Curs, Harper, Park, The Role of Emergency Financial Relief Funding in Improving Student Outcomes (EdWorkingPaper 24-991) (2024-07)
Summary
EdWorkingPaper 24-991, dated July 2024, by Bradley R. Curs, Casandra E. Harper and Sangmin Park of the University of Missouri, on emergency financial relief for low-income students. It examines a one-time relief program among Pell Grant eligible undergraduates in Spring 2015. The paper states that the university waived $214,308 in outstanding balances for 797 students. It reports that the relief can improve retention and graduation rates, was most effective among first-generation college students, and did not improve GPA. It also reports that students receiving the relief were more likely to receive a financial hold in a subsequent semester.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“Curs, Harper, Park, The Role of Emergency Financial Relief Funding in Improving Student Outcomes (EdWorkingPaper 24-991) (2024-07).”
Original at edworkingpapers.com. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/curs-harper-park-the-role-of-emergency-financial-relief-funding-in-imp-af539b1077/
, SHA-256 4268262da7173622baddbdca54d2e4ed5ab6573dd6753daa5a54f64d9c2dc843.Full text
EdWorkingPaper No. 24-991
The Role of Emergency Financial Relief
Funding in Improving Low-Income Students’
Academic and Financial Outcomes Across
Demographic Characteristics
Bradley R. Curs Casandra E. Harper Sangmin Park
University of Missouri University of Missouri University of Missouri
This quasi-experimental study examined the effectiveness of a one-time emergency financial relief program
among Pell Grant eligible undergraduate students in Spring 2015 pursuing their first bachelor’s degree across
academic and financial outcomes. The academic outcomes included retention to the next semester, degree
completion, attempted credit hours, and grade point average. The financial outcome captured whether students
received a stop registration hold due to an unpaid financial balance in the semester after receiving the
emergency relief. The results reveal that financial relief applied to low-income students’ accounts can improve
their retention and graduation rates. The financial relief was most effective among first-generation college
students, resulting in a complete elimination of the retention gap for first-generation students. The emergency
relief did not improve GPA or substantially change the number of credits earned. A concerning finding was that
students receiving this emergency support were more likely to receive a financial hold in a subsequent semester
and that effect was stronger among students of color (Black/African American, Hispanic/Latine, Asian,
Multiracial, American Indian/Alaska Native), males, and first-generation college students.
VERSION: July 2024
Suggested citation: Curs, Bradley R., Casandra E. Harper, and Sangmin Park. (2024). The Role of Emergency Financial Relief
Funding in Improving Low-Income Students’ Academic and Financial Outcomes Across Demographic Characteristics.
(EdWorkingPaper: 24-991). Retrieved from Annenberg Institute at Brown University: https://doi.org/10.26300/wgyh-ek62
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
The Role of Emergency Financial Relief Funding in Improving Low-Income Students’ Academic and
Financial Outcomes Across Demographic Characteristics
Bradley R. Curs, Casandra E. Harper, & Sangmin Park
Department of Educational Leadership & Policy Analysis, University of Missouri
Author Note
Bradley R. Curs https://orcid.org/0000-0003-3356-127X
Casandra E. Harper https://orcid.org/0000-0001-7889-7952
Sangmin Park
We have no known conflict of interest to disclose.
Institutional data were obtained in partnership with the Office of the Provost, Student Financial
Aid, Enrollment Management, and the Office of Cashier’s at the University of Missouri. The research
reported in this article was made possible in part by a grant from the Spencer Foundation (#202300145).
The views expressed are those of the authors and do not necessarily reflect the views of the Spencer
Foundation.
Correspondence concerning this article should be addressed to Bradley R. Curs, University of
Missouri, 303 Townsend Hall, Columbia, MO 65211. Email: cursb@missouri.edu
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Abstract
This quasi-experimental study examined the effectiveness of a one-time emergency financial relief
program among Pell Grant eligible undergraduate students in Spring 2015 pursuing their first bachelor’s
degree across academic and financial outcomes. The academic outcomes included retention to the next
semester, degree completion, attempted credit hours, and grade point average. The financial outcome
captured whether students received a stop registration hold due to an unpaid financial balance in the
semester after receiving the emergency relief. The results reveal that financial relief applied to low-
income students’ accounts can improve their retention and graduation rates. The financial relief was
most effective among first-generation college students, resulting in a complete elimination of the
retention gap for first-generation students. The emergency relief did not improve GPA or substantially
change the number of credits earned. A concerning finding was that students receiving this emergency
support were more likely to receive a financial hold in a subsequent semester and that effect was
stronger among students of color (Black/African American, Hispanic/Latine, Asian, Multiracial, American
Indian/Alaska Native), males, and first-generation college students.
Keywords: emergency funds, financial relief, low-income, first-generation college students,
retention
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
The Role of Emergency Financial Relief Funding in Improving Low-Income Students’ Academic and
Financial Outcomes Across Demographic Characteristics
College students are increasingly unable to cover necessary expenses (Goldrick-Rab et al., 2022;
Grodsky & Jones, 2007; Page & Scott-Clayton, 2016), and most college students report feeling stressed
about finances (Heckman et al., 2014). In addition to the expected expenses (e.g., tuition, fees,
textbooks), unexpected and emergency costs can be devastating to students’ progress (Evans et al.,
2019), sometimes leading to college departure (Krueger et al., 2016; Lowman et al., 2021). More than
40% of first-time enrolled students in four-year institutions do not complete a bachelor’s degree within
six years (NCES, 2020), and millions will not complete their degree at all (Engle & Tinto, 2008; Shapiro et
al., 2017). Some institutions have tried to assist students by providing emergency relief with emergency
federal (Goldrick-Rab et al., 2021) or institutional funds (Evans et al., 2020; Krueger et al., 2016), but
little empirical research exists to determine whether such efforts are effective in preventing future
financial crises for students or improving outcomes such as time to degree (Anderson et al., 2020;
Lowman et al., 2021), and specifically for students of color, low-income, and first-generation college
students.
Students of color, low-income, and first-generation college students experience additional
barriers related to financial needs and support. Specifically, more than one third of Hispanic students
were unaware of any financial aid source (O’Connor et al., 2010), and Black students—especially those
who are first in their families to attend college—are less likely to know about financial aid options
(Ramirez-Mendoza & Jones, 2020). Similarly, students from low-income families are less likely to receive
high-quality information about financial aid opportunities and are less likely to file an application for
financial aid (Goldrick-Rab et al., 2009; Long, 2008). Given rising college costs, disadvantages in
accessing financial aid information could mean the difference between degree completion or attrition
(Chen & DesJardins, 2010).
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
As word of mouth is still the primary method for disseminating financial aid information
(Krueger et al., 2016), students who need emergency funding can easily miss information and deadlines.
The recent pandemic, while setting back learning for all students, exacerbated hardships for low-income
and students of color (Harper, 2020). Economic hardships, such as job losses, have left students of color
and low-income students struggling to pay for necessities and academic needs (Amour, 2020; Curs et al.,
2022). Some universities responded with small emergency financial awards; Canadian institutions in
Ontario, for example, offered awards of $500-$1,500 through the Bursar’s office (El Masri & Sabzalieva,
2020). Therefore, it is necessary to investigate whether emergency financial aid is helpful in improving
postsecondary outcomes for students, particularly by race and income.
During the 2015 summer semester, the University of Missouri delivered a one-time financial
relief program that waived eligible small outstanding financial balances for students. Eligible outstanding
balances included expenses related to direct academic expenses (e.g., tuition, fees, course materials)
but did not include other non-academic expenses (e.g., clothing purchases at the MU Bookstore).
Students were unaware of this program and there was no application process as the balance waivers
were automatically applied to eligible student’s accounts. In total, the university waived $214,308 in
outstanding balances for 797 students. The average waived balance was $269 and ranged from $10 to
$4,000.
Specifically, our research questions are:
1) Was receiving a balance waiver through this one-time financial relief program associated with
subsequent academic and financial outcomes?
2) Was the association between the financial relief program and student outcomes different
across demographic characteristics, such as race, sex, first-generation status, or year in school?
This study contributes to our understanding of financial relief programs that institutions might
implement themselves, and this research informs the current CARES Act (Coronavirus Aid, Relief, and
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Economic Security) and HEERF (Higher Education Emergency Relief Funds) funding that serves a similar
function on a larger scale (Schmidt & Weissman, 2021)
Relevant Literature
Students have increasing financial, unmet, and emergency needs. The rising cost of college is a
major concern for students and their families (Mitchell et al., 2019; Zaloom, 2019). An increasing
proportion of students are experiencing unmet need—where the expected cost of college exceeds what
families can pay or cover through financial aid scholarships, grants, or loans (Walizer, 2018). Nearly
three in four students have an unmet need and in public, four-year colleges the average size of that gap
is $9,000 (Walizer, 2018). The unmet need among students of color and low-income students is even
greater (Walizer, 2018; Lopez, 2013). Black students, regardless of the institution type that they attend,
are found most likely to have unmet need compared to their peers from other racial/ethnic groups
(Walizer, 2018).
Institutional relief funding programs to support students in financial need, which more than 70%
of higher education institutions have, can take on different forms (e.g., vouchers, scholarships, loans,
food access, restricted and unrestricted grants). These programs serve a variety of objectives, such as
humanitarian aid, removing barriers to success, removing financial holds so that students can graduate,
or to retain students (Krueger et al., 2016). Institutions are most likely to rely on donors and their own
operating budget to procure these funds, but students’ needs tend to exceed funding availability
(Krueger et al., 2016).
Broadly, research on emergency financial aid programs has not found strong evidence of effects
on student outcomes. Evans et al. (2019, 2020) studied institutional relief for low-income community
college students and found that emergency aid, even when paired with a case management coaching
and mentoring program, did not significantly improve student success (e.g., course enrollment, credits
earned, GPA, or degree completion). A recent 11-institution study of public universities using
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
completion grants to assist students with unmet need through randomly applied small grants ($1,200 on
average), resulting in a modest decrease in credit card usage to pay for college but no change in
graduation rates (Goldrick-Rab et al., 2022). The state of Wisconsin tried a lottery-based, random
assignment of one-time funds to increase degree completion and graduate school enrollment, which the
program failed to improve (Anderson et al., 2020). There was some reduction in students’ time-to-
degree and a slight increase in the number of STEM degrees earned during the program (Anderson et
al., 2020). Among students attending HBCU institutions, there is preliminary evidence that emergency
funding is most helpful in reducing stress, helping with educational expenses for class, and maintaining
enrollment, according to students’ own perceptions (Dahl et al., 2022).
The initial evidence about one-time relief programs is that these funds do not significantly
predict student persistence into the following semester (Lowman et al., 2021). Low-income students are
particularly impacted by emergency events, such as COVID-19 (McKinnon-Crowley, 2023), and can
particularly benefit from emergency aid (Rodríguez-Planas, 2020). Most students surveyed about
receiving these emergency funds reported how this aid was helpful to them personally and
academically, communicated care and concern from their institution, and believed that the money
would increase their chance of graduating (Goldrick-Rab et al., 2021). Administrators have similar
perceptions about the benefits of such funds helping students stay enrolled (Taylor & Melidona, 2021).
The financial relief program we are studying is small (i.e., awards were less than $4,000) and was
a surprise to students. Thus, we will utilize the broad concept of scarcity as a theoretical framework for
our research study. Generally, when people are faced with a scarcity of important resources (e.g.,
money, time), they are more likely to focus on immediate needs and have reduced bandwidth for other
productive activities (Mullainathan & Shafir, 2013). In our study, students who received relief funds for
small financial balances may be less likely to worry about finances and might then be able to focus their
attention on their academic success. Previous research has shown that financial stress is increasing and
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
pervasive (Heckman et al., 2014) and has negative associations with academic outcomes (e.g., Baker &
Montalto, 2019; Goldrick-Rab, 2016; Joo et al., 2008).
Research Methodology
In this section we describe the research methodology we utilized to estimate the relationship
between receipt of the financial relief program and academic and financial outcomes. We begin by
describing the data sources and variables utilized in our analysis. Next, we describe the target
population and restrictions made to create the analytical sample - Pell Grant eligible students at the
University of Missouri that were enrolled in the spring 2015 semester. We then describe our preferred
empirical strategy, propensity score matching, and the decisions we made within the estimation
process. We also discuss alternative estimation strategies utilized to demonstrate the robustness of our
approach. Within our analysis, we also test for differential responses to the financial relief program
across student demographic characteristics (e.g., race/ethnicity, sex, first-generation status, and year in
school).
Data Sources and Analytic Variables
The data utilized for this study come from University of Missouri administrative records.
Demographic information was collected from application records. Academic outcomes were collected
from semester-level academic records. Treatment status was collected from financial aid records.
Finally, registration holds were collected from a financial transaction data file that contains all records of
official university correspondence related to student financial issues (e.g., past due balances and/or
registration holds). All data sources were merged using a student-level unique identifier created for
research purposes to protect the anonymity of the student records. Descriptive statistics of all variables
utilized in the analysis for the analytic sample are provided in Table 1.
Dependent Variables
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
We investigated the effect of receiving the financial relief award on five dependent variables.
The first dependent variable was spring to fall semester retention. Specifically, we classified a student as
having been retained if they attempted and completed at least one 3-credit course in the subsequent
fall semester. For the analytical sample, the mean spring to fall retention rate was 91%.
The second dependent variable was completion of a bachelor’s degree. Specifically, we classified
students as completing their bachelor’s degree if they received their bachelor’s degree from the
University of Missouri by the Spring 2022 semester. Thus, our completion variable is censored at seven
years beyond the financial relief treatment. For the analytical sample, the mean degree completion rate
was 83%.
The third dependent variable was the number of credit hours attempted in the fall 2015
semester, the semester immediately after the implementation of the financial relief program. For the
analytical sample (removing those that were not retained into the fall 2015 semester), the mean
number of credit hours attempted in the fall 2015 semester was 13.2. This average credit load implies
that the average student was completing a full-time course load (defined as 12 credits or more) but was
not attempting enough credits to maintain a 4-year graduation pace (15 per semester).
The fourth dependent variable is the grade point average for fall 2015 coursework. The mean
(removing those that were not retained into the fall 2015 semester) fall 2015 grade point average was
3.0, on a 0.0 to 4.0 scale.
The fifth dependent variable measures whether a student receives a stop registration hold due
to financial delinquency during the fall 2015 semester. Stop registration holds are placed on a student’s
account if they have an outstanding financial balance that is greater than 60 days past due. Until the
outstanding balance is taken care of, students face penalties that include the prevention of future class
enrollment, requests for transcripts, and receipt of a diploma. For the analytical sample (removing those
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
that were not retained into the fall 2015 semester), 21% of students received a stop registration hold
during the fall 2015 semester.
Treatment Variable
The treatment indicator variable represents whether a student received the financial relief
award. Within the analytical sample, 6% of the students received this award. Within the analytic sample,
the average balance that was waived (for those that received an award) was $198 and ranged from $10
to $4,000.
Covariates
We include covariates in our regression and matching models to control for factors related to
prior academic achievement, prior financial factors, and how many years the student has been enrolled
at the institution. Our rationale for covariate selection is described below when we describe our
empirical strategy.
With respect to prior academic achievement, we include the student’s cumulative grade point
average through spring 2015, their spring 2015 grade point average, and their spring 2015 number of
credit hours attempted. With respect to financial factors, we include information about the student’s
2014-2015 academic year financial aid packaging including: their expected family contribution (EFC), and
indicator variable as to whether they received institutional needs-based financial aid, and an indicator
variable as to whether they received institutional merit-based financial aid. To capture prior financial
troubles, we include indicator variables for whether they had a stop registration holds in spring 2015 or
fall 2014 semesters, and variables that measure the balance sizes associated with those holds. Finally,
we include indicator variables for whether the spring 2015 semester was their second, third, or fourth
year at the institution (first year serves as the reference).
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Demographic Variables
We utilize three demographic factors to test whether the treatment differentially affects
students. First, we test for differences between students of color and white students. Specifically, we
define students of color as students who self-reported their race/ethnicity on their application for
admissions as Black/African American, Hispanic/Latine, Asian, Multiracial, or, American Indian/Alaska
Native. Unfortunately, sample sizes for treated groups restricts the statistical power to analyze
treatment effects for more segmented race/ethnicities.
Second, we test for differences across sex/gender. At the time of the financial relief program, in
the application for admissions the University of Missouri asked students for their self-reported gender
but provided sex (male or female) categories as the response options. Thus, we utilize the binary sex
categories of male and female in our statistical reporting.
Third, we test for differences across first-generation status. The University of Missouri classifies
a student as a first-generation college student based upon their response to a question on their
admissions application asking, “Prior to your 18th birthday, did you either reside with or receive support
from a parent who has a bachelor’s degree from a college or university?” Students responding “no”
were classified as first-generation.
Fourth, we test for differences across the number of years a student was enrolled at the
University of Missouri. We categorize students into years based upon when they first enrolled at the
University of Missouri. Thus, we define a first-year student as someone who first enrolled at the
University of Missouri during the academic year (fall 2014 to summer 2015) of the treatment (Summer
2015). A second year is defined as having started during the academic year prior to the treatment (fall
2013 to summer 2014). Third- and fourth-year students are similarly defined as second-year students.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Sample Characteristics
The base population for this study includes all Pell Grant eligible undergraduate students who
were enrolled at the University of Missouri during the spring 2015 semester (n = 12,287). To create an
analytic sample appropriate to analyze the effect of the financial relief program, we made a few sample
restrictions. First, we removed all students who were ineligible for the financial relief treatment as they
graduated during the spring 2015 or summer 2015 semesters (n = 1,772). Second, we removed students
who started their bachelor’s degree program prior to 4 years before the spring 2015 semester (n = 347).
Thus, the potential analytical sample for this study was 10,168 Pell Grant eligible students who were
within four years of starting their degree program. The final analytic sample 9,653 after the trimming
procedure described in the following section.
Empirical Strategy
Our preferred estimation strategy employs a propensity score matching estimator to estimate
the effects of the treatment on academic and financial outcomes. This quasi-experimental procedure
allows us to make causal estimates of the treatment effect under two primary assumptions: the
unconfoundedness assumption and the overlap assumption (Caliendo & Kopeinig, 2008; Imbens, 2015).
The following three steps describe our approach to making the decisions necessary to estimate the
propensity score matching treatment effect within the context of the two primary assumptions.
Step One – Choose the Covariates and Estimate the Propensity Score
Following the guidance of Imbens (2015), the first step of our propensity score matching
strategy is to choose the set of covariates to include in the propensity score model. Our choice of
covariates was guided by the unconfoundedness assumption, which states “given a set of observable
covariates X which are not affected by treatment, potential outcomes are independent of treatment
assignment” (Caliendo & Kopeinig, 2008, p. 35). Thus, all constructs that affect both treatment status
and the outcome must be observed and accounted for by the researchers. This assumption is inherently
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
untestable, and scholars have stated that the most credible strategies for satisfying this assumption
come through the richness of the data and detailed understanding of the treatment selection process
(Blundell et al., 2005; Imbens, 2015; Steiner et al., 2010).
We utilized the detailed administrative longitudinal data on both academic constructs (i.e.,
cumulative GPA, spring 2015 GPA, spring 2015 credit hours passed) and financial outcomes (e.g., EFC,
spring 2015 registration hold indicator and amounts, fall 2014 registration hold and amounts, and
indicators for receipt of institutional need- and merit-based grants) to account for factors likely to be
related to both treatment status and subsequent academic outcomes. We also include indicators of
whether a student is in their second, third, and fourth year at the University of Missouri (students in
their first year are the reference group).
Table 2 presents estimation of the propensity score model for the full sample (column 1) and
the sample used for treatment effects after the trimming process in step 2 (column 2). With respect to
academic factors, the probability of receiving treatment was positively associated with cumulative GPA
and the number of credit hours passed in spring 2015 (the semester preceding the financial treatment).
With respect to financial factors, the treatment was negatively related to a student’s EFC and positively
related to students having received a stop registration hold in the spring 2015 and fall 2014 semesters.
First year students were more likely to receive the treatment when compared to second-, third-, or
fourth-year students.
Step Two – Trim the Sample
The overlap assumption “ensures that persons with the same X values have a positive
probability of being both participants and nonparticipants” (Caliendo & Kopeinig, 2008, p. 35). After
estimating the propensity scores on the full sample, the overlap assumption can be tested through
inspection of the overlap of propensity scores for the treated and untreated groups. Figure 1 presents
the propensity score distributions for the treated and untreated groups. The vertical dashed lines
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
represent the boundaries for which there is an overlap between the propensity scores for the treated
and untreated groups. Following guidance from the propensity score literature (e.g., Caliendo &
Kopeinig, 2008, Imbens, 2015) we removed observations with propensity scores outside of this range of
overlap. Specifically, we removed 158 observations whose propensity score was less than 0.0062 and
357 observations with propensity scores above 0.3757. Propensity scores were then re-estimated
(presented in column 2 of Table 2) on the final analytic sample of 9,653 observations.
Step Three – Estimate the Treatment Effects
After choosing the covariates for the propensity score model and trimming the sample, we
estimated the treatment effects using a propensity score matching estimator. Our strategy utilized a
single nearest neighbor from the comparison group to match to each treated observation. Specifically,
average treatment effects using propensity score matching were estimated using the teffects psmatch
command in Stata SE 17 (StataCorp, 2021).
Figure 2 presents the standardized differences for each covariate used in the analysis between
the treated and untreated groups. Circles represent the standardized differences in the raw data and
triangles for the matched samples. For the raw data, concerning standardized differences (i.e., greater
than 0.25) were observed for variables that measured a student’s EFC and whether they had a
registration hold in spring 2015. After matching, the differences were reduced (all less than 0.1)
providing evidence that the matching process reduced potential concerns of the validity of the
unconfoundedness assumption.
Alternative Estimation Strategies. Imbens (2015) suggests providing estimates from alternative
treatment effects estimators to check for robustness across estimation techniques. Thus, we also
present estimates from two alternative treatment effects estimation strategies. First, we present results
using a linear regression with covariate adjustment strategy. All covariates described in the covariates
section above will be included in the linear regression specification.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Second, a nearest neighbor matching design using a Mahalanobis distance allows us to compare
similar students based upon observable characteristics who did and did not receive the financial relief
treatment. Specifically, we forced matches to be exact based upon the year of enrollment for each
student (first, second, third, or fourth year of enrollment) to make sure that treated students were
matched to non-treated students who were in similar places in their academic careers. Second, we
included all the covariates previously described within our matching process to help predict whether a
student received the treatment. The nearest neighbor matching estimates were obtained through the
teffects nnmatch command in Stata SE 17 (StataCorp, 2021).
Results
Table 3 presents the results with respect to our first research question, what was the effect of
receiving a balance waiver through this financial relief program on subsequent academic and financial
outcomes? Results are presented for all five dependent variables for three analytical treatment effects
methods; (a) propensity score matching (PSM); (b) linear regression with covariate adjustment (OLS);
and (c) nearest neighbor matching (NNM). Interpretation of coefficients will be focused on our preferred
estimation strategy – propensity score matching, with comparison to the other two specifications (OLS
and NNM) when relevant.
The spring 2015 to fall 2015 retention rate for the comparison group (i.e., students who did not
receive financial relief treatment) was 91.0%. The financial relief treatment was estimated to be
positively associated with spring-to-fall retention for all three empirical specifications. The estimate
from the PSM specification indicates that students who received the financial relief treatment were 4.2
percentage points (p < .01) more likely to be retained (i.e., their likelihood of retention increased from
91.0% to 95.2%). Across our three specifications, the financial relief treatment effect estimate was
robust, ranging from 3.7 (OLS) to 5.3 (NNM) percentage points.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
The financial relief treatment was also positively associated with the likelihood of graduating
with a bachelor’s degree from the University of Missouri. The graduation rate for the comparison group
(i.e., students who did not receive the financial relief treatment) was 82.6%. The PSM treatment effect
estimate indicates that receipt of the financial relief treatment increased a student’s likelihood of
graduation by 5.3 percentage points (p < .01), to an estimated graduation rate of 87.9%. The treatment
effect on graduation ranged across our empirical specifications from 2.9 (OLS) to 5.8 (NNM) percentage
points.
For the final three dependent variables, the analytical sample was reduced to students who
were retained (and thus enrolled) during the fall 2015 semester (n = 8,814). The fall 2015 credit hours
attempted for untreated students was 13.2 credit hours. The estimate from our preferred specification
(PSM) was positive (0.21) but not statistically significant (p > .1) indicating that the financial relief
treatment was not associated with credit hours attempted in the subsequent semester. The other two
specifications produced positive (0.27 for OLS and 0.34 for NNM) and statistically significant (p < .05)
effects of the financial relief treatment on credit hour attempted. However, the practical significance of
these effects estimates was small, as they indicate a change in credit hours attempted of less than a
third of a credit hour.
The financial relief program was not found to be practically (i.e., point estimates are very small)
or statistically (p > .1) associated with a student’s GPA in the following semester for all three
specifications. The comparison group’s mean fall 2015 semester GPA was 3.04 and the treatment effect
point estimates ranged from -0.03 (OLS) to 0.03 (NNM).
The financial relief treatment was positively associated with the likelihood a student receives a
registration hold due to a past due balance of at least 60 days in the fall 2015 semester. For context, the
registration hold rate for students who did not receive treatment was 20%. The PSM treatment effect
estimate indicates that receipt of the financial relief treatment increased a student’s likelihood of a fall
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
2015 registration hold by 4.7 percentage points (p < .05), or an estimated registration hold rate of
24.7%. The treatment effect on registration holds rates ranged across our empirical specifications from
3.8 (OLS) to 5.1 (NNM) percentage points.
Heterogeneity in Treatment Effect Estimates by Demographic Factors
Tables 4 and 5 present the results with respect to our second research question, was the effect
different across demographic characteristics, such as race/ethnicity, sex, first-generation status, or year
in school? Results of heterogeneous treatment effects are presented for all five dependent variables for
when samples were separated by demographic factor. For brevity, the results are only presented for the
propensity score matching estimation strategy.
Race/Ethnicity
Table 4, panel A, presents the propensity score estimates of the treatment effect for students of
color and white student separately. Point estimates indicate that the financial relief program had a
smaller association with retention for students of color students (3.7 percentage points, p < .1)
compared to white students (4.8 percentage points, p < .01). However, the 95% confidence intervals for
each coefficient overlapped indicating that these estimates may not be statistically significantly
different. Similarly, the estimated effect sizes on bachelor’s degree receipt were larger for white
students (5.2 percentage points, p < .05) when compared to students of color students (3.8 percentage
points, p > .1). Combined, the heterogeneity in treatment effects would indicate that the financial relief
program was beneficial for both students of color and white students with respect to retention and
graduation. However, the point estimates would indicate that the program may exacerbate already
existing gaps in retention and graduation between these groups.
Consistent with the results from the full sample, the financial relief program’s estimated effects
on academic outcomes were either practically small or statistically insignificant when separated by
race/ethnicity. The financial relief program had a larger association in increasing credit hours attempted
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
for white students (0.39, p < .05) than students of color (0.12 hours, p > .1), although the effect sizes
remain practically insignificant given their small sizes (i.e., less than a third of a credit hour for white
students). There were no statistically significant estimates with respect to fall 2015 GPA for either
students of color or white students.
The financial relief program was positively related to registration holds for students of color
students (11.0 percentage points, p < .05) but not for white students (2.3 percentage points, p > .1).
Sex/Gender
The financial relief treatment was estimated to increase retention (5.1 percentage points, p <
.01) and bachelor’s degree recipients (3.5 percentage points, p < .05) for female students. However, for
male students the financial relief treatment was not associated with changes in retention (p > .1) or
graduation (p > .1). The financial relief treatment was estimated to increase credit hours attempted for
male students (0.54 hours, p < .05) while not significantly increasing credit hours attempted for female
students (p > .1). There were no statistically significant estimates with respect to fall 2015 GPA for either
female or male students. A positive association with fall 2015 registration holds was found for male
students (10.0 percentage points, p < .05) but not female students (p > .1).
First-Generation
The positive effect of the financial relief program on retention was the largest for first-
generation students (7.6 percentage points, p < .01), almost triple the size when compared to
continuing-generation students (2.4 percentage points, p < .1). In fact, the treatment was associated
with a complete elimination of the retention gap for first-generation students. For untreated students,
first-generation students had a spring 2015 to fall 2015 retention rate of 88.7% compared to a 92.5% for
continuing-generation students. With the financial relief treatment, the retention rates were 96.3% for
first-generation students and 94.9% for continuing-generation students. The treatment effect of the
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
financial relief program on graduation rates was similar for first-generation students (6.0 percentage
point, p < .1) and continuing-generation students (5.8 percentage points, p < 0.1).
The financial relief treatment was estimated to increase credit hours attempted continuing-
generation students (0.46 hours, p < .05) while not significantly increasing credit hours attempted first-
generation students (0.42, p > .1). There were no statistically significant estimates with respect to fall
2015 GPA for either first-generation or continuing-generation students. A positive association with fall
2015 registration holds was found for first-generation students (10.7 percentage points, p < .05) but not
continuing-generation students (p > .1).
Year in School
Table 5 presents the results for the propensity score matching specification separately by the
year of enrollment at the University of Missouri. The estimates of the effect of the financial relief
program on retention were found to be largest for fourth-year students (i.e., those that had finished
their fourth year and were transitioning into their fifth year) with an estimated increase in retention of
7.3 percentage points (p < .1), followed by second-year students (6.1 percentage points, p < .1), first
year students (5.8 percentage points, p < .01), and third-year students (3.4 percentage points, p < .05).
The estimates of the financial relief treatment on graduation were similar across academic years,
ranging from 3.6 percentage points for second-year students to 4.8 percentage points for third-year
students, although the points estimates were only statistically significant for third-year students (p < .01)
and fourth-year students (p < .1).
The estimated financial relief treatment effects on fall 2015 credit hours attempted across
academic year show quite a bit of variation. The treatment was associated with increased credit hours
for first-year students (0.5 credit hours, p < .01) and third-year students (0.9 credit hours, p < .01). For
third-year students this translates into almost a third of a typical 3-credit course. Alternatively, receipt of
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
the financial relief treatment was associated with a 1.2 credit hour reduction for fourth-year students.
The variance in treatment effects across academic years is puzzling.
Consistent with the previous findings, the financial relief treatment was not found to be related
to GPA for any of the academic year subgroups. Interestingly, there were no statistically significant
treatment effects on fall 2015 registration holds for any academic year subgroups.
Heterogeneity in Treatment Effects by Treatment Amounts
Given the relative robustness of the findings regarding the treatment effect for retention, we
wanted to explore whether the treatment effect varied by the dollar amount. To estimate whether the
treatment effect of the financial relief program varied by the size of the financial relief award we
estimated dose response functions (Bia & Mattei, 2008; Hirano & Imbens, 2005). The dose response
function estimates the average treatment effect for a particular treatment level (i.e., ATE(t)) within a
generalized propensity score framework. We estimated the dose response function using the same set
of control variables as our PSM models using the ctreatreg function in Stata 17 (Bia & Mattei, 2008).
For retention, the dose response function approach estimated an average treatment effect of
0.51 (p < .01), which falls within the range of the PSM (0.42) and NNM (0.53). Figure 3 presents the dose
response function, in which the solid line can be interpreted as the average treatment effect for a given
treatment level and the dashed lines as the 95% confidence interval of those estimates. The dose
response function would indicate the average treatment effect of the financial relief program on
retention was near zero for treatment levels below about $800. At $800 we see an increasing average
treatment effect that peaks at about $2,600 and starts to decline at about $3,200. This would suggest
that the financial relief program was most successful at influencing retention for students that received
between $800 and $3,200 in their relief award. However, the 95% confidence intervals include zero for
all values treatment likely due to the small sample size of the treated group. Thus, caution should be
used when interpreting the relative effectiveness of the financial relief award across treatment values.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Limitations
One limitation of the study is the limited scope of a unique, one-time program at one institution.
The program is relatively low-cost and easy to implement, which institutions might find beneficial. Still,
additional research is needed to determine whether these results mirror other large-scale financial relief
programs or are replicable on other campuses.
A second limitation is that we were not involved in the program's administration and do not
know if students were notified of the financial relief awards or what those notifications may have said,
which may have influenced how the students felt about the awards. Without qualitative data, we are
unable to speak to their reactions or states of mind about the relief awards or how they perceived their
financial situations in general.
A third limitation is that in the semester after this financial relief funds were dispersed, the
University of Missouri experienced protests against racism and inequitable resources for students of
color, particularly Black and African American students, resulting in national attention and major
administrative leadership changes (American Council on Education, 2018). If the events of fall 2015
differentially affected students who did and did not receive the treatment, this would be a concern with
our design. To minimize some of these concerns, we also present estimates separately by race to
account for the fact that students of different races may have experienced the events in fall 2015
differently.
Finally, as with any quasi-experimental research methodology, we rely upon the assumptions of
the model to allow causal interpretations of the treatment effect estimates. In our case,
unconfoundedness assumption is the key assumption to whether our estimates are internally valid. Our
approach relies upon the richness of the administrative data utilized in this study, including detailed
academic and financial records to satisfy the requirement that sufficient observed factors can remove
selection on the unobservable factors.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Discussion
College students are increasingly unable to cover necessary expenses (Grodsky & Jones, 2007;
Page & Scott-Clayton, 2016), and 70% of college students report feeling stressed about finances
(Heckman et al., 2014). In addition to the expected expenses (e.g., tuition, fees, textbooks), unexpected
and emergency costs can be devastating to students’ progress (Evans et al., 2020), sometimes leading to
college departure (Krueger et al., 2016; Lowman et al., 2021). Students of color and low-income college
students experience additional barriers related to financial needs and support (O’Connor et al., 2010;
Ramirez-Mendoza & Jones, 2020), including higher incidences of past due financial balances that lead to
registration holds and other sanctions (Curs et al., 2022). Some institutions have tried to assist students
by providing emergency relief with emergency federal (Goldrick-Rab et al., 2021) or institutional funds
(Krueger et al., 2016), but little empirical research exists to determine whether such efforts are effective
in preventing future financial crises for students or improving outcomes such as time to degree
(Anderson et al., 2020; Lowman et al., 2021), and specifically for students of color, low-income, and first-
generation college students.
Despite administrator perceptions of the benefits of emergency funding for students (Taylor &
Melidona, 2021), previous research has mainly found that emergency relief does not lead to
improvements in degree completion (Anderson et al., 2020; Evans et al., 2019, 2020; Goldrick-Rab et al.,
2022) or persistence into the following semester (Lowman et al., 2021). Although students’ perceptions
of such support can be positive and seen as an indication of institutional care and a boost toward
graduation (Goldrick-Rab et al., 2021). Our results revealed that a surprise financial relief program that
reduced outstanding balances for Pell eligible students was positively related to students’ retention and
graduation rates. This contributes to the growing research on whether institutional relief funds (Krueger
et al., 2016) have any effect on student outcomes (Anderson et al., 2020; Lowman et al., 2021).
Improving retention and graduation rates are longstanding concerns within higher education and these
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
results offer some promise about being able to improve those outcomes at a relatively small cost. At the
same time, waiving past due balances is helping students return to the university and persist, but it is
not impacting their ability to afford college and avoid financial holds in future semesters.
A concerning finding related to receiving the emergency funding was that students were more
likely to receive a financial hold in a subsequent semester and that effect was stronger among students
of color (Black/African American, Hispanic/Latine, Asian, Multiracial, American Indian/Alaska Native),
males, and first-generation college students. It is unclear why the emergency aid led to greater financial
hardship in a subsequent semester. Future research is needed to determine whether students
developed an expectation that future balances would be similarly waived or perhaps they overestimated
the benefit of the award and were less careful with their spending, leading to financial challenges in a
subsequent semester.
Financial concerns tend to be higher among students of color (O’Connor et al., 2010; Ramirez-
Mendoza & Jones, 2020) and low-income college students (Goldrick-Rab et al., 2009; Long, 2008),
particularly post-pandemic (Amour, 2020; Harper, 2020). The results of this study offer evidence that
the relief intervention had a larger retention and graduation effect among white students when
compared to students of color, although both groups benefitted. All the students in the study were low-
income, so these results offer evidence of an effective intervention for students who tend to have the
highest unmet financial need (Walizer, 2018; Lopez, 2013). The intervention was also more effective on
retention and graduation among first-generation college students, with increases of 7.6% in the
retention rate to fall semester and 6% in the graduation rate for this group. These results are promising
given the high attrition rates among first-generation college students (Ishitani, 2003) and their unique
college experiences (Garriott et al., 2023). Nationally, the bachelor’s degree attainment rate for first-
generation college students is about half that of continuing-generation college students (20% vs. 42%,
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
respectively), with cost cited as the biggest factor preventing degree completion (Redford & Hoyer,
2017).
Our findings had a negligible relationship to subsequent term academic outcomes, including
GPA and credit hours attempted. This is consistent with past findings that also found that emergency
institutional relief did not improve GPA or substantially change the number of credits earned (Evans et
al., 2017, 2019). These results are contrary to past research on merit aid funding of similar amounts
(Curs & Harper, 2012), where students would be notified in advance of their awards and would have an
incentive to maintain a certain GPA to retain the award in subsequent years. Since emergency relief
funding is a surprise that clears a past due balance and holds no promise of future relief, it might make
sense that students’ GPAs were unaffected.
Implications for Research and Practice
This financial relief program was effective in improving the retention rates of the full sample of
low-income students and was most effective for first-generation college students. The program resulted
in a temporary elimination of the retention gap, particularly among first-generation college students.
The ease of the program from the students’ perspectives, given that they did not have to apply for
funding—which low-income students are less likely to do (Goldrick-Rab et al., 2009; Long, 2008)—might
be something to consider maintaining in future iterations of the program. Staff may have also found this
to be helpful in that applications did not need to be collected or reviewed. Finding ways to continue the
program through advancement or other institutional funds would be helpful while also determining
whether similar, relatively small, amounts of continued support would help address the graduation gap
that persists long-term and is left unaddressed by this short-term program. Donors and administrators
deciding how to apply similar funds might be moved by the effectiveness that $214,308 had for nearly
800 students. As a reminder, the average waived balance was only $269 and ranged from $10 to $4,000.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Further research is needed to determine whether there is an ideal range of balance waivers that
institutions should consider to maximize the effectiveness of these funds.
Returning to our theory of scarcity and the hope that the program would have allowed students
to worry less about finances and improve their academic attention and success (Mullainathan & Shafir,
2013), the results did not find such a connection. Since the program had no meaningful effect on grades
or credits earned, this program should not be relied upon for academic outcomes. It may be that the
funds were not substantial enough for the students to feel a sense of relief about their finances. Pairing
this intervention with other forms of support might be another strategy to address the increased risk of
financial trouble these students experienced in a subsequent semester. Related to that, further research
is needed to determine why this financial relief led students to be more likely to run into financial
difficulties. Qualitative research would help discover students’ thought processes and reactions to
receiving this support and how their behaviors or decisions changed after receiving the funds.
Understanding this from students would help determine what kinds of further support would be helpful
(e.g., financial literacy programs, budgeting support, tuition payment plans).
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
References
American Council on Education. (2018). Speaking truth and acting with integrity: Confronting challenges
of campus racial climate. https://pullias.usc.edu/wp-content/uploads/ 2023/05/Speaking-Truth-
and-Acting-with-Integrity-link-to-download-from-ACE.pdf
Amour, M. S. (2020, December 10). Low-income community college students most likely to cancel plans.
Inside Higher Ed. https://www.insidehighered.com/news/2020/12/10/ analysis-low-income-
community-college-students-most-likely-report-canceling-college
Anderson, D. M., Broton, K. M., Goldrick-Rab, S., Kelchen, R. (2020). Experimental evidence on the
impacts of need-based financial aid: Longitudinal assessment of the Wisconsin Scholars Grant.
Journal of Policy Analysis and Management, 39(3), 720-739. https://doi.org/10.1002/pam.22190
Baker, A. R., & Montalto, C. P. (2019). Student loan debt and financial stress: Implications for academic
performance. Journal of College Student Development, 60(1), 115-120.
https://doi.org/10.1353/csd.2019.0008
Bia, M., & Mattei, A. (2008). A Stata package for the estimation of the dose-response function through
adjustment for the generalized propensity score. The Stata Journal, 8(3), 354-373.
https://doi.org/10.1177/1536867X0800800303
Blundell, R., Dearden, L., & Sianesi, B. (2005). Evaluating the effect of education on earnings: Models,
methods and results from the national child development survey. Journal of the Royal Statistical
Society: Series A (Statistics in Society), 168(3), 473-512. https://doi.org/10.1111/j.1467-
985X.2004.00360.x
Caliendo, M., & Kopeinig, S. (2008). Some practical guidance for the implementation of propensity score
matching. Journal of Economic Surveys, 22(1), 31-72. https://doi.org/10.1111/j.1467-
6419.2007.00527.x
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Chen, R., & DesJardins, S. L. (2010). Investigating the impact of financial aid on student dropout risks:
Racial and ethnic differences. The Journal of Higher Education, 81(2), 179–208.
https://doi.org/10.1080/00221546.2010.11779048
Curs, B., & Harper, C. (2012). Financial aid and first-year collegiate GPA: A regression-discontinuity
approach. Review of Higher Education, 35(4), 627-649. https://doi.org/10.1353/rhe.2012.0040
Curs, B. R., Harper, C. E., & Kumbal, J. (2022). Institutional inequities in the prevalence of registration
sanctions at a flagship public university. Journal of Diversity in Higher Education. Advance online
publication. https://doi.org/10.1037/dhe0000432
Dahl, S., Strayhorn, T., Reid, M. Jr, Coca, V., & Goldrick-Rab, S. (2022, January). Basic needs insecurity at
Historically Black Colleges and Universities: A #RealCollegeHBCU report. The Hope Center for
College, Community, and Justice and the Center for the Study of HBCUs.
https://files.eric.ed.gov/fulltext/ED620138.pdf
El Masri, A. & Sabzalieva, E. (2020). Dealing with disruption, rethinking recovery: Policy responses to the
COVID-19 pandemic in higher education. Policy Design and Practice, 3(3), 312-333.
https://doi.org/10.1080/25741292.2020.1813359
Engle, J., & Tinto, V. (2008). Moving beyond access: College success for low-income, first-generation
students. Pell Institute for the Study of Opportunity in Higher Education.
https://eric.ed.gov/?id=ED504448
Evans, R., Stansberry, D., Bullington, K., & Burnett, D. (2020). First in college: A qualitative exploration of
experiences of first-generation students. Inquiry: The Journal of the Virginia Community
Colleges, 23(1), 1-32. https://commons.vccs.edu/inquiry/vol23/iss1/5
Evans, W. N., Kearney, M. S., Perry, B. C., & Sullivan, J. X., (2019). A cautionary tale about emergency
financial assistance without services. AEA Papers and Proceedings, 2019(109), 218-222.
https://doi.org/10.1257/pandp.20191014
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Evans, W. N., Kearney, M. S., Perry, B., & Sullivan, J. X. (2020). Increasing community college completion
rates among low‐income students: Evidence from a randomized controlled trial evaluation of a
case‐management intervention. Journal of Policy Analysis and Management, 39(4), 930-965.
https://doi.org/10.1002/pam.22256
Garriott, P., Ko, S.-J., Grant, S. B., Jessen, M., & Allan, B. (2023). When race and class collide: Classism
and social-emotional experiences of first-generation college students. Journal of College Student
Retention: Research, Theory, and Practice, 25(3), 509-532.
https://doi.org/10.1177/1521025121995483T
Goldrick-Rab, S. (2016). Paying the price: College costs, financial aid, and the betrayal of the American
dream. The University of Chicago Press.
Goldrick-Rab, S., Baker-Smith, C., York, T., Clark, K., Webber, D., & Perkins, C. (2022, December).
Affording degree completion: An experimental study of completion grants at accessible public
universities. The Hope Center for College, Community, and Justice.
https://hope.temple.edu/sites/hope/files/media/document/version%204%2C121222.pdf
Goldrick-Rab, S., Harris, D. N., & Trostel, P. A. (2009). Why financial aid matters (or does not) for college
success: Toward a new interdisciplinary perspective. In J. C. Smart (Ed.), Higher education:
Handbook of theory and research (pp. 1–45). Springer.
https://doi.org/10.1007/978-1-4020-9628-0_1
Goldrick-Rab, S., Hacker, N. L., Kienzl, G., Price, D. V., & Curtis, D. (2021, October). When care isn’t
enough: Scaling emergency aid during the pandemic. The Hope Center for College, Community,
and Justice. https://hope.temple.edu/sites/hope/files/media
/document/EA_10_26_2021_RA_final.pdf
Grodsky, E., & Jones, M. T. (2007). Real and imagined barriers to college entry: Perceptions of cost.
Social Science Research, 36(2), 745–766. https://doi.org/10.1016/j.ssresearch.2006.05.001
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Harper, S. R. (2020). COVID-19 and the racial equity implications of reopening college and university
campuses. American Journal of Education, 127(1), 153–162. https://doi.org/10.1086/711095
Heckman, S., & Grable, J. (2011). Testing the role of parental debt attitudes, student income,
dependency status, and financial knowledge have in shaping financial self-efficacy among
college students. College Student Journal, 45(1), 51–64.
Heckman, S., Lim, H., & Monatalto, C. (2014). Factors related to financial stress among college students.
Journal of Financial Therapy, 5(1), 19-39. https://doi.org/10.4148/1944- 9771.1063.
Hirano, K., & Imbens, G. W. (2005). The propensity score with continuous treatments. In Applied
Bayesian Modeling and Causal Inference from Incomplete-Data Perspectives: An Essential
Journey with Donald Rubin's Statistical Family (pp. 73-84). Wiley-Blackwell.
https://doi.org/10.1002/0470090456.ch7
Imbens, G. W. (2015). Matching methods in practice: Three examples. Journal of Human Resources,
50(2), 373-419. https://doi.org/10.3368/jhr.50.2.373
Ishitani, T. T. (2003). A longitudinal approach to assessing attrition behavior among first-generation
students: Time-varying effects of pre-college characteristics. Research in Higher Education,
44(4), 433-449. https://doi.org/10.1023/A:1024284932709
Joo, S.-H., Durband, D., & Grable, J. (2008). The academic impact of financial stress on college students.
Journal of College Student Retention, 10(3), 287-305. https://doi.org/10.2190/CS.10.3.c
Krueger, K., Parnell, A., & Wesaw, A. (2016). Landscape analysis of emergency aid programs. National
Association of Student Personnel Administrators. https://www.naspa.org/ report/landscape-
analysis-of-emergency-aid-programs
Long, B. T. (2008). What is known about the impact of financial aid? Implications for policy. National
Center for Postsecondary Research. https://eric.ed.gov/?id=ED501555
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Lopez, J. I. (2013). Unmet need and unclaimed aid: Increasing access to financial aid for community
college students. New Directions for Community Colleges, 2013(164), 67-74.
https://doi.org/10.1002/cc.20082
Lowman, J., Choroszy, M. N., Beattie, J., & Gonzales, L. (2021). The academic impact of CARES Act
funding and peer mentoring on first generation students at the University of Nevada, Reno.
Journal of Higher Education Management, 36(1), 117–127.
McKinnon-Crowley, S. (2023). Financial aid access during the 2020 pandemic period. Journal of Student
Affairs Research and Practice, 60(1), 17-30. https://doi.org/10.1080/19496591.2022.2096407
Mitchell, M., Leachman, M., & Saenz, M. (2019). State higher education funding cuts have pushed costs
to students, worsened inequality. Center on Budget and Policy Priorities.
https://www.cbpp.org/research/state-budget-and-tax/state-higher-education-funding-cuts-
have-pushed-costs-to-students
Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too little means so much. Macmillan.
National Center for Education Statistics. (2020). The NCES Fast Facts. National Center for Education
Statistics. https://nces.ed.gov/fastfacts/display.asp?id=40
O’Connor, N., Hammack, F. M., & Scott, M. A. (2010). Social capital, financial knowledge, and Hispanic
student college choices. Research in Higher Education, 51(3), 195–219.
https://doi.org/10.1007/s11162-009-9153-8
Page, L. C., & Scott-Clayton, J. (2016). Improving college access in the United States: Barriers and policy
responses. Economics of Education Review, 51, 4–22. https://doi.org/
10.1016/j.econedurev.2016.02.009
Ramirez-Mendoza, J., & Jones, T. (2020). Using professional judgment in financial aid to advance racial
justice and equity. In Education Trust. https://eric.ed.gov/?id=ED610822
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Redford, J., & Hoyer, K. M. (2017). First-generation and continuing-generation college students: A
comparison of high school and postsecondary experiences (NCES 2018-009). U.S. Department of
Education, National Center for Education Statistics. https://nces.ed.gov/pubs2018/2018009.pdf
Rodríguez-Planas, N. (2022). Hitting where it hurts most: COVID-19 and low-income urban college
students. Economics of Education Review, 87, 1-18. https://doi.org/10.1016/j.
econedurev.2022.102233
Schmidt, J., & Weissman, E. (2021). Strengthening emergency aid programs: Lessons from the CARES Act
and the higher education emergency relief fund. NASFAA, NASPA, & MDRC.
https://www.mdrc.org/publication/strengthening-emergency-aid-programs
Shapiro, D., Dundar, A., Huie, D., Wakhungu, P. K., Yuan, X., Nathan, A., & Hwang, Y. (2017). A national
view of student attainment rates by race and ethnicity – fall 2010 cohort (Signature Report No.
12b). National Student Clearinghouse Research Center.
https://www.luminafoundation.org/files/resources/signature12-raceethnicity.pdf
StataCorp. (2021). Stata 17 Causal Inference and Treatment-Effects Estimation Reference Manual. Stata
Press. https://www.stata.com/bookstore/causal-inference-and-treatment-effects-reference-
manual/
Steiner, P. M., Cook, T. D., Shadish, W. R., & Clark, M. H. (2010). The importance of covariate selection
in controlling for selection bias in observational studies. Psychological Methods, 15(3), 250-267.
https://doi.org/10.1037/a0018719
Taylor, M., & Melidona, D. (2021, November 8). First look: Use of higher education emergency relief
funds (HEERF) at U.S. colleges and universities. American Council on Education.
https://www.acenet.edu/Research-Insights/Pages/Senior-Leaders/Presidents-Survey-
HEERF.aspx
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Walizer, L. (2018). When financial aid falls short: New data reveal students face thousands in unmet
need. Center for Law and Social Policy (CLASP).
https://www.clasp.org/sites/default/files/publications/2018/12/2018whenfinancialaidfallsshort.
pdf.
Zaloom, C. (2019). Indebted: How families make college work at any cost. Princeton University Press.
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Table 1
Descriptive Statistics of the Analytic Sample
Mean S. D.
Dependent Variables
Retained (Fall 2015) 0.91 0.28
Graduated (Ever) 0.83 0.38
Credit Hours (Fall 2015) a 13.23 3.06
GPA (Fall 2015) a 3.04 0.79
Registration Hold (Fall 2015) a 0.21 0.41
Treatment Variables
Treatment Receipt (0,1) 0.06 0.25
Treatment Amount ($) b 197.87 413.47
Covariates
Cumulative GPA 3.03 0.58
GPA – SP15 2.99 0.80
Hours Passed – SP15 12.99 3.34
Expected Family Contribution ($1,000s) – 14/15AY 8.85 8.35
Registration Hold (0,1) – SP15 0.35 0.48
Registration Hold Amount ($1,000s) – SP15 0.33 1.05
Registration Hold (0,1) – F14 0.21 0.41
Registration Hold Amount ($1,000s) – F14 0.37 1.16
Received Institutional Merit Aid (0,1) – 14/15AY 0.39 0.49
Received Institutional Needs-Based Aid (0,1) – 14/15AY 0.34 0.47
First-Year Student 0.38 0.49
Second-Year Student 0.29 0.45
Third-Year Student 0.24 0.43
Fourth-Year Student 0.09 0.28
N 9,653
a
Calculated based upon the sample of students that were retained (n = 8,814).
b Calculated based upon the sample of students that received the treatment (n = 624).
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Table 2
Propensity Score Model Estimates
Full Trimmed
Sample Sample
Cumulative GPA 0.675*** 0.686***
(0.135) (0.135)
GPA – SP15 0.016 0.016
(0.104) (0.104)
Hours Passed – SP15 0.032** 0.032**
(0.016) (0.016)
Expected Family Contribution ($1,000s) – 14/15AY -0.071*** -0.072***
(0.007) (0.007)
Registration Hold (0,1) – SP15 1.189*** 1.193***
(0.098) (0.098)
Registration Hold Amount ($1,000s) – SP15 0.006 0.005
(0.039) (0.039)
Registration Hold (0,1) – F14 0.449*** 0.461***
(0.120) (0.120)
Registration Hold Amount ($1,000s) – F14 -0.023 -0.024
(0.040) (0.040)
Received Institutional Merit Aid (0,1) – 14/15AY 0.119 0.110
(0.092) (0.092)
Received Needs-Based Aid (0,1) – 14/15AY -0.099 -0.104
(0.093) (0.093)
Second Year Student -0.392*** -0.395***
(0.103) (0.104)
Third Year Student -0.757*** -0.758***
(0.119) (0.120)
Fourth Year Student -0.438*** -0.441***
(0.161) (0.161)
N 9,814 9,653
Note. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Table 3
Treatment Effects Estimates of the Financial Relief Award by Outcome
Retained Graduated Credit GPA Registration
(Fall 2015) (Ever) Hours (Fall 2015) Hold
(Fall 2015) (Fall 2015)
Propensity 0.042*** 0.053*** 0.205 -0.024 0.047**
Score Matching (0.014) (0.020) (0.137) (0.037) (0.020)
OLS 0.037*** 0.029** 0.274** -0.029 0.038**
w/ covariates (0.011) (0.014) (0.121) (0.026) (0.016)
Nearest Neighbor 0.053*** 0.058*** 0.335** 0.028 0.051**
Matching (0.008) (0.014) (0.141) (0.031) (0.023)
Outcome Mean 0.91 0.83 13.22 3.04 0.20
Of Untreated
N 9,653 9,653 8,814 8,814 8,814
Note. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Table 4
Heterogeneity of Propensity Score Matching Treatment Effects by Selected Demographics
Retained Graduated Credit GPA Registration
(Fall 2015) (Ever) Hours (Fall 2015) Hold
(Fall 2015) (Fall 2015)
Panel A: Differences across Race and Ethnicity
Students of Color 0.037* 0.038 0.116 0.082 0.110**
(0.019) (0.036) (0.227) (0.068) (0.047)
Untreated Mean 0.907 0.797 13.07 2.841 0.267
N 2,380 2,380 2,171 2,171 2,171
White Students 0.048*** 0.052** 0.387** -0.023 0.023
(0.011) (0.022) (0.156) (0.043) (0.026)
Untreated Mean 0.914 0.839 13.27 3.109 0.181
N 7,111 7,111 6,515 6,515 6,515
Panel B: Differences across Sex
Female Students 0.051*** 0.035** -0.049 -0.084 0.021
(0.011) (0.018) (0.176) (0.053) (0.030)
Untreated Mean 0.917 0.851 13.55 3.176 0.189
N 5,099 5,099 4,692 4,692 4,692
Male 0.026 0.009 0.537** -0.005 0.100**
Students (0.025) (0.028) (0.243) (0.062) (0.039)
Untreated Mean 0.905 0.801 12.85 2.894 0.220
N 4,313 4,313 3,913 3,913 3,913
Panel C: Difference across First Generation Status
First Generation 0.076*** 0.060* 0.418 -0.012 0.107**
Students (0.015) (0.033) (0.275) (0.060) (0.051)
Untreated Mean 0.887 0.787 13.08 2.953 0.211
N 3,755 3,755 3,349 3,349 3,349
Continuing Gen. 0.024* 0.058*** 0.458** -0.031 0.000
Students (0.014) (0.019) (0.194) (0.052) (0.027)
Untreated Mean 0.925 0.851 13.30 3.091 0.196
N 5,900 5,900 5,470 5,470 5,470
Note. All estimates were estimated on subsamples using the propensity score matching estimation
strategy. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Table 5
Heterogeneity of Propensity Score Matching Treatment Effects by College Year
Retained Graduated Credit GPA Registration
(Fall 2015) (Ever) Hours (Fall 2015) Hold
(Fall 2015) (Fall 2015)
First-Year 0.058*** 0.047 0.531*** 0.030 0.009
Students (0.015) (0.029) (0.155) (0.051) (0.030)
Untreated Mean 0.905 0.758 13.26 3.013 0.183
N 3,663 3,663 3,333 3,333 3,333
Second-Year 0.061* 0.036 -0.424 -0.048 0.018
Students (0.032) (0.059) (0.261) (0.086) (0.054)
Untreated Mean 0.917 0.847 13.41 3.043 0.195
N 2,711 2,711 2,495 2,495 2,495
Third-Year 0.034** 0.048*** 0.914*** -0.022 0.027
Students (0.014) (0.017) (0.299) (0.086) (0.069)
Untreated Mean 0.937 0.902 13.41 3.169 0.211
N 2,161 2,161 2,028 2,028 2,028
Fourth-Year 0.073* 0.047* -1.202*** -0.072 0.170
Students (0.043) (0.025) (0.407) (0.245) (0.122)
Untreated Mean 0.857 0.866 11.84 2.826 0.317
N 724 724 619 619 619
Note. All estimates were estimated on subsamples using the propensity score matching estimation
strategy. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Figure 1
Overlap of Propensity Scores
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Figure 2
Standardized Differences for Selected Covariates
FINANCIAL RELIEF FUNDING AND STUDENT OUTCOMES
Figure 3
Dose Response Function of Treatment Amount on Retention
File and source
- File
- the-role-of-emergency-financial-relief-funding-edworkingpaper-24-991.pdf
- Size
- 934,155 bytes
- SHA-256
- 4268262da7173622baddbdca54d2e4ed5ab6573dd6753daa5a54f64d9c2dc843
- Original
- edworkingpapers.com