How Did Colleges Disburse Emergency Aid During COVID-19? An Implementation Analysis of the Higher Education Emergency Relief Fund

Abstract

To address the unprecedented challenges of the COVID-19 pandemic, Congress authorized the Higher Education Emergency Relief Fund (HEERF I) in March 2020 with over $6 billion allocated for emergency financial aid. In this paper, we utilize the administrative burden framework to analyze HEERF I implementation for a stratified random sample of colleges, focusing on the implications for equity. We find that disbursement policies varied along two dimensions: (1) whether they imposed burdens on students by requiring applications and proof of hardship and (2) whether they targeted needy students and varied the amount of aid according to need. When we examine sectoral differences, we find that private for-profit colleges were more likely to place higher burden on students, whereas public and minority-serving institutions were more likely to reduce burden.

In March 2020, Congress passed the Coronavirus Aid, Relief, and Economic Security Act (CARES Act), which authorized the Higher Education Emergency Relief Fund (HEERF I). The broad goal of HEERF I was to support institutions of higher education (IHEs) and students during the COVID-19 pandemic to facilitate continuity of education. HEERF I authorized the Department of Education (ED) to make available over $6 billion in direct support to students in the form of emergency grants. ¹ Although the legislative goal to support students and IHEs was clear, the fluctuating and contradictory administrative guidance left IHEs to determine how they would disburse the aid to students.

Beginning in late March 2020, IHEs throughout the United States rapidly developed systems and procedures to distribute the HEERF I aid that they received “to provide emergency financial aid grants to students for expenses related to the disruption of campus operations due to the coronavirus [pandemic]” (Department of Education, 2020a). When deciding how to distribute this aid, IHEs confronted several competing pressures: How quickly should they distribute aid, and how much should each student receive? Should they require students to complete an application and provide proof of hardship so that they can target aid to those who need help the most? Should only “needy” students receive aid, or should all students be eligible for aid given that all students were affected by COVID-19?

Each of these decisions created important variation across IHEs that could be consequential for student outcomes. Although these decentralized implementation processes are not uncommon in higher education, they are often left unexplored by scholars in education policy who historically have treated implementation as a “black box” (Century & Cassata, 2016). However, in line with the growing focus on administrative burden in the implementation of college financial aid (Bell, 2020; Deming & Dynarski, 2009; Dynarski & Scott-Clayton, 2008; Page et al., 2020; Rosinger et al., 2021; Scott-Clayton & Schudde, 2016), we argue that some of the most crucial decisions for equity in access to aid are made in the process of decentralized implementation. Building on the existing reports that demonstrate the importance of implementation in the disbursement of previous smaller-scale emergency aid programs (Evans et al., 2019; Goldrick-Rab & Cady, 2017; NASPA, 2020; Reos Partners, 2018), we investigate how the largest at-scale emergency aid program (HEERF I) was disbursed to students with a particular focus on equity and administrative burden. We leverage the definition of equity established by the National Association for Public Administration: “the fair and equitable distribution of public services, and implementation of public policy, and the commitment to promote fairness, justice, and equity in the formation of public policy” (Svara & Brunet, 2005, p. 256). In line with this definition, we conceptualize equitable disbursement policies as those that allocate aid to needy students without imposing undue administrative burdens. Although some documentation may be helpful in targeting aid to students who need help the most, disbursement policies that impose administrative burdens have also been shown to disproportionately reduce access to aid for disadvantaged students that would run counter to equity goals (e.g., Dynarski & Scott-Clayton, 2008; Page et al., 2020; Rosinger et al., 2021; Scott-Clayton & Schudde, 2016). Because disadvantaged students may be less likely to know about the program and understand the program requirements and less able to cope with compliance demands, we focus on the number of hurdles in the process of applying for HEERF funds to shed light on the potential for inequitable distributions that shut out students who need help the most.

To examine how IHEs disbursed the student portion of HEERF I funds, we conducted an analysis of HEERF I implementation across a stratified random sample of institutions by sector, maximum length of study, highest degree offered (e.g., associate, bachelor’s degree, master’s degree, or doctorate), and type of IHEs (e.g., special focus). This study was originally an effort by authors as members of the General Services Administration’s Office of Evaluation Sciences (OES) team to provide a preliminary overview of how IHEs were implementing the HEERF I program and proposed avenues to evaluate HEERF I for the U.S. Department of Education (ED). ² As part of this effort, we analyzed the implementation strategies across a stratified random sample of IHEs so that we could better characterize variation in disbursement policies by institutional sector and type. ³

In this analysis, we had two main research questions:

1. What are the different strategies for HEERF I aid disbursement, and how do these strategies align with best practices for equity?

2. How do the types of disbursement policies vary across different types of institutions?

In our investigation into the first question, we find that there were two main dimensions upon which institutional disbursement policies vary: the level of administrative burden ⁴ and the degree to which institutions targeted aid based on financial need. For example, although some institutions required students to apply and submit documentation proving financial hardship, increasing the level of administrative burden for students, other institutions adopted automatic disbursements that shifted burden away from students and onto institutions. We also find that while some IHEs provided funding to all students, other IHEs took a more targeted approach that considered whether the student was low income or disadvantaged prior to the pandemic either based on information from the Free Application for Federal Student Aid (FAFSA) or based on information collected in the HEERF I application processes. Considering the literature on the efficiency of targeted financial aid and the deleterious effects of administrative burden on student access and outcomes (Bell and Smith, 2022; Bell et al., 2021; Dynarski & Scott-Clayton, 2008; Dynarski et al., 2018; Heller & Marin, 2002), both of these dimensions likely have serious implications for equity in access to HEERF I funds.

For our second research question, we find that there was significant systematic variation across institutional sectors that could also have important implications for equity in access and outcomes. Specifically, private for-profit IHEs were more likely to require students to overcome a higher level of burden prior to accessing HEERF I funds. Given that private for-profit IHEs serve a disproportionate number of students of color and low-income students relative to public institutions, these burdens may have prevented students who were hardest hit by the pandemic from accessing the emergency relief funds (Espinosa et al., 2019). On the other hand, we find that public institutions and minority-serving institutions were less likely to impose burdens on students, which aligns with their public service mission and best practices for equity.

Although these findings provide significant insights into the implementation of the student portion of HEERF I, there are also important limitations to our data and sample that should be factored into the interpretation of the results. First, we only examine publicly reported information about HEERF I disbursement policies as of September and October 2020. Therefore, we capture certain, publicly posted information about HEERF I disbursement policies but are unable to examine the policies that were not released publicly or any differences between posted information and actual policies. Second, we are unable to document how disbursement policies may have changed over time, which is a fruitful area for future researchers to examine.

In the remaining sections, we begin by describing the details of the CARES Act and how HEERF I aid was designed to mitigate the negative impacts of COVID-19. Then, we present a theoretical framework that helps us conceptualize the different levels of burden across disbursement policies. We conclude by proposing a series of research questions for scholars to investigate in the future to generate more evidence on the implementation and effects of HEERF I.

Policy Background on HEERF I

A Window into How COVID-19 Impacted Students

The COVID-19 pandemic caused multiple disruptions that made it more difficult for students to continue their education and appears to have caused some students to shift focus away from their education and drop out of school. A survey of Arizona State University students showed that 69 percent of students report that they have been significantly negatively affected by the COVID-19 pandemic (Aucejo et al., 2020). Recent nationwide surveys indicate that 49% of students pursuing a bachelor’s degree and 56% of students pursuing an associate’s degree believe COVID-19 will negatively impact their likelihood of graduation (Marken, 2020). This could be due to expenses directly related to the cost of attendance such as tuition and fees, but it also is likely driven by basic needs insecurity, in which students experience food and housing insecurity and even homelessness (Goldrick-Rab et al., 2020; The Hope Center for College, Community, and Justice, 2021). Recent nationwide surveys estimate in 2020 among close to 200,000 students reveal over half of students reporting basic needs insecurity, with students at two-year colleges reporting higher basic needs insecurity (61%) than four-year college students (53%) (The Hope Center for College, Community, and Justice, 2021).

These statistics, along with the disproportionate drop in enrollment in community colleges in fall 2020 (Sedmak, 2020), make it clear that while many students have been negatively impacted by the COVID-19 pandemic, the negative effects appear to be concentrated in certain sectors of higher education that serve more disadvantaged students (Baker, 2020; Daugherty, 2020). Research shows that disadvantaged students have limited capacity to overcome financial shocks (Evans et al., 2019), which explains why the COVID-19 pandemic appears to have exacerbated existing inequality, hitting low-income students, students of color, and first-generation college students the hardest (Baker, 2020).

The CARES Act and HEERF I Aid

In the wake of the COVID-19 pandemic, lawmakers sought to support students facing hardship through the passage of the CARES Act, which authorized over $6 billion for the disbursement of HEERF I emergency aid directly to students. The goal of the student portion of HEERF I aid was to help cover expenses related to campus disruptions caused by the COVID-19 pandemic, including expenses related to the cost of attendance, such as food, housing, health care, course materials, technology, and childcare; however, there were multiple sets of guidance provided, sometimes conflicting, that may have increased variability in IHEs disbursement and documentation decisions. There were two factors that determined how much formula funding colleges were awarded: “(1) 75% of the funds were awarded to IHEs based on the share of full-time equivalent (FTE) enrollment of Pell grant recipients who were not enrolled exclusively in distance education prior to the coronavirus emergency, relative to the total FTE enrollment of such individuals in all IHEs, and (2) 25% of the funds were awarded to IHEs based on each IHE’s share of FTE enrollment of students who were not Pell rant recipients and who were not enrolled exclusively in distance education prior to the coronavirus emergency, relative to the total FTE enrollment of such individuals in all IHEs” (U.S. Department of Education, 2020). After the passage of the CARES Act legislation in March 2020, the first set of administrative guidance from the U.S. Department of Education (ED) was released in April 2020. The administrative guidance for HEERF I under Secretary Betsy Devos specified that students must meet Title IV eligibility ⁵ and clarified the allowable expenses for the student share of HEERF I funds. In May 2020, however, the U.S. Department of Education announced they would not enforce previous guidance that required Title IV eligibility for students. In June 2020, ED published a preview of an interim final rule which reinstated the position that only Title IV–eligible students could receive funding but clarified that it would not be retroactively enforcing the Title IV eligibility interpretation prior to the publication of the rule in the Federal Register. With this fluctuating administrative guidance, IHEs were left with unclear expectations from ED regarding how they should implement HEERF I, creating significant variation in institutional approaches to disbursement.

Institutional HEERF I funds became available in early April 2020, with institutions beginning disbursements to students shortly thereafter. HEERF I resulted in the disbursement of hundreds of thousands of one-time cash grants, typically between $250 and $750 per student in our sample of IHEs, to students struggling with the adverse effects of the COVID-19 pandemic. The idea behind HEERF I was that these emergency aid grants, if distributed in a timely and easily accessible manner, could help students overcome unexpected financial shocks and enable them to continue making academic progress toward a degree. However, the existing literature on emergency financial aid programs is scarce and focuses on smaller-scale programs mostly at single institutions, which leaves us with unclear predictions for how emergency aid may have impacted students in the COVID-19 pandemic.

Previous studies suggest that the design and implementation of emergency aid matters; in fact, scholars find that emergency aid may have no effects on students unless it is combined with nonfinancial “wrap-around services,” such as counseling, child care, and mentoring (Evans et al., 2020; Miller et al., 2019). A randomized control trial (RCT) on the impacts of emergency aid found that “Stay the Course” emergency grant aid—which served Tarrant County College students who were Pell eligible or fell below 200 percent of the poverty line and were enrolled in at least nine credit hours, had a 2.0 GPA, and accumulated less than 30 credit hours—did not significantly improve academic outcomes. The authors speculate this was because of the lack of nonfinancial support systems, but it also could be due to the burdens imposed on students in program implementation (Evans et al., 2019). In a follow-up study, the same authors found that when paired with personalized case management services, emergency aid increased credits earned and the likelihood of completing an associate’s degree for community college students (Evans et al., 2020). In the implementation of the “Stay the Course” program, students were required to demonstrate that they were in dire financial need and prove that the financial obstacle would hinder their ability to graduate, which imposes administrative burdens on a vulnerable subset of students. For instance, students may face burdens in the form of psychological costs including stigma and shame in having to demonstrate financial need or struggle to meet the compliance costs associated with documenting financial hardships (Bell & Smith, 2022). This case makes it clear that implementation is consequential to the effectiveness of emergency aid and that imposing burdens in program design and implementation as well as providing key levers for reducing burden—such as providing trained social workers to help navigate the application process—may be key to programmatic success (Bell & Smith, 2022). Although this preliminary evidence provides some insight into the effectiveness of emergency aid, more rigorous research is needed to better understand how and when emergency aid could impact student outcomes in a broader variety of institutional contexts. In this article, we set the stage for this future work by analyzing variation in HEERF I disbursement policies and the associated burdens these policies create for students and IHEs.

Theoretical Framework: Administrative Burden and Access to HEERF I Aid

We leverage the administrative burden framework to situate why variation in disbursement strategies may matter for student outcomes. ⁶ A growing body of research demonstrates that administrative burdens (i.e., learning, compliance, and psychological costs), otherwise known as sludge, ordeals, or friction costs, undermine access to government programs (Bhargava & Manoli, 2015; Christensen et al., 2020; Dynarski & Scott-Clayton, 2008; Heinrich, 2016; Herd & Moynihan, 2018; Sunstein, 2019; Sunstein & Gosset, 2020). This is because in order to access many public benefit programs students have to (1) learn about the program and correctly identify whether they are eligible; (2) put together confusing documentation requirements that require time, effort, and coordination with parents; and (3) overcome any stigma or frustration that occurs in the process of applying (Dynarski & Scott-Clayton, 2013).

In the context of financial aid programs, administrators must make a variety of choices when establishing a new program that may seem inconsequential at first but can significantly shape the allocation of benefits and burdens. Administrators can choose, for instance, whether to allocate a higher level of burden on themselves by taking on the responsibility for identifying eligible students and proactively reaching out to them. Alternatively, they can shift the burden onto students to prove they are eligible and take a passive approach to information dissemination, which may have important implications for students’ ability to access financial aid (Bell et al., 2020).

When designing HEERF I disbursement policies, administrators could make several choices that imposed different costs and burdens on students. Institutions had to weigh serious tradeoffs between efficiency and equity. For instance, institutions could shift burdens away from students by using existing administrative data or low-burden applications to identify and disburse emergency aid without requiring students to provide extensive documentation proving financial hardship. The FASFA can serve as a proxy for student financial need but relying on FASFA data comes with significant issues—it is a static point in time with incomplete data provided by EFC calculations (Goldrick-Rab, 2016). Some colleges like Compton College are utilizing innovative tools like Edquity to target financial need more effectively while still minimizing student burden, which leverages an algorithm to efficiently sort through student applications for emergency aid and determine which students are the highest priority for funding. Initial results from that pilot program show that students who received emergency grants were twice as likely to graduate, but more data is still needed to holistically assess the program tool (Anderson, 2021). However, this process takes time and administrative capacity that institutions serving predominantly disadvantaged students with the most needs often do not have.

Alternatively, IHEs could have required students to apply for funds (which imposes a burden) and provide proof that they incurred an eligible cost (another burden). These policies can serve multiple purposes, one of which is to target the aid to students who face the greatest hardship. Although this may seem equity enhancing, there may be unintended consequences for equity in access that undermine the intentions of these targeting policies. Moreover, these targeted policies may garner less political support, which could threaten the long-term viability of the program (Bell, 2020). Additionally, many institutions, especially in the private for-profit sector, may adopt these policies to prepare documentation in case they are subject to federal scrutiny or audits. However, by requiring students to complete an application, IHEs impose learning costs, which require students to seek out information on eligibility and correctly identify whether they qualify for the aid. Additionally, by requiring documentation of financial hardship caused by COVID-19, IHEs imposed compliance costs and psychological costs, especially among students who may feel ashamed of their financial situation and may not have access to a printer, scanner, or reliable internet. ⁷ Therefore, the attempt to target financial aid to the neediest students or the need for documentation to avoid lawsuits and comply with accountability pressures may create high burden environments for students that undermine equity in access.

This is important because if administrators opt for, or circumstances require them to impose, a higher level of burden or complexity by requiring students to fill out lengthy application forms and/or submit income documentation, then existing evidence suggests that less disadvantaged students will be able to access key financial aid that may be the difference between continuing their education and dropping out (Dynarski & Scott-Clayton, 2008, 2013; Kennedy et al., 2007; McKinney & Novak, 2013; Scott-Clayton & Schudde, 2016). To improve the process, existing emergency aid administrators and researchers suggest advertising, promoting affirming messages to students, and making sure there are low levels of burden on students in the application process as well as providing aid in a timely manner (Evans et al., 2019; Goldrick-Rab & Cady, 2017; Weissman & Schmidt, 2020). In addition, the large body of research on traditional financial aid programs demonstrates that the most effective financial aid programs have simplified these application processes with low barriers to access (Anderson, 2020; Clotfelter et al., 2018; Goldrick-Rab, 2016). This ensures that students who need help the most will be able to access the aid, which is not always the case for programs that impose significant burdens on students in application processes. In light of the recent literature on frictions, sludge, or administrative burdens (Baekgaard et al., 2021; Christensen et al., 2020; Herd & Moynihan, 2018; Sunstein, 2019), which demonstrates that hurdles in application processes reduce equity by restricting access for marginalized groups, we argue that better understanding these implementation decisions in the largest emergency aid program ever to be implemented in the United States (HEERF I) is both theoretically important for future evaluations of HEERF I and normatively important for the advancement of equity in crises like the COVID-19 pandemic.

Identifying Our Sample

In September 2020, we reviewed the public websites of 340 institutions for any publicly available information about how these institutions distributed HEERF I aid. To generate this sample, we began by randomly selecting 10 institutions from 26 categories defined by characteristics that are likely to influence the composition of HEERF I disbursement policies. Specifically, we construct a stratified sample by ownership status (private-non-profit, private-for-profit, public), maximum length of study (two-year, four-year, or four-year plus), degree type (majority certificate granting, associate degree granting, bachelor’s degree granting, etc.), and special focus schools (e.g., health-care focus schools, etc.). Because the CARES Act provided additional funds to minority-serving institutions (MSI), we also include a significant sample of MSI institutions (n = 80).

We stratify by sector because public institutions, private nonprofits, and private for-profit institutions differ in sources of funding, external reporting, and accountability requirements. Furthermore, we stratify by maximum length of study and degree type to account for institutional resources and enrollment types (e.g., a two-year certificate and AS-degree-granting institutions are primarily community colleges with a high percentage of part-time enrollment whereas four-year plus, doctoral-granting institutions are larger universities with potential financial and institutional resources and capacity).

To identify these institutions and their characteristics, we use data from the 2019 Integrated Postsecondary Data System (IPEDS), Institutional Characteristics (IC), Fall Enrollment (FE), Completers (C), and Student Financial Aid (SFA) and the Net Price (NP) surveys. Our final sample includes 289 institutions that provided information on their disbursement strategies on their website. Approximately 15 percent of our original sample (n = 51) did not have any information on their disbursement strategies online. ⁸ We encourage caution in extrapolating the results from our implementation analysis to all IHEs, and we conduct sensitivity analysis with inverse probability weights to enhance the representativeness of our sample in the analysis (i.e., these weights reweight our sample to better approximate the actual population of IHEs in the United States). ⁹

There are several potential reasons why an institution may choose not to post information about any HEERF disbursement policy. As seen in online appendix Table S1, we see that IHEs with smaller enrollments are significantly less likely to post information either compared to those IHEs in the surveyed sample or those in the universe of IHEs in the United States that responded to the IC, FE, C, SFA, and NP 2019 IPEDS surveys, excluding the IHEs whose policies we reviewed (i.e., the “in-sample” IHEs). The extent to which enrollment proxies for institutional capacity, it could be that these IHEs simply lack the capacity or the ability to regularly update their website to post any policies, and they chose to communicate any policies to students in some other fashion. It may also be that these smaller institutions may have decided how to allocate HEERF funds to students, and they did not find it necessary to communicate any policy or justify their decisions publicly.

Table 1 depicts our sample by length of study (4 years or more vs. less than four years) and sector. ¹⁰ In Table 1, we also compare our sample to the universe of IHEs in the United States that responded to the IC, FE, C, SFA, and NP 2019 IPEDS surveys, excluding the in-sample IHEs. We see that our sample contains more public institutions and significantly fewer private, for-profit IHEs. We also see that our sample consists of institutions with a larger overall enrollment and more white students, Asian students, and multiracial and Native American students than the average IHE in the United States. The average net price for the IHEs in our sample is more than the average IHE in the United States, and students at the IHEs in our sample receive more in Pell grants (approximately $400) than the average IHE in our sample.

Table 1.

Descriptive Statistics for In-Sample IHE and Out-of-Sample IHE

	In-Sample		Out-of-Sample		|Difference|	P-Value
	Mean	Standard Dev.	Mean	Standard Dev.
% Public institution	0.47	[0.50]	0.30	[0.46]	0.17**	0.00
% Private, nonprofit institution	0.31	[0.47]	0.29	[0.46]	0.02	0.36
% Private for-profit	0.21	[0.41]	0.41	[0.49]	0.20**	0.00
% Four-year or greater	0.67	[0.47]	0.43	[0.49]	0.24**	0.00
Average enrollment	7497.50	[11,363.57]	4098.08	[9,663.63]	3399**	0.00
Student Demographics
% White	0.44	[0.28]	0.49	[0.27]	0.05**	0.01
% Black	0.18	[0.23]	0.18	[0.21]	0.00	0.60
% Hispanic	0.18	[0.22]	0.17	[0.22]	0.01	0.63
% Asian	0.06	[0.08]	0.04	[0.08]	0.02**	0.00
% Other/multiracial/Native American	0.14	[0.14]	0.12	[0.14]	0.02**	0.00
% of Students part-time	0.27	[0.27]	0.28	[0.24]	0.01	0.76
% of Students exclusively online	0.14	[0.18]	0.17	[0.23]	0.03	0.14
% of Students partially online	0.20	[0.17]	0.22	[0.19]	0.02	0.40
% Undergraduate students receiving Pell	0.45	[0.22]	0.46	[0.22]	0.01	0.54
Average Pell Amount Awarded to Undergraduate Students ($)	$4,488	[807.08]	$4,369.24	[847.63]	118.76*	0.02
Average Net Price – Students Awarded Grant or Scholarship Aid in 2018-2019 ($)	$10,622	[4,760]	$9,864	[4,960]	758	0.10
Current Assets ($ per Student)	$107,858	[464,063]	$97,339	[385,350]	$10,519	0.69
Number of Institutions	289		5,983

Notes. Mean is reported; standard deviation shown in brackets. * p<0.05, ** p<0.01. The out-of-sample number of institutions reflects those institutions that responded to the institutional characteristics and fall enrollment survey.

The compositional differences between our sample and the full universe of IHEs in the United States that provided is caused primarily by (1) the fact that we used a stratified sample that does not account for the size of each strata in the universe of IHEs and (2) we over-sampled MSIs. Therefore, although this sample ensures we have a diverse sample of institutions to compare disbursement policies, we acknowledge the limited external validity when extrapolating to the universe of IHEs in the United States.

Implementation Analysis: Quantifying Administrative Burden

We reviewed the publicly posted HEERF I policies for the 289 institutions in our final sample and identified multiple features of disbursement policies that alleviate or exacerbate administrative burden for student applicants. As a condition of receiving HEERF I grants, IHEs were required to publicly post information about the methods used to determine which students would receive aid, the amount of aid students would receive, and any instructions to students about accessing aid (Department of Education, 2020b). As part of an effort to provide ED with information on how IHEs were designing their disbursement policies and, specifically, how IHEs were imposing burdens on students by having them apply for aid, we examined publicly posted information about the HEERF I program on 30 randomly selected IHEs in September 2020. We found that, in addition to whether IHEs required students to apply for aid, there were major differences across institutions, including how they used preexisting information from the FAFSA, how much they disbursed aid, and the key components of the application.

ED was interested in understanding how IHEs were disbursing aid and any barriers being put onto students or parents in this process. Based on conversations between OES and ED, the following main variables of interest were identified: (1) whether IHEs required students to apply to receive aid; (2) if IHEs required students to provide evidence of hardship before receiving aid; and (3) if IHEs automatically disbursed aid. By requiring students to complete an application and submit documentation, IHEs may be imposing several types of administrative burdens. Students may feel ashamed of their financial position (psychological costs), or they may be overwhelmed by documentation requirements (compliance burden). If IHEs automatically disbursed aid to qualified students (via direct deposit) or provided direct deposit forms for students, burdens on students might be reduced. We also identified if (4) IHEs used preexisting financial information to identify specific types of students who qualify for aid, e.g., students with an expected family contribution (EFC) below a specific level, etc. Lastly, we identified (5) if the IHE had a policy where the amount of aid varied by need, i.e., varied by Pell status or EFC.

Using this information, the researcher team systematically reviewed the public websites and identified if, in our final sample of 289 IHEs that posted information regarding their HEERF disbursement policies, they posted information about the five main outcomes specified previously. All websites were reviewed by two research assistants (RAs) who examined the following website pages of each IHE: the main web page of the IHE (which was reviewed for any reference to information about COVID-19, campus shutting down, or financial aid) and, if available, the Student Services and Dean of Student web pages (and any links on this page), the Financial Aid and Registrar web page (and any links on this page), and the IHE’s Facility (or student dorm) web page. Each RA also did a series of Google searches that included the college’s name and keywords, including “emergency aid” and “COVID-19.” Each RA flagged parts of the website using HTML links, and one of the authors of this paper confirmed the presence of the information. ¹¹

In our analysis, we examine predictors of each of these five policies individually as well as empirically or theoretically relevant combinations of policies. We code seven new binary policy variables if the following combination of policies are present: (1) did students need to apply and provide evidence of hardship; (2) did students need to apply and did the amount of aid vary; (3) did students have to apply, provide evidence of hardship, and did the amount vary; (4) did IHEs use existing information to target students and automatically disburse aid; (5) did IHEs use existing information to target students and did the amount of aid vary; (6) did IHEs use existing information to target students, require an application, and did the amount of aid vary; and (7) did IHEs use existing information to target students, require an application, and not vary the amount of aid? Theoretically, the last policy combination presents a case where the institutions are imposing a high level of administrative burden and not varying the amount of aid according to student need, which presents the disbursement policy that falls furthest from best practices for equity. We present the summary statistics of the main policies and combinations of policies in our sample in online appendix Table S2. We see that many IHEs publicly posted that they varied the amount of aid; however, only approximately 40 percent of the sample reported that they automatically disbursed aid.

It is important to acknowledge that, given the variety of policies and the variations in how this information is presented, we could not capture all aspects of HEERF I disbursement strategies. Moreover, simply because an institution did not publicly acknowledge a specific policy does not mean that the element was not a part of their disbursement policy. Many IHEs may have distributed aid through automatic disbursements; however, they simply did not publicly acknowledge this fact in their external policy. Thus, our analysis can be thought of as capturing the lower-bound average in the use of a specific policy. We encourage future researchers to build on our study by utilizing mixed-methods approaches to better capture the nuances of HEERF I implementation that could not be captured by examining publicly available websites. Despite these limitations, the policies we examine are very common policies addressed in almost all the public documentation we reviewed, and these policies will directly correlate with the level of administrative burden students experienced when trying to obtain HEERF I aid.

Methods

Given the descriptive nature of the paper, we primarily visualize average differences in disbursement policies across different types of institutions. We also use a series of simple ordinary least squares (OLS) regressions where we regress a binary policy indicator on one or a series of IHE characteristics—for example, sector status, length of study, MSI status, etc. The model is as follows, where i indexes for each institution:

$$y_i={\beta }_0+{\beta }_k{\chi }_i+{\epsilon }_i$$

Where y_i is the disbursement policy or policies of interest and χ_i represents a vector of institutional characteristics. In most specifications, we will only include a single covariate of interest (e.g., sector, or average length of study), but we also discuss the robustness of the results to the inclusion of additional covariates that account for differences in student enrollment and diversity. Given the number of hypotheses we are testing, we presentp-values adjusted using the free step-down resampling methodology of Westfall and Young (1993), as well as Bonferroni-Holm corrected p-values, to adjust for multiple hypothesis testing.

Results: Administrative Burden in Disbursement Policies, by Institutional Characteristics

We begin by discussing the institutional characteristics that are associated with each of the five individual policies (presented in Figures 1–3 and in columns 1–5 of Table 2) before discussing the results for each combination of disbursement policies (presented in columns 6–12 in Table 2). In Figure 1, we present the percentage of institutions that required students to apply for aid and provide evidence of financial hardship and that automatically disbursed aid to students. We see that private for-profit IHEs were more likely to require students to apply for aid (approximately 70 percent) than private nonprofit and public IHEs (approximately 50 percent). Similarly, private for-profit IHEs in our sample were more likely to require students to submit evidence of financial hardship (approximately 66 percent) whereas less than half of the private nonprofit and public IHEs place this burden on students (approximately 46–47%). In contrast, public institutions were significantly more likely to automatically disburse aid than either private for-profit or private nonprofit IHEs. The results in panel A of Table 2 confirm that private for-profits were significantly more likely to require students to apply for aid and provide evidence of hardship than public IHEs, and both nonprofit and private for-profit IHEs were significantly less likely to publicly announce that they would automatically disburse aid. ¹²

Figure 1.

Student Burden-Related HEERF Policies by Sector.

Figure 2.

Student Burden-Related HEERF Policies, by Four-Year Status.

Figure 3.

Student Burden-Related HEERF Policies, by Minority-Serving Status.

Table 2.

Conditional Correlations Between Sectoral Differences, Year of Education, or MSI Status and Key Disbursement Policy Features

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
	Students Apply for Aid	Students Need to Provide Evidence	Use of Automatic Disbursements	Did Colleges Target Students	Did Amount Vary by Student Need	Amount Varied and Automatic Disbursement	Colleges Targeted Students and Amount Varied	Students Need to Apply and Provide Evidence	Students Need to Apply and Amount Varies	Students need to Apply, Provide Evidence, and Amount Varies	College Targets, Requires App, and Amts Vary	College Targets, Req App, and Same Amt. for All
Panel A: Sectoral Differences (Reference Category = Public IHEs)
Private nonprofits (PNP)	0.00	0.02	−0.50**	−0.01	−0.14*	−0.40**	−0.04	0.02	−0.05	−0.02	0.05	0.04
	(0.07)	(0.07)	(0.06)	(0.06)	(0.06)	(0.06)	(0.07)	(0.07)	(0.07)	(0.07)	(0.06)	(0.03)
	[1.000]	[1.000]	[0.001]	[1.000]	[0.185]	[0.001]	[0.999]	[1.000]	[0.999]	[1.000]	[0.999]	[0.854]
	{1.000}	{1.000}	{0.001}	{0.999}	{0.293}	{0.001}	{1.000}	{1.000}	{1.000}	{1.000}	{1.000}	{1.000}
Private for-profits (PFP)	0.16*	0.20**	−0.60**	0.14*	−0.14+	−0.46**	−0.01	0.13+	0.03	0.01	0.18*	0.12*
	(0.07)	(0.07)	(0.06)	(0.07)	(0.07)	(0.06)	(0.08)	(0.08)	(0.08)	(0.08)	(0.07)	(0.05)
	[0.311]	[0.103]	[0.001]	[0.345]	[0.345]	[0.001]	[1.000]	[0.485]	[1.000]	[1.000]	[0.631]	[0.068]
	{0.638}	{0.180}	{0.001}	{0.719}	{0.720}	{0.001}	{1.000}	{1.000}	{1.000}	{1.000}	{1.000}	{0.095}
Panel B: Four Years or Higher (Reference Category = Two-year IHEs)
4-yr or Higher (4yearPlus)	0.04	0.13*	0.28**	0.12+	0.10+	0.30**	0.13*	0.14*	0.09	0.14*	0.11+	−0.00
	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.05)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.03)
	[0.869]	[0.222]	[0.001]	[0.222]	[0.283]	[0.001]	[0.192]	[0.159]	[0.426]	[0.147]	[0.908]	[0.957]
	{1.00}	{0.315}	{0.001}	{0.316}	{0.380}	{0.001}	{0.295}	{0.268}	{0.592}	{0.251}	{1.000}	{1.000}
Panel C: Minority-Serving Status (Reference Category = All Other IHEs)
Minority-Serving Institution (MSI)	−0.12+	−0.06	0.55**	−0.22**	0.10+	0.42**	−0.07	−0.04	−0.03	0.01	−0.15**	−0.06+
	(0.07)	(0.07)	(0.05)	(0.07)	(0.06)	(0.06)	(0.07)	(0.07)	(0.07)	(0.07)	(0.06)	(0.03)
	[0.243]	[0.583]	[0.001]	[0.005]	[0.288]	[0.001]	[0.583]	[0.707]	[0.755]	[0.850]	[0.018]	[0.350]
	{0.540}	{1.000}	{0.001}	{0.005}	{0.598}	{0.001}	{1.000}	{1.000}	{1.000}	{1.000}	{0.213}	{0.752}

Notes. Robust standard errors in parentheses. **p < 0.01, *p < 0.05, +p < 0.1. N = 289 for all observations. Brackets [] indicate adjusted p-value using the free step-down resampling methodology of Westfall and Young (1993), curly brackets {} indicate the Bonferroni-Holm corrected p-values. For all p-values, we treated each panel—PNP/PFP, 4yearPlus, and MSI, respectively—as separate families when calculating p-values and the family-wise error rate.

These results collectively suggest that private for-profit IHEs are more likely to place burdens on students compared to public and private nonprofit IHEs. This is concerning, especially because students of color make up a disproportionate share of the student population at private for-profit institutions (de Brey et al., 2019). This means that students hit the hardest by the pandemic—namely Black and Hispanic students—were also more likely to be subject to higher levels of administrative burden in accessing HEERF I aid. This aligns with recent research that calls attention to the disproportionate burden in programs that serve predominantly disadvantaged populations (Herd & Moynihan, 2018).

In Figure 2, we find that four-year colleges were equally likely to require students to apply for aid (58 percent compared to 54 percent), but they were significantly more likely to require students to provide evidence of hardship and automatically disburse aid. This may be because four-year institutions are more likely to have the capacity to support automatic disbursements that vary according to student need, whereas two-year institutions may have less administrative capacity.

Lastly, as we show in Figure 3 and panel C of Table 2, minority-serving institutions ¹³ were less likely than non-MSIs to require students to apply and require them to provide evidence of hardship, and they were much more likely to automatically disburse aid to students. Thus, MSIs made it easier for students to obtain funds and were more likely to vary the amount of aid received based on need, but students still needed to be proactive to obtain assistance. ¹⁴

When we examine combinations of disbursement policies in columns 6–12, many of the same conclusions emerge. Consistent with columns 3 and 4, we find that private institutions—both public nonprofits and for-profits—were less likely to use automatic disbursements with varying aid amounts, compared to public institutions (see column 6 of panel A in Table 2). We also find that private, for-profit institutions were more likely to use existing information to target students, have students apply for aid, and require students to provide evidence of hardship or financial need (regardless of whether the aid amount varies or not). This is consistent with the results in column 1 and column 2 of panel A. In panel B, we find that four-year IHEs were more likely to vary aid amounts in combination with requiring applications and requiring evidence of hardship compared to two-year IHEs. In panel C, we find that MSI IHEs were less likely to require applications and use existing information from the FAFSA to target students, regardless of whether the amount of aid varied.

If we include a vector of covariates accounting for student enrollment and diversity in the OLS regressions presented in Table 2, we find that many of our conditional correlations between the disbursement policies and sector (for-profit and nonprofit status), as well as minority-serving status, remain statistically significant. However, many of the relationships between length of study—notably the relationships between length of study and the policy that students need to provide evidence, the use of automatic disbursements, and the combinations of these two policies—become statistically insignificant. These results are available upon request.

To account for the fact that our sample is not representative of the population of IHEs in the United States, we weight our sample using inverse probability weighting. This inverse probability weights will reweight our sample to better approximate the actual population of IHEs in the United States. In online appendix Table S3, we reweight our surveyed sample by these inverse probability scores and compare them to the full universe of IHEs in the United States. Comparing Table 1 and online appendix Table S3, we find that this weighting does improve the representativeness of our sample when compared to the universe of IHEs in the United States. We present the results of these weighted regressions in online appendix Table S4, and we do not find large differences between the results in this table and those found in Table 2.

Discussion

Understanding the significant variation in program design and implementation can help higher education policy scholars build evidence on the impacts of emergency aid programs like HEERF I on student outcomes during the COVID-19 pandemic. In our investigation into the disbursement policies for HEERF I aid, we found significant variation in the level of burden and the use of targeting across IHEs. Institutions with public service missions, notably public IHEs and MSIs, were the most likely to reduce burden on students in applying for HEERF I aid. Compared to private for-profit institutions, students at these institutions have lower compliance costs—that is, they are less likely to require students to apply for aid and thus less likely to require them to demonstrate evidence of hardship. Moreover, public institutions were more likely to vary the amount of aid based on need, which could enhance equity by increasing the allocation of aid to students in need (Bell, 2019; Deming & Dynarski, 2009; Heller & Marin, 2002).

Conversely, IHEs in the private for-profit sector were the most likely to impose additional burdens on students seeking HEERF I emergency aid. One potential reason for this behavior is that private for-profit IHEs are often subjected to greater and more substantial reporting requirements than public and private nonprofit IHEs (e.g., gainful employment). If these IHEs expected that their disbursement decisions would be subject to greater oversight and a potential audit, then they may have required students to supply more paperwork to justify these disbursement decisions. However, in doing so, these institutions increase the compliance burdens for students. This highlights the need to strike a balance between accountability mechanisms, like audits of private for-profit colleges and detailed reporting requirements and reducing administrative burdens for students with the greatest need. Another potential reason for for-profits imposing additional burdens is the contradictory guidance from the Department of Education during the rollout of HEERF, which could have led to confusion in what documentation was required in the implementation process. This possibility highlights the importance of consistent and clear guidance to colleges about what type of documentation is necessary so that institutions do not feel the need to arbitrarily collect additional documents from students during a time of crisis.

Our implementation analysis of the student portion of HEERF I takes the first step toward a broader research agenda that extends existing literature on emergency financial aid and informs the design of future emergency aid efforts at the federal level. Although we focused on HEERF I aid to students passed in March 2020, there is much more work to be done to explore the institutional aid portion of HEERF I and the implementation of the second round of relief funding passed December 2020 (HEERF II) (Department of Education, 2021). Moreover, our results characterizing the variation in disbursement policies set the stage for future research on (1) the determinants of disbursement policy decisions across IHEs, (2) how disbursement policies influenced student access to aid, and (3) the impact of HEERF aid on student outcomes. For example, did prepandemic institutional resources (e.g., financial reserves) or financial systems (e.g., more technologically advanced or adaptive payment systems) influence how institutions responded to Covid and, relatedly, how these IHEs chose to disburse aid? Or, as noted previously, does the political or legal environment of certain IHEs influence their disbursement decisions? Future research could also explore variations in HEERF disbursement policies and how these policies affected student aid access among a larger or more narrowly focused sample of IHEs (e.g., HBCUs, HSIs, TCUs, AANAPISIs, exclusively flagship public, or two-year community college). Given our limited sample size of each type of MSI, future research could generate additional insights by exploring the differences in disbursement policies and aid access among students across HBCUs, HSIs, TCUs, and AANAPISIs.

Next, examining student reactions and access to HEERF across different disbursement models is of paramount importance in establishing whether the aid reached students who needed help the most. After understanding the consequences of implementation decisions on who receives aid, an obvious next step is assessing how effective HEERF aid is at improving student outcomes, such as retention and graduation. Examining the average effect of aid on academic outcomes for all students and how aid may have had different effects on specific groups of students (e.g., low-income students, first-generation students) could help fill an evidence gap on the impacts of emergency student aid. Given how IHEs chose to distribute aid, researchers may be able to exploit eligibility thresholds based on estimated family contribution (EFC) to implement a regression discontinuity (RD) design, which would identify the causal impacts of aid for students near the threshold.

Lastly, another helpful direction for future study is how the availability of other sources of aid and wrap-around services were affected by HEERF and may moderate the effect of HEERF on student outcomes. Although HEERF was intended to go directly to students with IHEs not taking the funding into account for other institutional decision-making, it is possible that both IHEs and external funders shifted other aid because of the influx of HEERF money. Some IHEs made additional emergency assistance available for students who were not Title IV eligible—for example, for noncitizen students. Other institutions may have utilized the other half of HEERF money allocated for institutional expenses to build technical capacity and student support systems. Ultimately, the impacts of emergency aid such as HEERF are likely dependent on the suite of resources available to students. Understanding the conditions under which a program like HEERF could supplement existing resources to meaningfully impact student outcomes may help IHEs and policymakers design and implement effective programs to support students during emergencies.

Authors

ELIZABETH BELL, PhD, is an assistant professor at the Askew School of Public Administration and Policy at Florida State University, 627 Bellamy Building 113 Collegiate Loop, Tallahassee, FL 32306; ebell3@fsu.edu . Her research interests are in policy design and implementation, with a focus on education policy and social equity.

DAVID J. SCHWEGMAN, PhD, is an assistant professor of public administration and policy at the School of Public Affairs at American University, 4400 Massachusetts Ave NW #270, Washington, DC 20016; schwegma@american.edu . His research interests are in state and local public finance and education finance.

MICHAEL HAND, PhD, is a behavioral science specialist at the U.S. General Services Administration, Office of Evaluation Sciences, 1800 F Street, NW Washington, DC 20405; michaels.hand@gsa.gov .

MICHAEL DIDOMENICO, MPP, leads the GSA Evaluation Division at the U.S. General Services Administration, Office of Evaluation Sciences, 1800 F Street, NW Washington, DC 20405; michael.didomenico@gsa.gov .