Financial Services Review, 33(1) 179 Does Overspending Harm Retirement Preparation? Christina Lynn,1 Stuart Heckman,2 Michael Kothakota,3 and Derek Lawson4 Abstract This study addressed the research question of how overspending is related to retirement preparation. A commonsense answer to this question is this: overspending should negatively impact retirement preparation. However, the existing body of knowledge does not provide evidence to support or deny this assumption. The Behavioral Life Cycle Hypothesis was tested as a theoretical framework to answer this research question, providing valuable insight. Three data sets were used, including the Survey of Household Economic Decisionmaking (SHED), the Survey of Consumer Sciences (SCF), and the National Financial Capability Study (NFCS), to conduct logit and OLS regressions in testing the hypotheses. Because the overspending measurements were only negatively related to retirement preparation in a little over half the analyses, the results point to a new cultural norm where one’s overspending behavior does not necessarily reflect one’s retirement preparation behavior. Results provide support for policy actions related to tightening credit card policies, exposing a lack of awareness on overspending, providing practical approaches for avoiding overspending behavior, and the value of using multiple data sets as a robustness check. Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License Recommended Citation Lynn, C., Heckman, S., Kothakota, M., & Lawson, D. (2025). Does overspending harm retirement preparation? Financial Services Review, 33(1), 179-204. Introduction Does the average American family feel stretched so thin they cannot afford to contribute to their retirement account? Is overspending to blame? This study examines the intricate relationship between overspending and retirement preparation using data from multiple sources, including the Survey of Household Economics and Decisionmaking (SHED), the Survey of Consumer Finances (SCF), and the National Financial Capability Study (NFCS). Guided by 1 Corresponding author (christinalynn0809@gmail.com). Kansas State University, Manhattan, KS, USA 2 Kansas State University, Manhattan, KS, USA 3 Kansas State University, Manhattan, KS, USA 4 Kansas State University, Manhattan, KS, USA the Behavioral Life Cycle Hypothesis (BLCH) (Shefrin & Thaler, 1988), this research explores the dynamics of overspending and its impact on retirement preparation behaviors among average working adults in the United States. Advisors may feel stumped when advising clients who habitually overspend. This issue is alarming, as many households are ill-prepared for retirement (Board of Governors of the Federal Reserve System, 2021). Additional challenges, such as longevity risk (Lim & Lee, 2021), rising https://creativecommons.org/licenses/by-nc/4.0/ https://creativecommons.org/licenses/by-nc/4.0/ Financial Services Review, 33(1) 180 inflation (Bennett, 2021), and a general lack of consumer awareness (Fan et al., 2022), exacerbate these concerns for advisors. While extensive research exists on retirement preparation, the specific effects of overspending remain underexplored. The assumed negative relationship between overspending and retirement readiness appears logical, yet academic evidence is scant. This study seeks to address the critical question: Does overspending significantly undermine retirement preparation? Background Review and Theoretical Orientation Retirement Preparation Retirement preparation involves actions taken throughout one's working career to ensure financial wellbeing in retirement (Muratore & Earl, 2010; Muratore & Eckert, 2004). Saving for the future is a vital aspect of most households' personal finances. Given that retirement planning is highly individualized, there is no universal formula. The key to successful retirement preparation lies in adequately planning for one's goals and needs (Adams & Rau, 2011). However, inadequate retirement savings (Board of Governors of the Federal Reserve System, 2021; Oakley & Kenneally, 2019), longevity risk (Lim & Lee, 2021), inflation risk (Bennett, 2021), and a lack of consumer awareness (Fan et al., 2022) indicate a broader problem in retirement readiness. This study aims to better understand why the average household is underprepared for retirement, despite the critical importance of retirement planning. Overspending There is no consensus in the literature on why overspending, a self-harming financial behavior, occurs. Possible explanations include spending preferences influenced by social norms and cultural trends (Cynamon & Fazzari, 2008), social comparison (Pahlevan Sharif et al., 2022; Watson, 2003), materialism (Kimiyagahlam et al., 2019), and budgeting behaviors (Choe & Kan, 2021; Sui et al., 2021). Other researchers highlight the concept of the "pain of payment" to explain overspending (Achtziger, 2022; Choe & Kan, 2021; Rick, 2018). "Pain of payment" refers to the discomfort associated with spending money, which can deter overspending when the anticipated financial consequences are significant (Choe & Kan, 2021). Despite these insights, little is known about the relationship between overspending and retirement preparation behavior. Behavioral Life Cycle Hypothesis The BLCH recognizes how people often act irrationally when it comes to money (Shefrin & Thaler, 1988). In contrast to the flat consumption line of the LCH, the consumption line under the BLCH model may not be straight. Consumption may be higher in pre-retirement, due to low levels of self-control leading to overspending, and then decline in later years due to deficient savings, as depicted in Figure 1. A dual preference framework explains the opposing forces of the doer and the planner preferences in consumption choices. A doer preference is associated with a short-term orientation, whereas planner preferences are associated with a long-term orientation. Among the three behavioral factors identified by the BLCH, self-control, framing, and mental accounting (Shefrin & Thaler, 1988), this study isolates self-control as the behavioral influence impacting consumption decisions. Lynn et al. 181 Figure 1. BLCH and LCH Conceptual Models of Consumption Note: Adapted from www.economicshelp.org LCH model. Hypothesis This study proposes the following hypothesis: H1: Overspending is negatively associated with retirement preparation. Methodology Data Sets This study utilized three separate cross-sectional datasets: the 2019 SHED, the 2019 SCF, and the 2018 NFCS. Using these nationally representative surveys adds to the credibility of the findings because each survey offers similar yet different measures of the constructs, enhancing the study’s robustness. The samples were limited to full-time workers, assuming that individuals who were not working would not be actively preparing for retirement. This restriction was based on a respondent’s employment status in each data set, ignoring the employment status of a spouse’s working status. This restriction reduced the SHED sample size from 12,173 to 6,651 respondents, the SCF sample size from 5,777 to 3,361, and the NFCS sample size from 5 A data limitation of this measurement was that researchers have operationalized it both as spending less than income (e.g., Borsch-Supan & Lusardi, 2003) as well as a retirement preparation measurement (e.g., Ferdous et al., 2010; Heckman & Hanna, 2015; Kim 27,091 to 10,800. Operationalizations of each measure were based upon data set availability. Overspending Variables From an advisor’s perspective, overspending is straightforward to measure: it occurs when spending exceeds income. To evaluate overspending, one subjective and three objective measures are utilized, offering different perspectives and enabling a comprehensive analysis of the phenomenon. Spending More Than Income. Operationalizing overspending with spending more than income (SMTI) was consistent with prior research (Borsch-Supan & Lusardi, 2003).5 If income exceeds expenditures, the household was assumed to be saving. If expenditures exceed income, the household was assumed to be overspending. The survey question used to operationalize SMTI was similar in all three data sets, and read as, “Over the past year, would you say your household’s spending was less than, more than, or equal to your household’s income?” & Hanna, 2017; Yuh & Hanna, 2010). Because retirement preparation was not operationalized in this way in the current study, it was assumed to not cause any methodological concerns, but still warranted disclosure. http://www.economicshelp.org/ Financial Services Review, 33(1) 182 Only respondents that selected SMTI were treated as overspending. The 22 responses in the SHED and 290 in the NFCS of “don’t know” or “prefer not to say” were treated as missing data. Revolving Credit Card Debt. Experts suggest that credit cards facilitate overspending because they were easy to obtain and use, they reduce the pain of payment, and the credit limit and minimum payment due displayed on statements may act as anchors that induce higher levels of consumption Gärling & Ranyard, 2020. Revolving credit card debt (RCCD) was also considered financially damaging because of the interest and fees charged. All three surveys had similar questions to operationalize RCCD. If a respondent indicated that they had a credit card and did not pay off the balance every month, they were treated as RCCD. The measure had three categories: (a) credit card revolvers, (b) non- credit card revolvers, and (c) non-credit card holders. The 23 “don’t know” or “prefer not to say” observations in the SHED and 190 in the NFCS were treated as missing data. Alternative Financial Service Usage. Alternative financial services (AFS) usage was a third measure of overspending. AFS were high- cost debt instruments, such as payday loans or tax refund anticipation loans, (Robb et al., 2015) used to “…fund purchases of desired consumer products” (Gärling & Ranyard, 2020, p. 272). Because AFS products were egregiously financially disadvantageous to the consumer, utilizing them indicated spending beyond one’s budget. With the SHED, if a respondent indicated they had used either a payday loan, auto title loan, or a pawn shop loan, or a tax refund advance in the past 12 months, they were treated as overspending. The 26 “refused to answer” observations were treated as missing data. The SCF assesses whether a respondent had used a payday loan in the past year. In the NFCS, if a respondent had used either an auto title loan, payday loan, tax refund advance, pawn shop loan, or used a rent-to-own store in the past five years, they were treated as utilizing AFS. The 266 responses of “don’t know” or “prefer not to say” were treated as missing data. Absence of Emergency Fund. The absence of an emergency fund (AEF) was the final binary measurement of overspending. Maintaining an emergency fund was considered a critical component of managing personal finances to weather unexpected expenses or income dips (Farrell et al, 2019). The SHED question “Have you set aside emergency or rainy day funds that would cover your expenses for 3 months in the case of sickness, job loss, economic downturn, or other emergencies?” operationalizes AEF, where a “no” response indicated AEF. The 17 “refused to answer” observations were treated as missing data. The SCF question that captures AEF reads, “If tomorrow you experienced a financial emergency that left you unable to pay all of your bills, how would you deal with it?” If a respondent answered anything other than “spend out of savings or investments,” they were treated as AEF. The question, “Have you set aside emergency or rainy-day funds that would cover your expenses for 3 months, in case of sickness, job loss, economic downturn, or other emergencies?” was used to operationalize AEF in the NFCS, where “no” responses were treated as AEF. The 406 “I don’t know” and “prefer not to say” responses were treated as missing data. Retirement Preparation Variables Retirement preparation was a challenging construct to measure objectively because individuals have unique goals and requirements that necessitate adjustments to life events. To address this challenge, four measurements were used to operationalize retirement preparation: three were objective measures and one was subjective. Retirement Account Ownership. Retirement account ownership was assumed to be a viable proxy for retirement preparation because researchers (e.g., Lim & Lee, 2021; Oakley & Kenneally, 2019; Sturr et al., 2021) consider defined contribution plans to be the foundational vehicle for funding retirement income. In the SHED, if a respondent indicated owning any of the following account types, they were treated as owning a retirement account: 401(k), 403(b), Keogh, other defined contribution plan through an employer, pension with a defined benefit, IRA, Roth IRA, savings outside a retirement account, Lynn et al. 183 a business or real estate that will provide income in retirement, or other retirement savings. The 529 missing observations were treated as missing data. In the SCF, if a respondent had any of the following types of accounts, they were treated as owning a retirement account: Keoghs, pensions, retirement, or tax-deferred savings plans. In the NFCS, if a respondent owned any of the following types of accounts, they were treated as owning a retirement account: A pension plan, a Thrift Savings Plan (TSP) or a 401(k), an IRA, Keogh, Simplified Employee Pension (SEP), or any other type of retirement account that they had set up themselves or through an employer. The 846 “don’t know” or “prefer not to say” responses were treated as missing data. Active Contribution to a Retirement Account. Active contribution to a retirement account (ACRA) was reflective of active household retirement savings, and thus is assumed to be a viable operationalization of retirement preparation (Sturr et al., 2021). In the SCF, if anyone in the household made contributions to IRAs or Keoghs in the previous year or contributed to a traditional pension, 401(k), 403(b), TSP, profit sharing plan, supplemental retirement annuity, cash balance plan, portable cash option plan, SEP, Simplified Incentive Match Plan for Employees (SIMPLE), money purchase plan, stock purchase plan (ESOP), 457 plan, or other retirement account, they were treated as ACRA. If respondents in the NFCS regularly contributed to a Thrift Savings Plan (TSP), 401(k), or IRA, they were treated as “yes” for ACRA. The 4,024 missing observations are treated as missing data. The SHED did not measure ACRA. Retirement Account Assets. Retirement account assets require a large enough balance to sustain steady distributions for the duration of one’s retirement. While there was no one formula to calculate the “right” amount of retirement account assets, retirement account assets can represent retirement preparation because they are a source of someone’s future retirement income 6 In the SHED, respondents were asked if they considered themselves to be retired, which was separate from the employment status measure used to apply the sample restriction. Those who considered stream. In the SHED, retirement account assets were captured with the question: “Approximately how much money do you currently have saved for retirement?” The 1,702 missing observations were treated as missing data. In the SCF, retirement account assets were measured by the combination of Roth IRAs, roll-over IRAs, regular or other IRAs, Keoghs, and employer- sponsored plans balances. The log of retirement account assets was used to account for skewness associated with this measure. The NFCS did not measure retirement account asset balance. Perceived Retirement Preparation. Perceived retirement preparation was measured by a subjective question capturing how well a respondent felt financially prepared for retirement. In the SHED, respondents were treated as “yes” for perceived retirement preparation if they answered yes to the question, “Do you think that your retirement savings plan is currently on track?” The 1,481 missing observations were treated as missing data.6 The SCF question that captures perceived retirement preparation reads “How would you rate the retirement income you receive (or expect to receive) from all sources?” Respondents could respond on a five-point Likert type scale with a score of one indicative of the lowest level of perceived retirement preparation and five indicating the highest level of perceived retirement preparation. The NFCS question that captured perceived retirement preparation reads, “I worry about running out of money in retirement.” Respondents could respond on a seven-point Likert-type scale, which was reverse coded to directionally align with measure in the SHED and SCF, such that one represents the lowest level of perceived retirement preparation and seven represents the highest. The 211 “don’t know” and “prefer not to say” responses were treated as missing data. Control Variables Where possible, the following socio-economic and behavioral characteristics were controlled for: Age, marital status, race, gender, education themselves to be retired were not asked the perceived retirement preparation question, which contributed to the missing observations of this measure. Financial Services Review, 33(1) 184 of a respondent, parental education, income, health, objective and subjective financial knowledge, risk tolerance. Controlling for these factors which were known to be associated with retirement preparation (e.g., Muratore & Earl, 2010; Ozgen & Esiyok, 2020), help prevent their effects from being captured in the error term. Analyses Regressions were used in each data set to analyze the relationship between overspending (the independent variables) and retirement preparation (the dependent variables). Listwise deletion was used to handle missing data in the SHED and NFCS. The Repeated-Imputation Inference (RII) method as well as bootstrapped standard errors, was used to handle missing data in the SCF. Figure 2. Empirical Model Results Descriptive Statistics Overspending Descriptive Statistics. Table 1 displays the overspending descriptive statistics. In the SHED, only about one-sixth of the sample considered themselves to be overspending, as measured by SMTI (16%). Even fewer were overspending as measured by AFS usage (5%). Over half the respondents, however, were overspending as measured by RCCD (52%), and almost half the sample overspent as measured by AEF (47%). In the SCF, very few respondents were overspending as measured by SMTI (4%) and AFS usage (3%). Over one-third of the sample was overspending as measured by RCCD (38%). Half the sample overspent as measured by AEF (50%). In the NFCS, one-fifth the sample overspent, as measured by SMTI (20%). About one-third of the sample overspent, as measured by AFS usage in the past five years (32%). Almost half the sample overspent, as measured by AEF (46%), and over a third of the sample overspent, as measured by RCCD (36%). Lynn et al. 185 Retirement Preparation Descriptive Statistics. Retirement preparation descriptive statistics were also listed in Table 1. In the SHED, over four-fifths of the sample owned a retirement account (84%), but a smaller portion felt on track for retirement (52%). Almost half of the sample had under $50,000 saved for retirement (46%), whereas almost a fourth of the sample had under $10,000 saved for retirement (24%). In the SCF, the majority of respondents owned a retirement account (64%), with about half of respondents ACRA (49%). On a scale from zero to $16,400,000, the median retirement account balance was $288,783. The average respondent reported moderate perceived retirement preparation (M = 3.03, SD = 1.25), on a scale from one to five. In the NFCS, four-fifths of the sample owned a retirement account (80%) and even more indicated ACRA (88%). Perceived retirement preparation had a mean of 3.21 on a scale of one to seven. Regression Results: Retirement Account Ownership The full list of results was displayed in Table 2. The SHED results show those who SMTI had 31% lower odds of owning a retirement account (OR = 0.69, p < 0.001); those who RCCD had 22% lower odds of owning a retirement account (OR = 0.78, p < 0.001); those with AEF had 65% lower odds of owning a retirement account (OR = 0.35, p < 0.001). The SCF results reveal that those who RCCD had 26% lower odds of owning a retirement account (OR = 0.74, p < 0.001); those with AEF had 30% lower odds of owning a retirement account (OR = 0.70, p < 0.001). The NFCS results show that those who RCCD had 33% lower odds of owning a retirement account (OR = 0.67, p < 0.001), and those with AEF had 53 percent lower odds of owning a retirement account (OR = 0.47, p < 0.001). Financial Services Review, 33(1) 186 Table 1. Descriptive Statistics SHED, N = 6,651 SCF, N = 3,361 NFCS, N = 10,800 Variable n % n M/% SD min max n M/% SD min max Retirement preparation Owns a retirement account 5,104 83.50% 2,135 63.53% - - - 8,012 80.49% - - - Actively contributes to a retirement account - - 1,652 49.15% - - - 5,954 87.87% - - - Retirement account assets - - 3,361 $2,000 $288,783 0 $16,400,000 - - - - - <$24,999 1,732 35.00% - - - - - - - - - - $25,000-$249,999 1,902 38.43% - - - - - - - - - - $250,000- $1,000,000+ 1,314 26.57% - - - - - - - - - - Perceived retirement preparation: on track 5,170 51.59% 3,361 3.03 1.25 1 5 10,589 3.21 1.93 1 7 Overspending Spends more than income 1,034 15.60% 123 3.67% - - - 2,104 20.02% - - - Revolves credit card debt 3,437 51.87% 1,288 38.31% - - - 3,766 35.50% - - - Uses AFS 351 5.30% 101 3.01% - - - 3,356 31.86% - - - Absence of Emergency Fund 3,143 47.37% 1,678 49.93% - - - 4,808 46.26% - - - Note. Analyses were weighted. Data from the 2019 SHED, 2019 SCF, and 2018 NFCS. Samples restricted to only those working as a paid employee. Lynn et al. 187 Table 2. Binary Logistic Regression of Retirement Account Ownership SHED, n = 5,069 SCF, n = 3,361 NFCS, n = 9,114 Variable B SE OR B SE OR B SE OR Intercept 0.39 0.34 1.48 -7.43*** 0.87 0.00 -1.14*** 0.3 0.32 Spends more than income -0.36* 0.14 0.69 0.03 0.22 1.03 0.02 0.09 1.02 Revolves credit card debt -0.25*** 0.04 0.78 -0.30*** 0.07 0.74 -0.40*** 0.06 0.67 Uses AFS 0.14 0.23 1.15 0.14 0.19 1.15 -0.09 0.09 0.91 Absence of Emergency Fund -1.05*** 0.15 0.35 -0.35*** 0.07 0.70 -0.75*** 0.08 0.47 Control variables Age (under 35) - - - 0.01** 0.00 1.01 - - - 35-44 0.54*** 0.16 1.71 - - - 0.28** 0.10 1.32 45-54 0.74*** 0.18 2.10 - - - 0.46*** 0.11 1.59 55-64 0.92*** 0.20 2.51 - - - 0.72*** 0.15 2.06 65 or older 1.06* 0.43 2.90 - - - 0.42 0.30 1.53 Marital status (married) living with partner or never married -0.49*** 0.14 0.61 0.10 0.10 1.11 -0.32*** 0.09 0.72 separated or divorced -0.29 0.18 0.75 -0.01 0.11 0.99 -0.28* 0.13 0.75 widowed 1.53 0.96 4.63 - - - -0.74** 0.27 0.48 Race of respondent (White) - - - - - - -0.01 0.08 0.99 Black 0.10 0.18 1.11 -0.05 0.08 0.95 - - - Hispanic -0.12 0.17 0.95 -0.45*** 0.13 0.64 - - - Asian or other -0.05 0.24 0.87 0.10 0.17 1.11 - - - Gender of respondent (male) 0.15 0.12 1.17 0.12 0.10 1.13 0.38*** 0.08 1.45 Education (high school degree) no high school degree -0.79** 0.27 0.45 -0.76*** 0.20 0.47 -0.69 0.37 0.50 Financial Services Review, 33(1) 188 some college 0.18 0.15 1.20 0.13 0.09 1.14 0.12 0.11 1.12 bachelor's degree 1.03*** 0.18 2.80 0.62*** 0.12 1.86 0.61*** 0.15 1.85 advanced - - - 0.90*** 0.13 2.46 0.55** 0.19 1.74 Parental education (high school degree) no high school degree 0.01 0.23 1.01 -0.36** 0.12 0.70 0.02 0.19 1.02 some college 0.06 0.16 1.06 -0.08 0.11 0.92 -0.04 0.11 0.96 bachelor's degree -0.14 0.17 0.87 0.03 0.10 1.03 -0.48*** 0.13 0.62 advanced degree 0.13 0.22 1.14 -0.39* 0.17 0.67 Risk tolerance 0.08** 0.02 1.08 0.02 0.02 1.02 0.10*** 0.02 1.10 Income (under $25,000) - - - 0.64*** 0.07 1.90 - - - $25,000-$49,999 0.70*** 0.18 2.02 - - - 0.67*** 0.12 1.94 $50,000-$74,999 1.00*** 0.19 2.71 - - - 1.27*** 0.13 3.56 $75,000-$99,999 0.91*** 0.22 2.46 - - - 2.00*** 0.17 7.39 $100,000-$149,999 1.43*** 0.24 4.20 - - - 2.15*** 0.18 8.63 $150,000 or more 1.40*** 0.30 4.04 - - - 2.43*** 0.27 11.36 Objective financial knowledge 0.34*** 0.06 1.41 0.17*** 0.04 1.19 0.18*** 0.03 1.20 Subjective financial knowledge - - - -0.01 0.02 0.99 0.01 0.03 1.01 Financial attitude - - - - - - 0.06** 0.02 1.06 Health (excellent) fair -0.14 0.12 0.87 -0.16 0.09 0.85 - - - poor 0.27 0.44 -1.18** 0.34 0.31 - - - Model Fit Statistics C-statistic 0.88 Log likelihood -1,772 C-statistic 0.83 Pseudo R2 0.32 Pseudo R2 0.20 Pseudo R2 0.24 Note. Analyses were weighted. SCF income was logged. Data from the 2019 SHED, 2019 SCF, and 2018 NFCS. Restricted samples to only those working full-time. * p < .05; ** p < .01; *** p < .001. Lynn et al. 189 Retirement Results: Active Contribution to a Retirement Account The SCF results (Table 3) indicate that those who RCCD had 21% lower odds of ACRA (OR = 0.79, p < 0.001). The NFCS results show that those who reported AEF were associated with 37% lower odds of ACRA (OR = 0.63, p < 0.001). Table 3. Binary Logistic Regression of Active Contributions to a Retirement Account SCF, n = 3,361 NFCS, n = 6,351 Variable B SE OR B SE OR Intercept -6.16*** 0.80 0.00 0.95*** 0.44 2.57 Spends more than income 0.06 0.22 1.06 -0.16 0.13 0.86 Revolves credit card debt -0.24*** 0.06 0.79 -0.11 0.13 1.00 Uses AFS -0.02 0.18 0.98 -0.46 0.11 0.89 Absence of Emergency Fund -0.12 0.07 0.89 -0.11*** 0.11 0.63 Control variables Age (under 35) -0.01* 0.00 0.99 - - - 35-44 - - - 0.08 0.14 1.09 45-54 - - - 0.00 0.14 1.00 55-64 - - - -0.35* 0.15 0.71 65 or older - - - -0.83** 0.24 0.44 Marital status (married) living with partner or never married 0.19* 0.08 1.21 0.13 0.12 1.15 separated or divorced 0.02 0.10 1.02 0.01 0.16 1.01 widowed - - - -0.37 0.34 0.69 Race of respondent (White) 0.01 0.11 1.01 Black -0.09 0.09 0.91 - - - Hispanic -0.34** 0.12 0.71 - - - Asian or other 0.28* 0.12 1.32 - - - Gender of respondent (male) 0.23* 0.09 1.26 0.03 0.10 1.03 Education (high school degree) no high school degree -0.76*** 0.19 0.47 -0.73 0.58 0.48 some college -0.09 0.08 0.91 -0.07 0.17 0.93 bachelor's degree 0.26** 0.09 1.30 -0.17 0.18 0.84 advanced 0.16 0.13 1.17 -0.50* 0.21 0.60 Parental education (high school degree) no high school degree -0.33** 0.10 0.72 -0.29 0.23 0.74 some college 0.10 0.10 1.11 0.26 0.16 1.30 bachelor's degree -0.04 0.08 0.96 0.07 0.15 1.07 Financial Services Review, 33(1) 190 advanced degree - - - 0.20 0.19 1.23 Risk tolerance 0.01 0.02 1.01 0.03 0.02 1.03 Income 0.55*** 0.06 1.73 - - - $25,000-$49,999 - - - 0.60* 0.24 1.82 $50,000-$74,999 - - - 0.95*** 0.25 2.58 $75,000-$99,999 - - - 1.10*** 0.25 3.01 $100,000-$149,999 - - - 1.53*** 0.27 4.60 $150,000 or more - - - 2.18*** 0.32 8.84 Objective financial knowledge 0.11* 0.04 1.12 0.00 0.04 1.00 Subjective financial knowledge -0.01 0.01 0.99 0.07 0.05 1.08 Financial attitude - - - -0.06 0.03 0.94 Health (excellent) fair -0.13 0.09 0.88 - - - poor -0.60 0.31 0.55 - - - Model Fit Statistics Log likelihood -2,096 C-statistic 0.69 Pseudo R2 0.10 Pseudo R2 0.06 Note. Analyses were weighted. SCF income was logged. Data from the 2019 SCF, and 2018 NFCS. Restricted samples to only those working full-time. * p < .05; ** p < .01; *** p < .001. Regression Results: Retirement Account Balance The SHED results (Table 4) reveal that RCCD (B = -0.29, p < 0.001) and AEF (B = -0.92, p < 0.001) had a significant negative relationship with retirement account balance. SMTI and AFS usage did not have a significant relationship with retirement account balance. The SCF results indicate that RCCD (β = -0.05, p < 0.001) and AEF (β = -0.08, p < 0.001) had significant negative relationships with retirement account balance. Lynn et al. 191 Table 4. Ordinal (SHED) and OLS (SCF) Regressions of Retirement Account Asset Balance SHED, n = 4,207 SCF, n = 3,361 Variable B SE B SE B β Intercept - - -13.35*** 1.33 - Spends more than income -0.18 0.12 -0.09 0.39 0.00 Revolves credit card debt -0.29*** 0.05 -0.64*** 0.12 -0.05 Uses AFS -0.27 0.28 -0.24 0.39 -0.01 Absence of Emergency Fund -0.92*** 0.09 -0.98*** 0.15 -0.08 Control variables Age (under 35) - - 0.05*** 0.01 0.12 35-44 1.74*** 0.13 - - - 45-54 2.63*** 0.14 - - - 55-64 3.21*** 0.15 - - - 65 or older 3.39*** 0.18 - - - Marital status (married) living with partner or never married -0.21* 0.11 0.27 0.17 0.01 separated or divorced -0.30* 0.13 0.28 0.21 0.02 widowed -0.43 0.26 - - - Race of respondent (White) Black -0.83*** 0.15 -0.54* 0.22 -0.03 Hispanic -0.42** 0.14 -0.90** 0.28 -0.05 Asian or other 0.07 0.16 0.43 0.31 0.02 Gender of respondent (male) -0.20* 0.08 0.08 0.15 0.01 Education (high school degree) no high school degree -0.07 0.46 -0.80** 0.31 -0.03 some college 0.01 0.13 -0.01 0.20 0.00 bachelor's degree 0.42 0.14 1.20*** 0.21 0.09 advanced - - 1.47*** 0.22 0.11 Parental education (high school degree) no high school degree -0.33 0.17 -0.82*** 0.23 -0.05 some college -0.04 0.12 -0.13 0.21 -0.01 bachelor's degree or beyond -0.02 0.12 0.08 0.18 0.01 advanced degree -0.04 0.12 - - - Risk tolerance 0.14*** 0.02 0.13*** 0.03 0.06 Income (under $25,000) - - 1.35*** 0.12 0.43 $25,000-$49,999 0.53 0.34 - - - $50,000-$74,999 1.27*** 0.33 - - - $75,000-$99,999 1.70*** 0.34 - - - Financial Services Review, 33(1) 192 $100,000-$149,999 2.19*** 0.34 - - - $150,000 or more 2.87*** 0.35 - - - Objective financial knowledge 0.19*** 0.05 0.51*** 0.09 0.07 Subjective financial knowledge - - 0.05 0.03 0.02 Health (excellent) fair -0.16 0.08 -0.34 0.18 -0.03 poor -0.22 0.48 -1.46* 0.64 -0.03 Model Fit Statistics Pseudo R2 0.34 R2 0.30 Adjusted R2 0.29 Note. Analyses were weighted. SCF income was logged. Data from the 2019 SHED, 2019 SCF, and 2018 NFCS. Restricted samples to only those working full-time. * p < .05; ** p < .01; *** p < .001 Regression Results: Perceived Retirement Preparation The SHED results (Table 5) show that those who SMTI had 43% lower odds of feeling on track for retirement (OR = 0.57, p < 0.001); those who RCCD had 23% lower odds of feeling on track for retirement (OR = 0.77, p < 0.001); and AEF had 65% lower odds of feeling on track for retirement (OR = 0.35, p < 0.001). The SCF results indicate that SMTI (β = -0.06, p < 0.001), RCCD (β = -0.05, p < 0.001), AFS usage (β = - 0.04, p < 0.001), and AEF (β = -0.06, p < 0.001) had a significant negative relationship with perceived retirement preparation. The NFCS results indicate that only those who utilized AFS were more likely to perceive retirement preparation negatively (β = -0.03, p = 0.003).7 7 Additional robustness checks were conducted, including (a) an unweighted sample, (b) multiple imputation, (c) excluding non-credit card holders, (d) an alternative STTP, and (e) and alternative measure for SMTI. Interaction effects were tested between overspending and income with retirement preparation, due to the suspicion that the relationship between overspending and retirement preparation may have depended on the level of income. To explore the reliability of overspending and retirement preparation constructs, additive scales, and confirmatory factor analyses (CFA) were employed. While the robustness checks reveal variation between data sets and various significant interaction effects between overspending and income with retirement preparation, there is not enough evidence to cause concern over the validity of the primary analyses. The results of robustness checks are available upon request. Lynn et al. 193 Table 5. Logistic (SHED) and OLS (SCF and NFCS) Regression of Perceived Retirement Preparation SHED, n = 4,317 SCF, n = 3,361 NFCS, n = 9,524 Variable B SE OR B SE B β B SE B β Intercept -0.42 0.30 0.66 0.65* 0.27 - 6.27*** 0.16 - Spends more than income -0.55*** 0.12 0.57 -0.48*** 0.11 -0.06 -0.04 0.05 -0.01 Revolves credit card debt -0.26*** 0.05 0.77 -0.15*** 0.02 -0.05 0.01 0.03 0.00 Uses AFS 0.21 0.24 1.23 -0.36** 0.12 -0.04 -0.14** 0.05 -0.03*** Absence of Emergency Fund -1.06*** 0.09 0.35 -0.16*** 0.04 -0.06 0.02 0.04 0.01 Control variables Age (under 35) - - - -0.00 0.00 0.00 - - - 35-44 0.13 0.12 1.14 - - - -0.26*** 0.05 -0.06*** 45-54 0.09 0.13 1.09 - - - -0.37*** 0.05 -0.08*** 55-64 0.36** 0.13 1.43 - - - -0.34*** 0.06 -0.06*** 65 or older 0.32 0.21 1.38 - - - -0.20 0.10 -0.02 Marital status (married) living with partner or never married -0.24* 0.11 0.79 0.08* 0.04 0.03 0.05 0.05 0.01 separated or divorced -0.32 0.14 0.73 0.04 0.05 0.01 0.01 0.06 0.00 widowed 0.13 0.47 1.14 - - - 0.07 0.16 0.00 Race of respondent (White) - - - - - - -0.04 0.04 -0.01 Black -0.10 0.15 0.91 0.22*** 0.06 0.06 - - - Hispanic -0.57*** 0.15 0.99 0.03 0.08 0.01 - - - Asian or other -0.01 0.18 0.56 -0.21** 0.06 -0.04 - - - Gender of respondent (male) -0.03 0.09 0.97 -0.12** 0.04 -0.04 -0.08* 0.04 -0.02* Education (high school degree) no high school degree -0.61 0.39 0.55 0.08 0.11 0.00 0.16 0.23 0.01 some college -0.02 0.13 0.98 -0.04 0.06 0.00 0.02 0.06 0.01 bachelor's degree 0.32* 0.13 1.37 0.12 0.07 0.04 0.07 0.07 0.02 Financial Services Review, 33(1) 194 advanced degree - - - 0.09 0.07 0.03 0.02 0.08 0.00 Parental education (high school degree) no high school degree -0.05 0.20 0.95 -0.10 0.07 -0.02 0.09 0.10 0.01 some college -0.11 0.12 0.90 -0.02 0.07 0.00 0.03 0.06 0.01 bachelor's degree or beyond 0.04 0.12 1.04 0.01 0.05 0.00 0.08 0.06 0.02 advanced degree 0.01 0.13 1.01 - - - 0.11 0.07 0.02 Risk tolerance 0.12*** 0.02 1.12 0.06*** 0.01 0.12 -0.01 0.01 -0.01 Income - - - 0.19*** 0.02 0.27 - - - $25,000-$49,999 0.79*** 0.22 2.19 - - - -0.02 0.09 -0.00 $50,000-$74,999 0.98*** 0.21 2.66 - - - -0.07 0.09 -0.01 $75,000-$99,999 1.06*** 0.22 2.89 - - - -0.08 0.09 -0.02 $100,000-$149,999 1.08*** 0.21 2.95 - - - 0.07 0.10 0.01 $150,000 or more 1.33*** 0.23 3.79 - - - 0.34** 0.11 0.05 Objective financial knowledge 0.01 0.05 1.01 -0.01 0.03 -0.01 -0.03* 0.01 -0.02 Subjective financial knowledge - - - 0.07*** 0.01 0.11 0.05** 0.02 0.03 Financial attitude - - - - - - -0.62*** 0.01 -0.63 Health (excellent) fair -0.49*** 0.09 0.61 -0.24*** 0.05 -0.09 - - - poor -1.03* 0.42 0.36 -0.73*** 0.19 -0.07 - - - Model Fit Statistics C-statistic 0.81 R2 0.16 R2 0.42 Pseudo R2 0.23 Adjusted R2 0.15 Note. The analyses were weighted. SCF income was logged. Data from the 2019 SHED, 2019 SCF, and 2018 NFCS. Samples restricted to only those working full-time. * p < .05; ** p < .01; *** p < .001. Lynn et al. 195 Discussion Out of all the retirement preparation analyses conducted, a significant negative relationship was found in many but not all the analyses, partially supporting H1. The lack of consistency in results across the data sets temper the conclusions drawn. Retirement Account Ownership Overspending, as measured by SMTI, was found to have a significant negative relationship with retirement account ownership in the SHED analysis, but not in the SCF or NFCS analyses. RCCD was significant negatively related to retirement account ownership in the SHED, SCF, and NFCS analyses. AFS usage was not related to retirement account ownership in the SHED, SCF, or NFCS analysis. AEF was negatively associated with retirement account ownership in both the SHED, SCF, and the NFCS analyses. Because some overspending measures, but not all, exhibited a significant negative relationship with retirement account ownership, only partial support for H1 was noted. Active Contributions to a Retirement Account Overspending, as measured by SMTI and AFS usage, was not related to ACRA in the SCF or NFCS analyses. RCCD was found to have a significant relationship with ACRA in the SCF analysis, but not the NFCS analysis. AEF was negatively associated with ACRA in the NFCS analysis, but not in the SCF analysis. As such, H1 was partially supported. 8 A possible explanation for only about half of the overspending and retirement preparation analyses being significant is that overspending has become normalized in American culture. A few ideas are provided here to describe what normalized overspending behavior could look like. First, respondents may not realize they are spending more than their income. Second, credit cards could be considered necessary for some households to manage their cashflow, and thus does not significantly relate to retirement preparation behavior. Third, AFS usage could be seen by some as an acceptable financing resource to occasionally bridge cashflow gaps when needed and does not correspond to a shift in one’s retirement preparation behavior. Fourth, an emergency Retirement Account Balance Overspending, as measured by SMTI and AFS usage, were not related to a respondent’s retirement account balance in the SHED analysis or the SCF. RCCD was negatively related to retirement account balances in the SHED and SCF analyses. AEF was negatively associated with retirement account balances in both the SHED and SCF analyses. Once again, H1 was only partially supported. Perceived Retirement Preparation Overspending, as measured by SMTI and credit card revolving, was found to have a significant negative relationship with perceived retirement preparation in the SHED and SCF analyses, but not the NFCS analysis. Overspending, as measured by AFS usage, was negatively associated with perceived retirement preparation in the SCF and NFCS analyses, but not the SHED analysis. Overspending, as measured by AEF, was negatively related to perceived retirement preparation in both the SHED and SCF analyses, but not the NFCS analysis. Because some, but not all, overspending showed a significant negative relationship with perceived retirement preparation, H1 could only be partially supported.8 Implications Borrowing using credit cards is incredibly easy for the average consumer, both for those who can afford it and those who cannot. One policy implication is to make it more challenging to qualify or transact using consumer credit (Gärling & Ranyard, 2020). For example, lenders could fund may seem out of reach or not necessary for cash flow management by some but does not necessarily mean a household is not preparing for retirement. Fifth, overspending could be occurring simultaneously while preparing for retirement. Sixth, automatic enrollment and automatic contribution increases by employers could be the cause of a significant negative relationship between overspending and retirement preparation where the employer, rather than the respondent, is responsible for the respondent’s retirement preparation behavior. In sum, the analyses may reveal a new normal where overspending does not necessarily reflect a respondent’s retirement preparation behavior. Financial Services Review, 33(1) 196 require setting credit limits with reference to a client’s respective affordability level or limit the number of credit cards that may be issued to an individual (Gärling & Ranyard, 2020). Because credit limits and minimum payments have been found to unintentionally act as anchor points (Gärling & Ranyard, 2020; Tescher and Stone, 2022), tightening the minimum payment due formula could potentially help households get out of credit card debt faster than current standards (Tescher & Stone, 2022). An industry implication to help promote the use of emergency funds is to implement the automatic enrollment of employees in savings accounts linked to their 401(k)s, which may potentially be matched by employers. This is now possible due to the recent passing of the Secure Act 2.0 (Senate Finance Committee, 2022). Studies have shown that automatic enrollment increases participation rates (see Beshears et al., 2010). Advisors can also make clients aware of this new opportunity. Client overspending behavior is an occasional obstacle advisors face. Despite a financial planner’s best efforts, it can seem like some clients at risk of running out of money fail to realize the consequences of overspending. One way to indirectly assist individuals who overspend is to focus on augmenting self-control levels, rather than focusing on cutting back overspending. This may be particularly impactful in youth. At the practitioner level, as a value-add to clients, advisors could provide clients with young children or grandchildren educational games or activities that teach self-control in personal finances. One such example is the CASHFLOW® Game. Some have suggested that potential life hacks may help increase levels of self-control when it comes to overspending. Gärling and Ranyard (2020) highlighted techniques originally tested by Karlsson (2003) as plausible modern solutions to help bolster self- control and thereby curbing overspending. Additionally, financial practitioners could coach their clients to limit themselves to owning only one credit card, rather than multiple credit cards, to simplify financial decisions where borrowers often make mistakes (Gärling & Ranyard, 2020). Alternatively, budgeting could be promoted as a tool to help prevent overspending. Simplistic budgeting techniques should be promoted over complex ones (Kim & Choi, 2018), such as minimizing the number of categories withing a budget (Kim, 2022). Budgeting using mental accounting could also be encouraged, as researchers have found it may help improve savings and consumption decisions (Sui et al., 2021; Xiao & O’Neill, 2018). Xiao and O’Neil (2018) also made the point that simply suggesting the practice of budgeting is not enough, because households may not know how to do so. Households would benefit from advisors teaching them budgeting techniques and being shown projections of the hypothetical impact of prioritizing savings in the budget. Another practitioner implication is to ensure that all clients receive education on overspending. Clients may not realize that RCCD or AEF is a form of overspending. They may also not realize the negative implications of AFS usage or SMTI. Advisors should educate clients and prospects regarding the financial dangers of normalizing overspending. In this regard, because overspending may be a manifestation of a deeper psychological issue, it is important for advisors to be self-aware of their own limitations in providing help (Lutter, 2022) . Conclusion The results from this study suggest that overspending is negatively related to retirement preparation. Logically, it seems obvious that overspending should be negatively related to retirement preparation. The BLCH was found to be a valuable theoretical framework for the study of the relationship between overspending and retirement preparation. Because the overspending measurements were negatively related to retirement preparation in a little over half the analyses, it is possible that a new cultural norm where one’s overspending behavior does not necessarily mean that the person is not also preparing for retirement may be existence. This possibility is worthy of future study. References Achtziger, A. (2022). Overspending, debt, and poverty. Current Opinion in Psychology, 46, 101342. Lynn et al. 197 https://doi.org/10.1016/j.copsyc.2022.1013 42 Adams, G. A., & Rau, B. L. (2011). Putting off tomorrow to do what you want today: Planning for retirement. American Psychologist, 66(3), 180–192. https://doi.org/10.1037/a0022131 Bennett, J. N. (2021). A dollar’s worth: Inflation is real. In PAGE ONE Economics (Issue October). https://doi.org/10.5860/choice.51-1007 Beshears, J., Choi, J. J., Laibson, D., & Madrian, B. C. (2010). The impact of employer matching on savings plan participation under automatic enrollment. In David Wise (Ed.), Research Findings in the Economics of Aging (pp. 311–327). University of Chicago Press. Board of Governors of the Federal Reserve System. (2021). Economic well-being of U.S. households in 2020 (5). Borsch-Supan, A., & Lusardi, A. (2003). Measuring life-cycle saving. In Borsch- Supan, A. (Ed.), Life-cycle savings and public policy: A cross-national study of six countries. Academic Press. Choe, Y., & Kan, C. (2021). Budget depreciation: When budgeting early increases spending. Journal of Consumer Research, 47(6), 937– 958. https://doi.org/10.1093/jcr/ucaa049 Cynamon, B. Z., & Fazzari, S. M. (2008). Household debt in the consumer age: Source of growth--Risk of collapse. Capitalism and Society, 3(2). https://doi.org/10.2202/1932-0213.1037 Fan, L., Stebbins, R., & Kim, K. T. (2022). Skint: Retirement? Financial hardship and retirement planning behaviors. Journal of Family and Economic Issues, 43, 354–367. https://doi.org/10.1007/s10834-021-09779- z Farrell, Diana, Fiona Greig, and Chenxi Yu. (2019). “Weathering volatility 2.0: A monthly stress test to guide savings.” JPMorgan Chase Institute. https://institute.jpmorganchase.com/institut e/research/household-income- spending/report-weathering-volatility-2.0. Ferdous, N., Pinjari, A. R., Bhat, C. R., & Pendyala, R. M. (2010). A comprehensive analysis of household transportation expenditures relative to other goods and services: An application to United States consumer expenditure data. Transportation, 37(3), 363–390. https://doi.org/10.1007/s11116-010-9264-2 Gärling, T., & Ranyard, R. (2020). The psychological perspective on the antecedents and consequences of consumer borrowing. In Psychological Perspectives on Financial Decision Making (pp. 267– 290). Springer International Publishing. https://doi.org/10.1007/978-3-030-45500- 2_12 Heckman, S. J., & Hanna, S. D. (2015). Individual and institutional factors related to low-income household saving behavior. Journal of Financial Counseling and Planning, 26(2), 187–199. https://doi.org/10.1891/1052- 3073.26.2.187 Karlsson, N. (2003). Consumer self-control strategies: An empirical study of their structure and determinants. Journal of Consumer Policy, 26, 23–41. Kim, G. J., & Hanna, S. (2017). Do self-control measures affect saving behavior? Journal of Personal Finance, 16(2), 7–20. Kim, M. J. (2022). Two sides of the same coin: The simultaneous effects of spending and saving needs on budget estimation. Journal of Financial Services Marketing, 0123456789. https://doi.org/10.1057/s41264-021-00132- x Kim, Y., & Choi, S. M. (2018). The effects of breadth of product categories on budgeting. Advances in Consumer Research, 46, 440– 443. Kimiyagahlam, F., Safari, M., & Mansori, S. (2019). Influential behavioral factors on retirement planning behavior: The Case of Malaysia. Journal of Financial Counseling https://doi.org/10.5860/choice.51-1007 Financial Services Review, 33(1) 198 and Planning, 30(2), 244–261. https://doi.org/10.1891/1052- 3073.30.2.244 Lim, H. N., & Lee, J. M. (2021). Retirement income sources and subjective financial well-being: A comparison of retirees and non-retirees. Journal of Financial Counseling and Planning, 32(3), 517–534. https://doi.org/10.1891/JFCP-19-00101 Lutter, S. (2022, August 25). Psychology of financial planning webinar. CFP Board. https://www.cfp.net/events/2022/08/the- psychology-of-financial-planning-webinar Muratore, A. M., & Earl, J. K. (2010). Predicting retirement preparation through the design of a new measure. Australian Psychologist, 45(2), 98–111. https://doi.org/10.1080/0005006090352447 1 Oakley, D., & Kenneally, K. (2019). Retirement insecurity 2019: Households’ view of the retirement crisis (Issue March). Ozgen, O., & Esiyok, E. (2020). Consumer ethics, materialism and material satisfaction: A study on Turkish adolescent consumers. International Journal of Consumer Studies, 44(1), 14–24. https://doi.org/10.1111/ijcs.12541 Pahlevan Sharif, S., She, L., Yeoh, K. K., & Naghavi, N. (2022). Heavy social networking and online compulsive buying: The mediating role of financial social comparison and materialism. Journal of Marketing Theory and Practice, 30(2), 213–225. https://doi.org/10.1080/10696679.2021.190 9425 Rick, S. (2018). Tightwads and spendthrifts: An interdisciplinary review. Financial Planning Review, 1(1–2), e1010. https://doi.org/10.1002/cfp2.1010 Senate Finance Committee. (2022). Secure 2.0 Act of 2022. https://www.finance.senate.gov/imo/media /doc/Secure percent202.0_Section percent20by percent20Section percent20Summary percent2012-19-22 percent20FINAL.pdf Shefrin, H., & Thaler, R. (1988). The behavioral life-cycle hypothesis. Economic Inquiry, 26(4), 609–643. Sturr, T., Lynn, C., & Lawon, D. (2021). Financial self-efficacy and retirement preparation. Journal of Financial Planning, June, 86–98. https://doi.org/10.1002/9781119203124.ot h1 Sui, L., Sun, L., & Geyfman, V. (2021). An assessment of the effects of mental accounting on overspending behaviour: An empirical study. International Journal of Consumer Studies, 45(2), 221–234. https://doi.org/10.1111/ijcs.12613 Tescher, J., & Corey Stone. (2022). Revolving debt’s challenge to financial health and one way to help consumers pay it off. The Brookings Institution. https://www.brookings.edu/research/revolv ing-debts-challenge-to-financial-health- and-one-way-to-help-consumers-pay-it-off/ Watson, J. J. (2003). The relationship of materialism to spending tendencies, saving, and debt. Journal of Economic Psychology, 24(6), 723–739. https://doi.org/10.1016/j.joep.2003.06.001 Xiao, J. J., & O’Neill, B. (2018). Mental accounting and behavioural hierarchy: Understanding consumer budgeting behaviour. International Journal of Consumer Studies, 42(4), 448–459. https://doi.org/10.1111/ijcs.12445 Yuh, Y., & Hanna, S. D. (2010). Which households think they save? Journal of Consumer Affairs, 44(1), 70–97. https://doi.org/10.1111/j.1745- 6606.2010.01158.x Lynn et al. 199 Appendix Table A1 Sample Demographics of the SHED Sample (N = 6,651) unweighted weighted Variable n M/% SD n M/% SD min max Age Under 35 2,039 30.66% - 2,328 35.00% - - - 35-44 1,306 19.64% - 1,454 21.87% - - - 45-54 1,268 19.06% - 1,158 17.41% - - - 55-64 1,490 22.40% - 1,310 19.69% - - - 65+ 548 8.24% - 401 6.02% - - - Marital Status Married 3,700 55.63% - 3,655 54.96% - - - Living w/ partner, single, never married 2,094 31.48% - 2,241 33.69% - - - Separated/divorced 741 11.14% - 651 9.79% - - - Widowed 116 1.74% - 103 1.55% - - - Race White 4,495 67.58% - 4,145 62.32% - - - Black 713 10.72% - 797 11.98% - - - Hispanic 882 13.26% - 569 8.56% - - - Other 561 8.43% - 1,141 17.15% - - - Gender Male 3,605 54.20% - 3,426 51.52% - - - Female 3,046 45.80% - 3,225 48.49% - - - Education of respondent No high school degree 159 2.39% - 420 6.32% - - - High school degree 1,244 18.70% - 1,614 24.27% - - - Some college 1,990 29.92% - 1,875 28.19% - - - Financial Services Review, 33(1) 200 Bachelor’s degree or higher 3,258 48.99% - 2,742 41.22% - - - Parental education (higher of mother or father's) No high school degree 544 8.52% - 648 10.16% - - - Highschool degree 1,855 29.06% - 1,856 29.09% - - - Some college 1,642 25.72% - 1,618 25.34% - - - Bachelor’s degree 1,272 19.93% - 1,208 18.93% - - - Advanced degree 1,070 16.76% - 1,052 16.49% - - - Income $0-$24,999 620 9.32% - 564 8.47% - - - $25,000-$49,999 1,148 17.26% - 1,088 16.35% - - - $50,000-$74,999 1,216 18.28% - 1,129 16.97% - - - $75,000-$99,999 1,079 16.22% - 998 15.01% - - - $100,000-$149,999 1,395 20.97% - 1,372 20.63% - - - $150,000 or more 1,193 17.94% - 1,501 22.56% - - - Health Excellent 3,208 54.94% - 3,196 57.74% - - - Fair 2,550 43.67% - 2,558 43.81% - - - Poor 81 1.39% - 85 1.45% - - - Objective financial knowledge 6,651 1.93 1.08 6,651 1.84 1.10 0 3 Risk tolerance 6,628 4.59 2.53 6,628 4.58 2.56 0 10 Note: 2019 SHED. Restricted sample to only those working as a paid employee. Lynn et al. 201 Table A2 Sample Demographics of the SCF Sample (N = 3,361) unweighted weighted Variable n M/% SD n M/% SD min max Age 3,361 47.46 13.40 3,361 43.96 13.02 18 90 Marital Status Married 1,913 56.91% - 1,644 48.90% - - - Living with partner or never married 924 27.49% - 1,136 33.78% - - - Separated/divorced 42 13.64% - 509 15.15% - - - Widowed 66 1.96% - 73 2.16% - - - Race - - - - White 2,333 69.41% - 2,169 64.54% - - - Black 436 12.98% - 541 16.09% - - - Hispanic 362 10.78% - 427 12.71% - - - Asian/other Households 229 6.82% - 224 6.66% - - - Gender Male 2,168 64.50% - 1,935 57.57% - - - Female 1,193 35.50% - 1,426 42.43% - - - Education of respondent No high school degree 190 5.66% - 217 6.46% - - - High school degree 576 17.14% - 673 20.01% - - - Some college 809 24.07% - 960 28.56% - - - Bachelor’s degree 938 27.91% - 899 26.73% - - - Advanced degree 848 25.23% - 613 18.23% - - - Parental education (higher of mother or father's) No high school degree 399 11.87% - 441 13.11% - - - Highschool degree 1,003 29.84% - 1,059 31.50% - - - Some college 568 16.90% - 623 18.53% - - - Financial Services Review, 33(1) 202 Bachelor’s degree or beyond 1,391 41.39% - 1,239 36.85% - - - Health Excellent 2,819 83.88% - 2,754 81.93% - - - Fair 503 14.98% - 562 16.73% - - - Poor 39 1.15% - 45 1.34% - - - Income (log) 3,361 $1,272,932 $13,000,000 3,361 $76,359 $557,981 0.00 $704,000,000 Objective financial knowledge 3,361 2.38 0.81 3,361 2.23 0.85 0.00 3.00 Subjective financial knowledge 3,361 7.42 1.99 3,361 7.11 1.98 0.00 10.00 Risk tolerance 3,361 5.27 2.54 3,361 4.78 2.51 0.00 10.00 Note: 2019 SCF. Restricted sample to only those working full-time. Lynn et al. 203 Table A3 Sample Demographics of the NFCS Sample (N = 10,800) unweighted weighted Variable n M/% SD n M/% SD min max Age Under 35 3,540 32.78% - 3,757 34.79% - - - 35-44 2,648 24.52% - 2,596 24.04% - - - 45-54 2,532 23.44% - 2,452 22.70% - - - 55-64 1,757 16.27% - 1,694 15.68% - - - 65+ 323 2.99% - 301 2.79% - - - Marital Status Married 5,984 55.41% - 5,798 53.69% - - - Single 3,499 32.40% - 3,750 34.72% - - - Separated/divorced 1,169 10.82% - 1,110 10.28% - - - Widowed/widower 148 1.37% - 142 1.31% - - - Race White 7,737 71.64% - 6,488 60.07% - - - Non-White 3,063 28.36% - 4,312 39.93% - - - Gender Male 5,680 52.59% - 6,292 58.26% - - - Female 5,120 47.41% - 4,508 41.74% - - - Education of respondent No high school degree 88 0.81% - 92 0.85% - - - High school degree 2,009 18.60% - 2,281 21.12% - - - Some college 3,892 36.04% - 4,203 38.92% - - - Bachelor’s degree 3,021 27.97% - 2,599 24.06% - - - Advanced degree 1,790 16.57% - 1,625 15.04% - - - Parental education (higher of mother or father's) No high school degree 482 4.52% - 576 5.40% - - - Financial Services Review, 33(1) 204 Highschool degree 2,941 27.59% - 3,064 28.75% - - - Some college 3,064 28.75% - 3,174 29.78% - - - Bachelor’s degree 2,643 24.80% - 2,425 22.75% - - - Advanced degree 1,529 14.34% - 1,420 13.32% - - - Income $0-$24,999 846 7.83% - 966 8.94% - - - $25,000-$49,999 2,515 23.29% - 2,540 23.52% - - - $50,000-$74,999 2,401 22.23% - 2,446 22.64% - - - $75,000-$99,999 2,010 18.61% - 1,964 18.19% - - - $100,000-$149,999 1,958 18.13% - 1,877 17.38% - - - $150,000 or more 1,070 9.91% - 1,008 9.33% - - - Objective financial knowledge 10,800 3.24 1.63 10,800 3.15 1.64 0 6 Subjective financial knowledge 10,591 5.22 1.27 10,591 5.23 1.29 1 7 Risk tolerance 10,584 5.59 2.57 10,584 5.71 2.62 1 10 Financial attitude 10,591 4.67 1.94 10,682 4.68 1.94 1 7 Note: 2018 NFCS. Restricted sample to only those working full time. Statement of Conflict of Interest One of the authors of this paper (Michael Kothakota) worked as the Head of Research at the Certified Financial Planner Board of Standards, Inc. and was a member of its Executive Leadership at the time of submission. He received no compensation for his role in preparing this manuscript.