Advertisement

Journal of Economics, Race, and Policy

, Volume 1, Issue 2–3, pp 176–195 | Cite as

Business Cycles and HBCU Appropriations

  • Alberto OrtegaEmail author
  • Omari H. Swinton
Original Article

Abstract

The demand for postsecondary education has steadily increased since the 1970s. However, state government’s pecuniary investments in higher education have steadily declined over this same period, a decline that is magnified during recessions. This paper examines the effect of state level business cycle measures on the funding of Historically Black Colleges and Universities (HBCUs), relative to other schools. We find that when controlling for school characteristics, HBCU funding is very responsive to the business cycle—particularly downturns. Generally, HBCUs receive less funding than other institutions, but this result is exacerbated during periods of economic downturn. Our results are driven by 4-year institutions—the most common type of HBCU—and dissipates when only considering 2-year institutions. Additionally, when examining the results within each state, we find that during periods of increased unemployment, Kentucky, Maryland, North Carolina, Ohio, South Carolina, and Virginia provide fewer funds to 4-year HBCUs. We also find that Florida, Kentucky, Louisiana, Pennsylvania, Tennessee, Texas, and West Virginia do not appropriate fewer funds to 4-year HBCUs within their states, independent of any unemployment effects. The remaining states have provided less funding for 4-year HBCUs, in general, over the time of our sample. We also examine states that have at least one public 2-year HBCU. The results suggest that Alabama appropriates more funds to 2-year HBCUs, relative to other 2-year schools, during economic downturns and otherwise; South Carolina provides fewer funds to HBCUs during periods of higher unemployment. In discussions with each states’ higher education executive association, we find that the mechanism through which funding disparities persist are difficult to assess due to the influence and autonomy state governors and legislatures possess in determining appropriations.

Keywords

Business cycles Higher education finance 

JEL classification

H52 H75 I22 E24 

Introduction

There has been a steady increase in the demand for postsecondary education since the 1970s that has not been matched by state government spending on higher education—a phenomenon that is exacerbated during economic downturns. The average support for higher education remains below pre-2007 recession levels,1 while access to educational opportunities for all students has greatly increased. The benefits of obtaining a higher education are evidenced throughout literature (Autor et al. 2008; Card 2001; Jaeger and Page 1996). Many fear that continued cuts to education funding coupled with steady increases in the demand for postsecondary education will result in perpetual hikes in tuition, which are already rising at a rate faster than inflation and family wages (Desrochers and Wellman 2011). Some argue that lower appropriations can result in fewer available courses and full-time faculty, which may potentially decrease the quality of education. A lack of investment and support for public institutions threatens to put higher education out of reach for a larger share of the population. Given that state funding of public Historically Black Colleges and Universities (HBCUs) has been shown to be lower than that of other institutions (Sav 2010), we investigate whether these inequities in state funding persist during periods of economic downturns. This paper examines whether institution type, specifically congressional designation as an HBCU, results in differential funding decreases within a state.

HBCUs were founded to provide educational opportunities to African Americans when other postsecondary education institutions were closed, restricted, or discouraged. Today, HBCUs award nearly one-fifth of all bachelor degrees earned by African Americans (Palmer and Gasman 2008). This is significant given that HBCUs represent only 3% of higher learning institutions in the nation. Some scholars argue that HBCUs educate many African American and other diverse students, regardless of academic preparation, test scores, and socioeconomic status; some of whom might otherwise not be able to pursue a degree at the collegiate level (Freeman and Cohen 2001). Given that HBCUs are effective at enrolling, retaining, and graduating African Americans (Minor 2008), there are public concerns regarding budget cuts or a lack of support for these institutions. For instance, Sav (2010) finds that although the discrepancy in state funding has decreased over the last three decades, the disparities in the funding of HBCUs remain.

To the extent that higher education cuts are most dire during recessionary periods, it is of interest to determine whether inequities in funding persist during such periods. Humphreys (2000) finds that state appropriations to higher education decrease substantially during recessionary downturns. In addition, enrollment during recessionary periods has been found to be countercyclical (Johnson 2013; Humphreys 2000; Betts and McFarland 1995). Together, these outcomes create pressures on colleges and universities that may affect the quality of education—pressures that can be exacerbated at HBCUs due to historical funding levels proportionately lower than their counterparts. Moreover, Clelan and Kofoed (2017) finds that during recessionary periods states cut back their need based student financial aid, which place a greater financial burden on families. To prevent a drop in sorely needed student financial aid, and thus, in student enrollment, families seek elusive increases in state and federal aid. Inevitably, families of least advantage and underrepresented minorities are most adversely impacted.

Using macroeconomic indicators similar to those used in Humphreys (2000) and McGranahan (1999) along with linear regression model setup similar to Sav (2010), we examine the effect business cycle fluctuations on the funding of higher education institutions. We add to the literature by considering that funding across all types of colleges/universities is heterogeneous during economic downturns and otherwise. We find that on average HBCUs receive less funding in general than other types of institutions. During downturns, the funding appropriated to HBCUs decreases at a larger rate than other institutions, a result that is driven by inequities among 4-year institutions. We also examine whether unequal funding persists within each state and find that some states have not experienced differential funding effects between HBCUs and other institutions. Further, we explore some of the possible mechanisms through which these differences persist. Although most states with HBCUs have funding formulas, discussions with each state’s higher education executive association reveal that state funding to higher education, independent of a funding formula, is often influenced by state governor and legislature preferences.

Data and Methodology

Revenues raised by institutions of higher education arise from a variety of sources including tuition and fees, government grants, as well as private donations. However, in this paper, we focus on the funding appropriated by state government. The higher education data used in our analysis comes from the Delta Cost Project Database, which is derived from the Integrated Postsecondary Education Data System (IPEDS) longitudinal database. The IPEDS surveys postsecondary institutions under the auspices of the National Center for Education Statistics and is comprised of data for the academic years 1986–1987 through 2014–2015 and contains information on college and university finances, enrollment, staffing, completions, and student aid. Given that the financing of higher education takes place in a very enigmatic setting with many moving parts, it is imperative to consider institutional level covariates to help explain state funding. Thus, we model our dependent variable of interest, state appropriations, yist (in millions) to public institution i in state s at year t, as

$$ {y}_{is t}=\alpha +{\beta}_1\mathrm{HBC}{\mathrm{U}}_{it}+{\beta}_2\mathrm{Unem}{\mathrm{p}}_{s\left(t-1\right)}+\delta {\left(\mathrm{HBCU}\times \mathrm{Unemp}\right)}_{is\left(t-1\right)}+{\gamma}_s+{\varphi}_t+ X\beta +{u}_{is t} $$
where HBCUit is an indicator for whether the institution is an HBCU, Unemps(t − 1) represents a state’s unemployment rate in year (t − 1), and γs and φt are state and year fixed effects, respectively. Our variable of interest is the interaction between HBCUit and Unemps(t − 1), where the effect of being an HBCU in a state during an increase in the unemployment rate in year (t − 1) is captured by the coefficient δ. In addition to lagged values of state unemployment, we also examine the effects of state income per capita in year (t − 1) on yist. As in Humphreys (2000), we use the lagged values of the business cycle measures to explain observed variation in appropriations. Using measures in time t may lead to simultaneity bias and incorrect estimates. The variable X contains university/college and state level covariates that help explain state appropriations to public institutions. Table 1 presents summary statistics for the variables used in our analysis. We separately provide the means for both HBCU and other institutions. A list of the HBCUs used in our analysis as well as other institutions in states with an HBCU (henceforth, HBCU states) is listed in the Appendix. The dependent variable, state appropriations, is revenues received by the institution through acts of a state legislative body. Non-HBCU institutions receive, on average, about $17 million more dollars from their respective states; however, their enrollment consists of about 2000 more students, captured by the Full-Time Equivalent (FTE) enrollment.2 When restricted to only 4-year schools, the difference in appropriations is more pronounced. Four-year HBCUs have received about $1200 fewer per-student than other types of 4-year institutions over the time of our sample. It is also important to note the wider variance in the amount of state dollars received by other institutions in comparison to HBCUs. We also include FTE enrollment given that states are most likely to support institutions that attract more students and produce more credit-hours. We also consider revenues from auxiliary enterprises, including monies collected from residence halls, student health services, and intercollegiate athletics. Many of these activities produce positive externalities (e.g., sports arenas) and thus may be subsidized, in part, by state funds.
Table 1

Summary statistics

 

HBCUs

Other institutions

Mean

Std. Dev.

Mean

Std. Dev.

(1)

(2)

(3)

(4)

Institution

 State appropriations

35.94

25.27

52.77

117.57

 FTE enrollment

3518.45

2210.58

5656.71

8923.61

 Auxiliary rev.

11.47

9.31

18.86

53.95

 Federal appropriations

22.01

15.72

26.78

83.10

 Private earnings

1.45

2.20

9.68

49.84

 Tuition rev. share

0.24

0.09

0.28

0.15

State level

 

HBCU states

Non-HBCU states

 Unemployment rate

6.39

2.04

5.96

2.03

 Democratic governor

0.55

0.49

0.50

0.50

 Income per capita

36,112.73

8105.66

37,746.42

7674.67

 Number of institutions

47

1693

N

1557

46,408

All pecuniary measures are in millions unless otherwise indicated. The first panel provides institution level summary statistics for HBCUs and other institutions. The second panel provides state level demographic summary statistics for states that have at least one HBCU and states that do not have any HBCUs

Table 1 reveals that auxiliary revenues are much larger for other institutions than for HBCUs—a result consistent with the discrepancy in enrollment. Moreover, postsecondary institutions also receive funds from the federal government. This measure accounts for the potential of crowding-out state investment (Knight 2002). Other institutions receive about $27 million in federal funds, whereas HBCUs receive $22 million (with a much larger standard deviation for other institutions). This difference is much smaller than that of state funds. We also include private earnings in our specification. This captures the total amount of revenue coming from affiliated entities, private gifts, grants, contracts, investment returns, and endowment earnings. Similar to the federal funds argument, private funds may crowd-out state funds. Thus, fewer private revenues may result in increased state involvement. A college or university’s reliance on tuition revenues and their share of revenues from net tuition are other measures that allow us to control for how independent an institution is. It is reasonable to expect an inverse relationship between tuition reliance and public appropriations. An extreme example would be where completely private institutions’ tuition prices are typically higher than their public counterparts. Lastly, we control for the sector of the institution (i.e. 4-year versus 2-year).

State level variables were collected from the University of Kentucky Center for Poverty Research (2016), where we include a dichotomous variable equal to 1 if the governor is a Democrat and zero otherwise. Democratic governors have been shown to be more prone to public spending—particularly, education spending (Beland 2015; Tandberg 2010; Kane 2003). For our variables of interest, we consider two variables that reflect a state’s business cycle: (1) the unemployment rate and (2) real per capita income. We focus on state, rather than national, business cycles since a state’s economy and subsequent budget are most likely to affect its funding decision across institutions.

Our baseline analysis will be partitioned in two ways: (1) we consider the average effect of state business cycles on funding to HBCUs when compared to other institutions across all states and (2) we then restrict the analysis to states in which there are HBCUs. This distinction is important because a legislature’s decision to fund public institutions within a state may depend on the type of institutions within their jurisdiction. Moreover, the summary statistics from Table 1 indicate that HBCUs states’ demographics are slightly different than that of other states. Figure 1 depicts the states in which at least one HBCU resides. HBCUs are primarily located in southern states.
Fig. 1

HBCU states

Results

Main Findings

We start by examining state recessionary effects on HBCUs across states presented in Table 2. Columns 1 and 2 consider institutions in all 50 states while, columns 3 and 4 restrict our analysis to states with at least one HBCU. The coefficient on HBCU is negative across all specifications. This finding coincides with that of Sav (2010) where an institution’s HBCU status results in fewer state funds, all else being equal. When examining our variable of interest, the interaction between HBCU status, and the unemployment rate, we notice a similar negative pattern. During periods of higher unemployment, HBCUs receive fewer funds. When examining schools across all states, a 1% increase in the unemployment rate results, on average, in approximately $820,000 fewer dollars in state revenues for HBCUs, column 1. This result persists when restricting the analysis to HBCU states, column 3. However, the magnitude of the decrease is lower—about $464,000 fewer dollars in state revenues during periods of increased unemployment. We also consider fluctuations in a state’s income per capita. Columns 2 and 4 both indicate that increases in a state’s income increase HBCU funding. A 1% increase in state income per capita results in an increase of about $346,000 for HBCUs. However, this increase does not outweigh the decrease in funding to HBCUs independent of business cycle.
Table 2

The effect of business cycle on HBCUs

 

(1)

(2)

(3)

(4)

HBCU × Unemp_(t − 1)

− 0.820a (0.352)

 

− 0.464a (0.220)

 

Unemp_(t − 1)

− 0.279 (1.264)

 

0.594 (1.441)

 

HBCU × income per capita_(t − 1)

 

0.096a (0.043)

 

0.092a (0.040)

Income per capita_(t − 1)

 

0.123b (0.0446)

 

0.0652 (0.106)

HBCU

− 16.65b (4.596)

− 45.97b (14.00)

− 12.39b (3.168)

− 42.13b (12.33)

FTE enrollment

0.690b (0.0533)

0.690b (0.0531)

0.736b (0.0646)

0.736b (0.0646)

Auxiliary rev.

0.373b (0.0868)

0.373b (0.0870)

0.225a (0.0955)

0.225a (0.0946)

Federal appropriations

0.380b (0.0898)

0.380b (0.0898)

0.515b (0.0679)

0.517b (0.0679)

Private earnings

− 0.0365 (0.125)

− 0.0356 (0.125)

− 0.111 (0.144)

− 0.111 (0.144)

Tuition rev. share

− 134.8b (19.30)

− 134.5b (18.80)

− 137.3b (16.92)

− 135.2b (17.20)

Democratic governor

− 1.099

(1.805)

(6.331)

− 1.451

(1.930)

(11.42)

2.754b

(0.841)

(7.370)

2.652b

(0.911)

(14.65)

2-year college dummy

− 23.42b (5.399)

− 23.21b (5.406)

− 17.17b (4.158)

− 16.81b (4.209)

Constant

57.33b (13.19)

17.19 (15.03)

41.72b (12.69)

26.58 (29.13)

School controls

Yes

Yes

Yes

Yes

State FE

Yes

Yes

Yes

Yes

Year FE

Yes

Yes

Yes

Yes

Observations

31,183

31,183

16,120

16,120

R-squared

0.855

0.855

0.864

0.865

The dependent variable is inflation-adjusted state appropriations in 2015 (in millions) dollars. The variables of interest are the interaction between a dummy for HBCU and the unemployment rate and interaction between HBCU and income per capita. Columns 1 and 2 consider publicly funded colleges or universities in all 50 states. Columns 3 and 4 restrict the analysis to states that have an HBCU. Control variables included are total full-time equivalent students enrolled, revenues from auxiliary enterprises, federal appropriations, revenue from affiliated entities, private gifts, grants, and contracts and endowment earnings, share of operating revenues from net tuition, an indicator equal to one if the college or university is a 2-year college and zero if a 4-year college, and an indicator equal to one if the state governor was a Democrat (zero otherwise). All regressions contain state and year FE. Standard errors are in parenthesis and clustered at the state level. Income per capita and full-time enrollment estimates are scaled by 100

aStatistically significant at 5%

bStatistically significant at 1%

We also consider the possibility that state legislature preferences in funding 4-year colleges differ from that of 2-year institutions. Table 3 restricts our analysis to 4-year institutions. Our unemployment measure is consistent with our previous findings. Columns 1 and 3 suggest that during periods of higher unemployment HBCUs receive significantly fewer funds whether we consider all states or only states with HBCUs. Moreover, the effects in Table 3 are more pronounced—a 1% increase in a state’s unemployment rate results in approximately two million fewer dollars for an HBCU. This amounts to about $514 fewer per 4-year HBCU student. Interestingly, the income per capita result dissipates whether considering all states or only HBCU states. This result is consistent with Humphreys (2000), who finds that the effect on appropriations is larger during downturns than during expansionary periods. Tables 2 and 3 indicate that 4-year HBCUs are subject to larger funding cuts during downturns when compared to other 4-year institutions.
Table 3

Business cycle effects, 4-year colleges

 

(1)

(2)

(3)

(4)

HBCU × Unemp_(t − 1)

− 1.979b (0.541)

 

− 1.739b (0.419)

 

Unemp_(t − 1)

− 0.169 (1.835)

 

0.891 (2.204)

 

HBCU × income per capita_(t − 1)

 

0.124 (0.0836)

 

0.104 (0.0728)

Income per capita_(t − 1)

 

0.212b (0.000779)

 

0.156 0.00186)

HBCU

− 31.54b (6.852)

− 65.02a (28.18)

− 27.73b (6.592)

− 54.94a (25.15)

Constant

100.2b (19.10)

32.39 (28.46)

74.58b (20.15)

32.73 (52.01)

School controls

Yes

Yes

Yes

Yes

State FE

Yes

Yes

Yes

Yes

Year FE

Yes

Yes

Yes

Yes

N

13,142

13,142

6935

6935

R-squared

0.870

0.870

0.878

0.878

The dependent variable is inflation-adjusted state appropriations in 2015 (in millions) dollars. The variables of interest are the interaction between a dummy for Historically Black College or University (HBCU) and the unemployment rate, and interaction between HBCU and income per capita. Columns 1 and 2 consider publicly funded 4-year colleges or universities in all 50 states; columns 3 and 4 restrict the analysis to states that have an HBCU. Control variables included are total full-time equivalent students enrolled, revenues from auxiliary enterprises, federal appropriations, revenue from affiliated entities, private gifts, grants, and contracts and endowment earnings, share of operating revenues from net tuition, and an indicator equal to one if the state governor was a Democrat (zero otherwise). All regressions contain state and year FE. Standard errors are in parenthesis and clustered at the state level. Income per capita estimates are scaled by 100

aStatistically significant at 5%

bStatistically significant at 1%

Due to the magnitude of the unemployment effect for 4-year colleges, it is of interest to examine whether our results persist for 2-year schools. An understanding of the components that affect these types of institutions is imperative due to their prevalence in the USA. Two-year colleges represent about 51.6% of all public undergraduate enrollment. Table 4 presents our results for 2-year colleges. Columns 1 and 2 show the result across all states, while columns 3 and 4 restrict the results to states that have a 2-year HBCU in our sample. These states are Alabama, Louisiana, Mississippi, and South Carolina. The latter estimates should be interpreted with caution given that over 80% of the HBCUs in our sample are 4-year schools.3 The findings in Table 4 indicate that it is 4-year colleges that drive the decrease in state funding during recessionary periods, and otherwise, for HBCUs. The estimates of the HBCU indicator in Table 4 are now imprecisely estimated. Put another way, when restricting the analysis to 2-year colleges, we do not see unequal levels of funding resulting from being designated an HBCU, all else equal. The results in columns 1 and 2 suggest that when comparing 2-year colleges across all states, we do not see a disparity in state funding in response to the business cycles or otherwise—a conclusion that differs from previous studies (e.g., Sav 2010). Interestingly, when looking only at states with a public 2-year HBCUs, there seems to be an increase in funding in response to increases in state unemployment rate. A 1% increase in the unemployment rate leads to about $119,000 more dollars to 2-year public HBCUs relative to other HBCUs. Over the time of our sample, this is about an extra $60 per public 2-year HBCU student. Although this result considers a small number of 2-year HBCUs among only four states, it indicates that the inequity in funding is largest among the most common type of HBCUs—4-year HBCUs.
Table 4

Business cycle effects, 2-year colleges

 

(1)

(2)

(3)

(4)

HBCU × Unemp_(t − 1)

0.080 (0.116)

 

0.119b (0.057)

 

Unemp_(t − 1)

− 0.671 (0.494)

 

− 0.277 (0.747)

 

HBCU × (income per capita)_(t − 1)

 

0.015 (0.024)

 

0.011 (.001)

Income per capita_(t − 1)

 

0.052b (0.021)

 

− 0.065 (0.060)

HBCU

0.484 (2.353)

− 4.720 (5.888)

1.956 (2.041)

− 2.410 (2.849)

Constant

13.88c (4.374)

− 7.185 (5.812)

13.93a (4.813)

28.66 (15.85)

School controls

Yes

Yes

Yes

Yes

State FE

Yes

Yes

Yes

Yes

Year FE

Yes

Yes

Yes

Yes

N

18,041

18,041

1551

1551

R-squared

0.781

0.781

0.807

0.808

The dependent variable is inflation-adjusted state appropriations in 2015 (in millions) dollars. The variables of interest are the interaction between a dummy for Historically Black College or University (HBCU) and the unemployment rate and interaction between HBCU and income per capita. Columns 1 and 2 consider publicly funded 2-year colleges or universities in all 50 states; columns 3 and 4 restrict the analysis to states that have at least one public 2-year HBCU. Control variables included are total full-time equivalent students enrolled, revenues from auxiliary enterprises, federal appropriations, revenue from affiliated entities, private gifts, grants, and contracts and endowment earnings, share of operating revenues from net tuition, and an indicator equal to one if the state governor was a Democrat (zero otherwise). All regressions contain state and year FE. Standard errors are in parenthesis and clustered at the state level. Income per capita estimates are scaled by 100

aStatistically significant at 10%

bStatistically significant at 5%

cStatistically significant at 1%

Examining Each State

Policy Discussion

Given the large differences in funding, we consider the mechanism through which this inequality persists. Each state in the sample has a unique set of rules to determine the level of funding that each institution receives. We contacted each state’s higher education executive association to get a clearer indication as to how funding in each state is determined.4 What we found varied from state to state and sometimes from year to year. The following gives a brief outline of the main determinations from each state.5 Funding in Alabama is determined by a base level of funding that might increase in the following year. The total allocation, however, is determined by the state legislature. Arkansas is moving to a productivity funding model where their overall funding level is set by the legislature and governor. In Delaware, each institution makes its own budget request and the legislature makes the final decision. In Florida, the higher education system provides guidance on the amount or baseline that they request from the state. Each institution may request additional money outside the board, but this is subject to legislative approval. In Georgia, a funding formula is used to request a lump sum of funds, but the governor and legislature have liberties in adjusting the formula. After this, the lump sum is divided by the board of regents. In Kentucky, a general budget request is made by Council on Postsecondary Education who tries to get unanimous approval from member presidents. The funding is based on a base plus system but is moving to a performance model going forward where the governor and general assembly ultimately make the funding decision. In Louisiana, using each institution’s request, the Board of Regents sends a budget for approval by the legislature. In Maryland, the budget that comes from the governor determines the overall allocations of funds to each public institution in the state. In Missouri, the general assembly appropriates funding using a core amount based on the prior year. The state is moving toward a performance funding model. In North Carolina, funding decisions are made by the state legislature with a portion of funding resulting from a formula and the rest subject legislative discretion. In Ohio, the funding formula has recently changed. In our conversation with Ohio’s higher education executive association, there was an acknowledgement of the risk factors associated with unequal funding, which has led to the supplemental funding of Central State University. In Oklahoma, the legislature gives one lump sum amount of funding to the state system of Higher Education. The amount is then split among the schools using a base level appropriation. In Pennsylvania, the state uses an allocation formula to determine the level of funding for each institution. In South Carolina, the general assembly determines the funding levels. In Tennessee, the governor and general assembly use formulas to determine the overall budget and submit that to the Higher Education Commissions. In Texas, the formula advisory committee along with the commission on higher education provides funding recommendations that are then subject to approval from the legislative budget board. In Virginia, a funding model is used as a guideline to give budget recommendation to the legislature; however, the legislature ultimately determines overall funding. In West Virginia, each school submits a budget to the policy commission. The budget is either approved or sent back. Once a budget is approved, it is sent to the legislature for final approval.

Acknowledging the heterogeneity in the state funding mechanisms, we consider whether all HBCU states have perpetuated unequal funding, during downturns or otherwise, over time. The next section discusses within state results for states with at least one public HBCU.

State-by-State Analysis

Although 4-year HBCUs experience less funding both during and outside of economic downturns, the reasons for funding differences across states may vary based on state preferences, funding mechanisms, etc. Table 5 examines each state individually to identify differences in funding that result from distinct state heterogeneity. Of the 19 states that have HBCUs, 12 states provide significantly fewer state funds to their 4-year HBCUs relative to other 4-year institution—independent of business cycle effects. Table 5 illustrates that over time some states have managed to fund 4-year HBCUs more comparably to other institutions. Florida, Louisiana, Pennsylvania, Tennessee, Texas, and West Virginia do not differentially fund HBCUs, irrespective of any unemployment effects. Although all the point estimates for the HBCU dummies are negative, they are not statistically significant for these six states. We also find that during periods of increased unemployment, Kentucky, Maryland, North Carolina, Ohio, South Carolina, and Virginia provide fewer funds for 4-year HBCUs. However, Kentucky, which has only one 4-year HBCU, seems to have provided more funds to its HBCU over time, a result that although marginally significant outweighs the negative effect during downturns.
Table 5

Business cycle effects within HBCU states, 4-year colleges

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

Alabama

Arkansas

Delaware

Florida

Georgia

Kentucky

Louisiana

Maryland

Mississippi

Missouri

North Carolina

Ohio

Oklahoma

Pennsylvania

South Carolina

Tennessee

Texas

Virginia

West Virginia

HBCU × Unemp_(t − 1)

− 1.134 (0.735)

− 0.239 (0.917)

0.150 (0.589)

− 1.758 (1.745)

− 1.041 (0.730)

− 3.055b (1.489)

− 0.0223 (0.283)

− 3.945c (1.423)

− 0.332 (0.596)

− 1.716 (2.028)

− 2.553b (1.100)

− 4.014a (2.433)

1.132 (1.236)

− 1.415 (0.892)

− 2.123b (0.973)

− 1.793 (1.282)

− 3.397 (2.358)

− 2.346a (1.242)

− 0.597 (0.961)

Unemp_(t − 1)

0.931 (20.50)

17.80a (10.44)

0.899 (3.520)

1.556 (10.82)

− 6.434 (14.22)

35.79 (33.76)

6.543 (38.36)

3.553 (6.055)

− 0.951 (38.99)

− 57.97 (85.14)

17.75b (7.475)

− 24.64 (210.0)

33.64 (85.64)

59.83a (30.89)

3.199 (12.67)

126.7 (244.2)

20.17 (72.00)

12.34 (62.54)

− 3.694 (20.29)

HBCU

− 20.78c (5.234)

− 27.71c (6.424)

− 237.4c (26.56)

− 18.54 (11.29)

− 27.01c (5.337)

24.65a (12.93)

− 6.225 (3.981)

− 30.39c (8.165)

− 13.67c (5.050)

− 39.59c (12.84)

− 36.31c (6.590)

− 54.17c (19.05)

− 11.40a (6.709)

− 5.943 (5.435)

− 31.47c (9.471)

− 10.09 (11.13)

− 17.01 (15.21)

− 22.80c (5.678)

− 10.56 (7.602)

Constant

75.81 (155.6)

− 95.66 (68.61)

287.3c (39.95)

15.18 (80.79)

221.3a (129.1)

44.58 (177.6)

− 22.40 (276.4)

43.80a (26.53)

63.58 (342.6)

416.1 (554.0)

− 13.50 (43.52)

203.3 (1541)

− 93.89 (445.4)

− 437.4a (239.5)

110.3 (108.8)

− 1464 (1856)

− 81.83 (470.8)

− 24.87 (374.0)

102.9 (264.9)

N

390

286

53

469

578

214

398

346

220

270

452

373

378

542

347

209

694

429

287

R-squared

0.904

0.968

0.999

0.927

0.887

0.871

0.941

0.945

0.909

0.970

0.949

0.917

0.926

0.925

0.840

0.970

0.940

0.885

0.931

The dependent variable is inflation-adjusted state appropriations in 2015 (in millions) dollars. The variable of interest is the interaction between a dummy for Historically Black College or University (HBCU) and the unemployment rate. Each column presents the results for all 4-year public colleges or universities in each state that contains at least one HBCU, respectively. Control variables included are total full-time equivalent students enrolled, revenues from auxiliary enterprises, federal appropriations, revenue from affiliated entities, private gifts, grants, and contracts and endowment earnings, share of operating revenues from net tuition, and an indicator equal to one if the state governor was a Democrat (zero otherwise). All regressions contain year FE. Standard errors are in parenthesis

aStatistically significant at 10%

bStatistically significant at 5%

cStatistically significant at 1%

Due to the increase in funding for 2-year public HBCU, during increases in state unemployment, we also examine effects within each of these states. The results in Table 6 show the effects of state level unemployment and HBCU status on funding over time in Alabama, Louisiana, Mississippi, and South Carolina. Although Alabama unequally funds 4-year HBCUs, relative to other HBCUs (see Table 5), column 1 of Table 6 suggests that they fund 2-year public HBCUs much more than other 2-year institutions. In addition to generally providing more funds to 2-year HBCUs, they also provide more state funds during periods of increased unemployment; a 1% increase in the unemployment rate results in about $277,000 more for 2-year HBCUs relative to other public 2-year schools.
Table 6

Business cycle effects within HBCU states, 2-year colleges

 

(1)

(2)

(3)

(4)

Alabama

Louisiana

Mississippi

South Carolina

HBCU × Unemp_(t − 1)

0.277b (0.124)

− 0.0989 (0.482)

− 0.0459 (0.124)

− 0.568b (0.237)

Unemp_(t − 1)

− 0.765 (0.697)

− 9.571 (13.24)

− 2.525 (1.859)

1.394 (0.904)

HBCU

4.516c (0.887)

− 8.294a (4.416)

− 1.319 (1.050)

− 1.971 (1.553)

Constant

8.681a (5.075)

94.07 (90.92)

27.98a (16.20)

2.334 (6.522)

Observations

571

180

309

491

R-squared

0.881

0.865

0.910

0.858

The dependent variable is inflation-adjusted state appropriations in 2015 (in millions) dollars. The variable of interest is the interaction between a dummy for Historically Black College or University (HBCU) and the unemployment rate. Each column presents the results for all 2-year public colleges or universities in each state that contains at least one 2-year HBCU, respectively. Control variables included are total full-time equivalent students enrolled, revenues from auxiliary enterprises, federal appropriations, revenue from affiliated entities, private gifts, grants, and contracts and endowment earnings, share of operating revenues from net tuition, and an indicator equal to one if the state governor was a Democrat (zero otherwise). All regressions contain year FE. Standard errors are in parenthesis

aStatistically significant at 10%

bStatistically significant at 5%

cStatistically significant at 1%

In contrast, South Carolina has provided fewer funds to its only 2-year HBCU over time relative to other 2-year institutions. This result may stem from the fact that there are six 2-year public HBCUs in Alabama and only one in South Carolina. Thus, there may be more awareness of these schools or more political will to fund these types of schools in Alabama. However, our results from our main specification are unchanged when controlling for the number of HBCUs within a state. Mississippi which appropriates fewer funds to 4-year HBCUs (Table 5) does not seem to perpetuate this inequity among its 2-year schools. Louisiana provides fewer funds for 2-year HBCU independent of any business cycle effects.

In the Appendix, we reproduce Tables 5 and 6 using income per capita as a measure of macroeconomics conditions. These findings suggest that for 2-year HBCUs, only South Carolina is responsive to changes in income per capita. They respond in a procyclical manner, coinciding with the result in Table 5. For 4-year HBCUs, we see that some states are very responsive to increases in state income. Alabama, Florida, Kentucky, Maryland, Mississippi, Missouri, Ohio, Pennsylvania, South Carolina, Tennessee, Texas, Virginia, and West Virginia all increase funding to public 4-year HBCUs during times of increased state income. However, these increases are substantially lower than what HBCUs receive independent of macroeconomic conditions, all else equal. Our income per capita results (Table 10 of the Appendix) indicate that although Kentucky, Maryland, Ohio, South Carolina, and Virginia provide fewer funds to 4-year HBCUs during periods of increased state unemployment (as shown in Table 5), funding is increased during periods of higher state income per capita. Interestingly, 4-year HBCUs in North Carolina receive fewer funds during periods of higher unemployment but do not receive more funds during expansionary periods.

Potential Mechanisms

A potential mechanism that may help explain the variation in funding is state-adopted funding formulas. Unfortunately, data on funding formula specifics for each state over the time of our study are not available. The only data that documents funding formulas for every state comes from a 2012 report conducted by SRI International (2012). Over many weeks of interviews and phone calls with state officials, SRI International gathered information on each state’s funding mechanism for the year 2012. With this information, we can identify which states use a funding formula. Our results are robust to controlling for whether a state has had a funding formula. We acknowledge that this specification is likely incorrect, given that we do not exactly have changes in funding information before or after 2012. Of the 19 states with an HBCU, 14 states utilize some type of formula to determine higher education funding: Alabama, Arkansas, Florida, Georgia, Louisiana, Maryland, Mississippi, North Carolina, Ohio, Pennsylvania, South Carolina, Tennessee, Texas, and Virginia. State funding formulas typically consist of a variety of budgetary functional areas ranging from instruction and student services to Operation and Maintenance of Physical Plant (SRI international 2012). Of the 14 HBCU states that SRI International categorizes as having a funding formula, nine states provide significantly fewer funds for HBCUs. Of the seven states that do not significantly provide fewer funds to HBCUs, five have a funding formula. When examining the states that unequally fund HBCUs during downturns, five out of the six states that provide significantly less funding to HBCUs have some sort of funding formula. The exception being Kentucky, where a formula is not in place but during higher periods of unemployment Kentucky’s one HBCU is provided with fewer state funds. However, the HBCU dummy on Kentucky is positive meaning that although funding for HBCUs during downturns is higher the one HBCU in Kentucky has received more funds relative to other 4-year institutions. Therefore, Maryland, Ohio, North Carolina, South Carolina, and Virginia unequally fund HBCUs during downturns (and otherwise), and all have had funding formulas. This result does seem odd because the presence of a “formula” indicates a systematic distribution of funds where one would expect inequities to be minimized after controlling for school characteristics. However, the complexity of funding formulas varies widely from state to state. Many states that use a formula also use non-formula funds to supply operations and maintenance to special programs (SRI International 2012). Thus, it is difficult to attribute the differences in funding to these formulas.

Moreover, as discussed in the previous section, many state formulas allow for a considerable amount of autonomy as it pertains to funding. Many states’ funding of postsecondary institutions, formula or not, is subject to changes and approval of the state legislatures. Given the heterogeneity in preferences across state policy makers, it is difficult to attribute the uneven levels of funding to any set mechanism. However, the literature on state politics and funding levels suggest a difference between Republican and Democratic governors in preferences over funding and other socioeconomic outcomes. Political parties can be seen as having distinct preferences and thus influencing the mechanism by funds are appropriated. For example, Beland and Oloomi (2017) find that under Democratic governors, states spend more on health and education. Since governors have considerable influence over state budgets, we also consider whether differences in political parties may be a driver in our results. Our results are consistent across gubernatorial party lines; when restricting our analysis to states with either a Republican or Democratic Governor the funding, inequality persists among 4-year HBCUs.6 However, there is a stark contrast in the magnitude of these estimates—the disparity in appropriations is larger when a Republican governor is present, a result consistent with the previous literature.

Conclusion

We examine the funding of public HBCUs over a 29-year period. We find that when controlling for school characteristics, HBCUs receive less funding than other institutions. Additionally, when there is an economic downturn, the funding for HBCUs relative to other institutions becomes more uneven. Interestingly, our results are driven by 4-year institutions—the most common type of HBCU. When examining the results within each state specifically, we find that Florida, Kentucky, Louisiana, Pennsylvania, Tennessee, Texas, and West Virginia do not provide fewer funds HBCUs, independent of any business cycle effects. We also find that during periods of increased unemployment, Kentucky, Maryland, North Carolina, Ohio, South Carolina, and Virginia provide fewer funds for 4-year HBCUs. The remaining states appropriate less funding for HBCUs in general. One exception is Kentucky which has provided more funds for its only 4-year HBCU over the time of our sample—a result that outweighs the fact that their HBCU receives fewer funds during downturns. Moreover, among the four states that have a public 2-year HBCU, we find that Alabama appropriates more funds 2-year HBCUs, relative to other 2-year schools, during economic downturns and otherwise; South Carolina provides fewer funds to HBCUs during periods of higher unemployment. Two-year HBCUs are not differentially funded during downturns in Louisiana nor in Mississippi; however, Louisiana does generally appropriate fewer funds to 2-year HBCUs.

Given the student populations that are served by HBCUs, it is discouraging that funding disparities exist. Although most states in the sample use funding formulas, there appears to be some systematic undervaluing of HBCUs given that we control for university specific characteristics. Without examining all the funding decisions for each state for every year, it is hard to determine exactly what the prevailing issues might be. However, a common theme in our talks with state higher education representatives was that of gubernatorial and state legislative autonomy pertaining to higher education funding, even when a state formula is present. Many state formulas, which can be changed year-to-year, are in place simply as a suggestion or can be supplemented with additional funds subject to legislative approval. The persistently lower levels of funding present in our results may stem from seemingly ad hoc funding regimes that are likely influenced by state legislative preferences. Such preferences may be influenced by the fact that HBCUs primarily serve a segment of the population that is underrepresented in state legislatures. Irrespective of the funding mechanism, it is important for each state’s funding authority to recognize that these funding differentials persist.7

Footnotes

  1. 1.
  2. 2.

    Full-time equivalent enrollment is calculated by adding fall full-time enrollees plus an institution’s part-time enrollment scaled by factors determined by the U.S. Department of Education to account for the less than full-time status.

  3. 3.

    There are only eight 2-year HBCUs in our sample.

  4. 4.

    Mississippi is the only state to not respond to our inquiries.

  5. 5.

    The contact person for each institution is available upon request.

  6. 6.

    Results are available upon request.

  7. 7.

    In our discussions with the Ohio state higher education commission, this was acknowledged and recognized with additional funding to Central State University. It is not clear given the years that we have in this sample, if this has led to more equitable funding in Ohio.

Notes

Compliance with Ethical Standards

Conflict of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

References

  1. Autor DH, Katz LF, Kearney MS. Trends in US wage inequality: revising the revisionists. Rev Econ Stat. 2008;90(2):300–23.CrossRefGoogle Scholar
  2. Beland LP, Oloomi S. Party affiliation and public spending: evidence from US governors. Econ Inq. 2017;55(2):982–95.CrossRefGoogle Scholar
  3. Betts JR, McFarland LL. Safe port in a storm: the impact of labor market conditions on community college enrollments. J Hum Resour. 1995:741–65.CrossRefGoogle Scholar
  4. Card D. Estimating the return to schooling: progress on some persistent econometric problems. Econometrica. 2001;69(5):1127–60.CrossRefGoogle Scholar
  5. Clelan EB, Kofoed MS. The effect of the business cycle on freshman financial aid. Contemp Econ Policy. 2017;35(2):253–68.CrossRefGoogle Scholar
  6. Desrochers DM, & Wellman JV. Trends in college spending 1999-2009. Where does the money come from? Where does it go? What does it buy? A Report of the Delta Cost Project. Delta Project on Postsecondary Education Costs, Productivity and Accountability.2011.Google Scholar
  7. Freeman K, Cohen RT. Bridging the gap between economic development and cultural empowerment: HBCUs’ challenges for the future. Urban Education. 2001;36(5):585–96.CrossRefGoogle Scholar
  8. Humphreys BR. Do business cycles affect state appropriations to higher education? Southern Econ J, 2000 398–413.CrossRefGoogle Scholar
  9. Jaeger DA, & Page ME. Degrees matter: new evidence on sheepskin effects in the returns to education. The review of economics and statistics, 1996 733-740CrossRefGoogle Scholar
  10. Johnson MT. The impact of business cycle fluctuations on graduate school enrollment. Econ Educ Rev. 2013;34:122–34.CrossRefGoogle Scholar
  11. Kane TJ. State fiscal constraints and higher education spending: the role of Medicaid and the business cycle. 2003Google Scholar
  12. Knight B. Endogenous federal grants and crowd-out of state government spending: theory and evidence from the federal highway aid program. Am Econ Rev. 2002;92(1):71–92.CrossRefGoogle Scholar
  13. McGranahan L. State budgets and the business cycle: implications for the federal balanced budget amendment debate. Economic Perspectives-Federal Reserve Bank of Chicago. 1999;23:2–17.Google Scholar
  14. Minor JT. Contemporary HBCUs: considering institutional capacity and state priorities. University of Michigan. 2008Google Scholar
  15. Palmer R, Gasman M. "It takes a village to raise a child": the role of social capital in promoting academic success for African American men at a Black college. J Coll Stud Dev. 2008;49(1):52–70.CrossRefGoogle Scholar
  16. Sav GT. Funding historically black colleges and universities: progress toward equality? Journal of Education Finance. 2010;35(3):295–307.CrossRefGoogle Scholar
  17. SRI International. State methods for funding higher education: Nevada Legislature’s Committee to Study the Funding of Higher Education. SRI International. 2012Google Scholar
  18. Tandberg DA. Politics, interest groups and state funding of public higher education. Res High Educ. 2010;51(5):416–50.CrossRefGoogle Scholar
  19. University of Kentucky Center for Poverty Research. 2016. UKCPR National Welfare Data, 1980–2015. Gatton College of Business and Economics, University of Kentucky, Lexington, KY. http://www.ukcpr.org/data (accessed <October 29th, 2017>)

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Economics DepartmentWhitman CollegeWalla WallaUSA
  2. 2.Howard UniversityWashingtonUSA

Personalised recommendations