IARFC Journal of Personal Finance Volume 20 Issue 1 2021 by IARFC

Volume 20 Issue 1 2021 www.journalofpersonalfinance.com

Journal of Personal Finance

Techniques, Strategies and Research for Consumers, Educators and Professional Financial Consultants

IARFC

Are You

Ethics Approved?

How to Obtain an Ethics Approved Status • • • • •

Apply for an IARFC designation/credential Adhere to IARFC Code of Ethics Clear record for past 5 years Successful completion of IARFC Ethics Exam Verified through FINRA, State Licensing records, and internet searches

www.iarfc.org

Journal of Personal Finance

Journal of Personal Finance Volume 20, Issue 1 2021 Editor Craig W. Lemoine, Ph.D., MRFC®, CFP® University of Illinois

Editorial Board

Sarah D. Asebedo, Ph.D., CFP® University of Arizona Daria J. Auciello Newfeld, Ph.D. Albright College H. Stephen Bailey, Ph.D., MRFC® HB Financial Resources, Ltd./IARFC David Blanchett, Ph.D., CFA®, CFP® Morningstar Investment Management, LLC Swarn Chatterjee, Ph.D. University of Georgia Yuanshan Cheng, Ph.D., CFA, CFP®, FRM Winthrop University Preston D. Cherry, Ph.D., CFP® University of Wisconsin — Green Bay Chia-Li Chien, Ph.D., CFP®, PMP® California Lutheran University Ronnie Clayton, Ph.D. Jacksonville State University Ben Cummings, Ph.D., CFP®, RFC® Utah Valley University Dale L. Domian, Ph.D., CFA®, CFP® York University Lu Fan, Ph.D., CFP® University of Missouri

Michael S. Finke, Ph.D., CFP® The American College of Financial Services Joseph W. Goetz, Ph.D. University of Georgia Michael A. Guillemette, Ph.D., CFP® Texas Tech University Tao Guo, Ph.D., CFP®, CFA William Paterson University Douglas A. Hershey, Ph.D. Oklahoma State University Vera Intanie Dewi, Ph.D. Universitas Katolik Parahyangan Carrie L. Johnson, Ph.D., AFC® North Dakota State University Kyoung Tae Kim The Ohio State University Douglas J. Lamdin, Ph.D. University of Maryland, Baltimore County (UMBC) David Nanigian, Ph.D., CFP® Dr. Thomas Warschauer Endowed Professor of Finance Fowler College of Business, San Diego State University

Blain Pearson, Ph.D., CFP® Kansas State University Wade D. Pfau, Ph.D., CFA The American College of Financial Services Nilton Porto, MBA/Ph.D. University of Rhode Island Abed G. Rabbani, Ph.D., CFP® University of Missouri Brandon Renfo, Ph.D., CFP®, RICP®, EA Eastern Texas Baptist in the Fred Hale School of Business Laura Ricaldi, Ph.D., CFP® Utah Valley University Donald Bruce Ross III, Ph.D., AFC® University of Kentucky Zack Taylor, Ph.D. Trellis Company Jacob Tenney, Ph.D., CFP® University of Charleston Sandra Timmermann, Ed.D. The American College of Financial Services Walt Woerheide, Ph.D., ChFC®, RICP® Jing Jian Xiao, Ph.D. University of Rhode Island

Mailing Address:

Disclaimer: The Journal of Personal Finance is

General Editorial Policy: It is the editorial policy of this Journal to only publish content that is original, exclusive, and not previously copyrighted.

IARFC® Journal of Personal Finance 146 N. Breiel Blvd Middletown, OH 45042

Postmaster: Send address changes to

IARFC® Journal of Personal Finance 146 N. Breiel Blvd Middletown, OH 45042

Permissions: Requests for permission to make copies or to obtain copyright permissions should be directed to the Editor at editor@iarfc.org. Designations/Credential Inquiries:

For more information about the Registered Financial Consultant, Master Registered Financial Consultant, or Registered Financial Associate, or to find a consultant, please visit www.iarfc.org.

intended to present timely, accurate, and authoritative information. The editorial staff of the Journal is not engaged in providing investment, legal, accounting, financial, retirement, or other financial planning advice or service. Before implementing any recommendation presented in this Journal readers are encouraged to consult with a competent professional. While the information, data analysis methodology, and author recommendations have been reviewed through a peer evaluation process, some material presented in the Journal may be affected by changes in tax laws, court findings, or future interpretations of rules and regulations. As such, the accuracy and completeness of information, data, and opinions provided in the Journal are in no way guaranteed. The Editor, Editorial Board, and the Board of the International Association of Registered Financial Consultants specifically disclaim any personal, joint, or corporate (profit or nonprofit) liability for loss or risk incurred as a consequence of the content of the Journal.

Subscription requests should be addressed to: IARFC® Journal of Personal Finance 146 N. Breiel Blvd Middletown, OH 45042 editor@iarfc.org (800) 532-9060 Subscription Rates, Includes 1 year, 2 issues Individual: Members $45; Non-Members $65 Institutional: $120, 3 copies, each issue Single Copies: Members $25, Non-Members $35 Digital Download: $20 (from store.iarfc.org) Add $15/issue for delivery outside the U.S. The Journal of Personal Finance is published in the U.S. in the months of March and October by the International Association of Registered Financial Consultants (IARFC®). ISSN 1540-6717 (Print); 2638-3217 (Online)

Volume 20 • Issue 1 2021

Call for Papers Journal of Personal Finance The Journal of Personal Finance is seeking high-quality manuscripts that add to the growing literature in personal finance and household financial decision making. The editor is looking for original research that uncovers new insights — research that will have an impact on professional financial advice provided to individuals. Potential topics include: •

Individual financial decision making

•

Household portfolio choice

•

Retirement planning and income distribution

•

Household risk management

•

Life cycle consumption and asset allocation

•

Investment research relevant to individual portfolios

•

Household credit use

•

Professional financial advice and its regulation

•

Behavioral factors related to financial decisions

•

Financial education and literacy

•

Wealth management

•

Diversity and inclusion within financial services

Please check the “Submission Guidelines” on the Journal’s website (www.journalofpersonalfinance.com) for more details about submitting manuscripts for consideration. The Journal of Personal Finance is committed to providing high-quality article reviews in a blind, single-reviewer format within 60 days of submission.

Editorial Board The Journal is seeking qualified members for the Editorial Board. Those interested in joining the Editorial Board should send their current CV to the editor at the email address below.

Contact Craig W. Lemoine, Ph.D., MRFC®, CFP® Editor jpfeditor@iarfc.org www.journalofpersonalfinance.com

Volume 20 • Issue 1 2021

Contents Determinants of Parents’ College Education Saving Decisions �� 8 Thomas Korankye, Ph.D., CFP® Charlene M. Kalenkoski, Ph.D., CFP® This study examines the factors associated with the decisions of U.S. households to save for the college education of their children using state-level data from the 2015 U.S. National Financial Capability Study. The results suggest that financially fragile households and those characterized by low income, low education, more children, no health insurance, and no homeownership are less likely to have college savings. Households who are willing to take financial risks and those with higher perceived financial knowledge are more likely to save for the college education of their children. Whereas bond- and loan-pricing literacy are the only components of financial literacy that are associated with a higher probability of college savings, the overall financial literacy score is associated with a lower likelihood of saving for college. Student Loan Debt Letters: How Colleges Communicate Debt with Students �� 26 Zachary Taylor, Ph.D. Gretchen Holthaus Karla Weber As the student loan debt crisis has continued to gain national attention from higher education leaders, education policymakers, and the media, states have begun mandating that institutions send student loan debt letters to any current or former student with outstanding student loan debt. Preliminary studies of the effectiveness of student loan debt letters have been mixed, but these studies have not analyzed how institutions have composed student loan debt letters at the word-, sentence-, and document-level. As a result, this study gathered six student loan debt letters sent by different institutions across the United States and analyzed these letters for readability, cohesion, and lexical diversity. Results suggest student loan debt letters have been written in drastically different ways and do not share common vocabulary, possibly confusing the debt repayment process for students. Implications for research and practice are addressed. The Relationship Between Home Equity and Retirement Satisfaction �� 40 Blain Pearson, Ph.D., CFP® Donald Lacombe, Ph.D. As life expectancies continue to climb, home equity may increasingly become a much-needed resource to finance late-life consumption. This study hypothesizes that a higher ratio of home equity relative to net worth creates resource constraints, resulting in disutility for individuals who are in retirement. The findings suggest that increases in a retiree’s ratio of home equity relative to net worth is associated negatively with a satisfactory retirement experience. The ensuing discussion highlights two important issues. The first issue is that, for many retirees, home equity is inefficient in promoting retiree well-being. The second issue is that retirees may have limited knowledge of the available tools to access home equity. Thus, arguments for increased efforts to promote the responsible utilization of home equity as a part of an individual’s plan for retirement are discussed. Investment Strategies During the Great Recession: Who Remains Calm, and Who Panics?—The Role of Financial Planners �� 51 Shan Lei, Ph.D., CFA, CFP® This study uses a proprietary dataset to investigate factors related to investment strategies chosen in the wake of the “Great Recession” of 2007 to 2009 and discussed the role of financial planners in particular. This study finds support for a positive relationship between investors using financial planners and following a disciplined investment strategy or portfolio rebalancing strategy. Additionally, this study also finds that using financial planners reduces the likelihood of holding the losers too long in the down market which might hurt individual investors' abilities to achieve long-term financial goals. Personal characteristics, such as gender, age, race, personal saving rate, and investable assets may present an essential part in shaping individuals' investment decisions in a recession. Potential investment bias associated with each investment strategy is discussed. Suggestions are also provided to overcome or accommodate these biases.

Journal of Personal Finance

Mortality Salience Lowers Preferred Retirement Asset Decumulation Rates �� 67 Yi Liu, Ph.D. Russell James, J.D., Ph.D., CFP® Standard life-cycle economic theory suggests that people should spend down assets during retirement at a rate maximizing their lifetime consumption. However, actual retiree behavior exhibits asset decumulation at much slower rates, not at all, or even continued accumulation. One potential factor is that decumulation requires personal mortality estimations. Previous research finds that personal mortality reminders (1) are aversive and (2) increase focus on impacting those who will survive. Correspondingly, recent research has found that (1) annuity purchases are reduced due to associated personal mortality reminders and (2) mortality reminders increase relative preference for annuities that pay less income but provide a greater bequest provision. This study investigates whether mortality salience will also influence the broader issue of an individual’s asset decumulation decisions in retirement. Using a randomly assigned experimental test, we find that increasing mortality salience increases the desire to retain assets in retirement, reducing the preferred spending rates in retirement. Understanding the role of mortality salience on decisions about asset decumulation in retirement can be beneficial to academic researchers and financial planners.

Volume 20 • Issue 1 2021

Editor's Notes I am excited to introduce the latest issue of the Journal of Personal Finance. Our Spring, 2021 edition reflects traditional touch points across a client’s life-cycle diving into student debt, home equity, education savings decisions, investment strategies, and mortality salience. Our latest issue enriches professional and academic discussions for new graduates, those building wealth, and retirees. Our first article examines determinants of parents’ college education saving decisions. Utilizing data from the 2015 U.S. National Financial Capability Study, Thomas Korankye and Charlene M. Kalenkoski find that financial fragile households are less likely to engage in saving for children’s higher education. Households willing to take risks were more likely to save; however overall financial literacy scores and the decision to save for college were negatively related. In our second article Zachary Taylor, Gretchen Holthaus and Karla Weber explore Student Debt Letters, analyzing letters for readability, cohesion, and lexical diversity. Letters varied greatly by institution leading to potential confusion and less than optimal outcomes. Blain Pearson and Donald Lacombe investigate the relationship between home equity and retirement satisfaction in our third piece. The authors uncover two driving issues surrounding retirees and home equity. The first issue is that, for many retirees, home equity is inefficient in promoting retiree well-being. The second issue is that retirees may have limited knowledge of the available tools to access home equity. In our forth contribution, Shan Lei studied behavior during the great recession. Who remains calm and who panics? Financial planners and personal characteristics help shape how we behave in times of financial stress. Disciplined financial strategies can help protect investors from bias and potentially catastrophic decisions. Our last article in this issue ends with a discussion of mortality salience. Yi Liu and Russell James investigate whether mortality salience will also influence the broader issue of an individual’s asset decumulation decisions in retirement. Using a randomly assigned experimental test, the authors find that increasing mortality salience increases the desire to retain assets in retirement, reducing the preferred spending rates in retirement. This set of five articles span saving for college to considering mortality. Together they provide a fascinating exploration across a client’s life cycle. We hope you enjoy this issue of the Journal of Personal Finance. Sincerely, Craig W. Lemoine, Ph.D., MRFC®, CFP® Editor

Journal of Personal Finance

Determinants of Parents’ College Education Saving Decisions

Thomas Korankye, Ph.D., CFP® Charlene M. Kalenkoski, Ph.D., CFP®

Abstract This study examines the factors associated with the decisions of U.S. households to save for the college education of their children using state-level data from the 2015 U.S. National Financial Capability Study. The results suggest that financially fragile households and those characterized by low income, low education, more children, no health insurance, and no homeownership are less likely to have college savings. Households who are willing to take financial risks and those with higher perceived financial knowledge are more likely to save for the college education of their children. Whereas bond- and loan-pricing literacy are the only components of financial literacy that are associated with a higher probability of college savings, the overall financial literacy score is associated with a lower likelihood of saving for college.

Keywords Savings for college, financial fragility, financial literacy, post-secondary education

Volume 20 • Issue 1 2021

INTRODUCTION Parental savings for the college education of children play an essential role in the American economy. Research shows that they are linked significantly to educational outcomes such as children’s four-year college attendance (Song and Elliot, 2012), college graduation (Nam & Ansong, 2015), and students’ reliance on loans for college (Elliot et al., 2014). Lemieux (2006) suggests that postsecondary education is an important factor in explaining wage inequality. The 2016 Survey of Consumer Finances also reports that the median income of households with a college education is more than twice the median income of households with less than a college education (Bricker et al., 2017). Bogan (2015) suggests that the inability of parents to fund the college education of their children influences their children’s potential future earning capacity and human capital development adversely. Despite the benefits of saving for college, Lusardi (2011) notes that most Americans do not save for the college education of their children. Meanwhile, the cost of college education continues to increase (Schell-Olsen, 2018). At the same time, the amount of student-loan debt has exceeded $1 trillion, serving as a threat to the American economy because of potential repayment difficulties (Federal Reserve Bank of New York, 2018; U.S. Department of Education, 2017). To inform policy formulation and implementation, a number of research papers have examined some of the factors that could influence parental savings for college. For example, Dondero and Humphries (2016) study the factors that predict parental college education saving decisions of Asian and Latino immigrants. Bogan (2015) examines the effect of family composition on college savings. Hossler and Vesper (1993) consider the factors associated with parental savings for the college education of a sample of 182 students from 21 high schools in the state of Indiana. Steelman and Powell (1991) use the 1980 Parent Survey of the High School and Beyond to study the willingness of parents to provide financial support for their children’s higher education. As important as these studies are, it appears that current research that broadly assesses the saving behavior of parents for college is sparse. The purpose of the current study is to examine the determinants of the saving decisions of parents for the college education of their children. The present paper extends the literature in two main ways. First, it uses a nationally representative survey data from the 2015 National Financial Capability Study (NFCS) to examine parents’ saving decisions

for college from a broader perspective than what other researchers have done previously. This study is broader than previous studies in the sense that it examines the effects of several factors, including financial literacy, financial-planning horizon, financial-risk taking, subjective financial knowledge, student debt, and financial fragility on college-saving decisions of U.S. parents. Previous studies of this nature have not examined these essential factors to a greater extent for the U.S. population. As a result, researchers, practitioners, and policymakers have insufficient results from empirical research to aid their understanding of the roles these factors play in influencing the behavior of U.S. parents towards saving for college. Second, the current paper performs sensitivity analyses for the components of financial literacy. These financial literacy components include numeracy, inflation, bond pricing, loan pricing, mortgage payments, and risk diversification. This approach uncovers the elements of financial literacy that financial educators can focus on to enhance the saving decisions of parents for the college education of their children. After carrying out the empirical analyses, the results show that homeownership, being married, educational attainment of parents, household income, and ownership of health insurance are associated positively with the probability of saving for college. The study finds parents’ financial-risk behavior and subjective financial knowledge to be positive determinants of saving for college. The study also finds that financial fragility, being white, and the age of parents are associated negatively with saving for college. The results further show that parents with four or more children are less likely to save for the college education of children compared with parents with one child. Regarding financial literacy, the results suggest that bond- and loan-pricing literacy are associated positively with college savings, while compound-interest, inflation, mortgagepayment, and risk-diversification literacy have negative relationship with college savings.

LITERATURE REVIEW The economic organization of the household is such that households make consumption and saving decisions to maximize their satisfaction. A household may choose to invest part of current savings in family members, for example, through saving for children’s college education (Becker, 1964; Bryant & Zick, 2005). Parental savings for children’s college education is a form of human capital investment. According to the human capital theory, investing in education enhances

Journal of Personal Finance

the productive capacities of individuals. To maximize future wealth, a household would want to invest in children’s college education to the extent that the marginal cost of schooling equals the marginal benefit of schooling. The literature documents several factors that could influence the financial behaviors of households. This section reviews household characteristics pertaining to the saving decisions of parents for the postsecondary education of their children.

than non-couples (Xiao & Noring, 1994). Grissett and Furr (1994) indicate that children from divorced homes are less likely to receive financial assistance from their family for their postsecondary education. The ability of married couples to save for college education may stem from the fact that they are expected to have more significant financial resources than those who are not married owing to factors such as economies of scale and opportunities for specialization.

One factor that has been identified to influence college savings is the age of the parent. Yilmazer (2008) shows that the relationship between the age of the household head and the amount that a household sets aside for children’s college is positive. The findings indicate that a household’s propensity to save for college is likely to increase as the age of the parent increases. The desire of the head of household to smooth consumption over the life cycle may explain this finding.

The effect of the financial-planning horizon on financial behaviors other than college education savings has been studied extensively in the literature (Hong et al. 2014; Finke, 2006; Frederick et al., 2002; and Bernheim et al., 2001). Gutter et al. (2012) show that longer financial-planning horizon relates negatively to the likelihood of having only a savings account among low-to-moderate income households. However, in the same study, the authors do not find a statistically significant effect of the planning horizon on having both savings and investment accounts. In another study, Ashraf et al. (2006) show that individuals with low discount rates tend to exhibit positive saving behaviors such as participating in a commitmentsavings program. Using a couples’ data from the 1998 Survey of Consumer Finances, Devaney and Chien (2002) point out that a longer planning horizon is a likely determinant of having a savings goal in the form of postsecondary education.

Another factor is the race of parents. The literature observes that both minority and white parents endeavor to save for their children’s college education (Lee et al., 1997; Steelman & Powell, 1993). Steelman and Powell (1993) note that white parents are more likely to expect their offspring to shoulder part of the cost of funding their college education. On the contrary, the authors find that minority parents are more likely to hold the opinion that funding for their children’s education is both the responsibility of the government and themselves. Research shows that educational attainment of the parent also influences the ability of individuals to make financial decisions (Lusardi & Mitchell, 2014). Gutter et al. (2012) find that individuals with less education have a low propensity to own both investment and savings accounts. Using data from the 1992 Survey of Consumer Finances, Lee et al. (1997) find that households with low education are less likely than families with more education to save for the postsecondary education of their children. Another determinant of college-saving decisions is the family composition of the household. Bogan (2015) uses data from the 2007-2009 Survey of Consumer Finances and finds that the presence of an elderly dependent reduces a household’s propensity to save for the college education of children. In the same study, Bogan (2015) finds that the number of children does not have a statistically significant association with having a college savings account. However, Devaney and Chien (2002) report that families with more children are more likely to save for college. For married couples, research shows that they are more likely to save for children’s postsecondary education

The financial-risk-taking behavior of parents is another important factor to consider in examining the college-savings behavior of households. People’s attitudes toward risk reflect their risk preferences, and is known to influence financial behaviors of households (Gilliam et al. 2010; Wang & Hanna, 2007). Financial-risk-taking behavior also has been found to influence household net worth positively (Finke & Houston, 2003). The relationship between financial-risk taking and financial behaviors such as retirement savings, insurance ownership, portfolio formation, and stock investments have been examined extensively in the literature. However, the effect of parents’ financial-risk-taking behavior on college savings has hardly been examined in the literature. Several studies show that financial knowledge also influences individuals’ ability to make financial decisions (Lusardi & Mitchell, 2014; Agnew et al., 2013; Lusardi et al., 2010; and Lusardi, 2008). For instance, it has been established that financial literacy influences retirement planning (Agnew et al., 2013; Brown, 2013; Van Rooij et al., 2011a; Lusardi &Mitchell, 2011), stock market participation (Brown, 2013; Van Rooij et al., 2011b), credit card usage (Mottola, 2013 and Norvilitis et al., 2006), and borrowing behavior (Gathergood, 2012; Lusardi, 2008).

Volume 20 • Issue 1 2021

A study by Allgood and Walstad (2011) ascertain that having low subjective financial knowledge is associated with a low probability of paying credit-card debt fully. Alyousif and Kalenkoski (2017) find that individuals with low subjective financial knowledge are less likely to seek financial advice. Research shows that individuals who do not seek financial advice tend to suffer significant losses from low portfolio diversification (Gaudecker, 2015) and may tend to have low wealth. In their study, where subjective financial knowledge and objective financial knowledge are found to be highly correlated, Perry and Morris (2006) find that household saving, budget, and spending decisions are influenced positively by subjective financial knowledge. Although financial literacy and subjective financial knowledge are known to influence financial behavior, neither of these two factors has been examined in the literature regarding college savings. Studies as well show that household income can influence the college-savings behavior of parents (Lefebvre, 2004; Acemoglu & Pischke, 2001). High-income earners and the wealthy are more likely to save for college than low-income earners and the less affluent (Yilmazer, 2008). Lefebvre (2004) explains that income alone does not account for the disparity in the saving behavior of high- and low-income earners. Rather, factors such as academic performance, academic program awareness, and the age of the child are important in explaining the saving disparity between high-and low-income groups. The literature also documents the relationship between homeownership and college savings. Lefebvre (2004) finds that homeowner parents without a mortgage are more likely to save for college than parents residing in either rental housing units or houses with mortgages. The author further finds that parents who own their homes free and clear are likely to save substantially more in dollar terms for their offspring’s education than their counterparts who reside in rental or mortgaged housing units. Concerning health status, research shows that parents with poor health status are less likely to save for their kids’ college education than healthy parents (Bogan, 2015; Devaney & Chien, 2002). The relationship between the employment status of parents and college savings also has been examined to some extent in the literature. Devaney and Chien (2002) point out that having a full-time job increases the probability of having children’s college education as a goal for saving. However, the authors find that being self-employed is not a statistically significant predictor of having college education as a saving goal.

Student-loan debt of parents also is another factor that may influence college education savings. Using data from the 2012 National Longitudinal Survey of Youth (NLSY79), Martin et al. (2020) observe that parents who have student debt on their balance sheet are less likely to save for the education of their children. Steelman and Powell (1991) report that parents who received parental financial support for their college education are more likely to pay for the college education of their children. However, parents who did not receive parental financial support for their college education are less likely to save for the postsecondary education of children. Another factor that may influence college savings is the ownership of health insurance. Several studies, including Starr-McCluer (1996), Hsu (2013), Qiu (2016), and Bogan (2015), examine the association between health-insurance ownership and savings behavior. Qiu (2016) shows that health insurance relates positively to both the likelihood and the amount of stock ownership among households. Starr-McCluer (1996) also indicates that health insurance influences the savings behavior of U.S. households positively. Bogan (2015) includes health insurance as a control variable in the study involving family composition and savings for offspring education and finds a positive effect on college savings. In theory, the demand for college savings is subject to financial constraints. Having health insurance may cause a reduction of these constraints by reducing excessive out-of-pocket health costs. Financial fragility, which represents the vulnerability of a person to economic shocks, also may relate to college savings. However, the effect of financial fragility on the college-saving behavior of parents has not been examined in the literature on college savings specifically. Alyousif and Kalenkoski (2017) show that financially fragile individuals are less likely to seek savings or investment advice, but are more likely to seek help relating to debt counseling. Accordingly, one might suggest that financial fragility makes it difficult for individuals to incorporate long-term decisions such as college savings into their decision-making process. Using data from the NLSY79, Sorokina (2013) reveals that youths from liquidity-constrained households have a low probability of attending college. Lusardi et al. (2011) point out that financial fragility is prevalent among individuals with low levels of education and financial literacy. Overall, existing studies examine the effect of some household characteristics on college-education-savings behavior. Empirical studies that comprehensively assess household factors predicting parental willingness to save for college are sparse. The present study uses a nationally representative

Journal of Personal Finance

data set to examine broadly the factors that could influence parents’ college-education-saving behavior from demographic, preference, financial knowledge, and economic perspectives. In particular, the current paper sheds light on the effects of financial fragility, financial-planning horizon, financial-risktaking behavior, student-loan debt of parents, subjective financial knowledge, and financial literacy on parents’ saving decisions for college. The literature related to college savings has not examined these factors sufficiently.

that equals 1 if the race of the respondent is white and 0 if the respondent is nonwhite. The highest educational level of the respondent is represented by three dummy variables comprising some college, a college degree, and a postgraduate degree. The reference category is high school or less. Married is a dichotomous variable that equals 1 if the respondent is married and 0 otherwise. The number of financially dependent children is classified into three dummies, comprising one, two, and three financially dependent children. The reference category is having four or more financially dependent children.

DATA

The preference factors (financial-planning horizon and financial-risk taking) are measured as follows. For the planninghorizon variable, the survey question asks respondents to indicate the time period that they consider most important in planning their household’s saving and spending. The responses range from “the next few months” to “longer than 10 years.” Due to limited observations in the response categories, this study classifies the financial-planning horizon variable into two categories, comprising short-term and long-term planning horizons. The short-term planning-horizon variable includes the following responses: “next few months”, “the next year”, and “the next few years.” The long-term horizon variable includes “the next 5 to 10 years” and “longer than 10 years.” The planninghorizon variable is coded as 1 if the responded has a long-term horizon and 0 otherwise.

The current study uses the restricted version of the statelevel data set from the 2015 National Financial Capability Study (NFCS). The NFCS is funded by the Investor Education Foundation of the United States’ Financial Industry Regulatory Authority (FINRA). This survey started in 2009 and since has been conducted every three years. The 2015 survey collected data about the financial capability, socioeconomic characteristics, financial attitudes, and financial behaviors of 27,564 adults nationwide through the use of online questionnaires. Given that the present study focuses on parental savings for children’s college education, the analyses exclude individuals who have no children and those without financially dependent children. The number of observations for the empirical analyses is therefore 8,650 parents aged 18 and over. Survey weights are applied to make the data nationally representative. The dependent variable represents the saving decision of parents for the college education of their children. The survey question for constructing this variable is, “Are you setting aside any money for your children’s college education?” The value for the dependent variable equals 1 if a household answers “yes” and 0 if a household answers “no.” The explanatory variables include demographic, preference, financial knowledge, and economic variables. The demographic variables include age of the respondent, race of the respondent, educational attainment of the respondent, married status of the respondent, and number of children in the household. The life-cycle theory of savings suggest an inverted U-shape relationship between age and savings (Ando & Modigliani, 1963; Browning & Crossley, 2001). Accordingly, this study includes the age of the respondent as a quadratic continuous variable to capture the non-linear relationship that may exist between age and the decision to save for college over the life cycle. White is a dichotomous variable

The financial-risk-taking variable is constructed based on the responses to the question, “When thinking of your financial investments, how willing are you to take risks?” The responses range from 1 (“not at all willing”) to 10 (“very willing”). Due to the limited number of observations pertaining to some of the original response categories, the current study recodes the responses to the financial-risk-taking question to three categories. The first category measures low financial-risk taking and includes responses ranging from 1 to 3. The second category measures medium-risk taking and includes responses ranging from 4 to 7. Finally, individuals with responses ranging from 8 to 10 are classified as having high financial-risk-taking behavior. Thus, two dummies are used for the risk-taking variable, with the reference category being low-risk taking. The financial-knowledge variables also are constructed as follows. Concerning the financial-literacy variable, the 2015 NFCS uses six financial-literacy questions to assess respondents’ financial knowledge regarding compound interest or numeracy, inflation, bond pricing, loan pricing, mortgage payments, and risk diversification. In the current

Volume 20 • Issue 1 2021

study, financial literacy is measured first in terms of the overall score obtained by the respondents. Second, financial literacy is assessed based on its components. A correct response to each financial-literacy component is coded as 1 and an incorrect response is coded as 0. To assess subjective financial knowledge, the NFCS asks the question, “On a scale from 1 to 7, where 1 means very low and 7 means very high, how would you assess your overall financial knowledge?” Because there are limited number of observations in the first and second response categories, the present paper follows the approach used by Alyousif and Kalenkoski (2017) and recodes the responses into three categories. Individuals who provided responses ranging from 1 to 3 are classified as having low subjective financial knowledge. Those who answered 4 on the scale are classified as having a medium level of subjective financial knowledge. A high subjective financial knowledge is assigned when a respondent chooses any value that ranges from 5 to 7 on the measurement scale. Eventually, this study uses two dummies for the subjective-financialknowledge variable. The reference category is low subjective financial knowledge. Finally, the economic variables (household income, employment status, homeownership, student loan, health insurance ownership, and financial fragility) are constructed as follows. The NFCS provides household income as a categorical variable and therefore this study measures the income variable as three dummies, ranging from “less than $50,000” to “$100,000 to less than $150,000.” The reference category is “$150,000 or more.” Homeownership is a dichotomous variable taking a value of 1 if the respondent owns a house and 0 otherwise. The respondent’s employment status is assessed using seven dummies, comprising self-employed, employed full-time, employed part-time, full-time student, permanently sick or disabled, unemployed, and retired. The reference category is homemaker. The student-loan-debt variable is constructed as a dummy variable that takes a value of 1 if the respondent has a current student loan for either self or partner and 0 otherwise. Health-insurance ownership is a dichotomous variable that equals 1 if the respondent has health insurance and 0 otherwise. Similar to the approach used by Alyousif and Kalenkoski (2017), the present study measures financial fragility as a sum of the responses to five questions in the 2015 NFCS. These five questions pertain to whether the respondent has too much debt, difficulty in covering expenses, a three-month emergency fund, excess spending over income, and the ability

to raise an amount of $2,000 if it becomes necessary. This study measures the five components constituting financial fragility as follows. A value of 1 is assigned if a respondent agrees or strongly agrees to have too much debt and 0 otherwise. Respondents who say they find it very difficult or somewhat difficult to pay all their expenses are assigned a value of 1 and 0 otherwise. Not having a 3-month emergency fund takes a value of 1 and 0 otherwise. Respondents who say they are unable or probably unable to raise $2,000 when an unexpected circumstance arises are coded as 1 and 0 otherwise. Finally, respondents whose spending exceeds income are coded as 1 and 0 otherwise. In summary, the financial fragility variable is represented by five dummies ranging from one to five types of fragility. The reference category is “no financial fragility.” Table 1 presents the descriptive statistics for the dependent and explanatory variables, including the t-test results for households with college savings versus those without college savings. Approximately 45 percent of the sampled households have set aside some money for the college education of their children. The average age for the overall sample is 41. Among those with college savings, the mean age is 39, while the mean age among respondents without college savings is 43. About 71% of households are white. The percent of households with college savings who are white is 68%, but that of those without college savings is 73%. Among all households, there are 25% with high school or less education, 42% with some college education, 20% with a college degree, and 13% with a postgraduate degree. Among households without college savings, about 31% have a high school diploma or less, 46% have some college education, 15% have a college degree, and 8% have a postgraduate degree. With respect to households with college savings, about 17% have a high school diploma or less, 37% have some college education, 28% have a college degree, and 18% have a postgraduate degree. Table 1 also shows that 72% of all the sampled households are married. The percent of households with college savings who are married is 78%, exceeding the percent of married households without college savings by 10%. More than two thirds of all households have a combination of one and two financially dependent children. Approximately 27% of all households, 31% of households with college savings, and 24% of households without college savings have a long-term financial-planning horizon. The percent of all households with low, medium, and high financial-risk taking are 23%, 47%, and 30% respectively. Among households with college savings, 12%, 46%, and 42% have low, medium, and high financial-risk taking respectively.

Journal of Personal Finance

Table 1 further shows that about 82% of all households have high subjective financial knowledge. The percent of households with high subjective financial knowledge increases to nearly 90% among households with college savings. The overall financial literacy score is 3.21, 3.33, and 3.11 among all households, households with college savings, and households without college savings respectively. Most households answer the compound interest, mortgage payment and inflation financial-literacy questions correctly. However, the majority of households answers the literacy questions pertaining to bond and loan pricing wrongly. More than two thirds of the sampled households have household annual income below $100,000. Over two thirds of the sampled households also are homeowners. Among all households, about one third have student-loan debt, 90% own health insurance, 16% are homemakers, 54% have full-time employment, 8% are selfemployed, 5% are retired, and 3% are permanently sick or disabled. Only 20% of all households have no financial fragility, about 18% households report having one, two, or four type(s) of financial fragility, 20% have three types of financial fragility, and 7% have five types of financial fragility.

MODEL The current study estimates a probit model to examine the saving decisions of parents for college. The probit model is appropriate because the dependent variable takes values of 1 and 0. It is stated as follows.

Where, Savei* is the latent variable indicating the net benefits parents receive from saving for the college education of their children. The subscript i refers to the parent respondent. Savei is the observed variable that takes a value of 1 if a parent has college savings and 0 otherwise. The intercept is β0 and the slope parameters are βj, where j=1,2,3,…,15 for each explanatory variable. The error term is μi and it is assumed to follow the standard normal distribution. The term Xi is a matrix of the explanatory variables relating to the parent. These explanatory variables include age, white, highest educational level attained, married, number of financially dependent

children, financial-planning horizon, financial-risk taking, subjective financial knowledge, and financial literacy. The rest include the household’s annual income, employment status, homeownership, presence of student-loan debt, healthinsurance ownership, and financial fragility. The current study tests several hypotheses. The effect of being married is expected to be positive. Becker’s (1973) economic theory of marriage suggests that individuals marry because they believe that marriage will make them better off than being single. The output from being married is therefore expected to be greater than the output from being single owing to factors such as economies of scale, risk sharing, and opportunities for specialization. Hence, being married is expected to increase the probability that a household will have college savings. The level of educational attainment of parents is expected to have a positive relationship with college savings. The more education that individuals have, the more they are able to make better financial decisions. Highly educated individuals may desire to have highly educated offspring compared to less-educated ones (Becker, 1994) because they understand the net benefits of education. As the level of parents’ education increases, the likelihood that they would invest in the education of their children may increase. The effect of age is expected to be ambiguous as individuals go through the different phases of the life cycle. Young adults have more liquidity constraints and less wealth than older adults. Parents reaching the decumulation phase of the life cycle may focus more on other financial goals and are less likely to save for the college education of their dependent children. White, as a proxy for preferences, is expected to have a negative relationship with college savings. As discussed in the literature review, white parents have greater expectations than nonwhites that their children would absorb part of the cost of college education. Therefore, whites are expected to save less for the postsecondary education of their children than nonwhites. Financial-planning horizon is expected to be positive association with college savings. As the financial-planning horizon increases, the more individuals are willing to substitute current consumption for future consumption. Thus, individuals with long-term financial-planning horizons may have the desire to save for college to smooth the consumption of the household over the life cycle.

Volume 20 • Issue 1 2021

Financial literacy is expected to have a positive association with college savings. Financially literate individuals possess the ability to make better informed financial decisions than the less financially literate. Subjective financial knowledge is hypothesized to have a negative association with college savings. People with high levels of subjective financial knowledge may be overconfident about their financial skills and are therefore susceptible to making financial mistakes. Financial-risk taking is hypothesized to have a positive relationship with college savings. As financial-risk taking increases, household wealth (the combination of human capital and net worth) increases as well, increasing the likelihood that such households would save for college. The effect of the number of financially dependent children is hypothesized to be negative. The human capital theory predicts that household’s spending per child decreases with number of children (Becker, 1994). As the number of children increases, the household’s financial resources are constrained. Thus, the probability that a household will save for children’s college education is expected to decrease as the quantity of children increases. The effect of the household’s annual income is expected to be positive. Higher income increases the financial resources available for parents to save. As household income increases, economic theory suggests that the demand for normal goods, including child services, will increase. According to Becker’s (1960) quantity-quality model of fertility, a rise in income will likely increase child quality. Thus, parents will tend to increase the human capital investments they make in their children as income increases. The effect of a parent’s employment status is expected to be ambiguous. Employed individuals have more financial resources than the unemployed. Self-employed individuals could have more cash-flow constraints, especially during the early stages of the business, than full-time employees. Homeownership is expected to have a positive association with college savings. Owning a home signals wealth endowment. Cash flow constraints are less, particularly when the household owns the home outright. A parental student loan is expected to have a negative association with college savings. Servicing the principal and interest components of the student-loan debt constraints the spending plans of the household.

Health-insurance ownership is hypothesized to have a positive relationship with college savings. Owning health insurance could free-up financial resources that would have been used for excessive out-of-pocket costs, enabling the household to take advantage of investment opportunities such as saving for college. The effect of financial fragility is expected to be negative. High levels of financial fragility signal financial resource constraints and a decline in the ability to invest in the human capital of children.

RESULTS The marginal effects and standard errors for the probit model are shown in Table 2. Column “A” of Table 2 shows the main results for the current study. Column “B” shows the results for the sensitivity analyses using the components of financial literacy. The analyses that follow are based on the main results contained in column “A”. The last two paragraphs in this section present the sensitivity analyses. Table 2 shows that the relationship between married and college savings is positive and statistically significant. Consistent with the research hypothesis and the findings of Xiao and Noring (1994), the results suggest that married households have a greater probability of saving for college compared to non-married households. Specifically, this study shows that being married increases the probability that a household will save for college by 0.03. The relationship between married and college savings is not surprising because the output from being married is expected to be greater than the output from being non-married. The results for parental education show that parents with some college education are more likely to save for college compared to those with less than a high school education. The results also show that having a college degree increases the probability that a household wills save for college by 0.14 compared to households with less than a high school education. Similarly, the study finds that households with a postgraduate degree have a greater probability of saving for college compared to households with less than a high school education. These results are consistent with the hypothesized relationship between parental education and college savings. The results also are consistent with the findings by Lefebvre (2004) who ascertains that, parents with less education are less likely to save for the college education of their children compared with parents who have more education. The findings suggest that

Journal of Personal Finance

highly educated individuals prefer highly educated offspring to less educated ones because they understand the net benefits of education. In Table 2, the results for the age of the parent show that age and college savings are related inversely, suggesting that an increase in age decreases the probability that a household will save for college. Although the magnitude of the relationship between age and college savings is small, the results suggest that young households may have college savings as a savings goal. The current result disagrees with that of Yilmazer (2008) who observes that the amount that a household saves for the postgraduate education of offspring is likely to increase as the age of the household head increases. The differences in the findings of the current study and that of Yilmazer (2008) may be attributed to the fact the later focuses on the amount of college savings, while the current study focuses on the incidence of college savings. Consistent with the hypothesis for white race, the relationship between white and college savings is negative and statistically significant. The results suggest that being white decreases the probability of saving for the college education of children by 0.10 compared with being nonwhite. This result is consistent with the findings by Steelman and Powell (1993) that white parents, unlike nonwhites, expect their children to absorb part of the cost of funding college education. Theoretically, the effect of incorporating household-spending decisions into saving decisions for college may result in white parents being less likely to save for the college education of their children than nonwhite parents. Regarding the number of financially dependent children, Table 2 shows statistically significant negative results for parents with four or more children. The results imply that parents with four or more children are less likely to save for the college education of their children compared with parents who have one child. This result supports human capital theory, but is at variance with the results by Devaney and Chien (2002) who ascertain that families with more children are more likely to save for the college education of their children. In the current study, the results indicate that cash-flow constraints could be a hindrance to the ability of parents with more children to save for the college education of their children. With more children, parents spend less per child as they apportion their total consumption on education for the number of children (Becker, 1994).

The financial-risk variables show statistically significant relationships with college savings. Compared with parents with a low financial risk, the results show that parents with medium financial risk are more likely to save for college. The results also show that having high financial risk increases the probability that a household will save for college by 0.20 compared to having low financial risk. This effect is not only statistically significant, but also practically important. Financial-risk taking is known to be associated with household wealth, and this could explain the positive association between risk taking and the decision to save for the college education of children. The results for household income show that having a household income in the $100,000 to $150,000 range increases the probability that a household will save for college compared to having a household income below $50,000. Similarly, the probability that a household with $150,000 or more annual income will save for college is 0.13 higher compared to a household with annual household income under $50,000. This result is similar to that of Yilmazer (2008) who finds that high-income earners are more likely to save for the college education of children than low-income earners. The results show increase in child quality as income increases because the probability of investing in the human capital development of children increases when household income is at least $100,000. Homeownership, an indicator of wealth endowment and socioeconomic status in society, is statistically significant and is associated with a 0.10 higher probability of saving for college. This result is consistent with the initial hypothesis for homeownership. The result also moves in tandem with the findings by Lefebvre (2004) who show that homeownership positively influences college savings. All else equal, cash flow constraints are less for households who own homes and therefore they tend to have a higher probability of saving for college than families who do not own homes. As earlier hypothesized, the results for the current study show that health-insurance ownership is associated positively with college savings. Specifically, having health insurance increases the probability that a household will save for college by 0.10 compared to not having health insurance. Owning health insurance has been found to influence household saving behavior positively (Starr-McCluer, 1996; Bogan, 2015) and the findings from the current study do not suggest otherwise. Intuitively, the presence of health insurance may limit the out-of-pocket health costs a household may incur, making it possible for the household to have additional financial resources for college savings.

Volume 20 • Issue 1 2021

The findings in Table 2 also show that the employment status of the household may be important in explaining the collegesaving decision of the household. The results show that being self-employed increases the probability of saving for college by 0.06 compared to a homemaker. The study also finds similar results for a household where the respondent has a full-time employment status. The results suggest that the opportunities for saving for college may be greater for full-time and selfemployed households than households where the respondent is a homemaker. Consistent with the hypothesis for this study, financial fragility is associated with a lower probability of saving for the college education of children. The results show that having one type of financial fragility decreases the probability of saving for college by 0.09 relative to having no financial fragility. Relative to having no financial fragility, the results also indicate that having two or three types of financial fragility decreases the probability that a household will have college savings by 0.13 and 0.22 respectively. Compared to having no financial fragility, Table 3.2 further shows that having four or five types of financial fragility decreases the probability that a household will save for college by 0.30 and 0.32 respectively. Relative to their respective means, these statistically significant results also are substantively meaningful. Financial fragility makes it difficult for individuals to incorporate long-term decisions into their decision-making process. The findings from this study suggest that high levels of financial fragility signal financialresource constraints and a decline in the ability to invest in the human capital of children. The results for subjective financial knowledge show that parents with high-subjective financial knowledge have a higher probability of saving for college than those with low-subjective financial knowledge. The results disagree with the expectation that people with high levels of subjective financial knowledge are likely to be overconfident about their financial skills and are therefore susceptible to making financial mistakes. The results, however, are similar to the findings by Perry and Morris (2006) that, high-subjective financial knowledge influences financial decisions positively. The findings also show that the overall financial literacy score of parents is statistically significant, but related negatively to saving for the college education of children. Although the marginal effect of the total financial literacy score is not substantively meaningful, the result contradicts the initial hypothesis. Financial literacy is known to be associated

positively with financial behaviors such as retirement planning, stock market participation, credit card usage, and borrowing behavior (Agnew et al., 2013; Brown, 2013; Mottola, 2013; Gathergood, 2012; Van Rooij et al., 2011a; Van Rooij et al., 2011b; Lusardi &Mitchell, 2011; Lusardi, 2008; and Norvilitis et al., 2006). It could therefore be the case that individuals who are more financially literate would prioritize saving for retirement over children’s education. One also may attribute the observed relationship between financial literacy score and college savings to social desirability bias, a phenomenon that may arise from the tendency for survey respondents to provide responses that make them look good in the presence of other people (Fisher, 1993). However, the results for a restricted model (not shown) involving overall financial literacy score and college savings show a statistically significant positive relationship. The positive relationship changes to negative when the other variables used in this study are added to the model. Theoretically, investments in human capital are different from capital market investments. Individuals with high financial literacy may tend to be more knowledgeable about alternative sources of financing the college education of their children than those with low levels of financial literacy. Therefore, less financially literate parents with the goal of maximizing their utility from investing in children would be more likely to save for the college education of their children. In Column “B” of Table 2, the results from the sensitivity analyses show that the compound interest and inflation components of financial literacy have a statistically significant negative relationship with parental saving for college. The mortgage-payment and risk-diversification components also have a statistically significant negative association with college savings. The bond- and loan-pricing literacy components (cumulatively referred to as asset-pricing literacy) are the only components that have positive association with college savings. Interestingly, the asset-pricing literacy components have the lowest mean scores. Although these components of financial literacy have statistically significant effects, the magnitude of their association with college savings is not large.

CONCLUSION Saving for the college education of offspring remains one of the challenges confronting American households. The present study has examined the factors associated with the

Journal of Personal Finance

decision of parents to save for the college education of their children. The study utilizes a nationally representative data set from the restricted version of the 2015 National Financial Capability Study and performs sensitivity analyses in addition to estimating the main probit model. The current study finds saving for college is associated negatively with the number of dependent children and financial fragility, indicating that increased household spending constraints decrease the utility derived from investing in children. The results for age show that an increase in age has a negative effect, suggesting that college savings may not be a prority for older parents. Being white is negatively associated with the probability of saving for college, suggesting that the wealth disparity between whites and nonwhites is less likely to persist. Health-insurance ownership has a positive effect, signifying that health insurance reduces the likelihood for households to overspend on health care thereby enabling households to save for college. Being married increases the probability of saving for college, suggesting that married households have greater output than non-married households. Parents who own homes, have high incomes, and possess high education have a greater probability of saving for college than parents who do not own a home, have low income, and have low education respectively. The positive relationship between income and college savings suggests that children are normal goods (in economic terms). The results for parental education indicate that parents with high education prefer high-quality to low-quality children. The results for homeownership imply that a wealth endowment and socioeconomic status influence college savings positively.

Overall, to increase the likelihood of more parents saving for college education, policies that aim to encourage parents to save for the college education of their children may target those with low income, low education, four or more children versus one child, no health insurance, and no homeownership. The other households to target are older adults, non-married households and the financially fragile. Financial practitioners interested in helping more households to save for college could develop programs and strategies that would be attractive to those who are less likely to have college savings. One such approach could be offering parents financial education that focuses on helping them appreciate the benefits of saving for college and the costs of doing otherwise. For financially constrained households, financial practitioners could provide pro bono financial education to them as part of their philanthropic activities to their communities. The results also underscore the need for financial planners to examine the financial-risk behavior and the subjective financial knowledge of parents given that these two factors also are associated with higher probability of saving for college. Financial practitioners could encourage parents with medium and high financial-risk preferences as well as those with high subjective financial knowledge to consider adding college savings to their financial goals. One of the limitations of this study is that the NFCS does not contain the amount of college savings for analysis. Another limitation is that the study does not include information on the children because such data also is not available in the NFCS. An example of such omitted variable is the ability of children. Nonetheless, the study sheds light on several factors that are important for policymakers, practitioners, and academics.

The study finds that the overall financial literacy score and the decision to save for college are negatively related. The sensitivity analyses suggest the compound interest, inflation, mortgage payment, and risk diversification components of financial literacy could drive the inverse relationship between the overall literacy score and college savings. Bond- and loan-pricing literacy components are the only components of financial literacy that are associated with higher probability of college savings.

Volume 20 • Issue 1 2021

TABLES Table 1 Summary Statistics Has College Savings Overall

Yes

Dependent variable College savings (1=Yes)

0.4442 (0.0054)

Explanatory variables Demographics Age

41.0928 (0.1309)

39.0916 (0.2043)

42.6921 (0.2142)

***

White

0.7070 (0.0050)

0.6771 (0.0093)

0.7310 (0.0079)

***

High school or less

0.2490 (0.0047)

0.1711 (0.0077)

0.3112 (0.0081)

***

Some college

0.4218 (0.0054)

0.3708 (0.0094)

0.4626 (0.0085)

***

Parental education level

College

0.2035 (0.0044)

0.2771 (0.0079)

0.1446 (0.0052)

***

Postgraduate

0.1257 (0.0036)

0.1809 (0.0066)

0.0817 (0.0040)

***

0.7232 (0.0049)

0.7833 (0.0080)

0.6752 (0.0080)

***

One

0.4302 (0.0054)

0.4084 (0.0093)

0.4476 (0.0084)

***

Two

0.3552 (0.0052)

0.3899 (0.0092)

0.3275 (0.0079)

***

Three

0.1419 (0.0038)

0.1379 (0.0065)

0.1451 (0.0059)

Four or more

0.0726 (0.0028)

0.0637 (0.0046)

0.0797 (0.0046)

Long-term planning horizon

0.2688 (0.0048)

0.3108 (0.0087)

0.2352 (0.0071)

***

Short-term planning horizon

0.7312 (0.0048)

0.6892 (0.0087)

0.7648 (0.0071)

***

Low

0.2312 (0.0046)

0.1147 (0.0059)

0.3243 (0.0080)

***

Medium

0.4735 (0.0054)

0.4643 (0.0094)

0.4809 (0.0084)

High

0.2953 (0.0050)

0.4210 (0.0094)

0.1948 (0.0067)

***

Low

0.0586 (0.0026)

0.0267 (0.0033)

0.0842 (0.0047)

***

Medium

0.1237 (0.0036)

0.0762 (0.0049)

0.1617 (0.0063)

***

High

0.8177 (0.0042)

0.8971 (0.0058)

0.7541 (0.0073)

***

Overall financial literacy score

3.2080 (0.0173)

3.3263 (0.0308)

3.1136 (0.0263)

***

Compound interest

0.7612 (0.0046)

0.7516 (0.0083)

0.7690 (0.0072)

Inflation

0.5578 (0.0054)

0.5397 (0.0094)

0.5722 (0.0084)

Bond pricing

0.2819 (0.0049)

0.3383 (0.0089)

0.2368 (0.0070)

***

Married Financially dependent children

Preference factors Financial-planning horizon

Financial-risk taking

Financial knowledge Subjective financial knowledge

Financial literacy

Journal of Personal Finance

Has College Savings Loan pricing

Overall

Yes

0.3621 (0.0052)

0.4079 (0.0092)

No 0.3254

(0.0078)

***

Mortgage payment

0.7891 (0.0044)

0.7936 (0.0078)

0.7855 (0.0073)

Risk diversification

0.4560 (0.0054)

0.4952

(0.0094)

0.4247 (0.0083)

***

Less than $50,000

0.3798 (0.0053)

0.2556 (0.0086)

0.4791 (0.0085)

***

$50,000 to less than $100,000

0.3875 (0.0053)

0.4093 (0.0092)

0.3700 (0.0081)

***

$100,000 to less than $150,000

0.1604 (0.0040)

0.2210 (0.0077)

0.1120 (0.0052)

***

$150,000 or more

0.0723 (0.0028)

0.1141 (0.0057)

0.0389 (0.0031)

***

Economic factors Household annual income

Homeownership

0.6766 (0.0051)

0.7908 (0.0079)

0.5853 (0.0084)

***

Student-loan debt

0.3101 (0.0050)

0.3320 (0.0088)

0.2926 (0.0075)

***

Health-insurance ownership

0.8972 (0.0033)

0.9369 (0.0049)

0.8655 (0.0058)

***

Self-employed

0.0753 (0.0029)

0.0836 (0.0053)

0.0687 (0.0044)

Full-time

0.5439 (0.0054)

0.6400 (0.0091)

0.4671 (0.0084)

***

Part-time

0.0831 (0.0030)

0.0781 (0.0050)

0.0871 (0.0046)

Homemaker

0.1559 (0.0040)

0.1184 (0.0060)

0.1858 (0.0065)

Employment status

***

Student

0.0238 (0.0017)

0.0213 (0.0028)

0.0259 (0.0026)

Permanently sick or disabled

0.0265 (0.0018)

0.0099 (0.0021)

0.0398 (0.0033)

***

Unemployed

0.0365 (0.0020)

0.0203 (0.0029)

0.0494 (0.0039)

***

Retired

0.0549 (0.0025)

0.0282 (0.0031)

0.0762 (0.0046)

***

None

0.1982 (0.0043)

0.2996 (0.0085)

0.1172 (0.0053)

***

One

0.1767 (0.0042)

0.2299 (0.0081)

0.1342 (0.0057)

***

Two

0.1785 (0.0042)

0.2023 (0.0075)

0.1595 (0.0062)

***

Financial fragility

Three

0.1962 (0.0043)

0.1588 (0.0069)

0.2261 (0.0071)

***

Four

0.1813 (0.0042)

0.0854 (0.0053)

0.2579 (0.0075)

***

Five

0.0691 (0.0028)

0.0240 (0.0026)

0.1051 (0.0051)

***

Notes: Author’s analysis using restricted data set from the 2015 FINRA Foundation NFCS. Mean values are shown alongside the standard errors in parenthesis. Survey weights are applied. *** indicates significance at the 1% level; ** indicates significance at the 5% level. N=8,650

Volume 20 • Issue 1 2021

Table 2 Probit Model of Parents’ College Saving Decisions A: Main Model

B: Sensitivity Model

Marginal Effects (Standard Error)

Age (quadratic)

-0.0075*** (0.0005)

-0.0072*** (0.0005)

White

-0.0525*** (0.0107)

-0.0491*** (0.0106)

Demographic factors

Married (versus not married)

0.0247**

(0.0119)

0.0249**

(0.0119)

Parental education level (versus High school or less) Some college

0.0406*** (0.0123)

0.0449*** (0.0122)

College

0.1410*** (0.0155)

0.1425*** (0.0155)

Postgraduate

0.1339*** (0.0182)

0.1335*** (0.0182)

Financially dependent children (versus one) Two

0.0135

(0.0108)

0.0160

(0.0108)

Three

-0.0078

(0.0145)

-0.0069

(0.0144)

Four or more

-0.0335*

(0.0191)

-0.0327*

(0.0191)

0.0081

(0.0112)

0.0108

(0.0112)

Preference factors Long-term financial planning horizon (versus short-term) Financial-risk taking (Versus low) Medium

0.0867*** (0.0124)

0.0836*** (0.0124)

High

0.1972*** (0.0151)

0.1834*** (0.0153)

Medium

0.0123

0.0167

High

0.0740*** (0.0229)

Financial knowledge Subjective financial knowledge (versus low) (0.0257)

(0.0256)

0.0734*** (0.0229)

Financial literacy Overall financial literacy score

-0.0152*** (0.0033)

Compound interest

-0.0419*** (0.0120)

Inflation

-0.0511*** (0.0110)

Bond pricing

0.0350*** (0.0110)

Loan pricing

0.0173*

(0.0100)

Mortgage payment

-0.0223**

(0.0166)

Risk diversification

-0.0181*

(0.0105)

-0.0129

(0.0123)

Economic factors Household annual income (versus Less than $50,000) $50,000 to less than $100,000

-0.0103

(0.0123)

Journal of Personal Finance

A: Main Model

B: Sensitivity Model

Marginal Effects (Standard Error)

$100,000 to less than $150,000

0.0728*** (0.0168)

0.0776*** (0.0168)

$150,000 or more

0.1326*** (0.0220)

0.1290*** (0.0220)

Homeownership

0.1029*** (0.0119)

0.0997*** (0.0119)

Student-loan debt

0.0150

0.0130

Health insurance ownership

0.0882*** (0.0161)

0.0896*** (0.0162)

Self-employed

0.0630*** (0.0212)

0.0674*** (0.0211)

Full-time

0.0455*** (0.0145)

0.0471*** (0.0145)

Part-time

0.0301

0.0289

(0.0112)

Employment status (versus homemaker)

Student

(0.0204)

0.0084

(0.0331)

0.0045

(0.0331)

-0.0299

(0.0359)

-0.0262

(0.0358)

Unemployed

0.0034

(0.0291)

0.0017

(0.0291)

Retired

0.0018

(0.0290)

0.0038

(0.0289)

Permanently sick or disabled

Financial fragility (versus none) One

-0.0863*** (0.0141)

-0.0858*** (0.0141)

Two

-0.1272*** (0.0142)

-0.1283*** (0.0142)

Three

-0.2150*** (0.0139)

-0.2164*** (0.0138)

Four

-0.2961*** (0.0143)

-0.2941*** (0.0143)

Five

-0.3216*** (0.0156)

-0.3200*** (0.0157)

8,650

Number of Observations

Volume 20 • Issue 1 2021

REFERENCES Acemoglu, D., & Pischke, J. S. (2001). Changes in the wage structure, family income, and children's education. European Economic Review, 45(4-6), 890-904. Agnew, J. R., Bateman, H., & Thorp, S. (2013). Financial literacy and retirement planning in Australia. Numeracy: Advancing Education in Quantitative Literacy, 6(2). Allgood, S., & Walstad, W. (2011). The effects of perceived and actual financial knowledge on credit card behavior. IDEAS Working Paper Series from RePEc. Alyousif, M. H., & Kalenkoski, C. M. (2017). Who seeks financial advice? Financial Services Review, 26(4). Ando, A., & Modigliani, F. (1963). The “life cycle" hypothesis of saving: Aggregate implications and tests. The American Economic Review, 53(1), 55-84. Ashraf, N., Karlan, D., & Yin, W. (2006). Tying Odysseus to the mast: Evidence from a commitment savings product in the Philippines. The Quarterly Journal of Economics, 121(2), 635-672. Becker, G. S. (1994). Human capital revisited. In Human capital: A theoretical and empirical analysis with special reference to education (3rd Edition) (pp. 15-28). The University of Chicago Press. Becker, G. S. (1974). A theory of social interactions. Journal of Political Economy, 82(6), 1063-1093. Becker, G. S. (1973). A theory of marriage: Part I. Journal of Political economy, 81(4), 813-846. Becker, G. S. (1960). An economic analysis of fertility. In demographic and economic change in developed countries (pp. 209-240). Columbia University Press. Bernheim, B. D., Skinner, J., & Weinberg, S. (2001). What accounts for the variation in retirement wealth among US households? American Economic Review, 91(4), 832-857. Bogan, V. L. (2015). Household asset allocation, offspring education, and the sandwich generation. American Economic Review, 105(5), 611-615. Bricker, J., Dettling, L. J., Henriques, A., Hsu, J. W., Jacobs, L., Moore, K. B., ... & Windle, R. A. (2017). Changes in US family finances from 2013 to 2016: Evidence from the Survey of Consumer Finances. Federal Reserve Bulletin, 103(1). Brown, M., & Graf, R. (2013). Financial literacy and retirement planning in Switzerland. Numeracy, 6(2), 6. Browning, M., & Crossley, T. F. (2001). The life-cycle model of consumption and saving. Journal of Economic Perspectives, 15(3), 3-22. Devaney, S. A., & Yi-Wen, C. (2002). Children's education as the most important savings goal. Journal of Family and Consumer Sciences, 94(1), 64. Dondero, M., & Humphries, M. (2016). Planning for the American dream: The college-savings behavior of Asian and Latino foreignborn parents in the United States.” Population Research and Policy Review, 35(6), 791-823. Elliott, W., Lewis, M., Grinstein-Weiss, M., & Nam, I. (2014). Student loan debt: Can parental college savings help? Federal Reserve Bank of St. Louis Review, 96(4), 331-357. Federal Reserve Bank of New York (2018). Household debt and credit report (Q2 2018), Center for Microeconomic Data. Retrieved from https://www.newyorkfed.org/microeconomics/hhdc.html. November 13, 2018. Finke, M. S. (2006). Time orientation and economic decision making. In understanding behavior in the context of time (pp. 120-134). Psychology Press. Finke, M. S., & Huston, S. J. (2003). The brighter side of financial risk: Financial risk tolerance and wealth. Journal of Family and Economic Issues, 24(3), 233-256.

Journal of Personal Finance

Fisher, R. J. (1993). Social desirability bias and the validity of indirect questioning. Journal of Consumer Research, 20(2), 303-315. Frederick, S., Loewenstein, G., & O'donoghue, T. (2002). Time discounting and time preference: A critical review. Journal of Economic Literature, 40(2), 351-401. Gathergood, J. (2012). Self-control, financial literacy, and consumer over-indebtedness. Journal of Economic Psychology, 33(3), 590-602. Gaudecker, H. (2015). How does household portfolio diversification vary with financial literacy and financial advice? Journal of Finance, 70(2), 489-507. Gilliam, J., Chatterjee, S., & Grable, J. (2010). Measuring the perception of financial risk tolerance: A tale of two measures. Journal of Financial Counseling and Planning, 21(2), 30-43,82-83. Gutter, M. S., Hayhoe, C. R., DeVaney, S. A., Kim, J., Bowen, C. F., Cheang, M., ... & Mauldin, T. (2012). Exploring the relationship of economic, sociological, and psychological factors to the savings behavior of low‐to moderate‐income households. Family and Consumer Sciences Research Journal, 41(1), 86-101. Grissett, B., & Furr, L. A. (1994). Effects of parental divorce on children's financial support for college students. Journal of Divorce & Remarriage, 22(1-2), 155-166. Hong, E. O., & Hanna, S. D. (2014). Financial planning horizon: A measure of time preference or a situational factor? Journal of Financial Counseling and Planning, 25(2), 184. Hossler, D., & Vesper, N. (1993). An exploratory study of the factors associated with parental saving for postsecondary education. The Journal of Higher Education, 64(2), 140-165. Hsu, M. (2013). Health insurance and precautionary saving: A structural analysis. Review of Economic Dynamics, 16(3), 511-526. Jalbert, T., Stewart, J. D., & Johnson, G. (2010). The college or retirement decision. Journal of Personal Finance, 9. Lee, S., Hanna, S., & Siregar, M. (1997). Children's college as a saving goal. Journal of Financial Counseling and Planning, 8(1), 33. Lefebvre, S. (2004). Saving for postsecondary education. Perspectives on Labour and Income 5(7), 5-12. Lemieux, T. (2006). Postsecondary education and increasing wage inequality. American Economic Review, 96(2), 195-199. Lusardi, A., & Mitchell, O. S. (2014). The economic importance of financial literacy: Theory and evidence. Journal of Economic Literature, 52(1), 5-44. Lusardi, A. (2011). Americans' financial capability (No. w17103). National Bureau of Economic Research. Lusardi, A., & Mitchell, O. S. (2011). Financial literacy and retirement planning in the United States. Journal of Pension Economics & Finance, 10(4), 509-525. Lusardi, A., Schneider, D. J., & Tufano, P. (2011). Financially fragile households: Evidence and implications (No. 17072). National Bureau of Economic Research, Inc. Lusardi, A., Mitchell, O. S., & Curto, V. (2010). Financial literacy among the young. Journal of Consumer Affairs, 44(2), 358-380. Lusardi, A. (2008). Financial literacy: An essential tool for informed consumer choice? (No. w14084). National Bureau of Economic Research. Manly, C. A., Wells, R. S., & Bettencourt, G. M. (2017). Financial planning for college: Parental preparation and capital conversion. Journal of Family and Economic Issues, 38(3), 421-438. Martin Jr, T. K., Augustin, L. A. V, Ricaldi, L. C. & Nunez, J. (2020). The effect of student Loans on parental views of education financing. Journal of Financial Planning, 33(6), 46-57. Mottola, G. R. (2013). In our best interest: Women, financial literacy, and credit card behavior. Numeracy, 6(2), 4. Nam, J., & Ansong, D. (2015). The effects of a dedicated education savings account on children's college graduation. Economics of Education Review, 48, 198-207.

Volume 20 • Issue 1 2021

Norvilitis, J. M., Merwin, M. M., Osberg, T. M., Roehling, P. V., Young, P., & Kamas, M. M. (2006). Personality Factors, Money Attitudes, Financial Knowledge, and Credit‐Card Debt in College Students 1. Journal of Applied Social Psychology, 36(6), 1395-1413. Perry, V. G., & Morris, M. D. (2005). Who is in control? The role of self‐perception, knowledge, and income in explaining consumer financial behavior. Journal of Consumer Affairs, 39(2), 299-313. Qiu, J. (2016). Precautionary saving and health insurance: A portfolio choice perspective. Frontiers of Economics in China, 11(2), 232. Schell-Olsen, S. (2018). Opinion: Why is college so expensive? The News Record, University of Cincinnati. Retrieved from http://www.newsrecord.org/opinion/opinion-why-is-college-so-expensive/article_d7d84102-e438-11e8-9d64-fb2ed548bca4.html. November 13, 2018. Song, & Elliott. (2012). The effects of parents' college savings on college expectations and Hispanic youth's four-year college attendance. Children and Youth Services Review, 34(9), 1845-1852. Sorokina, O. V. (2013). Parental credit constraints and children’s college education. Journal of Family and Economic Issues, 34(2), 157-171. Starr-McCluer, M. (1996). Health insurance and precautionary savings. The American Economic Review, 86(1), 285-295. Steelman, L. C., & Powell, B. (1993). Doing the right thing: Race and parental locus of responsibility for funding college. Sociology of Education, 223-244. Steelman, L. C., & Powell, B. (1991). Sponsoring the next generation: Parental willingness to pay for higher education. American Journal of Sociology, 96(6), 1505-1529. U.S. Department of Education (2017). U.S. Department of Education releases national student loan fy 2014 cohort default rate. Retrieved from https://www.ed.gov/news/press-releases/us-department-education-releases-national-student-loan-fy-2014cohort-default-rate. December 2, 2018. Van Rooij, M. C., Lusardi, A., & Alessie, R. J. (2011a). Financial literacy and retirement planning in the Netherlands.” Journal of Economic Psychology, 32(4), 593-608. Van Rooij, M., Lusardi, A., & Alessie, R. J. (2011b). Financial literacy and stock market participation. Journal of Financial Economics, 101(2), 449-472. Wang, C., & Hanna, S. D. (2007). The risk tolerance and stock ownership of business owning households. Financial Counseling and Planning. 18(2), 3–18. Xiao, J. J., & Noring, F. E. (1994). Perceived saving motives and hierarchical financial needs. Financial Counseling and Planning, 5(1), 25-44. Yilmazer, T. (2008). Saving for children’s college education: An empirical analysis of the trade-off between the quality and quantity of children. Journal of Family and Economic Issues, 29(2), 307-324.

Journal of Personal Finance

Student Loan Debt Letters: How Colleges Communicate Debt with Students

Zachary Taylor, Ph.D. Gretchen Holthaus Karla Weber

Abstract As the student loan debt crisis has continued to gain national attention from higher education leaders, education policymakers, and the media, states have begun mandating that institutions send student loan debt letters to any current or former student with outstanding student loan debt. Preliminary studies of the effectiveness of student loan debt letters have been mixed, but these studies have not analyzed how institutions have composed student loan debt letters at the word-, sentence-, and document-level. As a result, this study gathered six student loan debt letters sent by different institutions across the United States and analyzed these letters for readability, cohesion, and lexical diversity. Results suggest student loan debt letters have been written in drastically different ways and do not share common vocabulary, possibly confusing the debt repayment process for students. Implications for research and practice are addressed.

Keywords student debt letters, financial aid, student loans, higher education

Volume 20 • Issue 1 2021

THE PROMISE OF COLLEGE STUDENT LOAN DEBT LETTERS: HOW INSTITUTIONS DELIVER DEBT-RELATED MESSAGING TO COLLEGE STUDENTS Legislators have increasingly become concerned about their state’s economic well-being with regard to student loan debt, as state leadership has called for progressive action in reducing loan burdens among students (Ahlman & Gonzalez, 2019). One response to these calls has been to institute student loan debt letters at universities (Reschke, 2017). As a low-cost effort associated with financial education, many schools have opted to mail or email students summary information on their federal student loan debt. Often referred to as “debt letters,” these letters typically include information on total debt accrued, estimated monthly payments, and borrowing eligibility that remains (Darolia, 2016). Possibly due to the low cost and ease of implementation, debt letters quickly gained attention as a possible solution to the “student loan debt crisis” (Quinton, 2016b, para. 1). Legislators in twelve states to-date have since proposed or passed laws requiring debt letters to be sent by colleges and universities (Attigo, 2018). A new resolution has also been proposed to make debt letters a nationwide requirement. House Resolution 1429, known as the Letter of Estimated Annual Debt for Students Act of 2017, would require all institutions accepting federal aid to send annual letters to borrowers estimating loan debt and future payments (Reschke, 2017). These proposals were initiated before research on debt letter outcomes was out, however. Today, research suggests substantial differences in the effectiveness of debt letters and the ways in which they are structured. Variance in the content of debt letters is associated with differences in outcomes associated with student loan borrowing, academic performance, and student retention from semester and year. Research pertaining to the effectiveness of debt letters is explored in the literature review that follows.

PURPOSE OF THE STUDY Researchers studying the effects of student loan debt letters have found a wide range of outcomes among institutions implementing them. Although differences in debt letter outcomes have been noted, the reasons for these differences have not been fully examined. While some institutions appear to have had success with student debt letters, other institutions

have not experienced any significant outcomes. Policy makers have looked to institutions who have experienced success when formulating recommendations to require student loan debt letters at U.S. universities and colleges. More information is needed to make effective recommendations and policy decisions related to them, however. While a range of debt letter outcomes have been documented, linguistic differences between letters that may impact debt letter effectiveness have not been fully explored. Although debt letters may be similar in their purpose, the phrasing, length and content of debt letters may vary substantially between institutions. Evaluating linguistic differences in debt letters between institutions may shed additional light on the differences in their effectiveness. A linguistic analysis is presented to further inform debt letter formation and policy decisions related to their creation, which are developing expeditiously.

GUIDING RESEARCH QUESTIONS This research looks at the linguistic differences in student loan debt letters made publicly available across six post-secondary institutions. Guiding research questions in this study were: 1. Does the reading difficulty of student loan debt letters vary substantially between postsecondary institutions? 2. Does the length of student loan debt letters vary substantially between postsecondary institutions? 3. Does the content of student loan debt letters vary substantially between postsecondary institutions?

SIGNIFICANCE OF THE STUDY This review of student loan debt letters utilized by postsecondary institutions will serve two important functions. First, this review will examine case study research on debt letter outcomes in aggregate, pointing to more generalizable findings across institutions. Second, this study will further contribute to the literature on student loan debt letters by analyzing linguistic differences between institutional student loan debt letters. These research findings will be presented along with suggestions for improvements to loan debt letters. These findings may also influence future policy formation pertaining to student loan debt letters.

Journal of Personal Finance

REVIEW OF THE LITERATURE This literature review will provide synopses of three studies that were conducted to evaluate the effectiveness of student loan debt letters in the United States. While these research findings are limited in their generalizability when considered individually, together the three studies may provide a more complete understanding of debt letters and their associated outcomes.

RESEARCH ON DEBT LETTER OUTCOMES Indiana University Debt Letters As one of the first universities to initiate debt letters in 2012, Indiana University is often looked to as a success story associated with this initiative (Darolia, 2016). Within two years of instituting debt letters, student loan borrowing at the institution had been reduced by about $44 million, or 16% overall (IU Newsroom, 2015b). By 2018, federal and private loan borrowing at IU had decreased by 19%, or more than $126 million total (McRobbie, 2018). These results garnered a great deal of media attention at the time, with articles published in national news sources such as The Wall Street Journal (Korn, 2017), CNN Money (Quinton, 2016b), and Yahoo Finance (Woodruff, 2015), among others. Some of these articles asked, “Could this simple solution solve the student loan crisis?” (Quinton, 2016b). While the debt letter solution may have seemed simple to outsiders, a much more complex financial literacy initiative was in place at Indiana University. During this timeframe, the university developed a “multifaceted financial literacy program and started adopting policies to increase student financial assistance and promote on-time graduation” (IU Newsroom, 2016a). In addition to debt letters, IU also offered peer-to-peer financial counseling, a podcast on personal finance, a website with quizzes and loan calculators, a full-time enrollment campaign, and changed the financial aid loan acceptance process to make it easier to decline loans (Quinton, 2016a). Because the debt letters were part of larger efforts to reduce loan borrowing, it is difficult to determine the effects from the letter alone, as other initiatives were also implemented during this timeframe. Administrative focus on financial education efforts was also exceptionally high at Indiana University during this time. Indiana University’s president mentioned the university’s work on financial literacy as a priority in every state of the union address from 2011 to 2018 (McRobbie, 2018). The

president also chose to award the senior vice president and chief financial officer at Indiana University, MaryFrances McCourt, with the President’s Medal for Excellence for her work on student affordability and her oversight of the IU Office of Student Financial Literacy in 2016 (IU Newsroom, 2016b). In addition to a high level of institutional focus on financial education, the university also led a national initiative on financial literacy by co-founding the Higher Education Financial Wellness Association, formerly known as the National Summit on Collegiate Financial Wellness (IU Newsroom, 2015a). Due to the comprehensive financial literacy efforts in place at Indiana University, as well as the administration’s extraordinary focus on the subject, the loan debt reduction experienced at IU may not be causally linked to student loan debt letter initiatives alone. To determine the effects of loan debt letters, it is beneficial to turn to other institutions that have implemented similar stand-alone initiatives for further examination.

Montana State University Debt Letters Debt letters similar to Indiana University’s were implemented by Montana State University in 2012 and reviewed by Stoddard, Urban, and Schmeiser in 2017. Montana State’s letter differed in that it included debt thresholds at which point letters would be sent to some, but not all, students. Freshmen who had more than $6,250 in student loans, sophomores with more than $12,000, juniors with more than $18,750, and any student with more than $25,000 in debt received a letter at Montana State. Students were provided with incentives to meet with financial planners and career coaches. Montana State’s debt letters also included strategies to reduce borrowing and work towards a timely graduation. In particular, federal Satisfactory Academic Progress regulations were outlined, informing students of the need to pass 67% of courses each semester to continue to receive federal funding. Information was also shared on the university’s banded tuition program, in which students do not incur any additional tuition charges after enrolling in 12 credit hours a semester. By charging the same amount for 12 and 15 credit hours, for example, the university sought to increase credit hours completed, leading to higher on-time graduation rates. Additionally, benefits to earning a college degree were outlined by Montana State, including lower average unemployment rates and better long-term health outcomes. To study the outcomes of Montana State’s debt letter, Stoddard, Urban, and Schmeiser (2017) used a difference-in-differences approach, using the University of Montana as a comparison site, where no student debt letters were sent. In this study, the

Volume 20 • Issue 1 2021

researchers did not find a significant reduction in the amount of student loans borrowed due to the debt letters. However, the researchers did find positive academic effects associated with the debt letters. Receiving a letter increased average grade point averages for the semester, as well as the number of credit hours completed. These effects continued into the following semester and year. Students receiving debt letters also experienced higher retention rates by semester and year compared to their peers who did not receive the letters at the University of Montana. The authors of the report argue that the academic successes students experienced may be attributable to the information provided about Satisfactory Academic Progress. While student loan debt did not significantly decrease, there were other, unintended positive outcomes associated with the letters. Montana State’s outcomes suggest that outlining Satisfactory Academic Progress and other benefits to completing coursework towards a timely graduation are important to include in student loan debt letters.

University of Missouri Debt Letters

letter, and another four reported being unsure. Additionally, two out of four students in a control group stated that they had received the debt letter, when they in fact had not. Overall, the debt letters sent at the University of Missouri did not appear to be particularly memorable for students. One concern about sending debt letters is that they may potentially discourage students who need loans to complete their education from utilizing them (Quinton, 2016a). Research demonstrates that students who are averse to borrowing, and that have unmet need of $2,000 or more during their first year of college, are less likely to complete their degree, for example (Institute for Higher Education Policy, 2008). The researchers at the University of Missouri, therefore, looked for any negative completion outcomes associated with sending debt letters to students. They found no negative outcomes associated with sending debt letters to students, however. Students receiving debt letters were no more likely to withdraw from courses, change their major, leave the university, or change the number of hours they worked in work-study positions (Darolia & Harper, 2017).

Although the researchers were unable to determine any harm that had been caused by the letters, they did find that they Another study of debt letters was produced by Darolia and may not be the most effective approach to addressing student Harper in 2017 at the University of Missouri. Debt letters sent by loan debt either. Half of students who received an emailed the university differed from other debt letters in that they only debt letter reported that they believed that it was the best provided factual information about loan debt and estimated approach, while the other half that were interviewed did not repayments. Unlike Montana State and Indiana University, recommend debt letter emails, believing that students skimmed other financial education resources were not promoted or overlooked them (Darolia & Harper, 2017). The researchers simultaneously, and students were not outwardly encouraged found that students who receive frequent communication about to reduce their borrowing. Debt letters at the University their finances may decrease their attention to any one message. of Missouri were not written with the intent to increase or Some students even reported purposefully avoiding paying decrease loan-borrowing behavior, but rather to provide attention to their student debt. In interviews, students suggested factual information. that other approaches such as tweets, texts, songs/videos, Darolia and Harper (2017) found in their 2017 review that presentations/budgeting classes, letters sent to parents, or onesending a debt letter at the University of Missouri did not lead on-one financial or academic advising may be more beneficial. It to a change in the amount students borrowed or the likelihood is important to note that, overall, students who were interviewed that they would borrow. Although Missouri’s debt letter did about debt letters referred to their lack of understanding, not a not alter borrowing behavior, it did induce more information lack of data as hindering their financial decision-making. seeking among some students. The researchers found that students receiving debt letters were two percent more likely to Review of Debt Letter Findings To-Date seek a meeting with a financial aid officer. Together, the three studies at Indiana University, Montana State University, and the University of Missouri suggest that Interviews conducted by Darolia and Harper (2017) with debt debt letters by themselves may not be effective in reducing letter recipients demonstrated that students did not find student loan debt, but as part of larger financial education the letters particularly distinguishable from others sent by programs, they may be beneficial. When students are provided the financial aid office or other offices on campus. Out of 23 with information on additional resources they may access, students interviewed, just nine remembered receiving the debt they are more likely to engage in help-seeking behavior

Journal of Personal Finance

(Darolia & Harper, 2017). Experimenting with other methods of communicating student debt information, such as through academic courses, presentations, and the use of social media is also recommended (Darolia & Harper, 2017). Including information on Satisfactory Academic Progress and other incentives to graduate on-time are important to include in student debt letters as well (Stoddard et al., 2017). The most recent study to date, McKinney’s (2020) dissertation, focused on the effect of a debt letter on community college student decision making, including decisions on enrollment, academic programs, and borrowing. McKinney (2020) ultimately found that community college students who received a debt letter were more likely to borrow less of a total percentage of their available loan amount than students who did not receive the letter. McKinney (2020) employed t-testing without regressing individual student characteristics, leaving room for future research to explore how student identity (age, race, gender, etc.) may impact student debt letter interpretation and subsequent action. Moreover, McKinney’s (2020) findings suggest that student loan debt letters may influence student borrowing, but that student loan debt letters may be one source of information that a student uses to decide how to manage their educational debt.

Readability of Financial Aid Information U.S. institutions should be aware of how their student loan debt letters are written and whether their students are able to read and fully comprehend the letters they send. Letters may be written differently between actors and institutions. The reading level of financial aid award documents, for example, vary greatly in their reading grade level between institutions (Taylor & Bicak, 2019). Many financial aid offices craft financial aid websites at a higher reading level than students are currently at (Taylor & Bicak, 2019). Research demonstrates that the average U.S. resident reads at the 7th grade reading level (Clear Language Group, 2020), and just 37% of high school graduates read at the 12th grade level in the United States (National Assessment Governing Board, 2020). Further compounding the difficulties students may face in reading text above their reading grade level is the fact that students experiencing stress or anxiety read and comprehend material differently than students who are not experiencing these symptoms (Rai et al., 2015). Students experiencing financial stress may, therefore, exhibit different outcomes than students who are not experiencing high levels of financial stress when reading student loan debt letters. Additionally,

the origin and duration of stress and/or anxiety a student experiences may impact cognitive function and reading comprehension (Sandi, 2013). Students with prolonged levels of financial stress, therefore, may read and comprehend information differently than students who have experienced shorter periods of stress. Research demonstrates that reading comprehension also depends on the difficulty and the familiarity of the reading task at hand (Plieger et al., 2017). Some acute stress in a wellrehearsed task may actually improve comprehension, while long-term stress or anxiety may worsen comprehension (Rai et al., 2015). New reading tasks that are unfamiliar to students may reduce comprehension overall, while tasks that are familiar or have been experienced in similar ways before may result in higher levels of reading comprehension. Upon receipt of a debt letter, college students may be reading debt-related or financial-related content for one of the first times in their life, as new financial situations can be stressful to experience and discuss, financial-related information should be written as simply as possible without removing critical information from the material (Taylor, 2019; Taylor & Bicak, 2019). Together, the current literature suggests that reading comprehension may be an important element to consider when evaluating the effectiveness of student loan debt letter interventions. Factors that impact reading comprehension are necessary to consider when crafting debt letters and policies associated with them. Studying the linguistic differences between institutional debt letters may help contribute to our understanding of the differences experienced in outcomes between institutions. It may also help provide guidance for future policy decisions pertaining to debt letters.

Research Related to Debt and Borrowing Among College Students Although an exhaustive review of all debt-relate research regarding college students is far beyond the scope of this study, it is important to highlight several main trends in how institutions inform their students of their debt and borrowing habits, in addition to educational interventions to alert students to financial concepts and debt knowledge. Longitudinal studies of college student debt and specific borrowers have found that low-income students are particularly debt averse and avoid college borrowing at many costs (Boatman et al., 2017), while particular students have gaps in knowledge regarding student loans and borrowing, including students of Color (Ahlman & Gonzalez, 2019; Elliot &

Volume 20 • Issue 1 2021

Lewis, 2015), low-income students (Ahlman & Gonzalez, 2019; Boatman et al., 2017; Huelsman, 2015), and first-generation college students (Lee & Mueller, 2014). Aside from written communication such as debt letters, many institutions of higher education offer financial literacy classes and financial wellness programming to help students understand debt and borrowing (Beale & Cude, 2017; Britt et al., 2011; Leighty, 2018). Moreover, national efforts have been made to simplify the financial aid and loan borrowing process, including simplifications of the Free Application for Federal Student Aid (FAFSA) (Taylor, 2019) and Federal Student Aid introducing a simplified College Financing Plan (formerly known as the Financial Aid Shopping Sheet) (U.S. Department of Education, 2020). However, to date, there has been no universal or standardized financial aid award letter system, debt letter system, or any other mechanism to describe and detail a students’ debt and pathways toward completing a degree and successfully repaying that debt. As a result, this study fills an important gap in the research to explain how several institutions are directly communicating debt to students and whether these communications could or should be tied to student behaviors, such as reduced borrowing, academic pathway changes, enrollment changes, or other debt-related reactions.

RESEARCH METHODOLOGY Because outcomes vary greatly between institutions attempting to achieve the same positive debt letter outcomes, linguistic differences that exist between letters are worthy of further study. To date, no research has addressed the linguistic elements comprising student loan debt letters, nor has prior research articulated the manner in which debt letters were written before they were delivered to students. As a result, this exploratory study articulates how debt letters are written using several linguistic techniques in hopes that institutions may consider how debt letters are written before they are delivered to students.

Linguistic Analysis of Student Debt Letters Although debt letter outcomes vary greatly between institutions, no extant research has examined deeper level linguistic differences within these letters. The reading grade level, word count, and complexity of the letters across six institutions are therefore examined in this study. Although extant research has suggested financial aid-related materials

may be difficult to read for prospective student audiences (Taylor & Bicak, 2019), no studies have focused on the difficulty of reading a student loan debt letter. Such a gap in the literature is fathomable considering the relatively short period of time in which debt letters have been written and disseminated to student audiences. This preliminary study seeks to understand how difficult debt letters are to read, how long they are, and how semantically-diverse these letters are, measured by token-type ratio (Jurafsky & Martin, 2014). In this study, a linguistic analysis was conducted on the three debt letters previously studied (Indiana University, Montana State, and the University of Missouri), in addition to three supplementary student loan debt letters that have been made publicly available (Columbus State Community College, Sam Houston State University, and Western Governors University). The research team contacted these institutions and learned that each letter was sent semesterly, but the institutions did not provide further detail on frequency or mode (email, physical letter, etc.). Moreover, it is unclear what the intended effect of each letter was by institution, as no prior research has explored the effect of debt letters or has interviewed institutional professionals about their anticipated effects of sending a student debt letter, whether that effect being reduced borrowing, a change in academic pathway, or another loan-related behavior, such as consolidation or repayment. Together, these six student loan debt letters were evaluated on reading grade level, word count, and token-type ratio of the words used within the letters. The research team employed Readability Studio—a quantitative linguistic analysis software program—to examine these six letters student debt letters which have been made publicly available. Grade-level readability measures the difficulty of sentence structure and word choice, producing a reading grade comprehension level required to read a particular text (Carver, 1974; Sticht, 1970). Grade-level readability measures have been used extensively by the U.S. Department of Defense and have been tested for validity and reliability throughout history (Carver, 1974; Fry, 1987; Johnson, 1972; Kniffin, 1979; McClure, 1987; Sticht, 1970; Sticht & Zapf, 1976). The readability measures employed in this study are the Automated Readability Index, Flesch-Kincaid Grade Level Test, GunningFog Index, and Simple Measure of Gobbledygook (SMOG). These four measures have been used to evaluate higher educationrelated materials in prior research (Taylor, 2018a, 2018b, 2019).

Journal of Personal Finance

The research team then employed Readability Studio to merge the six texts in order to perform the token-type analysis and the corpora analysis of word frequency (Table 2). Token-type ratio is an important metric of text—and student debt letters— as the token-type ratio of a text indicates the differentiation of diction of the text. Specifically within financial aid-related diction, there may be difficult words that institutions must use in their debt letters, such as unsubsidized or accrual. However, repeating these words multiple times in a text, driving down the token-type ratio of the text, may prove beneficial to students’ reading comprehension of debt letters, allowing these borrowers to better understand their debt and how to repay it. Ultimately, by merging the texts and analyzing both the corpus and each separate text, the research team was able to identify the most frequently used terms across the entire corpus and each text, providing insight into the similarities and/or differences between each text and the corpus.

FINDINGS The linguistic analysis of student loan debt letters (n=6) reveals variance in the reading grade levels, the word count, and the complexity of the language used in debt letters across six institutions. Utilizing findings from this research, as well as those in the current literature, this article then outlines recommendations to improve institutional debt letters. These recommendations are provided with the intent to inform policy implementation. Table 1 includes descriptive statistics pertaining to the readability of six debt letters currently available to the general public as of March 2019:

Table 1 Linguistic analysis of debt letters, by readability, word count, and token-type ratio Institution

Readability*

Word count

Token-type ratio

Columbus State Community College

13th-grade

218

57.3%

Indiana University, Bloomington

12.5th-grade

646

39.3%

Montana State University

12.8th-grade

544

47.9%

Institution

Readability*

Word count

Token-type ratio

Sam Houston State University

11.8th-grade

253

49.8%

University of Missouri, Columbia

12.2nd-grade

238

51.6%

Western Governors University

11.9th-grade

154

65.5%

*Note: Readability levels calculated using four common English-language readability measures that produce gradelevel readability levels (e.g., a 13th-grade level text would likely require 13 years of English reading instruction to comprehend the text); These measures included the Automated Readability Index, Flesch-Kincaid Grade Level Test, Gunning-Fog Index, and Simple Measure of Gobbledygook (SMOG). Data in Table 1 suggest that debt letters—from institution to institution—may be written much differently in terms of readability level, word count, and token-type ratio. For instance, the simplest debt letter to read was Sam Houston State University’s debt letter at the 11.8th-grade level and in 253 words. Inversely, the most difficult debt letter to read was Columbus State Community College’s at the 13th-grade level in 218 words. This short study suggests there may be a readability gap between debt letters at different institutions. Moreover, the longest debt letter was Indiana University, Bloomington, which was written in 646 words. Western Governors University wrote their debt letter in 154 words, representing a nearly 500-word difference. There were also marked differences in token-type ratio, a measure of semantic diversity in a text. The most semantically-diverse debt letter was Western Governors University at 65.5%, while Indiana University, Bloomington’s debt letter was written at a 39.3% token-type ratio, representing nearly a 30% difference in semantic diversity from letter to letter. Table 2 outlines a corpus analysis of content words of all six debt letters and each debt letter individually to articulate intrasemantic and inter-semantic diversity of the six letters. Corpus analysis is a method for conducting in-depth investigations of linguistic phenomena which are retrieved from authentic communications that are digitally stored and made available for access, retrieval and analysis.

Volume 20 • Issue 1 2021

Table 2 Corpus analysis identifying ten most frequent content words in debt letters, by individual letter and corpus (n=6)

Corpus

Columbus State Community College

Indiana University, Bloomington

Montana State University

loans (52)

student (7)

loans (24)

financial (11)

loan (44)

loan (5)

loan (11)

loan (9)

student (39)

amount (5)

federal (8)

loans (8)

debt (28)

federal (4)

included (8)

student (7)

federal (21)

loans (4)

debt (8)

debt (7)

information (20)

CSCC (3)

information (7)

coach (6)

repayment (18)

borrowed (3)

estimates (7)

please (4)

financial (18)

affect (2)

letter (6)

future (4)

interest (11)

maximum (2)

student (5)

education (4)

borrowing (11)

campus (2)

interest (4)

repayment (3)

Sam Houston State University

University of Missouri, Columbia

Western Governors University

loan (12)

loans (6)

student (6)

student (12)

repayment (5)

borrowing (4)

debt (9)

loan (5)

loans (4)

loans (6)

interest (5)

time (2)

federal (3)

information (5)

term (2)

information (3)

debt (3)

limit (2)

repayment (3)

direct (3)

loan (2)

estimated (2)

student (2)

paying (2)

insert (2)

accruing (2)

aid (2)

school (2)

Results in Table 2 suggest that across all debt letters, the terms “loans” and “loan” were the most frequent, followed by “student” and “debt.” However, across all six texts, there were many different terms appearing most frequently. For instance, Columbus State Community College’s debt letter frequently mentioned “amount,” but this term was not frequent in any other debt letter. Similarly, even though the type of letters analyzed were debt letters, “debt” was not a common

term in Columbus State Community College’s or Western Governors University’s debt letter. Many unique terms were included in Montana State University’s debt letter, including “coach,” “please,” “future,” and “education”—none of these terms frequently appeared in any other debt letter. In addition, only the Columbus State Community College debt letter mentioned the name of the institution in the debt letter—“CSCC” was mentioned three times—whereas no other institution made frequent mention of their name. Similarly, there were very few mentions of “education,” “school,” or “institution”—these words were used very infrequently, if at all. From here, results in Table 2 suggest that—across the corpus and each individual letter—the semantics may drastically differ from debt letter to debt letter.

DISCUSSION In reviewing student loan debt letters, there are several notable findings. First, it must be noted that the research team is unclear on the mode of dissemination of the debt letters in this study. As a result, it is unclear whether institutions emailed these letters, mailed them to physical addresses, texted students with a link to their letter, or some other form of digital or print communication. In addition, we are unclear the goal of sending the student debt letter, save for the fact that institutions wanted students to be aware of their debt. Whether institutions want students to change their borrowing habits or academic pathways is unclear from the data presented in this study. However, what is clear, is that institutions convey similar information in very different ways, and many student loan debt letters may be unreadable and unintelligible for students of average reading ability, making it difficult for students to understand their student loan debt and act upon it in whatever way they deem appropriate. From here, the primary finding of this study is that different institutions may be communicating debt in confusing and unfamiliar ways . If a student is unable to understand their debt, it is difficult to argue that the student is an informed borrower and understands their pathways to either repayment, debt consolidation, or some other debt-related action. For instance, consider this section of a debt letter in this study, written at near the 18th grade reading level with a high token-type ratio (0.60): Receiving the maximum amount of loans each year can add up quickly - which can result in less loan availability for the continuing your education and affect how much you owe the federal government when repayment begins.

Journal of Personal Finance

Debt-related ideas such as loan maximums, loan availability, and repayment may be too complex for students understand on their own, and the language used in the above section does little to simplify this process. Instead, the above section could be re-written as: If you take the maximum amount of loans you can, you may run out. This means you have less in loans available to you later. Also, the more you take out, the more you have to repay. Here, this section is written at the 7th grade level and with a token-type ratio of 0.32—appropriate for all college students and their support networks—and conveys the same basic information in simpler terms. Moreover, the re-written section amounts to 37 total words, whereas the original text was 36 words—by splitting the content into three sections, driving down the token-type ratio, and only increasing word count by one word, the section is roughly 10 grade levels easier to read. As a result, students may better able to understand their student loan debt and pathways forward, even if it is unclear what the institutional goal is for sending the student their student loan debt letter. Next, it is clear that student loan debt letters may feature different terms across different institutions, yet all six institutions in this study allow for Title IV student borrowing and the completion of the FAFSA, meaning that all six institutions work with Federal Student Aid to award similar types of loans and aid packages. Therefore, it is curious that each institution wrote their letter in different ways, even though aid was coming from the same source. As a result, this study makes an important finding that, indeed, student loan debt letters are written differently by different institutions, and that institutions not only write debt letters at different readability levels, but they also use different terminology to explain student debt to their students. For instance, consider that several letters did not mention student loan repayment. In these cases, it is hard to argue that the institutional goal of the debt letter was to inform students on how to repay their debt. Future research should investigate what institutional goals actually are for sending student loan debt letters, beyond simply making students aware of what their debt is and what types of loans they may hold. If a student holds federal student loan debt, students are always able to check their National Student Loan Data System (NSLDS) account and research the most up-to-date information regarding their student loans. A student loan debt letter seems to be a repetition of this information nudged to a

student, instead of a student having to fetch the information themselves from the NSLDS website. This type of nudge may be important for students to understand what debt they hold, but without insight from institutions as to the purposes and the outcomes of sending a debt letter, it seems that sending this information may be repetitive and already available to students in real time through the NSLDS portal. It should be noted that student loan debt letters have positively impacted both financial and academic student success outcomes. There are key elements that may be included in debt letters to strengthen those student success outcomes. These include information about Satisfactory Academic Progress and information on how to contact a professional for further guidance on student loans. Making contact information for additional guidance readily accessible at the top of the debt letter is recommended, as this study suggests that many institutions did not include contact information in their debt letter or did not mention their institution’s name in the debt letter (see Table 2). Student debt letters also appear to be more effective when integrated within a systematic and wide-reaching financial literacy program (IU Newsroom, 2015a, 2015b). Because debt letters may encourage additional help-seeking behavior, it is difficult to know the extent of the outcomes experienced. Systems that are prepared to help students gather additional information pertaining to their student loan debt may extend the outcomes experienced from receiving debt letters. Links and contact information for additional offices or services on campus are important to include in debt letters, as this information allows students to seek out resources for assistance. However, as previously discussed, if students are unable to read their debt letter because it is too difficult as this study suggests (see Table 1), it is unclear whether students can act upon the information in the debt letter. Debt letters may benefit from having students engaged in the process of writing them, as readability and comprehension of financial aid information may be higher when students themselves are asked to help craft the content: Students may write in intelligible, student voices. As a result, universities may wish to consider asking students to draft or audit debt letters. The types of information that students would find beneficial to include in a debt letter may differ from the information that professional staff believe to be helpful, as the debt letters in this study may benefit from a reduction in readability difficulty and token-type ratio (see Table 1). By allowing students to

Volume 20 • Issue 1 2021

analyze and simplify debt letters, without removing critical content, institutions may be able to communicate more clearly with their student borrowers. Reducing the amount of information that students need to read to gain the necessary information on their student loans may also be beneficial. Utilizing bullets, charts, videos, or images may help reduce the length of letters students are asked to read. Reducing the amount of text by using active, rather than passive language, can also communicate messages more efficiently (Taylor, 2019; Taylor & Bicak, 2019. Although this study did not analyze debt letters that included complex charts or visual graphics to help explain student debt, such features may be beneficial for helping students understand complex financial topics, such as student loan borrowing. Students may benefit from clearer language and structure of debt letters. Direct and actionable steps may be provided for students to follow to reduce confusion. Additional steps can be taken to ensure that the language utilized in student loan debt letters is devoid of acronyms or jargon that students may be unfamiliar with. When unfamiliar words are used, they may be situated in sentences that provide context clues for readers who are unfamiliar with them. As data in Table 1 suggests, token-type ratios indicate that student loan debt letters may use niche, financial terms only once or twice in debt letter, failing to allow students to find context clues within the debt letter to help them understand it, a common way of increasing reading comprehension (Taylor, 2019; Taylor & Bicak, 2019). The timing of when debt letters are sent is also important to consider. Because students experiencing a slightly heightened level of stress read more proficiently, times of the year where students are experiencing some, but not overwhelming stress, may be the best times to send debt letters. Conversely, sending debt letters during finals week, midterms, or other periods of time when students are facing undue amounts of stress would not be recommended. Here, it is important to understand that a limitation of this study is that the research team was not clear on specifically when the debt letter was sent, beyond the fact that these letters were sent each semester. The time of day, the day of the week, and the mode of communication of the debt letter is critical to understand how debt letters may influence behavior. Moreover, the research team was not privy to the process of how institutions wrote the debt letter and whether or not institutions were trying to match their students perceived academic ability or audience in writing the letter. From here, future research should

investigate how debt letters are written, who is involved in that writing process, and specifically how debt letters are delivered. Research also suggests that students with an increased likelihood of experiencing stress—financial stress in particular— may not read and comprehend information in debt letters as effectively as students who are not experiencing financial stress (Rai et al., 2015; Sandi, 2013; Taylor, 2019; Taylor & Bicak, 2019). Continued steps may be taken to examine the effects of debt letters on populations experiencing higher levels of stress. Because reading and comprehension improve with practice reading similar items over time (Carver, 1974; Fry, 1987; Johnson, 1972; Taylor, 2019), sending out similar debt letters annually or each semester may improve comprehension of such letters, impacting a student’s ability to act upon that letter. Finally, universities implementing debt letters may wish to conduct research on the outcomes associated with this initiative. Contacting the university institutional research office may be an effective way to assess and continue to improve student loan debt letters. There still may be many unintended consequences associated with debt letters that are still unknown. Researching the outcomes associated with debt letter interventions ensures that universities are taking steps to avoid causing any unforeseen harm. Yet it is important to remember that debt letters are a relatively novel and new intervention to help college students understand their debt, and much more future research is needed to understand how to best communicate with college students and help them develop a strong sense of financial literacy and wellness.

CONCLUSION Although prior research has started to analyze the effectiveness of student debt letters (Darolia & Harper, 2017; McKinney, 2020; Stoddard et al., 2017), the study at hand demonstrated that not only does the effect of delivering student debt letters vary from institution to institution, but the way in which debt letters are written are also drastically different from institution to institution. From here, higher education researchers interested in student financial aid, student borrowing, institutional communication, and institutional retention and graduation initiatives ought to further investigate how student debt letters are written and whether the nature of the written composition of the letter affects how students react to the letter, thus influencing their future educational decisions.

Journal of Personal Finance

If there is indeed a student debt crisis (Quinton, 2016b), all aspects of student borrowing should come under the microscope to learn how both the act of borrowing and communication about borrowing may affect students and their decision making. Today, attending a postsecondary institution in the United States may cost a student hundreds of thousands of dollars. Surely, a student debt letter—a few pennies worth of paper and a stamp—is valuable enough to analyze and refine in order to provide students with the most accurate and instructional information possible. However,

institutions need be aware of how they are writing a debt letter in any format and when they are specifically sending the debt letter could affect a student's comprehension of their debt and their ability to act upon it. Ultimately, college students and their families who are just beginning to develop their nascent sense of financial literacy and wellness should be given every opportunity, in the simplest forms possible, to better understand their debt to make them a more savvy borrower and successful college student.

Volume 20 • Issue 1 2021

REFERENCES Ahlman, L., & Gonzalez, V. (2019). Casualties of college debt: What data show and experts say about who defaults and why. Retrieved from: http://hdl.handle.net/10919/99788 Attigo. (2018). Definitive guide to student debt letters. Retrieved from: https://www.attigo.com/definitive-guide-to-debt-letters Beale, E.M., & Cude, B.J. (2017). College students’ attitudes toward debt. International Journal of Undergraduate Research and Creative Activities, 9(5), 1-12. https://doi.org/10.7710/2168-0620.1099 Blalock, C. (2017). African American graduates' experiences of managing college debt. [Doctoral dissertation, Walden University]. Retrieved from: https://scholarworks.waldenu.edu/dissertations/4422/ Boatman, A., Evans, B.J., & Soliz, A. (2017). Understanding loan aversion in education: Evidence from high school seniors, community college students, and adults. AERA Open, 3(1), 1-16. https://doi.org/10.1177%2F2332858416683649 Britt, S. L., Grable, J. E., Cumbie, J., Cupples, S., Henegar, J., Schindler, K., & Archuleta, K. L. (2011). Student financial counseling: An analysis of a clinical and non-clinical sample. Journal of Personal Finance, 10(2), 95-121. Brown, A., Collins, J. M., Schmeiser, M., & Urban, C. J. (2014). State mandated financial education and the credit behavior of young adults (Federal Reserve Board Working Paper 2014-68). Washington D.C.: Federal Reserve Board. Retrieved from: https://www.federalreserve.gov/pubs/feds/2014/201468/201468pap.pdf Carver, R. P. (1974). Two dimensions of tests: Psychometric and edumetric. American Psychologist, 29(7), 512-518. Retrieved from http://psycnet.apa.org/journals/amp/29/7/512/ Clear Language Group. (2020). Readability: What is readability? Retrieved from: http://www.clearlanguagegroup.com/readability/ Darolia, R. (2016). An experiment on information use in college student loan decisions (Federal Reserve Bank of Philadelphia Working Paper 16-18). Retrieved from: https://ideas.repec.org/p/fip/fedpwp/16-18.html Darolia, R., & Harper, C. (2017). Information use and attention deferment in college student loan decisions: Evidence from a debt letter experiment. Educational Evaluation and Policy Analysis, 40(1), 129-150. https://doi.org/10.3102%2F0162373717734368 Elliott, W., & Lewis, M. (2015). Student debt effects on financial well‐being: Research and policy implications. Journal of Economic Surveys, 29(4), 614-636. https://doi.org/10.1111/joes.12124 Fry, E. (1987). The varied uses of readability measurement. Journal of Reading, 30(4), 338-343. https://www.jstor.org/stable/40032863 Huelsman, M. (2015). The debt divide: The racial and class bias behind the "new normal" of student borrowing. Retrieved from: http://hdl.handle.net/10919/90810 Institute for Higher Education Policy. (2008). Student aversion to borrowing. Who borrows and who doesn’t? Retrieved from: https://files.eric.ed.gov/fulltext/ED503684.pdf IU Newsroom. (2015a, June 25). Indiana University to host national summit on student financial wellness. Indiana University. Retrieved from: http://archive.news.iu.edu/releases/iu/2015/06/financial-wellness-summit.shtml IU Newsroom. (2015b, June 3). IU leader testifies about success of university’s financial literacy efforts. Indiana University. Retrieved from: http://archive.news.iu.edu/releases/iu/2015/06/financial-literacy-testimony.shtml IU Newsroom. (2016a, September 1). Indiana University initiatives continue to pay off in reduced student borrowing. Indiana University. Retrieved from: http://archive.news.iu.edu/releases/iu/2016/09/student-loan-reductions.shtml IU Newsroom. (2016b, March 11). MaryFrances McCourt awarded Indiana University President’s Medal. Indiana University. Retrieved from: http://archive.news.iu.edu/releases/iu/2016/03/mccourt-presidents-medal.shtml Johnson, K. H. (1972). An analysis of the relationship between readability of Air Force procedural manuals and discrepancies involving non-compliance with the procedures (Master’s thesis). Available from ERIC database. (UMI No. ED070941)

Journal of Personal Finance

Jurafsky, D. & Martin, J.H. (2014). Speech and language processing: An introduction to natural language processing, computational linguistics, and speech recognition. Retrieved from: http://www.cs.colorado.edu/~martin/SLP/Updates/1.pdf Kniffin, J. D. (1979). The new readability requirements for military technical manuals. Technical Communication, 26(3), 26-3. Retrieved from: https://www.jstor.org/stable/43086724?seq=1 Korn, M. (2017, July 12). I owe that much? Having student-loan data leads to drop in borrowing. The Wall Street Journal. Retrieved from: https://www.wsj.com/articles/states-require-more-disclosure-on-student-loans-1499798161 Lee, J., & Mueller, J.A. (2014). Student loan debt literacy: A comparison of first-generation and continuing-generation college students. Journal of College Student Development, 55(7), 714-719. https://muse.jhu.edu/article/558257 Leighty, A. (2018). Implications of student debt on financial wellness: Can universities help? [Thesis, Butler University]. Retrieved from: https://digitalcommons.butler.edu/ugtheses/427/ McClure, G. (1987). Readability formulas: Useful or useless? IEEE Transactions on Professional Communication, 30(1), 12-15. http://dx.doi.org/10.1109/TPC.1987.6449109 McKinney, K.P. (2020). Student loan debt for community college transfer students and how debt information letters impact future borrowing decisions. [Doctoral dissertation, Mississippi State University]. Retrieved from: https://hdl.handle.net/11668/19458 McRobbie, M. (2018, October 16). Progress and growth in the last year of Indiana University’s second century. Indiana University Office of the President: State of the University. Retrieved from: https://president.iu.edu/speeches/state-of-university/2018.html National Assessment Governing Board. (2020). The nation’s report card. Retrieved from: https://www.nationsreportcard.gov/ Plieger, T., Felten, A., Diks, E., Tepel, J., Mies, M., & Reuter, M. (2017). The impact of acute stress on cognitive functioning: A matter of cognitive demands? Cognitive Neuropsychiatry, 22(1), 69-82. https://doi.org/10.1080/13546805.2016.1261014 Quinton, S. (2016a, May 19). What happens when you warn students about their loan debt? The Pew Charitable Trusts. Retrieved from: http://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2016/05/19/what-happens-when-you-warn-studentsabout-their-loan-debt Quinton, S. (2016b, May 24). Could this simple solution solve the student loan crisis? CNN Money. Retrieved from: http://money.cnn.com/2016/05/24/pf/college/student-loan-letter/index.html Rai, M. K., Loschky, L. C., & Harris, R. J. (2015). The effects of stress on reading: A comparison of first-language versus intermediate second-language reading comprehension. Journal of Educational Psychology, 107(2), 348-363. https://doi.org/10.1037/a0037591 Reschke, M. (2017, March 20). Bill would make IU student debt initiative a nationwide requirement. IU Bloomington Newsroom. Retrieved from: http://archive.news.indiana.edu/releases/iub/iu-in-the-news/2017-03-20-debt-bill.shtml Sandi, C. (2013). Stress and cognition. WIREs Cognitive Science, 4(3), 245-261. https://doi.org/10.1002/wcs.1222 Sticht, T. G. (1970, October). Literacy demands of publications in selected military occupational specialties (Research Report No. 25-70). Retrieved from https://eric.ed.gov/?id=ED044615 Sticht, T. G., & Zapf, D. W. (1976, September). Reading and readability research in the armed forces (Research Report No. 76-4). Retrieved from http://eric.ed.gov/?id=ED130242 Stoddard, C., Urban, C., & Schmeiser, M. (2017). Can targeted information affect academic performance and borrowing behavior for college students? Evidence from administrative data. Economics of Education Review, 56, 95-109. Retrieved from: https://ideas. repec.org/a/eee/ecoedu/v56y2017icp95-109.html Taylor, Z. W. (2018a). ¿Comprehenderán mis amigos y la familia? Analyzing Spanish translations of admission materials for Latina/o students applying to 4-Year institutions in the United States. Journal of Hispanic Higher Education, 1-15. https://doi.org/10.1177/1538192718775478

Volume 20 • Issue 1 2021

Taylor, Z.W. (2018b). Intelligibility is equity: Examining the readability of international undergraduate admissions materials. Higher Education Quarterly, 72(2), 160-169. https://doi.org/10.1111/hequ.12155 Taylor, Z.W. (2019). Six easy steps: Do aspiring college students understand how to apply for financial aid? Journal of Student Financial Aid, 48(3), 1-17. https://ir.library.louisville.edu/jsfa/vol48/iss3/1/ Taylor, Z.W. & Bicak, I. (2019) What’s the FAFSA? An Adult Learner Knowledge Survey of Student Financial Aid Jargon. Journal of Adult and Continuing Education. https://doi.org/10.1177/1477971418824607 U.S. Department of Education. (2020). The college financing plan. Retrieved from: https://www2.ed.gov/policy/highered/guid/aidoffer/index.html Woodruff, M. (2015, May 6). 5 genius ways colleges are tackling the student debt crisis. Yahoo Finance. Retrieved from: https://finance.yahoo.com/news/5-genius-ways-colleges-are-tackling-the-student-debt-crisis-194429389.html

Journal of Personal Finance

The Relationship Between Home Equity and Retirement Satisfaction

Blain Pearson, Ph.D., CFP® Donald Lacombe, Ph.D.

Abstract As life expectancies continue to climb, home equity may increasingly become a much-needed resource to finance late-life consumption. This study hypothesizes that a higher ratio of home equity relative to net worth creates resource constraints, resulting in disutility for individuals who are in retirement. The findings suggest that increases in a retiree’s ratio of home equity relative to net worth is associated negatively with a satisfactory retirement experience. The ensuing discussion highlights two important issues. The first issue is that, for many retirees, home equity is inefficient in promoting retiree well-being. The second issue is that retirees may have limited knowledge of the available tools to access home equity. Thus, arguments for increased efforts to promote the responsible utilization of home equity as a part of an individual’s plan for retirement are discussed.

Keywords financial planning, home equity, retirement adequacy, retirement income

Volume 20 • Issue 1 2021

INTRODUCTION The number of U.S. individuals that are aged 65 and older is expected to grow to 83.7 million by 2050 (Ortman et al., 2014). As the older segment of the U.S. population continues to grow and transitions into retirement, capital accessibility to finance consumption is a consideration that many retirees face. When compared to other resources that can finance consumption during retirement, home equity is often an overlooked resource (Munnell, 2015; Neuwirth et al., 2017). One explanation for why home equity is an overlooked resource to finance consumption during retirement is that home equity is relatively illiquid when compared to other types of assets. A high concentration of assets allocated in the form of home equity creates relative resource constraints for retirees. The resulting resource constraints may lead to a reduced ability for retirees to enjoy their saved resources. Thus, resource constraints that are byproducts of “house-rich” retirees may prevent many retirees from achieving a fulfilling retirement experience. The findings of this study suggest that high concentrations of home equity during retirement are associated negatively with retirement satisfaction. While home ownership may provide a sense of stability and a set of positive emotions (Elsinga & Howkstra, 2005; Foye et al., 2018), home equity is a relatively unproductive asset in the promotion of a positive retirement experience. The ensuing discussion highlights the lack of a theoretical justification for retirees to hold their assets in the form of home equity and evaluates the available options for the responsible utilization of retiree home equity as a resource to increase retiree well-being.

BACKGROUND Over 95% of Americans that are age 75 or older want to remain in their current residence as long as possible (Venti & Wise, 2000). An AARP (2000) study shows that individuals over the age of 45 (65) have an 80% (82%) desire to remain living in their current residence. Using structural modeling, Nakajima & Telyukova (2020) show that homeowners are reluctant to use home equity because they prefer to stay in their home and cannot easily borrow against it. The desire for residential continuity may explain why older Americans continually build substantial equity in their homes (Venti & Wise, 1989; Feinstein & McFadden, 1989; Megboluge, 1997; Venti & Wise, 2000; Munnell et al., 2007; Sabia, 2008;), and why the majority of older Americans own their home (U.S. Census Bureau, 2000)

and 85% have no lien on their home (U.S. Census Bureau, 2005). Continuing to build home equity in late-life, however, is counterintuitive to the consumption over the life-cycle hypothesis (Modigliani & Brumberg, 1954). The consumption over the life cycle hypothesis suggests that all saved resources are accumulated for the purposes of financing late-life consumption. However, much of the current literature provides evidence suggesting that individuals nearing late-life exhibit a pattern of spending declines (Hurd & Rohwedder, 2003; Haider & Stephens 2007; Fisher et al., 2008; Aguila et al., 2011; Olafsson & Pagel, 2018). Utilizing data from the American Housing Survey, Davidoff (2004) shows that homeowners over 75 routinely spend less year over year compared to younger homeowners. The declines in spending during retirement have been shown not to be driven by socio-economic status (Battistin et al., 2009). Researchers have argued that changes in household production, as a result of shifts in the opportunity cost of time after leaving the labor force, and decreases in work-related expenses help explain the reduction in spending during retirement (Hurt, 2008; Aguiar & Hurt, 2013). Thus, while spending may decline during retirement, consumption remains constant. This viewpoint, however, does not help explain individuals’ reduction in leisure spending and their increases of liquid savings during retirement (Olafsson & Pagel, 2018). Individuals who utilize their saved resources to finance consumption during their retirement may perceive the ability to spend certain assets differently. For example, retirees may perceive their home equity or gains in their home equity as having less discretionary spending power than their stock assets or gains in their stock assets. Thus, retirees may be resistant to utilize home equity because retirees may mentally account their home equity, relative to their other assets, as an asset that should not be utilized for consumption (Thaler, 1990). The resistance to utilize home equity to finance late-life consumption has been evidenced by the research showing home equity tends to be the last asset consumed over individuals’ life course (Chen & Jensen, 1985; Butrica & Mudrazija, 2016; Mayer, 2017). In the case of reverse mortgages, homeowners, generally, utilize reverse mortgages only as a last resort (Leviton, 2002; Jester et al., 2006). An explanation for retirees’ resistance to utilize home equity is that retirees may view home equity as consumption insurance, and this viewpoint of home equity may affect asset consumption differently when compared to other

Journal of Personal Finance

assets (Piazzesi et al., 2004; Lustig and Van Nieuwerburgh, 2005; Agarwal & Qian, 2017; Bravo et al., 2019; Pearson, 2020). Retirees may recognize the need for saving for unanticipated health shocks (Fronstin & VanDerhei, 2017) or the need for savings in the event that long-term care is required in late-life (Hurd et al., 2014; Harrington et al., 2017). If these motivations exist, it suggests retirees anticipate that they will be able to access home equity when these events occur and may provide an explanation for the timing of home equity utilization.

systemic. Retirees may have limited knowledge of the available tools to unlock home equity. Thus, retirees may overlook home equity as an opportunity to improve their retirement experience. Arguments for increased efforts to promote the utilization of home equity as a part of an individual’s plan for retirement are made.

With respect to the tools available to access home equity, limited product awareness or knowledge may also offer an explanation for why retirees do not utilize their home equity to finance consumption. Research posits that even if retirees are aware of the available tools, retirees are unlikely to have knowledge of the tools’ operational terms (Davidoff et al., 2017; Dillingh et al., 2013). This lack of knowledge may translate into misconceptions about home equity access tools. Leviton (2002) shows that product misconceptions help explain the low demand for reverse mortgages.

The data utilized are collected from the 1992-2016 RAND HRS longitudinal file of the Health and Retirement Study (HRS).1 The HRS is sponsored by the National Institute on Aging (grant number NIA U01AG009740) and is conducted by the University of Michigan. This data and other information provided by the HRS are collected through survey questions and recorded responses. The HRS has a participation sample of approximately 20,000 and collects data at both the respondent and household levels. The data that are collected are published every 2 years. The HRS collects data from survey participants on a variety of topics related to health and retirement. The purpose of the data collection is to provide data for research on health and aging in the United States. The sample is limited to the subset of HRS respondents that respond ‘retired’ when asked, “Are you working now, temporarily laid off, unemployed and looking for work, disabled and unable to work, retired, a homemaker, or what?” Observations that report responses other than “retired” and missing values are dropped. There are 60,825 retirees that are studied over 12 waves (24 years).

Although retirees may be reluctant to utilize home equity for consumption, research has suggested that retirees can improve the chances that they will make it to the end of retirement in an improved fiscal position through the use of home equity. Artle and Varaiya (1978) show that relaxing liquidity constraints by borrowing against home equity in retirement can smooth late-life consumption. Strategic use of a reverse mortgage can improve retirement outcomes (Yogo, 2009; Mitchell & Piggott, 2004; Pfau, 2015; Nakajima & Telyukova, 2017). Using Monte Carlo simulations, Pfeiffer et al. (2014) show that using Home Equity Conversion Mortgage (HECM) can drastically increase the likelihood that a retiree has the financial capability to make it through retirement. Mayer & Simons (1994) show that reverse mortgages would allow over 1.4 million elderly persons to raise their incomes above the poverty line. Financing consumption through the use of home equity can be achieved through the uses of reverse mortgages, home equity lines of credit (HELOCs), refinancing existing mortgage debt balances to a longer payoff period, and cash out refinancing. Two important issues are highlighted in this study. The first issue is that, for many retirees, home equity is inefficient in promoting retiree well-being. Building home equity in late-life is counterintuitive to the consumption life-cycle stage in the consumption over the life-cycle hypothesis. The second issue is 1.

DATA

The dependent variable is retirement satisfaction. Table 1 provides the retirees’ retirement satisfaction levels. The 60,825 retirees in this study rank their level of retirement satisfaction between 1 and 3, with 3 representing ‘Very satisfied,’ 2 representing ‘Moderately satisfied,’ and 1 representing ‘Not at all satisfied.’ ‘Moderately satisfied’ and ‘Not at all satisfied’ responses are combined into one satisfaction ranking, ‘Moderately & Not at all satisfied.’ ‘Very satisfied’ and ‘Moderately & Not at all satisfied’ and have a reported response of 36,358 (59.77%) and 24,467 (40.23%), respectively. The independent variable of interest is estimated from the retirees’ ratio of home equity to net worth (HEtNW). The HEtNW is estimated by the equation below:

HEtNW =

Home Equity Net Worth

RAND HRS 2016 Fat File (E2A)]. Produced by the RAND Center for the Study of Aging, with funding from the National Institute on Aging and the Social Security Administration. Santa Monica, CA ([May 2019]).

Volume 20 • Issue 1 2021

The net worth denominator is created by summing the retirees’ total assets and subtracting the retirees’ total liabilities. The home equity numerator is created by subtracting the retirees’ mortgage debt associated with the retirees’ primary residence from the value of the retirees’ primary residence. The retirees’ average home equity is $170,817. The retirees’ average HEtNW is 51%. The level of HEtNW suggests that roughly half of the retirees’ total net worth is held as home equity. The level of HEtNW is in line with the findings from Hanewald et al. (2016), who shows that U.S. households’ primary residence equity value for those aged 65 and over comprises on average (median) 49 percent (52 percent) of their total net worth. Table 2 provides the descriptive statistics of the retirees. The variables married, white, and male are coded as “1” if respondent is married, white, and male in a given wave, respectively. A “0” is coded otherwise. Age, income, and net wealth are continuous measures. 67.2% of the retirees are married, 84.1% of the retirees are white, and 44.3% of the retirees are male. Average age is 72 and average years of education is 12.8. Average income is $52,640 and average net worth is $563,605.

METHODS To test the hypothesis, a random-effects probit regression model with a Mundlak correction is estimated on an unbalanced panel: SATit* = β0i + β1HEtNWit + βjDVit + βkmeantv(DVit) + αi + eit SATit = 0 if SATit* < μ1 (Not at all satisfied) SATit = 0 if μ1 ≤ SATit* < μ2 (Moderately satisfied ) SATit = 1 if μ2 ≤ SATit* (Very satisfied ) Where SATit* is a latent measure of retiree i’s satisfaction in wave t. The unknown thresholds, μ1and μ2, are estimated and associated with responses to the question, "All in all, would you say that your retirement has turned out to be very satisfying, moderately satisfying, or not at all satisfying?" The variable HEtNWit measures retirees’ ratio of home equity to net worth. DVit is a vector of demographic variable. have intragroup correlation. Utilizing the Mundlak approach for random-effects modelling (Mundlak, 1978), the vector 2.

meantv(DVit) is included, which is a vector of panel-level means of the time-varying co-variates.2 The Mundlak correction controls for intragroup correlation between the independent variables. β0 represents the y-intercept of the model. β1 is the coefficient associated with the explanatory variable HEtNWit. βj is a vector of coefficients associated with the DVit matrix. βk is a vector of coefficients associated with the meantv(DVit) matrix. The error term is assumed to follow the standard normal distribution. The variable HEtNWit is measured continuously. Increases in HEtNWit suggests that retirees have a higher percentage of their net worth in the form of home equity. Home equity is considered illiquid. Holding a higher percentage of net worth in the form of illiquid assets restricts access to capital to finance consumption. The lack of access to resources to finance consumption is a resource constraint. An increase in resource constraints is expected to result in disutility. Thus, HEtNWit is expected to have a negative association with SATit*. The demographic variables included are married, net worth, income, age, age2, white, education, male, and health. Married is an indicator variable coded as a “1” if the retiree is married and coded as a “0” otherwise. White is an indicator variable coded as a “1” if the retiree is white and coded as a “0” otherwise. Male is an indicator variable coded as a “1” if the retiree is male and coded as a “0” otherwise. Age, income, and net worth are continuous measures. Education is measured continuously as the number of years of education the respondent has completed. Health is a series of dummy variables that corresponds to health status as being poor, fair, good, very good, and excellent. The reference category, poor, is the health status that the other health statuses are compared against.

RESULTS The average marginal effects and robust standard errors from the random-effects probit regression are reported in Table 3. An increase in HEtNW is associated negatively with retirement satisfaction. The average marginal effect is -0.0992. This result is statistically and economically significant. Holding a larger percentage of assets in the form of home equity creates resource constraints for retirees who may need access to capital to promote a fulfilling retirement experience.

The panel-level average of the time-varying covariates are regressed on the time-varying means and covariates on SATit. A robust estimator of the variance-covariance matrix is used. There is evidence that the panel-level means are jointly zero (χ² = 1,025.3)

Journal of Personal Finance

DISCUSSION Many retirees are likely to have worked and saved for numerous years for the opportunity to enter retirement. If retirement was an expected event, many retirees likely made this transition with the expectation of having a fulfilling retirement experience. However, as noted by the regression results, a higher ratio of home equity increases retirees’ vulnerability to experiencing a less-than-optimal retirement. The findings suggest that the less-than-optimal retirement experience can be explained by the resource constraints that high concentrations of home equity imposes. Access to financial resources is paramount in replacing employment related losses that occur when transitioning into retirement. Retirees experience the loss of a sense of purpose, role identity, social circle, and steady routine when leaving employment. To facilitate the replacement of the losses that occur when entering retirement, retirement will require a reorganization of retirees’ sense of purpose, role identity, social circle, and steady routine. One way of facilitating this reorganization is through the increased participation in activities the retiree finds enjoyable. Regular participation in activities helps in promoting the establishment of a steady routine. Social participation creates social capital through network links, which may provide retirees access to a social support system throughout retirement. Participation in volunteer work provides an opportunity for retirees to maintain self-esteem and reestablish their sense of purpose (van Willigen, 2000). Home equity is illiquid. Thus, home equity eliminates retirees’ ability to utilize their saved assets to participate in enjoyable retirement experiences and establish an adequate roleidentity post-employment. Creating liquidity by unlocking home equity offers retirees the ability to consume their saved resources. If retirees are granted access to their home equity for consumption purposes, they may be able to lessen the negative byproducts associated with the transition into retirement. Thus, the responsible consumption of home equity may help in mitigating the negative consequences of transitioning from employment to retirement and be an additive factor in promoting retiree well-being.

RETIREE OPTIONS FOR CREATING LIQUIDITY WITH HOME EQUITY The individuals in this study’s sample are fully retired, and the average age is 72. Given the older age of the retirees, it is unlikely the retirees will seek employment for the purposes of labor income and will rely on non-labor income sources to finance consumption. In the absence of non-labor income sources, such as Social Security income, pension income, and annuity income, that meet the retirees’ consumption needs, retirees utilize saved assets to finance consumption during retirement. Home equity is a saved asset; however, home equity is illiquid and prevents retirees’ from utilizing this asset to promote retiree well-being. One way of unlocking home equity is to downsize from the retirees’ current home to a less-expensive home. By downsizing to a less-expensive home, retirees are able to unlock the equity in their current home. Transaction costs and the desire for residential continuity may limit downsizing as an effective solution for unlocking home equity for many retirees. Other options for retirees to consider include reverse mortgages, home equity lines of credit (HELOCs), refinancing existing mortgage debt balances to a longer payoff period, and cash out refinancing. Reverse mortgages allow individuals over the age of 62 to access home equity, either as a lump sum payment or a fixed monthly payment. An additional benefit of utilization of a reverse mortgage is the elimination of mortgage payments. The elimination of mortgage payments decreases the amount of required expenses from the retirees’ budget. The combined benefits of converting home equity into forms of cash payments and the elimination of mortgage debt greatly enhances liquidity for retirees. The major disadvantage of a reverse mortgage is that home ownership is transferred to the lending institution upon the death of the retirees. Thus, bequest motives may provide an explanation for the low demand for reverse mortgages (Rasmussen et al., 1995; Chiang & Tsai, 2016). Home equity lines of credit (HELOCs) allow retirees to “withdraw” from home equity in the form of a loan with required principal and interest payments. The major advantage of HELOCs is that retirees’ residences are not forfeited upon death, allowing any remaining home equity to pass to the beneficiaries of the retirees. HELOCs may provide benefit to retirees for largeunexpected expenses, but the required payments may result in

Volume 20 • Issue 1 2021

increased pressure on retirees’ budget constraint. The increase in the number of required payments that result from HELOC usage may produce long-term disutility, potentially making HELOCs a less-than-optimal option for retirees. Refinancing the existing mortgage debt balance of retirees can increase retiree free cash-flow by extending the repayment term of the mortgage. Extending the repayment term of the mortgage on an existing mortgage debt amount is expected to lower monthly payments. A lower monthly mortgage payment, holding income constant, increases retirees’ free cash-flow. The two primary disadvantages are the costs to refinance and the new amortization schedule of payments, as a larger percentage of the retirees’ repayment will be allocated to interest instead of principal the post-refinance. Cash out refinancing allows retirees to refinance to a debt amount that is larger than retirees’ existing mortgage balance. By refinancing to a debt amount that is larger than the existing mortgage balance, retirees are able to unlock their home equity for consumption. Assuming there are little differences in the terms of the original and new mortgage, retirees who refinance back to their original mortgage debt may be able to unlock their home equity with little effect on their budget. Potential home price appreciation may also allow retirees to access their home’s growth for consumption. However, cash out refinancing beyond the original mortgage balance may severely increase retirees’ mortgage payment amount. The increase in retirees’ mortgage payment amount may not be an optimal long-term solution for liquidity-constrained retirees.

CONCLUSION The transition into retirement is a major life course event with effects felt in many life domains. The use of saved assets can help facilitate the development of a post-retirement lifestyle, which is paramount in promoting a satisfactory retirement experience. This study illustrates that home equity is an ineffective asset in the development of a post-retirement lifestyle because of the illiquid nature of the asset. The findings suggest that a higher ratio of home equity relative to net worth creates resource constraints for retirees, resulting in a less-than-optimal retirement experience. The ensuing discussion highlights that retirees who consider responsibly utilizing home equity as an asset for consumption provides retirees with an opportunity for a more fulfilling retirement experience. The discussion also highlights the advantages and disadvantages of utilizing home equity through the uses of reverse mortgages, HELOCs, refinancing existing mortgage debt balances to a longer payoff period, and cash out refinancing. The effective use of assets during retirement promotes a more satisfactory retirement experience. Efforts going forward should highlight the positive benefits of responsibly utilizing home equity as a tool for retirement planning. Increases in the public’s awareness of the available tools to unlock home equity during retirement may emphasize the positive benefits home equity can offer during an individual’s retirement.

Journal of Personal Finance

TABLES Table 1 Cumulative Retirement Satisfaction of Retirees Frequency

Percent

Very Satisfied

36,358

59.77

Moderately & Not at All Satisfied

24,467

40.23

Total

60,825

100

Mean

Standard Deviation

$170,816.70

271,399

HEtNW

0.5091

0.2839

Age

72.2691

8.8409

Married

0.6716

0.4696

Data from the Health and Retirement Survey N = 60,825

Table 2 Descriptive Statistics of Retirees Home Equity

Income ($10k)

$52,640

8.0552

Net Worth ($10k)

$563,605

111.6958

Health

3.0875

1.0901

White

0.8414

0.3658

Male

0.4426

0.4967

12.8482

2.9967

Years of Education Data collected from the Health and Retirement Study N = 60,825

Married, White, Male are coded as “1” if respondent is married, white, and male, and a “0” otherwise.

Table 3 Random-Effects Probit Regression Results Marginal Effects

Robust Standard Errors

HEtNW

-0.0992***

0.0086

Age

0.01948***

0.0037

Age2

-0.0002***

0.0000

Married

0.0279**

0.0083

Log(Income)

0.0059*

0.0029

Wealth

0.0000

Volume 20 • Issue 1 2021

Marginal Effects

Robust Standard Errors

-0.0062

0.0077

Education

0.0017

0.0011

Male

0.0079

0.0068

Fair

0.0388***

0.0077

Good

0.0824***

0.0083

Very Good

0.1322***

0.0097

Excellent

0.1448***

0.0120

White

Health (poor as base outcome)

Data collected from the Health and Retirement Study N = 60,825 Rho = 0.5904 Wald χ² = 3,701.03 Significance is defined as follows: * significant at p < 0.05; ** significant at p < 0.01; *** significant at p < 0.001

Journal of Personal Finance

REFERENCES AARP. 2000. “Fixing to Stay: A National Survey on Housing and Home Modification Issues.” Retrieved March 12, 2020, from http://research.aarp.org/lil/home_mod.html Agarwal, S., & Qian, W. (2017). Access to home equity and consumption: Evidence from a policy experiment. Review of Economics and Statistics, 99(1), 40-52. Aguiar, M., & Hurst, E. (2013). Deconstructing life cycle expenditure. Journal of Political Economy, 121(3), 437-492. Aguila, E., Attanasio, O., & Meghir, C. (2011). Changes in consumption at retirement: evidence from panel data. Review of Economics and Statistics, 93(3), 1094-1099. Artle, R., & Varaiya, P. (1978). Life cycle consumption and homeownership. Journal of Economic Theory, 18(1), 38-58. Battistin, E., Brugiavini, A., Rettore, E., & Weber, G. (2009). The retirement consumption puzzle: evidence from a regression discontinuity approach. American Economic Review, 99(5), 2209-26. Bravo, J. M., Ayuso, M., & Holzmann, R. (2019). Making use of home equity: The potential of housing wealth to enhance retirement security. Butrica, B. A., & Mudrazija, S. (2016). Home Equity Patterns among Older American Households. Washington, DC: Urban Institute. Chiang, S. L., & Tsai, M. S. (2016). Analyzing an elder’s desire for a reverse mortgage using an economic model that considers house bequest motivation, random death time and stochastic house price. International Review of Economics & Finance, 42, 202-219. Chen, A., & Jensen, H. H. (1985). Home equity use and the life cycle hypothesis. Journal of Consumer Affairs, 19(1), 37-56. Davidoff, T. (2004). Maintenance and the Home Equity of the Elderly. Fisher Center for Real Estate and Urban Economics Paper, (03-288). Davidoff, T., Gerhard, P., & Post, T. (2017). Reverse mortgages: What homeowners (don’t) know and how it matters. Journal of Economic Behavior & Organization, 133, 151-171. Dillingh, R., Prast, H., Rossi, M., & Brancati, C. U. (2013). The psychology and economics of reverse mortgage attitudes: evidence from the Netherlands. Center for Research on Pensions and Welfare Policies. Elsinga, M., & Hoekstra, J. (2005). Homeownership and housing satisfaction. Journal of Housing and the Built Environment, 20(4), 401-424. Feinstein, J., & McFadden, D. (1989). The dynamics of housing demand by the elderly: Wealth, cash flow, and demographic effects. In The economics of aging (pp. 55-92). University of Chicago Press. Fisher, J. D., Johnson, D. S., Marchand, J., Smeeding, T. M., & Torrey, B. B. (2008). The retirement consumption conundrum: Evidence from a consumption survey. Economics Letters, 99(3), 482-485. Fronstin, P., & VanDerhei, J. (2017). Savings Medicare Beneficiaries Need for Health Expenses: Some Couples Could Need as Much as $350,000. EBRI Notes, 38(1). Haider, S. J., & Stephens Jr, M. (2007). Is there a retirement-consumption puzzle? Evidence using subjective retirement expectations. The review of economics and statistics, 89(2), 247-264. Hanewald, K., Post, T., & Sherris, M. (2016). Portfolio choice in retirement—what is the optimal home equity release product?. Journal of Risk and Insurance, 83(2), 421-446. Harrington, C., Jacobsen, F. F., Panos, J., Pollock, A., Sutaria, S., & Szebehely, M. (2017). Marketization in long-term care: a crosscountry comparison of large for-profit nursing home chains. Health services insights, 10, 1178632917710533. Hurd, M., & Rohwedder, S. (2003). The retirement-consumption puzzle: Anticipated and actual declines in spending at retirement (No. w9586). National Bureau of Economic Research.

Volume 20 • Issue 1 2021

Hurd, M. D., Michaud, P. C., & Rohwedder, S. (2014). The lifetime risk of nursing home use. In Discoveries in the Economics of Aging (pp. 81-109). University of Chicago Press. Hurst, E. (2008). The retirement of a consumption puzzle (No. w13789). National Bureau of Economic Research. Leviton, R. (2002). Reverse mortgage decision-making. Journal of Aging & Social Policy, 13(4), 1-16. Lustig, H. N., & Van Nieuwerburgh, S. G. (2005). Housing collateral, consumption insurance, and risk premia: An empirical perspective. The Journal of Finance, 60(3), 1167-1219. Mayer, C. J. (2017). Housing, Mortgages, and Retirement. Evidence and Innovation in Housing Law and Policy, 203. Mayer, C. J., & Simons, K. V. (1994). Reverse mortgages and the liquidity of housing wealth. Real Estate Economics, 22(2), 235-255. Megboluge, I. (1997). Residential real estate in the age of information technology. Housing Finance International, 11, 22-26. Mitchell, O. S., & Piggott, J. (2004). Unlocking housing equity in Japan. Journal of the Japanese and International Economies, 18(4), 466-505. Modigliani, F., & Brumberg, R. (1954). Utility analysis and the consumption function: An interpretation of cross-section data. Franco Modigliani, 1(1), 388-436. Mundlak, Y. (1978). On the pooling of time series and cross section data. Econometrica: journal of the Econometric Society, 69-85. Munnell, A. H., Soto, M., & Aubry, J. P. (2007). Do people plan to tap their home equity in retirement?. Chestnut Hill, MA: Center for Retirement Research at Boston College. Munnell, A. H. (2015). Falling short: The coming retirement crisis and what to do about it. Issues in Brief, 15-7. Nakajima, M., & Telyukova, I. A. (2017). Reverse mortgage loans: A quantitative analysis. The Journal of Finance, 72(2), 911-950. Nakajima, Makoto, and Irina A. Telyukova. "Home equity in retirement." International Economic Review 61, no. 2 (2020): 573-616. Neuwirth, P., Sacks, B. H., & Sacks, S. R. (2017). Integrating Home Equity and Retirement Savings Through the “Rule of 30.” Journal of Financial Planning, 30(10), 52-62. Olafsson, A., & Pagel, M. (2018). The retirement-consumption puzzle: New evidence from personal finances (No. w24405). National Bureau of Economic Research. Ortman, J. M., Velkoff, V. A., & Hogan, H. (2014). An aging nation: the older population in the United States (pp. 25-1140). Suitland, MD, USA: United States Census Bureau, Economics and Statistics Administration, US Department of Commerce. Pearson, B. (2020). Demographic Variations in the Perception of the Investment Services Offered by Financial Advisors. Journal of Accounting & Finance, 20(3), 127-139. Pfau, W. D. (2015). Incorporating home equity into a retirement income strategy. Available at SSRN 2685816. Shaun Pfeiffer, C Angus Schaal, & John Salter. (2014). HECM Reverse Mortgages: Now or Last Resort? Journal of Financial Planning, 27(5), 44. Piazzesi, M., Schneider, M., & Tuzel, S. (2007). Housing, consumption and asset pricing. Journal of Financial Economics, 83(3), 531-569. Rasmussen, D. W., Megbolugbe, I. F., & Morgan, B. A. (1995). Using the 1990 public use microdata sample to estimate potential demand for reverse mortgage products. Journal of Housing Research, 1-23. Sabia, J. J. (2008). There's no place like home: A hazard model analysis of aging in place among older homeowners in the PSID. Research on Aging, 30(1), 3-35. Thaler, R. H. (1990). Anomalies: Saving, fungibility, and mental accounts. Journal of economic perspectives, 4(1), 193-205. U.S. Census Bureau. 2000. “United States Census 2000.” Retrieved March 16, 2020, from http://www.census.gov/main/www/ cen2000.html

Journal of Personal Finance

U.S. Census Bureau. 2005. American Housing Survey (AHS). Retrieved March 12, 2020, from http://www.census.gov/hhes/www/ housing/ahs/ahs.html van Willigen, M. (2000). Differential benefits of volunteering across the life course. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 55(5), S308-S318. Venti, S. F., & Wise, D. A. (1989). Aging, moving, and housing wealth. In The economics of aging (pp. 9-54). University of Chicago Press. Yogo, M. (2016). Portfolio choice in retirement: Health risk and the demand for annuities, housing, and risky assets. Journal of Monetary Economics, 80, 17-34.

Volume 20 • Issue 1 2021

Investment Strategies During the Great Recession: Who Remains Calm, and Who Panics?—The Role of Financial Planners

Shan Lei, Ph.D., CFA, CFP®

Abstract This study uses a proprietary dataset to investigate factors related to investment strategies chosen in the wake of the “Great Recession” of 2007 to 2009 and discussed the role of financial planners in particular. This study finds support for a positive relationship between investors using financial planners and following a disciplined investment strategy or portfolio rebalancing strategy. Additionally, this study also finds that using financial planners reduces the likelihood of holding the losers too long in the down market which might hurt individual investors' abilities to achieve long-term financial goals. Personal characteristics, such as gender, age, race, personal saving rate, and investable assets may present an essential part in shaping individuals' investment decisions in a recession. Potential investment bias associated with each investment strategy is discussed. Suggestions are also provided to overcome or accommodate these biases.

Keywords financial planners, investment strategies, Great Recession, investment bias

Journal of Personal Finance

INTRODUCTION As the latest and most influential global economic crisis in the past 40 years (Rogoff & Chainey, 2019), the "Great Recession" of 2007 to 2009 has widespread effects on the entire economy and millions of peoples' lives. The memory is still fresh for most people. The most direct impact was a higher unemployment rate and lower family income. Data has shown that more than 6% of jobs evaporated and household annual income declined nearly $3,000 from 2007 to 2009 (The State of Working America, 2018). The S&P 500 Index lost over 56% during that period. In the 2008 stock market crash, Dow Jones experienced its biggest drops in history – 777.68 points drop in intra-day trading since the Great Depression (Amaded, 2018; NBCNews. com, 2018). At the household level, many families experienced great losses in financial portfolios (Hurd & Rohwedder, 2010). According to the Survey of Consumer Finances, the percentage of stock and business equity investment in household portfolios decreased by almost 5% during the Great Recession (Bricker, Bucks, Kennickell, Mach, & Moore, 2011). Investors responded to this adversity by making changes in their financial portfolios. First, some sold their depreciated assets immediately. Hurd and Rohwedder (2010) found that more investors reduced stock holdings in their retirement accounts in response to the unpleasant investment environment. In contrast, others maintained their strategies and took a long-term approach. The effects of the crisis has been still ongoing. According to Fidelity Investments’ "Ten Years Later" Analysis, investors’ confidence about the financial market is still shaky. Around 25% of respondents have even reported switching to a conservative investment strategy as a result of the crisis (Fidelity Investments, 2017). Contemplating the increase in individual responsibilities toward complex and volatile financial markets, seeking professional help is a plausible and efficient approach (Willis, 2008). Facing unpleasant major life changes, such as economic shock in financial turmoil, many investors are easily affected by behavioral biases during their investment decisionmaking process (Hayes, 2019). Would working with financial professionals help investors obtain a better investment strategy during the crises? Drawing on the literature regarding investment decision making and strategies, this paper uses data collected during the Great Recession of 2007 to 2009 to investigate investors’ investment strategies in the down market. Financial crises happen in cycles. When red flags were raised, learning from

those who panicked and who kept calm in the wake of the crisis could help investors identify a favorable investment strategy and prepare for the next economic turmoil. This paper also attempts to shed light on factors related to investment decision-making and strategy choice during the crisis. Moreover, the role of the financial planners in investment strategy choice has been specifically examined.

USE OF FINANCIAL ADVISORS AND INVESTMENT STRATEGIES There were many pieces of evidence in prior research that have confirmed the role of financial advisors in individual investors' investment decision making. The use of financial advisors has been found to be positively associated with investors’ risky assets ownership, such as stocks (Direr & Visser, 2013; Georgarakos & Inderst, 2014; Montford & Goldsmith, 2016). Investors using financial advisors have also been found to have a better portfolio choice, evidenced by a more diversified portfolio allocation and/or better portfolio performance. For example, using the data collected from a national operated relationship bank from April 2003 to August 2007 in Netherland, Kramer (2012) found that clients working with financial advisors hold a more diversified portfolio with "more mutual funds, more index funds, less domestic equity, more asset classes, and more common equity positions". In a similar vein, using a proprietary dataset, Kramer and Lensink (2012) examined 5,500 Dutch households’ stock portfolios. Their results revealed that using financial advisors helped consumers improve their portfolio decision-making, evidenced by higher portfolio returns compared to self-advised participants. Lei and Yao (2016) echoed with prior research findings by examining the 2013 Survey of Consumer Finances. They found that households who used financial planners to assist their investment decision-making were more likely to have better projected households’ portfolio performance, measured by Sharpe Ratio. Further, prior research provided evidence of the long-term effect of using financial advisors on investors’ behavior and decision making. Investors who worked with financial advisors were found to focus on long-term retirement goals (Marsden, Zick, & Mayer, 2011) and were more likely to commit to the long-term investment goals even in a financial crisis (Winchester, Huston, & Finke, 2011). What was not well developed was the role of financial advisors in shaping investment decisions and choosing investment strategies in the crisis. It was worthwhile to be noted that

Volume 20 • Issue 1 2021

these few empirical studies showed that the use of financial advisors contributed to the investors' investment decision making, evidence by “maintaining an optimal portfolio composition’ during the Great Recession (Silverblatt, 2010). The investors who worked with financial planners presented “positive behavior responses” in the down economic market. They did not decrease their risky investment and kept saving regularly (Marsden, Zick, & Mayer, 2011).

INVESTORS’ CHARACTERISTICS AND INVESTMENT DECISIONS/STRATEGY CHOICE Prior research has demonstrated a positive relationship between wealth, income, and investment decisions/strategy choice. Consider the work of Calvet, Campbell, and Sodini (2009). They found that Swedish investors adjusted their portfolios based on personal returns rather than market returns, as evidenced by the rare adjustment of portfolios to hold riskier assets in bear markets. Specifically, households with higher income and more wealth were more likely to enter, and less likely to exit, a risky financial market during a bear market. In another recent study, Guillemette, Blanchett, and Finke (2018) found evidence to suggest that investors with higher wealth levels were more likely to follow the predetermined long-term investment strategies. A set of other investors’ characteristics was also found to be related to investment decisions/strategy choice in a down market, such as age, gender, and educational achievement. By examining 116,471 individual trades on the Taiwan Stock Exchange Electronic Sector Futures between January 2003 and December 2007, Cheng, Lee, and Lin (2013) concurred with previous studies that such factors as age, gender, trading tenure, and market conditions were related to the investment strategy of doing nothing in the hope that the return would revert to the normal level. For example, they found that women were more likely to hold onto a loss for too long, while men were more likely to execute a more active strategy evidenced by a larger trading volume (Barber & Odean, 2000). More recently, by analyzing transaction data in the Turkish stock market in 2011 during the global financial crisis period, Tekcea, Ylmaza, and Bildikb (2016) found that female, older, and wealthier Turkish stock investors were more likely to have a “do nothing” strategy in the down market.

By examining the SAVE data set, focusing on German households from 2007 to 2009, a global financial crisis period, Bucher-Koenen and Ziegelmeyer (2011) found evidence to suggest that financial education was related to investor decision-making for portfolio rebalancing during the financial crisis. They concluded that financially unsophisticated households in Germany were more likely to realize the loss. However, selling depreciated assets during the crisis was more detrimental to their financial wealth in the long run. When people saw their portfolios losing values in the crisis, some people held the losers too long on regret, some sold the losers too quickly on fear, while those who believed they could take advantage of the down market saw the low prices as a buying opportunity. Would working with financial planners help individual investors develop an investment strategy in line with their predetermined investment policies instead of out of their unwarranted faith in their intuitive judgments? This study distinguished itself in two ways from related research. First, different from many prior studies that used a broad definition of finance advisors (Bluethgen et al., 2008; Gerhardt & Hackethal, 2009; Kramer, 2012; Lusardi & Mitchell, 2011), this study examined the role of a specific group of financial advisors: financial planners. Second, this study used a dataset collected during the "Great Recession", which allowed it to examine the role of financial planners in investors' decision making during the down market.

CONCEPTUAL FRAMEWORK Expected Utility Theory has suggested that rational individuals would make a decision that maximized their expected utility under risk (Von Neumann, 1944). In terms of whether using a financial planner to help with their investment decision making, it mainly depends on whether this decision could help increase their marginal expected utility net of cost (Lei & Yao, 2016). Information Economics has indicated that information is a very special good (Rothschild & Stiglitz, 1976; Spence, 1974; Stigler, 1961), which requires expertise and experience, to increase the search and use efficiency (Bakos & Brynjolfsson, 1999). Compared to individual investors, professional financial planners have a set of advantages, such as “solid financial education as well as an information lead over the customers due to dealing with financial markets extensively on a daily basis” (Fischer, & Gerhardt, 2007, pp19). Thus, as one of the most important information sources for the consumers, financial planners could

Journal of Personal Finance

benefit investors’ investment decision-making process through processing and integrating information more efficiently, known as information bundling (Yao & Lei, 2018). Based on the above-mentioned conceptual framework and prior literature, it is hypothesized that the use of financial planners can benefit investors in choosing a better investment strategy and/or avoiding an investment strategy that might have potential long-term negative consequences.

METHODOLOGY Data This study used data taken from the 2008 FPA-Ameriprise Financial Value of Financial Planning Research Study (FPAAmeriprise Study) to analyze the disparity in investment strategies during the Great Recession in the United States. The data was collected by an independent market research firm between June 27, 2008, and July 18, 2008, through an online survey randomly during the Great Recession, which was considered to last from December 2007 to June 2009 (NBER, 2010). This study used weights provided by the FPA- Ameriprise Study to adjust for the oversampling of wealthy respondents and respondents' online propensity (Financial Planning Association, 2014), which followed the prior research practice using such datasets (Lei & Yao, 2015). This dataset contained a large array of information on respondents’ demographics, financial status, and their expectations and attitudes. It asked the respondents whether using financial planning services as well as recorded the respondents' replies regarding investment strategies adopted in the face of the recession. The final sample size in this study was 3,022.

Models In the survey, the respondents were asked: “How would you best characterize your own financial approach during down market conditions 1) actively advantage of down market conditions; 2) sit on sidelines, not invest new money and try to ride it out until it is over; 3) try to minimize the hit to my portfolio with some investment changes; 4) stay the course, continuing to invest at the same rate as I did prior to the market downturn?" The survey was conducted during the financial crisis, so it could be inferred that the survey question would like to understand if and how the economic shock would change investors' predetermined investment strategies. Since

the four responses were nominal and presented no obvious order, multinomial logistic regression was used to comprehend investment strategy choice while Strategy 4 was used as the reference strategy in the regression. Specifically, the survey asked respondents if they were working with financial planners. The survey described the definition of financial planners to help respondents identify and confirm the financial service they received was financial planning service, which might include investment planning, retirement planning, education planning, financial management, risk management, tax planning, estate planning, or a combination of two or more types of above-mentioned services. In the model, household income, household investable assets as well as business ownership were included (Calvet, Campbell, & Sodini, 2009; Yao & Lei, 2016) to account for the respondents’ financial and economic status. A set of individual demographics was also included to control for age, education (including financial education), race, and gender (Fernandes, Lynch Jr, & Netemeyer, 2014; Murendo & Mutsonziwa, 2017). This study also controlled for such psychological factors as confidence level about personal financial future.

RESULTS Descriptive Statistics Table 1 showed the coding of each variable used in the analysis. Table 2 presented the resulting descriptive statistics for each variable for the entire sample as well as the sample characteristics organized by the four different investment strategies during down market conditions. On average, nearly 14% of the respondents followed Strategy 1 to take the initiative in the down market. Nearly 19% of the respondents chose to do nothing, noted as Strategy 2 in this study. 23% of the respondents made some investment changes to minimize loss during the recession, noted as Strategy 3 while around 44% of the respondents chose Strategy 4 to maintain their existing investment policies. A majority of the respondents (97.9%) in the sample had household income from $50,000 to $250,000 and had investable assets less than $500,000 (89.7%). Across the sample, nearly 10% of respondents reported having no savings at all. Additionally, many respondents were equipped with an emergency fund (72.1%) and used financial planning services (61.4%). In terms of the self-reported confidence about the personal financial future, respondents indicated being very/fairly (60.7%) or extremely

Volume 20 • Issue 1 2021

(10.4%) confident. The sample as a whole was skewed towards higher levels of educational attainment. Specifically, nearly 45% of respondents reported having a college degree or higher. Further, respondents claimed to understand financial knowledge very well in the sample. It was intriguing to find that respondents in the highest income group (household income greater than $250,000) turned out to be the largest group (28.9%) among those who followed Strategy 1. A similar pattern was found in terms of household investable assets, as evidenced by the results that respondents in the highest investable assets group (household investable assets greater than $1,000,000) were the largest group who executed active strategy (Strategy 1). Notably, the respondents in the lowest investable assets group (household investable assets less than $25,000) were the largest group who executed the "do nothing" strategy (Strategy 2). This finding was consistent with prior research that wealthy investors were more likely to enter into the down market while less wealthy investors were more likely to hold on to losers (Calvet, Campbell, & Sodini, 2009). More problematically, most respondents who reported having no savings at all claimed to follow Strategy 2. These investors who saved more than 10% of their annual income made the most use of Strategy 1. In general, Strategy 3 was chosen the least by the youngest respondents (younger than 35) and was chosen the most by respondents aged from 55 to 64. Female respondents tended to follow Strategy 2. However, Strategy 2 was found to be the most followed strategy among those with the lowest educational attainment. Interestingly, respondents who claimed to understand financial knowledge well more often decided to follow the more aggressive Strategy 1.

Multivariate Analysis A multicollinearity test was conducted to identify possible significant multicollinearity between independent variables. Variance inflation factor (VIF) statistics for all the independent variables were less than 10, indicating that multicollinearity was not a concern (Allison, 2012; Freund & Wilson, 1998). Findings from the multivariate analysis identified a negative correlation between using financial planning services and less favorable investment strategy, which confirmed the hypothesis. For example, compared to strategy 4, respondents who used financial planning services were found to be less likely to follow Strategy 2 (relative risk ratio= 0.691). Interestingly, it also has been noted that investors using financial planning services were more likely to use Strategy 3 over Strategy 4 (relative risk

ratio= 1.949). Table 3 also showed that wealthier respondents were more likely to choose Strategy 1 or 3 than Strategy 4. For example, for respondents with more than 1 million investable assets, the relative risk for choosing Strategy 1 in a down market relative to Strategy 4 would be expected to increase by a factor of 1.1 given the other variables in the model were held constant, compared to those with less than $25,000 investable assets. Additionally, compared to the respondents with no savings, savers preferred to use Strategy 4 over Strategy 2 or 3. While for respondents who saved between 8% to 10% of their annual after-tax income, the relative risk for choosing Strategy 1 vs. 4 would be expected to decrease by a factor of nearly 50%, respondents with a personal annual savings rate between 8% to 10% and greater than 10% showed a less tendency to use Strategy 3. Further, respondents with emergency funds were found to be more likely to follow Strategy 1 vs. Strategy 4 (relative risk ratio=1.541) but less likely to follow Strategy 2 vs. Strategy 4 (relative risk ratio=0.685), while business owners were more likely to choose Strategy 1 vs. Strategy 4 (relative risk ratio=1.545). Generally, compared to the youngest respondent (<35), older respondents were less likely to follow Strategy 1 or 2 over Strategy 4. The results also showed that much older respondents (> 55) tended to choose Strategy 3 over Strategy 4. Further, female respondents were found to be less likely to follow Strategy 1 over Strategy 4 under down market conditions but showed a significantly greater tendency to choose Strategy 2 over Strategy 4. Hispanic respondents were also found to be more likely to choose Strategy 2 than Strategy 4, compared to the White cohorts. Consistent with prior literature (Barber & Odean, 2001; Kuo & Lin, 2013), findings from this study also illustrated that investors with higher confidence levels preferred an active investment strategy. For example, for respondents who claimed to be extremely confident about their financial futures, relative to Strategy 4, the relative risk for choosing Strategy 1 was expected to increase by a factor of 2.2. Meanwhile, respondents who claimed to understand financial knowledge well also tended to choose Strategy 1 over Strategy 4 when facing unfavorable market conditions.

Further Analysis of the Impact of Use of Financial Planners Contemplating the possibility of the use of financial planners might be endogenous, a propensity matching scoring technique was used to attempt to address the selection bias

Journal of Personal Finance

following previous research (Kim, Pak, Shin, & Hanna, 2018; Robinson & Sanderford, 2016). First, PROC LOGISTIC options in the SAS software was used to calculate and estimate the predicted probability (i.e. the propensity score) of using financial planners for each observation in the data set through a Probit regression of use of financial planners on its related factors (Cummings, & James III, 2014; Elmerick, Montalto & Fox, 2002; Finke, Huston, & Winchester, 2011). Propensity score matching methods allowed the same probability of choosing to use financial planners between the group who used financial planners and the group who did not, thus any difference in investment strategies can be attributed to the effect of using financial planners. When generating propensity scores, the same set of variables was included as in the multinomial logistic regression (shown in Table 3). Then, a separate multinomial logistic regression analysis was conducted on the matched sample. The regression estimates (relative risk ratio=0.72 and 2.0 respectively for Strategy 2 and 3, shown in Table 4) carried the same sign and similar magnitudes in the baseline multinomial logistic regression (relative risk ratio=0.69 and 1.9 respectively for Strategy 2 and 3, shown in Table 3), confirming the positive role of the use of the financial planners in helping investors with better investment decisionmaking in this balanced sample, which shared the similar personal features in the investors who used financial planning service and those who did not.

DISCUSSION AND IMPLICATION Results from this study showed that the personal characteristics of investors could partially describe investment strategy decision-making during a recession. As shown in Table 2, compared to the investors who used other investment strategies during the down market condition, those who held a loser in a down market had lower income, fewer investable assets, with zero or lower saving rate and no emergency fund. They were also disproportionately female and had the lowest educational achievement. On the contrary, those who were more aggressive in their investment were most confident about their financial future and claimed to understand financial issues well. They were mostly business owners, had the highest annual savings rates, the highest income, and investable assets. This study added to the existing literature by discussing the factors regarding investors’ decision-making in investment strategy during down market conditions. Four different investment strategies were examined. Strategy 4 was to

maintain the predetermined investment policy while the remaining three strategies had changes in different ways and levels. Strategy 1 was to actively take advantage of the down market conditions. Strategy 2 was to stop all the investments, known as the "do nothing" strategy. Strategy 3 was to make some minor changes to try to minimize the loss. While most people were fearful in a crisis, investors conducting Strategy 1 (actively take advantage of the down market conditions) were investing in a crisis, which was risky since timing to identify the bottom of the crisis required professional knowledge and luck as well (Haynes, 2019). If investors overestimated their knowledge, abilities, and luck, they were very likely to have overconfidence bias. Considering the lower performance associated with overconfidence bias (Barber & Odean, 2000) and its effect on the volatility of stock market returns (Abbes, 2013), investors who were found to be more likely to exhibit this bias (such as men) (Barber & Odean, 2001) should consider applying more prudence in their investment strategies. It might not be easy to overcome such emotional biases as overconfidence bias in the investment decision-making, however, professional experts recommended that investors should understand that good investment performance was a probabilistic activity. Making a detailed record of trading and decision-making could help investors to be aware of potential biases. Seeking advice from professionals was also suggested to help identify biases by exploring the decision-making process from an outside perspective (Pompian, 2017, p.50). Results indicated that investors who followed Strategy 3 changed their current strategy to minimize loss. The description of Strategy 3 in the survey made it unclear how investors reached this goal. Investors might have sold losers very soon, which was a possible sign of myopic loss aversion. It was also possible that investors chose a portfolio rebalance strategy to help maintain their target portfolio allocation or sold the losers to avoid further losses. Investors working with financial planners were found to be more likely to try to minimize the loss (Strategy 3) rather than following the “do nothing” strategy (Strategy 2)(in a separate regression using Strategy 2 as a reference, relative risk ratio =2.44, p<0.001). If it was the case that these investors sold losers too soon, one possible explanation was that making decisions for others did not help with eliminating loss aversion in individual investors' investment decision-making process (Vieider, Villegas-Palacio, Martinsson & Mejia, 2016). These findings indicated that financial professionals should be aware of and discuss their

Volume 20 • Issue 1 2021

investment biases when making decisions for their clients. If it was the case that investors were following a portfolio rebalance strategy to help minimize the effect of the loss, this strategy could be considered as a preferred investment strategy in a turbulent market over a chasing strategy (Beach & Rose, 2005; Bianchi, 2018). Practical evidence also showed that good financial professionals always employed the portfolio rebalancing strategy out of long-term perspectives, which was consistent with clients' long-term goals (Crandall, 2020; Friedburg, 2018). The findings in this study might indicate that financial professionals were helping their clients with a more favorable investment strategy, evidenced by the result that working with financial planners increased the probability of choosing Strategy 3 over 4. As academic research and financial advice suggested, a disciplined investment strategy (Strategy 4) could help keep investors on track for their long-term financial goals (Beach & Rose, 2005) and minimize the impact of the investment biases (Chaudhary, 2013). Fidelity Investments "Ten Years Later" Analysis confirmed this long-term strategy by showing that those investors who continued to follow their investment policies in their 401(k) plans and IRAs had significantly higher account balances compared to the amount during the crisis (Fidelity Investments, 2017). As evidenced in this study, investors who had better saving habits and worked with financial planners are found to maintain their current strategy during times of recession rather than changing to a "do nothing" strategy. Unfortunately, this did not hold for all demographics. Women and Hispanic investors were less likely to choose Strategy 4. Instead, they chose to stop the investment and did nothing facing the loss. The descriptive statistics in this study also

echoed previous research (Fernandes, Lynch Jr, & Netemeyer, 2014; Gaudecker, 2015; Lusardi & Mitchell, 2011) that lower education levels were associated with the adoption of Strategy 2 under down market conditions (shown in Table 2). Financial planners should provide more education to women, and Hispanic clients in following a disciplined investment strategy or appropriate portfolio rebalance strategy, especially when they were confronted with unpleasant life or economic changes. The use of financial planning services might assist investors with better decision-making as evidenced by some empirical research findings (Lei & Yao, 2016; Kramer, 2012) as well as confirmed in this study. Working with clients to design and implement a disciplined approach and a portfolio rebalancing strategy might help guide them to make a rational decision in the face of losses. A timely and detailed review of the reasoning behind the trading for investors who want to take advantage of the down market would demonstrate any false confidence. Globalization made the economy in every country interrelated more than before. A direction for further research would be to collect data from diverse datasets from various countries to examine the behaviors directly linked to some behavior biases. Though it was expected that investors in different economic environments might share some common investment behaviors, different levels of development and financial systems might play a role or be considered as a constraint in investors' decision-making process. With financial crises growing more "contagious" in recent years, financial professionals and policymakers needed to create programs or procedures to help investors choose rational investment strategies when facing the crisis.

Journal of Personal Finance

TABLES Table 1 Summary of the Variables Used in the Empirical Model Variable Name

Variable Type

Measure type

Variable description

Investment Strategy

Dependent Variable

4-category Nominal

Strategy 1, Strategy 2, Strategy 3, Strategy 4 [reference]

Use financial planning service

Independent Variable

Dichotomous

Yes=1; No=0

Household income

Independent Variable

5-level Categorical

< $50,000[reference], $50,000-99,999; $100,000-149,999; $150,000-249,999; ≥ $250,000

Household investable assets

Independent Variable

5-level Categorical

< $25,000 [reference], $25,000-$99,999, $100,000-$499,999, $500,000-$999,999; ≥ $1,000,000

Personal annual saving rate

Independent Variable

5-level Categorical

No saving [reference], 1%-3%, 4%-7%, 8%-10%, > 10%

Emergency fund

Independent Variable

Dichotomous

Yes=1; No=0

Business owner

Independent Variable

Dichotomous

Yes=1; No=0

Age

Independent Variable

5-level Categorical

Younger than 35 years old [reference], 35 to 44 years old, 45 to 54 years old, 55 to 64 years old, 65 years and older

Gender

Independent Variable

2-level Categorical

Male [reference], female

Education

Independent Variable

3-level Categorical

High school/GED or less [reference], Some college or associate, Bachelor's degree or higher

Race

Independent Variable

4-level Categorical

Non-Hispanic white [reference], Black/ African-American, Hispanic, and other

Financial knowledge

Independent Variable

3-level Categorical

Agree; disagree; nor agree nor disagree [reference] with the following statement: I understand financial related issues

Confidence about personal financial future

Independent Variable

3-level Categorical

Not at all confident [reference], somewhat confident, fairly/very confident, extremely confident

Volume 20 • Issue 1 2021

Table 2 Demographic Profile Variables

All (N=3,022)

Investment Strategy

Strategy 1 Strategy Strategy (N=417) 2 (N=571) 3 (N=694)

Strategy 4 (N=1,340)

13.8

18.9

23.0

44.3

61.4

64.5

47.1

77.5

58.2

2.1

1.5

3.4

2.7

1.5

Household income: $50,000-99,999

33.2

25.7

42.2

29.6

33.6

Household income: $100,000-149,999

25.5

22.5

25.9

26.0

Household income: $150,000-249,999

20.8

21.5

16.8

22.4

21.6

Household income: ≥ $250,000

18.3

28.9

11.8

19.4

17.3

Household investable assets: < $25,000

33.8

11.7

31.2

11.5

22.1

Household investable assets: $25,000-99,999

31.1

13.5

13.4

12.3

17.7

Household investable assets: $100,000-499,999

24.8

21.2

20.7

24.7

23.7

Household investable assets: $500,000-999,999

5.8

19.0

14.5

19.1

15.5

Household investable assets: ≥ $1,000,000

4.6

34.7

20.2

32.4

21.1

Personal annual saving rate: no saving

9.7

6.5

20.5

10.1

6.0

Use financial planning service Financial Situations Household income: < $50,000

Personal annual saving rate: 1%-3%

12.4

7.0

20.0

11.0

11.6

Personal annual saving rate: 4%-7%

19.5

17.0

19.3

21.5

19.3

Personal annual saving rate: 8%-10%

21.1

17.0

22.3

23.4

Personal annual saving rate: > 10%

37.3

52.5

23.3

35.2

39.7

Emergency fund

72.1

83.2

54.8

78.5

72.6

Business owner

19.4

27.3

18.0

20.6

16.8

14.7

17.0

18.7

7.4

16.0

Demographics Age: younger than 35 years old Age: 35 to 44 years old

13.8

15.1

12.1

12.0

15.1

Age: 45 to 54 years old

21.9

22.8

17.7

20.9

23.9

Age: 55 to 64 years old

33.3

30.7

29.3

40.8

32.0

Age: 65 years and older

16.4

14.4

22.2

19.0

13.1

Female

41.5

28.5

48.3

44.8

41.0

4.5

2.6

6.5

4.5

4.3

Educ: some college or associate

50.7

50.4

52.7

51.7

49.3

Educ: college degree or higher

44.8

47.0

40.8

43.8

46.3

Race: White non-Hispanic

Educ: high school diploma/GED or less

86.1

81.8

85.5

87.3

87.0

Race: Black/African-American

2.2

2.9

1.8

1.4

2.5

Race: Hispanic

4.3

5.5

5.3

3.9

3.7

Race: Other races

7.5

9.8

7.5

7.4

6.8

9.7

5.0

18.2

7.4

8.8

17.6

7.4

23.5

16.0

19.0

Respondents' expectations Financial knowledge: I don’t understand Fin: neither agree nor disagree

Journal of Personal Finance

Variables

All (N=3,022)

Fin: I understand Confidence about personal financial future: not at all

Strategy 1 Strategy Strategy (N=417) 2 (N=571) 3 (N=694)

Strategy 4 (N=1,340)

72.7

87.5

58.3

76.7

72.2

6.1

2.2

14.4

4.9

4.4

Confidence: somewhat

22.9

12.7

32.6

26.2

20.2

Confidence: fairly/very

60.7

61.6

47.8

60.8

65.8

Confidence: extremely

10.4

23.5

5.3

8.1

9.6

Note: Mean percentages are reported. Percentages may not add up to 100% due to rounding.

Table 3 Multinomial Logistic Model: factors related to investment strategy decision making Strategy 4 is the reference strategy in the Multinomial Logistic Model Strategy 1 Variables

Coef

Std.err

Intercept

-2.372***

0.740

Coef

Std.err

1.906***

0.492

-0.081

0.134

0.922

-0.369**

0.125

$50,000-99,999

-0.058

0.499

0.943

-0.291

$100,000-149,999

-0.134

0.501

0.874

$150,000-249,999

-0.123

0.504

≥ $250,000

0.078

0.507

$25,000-$99,999

0.258

0.232

$100,000-$499,999

0.378

$500,000-$999,999 ≥ $1,000,000

Relative Risk Ratio

Strategy 3 Coef

Std.err.

-1.215*

0.525

0.691

0.667***

0.122

1.949

0.361

0.747

-0.114

0.355

0.893

-0.281

0.364

0.755

-0.061

0.357

0.941

0.884

-0.484

0.373

0.616

-0.208

0.362

0.812

1.081

-0.582

0.387

0.559

-0.177

0.369

0.838

1.294

-0.092

0.189

0.912

0.282

0.195

1.326

0.224

1.459

0.178

0.181

1.195

0.539**

0.182

1.714

0.729**

0.239

2.074

0.207

0.208

1.230

0.540**

0.199

1.716

0.765**

0.242

2.149

0.473

0.211

1.604

0.848***

0.199

2.335

1%-3%

-0.485

0.319

0.616

-0.721***

0.213

0.486

-0.401

0.233

0.670

4%-7%

-0.281

0.281

0.755

-1.007***

0.206

0.365

-0.369

0.212

0.691

8%-10%

-0.645*

0.282

0.525

-1.062***

0.211

0.346

-0.547**

0.212

0.579

More than 10%

-0.259

0.263

0.772

-1.238***

0.203

0.290

-0.682***

0.204

0.505

0.432**

0.162

1.541

-0.378**

0.125

0.685

0.214

0.126

1.238

Use financial planning service

Relative Risk Ratio

Strategy 2

Relative Risk Ratio

Financial Situations Household income (ref: < $50,000)

Household investable assets (ref: < < $25,000)

Personal annual saving rate (ref: no saving)

Emergency fund

Volume 20 • Issue 1 2021

Strategy 1 Variables

Strategy 2

Strategy 3

Coef

Std.err

Relative Risk Ratio

Coef

Std.err

Relative Risk Ratio

Coef

Std.err.

Relative Risk Ratio

0.435**

0.143

1.545

0.279

0.147

1.322

0.100

0.129

1.105

35 to 44 years old

-0.286

0.216

0.752

-0.493*

0.207

0.611

0.374

0.220

1.453

45 to 54 years old

-0.419*

0.199

0.658

-0.527**

0.191

0.591

0.272

0.205

1.313

55 to 64 years old

-0.550**

0.193

0.577

-0.186

0.179

0.830

0.569**

0.196

1.767

Business owner Demographics Age: (ref: younger than 35 years old)

65 years and older

-0.581*

0.230

0.559

0.432*

0.203

1.540

0.669**

0.219

1.952

-0.420**

0.130

0.657

0.250*

0.114

1.284

0.116

0.103

1.123

some college or associate

0.441

0.361

1.554

-0.219

0.256

0.803

-0.102

0.251

0.903

college degree or higher

0.399

0.362

1.490

-0.366

0.259

0.694

-0.172

0.253

0.842

Black/AfricanAmerican

0.370

0.436

1.448

-0.506

0.522

0.603

-0.581

0.507

0.559

Hispanic

0.310

0.278

1.363

0.663*

0.267

1.941

0.083

0.261

1.086

Other races

0.312

0.500

1.366

-0.408

0.662

0.665

-0.569

0.653

0.566

0.477

0.324

1.611

0.136

0.198

1.146

-0.062

0.221

0.940

0.825***

0.217

2.282

-0.277*

0.143

0.758

0.230

0.142

1.259

Confidence: somewhat

-0.038

0.404

0.963

-0.365

0.221

0.694

-0.046

0.257

0.955

Confidence: fairly/ very

0.088

0.387

1.092

-1.039***

0.220

0.354

-0.583*

0.252

0.558

Confidence: extremely

0.796*

0.413

2.217

-1.314***

0.311

0.269

-0.954***

0.304

0.385

Female Educ: (ref: high school diploma/GED or less)

Race: (ref: White nonHispanic)

Respondents' expectations Financial knowledge: (ref: not agree nor disagree) Fin: disagree Fin: agree Confidence about personal financial future: (ref: not at all)

Likelihood Ratio Test: Chi-square=658.88, p<0.001 *** p<0.001, ** p<0.01, * p<0.05

Journal of Personal Finance

Table 4 Multinomial Logistic Model: factors related to investment strategy decision making – Matched Sample Strategy 4 is the reference strategy in the Multinomial Logistic Model Strategy 1 Variables

Coef

Std.err

Intercept

-1.492

0.879

Use financial planning service

-0.018

0.157

$50,000-99,999

-0.650

$100,000-149,999

Relative Risk Ratio

Strategy 2 Coef

Std.err

1.356*

0.677

0.983

-0.327*

0.150

0.576

0.522

-0.192

-0.490

0.578

0.612

$150,000-249,999

-0.568

0.584

≥ $250,000

-0.229

$25,000-$99,999

Relative Risk Ratio

Strategy 3 Coef

Std.err. Relative Risk Ratio

-1.303

0.722

0.721

0.716***

0.138

2.046

0.514

0.825

-0.232

0.525

0.793

0.045

0.519

1.046

-0.213

0.530

0.808

0.567

-0.218

0.531

0.804

-0.229

0.535

0.795

0.590

0.795

-0.423

0.553

0.655

0.021

0.543

1.021

0.294

0.266

1.342

-0.014

0.222

0.986

0.057

0.215

1.058

$100,000-$499,999

0.394

0.264

1.483

0.130

0.222

1.138

0.366

0.208

1.442

$500,000-$999,999

0.669*

0.309

1.952

0.250

0.279

1.283

0.300

0.262

1.350

≥ $1,000,000

0.759*

0.317

2.135

0.566

0.296

1.761

0.788**

0.266

2.200

1%-3%

-1.066*

0.437

0.344

-0.667*

0.304

0.513

-0.631

0.341

0.532

4%-7%

-0.468

0.376

0.627

-0.993***

0.298

0.371

-0.335

0.316

0.716

8%-10%

-1.119**

0.386

0.327

-1.155***

0.305

0.315

-0.703*

0.320

0.495

More than 10%

-0.668

0.359

0.513

-1.357***

0.293

0.258

-0.817**

0.310

0.442

Emergency fund

0.323

0.202

1.381

-0.425**

0.162

0.654

0.178

0.161

1.194

Business owner

0.612*

0.193

1.844

0.197

0.202

1.217

0.156

0.182

1.168

35 to 44 years old

-0.460

0.273

0.631

-0.485

0.272

0.616

0.253

0.263

1.288

45 to 54 years old

-0.535*

0.245

0.586

-0.545*

0.249

0.580

0.185

0.243

1.203

55 to 64 years old

-0.367

0.233

0.693

0.028

0.226

1.028

0.619**

0.231

1.857

65 years and older

-0.623*

0.295

0.536

0.620*

0.256

1.859

0.657*

0.267

1.929

-0.480**

0.183

0.619

0.421**

0.155

1.523

0.237

0.144

1.268

Financial Situations Household income (ref: < $50,000)

Household investable assets (ref: < < $25,000)

Personal annual saving rate (ref: no saving)

Demographics Age: (ref: younger than 35 years old)

Female Educ: (ref: high school diploma/GED or less)

Volume 20 • Issue 1 2021

Strategy 1 Variables

Strategy 2

Strategy 3

Coef

Std.err

Relative Risk Ratio

Coef

Std.err

Relative Risk Ratio

Coef

Std.err. Relative Risk Ratio

some college or associate

0.433

0.461

1.541

-0.259

0.322

0.772

0.167

0.349

1.182

college degree or higher

0.418

0.465

1.519

-0.358

0.329

0.699

0.092

0.354

1.096

Black/AfricanAmerican

0.546

0.522

1.727

-0.692

0.776

0.501

-0.294

0.585

0.745

Hispanic

0.588

0.382

1.800

0.385

0.406

1.469

0.432

0.369

1.540

Other races

1.159

0.700

3.188

0.058

0.901

1.060

-0.401

1.116

0.670

Fin: disagree

0.548

0.396

1.730

-0.011

0.270

0.990

0.040

0.293

1.041

Fin: agree

0.694**

0.266

2.001

-0.276

0.187

0.759

0.321

0.190

1.379

Confidence: somewhat

0.155

0.501

1.167

-0.185

0.309

0.831

0.081

0.326

1.085

Confidence: fairly/very

0.068

0.483

1.070

-0.778*

0.305

0.459

-0.551

0.319

0.577

Confidence: extremely

1.021*

0.520

2.776

-0.828*

0.417

0.437

-0.752

0.399

0.471

Race: (ref: White nonHispanic)

Respondents' expectations Financial knowledge: (ref: not agree nor disagree)

Confidence about personal financial future: (ref: not at all)

Likelihood Ratio Test: Chi-square=363.38, p<0.001 *** p<0.001, ** p<0.01, * p<0.05

Journal of Personal Finance

REFERENCES Abbes, M. B. (2013). Does overconfidence bias explain volatility during the global financial crisis? Transition Studies Review, 19(3), 291-312. Allison, Paul D. Logistic regression using SAS: Theory and application. SAS Institute, 2012. Amaded, Kimberly. (2018, October 13). Dow Highest Closing Records. Retrieved from https://www.thebalance.com/dow-jonesclosing-history-top-highs-and-lows-since-1929-3306174 Barber, B. M., & Odean, T. (2000). Trading is hazardous to your wealth: The common stock investment performance of individual investors. The Journal of Finance, 55(2), 773-806. Barber, B. M., & Odean, T. (2001). Boys will be boys: Gender, overconfidence, and common stock investment. The quarterly journal of economics, 116(1), 261-292. Beach, S. L., & Rose, C. C. (2005). Does portfolio rebalancing help investors avoid common mistakes? Journal of Financial Planning, 18(5), 56. Bianchi, M. (2018). Financial literacy and portfolio dynamics. The Journal of Finance, 73(2), 831-859. Bluethgen, R., Gintschel, A., Hackethal, A., & Mueller, A. (2008). Financial advice and individual investors' portfolios. Available at SSRN 968197. Bricker, J., Bucks, B. K., Kennickell, A., Mach, T. L., & Moore, K. (2011). Drowning or weathering the storm? Changes in family finances from 2007 to 2009 (No. w16985). National Bureau of Economic Research. Retrieved from https://www.nber.org/papers/w16985.pdf Bucher-Koenen, T., & Ziegelmeyer, M. (2011). Who lost the most? Financial literacy, cognitive abilities, and the financial crisis. Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1761512 Calvet, L. E., Campbell, J. Y., & Sodini, P. (2009). Fight or flight? Portfolio rebalancing by individual investors. The Quarterly journal of economics, 124(1), 301-348. Chaudhary, A. K. (2013). Impact of behavioral finance in investment decisions and strategies–a fresh approach. International Journal of Management Research and Business Strategy, 2(2), 85-92. Cheng, T. Y., Lee, C. I., & Lin, C. H. (2013). An examination of the relationship between the disposition effect and gender, age, the traded security, and bull–bear market conditions. Journal of Empirical Finance, 21, 195-213. Crandall. (2020). Why it's important to rebalance your portfolio. Retrived from https://www.ameriprise.com/research-marketinsights/financial-articles/investing/why-rebalance-your-portfolio/. Direr, A., & Visser, M. (2013). Portfolio choice and financial advice. Finance, 34(2), 35-64. Elmerick, S. A., Montalto, C. P., & Fox, J. J. (2002). Use of financial planners by US households. Financial Services Review, 11(3), 217. Fernandes, D., Lynch Jr, J. G., & Netemeyer, R. G. (2014). Financial literacy, financial education, and downstream financial behaviors. Management Science, 60(8), 1861-1883. Fidelity Investments. (2017, October 26). Lessons Learned 10 Years After the Global Financial Crisis Serve as Powerful Reminders for Investors. Retrieved from https://www.fidelity.com/about-fidelity/individual-investing/lessons-learned-10-years-after Financial Planning Association. The FPA® and Ameriprise® Value of Financial Planning study: Consumer Attitudes and Behaviors in a Changing Economy. (2014, May).Retrieved from http://chapters.onefpa.org/cinci/wp-content/uploads/ sites/3/2014/05/2008ValueofFinancialPlanningReport.pdf Finke, M. S., Huston, S. J., & Winchester, D. D. (2011). Financial Advice: Who Pays. Journal of Financial Counseling and Planning. 22(1), 19. Fischer, R., & Gerhardt, R. (2007, August). Investment mistakes of individual investors and the impact of financial advice. In 20th Australasian Finance & Banking Conference.

Volume 20 • Issue 1 2021

Freund, R. J.,Wilson, W. J. and Sa.P. (1998).Statistical modeling of a response variable. Regression Analysis; Academic Press: St. Louis, MI, USA. Friedberg B. (2018, May 8). Sometimes, Rebalancing Does More Harm Than Good. Retried from https://money.usnews.com/ investing/buy-and-hold-strategy/articles/2018-05-08/is-portfolio-rebalancing-necessary. Gaudecker, H. M. V. (2015). How does household portfolio diversification vary with financial literacy and financial advice?. The Journal of Finance, 70(2), 489-507. Georgarakos, D., & Inderst, R. (2014). Financial advice and stock market participation. Available at SSRN 1641302. Gerhardt, R., & Hackethal, A. (2009). The influence of financial advisors on household portfolios: a study on private investors switching to financial advice. Available at SSRN 1343607. Guillemette, M., Blanchett, D., & Finke, M. (2018). The effect of investment and withdrawal horizons on myopic loss aversion. Applied Economics Letters, 1-4. Hayes, A. (2019, June 15). Investing In Crisis, A High Risk-High Reward Strategy. Retrieved from https://www.investopedia.com/ articles/investing/041415/investing-crisis-high-riskhigh-reward-strategy.asp Hurd, M. D., & Rohwedder, S. (2010). Effects of the financial crisis and great recession on American households (No. w16407). National Bureau of Economic Research. Kim, K. T., Pak, T. Y., Shin, S. H., & Hanna, S. D. (2018). The relationship between financial planner use and holding a retirement saving goal: A propensity score matching analysis. Financial Planning Review, 1(1-2), e1008. Kramer, M. M. (2012). Financial advice and individual investor portfolio performance. Financial Management, 41(2), 395-428. Kramer, M., & Lensink, R. (2012). The impact of financial advisors on the stock portfolios of retail investors. Available at SSRN 2021883. Lei, S., & Yao, R. (2015). Factors Related to Making Investment Mistakes in a Down Market. Journal of Personal Finance, 14(2). Lei, S., & Yao, R. (2016). Use of Financial Planners and Portfolio Performance. Journal of Financial Counseling and Planning, 27(1), 92-108. Lusardi, A., & Mitchell, O. S. (2011). Financial literacy and planning: Implications for retirement wellbeing (No. w17078). National Bureau of Economic Research. Marsden, M., Zick, C. D., & Mayer, R. N. (2011). The value of seeking financial advice. Journal of family and economic issues, 32(4), 625-643. Montford, W., & Goldsmith, R. E. (2016). How gender and financial self‐efficacy influence investment risk taking. International Journal of Consumer Studies, 40(1), 101-106. Murendo, C., & Mutsonziwa, K. (2017). Financial literacy and savings decisions by adult financial consumers in Zimbabwe. International Journal of Consumer Studies, 41(1), 95-103. NBER. (2010, September 20). The latest announcement from the NBER's Business Cycle Dating Committee. Retrieved from https://www.nber.org/cycles/sept2010.html. NBCNews. (2018). 11 historic bear markets. Retrieved from http://www.nbcnews.com/id/37740147/ns/business-stocks_and_ economy/t/historic-bear-markets/#.W-nliy2ZNBwnbc Pompian, M. M. (2017). Behavioral Finance, Individual Investors, and Institutional Investors. Charlottesville, VA: CFA Institute. Robinson, S. J., & Sanderford, A. R. (2016). Green buildings: similar to other premium buildings?. The Journal of Real Estate Finance and Economics, 52(2), 99-116. Rogoff, K & Chainey, R (2019, April 30). An economist explains what happens if there’s another financial crisis. Retrieved from https:// www.weforum.org/agenda/2019/04/an-economist-explains-what-happens-if-there-s-another-financial-crisis/ Rothschild, M. & Stiglitz, J. (1976). Equilibrium in competitive insurance markets: An essay on the economics of imperfect information. The Quarterly Journal of Economics, 90, 629-649. doi: 10.2307/1885326

Journal of Personal Finance

Silverblatt, R. (2010). How to repair your damaged portfolio. U.S. News & World Report Money. Retrieved from http://money.usnews. com/money/personal-finance/mutual-funds/articles/2010/05/21/how-to-repair-your-damaged-portfolio.html. Spence, A. M. (1974). Market signaling: Informational transfer in hiring and related screening processes, pp. 200-201. Cambridge, Mass: Harvard University Press. Stigler, G. J. (1961). The economics of information. Journal of Political Economy, 69, 213-225. doi: 10.1086/258464 Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: a social psychological perspective on mental health. Psychological bulletin, 103(2), 193. Tekçe, B., Yılmaz, N., & Bildik, R. (2016). What factors affect behavioral biases? Evidence from Turkish individual stock investors. Research in International Business and Finance, 37, 515-526. The State of Working America. (2018). The Great Recession. Retrieved from http://stateofworkingamerica.org/great-recession/ Vieider, F. M., Villegas‐Palacio, C., Martinsson, P., & Mejía, M. (2016). Risk taking for oneself and others: A structural model approach. Economic Inquiry, 54(2), 879-894. Von Neumann, J., & Morgenstern, O. (1944). Theory of games and economic behavior. Princeton, New Jersey: Princeton University Press. Willis, L. E. (2008). Against financial literacy education. Iowa Law Review, 94, 08-10. Winchester, D. D., Huston, S. J., & Finke, M. S. (2011). Investor Prudence and the Role of Financial Advice. Journal of Financial Service Professionals, 65(4). Yao, R., & Lei, S. (2016). Prior investment outcomes and stock investment in defined contribution plans. Applied Economics Letters, 23(14), 999-1002. Yao, R., & Lei, S. (2018). Source of Information and Projected Household Investment Portfolio Performance. Family and Consumer Sciences Research Journal, 46(3), 219-237.

Volume 20 • Issue 1 2021

Mortality Salience Lowers Preferred Retirement Asset Decumulation Rates

Yi Liu, Ph.D. Russell James, J.D., Ph.D., CFP®

Abstract Standard life-cycle economic theory suggests that people should spend down assets during retirement at a rate maximizing their lifetime consumption. However, actual retiree behavior exhibits asset decumulation at much slower rates, not at all, or even continued accumulation. One potential factor is that decumulation requires personal mortality estimations. Previous research finds that personal mortality reminders (1) are aversive and (2) increase focus on impacting those who will survive. Correspondingly, recent research has found that (1) annuity purchases are reduced due to associated personal mortality reminders and (2) mortality reminders increase relative preference for annuities that pay less income but provide a greater bequest provision. This study investigates whether mortality salience will also influence the broader issue of an individual’s asset decumulation decisions in retirement. Using a randomly assigned experimental test, we find that increasing mortality salience increases the desire to retain assets in retirement, reducing the preferred spending rates in retirement. Understanding the role of mortality salience on decisions about asset decumulation in retirement can be beneficial to academic researchers and financial planners.

Key Words mortality salience, asset decumulation, financial planning, terror management theory

Journal of Personal Finance

INTRODUCTION Approximately 18% of the country is 65 or older (Federal Reserve Bulletin, 2017) and nearly 10,000 baby boomers retire every day (Kessler, 2014; Sherry et al., 2017). This makes the issue of asset decumulation in retirement particularly important for financial planning practitioners. Households may have accumulated significant retirement assets, but many of them do not spend down those assets according to standard economic expectations, often foregoing spending that could lead to a more comfortable retirement. Understanding what drives retirees’ asset decumulation decisions is thus an important issue for both financial planners and researchers. The standard life-cycle hypothesis predicts that individuals spend down their assets in their retirement phase. Specifically, individuals use their savings to finance their consumption in their post-retirement phase to maximize their utility subject to their budget constraint, smoothing their marginal utility of consumption over their life cycle (Alexandre et al., 2020; Ando & Modigliani, 1963; Scott et al., 2020; Yaari, 1965). This theory suggests that people accumulate wealth during their working years and then decumulate their wealth when they are older. Therefore, rational individuals would be expected to make optimal financial decisions, drawing down their financial portfolios to fund their retirement needs. However, much empirical evidence fails to support the standard life-cycle hypothesis that retirees will spend down their assets in this manner. On the contrary, several studies suggest that retirees with wealth spend down their financial assets very slowly, do not spend down their financial assets, or even continue to accumulate wealth as age increases (Browning et al., 2016; Haider et al., 2000; Love et al., 2009; Poterba et al., 2011). Due to the inconsistency between the prediction of the life-cycle model and retirees’ behaviors in the real world, more research is needed to understand the factors that drive retirees’ asset decumulation decisions. Emerging new experimental results suggest that mortality salience may be an issue in such decisions. Previous psychological (Pyszczynski et al., 1999) and economic (James, 2016) models of mortality salience suggest that personal mortality salience is (1) aversive and (2) increases focus on impacting those who will survive. Salisbury and Nenkov (2016) demonstrated that contemplating the purchase of an annuity triggered mortality salience and, matching the first reaction, such mortality salience decreased an individual’s desire to purchase an annuity at retirement. Matching the

second reaction, Williams and James (2019) demonstrated that mortality salience increased an individual’s relative desire to purchase a lower-income annuity with a bequest provision, rather than a higher income annuity with no bequest provision. This paper examines the effect of mortality salience on the broader issue of an individuals’ asset decumulation decisions during retirement. Using a randomly assigned experimental test, we find that increasing mortality salience decreases the desired asset decumulation rates in retirement. This matches the economic and psychological theory as well as our hypothesis. It also matches and adds to the previous findings regarding annuities. The previously documented aversion to annuities resulting from mortality salience might have been explained simply because annuities are a bet on one’s own longevity. As the anticipation of death becomes suddenly nearer, this bet becomes less attractive. But this response to the same intervention, under life cycle theory, would also predict an increased asset decumulation rate. We observe the opposite. Thus, in combination with past results, this further strengthens the case that understanding the special role of mortality salience can be important to academic researchers and financial planners.

LITERATURE REVIEW Mortality Salience Mortality salience is the conscious awareness of an individual’s own inevitable death. In experiments, the two predominant responses to mortality salience are expressions of (1) avoidance and (2) pursuit of lasting social impact or “symbolic immortality” (Burke et al., 2010). Symbolic immortality provides people with a form of continued existence in that they may leave a lasting impact on something outside of themselves, e.g., supporting children, family, society, nation, etc. (Arndt & Vess, 2008; Becker, 1973; Greenberg et al., 1997; Martin, 1999). A psychological model, referred to as Terror Management Theory, holds that avoidance and pursuit of symbolic immortality are mental defenses designed to manage the increased fear, anxiety, and despair that originate from personal death-related thoughts (Becker, 1973; Greenberg et al., 1986, Pyszczynski et al., 1999; Rank, 1989). An economic model from James (2016) derives these same responses to mortality salience from rational utility maximization. This model allows for utility from current and future personal consumption and

Volume 20 • Issue 1 2021

the current and future circumstances of others with whom the individual is concerned (Becker, 1974). Additionally, it allows for current anticipatory utility from expected future circumstances where such expectations are subject to rational optimism (Brunnermeier & Parker, 2005). Combining these two concepts, in particular with regard to subjective survival estimates, James (2016) proposed that felicity from current anticipation of future personal consumption depends on an individual’s subjective expectations about surviving in the future. Thus, this anticipated enjoyment can be increased by ignoring personal mortality (constrained by the impact on actual future consumption). This benefit from “rational” optimism is destroyed by mortality reminders, thus triggering an “avoidance” response to such reminders. Separately, the increased awareness of personal mortality reduces the anticipation value of investments in future personal consumption relative to investments in future social impact. This triggers an increased desire for investment in future social impact that are not subject to personal mortality, similar to the “symbolic immortality” notion from psychology. In simple terms, people will tend to respond to death reminders by (1) increasing avoidance of anything related to personal mortality and/or (2) becoming increasingly attracted to impacting those things that will live beyond them. Corresponding with the first of these predictions, Salisbury and Nenkov (2016) found that annuity decisions reminded people of their personal mortality and death reminders increased resistance to using annuities in retirement planning. Corresponding with the second of these predictions, Williams and James (2019) found that death reminders increased attraction to a bequest benefit in annuities, even at the expense of reduced income in retirement. Beyond the application to annuity decisions in retirement, this mortality-related model may have implications for retirement spending in general. James (2016, p. 75) suggested that an avoidance response would, “suggest a particular attraction to spending no more than current income (from assets or otherwise), as this is the highest level of spending that does not require contemplation of the timing of one’s own death.” Avoidance might also be reflected in a desire to spend down at a very slow rate such that the contemplated age of asset exhaustion was very far into the future. Additionally, the second response – pursuit of lasting social impact/symbolic immortality – would suggest an increased desire for leaving a bequest, also diminishing the desire to consume assets during retirement. This response aligns with the previous empirical

studies that mortality salience increases the desire for leaving bequests and other forms of prosociality (James, 2016; Jonas et al., 2002; Zaleskiewicz et al., 2015).

Mortality Salience and Consumption Previous literature has documented that mortality salience affects financial behavior in part because money plays a role of an existential anxiety buffer for death (Arndt et al., 2004; Kasser & Sheldon, 2000; Mandel & Heine, 1999; Solomon et al., 2004; Zaleskiewicz et al., 2013a; Zaleskiewicz et al., 2013b). For example, Zaleskiewicz, et al. (2013b) suggested that money provides protection against death anxiety and offers people a sense of security for future uncertainty. Saving money can be more effective than spending money in order to soothe the fear of death (Zaleskiewicz et al., 2013b). Thus, spending down retirement assets more slowly may maintain self-esteem. However, there is little prior research that looks at the impact of individuals’ mortality salience on their asset decumulation decisions in retirement. The goal of this experimental study is to investigate the effects of mortality salience on individuals’ asset decumulation decisions. The present paper adds to the existing literature by providing empirical evidence using experimental data that mortality salience affects individuals’ decisions regarding their preferred decumulation rate for their retirement assets.

Other Factors Affecting Asset Decumulation Asset decumulation decisions involve complex economic decision-making and are influenced by a number of factors. For example, life expectancies (Davies, 1981; De Nardi et al., 2009), bequest motives (Browning et al., 2016; De Nardi et al., 2010; Hurd, 2002), cognitive ability (Banks & Oldfield, 2007; Browning et al., 2016), health status (De Nardi et al., 2009; 2010; French et al., 2006), racial differences (Lindley et al., 1984), education (Bernheim & Scholz 1993; Hubbard et al., 1995), household size (Haider et al., 2000), age (Asher et al., 2017; Banks, Crawford & Tetlow, 2015; Furnham, 1985; Kopczuk, 2007) and gender (Alessie et al., 1999) can affect individuals’ saving and spending decisions. The empirical literature about marital status in asset decumulation is mixed. Married couples might save more than individuals, due to economies from resource sharing, and could have greater bequest desires (Rindfuss & VandenHeuvel, 1990). On the other hand, married couples might have additional spending desires, such as taking expensive vacations together (Waite, 1995).

Journal of Personal Finance

HYPOTHESIS Hypothesis 1: Participants randomly assigned to a mortality salience task will prefer lower rates of asset consumption (decumulation) during retirement compared with those assigned to a control task. Hypothesis 2: Greater self-reported mortality salience will be associated with lower desired rates of asset consumption (decumulation) during retirement.

METHODOLOGY Experimental participants were recruited from a national online panel, Amazon’s Mechanical Turk (MTurk), a popular crowdsourcing site that has been used extensively in experimental psychology research (Buchheit et al., 2019; Hauser & Schwarz, 2016; Smith et al., 2015). MTurk survey responses produce results similar to those generated from traditional non-probability samples (Goodman et al., 2013; Hauser & Schwarz, 2016). Participants were recruited on January 17, 2017 with the surveys being completed on the same day using the Qualtrics survey platform. The participants were recruited to complete a survey using the description “University survey on lifetime personal financial plans,” and were paid $0.75 to complete the experiment. Participants were limited to U.S. residents. To ensure the highest quality responses prior to taking the survey participants had to pass an attention check test. A densely worded paragraph followed a simple question with instructions in the paragraph to answer the question in a non-standard way. Of 1,559 possible participants, 192 were prevented from beginning the survey questions due to incorrectly following these instructions. The study was approved by the Human Subjects Institutional Review Board (IRB2016-1053) of the second Author’s affiliated university. Preliminary instructions informed participants, “Taking part in this research study is completely voluntary. You may skip questions. If you do not wish to participate in this study, feel free to exit out of the survey at any time by closing your internet browser.”

Data There were 1,222 respondents (out of 1,367 total potential respondents) who answered all questions with regard to each measured variable. Although respondents came from

a wide variety of backgrounds, the sample of experimental participants is not intended to be nationally representative. This study used SAS and STATA for analysis.

Intervention This experimental study randomly assigned participants to complete an essay related to imagining either their own death (the experimental group) or their own dental operation (the control group). On average, participants spent 10 minutes 35 seconds writing in response to the essay intervention. Such experimental and control manipulations are common in research using Terror Management Theory (Burke et al., 2010; Wisman et al., 2015). The death reminder essay instructions stated: “Imagine that you die tomorrow. Jot down, as specifically as you can, what you think will happen to you as you physically die and once you are physically dead.” The control group instructions stated: “Imagine that you have a painful dental operation done tomorrow. Jot down, as specifically as you can, what you think will happen to you as you are experiencing the pain.” After answering this, the experimental group was asked, “Also, please briefly describe the emotions that the thought of your own death arouses in you,” and the control group was asked, “Also, please briefly describe the emotions that the thought of dental pain arouses in you.”

Dependent Variable In the analysis, the preferred asset decumulation rate (spending down rate) in retirement was the dependent variable. This variable was determined by the respondent’s answer to the following question: “Suppose you were age 65 and had saved a large sum of money to use during retirement. Which of the following spending plans would you be most likely to use? Spend the interest earned from investments and enough of the principal each year so that the money would run out at age 75, 85, 95, 105, 115, 125, or spend only the interest earned from the investments each year so that the money would never run out.”

Main Explanatory Variables Mortality salience treatment and self-reported mortality salience are the two main explanatory variables. The selfreported mortality salience “Reported MS” variable was the numerical sum (0-15) of responses to the following three questions asked:

Volume 20 • Issue 1 2021

1. To what extent have you been thinking about death in the past several minutes? (0=Never, 1=Very Rarely, 2=Rarely, 3=Occasionally, 4=Frequently, 5=Very Frequently) 2. Please rate your level of agreement with the following phrase: The prior tasks in the survey reminded me of death. (0=Very Strongly Disagree, 1=Strongly Disagree, 2=Disagree, 3=Agree, 4=Strongly Agree, 5=Very Strongly Agree) 3. To what extent did the prior tasks in this survey evoke thoughts of death? (0=Never, l=Very Rarely, 2=Rarely, 3=Occasionally, 4=Frequently, 5=Very Frequently)

Other Explanatory Variables Besides the main explanatory variables, other control variables included age, income, and gender. Age was reported as the midpoint of each age category range but with a maximum of 90 and a minimum of 18 for the highest and lowest categories, respectively. Income was calculated as the midpoint of each range but with "more than $150,000" coded as $175,000. Male was a dichotomous variable that takes a value of 1 if the respondent is male and 0 if the respondent is female. Marital status was a dichotomous variable and respondents who were married were coded as “1.” Other responses such as widowed, divorced, separated, or never married were coded as “0” for marital status.

MODEL This paper uses an ordered probit regression model estimated via maximum likelihood. This model is appropriate to use for statistical analysis because this study has an ordinal dependent variable.

{

DCUi* = β0 + β1MS + βi’DEM + ε

DCUi* =

1 if DCUi* ≤ μ1

2 if μ1 < DCUi* ≤ μ2 3 if μ2 < DCUi* ≤ μ3 4 if μ3 < DCUi* ≤ μ4 5 if μ4 < DCUi* ≤ μ5 6 if μ5 < DCUi* ≤ μ6 7 if DCUi* > μ6

where DCUi* was a latent variable representing the desired asset decumulation preferences of an individual i. DCUi* was the observed decumulation variable that was ordered with seven spending rate categories. The variable MS denoted

mortality salience that took a value of 1 if the respondent was in the mortality salience treatment group and 0 otherwise; DEM was a matrix of demographic variables that included age, male, marital status, and income. The unobserved thresholds were μ1, μ2, μ3, μ4, μ5 and μ6. The term ε was an error term that follows the standard normal distribution. Additionally, β0 was the intercept with β1, β2, β3, β4, β5 and β6 reflecting differences in the net desired asset decumulation associated with changes in each of the explanatory variables, all else equal. Marginal effects were calculated to determine the magnitude of the effects on the observed dependent variable.

RESULTS Respondents indicated their preferred decumulation rate on a scale ranging from 1 (fastest decumulation, exhausting at age 75) to 7 (slowest decumulation, interest only). On this scale, the average decumulation preference was 3.50 for the mortality reminded group and 3.30 for the control group. On the previously described 15-point mortality salience scale, as shown in Table 1, the average was 10.26 for the mortality reminded group and 8.58 for the control group. In a two-tailed t-test both of these differences were significant at p<.05. This provides some initial evidence that the mortality reminder both increased mortality salience and reduced desired decumulation rates. The two groups were not significantly different in age (37.80 and 38.80, respectively) or income ($49,908 and $47,273, respectively). Table 2 shows results from an ordered probit model examining the marginal effects of the mortality salience treatment on the desired decumulation rate. The results indicate that the mortality salience treatment significantly increased interest in having a relatively slow rate of decumulation (exhaustion at age 105 or 115), or no decumulation at all. Conversely, the treatment significantly decreased interest in having a relatively fast rate of decumulation (exhaustion at age 75 or 85). Although the respondents were randomly assigned to the treatment and control groups, Table 3 reports these results when controlling for age, income, gender, and marital status. Including these control variable lowers the statistical significance of the relationships from Table 2, but the same general relationship remains. Such a reduction in significance is not unexpected when adding additional control variables with a sample of this size, especially given that some control variables are commonly associated with both the decumulate preferences (the outcome) (Alessie et al., 1999; Asher, Meyricke

Journal of Personal Finance

et al., 2017; Banks et al., 2015; Furnham, 1985; Kopczuk, 2007; Rindfuss & VandenHeuvel, 1990) and also with mortality salience (the explanatory pathway) (Fessler & Navarrete, 2005; Hirschberger et al., 2002). Respondents also reported their subjective feelings of mortality salience categorized on a 15-point scale. This provides a separate level of analysis potentially connecting mortality salience and desired decumulation rates. For example, the treatment may have been more effective at triggering mortality salience for some participants than for others. Conversely, the control task may have also triggered mortality salience for some participants. Using this separate approach, similar results emerge. Table 3 reports results from an ordered probit model examining the marginal effects of subjective feelings of mortality salience on the desired decumulation rate. Again, the results indicate that higher levels of reported mortality salience were associated with a significantly greater interest in using a relatively slow rate of decumulation (exhaustion at age 105, 115, or 125), or no decumulation at all. Conversely, greater self-reported mortality salience was associated with significantly lower interest in having a relatively fast rate of decumulation (exhaustion at age 75 or 85). As before, this association is also tested while controlling for age, income, gender and marital status. The results change little when adding these controls. These results are reported in Table 4. Thus, whether mortality salience is measured by the objective random assignment to an experimental treatment or the subjective self-report of experienced mortality salience, the associations are similar. Greater mortality salience is associated with a preference for decreased decumulation rates in retirement.

DISCUSSION Previous research showed an aversion to annuity purchases due to mortality salience (Salisbury & Nenkov, 2016). It also showed that inducing mortality salience further increased aversion to annuity purchases (Salisbury & Nenkov, 2016). Other research identified that inducing mortality salience increased the preference for annuities paying lower income but with a greater bequest benefit (Williams & James, 2019). These past results fit the theoretical predictions related to mortality salience from psychology and economics. However, one counter-argument to these previous results with annuities is that mortality reminders might simply lead

to a reduced longevity expectation. A reduced longevity expectation would make annuities relatively less desirable, as these pay only for the life of the owner. However, decumulation rates are not subject to the same argument. Based on standard life-cycle consumption models, if mortality reminders led to a reduced longevity expectation, then mortality reminders should be associated with an increased rate of asset decumulation in retirement. These results show the opposite. The influence of mortality salience on financial decision-making is attracting increasing interest among both practitioners and researchers. Existing research has found that the role of mortality salience is important in financial behaviors, including saving and spending. However, the effects of mortality salience on asset decumulation decisions in retirement, outside of annuity purchase decisions, are not well understood. Without understanding the previous economic and psychological theory, these results might seem to be counterintuitive. One might reasonably argue, for a consumption maximization perspective, that mortality awareness reminds people of death and should trigger an increased rate of spending since life feels shorter or uncertain. The alternative hypothesis is that, instead of this, mortality salience could encourage a willingness to invest in future social impact for loved ones, i.e., a bequest benefit, or increase the desire to retain wealth as a psychological defense to mortality awareness. Respondents who experienced the mortality salience intervention chose lower asset decumulation rates. This might relate to a desire for an increased sense of financial security, shielding against anxiety related to death. It might relate to an increased desire to leave a larger bequest amount that would improve other people’s lives. Uncovering this mechanism suggests potential implications for future explorations. For example, Salisbury and Nenkov (2016), found that mortality-salience related resistance to annuities could be reduced by changing annuity descriptions from death-related language to life-related language. Similarly, Williams and James (2019) found that changing annuity language to death-related language decreased the desire for annuity income in favor of an increased bequest provision. Thus, a future study might test how framing spending rates during retirement using death or life language could change people’s preferred retirement spending rates. An important caveat about the present work is that the average age of respondents was well below retirement age (approximately 38-39 years old) and the decision was

Volume 20 • Issue 1 2021

hypothetical. Mturk tends to recruit young participants (Arechar et al., 2017; Ipeirotis, 2010). Therefore, it might not accurately predict future retirees’ actual decisions regarding their spending down rates during their retirement. Other limitations of this study include a potential lack of generalizability of the findings resulting from a non-probability crowdsourcing sample. Future studies might use nationally representative samples or capture respondents’ actual saving, spending and asset decumulation decisions. Additionally, future studies could explore how individuals’ bequest motives mediate the relationship of mortality salience and asset decumulation rates, and whether those behaviors increase individuals’ financial satisfaction.

Implications Findings from this study have implications for financial planning practice. Completing “ideal” retirement spending rates on a spreadsheet may not be all that is necessary for effective practice. Understanding fundamental human tendencies, such as those related to mortality-related planning, may be equally important. Eliciting retirement spending preferences in a highly mortality-salient context, such as during estate planning, is likely to generate different results than in other contexts.

Presenting options in ways that take into account such responses to mortality reminders may help clients to achieve their goals. Financial planners are in a unique position to help clients to identify their needs for retirement and to match clients’ intentions with their actions. This is particularly so as personal mortality salience does not typically apply – or apply as strongly – when planning for another person’s lifetime spending. Through understanding the effect of mortality salience on asset decumulation decisions, financial planners can more knowledgeably engage their clients to pursue favorable long-term financial and life satisfaction outcomes in retirement spending.

CONCLUSION This study was designed to highlight the role of mortality salience in retirement decumulation decisions by using a randomly assigned experimental test. The results of this study support the hypothesis that mortality salience can increase individuals’ desire to retain assets in retirement. The study adds to other findings in Terror Management Theory as applied to financial planning. Along with previous research it suggests that mortality salience may be an important factor for a broad range of financial behaviors.

Journal of Personal Finance

TABLES TABLE 1 Descriptive Statistics: Mortality Salience Treatment Group and Dental Essay Control Group Variable

Mortality Salience Essay

Dental Essay

Reported MS [0-15]

10.26

8.58

Assets end at age 75

4.45%

4.88%

Assets end at age 85

21.58%

25.20%

Assets end at age 95

37.23%

37.72%

Assets end at age 105

18.62%

18.70%

Assets end at age 115

4.45%

3.25%

Assets end at age 125

1.98%

1.63%

Interest only

11.70%

8.62%

Age (mean)

37.80

38.80

Male (proportion)

44.39%

42.83%

Married (proportion)

44.22%

43.66%

$49,907.87

$47,273.48

Income (mean)

TABLE 2 Ordered Probit Model: The Marginal Effect of Mortality Salience Treatment on Asset Decumulation Rate by Age of Exhaustion Desired Decumulation Rate by Planned Age of Exhaustion Age 75 Mortality Salience Treatment

Age 85

Age 95

Age 105 Age 115

Age 125

Interest only

-0.0124**

-0.0304** -0.0040*

0.0161** 0.0053**

0.0028*

0.0225**

(0.0060)

(0.0142)

(0.0076)

(0.0014)

(0.0106)

(0.0023)

(0.0026)

Note: N =1222; *** p<.01 ** p<.05 *p<.1

TABLE 3 Ordered Probit Model: The Marginal Effect of Mortality Salience Treatment on Asset Decumulation Rate with Control Variables Desired Decumulation Rate by Planned Age of Exhaustion Age 75

Age 85

Age 95

Age 105

Age 115

Age 125

Interest only

Mortality Salience Treatment

-0.0106*

-0.0263*

-0.0033

0.0141*

0.0046*

0.0025*

0.0191*

(0.0059)

(0.0143)

(0.0022)

(0.0077)

(0.0026)

(0.0014)

(0.0105)

Age

0.0004*

0.0010*

0.0001

-0.0005*

-0.0002*

-0.0001*

-0.0007*

(0.0002)

(0.0005)

(0.0001)

(0.0003)

(0.0001)

(0.0004)

-0.0022**

-0.0056**

-0.0007*

0.0030**

0.0010**

0.0005**

0.0040**

Income (10k)

Volume 20 • Issue 1 2021

Desired Decumulation Rate by Planned Age of Exhaustion

Male

Married

Age 75

Age 85

Age 95

Age 105

Age 115

Age 125

Interest only

(0.0009)

(0.0022)

(0.0004)

(0.0012)

(0.0004)

(0.0002)

(0.0016)

-0.0111*

-0.0275*

-0.0035

0.0148*

0.0048*

0.0026*

0.0200*

(0.0060)

(0.0147)

(0.0022)

(0.0079)

(0.0026)

(0.0015)

(0.0107)

0.0012

0.003

0.0004

-0.0016

-0.0005

-0.0003

-0.0022

(0.0018)

(0.0043)

(0.0006)

(0.0023)

(0.0008)

(0.0004)

(0.0032)

Note: N = 1,204; *** p<.01 ** p<.05 *p<.1

TABLE 4 Ordered Probit Model: The Marginal Effect of Reported Mortality Salience on Asset Decumulation. Desired Decumulation Rate by Planned Age of Exhaustion

Reported MS (1-15)

Age 75

Age 85

Age 95

Age 105

Age 115

Age 125

Interest only

-0.0017**

-0.0043**

-0.0006*

0.0022**

0.0007**

0.0004**

0.0031**

(0.0008)

(0.0020)

(0.0003)

(0.0010)

(0.0004)

(0.0002)

(0.0015)

Note: N = 1,221; *** p<.01 ** p<.05 *p<.1

TABLE 4 Ordered Probit Model: The Marginal Effect of Reported Mortality Salience on Asset Decumulation with Control Variables Desired Decumulation Rate by Planned Age of Exhaustion Age 75

Age 85

Age 95

Age 105

Age 115

Age 125

Interest only

-0.0016**

-0.0041**

-0.0005*

0.0022**

0.0007**

0.0004*

0.0029**

(0.0008)

(0.0020)

(0.0003)

(0.0011)

(0.0004)

(0.0002)

(0.0014)

0.0004*

0.0009*

0.0001

-0.0005*

-0.0002*

-0.0001

-0.0007*

(0.0002)

(0.0005)

(0.0001)

(0.0003)

(0.0001)

(0.0004)

Income (10K)

-0.0023**

-0.0057**

-0.0007*

0.0031**

0.0010**

0.0005**

0.0042**

(0.0009)

(0.0022)

(0.0004)

(0.0012)

(0.0004)

(0.0002)

(0.0016)

Male

-0.0100*

-0.0255*

-0.0032

0.0135*

0.0044*

0.0024

0.0184*

(0.0059)

(0.0149)

(0.0022)

(0.0079)

(0.0026)

(0.0015)

(0.0108)

0.0011

0.0027

0.0003

-0.0014

-0.0005

-0.0003

-0.002

(0.0017)

(0.0044)

(0.0006)

(0.0023)

(0.0008)

(0.0004)

(0.0032)

Reported MS (1-15) Age

Married

Note: N = 1,203; *** p<.01 ** p<.05 *p<.1

Journal of Personal Finance

REFERENCES Alessie, R., Lusardi, A., & Kapteyn, A. (1999). Saving after retirement: evidence from three different surveys. Labour Economics, 6(2), 277-310. Alexandre, F., Bação, P., & Portela, M. (2020). Is the basic life-cycle theory of consumption becoming more relevant? Evidence from Portugal. Review of Economics of the Household, 18(1), 93-116. Ando, A., & Modigliani, F. (1963). The "life cycle" hypothesis of saving: Aggregate implications and tests. The American Economic Review, 53(1), 55-84. Arechar, A. A., Kraft-Todd, G. T., & Rand, D. G. (2017). Turking overtime: how participant characteristics and behavior vary over time and day on Amazon Mechanical Turk. Journal of the Economic Science Association, 3(1), 1-11. Arndt, J., Solomon, S., Kasser, T., & Sheldon, K. M. (2004). The urge to splurge: A terror management account of materialism and consumer behavior. Journal of Consumer Psychology, 14(3), 198-212. Arndt, J., & Vess, M. (2008). Tales from existential oceans: Terror management theory and how the awareness of our mortality affects us all. Social and Personality Psychology Compass, 2(2), 909-928. Asher, A., Meyricke, R., Thorp, S., & Wu, S. (2017). Age pensioner decumulation: Responses to incentives, uncertainty and family need. Australian Journal of Management, 42(4), 583-607. Banks, J., & Oldfield, Z. (2007). Understanding pensions: Cognitive function, numerical ability and retirement saving. Fiscal Studies, 28(2), 143-170. Banks, J., Crawford, R., & Tetlow, G. (2015). Annuity choices and income drawdown: evidence from the decumulation phase of defined contribution pensions in England. Journal of pension economics & finance, 14(4), 412-438. Becker, E. (1973). The Denial of Death. Simon and Schuster. Becker, G. S. (1974). A theory of social interactions. Journal of Political Economy, 82(6), 1063- 1093. Bernheim, B. D., & Scholz, J. K. (1993). Private saving and public policy. Tax Policy and the Economy, 7, 73-110. Browning, C., Huston, S. J., & Finke, M. S. (2015). Cognitive ability and post retirement asset decumulation. Journal of Family and Economic Issues, 37(2), 242-253. Brunnermeier, M. K. and Parker, J. A. (2005). Optimal expectations. American Economic Review, 95(4):1092-1118. Buchheit, S., Dalton, D. W., Pollard, T., & Stinson, S. (2019). Crowdsourcing intelligent research participants: a student versus MTurk comparison. Behavioral Research in Accounting. Earlyview https://doi.org/10.2308/bria-52340 Burke, B. L., Martens, A., & Faucher, E. H. (2010). Two decades of terror management theory: A meta-analysis of mortality salience research. Personality and Social Psychology Review, 14(2), 155-195. Davies, J. B. (1981). Uncertain lifetime, consumption, and dissaving in retirement. The Journal of Political Economy, 89(3), 561-577. De Nardi, M., French, E., & Jones, J. B. (2009). Life expectancy and old age savings. American Economic Review, 99(2), 110-15. De Nardi, M., French, E., & Jones, J. B. (2010). “Why do the Elderly Save? The Role of Medical Expenses.” Journal of Political Economy, 118(1), 39-75. French, E., De Nardi, M., Jones, J. B., Baker, O., & Doctor, P. (2006). “Right before the End: Asset Decumulation at the End of Life.” Economic Perspectives-Federal Reserve Bank of Chicago, 30(3), 2. Federal Reserve Bulletin. (2017). Changes in U.S. Family Finances from 2013 to 2016: Evidence from the Survey of Consumer Finances [September 2017]. Available at https://www.federalreserve.gov/publications/files/scf17.pdf Fessler, D. M., & Navarrete, C. D. (2005). The effect of age on death disgust: Challenges to terror management perspectives. Evolutionary Psychology, 3(1), 279-296.

Volume 20 • Issue 1 2021

Fisher, P. (2010). Gender differences in personal saving behaviors. Journal of Financial Counseling and Planning, 21(1). Goodman, J. K., Cryder, C. E., & Cheema, A. (2013). Data collection in a flat world: The strengths and weaknesses of Mechanical Turk samples. Journal of Behavioral Decision Making, 26(3), 213-224. Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences of a need for self-esteem: A terror management theory. In R. F. Baumeister (Ed.), Public self and private self (pp. 189-212). New York: Springer-Verlag. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self-esteem and social behavior: Empirical assessments and conceptual refinements. In M. P. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 29, pp. 61-139). New York: Academic Press. Haider, S., Hurd, M., Reardon, E., & Williamson, S. (2000). Patterns of dissaving in retirement. AARP Public Policy Institute. Hauser, D. J., & Schwarz, N. (2016). Attentive Turkers: MTurk participants perform better on online attention checks than do subject pool participants. Behavior Research Methods, 48(1), 400-407. Hirschberger, G., Florian, V., Mikulincer, M., Goldenberg, J. L., & Pyszczynski, T. (2002). Gender differences in the willingness to engage in risky behavior: A terror management perspective. Death studies, 26(2), 117-141. Hurd, M. D. (2002). Are Bequests Accidental or Desired? Santa Monica, CA: RAND Corporation, 2002. https://www.rand.org/pubs/ drafts/DRU3010.html. Hubbard, R. G., Skinner, J. & Zeldes, S. P. (1995). “Precautionary Saving and Social Insurance.” Journal of Political Economy, 103, 360-399. Ipeirotis, P. G. (2010). Demographics of mechanical turk. NYU Working Paper No. CEDER-10-01. Available at SSRN: https://ssrn.com/ abstract=1585030 James, R. N. (2016). An Economic Model of Mortality Salience in Personal Financial Decision Making: Applications to Annuities, Life Insurance, Charitable Gifts, Estate Planning, Conspicuous Consumption, and Healthcare. Journal of Financial Therapy. 7(2):62-82. Jonas, E., Schimel, J., Greenberg, J., & Pyszczynski, T. (2002). The Scrooge effect: Evidence that mortality salience increases prosocial attitudes and behavior. Personality and Social Psychology Bulletin, 28,1342-1353. Kasser, T., & Sheldon, K. M. (2000). Of wealth and death: Materialism, mortality salience, and consumption behavior. Psychological science, 11(4), 348-351. Kessler, G. (2014). Do 10,000 baby boomers retire every day? Washington Post, July, 24. Kopczuk, W. (2007). Bequest and tax planning: Evidence from estate tax returns. The Quarterly Journal of Economics, 122(4), 1801-1854. Love, D. A., Palumbo, M. G., & Smith, P. A. (2009). The trajectory of wealth in retirement. Journal of Public Economics, 93(1-2), 191-208. Lindley, J. T., Selby, E. B., & Jackson, J. D. (1984). Racial discrimination in the provision of financial services. The American Economic Review, 74(4), 735-741. Mandel, N., & Heine, S. J. (1999). Terror management and marketing: He who dies with the most toys wins. NA - Advances in Consumer Research Volume 26, eds. Eric J. Arnould and Linda M. Scott, Provo, UT: Association for Consumer Research, 527-532. Martin, L. L. (1999). ID compensation theory: Some implications of trying to satisfy immediate-return needs in a delayed-return culture. Psychological Inquiry, 195-208. Poterba, J., Venti, S., & Wise, D. (2011). “The Composition and Drawdown of Wealth in Retirement.” Journal of Economic Perspectives, 25(4), 95-118. Pyszczynski, Greenberg, & Solomon (1999) Creativity and terror management: Evidence that creative activity increases guilt and social projection following mortality salience. Journal of Personality and Social Psychology. 77(1):19-32. Rank, O (1989). Art and Artist: Creative Urge and Personality Development. New York: Knopf.

Journal of Personal Finance

Rindfuss, R. R., & VandenHeuvel, A. (1990). Cohabitation: A precursor to marriage or an alternative to being single?. Population and development review, 703-726. Salisbury, L. C., & Nenkov, G. Y. (2016). Solving the annuity puzzle: The role of mortality Salience in retirement savings decumulation decisions. Journal of Consumer Psychology, 26(3), 417-425. Scott, J. S., Shoven, J. B., Slavov, S. N., & Watson, J. G. (2020). Can low retirement savings be rationalized? (No. w26784). National Bureau of Economic Research, https://www.nber.org/papers/w26784 Smith, N. A., Sabat, I. E., Martinez, L. R., Weaver, K., & Xu, S. (2015). A convenient solution: Using MTurk to sample from hard-toreach populations. Industrial and Organizational Psychology, 8(2), 220–228. Sherry, A., Tomlinson, J. M., Loe, M., Johnston, K., & Feeney, B. C. (2017). Apprehensive about retirement: Women, life transitions, and relationships. Journal of Women & Aging, 29(2), 173-184. Solomon, S., Greenberg, J., & Pyszczynski, T. A. (2004). Lethal consumption: Death-denying materialism Psychology and Consumer culture: The Struggle for a Good Life in a Materialistic World, 127-146. Waite, L. J. (1995). Does marriage matter? Demography, 32(4), 483-507. Williams, J. & James, R. N., III (2019). Bequest provision preferences in commercial annuities: An experimental test of the role of mortality salience. Journal of Financial Counseling and Planning, 30(1), 1-16. Wisman, A., Heflick, N., & Goldenberg, J. L. (2015). The great escape: The role of self-esteem and self-related cognition in terror management. Journal of Experimental Social Psychology, 60, 121-132. Yaari, M. E. (1965). Uncertain lifetime, life insurance, and the theory of the consumer. The Review of Economic Studies, 32(2), 137-150. Zaleskiewicz, T., Gasiorowska, A., Kesebir, P., Luszczynska, A., & Pyszczynski, T. (2013a). Money and the fear of death: The symbolic power of money as an existential anxiety buffer. Journal of Economic Psychology, 36, 55-67. Zaleskiewicz, T., Gasiorowska, A., & Kesebir, P. (2013b). Saving can save from death anxiety: Mortality salience and financial decision-making. PloS one, 8(11), e79407. Zaleskiewicz, T., Gasiorowska, A., & Kesebir, P. (2015). The Scrooge effect revisited: Mortality salience increases the satisfaction derived from prosocial behavior. Journal of Experimental Social Psychology, 59, 67-76.

Volume 20 • Issue 1 2021

CE Exam for Members of the IARFC® Members of the IARFC® can earn CE credit by reading the Journal of Personal Finance (JPF). Two units of IARFC® CE credit will be awarded to members who achieve a 70% or higher on this multiple choice quiz. Only one submission per IARFC® member is allowed. Please read the articles in the JPF, and then take the quiz online. The questions are provided here for reference. A link to register for the quiz (or for quizzes on prior JPF issues), is available on the JPF website (www.journalofpersonalfinance. com). Upon registration, an email will be sent with a link to access the quiz. As of this printing, JPF Online CE quizzes cost $20 for each Volume, Issues 1 and 2. 1. Which of the following best describes “symbolic immortality”? a. Provides people with a form of continued existence in that they may leave a lasting impact on something outside of themselves, e.g., supporting children, family, society, nation, etc. b. Provides people an opportunity to avoid the contemplation of their own mortality c. The idea of cryogenically preserving one’s body so as to live again in the future d. Creating a popular emoji with religious connotations

4. If your client's personal mortality salience (awareness of their death) increased, which of the following behaviors are likely to result? a. Have an increased focus on impacting loved ones who will survive them b. Have a decreased desire for people to remember them c. Have an increased desire to continue to be reminded of their mortality at future meetings d. Have an increased desire to annuitize more assets if such annuities are described as paying “until you die”

2. What is an example of a psychological defense that personal mortality salience may generate? a. Pursuit of lasting social impact or symbolic immortality b. Increased desire for annuitization of all assets c. Increased desire for heirs to quicky spend any inheritance d. Increased desire to reduce any funds left at death

5. The key finding from The Relationship between Home Equity and Retirement Satisfaction provides evidence suggesting a: a. Positive relationship between relative home equity and retirement satisfaction b. Negative relationship between relative home equity and retirement satisfaction c. No statistically significant relationship between relative home equity and retirement satisfaction d. None of the above

3. What is an example of a contradiction observed between the Standard life-cycle economic theory and retirees’ consumption behaviors? a. The life-cycle hypothesis predicts that individuals should spend down their assets in their retirement phase. In reality, many retirees do not spend down their financial assets, or even continue to accumulate wealth even in advanced age. b. The life-cycle hypothesis predicts that individuals should accumulate their assets in their retirement phase. In reality, many retirees spend down their financial assets as they age. c. The life-cycle hypothesis predicts that individuals should accumulate their assets prior to their retirement phase. In reality, retirees have less wealth than those recent college graduates. d. The life-cycle hypothesis predicts that individuals should spend down their assets prior to their retirement phase. In reality, college students have less wealth than those reaching retirement age.

6. Pearson and Lacombe suggest that one explanation for retirees experiencing a less-than-optimal retirement experience is: a. Lack of homeownership b. Having a mortgage upon entering retirement c. Resource constraints that are a byproduct of high concentrations of home equity d. Utilizing home equity to finance consumption 7. As of December 2020, student loan debt letters have been mandated in a. All 50 states b. 46 states c. Have not been mandated anywhere in the U.S. d. Have been mandated in 12 states, including Texas and Indiana

Journal of Personal Finance

8. The average U.S. adult reads at what grade level? a. 12th grade level b. 3rd grade level c. 7th to 8th grade level d. 11th grade level 9. According to "The Promise of Student Loan Debt Letters" article, reading comprehension can be affected by a. Diet b. Stress c. Software Programs d. Employment Status 10. Which factors were found to be related to investment decisions/strategy choice in a down market? a. Age and gender b. Wealth and income c. Educational achievement d. All of the above 11. What characteristics do the investors who held to the losers in a down market have in common based on this research? a. Lower income, fewer investable assets, with zero or lower saving rate and no emergency fund b. More investable assets c. Higher educated d. Business owners

13. Regarding college savings, what is most likely to happen when a household’s income increases? a. Child quantity will increase b. Child quality will increase c. Student-loan debt will increase d. Children will be able to attend four-year in-state universities 14. One of the most likely consequences of financial fragility is? a. Increase in household spending constraints b. Increase in college savings c. Decrease in student-loan borrowings d. Consumption smoothing 15. Which of these factors are positive determinants of college savings? a. Household income, financial fragility, financial literacy b. Household income, age, financial literacy c. Household income, number of children, financial-risk taking d. Household income, bond pricing literacy, homeownership

12. What is the implication for the financial planners based on this research? a. Working with clients to design and implement a disciplined approach and a portfolio rebalancing strategy might help guide them to make a rational decision in the face of losses. b. Financial planners should provide education to their clients in following a disciplined investment strategy or appropriate portfolio rebalance strategy, especially when they were confronted with unpleasant life or economic changes. c. Financial professionals should be aware of and discuss the possible investment biases with their clients. d. All of the above

MRFC® Application Applicant Information

Exam

(please print or type)

______________________________________________________________________________________ Please provide your name only on the line below as you want it to appear on your Certificate.

The Certification staff will review all candidate applications submitted to determine if the candidate is eligible to sit for the MRFC exam

Business Information

and for completeness and payment of fees.

______ _______________________ __________________ __________________________ __________ Prefix First Name Middle Name Last Name Suffix ____________________________________________________________ __________________________ Business Name Preferred Salutation ____________________________________ __________ __________________ _______ ____________ Street Address Ste#/Apt City State Zip ____________________________________ _________________________ ________________________ Business Phone Fax Cell Phone __________________________________________________ ____________________________________ Business Email Address Primary Yes No Website

Home Information ____________________________________ __________ ____________________ ______ ___________ Street Address Ste#/Apt City State Zip ______________________________________________ ________________________________________ Home Phone Cell Phone ____________________________________________________ __________________________________ Home Email Address Birth date MM/DD/YY

Please send all mail to my:

Business Address

Home Address

Candidates will be notified of their eligibility to sit for the MRFC Exam. The candidate will have 90 days, after notice of application approval.

MRFC® Fee Schedule

Choose one

Nonrefundable Application Fee:

$100

Examination/Certification Fee:

$350

Association Fee (join)

$100

Current Association Member (continue) Total payment:

___________

Annual Recertification Fee

Experience

Commence on anniversary of passing MRFC Exam

Must have a minimum of four years of experience as a full-time practitioner in the field of financial planning or financial services: Full-time__________ Part-time___________

MRFC (Certification Only)

$350

MRFC (with Association Membership)

$450

Education Education criteria requires an applicant to assert and document achievement in any one of these areas: Education, Professional Designations/Credentials or Licensing.

Education Background School, City, State (Since High School)

1. Mail Application with payment to:

Graduated Yes

Payment Options

Major

Degree

IARFC P.O. Box 506, Middletown, OH 45042 2. Fax Application to: (513) 345-9479 (credit card only) 3. Email Application to: info@iarfc.org

(Evidence of license, diploma or documents may be requested. You need not submit evidence with the application.)

Check payable to: IARFC

Professional Designations/Credentials AAMS®

CFA

CFP®

ChFC®

CLU®

CPA

LUTCF®

RFC®

Credit Card: Visa, MC, Amex, or Discover

Other _______________________________________________________________________________

Licensing Broker/Dealer______________________________________ (Personal) FINRA CRD No. ____________ Securities Licenses: Insurance Licenses:

Series 6 and 63 Series 7 and 66

Series 7 and 63 Series 6 and 66 Series 65 Other_____________________________

Credit Card#

Exp. Date

Security Code

Life Health Variable Contracts Prop. & Casualty Other ____________________________________________________________

Primary Insurance Company (if any)________________________________________________________ Affiliated with an SEC Registered Investment Advisor (RIA)?

Yes

Signature

Name of RIA ___________________________________________________________________________ Code of Ethics (Applicants must subscribe and adhere to the IARFC Code of Ethics) I will at all times put my client’s interest above my own. I will maintain proficiency in my work through continuing education. When fee-based services are involved, I will charge a fair and reasonable fee based on the amount of time and skill required. I will abide by both the spirit and the letter of the laws and regulations applicable to financial planning services. I will give my clients the same service I would give myself in the same circumstances.

Questions relating to business and ethical conduct

Recommend a colleague for the MRFC

(If you check “Yes” to any of the following questions, please attach a written explanation.)

Yes

Have you ever been refused a surety bond or other form of employment security? Have you ever been denied or enjoined from selling or dealing in securities or from functioning as an Investment Advisor? Have you ever been arrested, indicted, or convicted for any felony or misdemeanor, except for minor traffic offenses? Have you ever been known personally by any other name, or have you ever conducted financial activities, conducted business or carried brokerage or bank accounts in any other name? Have you ever become insolvent, failed in business or compromised with creditors? If “Yes” – please provide the date, name, and location of court, disposition, liabilities, and assets.

Full Name

Address

City

Have you ever had a license, permit, certificate, registration or membership denied, suspended, revoked or restricted, or have you had an application of such type ever withdrawn for cause?

State, Zip

Have you ever been the subject of any order, judgement, decree or other sanction of a foreign court, foreign exchange, or have you ever been the subject of any action by a foreign or domestic governmental or regulatory agency?

Phone

Attestations (Applicants please read carefully) 1. 2.

10. 11.

12.

13.

I hereby certify that I have read and understand the foregoing statements and that my responses are true and complete to the best of my knowledge. I hereby apply for the MRFC credential and in consideration of my application, I submit myself to the jurisdiction of the Association and hereby verify that I agree to abide by all the provisions of the By-Laws and regulations of the Association as they are and may be amended. I agree to comply with all such requirements, subject to right of appeal as provided by law. I agree that any decision as to the result of any exam(s) that I may be required to pass or annual Continuing Education (CE) requirements will be accepted by me as final. I further agree that neither the Association nor its trustees, directors, officers, or employees shall be liable to me for action taken or omitted in official capacity or in the scope of employment, except as otherwise provided in the statutes, By-Laws, or the Association’s regulations. I hereby certify that I have a sound record of business integrity with no suspension or revocation of any professional licenses, and I hereby subscribe to the IARFC Code of Ethics, a copy of which I have read and understand. It is agreed and understood that any material misrepresentation of facts or information given in this or subsequent application or renewal may be cause for immediate revocation of the MRFC credential and all its privileges, without refund of any dues or fees paid. I understand that failure to disclose any regulatory event, including suspensions or revocations, may disqualify me from initially obtaining the MRFC credential or could result in revocation of the credential. As an applicant for registration, I understand and agree that my MRFC credential will not become effective until I have met all the eligibility requirements and had have successfully passed the MRFC exam. I understand that the MRFC credential remains the property of the MRFC Certification Board, (MCB) and must be destroyed or returned to the MCB should my right to display the credential be suspended or terminated. I understand that the continuation of the MRFC credential requires the successful awarding of forty (40) units of financial services focused CE credits — of which two (2) units every year must be related to Professional Ethics commencing the January of the year following initial acceptance. I understand this application is valid for sixty (60) days from the date of receipt by MCB’s home office and I have ninety (90) days upon application approval to schedule the MRFC exam. I authorize the Association to make available to any federal, state or municipal agency, or any securities or commodities industry self-regulatory organization, any information they may have concerning me or to request confirmation of my status, and I release those organizations, employees and agents, from any and all liability of whatever nature by reason of furnishing such information. I further agree that my contact information contained in this application be divulged to interested parties as part of the member profile on the IARFC website for the benefit of members and the public. I understand that except for my certification status, written authorization by me is required to release my information.

How did you learn about the MRFC? Linkedin

Facebook

Twitter

Broker/Dealer Direct Mail IARFC Website

Exhibit

Referral

Other __________________________________

Referred by (if applicable)

Full Name

City, State

International Association of Registered Financial Consultants P.O. Box 506 Middletown, OH 45042-0506 Phone: (800) 532-9060 Fax: (513) 345-9479 Email: mrfc@iarfc.org Website: iarfc.org

I attest that I have read and understand the above, that the information I have provided is complete and accurate to the best of my knowledge and belief, and I further understand that my MRFC credential may be revoked if I have provided any false or incomplete information.

Signature of Applicant (required)

Date

1/18/21

Call for Papers

To Financial Educators The Journal of Personal Finance encourages high quality submissions that add to the growing literature in personal finance. Since this literature spans a number of disciplines, authors are encouraged to conduct a thorough review of literature prior to submission. We are looking for original research that uncovers new insights in personal finance — research that will have an impact on advice provided to individuals. It is the goal of the editor to provide timely reviews (less than 60 days) and decisions to authors. To submit manuscripts to the IARFC for publication. Visit https://www.iarfc.org/publications/journal-of-personal-finance for submission guidelines or contact jpfeditor@iarfc.org.

Journal of Personal Finance International Association of Registered Financial Consultants - IARFC 146 N. Breiel Boulevard P.O. Box 506 Middletown, Ohio 45042

lan P g n i Winn e h t the for Look blished in urnal! pu he Jo t f o sue Fall Is

IARFC NATIONAL FINANCIAL PLAN COMPETITION

Corporate Sponsors Supporting the Next Generation of Financial Consultants

To learn more call (800) 532-9060 or visit www.IARFC.org