publisher colophon
abstract

State-level performance-based funding (PBF) policies are an increasingly common way to allocate funds to public colleges and universities. While a growing body of research has examined whether these policies are effective in improving student outcomes, little is known about how colleges respond to PBF policies. In this paper, we examine whether two-year and four-year colleges subject to PBF change their patterns and allocations of revenues, expenditures, and financial aid. We find limited evidence that colleges facing PBF receive different levels of revenue or reallocate some funds to different expenditure categories. Notably, colleges subject to PBF receive less Pell Grant revenue than colleges not subject to PBF, suggesting potential strategic behaviors targeting students from higher-income families.

introduction

Pressures for accountability in public higher education have increased dramatically over the past three decades due to concerns over the price of college and institutional productivity. Inflation-adjusted tuition and fees at public four-year colleges have risen about four percent per year over the past three decades, while increases at public two-year colleges have been about three percent per year (Baum and Ma 2013). In the meantime, the average graduation rate of students attending the average public four-year college increased by only four percentage points over the last decade and community college graduation rates increased by just 0.5% (Snyder and Dillow 2013).

States have primarily funded their higher education systems through appropriations based on the number of students enrolled and institutional missions. A report by SRI International (2012) showed that 27 states allocated [End Page 302] instructional funds to colleges based on a formula, while the remaining states used either “base plus” budgeting (giving all colleges the same percentage increase or decrease) or allocated funds as the legislature and governor saw fit. All but five states using a formula allocated funds based on enrollment, emphasizing the traditional importance of maximizing enrollment to maximize revenue.

There has been a push to align state funding of public higher education with outcomes (such as credits completed, retention, and graduation rates) over the past three decades. These accountability systems haven taken three forms (McLendon et al. 2006). The weakest form is performance reporting, in which states require public institutions to make outcome data public. Performance budgeting, a stronger form of accountability, allows state legislatures to consider outcomes in funding decisions. The strongest form of accountability is performance-based funding (PBF), in which a portion of appropriations is directly tied to outcomes. This study focuses on PBF, which is now in use in approximately 35 states (National Conference of State Legislatures 2015).

While there has been a body of research examining whether state-level PBF programs have affected student outcomes, there has been no empirical examination to this point of whether PBF policies affect both institutional expenditure and revenue patterns. Given the growing prevalence of PBF policies and continued discussions in additional states, it is important to examine whether PBF is having its intended effects. In this paper, we use institutional-level revenue and expenditure data merged with state-level characteristics on performance funding policies and other higher education practices to examine the following research questions:

  1. (1). Do PBF programs affect total institutional revenue and expenditure levels, or resource allocations within these categories?

  2. (2). Do colleges subject to PBF policies change their institutional financial aid practices compared to colleges not subject to PBF?

the state of pbf

Tennessee was the first of two states to develop performance-based funding systems in the late 1970s, with a substantial number of other states adding small PBF programs (now known as PBF 1.0) during the 1990s (Dougherty and Natow 2015). These programs represented the first effort to tie at least some funding to outcomes, but their effectiveness was limited due to the large number of measures used, data limitations, and an unclear set of policy outcomes (e.g., Layzell 1999). More than half of the PBF systems were dropped in the early 2000s due to budgetary pressures and concerns about the programs’ effectiveness (Dougherty et al. 2012). [End Page 303]

A new wave of performance-based funding systems (PBF 2.0) began spreading in the mid-2000s, moving funding into an institution’s main appropriation instead of acting as a bonus such as under PBF 1.0. This is designed to make the systems more stable through recessions (e.g. Dougherty and Reddy 2011). The amount of funding tied to outcomes has also increased, making up the majority of funding in Tennessee and Ohio and more than five percent of funding in several other states (Dougherty et al. 2014a).1 The range of outcomes measured using PBF 2.0 varies by institutional sector and mission, and can include general outcomes such as graduation rates, progress outcomes such as retention and credit completion, outcomes for student subgroups like Pell recipients and minority students, and high-need indicators including STEM degrees and placement in urban school districts (Friedel et al. 2013).

There is no clear agreement about the number of states that are currently using PBF, in part due to disagreements over what exactly counts as PBF. The National Conference of State Legislatures (2015) counted 30 states as using PBF for at least some sector of higher education in January 2015, as defined as having a system in place that ties at least some funding to performance indicators. An additional four states had approved a PBF plan, although not all details had been finalized. Snyder (2015) counted 35 states as having either developed or implemented plans for what she called “outcomes-based funding” models. This compares to the Lumina Foundation’s (2014) count of 16 states using outcomes-based funding in the 2013–14 fiscal year, where states are included if funds are appropriated and at least 25% of appropriations are tied to outcomes. When combining the three data sources, only eight states have not had at least formal discussions of adopting PBF.

The designs of performance-based funding systems vary considerably across states. A report by Snyder (2015) divided states with outcomes-based funding systems into four classifications. These classifications include the percentage of state funding allocated to outcomes, whether the desired outcomes are differentiated by mission and sector, whether extra weight is given to the outcomes of disadvantaged students, and whether the state has an explicit college completion goal. Under Snyder’s typology, Ohio and Tennessee have the most advanced funding systems, with Arkansas, Indiana, and Nevada falling into the next tier, and most systems currently in use falling into the “rudimentary” category. There does not appear to be a strong relationship between the strength of a performance funding system and metrics such as partisan political control, [End Page 304] state appropriations for higher education, or the size or types of state financial aid programs.

the literature on pbf

There is a small, but growing, body of literature examining the impacts of state PBF policies on institutional outcomes. The results with respect to graduation rates and degree completions are mixed. Several institutional-level and state-level analysis have found a range of generally modest effects (e.g, Hillman et al. 2014, 2015; Sanford and Hunter 2011; Shin 2010). Analyses of PBF programs using data from multiple states have found negative impacts on graduation in the four-year sector (Rabovsky 2012; Rutherford and Rabovsky 2014), a mix of positive, negative, and insignificant across states in the two-year and four-year sectors (Tandberg and Hillman 2013; Tandberg et al. 2014), and some evidence that longstanding PBF programs increase graduation rates (Tandberg and Hillman 2014).

The logic behind PBF programs is that colleges and universities are using their resources inefficiently, spending too much money on areas such as research and auxiliary enterprises at the expense of core instructional activities. This fits with the body of literature examining rising levels of spending on amenities in higher education (e.g., Armstrong and Hamilton 2013; Jacob et al. 2013). Resource dependence theory (Pfeffer and Salancik 1978) would suggest that colleges would be responsive to policy regimes that tie necessary funding to explicit performance metrics, and interviews with key state-level leaders in three PBF states suggest that colleges are reacting to the new incentive structures (Dougherty et al. 2014a). However, there has been little empirical research on the mechanisms behind how PBF programs could be influencing outcomes. Rabovsky (2012) examined the impact of state-level PBF programs from 1999 to 2009 and found that PBF programs had relatively small, but statistically significant, effects on institutional spending patterns. He found that the percentage of educational expenditures allocated to research dropped by 0.34 percentage points in PBF states, while instruction’s portion of total allocations increased by 0.89 percentage points.

Another important aspect of PBF policies is whether institutions attempt to change the amount and sources of revenue in order to meet the designated outcomes. Colleges could seek to devote a higher percentage of resources to instruction, while simultaneously working to increase revenue in order to fund non-instructional priorities at the same levels as before. In order to maximize revenue, institutions can raise tuition and fees, although the extent to which this is possible is often determined by state legislatures or other governing bodies. They can work to strategically allocate financial aid in order to minimize aid [End Page 305] expenditures and maximize the net price. An additional source of revenue at some public institutions is private donations, as manifested through institutional endowments. The impact of PBF on these potential revenue sources has yet to be examined to the best of our knowledge.

Colleges can simultaneously work to meet outcome thresholds in PBF systems and attempt to increase revenue by changing the composition of the student body or by altering admissions standards. While PBF 1.0 models, which used average outcomes as metrics, gave colleges an incentive to become more selective (Shulock 2011), PBF 2.0 models that take subgroup performance into account may provide incentives to recruit first-generation students, those in need of some remediation, and Pell Grant recipients under certain circumstances (McKinney and Hagedorn 2015). But even these more advanced systems may still be subject to gaming. Umbricht, Fernandez, and Ortagus (forthcoming) examined Indiana’s PBF 2.0 system and found that colleges became more selective in response to PBF and did not increase degree completions compared to neighboring states. Additionally, PBF systems may provide incentives to colleges to weaken standards in order to increase completion rates (Dougherty et al. 2014b), although PBF 2.0 systems are still too new to rigorously evaluate that possibility.

data, sample, and methods

To explore whether state performance-based funding systems affected colleges’ revenue and expenditure patterns, we used a ten-year panel of state and sector-specific PBF policies combined with per-student revenue and expenditure data and state-level financial and political characteristics over the corresponding period.

Data

Our analysis included data from multiple sources for all public two-year and four-year institutions from the 2003–04 through the 2012–2013 academic years. Data on the majority of institutional-level financial measures are from the Integrated Postsecondary Education Data System (IPEDS). These measures were adjusted for inflation into 2012 dollars using the Consumer Price Index and trimmed (Winsorized) at the 1st and 99th percentiles (Tukey 1962). Per-full time equivalent (FTE) revenue measures included total revenue, net tuition and fee revenue, state and local appropriations, and revenue from auxiliary enterprises such as residence halls, dining facilities, and athletics. Per-FTE expenditure measures included total expenditures, instruction, research, public service, [End Page 306] student services, institutional support (a proxy for administrative expenditures), and auxiliary enterprises.2 For a small number of institutions which report revenue and expenditures at the system level instead of the institutional level (see Jaquette and Parra (2014) for more details), we allocated revenues and expenditures equally on a per-FTE basis across branch campuses.

Because colleges may respond to performance-based funding systems by trying to recruit students with a higher likelihood of succeeding by using grant aid, we examined per-FTE institutional grant aid measures from IPEDS. Grant aid is divided into funded grants (which come from dedicated resources such as the endowment or annual donations) and unfunded grants (which are forgone tuition). We also calculated a tuition discount rate, which is the ratio of unfunded grant aid to gross tuition revenue before discounts are considered. However, IPEDS data do not distinguish between merit-based aid and need-based aid, obscuring any changes in the types of aid offered to students that could be attributed to PBF. For example, we cannot observe whether colleges in states with PBF systems offer additional institutional merit aid in order to attract students with a higher probability of success. The final set of institutional financial characteristics reflects the potential ability of the college to raise additional funds and how students are paying for their education. Listed tuition and fees and per-FTE endowment values are from IPEDS, while per-FTE Pell Grant and student loan revenue are from the U.S. Department of Education’s Office of Federal Student Aid.

We included three sets of state-level factors that could potentially influence the effects of PBF policies. The first set of factors includes measures of state grant aid for undergraduate students (adjusted for inflation) and the percentage of aid based on financial need instead of merit from annual surveys administered by the National Association of State Student Grant and Aid Programs. The grant aid amount was divided by the number of adults between ages 18 and 24 in the state (from the Census Bureau) and adjusted for inflation to create a measure of state financial aid effort in addition to funding via state appropriations. The second set of factors is based on a set of surveys on state-level tuition and fee policies from the State Higher Education Executive Officers Association. These surveys were given to state fiscal officers in the 2002–03, 2005–06, and 2010–11 academic years during the period of this study, with the most recent year of data used in between survey waves. We used measures of whether a state had imposed tuition and/or fee caps in the previous three years. The third set of factors represents partisan control of the state House, Senate, and governor’s office during the period of the study, with Nebraska excluded on account of its [End Page 307] unique unicameral legislature. These data came from a database compiled by Carl Klarner at Indiana State University, supplemented with information from the National Conference of State Legislatures.

Data on state-level and sector-level PBF programs came from a number of sources, as no central historical database of PBF policies currently exists. We compiled a dataset based on prior research (Dougherty and Reddy 2013; Gurbonov 2013; Hillman et al. 2013; Lumina Foundation 2014; National Conference of State Legislatures 2014), supplemented with searches on state and system websites. When sources disagreed about whether a state had a PBF system in a given year, we typically counted a state as having a system to maintain consistency. Further, our study did not make a distinction between PBF 1.0 and PBF 2.0 systems, as there is no agreed-upon distinction between the two systems and most states allocated less than 5% of funding to performance metrics during the period of this study. However, dropping Tennessee and Ohio (the two states with the highest percentage of funds allocated based on performance) does not change the findings reported later in this article.3

Sample

Institutions are included in our sample if they are public, offer associate’s degrees or bachelor’s degrees, participated in federal financial aid programs, and were active during the 2012–13 academic year. This notably excludes many public technical colleges which grant certificates instead of associate’s degrees, but the different lengths of their academic programs and varying tuition structures make comparisons with community colleges difficult. The analytic sample consists of 1,649 public colleges and universities in the United States, including 538 public four-year institutions and 1,113 public two-year institutions in all 50 states.

Summary statistics of the four-year and two-year colleges in our sample can be found in Table 1. The median four-year college received $21,172 in revenue per FTE in the 2012–13 academic year, with $6,237 from net tuition, $5,285 in state and local appropriations, $2,491 in auxiliary enterprises, and the remainder from other sources such as research and hospitals. Two-year colleges received $11,959 in revenue, with $1,927 coming from tuition and fees, $4,792 from appropriations, and $362 in auxiliary enterprises. Median per-FTE expenditures in both sectors were slightly lower than median revenues, with less than half of total expenditures ($7,564 at four-year colleges and $4,882 at two-year colleges) devoted to instruction. Student services and institutional support were substantial revenue sources in both sectors ($3,868 combined for the median [End Page 308]

Table 1. Summary statistics of the institutions in the dataset (2012-academic year)
Click for larger view
View full resolution
Table 1.

Summary statistics of the institutions in the dataset (2012-academic year)

[End Page 309]

four-year college and $2,758 for the median two-year college), while auxiliary enterprises were a larger expenditure at four-year colleges ($2,913) than two-year colleges ($618).

Institutional grant aid plays a far larger role at four-year colleges, with a median tuition discount rate of 11.8% compared to 2.2% at two-year colleges; the majority of grants are forgone revenue instead of funded by the endowment or other sources. Given higher listed tuition prices at four-year colleges ($7,002 versus $2,987), it is not surprising that the median four-year college received more than twice the tuition revenue from student loans than the typical two-year college ($5,765 versus $2,062). However, the median community college received more Pell revenue per FTE than the typical four-year college ($2,336 versus $1,553).

Across four-year and two-year institutions, the typical public college operates in a state where grant aid is based entirely on financial need instead of prior academic achievement. The typical state spent roughly $300 in grant aid per young adult, with substantial variation across states. Roughly half of all colleges are in states with tuition caps in place, while about one in four colleges operated in states with a fee cap. A majority of colleges are in states with Republican legislative or executive control in 2012, although many of these states had Democratic control of at least one body. Only 32% of four-year colleges and 42% of two-year colleges are located in states with active performance-based funding systems as of the 2012–13 academic year, although an additional 15% of colleges in each sector had been subject to PBF at some point over the previous decade.

We next compared two-year and four-year colleges in states with PBF systems in the 2012–13 academic year to colleges in states without PBF (Table 2). As there are regional differences in the states that have adopted PBF (with more Southern states and fewer Northeastern states having active PBF policies), there are also differences in institutional resources and state characteristics by the presence of PBF. Four-year colleges in states with PBF receive about $4,000 less in revenue per FTE than colleges in non-PBF states, with a difference of about $2,000 for community colleges. About $1,000 of the difference is due to higher state and local appropriations in non-PBF states, but net tuition revenue is not significant at p<.05. Notably, there is no statistically significant difference in instructional expenditures between PBF and non-PBF states, although there are consistent differences in student service and institutional support spending.

There are some differences in state-level characteristics between colleges located in states with PBF policies in 2012–13 to colleges in states without PBF policies. Four-year colleges in states with PBF had a lower percentage of state grant aid awarded via financial need (69%) than states without PBF (82%). Community colleges in states with PBF were twice as likely to be operating [End Page 310]

Table 2. Comparison of PBF and non-PBF states, 2012-academic year
Click for larger view
View full resolution
Table 2.

Comparison of PBF and non-PBF states, 2012-academic year

[End Page 311]

under a fee cap (41% versus 19%). Finally, PBF states were more likely to have Republican legislative or executive control, with the magnitude of the difference generally around 20 percentage points.

Methods

Our study utilized panel regressions with institutional and year fixed effects to examine whether PBF policies affect institutional expenditure or revenue patterns.4 The model (run separately for two-year and four-year institutions) is the following for college j in year t:

inline graphic

Outcome represents the per-FTE expenditure or revenue pattern of interest in year t. PBF indicates the presence of a performance-based funding policy one year prior to the outcomes being measured. This one-year lag reflects the typical decision-making process in higher education, as institutional budgets are often set several months in advance of the beginning of the following fiscal year. Therefore, a performance-based funding policy in effect in a given year is likely to influence institutional behavior the most in the following year. We estimated models with both no lag period and a two-year lag as robustness checks and present these results in the Appendix.

StateFin reflects state-level financial characteristics including the percentage of state grant aid based on need, the average amount of grant aid available per young adult, and whether the state had a tuition and/or fee cap in the last three years, StatePol represents state-level partisan political control, is a set of year-level fixed effects, is the institutional-level effect, and is an idiosyncratic error term. We added the variables in three blocks, with the first model including just the PBF indicator and year/institution fixed effects, the second model adding state financial characteristics, and the third model adding state political characteristics. Note that research expenditures, public service expenditures, and per-FTE endowments are only examined for four-year institutions since they are typically not applicable in the public two-year sector.

Limitations

The most substantial limitation of this study, and of research on performance-based funding in general, is that it is difficult to agree upon a definition of what should count as performance-based funding. Because funding for implemented systems is not always provided by state legislatures, it can be difficult to determine whether a PBF policy should be considered ‘in effect’ in a given year. The various [End Page 312] types of PBF policies, including the percentage of funds tied to outcomes, whether colleges have some ability to choose their own measures, and whether different weights are given to the success of different types of students, also complicate any analysis of the effectiveness of PBF policies at the national level.

An additional limitation of considering revenue and expenditure data from IPEDS is that the U.S. Department of Education’s guidance to college regarding what should be included in each revenue and expenditure category is often unclear. For example, information technology expenditures used to support student services may be classified under student services or institutional support, depending on how a college chooses to construct a budget (National Center for Education Statistics n.d.). As such, the potential effects of PBF on any particular revenue or expenditure category should be interpreted with caution.

results

We first examined the relationship between the presence of a PBF policy and institutional-level revenue, expenditures, grant aid, and institutional financial characteristics among four-year institutions using a one-year lag between the presence of a PBF policy and examining outcomes (Table 3). Across each of the three blocked regression models, there was no statistically significant relationship between the presence of PBF and total revenue or total expenditures after including year and institution fixed effects. After controlling for state financial and political characteristics (Model 3), the presence of PBF is associated with a small and marginally significant $107 increase in state and local appropriations (p<.10). This result is larger and more highly significant when no lag period is used in alternative models (Appendix 1), but becomes negative and insignificant when using a two-year lag period This suggests that as states adopt PBF, they are making at least some additional funds available to institutions instead of completely including PBF in the base budget, but these funds fade away over time. However, PBF 2.0 policies seek to include the entire portion of appropriations allocated to PBF in the base budget, so this result may not hold in future years.

Across all expenditure categories, there is no evidence of a relationship between the presence of PBF policies and changes in expenditures when using a one-year lag. When no lag period is used (Appendix 1), the presence of PBF is associated with a $161 decline in instructional expenditures, a $34 decrease in student services expenditures, and a $63 increase in auxiliary enterprise expenditures (all p<.05). However, these coefficients are all significant in the opposite direction with a two-year lag, suggesting a shift to higher instruction and student services spending and less auxiliary enterprise spending in states with PBF.5 [End Page 313]

Table 3. Regression results examining relationship between PBF and institutional financial characteristics (four-year colleges).
Click for larger view
View full resolution
Table 3.

Regression results examining relationship between PBF and institutional financial characteristics (four-year colleges).

There is some evidence that four-year colleges in states with PBF policies changed their recruitment strategies in order to recruit different types of students. Colleges subject to PBF in the previous year spent $43 more in unfunded grant [End Page 314] aid than colleges not subject to PBF (p<.05), or approximately 5% of the median college’s unfunded grant aid. Colleges facing PBF received approximately $30 less per FTE in Pell Grant revenue than colleges in non-PBF states, representing about 2% of typical per-FTE Pell Grant revenue in the 2012–13 academic year. This appears to represent a slight shift toward enrolling students from higher-income families, which could be a function of enrollment management policies or student preferences between public four-year colleges in states subject to PBF compared to other institutions. This result holds using a two-year lag period as well as a one-year lag (Appendix 1).

Table 4 contains the results examining the relationship between PBF policies and institutional financial characteristics for two-year public colleges. Similar to the four-year sector, there is no statistically significant relationship between being subject to a PBF policy and total per-FTE revenue or expenditure after including year and institution fixed effects. There is also some suggestive evidence that being subject to PBF might result in slightly lower net tuition and fee revenue after including state-level characteristics in Model 3 ($56 difference, p<.10), as well as lower listed tuition and fees ($19 difference, p<.10). Colleges facing PBF also received $26 less in auxiliary enterprise revenues than colleges without PBF (p<.05). These results generally held using a two-year lag (Appendix 2), although total revenue is positive and significant when using a two-year lag ($670 difference, p<.05).

Although student services expenditures at colleges with PBF were slightly higher ($24 difference, p<.10), there were no other differences in expenditures by PBF status using the preferred one-year lag. Using a two-year lag (Appendix 2), total expenditures and instructional expenditures were also positive and significant. Consistent with the results from four-year colleges, two-year colleges subject to PBF received approximately $39 less in Pell Grant revenue per FTE (p<.01), suggesting again the possibility of targeting students with less financial need.

discussion and future work

State performance-based funding systems have increased in number and complexity over the last decade, in addition to having greater financial stakes tied to meeting performance metrics. Yet most research examining the implications of PBF has focused on whether these policies have affected graduation rates or the number of completions without considering whether colleges might be reassessing their financial priorities in light of these policies. In this study, we explored whether colleges in states with PBF policies have worked to change or reallocate revenues and expenditures across different categories as a mechanism to improve outcomes as well as whether colleges might be adjusting their [End Page 315] financial aid practices in order to recruit certain types of students that may result in extra funds.

Table 4. Regression results examining relationship between PBF and institutional financial characteristics (two-year colleges).
Click for larger view
View full resolution
Table 4.

Regression results examining relationship between PBF and institutional financial characteristics (two-year colleges).

We find some evidence that four-year colleges subject to PBF policies received somewhat more in appropriations than those not subject to PBF, but this finding fades out when a longer time horizon is used. Two-year colleges facing PBF charged slightly lower tuition and received less tuition revenue, but also spent slightly more on student services than other colleges. Both two-year and four-year colleges subject to PBF received less Pell Grant revenue per student—a key sign of enrollment of students from lower-income families. The increase in [End Page 316] unfunded grant aid, which is often merit-based, at four-year colleges in PBF states relative to other colleges also suggests colleges may be trying to recruit more students from higher-income families via the strategic use of institutional financial aid.

These results suggest that public colleges that face at least some funding being tied to outcomes do change their expenditure patterns and potentially even the composition of their student body. Further research is needed to explain whether these observed relationships are truly a result of institutional behaviors, or whether they are a result of unobserved factors common to colleges in states with performance-based funding systems. In particular, additional qualitative research is needed to explore the extent to which the desire to reach the outcome levels required for funding under PBF is an implicit or explicit part of a college’s budgeting process. This is particularly true for states in which colleges can influence or directly choose some of the metrics through which their performance will be assessed.

An additional area for future research is examining whether the amount of funds at stake in PBF programs influences colleges’ financial decisions. During the period of time examined in our analyses, most colleges had between one and five percent of total appropriations tied to various performance metrics, occasionally with “stop loss” provisions that limited annual changes in funding based on performance. Going forward, it is possible that more states follow Tennessee and Ohio by tying a higher percentage of state funding to outcome metrics. This means that future research should consider whether increasing the percentage of state funding tied to performance by one percent has any effects on how colleges allocate resources or recruit students.

Further research is also needed to examine whether PBF policies result in changes in the characteristics of a public college’s student body, which we find some potential evidence of in this research. Further research should consider the key characteristics of race/ethnicity, gender, family income, and academic preparation. There are incentives in some states’ PBF systems (particularly under PBF 1.0) for colleges to become more selective and enroll students with a higher probability of success, which could have implications for college access and equity. While interviews with key stakeholders suggest colleges may be acting to becoming more selective where possible (Dougherty et al. 2014b) and research from Indiana suggests colleges are becoming more selective in response to the state’s PBF 2.0 system (Umbricht et al. 2015), empirical research is needed to determine whether a substantial number of colleges are acting in this manner. Additionally, further research should examine whether these attempts to become more selective actually result in colleges gaining a larger share of available performance funding dollars. [End Page 317]

Robert Kelchen and Luke J. Stedrak

Robert Kelchen, PhD, is the Assistant Professor of Higher Education and Luke J. Stedrak, EdD, is the Assistant Professor of Education Leadership, Management and Policy at Seton Hall University in South Orange, NJ.

Appendix 1. Regression results examining relationship between PBF and institutional financial characteristics with different lag periods (four-year colleges)

SOURCES: See Table 1. NOTES: (1) State financial characteristics include the amount of state grant aid (merit and need), as well as whether a tuition or fee cap existed in the last three years. (2) State political characteristics include partisan control of the state House, Senate, and governor’s office. Nebraska is excluded on account of its unicameral legislature. (3) *represents p<.10, **represents p<.05, and ***represents p<.01.
Characteristic No lag 1-year lag 2-year lag
Coef. (SE) Coef. (SE) Coef. (SE)
Per-FTE revenues ($)
    To t a l 275 (300) 362 (283) 499* (280)
    Tuition and fees -87* (46) -44 (43) 35 (43)
    State/local appropriations 314*** (66) 107* (62) -93 (61)
    Auxiliary enterprises 34 (27) -33 (26) -121*** (25)
Per-FTE expenditures ($)
    To t a l 86 (236) 30 (222) 31 (220)
    Instruction -161** (62) 23 (59) 104* (59)
    Research -63 (52) -22 (50) 26 (49)
    Public service -6 (21) 3 (19) 13 (19)
    Student services -34** (15) 6 (15) 35** (14)
    Institutional support 1 (33) 37 (31) 36 (30)
    Auxiliary enterprises 63** (32) -15 (30) -69** (30)
Institutional grant aid
    Funded grants ($) 1 (8) 2 (7) 1 (7)
    Unfunded grants ($) 16 (18) 43** (17) 74*** (17)
    Tuition discount rate (pct) -0.2 (0.2) 0.4* (0.2) 0.8*** (0.2)
Institutional fnancial characteristics
    Listed tuition and fees ($) -6 (36) -14 (34) -11 (33)
    Student loans per FTE ($) 11 (40) 22 (37) 11 (37)
    Pell Grants per FTE ($) -31*** (11) -32*** (11) -33*** (11)
    Per-FTE endowment ($) -207 (235) -156 (221) -17 (219)
Maximum sample size 532 532 532
Year and institutional FEs? Yes Yes Yes
State fnancial characteristics? Yes Yes Yes
State political characteristics? Yes Yes Yes

[End Page 318]

Appendix 2. Regression results examining relationship between PBF and institutional financial characteristics with different lag periods (two-year colleges).

SOURCES: See Table 1. NOTES: (1) State financial characteristics include the amount of state grant aid (merit and need), as well as whether a tuition or fee cap existed in the last three years. (2) State political characteristics include partisan control of the state House, Senate, and governor’s office. Nebraska is excluded on account of its unicameral legislature. (3) *represents p<.10, **represents p<.05, and ***represents p<.01.
Characteristic No lag 1-year lag 2-year lag
Coef. (SE) Coef. (SE) Coef. (SE)
Per-FTE revenues ($)
    Total -78 (346) 199 (332) 670** (325)
    State/local appropriations -97 (91) -17 (87) 99 (85)
    Tuition and fees -52* (30) -56* (29) -61** (28)
    Auxiliary enterprises 3 (14) -26** (13) -25* (13)
Per-FTE expenditures ($)
    Total -68 (326) 184 (313) 588* (306)
    Instruction -38 (98) 71 (94) 198** (92)
    Student services -7 (15) 24* (15) 44*** (14)
    Institutional support -31 (34) -27 (32) -15 (32)
    Auxiliary enterprises -2 (14) -10 (13) -16 (13)
Institutional fnancial characteristics
    Listed tuition and fees ($) 26** (10) -19* (9) -61*** (10)
    Student loans per FTE ($) 70*** (26) 22 (25) 29 (25)
    Pell Grants per FTE ($) -11 (12) -39*** (12) -32*** (11)
Maximum sample size 1,106 1,106 1,106
Year and institutional FEs? Yes Yes Yes
State fnancial characteristics? Yes Yes Yes
State political characteristics? Yes Yes Yes

[End Page 319]

Bibliography

Armstrong, Elizabeth A., and Laura T. Hamilton. Paying for the Party: How College Maintains Inequality. Cambridge, MA: Harvard University Press, 2013.
Baum, Sandy, and Jennifer Ma. Trends in College Pricing. Washington, DC: The College Board, 2013.
Dougherty, Kevin J., Sosanya S. Jones, Hana Lahr, Rebecca S. Natow, Lara Pheatt, and Vikash Reddy. “Performance Funding for Higher Education: Forms, Origins, Impacts, and Futures.” The ANNALS of the American Academy of Political and Social Science 655, no. 1 (September 2014a): 163-184.
Dougherty, Kevin J., Sosanya S. Jones, Hana Lahr, Rebecca S. Natow, Lara Pheatt, and Vikash Reddy. Implementing Performance Funding in Three Leading States: Instruments, Outcomes, Obstacles, and Unintended Impacts. New York, NY: Community College Research Center Working Paper No. 74, 2014b.
Dougherty, Kevin J., Sosanya S. Jones, Hana Lahr, Rebecca S. Natow, Lara Pheatt, and Vikash Reddy. The Political Origins of Performance Funding 2.0 in Indiana, Ohio, and Tennessee: Theoretical Perspectives and Comparisons with Performance Funding 1.0. New York, NY: Community College Research Center Working Paper No. 68, 2014.
Dougherty, Kevin J., and Rebecca S. Natow. The Politics of Performance Funding for Higher Education: Origins, Discontinuations, and Transformations. Baltimore, MD: Johns Hopkins University Press, 2015.
Dougherty, Kevin J., Rebecca S. Natow, and Blanca E. Vega. “Popular but Unstable: Explaining Why State Performance Funding Systems in the United States Often Do Not Persist.” Teachers College Record 113, no. 3 (November 2012): 1-41.
Dougherty, Kevin J. and Vikash Reddy. The Impacts of State Performance Funding Systems on Higher Education Institutions: Research Literature Review and Policy Recommendations. New York, NY: Community College Research Center Working Paper No. 37, 2011.
Dougherty, Kevin J. and Vikash Reddy. “Performance Funding for Higher Education: What are the Mechanisms? What are the Impacts?” ASHE Higher Education Report 39, no. 2 (2013): 1-134.
Friedel, Janice N., Zoe M. Thornton, Mark M. D’Amico, and Stephen G. Katsinas. Performance-Based Funding: The National Landscape. Tuscaloosa, AL: The University of Alabama Education Policy Center, 2013.
Gorbunov, Alexander V. Performance Funding in Public Higher Education: Determinants of Policy Shifts. Unpublished doctoral dissertation, Vanderbilt University, 2013.
Hillman, Nicholas W., David A. Tandberg, and Alisa H. Fryar. Evaluating the Impacts of “New” Performance Funding in Higher Education. Educational Evaluation and Policy Analysis, 37, no. 4 (December 2015): 501-519.
Hillman, Nicholas W., David A. Tandberg, and Jacob P. K. Gross. “Performance Funding in Higher Education: Do Financial Incentives Impact College Completions?.” The Journal of Higher Education 85, no. 6 (November/December 2014): 826-857.
Jacob, Brian, Brian McCall, and Kevin M. Stange. College as Country Club: Do Colleges Cater to Students’ Preferences for Consumption? Cambridge, MA: NBER Working Paper 18745, 2013.
Jaquette, Ozan, and Edna E. Parra. “Using IPEDS Data for Panel Analyses: Core Concepts, Data Challenges, and Empirical Applications.” In Higher education: Handbook of theory and research (Vol. 29), edited by Michael B. Paulsen, 467-533. Dordrecht, The Netherlands: Springer, 2014.
Layzell, Daniel T. “Linking Performance to Funding Outcomes at the State Level for Public Institutions of Higher Education: Past, Present, and Future.” Research in Higher Education 40, no. 2 (April 1999): 233-246.
Lumina Foundation (2014). “How Much are States Investing in Outcomes-Based Funding?.” Strategy Labs. Last modified December 14, 2014. http://strategylabs.luminafoundation.org/wp-content/uploads/2013/03/OBF-Table_Notes.pdf. [End Page 320]
McKinney, Lyle, and Linda S. Hagedorn. Performance-Based Funding for Community Colleges in Texas: Are Colleges Disadvantaged by Serving the Most Disadvantaged Students? Bryan, TX: Greater Texas Foundation, 2015.
McLendon, Michael K., James C. Hearn, and Russ Deaton. “Called to Account: Analyzing the Origins and Spread of State Performance-Accountability Policies for Higher Education.” Educational Evaluation and Policy Analysis 28, no. 1 (Spring 2006): 1-24.
National Center for Education Statistics. “Integrated Postsecondary Education Data System-Glossary.” Retrieved March 26, 2015. http://nces.ed.gov/ipeds/glossary/.
National Conference of State Legislatures. 2015. “Performance-Based Funding for Higher Education.” Retrieved June 24, 2015. http://www.ncsl.org/research/education/performance-funding.aspx.
Pfeffer, Jeffrey, and Gerald R. Salancik. The External Control of Organizations: A Resource Dependence Perspective. New York, NY: Harper Row, 1978.
Rabovsky, Thomas M. “Accountability in Higher Education: Exploring Impacts on State Budgets and Institutional Spending Patterns.” Journal of Public Administration Research and Theory 22, no. 4 (October 2012): 675-700.
Rutherford, Amanda, and Thomas Rabovsky. “Evaluating Impacts of Performance Funding Policies on Student Outcomes in Higher Education.” The ANNALS of the American Academy of Political and Social Science 655, no. 1 (September 2014): 185-208.
Sanford, Thomas, and James M. Hunter. “Impact of Performance-Funding on Retention and Graduation Rates.” Education Policy Analysis Archives 19, no. 33 (November 2011): 1-30.
Shin, Jung C. “Impacts of Performance-Based Accountability on Institutional Performance in the U.S.” Higher Education 60, no. 1 (July 2010): 47-68.
Shulock, Nancy. Concerns About Performance-Based Funding and Ways that States are Addressing the Concerns. Sacramento, CA: California State University Sacramento Institute for Higher Education Leadership & Policy, 2011.
Snyder, Martha. Driving Better Outcomes: Typology and Principles to Inform Outcomes-Based Funding Models. Washington, DC: HCM Strategists, 2015.
Snyder, Thomas D., and Sally A. Dillow. Digest of Education Statistics 2012. Washington, DC: Institute of Education Sciences, National Center for Education Statistics, 2013.
SRI International. States’ Methods of Funding Higher Education. Menlo Park, CA: Author, 2012.
Tandberg, David A., and Nicholas W. Hillman. State Performance Funding for Higher Education: Silver Bullet or Red Herring? Madison, WI: WISCAPE, 2013.
Tandberg, David A., and Nicholas W. Hillman. “State Higher Education Performance Funding: Data, Outcomes, and Policy Implications.” Journal of Education Finance 39, no. 3 (Winter 2014): 222-243.
Tandberg, David A., Nicholas Hillman, and Mohamed Barakat. “State Higher Education Performance Funding for Community Colleges: Diverse Effects and Policy Implications.” Teachers College Record 116, no. 12 (2014): 1-31.
Tukey, John W. “The Future of Data Analysis.” The Annals of Mathematical Statistics 33, no. 1, (1962): 1-67.
Umbricht, Mark. R., Frank Fernandez, and Justin C. Ortagus. “An Examination of the (Un) Intended Consequences of Performance Funding in Higher Education.” Educational Policy (forthcoming). [End Page 321]

Footnotes

1. In a sense, revised systems in Tennessee and Ohio represent a third wave of PBF systems, in which a large percentage of funds are tied to performance indicators instead of less than 10% of overall appropriations. This shift happened outside the period of this study, but future work should consider whether higher-stakes systems should be considered to be a separate PBF 3.0 wave.

2. Research and public service expenditures were excluded for two-year public colleges due to many colleges reporting zero or trivial expenditures in these categories.

3. Regression coefficients from models excluding Tennessee and Ohio are available upon request from the first author.

4. Results from models without institutional fixed effects are available upon request from the first author.

5. Results of a falsification test using a one-year lead period (using the PBF policy in place one year after outcomes were measured) are similar to the results with no lag or lead period, and are available upon request from the first author.

Additional Information

ISSN
1944-6470
Print ISSN
0098-9495
Pages
302-321
Launched on MUSE
2016-04-06
Open Access
No
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.