The Effect of Student Loan Debt on Emergency Savings and the Moderating Role of Financial Knowledge: Evidence from the U.S. Survey of Household Economics and Decisionmaking

Abstract

This study examines data from the U.S. 2018 and 2019 Survey of Household Economics and Decision making (SHED) to understand the association between student loan debt and emergency-saving decisions, including the moderating role of financial knowledge. Controlling self-selection bias through a propensity score and coarsened exact matching approach, the findings reveal that individuals with student loan debt are less likely to save for financial emergencies. The findings also show that financial knowledge is positively associated with a higher likelihood of having emergency savings. Furthermore, the results from the moderating analysis indicate a statistically significant interaction effect. Based on the empirical results and the corresponding interaction plots, the findings suggest that targeted financial education may lead to improved financial outcomes for student loan borrowers, rather than assuming that such education occurred prior to a loan application.

Full text

Citation: Korankye, Thomas, Blain
Pearson, and Peter Agyemang-Mintah.
2024. The Effect of Student Loan Debt
on Emergency Savings and the
Moderating Role of Financial
Knowledge: Evidence from the U.S.
Survey of Household Economics and
Decisionmaking. Journal of Risk and
Financial Management 17: 420.
https://doi.org/10.3390/jrfm17090420
Academic Editor: Thanasis Stengos
Received: 7 August 2024
Revised: 6 September 2024
Accepted: 19 September 2024
Published: 21 September 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Journal of
Risk and Financial
Management
Article
The Effect of Student Loan Debt on Emergency Savings and the
Moderating Role of Financial Knowledge: Evidence from the
U.S. Survey of Household Economics and Decisionmaking
Thomas Korankye 1, * , Blain Pearson 2and Peter Agyemang-Mintah 3
1Personal and Family Financial Planning Program, Norton School of Human Ecology, University of Arizona,
Tucson, AZ 85721, USA
2Department of Finance and Economics, Coastal Carolina University, Conway, SC 29526, USA
3College of Business, Zayed University, Abu Dhabi P.O. Box 144534, United Arab Emirates
*Correspondence: korankye@arizona.edu
Abstract:
This study examines data from the U.S. 2018 and 2019 Survey of Household Economics and
Decision making (SHED) to understand the association between student loan debt and emergency-
saving decisions, including the moderating role of financial knowledge. Controlling self-selection
bias through a propensity score and coarsened exact matching approach, the findings reveal that
individuals with student loan debt are less likely to save for financial emergencies. The findings also
show that financial knowledge is positively associated with a higher likelihood of having emergency
savings. Furthermore, the results from the moderating analysis indicate a statistically significant
interaction effect. Based on the empirical results and the corresponding interaction plots, the findings
suggest that targeted financial education may lead to improved financial outcomes for student loan
borrowers, rather than assuming that such education occurred prior to a loan application.
Keywords:
coarsened exact matching; emergency fund; emergency savings; financial literacy; propensity
score matching; student loan debt
JEL Classification: G51; G53
1. Introduction
An important lesson from the COVID-19 pandemic is that financial emergencies are
inevitable (McDonald 2021). The unexpected occurrence of widespread job losses, health
shocks, and mortality of loved ones during the pandemic echoes the belief that some life
events are rarely predictable. Consequently, this requires households to have adequate
liquid funds available to weather the financial burdens of such events. It is a commonly
accepted practice for financial service professionals to recommend having three to six
months of fixed expenses held in the form of cash or cash equivalents to allow households
to weather financial emergencies. This may help households avoid the risk of predatory
loans, incurring transaction costs to convert illiquid assets into liquid funds, or liquidating
long-term investments, such as stocks or retirement assets, at inopportune times.
The absence of emergency funds could jeopardize the financial health of affected
individuals. Evidence suggests that the lack of emergency savings results in increased
vulnerability to short-, medium-, and long-term financial insecurity risks (Adams and
West 2015;Hasler et al. 2018). Other studies identify rent or mortgage payment problems,
high-cost borrowings, food insecurity, energy insecurity, and psychological distress result-
ing from sleeplessness and poor health as some of the risks associated with the absence
of adequate emergency savings (Brobeck 2008;Gjertson 2016;Lusardi et al. 2011). Con-
versely, studies have shown that having emergency savings provides financial protection
for families. For instance, among low-income households, Gjertson (2016) and Mills and
J. Risk Financial Manag. 2024,17, 420. https://doi.org/10.3390/jrfm17090420 https://www.mdpi.com/journal/jrfm

J. Risk Financial Manag. 2024,17, 420 2 of 18
Amick (2010) report that having emergency savings is associated with a low likelihood
of experiencing financial hardships. Studies also show that holding emergency savings is
positively associated with financial satisfaction (Korankye and Kalenkoski 2021a;Tharp
et al. 2020;Woodyard and Robb 2016). For example, Korankye and Kalenkoski (2021a)
show that having emergency savings among U.S. adults is associated positively with higher
levels of financial well-being compared with those without such funds.
Despite the benefits of having emergency savings, reports indicate that adequate
emergency fund holdings are limited among American adults (Lin et al. 2022;Ratcliffe et al.
2022). The Consumer Financial Protection Bureau’s Making Ends Meet survey shows that
almost 25% of Americans have no emergency savings, while 49% have emergency funds
that may not last for a month (Ratcliffe et al. 2022). When asked whether respondents have
emergency funds to cover expenses for three months, the FINRA Foundation National
Financial Capability Study found that 47% of Americans responded negatively (Lin et al.
2022). The findings from these survey reports suggest that many American adults are not
adequately prepared to handle financial emergencies.
Existing studies identify several factors that could explain emergency-saving behavior.
For instance, studies show that people with high financial literacy have a greater propensity
to own emergency savings than those with low financial literacy (Babiarz and Robb 2014;
Fan and Zhang 2021). Studies also show that seeking financial advice motivates individuals
to have emergency funds (Alyousif and Kalenkoski 2017;Marsden et al. 2011). Those who
receive financial education from school and work are also more likely to hold emergency
savings (Fan and Zhang 2021). Using data from the U.S. National Financial Capability
Study, Despard et al. (2020) ascertain that homeownership, the presence of savings accounts,
financial confidence, and subjective financial knowledge positively correlate with the
likelihood of owning an emergency fund. As important as these studies are, not much
is known about the influential role of student debt in shaping individuals’ emergency-
saving decisions.
The present study aims to expand our understanding of the factors influencing emer-
gency savings, with a specific focus on student loan debt. Generally, the literature suggests
that student loans are essential for helping financially constrained individuals obtain post-
secondary qualifications and expand opportunities for higher-income occupations. Low
federal and state-level funding targeted to colleges, coupled with rising tuition rates (Fed-
eral Reserve Bank of New York 2021;Ma et al. 2018), have led to increased student loan
usage in the United States, resulting in adverse outcomes for individuals and families, all
else being equal (Cho et al. 2015;Korankye and Kalenkoski 2021b;Korankye 2023;Pearson
and Lee 2022).
Two research questions are central to the current study. First, the study seeks to
answer the question, “How does student loan indebtedness relate to the emergency-saving
decisions of individuals?” Second, “Does financial knowledge matter in explaining the
relationship between student debt and emergency-saving decisions?” To address the re-
search questions, the current study moves beyond correlational research to account for
self-selection bias and endogeneity issues through propensity score and coarsened exact
matching. This approach allows inference of a causal relationship between student debt
and emergency-saving decisions. In the end, policymakers, financial educators, and finan-
cial service professionals will be better positioned to appreciate the empirical relationship
between student debt and individuals’ emergency-saving behavior for timely intervention.
2. Conceptual Framework
The study uses Sherraden’s (2013) conceptual framework on financial capability to
examine student debt, financial knowledge, and emergency-saving decisions. The frame-
work considers financial knowledge, skills, and access to financial products as inputs that
influence financial outcomes, such as emergency savings. Despard et al. (2020) have used
Sherraden’s (2013) conceptual framework to examine the role of financial capability in ex-
plaining the lack of emergency funds among U.S. households. While emphasizing financial

J. Risk Financial Manag. 2024,17, 420 3 of 18
knowledge, the present study incorporates student loan debt into Sherraden’s (2013) model
to examine the possible role of student debt in explaining the emergency-saving decisions
of Americans.
2.1. Student Loans and Associated Outcomes
Existing studies have shown that student debt influences several household decisions,
including stock ownership, homeownership, college savings, and retirement savings (Ko-
rankye and Guillemette 2021;Martin et al. 2020;Mountain et al. 2020). Cho et al. (2015)
provide a review of several studies on student debt outcomes. Following that review, a
longitudinal study of data from 2011 to 2017 using the U.S. Panel Study of Income Dy-
namics show that households with student debt are less likely to hold stock investments
in non-retirement accounts (Korankye 2023;Korankye and Guillemette 2021). Martin
et al. (2020) use data from the 2012 National Longitudinal Survey of Youth (NLSY79) to
show that parents who hold student debt on their balance sheets have a low probability
of saving for their children’s education. Focusing on millennials and utilizing data from
the 2016 UF/VT Millennial Housing and Student Debt Survey, Mountain et al. (2020)
demonstrate that the burden associated with student debt is negatively associated with
millennial homeownership.
Given that having emergency savings could help individuals weather financial shocks,
expanding the existing literature to examine the link between student debt and the likeli-
hood of holding emergency savings is warranted. As the studies reviewed above indicate,
the burden associated with student debt can create financial stressors that limit borrowers’
ability to make significant financial decisions that could adversely impact their well-being
(Kim et al. 2021). The social stress theory suggests that financially constrained individuals
tend to report higher levels of stress, which is associated with lower levels of well-being
(Mossakowski 2014).
From a standard economic theory perspective, student debt constitutes an investment
in human capital. With all else equal, investment in human capital generates more future
earnings, consumption, and wealth (Becker 1962). With this logic, a rational decision
maker would only invest in their human capital if the cost of the investment is lower than
the expected payoff. Depending on intertemporal preferences, student debt holders may
choose to consume more during the current period while paying off higher interest debt
rather than having financial assets in emergency savings, holding interest rates constant.
The consumption basket of those without student debt excludes student debt repayment
obligations. With all else equal, this makes it less financially burdensome for them to
allocate resources to other necessities, including holding an emergency fund.
2.2. Effect of Financial Knowledge
Financial knowledge and skills, which could be called financial literacy, form part of
the inputs in Sherraden’s (2013) framework. Financial literacy has been shown to relate
positively to emergency savings (Babiarz and Robb 2014;Bhutta et al. 2022). For instance,
using the 2016 and 2019 Survey of Consumer Finances and the Big Three financial literacy
questions, Bhutta et al. (2022) show that financially literate individuals are more likely
to have liquid savings. Through sensitivity analysis, Bhutta et al. (2022) observe that
the magnitude of the effect of financial literacy on liquid savings is large for respondents
correctly answering all three financial knowledge questions on inflation, diversification,
and interest rates.
Although financial knowledge is essential, findings from existing studies have not
always been consistent. Using datasets from the 2009 to 2018 National Financial Capability
Study, Despard et al. (2020) show that objective financial knowledge is not a statistically
significant predictor of emergency savings. Instead, Despard et al. (2020) find subjective
financial knowledge to be a statistically significant contributor to the likelihood of having
emergency savings. A related study by Babiarz and Robb (2014) observes a positive
correlation between subjective financial knowledge and emergency savings. However,

J. Risk Financial Manag. 2024,17, 420 4 of 18
unlike Despard et al. (2020), Babiarz and Robb (2014) observe a statistically significant
positive correlation between objective financial knowledge and emergency savings.
While studies, such as Babiarz and Robb (2014), Bhutta et al. (2022), and Despard et al.
(2020), examine the direct relationship between financial literacy and emergency savings,
other studies examine the moderating role of financial literacy among financial behaviors
and decisions. For instance, Fan and Zhang (2021) use data from the 2018 U.S. National
Financial Capability Study to examine the mediating role of financial literacy between
the sources of financial education and emergency-saving decisions. Their findings show
that financial literacy mediates the association between financial education and holding
emergency funds.
Adding to the mediation study of Fan and Zhang (2021), the current study considers
the moderation role of financial knowledge between student debt and emergency savings.
This study believes that it is insufficient to examine the effect of student debt on emergency
savings without considering pathways that enhance positive behavior. Including financial
knowledge in the moderation analysis could demonstrate the effectiveness of financial
knowledge in empowering individuals to make utility-maximizing decisions. By using
this approach, the present study illustrates the associated effect of borrowing to pay for
college education on a household’s ability to meet contingent financial obligations and how
financial knowledge could change the dynamics for affected families.
2.3. Hypotheses
The current study tests the following hypotheses based on the conceptual framework
and existing studies:
H1. Holding student debt negatively relates to the emergency-saving decisions of individuals.
H2.
Financial knowledge moderates the association between holding student debt and emergency-
saving decisions.
3. Methods
3.1. Data
The study uses pooled cross-sectional data from the 2018 and 2019 Survey of House-
hold Economics and Decisionmaking (SHED). Designed to be nationally representative, the
SHED is administered by the U.S. Federal Reserve Board annually to explore the economic
well-being of households in the United States. The 2018 and 2019 waves are selected to
examine the research topic before the onset of the COVID-19 pandemic. In addition, these
are the two waves containing all the variables of interest for the current study. These two
waves are part of the National Endowment for Financial Education/Knology’s “Financial
Education Database”, which is streamlined for consistency across multiple years (Dwyer
et al. 2021).
3.2. Dependent Variable
The dependent variable for this study is that of emergency savings. The survey asks the
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?” An
affirmative response is coded as 1, and otherwise the response would be 0.
3.3. Key Explanatory Variable
This study includes two key explanatory variables. The first variable is student loan
debt, coded as 1 if an individual reports the presence of student debt on their balance
sheet, and otherwise the response would be 0. The second is financial knowledge, obtained
from the Big Three financial literacy questions covering inflation, interest rates, and risk
diversification (Lusardi 2008;Lusardi and Mitchell 2017). A correct answer to each question
is coded as 1, 0 otherwise. The study sums the correct responses to each question to obtain

J. Risk Financial Manag. 2024,17, 420 5 of 18
the objective financial knowledge score with values ranging from 0 to 3. The 2018 and 2019
SHED do not have information on subjective financial knowledge and the other financial
literacy questions.
3.4. Other Explanatory Variables
The study includes control variables relating to gender (female), age, race (White),
educational attainment, marital status (married), household size, physical health status,
homeownership status, household income, employment status, financial risk preference,
credit card debt, bank account ownership, and survey year fixed effects. The present study
includes these variables based on Sherraden’s (2013) conceptual framework and findings
from the existing literature (Babiarz and Robb 2014;Bhutta et al. 2022;Despard et al. 2020;
Fan and Zhang 2021).
Age, household size, and risk preference are measured as continuous variables. The
survey’s question for measuring financial risk preference is, “On a scale from zero to ten,
where zero is not at all willing to take risks and ten is very willing to take risks, what
number would be on the scale?”
Female, White, married, homeownership, credit card debt, and bank account owner-
ship are measured dichotomously, with 1 indicating a yes response and 0 otherwise. The
categorical variables from the SHED are educational attainment, physical health status,
household income, and employment status.
3.5. Model
The paper estimates four logistic hierarchical regression models based on the study’s
objectives. The first is a restricted model that includes the key explanatory variable of
student loan debt. The second model estimates the relationship between student debt and
the emergency-saving decisions of individuals, including all the control variables except
financial knowledge. The third model adds financial knowledge to the second model. The
fourth is the interaction model that examines the moderating role of financial knowledge
between student debt and emergency-saving decisions. The four models just described are
stated as follows:
Pr (Y= 1) = Φ(β0+β1Sdebt) (1)
Pr (Y= 1) = Φ(β0+β1Sdebt + βXX) (2)
Pr (Y= 1) = Φ(β0+β1Sdebt + β2FinKnow + βXX) (3)
Pr (Y= 1) = Φ(β0+β1Sdebt + β2FinKnow + β3Sdebt ×FinKnow + βXX) (4)
where Yis the dependent variable representing the emergency-saving decisions of in-
dividuals. Sdebt and FinKnow connote student debt and objective financial knowledge,
respectively. The matrix, X, represents the control variables, including gender, age, race,
education, marital status, household size, health status, homeownership, income, employ-
ment status, financial risk, credit card debt, bank account ownership, and survey year. The
β’s are the odds ratios to be estimated.
3.6. Accounting for Self-Selection Bias
The uptake of student loans is considered non-random (Cho et al. 2015). This suggests
that standard regression estimates could be biased due to the presence of self-selection bias
associated with student loans. For instance, it has been shown that holders and non-holders
of student loan debt could differ regarding financial knowledge, personality traits, gender,
and race/ethnicity (Korankye 2024;Li 2021;Liu et al. 2023). Therefore, using controlled
experimental data to establish a causal relationship between student debt and emergency
savings will be ideal. However, the current study uses survey data for empirical analyses.
Another strategy is to utilize an instrumental variable technique, but finding a suitable
instrument in this particular case is compelling.

J. Risk Financial Manag. 2024,17, 420 6 of 18
To address the endogeneity issues associated with student debt, the current study
initially performs a propensity score matching analysis (and later coarsened exact match-
ing), utilizing one-to-one matching with replacement. This analysis method contrasts the
treatment sample (student debt holders) from the control sample (those without student
debt) by conditioning on covariates. The estimators use propensity scores to match each
respondent to comparable observations, thereby eliminating the self-selection effect and
allowing causal inferences (Lanza et al. 2013). The current study estimates the average effect
of the treatment (holding student debt) on emergency-saving decisions in the population.
Studies, such as Korankye et al. (2023), Kim et al. (2018), and Marsden et al. (2011), estimate
a similar model to handle similar perceived self-selection problems. Abadi and Imbens
(2016) provide more information about propensity score matching.
4. Results
4.1. Descriptive Statistics
Table 1contains the summary statistics for the full sample and the t-test results
for those with and without emergency savings. The statistics here pertain to the raw
data that have not been exposed to any of the matching techniques. Almost 55% of the
respondents hold emergency funds that could cover three months of expenses when
financial emergencies occur. About 21% of the respondents have student debt on their
balance sheets. There are more student debtholders without emergency savings than those
with emergency funds (29% versus 14%), suggesting that student debtholders have a lower
likelihood of owning emergency funds. The average score for financial knowledge is 1.9 out
of 3 for the full sample. However, those with emergency savings have an average score
of 2.18 compared to 1.5 for respondents without emergency savings. The average score
for financial risk is 4 out of 10. The t-test results show that those with emergency savings
have higher risk tolerance than those without emergency funds (4.61 versus 3.52). The
average age for the full sample is 52, and there are approximately 50% females, 70% White,
39% college-educated respondents, and 31% respondents with a household income of
$100,000 or more.
Table 1. Summary Statistics for the Dependent and Explanatory Variables.
Full Sample aHave Emergency Savings aNo Emergency Savings a
Dependent variable: Emergency savings 0.5437
Main explanatory variable:
Student loan debt (1 = Yes) 0.2076 0.1372 0.2915 ***
Other explanatory variables:
Gender (1 = Female) 0.4971 0.4590 0.5425 ***
White (Yes = 1) 0.6986 0.7567 0.6293 ***
Married (Yes = 1) 0.5547 0.6442 0.4482 ***
Age (continuous) 52.2750 56.3383 47.4341 ***
Education
Less than high school 0.0468 0.0203 0.0784 ***
High school 0.2409 0.1939 0.2969 ***
Some college 0.3223 0.2929 0.3573 ***
Bachelor’s degree or higher 0.3900 0.4929 0.2675 ***
Household Size (1 to 12) 2.4983 2.3287 2.7003 ***
Health status
Poor 0.0277 0.0130 0.0452 ***
Fair 0.1237 0.0850 0.1697 ***
Good 0.3414 0.3115 0.3770 ***
Very good 0.3511 0.4198 0.2692 ***
Excellent 0.1033 0.1239 0.0787 ***
Household income

J. Risk Financial Manag. 2024,17, 420 7 of 18
Table 1. Cont.
Full Sample aHave Emergency Savings aNo Emergency Savings a
Less than $50,000 0.4034 0.2742 0.5574 ***
$50,000 to less than $100,000 0.2881 0.3010 0.2728 ***
$100,000 to less than $150,000 0.1635 0.2091 0.1091 ***
$150,000 or more 0.1450 0.2157 0.0607 ***
Employment status
Unemployed 0.0440 0.0150 0.0818 ***
Not working—Disabled 0.0455 0.0150 0.0818 ***
Self-employed 0.0772 0.0804 0.0734 *
Paid employee 0.5236 0.5028 0.5484 ***
Retired 0.2570 0.3362 0.1627 ***
Other 0.0527 0.0422 0.0652 ***
Homeownership (Yes = 1) 0.6664 0.8070 0.4989 ***
Health insurance ownership (Yes = 1) 0.5365 0.6007 0.4601 ***
Bank account ownership 0.9482 0.9869 0.9020 ***
Credit card debt 0.3780 0.2931 0.4791 ***
Financial risk preference (0 to 10) 4.1111 4.6094 3.5175 ***
Financial knowledge (0 to 3) 1.8730 2.1822 1.5047 ***
N 34,354 18,677 15,677
Notes: The data source is the 2018 and 2019 SHED. The t-tests are performed for households with emergency
savings versus those without college savings. * p< 0.05; *** p< 0.001. The mean values are shown, but not
the standard errors. Year dummies are included but not shown.
a
All values are percentages, except where
indicated otherwise.
Given that this study utilizes propensity score matching, Table 2shows the summary
statistics for the matched and unmatched samples. The means for the treated and control
groups in the matched sample are similar, a situation that is not observed for the unmatched
sample shown in Table 2. For instance, the average age for those with student debt (treat-
ment group) is 41, and that of those without student debt (control group) is approximately
41. The average scores for financial risk and knowledge are about 4 and 2, respectively, for
each treatment and control group.
Table 2. Summary Statistics of Matched and Unmatched Samples—Propensity Score Matching.
Unmatched Sample a
Propensity
Score Logit
Coefficients
Matched Sample a
Has
Student
Debt
No
Student
Debt
% Bias
Has
Student
Debt
No
Student
Debt
% Bias
P score 0.3669 0.1659 116.0 0.3666 0.3666 0.0
Gender (1 = Female) 0.5550 0.4820 14.6 0.2208 *** 0.5547 0.5406 2.8
White (Yes = 1) 0.5968 0.7252 −27.4 −0.3001 *** 0.5972 0.5879 2.0
Married (Yes = 1) 0.5109 0.5662 −11.1 0.2252 *** 0.5110 0.5322 −4.3
Age (continuous) 41.159 55.187 −91.3 −0.0422 *** 41.171 41.044 0.8
Education
Less than high school
High school 0.1071 0.2759 −43.9 0.1173 0.1072 0.1086 −0.4
Some college 0.3574 0.3131 9.4 1.2311 *** 0.3577 0.3749 −3.7
Bachelor’s degree or higher 0.5146 0.3574 32.1 1.6282 *** 0.5142 0.4912 4.7
Household Size (1 to 12) 2.862 2.403 30.2 0.0605 *** 2.8619 2.9036 2.7

J. Risk Financial Manag. 2024,17, 420 8 of 18
Table 2. Cont.
Unmatched Sample a
Propensity
Score Logit
Coefficients
Matched Sample a
Has
Student
Debt
No
Student
Debt
% Bias
Has
Student
Debt
No
Student
Debt
% Bias
Health status
Poor
Fair 0.1082 0.1277 −6.0 −0.0815 0.1083 0.1094 −0.3
Good 0.3424 0.3412 0.3 −0.1241 * 0.3425 0.3309 2.4
Very good 0.3379 0.3546 −3.5 −0.2670 *** 0.3382 0.3309 1.5
Excellent 0.1082 0.1020 2.0 −0.3171 *** 0.1083 0.1047 1.2
Household income
Less than $50,000
$50,000 to less than $100,000 0.3092 0.2826 5.8 −0.1140 ** 0.3091 0.2952 3.0
$100,000 to less than $150,000 0.1844 0.1580 7.0 −0.1185 * 0.1845 0.1862 −0.4
$150,000 or more 0.1317 0.1485 −4.8 −0.3211 * 0.1318 0.1360 −1.2
Employment status
Unemployed
Not working—Disabled 0.0327 0.0489 −8.2 −0.2656 ** 0.0327 0.0348 −1.1
Self-employed 0.0648 0.0804 −6.0 −0.1726 * 0.0648 0.0678 −1.1
Paid employee 0.7286 0.4699 54.7 0.0765 0.7284 0.7171 2.4
Retired 0.0618 0.3082 −66.9 −0.6283 *** 0.0619 0.0630 −0.3
Other 0.0543 0.0523 0.9 −0.3525 *** 0.0543 0.0596 −2.4
Homeownership (Yes = 1) 0.5164 0.7057 −39.6 −0.4082 *** 0.5168 0.5203 −0.7
Health insurance ownership (Yes = 1)
0.6451 0.5081 28.0 0.1099 ** 0.6450 0.6372 1.6
Bank account ownership 0.9478 0.9483 −0.2 −0.1585 * 0.9478 0.9537 −2.7
Credit card debt 0.5202 0.3407 36.9 0.8603 *** 0.5199 0.5179 0.4
Financial risk preference (0 to 10) 4.2227 4.0819 5.5 −0.0218 ** 4.2223 4.2656 −1.7
Financial knowledge (0 to 3) 1.7877 1.8954 −9.9 −0.0434 ** 1.7887 1.7689 −1.8
N 7132 27,222 34,354 7127 3253
Notes:
a
All values are percentages, except where indicated otherwise. * p< 0.05; ** p< 0.01; *** p< 0.001.
One-to-one matching with replacement is used. This could explain why the treated group (has student debt) is
larger than the control group (no student debt), as control units can be matched to multiple treated units, leading
to more treated units than unique control units in the final sample.
In addition to the summary statistics, Table 2contains the propensity score logistic
coefficient estimates for student loan indebtedness. The coefficient estimates support the
notion that student debtholders differ from those without such debt. These debt holders
are more likely to be females, non-White, and college degree holders, for instance. They
also are less likely to have high risk tolerance, financial knowledge, and household income.
4.2. Overlap Test and Covariate Balance
Based on recommendations from studies, such as Drukker (2016), Garrido et al. (2014),
and Stuart (2010), this study performs overlap and balance tests to check the unbiasedness
of the propensity score estimates. The overlap test results, which assume that every
respondent could receive any of the two treatment levels, are presented in Figure 1. The

J. Risk Financial Manag. 2024,17, 420 9 of 18
figure shows a good overlap because a comparison group has a similar propensity score
for every individual receiving treatment.
J. Risk Financial Manag. 2024, 17, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jrfm
Article
The Effect of Student Loan Debt on Emergency Savings and the
Moderating Role of Financial Knowledge: Evidence from the
U.S. Survey of Household Economics and Decisionmaking
Thomas Korankye
1
*, Blain Pearson
2
and Peter Agyemang-Mintah
3
Figure 1. Distribution of propensity scores across treatment and control groups.
Commented [M1]: Please added “0” before the
decimal sign. E.g., “.2” should be “0.2”
Figure 1. Distribution of propensity scores across treatment and control groups.
The results of the covariate balance test in the form of standardized mean differences
(% bias) are also shown in Table 2. An imbalance may exist if the standardized mean
differences exceed 10% (Benedetto et al. 2018). From Table 2, an imbalance exists in the
unmatched sample. However, the matched sample of interest does not appear imbalanced
because all the standardized mean differences fall below 10%. These findings provide
evidence of random assignment of the treatment, allowing a causal inference from the
empirical estimates.
4.3. Regression Results
Table 3contains the logistic regression estimates of emergency savings on student
loan debt using propensity score matching. The results for Model 1, the restricted model
with only student debt as the explanatory variable, indicate that student debtholders
have 0.5579 (p< 0.001) times the odds of having emergency savings as compared to those
without student debt. The results for Model 2, the model with student debt and other
predictor variables except for financial knowledge, show that student debtholders have
0.4826 (
p< 0.001
) times the odds of having emergency savings when compared to non-
student debtholders. The results for the full model without interaction terms (Model 3)
show that the odds of having emergency savings for student-debt holders is estimated
to be 0.4808 (p< 0.001) times the odds of holding emergency savings for those without
student debt. Table 3(Model 3) also shows that the odds ratio for financial knowledge is
1.1176 (p< 0.001), suggesting that each additional increase in financial knowledge score is
associated with an 11.76% increase in the odds of having emergency savings.

J. Risk Financial Manag. 2024,17, 420 10 of 18
Table 3.
Logistic Regression Estimates of Emergency Savings on Student Loan Debt—Propensity
Score Matching.
Model 1 Model 2 Model 3 Model 4
Explanatory Variables:
Treatment: Student loan debt
(1 = Yes) 0.5579 *** (0.0191) 0.4826 *** (0.0190) 0.4808 *** (0.0189) 0.5834 *** (0.0449)
Financial knowledge (0 to 3) - - 1.1176 *** (0.0227) 1.1757 *** (0.0315)
Interaction: Student debt ×Financial
knowledge - - - 0.9004 ** (0.0324)
Gender (1 = Female) - 0.9803 (0.0391) 1.0278 (0.0421) 1.0266 (0.0420)
White (Yes = 1) - 0.8783 ** (0.0354) 0.8569 *** (0.0348) 0.8590 *** (0.0349)
Married (Yes = 1) - 0.9897 (0.0455) 0.9902 (0.0456) 0.9926 (0.0457)
Age (continuous) - 1.0041 * (0.0017) 1.0031 (0.0017) 1.0029 (0.0017)
Education (versus less than high school)
High school - 1.5304 * (0.2597) 1.5194 ** (0.2578) 1.5157 ** (0.2572)
Some college - 1.6670 ** (0.2702) 1.6099 ** (0.2612) 1.5995 ** (0.2595)
Bachelor’s degree or higher - 2.5927 *** (0.4221) 2.4126 *** (0.3939) 2.3957 *** (0.3911)
Household Size (1 to 12) - 0.8688 *** (0.0116) 0.8700 *** (0.0116) 0.8705 *** (0.0116)
Health status (versus poor)
Fair - 0.7089 *** (0.0616) 0.7131 *** (0.0621) 0.7088 *** (0.0618)
Good - 0.9979 (0.0684) 1.0045 (0.0690) 1.0035 (0.0690)
Very good - 1.3490 *** (0.0927) 1.3502 *** (0.0930) 1.3501 *** (0.0931)
Excellent - 1.5586 *** (0.1312) 1.5672 *** (0.1321) 1.5640 *** (0.1319)
Household income (versus less
than $50,000)
$50,000 to less than $100,000 - 1.4439 *** (0.0738) 1.4221 *** (0.0729) 1.4199 *** (0.0727)
$100,000 to less than $150,000 - 1.9143 *** (0.1186) 1.8627 *** (0.1160) 1.8650 *** (0.1162)
$150,000 or more - 3.2195 *** (0.2358) 3.0891 *** (0.2277) 3.0992 *** (0.2287)
Employment status (versus unemployed)
Not working–Disable - 0.8765 (0.1384) 0.8823 (0.1396) 0.8770 (0.1387)
Self-employed - 1.2568 (0.1466) 1.2467 (0.1457) 1.2356 (0.1446)
Paid employee - 1.2065 * (0.1119) 1.2193 * (0.1135) 1.2137 * (0.1131)
Retired - 2.8925 *** (0.3648) 2.9069 *** (0.3672) 2.9087 *** (0.3676)
Other - 1.1024 (0.1354) 1.1112 (0.1368) 1.1053 (0.1362)
Homeownership (Yes = 1) - 1.7884 *** (0.0834) 1.7696 *** (0.0827) 1.7680 *** (0.0826)
Health-insurance ownership
(Yes = 1) - 1.1774 ** (0.0570) 1.1636 ** (0.0564) 1.1629 ** (0.0564)
Bank account ownership - 3.3621 *** (0.4254) 3.1602 *** (0.4010) 3.1286 *** (0.3966)
Credit card debt - 0.4125 *** (0.0167) 0.4125 *** (0.0168) 0.4141 *** (0.0168)
Financial risk preference (0 to 10) - 1.1223 *** (0.0092) 1.1148 *** (0.0093) 1.1143 *** (0.0093)
N 10,380 10,380 10,380 10,380
Notes: The data source is the 2018 and 2019 SHED. Odds ratios are shown alongside standard errors, which are in
parentheses. * p< 0.05; ** p< 0.01; *** p< 0.001. Year dummies are included but not shown.
The results for Model 4, which contains the interaction term, show that the interaction
between student debt and financial knowledge is statistically significant. However, the

J. Risk Financial Manag. 2024,17, 420 11 of 18
odds ratio for the interaction term is less than one. Considering the interaction slopes,
the odds of having emergency savings increases by a factor of 1.1757 for each additional
financial knowledge score if the individual does not hold student debt and by a factor of
1.0586 (that is, 1.1757
×
0.9004) for every level of financial knowledge if the individual has
student debt. Figure 2displays the interaction plot, which shows the predictive margins
at 95% confidence intervals. The figure shows that the likelihood of holding emergency
savings is higher for those without student debt at each level of financial knowledge
compared to those with student debt. Figure 2also suggests that the difference in the
likelihood of holding emergency savings between student debtholders and non-student
debtholders widens as the financial knowledge score increases.
J. Risk Financial Manag. 2024, 17, x FOR PEER REVIEW 2 of 4
Figure 2. Interaction plot—propensity score matching, predictive margins with 95% confidence
intervals.
Commented [M2]: Please added “0” before the
decimal sign. E.g., “.3” should be “0.3”
Figure 2.
Interaction plot—propensity score matching, predictive marginswith 95% confidence intervals.
Considering the control variables, factors, such as household size, educational at-
tainment, health status, household income, retirement status, homeownership, health
insurance, bank account ownership, credit card debt, and financial risk preference, are
statistically significant predictors of emergency savings across Models 2, 3, and 4. For
instance, in Model 4, college-educated individuals have 139.57% higher odds of holding
emergency savings than those with less than high school education. Increasing household
size by one person is associated with 12.95% lower odds of having emergency savings.
Homeownership and bank account ownership are associated with 76.80% and 212.86%
higher odds of having emergency savings, respectively, compared to individuals without
homeownership or bank account ownership. A one-unit increase in risk preference is
associated with an 11.43% higher likelihood of having emergency savings.
4.4. Sensitivity Analysis with Coarsened Exact Matching
This study performs sensitivity analysis using coarsened exact matching as an alter-
native statistical technique to support the findings from the propensity score matching.
Coarsened exact matching is a matching method for making causal inferences from ob-
servational data. It is noted for estimating improved causal effects through an imbalance
reduction in covariates between the treated and untreated samples (Blackwell et al. 2009;
King and Nielsen 2019). Baseline covariates are coarsened based on recommendations by
Iacus et al. (2009). The study directly matched on gender (female), race (White), marital
status (married), educational attainment, household income, retirement status, homeowner-
ship, health insurance, bank account ownership, and credit card debt. Age is split into adult

J. Risk Financial Manag. 2024,17, 420 12 of 18
(18 to 39), middle-aged adult (40 to 59), and senior adult (>60). Household size is divided
into small (
≤
2) and high (
≥
3) based on the median. Financial knowledge is composed of
low (<3) and high (=3), and risk preference is divided into low (
≤
3), medium (4 to 6), and
high (≥7).
In carrying out the analysis, the study employed the “CEM” algorithm in Stata and
k-2-k matching leading to an equal number of 2402 matched observations for the treated
and untreated samples. An examination of the differences in means and the quantiles of
the distributions of the treated and untreated groups shows a balance for the marginal and
joint distributions of the covariates, as suggested by Blackwell et al. (2009).
Table 4shows the logistic regression estimates of emergency savings on student loan
debt using coarsened exact matching for Models 1 to 4. The odds of having emergency
savings for student debtholders relative to those without student debt is estimated to
be 0.5896 (p< 0.001) for Model 1, 0.5154 (p< 0.001) for Model 2, and 0.5140 (p< 0.001)
for Model 3. Table 4(Model 3) also shows that the odds ratio for financial knowledge is
1.1327 (
p< 0.001
), suggesting that each additional increase in financial knowledge score is
associated with a 13.27% increase in the odds of having emergency savings. The interaction
term in Model 4 is statistically significant (odds ratio of 0.8606 with p < 0.01). More
specifically, the interaction results indicate that the odds of having emergency savings
increase by a factor of 1.2194 per every one-unit increase in financial knowledge score if the
individual does not have student debt, and by a factor of 1.0494 (that is, 1.2194
×
0.8606)
for every level of financial knowledge if the individual has student debt.
Comparatively, the results obtained through coarsened exact matching are similar
to those of the propensity score matching. That is, student debt negatively impacts the
emergency-saving decisions of individuals, although increases in financial knowledge
provide some gains to emergency-saving decisions among individuals with student debt.
Figure 3, which contains the interaction plot for the coarsened exact matching, amplifies
these results.
J. Risk Financial Manag. 2024, 17, x FOR PEER REVIEW 2 of 4
Figure 2. Interaction plot—propensity score matching, predictive margins with 95% confidence
intervals.
Commented [M2]: Please added “0” before the
decimal sign. E.g., “.3” should be “0.3”
Figure 3.
Interaction plot—coarsened exact matching, predictive margins with 95% confidence intervals.

J. Risk Financial Manag. 2024,17, 420 13 of 18
Table 4.
Logistic Regression Estimates of Emergency Savings on Student Loan Debt—Coarsened
Exact Matching.
Model 1 Model 2 Model 3 Model 4
Explanatory Variables:
Treatment: Student loan debt
(1 = Yes) 0.5896 *** (0.0343) 0.5154 *** (0.0348) 0.5140 *** (0.0347) 0.6845 ** (0.0944)
Financial knowledge (0 to 3) - - 1.1327 ** (0.0448) 1.2194 *** (0.0615)
Interaction: Student debt ×Financial
knowledge - - - 0.8606 * (0.0544)
Gender (1 = Female) - 0.9394 (0.0675) 0.9951 (0.0739) 0.9951 (0.0740)
White (Yes = 1) - 0.9891 (0.0773) 0.9499 (0.0753) 0.9481 (0.0751)
Married (Yes = 1) - 1.1727 (0.1207) 1.1822 (0.1219) 1.1810 (0.1218)
Age (continuous) - 1.0109 ** (0.0034) 1.0095 ** (0.0034) 1.0095 ** (0.0034)
Education (versus Less than high school)
High school - 4.0060 * (2.2624) 3.8626 * (2.1795) 3.8205 * (2.1571)
Some college - 4.8146 * (2.6943) 4.5045 * (2.5198) 4.4257 * (2.4776)
Bachelor’s degree or higher - 7.0977 *** (3.9941) 6.2752 ** (3.5360) 6.1977 ** (3.4948)
Household Size (1 to 12) - 0.8236 *** (0.0214) 0.8237 *** (0.0214) 0.8250 *** (0.0214)
Health status (versus poor)
Fair - 0.6904 (0.1571) 0.6778 (0.1547) 0.6740 (0.1539)
Good - 0.8369 (0.1128) 0.8362 (0.1128) 0.8365 (0.1126)
Very good - 1.2317 (0.1658) 1.2232 (0.1648) 1.2276 (0.1652)
Excellent - 1.4863 * (0.2353) 1.4780 * (0.2341) 1.4754 * (0.2334)
Household income (versus Less than
$50,000)
$50,000 to less than $100,000 - 1.4484 *** (0.1525) 1.3947 ** (0.1480) 1.3902 ** (0.1475)
$100,000 to less than $150,000 - 1.5241 ** (0.2009) 1.4475 ** (0.1927) 1.4464 ** (0.1927)
$150,000 or more - 2.8165 *** (0.4151) 2.6311 *** (0.3921) 2.6487 *** (0.3953)
Employment status (versus unemployed)
Not working–Disabled - 0.4015 ** (0.1284) 0.3986 ** (0.1278) 0.3995 ** (0.1281)
Self-employed - 1.0184 (0.2168) 1.0065 (0.2148) 1.0013 (0.2136)
Paid employee - 0.8908 (0.1572) 0.8936 (0.1582) 0.8894 (0.1573)
Retired - 2.0200 ** (0.4737) 1.9998 (0.4697) 1.9981 ** (0.4688)
Other - 1.0104 (0.2215) 1.0155 (0.2234) 1.0112 (0.2221)
Homeownership (Yes = 1) - 1.6915 *** (0.1714) 1.6897 *** (0.1716) 1.6867 *** (0.1713)
Health-insurance ownership
(Yes = 1) - 1.4626 *** (0.1492) 1.4403 *** (0.1473) 1.4380 *** (0.1469)
Bank account ownership - 4.4484 *** (1.4650) 4.3714 *** (1.4397) 4.3167 *** (1.4200)
Credit card debt - 0.3682 *** (0.0273) 0.3687 *** (0.0274) 0.3675 *** (0.0274)
Financial risk preference (0 to 10) - 1.1226 *** (0.0180) 1.1132 ** (0.0181) 1.1140 *** (0.0181)
N 4804 4804 4804 4804
Notes: The data source is the 2018 and 2019 SHED. Odds ratios are shown alongside standard errors which are in
parentheses. * p< 0.05; ** p< 0.01; *** p< 0.001. Year dummies are included but not shown.

J. Risk Financial Manag. 2024,17, 420 14 of 18
5. Discussion, Limitations, and Implications
5.1. Discussion
The life cycle theory of savings and consumption suggests that individuals smooth
consumption over time to maximize their financial well-being. Emergency savings can
help individuals smooth consumption because it makes liquid funds available to meet
short-term financial needs that may arise unexpectedly. Thus, an individual’s ability to hold
emergency funds equal to at least three times their monthly expenses is generally accepted
as necessary for minimizing the negative impacts of idiosyncratic and systematic financial
shocks on their well-being. This study utilized SHED data to examine the impact of student
debt on holding emergency savings, including the moderating role of financial knowledge.
The results from the current study concur with prior findings (i.e., Lin et al. 2022;
Ratcliffe et al. 2022), demonstrating that many individuals do not have emergency savings
that are sufficient to cover their regular expenses for at least three months in the event of
a financial emergency. Using propensity score and coarsened exact matching to account
for self-selection bias, the findings also show that student debt may negatively impact
the emergency-saving capacity of individuals. This finding confirms the hypothesis of
the current study. Additional analysis for the interaction effect of financial knowledge
and student debt is significant statistically, with an overall positive slope for student
debtholders. However, the relationship is not as strong as expected when considering only
the main effects.
These results indicate that individuals with student debt have a lower likelihood
of saving for financial emergencies than those without student debt. The finding that
student debt decreases the possibility of having emergency savings to cover three months
of expenses in the case of emergencies lends support to the growing body of research
highlighting the consequences of student debt on financial well-being (Cho et al. 2015).
The findings also indicate that incorporating student debt into Sherraden’s (2013) model
could provide further insights into the emergency-saving behavior of individuals. From
these results, one can infer that student debt may create a stressor to the detriment of
borrowers’ financial security and wellness during economic shocks. Mossakowski (2014)
has suggested that this stressor can be detrimental to mental health.
Gjertson (2016) found that those with emergency savings have a low likelihood of
experiencing overall and specific hardships like food and energy insecurity. This study
demonstrates that carrying student debt increases the likelihood of lower levels of emer-
gency savings, and therefore a higher likelihood of the financial insecurities studied by
Gjertson and two other studies (Korankye and Kalenkoski 2021a,2021b).
5.2. Limitations
The SHED dataset lacks information on the dollar values of emergency savings and
student loan debt for further analysis. Additionally, the question of whether an individual
has emergency savings or not uses three months as the generally accepted threshold.
Although this minimum threshold is prevalent among some financial educators/advisors,
it ignores those smaller amounts of emergency savings that could cushion individuals
(particularly those in low-income brackets) against mild financial emergencies; a more
nuanced understanding of the tipping point between financial security and a cascade
of negative effects was used by Gjertson (2016). Furthermore, the Big Three financial
knowledge questions are the only information available in the database, limiting the ability
of this study to explore the causal relationships between loan debt, risk, and stress. Other
questions relating to asset pricing, mortgage payment, and subjective financial knowledge
are unavailable in the SHED for analysis and could enhance the understanding of the
reported effects. Finally, the study lacks data on certain variables, such as household
wealth and other forms of debt. Nonetheless, studies have shown that the propensity score
methods utilized in this study can be used to control for confounding (Yang et al. 2019).

J. Risk Financial Manag. 2024,17, 420 15 of 18
5.3. Implications
The findings from the current study suggest that it is essential for student loan
debtholders to develop positive financial skills associated with saving for emergencies.
Although student debt service can impose financial restrictions on holders, it is necessary
to contextualize debt as part of an individual’s overall financial landscape. Doing so can
reduce the risk of narrow framing, where individuals focus on one aspect of their finances
to the detriment of the larger picture of which student debt is just one component of all debt.
Financial educators who consider emphasizing the importance of consumption smooth-
ing across the life cycle in their interactions with student debtholders may aid in their
students’ financial outcomes. For instance, financial educators who show how a financial
budget encompassing the relevant aspects of cashflows and future debt accumulation
could help student debtholders consider emergency savings as a necessity. Whether in
the classroom, workplace, or another community setting, providing financial education is
likely one of the most viable solutions to help build emergency savings.
Financial stress and lower mental well-being could create default risks for both student
debtholders and the institutions that seek to reduce defaults on their loans. These find-
ings suggest that financial education programs should include direct instructions on how
debtholders can participate in income-driven repayment plans. Furthermore, private stu-
dent loan lenders could consider the value of reducing the chances of a default by adopting
income-driven repayment plans that can reduce the stress on borrowers, thereby ensuring
a lower likelihood of a default occurring due to a financial emergency in that borrower’s
life. Ultimately, measures such as these could motivate student debtholders to gradually
build sufficient emergency funds to cover at least three months of their monthly expenses
during financial emergencies, which leads to a long-term lower risk of default occurrence.
The results on financial knowledge and bank account ownership, for instance, em-
phasize their usefulness in motivating individuals to save for financial emergencies. The
findings on these factors, which lend support to the literature (Babiarz and Robb 2014;
Despard et al. 2020;Fan and Zhang 2021), suggest that individuals may be more likely
to save for financial emergencies when they have more financial knowledge and are in-
cluded in the financial system through account ownership. Thus, financial education
efforts geared towards enhancing financial acumen and inclusiveness represent steps in the
right direction.
The interaction effect results suggest that financial knowledge could play a role in
helping student debtholders to hold emergency savings that would be sufficient to meet
at least three months of expenses during economic shocks. However, the large gap in the
likelihood of holding emergency savings between student debtholders and those without
such debt suggests the need for policy and curriculum interventions during the period
leading up to application, during loan terms, when accepting a loan, and early in a student’s
educational career as they are growing their debt burden through student loans. This policy
suggestion is recommended for borrowers of both private and federal loans. The existence
of such intervention warrants intensification.
The federal government may consider a threshold that allows student debt holders
to build emergency savings before they start paying off their student loans. For instance,
a program that allows student loan holders an exemption for their student loan interest
requirements while building their emergency savings would allow borrowers to build
emergency savings with lessened financial stress. Additionally, a delayed interest accumu-
lation stage upon graduation, such as 6-month or 1-year, would provide borrowers the time
to build their emergency funds before being subject to interest charges. Initiatives such as
these could improve the financial resilience of affected individuals and families as they
save for unforeseen future financial circumstances. Having a financially resilient populace
is more likely beneficial for stakeholders, including the federal government, compared with
a non-financially resilient populace.

J. Risk Financial Manag. 2024,17, 420 16 of 18
Author Contributions:
Conceptualization, T.K.; methodology, T.K., B.P. and P.A.-M.; software, T.K.;
validation, T.K., B.P. and P.A.-M.; formal analysis, T.K. and B.P.; investigation, T.K., B.P. and P.A.-M.;
resources, T.K., B.P. and P.A.-M.; data curation, T.K.; writing—original draft preparation, T.K. and
B.P.; writing—review and editing, T.K., B.P. and P.A.-M.; visualization, T.K., B.P. and P.A.-M.; funding
acquisition, T.K. All authors have read and agreed to the published version of the manuscript.
Funding:
This research was funded by the National Endowment for Financial Education and Knology
through the Financial Education Database Training Fellowship awarded to Thomas Korankye.
Data Availability Statement:
Publicly available datasets were analyzed in this study. This data can
be found here: [https://gitlab.com/knologyresearch/KnologyFinEdStateSpending], accessed on
20 June 2022.
Acknowledgments:
Thomas Korankye is grateful to John Fraser, Joseph Dwyer, and Shaun Field
from Knology and the 2022 Fellowship cohort for their support. The authors also thank the partic-
ipants of the 2023 Academy of Financial Services conference in Phoenix, Arizona (USA), for their
insightful feedback.
Conflicts of Interest: The authors declare no conflict of interest.
References
Abadi, Alberto, and Guido W. Imbens. 2016. Matching on the estimated propensity score. Econometrica 84: 781–807. [CrossRef]
Adams, Deborah, and Stacia West. 2015. Asset building among low income adults: An exploratory study with participants in an
emergency savings program. Journal of Community Practice 23: 436–61. [CrossRef]
Alyousif, Maher, and Charlene M. Kalenkoski. 2017. Asking for Action: Does Financial Advice Improve Financial Behaviors?
SSRN 2954773. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2954773 (accessed on 1 May 2020).
Babiarz, Patryk, and Cliff A. Robb. 2014. Financial literacy and emergency saving. Journal of Family and Economic Issues 35: 40–50.
[CrossRef]
Becker, Gary. 1962. Investment in human capital: A theoretical analysis. Journal of Political Economy 70: 9–49. [CrossRef]
Benedetto, Umberto, Stuart J. Head, Gianni D. Angelini, and Eugene H. Blackstone. 2018. Statistical primer: Propensity score matching
and its alternatives. European Journal of Cardio-Thoracic Surgery 53: 1112–17. [CrossRef]
Bhutta, Neil, Jacqueline Blair, and Lisa Dettling. 2022. The smart money is in cash? Financial literacy and liquid savings among U.S.
families. Journal of Accounting and Public Policy 42: 107000. [CrossRef]
Blackwell, Matthew, Stefano Iacus, Gary King, and Giuseppe Porro. 2009. cem: Coarsened exact matching in Stata. The Stata Journal
9: 524–46. [CrossRef]
Brobeck, Stephen. 2008. Understanding the Emergency Savings Needs of Low-and Moderate-Income Households: A Survey-Based Analysis of
Impacts, Causes, and Remedies. Washington, DC: Consumer Federation of America.
Cho, Soo Hyun, Yilan Xu, and D. Elizabeth Kiss. 2015. Understanding student loan decisions: A literature review. Family and Consumer
Sciences Research Journal 43: 229–43. [CrossRef]
Despard, Mathieu R., Terri Friedline, and Stacia Martin-West. 2020. Why do households lack emergency savings? The role of financial
capability. Journal of Family and Economic Issues 41: 542–57. [CrossRef]
Drukker, David M. 2016. Estimating treatment effects from observational data using teffects, stteffects, and eteffects. Paper presented
at the United Kingdom Stata Users’ Group Meetings, London, UK, September 8–9; Available online: http://fmwww.bc.edu/
repec/usug2016/drukker_uksug16.pdf (accessed on 15 October 2022).
Dwyer, Joseph Torre, Shaun Field, and Bennett Attaway. 2021. KnologyFinEdStateSpending: A Database of Financial Education
Activities in the U.S. by State and Year. R Package Version 1.0.0. Knology. Available online: https://gitlab.com/knologyresearch/
KnologyFinEdStateSpending (accessed on 20 June 2022).
Fan, Lu, and Lini Zhang. 2021. The influence of financial education sources on emergency savings: The role of financial literacy. Family
and Consumer Sciences Research Journal 49: 344–61. [CrossRef]
Federal Reserve Bank of New York. 2021. Quarterly Report on Household Debt and Credit. Quarter 1, 2021: Q1. New York: Center for
Microeconomic Data.
Garrido, Melissa M., Amy S. Kelley, Julia Paris, Katherine Roza, Diane E. Meier, R. Sean Morrison, and Melissa D. Aldridge. 2014.
Methods for constructing and assessing propensity scores. Health Services Research 49: 1701–20. [CrossRef]
Gjertson, Leah. 2016. Emergency saving and household hardship. Journal of Family and Economic Issues 37: 1–17. [CrossRef]
Hasler, Andrea, Annamaria Lusardi, and Noemi Oggero. 2018. Financial Fragility in the U.S.: Evidence and Implications. Washington, DC:
Global Financial Literacy Excellence Center, The George Washington University School of Business.
Iacus, Stefano, Gary King, and Giuseppe Porro. 2009. CEM: Software for coarsened exact matching. Journal of statistical Software
30: 1–27. [CrossRef]

J. Risk Financial Manag. 2024,17, 420 17 of 18
Kim, Kyoung Tae, Jae Min Lee, and Jonghee Lee. 2021. Student loans and financial satisfaction: The moderating role of financial
education. Financial Counseling and Planning 32: 266–79. [CrossRef]
Kim, Kyoung Tae, Tae-Young Pak, Su H. Shin, and Sherman D. Hanna. 2018. The relationship between financial planner use and
holding a retirement saving goal: A propensity score matching analysis. Financial Planning Review 1: e1008. [CrossRef]
King, Gary, and Richard Nielsen. 2019. Why propensity scores should not be used for matching. Political Analysis 27: 435–54. [CrossRef]
Korankye, Thomas. 2023. Student loan debt and US married households’ stock investment decisions. Economic Analysis Letters 2: 13–18.
[CrossRef]
Korankye, Thomas. 2024. Student loan debt in retirement: Identifying the correlates and implications for policy, practice, and research.
Managerial Finance 50: 791–810. [CrossRef]
Korankye, Thomas, and Charlene M. Kalenkoski. 2021a. Student loan debt and financial well-being of the borrower: Does it matter
whom the debt is for? Journal of Personal Finance 20: 74–88.
Korankye, Thomas, and Charlene M. Kalenkoski. 2021b. The effect of households’ student debt on life satisfaction. Journal of Family and
Economic Issues 42: 757–72. [CrossRef]
Korankye, Thomas, and Michael Guillemette. 2021. Student debt and stock-ownership decisions of U.S. households. Applied Economics
Letters 28: 387–90. [CrossRef]
Korankye, Thomas, Blain Pearson, and Hossein Salehi. 2023. Financial advice use and saving for children’s college education: A
propensity score matching approach. Journal of Financial Counseling and Planning 34: 96–111. [CrossRef]
Lanza, Stephanie T., Julia E. Moore, and Nichole. M. Butera. 2013. Drawing causal inferences using propensity scores: A practical
guide for community psychologists. American Journal of Community Psychology 52: 380–92. [CrossRef]
Li, Qian. 2021. Student loan burden, financial literacy, and financial behavior-evidence from the 2018 FINRA survey. Journal of
Accounting & Finance 21: 2158–3625.
Lin, Judy T., Christopher Bumcrot, Garry Mottola, Olivia Valdes, Robert Ganem, Christine Kieffer, Annamaria Lusardi, and Gerri
Walsh. 2022. Financial Capability in the United States: Highlights from the FINRA Foundation National Financial Capability Study,
5th ed. Rockville: FINRA Investor Education Foundation. Available online: www.FINRAFoundation.org/NFCSReport2021
(accessed on 12 July 2022).
Liu, Yi, Thomas Korankye, and Blain Pearson. 2023. Personality traits and student loan holding for self and for children among baby
boomers. Journal of Financial Counseling and Planning 34: 415–29. [CrossRef]
Lusardi, Annamaria. 2008. Financial Literacy: An Essential Tool for Informed Consumer Choice? No. w14084. Cambridge: National Bureau
of Economic Research.
Lusardi, Annamaria, and Olivia. S. Mitchell. 2017. How ordinary consumers make complex economic decisions: Financial literacy and
retirement readiness. Quarterly Journal of Finance 7: 1750008. [CrossRef]
Lusardi, Annamaria, Daniel J. Schneider, and Peter Tufano. 2011. Financially Fragile Households: Evidence and Implications. No. w17072.
Cambridge: National Bureau of Economic Research.
Ma, Jennifer, Sandy Baum, M. Pender, and C. J. Libassi. 2018. Trends in College Pricing 2018. New York: The College Board.
Marsden, Mitchell, Cathleen D. Zick, and Robert N. Mayer. 2011. The value of seeking financial advice. Journal of Family and Economic
Issues 32: 625–43. [CrossRef]
Martin, Terrence K., Jr., Lua A. V. Augustin, Laura C. Ricaldi, and Jose Nunez. 2020. The effect of student Loans on parental views of
education financing. Journal of Financial Planning 33: 46–57.
McDonald, Wren. 2021. 15 Personal-Finance Lessons We Can All Learn from the Year of COVID-19. The Wall Street Journal. Available
online: https://www.wsj.com/articles/15-personal-finance-lessons-we-can-all-learn-from-the-year-of-covid- 19-11611493201
?st=s27owrhy7oy6h5z&reflink=share_mobilewebshare (accessed on 12 July 2022).
Mills, Gregory, and Joe Amick. 2010. Can Savings Help Overcome Income Instability? Perspectives on Low-Income Working Families
Brief, 18. Washington, DC: The Urban Institute.
Mossakowski, Krysia N. 2014. Stress and mental health. In The Wiley-Blackwell Encyclopedia of Health, Illness, Behavior, and Society. Edited
by William C. Cockerham, Robert Dingwall and Stella R. Quah. Hoboken: Wiley-Blackwell.
Mountain, Travis P., Xiang Cao, Namhoon Kim, and Michael S. Gutter. 2020. Millennials’ future homeownership and the role of
student loan debt. Family and Consumer Sciences Research Journal 49: 5–23. [CrossRef]
Pearson, Blain, and Jae Min Lee. 2022. Student debt and healthcare service usage. Journal of Financial Counseling and Planning 33: 183–93.
[CrossRef]
Ratcliffe, Caroline, Brianna Middlewood, Melissa A. Z. Knoll, Grant Guillory, and Misha Davies. 2022. Emergency Savings and Financial
Security: Insights from the Making Ends Meet Survey and Consumer Credit Panel. Consumer Financial Protection Bureau Office of
Research Reports Series, (22-1). Washington, DC: Consumer Financial Protection Bureau.
Sherraden, Margaret. 2013. Building blocks of financial capability. In Financial Capability and Asset Development: Research, Education,
Policy, and Practice. Edited by Julie Birkenmaier, Margaret Sherraden and Jami Curley. New York: Oxford University Press,
pp. 3–43. [CrossRef]
Stuart, Elizabeth A. 2010. Matching methods for causal inference: A review and a look forward. Statistical Science 25: 1–21. [CrossRef]
Tharp, Derek T., Martin Seay, Cherie Stueve, and Somer Anderson. 2020. Financial satisfaction and homeownership. Journal of Family
and Economic Issues 41: 255–80. [CrossRef]

J. Risk Financial Manag. 2024,17, 420 18 of 18
Woodyard, Ann Sanders, and Cliff A. Robb. 2016. Consideration of financial satisfaction: What consumers know, feel and do from a
financial perspective. Journal of Financial Therapy 7: 4. [CrossRef]
Yang, Jeff Y., Michael Webster-Clark, Jennifer L. Lund, Robert S. Sandler, Evan S. Dellon, and Til Stürmer. 2019. Propensity score
methods to control for confounding in observational cohort studies: A statistical primer and application to endoscopy research.
Gastrointestinal Endoscopy 90: 360–69. [CrossRef]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.