Assessment. UNROD Information Paper No. 21, Dacca. Srivastava, H. S. (1968). The History of Indian Famines, Development of Famine Policy (1858-1918). Sri Ram Mehra and Co., Agra. Steimer, G. A. (1969).
Top Management Planning.
MacMillan, New York.,
Steinitz, C., Parker, P., and Jordan, L. (1976). Hand-drawn Overlays: Their History and Prospective Uses. Landscape Architecture, Vol. No. 5, pp. 444-455
66,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
340 Stepanek, J. (1975). Putting Relief in a Development Context. Memorandum, USAID, Dacca. Stuart, H. and Stevenson, S. (1969). Physical Growth and Development in Nelson W. Textbook of Pediatrics, 9th Ed., pp. 12-61. Saunders, Philadelphia. Susser, M. (1973). Causal Thinking in the Health Sciences. University Press, New York.
Oxford
Taylor, A. E. (1947). Famine. Unpublished Manuscript, Food Research Institute, Stanford University, Stanford, California. Thackeray, W. M. (1875). Irish Sketch Book, The Works of William Makepeace Thackeray, Vol. XVIII, Smith and Elder & Co., London. Thomas, D. S. (1941). Social and Economic Aspects of Swedish Population Movements. Macmillan, New York. Thomas, J. W. (1972). Public Policy in the Reconstruction and Development of Rural Bangladesh. Paper read at the Bengal Studies Conference, University of Toronto. Thorpe, E. W. and Tweedie, A. D. (1956). The New South Wales Floods of February, 1955. Australian Geographer, Vol. , pp. 3-13.
6
Thraves, B. D. (1976). Winnipeg Futures: An Application of Delphi Forecasting to the Processes of Urban Growth. D. A. Thesis, University of Manitoba. Timms, R. (1976). Dacca.
Unpublished memorandum from the Director of CORR,
Toksoz, M. N. (1975). The Subduction of the Lithosphere. American, Vol. 233, No. 5, pp. 88-95. Turnbull, C. (1974).
The Mountain People.
Scientific
Picador, Pan Books, London.
Turner, F. B. (1926). Flood and Famine in North China. Journal of the Royal Asiatic Society of Great Britain and Ireland, North China Branch, Vol. LVII, pp. 1-18. Turton, D. (1977). Response to Drought: The Mursi of Southwestern Ethiopia. Human Ecology in the Tropics, Eds., Gartick, J. P. and Dacy, R.W.J., Halsted Press, New York, pp. 165-192. United States Agency for International Development (USAID, 1974). Drafted by J. Shepard and G. Walker, Addis Abbaba, August th.
8
U.S. Agency for International Development (1977). U.S. Economic Assistance to Bangladesh, Dacca, Bangladesh, Sept. 30th.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
341 Van de Ven, A. H. and Delbecq, A. (1971). Nominal versus Interacting Groups for Committee Decision-Making Effectiveness. Journal of the Academy of Management, June 14 (2). Verghese, B. G. (1967). Committee.
Beyond the Famine, New Delhi:
Bihar Relief
Visvader, H. and Burton, I. (1974). Natural Hazards and Hazard Policy in Canada and the United States, in Natural Hazards, Local, National, and Global, Ed., Gilbert F. White, Oxford University Press, London. Waddell, E. (1974). Frost Over Niugini: New Guinea, 8 (4), pp. 39-49.
A Retrospect on Bungled Relief.
Waddell, E. (1975). How the Enga Cope with Frost: Responses to Climatic Perturbations in the Central Highlands of New Guinea. Human Ecology, Vol. 3, No. 4, pp. 249-273. Waksmundzki, K. A. (1972). A Composite Map of Perturbations and Devastations in a Geographical Environment, Paper No. 59, IGU, Man and Environment Commission, Calgary, July 23-30. Waterlow, J. C. (1973). Note on the Assessment and Classification of Protein-Energy Malnutrition in Children. The Lancet, pp. 87-89. Weaver, W. T. (1970). Delphi as a Method for Studying the Future: Teaching Some Underlying Assumptions. Educational Policy Research Center, Syracuse, New York. Welty, G. (1972). A Critique of the Delphi Technique. Proceedings of the American Statistical Association; Washington, D.C. White, G. F,, Ed. (1961). Papers on Flood Problems, Research Paper No. 70, Department of Geography, University of Chicago. White, G. F. (1973). Natural Hazards Research, in Directions in Geography, Ed., Chorley, R. J. Methuen, London. Wilson, J. T. (1972). Maoâ&#x20AC;&#x2122;s Almanac: 3,000 Years of Killer Earth quakes. Saturday Review, February 19th, pp. 60-64. Wingert, E. A. (1974). Subtraction of ERTS Images of Bangladesh River Islands, Personal Communication, Department of Geography, University of Hawaii. Wiseberg, L. S. (1975). The Statistics Jungle: Measuring War, Plague, Fire, and Famine. Society, Vol. 12, No. 5, pp. 53-60. Wood, A. P. (1976). The Resettlement of Famine Victims in Illubor Province, Ethiopia. Working Paper, No. 28, African Mobility Project, University of Liverpool.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
342 Woodham-Smith, C. (1963). The Great Hunger, Ireland, 1845-49. Hamish Hamilton, London. World Food Program (1976). Vulnerable Group Feeding in Distressed Areas. Project No. 2227, Dacca, Bangladesh. Wright, P. and Davies S. (1974). Fertility Behaviour of a Minority Population: The Catholics in Northern Ireland. Ekistics, Vol. 37, No. 221, pp. 249-256. Wrigley, E. A. (1968). Mortality in pre-industrial England: The Example of Colyton, Devon, over Three Centuries. Daedalus, Vol. 97, No. 2, pp. 246-289. Zaidi, S.M.F. (1970). The Village Culture in Transition. Center Press, Honolulu.
East-West
Zaidi, S.M. Hafeez, and Schuler, E. A. (1962). Reactions to Disaster in an East Pakistan Village, in J. Owen, Ed., Sociology in East Pakistan. Dacca, Sociology Society of Pakistan.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BDEGJi
ABCD,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89°
ABDGJ 54b
- % X \ ABEGJ
K a lig an j
ABDEGJ
-*BDEG>/~
- '67" " ■:-La Ifnanl rhaT —®
ABCD, Saidpur
ADFGJ .Par bat ipu
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
P ub l is h ed Oct 197 8
THANAS
m
S
VERY LIABLE TO FAMINE
3
FI -Dr-PP-FD-LAE-LCY -PLT-RE-NIAI
i
FI-Dr-PP-FD.........LCY-PLT-RE-MAI
i
i
e
r a
\bdegj 53b Nageswar i
THANAS LIABLE TO
i
i
|
1 64 I
63a |
F l-D r—
FD.........L
F l-D r— -FD-LAE— D r-P P -F D .........
FI Or------ FD-LAE-LCY-PLT-RE-MA1
L 636 ]
FI-Dr-PP-FD .........LCY-PLT-RE
1 62
FI-Dr-PP-FD......... LCY..........RE......... Cy
1 61a |
F I............. Fd-LAE-
FI-Dr-PP-FD .........LCY..........RE
|
61b |
Fl-D r-PP-FD -—
F I-D r— -FD-LAE-LCY......... RE
|
59a |
F l- D r - — FD.........L
|
___Dr—
FD-LAE-L
-
FI-D r-PP-FD -............... PLT-RE-MAI
I
59b 1
F l-D r— -FD.........L
FI-Dr-PP-FD.........LCY........................ Cy
|
58a
J
F l-D r..................... 1
F I-D r—
FD.........LCY-PLT-RE..........Cy
1 58b 1
F l-D r—
IM W j
F I-D r—
FD.........LCY-PLT-RE-MAI
|
F l-D r— -FD-LAE
hY-J&VJ
F| -D r—
FD.........LCY-PLT-RE
1 57a J
F I — -PP-FD......... 1
ET m H I
F l-D r— -FD-LAE.........PLT-RE
1 57b 1
F l-D r— -FD......... 1
FI -D r—
1 55
— D r— -FD......... 1
I
•7.3-r 1
h ii i l
FD.........LCY-PLT-RE
58c |
|
FD......... .
m
— Dr-PP-FD.........LCY-PLT-RE..........Cy
1 54a |
— Dr-PP-FD......... 1
lir o + l
FI-Dr-PP-FD.................PLT— -MAI
1 54b |
F l-D r— -FD.........
I
FI-D r-P P ...............
!• 6 7 -
1 F |-D r— -FD-LAE—
r. *rL
ll
ci
n ..
cn
-PLT— -MAI
inv d i t
mia i
53a |
1 53b 1
F l-D r............. LAE-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BANGLAD AREAS LIABLE TO S c a le 1 : 5 0 0 ,0 0 0 1 inch equals approximately 8 miles 10^
0
____ 0
10
10
20
20
30
40
1976 THANAS Li ABLE TO FAMINE LCY-PLT
LCY-PLT LCY-PLT
FD-LAE-LCY-PLT -Fd-LAE FI-Dr-PP-FD-. . . . . . Fl-Dr — FI Jlr—
Fl-D r
FD-LAE
-LCY-PLT I 47a 1 FI-Dr-PP1 47b "1 Fl-Dr FDF W ] Fl-Dr— FD FI-Dr-PP 1 45a 1 Fl-Dr
PLT-RE-
FI-Dr-PP.. Fl-Dr—
COmPGNENTS GF FAMINE LIABILITY Flooding
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ADESH E TO FAMINE 5 0 0 ,0 0 0 oxinately 8 miles
20 !0
30 30
40
40 Mi les 50 Kilometers
â&#x20AC;˘76 SOME REMEDIAL POLICIES FOR THE COMPONENTS OF FAMINE LIABILITY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ilNE li les
DIAL POLICIES FOR THE COMPONENTS OF FAMINE LIABILITY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABCD.
ADFGJ
ADFGJ
ADFGJ ADFGJ
abdgj Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABDGHj 56b I taÂťanganj!
\BDGH
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LCY-PLT-RE
1 47b 1 Fl-Dr— FDI 47c"j Fl-Dr— -FD1 47d | FI-Dr-PP. .
FD-LAE
•D
LCY-PLT-RE PLT
Cy
1 46 1 I 45a 1 j 45b I rW~\
Fl-Dr. . . . Fl-Dr— Fl-Dr— Fi-ur—
MAI
OF FAMINE LIABILITY LCY-PLT LCY-PLT CY-PLT CY-PLT
Cy
FI Dr PP FD LAE LCY PLT RE Cy MAI
Flooding Drought Population Pressure Food Deficit Lack of Alternative Employment Low Crop Yields Poor Land Transport River Erosion Cyclones Maldistribution of Agricultural Inputs
“rt ’ssyssiwss s.*-.'
Priority for aid to these ci
ADEG EFG
AEFG
AEFG
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Tii i L £ r X
,.r K r r ^ ,,i.n5,5 the Barirtd tract, water trans liable to famine are riverine
”::ri«T3Jd,sz?r^Z'i\
.ra y S M B t
iU*rsa'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ADFGJ
ADFGJ
ABDGJ ABDGJ BDFH. ABDJ
ABDJ
damages during year of severe flooding (BWAPDA/IBRD 1972)
(16.4%,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABDGJ ADFGJ 1 49c | Adamd igh!
ABDFJ
ADFGJ
JABDGJj
ABDGJ ABFJ
Bagnu
ABDGJ
DJ
'ABDJ
ABDGJ
ABDJ Ul iapara
ABDGJ
ABD,
ABCDGHJ ABDJ Baraigram
ABDGJ
ABDJ
BCDHJi
ABD ABDJ/ Santhi
ABDJ
ABD
ABD,
ABDE
ABDE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACDH
ACD(
ABDGH Nagarpur
ABDEt .ABDEt
DACCA CITY.
ABDEH Si ngai
ACDhNi ABDGH
49b Kerani ganj
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AEFG
AEFG
ACDG
ACDG ACDG
ACDH
ACDH Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
HAIL HAOR DISTRIBUTION
m
CLASSiFiCAT
AREAS VERY LIABLE TO FAMINE
AREAS LIABL
, Running Mean (10 unit peri Frequency cf Index Values on 5km sq, Nat
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
msmiamoH m cus?i^cation of indices
AREAS VERY LIABLE TO FAMINE
AREAS LIABLE TO FAMINE
Hean (10 unit period)
, Running
Frequency of Index Values on 5km sq. National Grid
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
damages during year of severe flooding
16.4% FLOODING AREAS
probabi I i ty)(of week period during the critical aman seedling period June 13 th-Sept 15th
"J}3*
r* 3 %
% DROUGHT AREAS
i 1074 Populati Density whicigreater than
\
mean. (Bangladesh Census, 1975)
13.7% POPULATION PRESSURE
areas . (Food Di rectorate, 1977.)
13.4%; FOOD DEFICIT Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1*1 ABD
ABD
ABDE
ABD
BDEG ABCj ^i 45b
,
Naga rKa nd al
BDEG
,ABD, ABD ABDFG
ABDFI BDFG
BCDF
BDFG I
BDFGHI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kursdnsgsr
/ acdh 49 b / Dabidwai
« «
J?
j
.
lADFG I , ? Wshuram
,- # B e o w
, jlamganj V
- 1 #
ABCHI
i 53a ^ Senbag
CO.
ABCHI
53a Begumganj
s BDFGHI^ 55 *
)
BDFGHI(
55 Icomparigani#
Sonagazi
ABDGI
CP\
45a mi rsarai
BDFGHI 55 Ramgat i
V
o «/>
%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Frequency c? Index Values on 5km sq. National Grid
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LACK OF ALTERNATIVE EMPLOYMENT AREAS
(IBRD Land
LOW CROP YIELD AREAS
COMPONENTS OF THE FAMINE LIABILITY INDEX Approximate inland cyclone risk (grammar, H. 1975)
6.9% POOR LAND TRANSPORT AREAS
CYCLONE AREAS
(SoiI Survey
5.6%, RIVER EROSION _____ AREAS
MALDISTRIBUTION OF AGRICULTURAL INPUTS AREAS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BDEFGI Kali gap]
BDFG I (Jacopo
BDEG I Syannaga
NOMINAL GROUPS > TECHNICAL EXPER1 AID DISTRIBUTOR!
A Project for the Ministry of Relief and Rehabilitation Government of Bangladesh Supported by the United States of America Agency for International Development Implemented by the Johns Hopkins University International Center for Medical Research
AID 0 ISTIRBUTORS
TECHNICAL EXPERT
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
NOMINAL GROUPS WHO WEIGHTED THE FINAL COMPONENTS OF THE FAMINE LIABILITY INDEX: TECHNICAL EXPERTS: S . A L A M , H .8 R AM M ER , D . CATLING, P . VAN DEN GRAAF, M. E LA H I , A . M A L E K , J . STEPANEK, R . TARAFOAR, I AID DISTRIBUTORS: F . ABED, S . ADHIKHARI, M . MALAKER, P . MYERS, A . RASHEEO, R . LeROCHE, W. THOMAS, G . WARNER, A . ZAMAI SENIOR RETIRED M . AHMED, I . AHMED, M . A L I , R . AMIN. S . HALIM, F . RAHMAN, H . RAHMAN, M . RAHMAN, M . AHMED (K . ISI GOVERNMENT OFFICERS: GROUPS SUGGESTING:
fa) THE INITIAL INDICATORS OF AREAS LIABLE TO FAMINE IN BANGLADESH: (b) POSSIBLE FAMINE PREVENTOR PLANS FOR THOSE AREAS: N . BABURCHI, S . BA BURCH i, A . B EPARY, E . BEPARY, F , BEPARY, G . B E P A R Y , N . B E P A R Y , O . B E P A R Y , W . Bl I. FAKIR, H . G H A Z I , I . G H AZ I , j . G H A Z I . A . HOWLADER, B . HOWLADER, ME. HOWLADER, MO . HOWLADER, N. V . HOWLADER, A . KHAN, H . LAKURIA, A . HAL, A . MANN AN. A . BATABAR, A . MATBER, A . M A Z I , J . M A Z I , Y R . M O L L A , A .M O R O L , R . MRIDHA, S . MRIDHA, YA . MRIDHA, YO. MRIDHA. A . SARDAR, H . S A R K A R , S . SARKEI A . TALUKDER . SENIOR RETIRED A. AHAMAD, FAI. AHMED, J. AHMED, MO . AHMED, MA. AHMED. H . A L I , M . A L I , A . AHN S AR I , R . AMIN, H. GOVERNMENT OFFICERS: M . FAIZ, M. HALIM, E . H A Q U E , R. HAGUE, S. HAGUE, D . HOSSAIN, H . H O S S A I N , N . HUQUE, R . H U O U E , 0. M . K A D I R , A. KARIM, A . K H A L A Q U E , Y . KHAN, A .MANSURUZZAMAN, M. MUTALIB, M. S?SZ!BULLA. B. NIZAM, I FAZ. RAHMAN. H . RAHMAN. M . RAHMAN, S .S A I E D , A . SALAM. M . S I O D I Q U I , A . T A H L R , M . YAJUDDIN, H . TJ AID 01ST IRBUTORS: F . A B E D , S . ADH I KA R I , I. AHMED, T . B A L K E , R . BURKHARD, 0 . CAMPBEL, A . CHOWDHURY, K . OE HILDE, S. I H. JOAST, T. JOHANNS ON, J . KAY, D . KERR, M. KHAN, R. LEROCHE, M . MALAKER, B . MONDOL, R. MYERS, J V. PEDERSON, A . QUADIR, A. QUASHEM, B . RAHMAN, A. RASHEED, A. RASHEEO, G. SCHROMBGENS, B . SIDDIQI 8S . TSUKAHARA, G . WARNER, A .Z A MA N , BKEDARGANJ THANA VILLAGE ELDERS:
TECHNICAL EXPERTS:
C . AALL, R .A G A R W A L A , R . AHMAD, E . A H S A N , S . A L A M , M. ASAFUDDOWLA, W. BEGLEY, A.EIHUIVA. H . BRAMI N. CHOWDHURY, E . CLAY, B. ELAHI, E . HELSING, D . HENDERSON, A . HOSSAIN, ». HOSSAIN, F . HUG, tt. HUG, A . M A L E K , M. MOWLA, K. PITMAN, A . Q U O D U S , H .RAHMAN, M. RAHMAN, H . RAHMATULLAH, P . SATTERTHWAlTE, R . TOURNIER, P . VAN DEN GRAAF, 0 . VORAN, •THE MANY OTHER PERSONS INDIRECTLY INVOLVED WITH THE COMPILATION OF THIS MAP ARE FULLY ACKNOWLEDGED IN THE ACCOMPANYING REPOR'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BDFGHI
92°
£>
5TEPANEK, R . TARAFDAR, M . RAHMATULLAH (K . IS U M ) . DAS, G. WARNER, A . ZAMAN (K . ISLAM) . RAHMAN, M . AHMED (K . ISLAM) .
tR,
H.SARKAR,
S . SARKER,
A S. S IK O AR ,
A M . SIKOAR,
N . H U Q U E , R . H U Q U E , 0 . ISLAM, A . J A M E L Y , A . K A B I R , , NAZIRULLA, B. NIZAM, M .R A HA M AN , FAt. RAMAN, W E R . M. TAJUDDIN, H . TALUKDAR, M . W A H E D , 0 . ZAHUR . IHURY, K. DE WILDE, S . DRYER, 3 . FERNANDO, A. JENKINS, B . M O N D O L , R. MYERS, J . O ’ROURKE, I . PAGE, A . PATHAM, SCHROMBGENS, B . S I DD I QU I, R . THOMAS, R . TIMM, .EY, A . B H U I Y A . H . BRAMMER, E. CAESAR, B . C A T L I N 6 , HOSSAIN, F . H U Q , M . H U O , A . ISLAM, T . JACKSON, K. JORDAN. ILLAH, P . SATTERTHWAlTE, J . STEPANEK, R. TARAFDAR, IN THE ACCOMPANYING REPORT .
C o m p ilatio n and Cartography Bruce Currey Departments of Geography & In t e r n a tio n a l H ealtl U n iv e rs ity of Hawai i Presently: Consultant, Food Systems Group, Resource Systems Institute, East-West Ce Honolulu, Hawaii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHITTAGONG R O R T p l
ABFGi 45a Patiya
BDFGHI
ABFGI
ABDGI
ABD)
ABD I
NOTE:
Narihal iinjera (St. Martin C o m p ila tio n and Cartography Bruce Currey Departm ents o f Geography & In t e r n a t io n a l H e a lth U n iv e r s ity o f Hawaii
Presently: Consultant, Food Systems Group, Resource Systems Institute, East-West Cenl Honolulu, Hawaii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.