Pi Sigma Alpha Spring 2016 Journal by psajournalou

Vol. XVI No. 1

Spring 2016

The Pi Sigma Alpha Undergraduate Journal of Politics (ISSN 1556-2034) is published bi-annually by the Nu Omega Chapter of Pi Sigma Alpha, Oakland University, Department of Political Science, 418 Varner Hall, Rochester, MI 48309-4488. The journal is funded by Pi Sigma Alpha, the National Political Science Honor Society, 1527 New Hampshire Avenue, NW, Washington, DC 20036, http://www.pisigmaalpha.org. The Pi Sigma Alpha Undergraduate Journal of Politics was founded in the spring of 2001 by Delta Omega Chapter of Pi Sigma Alpha at Purdue University, under the name The American Undergraduate Journal of Politics and Government. With the sponsorship of Pi Sigma Alpha, the National Political Science Honor Society, the name of the Journal was changed to The Pi Sigma Alpha Undergraduate Journal of Politics as of the Fall 2004 edition. Electronic editions of the Journal are available online at http://www.psajournal. org. For further information, please contact Dr. Terri Towner at Oakland University (towner@oakland.edu). All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the editors and faculty advisors of The Pi Sigma Alpha Undergraduate Journal of Politics. The Pi Sigma Alpha Undergraduate Journal of Politics and content appearing therein is copyrighted by Pi Sigma Alpha. While holding these rights, Pi Sigma Alpha does not exert editorial or other control over the content of the Journal or the decisions or actions of its staff in the course of normal business operations. As such, Pi Sigma Alpha neither asserts nor accepts responsibility for the content or actions of staff of the publication in the normal course of business as the customs and usages of the law allow. All assertions of fact and statements of opinion are solely those of the authors. They do not necessarily represent the views of Pi Sigma Alpha, the National Political Science Honor Society, the Editorial Board, the Advisory Board, the Faculty Advisors, Oakland University, or its faculty and administration. COPYRIGHT ÂŠ 2016 PI SIGMA ALPHA. ALL RIGHTS RESERVED

The Pi Sigma Alpha Undergraduate Journal of Politics Spring 2016 Volume XVI

Thirty-First Edition

Alexa James Mikaela Strech Molly Knappenberger Dr. Terri Towner

Number 1

Outreach Editor Scheduling/Content Editor Technology Editor Faculty Advisor

Editorial Board

Jana Al-Jarrah Tanir-Vefa Avci Brittany Awadish William Cochran Marissa Coloske Adam George Zuzanna Gos Megan Hoberg

Kalsoom Hussain Jessica Jones Shanna Johnson Jessica Kaljaj Jeffrey Schuett Amanda Stevens Nicholas Walter

Advisory Board

Dr. Robert Alexander Dr. Nicole Asmussen Dr. Cristian Cantir Dr. Rosalee Clawson Dr. Matthew Dabros Dr. David Dulio Dr. Natasha Duncan Dr. Alan Epstein Dr. Megan Hershey Dr. Jennifer Hopper Dr. Dwaine Jengelley

Dr. John Klemanski Dr. Paulette Kurzer Dr. Laura Landolt Dr. Daniel O’Neill Dr. D’Andra Orey Dr. Diana Orces Dr. Christine Pappas Dr. Jo Reger Dr. Robert Spitzer Dr. Harry “Neil” Strine IV Dr. Peter Trumbore

Editorâ&#x20AC;&#x2122;s Preface to the Spring Edition The Pi Sigma Alpha Journal of Undergraduate Politics would first and foremost like to acknowledge all those individuals and institutions which make the publication of this Journal possible semester after semester and year after year. The Journal has continued to grow in terms of submissions, quality, and prestige. Admissions to the Spring 2016 edition were both vast in number and constituted a diverse array of topics. We greatly appreciate all those who have submitted their work to the Journal in hopes of being published. The articles published herein exemplify a high quality sample of the types of undergraduate research being conducted across the country. Although the publication is a completely student-run endeavor, the efforts of the student Editorial Board are guided and supported by a number of individuals and institutions which we would like to thank. First, we would like to thank the Pi Sigma Alpha Executive Council and Executive Committee whose vision and financial support has maintained the quality and direction of the journal. Second, we would like to thank the Faculty Advisory board: the thorough and constructive reviews provided by the members of this board have ensured the articles published herein meet a consistent standard of quality. Finally, we extend tremendous thanks to Editorial Board Faculty Advisor Terri Towner, who has clocked countless hours to ensure the integrity of the journal continues to exceed the standards of excellence set by the editors of its previous editions. The Editorial Board at Oakland University is proud to present the Spring Edition which contains a well-rounded set of articles with varied methodological approaches and topical matter. The publishing process for the Spring Edition followed a relatively smooth path from submission to publication, and the Nu Omega Chapter and Oakland University wish the readers of this edition a similarly enjoyable time. Best, The Editors

Submission of Manuscripts The Journal accepts manuscripts from undergraduates of any class and major. Members of Pi Sigma Alpha are especially encouraged to enter their work. We strive to publish papers of the highest quality in all areas of political science. Generally, selected manuscripts have been well-written works with a fully developed thesis and strong argumentation stemming from original analysis. Authors may be asked to revise their work before being accepted for publication. Submission deadlines are October 1st for the Fall edition and February 1st for the Spring edition. Manuscripts are accepted on a rolling basis; therefore early submissions are strongly encouraged. To submit your work please email psajournalou@gmail.com with an attached Word document of the manuscript. Please include your name, university and contact details (mailing address, email address, and phone number). If possible, include how you heard about the Journal. Submitted manuscripts must include a short abstract (approximately 150 words), citations and references that follow the APSA Style Manual for Political Science. Please do not exceed the maximum page length of 35 double-spaced pages. The Journal is a student-run enterprise with editors and an Editorial Board that are undergraduate students and Pi Sigma Alpha members at Oakland University. The Editorial Board relies heavily on the help of our Advisory Board consisting of political science faculty from across the nation, including members of the Pi Sigma Alpha Executive Council. Due to the time committed to the manuscript review process, we would like to remind students to submit to only one journal at a time. Please direct any questions about submissions or the Journalâ&#x20AC;&#x2122;s upcoming editions to our editors at psajournalou@gmail.com.

Table of Contents The Influence of Ethnic Minority Demographics on Provincial Preferential Policy Making in the Chinese College Admission System

A Counterweight to Populism: Judicial Legitimacy and Popularity in Latin America

Las Ministras: Gender Quota Laws and Female Representation in Latin American Cabinets and Legislatures

Public Opinion of Labor Unions in the United States

Fact Checking in the 2016 Presidential Campaign: Exploring the Effectiveness of Fact Checking Statements by the Candidates Jennifer Kahn and Alexis Wagner, Arizona State University

Bing Sun, University of Michigan

Mishella Romo, Montclair State University

Kristine E. Adams, Saint Anselm College

Baili Volluz, Saint Louis University

The Influence of Ethnic Minority Demographics on Provincial Preferential Policy Making in the Chinese College Admission System Bing Sun University of Michigan This study focuses on the various provinces’ preferential policies toward ethnic minorities in the Chinese college admission system, specifically in terms of bonus points awarded in the National College Entrance Examination. It explores how demographics of ethnic minorities in provinces influence provincial governments’ preferential policy-making. By examining 31 Chinese provinces’ population, education, economic data, and bonus point policies, this study finds that a higher percentage of minority population likely prompts provincial governments to make more generous and detailed policies. However, whether autonomous counties are significantly disadvantaged educationally (lower student enrollment rates in regular secondary school) and financially (lower GDP per capita) compared to regular counties does not seem to affect preferential policy making. Two noticeable outliers in the study are the Sichuan province and Xizang (Tibet) Autonomous Region.

Introduction

gaokao 高考). Athough different provinces administer

different exams, most of the exams have 750 points

Very few, if any, deny the role higher

education attainment plays in upward social mobility. Many also believe that ethnic and racial minorities are disadvantaged in the process. Modern states with multiracial and multiethnic societies have widely adopted positive policies to redress historic inequalities among ethnic groups and the potential for ethnic conflict (Jalali and Lipset 1992). The People’s Republic of China, ethnically dominated by the Han – constituting 91.6 percent of the population – but has 55 officially recognized ethnic minority groups, is no exception (National Bureau of Statistics of China 2011).

Chinese college admission is based entirely on

the National College Entrance Examination (NCEE,

in total. In addition to the raw score received from the test, some students may also receive bonus points because they are student athletes, nationwide academic competition winners, honors students (sanhao xuesheng 三好学生), children of overseas Chinese (huaqiao 华侨), Taiwanese students, and ethnic minorities.

The national policy that universities are

supposed to “relax” their admission standards to an appropriate extent for ethnic minority students has legal bases (Postiglione 1999). Participation of ethnic minorities in higher education has been promoted in three ways (Min 1997). First, ethnic minorities are given enrollment priority – the admission standards

8 in terms of scores on the NCEE have been adjusted by

power that allows ARs to make laws or regulations

the government. Bonus points awarded to minority

that diverge from the central government with

applicants, however, have been steadily decreasing. The

higher level permission. This type of entity exists

government has been encouraging minority students

for smaller administrative units as well. After AR

to engage in “self-strengthening” (ziqiang 自强), with

(from the highest to the lowest level of administrative

the hope that bonus points based on minority status

subdivision), there are autonomous prefecture,

will eventually not be needed. Second, beginning

autonomous county, and autonomous township.

in 1980, some national universities have established

Scholars, policy makers, and average citizens have

special preparatory classes for minority groups due

discussed consistently the equality of educational

to a Ministry of Education of China’s requirement.

opportunity as a social issue. However, in China’s

Ethnic minority students enrolled in the university

case with the specific regard to ethnic minorities, the

with scores lower than the minimum are required to

equality of educational opportunity is rarely discussed

go through one or two years of preparatory studies

in a rigorous way supported by data. Focusing on

before being integrated into the regular student

higher education, I explore why preferential policies

body. Third, nationality institutes that mostly admit

towards ethnic minorities differ across provinces

ethnic minority students have been established. Even

in Chinese college admissions. Specifically, why do

with all these efforts, minorities are still not well-

provinces have different bonus points for ethnic

represented in higher education, especially in top-

minorities in the NCEE? Some may argue that one

tier universities. In this research, I focus on the first

point in NCEE does not matter that much. However,

method mentioned above. Thus, “preferential policies”

since college admission is solely based on points, one

in the following text only refers to the practice of

point might put a student right on the cutoff to be

bonus points if not specified. With the general trend

admitted to the ideal school. At the same time, because

of decentralization in educational governance (Mok

many applicants are taking the exam each year, a

2002), provincial governments in China have deployed various preferential policies toward ethnic minorities – offering a different amount of points based on different criteria.

Although they only constitute 8.49% of the

population, ethnic minorities not only demonstrate their uniqueness in the college admission system, but also through China’s administrative subdivision. An autonomous region (AR, zizhiqu 自治区) is a minority entity that has a higher population of particular minority ethnic groups. AR is a first-level administrative subdivision (same as a province) but with limited administrative, political, and legal autonomy (Sautman 2010). There is a modification

one point increase sometimes can put an applicant in front of another 600 people in the overall ranking in a province. Since most of the provinces offer at least five bonus points to ethnic minority students, they have a huge advantage in the admission process, which indicates the importance of deepening our understanding of provinces’ preferential policies. In the following section, the literature review lays the foundation of the concept of equality of educational opportunity that originated in the U.S., and then presents the big picture of education and ethnic minorities in China, specifically discussing higher education and preferential policies.

Pi Sigma Alpha Undergraduate Journal of Politics

Educational Inequality and Ethnic Minorities in

of these definitions are different. They agree that the

China

negative definition, the foundation definition, the

Equality of Educational Opportunity

definition, and the maximum-variance-ratio-definition

Throughout the history of U.S. education,

competition definition, the equal-dollars-per-pupil are definitely dealing with educational inputs. In

scholars have provided various conceptualizations

addition, the minimum attainment definition counts

of equality of educational opportunity, but failed

towards the output category. Takase (1999, 49) claims

to agree on a universally applicable definition.

that the definition of leveling –“resources should be

Nevertheless, two schools of thought, originating

allocated in inverse proportion to students’ ability

with Thomas Jefferson and Horace Mann, mostly

based on the assumption that students should as

dominate researchers’ ideas of what constitutes an

nearly as possible leave school with an equal chance

equal educational opportunity. The cleavage between

of success” – talks about outputs. However, in his

the two camps is this: do scholars define equality

overview, Nwabuogu (1984) identifies two other

based on input or output of education? The trend from

definitions concerning outputs: the full-opportunity

Jefferson places emphasis on input, whereas the trend

definition – all persons are to be given full opportunity

from Mann is concerned with equal output. Jefferson

to develop their abilities to their limit, and the

emphasized the equality of all men and their right to

classification definition – equality for all within

be educated (Ulich 1965). Mann stressed that equal

a classification based on the general idea of “the

attainment, or outcome, is possible. Scholars do not

equal treatment of equals”. One take away from this

necessarily separate the two from each other, and many comparison is that the two schools of thought can of them incorporate both concepts into their studies.

sometimes be intertwined when talking about more

James Coleman (1968, 16) identifies the following four

specific policies.

types of inequalities: (1) “differences of community’s

input to the school”; (2) “different racial composition

(1978) assess equality of educational opportunity

of the school”; (3) “various intangible characteristics

according to the following four standards: (1) equality

of the school as well as the factors directly traceable

of educational access; (2) equality of educational

to the community inputs to the schools”; and (4)

participation; (3) equality of educational results; and

“consequences of the school for individuals with equal

(4) equality of educational effects on life changes.

backgrounds and abilities.” The first and the third

Clearly, the four criteria fit into the two main camps

definitions are directly related to Jefferson’s idea of

– the former two focusing on inputs, and the latter

input. The last definition focuses on Mann’s ideal of

two on outputs. Nwabuogu (1984) presents these

equality of result but at the same time works on the

four standards in a step-wise diagram showing

premise of same individual input.

the evolution of measures of equal educational

opportunity over time, which is in accord with the

Wise (1968) proposes nine definitions of equal

Levin (1976) and Garms, Guthrie, and Pierce

educational opportunity, each of which is concerned

sequence presented above. The first criterion is the

with either inputs or outputs of students. One might

most traditional, more easily-achievable measure of

be surprised that different researchers’ categorizations

equal educational opportunity, and the last criterion

Sun

is the most advanced, least attainable measure. As society prospers politically, economically, and socially, the move towards equalizing resource inputs and individual achievements begin to emerge. Similarly, Green (1980) advances the two views, calling them “the best principle” and the “equal principle”. The former shares Jefferson’s ideal, claiming the right to receive individualized education that is the best for each child, whereas the latter aligns with Mann’s ideal of the right to receive an equal education for everyone. The two principles are not necessarily mutually exclusive, according to Kurosaki (1995), which is consistent with the observation about the Wise (1968) definitions and the time process Nwabuogu (1984) presents.

From the scholarly works mentioned above,

one can see that the two trends clearly focus on different criteria of equal educational opportunity. However, researchers have been incorporating both ideas into their conceptualizations. Thus, when examining specific policies, scholars should not overlook either side. An approach that balances the two competing values may render better results. Although the literature discussed above does not have a specific focus on higher education, nor is it related to China, and Chinese scholarly works discussed in the following section do not refer back to the U.S. originated concepts. Scholars still have to keep the two schools of thought in mind when studying the preferential policies in education in China, especially when making policy recommendations. In fact, China’s policies to encourage ethnic minorities to receive all levels of education seem to be a combination of input and output focuses. The government wants to ensure minorities’ right to receive education, and hope that ethnic minorities can achieve highest education attainment as well.

Education and Ethnic Minorities in China The General Picture Access to schooling has never been more widespread in China; when compared to other developing countries, China’s pace in making education available to its huge population has been astonishingly fast (Postiglione 2006). The irony is that while the socialist market economy has increased the educational choices available, it has also made these choices more a function of income, gender, and ethnicity compared to the pre-reform, planned economy period. Educational inequalities continue to widen (Postiglione 2006). Although economic development and policies aiming to promote minority education have increased access to formal schooling since 1949, the educational attainment of ethnic minorities still remains behind the Han majority (National Bureau of Statistics of China 2011). Minority groups in China are diverse with regard to cultural practices and historical experience with the larger Han society (Gladney 1996; Harrell 1995; Mackerras 1994, 1995). Studies have suggested that ethnic groups develop unfavorable attitudes toward education if they perceive that the school system is incompatible with aspects of their own cultures, or if they do not observe tangible returns to education among members of their own communities (Hansen 1999; Harrell and Mgebbu 1999). According to Hansen (1999), one big problem of minority school education is that it is entirely based upon Chinese language and history, allowing no room for the transmission of cultural values of ethnic minorities and denying the significance of the minorities’ own languages, histories, religions, and cultural values. There is a striking contradiction where the government preaches the constitutional equality of nationalities (minzu 民族), while impressing on minority

Influence of Ethnic Minority Demographics

students’ immense feelings of cultural inferiority. The Chinese government educators and many Chinese intellectuals praise education as a means of civilizing the “backward”, which contradicts the outspoken message of national equality. This is one of the reasons the Chinese government fails to popularize Chinese education effectively to some minority groups. On the contrary, in other regions where schools that teach Confucian values and Chinese classics were established early and the history of Han influence was long (for example, Naxi in Lijiang, Yunnan), the popularization effort is very successful. Nevertheless,

also confirmed the role of poverty and geography in contributing to educational disparities by ethnicity. Wang (2009) states clearly the policy’s positive impacts, which include: (1) increasing diversity at universities in terms of both culture and languages at a minimal financial cost; (2) significantly broadening minorities’ access to college education with little social cost; and (3) showcasing great symbolic significance for ethnic equality and the cohesion of the nation. Some scholars criticize this policy, however. Wang (2010) lists a series of problems the current preferential policies cause: (1) minority students

being assimilated by education, minority students’ level of understanding of their own ethnic culture is worrisome (Sun 2012). Consistent with the theory presented before, another important reason some minorities choose not to receive education is that they fail to see significant economic or social advantage in spending money on school education. It certainly does not help that most of the estimated return to education in China is considerably lower than the world average (Han 2014).

have less learning enthusiasm and confidence; (2) policies create waste in education and human capital; (3) companies and organizations have bias against minority graduates; and (4) admission process is less fair due to opportunistic tendency. For example, there were numerous scandals that applicants falsely reported or deliberately changed their ethnic identities in order to obtain the bonus points (Sun 2012; Wang 2009). Ma (2009) adds that the adjustment of admission standards is the primary factor responsible for the decline of educational quality in both secondary schools and universities. Other researchers are concerned with the concept of fairness or equality. Yan and Li (2012) explain the problem in terms of “reverse discrimination”. They point out that under the current “separate-provincial admission” (fensheng luqu 分省录取) model, the preferential policies might hurt Han students’ opportunity to get into college. The current circumstance is that within one province, access to educational resources is fairly equal for different ethnic groups, while most inequalities occur among different provinces. Under the separate-provincial admission model, the real competition for admissions is within a province. Han students in provinces with big minority populations, without preferential admissions, are

Higher Education and Preferential Policies Most Chinese scholars recognize the necessity and benefits of preferential policies towards ethnic minorities (Jin and Wang 2007; Long 2010; Sun 2012; Tang 2003; Teng and Ma 2009; Wang 2007; 2009). They argue that the policies function as an effective guarantee of more equal access to education right, opportunities, and resources for ethnic minority students whose educational performance is comparatively poor. At the same time, scholars recognize that the unequal opportunity for ethnic minorities is due to economic reasons and the lack of educational resources in some regions. Thus preferential admission is a reasonable remedial policy to address the inequality. Hannum (2002)

12 disadvantaged compared to their minority classmates because those provinces generally do not provide good educational resources due to economic constraints. Also, in regard to equality, Zhou (2009) states that China’s approach to national equality, borrowed from the Soviet Union, causes two major problems. First, the approach emphasizes equality of groups while it ignores differences between groups. For example, different groups’ proportional representation in universities is not considered. Other studies also mention that some ethnic groups, such as Mongol, Manchu, and Hui, have exceeded the average level of educational attainment of the country (Sun 2012). Second, the approach focuses on equality among groups while it ignores inequality within a group. For instance, larger quotas go to urban minority residents than rural minority residents. In short, the Chinese concept of “minority education” tends to regard all nationalities as one relatively homogenous group of people in need of more or less uniform special considerations within education (Hansen 1999). Despite the controversy, studies offer very similar policy recommendations. Many researchers recognize that economic reasons and the lack of educational resources cause unequal educational opportunities for ethnic minorities (Long 2010; Tang 2003; Wang 2007; Xiao and Liu 2014). They recommend that the government lower tuition costs and provide special financial aid to help minority college students (Jin and Wang 2007; Long 2010; Tang 2003). In the long run, the government should level the access to basic education, making minority applicants more competitive in the NCEE which will lead to more equal access to higher education (Wang 2010; Yan and Li 2012). These considerations show that China’s ideal of equality of educational opportunity is in accordance with Jefferson’s concept of input, which according to Nwabuogu (1984), is the most traditional, more easily-

achievable measure of equal educational opportunity. Other studies suggest that the government should improve the preferential policies by taking into consideration the differences in region, socioeconomic status, and access to basic education (Teng and Ma 2009; Sun 2012). From the articles discussed above, one may notice that most scholars examine the preferential policies on the national level, ignoring the fact that preferential policies vary widely among provinces. My research focuses on comparing different provinces and explaining why they have different preferential policies in terms of bonus points for ethnic minorities in the NCEE. I contend that the distinct demographics of ethnic minorities in each province’s jurisdiction influence provincial governments’ preferential policy making. In the next section, I explain my theory and hypotheses in a more detailed fashion. Theory and Hypotheses Why do preferential policies towards ethnic minorities differ across provinces in Chinese college admissions? Under the provincial enrollment model, the real competition in admissions is within each province (Yan and Li 2012). One might assert that this makes comparisons between provinces meaningless because they do not compete with each other. However, this is exactly the reason why it is important to explain the difference in policies. Although all provincial governments build on the central government’s foundational policy, each government acts solely on specific preferential policy making. I theorize that the distinct demographics of ethnic minorities in each province’s jurisdiction affect the preferential policies made by provincial governments. Specifically, I propose the following four hypotheses: Hypothesis 1: Greater percentage of bonus points out of the total points are given to

Pi Sigma Alpha Undergraduate Journal of Politics ethnic minority students when a province has a bigger percentage of minority population. Hypothesis 2: Greater percentage of bonus points out of the total points are given to ethnic minority students when, within that province, the student enrollment rates in regular secondary schools of autonomous counties are significantly lower than other regular counties. Hypothesis 3: Greater percentage of bonus points out of the total points are given to ethnic minority students when, within that province, the GDP per capita of autonomous counties are significantly lower than other regular counties. Hypothesis 4: A province offers more types of bonus points for ethnic minority students when a province has bigger percentage of minority population.

The first three hypotheses share the same

dependent variable: the percentage of bonus points out of the total points given to ethnic minority student. Hypothesis 1 is based on the assumption that when a group constitutes a bigger percentage of the population, the government needs to consider minority interest more in policy making to prevent turbulence. Keeping the stability of the society (weiwen,

) is one of the Chinese government’s

top priorities. This is also true for provincial governments, especially for provinces that have higher percentages of ethnic minority populations. For example, in Xizang (Tibet) and Xinjiang Autonomous Regions, there has been a series of violent disturbances due to the tension between Tibetans/Uyghurs and the government. In those provinces, the government’s

preeminent goal is to pacify the unrest. Offering more bonus points in the NCEE and letting more minority students receive mainstream higher education is one response to lessen conflict. Although Tibet and Xinjiang may be the most extreme examples, there is no doubt that the bonus points serve as a “carrot” to bring more minorities to the government’s side. Plus, many students go back to their hometown to serve as government officials, which is even more effective for the overall objective of a stable society. To clarify, this explanation mainly addresses why provinces’ preferential policies differ in the first place; it does not necessarily explain the trajectory of one province’s policy development over time. For example, one might expect that after an ethnic unrest, the provincial government, in order to keep stability, would award more bonus points to minority students. Nonetheless, the provincial government might not be that responsive to unrests for two reasons. First, by awarding more bonus points, it could send a wrong signal to other minorities both within and outside of the province that similar behavior will be “rewarded”. It might risk the stability of the bigger society, and the central government would be unlikely to approve such policy. Secondly, awarding more bonus points is in conflict with central government’s promotion of minority “self-strengthening” in recent years. Therefore, one needs to exercise caution when interpreting relevant results. Hypotheses 2 and 3 look into the difference between autonomous counties and regular counties in provinces. Autonomous counties, although not exclusively populated by ethnic minority citizens, usually have much higher percentages of ethnic minority citizens than regular counties (see Tables 1 and 2 in the Appendix for weighted percentage of ethnic minority population in autonomous counties of provinces). Both hypotheses are establishing

Sun

the fact that ethnic minorities are disadvantaged

can refer to different numbers of bonus points for

compared to Han. Specifically, Hypothesis 2

students living in areas where minorities are highly

focuses on whether minorities are disadvantaged

concentrated (such as autonomous townships), versus

in education by comparing the student enrollment

living sporadically in other rural or even urban areas.

rates in autonomous counties and regular counties.

For Jilin and Heilongjiang provinces, this refers to

Hypothesis 3 concentrates on whether minorities are

different numbers of bonus points that are given to

disadvantaged economically in general by comparing

students who use Mandarin or traditional ethnic

the GDP per capita in both types of counties. Note that

language during NCEE. There is a variety of criteria for

educational disadvantage and economic disadvantage

provinces in terms of offering different types or levels

assessed in these two hypotheses are probably inter-

of bonus points.

related since education and economic well-being are

generally intertwined, both promoting the other. I

directionality of the relationship: could the more

hypothesize that when ethnic minorities are more

generous preferential policies attract more ethnic

disadvantaged educationally and economically,

minorities to move to that province? This is highly

provincial governments will make strong preferential

unlikely due to the household registration system

policy and will award more bonus points. This aligns

(hukou zhidu 户口制度) in China. As a result of hukou

with the central government’s considerations to

or being registered with a specific local police station,

have these remedial policies in the first place, where

changing hukou is usually a difficult task, involving

ethnic minorities need special treatment due to their

onerous paperwork. This restricts one’s freedom to

disadvantaged background.

migrate across provinces, or even within province

between rural and urban areas. Many are deterred

Hypothesis 4 has a different dependent variable

One might raise a question about the

from the other hypotheses but shares the independent

by the complicated process. In addition, a student’s

variables with Hypothesis 1. I assume that when a

eligibility to take NCEE in a province is related to

province has a larger percentage of ethnic minorities, the ethnic population in that population is likely to be more diverse. The diversity here can entail a wide range of criteria: how many different ethnic minority groups reside within the province? Is the habitation pattern of ethnic minorities scattered or concentrated? How many ethnic minorities live in urban areas and in rural areas? What is the language of administration during the NCEE, Mandarin or traditional ethnic languages? When the ethnic population is more diverse, the provincial governments inevitably need to make more detailed policies to address this diversity. Thus, there will be more types or levels of bonus points awarded to ethnic minorities in those provinces. This

hukou. If hukou record shows that the student recently migrated to the place where he or she plans to take NCEE, the student risks the possibility of not being granted the permission to take the exam. Thus, it is reasonable to assume that it is the ethnic minority population that influences the provincial preferential policy making.

The four hypotheses all address the specific

demographic features of the ethnic minorities within a province, which leads to my overarching thesis statement that distinct demographics of ethnic minorities in each province’s jurisdiction affect a provincial government’s preferential policy making.

Influence of Ethnic Minority Demographics Research Design Definition and Measurement In this study, provinces are defined as the 31 provinces (or autonomous regions) in mainland China, excluding Hong Kong, Macau, and Taiwan due to their distinct college admission systems. The dependent variable for the first three hypotheses is what percentage of bonus points out of total points is given to ethnic minority students in each province. Although the NCEE’s total number of points is 750 for most provinces, several provinces use different total points. As a result, the dependent variable is defined in percentage scale. When examining these three hypotheses, I take into account only the highest possible bonus points an ethnic minority student can get in a province. In other words, different levels of bonus points within a province’s policy are omitted. The percentages are calculated according to a list of provinces’ preferential policies in 2014 compiled by the widely-known college application expert in China who goes by the pseudonym “Chenwu” (Chenwu 2014). One might argue that the simplification of only counting the maximum can skew the interpretation of the policy because the highest possible bonus points available might be given much less frequently in reality than lesser amounts of bonus points. I contend that it makes sense to examine the upper limit of bonus points given because it shows exactly how far the provincial governments can be pushed in policy making considering the demographics of ethnic minorities in their jurisdictions. In addition, the criteria for awarding bonus points are extremely clear – applicants get what they qualify for. Because the admission is so competitive, preferential policy and NCEE in general are always under strict public scrutiny. This further decreases the chance of governmental malpractice.

The dependent variable for the fourth hypothesis is how many different types or levels of bonus points are offered by a province’s preferential policy. “Type” refers to any specification in the preferential policy that addresses part of the ethnic minority applicants. For example, this can be multiple clauses that address the bonus points offered to different ethnic minority groups, or clauses that separate the policy towards students who take the NCEE in Mandarin from students who take it in their traditional ethnic language. “Level” refers to the different amount of bonus points given. However, if a province offers two types of bonus points at the same level (e.g., a policy that has two separate clauses, both awarding 10 points to two specific minority groups in the province) the province would still be assigned the value of two in this scale. In addition, if a section of the policy states that “when a Han student and an ethnic minority student have the same score, the latter is preferred in admission” (Chenwu 2014) without designating points-added, I count this as 0.5 in the scale. Note that this scale does not include the specific policies towards Han living in minority concentrated areas. I will show how the bonus points type/level is counted using Liaoning Province as an example. Liaoning’s preferential policy includes three clauses. First, “ethnic minority applicants who reside scatteredly in Han regions have admission priority when all else being equal” (Chenwu 2014). This counts as 0.5 in the scale. Second, “ethnic minority applicants from Kazuo, Fuxin Mongol Autonomous Counties and XinBin, Qingyuan, Fengcheng, Xiuyan, Kuandian, Beizhen,Beixi, Huanren Manchu Autonomous Counties are awarded five bonus points during admission” (Chenwu 2014) Since it is one clause that provides a single level of bonus points offered, it counts as one in the scale. Third, “Korean and Mongol

16 applicants who study at ethnic secondary schools with bilingual (Korean/Mongol and Mandarin) instruction are awarded ten bonus points during admission” (Chenwu 2014) This clause that is based on high schools’ languages of instruction indicates another level of bonus points awarded, thus is counted one in the scale. To add the three scores together, Liaoning’s bonus points type/level count is therefore 2.5. Note that if there is one additional clause that also offers 10 points to certain minority students who meet other criteria, it would be counted as another “type” of bonus point, and one should be added to the total type/level count. The independent variable for Hypotheses 1 and 4 is percentage of minority population of a province. These data come directly from China 5th Nation Province Population Census Data in 2000 (China Data Center 2000a). For Hypotheses 2 and 3, the independent variable is the calculated difference in percentage (∆%), examining how much lower the enrollment rates in regular secondary schools and the GDP per capita in autonomous counties are than in regular counties. I chose to calculate percentages instead of using numerical differences because the former more accurately demonstrates the extent of the disparity. For instance, two provinces may have the same numerical difference of 1,000 dollars in GDP per capita between two county types, but the baseline GDP per capita, GDP per capita of regular counties, may be 10,000 dollars apart, making the numerical difference measurement imprecise in judging the disparities. I will demonstrate the calculation involved in Hypotheses 2 and 3 again using Liaoning province as an example. Liaoning has 36 regular counties and eight autonomous counties. For Hypothesis 2, I calculate the weighted averages of student enrollment rates in regular secondary schools in these two county types. The weighted average for regular counties is

3.07 students per 10,000 people, while the weighted average for autonomous counties is 2.21 students per 10,000 people. I subtract the latter from the former, and the result indicates that for every 10,000 people, the autonomous counties have 0.86 fewer students enrolled in regular secondary schools than regular counties. This numerical difference is then divided by the weighted average of regular counties (3.07 per 10,000 people) in order to provide the difference in percentage. Thus, the autonomous counties’ enrollment rate of regular secondary schools in Liaoning province is 28.15% lower than regular counties’ rate. The 28.15% is the value of the independent variable for Hypothesis 2. Note that a negative percentage means that autonomous counties’ enrollment rate is actually higher than regular counties. Similarly, I calculate the GDP per capita of both regular counties and autonomous counties. For Liaoning province, the regular counties have a GDP per capita of 6,716.67 dollars, while the autonomous counties have a GDP per capita of 4,877.79 dollars. After subtraction, the result shows that the autonomous counties’ GDP per capita is 1,838.88 dollars lower than regular counties’. This amount is then divided by 6,717.67 dollars, the regular counties’ GDP per capita. Therefore, the GDP per capita in Liaoning’s autonomous counties is 27.38% lower than in its regular counties. The 27.38% is the value of the independent variable for Hypothesis 3. A negative province percentage indicates that autonomous counties have a higher GDP per capita compared to regular counties in that province. The county level data of population, enrollment, and GDP needed to conduct the test are provided by China 2000 County and District Population Census Data Assemblies (China Data Center 2000b). To maintain the time consistency of independent variables, I use the county level data in

Pi Sigma Alpha Undergraduate Journal of Politics the year of 2000 as well. For Hypotheses 2 and 3, data of the autonomous regions and provinces that do not have autonomous counties are excluded because no comparison can be made. I expect most of the calculated differences in percentage are positive since ethnic minorities are generally less educationally and economically advantaged. This does not mean that a negative value is impossible to appear for several specific ethnic groups actually do equally well or even better than Han in terms of family wealth and education attainment. For each hypothesis, I create a graph to show the relationship between the independent variable and the dependent variable. Positive correlations in graphs would demonstrate support for my hypotheses. Analysis and Results Figure 1 illustrates the relationship between the percentage of ethnic minorities in a provinceâ&#x20AC;&#x2122;s population and the percentage of bonus points out of the total points that are given to ethnic minority

students (See Table 3 in the Appendix for specific data point information). Taking most of the points in Figure 1 into account, a positive correlation between the variables can be observed. If ethnic minorities constitute a higher percentage of a provinceâ&#x20AC;&#x2122;s population, then that province is more likely to offer generous bonus points to ethnic minority students in NCEE. Although most of the data points are clustered in the lower left corner (x-axis 0%-40%; y-axis 0%3%), and many points are close to the y-axis, Figure 1 shows that as the percentage of ethnic minority population increases, the spread of the percentage of bonus points awarded shrinks and the data points are closer to the upper bound of this corner. There are two noticeable outliers on the graph â&#x20AC;&#x201C; Sichuan province at the upper left corner, and Xiziang (Tibet) Autonomous Region at the lower right corner. I will analyze the possible reasons for these provinces to be drastically different from the rest of the provinces. Overall, Figure 1 supports Hypothesis 1 that greater percentage of bonus points out of the total points are given to

Figure 1. Percentage of Bonus Points out of Total by Percentage of Ethnic Minority Population

Note: N=31 Provinces Sources: Chenwu (2014), China Data Center (2000a)

Sun

ethnic minority students when a province has a bigger percentage of minority population. For Hypotheses 2 and 3, I mentioned in the research design section that I would exclude all the autonomous regions and provinces that do not have autonomous counties. In the end, only fifteen provinces are included in these two analyses. Because most of the bonus points are numbers â&#x20AC;&#x201C; for example 5, 10, 20 and most of the provinces have 750 total points in NCEE â&#x20AC;&#x201C; the levels of percentage of bonus points are limited for these fifteen provinces. In fact, eight out of the fifteen provinces, more than half,

that in general, the average student enrollment rate in regular secondary schools is higher in regular counties than in autonomous counties. Similarly, average GDP per capita is higher in regular counties than in autonomous counties. However, when paying attention to the right side of the graphs, it is hard to discern any pattern. Sichuan province still stands out in the graphs due its incredibly generous preferential policy. On the negative side of the x-axis, Qinghai province is the outlier in Figure 2. Out of the seven autonomous counties in Qinghai, five of them are designated Hui and Tu autonomous counties. Sun

award 1.33% bonus points in the exam. There are only five different percentages of bonus points in total for these provinces. Thus, it is very hard to draw reliable conclusions from Figures 2 and 3 below. The fact that most data points fall in the right-hand side of the y-axis does meet my expectation. This means

(2012) mentions Hui as one of the ethnic minorities that has better education attainment then the nationâ&#x20AC;&#x2122;s average. Thus, part of the reason why Qinghai has a negative value may be that Hui and Tu students have better access to secondary education than regular counties or their families are more willing to

Figure 2. Percentage of Bonus Points out of Total by Percentage of Difference in Student Enrollment between Regular and Autonomous Counties

Note: N=15 Provinces Sources: Chenwu (2014), China Data Center (2000a, 2000b)

Influence of Ethnic Minority Demographics

send them to schools. Since Qinghai province has a very high percentage of ethnic minority population (45.97%), the bar is set relatively low compared to a province that has more than 90% of Han population with a very high enrollment rate. The data prove this point because Qinghai actually has the lowest regular county enrollment rate among the fifteen provinces. The average regular county enrollment for all provinces included is 3.43 for every 10,000 people, whereas Qinghai’s enrollment is 0.89 for every 10,000 people. For the autonomous counties, Qinghai’s enrollment (1.71 per 10,000 people) is only slightly

in percentage is -21.70%. Neither province has a big ethnic minority population (Jilin 9.15%, Hubei 4.36%). Both autonomous counties in Hubei province are designated as Tujia Autonomous County, which means Tujua people are slightly better off economically than the general population in Hubei. For Jilin, the three autonomous counties are Manchu, Mongol, and Korean. The first two minorities are mentioned by Sun (2012) as exceeding the national average higher education attainment, and Koreans have the anecdotal reputation of being well-educated, too. Considering the widely recognized reciprocal relationship between

above the average of all provinces (1.67 per 10,000 people). In Figure 3, the negative outliers are Jilin and Hubei province. Hubei is really close to the borderline where the calculated difference in GPD per capital in percentage is only -2.70%, whereas Jilin is a much more extreme case because its difference

education and financial success, it is not surprising that the value for Jinlin is negative. Figure 4 presents a strong positive relationship between percentage of a province’s ethnic minority population and the number of types/levels of bonus points a province offers. If ethnic minorities constitute a higher percentage of a province’s population, that

Figure 3. Percentage of Bonus Points out of Total by Percentage Difference in GDP per capita between Regular and Autonomous Counties

Note: N=15 Provinces Sources: Chenwu (2014), China Data Center (2000a, 2000b)

20 province is more likely to create detailed preferential policies that award several different bonus points amounts, distinguishing students who are from different minority groups, different cities/townships/ counties, rural/urban areas, or who use Mandarin/ traditional ethnic language in NCEE. Provinces employ various criteria to subdivide their ethnic minority applicants. Unlike Figure 1, one can spot a clear trend even in the lower left corner in Figure 4. Overall, Figure 4 provides support for Hypothesis 4. However, this figure has the same outliers as Figure 1, with Sichuan province in the upper left corner, and Xizang (Tibet) Autonomous Region in the lower right corner. It is very interesting that two figures present the same outliers – Sichuan and Tibet. Again, Hypothesis 1 and 4 share independent variables – percentage of province’s ethnic minority population, while having different dependent variables – percentage of bonus points granted and number of types/levels of bonus points granted. Demographically, Tibet has the highest

ethnic minority population percentage 93.93%, most of them Tibetan, leading the second place Xinjiang Uyghur Autonomous Region by more than 30%. It shares very little with the other Chinese provinces regarding culture. The Tibetan culture heavily relies on Buddhism. However, unlike the rest of China which is mainly influenced by Han-Chinese Buddhism; Tibet is dominated by the Tibetan Buddhism division. Religion is a much more important part in life for Tibetans than for average Chinese. Not surprisingly, the most respected people there are monks. As previously mentioned in the literature review section, some scholars contend that ethnic groups develop unfavorable attitudes toward education if they perceive that the school system is incompatible with aspects of their own cultures (Hansen 1999; Harrell and Mgebbu 1999). Hansen (1999) points out the problem precisely: even in areas where minorities are heavily concentrated, school education is still entirely based upon Chinese history and mainstream values, allowing very little if any room for distinct

Figure 4. Number of Types/Levels of Bonus Points by Percentage of Ethnic Minority Population

Note: N=30 Provinces Sources: Chenwu (2014), China Data Center (2000a)

Pi Sigma Alpha Undergraduate Journal of Politics languages, histories, religions, and cultural values of ethnic minorities. Therefore, regular education is not a very appealing choice. Many Tibetan families prefer to send their sons to monasteries than to let them go to regular secondary schools (Bass 1998; Kolas and Thowsen 2012). In 2000, when compared to other provinces, Tibet had the lowest regular secondary school enrollment (5.52 per 10,000 people), the fewest secondary schools and higher education institutions in China (0.55 and 0.66 for 10,000 people respectively). All these indicate that not many Tibetan students actually take the NCEE to pursue further education, making little sense for the provincial government to develop generous or detailed preferential policies. Sichuan province, the other outlier, has a relatively low ethnic minority population (5%), but offers ethnic minority students very high numbers of bonus points, and has a very detailed preferential policy (Chenwu 2014). Sichuan is unique because despite its small minority population, there are three autonomous prefectures and four autonomous counties within the jurisdiction of the province. It has more ethnic administrative units than any other province that has a similar percentage of ethnic minority population. The three autonomous prefectures (Garzê Tibetan Autonomous Prefecture, Liangshan Yi Autonomous Prefecture, Ngawa Tibetan and Qiang Autonomous Prefecture) cover more than 60% of the land in Sichuan (Sichuan Yearbook Office 2014). Thus, the ethnic minority population in the autonomous administrative units is more spread out than usual. The geographical distribution may result in additional difficulty in providing access to education for ethnic minority students. This may be the reason why Sichuan provincial government decided to offer very generous bonus points to ethnic minority applicants. Some might state that the fact of Tibetans constituting a large portion of the minority

population renders Sichuan government’s policy counter-intuitive given my argument on why Tibet is an outlier – Tibetans generally lack the interest in higher education. However, it is worth pointing out that while Tibetans are a prominent minority group is Sichuan, they only count for 30.81% of Sichuan’s ethnic minority population; in Tibet, 98.76% of the minority population is Tibetan (China Data Center 2000a). The much higher level of diversity in Sichuan’s minority population could justify provincial government’s policy making. For Hypothesis 4, Sichuan’s preferential policy is more detailed because it has separate clauses for ethnic minority students in different geographical areas, and for each clause, it offers different levels of bonus points for students who are applying to firsttier universities (yiben yuanxiao/benyipi yuanxiao 一本院校/本一批院校) and non-first-tier universities (Chenwu 2014). The wide geographic spread of the ethnic minority population can explain why Sichuan provincial government wants to establish clear standards on students in which areas get bonus points. Interestingly, Sichuan is the only province that differentiates the level of bonus points based on whether or not the student is applying to first-tier universities. Specifically, first-tier university applicants receive lower bonus points than other applicants. However, admission to first-tier universities is the most competitive in terms of the NCEE score required. This practice could be Sichuan provincial government’s effort to offer preferential treatment to ethnic minority applicants but still maintain some fairness in the most competitive part of the admission. Because Tibet is identified as an outlier, and Gansu and Qinghai were part of the historic homeland of Tibetans, it is natural to pay closer attention to the results of these two provinces. In Figures 1 and 4, Gansu is in the lower left corner; it does not demonstrate uniqueness like Tibet

Sun

does. This should not be surprising given the small minority population in Gansu nowadays (8.75%). Qinghai, on the other hand, has a very high percentage of ethnic minority population (45.97%). It is harder to explain why Qinghai provincial government did not behave the same way the Tibetan government did, making simple preferential policies, due to the fact that Tibetans account for almost half of the minority population (49.01%). It could be because there is almost a quarter of the population in the province that are non-Tibetan minorities. Conclusion This research explores the relationship between demographic characteristics of ethnic minorities in provinces and the preferential policies towards ethnic minority students in Chinese college admission system, in the form of bonus points awarded in the NCEE. The results suggest that when a province has a larger percentage of minority population, it is likely to offer greater percentage of bonus points out of the total points. It also tends to have a more detailed policy that includes more types and levels of bonus points. Nevertheless, Sichuan and Tibet are the two outliers in these analyses. The potential reasoning for Tibet being an outlier is that the cultural incompatibility makes regular education a less attractive path for Tibetan students. Detailed and generous preferential policies might be useless, so the government did not create them. The wide geographical spread of ethnic minority population may be the reason that Sichuan is an outlier indicating that the provincial government wants to compensate for the difficulty of providing access to education for ethnic minority students added by their geographical spread. The provincial government’s effort to maintain some fairness in the admission of the most competitive universities also contributes to the elaborate policy. The other two hypotheses are not

supported based on the figures. How much the student enrollment rates in regular secondary schools and the GDP per capita in regular counties are higher than autonomous counties do not seem to affect provincial government’s preferential policy making. One important limitation of this study is the time consistency of the data sources of different variables. The dependent variables, percentage of bonus points out of the total points and types or levels of bonus points, are generated according to policies in 2014, while all the independent variables are in the year of 2000 due to the difficulty of finding comprehensive 2010 Census data. However, the preferential policies towards ethnic minorities remain relatively stable over the years, unlike preferential policies towards awards recipients of academic competitions or certified national standard athletes, due to central government’s firm stand on the importance of these remedial policies. Therefore, this study can still provide some insights on why provinces have different preferential policies towards ethnic minorities in college admissions. It is also worth pointing out that this is a specific case study of the Chinese college admissions system. The results of this study are not likely to be generalizable to other populations or countries due to China’s uniqueness as a country, a state, and a nation. One might also raise the question that whether different ethnic groups across provinces receive different treatment or whether geographic spread accounts for different treatment across provinces. In other words, is the policy ethnic-sensitive or geographic distribution-sensitive? Unfortunately, it is hard to provide one single answer for all the provinces. First, the two criteria mentioned are usually intertwined with one another. Second, provinces/ ARs really employ different approaches to address the bonus points issue. Some divided the clauses based on

Influence of Ethnic Minority Demographics specific ethnic identities, and some divided the clause based on geographical boundaries. Therefore, it is difficult to reach a specific conclusion. Given that the Chinese government’s firm stance on having remedial policies towards ethnic minority students is unlikely to change in the near future, these preferential policies will remain for a period of time, but not necessarily without any small adjustment. Therefore, it is crucial to understand what may affect provinces’ preferential policy

making. Future studies can explore the impact of other provincial characteristics on preferential policy making. This paper only focuses on one specific aspect of preferential policies – bonus points. Other forms of preferential policies towards ethnic minorities mentioned in the introduction section also need to be examined closely. Due to the lack of scholarly works on this topic that are supported by data, there is a lot of room for researchers to narrow the gap between tentative theories and actual circumstances.

Appendix Table 1. Provinces’ Data on Bonus Points Percentage and GDP per capita of Regular and Autonomous Counties

Province

Countries General Information GDP per capita (dollars) Bonus Points Minority DifferPercentage Regular Population in Numerical ence in out of To- County Autonomous Autonomous Regular Autonomous Difference Percenttal Points Count County Count Counties Counties Counties (r) age (r%)

Hebei

1.33%

132

66.33%

6944.73

4242.17

2702.56

38.92%

Liaoning

1.33%

61.99%

6716.67

4877.79

1838.88

27.38%

Jilin

1.33%

22.04%

5361.23

6524.72

-1163.49

-21.70%

Heilongjiang

1.33%

19.65%

4827.84

3970.71

857.13

17.75%

Zhejiang

1.33%

10.79%

12500.51

4615.38

7885.12

63.08%

Hubei

1.33%

61.99%

5167.34

5306.78

-139.44

-2.70%

Hunan

2.67%

72.26%

4286.73

3145.18

1141.54

26.63%

Guangdong

1.33%

36.51%

7720.34

3645.81

4074.52

52.78%

Hainan

2.22%

50.73%

5684.52

4208.78

1475.74

25.96%

Chongqing

1.33%

69.61%

3687.83

2036.17

1651.67

44.79%

Sichuan

6.67%

137

47.90%

3893.51

2450.32

1443.19

37.07%

Guizhou

2.67%

62.40%

2400.82

1559.84

840.98

35.03%

Yunnan

2.67%

58.95%

3362.17

2527.51

834.66

24.83%

Gansu

2.67%

62.79%

2458.29

1406.58

1051.72

42.78%

Qinghai

4.67%

53.28%

3891.03

2947.37

943.66

24.25%

Average:

50.48%

5260.24

3564.34

1695.90

29.12%

Note: N=15 Provinces Source: Chenwu(2014), China Data Center (2000a, 2000b)

24 Table 2. Provincesâ&#x20AC;&#x2122; Data on Minority Population, Bonus Points Percentage and Bonus Points Type/Level Bonus Points PerEthnic Minority centage out of Total Bonus Points Type/ Province Population Points Level Count Shanxi 0.32% 0.00% 0.5 Beijing 4.31% 0.67% 1 Tianjin 2.71% 0.67% 1 Anhui 0.67% 0.67% 1 Henan 1.25% 0.67% 1 Shanghai 0.63% 0.83% 1 Hebei 4.35% 1.33% 2 Inner Mongolia 20.83% 1.33% 3.5 Liaoning 16.06% 1.33% 2.5 Jilin 9.15% 1.33% 2 Heilongjiang 4.89% 1.33% 2 Zhejiang 0.85% 1.33% 1 Fujian 1.71% 1.33% 1.5 Shandong 0.70% 1.33% 1 Hubei 4.36% 1.33% 1 Guangdong 1.49% 1.33% 1.5 Chongqing 6.47% 1.33% 2.5 Xizang(Tibet) 93.93% 2.67% 2 Shaanxi 0.50% 1.33% 1.5 Jiangsu 0.35% 2.08% 2 Hainan 17.38% 2.22% 3.5 Jiangxi 0.31% 2.67% 2 Hunan 10.13% 2.67% 3 Guangxi 38.37% 2.67% 4 Guizhou 36.56% 2.67% 2 Yunnan 33.41% 2.67% 6 Gansu 8.75% 2.67% 1.5 Qinghai 45.97% 4.67% 5 Ningxia 34.56% 2.67% 4 Sichuan 5.00% 6.67% 5.5 Xinjiang 59.42% 6.67% 4 Note: N=31 Provinces Source: Chenwu (2014), China Data Center (2000a)

Pi Sigma Alpha Undergraduate Journal of Politics

Table 3. Provinces’ Data on Bonus Points Percentage and Student Enrollment in Regular Secondary Schools of Regular and Autonomous Counties

Province

Student Enrollment in Regular Secondary Countries General Information Schools (out of 10,000 people) Bonus Points Minority DifferPercentage Regular Population in Numerical ence in out of To- County Autonomous Autonomous Regular Autonomous Difference Percenttal Points Count County Count Counties Counties Counties (r) age (r%)

Hebei

1.33%

132

66.33%

3.520542

3.261609

0.258934

7.35%

Liaoning

1.33%

61.99%

3.073269

2.208274

0.864995

28.15%

Jilin

1.33%

22.04%

3.312769

2.773078

0.539691

16.29%

Heilongjiang

1.33%

19.65%

3.178031

1.510000

1.668031

52.49%

Zhejiang

1.33%

10.79%

4.026878

0.820000

3.206878

79.64%

Hubei

1.33%

61.99%

4.731167

2.379582

2.351584

49.70%

Hunan

2.67%

72.26%

4.761865

1.709365

3.052499

64.10%

Guangdong

1.33%

36.51%

5.80429

0.998078

4.806212

82.80%

Hainan

2.22%

50.73%

3.050794

1.736171

1.314623

43.09%

Chongqing

1.33%

69.61%

4.86783

2.245269

2.622561

53.88%

Sichuan

6.67%

137

47.90%

3.868035

0.385054

3.482981

90.05%

Guizhou

2.67%

62.40%

2.469349

1.468862

1.000487

40.52%

Yunnan

2.67%

58.95%

2.071385

1.237729

0.833656

40.25%

Gansu

2.67%

62.79%

1.79589

0.645683

1.150207

64.05%

Qinghai

4.67%

53.28%

0.893931

1.709197

-0.81527

-91.20%

Average:

50.48%

3.428402

1.67253

1.755871

41.41%

Note: N=15 Provinces Source: Chenwu(2014), China Data Center (2000a, 2000b)

References Bass, Catriona. 1998. Education in Tibet: Policy and Practice since 1950. London : Zed Books; Distributed in the US by New York: St. Martin’s Press. Chenwu. 2014. “Qianqibaiguai de Shaoshu Minzu Gaokao Jiafen Zhengce [The Various Strange Preferential Policies towards Ethnic Minorities

in College Admissions].” December 28. http:// blog.sina.com.cn/s/blog_53fe981c0102ve12. html (Accessed April 7, 2016). China Data Center. 2000a. “China 5th Nation Province Population Census Data in 2000”. http:// chinadataonline.org/ (Accessed April 7, 2016) China Data Center. 2000b. “China 2000 County & District Population Census Data Assemblies.”

Sun http://chinadataonline.org/ (Accessed April 7, 2016)

Coleman, James S. 1968. “The Concept of Equality of Educational Opportunity.” Harvard Educational Review 38(1): 7-22. Garms, Walter I., James W. Guthrie, and Lawrence C. Pierce. 1978. School Finance: The Economics and Politics of Public Education. Englewood Cliffs, N.J: Prentice-Hall. Gladney, Dru C. 1996. Muslim Chinese: Ethnic Nationalism in the People’s Republic. Cambridge, Mass.: Council on East Asian Studies, Harvard University; Distributed by Harvard University Press. Green, Thomas F. 1980. Predicting the Behavior of the Educational System. Syracuse, N.Y.: Syracuse University Press. Han, Jin. 2014. “An Empirical Study on Return to Education in Urban China.” Master’s thesis. Clemson University. Hannum, Emily. 2002. “Educational Stratification by Ethnicity in China: Enrollment and Attainment in the Early Reform Years.” Demography 39(1): 95-117. Hansen, Mette Halskov. 1999. Lessons in Being Chinese: Minority Education and Ethnic Identity in Southwest China. Seattle: University of Washington Press. Harrell, Stevan. 1995. Cultural Encounters on China’s Ethnic Frontiers. Seattle: University of Washington Press. Harrell, Stevan, and Lunzy Mgebbu.1999. “Folk Theories of Success Where Han Aren’t Always the Best.” In China’s National

Minority Education: Culture, Schooling, and Development, ed. Gerard Postiglione. New York: Garland Press, 213-241. Jalali, Rita, and Seymour Martin Lipset. 1992. “Racial and Ethnic Conflicts: A Global Perspective.” Political Science Quarterly 107(4): 585-606. Jin, Donghai, and Ailan Wang. 2007. “Guanyu Woguo Shaoshu Minzu Xuesheng Gaokao Youhui Zhaosheng Zhengce Ruogan Wenti de Sikao [Development, Characteristics, and Consideration on the Problems of Preferential Enrollment Policy of University Entrance Examination or Minority Students in China].” Xibei Shida Xuebao (Shehui Kexue Ban) 44(1): 85-88. Kolas, Ashild, and Monika P. Thowsen. 2012. On the Margins of Tibet: Cultural Survival on the Sino-Tibetan Frontier. Seattle: University of Washington Press. Kurosaki, Isao. 1995. Gendai Nihon No Kyōiku to Nōryoku Shugi: Kyōtsū Kyōiku Kara Atarashii Tayōka E [Japanese Education Today and Its Ability Centrism]. Tōkyō: Iwanami Shoten. Levin, Henry M. 1976. “Educational Opportunity and Social Inequality in Western Europe.” Social Problems 24(2): 148-72. Long, Yaohai. 2010. “Guanyu Woguo Gaokao Minzu Jiafen Zhengce de Sikao [Thoughts on the Ethnic Minority Bonus-Point Policy for National College Entrance Examination],” Minzu Luntan, 9: 10-11. Ma, Rong. 2009. “Issues of Minority Education in Xinjiang, China.” In Affirmative Action in China and the U.S.: A Dialogue on Inequality and Minority Education. International &

Influence of Ethnic Minority Demographics Development Education, eds. Minglang Zhou and Ann Maxwell Hill. New York: Palgrave Macmillan, 179-98. Mackerras, Colin. 1994. China’s Minorities: Integration and Modernization in The Twentieth Century. Hong Kong; New York: Oxford University Press. Mackerras, Colin. 1995. China’s Minority Cultures: Identities and Integration since 1912. New York: St. Martin’s Press. Min, Weifang. 1997. “China.” In Asian Higher Education: An International Handbook and Reference Guide, eds. Gerard Postiglione and Grace Mak. Westport, Conn.: Greenwood Press, 37-55. Mok, Ka-Ho. 2002. “Policy of Decentralization and Changing Governance of Higher Education in Post-Mao China.” Public Administration and Development 22(3): 261-273. National Bureau of Statistics of China. 2011. “Tabulation on the 2010 Population Census of the People’s Republic of China.” http://www. stats.gov.cn/tjsj/pcsj/rkpc/6rp/indexch.htm (Accessed April 23, 2016) Nwabuogu, Michael N. 1984. “On the Meaning and Application of Equal Educational Opportunity: A Review Article.” Journal of Education Finance 10(1): 64-82. Postiglione, Gerard A. 1999. China’s National Minority Education: Culture, Schooling, and Development. New York: Falmer Press. Postiglione, Gerard A. 2006. Education and Social Change in China: Inequality in a Market Economy. Armonk, N.Y.: M.E. Sharpe.

Sautman, Barry. 2010. “Scaling Back Minority Rights? The Debate about China’s Ethnic Policies.” Stanford Journal of International Law 46(1): 51120. Sichuan Yearbook Office. 2014. Sichuan Yearbook 2014. Sichuan: Sichuan Yearbook Office Sun, Ping. 2012. “Shaoshu Minzu Gaokao Jiafen Youhui Zhengce Cunzai Wenti ji qi Tiaoshi yi Qiandongnanzhou wei Gean [The Problems and Adjustments of Ethnic Minority Bonus Points Preferential Policy – A Case Study on Qiandongnan Miao and Dong Autonomous Prefecture.]” Master’s thesis. Southwest University. http://d.wanfangdata.com.cn/ Thesis/Y2085078 (Accessed April 25, 2016). Takase, Noriko. 1999. “Transition of the Concept of Equality of Educational Opportunity: The Case of High School Reform in Japan.” Ph.D. diss. University of Colorado at Denver. Tang, Ying. 2003. “Zhongguo Shaoshu Minzu Gaodeng Jiaoyu Jihui Jundeng Wenti Yanjiu [On Equal Opportunity of Higher Education for Chinese Ethnic Minorities.]” Jiaoyu yu Jingji 2: 17-20. Teng, Xing, and Xiaoyi Ma. 2009. “Zhongguo Gaodeng Jiaoyu de Shaoshu Minzu Youhui Zhengce yu Jiaoyu Pingdeng [Ethnic Minority Preferential Policies and Education Equality in Chinese Higher Education.]” Taiwan Yuanzhumin Yanjiu Luncong 5: 191-208. Ulich, Robert. 1965. Education in Western Culture. New York: Harcourt, Brace & World. Wang, Jing. 2010 “Gaoxiao Shaosheng zhong Minzu Youhui Zhengce Dailai de ‘Nixiang Qishi’ Wenti Yanjiu [A Study on the Reversed Discrimination from Minority Preferential

28 Policy in College Admissions.]” Hunan Yike Daxue Xuebao (Shehui Kexue Ban) 12(1): 192194 Wang, Tiezhi. 2007. “Preferential Policies for Ethnic Minority Students in China’s College/ University Admission.” Asian Ethnicity 8(2): 149-163. Wang, Tiezhi. 2009. “Preferential Policies for Minority College Admission in China: Recent Developments, Necessity and Impact.” In Affirmative Action in China and the U.S.: A Dialogue on Inequality and Minority Education. International & Development Education, eds. Minglang Zhou and Ann Maxwell Hill. New York: Palgrave Macmillan, 71-82. Wise, Arthur E. 1968. Rich Schools, Poor Schools; the Promise of Equal Educational Opportunity. Chicago: University of Chicago Press. Xiao, Jin, and Zeyun Liu. 2014. “Inequalities in the Financing of Compulsory Education in China: A Comparative Study of Gansu and Jiangsu Provinces with Spatial Analysis.” International Journal of Educational Development 39: 250263. Yan, Sesheng, and Meixia Li. 2012. “Fensheng Luqu Moshi xia Gaokao Shaoshu Minzu Jiafen Zhengce Tanxi [The Policy of Minorities’ Extra Scores under Provincial Enrollment Mode in College Entrance Examination.]” Zhongguo Kaoshi 7: 54-59. Zhou, Minglang. 2009. “Tracking the Historical Development of China’s Positive and Preferential Policies for Minority Education: Continuities and Discontinuities.” In Affirmative Action in China and the U.S.: A

Dialogue on Inequality and Minority Education. International & Development Education, eds. Minglang Zhou and Ann Maxwell Hill. New York: Palgrave Macmillan, 47-70.

A Counterweight to Populism: Judicial Legitimacy and Popularity in Latin America Mishella Romo Montclair State University Populism, a resilient phenomenon in Latin America, has enabled the inclusion of politically alienated masses. However, it has simultaneously used popular support to legitimize the weakening of institutions, such as the judiciary, that ensure a balance of power. This has been seen in populist leaders’ use of popularity to gather support for referendums to approve constitutional amendments, and replacements, which have historically increased executive power. Thus, populism’s force has shown to be resilient because of the support and legitimization it obtains from the citizenry. This research’s aim is to understand how court empowerment and independence is possible in the face of such a powerful anti-institutional force. This research argues that the answer may be found in the same mechanisms that enable populism: popularity and legitimacy. To make the case for the judiciary, an institutional approach is applied and the role populism has played in the Ecuadorian judiciary is examined. Then, the reforms of the judiciary and Public Defender’s Office in Brazil are analyzed followed by an examination of the judicial empowerment seen in Mexico by means of public relations. Both of these pro-judiciary efforts show how judiciaries, through direct interaction and communication with the populace, can build the necessary legitimacy and public support that populism diminishes.

Introduction

Since its inception, populism has appeared to

be a more responsive form of governability (Hellinger 2011, 421). With its ability to include those who have been alienated from politics, populist leaders have been particularly resilient in Latin America; a region with a historical record of weak citizenship rights and unequal enforcement of the rule of law (De La Torre 2010, 124). The people thus turn to a leader, and although populism has been able to respond to citizen demands for inclusion, it has been criticized for

and dismissed eight different presidents. Although the Ecuadorian people have shown a willingness to remove those who no longer represent their interests, they have also shown a continued and present susceptibility to legitimate populist leaders; leaders who paradoxically perpetuate the weakness of rule of law and the institutions the people feel alienated from in the first place, particularly the court.

This study alludes to broader issues regarding

political inclusion and the perpetual weak enforcement of rule of law that populism exacerbates in Latin America. Given its history with populism since the

diminishing the necessary independence of democratic 1940’s, Ecuador is an appropriate case to study (De La institutions (Conaghan and De La Torre 2008; De Torre 2010, 8). Since populism attacks the judiciary, La Torre 2010; Weyland 2013). Populism in Ecuador has not been the exception; however, it has been unique. From 1996 to 2006, the country welcomed

and disables it from independently and ambitiously defending the rule of law (Weyland 2013, 23), this

Romo

research seeks to understand how court empowerment

use constituent power to simultaneously weaken

and independence is possible in the face of the

judicial independence in countries like Ecuador.

powerful anti-institutional force: populism. This

Weyland (2013, 23) has argued that populists attack

paper argues that the answer may be found in the

the rule of law and establish arbitrariness in their use

same mechanisms that enable populism, specifically

of “discriminatory legalism,” which is the application

popularity and legitimacy. In order to provide evidence

of “legal authority in discretionary ways to promote

for the latter claim, it is important to conceptualize

their cronies and allies while punishing or intimidating

both populism and legitimacy, explore the detrimental

critics and opponents in politics and society.”

effect populism has had on Ecuador, and then

point to pro-judiciary efforts in Brazil and Mexico.

empowerment, the populist gains the people’s trust and

Brazil and Mexico have gathered the legitimacy and

support. This allows for his or her ability to appropriate

popularity through reforms for career safety and the

the will of such “people” (De La Torre 2010, 149). This

strategic use of public relations to increase power

appropriation is problematic since the notion of the

and independence. Both of these cases will show how

“people” is ambiguous and undefined, yet the populist

populism can be countered with increased judicial

uses this concept and the attached constituent power

legitimacy.

to concentrate excessive authority, and legitimate its

Populism

influence over institutions tailored to limit such power.

Most scholars conceptualize populism differently since it arises under different contexts and sides of the ideological spectrum (Weyland 2013). Nevertheless, Panizza (2016, 194) argues that populism has an “ineliminable core.” At the core is an “anti-status quo appeal that draws a political frontier between two antagonistic camps, ‘the people’ and its enemies or its ‘Other.’”1 This frontier arises when the unfulfilled demands of “the people” become one, and a “homogenizing function is carried out by a pure name: the name of the leader” (Laclau 2005, 38). Through the use of charisma, antagonistic rhetoric, and the creation of an unmediated relationship with the “people,” the leader includes the masses into politics (De La Torre 2010). This has been evident in their use of referendums, constituent assemblies, and constitutional replacements. Although these “include” the masses into politics, they ignore the need to repair weak enforcement of rule of law and instead

Nevertheless, through inclusion and

Populism, therefore, becomes a political strategy (Weyland 2001, 18). The populist strategy, therefore, works from the top-bottom and from the bottom-up. While the populist connects with, and includes the “people,” the “people” also empower the leader and support his or her actions (Hellinger 2011, 142).

Separate from these outcomes, two things

can be learned from this strategy. The first is that the use of “popularity” and “inclusion” increases political trust, institutional support, and legitimacy for the populist. Second, populism, irrespective of its negative tone and ambitious desire to consolidate power, is nonetheless a form of leadership. This leadership is in popular demand since it has been responsive to usually unheard constituent demands (Laclau 2005; Panizza 2005, 2009). Perhaps, since this leadership is what the citizenry desires, this “political strategy” is not what is detrimental to democracy per se, especially, if it can be inclusive. The danger lies in the

A Counterweight to Populism

populists’ appropriation of the people’s will to change

high support and trust. These data showed that out of

the political order (De La Torre 2010, 149). Hence,

the twelve institutions, the president enjoyed a third

the leader consolidates power to set up new legal

spot given his average of 62.8 points of support/trust,

frameworks that enjoy enforcement via the legitimacy

while the Court of National Justice landed a tenth spot

it gains from the populace. This excess concentration

with its average score of 41.8 points of support/trust.

of authority perpetuates a subservient judiciary to the

executive (Weyland 2013, 23).

is to elect judges. However a degree of the court’s

Legitimacy

normative legitimacy is based on the notion that the

Legitimacy is “the authority to govern.” The

source of this right, however, comes from the people (Caldeira and Gibson 1994, 357). For this research, legitimacy is understood as a normative concept, as people delegate power to an institution because they have a deep belief in its role (Madeira 2014, 55). Therefore, there is trust, and what results from this trust is a “diffuse support,” which is a willingness to support the institution when other branches attempt to weaken it (Caldeira and Gibson 1995, 363). Not only does the populist enjoy this legitimacy through elections, which already express public trust and a commitment for institutional support, but as the case of Ecuador will show, he or she uses it to attack judicial legitimacy.

Since the court lacks the “direct electoral

connection” with the people, legitimacy suffers (Caldeira and Gibson 1995, 356), and the court lacks public support and popularity. This notion is supported in data presented by the Barómetro de Las Americas for the Latin American Public Opinion Project (2012) which surveyed and measured Ecuadorians’ public support and trust of twelve institutions in the country (Barómetro de Las Americas 2012, 123-124).2 After surveying respondents, the Barómetro de Las Americas scaled the responses as points from 0-100, with 0 points indicating low support and trust and 100 representing

It may be tempting to argue that the solution

court is legitimate because it is politically insulated, and is instead committed to the law as well as its fair application (Staton 2010, 6). However, this normative construct of court legitimacy in Latin America is currently under attack since the populist controls the decision-making. At times, the court has even shown to be an empowered political agent that is simultaneously unable to assert its independence or serve its purpose as the fair protector of the law (Weyland 2013, 123). Under these circumstances, someone will always be above the law, among these actors is the president.

The court must consequently rebuild its

popularity and its normative legitimacy to defend its independence. This research constructs such popularity and normative legitimacy in two ways: internally and externally. Horan and Meinhold (2012, 236) characterize “internal legitimacy” as the court’s ability to deal “with performance problems, such ensuring there are credentials of the justices, physical resources of the court, or an increase in professionalism.” On the other hand, “external legitimacy” is built through the court’s “relationships with other institutions and citizens” (Horan and Meinhold 2012, 236). These forms of legitimacy could increase public trust, and the court’s popularity. This in turn can construct the institutional commitment by the people to defend its role and independence. This empowerment could enhance the consolidation

32 of democracy that Ecuador seeks. A consolidation in which the court is able to imitate the responsiveness that populism creates, while also enabling the fairness that the rule of law should ensure. The Ecuadorian Experience: How Populism Affects the Judiciary Given that Ecuador has had eight presidents from 1996 to 2006, it is accurate to say that they have experienced what institutionalists call “interrupted presidencies.” Despite its institutional instability and encounter with populism, Ecuador remains an electoral democracy capable of change and reform. Some scholars even argue that parliamentary government would be a better alternative, because by design its organization of power is less conflict prone (Linz and Valenzuela 1994; Valenzuela 2004, 17). As Valenzuela (2004) acknowledges, however, Latin America has become permanently comfortable with modeling the U.S. presidential system. Consequently, Valenzuela (2004, 18) suggests that Latin America should have the prudence to design a system that best represents their context and circumstances. Other scholars follow this argument, but write that “reform” is better than a break from the presidential system altogether (Mainwaring and Scully 2010). Being mindful of the unlikeliness of an institutional breakaway, what seems more likely is institution building, and this paper does so through the construction of legitimacy and popularity for the court. Institution building is crucial for a country like Ecuador, which has historically and is currently under the leadership of a populist leader. In 2011, President Rafael Correa pushed for a referendum for a constitutional amendment that would “reform” the judiciary among other things. The reform created a

three-member Transitional Council of the Judiciary who would elect the new five member “Council of the Judiciary,” which would then be in charge of appointing judges for both the National Court of Justice and lower courts. Despite the illusion of reform, little changed for the better. The transitory council was not free of executive influence since it was elected by the executive/legislative branches (which the executive controlled) and a citizen’s branch. The president had direct influence in the election of the council and Human Rights Watch (2014a, para. 6) consequently mentioned that this process lacked “objectivity.” Human Rights Watch (2014a, para.14) also noted that “during the 18 months of the Transitional Council of the Judiciary, its members appointed 616 judges, suspended 182, and removed 244.” This accounts for the removal or suspension of more than 65% of judges. The Council justified most of these removals by arguing that judges had “violated a vaguely worded article in the Organic Code of the Judicial Function that forbids judicial officials to act with criminal intent, evident negligence, or inexcusable error” (Human Rights Watch 2014a, para.15). Conveniently, the Council continuously terminated judges based on judicial interpretation that they vaguely justified as this inexcusable error. Given this vague justification for removal and his newly established influence over the new council, President Correa was, and is currently able, to pick and choose when to remove judges that do not acquiesce with his agenda. According to a Due Process of Law Foundation (DPLF) report authored by legal scholar Luis Pásara, these judges have ultimately been removed because they have “failed to comply with the policies demanded by the executive branch” (Pásara 2015, 7). This has ultimately weakened “judicial independence” (Human Rights Watch 2014b, para. 1).

Pi Sigma Alpha Undergraduate Journal of Politics After more than 65% of removals, judges think twice before publishing an opinion that is antigovernment, instead they will prudently decide cases. But how could this attack on the judiciary not face scrutiny? Although this amendment further weakened checks and balances in the country, it was not subject to scrutiny because the president pursued the reform through the strategic use of a popular referendum. In 2011, Ecuador’s National Electoral Council reported that the judicial reform passed with a 52% approval for “yes” (Consejo Nacional Electoral 2011, 3 and 7). Therefore, when facing international criticism for his actions, President Correa noted “they will say we want to get our hands on the courts. Yes, we want to get our hands [on them]. For the sake of the Ecuadorian people!” (Human Rights Watch 2014a, para 20). This exposes the populist’s “appropriation of the people’s will,” that De La Torre (2010, 149) warns about, and also how such appropriation occurs. Although “the people” approved a referendum, it was for reform, not for a further weakening of the court. Nevertheless, through antagonizing discourse and populist mechanisms that include the masses into politics, such as a referendum, President Correa was able to pack the Judicial Council, fire judges who did not acquiesce to his agenda setting, and shield himself from accountability for his assault on the judiciary’s independence because of the legitimacy stemming from the referendum. Ecuador’s case and the Ecuadorian people’s involvement in this “reform” of the judiciary shows two contradicting results. One is a distrust and alienation from the court that is exacerbated by populism. Second, there is still an underlying normative belief in the court’s role (Madeira 2014). The judiciary’s biggest hurdle is to gain acceptance and public trust, especially because this “strategy of prudence” in decision-making

caused by the “inexcusable error” policy (Human Rights Watch 2014a, para.15) in Ecuador cannot construct judicial power in the long run (Staton 2010, 17). Consequently, the court needs to rehabilitate its power and independence through popularity and legitimacy, since this entails a “deep belief among a population in the court” (Caldeira and Gibson 1995; Staton 2010, 13). This legitimacy must be acquired at the internal and external level, as it must prove its ability to respond to citizen demands from the topbottom and from the bottom up, as populism does. Theoretical Argument This research constructs legitimacy internally and externally because it is a practical way of affirming the “normative” essence of the court’s legitimacy that Madeira (2014, 55) discusses, but it does so in a manner that is less abstract and best resembles populism’s political strategy. It directly shows how to build power through popularity, and the public trust for institutional commitment that populism champions. Unlike the populist, however, the judiciary’s internal and external legitimacy will rely on constituent power to create independence, not destroy or abuse it. Internal legitimacy will be evident in the case of Brazil. The independence of the judges and Public Defender’s Office in Brazil, an arm of the judiciary, will show that judicial empowerment can be made possible through bottom-up politics. Here, the civil society groups, as well as members within the judiciary pressured the government to pass a constitutional amendment (Emenda Constitucional No. 45 2004) that would guarantee career security for judges and public defenders (Madeira 2014, 53 and 55). Thus, through bottom up politics, legal professionalism free legal representation for citizens was established. This created

Romo

a starting point at the most basic level in the system, which increased the public trust and institutional commitment necessary for potential independence. Subsequently, the case of Mexico will show how external legitimacy can be created. The Mexican courts have constructed judicial power through the creation of direct relationships with the citizenry through means of public relations (Staton 2010). Using strategic public communication to increase popularity, awareness, transparency, and education, the Mexican court shows how direct interaction is possible and able to increase legitimacy externally.

attacking the court’s independence and power, the populist concentrates power, and disables the equal and fair enforcement of the rule of law (Weyland 2013). An empowered and independent court creates a balance of power, protects the rule of law, and diffuses the responsibility of decision-making, thereby creating a more stable government and policies (O’Donnell 1994, 62). In the broader sense, an empowered and independent court is especially important because it establishes the stability necessary to curb the previous political instability that has

This is especially important because it allows citizens to politically identify with the court as an institution that can be popular and inclusive. Although the court’s justices cannot be elected, the use of public relations can breed a public enforcement model that help a populace identify with a court they sometimes feel alienated from, and consequently, express institutional commitment for it (Staton 2010, 14). This will enable compliance of rulings, which would affirm judicial independence (Staton 2010, 25).

been manifested in interrupted presidencies, or the resilience of populism in countries like Ecuador. Horizontal accountability and rule of law therefore allow for a quality of democracy (Kapiszewski and Taylor 2008, 279; Linz and Stepan 1996, 10).

Why does the Court Matter?

institutional actions. The purpose is to understand

The cases of Brazil and Mexico will show how the court can empower itself to attain the role necessary to enable what O’Donnell (1994, 61) characterizes as “horizontal accountability.” This horizontal accountability acts as a check on the executive that keeps a democracy from becoming “delegative,” where the ruler governs as “he or she sees fit” (O’Donnell 1994, 59). Horizontal accountability allows for a “republican dimension of democracy [which creates] a careful distinction between the spheres of public and private interests of office holders” (O’Donnell 1994, 61). This differentiation is important because without it even the inclusion that populism enables in Latin America becomes self-defeating. In

how to strengthen the judiciary in a country, such as

Methodological Approach

This research applies an institutional approach

of analysis and examines how populism, and its use of constituent power, can combine to legitimize its anti-

Ecuador, with a chronic susceptibility to populism (Conaghan and De La Torre 2008; De La Torre 2010). As was previously mentioned, Ecuador’s continuous encounter with populism is interesting in that it shows Ecuadorians’ susceptibility to populism, but also their potential and political will to mobilize in favor or against an actor who will represent their interests.

The literature review below discusses how

the relationship between populism and constituent power – even at the extremes – is still under the jurisdiction of a court that holds a degree of authority. This is done through a look at populism at such a degree in Venezuela. The review shows how the court

A Counterweight to Populism is therefore still a necessary actor, but an actor which

363) necessary for Ecuador’s judiciary to counter

can be empowered (i.e., able to decide a case and enjoy

populism through the aid of constituent power.

compliance of its ruling), yet still lack independence. As Madeira (2014, 55) mentions, sometimes the judiciary is granted power only so that it could fulfill the proposed agenda of another actor, such as the executive. Scholars have already begun to explore the conceptual difference between judicial power and judicial independence (Kapiszewski and Taylor 2008, 745, 750; Staton 2010, 9). This paper will follow this analysis in order to underscore the importance of building legitimacy. Only legitimacy building can increase both the power and independence necessary for the court to counter populism’s resilient dominion over it and its abuse of the rule of law. Case Selections: Brazil and Mexico

Subsequently, this research points to the

successful pro-judiciary efforts in Brazil and Mexico to increase legitimacy internally and externally. Similar to Ecuador, Brazil and Mexico also share the important similarity and independent variable of being “presidential systems.” Therefore, these countries are practical for comparative analysis as “most similar cases” (Seawright and Gerring 2008, 304). However, Brazil and Mexico are also “influential cases.” The “goal of this style of case study is to explore cases that may be influential vis-à-vis some larger crosscase theory” (Seawright and Gerring 2008, 303). This larger-cross case theory is on “legitimacy” and the belief that individuals delegate power because there is a normative element of trust in the institution (Caldeira and Gibson 1995). The aim is to use the internal and external legitimacy present in Brazil and Mexico to understand the potential for the growth of the legitimacy which engenders the popularity and “institutional support” (Caldeira and Gibson 1995,

Literature Review The literature review particularly focuses on populism’s extreme relationship with constituent power in Venezuela, ultimately exposing two things: First, “the people” as a concept is not clearly defined by the populist, but in spite of the nebulous nature of “the people,” there is power behind the use of this concept. Second, even when the relationship between the populist and constituent power is at the extreme, it still relies on court opinions to legitimize the will that is being appropriated or imposed. Examining populism in the extreme gives better insights into “what populism is and is not” (Spanakos 2008, 543). This consideration will help explain the limits of populism, and thus, why it can be challenged in Latin America through the same mechanisms it relies on – legitimacy and popularity – both of which come from “the people.” Populism’s Political Frontier: Constituent Power and the Judiciary According to Panizza (2016, 196), “populism is about the constitution of popular identities, [and is] is best understood as a mode of political identification.” A political identification between these popular identities (i.e., the people) and the leader breeds a political relationship contingent of trust and charisma. When strong enough, this relationship enables the populist to construct a “political frontier,” which “tends to divide society into two irreconcilable camps” (Panizza 2016, 196). Therefore, the leader and the political identities are on one side, while the institutions that the populist antagonizes are on the other side. This political frontier may seem like a permanent state; however, Panizza (2016, 211)

36 acknowledges that these political identities (i.e., the people) and their support for an institution or a leader are subject to change because they are ultimately ambiguous groups. These identities (i.e., people) could also only form together to challenge the status quo momentarily (Hardt and Negri 2004; Panizza 2016, 201; Spanakos 2013). This mention is critical as it implies that the people can never truly be defined, because their organization is not permanent. It also suggests that populism is not powerful merely because a leader convinces a population to follow him or her, but also because it is a mechanism by which an undefined people can unite as an “institution” to have their political will enacted through such leader. As Hellinger (2011, 142) notes, power and charisma is therefore “something that an eager population projects on the leader.” More recent literature extends on the notion of constituent power as an institution which can legitimate and de-legitimate institutions (CiccarielloMaher 2013; Hardt and Negri 2004; Spanakos 2013; Spanakos 2016). Given that Ecuador’s interrupted presidencies occurred partly because the people pressured congress with protests demanding that their voices be heard, exploring this literature is necessary and relevant. As in Venezuela, Ecuador’s populace has used constituent power to remove those who no longer represent their interests (Basabe-Serrano and PolgaHecimovich 2013, 156). Spanakos (2016, 115) makes note of extreme constituent power and its relationship with populism in Venezuela and characterizes this form of power as a “polemic institutionality/ies,” which seek direct inclusion into political decision making and reject third party interference with the enactment of their will.3 This literature places populism on a separate context because it notes that polemic constituent

institutionalities “are [sometimes] explicitly ephemeral and measured by their ability to facilitate and enable citizenship [and change] in a specific sphere” (Hardt and Negri 2004; Spanakos 2016, 117). This suggests that populism is resilient because it does not solely focus on knowing a polity, but rather because it is able to include and know generally what such polities demand. Therefore, the people, and their constituent power, seem to favor the populist more, because the populist is not “adverse” to their mission to be included (Spanakos 2016, 114). Once populists and these polemic institutionalities unite, however, they can even override formal institutions, such as constitutions or judiciaries (Spanakos 2016, 117). This was evident in 1999 when President Chavez called for a two question4 Referéndum Consultivo [Consultative Referendum] which aimed to convene a Constitutional Assembly that would rewrite the 1961 Constitution (Consejo Nacional Electoral 1999, 1). This was problematic because neither the president nor the people were the “constitutionally appointed agent[s]” able to initiate the process of changing the Constitution (Spanakos 2016, 118). As per the Constitution of the Republic of Venezuela (Art. 245 1961), the legislature had this right. As a result, the constitutionality of the referendum was taken to Supreme Court of Justice (Ibáñez 2014). With the court’s favorable interpretation of the call for the referendum, a common populism mechanism, was made constitutional (Opinion No. 17 1999). The court held that “constituent power” is “previous and superior to an established juridical order” (Opinion No. 17 1999, 9-10). Spanakos (2016, 117) mentions that the court’s ruling here established that the “invocation of the will of the sovereign people empowered the president to take on procedures

Pi Sigma Alpha Undergraduate Journal of Politics constitutionally reserved for the Congress,” and that is what also “[allowed] for [a] constitutional ruling which was [also] contra to the constitutional article 250 whose subheading proclaimed ‘the inviolability of the Constitution.’” Hence, the court became an instrument of interpretation, which interpreted the case in a way which formalized “constituent power” as a force that is “supra-constitutional” (Ibáñez 2014, para. 7). Therefore, it placed the referendum set forth by the highly populist leader and the empowered popular will above the established legal order (Spanakos 2016, 117). Despite this extreme case in which constituent power adamantly rejected a third-party “arbiter,”5 the people still relied on this third party (i.e., the court) as an “interpreter” to enable their will. This shows that although “the people” may hold the power, they do not always hold the political authority. Despite the fact that this literature describes populism in Venezuela at an “extreme moment,” it is helpful in that it shows the relationship between populism, constituent power, and the court. Even in this extreme moment where constituent power and populism triumphed with the referendum and the subsequent constituent assembly (Consejo Nacional Electoral 1999, 1), the court was still a necessary actor in politics. An actor which through procedure and interpretation stamped the “sovereignty” and “will” of these “polemic institutionalities,” the populist, and the constituent power with a formal seal of approval. Thus, research on how to empower its legitimacy is necessary. A question that follows is: why courts, even though they are critical as was evident in Venezuela, act so “prudently” in their decision-making? Perhaps the obvious answer is that the courts did so because they have been highly politicized and under pressure of populist forces. But, also because of their fear of non-compliance (Staton 2010). The problem of non-compliance has yet to be resolved.

One thing is important, however: even in extreme cases the “procedure of interpretation” is crucial to the enhancement of populist interests and extreme constituent power. Thus, the court always holds a certain degree of legitimacy. It is populism that weakens legitimacy and attacks public trust. It is only the constituency who holds to power to change this and legitimize the court again. Although scholars in comparative politics have conceptualized populism and identified the institutional repercussions of it, such as the weakening of a judiciary (Conaghan and De La Torre 2008, De La Torre 2013; Weyland 2013; O’Donnell 1994; Panizza 2009), few have considered alternatives in which these same democratic institutions can counter this force. Some scholars have even only examined legitimacy in presidential systems in a “dual context,” namely between the executive and legislative (Linz and Valenzuela 1994, 6). As a result, scholars have ignored the important role that court legitimacy can play for its necessary empowerment and independence. This may be because the court is not seen as a strong political actor that can counter populism. The literature review, however, has shown that the court is necessary. In addition, the empowering and independence of the court must be considered as it enables for opportunities that even populism cannot retain, namely long-term stability, fairness, and accountability (O’Donnell 1994). How does the court prove that their role is capable? The cases of Brazil and Mexico speak to this problem. Case Study I: Brazil Building Internal Legitimacy: A Bottom-Up Strategy for Legal Professionalism

The case of Brazil shows that the judiciary and

its structure can be inclusive from the bottom-up. This

Romo

process reflects “internal legitimacy,” which Horan

establish autonomy and “career security” as a legal

and Meinhold (2012, 236) defined as the court’s ability

guarantee for public defenders to work freely (Madeira

to deal “with performance problems such credentials

2014, 55). The laws in Brazil (Complimentary Act No.

of the justices, the physical resources of the court…

312, Art. 4; Constitution of the Federal Republic of

or the increase in professionalism.” However, this

Brazil, Art. 134 1988; Emenda Constitucional No. 45

internal legitimacy (i.e. legal professionalism) is also

2004) were important because they were an addition

best established with the support of the populace.

to the career security that judges also enjoy from the

Through the mobilization of civil society ranging from

constitution (Constitution of the Federal Republic of

individual actors, to professionals such as lawyers and

Brazil, Art. 95 1988).

judges, deeply committed to democratization these

groups successfully pressured and demanded legal

public defenders not only diminish the amount of

action that ensured institutional strength (Madeira

external influence of the judiciary, but they increase

2014, 55). The 2004 Amendment No. 45 (Emenda

the quality of services since they help ensure a real

Constitucional No. 45 2004) formally established the

fiduciary relationship between the judiciary and

independence and autonomy of public defenders by

individual. This will increase the normative public

granting them a constitutional guarantee to career and

trust in the branch because it shows a commitment

financial security (Constitution of the Federal Republic

from the judiciary to be responsive to citizen needs.

of Brazil, Art. 134 §1 and §2 1988).

This defeats common assumptions that the courts

are often established elites that cannot be inclusive or

Taking it a step further, Complimentary Act

These constitutional guarantees for judges and

No. 312 was passed in Brazil as a supplementary

cannot create internal legitimacy. Instead, the Brazilian

law, which clearly stipulated the objectives of the

case shows the opposite – inclusion at the most basic

Public Defender’s Office. Among these objectives

level - through free legal defense and professionalism.

was the “affirmation of the democratic rule of law”

(Complimentary Act No. 312, Art. 4). Scholars

limitations. Despite the “de jure” independence

characterize these measures as “de jure independence,”

incorporated in the Constitution of the Federal

which is independence specifically granted by law

Republic of Brazil (Art. 95 and Art.134 1988) by

(Madeira 2014, 53-55). Thus, given the efforts of civil

Amendment #45 (Emenda Constitucional No. 45

society and professionals in the judiciary, the explicit

2004), the Public Defender’s Office still battles to attain

establishment of a legal professionalism (i.e., internal

consistent “de facto independence.” This means that

legitimacy) and committed citizen inclusion with the

the Public Defender’s Office has struggled to generate

guidance of the rule of law became possible.

“their own budgets, setting wages, creating posts

Increase of Internal Legitimacy: De Jure and De Facto

and holding competitive civil-service examinations”

Independence

The benefits of this bottom-up initiative from

regular citizens and judicial actors is that it pushed for the establishment of a legal framework that would

Madeira (2014, 63) admits that there are

(Madeira 2014, 57). Despite this, citizens have still been represented and included by the judiciary, which will remain stable and will continue to grow. In effect, “de jure autonomy, or gaining autonomy in legal terms, has the effect of driving change” even though it

A Counterweight to Populism is not entirely consolidated in the budgetary context

Applying the Brazilian Experience to Ecuador

(Madeira 2014, 63). De jure independence ensures stability. This stability is mainly attributable to the aforementioned initiatives that were made possible by civil society efforts and professionals in the judiciary.

To a certain extent this pro-judiciary belief has been partially present in Ecuador, namely because a constituent assembly created the Public Defender’s Office in the constitution of 2008 (Defensoria Publica

Public Trust Increasing due to Increased Internal

n.d.). Although the constituent assembly is less

Legitimacy

politicized than the usual “partidocracia,” which

governed the legislative branch, there were limits to the

Ultimately, “the independence of the Public

Defender’s Office is a way in which to formally capacitate the judicial branch” (Madeira 2014, 54). Ingram (2012, 440) supports the latter claim, arguing that the institutionalization of the judiciary is not just a “judge” related process. Instead, it is a way to empower the structure from the bottom to the top. Consequently, those who often felt alienated from politics now have felt included by the judiciary’s ensured legal defense and fiduciary commitment. This increases the trust necessary for the “institutional commitment” of the citizenry to the court that Caldeira and Gibson (1995, 357) note is important.

The increase of this trust driven by this

reform is evident in the response of an interviewee in Madeira’s (2014, 57) study when he says “‘it is the monopoly of the State to accuse’, you see… today we have a model of [one] that defends.” Clearly, through the increase in external legitimacy enabled by the creation of legal professionalism of the judiciary and the Public Defender’s Office, public perception of the judiciary has changed. This creates a firm foundation and is a direct example of democratic consolidation, and to a closer extent, of justice. A deep belief in the courts enables reform that can ensure legal guarantees for change. This results from the strengthening of the judiciary’s internal as well as external legitimacy. This alternative can help defeat populism, which consistently antagonizes and labels the judiciary as purely elite and non-inclusive.

judicial independence that came from the assembly’s creation of the Public Defender’s Office in Ecuador. Unlike Brazil, Ecuador has not established mechanisms to depoliticize this office for it to be entirely “autonomous” as the Constitution of the Republic of Ecuador (Art. 192. 2008) claims it to be. Although the assembly showed a “belief ” in the judiciary’s role in the 2008 constitution by establishing the Public Defender’s Office, there is no language in the Constitution of the Republic of Ecuador (Art. 192. 2008). There is also no amendment that provides the “de jure independence,” such as the one Madeira (2014, 53) described, and the one which Brazil established with career security. Given the aforementioned Ecuadorian situation with removal of judges for an “inexcusable error,” (Human Rights Watch 2014b, para.15) Ecuadorians should – through their “belief in the judiciary” – empower the court with similar legal guarantees, such as “career security,” that Brazil set forth. Mainly because this development would act as a critical shield from politics and populism.

In the context of “de facto” independence,

Ecuador has taken some steps to increasing funding every year (Defensoria Publica 2014a). However, there are 737 public defenders for a population close to sixteen million people (Defensoria Publica 2014b; World Bank 2014). Therefore, there is much left to be resolved. Nonetheless, Madeira (2014, 66) posits that “the institutionalization and implementation

40 of independence [takes] place in a slow and varied

considers the judiciary’s role to address popular will,

fashion.” The benefit of this is that Ecuador has a

but through direct communication and the rule of law.

democratic starting point, which can eventually lead to

the desired outcome seen in Brazil, specifically a more

The 2000 Zedillo case is described in further detail

consolidated judiciary.

below.), the Mexican court was granted constitutional

review powers (Staton 2010, 48). To enforce such

In noting consolidation, this research hints to

In the wake of the Zedillo reforms of 1994 (i.e.,

independence. Independence, such as the one seen in

power, the court devised a way to construct external

Brazil, is important because it prevents “arbitrariness

legitimacy through public relations. The court created

by governments,” like those of populists, and allows

a press office, which was responsible for keeping track

for more “social stability” (Madeira 2014, 53; Rios-

of coverage of its decisions in the newspaper. The

Figueroa and Staton 2014). All of these are a result

office not only held programs to connect with the

from the inclusiveness the judiciary and its organs

citizenry, and hand out pamphlets informing them of

can provide. Ultimately, de jure independence of

the internal judicial structure, but the president and

this judiciary’s organ can increase public trust. With

minister of the court the made personal efforts to meet

public trust comes public support, which can help

the legal and popular press (Staton 2010, 58). By doing

curb populist control over it. Separate from internal

this, the court “put a personal face on the Supreme

legitimacy, the Ecuadorian judiciary must always

Court” (Staton 2010, 58 and 61). In direct engagement

establish external legitimacy. In addition, it is critical

between ministers and the people, the court blazed

that the judiciary expand its popularity, especially if it

the trail for stronger relationships in which political

wants to continue to curb future populist attacks.

identification could be possible.

Case Study II: Mexico Constructing External Legitimacy: Judicial Popularity for Judicial Empowerment Public Relations and the Public Enforcement Model in Mexico

In order to further defeat the assumption that

courts are unable to engage directly with citizens, Ecuadorian courts must enhance their “external legitimacy.” External legitimacy is defined as “court relationships with other institutions and citizens” (Horan and Meinhold 2012, 236). This external legitimacy can be seen in Mexico. Staton (2010, 14) conducted an empirical study in Mexico in which he noted that the high courts of Mexico empowered their branch by engaging in public relations and strategic communication with citizens. Thus, Staton (2010)

Adding to this, the court’s office was

responsible for issuing press releases of its decisions in its offices and to the media (Staton 2010, 85). Through public relations, the court communicated with the public, increased literacy and transparency, and combatted inaccurate reporting of its rulings (Staton 2010, 59). The strategy’s impact increased external legitimacy because it directly created a relationship with increased public trust between the public and the court. The latter results in what Staton (2010, 13-14) calls a “public enforcement mechanism.” This mechanism is what pressures the executive to respect the authority of the court to decide a case and comply with its rulings. Although Staton (2010) agrees not every decision will be desired by the citizenry, they will nonetheless support it because they will develop a:

Pi Sigma Alpha Undergraduate Journal of Politics “Normative

commitment

the

debt” had “turned public,” and that President Zedillo

appropriateness of the constitutional review

had overstepped his authority (Staton 2010, 66).

mechanism. The normative commitment to

law trumps policy interest [of the people]…

Chamber of Deputies, and ordered President Zedillo

on this account, the fundamental problem is

to provide the Mexican Chamber of Deputies with

solved by public beliefs in judicial legitimacy”

all the necessary information on the transactions

(Staton 2010, 13).

and the actors involved in this scandal (Controversia

Ultimately under the approach designed by

the Staton (2010), courts still “confront the problem of policy making, but the public serves as a baseline source of power” (Staton 2010, 13-14), and citizens pressure their representatives to respect judicial authority and decisions. This is all part of the public enforcement mechanism (Staton 2010, 13-14). Public Enforcement in Action: The Zedillo Case

In order to provide evidence for the success

In its ruling the court agreed with the Mexican

Constitucional 26/99 2000, 217). Many considered the decision a confirmation of the court’s power and ability to decide cases independent of the executive. Staton (2010, 3) notes that this case was historic because it was “the first time in which the Supreme Court challenged the power of the presidency in a case of such magnitude.” This power was evident because in spite of the political repercussions that compliance would have on him, President Ernesto Zedillo complied with the court’s decision. Staton (2010, 66)

of the “public enforcement” model, Staton (2010, 66)

notes that compliance was due to fact that “the public

discusses the Zedillo case of 2000, which is formally

costs of defiance were enormous for Zedillo.” Hence,

known as the Controversia Constitucional 26/99. Cámara de Diputados vs. Poder Ejecutivo Federal (Caso Banco Unión). In this case, the Supreme Court, a direct beneficiary of the Ernesto Zedillo reforms, ruled against Zedillo, and publicized the case in order to ensure its enforcement. In midst of the 1994 peso crisis, the Zedillo administration moved to “assume the debts of several failing banks [that were] under the Banking Fund for Savings Protection [Fondo Bancario de Protección al Ahorro (FOBAPROA)]” (Staton 2010, 65). However, when the government assumed these debts, it turned private debt into public debt that would amount to $65 million dollars (De La Garza 1998). In gratitude of the loan, the bankers allegedly used the same loan money to fund President Zedillo’s 1994 campaign (Dillon 1999). The Mexican Chamber of Deputies then demanded to see documents on the loans that were issued on the grounds that the “private

with public support stemming from public relations, the court was able to increase its presence and popularity through public relations in order to attain the power necessary for the compliance of President Zedillo. This is why courts should go public, because without popularity and proper communication with the citizenry, “noncompliance will go undetected” (Staton 2010, 25), and the judiciary will remain weak. The Public Enforcement Model’s Viability in Ecuador

Similar to Mexico, the possibility of Ecuador

increasing its external legitimacy and popularity to establish judicial power is not impossible. Ecuador has had a history of backing the judiciary against the strong executives in spite of its institutional weakness. In 2005, when President Lucio Gutierrez decided to replace all high judges for political reasons, Ecuadorians saw this as an act of “authoritarianism,”

Romo

The Financial Times reported in March 2016 that the independent judiciary has moved to investigate and to President Gutierrez’s removal from office (Basabepotentially hold ex-President Lula da Silva accountable Serrano and Polga-Hecimovich 2013, 156; Horan and for corruption and money laundering charges (Pearson Meinhold 2012, 156, 236). This track record, in which 2016). This shows that the internal legitimacy, and de constituent pressure has worked to remove even the jure independence in the Constitution of the Federal president from office, shows potential for Staton’s Republic of Brazil (Art. 95 and Art.134 1988), that (2010) public enforcement model. Even the electoral was attained through the mobilization of by civil results from the 2011 judicial referendum exposes the Ecuadorian people’s desire to eventually trust the court society strengthened the court’s legitimacy enough to ensure that its rulings fight for the “horizontal after reforming it (Consejo Nacional Electoral 2011, 3 and 7). The court’s use of strategic communication and accountability” necessary for a democracy that separates the leader’s personal interests from the public relations would make this possible. The latter public’s interests (O’Donnell 1994, 61). The Financial is the only way to create a popular relationship with Times article underscored that this was attributable the citizenry and ensure that when noncompliance by to the reforms of the constitution, and also mentions political actors occurs, the public is ready to pressure that Brazil’s judiciary has become a shining example of those who refuse to comply. independent and institutional strength (Pearson 2016). Conclusion In Mexico, the reliance on citizen support is evident in Staton’s (2010) public enforcement model. Given the detrimental, but strategic actions To put this in context, Mexico was able to democratize populism has taken against institutions such as the by empowering the court through public relations. court, it may seem unlikely that the judiciary can This empowerment forced the highest figure in act as a counterweight to the populist. However, the government, President Ernesto Zedillo, to comply literature review examining populism and constituent with a ruling (Controversia Constitucional 26/99 2000, power at the extreme in Venezuela, as well as the cases 217; Staton 2010, 66), that would have otherwise been of Brazil and Mexico, show that the court is a necessary ignored. This compliance affirms the independence actor that is capable of inclusion. Specifically, this of a court that is capable of ensuring that no one is discussion has shown that institutional empowerment above the law, because the law is enforced equally. This and subsequent independence must be established, becomes engrained in the political memory of citizens but that it comes directly from its reliance of citizen and possible because of external legitimacy. support and mobilization. An institution can only be When the Ecuadorian court is able to increase as powerful as the legitimacy and popularity it holds. its popularity and construct internal and external This was seen in Brazil through the pressure legitimacy for judicial independence and power, as that civil society and lawyers and judges applied on was the case in Brazil and Mexico, this research asserts the state to reform and strengthen the independence that their prospects will be positive. Fortunately, of the Public Defender’s Office (Madeira 2014, 55). Ecuador’s citizenry has previously shown an ability The success of Brazil’s increase in external legitimacy to mobilize and use their constituent power to make through legal professionalism has worked until 2016. resulting in widespread protests. This eventually led

A Counterweight to Populism

certain demands. With constituent power backing them, the courts can become not just empowered and independent, but capable of creating the political inclusion that citizens demand. Through the internal legitimacy shown in Brazil, a legal framework can be created to increase independence, and legal professionalism, which may also resemble the direct connection and use of law that the populist has created with the populace. This in turn can increase public support and remediate the alienation lack of public trust that citizens have historically had with the court (Barómetro de Las Americas 2012, 124).

both powerful and independent. This empowerment and independence will only occur if the court is able to construct the legitimacy necessary to become a more popular actor with which the citizenry can politically identify.

The case of Mexico echoes this notion, showing how through external legitimacy, particularly public relations, communication with the citizenry can be more personalized and politically inclusive. This communication can advance the popularity and public support necessary to pressure the executive into compliance, and thus empower the judiciary. This was evident in the public enforcement by Staton (2010), which “leverages a form of vertical accountability to provide a horizontal accountability” (Staton 2010, 16). Both forms of the necessary accountabilities are important since they limit arbitrary decision making of a leader, and make democracies fairer and less “delegative” (O’Donnell 1994, 59). As Weyland (2009, 14) notes, unlike courts, populists prefer “distributing favors than to guaranteeing rights.” This is alarming considering that after a favor a debt ensues. The result is a perpetual cycle of institutional instability that does not remedy the long term need for fair inclusion and equal enforcement of the law (De La Torre 2010). To avoid this, the populace must begin to trust the court more and the populist less. After all, the court must work by the rule of law and procedure rather than arbitrariness. However, the court’s ability to ambitiously enforce the rule of law equally will best work when the court is

etc.”

Notes Panizza (2016, 194) describes the “Other” as “the power holders, the oligarchy, the political elite, Washington insiders, Wall Street, the European Union, immigrants, and ethnic, religious or linguistic group, 1

Barómetro de Las Americas’s (2012, 124) barometer measured twelve institutions and adjusted the responses of the surveyed respondents on a scale from 0-100, with a 0 representing low support/trust and a 100 indicating high support/trust. The institutions surveyed and the corresponding results are the following: the Catholic Church (67.4 average), the military (65.8 average), the President (62.8 average), the Nacional Police (53.7 average), the media (52.7 average), elections (50.6 average), the National Electoral Council, (48.0 average) the Parliament (43.6 average), the justice system (42.9 average), the National Court of Justice (41.8 average), the Evangelic Church (35.0 average) and finally political parties (31.3 average). 2

Spanakos (2016,115) defines polemic institutionality as a “rejection of the intrinsic value of institutions; its openness to constituent power; its preference for experimentation and aversion towards ‘bureaucratisation’ of power; and lack of credibility in institutions serving as ‘neutral third party’ enforcers.” 3

The questions on the referendum were: 1) Do you convoke a National Constituent Assembly with the 4

44 purpose of transforming the State and creating a new juridical order that allows the effective functioning of a Social and Participatory Democracy? and 2) Do you authorize the President of the Republic through an Act of Government to set, listening to the opinion of political, social and economic sectors, the basis of the electoral process in which the members of the National Constituent Assembly will be elected? (Consejo Nacional Electoral 1999, 1; Spanakos 2016, 116). Spanakos (2016, 118) notes that a “neutral, third party enforcement and pro-government forces expected the Supreme Court to act, not as neutral arbiter, but as 5

interpreter of the will of the people (as both people in its totality and in its majority, which supported their positions), and thus, a part of the ‘process.’”

References

Ciccariello-Maher, George. 2013. We Created Chávez: A People’s History of the Venezuelan Revolution. Durham, NC: Duke University Press. Conaghan, Catherine, and Carlos De La Torre. 2008. “The Permanent Campaign of Rafael Correa: Making Ecuador’s Plebiscitary Presidency.” The International Journal of Press/Politics 13(3): 267-284. Complimentary Act No. 132. 2009. Federal Republic of Brazil. http://www.planalto.gov.br/ccivil_03/ leis/LCP/Lcp132.htm (Accessed April 26, 2016). Consejo Nacional Electoral. 1999. “Referendos Nacionales Efectuados en Venezuela (19992000).” República Bolivariana de Venezuela. http://www.cne.gob.ve/web/documentos/ estadisticas/e010.pdf (Accessed April 26, 2016).

Basabe-Serrano, Santiago, and John Polga-Hecimovich. Consejo Nacional Electoral. 2011. “Resultados Del Referéndum Y Consulta Popular 2011.” 2013. “Legislative Coalitions And Judicial Ecuador. http://www.funcionjudicial.gob.ec/ Turnover under Political Uncertainty: The Case www/pdf/baselegal/Consulta%20Popular%20 of Ecuador.” Political Research Quarterly 66(1): y%20Referendum.pdf (Accessed April 23, 154-166. 2016). Barometro de Las Americas for Latin American Public Constitution of the Republic of Venezuela. Art. 245 Opinion Project. 2012. “Cultura Politica de (1961). http://historico.tsj.gob.ve/legislacion/ La Democracia en Ecuador: Confianza en constitucion1961.pdf (Accessed April 26, Las Instituciones en Ecuador.” Vanderbilt 2016). University. http://www.vanderbilt.edu/lapop/ ecuador/Ecuador_Country_Report_2012_W. Constitution of the Federal Republic of Brazil. Art. 95 pdf (Accessed April 26, 2016). and Art. 134 (1988). http://www.planalto.gov. br/ccivil_03/Constituicao/Constituicao.html Caldeira, Gregory A., and James L Gibson. 1995. (Accessed April 28, 2016). “The Legitimacy of the Court of Justice in Constitution of the Republic of Ecuador. Art. 192 the European Union: Models of Institutional (2008). http://www.asambleanacional.gov. Support.” American Political Science Review 89 ec/documentos/constitucion_de_bolsillo.pdf (June): 356-376. (Accessed April 26, 2016).

Pi Sigma Alpha Undergraduate Journal of Politics Controversia Constitucional 26/99. Cámara de Diputados vs. Poder Ejecutivo Federal (Caso Banco Unión). 24 de Agosto 2000, Tribunal Pleno de la Suprema Corte de Justicia de la Nación, http://www.juridicas.unam.mx/ publica/librev/rev/arsiu/cont/24/act/act8.pdf (Accessed April 26, 2016). Defensoria Publica. n.d. “Reseña Histórica: La Defensoría Pública Llega para Equilibrar el Sistema de Justicial.” República del Ecuador. http://www.defensoria.gob.ec/index.php/ defensoria-publica/quienes-somos/resenahistorica (Accessed April 25, 2016). Defensoria Publica. 2014a. “Informe Annual de Labores.” República del Ecuador. http://www. defensoria.gob.ec/images/defensoria/rdc2015/ informe-gestion-2014.pdf (Accessed December 11, 2015). Defensoria Publica. 2014b. “Transparencia, Informacion de Rendicion de Cuentas.” República del Ecuador. http://www.defensoria. gob.ec/index.php/transparencia/informacionde-rendicion-de-cuentas/2014 (Accessed December 11, 2015). De la Garza, Paul. 1998. “Banking Scandal Could Cost Mexico $65 Billion,” The Chicago Tribune, September 1. http://articles.chicagotribune. com/1998-09-01/news/9809010119_1_biggestbanking-scandal-fobaproa-weakest-banks (Accessed April 25, 2016). De La Torre, Carlos. 2010. The Populist Seduction in Latin America. Athens, OH: Ohio University Press. De La Torre, Carlos. 2013. “El Tecnopopulismo de Rafael Correa. Es Compatible el Carisma con la

Tecnocracia?” Latin American Research Review 48(1): 24-43. Dillon, Sam. 2000. “Mexico Court Makes History By Siding With Congress,” The New York Times, August 24. http://www.nytimes. com/2000/08/25/world/mexico-courtmakes-history-by-siding-with-congress.html (Accessed December 19, 2015). Emenda Constitucional No. 45, 2004. República Federativa do Brasil, December 30. http:// www.planalto.gov.br/ccivil_03/Constituicao/ Emendas/Emc/emc45.htm (Accessed April 28, 2016). Hardt, Michael, and Antonio Negri. 2004. Multitude: War and Democracy in the Age of Empire. New York: Penguin. Hellinger, Daniel. 2011. Comparative Politics of Latin America: Democracy at Last? New York: Routledge. Horan, Jennifer E., and Stephen S. Meinhold. 2012. “Separation of Powers and The Ecuadorian Supreme Court: Exploring Presidential-Judicial Conflict in a Post-Transition Democracy.” The Social Science Journal 49:229-240. Ingram, Matthew C. 2012. “Crafting Courts in New Democracies: Ideology and Judicial Council Reforms in Three Mexican States.” Comparative Politics 44(4): 439-458. Human Rights Watch. 2014a. “Letter on Judicial Independence in Ecuador.” https://www. hrw.org/news/2014/01/29/letter-judicialindependence-ecuador (Accessed February 14, 2015). Human Rights Watch. 2014b. “Ecuador: Ensure Judicial Independence.” https://www.hrw.org/

Romo news/2014/01/29/ecuador-ensure-judicialindependence (Accessed February 14, 2015).

Ibáñez, Paul. 2014. “El 19 de Enero de 1999 fue Legitimado el Poder Originario que dio Surgimiento a la Constitución.” Agencia Venezolana de Noticias, January 18. http:// www.avn.info.ve/contenido/19-enero-1999fue-legitimado-poder-originario-que-diosurgimiento-constitución (Accessed April 24, 2016). Kapiszewski, Diana, and Matthew M. Taylor. 2008. “Doing Courts Justice? Studying Judicial Politics in Latin America.” Perspectives in Politics 6:741-767. Laclau, Ernesto. 2005. “Populism: What’s in a Name?” In Populism and the Mirror of Democracy, ed. Francisco Panizza, London: Verso, 32-49. Linz, Juan J., and Arturo Valenzuela. 1994. The Failure of Presidential Democracy: The Case of Latin America Vol. 2. Baltimore: Johns Hopkins University Press. Linz, Juan J., and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America and Post Communist Europe. Baltimore, MD: Johns Hopkins University Press. Madeira, Lígia Mori. 2014. “Institutionalization, Reform and Independence of the Public Defender’s Office in Brazil.” Brazilian Political Science Review 8(2): 48-69. Mainwaring, Scott, and Timothy R. Scully. 2009. Democratic Governance in Latin America. Stanford, CA: Stanford University Press. O’Donnell, Guillermo A. 1994. “Delegative Democracy.” Journal of Democracy 5(1): 55-69.

Opinion No. 17. 1999. Supreme Court of Justice of Venezuela on the Referendum of Convening a Constitution Assembly. January 19. http:// infovenezuela.org/democracy/attachmentsspanish/T1%20ST01%20N2%20Fallos%20 N17%20y%2018%20de%20la%20CSJ%20 sobre%20el%20R%20consultivo%20.pdf (Accessed April 25, 2016). Panizza, Francisco. 2005. “Introduction: Populism and the Mirror of Democracy.” In Populism and the Mirror of Democracy, ed. Francisco Panizza, 1-31. Panizza, Francisco. 2009. Contemporary Latin America: Development and Democracy Beyond the Washington Consensus. New York: Zed Books. Panizza, Francisco. 2016. “Populism, Social Democracy and the Tale of the ‘Two Lefts' in Latin America.” In Conceptualising Comparative Politics, eds. Anthony Petros Spanakos and Fransciso Panizza. New York: Routledge, 192211. Pasara, Luis. 2014. “Judicial Independence in Ecuador’s Judicial Reform Process.” Due Process of Law Foundation. http://www.dplf.org/sites/default/ files/indjud_ecuador_executivereport_eng.pdf (Accessed April 19, 2015). Pearson, Samantha. 2016. “Corruption Drama Puts Brazil’s Judges in the Spotlight,” The Financial Times, March 22. http://www.ft.com/ intl/cms/s/0/b3ba4d3a-f02d-11e5-a609e9f2438ee05b.html (Accessed April 14, 2016). Seawright, Jason, and John Gerring. 2008. “Case Selection Techniques in Case Study Research.” Political Research Quarterly 61(2): 294-308.

A Counterweight to Populism Spanakos, Anthony P. 2008. “New Wine, Old Bottles, Flamboyant Sommelier: Chavez, Citizenship, and Populism.” New Political Science 30(4): 521-544. Spanakos, Anthony P. 2013. “Latin America’s Left: Between Demos and Kratos.” Public Books. www.publicbooks.org/nonfiction/latinamericas-left-between-demos-and-kratos (Accessed December 11, 2015). Spanakos, Anthony P. 2016. “Institutionalities and Political Change in Bolivarian Venezuela.” In Conceptualising Comparative Politics, eds. Anthony Petros Spanakos and Fransciso Panizza. New York: Routledge, 113-124. Staton, Jeffrey K. 2010. Judicial Power and Strategic Communication in Mexico. New York: Cambridge University Press. Rios-Figueroa, Julio, and Jeffrey K. Staton. 2014. “An Evaluation of Cross-National Measures of Judicial Independence.” The Journal of Law, Economics, & Organization 30(1):104-137. Valenzuela, Arturo. 2004. “Latin American Presidencies Interrupted.” Journal of Democracy 15(4):5-19. Weyland, Kurt. 2001. “Clarifying a Contested Concept: Populism in the Study of Latin American Politics.” Comparative Politics 34(1):1-22. Weyland, Kurt. 2009. Populism and Social Policy in Latin America. Woodrow Wilson International Center for Scholars. Washington, D.C. 1-35. Weyland, Kurt. 2013. “The Threat from the Populist Left.” Journal of Democracy 24(3): 18-32. World Bank. 2014. “World Development Indicators: Population Data.” http://data.worldbank.org/

indicator/SP.POP.TOTL (Accessed December 14, 2015).

Las Ministras: Gender Quota Laws and Female Representation in Latin American Cabinets and Legislatures Kristine E. Adams Saint Anselm College Despite the existence of Machismo and Marianismo, Latin American states have high female representation in most political offices when compared to global averages. Many scholars have analyzed female representation in Central and South American legislatures, offering legislative, socioeconomic, and structural explanations for increased percentages of women. Few studies, however, have been conducted on women’s ascension to ministerial positions. By employing statistical analyses of twenty-one states in the Latin American region as well as in-depth case studies of Argentina and Brazil, this research evaluates well-enforced gender quota laws to determine if gender quota laws increase female representation in legislatures and cabinet positions in Latin America. Specifically, it is argued that implementation of these laws has increased female participation in legislatures, leading to an increased likelihood that women will be appointed to cabinet positions.

Introduction

Globally, there is an enormous gender gap

in political representation. On average, men hold nearly 80% of seats in legislatures and 85% of cabinet and ministerial positions (Inter-parliamentary Union 2015). A way to lessen this gap is to legitimize women as politicians, not women merely appointed to further women’s issues. Paradoxically, the best way to empower female politicians is through substantial political representation in these roles. A region that has had success in sustaining high levels of female representation in all levels of government is Latin America. Regionally, female representation in the legislature averages 27.1%, while women occupy 26.63% of cabinet positions (Inter-parliamentary Union 2015).

Despite these victories in the effort toward

significant female representation in politics, discriminatory cultural and legal restrictions halt women’s progress in Latin America. For example, in

the city of Córdoba, Argentina, it is illegal for a woman to drive a bus. Because it is culturally accepted that operating a bus is too difficult for female drivers, they are allowed to operate the city’s trolleys, due to the fact that electric cables control trolley routes. Although Argentina has deemed women to be inadequate navigators, the people of the country have no qualms with a woman guiding their country politically.

How is it that the very country – Argentina

– that prohibits women from driving busses has elected so many women to the legislature and appointed female cabinet members? Moreover, why has Latin America elected so many female politicians when “Machismo” and “Marianismo” reinforce the traditional roles of women? By definition, “Machismo” is the presumption that virility, strength, and entitlement to dominate are masculine attributes. “Marianismo” is the Hispanic veneration for feminine virtues, such as purity and moral strength, that empowers women in maternal roles (Stevens 1973),

Pi Sigma Alpha Undergraduate Journal of Politics ensuring that women are “primero que nada, madre”1 (Deutsch 1991, 268).

The above discrepancy suggests that additional

analysis of female representation in Latin American states is necessary and important. Understanding how women rise to high political office is essential to closing the gender gap and achieving equality. If scholars can pinpoint the factors increasing a woman’s chance of becoming a legislator or minister in Latin America, these practices may be applied worldwide. But what accounts for this phenomenon? In the following analysis, it is argued that the implementation of quota laws in Latin America has increased female participation in legislatures, resulting in an increased likelihood that women will be appointed to cabinet positions. Literature Review There is a vast body of scholarship regarding female representation in Latin America. From this literature, two distinct schools of thought emerge regarding gender quota laws. Specifically, scholars debate whether gender quota laws increase, or have no impact, on female representation in legislatures and cabinets. By definition, a quota law is legislation that mandates a certain percentage of women in a political party, congress, or on a ballot (Alnevall, Frankl, Freidenvall, Krook, and Nordlund 2015). Generally, gender quota laws do well where there is a long history of gender discrimination (Clayton 2015; Paxton and Hughes 2014). Numerous scholars address the first school of thought, asserting that gender quota laws increase female representation (e.g., Baldez 2007; Clayton 2015; Escobar-Lemmon, Schwindt-Bayer, and TaylorRobinson 2014; Fleschenberg 2007; Htun and Piscopo 2014; Htun and Power, 2006; Jalalzai 2014; Jones

2009; Kittilson and Schwindt-Bayer 2012; Krook and Norris 2015; Krook and O’Brien 2008; Murray 2013; O’Brien, 2015; Paxton and Hughes, 2014; Reynolds, 1999; Schwindt-Bayer, 2009; Schwindt-Bayer 2010; Schwindt-Bayer 2011; Viñas 2014). On average, since the implementation of quota laws in Latin America, parliaments have seen a ten percent increase in female representation (Baldez 2007). Within this argument, these scholars argue that certain electoral structures support the efforts of quota laws as they help improve female representation. One of these structures is electoral systems with proportional representation (PR) (Fleschenberg 2007; Hinojosa 2012; Kittilson and Schwindt-Bayer 2012; Krook and Norris 2014; Paxton and Hughes 2014; Schwindt-Bayer 2010). Theoretically, PR systems favor women because voters select parties rather than specific candidates. Additionally, these scholars claim that high district magnitude and closed electoral lists allow more women to get on the ballot without displacing a man (Fleschenberg 2007; Gray 2003; Htun and Piscopo 2014; Jones 2009; Kittilson and Schwindt-Bayer 2012; Paxton and Hughes 2014; Schwindt-Bayer 2010). Furthermore, this school of thought asserts that socioeconomic indicators, in conjunction with gender quota laws, increase female representation. Laborforce participation is argued to increase the number of female candidates due to the fact that work experience is often seen as qualification for political office (Iversen and Rosenbluth 2010). Additional indicators argued to increase female representation include high Gross Domestic Product or GDP, increased levels of female education, smaller gender gaps, and democratization (Hinojosa 2012; Iversen and Rosenbluth 2010). The second school of thought argues that gender quota laws are not an effective means of increasing female representation. As an affirmative

Adams

action effort, quotas are deemed undemocratic, although they strive to create a more representative legislature (Hinojosa 2012). Also, political parties evade quota laws and undermine efforts to increase female representation by taking advantage of soft enforcement of quotas and utilizing primaries (Baldez 2007; Hinojosa 2012). Moreover, it is common for political parties or states to adopt quotas as a means to legitimize themselves to the international community, with no real intention of following them (Baldez 2007; Casas-Acre and Saig 2015; Clayton 2015; Kittilson and Schwindt-Bayer 2012). Finally, quotas are critiqued for their inability to increase the amount of influence women have in the legislature, influential committees, and cabinet positions (Baldez 2007; Escobar-Lemmon and Taylor-Robinson 2014; Murray 2013), causing some to believe that they are unnecessary (Casas-Acre and Saig 2015).

Building on both lines of the literature, this

research argues that well-enforced gender quota laws provide women the greatest opportunity to hold a seat in congress or a cabinet. Although some argue that PR systems with closed lists affect female representation, empirical evidence has yet to show higher levels of female representation in these systems. Also, women across the globe have risen to power without these

laws and ministerial representation (Jalalzai 2014; Thames and Williams 2013). Therefore, this research addresses the following question: Do well-enforced gender quota laws increase female representation in legislatures and cabinets in Latin America? In general, it is argued that the adoption of well-enforced quota laws has increased female participation in legislatures, leading to a higher likelihood that women will be appointed to cabinet positions. First, it is hypothesized that Latin American states with well-enforced gender quota laws have increased female representation in legislatures (H1). Additionally, it is expected that high numbers of women in congress, due to quota laws, increase the number of women in cabinet positions (H2). Moreover, this analysis seeks to test several hypotheses for increased female representation proposed by other scholars (Gray 2003; Htun and Piscopo 2014; Schwindt-Bayer 2010). These variables include proportional representation electoral systems (A1), open lists (A2), democratization (A3), Gross Domestic Product or GDP (A4), labor ratios (A5), tertiary education ratios (A6), and a small gender gap within a state (A7). Variable Conceptualization and Operationalization

structures in place. Additionally, it is argued that if

The first independent variable measured is

political parties are mandated to adhere to gender

the presence of well-enforced quota laws. A â&#x20AC;&#x153;gender

quota laws, women will be able to overcome cultural

quotaâ&#x20AC;? is defined as legislation that places an increased

disadvantages and contest with men in elections.

percentage of women on a ballot or in congress.

For many women in Latin America, the majority

There are three types of quota laws: 1) reserved seats

of political advancements are made through the

quotas, or the designation of positions for women

legislature. Therefore, quota laws do not solely increase

in the legislature; 2) legal candidate quotas, where a

the number of women in Latin American congresses,

percentage of female candidates are legally mandated

but they also increase the probability that a woman will to be on a ballot; and 3) political party quotas, where hold a position in the cabinet. Finally, this work stands parties voluntarily place women on the ballot (Htun apart from existing literature in the respect that few

and Piscopo 2010). A well-enforced quota law is

scholars have analyzed the connection between quota

defined as legislation that is strictly upheld by the

Las Ministras government, or mandates the fair placement of women

mandated percentages, 3)the presence of voluntary political party gender quota laws, 4) the presence of on ballots. Both provisions do not need be in place to placement mandates for legislated quotas, 5) if the make the quota law effective. The more safeguards in quotas are well enforced, 5) the electoral system of place, however, the stronger the law. Therefore, wellthe state (whether they have open or closed lists), 6) enforced quota laws are measured by separating the concept into five different categories, allowing a test of female representation in lower and upper houses in the different aspects of the laws. Quota laws of different the legislature, 7) the percent of women in cabinet positions, 8) whether female representation in the states are differentiated by: 1) whether a legislated cabinet of a state is high or low when compared to the candidate quota exits; 2) the quota percent; 3) if the global average, 9) democratization, 10) gross domestic law applies to all elections; 4) if there are other forms product per capita with purchasing power parity, of quotas; and 5) if these quota laws are well enforced 11) female labor force participation rates, 12) female and ensure the fair placement of women on ballots. education levels, and 13) gender equality within each The second independent variable is “female state. By examining these variables, this research tests representation”. Female representation is defined as the proposed hypotheses (H1 and H2) and alternative women holding political office. Female representation hypotheses by utilizing multiple forms of statistical is measured at two levels: the legislature and cabinet analyses to control for other variables that may affect positions. At both levels, the percentage of women female representation at all levels of government. is measured as those who occupied these positions Finally, through multiple statistical analyses of the in two-year increments from September 2005 to hypothesized independent (e.g., gender quota laws) September 2015 in Latin American legislatures. This and dependent variables (e.g., female representation shows the change in female representation from in legislatures and on cabinets), the factors that congress to congress (Hinojosa 2012). increase female representation in Latin America can be Methodological Approach predicted. Second, by conducting a comparative analysis, First, a quantitative analysis of the Latin the findings of the regional statistical analysis through American region is conducted. Specifically, a time series data set of twenty-one Latin American states2 in an in-depth study of Brazil and Argentina utilizing Mill’s Method of Difference is explored (Mill 1884). two-year increments, ranging from 2005 to 2015, was These two states were selected due to their similar created by the author. Thirteen variables – obtained political structures, size, cultures of Machismo from the quotaProject (see Alnevall et al. 2015), The and Marianismo, and histories of dictatorial rule. World Bank (1990-2015), The World Factbook (2005Additionally, each of these countries has some sort of 2015), the Polity IV (see Marshall and Gurr 2013), legislated gender quota law. In total, the combination Global Gender Gap Index (see World Economic of statistical and comparative analyses is the most Forum 2006-2014), and Inter-parliamentary Union direct and thorough way to examine if there is a (2005-2015) databases – are included in the data set: significant empirical link between quota laws, more 1) the presence of legislated candidate gender quota female legislators in Latin America, and increased laws and their mandated percentages, 2) the presence female representation in cabinet positions. of subnational candidate gender quota laws and their

52 Female Representation before Quota Laws Despite the current levels of female representation in Latin American congresses, women have not consistently been involved in politics. Traditionally, politics was deemed as a man’s role in society. Traditionally, the ideal Latin American woman worked in the home, focusing on her maternal responsibilities (Deutsch 1991; Paxton and Hughes 2014; Stevens 1973). Marianismo, or admiration of the Virgin Mary through dedication to selfless motherhood, emphasized the woman’s role in the home (Paxton and Hughes 2014, 291). Machismo reinforced this, as women were “…expected to derive their identities from the males in their lives…” and “to derive fulfillment as the counterparts to these males, as daughters, wives, and mothers” (Paxton and Hughes 2014, 291). Women became significant political actors at the regional level during the military dictatorships of the 1970’s and 1980’s. As a reaction against these regimes, women protested the heinous acts of these military dictatorships. Women, who traditionally prescribed to their husband’s political opinions, suddenly became involved in activism. The Madres and the Abuelas de la Plaza de Mayo, an example of increased women’s participation, were instrumental in demanding governmental transparency and advocacy for human rights in Argentina during the “Process of National Reorganization” (Deutch 1991; Paxton and Hughes 2014). After democratic regimes were reinstated, however, women were unable to translate their activism into formal participation, despite the fact many states rewrote their constitutions to promote equality (Deutsch 1991). Several theories account for this sudden drop in female political participation. First, when democratization occurs, female political

participation initially declines. Due to high electoral competition in democratic regimes, fewer women were elected when compared to appointments of women by the ruling party and activist involvement during nondemocratic regimes (Fallon, Swiss, and Viterna 2012). With increasing democratic freedoms and elections, however, representation again increases (Fallon, Swiss, and Viterna 2012). Alternative Hypotheses 1-2: Political and Electoral Structures

As previously stated, after the military

dictatorships of the 1970’s and 1980’s, many states rewrote their constitutions to adopt more democratic principles (Deutsch 1991). One aspect of democratization in Latin America was the adoption of PR electoral systems. Minorities, including women, perform better in these systems due to the fact that voters select parties rather than candidates. These systems also create incentives for women to run for office and encourage political parties to nominate them. This translates into votes for women and potentially gaining seats (Schwindt-Bayer 2010). As a result, it is argued that women have a better chance of getting on ballot without displacing a man under a PR electoral system than other political systems, such as majoritarian systems (Paxton and Hughes 2014).

Despite these advantages, there is little concrete

evidence to support the first alternative hypothesis (A1), claiming that PR systems have increased female representation in Latin America. In her analysis of the efficacy of gender quota laws as a means to improve female representation in Latin America, SchwindtBayer (2010) found no relationship between PR systems and women in congress.

Hypothesis 1 supports the claim that

PR systems do not correlate to increased female

Pi Sigma Alpha Undergraduate Journal of Politics

representation in legislatures. To analyze the first alternative hypothesis (A1), the means of female representation in both houses of Latin American legislatures for all three electoral systems present in the sample are compared. In addition to PR systems, mixed member proportionality (MMP) systems and two round systems (TRS) were present. A MMP system is defined as a mixed electoral system in which representatives are elected through a proportional representational system and a plurality/majority system (International Institute for Democracy and Electoral Assistance [IDEA] 2015). Additionally, a

female representation in the legislature for each respective electoral system. Although the results of the comparison of means analysis for A1 yielded no statistically significant results, it is worth noting that the mean percentage of women in the lower house of the legislature is higher for both MMP and TRS systems (See Tables 1 and 2). In addition to PR systems, scholars claim that the presence of â&#x20AC;&#x153;closed listsâ&#x20AC;? increases female representation (A2) (Gray 2003; Htun and Piscopo 2014; Jones 2009). Closed lists are lists of candidates where political parties determine who is on the ballot,

TRS is defined as a plurality/majority electoral system where second elections are held if no candidate or party achieves a given level of votes, which most commonly is an absolute majority in the first round of elections. After closely analyzing these electoral systems, there is no significant difference in the means of

and voters cannot express a preference for a particular candidate (International IDEA 2015). Theoretically, women have an easier time getting on a ballot when they are guaranteed an opportunity to contest for a seat. The analysis in Table 3 shows that closed lists have no correlation to high female representation (A2).

Table 1. Proportional Representation Systems and Mixed Member Proportionality Systems Comparison of Means

Electoral System N 95 Women in Lower Proportional Representation House (%) Mixed Member Proportional System 19 36 Women in Upper Proportional Representation House (%) Mixed Member Proportional System 12

Std. Std. Error Sig. Mean Deviation Mean (2-tailed) 20.739 10.1740 1.0438 .160 24.537 10.4930 2.4073 17.428

10.3410

1.7235

25.508

16.0304

4.6276

.124

Mean Difference -3.7979 -8.0806

Table 2. Proportional Representation Systems and Two Round Systems Comparison of Means

Women in Lower House (%) Women in Upper House (%)

Electoral System Proportional Representation

N 95

Mean 20.739

Std. Deviation 10.1740

Two Round System

23.458

20.6523

Std. Error Sig. Mean (2-tailed) 1.0438 .451 5.9618

Proportional Representation

17.428

10.3410

1.7235

Two Round System

9.283

9.0444

3.6924

.084

Mean Difference -2.7194 8.1444

Adams

In a comparison of means test, states with PR systems and their mean legislative female representation for open lists and closed lists were analyzed. Despite higher levels of women in the lower house (22.126%) and the upper house (22.283%) for states that utilize closed lists, there is no statistically significant link between the means of women in the lower house. Conversely, a statistically significant difference is found between the percentages of women in the upper house in political systems that used closed lists. However, because the comparison of means test only included eighteen cases for open and closed list elections in

These findings support the existing scholarship in several ways. First, according to Hinojosa (2012), a state’s income is not related to its gender equality, for many poor states, such as Nicaragua and Cuba, have extremely high levels of female representation. In terms of labor force participation, Iversen and Rosenbluth (2010) claim that the relationship between labor force participation is inconsistent. The analysis did not invalidate all of the alternative hypotheses, as Table 4 shows a statistically significant correlation between low gender gaps (A7) and female representation in both houses of

the upper house, it cannot be concluded that this difference in means is statistically significant.

Latin American legislatures. Additionally, the results revealed a strong negative correlation between democratization (A3) and female representation in the lower/unicameral house of Latin American legislatures. An explanation for this phenomenon is that non-democratic regimes have high levels of female representation (Paxton and Hughes 2014). Although many women hold political office, women in these regimes do not possess real political power, “… because their mandates to govern are from controlling autocrats rather than from constituents” (Fallon, Swiss, and Viterna 2012, 395) Due to this, there are few women in the senate or cabinet positions. The latter is certainly the case in Latin America. In the past ten years, Cuba, the least democratic state in the sample, maintained a mean of 42.85% women in the lower house (Marshall and Gurr 2013). This average was higher than that of any other country in the dataset.

Alternative Hypotheses 3-7: Socioeconomic Development Indicators In addition to electoral structures, scholars argue that certain socioeconomic development indicators correlate to increased female representation in Latin America. Using the dataset compiled, a correlation analysis was conducted among variables measuring democratization (A3), GDP (A4), labor force participation (A5), tertiary education enrollment ratio (A6), and Global Gender Gap score (A7), and female representation in the both houses of Latin American legislatures. As reported in Table 4, there are no significant links between female representation and GDP (A4), labor force participation (A5), or tertiary school enrollment ratios (A6). Table 3. Open and Closed List Comparison of Means Open Lists v Closed Lists Women in Lower Open List House (%) Closed List Women in Upper Open List House (%) Closed List

N 42 54 18 18

Mean 19.062 22.196 12.572 22.283

Std. Deviation 9.5320 10.4972 4.6292 12.1986

Std. Error Sig. Mean (2-tailed) Mean Difference 1.4708 .130 -3.1344 1.4285 1.0911 .003 -9.7111 2.8752

Las Ministras The mean female representation in cabinet positions in the past ten years is significantly lower (24.38%), however, showing that although many women hold seats in the legislature, female representation is not substantive in Cuba (Inter-parliamentary Union 20052015). Hypothesis 1: Gender Quota Laws in Latin America It is argued that gender quota laws account for increased female representation in legislatures in Latin America. Female representation in legislatures rose significantly since the institution of gender quota laws in the region. Since the implementation of quotas in Latin American states, legislatures have seen an average increase in female representation by 8 to 10 percent (Baldez 2007). Table 5 shows that quota laws increased female representation in states by up to 33%. Other states, such as Brazil and the Mexican Chamber of Deputies, lost female participation after

the instatement of quota laws, while others, like the Chamber of Deputies in Paraguay and the Bolivian Senate, saw no immediate change. The negative connotation of gender quota laws and male politicians feeling threatened and subsequently lashing out toward women can account for this. By nature, gender quota laws strive to achieve â&#x20AC;&#x153;equality through positive discrimination and social group-based selectionâ&#x20AC;? (Fleschenberg 2007, 10). The opposition to quota laws claims that they are undemocratic and favor a small group of the electorate over the majority. However, the 2014 average female-to-male sex ratio for Latin American was 0.98, showing that women make up nearly as much of the electorate as men (World Economic Forum 2014). Despite their mixed results at implementation, it is argued that well-enforced legislative candidate quotas have increased female representation in Latin American legislatures. As a whole, there tends to

Table 4. Socioeconomic Development Correlations

Variables

1: Lower 2: Upper 3: Polity House House IV 4: GDP [N=126] [N=54] [N=126] [N=125]

5: Labor Ratio [N=118]

6: Tertiary 7: Gender Education Gap Ratio Score [N=109] [N=119]

1: Women in Lower House

2: Women in Upper House

.803**

3: Democratizationa (Polity IV Score)

-.293**

-.024

4: Gross Domestic Product (Billions)

-.080

.088

.073

5: Female to Male Labor Force Participation Ratio

-.168

-.049

.006

.113

6: Female to Male Tertiary Education Enrollment Ratio

.059

-.098

-.223*

-.144

.090

7: Global Gender Gap Scoreb

.568**

.552**

-.259**

-.051

.185*

.241*

Democratizationa: -10= autocratic, 10=most democratic. Global gender gap scoreb: 1= equality, 0=inequality *p < .05. **p < .01.

56 be a higher percent of female representation in the legislature where quotas exist (Htun and Piscopo 2012). An analysis suggests that this is the case in the Latin American region as a whole (see Figure 1), despite the fact that many Central American states without quota laws have high female representation in the legislature. Costa Ricaâ&#x20AC;&#x2122;s two-tiered implementation of a legislated candidate quota law and its subsequent enforcement policy emphasize the efficacy of wellenforced quotas as a means to increase female representation in the legislature. Before the adoption

2002, however, with the addition of an enforcement policy, female representation rose once again to 38.6% (Paxton and Hughes 2014). The statistical analysis supports the claims of the existing scholarship that legislated candidate gender quota laws positively affect female representation in Latin American legislatures. Table 6 reports a correlation matrix of the percent of women in the lower and upper houses of Latin American legislatures, the percentages of subnational quotas, the presence of legislated candidate quotas and their prescribed percentages, the presence of voluntary political party quotas, if non-compliance sanctions

of a quota law in 1994, only 15.8% of the national legislature was female. After the implementation of a 40% legislative candidate quota law, female representation rose to 19.8% (Paxton and Hughes 2014). Although female representation increased, it was not substantial due to the fact that political parties were not required to adhere to these laws. In

exist, and if there are placement mandates. As Table 6 illustrates, legislated candidate gender quota laws are a significant factor in increased female representation in Latin American legislatures, offering support for Hypothesis 1 (H1). There was a strong and positive correlation between the presence of legislated candidate gender quota laws and female

Table 5. The Implementation of Legislated Candidate Quota Laws Country Argentina Bolivia Brazil Costa Rica Dominican Republic Ecuador Mexico Panama Paraguay Peru Average Source: Htun 2002

Legislative Body Chamber Senate Chamber Senate Chamber Unicameral Chamber Unicameral Chamber Senate Unicameral Chamber Senate Unicameral

% Women Before Law 6 3 11 4 7 14 12 4 17 15 8 3 11 11 9

% Women After Law 30 36 12 4 6 35 16 15 16 16 10 3 18 18 17

Change (in % points) +24 +33 +1 0 -1 +21 +4 +11 -1 +1 +2 0 +7 +7 +8

Pi Sigma Alpha Undergraduate Journal of Politics

Figure 1. Womenâ&#x20AC;&#x2122;s Representation in Unicameral/Lower Houses with and without Quotas (2014)

Source: Htun 2002 representation in both houses of the legislature, meaning that states with these laws have more women in congress, Additionally, the higher percentage a legislated candidate gender quota law mandates, the higher the subsequent increase in women in legislatures. Placement and non-compliance sanctions also correlated to increased female representation in the legislature. Placement mandates lead to increased female representation in both houses of the legislature, while non-compliance sanctions only related to womenâ&#x20AC;&#x2122;s presence in the upper house. Examining these variables more closely, it is evident that female representation is much lower in states with quota laws without placement mandates and non-compliance sanctions (See Figures 2 and 3). Additionally, the mean female representation in the lower house for states with placement mandates is 29.65%, higher than the regional average of 27.1% (Alnevall et al. 2015). Subsequently, consistent with Hypothesis 1 (H1), quota laws are most effective when they are well enforced.

In addition, positive correlations are found between female representation in the lower house and the upper house, as shown in Table 6. Increased percentages of women in lower houses leads to increased opportunities for women to hold positions in the senate, showing the political progression of women through national Latin American congresses (See Table 6). Additionally, there was a positive correlation between subnational quotas and female representation in the both houses of the legislature. This implies that the quotas implemented at the municipal level tend to promote women to their respective legislatures, as well as national seats. These relationships show how women ascend to higher political positions once establishing themselves in the subnational legislatures and in the lower house of the national legislature. Hypothesis 2: Gender Quotas and Female Cabinet Representation This research argues that increased representation in legislatures, due to quota laws, will

Adams Table 6. Correlation Analysis: Quotas and Female Representation in Legislatures

Variable

1: Lower House [N=126]

2: Upper House [N=54]

3: LCQ (%) [N=126]

1: Women in Lower House (%)

2: Women in Upper House (%)

.803**

3: Legislated Candidate Quotas (Y/N)

.189*

.372**

4: Legislated Candidate Quotas (%)

.291**

.444**

.894**

*p < .05. **p < .01.

Figure 2. Efficacy of Quotas with Non-Compliance Sanctions

Figure 3. Efficacy of Quotas with Placement Mandates

4: LCQ (Y/N) [N=126]

Las Ministras

increase the likelihood that women will hold cabinet positions (H2). This claim supports the contagion theory of female representation, as outlined by Thames and Williams (2013). The authors claim that women’s representation in one governmental institution affects women’s representation in others like a contagion. Therefore, increasing women’s involvement in one branch of government will spread to others, due to the presence of more qualified candidates. Regionally, in 2005, women occupied 16.2% of cabinets in Latin American states while the current regional average is 26.63% (The World Factbook 2013-2015). Latin

represent women’s issues whether or not they wants to (Fleschenberg 2007; Franceschet 2005; Paxton and Hughes 2014). It is difficult for women to break out of these feminine positions due to the problem of gendered political institutions (Paxton and Hughes 2014). Gendered political institutions, defined as an organization that systematically reinforces gender differences by placing women in subservient positions or at a disadvantage, make it extremely difficult for women to hold cabinet positions, such as defense or chief of staff (Escobar-Lemmon and Taylor-Robinson 2009). Gender quota laws have been

America is an exemplum for female representation in cabinet positions as it far surpasses the global average of 17%. Much like female representation in the legislature, female cabinet representation continues to grow annually and the mean is projected to increase to 27.45% by 2017 (See Figure 4). Although there is a high level of female representation in Latin America, women in cabinet positions typically hold “soft portfolios,” or ministries of the government associated with femininity, such as women’s affairs, social services, and education (Paxton and Hughes 2014; Schwindt-Bayer 2010). Women represent the interests of female constituents more than men and are consequently expected to

blamed for the creation of gendered institutions and the marginalization of women. Additionally, male representatives who are hesitant to change often are seen as the reason that women are appointed to gender specific committees (Schwindt-Bayer 2010). Because some men feel threatened by the increasing number of women in politics, some of the political elite (who are predominantly male) isolate women to weaker positions to stop their progress. Despite these rare examples, the majority of male politicians are advocates for gender equality. In the analysis of female representation in cabinets in Latin America in Table 7, a strong positive correlation between the percentage of women in

Figure 4. Mean Women in Cabinet

60 the lower house and the percentage of women in cabinet positions in Latin America is found (H2). This evidence supports the claim that increased female representation in legislatures will lead to more women in ministerial positions. Although there is no significant correlation between female representation in the upper house and the percentage of women in cabinet positions, this does not nullify Hypothesis 2. Many of the states in the dataset are unicameral legislatures. Consequently, the most substantial correlations in the dataset in terms of female representation are in the lower house. Case Study: Argentina Looking to Argentina, women rise to power through the aforementioned method: first, by election to the legislature and then to cabinet positions by appointments. Women are able to climb the political ladder due to well-enforced quota laws have led to high levels of female representation across several branches of government. Historically, gender quotas have been a large part of Argentine culture. The first informal gender quota laws were utilized by Eva Peron, wife of President Juan Peron, in the 1950â&#x20AC;&#x2122;s. Her efforts resulted in 15% female representation in the Chamber of Deputies, higher than any other legislature in the world at the time (Paxton and Hughes 2014).

Additionally, Argentina was also the first state to adopt a legislated candidate quota in 1990. This law is called the Ley de Cupos, or â&#x20AC;&#x153;Law of Quotas,â&#x20AC;? and it ensures that a minimum of 30% of all political candidates must be women (Paxton and Hughes 2014). After the implementation of the Ley de Cupos, female representation in the Argentine lower house of congress, the Chamber of Deputies, rose from 4% to 21%. This success led to the implementation of quota laws for provincial and municipal elections (Paxton and Hughes 2014). Due to their success, many states took after Argentina, as countries all over the world began to adopt these affirmative action measures to increase female representation in their respective governments. Legislated candidate quota laws have maintained high levels of female representation in Argentina in the Chamber of Deputies. Currently, the Argentine Chamber of Deputies is 36.6% women and in the past ten years, the mean representation was 37.1% (Alnevall et al. 2015). Additionally, women make up 39% of the Argentine Senate (Alnevall et al. 2015). In the past ten years, the average representation has been 37.5%, showing consistent high female representation in the both houses (Alnevall et al. 2015). Well enforced quota laws account for this high level of female representation. According to Argentine

Table 7. Correlation: Female Representation in Legislatures and Cabinets

Variable

1: Lower House [N=126]

2: Upper House [N=54]

1: Women in Lower House (%)

2: Women in Upper House (%)

.803**

3: Female Cabinet Members (%)

.362**

.263

*p < .05. **p < .01.

3: Cabinet Members [N=126]

Pi Sigma Alpha Undergraduate Journal of Politics

Electoral Code, Article 60 (3), lists of candidates submitted by political parties that do not comply with the 30% mandate will be rejected (Alnevall et al. 2015). Enforcement provisions ensure that women make it on the ballot, but placement mandates ensure women are fairly placed on lists. The placement policy, as upheld by the constitution states: “for every two men, a party must nominate one woman. Parties with less than two seat must to place a woman in the top two candidates; parties renewing more than two seats must have a woman in the first three positions” (Alnevall et al. 2015). These laws are also

at all levels of government in Argentina (Alnevall et al. 2015). Much like female representation in the legislature, female representation in ministerial positions is also fairly high. On average, Argentina has maintained 26.62% female representation in cabinets in the past ten years, with 27% of ministerial positions currently held by women (“The World Factbook” 2005-15). Despite the efforts to increase female representation in the executive, many of these women hold soft portfolios (Schwindt-Bayer 2010). Female politicians hold ministerial positions in the Ministry

well enforced. If a list of candidates does not conform to the requirements, an electoral judge will contact the political party that violated the law and they must change the order of the candidates in the list within 48 hours. If the party still does not comply, the court will appoint women candidates who are next on the party list to the list of candidates (Alnevall et al. 2015). Additional quota legislation has helped to strengthen women’s representation in Argentina. Subnational quotas, although varied throughout Argentina’s many provinces, have been instrumental in promoting women to regional legislatures. Subnational elections are subject to the same 30% quota law enforced at the national level. Regional governments uphold these laws, rather than the national government. Voluntary political party quotas from Partido Justialista, Unión Cívica Radical, Alanzia para una República de Iquales have further strengthened the federal efforts of equality political representation between genders. The Partido Justialista, or the Peronist Party, is one of the largest political parties in Argentina. Combined with other political parties with voluntary quotas, these parties make up a significant part of the legislature, ensuring high levels of female participation in congress. Due to subnational and voluntary quotas, female representation is reinforced

of Culture, the Ministry of Health, the Ministry of the Interior, and the Ministry of Social Development. However, a woman currently holds the position of Minister of Security, showing that women are gradually breaking the norms of gendered institutions (CIA World Factbook 2005-15). The future for female representation in Argentina is very promising. The Ley de Cupos, the country’s legislated candidate quota law, ensures that women have opportunities to hold legislative positions because it is well-enforced and ensure the equal placement of women on electoral lists (Jones 2009). As the quantitative findings suggest, Argentina’s increased levels of female representation in national and subnational legislatures increase the likelihood that women will be appointed to the cabinet. Finally, it is likely that female representation in legislatures in cabinets will continue to grow due to the overwhelming success of gender quota laws. Case Study: Brazill Brazil serves as an example of the implementation of symbolic quota legislation. Because of this, women must rely on their peers and respective political party to transcend cultural and structural biases that hinder female representation. Brazil has a long-standing history of gender inequality. It was not

Adams

until 1988 that men and women were declared equal in the eyes of the government (Htun and Power 2006). As gender quota laws spread throughout the region, however, Brazil joined the trend. In 1996, Brazil adopted a 20% legislated candidate quota (Hepburn and Simon 2006). In 1998, mandatory quotas were applied at the federal and subnational levels of government (Htun and Power 2006). Two years later, the national legislated candidate quota was raised to 30% (Alnevall et al. 2015). Despite the presence of a legislated candidate quota law in the Brazilian Chamber of Deputies,

Finally, the very nature of elections to the

female representation remains relatively low (Jones 2009). There are several reasons for this occurrence. First, political institutions are fragile and political parties are weak in Brazil, which hinders the stateâ&#x20AC;&#x2122;s ability to be an agent of change in terms of gender equality (Htun and Power 2006). Consequently, there are no restrictions or penalties for political parties that do not adhere to the quota laws (Hepburn and Simon 2006). Also, there is a loophole in Brazilian electoral law that allows a political party to nominate a list of candidates without any women. Legally, the legislated candidate quota percentage applies to the overall number of positions on the list of candidates that a party may present in a district. For any given election, a party may nominate 150% to 200% of the possible elected positions (Htun and Piscopo 2014; Jones 2009). Consequently, it is absolutely legal for a political party to submit a list of candidates that excludes women entirely if it nominates candidates for 70% of the possible seats (Jones 2009). This aspect of the law was repealed in 2009, but the government has done little to intervene when the gender quota laws are not upheld (Htun and Piscopo 2014). Due to these many reasons, female representation in the Chamber of Deputies is only 9.9% (Alnevall et al. 2015).

percentages of women in the lower house of the

Chamber of Deputies in Brazil is not conducive to women, as these elections are quite competitive. For example, in the 2002 election cycle, 4,901 candidates ran for a mere 513 seats in the Chamber. There are twenty-seven electoral districts in Brazil and only seven of them nominated less than 100 candidates. Because parties win seats, not individuals, politicians of the same party compete among themselves for the votes their parties obtain (International IDEA 2005-2015) Typically, women do not do win where competition is high, further accounting for low legislature.

The 30% legislated candidate quota for the

senate is also largely ineffective in increasing female representation. Female representation in the senate is 14%, well under the global average of 22% (Alnevall et al. 2015). As is the case for the Brazilian Chamber of Deputies, the quota legislation that affects the senate is poorly enforced. For instance, if the minimum percentage of 30% of female representation is not met, candidates of the over-represented sex can be removed. However, these candidates cannot be replaced by women, which defeats the purpose of the legislation. Also, this law only applies if the party submits a list of candidates with the maximum number per constituency, 150-200% of the contested seats (Alnevall et al. 2015).

The 30% subnational legislated candidate

quotas that were instated at the state and municipal level in 1995 are ineffective for many of the same reasons as national legislated candidate quotas (Htun and Piscopo 2014). In municipalities and subnational elections in Brazil, there is no enforcement of these laws. This further shows that gender quota laws are mere gestures by elites to legitimize their parties and

Las Ministras

government in Brazil (Alnevall et al. 2015). Despite the

representation in legislatures and cabinets has doubled

present failures in their implementation of quota laws,

worldwide in the past twenty years, but few women

Brazil mandated in 2009 that 10% of the free broadcast

have held presidential offices. In Latin American,

time and 5% of public campaign funds of political

however, women have been able to occupy these offices

parties must be used solely to promote the female

with more success than their female counterparts in

participation in politics (Alnevall et al. 2015).

other regions, despite the Machismo and Marianismo

in Central and South America.

Due to the inefficiencies of Brazilian legislative

quota laws, female representation in the cabinet is low. Additionally, only 18% of cabinet positions are held by women, which is to be expected due to the lack of female representation in the legislature (Alnevall et al. 2015). Similar to Argentina, the majority of women in the cabinet hold soft portfolios, including the Ministry of Agriculture, Livestock and Supply, the Ministry of Education, the Ministry of the Environment, the Ministry of Social Development and Hunger Alleviation, the Secretariat for Promotion of Racial Equality, and the Secretariat for Womenâ&#x20AC;&#x2122;s Rights. Like Argentina, one woman holds a traditionally masculine institution: the Secretariat of Strategic Affairs (Alnevall et al. 2015).

Unlike Argentina, Brazilian female politicians

cannot rely on legislative experience to climb the political latter because women are unable to get elected. As previously mentioned, gender quota laws are ineffective and poorly enforced. This makes it very difficult for women to gain the necessary experience to become a viable candidate for appointment to the cabinet. Consequently, women must rely on political allies rather than holding a seat in the legislature to hold a seat in the cabinet.

To address this anomaly, it was hypothesized

that the implementation of well-enforced quota laws has increased female participation in legislatures, leading to an increased likelihood that women will be appointed to cabinet positions. In the quantitative and qualitative analyses, it was found that legislated candidate gender quota laws increased female representation in Latin American legislatures. Not only was the presence of these laws related to higher levels of female representation in both the upper houses and lower house of legislatures, but the high percentages that these laws entailed, the more women held seats in congress. Furthermore, the efficacy of these laws increased with the presence of enforcement policies, such as placement mandates and non-compliance sanctions. Finally, increased female representation in the legislature was significantly correlated with high percentages of women in ministerial positions.

The qualitative analyses supported the initial

findings. After analyzing Argentina and its respective quota legislation, it is evident that ascension through the legislature to cabinet positions is very likely for Argentine women. Quota laws with strict placement mandates and enforcement exist in all legislatures in

Conclusion

Argentina, insuring their efficacy and female political

the failures of quotas when poorly enforced. Because

Politicians, as the representatives of the

people of a state, are the real agents of change in a country. Because of this power, women have been excluded from legislative and ministerial roles. Female

participation. Brazil, on the other hand, demonstrates women have difficulty obtaining seats in the national legislature, women rely on political parties and allegiances to get appointed to cabinet positions.

Despite contrary works of other scholars,

Although the world is moving toward gender

this analysis was able to refute the findings of

equality in political representation, this analysis,

other alternative hypotheses for increased female

coupled with the existing literature on female

representation in legislatures. Gross domestic product,

representation, makes it apparent that there is still

labor force participation, and education levels have no

a long way to go until gender equality is achieved.

relationship to higher percentages of women holding

However, the implementation of gender quota laws

political office. However, a positive relationship

and a push toward gender equality in other aspects

between female representation and gender equality

of society may be able to help expedite this process.

was discovered as well as a negative relationship

Despite low levels of global female representation,

between democratization and female representation in

Latin America’s increased percentages of women at all

the lower house.

levels of government serve as an example for the rest

of the world and the topic of study for many years to

In terms of future scholarship, this analysis

shows a very interesting paradox that has received little

come as scholars try to address the gaps in the current

attention from the experts in the field: the relationship

scholarship.

between increased female representation in the

Notes

legislatures of non-democratic regimes. Although some scholars have claimed that women’s roles in autocratic governments, although common, are not substantial, it is essential to know why women are appointed or elected to these positions. In terms of electing women to legislatures and appointing women to leadership roles, democracies had much to learn from non-democratic regimes.

Roughly translates to “before all else, mothers”.

The states analyzed include Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, the Dominican Republic, Ecuador, El Salvador, Guatemala, Guyana, Haiti, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay, and Venezuela. 2

Additionally, further research should be

done on female presidents and prime ministers. As previously mentioned, few scholars have analyzed how women come to hold executive offices in Latin America, solely focusing on how women become ministers. There is a significant gap in the literature regarding the ascension of women to presidential roles. Although this analysis offers some explanation as to how previous women have risen to power and that gender quotas laws with an legitimate legal framework increase a woman’s likelihood of rising to the executive through representation in the cabinet, further research should explain why Latin America has seen more presidents than Europe and North America.

References Alnevall, Christina, Emma Frankl, Lenita Freidenvall, Mona Lena Krook and Anja Thorup Nordlund. 2015. “quotaProject: Global Database of Quotas for Women.” Stockholm University, InterParliamentary Union and International IDEA. http://www.quotaproject.org/index.cfm. (Accessed September 03, 2015) Baldez, Lisa. 2007. “Primaries vs. Quotas: Gender and Candidate Nominations in Mexico, 2003.” Latin American Politics and Society 49(3): 69-96.

Pi Sigma Alpha Undergraduate Journal of Politics

Casas-Arce, Pablo, and Albert Saiz. 2015. “Women and Power: Unpopular, Unwilling, or Held Back?” Journal of Political Economy 123(3): 641-669.

Gray, Tricia. 2003. “Electoral Gender Quotas: Lessons from Argentina and Chile.” Bulletin of Latin American Research 22(1): 52-78.

Clayton, Amanda. 2015. “Women’s Political Engagement Under Quota-Mandated Female Representation: Evidence from a Randomized Policy Experiment.” Comparative Political Studies 48(3): 333-369.

Hepburn, Stephanie, and Rita J. Simon. 2006. Women’s Roles and Statuses the World Over. Lanham, MD: Lexington Books.

Deutsch, Sandra McGee. 1991. “Gender and Sociopolitical Change in Twentieth-Century Latin America.” Hispanic American Historical Review 71(2): 259-306 Escobar-Lemmon, Maria, and Michelle Taylor-Robinson. 2009. “Getting to the Top: Career Paths of Women in Latin American Cabinets.” Political Research Quarterly 62(4): 685-699. Escobar-Lemmon, Maria, Michelle Taylor-Robinson, and Leslie Schwindt-Bayer. 2014. “Empirical Insights from Legislatures and Cabinets in Latin America.” In Representation: the Case of Women, eds. Escobar-Lemmon, Maria Escobar-Lemmon and Michelle. Taylor-Robinson. Oxford: Oxford University Press, 205-226. Fallon, Kathleen M., Liam Swiss, and Jocelyn Viterna. 2012. “Resolving the Democracy Paradox: Democratization and Women’s Legislative Representation in Developing Nations, 1975 to 2009.” American Sociological Review 77(3): 380-408. Fleschenberg, Andrea. 2007. “The Path to Political Empowerment? – Electoral Gender Quotas in South Asia.” Pakistan Journal of Women’s Studies 14(2): 1-27. Franceschet, Susan. 2005. Women & Politics in Chile. Colorado: Lynne Rienner Publishers.

Hinojosa, Magda. 2012. Selecting Women, Electing Women: Political Representation and Candidate Selection in Latin America. Philadelphia: Temple University Press. Htun, Mala C. 2002. “Women in Political Power in Latin America.” International IDEA, Women in Parliament, Stockholm, Sweden. http://www. idea.int/publications/wip/upload/Chapter1Htun-feb03.pdf. (Accessed September 18, 2015) Htun, Mala, and Jennifer M. Piscopo. 2010. Presence Without Empowerment? Women in Politics in Latin America and the Caribbean. Brooklyn, NY: SSRC. Conflict Prevention and Peace Forum, 1-24. http://webarchive.ssrc.org/pdfs/ Mala_Htun_and_Jennifer_M._Piscopo-Presence_without_Empowerment_CPPF_Briefing_ Paper_Dec_2010_f.pdf (Accessed September 18, 2015) Htun, Mala, and Jennifer M. Piscopo. 2014. Women in Politics and Policy in Latin American and the Caribbean. Working Papers on Women in Politics, Conflict Prevention and Peace Forum, 1-42. http://webarchive.ssrc.org/working-papers/ CPPF_WomenInPolitics_02_Htun_Piscopo.pdf (Accessed September 19, 2015) Htun, Mala, and Timothy J. Power. 2006. “Gender, Parties, and Support for Equal Rights in the Brazilian Congress.” Latin American Politics and Society 48(3):83-104.

Adams

International Institute for Democracy and Electoral Assistance (International IDEA). 2015. Electoral System Design: Electoral Systems Worldwide. http://www.idea.int/esd/world.cfm (Accessed October 17, 2015)

Krook, Mona Lena and Pippa Norris. 2014. “Beyond Quotas: Strategies to Promote Gender Equality in Elected Office.” Political Studies 62(1): 2-20.

Inter-Parliamentary Union. 2005-2015. Women in Parliament. http://www.ipu.org/wmn-e/classif-arc. htm (Accessed October 17, 2015)

Marshall, Monty G. and Ted Robert Gurr. 2013. Polity IV. Edited by The Center for Systemic Peace and the Integrated Network for Social Research. http://www.systemicpeace.org/polity/polity4. htm (Accessed October 17, 2015)

Inter-Parliamentary Union. 2015. “Sluggish Progress on Women in Politics Will Hamper Development.” http://www.ipu.org/press-e/pressre-

Mill, John Stuart. 1884. A System of Logic, Ratiocinative and Inductive: Being a Connected View of the Principles of Evidence and the Methods of Scien-

lease201503101.htm. (Accessed November 01, 2015) Iversen, Torben, and Frances McCall Rosenbluth. 2010. Women, Work, and Politics: The Political Economy of Gender Inequality. New Haven: Yale University Press. Jalalzai, Farida. 2014. “Gender, Presidencies, and Prime Ministerships in Europe: Are Women Gaining Ground?” International Political Science Review 35(5): 577-594. Jones, Mark P. 2009. “Gender Quotas, Electoral Laws, and the Election of Women: Evidence from the Latin American Vanguard.” Comparative Political Studies 42(1): 56-81. Kittilson, Miki Caul and Leslie A. Schwindt-Bayer. 2012. The Gendered Effects of Electoral Institutions: Political Engagement and Participation. Oxford: Oxford University Press. Krook, Mona Lena, and Diana Z. O’Brien. 2012. “All the President’s Men? The Appointment of Female Cabinet Ministers Worldwide.” Journal of Politics 74(1): 839-855.

tific Investigation. London: Longmans, Green, and Company. Murray, Rainbow. 2013. “Quotas, Citizens, and Norms of Representation.” Politics & Gender 9(3): 304309. O’Brien, Diana Z. 2015. “Rising to the Top: Gender, Political Performance, and Party Leadership in Parliamentary Democracies.” American Journal of Political Science 59 (4): 1022-1039. Paxton, Pamela M. and Melanie M. Hughes. 2014. Women, Politics, and Power: A Global Perspective. Thousand Oaks: SAGE Publications. Reynolds, Andrew. 1999. ‘‘Women in the Legislatures and Executives of the World.’’ World Politics 51(4): 547-572. Schwindt-Bayer, Leslie A. 2009. “Making Quotas Work: The Effect of Gender Quota Laws On the Election of Women.” Legislative Studies Quarterly 34(1): 5-28. Schwindt-Bayer, Leslie A. 2010. Political Power and Women’s Representation in Latin America. Oxford: Oxford University Press.

Las Ministras Schwindt-Bayer, Leslie A. 2011. “Women Who Win: Social Backgrounds, Paths to Power, and Political Ambition in Latin American Legislatures.” Politics & Gender 7(1): 1-33. Steinberg, Blema S. 2001. “The Making of Female Presidents and Prime Ministers: The Impact of Birth Order, Sex of Siblings, and FatherDaughter Dynamics.” Political Psychology 22 (1): 89-110. Stevens, Evelyn P. 1973.“Mariansimo: the Other Face of Machismo in Latin America.” The Female and Male in Latin America: Essays 1: 89-101. Thames, Frank, and Margaret S. Williams. 2013. Contagious Representation: Women’s Political Representation in Democracies around the World. New York: New York University Press. The World Bank. 1990-2015. Social Development Indicators. http://data.worldbank.org/indicator/ SE.ENR.TERT.FM.ZS/countries%5D (Accessed November 03, 2016). The World Factbook. 2005-2015. Washington, DC: Central Intelligence Agency, 2013. https:// www.cia.gov/library/publications/the-worldfactbook/index.html (Accessed November 03, 2016). Viñas, Silvia. 2014. “Latin Women Take the Helm.” World Policy Journal 31(1): 39-47. Waylen, Georgina. 1993. “Women’s Movements and Democratisation in Latin America.” Third World Quarterly 14(3): 573-589 World Economic Forum. 2006-2014. Gender Gap Index. http://www3.weforum.org/docs/GGGR14/ GGGR_CountryProfiles.pdf (Accessed November 03, 2016).

UNESCO, Institute for Statistics 2015. Country Profiles. http://www.uis.unesco.org/ DataCentre/Pages/country-profile. aspx?code=VEN&regioncode=40520 (Accessed October 20, 2016).

Public Opinion of Labor Unions in the United States Baili Volluz Saint Louis University The research seeks to better understand how Americans form opinions on labor unions. There are two main schools of thought on the topic: rational choice and sociological/environmental factors. This work adds sociological and cultural factors to rational choice to fill the gaps in rational choice theory, specifically when rational choice theory fails. By including both sociological rational choice factors, the model should create a more nuanced picture of opinion formation including relationships between income, occupation, political identification, and respondents’ opinion of unions. To examine the latter expectations, this research examines time series data of individual respondents by employing regression analysis. In addition, interviews of people in the workforce are conducted, asking about their opinions of labor unions. By utilizing both qualitative and quantitative approaches, this research shows that income, membership density, education, political party, occupation, change in gross domestic product, and union membership were significant in union opinion formation. Introduction Today, the American people today have less favorable opinions of labor unions, which have been in steady decline since the 1960s (Clawson and Clawson 1999, 97). The powerful unions of the 1940s and 1950s are a thing of the past. The 1940s and 1950s experienced an intense manufacturing period and union membership was highest among “male, blue-collar breadwinners working lifelong jobs in large firms” (Western and Rosenfeld 2012, par. 3). Specifically, union membership has declined from 29.3 percent in 1964 to 11.1 percent in 2014 (Hirsch, Macpherson, and Vroman 2001; U.S. Bureau of Labor Statistics 2015, par. 1). Labor unions supported policies such as the forty-hour workweek, an end to child labor, safe working conditions, better wages, and paid vacation. When at their most powerful, the increase in wages brought on by union membership resulted in a strong middle class during the 1960s and 1970s (Western and Rosenfeld 2012, par.3). State and federal minimum wages are also affected by labor unions. According to

Bartels (2008, 226) “the decline in union membership since the late 1940s has probably depressed the real value of the minimum wage by about 40%”. Simply put, this means that a person making minimum wage in the 1940s would be able to buy ten gallons of milk and a minimum wage worker of today would only be able to buy four for the same number of working hours. This affects the middle class in that if minimum wage workers cannot make a decent wage, they cannot expect to enter the middle class, causing the middle class to stagnate or shrink. Bartels (2008, 240) furthers his claim by saying, “the fate of the minimum wage has hinged even more crucially on the fortunes of the Democratic Party’s most important ally in this policy domain, organized labor”. Western and Rosenfeld (2012) argue that labor has to be both politically and economically powerful for the middle class to bounce back. Representation of workers in the political sphere has historically been carried out by union organizations, and, with the decline in union power and the supremacy of big business, one sees a distinct lack of representation

Pi Sigma Alpha Undergraduate Journal of Politics

of middle class interests. Labor unions have given workers many things they otherwise would not have. Many believe that unions created the middle class and led to many laws being passed that protect bluecollar workers. As unions disappear, however, the U.S. may see a decline in the middle class (Western and Rosenfeld 2012). With this decline in union presence in the U.S., so too has come a decline in approval of unions. Public support of labor unions has declined “…steadily since its mid-1950s high point” (Heshizer 1985, 225). This could be because of the decreasing presence of unions

factors influence attitudes toward unions in the U.S., examining two main schools of thought regarding how people form opinions (e.g., Downs 1957; Campbell, Converse, Miller and Stokes 1976; Heshizer 1985; Hogler and Henle 2011; Saporta and Lincoln 1995). Rational choice theory focuses on how selfinterest, specifically economic factors, affects opinion formation in an environment of uncertainty and scarce information. Rational choice fails when people act contrary to their self-interest where we believe they know what that interest is. When this occurs, theories of sociological factors work much better (Fatas-

in the labor force. In fact, “approval fell to an all-time low of 48% in 2009… Since then, labor union approval has remained at bare majority levels” (Dugan 2013, par. 3). There are many explanations for the declining approval of labor unions. Changes in the economic structure of the U.S. have caused a decline in industrial manufacturing jobs, resulting in fewer people holding union jobs. People may also see the potential issues unions can cause, such as historical corruption and other potential difficulties to business owners. Unions make the workforce more expensive and some argue have a negative effect on the economy (Clawson and Clawson 1999; Heshizer 1985; Hogler and Henle 2011). People might also be concerned about the few union members being isolated from some economic hardships while nonunion workers suffer (Dark 2011, 526). Union disapproval could be a result of unions disappearing, unions considered as irrelevant, or many other factors. The way people view unions is reflective of how they view their self-interest; thus, with less unions, there could be less self-interest tied to unions. Even if unions contributed to giving the American people many things, including the lifestyle of middle class, unions could be viewed to result in economic problems. This research seeks to understand what

Villafranca, Saura, and Vazquez 2011; Frank 2005). Sociological factors include upbringing, elite influence, and social influences. Sociological factors and rational choice theories, when combined, create a much clearer picture of public opinion influences than they provide on their own. These theories translate into the socioeconomic factors outlined in the hypotheses stated below. Previous Research People form their opinions toward labor unions in a variety of ways. There are two main schools of thought on how opinions form that is considered here: rational choice and sociological factors. Rational choice explanations focus on how people form opinions based on what they see as their self-interest; however, people make decisions based on factors other than their actual self-interest (Campbell et al. 1976; Downs 1957). This is also an issue because of the risk of tautology that comes with rational choice theories (i.e., anything an actor does can be justified as rational if the definition is used too broadly). Sociological factors focus on how cultural and environmental factors, such as upbringing, group influence, and elite influence, influence how people develop their opinions toward political issues and objects. Sociological factors

Volluz

can cause a separation of perceived and actual selfinterest, which will be defined below. Sociological factors can provide further clarity to people’s decision-making that includes elite influence and group influence (Campbell et al. 1976; Frank 2005; Moussaid Mehdi, Kaemmer, Analytis, and Neth 2013). First, this research provides further detail on rational choice theory then moves to sociological factors. Second, large economic forces, such as recessions and depressions, are explained as their own set of factors, which can be considered both group-centric and selfinterest. Theories of Self-Interest Rational choice theory is the basis for the selfinterest component of this argument. Self-interest assumes that there are many reasons that a rational actor makes choices, but that those choices are all the result of a rational cost-benefit analysis (Downs 1957, 4). The rational actor model is qualified later by Downs (1957) and Key and Cummings (1966). Key and Cummings (1966, 2) discuss elections, stating, “The voice of the people is but an echo. The output of an echo chamber bears an inevitable and invariable relation to the input”. This applies to union opinion in that rational actors may not be able to act exactly how they would like and must instead act in accordance with limited choices. Unfortunately, actors are forced to make decisions in an imperfect world. There are several factors that cause actors to act contrary to what we believe their interests to be. Key and Cummings (1966, 110) explain: “Differences in voters’ interest, in their range of information, in the orientation of their attention, in their firsthand experience, and in their exposure to communications produce enormous variation in their

perceptions of events and, consequently, in their appraisals of the alternatives posed by the electoral system” (see also Downs 1957). The uncertainty principle operates on a premise of limited information availability. A rational actor does a cost benefit analysis in making any given decision to either expend resources to acquire additional information or not. Any expenditure of energy to gather further information takes time. Time is a “scarce resource” that “…forces decision-makers to select only part of the total available information for use in making choices” (Downs 1957, 209, 219). Information combats uncertainty, which allows rational actors to make decisions. The rational basis with which actors act arises from the information actors receive from their surroundings (Downs 1957, 79). The average voter, who has no political science background and limited energy to spend on gathering political information, forms their opinions based on how they are doing under the current administration (Key and Cummings 1966, 111). This translates to opinion formation quite easily in that people form opinions in the real world with limited information and must, therefore, use what little information they have to decide what is best for them (Downs 1957, 207). People may make poor choices because they “… are not always aware of what the government is or could be doing, and often they do not know the relationship between government actions and their own utility incomes” (Downs 1957, 80). The latter could include how citizens feel about the advantages and disadvantages of labor unions. People must make decisions every day, and in doing so, must determine how much of their time, money, and brainpower they wish to spend on obtaining information. There is a shortcut to information gathering: ideology. Instead of being

Public Opinion of Labor Unions

forced to spend hours researching each political candidate, voters rely on the representative ideology of each candidate, basing their candidate selection purely on ideology. In this scenario, the vote choice creates a disconnect between the voter and the representatives (Downs 1957, 99). Citizens develop policy preferences based on “…their informative situation and individual beliefs on the sources of economic success” (Key and Cummings 1966, 59; Pitlik, Schwarz, and Brandl 2011, 273). The above explains part of the distinction between actual and perceived self-interest. If actors

to unions in that if people do not see unions as representative of their economic needs, then they may vote for candidates with policies that are less sympathetic to unions. It is well known that opinions toward unions are not the only driving force behind voter choices. There are many other factors that influence vote choice, such as religious beliefs, attitudes toward healthcare, and foreign policy opinions (Campbell et al. 1976; Key and Cummings 1966). Generally “…all that is required for an accurate determination of individual attitudes toward a specific policy measure is an assessment of the

are receiving biased information, they may act in accordance with what they are told rather than what they actually believe. Actual self-interest constitutes interests that directly benefit the respondent in obvious ways whereas perceived self-interest constitutes things that either indirectly benefit the respondent. Perceived self-interest can be hard to distinguish from actual self-interest in practice but is an important concept nonetheless. It is a matter of personal judgment, which is not always easy to glean from data alone (Key and Cummings 1966, 63). Pitlik, Schwarz, and Brandl (2011, 273) make the point that, “[w]hat matters eventually is what people believe to be in their selfinterest”. Americans will vote for the candidate they see as most beneficial to their current economic situation. Key and Cummings (1966, 3) argue that in elections “…people did, indeed, appear to vote their pocketbooks”. When asked about particular policy measures of interest in the 1960 election, people cared most about “domestic economic policy and unemployment… and close to one American in every five put those issues above even war or peace and relations with the Communist world” (Key and Cummings 1966, 131). The economy is clearly of great concern to the American people. This pertains

economic consequences for personal well-being” (Pitlik, Schwarz, and Brandl 2011, 271). Economic interests can be used to explain overall self-interest. Many voters do not have the time or interest to truly examine every policy measure and instead key off of particular past measures and proposed new measures (Key and Cummings 1966, 60). Rational choice is a good beginning to analyzing opinion formation but cannot give a complete depiction of opinion formation. In any respondent set, there are always those that do not conform to any set of norms, and, in addition to the outlier respondents, respondents are acting with limited, subjective information which can further muddy the waters (Key and Cummings 1966, 7, 70). Thus, the rational choice model “provides partial explanations of human behavior, and the theory should be judged on its ability to account for some aspects of human behavior, neither more nor less” (Blais 2000, 10). Theories of Social and Cultural Influence For even the most rational actor, sociological factors have a profound effect on personal opinions. This puts yet more focus on external factors rather than internal ones, such as self-interest. Upbringing, group membership, and elite influence are three

72 important subsets of this school of thought and are discussed below.

Upbringing is very important to opinion

formation because the influence of parents on their young children is stronger than the influence of others (Bardes and Oldendick 2012, 89). Bardes and Oldendick (2012, 88) state, “The process of learning about the political system is generally called political socialization”. Political socialization is “defined by Roberta Sigel [as]… ‘the process by which people learn to adopt the norms, values, attitudes, and behaviors accepted and practiced by the ongoing system’” (Bardes and Oldendick 2012, 88; Sigel 1965, 1, 2). This influences how a person behaves based upon what the political system required from them to sustain itself. Self-interest can partially originate from parental influence. For example, one may find unions beneficial to them because one’s father was a union carpenter and the value of the union to him was impressed upon one from a young age. Also, children are exposed to the views of their parents by sheer proximity more than any other source. Parental income is also a defining factor in childhood experiences, which may open or close off opportunities for them. Children are with their parents continually and are dependent upon them. It is no wonder why children value the opinions of their immediate family (Bardes and Oldendick 2012, 89; Sigel 1965, 1, 2). Bardes and Oldendick (2012, 89) refine this idea by stating, “the influence of the family is paramount in the political socialization of young children...mothers and fathers have maximum exposure with children, communicate with the child frequently, and, most important, experience maximum conditions of receptivity”. Being in a union household as a child may influence the information one has about unions and the perceptions one has about who union members are and what they do.

Citizens become involved with more groups as they mature, which can influence their attitudes and behaviors. Moussaid et al. (2013, 1) assert, “[W]hen facing the opinions of peers on a given issue, people tend to filter and integrate the social information they receive and adjust their own beliefs accordingly”. People talk to one another and this can cause opinions to fluctuate as one moves from group to group. In a group, individuals “…construct a network according to the relationships among them” (Wu, Hu, and He 2013, 3). This network of relationships is also a network of information. People are able to bounce ideas off of each other and ferret out which combination aligns best with their own views. There are two main types of groups: primary groups and secondary groups. Primary groups are family and friend groups whereas secondary groups are larger groups not based on personal relationships. One explanation of group interactions generally is based on the types of individuals that form groups. If group members have opinions that are too close in nature, then group members learn nothing from one another. If opinions are too far apart, there will not be a group consensus of opinion. The ability for opinion consensus also depends on the extremity of group members’ opinions (Wu, Hu, and He 2013). The types of groups that influence opinion formation the most are important. Citizens move among and between groups in their daily lives. Thus, the degree to which the opinions of others affects their own opinions is crucial to understanding how external factors can lessen rational, purely self-interest decision making. Secondary group influence “…can cause a large-scale, unidirectional shift in the partisan division of the national vote…” because of their size and political lobbying power (Campbell et al. 1976, 162). This influence is unique because it accounts for many mass political opinion movements that cannot be

Pi Sigma Alpha Undergraduate Journal of Politics accounted for otherwise through simple self-interest (Campbell et al. 1976, 162). Some examples of mass political movements include the Women’s movement, the Civil Rights movement, and the Labor movements of the early 20th century. These massive shifts in voting decisions can influence elections outcomes and shape the country’s future. This specifically pertains to unions because they are considered secondary groups. Union membership creates a special point of view for actors. Labor unions’ main purpose is to “force management to provide more liberally for the worker” (Campbell et al. 1976, 161). Even so, they take on a political role in their advocacy efforts. How members and nonmembers view these political acts is directly tied to the strength of opinion cohesiveness among union members. There is evidence that union members change their opinions of their union over time. Campbell et al. (1976, 178) explain: “The longer an individual has belonged to a union, the more likely he is to identify strongly with it, and we can find no other causative factors that begin to approach this relationship in strength”. One theory of group influence is twotiered. Groups influence outsiders because of their “wholeness”. Primarily, “groups have influence because we…think of them as wholes, and come to respond … to them in that form. In this sense, even people who are not members of a group may be influenced by the position that a group takes in politics” (Campbell et al. 1976, 161). For a union member, this may mean reading the Labor Tribune1 and reading the political section for which candidates the union endorses. For a church member this may mean discussing politics with their religious study group. For a bowler, this may mean discussing politics between frames during a tournament. Many people may look at groups as wholes and use them as “…reference points for the

formation of attitudes and decisions about behavior…” (Campbell et al. 1976, 161). Groups affect politics for insiders as well as outsiders. The characteristics attributed to some types of groups can have a distinct effect on politics – even when group identification in the traditional sense is absent. This is due to generalizations made about group members (Campbell et al. 1976, 166). For example, many pollsters see African Americans or Catholics as a specific voting block due to their perceived policy preferences based upon the group alone without regard to individual experiences. With traditional group identification among members, there is a heightened effect on decision making. Campbell et al. (1976, 166) assert “…shared membership provides a focus and direction for behavior that is lacking among non-group members who happen to be placed in the same life situation”. Group strength is an important factor in group influence. A “cohesive group” “…enjoy[s] high member identification…” (Campbell et al. 1976, 170). Member identification gives groups influence over members. As a union member, a citizen may see their membership as the single most important characteristic of their person. As such, they may vote based only upon the policies/candidate they see as beneficial to union survival. An apathetic group member is less likely to fall in line with group policy than a strongly identified member. Group identification strength is the “we feeling” members feel towards their group (Campbell et al. 1976, 168). This is important because “… a potent force in growth of group identification is simple contact and familiarity...” (Campbell et al. 1976, 179). Individuals within groups can be parceled out to account for group vote variation. Variation in group influence to each individual member may be attributed to life situations of members and length of participation (Campbell et al. 1976, 166). This is

Volluz

important because “the more highly identified a group member, the more likely he [is] to grant the group a right to engage in political activity” (Campbell et al. 1976, 177). If a member feels that the group is acting in their interests, they are more likely to tolerate the group lobbying for the group’s interests as a whole. The mandate for political participation must come from the members themselves or else upper group management risks mutiny. Engagement in political activity is important to promote group interests. Group legitimacy to members is important because there is “…evidence that the perceived relationship of the group to politics, like the relationship of the individual to the group, bears directly upon the strength of group forces in the field at the time of political decision” (Campbell et al. 1976, 172). Individuals are not exposed to the political system in a bubble of rationality nor in a bubble of only group influence (Campbell et al. 1976, 163, 164). Elite influence is an additional outside influence that matters in opinion formation. Elites have limited but sometimes striking influence on individual opinion because of their ability to frame issues in concise but often biased ways that are easy for the voting public to understand. Druckman and Nelson (2003, 741) also explain that elite framing is dependent upon the “…need to evaluate. Low- NE [Need to Evaluate] individuals and those who are highly knowledgeable appear more susceptible to elite framing effects… All of this suggests a more limited impact for elites than is often thought”. An individual’s “need to evaluate” is how much information the individual requires to make decisions and how recently that individual must have received the information. Elite framing is more prominently felt on the two extremes of the political spectrum: those that do not care and those that care a great deal. This is an important piece of the puzzle because it can also

derail the rationality of individuals acting in their interest. This may cause a person to misconceive the beneficence of labor unions to their purpose or to overestimate the union’s benefits. Elite framing can offer much insight into how individuals form their opinions; however, it can discount self-interest. In What’s the Matter With Kansas, Frank (2005) explains how American citizens often vote for one policy measure and end up receiving something else entirely due to the influence of elites. Some Americans vote for a candidate based on only one aspect of a platform, and often that is not economic in nature. Frank (2005, 7) argues, “Citizens vote their cultural interests but receive economic change against their interests instead”. For these citizens, rational choice theory does not work and instead, we must turn to group influence. For some Americans, economic factors do not cross their minds when voting (Frank 2005, 7). This may explain why some Americans hold views that may be considered economically contrary to their personal self-interest. A citizen may decide that economics do not matter nearly as much as religion, the environment, or another issue of current cultural concern. This is contrary to the economics of pocketbook voting, but makes sense if one considers how people may be influenced with biased and incomplete information by elites. Economic Factors Specific economic circumstances often influence public opinion, which may affect attitudes toward labor unions as well. Economic factors could be considered self-interest when one’s pocketbook is affected, but could also be considered a sociological factor when perceptions of organizations, namely unions, are changed by rhetoric that can create an “us-versus-them” mentality for both members and non-members. Heshizer (1985, 225) explains “the height of union approval occurred during the 1950s

Public Opinion of Labor Unions when prosperity, low inflation, and general confidence in all institutions prevailed”. Economic downturns are a tumultuous time for people and as such often change their opinions on economic matters. Historic economic disasters, such as the oil shocks of the 1970s and the recession of 2008, are important economic events that may influence opinion. In 2008, the mortgage bubble burst and the U.S. economy went into a tailspin with a three percent drop in GDP in two years (U.S. Bureau of Economic Analysis, see Table 1.1.1.). The economic change suggested a new hope to strengthen unions. Unfortunately, for public and private sector unions, the recession instead caused a backlash (Dark 2011, 526). There are many theories regarding why union support declined during the recession, one of which focuses on sociological factors, specifically, the economic environment of the day (Dark 2011; McCartin 2013). Instead of galvanizing workers throughout the U.S., the recession created ammunition for union opponents. The recession permitted them “… to deploy a potent new argument. They contended that unionized government workers now constituted an ‘elite’ whose pay and pensions were not sustainable in a world of rapidly diminishing expectations, lost benefits, and wage stagnation among private sector wage earners” (McCartin 2013, 56). This view gained footing in 2008 because of the composition of the American economy and the change in union membership rates and union composition (public versus private) (U.S. Bureau of Labor Statistics 2015). Public sector unions are those working for the government and private sector unions are those involved in work for private businesses. In the U.S., “…the vast majority of employees cannot realistically hope to ever become union members, the relative advantage of the unionized workforce can emerge as a potential source of envy and resentment on the part of those excluded from such benefits” (McCartin 2013, 58-9).

There is one view that American workers are divided between private sector non-union workers and public sector unionized workers and that they are somehow at odds with each other because of the benefits received by union members (McCartin 2013, 56). Perceptions of union members began to reflect possible resentment from workers as “…the economic crisis began to shape a mainstream consensus around the proposition that public sector union contracts had become too expensive, and that unsustainable and under- funded union-won pension plans were undermining the finances of cities and states” (McCartin 2013, 56). Unions helped insulate members from the recession’s harshest effects. Non-union workers were “…experiencing severe job losses, pay cuts, and benefit reductions…” and came “…to view unionized employees as a favored insider caste unfairly protected from the pain of economic contraction” (Dark 2011, 529). This may cause workers to see a separation between themselves and union members. This could diminish any possibility of non-members viewing unions as a group in solidarity with them. Unions could be cast as an insular, elitist, and isolated group far removed from non-members’ everyday lives. This coupled with union opinion and membership being at its highest during one of the most prosperous times in American history points to opinions varying in accordance with economic conditions. A Theory of Opinion Formation This research asserts that a specific set of factors work to shape a person’s opinion on labor unions. Respondents form opinions in a variety of ways but certain socioeconomic triggers make respondents more likely to view unions through a particular lens (Key and Cummings 1966, 5). The questions that follow are: What factors influence attitudes toward unions? More specifically, what are the predictors of

76 union disapproval over time?” Opinions are formed through perceptions of self-interest, upbringing, elite framing, and peer opinion, as well as on the basis of larger economic factors. Citizens may see unions as a benefit, hindrance, or be indifferent. Pitlik, Schwarz, and Brandl (2011, 288) states, “…our estimates indicate that subjective, perceived self-interest is a stable factor in opinion formation while the influence of ideology is dependent on the nature of the tested policy measure”. A respondent’s opinion of labor unions can be determined based on factors of perceived self-interest. Specifically, sociological factors combine with actual

rather than a perceived benefit linked to culturally or group-influence interests. Education level is important to consider because “Jobs with lower educational requirements are… more likely to be unionized” (Hundley 1988, 195). Education is important because union jobs are often an option for people who have a lower level of education. They offer good pay for skilled and unskilled manual labor (i.e., these are not the only unions – merely a large portion of private sector unions). People feel positively about institutions they feel are beneficial to their economic situation. Where a person lives has an impact on their

self-interest to create one’s perceived self-interest. Downs’ (1957) model outlines how any given person should act if that person is acting rationally, and why they may not act in ways we think are rational. If Downs (1957) is correct, a person that would benefit from labor unions should think higher of them than a person who does not if they have adequate, accessible information. For a rational actor, many factors influence their perception of their self-interest. This research considers many of those individual factors in the proposed hypotheses. First, and possibly the most important, is the theory that people have a “tendency to vote the pocketbook’” (Key and Cummings 1966, 3). For public opinion, this translates to people forming opinions based upon their economic interests. Income may affect a person’s opinions on labor unions. A person who is part of the middle or lower economic class may see labor union membership as an opportunity for a stable, industrialbased job. This would be in their self-interest in that a stable income could potentially benefit them economically. A person of a higher economic class is more likely to work in white-collar or managerial job that are less likely to be unionized. This would be considered a matter of rational choice in that it concerns direct economic benefit to the respondent

attitudes and behavior in many ways. There are two components regarding where someone lives. First, if a person lives in an industrialized area, he or she is more likely to hold a union job (or likely knows someone who does). According to the U.S. Bureau of Labor Statistics (2015, par. 19), half of the country’s union members live in seven states: California, New York, Illinois, Pennsylvania, Michigan, New Jersey, and Ohio. The self-interest component of region includes access to union jobs and affiliation economically by proximity. The second component of region is sociological. People who live in states that are more individualistic will view unions as: “…inconsistent with individual responsibility and adherence to traditional norms of public governance” (Hogler and Henle 2011, 137). The ideas of regional differences in culture and collectivism originates from the region’s history and social structure (Vandello and Cohen 1999, 283). This is a matter of cultural perception and a part of group influence. Socioeconomic factors, such as occupation, geographic location, and political party, are just a few examples of traits that influence attitudes toward unions. People form opinions partially based on how they were raised: whether they were rich or poor,

Pi Sigma Alpha Undergraduate Journal of Politics the regional area in which they lived, and type of employment among close friends and family. These same factors determine what people perceive as their self-interest (Charles and Hurst 2003). Many children end up in relatively similar economic situations as their parents (Charles and Hurst 2003). If one grew up surrounded by family members who benefited from labor unions, they may have a more favorably view unions. Partisanship can be a factor in opinion formation in that labor unions are normally an ally of the Democratic Party (Bartels 2008, 240). This has been true since the time of the New Deal when President Franklin Roosevelt passed 1935 The Wagner Labor Relations Act which secured bargaining and organization power to workers and fundamentally changed the industrial paradigm (Key and Cummings 1966, 32). The larger economic conditions surrounding an actor have much to do with how they view unions. If a person feels a complete disconnect from unions, including the belief that they are insulated from the damaging effects of a recession, that disconnect could turn from indifference to the unionâ&#x20AC;&#x2122;s cause to a resentment of the benefits reaped by union members. Instead of recessions creating worker solidarity, they may create an environment of resentment for nonmembers and an insulating effect for members. This is separate from income because, even when larger conditions fluctuate, some citizens will always make more money and have managerial jobs. Hypotheses The theory of rational choice translates into the following hypotheses. These are based on how each respondent will answer based on what their demographics tell us about what should be in their interest. The following hypotheses fall under rational choice theory:

1. Citizensâ&#x20AC;&#x2122; who have higher income levels will be more likely to have lower approval ratings of labor unions than citizens with lower income levels. 2. Citizens who live in densely unionized areas will have higher approval ratings of labor unions than those in less densely unionized areas. 3. Citizens who have lower levels of education will have higher approval ratings of labor unions than citizens with higher levels of education. The following hypothesis falls under economic theory. Economic theory involves system level economic forces, such as how recessions and depressions should affect opinion of labor unions. 4. Citizens will have lower approval ratings of labor unions during economic recessions than during good economic times. The following hypotheses fall under sociological theory, which details the demographic factors that could influence a respondent to act less in favor of their actual self-interest and more in line with their perceived self-interest. 5. Citizens who are of management level in their careers will have lower approval ratings of labor unions than those who are not management level. 6. Citizens who are more Democratic will have higher approval ratings of labor unions than citizens who are more Republican. 7. Citizens who are labor union members, or who know labor union members, will have higher approval ratings of labor unions than those citizens who are not union members.

Volluz

Data and Method

State Union Density data (see Hirsch, Macpherson, and Vroman 2001). The empirical analysis used Interviews with Union Members roughly 20,000 respondents. This research utilizes both qualitative and The dependent variable is the respondent’s quantitative methods to examine attitudes toward labor attitudes toward labor unions, which is measured unions. First, the qualitative aspect of this research by a “feeling thermometer” that asks respondents focuses on connecting the empirical data to real world to rate their opinion of labor unions on a scale of experience of union and non-union workers with the 0-100 from lowest to highest. All of the independent public and their coworkers. Interviews were conducted variables were tested for relationships with the with individuals who potentially had interactions with dependent variable using multivariate regression. labor unions in order to add context to the empirical The independent variables considered were income data. The author conducted the interviews in the level, partisanship, level of employment, union fall of 2015. Overall, a total of two respondents were membership, education, labor union density, and interviewed in Illinois. An Institutional Review Board economic environment (See Table 1). Income level approved confidential interviews regarding attitudes was measured using a six-level percentile range index toward labor unions. The complete questionnaire is provided by the ANES 2010 Time Series. Respondents provided in Appendix A. In general, the interview self-select into income groups that are then placed questions asked respondents about their personal into a percentile index to better span across time. The views of labor unions and labor union members base category used in the regression analysis for this as well as how the respondent thought others view variable was “0-16th percentile”. The income variable members and labor unions as a whole. The possible concerns how much income a respondent’s family interviewees were established contacts, including has in a given year. The highest level of income was family members, friends of family members, personal roughly six hundred percent of the Federal Poverty friends, and professors. Five people were contacted and Line and is roughly twice the median household two people volunteered to be interviewed. Interviews income for Americans (DeNavas-Walt and Proctor were recorded for accuracy and flow of conversation 2014, 7). The latter is important to note because in order to take detailed notes afterwards. These of the implications of labor unions for the “middle interviews act to tie the quantitative analysis to the class”. Labor unions can be a market stabilizer that real world through the opinions and experiences of have been shown to allow the middle class to flourish workers. (Western and Rosenfeld 2012). Regarding political Empirical Data

predispositions, political party ties can explain much about where a respondent’s opinion fits on the political Second, a quantitative data analyses was spectrum. For this, the party identification variable also employed to provide empirical evidence for the in the ANES Time Series (2010) dataset was utilized. proposed hypotheses. To examine attitudes toward Party identification is measured as a seven-point scale labor unions, data were drawn from four sources: the American National Election Survey (ANES) 2010 Time from 1 = Strong Democrat, 4 = Independent, to 7 = Strong Republican. For this categorical variable, Series Data, the U.S. Bureau of Economic Analysis (BEA) 1969-present, and the Georgia State University “independent” was used as the base category.

Public Opinion of Labor Unions

Next, level of employment is included as an independent variable, as it labor union membership is especially beneficial to lower level employees because labor unions provide bargaining power to those that otherwise would work without a contract. Higher, management level employees are less likely to view labor unions favorably because they either do not need labor unions in order to make a good wage, or see labor unions as unnecessarily tedious to deal with at work because labor unions can cause contractual headaches for management. The ANES Time Series (2010) asked respondents to select their occupation

in which “west” was used as the base category in the regression model. Gross Domestic Product (GDP) is also included as an independent variable from the Bureau of Economic Analysis “GDPPC”. A measure of GDP acts as an economic indicator to allow for analysis of how economic conditions affect public opinion of labor unions. Last, the empirical analysis controls for gender (0 = male; 1 = female).2 It has been established in voting and public opinion literature that there is a gender gap in voting behavior and issue importance between men and women (Norrander and Wilcox

category which were collapsed into the three levels of employment. These categories were management, non-management, and other with non-manager being the base category. In addition to level of employment, it is important to examine how union membership, particularly if an individual is a union member or not, influence union approval. In this research, union membership was measured as a categorical variable ranging from “no one” (no one in the family is a member) to “respondent plus” (the respondent and other family members are members). For this, the base category used was “no one” (no one a union member). I also include two independent variables, particularly Membership Density (percentage of union membership by state) and “Where did Respondent Grow up?” (region) to examine the link between geographical location and labor union opinion. These variables allow a test of whether labor union opinion is affected by the density of labor unions in the respondent’s area, as geography is an important component. Geography could also be a distorting factor in that labor unions are very concentrated geographically in the U.S. as well as differences in cultural perceptions of collectivity (Dark 2011, 527; Vandello and Cohen 1999). For “Where did Respondent Grow Up?”, this is a categorical variable

2008). According to Norrander and Wilcox’s (2008) research women should support unions more because of their acceptance of group membership as an ideal. Table 1 reports a matchup of the areas of interest discussed previously in the theoretical design and the variable names for variables analyzed. Analysis and Results Interview Analysis: Administrator and Labor Union Member The interviewees included a male administrator in education and a male union member in an industrial setting. The interviews allow a better understanding of what people in the real world think of unions and their members. See Appendix A for a list of the interview questions. The administrator had experienced unions on both sides of the bargaining table as a negotiator of contracts for the union as well as an administrator negotiating for the other side. The union member talked at length about what he saw as the potential benefits of unions saying, “… the more people that are unionized the less of an income gap we have and the more opportunities people have. It just it makes society better as a whole I think.” Both interviews gave logical, but fundamentally different opinions on why they think people see unions less

80 Table 1. Independent and Dependent Variables Variable Name Labor Union Feeling Thermometer Gender Labor Union Membership Education level Where did Respondent Grow up? (region) Family income Party identification Employment (manager, non-manager, or other) GDPPC (Percent Change in GDP) Membership density (union membership by state)

favorably. The administrator interviewed said: “There is in my mind …a general complacency with human beings in our country in the last 20 years. And … they just assume things are there. They assume they get these things. When you start really living there thinking that way you don’t see the value of a union.” While the union member explained “…if you don’t know anybody that’s union and the only the only thing you think about unions is the guy you see holding the road sign on the road then you probably I think that’s what a lot of people do. They see somebody holding a sign they think that’s union people making a bunch of money for doing nothin’. ” Both of these statements explain a rational choice view of the opinions of others. The union member also talked about the information part of rational choice and how information can alter opinions. He said, “I think the people…who know somebody that’s in a union or have family that’s in a union they see them in a positive way. But most people …who don’t have any association with it see ‘em in a negative way.” This is interesting because both opinions, while not a broad enough sample to make broad conclusions, point to average people recognizing rational self-interest in others and assuming that is how others operate.

Variable Type Dependent Control Independent Independent Independent Independent Independent Independent Independent Independent

Theory Rational Choice Group Influence Rational Choice Group Influence Rational Choice Rational Choice Group Influence Rational Choice Economic Conditions Self Interest

Empirical Analysis: Predictors of Union Attitudes In general, the quantitative analysis offers some evidence of a link between many of the independent variables and union attitudes. In the interviews the respondents both professed a belief that others acted in what the respondents felt were others’ best interests. Respondents made a connection between what people saw (or did not see) union members doing on the job and how those same people would perceive the group as a whole (Campbell et al. 1976, 161). This is particularly interesting because respondents came to these conclusions as laymen with no theoretical background given. A multivariate regression analysis was estimated, with the main independent variables (i.e., income, membership density percentage, the region where respondent grew up, education, percent change in GDP, respondent employment, party identification, union membership, and gender) regressed on labor union opinion. The regression analysis performed garnered the following results as shown in Table 2.3 See Appendix B for regression models separately testing the region and density variables. Several of the variables tested had negative relationships with the dependent variable. The income

Pi Sigma Alpha Undergraduate Journal of Politics

Table 2. Predictors of Union Opinion Union Thermometer

Coefficient

Standard Error

P>|t|

95% Confidence Level

2. 17 to 33 percentile 3. 34 to 67 percentile 4. 68 to 95 percentile 5. 96 to 100 percentile Membership Density Percentage Region Non-continental U.S. Midwest Northeast

*-2.48 *-6.27 *-8.03 *-10.53 *-0.50

0.57 0.52 0.57 0.91 0.02

0.000 0.000 0.000 0.000 0.037

-3.60 -7.30 -9.15 -12.31 -0.10

-1.36 -5.25 -6.90 -8.76 -0.00

2.80 -.060 *1.75 *2.01 *-1.49 *0.49 *0.49 -0.80 *-2.53

4.36 0.55 0.60 0.58 0.12 0.08 0.44 0.44 0.46

0.521 0.914 0.004 0.001 0.000 0.000 0.000 0.070 0.000

5.75 -1.15 0.56 0.87 -1.72 0.33 0.33 -1.67 -3.43

11.34 -1.03 2.931 3.16 -1.27 0.65 0.65 -0.06 -1.64

1. Strong Democrat *7.47 0.61 0.000 2. Weak Democrat *2.85 0.59 0.000 3. Independent - Democrat *2.05 0.67 0.002 5. Independent - Republican *-3.54 0.69 0.000 6. Weak Republican *-3.72 0.65 0.000 7. Strong Republican *-7.23 0.69 0.000 Union Membership No, not a union member *11.09 0.62 0.000 Yes, a union member *14.39 0.55 0.000 Yes, a member plus family members *15.94 1.28 0.000 Gender Female *2.14 0.35 0.000 Constant *64.74 1.03 0.000 *= statistically Significant at .05 Number of Observations= 23,016 2 R =.01359 Adjusted R2 = .0135

6.27 1.69 0.73 -4.89 -4.99 -8.57

8.66 4.00 3.36 -2.20 -2.45 -5.88

9.88 13.30 13.42

12.30 15.47 18.45

1.44 62.72

2.83 66.76

Income

South Education Percent Change in GDP Respondent Employment Other Manager Party ID

variable was one of the most noteworthy of the group. The negative relationship on the income variable was highest among the 96-100th percentile. Respondents of the highest income percentile the respondents were 13.8 points less favorable to unions than the lowest income percentile holding all other factors constant. The coefficient was negative for all percentile

ranges compared to the baseline lowest percentile but only reached double digits on the highest percentile range, which coincides with the â&#x20AC;&#x153;upperclassâ&#x20AC;? group mentioned previously. The empirical results provide evidence for Hypothesis 1. Indeed, there is a relationship between income and union attitudes, indicating that citizens with higher income

Volluz

are significantly linked to lower levels union approval as graphed in Figure 1 below (Campbell et al. 1976; Downs 1956; Key and Cummings 1966). Therefore, there is a significant link between respondents’ from lower income groups and union approval. The union density (membership percent by state) variable was statistically significant with each one unit increase in density having a 0.5 unit decrease in union opinion, holding all other variables constant. The latter is in contrast to the proposed Hypothesis 2, which asserted a positive relationship between union attitudes and union density (Campbell et al.

non-continental U.S. category had a positive but not significant relationship with the dependent variable. This may be due to the individualistic attitudes of Alaskans, but in Alaska there is a highly unionized workforce because of the types of jobs there. In Alaska, many people work in the oil extraction business which is highly unionized, also some commercial fishing is unionized. It could also be because the region variable grouped two very different cultures, which resulted in a non-significant relationship. The education variable also had a significant and negative coefficient, indicating that for every unit

1976; Downs 1956; Key and Cummings 1966). The latter relationship may exist because citizens may view unions as elitist. Next, region had a very interesting relationship with union attitudes. The only significant categories for this variable were the “South” and “North East” categories, each of which were positive relationships compared to the baseline “West” category. Respondents from the north east and south were, on average, 2 points and 1.8 points more favorable to unions, respectively to their western counterparts. The

increase in education level there is a 1.5 point decrease in union approval holding all other factors constant. Consistent with Hypotheses 3, this provides evidence that higher education levels are linked to lower levels of labor unions (Hundley 1988). The economic variable tested, the percent change in Gross Domestic Product, was positively and significantly related to union opinion. For every positive one unit increase in percent change GDP, respondents’ opinions of labor unions increased

Figure 1. Income Level and Labor Union Opinion

Public Opinion of Labor Unions

Figure 2. Percent Change in Gross Domestic Product and Labor Union Opinion

by .49 percent holding all other factors constant. The latter confirms Hypothesis 4, suggesting when economic times are good or getting better, opinions toward unions become more positive (Clawson and Clawson. 1999; Dark 2011). This finding goes against the conventional wisdom that the benefits unions can give workers during hard times would cause an increase in opinion rather than the decrease possibly caused by divisive political rhetoric and the insulating affects unions have against negative economic forces as discussed in the literature review. The results are illustrated in Figure 2. As Table 2 shows, there was a positive relationship between respondent employment and union attitudes. Respondents who were not management level in their jobs had a higher view of labor unions than did managers. A respondent who was a manager was 2.5 points less favorable to unions compared to non-managers holding all other factors constant. This result was expected and points to one’s occupation affecting one’s opinions of labor unions. This is consistent with Hypothesis 5 when viewed in light of the literature on group membership, familial

influence and rational choice. It follows rationally that some management may find labor unions irksome due to the continual adversarial relationship between unions and management. The results of the respondent occupation regression are presented in Figure 3. Table 2 also reports that a respondent’s party identification had a negative coefficient that showed a negative coefficient peak at the “Strong Republican” category (1 = strong Democrat to 7 = strong Republican). A respondent who was strong Republican was 7.2 points less favorable to unions compared to respondents who were independent, holding all other variables constant. This finding substantiates Hypothesis 6. One of the more striking coefficients was regarding the union membership. A respondent who was a member with other members in the house were 15.9 points more favorable to unions compared to respondents with no household union members holding all other factors constant. This finding aligns with the expected relationship represented by Hypothesis 7. A respondent who was a union member with no other members in the home was 14.4 points

84 Figure 3. Respondent Occupation and Labor Union Opinion

more favorable to unions compared to respondents with no household members, holding all other factors constant. Non-member respondents with family members in a union were 11.1 points more favorable to unions holding all other factors constant. This is also as expected with direct relationships with the union group causing vested interest in the group. Figure 4 illustrates the relationship between union membership and labor union opinion. Gender shows a positive relationship on the female category with a significant coefficient of roughly two. This suggests that more women than men think highly of labor unions. This correlation contrasts with prior research because men are expected to have a more positive regard for unions because of the high percentage of unionization for traditionally male dominated, manual labor jobs. This could mean that women approve of unions because of their increased tolerance for collectivism. Conclusion The quantitative results point to a trend in opinion formation that follows broad self-interest and cultural influence. The empirical results provide evidence for nearly all hypotheses proposed, except

Hypothesis 2 which asserted that higher union membership density by state was positively linked to union attitudes. Exposure to unions (Hypothesis 7) and income (Hypothesis 1) are the most important factors in opinion formation, at least when it comes to labor unions. However, more personal relationships with union members are positively linked to union opinion. It makes sense that less people approve of unions in 2016 than when labor unions were more prevalent in the U.S. The data showed that the more people in a respondentâ&#x20AC;&#x2122;s life who were involved in unions, the higher approval they had toward unions. The implications of this are important because if workers in the U.S. who would otherwise benefit from unionization, but who do not know about unions or who do not have sufficiently unbiased information on unions, could overlook their option to organize in their best interest. Unions could be good for some people who do not even know it. Furthermore, union growth in the U.S. will be difficult if the major indicator of opinion positivity is proximity to the very unions suffering from membership loss. Public opinion towards unions also reflects economic conditions as shown by percent change in GDP (Hypothesis 4). When economic times

Pi Sigma Alpha Undergraduate Journal of Politics

Figure 4. Union Membership and Labor Union Opinion

Note: Marginals represent the following: “No one” = no one is a union member in this household; “Not R” = Respondent is not a union member; “R” = Respondent is a union member; and “R+” = Respondent is a union member plus union members in household. are bad, this results indicate that citizens are likely to disapprove of unions. This is important to note because unions may benefit some workers during hard economic times while others feel left out. It is a help for the already unionized, but creates discord among outsiders. A person’s level of employment is also an important factor in determining union opinion (Hypothesis 5). If a person is a non-manager, this research suggests that non-managers are more likely to approve of labor unions than if they were a manager. This is very important because low-level employees are most likely to benefit from unionization. Indeed, unions could also have negative effects on some people. They could raise wages, resulting in greater expense for business as well as contribute to government corruption. Labor unions, whether they positively or negatively affect citizen’s opinion, are a political and social force to be reckoned with that divides the American public. The reasons people form their opinions on labor unions translates to a broader understanding of how people understand organizations as beneficial or not and why.

Last, there were several limitations to this research that require note. The qualitative analysis boasts only two interviews with union members and non-members. While the interviews provide additional context to the empirical results, more interviews are necessary to better understand the personal experiences of non-managers and managers. In addition, the empirical analysis relies on crosssectional survey data; therefore, causal links cannot be made between the predictors and union approval. Future research should seek to conduct a laboratory experiment on the influence of elite discourse, particularly media framing of labor union membership and workers, in a controlled environment. An experiment would allow researchers to control for enough extraneous factors to make a causal link between media framing and union opinion.

Notes The Labor Tribune is a newsletter sent through the mail to many active union members. It includes labor news, political op-eds, and lists of member deaths, and 1

Volluz

year of service awards. There is also a list of favorable candidates up for election at the state and federal level. Including age and religion variables in the regression model did not substantially change the results of the analysis in terms of the size and significance of the any of the effects. 2

I also ran two separate regressions due to the loss of data points when both occupation variables were estimated together. In addition, I used an ordinal Education variable but decided after running regression analysis with the variable categorically and 3

interval and finding little coefficient difference between categories that it was appropriate to run the variable as an interval because of issues with stretching the data too thin. I also ran the occupation variables (Father Occupation and Respondent Occupation) separately because when the two variables were run together the number of observations dropped to around 8,000. This was a size I felt too small considering the total number of ANES observations available was 55,000. Generally, the results were the same for all other variables in each regression except the Membership variable, which became insignificant when run with the Father Occupation variable. The results of the Father Occupation regression are available upon request from the author.

Public Opinion of Labor Unions

Appendix A Interview Questions

The researcher introduces herself and hands the participant the recruitment document. The researcher gives the participant time to read the document. 1. Do you have any questions? 2. In light of the information given to you, do you consent to an interview with me? 3. Personal view of unions a. What do you know about labor unions? i. (For clarification) when did you first learn about them? How? b. Have you had any interactions with labor unions? i. What were they? ii. What were they like? c. Based on these interactions, what are your impressions of labor unions? d. Are unions needed at all? i. For whom? ii. Why? e. Do you think unions are more important for society? i. Does the importance of unions for society depending on how the economy is doing? ii. (For clarification if necessary) when economic times are good, bad, in the middle, or at all? iii. How so? f. Do you think unions affect the economy? i. If so, how? g. Do you think unions affect business? i. If so, how do you think unions affect business? h. Do you think anyone benefits from unions? i. Who? ii. How? iii. Why? i. Do you believe unions have benefited you? i. How so? j. How do you think unions are faring in the US? i. (For clarification) 1. How are they faring economically? 2. How does the public see them? 3. How are membership rates today relative to the past?

k. How do you think union strength (ability of the union to pursue its interests) has changed over the years? l. Are union contracts necessary for all employees or just some types of workers? Workers being those not in a management position who work for hourly or salary pay. i. Which ones? 4. Personal view or members a. Do you know any union members personally? i. How do you know them? b. What do you think of union members? i. Generally? ii. Specifically? c. Is there a difference in how union members perform versus how non-members perform? 5. How others perceive unions a. Are there any stereotypes attached to unions or union members? i. What are they? ii. In your opinion are there any justified stereotypes associated with union membership? 6. How others perceive members a. How do you think others perceive union members? b. Is union membership stigmatized in the United States? i. How so? 7. Personal Questions a. Where did you grow up? i. Northeast ii. Northwest iii. Midwest iv. Southeast v. Southwest vi. Outside the United States vii. Prefer not to Answer b. What is your occupation? c. What is your level of education? d. How old are you? 8. Do you have anything else you would like to add? 9.Would you be willing to pass a recruitment flier to anyone you think would be good for me to talk to?

88 Appendix B Note: In order to further test for multi-collinearity between Region and Membership Density Percentage, two additional multivariate regressions were estimated. Table 1B reports the regressions results without Membership Density Percentage. Table 2B reports the regressions results without Region. Table 1B. Predictors of Union Opinion excluding Membership Density Percentage Union Thermometer

Coefficient Standard Error

P>|t|

95% Confidence Level

Income 2. 17 to 33 percentile 3. 34 to 67 percentile 4. 68 to 95 percentile 5. 96 to 100 percentile

*-2.54 *-6.52 *-8.36 *-10.62

0.57 0.51 0.56 0.89

0 0 0 0

-3.65 -7.52 -9.46 -12.36

-1.43 -5.51 -7.25 -8.87

Other Non-continental U.S. Midwest Northeast South

0.984 *6.25 -0.1 *1.57 *2.54 *-1.47 *0.46

1.37 2.35 0.544 0.5 0.544 0.11 0.08

0.474 0.008 0.852 0.007 0 0 0

-1.71 1.64 -1.17 0.43 1.47 -1.69 0.304

-3.68 10.85 -0.96 2.71 3.6 -1.24 0.62

-0.812 *-2.63

0.43 0.45

0.061 0

-1.66 -3.51

-0.04 -1.75

1. Strong Democrat *7.53 0.6 2. Weak Democrat *2.92 0.58 3. Independent - Democrat *2.16 0.66 5. Independent - Republican *-3.37 0.67 6. Weak Republican *-3.7 0.64 7. Strong Republican *-7.36 0.67 Union Membership No, not a union member *10.82 0.61 Yes, a union member *14.14 0.54 Yes, a member plus family members *15.91 1.26 Gender Female *2.27 0.35 Constant *63.7 0.89 obs.= 23,739 *= significant at .05 2 Adj R = .136 R2 = .1367

0 0 0.001 0 0 0

6.35 1.78 0.87 -4.69 -4.96 -8.69

8.71 4.06 3.45 -2.05 -2.46 -6.03

0 0 0

9.63 13.08 13.43

12.01 15.212 18.39

0 0

1.59 61.96

2.96 65.45

Region

Education Percent Change in GDP Respondent Employment Other Manager Party ID

Pi Sigma Alpha Undergraduate Journal of Politics

Table 2B. Predictors of Union Opinion excluding Region Union Thermometer

Coefficient Standard Error

P>|t|

95% Confidence Level

Income 2. 17 to 33 percentile *-2.53 0.57 3. 34 to 67 percentile *-6.03 0.52 4. 68 to 95 percentile *-8.12 0.57 5. 96 to 100 percentile *-10.44 0.9 Membership Density Percentage *-0.08 0.018 Education *-1.55 0.11 Percent Change in GDP *0.49 0.08 Respondent Employment Other -0.69 0.44 Manager *-2.42 0.45 Party ID 1. Strong Democrat *7.51 0.6 2. Weak Democrat *2.85 0.59 3. Independent - Democrat *2.03 0.67 5. Independent - Republican *-3.69 0.68 6. Weak Republican *-3.89 0.64 7. Strong Republican *-7.45 0.68 Union Membership No, not a union member *11.08 0.62 Yes, a union member *14.36 0.55 Yes, a member plus family members *15.85 1.28 Gender Female *2.13 0.35 Constant *66.73 0.83 obs.= 23,225 *= significant at .05 2 Adj R = .1339 R2 = .1346

0 0 0 0 0 0 0

-3.65 -7.33 -9.25 -12.21 -0.12 -1.77 0.33

-1.42 -5.53 -7 -8.67 -0.05 -1.32 0.65

0.116 0

-1.55 -3.31

0.17 -1.53

0 0 0.002 0 0 0

6.32 1.7 0.72 -5.03 -5.15 -8.79

8.7 4 3.33 -2.36 -2.63 -6.11

0 0 0

9.87 13.27 13.35

12.28 15.44 18.36

0 0

1.43 65.11

2.82 68.35

Volluz References

American National Election Studies. 2010. “Time Series Cumulative Data File.” Stanford University and the University of Michigan. www.electionstudies.org) (Accessed August 26, 2015). Bardes, Barbara A., and Robert W. Oldendick. 2012. Public Opinion: Measuring the American Mind. Lanham, Md: Rowman & Littlefield. Bartels, Larry M. 2008. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton: Princeton University Press. Blais, Andre. 2000. To Vote or Not to Vote. Pittsburgh: University of Pittsburgh Press.

Downs, Anthony. 1957. An Economic Theory of Democracy. Addison-Wesley Publishing Company. Druckman, James N., and Kjersten R. Nelson. 2003. “Framing and Deliberation: How Citizens’ Conversations Limit Elite Influence.” American Journal of Political Science. 4 (47): 729-745. Dugan, Andrew. 2013. “In U.S., Majority Approves of Unions, but Say They’ll Weaken,” Gallup Poll Briefing, August 30. http://www.gallup. com/poll/164186/majority-approvesunions-say-weaken.aspx?g_source=&g_ medium=search&g_campaign=tiles (Accessed June 16, 2016).

Campbell, Angus, Philip Converse, Warren Miller, and Donald Stokes. 1976. The American Voter. Chicago: University of Chicago Press.

Fatas-Villafranca, Francisco, Dulce Saura, and Francisco J. Vazquez. 2011. “A Dynamic Model of Public Opinion Formation.” Journal of Public Economic Theory 13 (3): 417–441.

Charles, Kerwin Kofi, and Erik Hurst. 2003. “The Correlation of Wealth across Generations.” Journal of Political Economy 111(6): 1155-1182.

Frank, Thomas. 2005. What’s the Matter with Kansas?: How Conservatives Won the Heart of America. New York: Henry Holt.

Clawson, Dan, and Mary Ann Clawson. 1999. “What Has Happened to the U.S. Labor Movement? Union Decline and Renewal.” Annual Review of Sociology 25 (August): 95-115.

Georgia State University. “Union Density Estimates by State, 1964-2015.” http://unionstats.gsu.edu/ MonthlyLaborReviewArticle.htm (Accessed June 1, 2016).

Dark, Taylor E. 2011. “The Economic Crisis and Organized Labor: Resentment over Solidarity.” New Political Science 33 (4): 525-539.

Heshizer, Brian. 1985. “Unions and Public Opinion: Why the Declining Relationship?” Labor Studies Journal 9(3): 254-270.

DeNavas-Walt, Carmen, and Bernadette Proctor. 2014. “Income and Poverty in the United States: 2013.” Current Population Reports: United States Census Bureau. https://www. census.gov/content/dam/Census/library/ publications/2014/demo/p60-249.pdf. (Accessed June 16, 2016).

Hirsch, Barry T., David A. Macpherson, and Wayne G. Vroman. 2001. “Estimates of Union Density by State.” Monthly Labor Review 124(7):51-55. Hogler, Raymond, and Christine Henle. 2011. “The Attack on Public Sector Unions in the United States: How Regional Culture Influences Legal Policy.” Labor Law Journal 62 (3): 136-144.

Public Opinion of Labor Unions Hundley, Greg. 1988. “Education and Union Membership.” British Journal of Industrial Relations 26 (2):195-201. Key, V. O., and Milton C. Cummings. 1966. The Responsible Electorate: Rationality in Presidential Voting, 1936-1960. Cambridge: Harvard University Press. McCartin, Joseph A. 2013. “Public Sector Unionism Under Assault: How to Combat the Scapegoating of Organized Labor.” New Labor Forum 22(3): 54-62. Moussaid, Mehdi, Juliane E. Kaemmer, Pantelis P. Analytis, and Hansjoerg Neth. 2013. “Social Influence and the Collective Dynamics of Opinion Formation.” PLoS ONE. 8(11): 1-24. https://arxiv.org/ftp/arxiv/ papers/1311/1311.3475.pdf (Accessed June 1, 2016). Norrander, Barbara, and Clyde Wilcox. 2008. “The Gender Gap in Ideology.” Political Behavior 30 (4): 503-523. Pitlik, Hans, Gerhard, Schwarz, Barbara Bechter, and Bernd Brandl. 2011. “Near Is My Shirt but Nearer Is My Skin: Ideology or Self-Interest as Determinants of Public Opinion on Fiscal Policy Issues.” Kyklos 64(2): 271-290. Saporta, Ishak, and Bryan Lincoln. 1995. “Managers’ and Workers’ Attitudes toward Unions in the U.S. and Canada.” Relations Industrielles / Industrial Relations 3(50): 550-567. Sigel, Roberta. 1965. “Assumptions about the Learning of Political Values.” Annals of the American Academy of Political and Social Science 361 (September):1-9.

U.S. Bureau of Labor Statistics. 2015. “Union Members- 2014.” USDL-15-0072. United States Department of Labor. http://www.bls.gov/news. release/pdf/union2.pdf. (Accessed June 16, 2016). U.S. Bureau of Economic Analysis. 2015. “National Income and Product Accounts Gross Domestic Product: First Quarter 2015 (Advance Estimate).” https://www.bea.gov/newsreleases/ national/gdp/gdpnewsrelease.htm. (Accessed May 25, 2016). Vandello, Joseph A., and Dov Cohen. 1999. “Patterns of Individualism and Collectivism across the United States.” Journal of Personality and Social Psychology 77(2): 279-292. Western, Bruce, and Jake Rosenfeld. 2012. “Workers of the World Divide: The Decline of Labor and the Future of the Middle Class,” Foreign Affairs, May/June. https://www.foreignaffairs.com/ articles/united-states/2012-04-20/workersworld-divide (Accessed June 1, 2016). Wu, Yue, Yong Hu, and Xiao-Hair He. 2013. “Public Opinion Formation Model Based on Opinion Entropy.” International Journal of Modern Physics C 24(11):1-13.

Fact Checking in the 2016 Presidential Campaign: Exploring the Effectiveness of Fact Checking Statements by the Candidates Jennifer Kahn and Alexis Wagner Arizona State University This study examines the impact of fact checking of presidential candidates’ statements in the 2016 nomination campaign. More specifically, we explore how the impact of fact checking varies with the source of the fact check and the party of the candidate. In conducting our study, we rely on an Internet survey experiment with 393 subjects. We utilize a classic pretest posttest experimental design with a three (i.e., liberal fact check source, conservative fact check source, neutral fact check source) by two (i.e., Republican candidate, Democratic candidate) by two (i.e., better known candidate, less known candidate) factorial design, producing 12 experimental conditions. Our results show that the source of a fact check does not consistently influence people’s impressions. Additionally, we find that the impact of fact checking on candidate evaluations was more powerful for the Republican candidates. Introduction Fact checking is a promising form of accountability journalism that serves to hold political candidates responsible for making misleading or non-factual claims. Fact checking began in the late 1980s when David Broder, a reporter for the Washington Post, claimed American journalists needed to investigate the claims made by candidates during campaigns. Broder believed journalists should question candidates on substance issues, hold political consultants accountable for the advertisements they create, and alert the public of inaccurate attacks (Adair and Holan 2011). The first fact checking began in 1990 when newspapers started to check the accuracy of political advertisements, called “ad watches,” as a regular part of campaign coverage. Presently, fact checking is more pervasive, with multiple news sources providing their own fact checks of statements made by electoral and non-electoral political actors. These fact checks

“aim to reduce the level of deception and confusion in U.S. politics” (Factcheck.org 2015). Despite fact checks being so prevalent, we are only beginning to understand how these fact checks influence the electorate. Background and Literature Scholarly attention examining the impact of fact checking is a relatively new area of research. Given the growing prevalence of fact checking in the political arena, it is important to systematically examine the rise of fact checking over time. Wintersieck and Fridkin (2016) conducted an extensive content analysis of fact checking organizations and newspaper fact checks from 2003 to 2012. The researchers, employing a LexisNexis search, identified and collected every newspaper and wire story mentioning fact checks in the headline, as well as searched the archives of fact checking organizations, producing a population of 7,008 fact-checking stories.

Pi Sigma Alpha Undergraduate Journal of Politics

Relying on systematic sampling with a random start, the researchers selected 1,267 fact check articles for analysis. Wintersieck and Fridkin (2016) coded for several dimensions including: 1) the subject of the fact check (e.g., the president, a senate candidate); 2) the type of facts being checked; 3) the source of the fact check; and 4) the accuracy of the fact being checked. The authors find that two-thirds of the fact checks examined were from two organizations: PolitiFact and Factcheck.org. Other notable findings include: fact checking is more common in presidential election years than in midterm election years, the

real issues and real articles, when possible, to increase external validity). These news articles included misleading statements by politicians reinforcing current and popular misconceptions. The statements were randomly assigned, with some participants receiving statements with corrective information and some receiving statements without the corrective information. After reading the statements, the participants answered a series of follow-up questions. The researchers conclude that correcting the statements of politicians did not reduce popular misperceptions and even strengthened misperceptions

prevalence of fact checking has increased over time, Republican claims are subject to scrutiny more often than Democratic claims, and statements of challengers are more likely to be fact checked and classified as false compared to those of incumbents. Wintersieck and Fridkin (2016) conclude that advertisements make up a majority of fact checking in non-presidential election years, while fact checking in presidential election years focuses more on the claims made in speeches and press releases. Their study demonstrates the growing importance of fact checking, in general, as well as the use of fact checking of candidate speeches during presidential campaigns. One of the first studies to examine the impact of fact checking was conducted by Nyhan and Reifler in 2010. The authors sought to uncover whether or not “corrective information embedded in realistic news reports succeeds in reducing prominent misperceptions” (Nyhan and Reifler 2010, 304). This research was conducted using a classical experimental design in an online format with undergraduate students as subjects. Subjects received a pretest and then participated in a distracting activity before being exposed to the experimental stimulus. The researchers conducted four iterations with the same basic design where the participants read newspaper articles (using

for some ideological groups. Following Nyhan and Reifler’s (2010) study, Garrett, Nisbet, and Lynch (2013) relied on a similar design to examine how corrective information, presented via a fact check, influence people’s beliefs. The authors employed a pretest-posttest control group design with a diverse sample of U.S. citizens, using an opt-in online panel. Participants were randomly assigned to one of several stimuli which presented the results of a fact check, but varied: 1) contradictory claims made by politicians in the corrective messages, 2) controversial claims made by politicians in the corrective messages, and 3) pictures accompanying the corrective messages. The results of their experiment show that the “effectiveness of a correction to an inaccurate political belief depends not just on the strength of the relevant evidence, but also on other, less relevant information included in the message” (Garrett, Nisbet, Lynch 2013, 630). Both Nyhan and Reifler (2010) and Garrett, Nisbet, and Lynch (2013) examine the impact of corrective statements presented in a text format online. However, Garrett, Nisbet, and Lynch (2013) find that corrective statements can be influential and the impact of these corrective statements depends on both the relevant and irrelevant content surrounding the messages being fact checked.

Kahn and Wagner

While the initial experimental work examined fact checking of politicians’ statements, Fridkin, Kenney, and Wintersieck (2015) focused their research on fact checking claims made in negative advertisements. The authors examined the impact of fact checking negative political advertisements with an online survey experiment with a representative sample of the U.S. population. They examined whether the practice of fact checking softens the impact of attack advertisements directed at voters. The authors find that fact checks are likely to sway voters’ acceptance or rejection of the claims made in a negative

accuracy of specific claims made by the legislators were examined. Nyhan and Reifler (2015a) find that the experimental treatment had a substantial effect in reducing the prevalence of negative ratings. In other words, legislators in the experimental conditions were less likely to make statements that were subsequently fact checked. This study demonstrates that elites are concerned about being criticized for making misleading statements. Furthermore, when legislators’ awareness of being fact checked is heightened, their likelihood of making questionable statements declines. In a final study on the impact of fact checking,

advertisement. In a follow-up study, Wintersieck, Fridkin, Kenney, Thompson, and Courey (2015) also examined negative advertisements, but focused on whether the source of the fact check (i.e., Fox News, PolitiFact, or MSNBC) influenced the message’s impact. The experimental results suggest that the source of the fact check does not influence receptivity to the fact check’s message. The researchers conclude that subjects welcomed the clarifying information from a fact check message, regardless of the source of the fact check. In a creative extension of the fact checking research, Nyhan and Reifler (2015a) explored whether the presence of fact checking alters the behavior of politicians. The researchers used a post-test only experimental design and randomly assigned 1,169 state legislators to one of three conditions: 1) letters sent to legislators reminding them of the risks to their reputation if they are caught making questionable statements; 2) letters sent to legislators stating campaign accuracy was being monitored; and 3) the control condition, where no letters were sent to legislators. The researchers were interested in examining whether these three conditions produced differences in whether a legislator received a negative rating by the PolitiFact affiliate and whether the

Nyhan and Reifler (2015b) examined how the public views fact checking and whether fact checking affects people’s general views of politics and government. In particular, the researchers sought to answer three research questions: 1) Will exposure to fact checking polarize people’s views of the practice? 2) Does fact checking affect levels of trust in politicians and will these effects be strongest among people who are already highly distrustful? And 3) how does fact checking affect political efficacy and do these effects vary with prior political knowledge? The researchers embedded a classical experimental design in a longitudinal public opinion survey conducted in five waves over the fall 2014 election campaign. In the initial wave, respondents were asked a number of questions about their perceptions and attitudes towards fact checking. The next waves consisted of three mini-waves where half of the respondents were randomly assigned to receive fact checking content and the other half received a placebo of nonpolitical information from press releases. The final wave consisted of a post-election survey. Nyhan and Reifler (2015b) conclude that approximately half of the subjects were unfamiliar with fact checking, but the researchers find an increasing favorability toward accountability and the practice of fact checking over

Fact Checking in the 2016 Presidential Campaign time. However, the researchers fail to find that fact checking: 1) polarizes people’s views of fact checking; 2) affects levels of trust in politicians; or 3) affects people’s level of political efficacy. Overall, researchers have begun to look at how fact checking of political rhetoric affects people’s views of the fact-checked statements. The results are inconsistent; with some studies showing that fact checking affects people’s assessments of the accuracy of the fact checked statements (e.g., Fridkin, Kenney, and Wintersieck 2015), while others (i.e., Nyhan and Reifler 2010) fail to find consistent effects. However,

the pervasiveness of fact checking is clearly increasing with every election cycle (Wintersieck and Fridkin 2016). Fact checking appears to limit the number of inaccurate statements made by politicians (Nyhan and Reifler 2015a); therefore, studying the impact of these fact checks is an increasingly important topic of study.

will make people less willing to accept fact check

As an example, a Democratic voter exposed to

a Fox News fact check questioning a statement made by Hillary Clinton may be less likely to accept the information presented in this fact check, compared to a MSNBC fact check of the same statement. In contrast, a Democratic voter exposed to a Fox News fact check questioning the statements of a Democratic challenger in a U.S. Senate race may be more accepting of the fact check since voters are likely to have less preexisting information, and less established views, about the senate candidate. In summary, we expect the level of prior information about a candidate information from a less friendly source (i.e., Fox News for Democrats, MSNBC for Republicans).

In our research, we explore several questions

regarding the impact of fact checking in presidential contests. To begin, we look at how the source of the

Expectations

fact check influences people’s ability to learn and

expect that citizens are more likely to learn from fact

In this research, we seek to contribute to the

retain the information presented in the fact check. We

fact checking literature by exploring whether the

checks emanating from trusted sources. For example,

source of the fact check is important in high profile

in the current 2016 presidential campaign, we expect

races. In particular, in a presidential contest, we

a person who identifies as a Democrat and strongly

expect source effects will be more powerful than

supports Hillary Clinton to be more persuaded by

in congressional races because citizens have more

a fact check from a trusted source, such as MSNBC,

information and more established views of presidential

compared to a less trusted source, such as Fox News.

contenders. In sub-presidential contests, fact checking,

regardless of source, has been found to influence

by news organizations influences people’s views of

people’s impressions (e.g., Fridkin, Kenney, and

the reliability of these news sources. We expect that

Wintersieck 2015). The source may be less important

liberals and conservatives will respond differently to

in congressional races since people have scant

fact checking by news organizations, depending on the

information about the politicians and all information,

source. More specifically, liberals viewing fact checks

regardless of source, may be viewed as important.

by MSNBC will view MSNBC as more reliable when

However, in presidential races where the electorate is

MSNBC is fact checking a Democratic candidate rather

more knowledgeable about the candidates, the source

than a Republican candidate. Similarly, conservatives

of the fact check may be more consequential.

viewing a fact check by Fox News will view Fox News

Second, we examine whether fact checking

96 as a more reliable news source when Fox News is

fact check source, conservative fact check source,

fact checking a Republican candidate rather than a

neutral fact check source) by two (i.e., Republican

Democratic candidate. In these examples, the news

candidate, Democratic candidate) by two (i.e., better

organization’s reliability is likely to be enhanced

known candidate, less known candidate) factorial

because the news organization is critiquing one of their

design, producing 12 experimental conditions.

“own” candidates, thereby increasing their credibility.

Subject Recruitment

Third, we hypothesize that fact check

messages will produce more negative evaluations of candidates when the source of the fact check and the targeted candidate share the same ideological profile. For instance, we expect a Fox News fact check of a Republican candidate will produce more negative impressions of the Republican candidate. In this example, Fox News is not expected to critically evaluate a Republican candidate, therefore, such criticisms are likely to be viewed as more credible and, thus, more persuasive.

Finally, and similar to earlier research (e.g.,

Fridkin, Kenney, and Wintersieck 2015), we expect that candidates who are fact checked will be viewed more negatively than candidates who have not been subjected to fact checking. Since fact checks often challenge the veracity of candidates’ statements, we hypothesize the fact checking messages will lead people to develop more negative evaluations of candidates, regardless of the source of the fact check. We turn next to a discussion of our experimental

In the fall of 2015, 393 undergraduate students were recruited from political science courses at Arizona State University. These students received course credit for their participation in an experimental lab. The subjects were randomly assigned to one of the 12 conditions. We checked to ensure that randomization across the 12 experimental conditions was successful and found the experimental conditions did not vary on key demographic dimensions (i.e., gender, party, ideology, and age), demonstrating the effectiveness of randomization. Since respondents were randomly assigned successfully to the various conditions, we can assume that at the start of the experiment, the 12 groups were similar. Consequently, if we find differences at the end of the experiment, these differences must be driven by the experimental stimulus. Table 1 shows the demographic characteristics of the subjects in the experiment. Although our sample is taken from political science courses, these students represent a wide range

method.

of majors (i.e., 54% of the subjects were political

Experimental Method

of political knowledge.1 Similarly, subjects are not

In our experiment, we examine whether fact checks of speeches made by presidential candidates in the 2016 presidential election affect people’s impressions of the candidates. We selected the 2016 presidential race because it is a high profile election with no incumbent running for re-election. We rely on an online survey experiment with a three (i.e., liberal

science majors); thus, they differ in their amount drawn exclusively from introductory courses, so the respondents differ in age. The ages of the respondents ranged from 17 to 53, with a mean of 21.36 and a standard deviation of 4.63. We recognize there are some limitations to our sample. In particular, a student sample may not be generalizable to a more diverse adult population.

Pi Sigma Alpha Undergraduate Journal of Politics

Table 1. Demographic Characteristics of Sample Description Age Male Party Identification Strong, Weak, or Leaning Democrat Independent Strong, Weak, or Leaning Republican Ideology Extremely Liberal, Liberal, or Somewhat Liberal Moderate Extremely Conservative, Conservative, or Somewhat Conservative Year in School Freshman Sophomore Junior Senior Political Interest Very much interested Somewhat interest Not much interested For example, 27% of our sample identifies with the Democratic Party, 18% identify with the Republican Party, and 52% identify themselves as Independents. In comparison, a recent Gallup poll of a representative sample of the country shows that 30% of all adults consider themselves Democrats, 28% consider themselves Republicans, and 39% consider themselves Independents (Gallup.com 2016). Therefore, our sample of undergraduate students over represents citizens identifying as Independents. Experimental Stimuli Given the 3 x 2 x 2 design, we develop the experimental stimuli by beginning with a fact check for the Republican candidates and a fact check for the Democratic candidates. In creating the two fact

Valid Percent or Mean 21.36 (mean) 60.1% 53.9% 14.5 31.6 48% 24.9 27 28% 20 27 25 41% 48 11

checks, we relied on actual fact checks of presidential candidates’ statements made during the early part of the 2016 nomination campaign. All of the fact checked statements were taken from FactCheck.org. The fact checked statements were changed so that the names of the candidates as well as the source of the fact check varied. The Democratic and Republican fact checks each include four statements: one labeled “false,” two labeled as “mostly false,” one labeled as “true” and are approximately the same length. In the Republican fact check, issues of immigration and business are discussed among the candidates’ claims, while the Democratic fact check examines the candidates’ claims regarding issues of immigration and education. See Appendix A for examples of the actual fact checks used in the experiment.

Kahn and Wagner

Table 2. Experimental Conditions Condition 1 2 3 4 5 6 7 8 9 10 11 12

Candidate Hillary Clinton Hillary Clinton Hillary Clinton Bernie Sanders Bernie Sanders Bernie Sanders Carly Fiorina Carly Fiorina Carly Fiorina Jeb Bush Jeb Bush Jeb Bush

In choosing the candidates for the fact checks,

Source FactCheck.org Fox News MSNBC FactCheck.org Fox News MSNBC FactCheck.org Fox News MSNBC FactCheck.org Fox News MSNBC

We asked the political interest question before the

we chose one widely known candidate and one lesser

knowledge battery questions because we did not want

known candidate to represent each political party. For

respondents who could not answer the knowledge

the Democrats, Hillary Clinton represents the widely

questions to conclude they were not interested in

known candidate whereas Bernie Sanders is considered politics. For the closed ended questions, we included the lesser-known candidate. For the Republicans, Jeb

“don’t know” options in order to prevent respondents

Bush represents the widely known candidate whereas

from guessing on a question in which they did not

Carly Fiorina represents the lesser known.2 The fact

know the answer. These “don’t know” options were

check for the two Democratic candidates – Hillary

subsequently coded as missing data for the purposes

Clinton and Bernie Sanders – is identical; we simply

of our analysis. The next set of questions assessed

changed the names of the candidates. Similarly, the fact

the respondents’ favorability ratings for a subset of

check for the two Republican candidates was identical,

presidential candidates.6 We asked questions regarding

except for the name of the candidates. In addition, for

six presidential candidates even though we were

each candidate we varied the source of the fact check,

only interested in four of these candidates so that the

alternating between FactCheck.org, Fox News, and

respondents’ attention was not focused on the four

MSNBC. We chose these three sources to represent

candidates in our stimulus, potentially increasing

a neutral source, a right-leaning source, and a left-

demand characteristics.7 We followed these questions

leaning source respectively.3 See Table 2 for an outline

with standard measures of partisanship (“Where would

of the experimental conditions.

you place yourself on the following scale? 1 = Strong

Questionnaire

Democrat, 2 = Weak Democrat, 3 = Independent

During the pretest questionnaire, we asked a series of questions assessing the respondents’ level of political interest4 and political knowledge.5

leaning Democrat, 4 = Independent, 5 = Independent leaning Republican, 6 = Weak Republican, 7 = Strong Republican, or 8 = don’t know?”) and ideology (“One

Fact Checking in the 2016 Presidential Campaign

way that people talk about politics in the United States

test if the respondents learned from the fact checks,

is in terms of ideology. The political views people hold

so we asked questions assessing the content of the fact

can be arranged on a scale ranging from 1 = extremely

check.14 At the end of the posttest questionnaire, we

liberal to 7 = extremely conservative. Where would you

included a series of standard demographic questions.15 Some threats to internal and external validity

place yourself on this scale or have you not thought about it enough to have an opinion?”) In addition,

in our experiment may have been present. With

we asked respondents to place eight political figures

regard to internal validity threats, subjects were able

on an ideological scale because we were interested in

to participate in the experiment over a six-day period

assessing respondents’ knowledge of these politicians.8

beginning October 30, 2015, and ending on November

Using the same seven point ideological scale,

4, 2016. Due to the experiment’s time frame,

we asked respondents to identify the ideological

subjects may have had an opportunity to discuss

leaning of major news sources, not only the sources

the experiment with other subjects (i.e., diffusion).

we included in our stimulus, in order to understand

However, since the subjects were participating in the

whether the respondents could correctly identify the political profile of Fox News, MSNBC, and FactCheck. org. Again, by asking about more news sources than those included in our stimuli, we do not focus respondents’ attention on the specific concerns of the experiment. Similarly, we asked respondents to rate the reliability of numerous news sources.9 Following exposure to the experimental stimuli, respondents were asked a series of questions in the posttest questionnaire. We measured three types of attitudes towards the candidates: favorability10, trait evaluations11, and issue competence assessments.12 We were interested in examining whether the fact check affected any of these dimensions of candidate assessments. For each of these questions, we randomized the ordering of the candidates’ names in order to minimize question-ordering effects. To determine if the respondents believed the information presented in the fact checks, we asked the respondents to rate how reliable they considered five

experiment online and were not debriefed immediately after their participation, diffusion is limited. With regard to external validity threats, since the subjects were participating in a study, they may have paid more attention to the fact check than they would have in a real life setting (i.e., reactive arrangements). The subjects included in our sample were selected from particular types of courses at a single university, so our sample may not be representative of the nation. Because students at a university are typically younger than the population, they may respond differently to the fact check (i.e., interaction of selection and stimulus). Overall, we believe our study has a great deal of internal validity given the successful randomization of subjects to various conditions. Similarly, we contend that the external validity of the study is relatively high as well since people commonly read political fact checks online. Results

news sources: the three sources used in the stimuli

Source Effect

and two others.13 With these questions, we are able to

establish if the stimulus influenced the respondents’

were more likely to learn from fact checks when the

views of these news organizations. We also wanted to

fact checks came from like-minded sources. In other

In our experiment, we examined if subjects

100 words, are liberals more likely to be persuaded by a

depending on the source of the fact check (see Table

liberal source and are conservatives more likely to

4). For both liberal and conservative respondents,

be persuaded by a conservative source. To examine

the source of the fact check did not influence what

this expectation, we controlled for the party of the

respondents learned from these fact checks.

candidate being fact checked. As the results in Table

3 show, our findings do not support our original

how liberals and conservatives viewed the news

expectation. We found that when conservatives are

sources. Specifically, did respondents’ assessments of

exposed to a fact check of a Republican candidate,

news sources sharing their ideological slant become

the conservative subjects’ ability to correctly identify

more or less positive when these news organizations

the information presented in the fact check did not

criticized candidates from their own party? To

differ significantly depending on whether the source

address this question, we began by analyzing whether

was considered conservative, liberal, or neutral. In

conservatives were more or less likely to view Fox News

other words, conservative respondents’ ability to retain

as reliable when Fox News fact checked a Republican

information presented in the fact checks did not vary

candidate, compared to a Democratic candidate. As

with the source of the fact check.

Table 5 shows, conservative respondents rate Fox

News as significantly more reliable when Fox News

We found the same pattern when examining

Second, we examined if fact checking impacted

liberal respondents – the accuracy of recall of

fact checked a Republican candidate, compared to

information presented in the fact check did not differ

a Democratic candidate. These results indicate that

Table 3. Conservative Respondents Exposed to Fact Check of Republican Candidate

Republican Candidates

N 57

95% Confidence Interval for Mean Mean Std. Error Lower Bound Upper Bound 2.23 .103 2.02 2.43

Democratic Candidates 45 Total 102 Note: F=1.50, n.s.

2.58 2.38

.164 .094

2.25 2.20

2.91 2.57

Table 4. Liberal Respondents Exposed to Fact Check of Democratic Candidate

Fact Check.org Fox News MSNBC Total Note: F=.041, n.s.

N 30

Mean 3.4667

Std. Error .16424

34 30 94

3.4412 3.5000 3.4681

.14741 .12457 .08381

95% Confidence Interval for Mean Lower Bound Upper Bound 3.1308 3.8026 3.1413 3.2452 3.3017

3.7411 3.7548 3.6345

Pi Sigma Alpha Undergraduate Journal of Politics

101

Table 5. Conservatives’ Reliability Assessments of Fox News as a News Source

Republican Candidates

N 57

95% Confidence Interval for Mean Mean Std. Error Lower Bound Upper Bound 2.23 .103 2.02 2.43

Democratic Candidates 45 Total 102 Note: F=3.53, p<.10

2.58 2.38

.164 .094

2.25 2.20

2.91 2.57

Table 6. Liberals’ Reliability Assessments of MSNBC as a News Source

Republican Candidates

N 87

95% Confidence Interval for Mean Mean Std. Error Lower Bound Upper Bound 2.62 .111 2.40 2.84

Democratic Candidates 94 Total 181

2.44 2.52

.096 .073

2.24 2.38

2.63 2.67

Note: F=1.59, n.s. conservative respondents seem to respect when Fox

more reliable when fact checking a Democratic

News offers critical commentary of a Republican

candidate, compared to a Republican candidate (see

candidate. While Fox News is expected to devote

Table 6). Overall, we find support for our expectation

negative coverage to Democrats given their ideological

that respondents will view news sources as more

slant, when Fox News scrutinizes candidates from

reliable when the news organization is fact checking a

“their own party,” they are rewarded with more

like-minded candidate.

positive reliability ratings. We also examined whether

conservatives were more or less likely to view MSNBC

evaluations of the candidates’ personal traits and

and FactCheck.org as reliable when MSNBC and

ability to deal with certain issues varied with the

FactCheck.org fact checked a Republican candidate,

source of the fact check. We hypothesize that fact

compared to a Democratic candidate, but we

check messages will produce more negative evaluations

uncovered no differences for these news outlets.

of candidates when the source of the fact checks and

the targeted candidate share the same ideological

After observing the results from the

Next, we examined whether respondents’

conservative reliability assessments, we sought

profile. Therefore, we expect to find respondents who

to determine whether we would find the same

view MSNBC fact checks of a Democratic candidate

pattern when looking at liberal assessments. Liberal

to rate the candidate more negatively than if the fact

respondents’ assessments of the reliability of MSNBC is

checks were authored by Fox News. In the same vein,

somewhat affected by whether MSNBC is fact checking

we expect to find respondents who view Fox News

a Democratic candidate or a Republican candidate.

fact checks of Republican candidates to rate the fact

Similar to the pattern with conservatives, we find that

checked candidate more negatively than if the fact

liberals view MSNBC as slightly (but not significantly)

check was authored by MSNBC.

Kahn and Wagner

102

see Fox News producing more negative views of Jeb

To test our hypothesis, we conducted four

analyses looking at whether evaluations of the

Bush, compared to MSNBC and Factcheck.org. The

candidates varied depending on whether the fact

story is different for Carly Fiorina, confirming our

checks were authored by FactCheck.org, Fox News,

expectations. In particular, people receiving the Fox

or MSNBC. The results show that across eight

News fact check consistently develop more negative

dependent variables (i.e., evaluations of the candidatesâ&#x20AC;&#x2122;

views of Fiorina, compared to people receiving fact

trustworthiness, knowledge, strength of leadership,

checks from the other news outlets.

compassion, ability to deal with education, terrorism,

immigration, and economic issues); impressions of the

check is more consequential for Republican candidates

Democratic candidates did not vary with the source of

than for Democratic candidates.16 In addition, we

the fact check message.

only find confirmation for our expectation for one of

the four candidates. For Carly Fiorina, we find that fact

However, among Republican candidates, the

These results show that the source of the fact

source of the fact check was more influential (see

check messages produce more negative evaluations of

Table 7 and Table 8). For instance, evaluations of

candidates when the source of the fact checks and the

Jeb Bush differed significantly across the source of

targeted candidate share the same ideological profile.

the fact check for four of the dimensions tested and

Fact Check Effect

evaluations of Carly Fiorina varied across the source of the fact check for three of the dimensions examined. While the source of the fact check mattered, we do not

In addition to examining the effect of the

source of the fact check, we examined whether subjects who viewed fact checks of the Democratic and

Table 7. The Impact of Fact Check Source on Evaluations of Jeb Bush

Bush Trustworthy F=2.67 p<.08

Bush Knowledgeable F=3.7 p<.03

Bush Strong Leader F=4.9, p<.01 Bush Competent on Education F=3.5 p<.04

FC.org FOX MSNBC Total FC.org FOX MSNBC Total FC.org FOX MSNBC Total FC.org FOX MSNBC Total

N 30 32 36 98 30 32 36 98 30 32 36 98 30 32 36 98

Mean 3.10 2.88 2.69 2.88 2.40 2.84 2.33 2.52 2.80 2.84 2.22 2.60 3.03 2.66 2.47 2.70

95% Confidence Interval for Mean Std. Error Lower Bound Upper Bound .121 2.85 3.35 .117 2.64 3.11 .131 2.43 2.96 .073 2.73 3.02 .170 2.05 2.75 .143 2.55 3.13 .126 2.08 2.59 .086 2.35 2.69 .169 2.45 3.15 .163 2.51 3.18 .150 1.92 2.53 .096 2.41 2.79 .162 2.70 3.37 .139 2.37 2.94 .152 2.16 2.78 .090 2.53 2.88

Fact Checking in the 2016 Presidential Campaign

103

Table 8. The Impact of Fact Check Source on Evaluations of Fiorina

Fiorina Trustworthy F=3.17 p<.05 Fiorina Competent on Economy F=3.56 p<.05 Fiorina Competent on Terrorism F=6.4 p<.01

FC.org FOX MSNBC Total FC.org FOX MSNBC Total FC.org FOX MSNBC Total

N 28 35 33 96 28 35 33 96 28 35 33 96

Mean 2.86 2.54 3.12 2.83 2.68 2.43 3.00 2.70 2.82 2.49 3.18 2.82

Republican candidates developed more negative views

95% Confidence Interval for Mean Std. Error Lower Bound Upper Bound .133 2.58 3.13 .198 2.14 2.95 .149 2.82 3.43 .099 2.64 3.03 .163 2.34 3.01 .165 2.09 2.76 .138 2.72 3.28 .093 2.51 2.88 .127 2.56 3.08 .155 2.17 2.80 .134 2.91 3.45 .086 2.65 2.99

assessments of how well the word “knowledgeable”

Overall, we find that people’s evaluations of the Republican candidates were more affected by the content of the fact check, compared to the Democratic candidates. For Bush and Fiorina, the fact check messages produced a total of six significant differences in evaluations. In comparison, fact checking of the Democratic candidates produced only one significant difference in evaluations.

describes Clinton are significantly lower for people

Conclusion

of the fact checked candidate. We analyzed evaluations of the candidates’ traits and issue competencies and failed to find that the fact check significantly influenced subjects’ views of the Democratic candidates for seven of the eight dependent variables examined. The one exception is that respondents’

who receive the Clinton fact check versus people who receive the Sanders fact check. (F=2.9, p<.09).

Likewise, we examined whether respondents

who viewed fact checks of Republican candidates developed more negative views of the candidate being fact checked. We found respondents’ evaluations of Bush’s trustworthiness and his ability to deal with immigration issues were negatively affected by the content of the fact check. Similarly, respondents’ evaluations of Fiorina’s trustworthiness, knowledge, and compassion ratings were negatively affected by the fact check, as were evaluations of her ability to deal with immigration issues.17

The results of our study suggest that the

source of a fact check does not consistently influence respondents’ views. First, people did not learn the substance of the fact check better if the fact check was authored by an ideologically similar source. We failed to find support for our expectation that conservatives would be more likely to learn the substance from fact checks disseminated by Fox News, while liberals would be better able to learn from fact checks emanating from MSNBC. However, the source of the fact check mattered for reliability assessments of the news organizations.

104 More specifically, liberals viewing fact checks by MSNBC viewed MSNBC as somewhat more reliable when MSNBC fact checked a Democratic candidate rather than a Republican candidate. Similarly, conservatives viewing a fact check by Fox News rated Fox News as significantly more reliable when Fox News fact checked a Republican candidate rather than a Democratic candidate. Consistent with our hypothesis, the news organization’s reliability was enhanced when these organizations critiqued their “own” candidates. We also examined whether respondents’

familiar with these candidates. Or, the student sample was more liberal than conservative and therefore, they may have been more open to criticisms of Republican candidates. Another possibility is that these particular Republican candidates were less viable than their Democratic counterparts and perhaps respondents were therefore more willing to accept the negative information given about these Republican candidates. With our current experiment, we cannot explain why the source of the fact check was more powerful for Republican candidates. However, we are the first study of fact checking to demonstrate that the source

evaluations of the candidates’ personal traits and ability to deal with certain issues varied with the source of the fact check. While we hypothesized that fact check messages will produce more negative evaluations of candidates when the source of the fact checks and the targeted candidate share the same ideological profile, we found only partial support for our expectations. For Carly Fiorina, the fact check authored by Fox News produced more negative impressions of the candidate. Finally, we explored whether candidates subject to fact checking were viewed more negatively than candidates who were not subject to fact checking. We found that people’s evaluations of the Republican candidates were more affected by the content of the fact check, compared to the Democratic candidates. That is, when Republican candidates were fact checked, people developed more negative impressions of their candidacies. Overall, our results indicate that the source of the fact check can be influential. We do find that fact checking – and the source of the fact check – is more consequential for Republican candidates than Democratic candidates. These results are intriguing, but the reasons for the party differences are not entirely clear. Perhaps people’s views of the Republican candidates are more malleable because they were less

of fact checking can be consequential. We encourage researchers to continue to explore the impact of the source of fact checks. In our highly partisan environment, with conservatives tending to listen to conservative media and liberals often relying on more liberal sources of news, the source of news messages may be particularly relevant. Our current research project is limited due to our reliance on student subjects who were younger, more informed, and more liberal than a more representative sample. Future research should examine the impact of the source of a fact check with a more nationally representative sample. Additionally, the impact of fact checking may differ depending on the type of message being fact checked (e.g., speeches, advertisements, debates). Therefore, future research should begin to explore the effects of different types of content being fact checked. It is important to appreciate the power of fact checking since fact checking is now an integral part of campaign coverage. We believe this experiment represents a first step in increasing our understanding of how the fact checking of presidential candidates’ statements impacts the public. In our electoral system, it is imperative that the news media provide citizens with reliable information about the competing

Pi Sigma Alpha Undergraduate Journal of Politics candidates for elective office. Therefore, we contend that it is essential to improve our understanding of the effectiveness of fact checking as a form of news coverage.

Notes The other majors included in our sample are: global studies (9%), social science major other than political science and global studies (6%), humanities (2%), natural science (3%), engineering (5%), business (14%), and other (22%). 1

Although he was widely known, Donald Trump was excluded as the widely-known Republican candidate because he is not a typical politician. In addition, at the time of designing the study (the study was conducted between October 30, 2015 and November 4, 2015), it was unclear if Trump would continue to pursue the Republican nomination. Additionally, Bernie Sanders was selected as the lesser-known Democratic candidate as opposed to Martin O’Malley because, although Sanders was arguably better known than O’Malley at the time, we were not confident that Martin O’Malley would continue to pursue the Democratic nomination. 2

According to our pretest results, among our respondents, the mean ideological rating for FactCheck.org is 3.8, the mean ideological rating for Fox News is 6.18, and the mean ideological rating for MSNBC is 2.59. The respondents were asked to place these sources on a scale ranging from 1 (extremely liberal) to 7 (extremely conservative). 3

Political interest was measured on a scale: “Some people don’t pay much attention to political campaigns. How about you? Would you say that you are very much interested, somewhat interested, or not much interested in political campaigns?” (1 = very 4

105

much interested, 2 = somewhat interested, 3 = not much interested, 4 = don’t know) Political knowledge was measured by the respondents’ ability to correctly answer four knowledge questions: “Do you happen to know what job or political office is now held by John Kerry?” “Whose responsibility is it to determine if a law is constitutional or not—is it the president, the Congress, or the Supreme Court?” “How much of a majority is required for the U.S. Senate and U.S. House to override a presidential veto?” “Do you happen to know which party has the most members in the House of Representatives in Washington D.C. 5

today?” If the respondent answered the question correctly, it was coded as 1 and if the respondent got the answer wrong it was coded as 0. Respondent knowledge ranged from 0 (getting 0 questions correct) to 4 (getting all 4 questions correct). The respondent’s favorability was measured on a tenpoint scale: “On a scale from 1 to 10, where 1 means you feel very unfavorable toward the candidate and 10 means you feel very favorable, how would you rate the following candidates?” 6

We randomized the order of the names of the political figures in the pretest questionnaire in order to limit question-ordering effects (e.g., primacy effects). 7

Both party identification and ideological leaning were collapsed into three categories. Party identification: (1) Strong, Weak, or Leaning Democrat; (2) Independent; (3) Strong, Weak, or Leaning Republica Ideology: (1) Extremely Liberal, Liberal, or Somewhat Liberal; (2) Moderate; (3) Extremely Conservative, Conservative, or Somewhat Conservative. 8

The reliability of the news organizations was measured on a four point scale with the question: “How reliable do you consider these news 9

Kahn and Wagner

106

organizations to be on a scale ranging from very reliable (4) to not reliable at all (1)?” Favorability were measured on a 10 point scale: “On a scale from 1 to 10, with 1 meaning that you feel very unfavorable toward the candidate and 10 meaning you feel very favorable toward the candidate, how would you rate the following candidates?” 10

The trait evaluations of the candidates were measured on a scale of how well the respondents felt the words trustworthy, knowledgeable, compassionate, strong leader described the candidate ranging from (1) 11

extremely well to (4) not well at all. The issue competence of the candidates was assessed by how competent the respondents felt the candidates were regarding the issues of immigration, education, the economy, and fighting terrorism ranging from (1) very competent to (4) very incompetent. 12

Appendix A: Stimuli Hillary Clinton/Bernie Sanders Stimulus Clinton’s/Sanders’ speech to Iowans: Fact-checking some key claims On Wednesday, Democratic candidate Hillary Clinton delivered a speech to supporters in Des Moines, Iowa. We fact checked some of the statements made in the speech. Clinton began her speech talking about immigration and asked the crowd, "How do Jeb Bush and Donald Trump differ on immigration? Spoiler alert: They don't." While Bush and Trump agree in some areas, they have also shown sharp contrasts on immigration policy. Bush and Trump agree on (1) Obama’s executive action: When we compare recent statements, Trump and Bush agree on providing legal status to certain immigrants and not citizenship, (2) Anchor babies: "I’ll use the word

anchor baby," Trump said Aug. 20. Bush on Aug. 19 called for "better enforcement so that you don't have these, you know, 'anchor babies,' as they're described, coming into the country." According spokespersons for the Bush and Trump campaign, there are at least two ways they differ: (1) Trump has called for mass deportations. Bush has repeatedly said that mass deportation is an unrealistic solution. (2) Trump calls for ending birthright citizenship and argues that it could be done despite the 14th amendment. Bush defended birthright citizenship. Our ruling: Clinton said, "How do Jeb Bush and Donald Trump differ on immigration? Spoiler alert: They don't." Bush and Trump agree in some areas, including getting rid of Obama’s executive action. But in other areas they disagree. Clinton’s overall message is misleading, even though they overlap on some policies. We rate this claim Mostly False. Clinton then highlighted her own family history as a way of demonstrating her commitment to immigration reform. Clinton related her personal family heritage to the struggles of undocumented immigrants trying to work in the United States. "I think if we were to just go around this room, there are a lot of immigrant stories," "All my grandparents, you know, came over here, and you know my grandfather went to work in a mill in Scranton, Pa., and worked there until he retired at 65. So I sit here and I think well you’re talking about the second, third generation. That’s me, that’s you." Buzzfeed’s, Andrew Kaczynski quickly pointed out that Clinton was wrong, primarily using census and military records from Ancestry.com. An unidentified Clinton spokesman admitted the error in a statement to BuzzFeed, saying Clinton’s "grandparents always spoke about the immigrant experience and, as a result she has always thought of them as immigrants." While it’s clear Clinton is incorrect, Megan Smolenyak a

Fact Checking in the 2016 Presidential Campaign genealogical expert said that seven of Clinton’s eight great-grandparents were born overseas into workingclass families. Clinton’s paternal grandfather, Hugh Rodham Sr., was born in England, but her three other grandparents were born in the United States. Even though three of Clinton’s grandparents were born in the United States, Smolenyak said they probably strongly identified with their immigrant parents. Our ruling: Talking about immigration in Iowa, Clinton said, "All my grandparents, you know, came over here." It’s clear from the evidence that not all of Hillary Clinton’s grandparents were immigrants. In fact, only one was. We rate Clinton’s claim False. Clinton began her speech with a discussion of education by pointing out inequities in educational opportunities. Clinton said, "Our schools are still segregated, in fact, more segregated than they were in the 1960s." According to a 2014 study by UCLA’s Civil Rights Project, schools are slightly less integrated now than they were in 1968. "It’s true that segregation for blacks is worse today than it was in 1968," said Gary Orfield, a UCLA professor of education and lead author of the study. In particular, densely populated cities of the Northeast and West are becoming more and more segregated. On the Pacific coast, Clinton’s claim is on the money: Latino students are now more isolated than black students and "more segregated than they’ve ever been," according to Orfield. School segregation has been intrinsically tied to the racial gaps in housing and income, leading to the re-emergence of the color line. Economic segregation, which disproportionately affects black and Latino students, is increasing, pointed out by Orfield. "Inequality is very related to the double segregation of low-income racial minorities and (their) isolation from the middle class, from the best teachers, the best curriculum. Our ruling: Hillary Clinton said, "In America today, our schools are more segregated than they were in

107

the 1960s." Overall, experts say and the data shows that the United States has reverted to pre-Civil Rights levels of segregation. She is correct when she says the country has fallen back from the high levels of diversity that existed from the 1970s to the early 1990s. On the whole, we rate her statement as True. Clinton then began discussing the importance of quality education for the future of the country. Clinton said, "Not one of the 15 GOP candidates has discussed how they'd address the rising cost of college." It is true that most of the GOP candidates have not offered any details about addressing the rising cost of a college education. However, Rubio has been talking about the issue since at least February 2014, when he made a major policy speech about controlling crippling college debt. Rubio’s plan includes several specific ideas including: (1) establishing a new accreditation process for nontraditional education, (2) requiring schools to tell students what kind of salary they can expect to earn with every degree, and (3) Basing loan repayment plans on the amount of money a graduate makes annually. In June, New Jersey Gov. Chris Christie laid out several ideas for reforming post-secondary education with an eye on reining in costs. Carly Fiorina suggested letting private banks compete for student loans. Sen. Rand Paul of Kentucky said in April 2015 he thought college tuition should be tax deductible. The remaining 11 candidates in the GOP field had not discussed how they would address the rising cost of college. Our ruling: Clinton said, "Not one of the 15 GOP candidates has discussed how they'd address the rising cost of college." While most of the Republican hopefuls haven’t had much to say on the issue, it’s not accurate to say none of them have. We rate Clinton’s statement as Mostly False. Jeb Bush/Carly Fiorina Stimulus Bush’s/Fiorina’s speech to Iowans: Fact-checking some key claims

108 On Wednesday, Republican candidate Jeb Bush delivered a speech to supporters in Des Moines, Iowa. We fact checked some of the statements made in the speech. Bush began the speech discussing issues related to immigration. Bush discussed how the flood of illegal immigrants creates economic hardships for United States’ citizens. Bush explained that hundreds of thousands of illegal immigrants have been placed in state and federal penitentiaries in the last year. We polled experts and asked government agencies and found no data that speaks to this claim conclusively. "The basic claim is at best unsustainable or more likely pure fiction. A fact created out of thin air," said Ramiro Martinez, a criminal justice professor at Northeastern University. "There is no substance to the Jeb Bush’s assertion since the data do not exist." The most recent comprehensive number we could find comes from the Bureau of Justice Statistics. In total, there were around 90,000 noncitizens incarcerated in federal and state prison systems at year-end in 2013. But this doesn’t tell us much; the noncitizen group includes both legal and illegal immigrants, and the report does not separate the two. But in any case, we know that there are definitely not "hundreds of thousands" of illegal immigrants in state and federal penitentiaries. We also wanted to note that every expert we polled said there is a consensus among scholars that undocumented immigrants are not more likely to commit crimes than U.S. citizens. Our ruling: Bush said "Hundreds of thousands of illegal immigrants are going to state and federal penitentiaries." The bottom line on Bush’s specific claim is that there’s no definitive number, or even a good ballpark figure, of how many illegal immigrants are currently incarcerated. Without any good information to back it up, Bush’s claim falls flat. We rate Bush’s claim as Mostly False. Bush later asserted that Mexico, unlike the United States, does not have birthright citizenship.

Bush went on to say that the United States is the only country in North America to recognize birthright citizenship. Mexico does offer birthright citizenship: The Mexican Constitution says the Mexican "nationality" is obtained by birth if the person is born "within the Republic’s territory whatever their parents’ nationality might be," among other circumstances. Article 37 of the Mexican Constitution adds that Mexican nationality by birth "shall never be revoked." Birthright citizenship also exists in Canada, Brazil and nearly every other country in Central and South America, according to a list of nations with birthright citizenship maintained by Numbers USA, which supports reduced immigration levels. Countries that offer birthright citizenship are located almost exclusively in the Western Hemisphere. No country in Europe or East Asia, for example, has a similar citizenship policy. Our ruling: The United States and Mexico, along with more than 30 countries around the world, predominantly in the Americas, offer birthright citizenship. We rate Bush’s claim as False. After discussing the problem of illegal immigration, Bush turned his attention to the failing economy. He began by saying "For the first time in 35 years, we have more businesses dying than we do starting," We took a closer look. We located a May 2014 report by the Brookings Institution, a Washington-based think tank, titled, "Declining Business Dynamism in the United States: A Look at States and Metros." In that report, coauthors Robert Litan and Ian Hathaway published data that supports Bush’s claim. They found that the rate of business failures held steady, except for an uptick during the Great Recession -- but they also found that the entry rate of new firms declined by nearly half between 1978 and 2011. "The precipitous drop since 2006 is both noteworthy and disturbing," the authors wrote. "Business deaths now exceed business births for the

Pi Sigma Alpha Undergraduate Journal of Politics first time in the 30-plus-year history of our data," the authors wrote. We contacted Litan to make sure Bush wasn’t twisting the statistic in any way, and Litan responded that Bush’s claim is accurate. Our ruling: Bush said, "for the first time in 35 years, we have more businesses dying than we do starting." The co-author of that report said he feels Bush has stated the claim accurately. We rate Bush’s claim True. Finally, Bush argued that the government stands in the way of lower and middle class Americans getting ahead. Bush said, "If low-income people work and make more money, they lose more in benefits than they would earn in salary," According to the Congressional Budget Office, there are very few instances when all 100 cents of an additional dollar earned would go to taxes and replacing lost benefits. In fact, among low- and moderate-income taxpayers, less than one percent lose 80 cents or more of every additional dollar earned. Those most affected would likely be individuals who earn near or just above the poverty level (about $20,000 in a three-person household), who are also enrolled in multiple benefit programs that are set to phase out with any additional income. This is not the norm. The vast majority of people face some increases in taxes and decreases in benefits when they make more money, but these individual would still take home substantially more in pay than they would under a lower salary. Our ruling: Bush said "If people work and make more money, they lose more in benefits than they would earn in salary." There are a few scenarios where a low-income could potentially lose more in benefits than he or she would gain by a slight increase in income. However, these cases are rare. Bush’s statement contains some element of truth but ignores critical facts that would leave a different impression. We rate Bush’s statement as Mostly False.

109

References Adair, Bill, and Angie Drobnic Holan. 2011. “Remembering David Broder and His Passion for Fact-Checking,” PolitiFact.com, March 11. http://www.politifact.com/truth-o-meter/ article/2011/mar/11/remembering-davidbroder/ (Accessed October 3, 2015). Fridkin, Kim, Patrick J. Kenney, and Amanda Wintersieck. 2015. “Liar, Liar, Pants on Fire: How Fact-Checking Influences Citizens’ Reactions to Negative Advertising.” Political Communication. 32 (1): 127-151. FactCheck.org. 2015. “About Us.” Annenberg Public Policy Center. http://factcheck.org/about (Accessed October 3, 2015). Gallup.com. 2016. “Party Affiliation.” http://www. gallup.com/poll/15370/party-affiliation.aspx (Accessed October 3, 2015). Garrett, R. Kelly, Nisbet, Erik C., and Lynch, Emily K. 2013. “Undermining the Corrective Effects of Media—Based Political Fact Checking? The Role of Contextual Cues and Naïve Theory.” Journal of Communication 63(4): 617-637. Nyhan, Brendan, and Jason Reifler. 2010. “When Corrections Fail: The Persistence of Political Misperceptions.” Political Behavior 32(2): 303330. Nyhan, Brendan, and Jason Reifler. 2015a. “Estimating Fact-Checking’s Effects.” http://www. americanpressinstitute.org/wp-content/ uploads/2015/04/Estimating-Fact-ChangesEffect.pdf (Accessed October 3, 2015). Nyhan, Brendan, and Jason Reifler. 2015b. “The Effect of Fact-Checking on Elites: A Field Experiment

110

Kahn and Wagner on U.S. State Legislators.” American Journal of Political Science 50(3): 628-640.

Wintersieck, Amanda, and Kim Fridkin. 2016. “Fact-Checking During Political Campaigns: Results from a Content Analysis.” The Praeger Handbook of Political Campaigning in the United States, ed. William L. Benoit. Santa Barbara: ABC-CLIO, LLC, 145-164. Wintersieck, Amanda, Kim Fridkin, Patrick Kenney, Josh Thompson, and Jillian Courey. 2015. “The Message Matters: The Influence of FactChecking on Evaluations of Political Messages.” Presented at the Annual Meeting of the American Political Science Association, San Francisco.

Fact Checking in the 2016 Presidential Campaign

111