Abstract
Regulatory enforcement is a policy decision in itself. Given the current federal commitment to deregulate the for-profit college market, state regulations will likely become increasingly important. This study examines the factors contributing to the enforcement of state-level proprietary college regulations. Using event history analysis, the authors test diffusion and innovation theories as potential explanations for patterns of enforcement. The findings suggest that geographic diffusion plays an important role in state lawsuits against for-profit colleges, although the specific mechanisms of diffusion may differ across regions. Moreover, loan repayment rates of for-profit students appear to contribute to the likelihood of a lawsuit being filed as states may seek to intervene when objective metrics suggest that students and state economic interests seem to be harmed. The article also suggests that market factors may play an influential role in state governments enforcing regulations on for-profit colleges.
For-profit colleges and universities (FPCUs) have a long history in the United States and have taken many forms. It was after these proprietary schools gained access to Federal Student Aid in the 1970s that the sector experienced rapid growth. With this growth came a number of scandals that involved fraud, profiteering, and deceptive recruitment tactics (Angulo, 2016). In response to exploitative FPCUs taking advantage of students, the 1992 reauthorization of the Higher Education Act (HEA) included several regulations seeking to safeguard students while still more are created through negotiated rulemaking (Natow, 2015; Serna, 2014). But not all regulations are equal because some are spelled out in the HEA whereas others are left to the discretion of the triad: accreditors, the U.S. Department of Education, and state governments.
Accrediting bodies are supposed to be a peer-reviewed process ensuring quality in educational enterprises, the Department of Education administers federal programs and ensures compliance with associated rules, and state governments provide the legal authority for an institution to operate in that given state (Pelesh, 2010). This triad is meant to ensure that students have access to quality postsecondary institutions. However, the Nunn Report of 1991, a U.S. Senate report produced just prior to the 1992 reauthorization of the HEA, pointed out failures of all three parts of the triad (Beaver, 2012). Driven by far-reaching reports of waste, fraud, and abuse in the Federal Student Aid programs, the report prompted Congress to pass stricter regulations on the postsecondary market, and particularly, for-profit colleges. In this report, state governments were specifically criticized for lax oversight of the market. This criticism serves as a focusing event and the starting point for the present study. The current iteration of the for-profit industry began around this time and is characterized by an increase in venture capital and organizational size (Tierney & Hentschke, 2007). These types of institutions have persisted through the past three decades and are often beholden to external financial stakeholders. The outside financial obligations have helped create an environment that induces exploitative behaviors at the expense of students. Many of the same consumer protection problems identified in the Nunn Report remain today, but the growth in student loan debt and default along with recent closures of for-profit colleges makes the triad even more important in regulating the higher education marketplace.
Background
Federal regulation of higher education, and the for-profit market in particular, has been a focal point of debate in the politics of education in recent decades. Regulations that monitor cohort default rates (CDRs), the proportion of revenue from Title IV sources (the 90/10 rule), and the employment status of program graduates (the Gainful Employment Rule) are all part of the federal regulatory apparatus. However, these rules do not come to fruition in a universal manner. Some are codified through the passage of new laws or the reauthorization of major acts, like the HEA. For example, for-profit colleges are barred from linking recruiters’ pay to the number of students they enroll. This “incentive compensation ban” is a federal statute first included in the 1992 reauthorization of the HEA, and then reaffirmed in each subsequent reauthorization (Beaver, 2012). Other rules, however, are created through a rulemaking process which reflects a federal agency’s ability to create or change legal regulations pertaining to programs administered by that agency (Kerwin & Furlong, 2010). One example of a rule resulting from such a process is the Gainful Employment Rule of 2011 which states that career-oriented programs must successfully place students into positions enabling them to pay off their federal education loans (Natow, 2015; Serna, 2014).
In addition to federal rules, state legislatures also create localized laws and regulations governing postsecondary institutions—some specifically targeting FPCUs. Localized rulemaking includes the authorization by the state for colleges to operate in a given jurisdiction. These authorizations are often supplemented by interstate articulation agreements that allow FPCUs to use the authorization by one state to operate in another. These agreements are often referred to as a State Authorization Reciprocity Agreement. State governments also generally have their own consumer protection laws that may be broadly applicable to FPCUs as companies, or specifically targeted at proprietary institutions. For example, California created the Bureau for Private Postsecondary Education which is tasked with ensuring educational standards and protecting the state and students from fraud through meaningful oversight (Taylor, 2013). In the academic literature, the role of the state has received minimal attention.
In the current political climate, state regulations may become increasingly important. Under Secretary DeVos’s supervision, the Trump administration has promised to roll back federal regulations on for-profit colleges and has already weakened the enforcement of some rules (Waldman, 2017). If federal regulations are dismantled or not enforced, state-level rules will likely begin to play a more important role in the regulatory environment for proprietary schools. However, the enforcement of state-level rules will be an important factor in how much impact these laws actually have. The enforcement of regulation as policy implementation is decidedly an important and nuanced understanding of this part of the policy process. Thus, we discuss the enforcement of regulation as an explicit and discrete aspect of policy.
The existence of regulations can act as a form of symbolism that signals a government’s intentions and commitment to the goals of a given law (Firth, 1973), and is specifically salient for education policy (Rosen, 2009). This symbolism is embodied by laws named after individuals (e.g., a gun control law named after James Brady), and can demonstrate a commitment to a specific cause. However, a lack of enforcement of a given regulation can also act as a symbolic gesture undermining the spirit of the law (Stone, 2011). For example, President Obama’s directive to limit the enforcement of immigration laws for students was a symbolic gesture that signified an appreciation of the contributions undocumented students make to American society. Enforcement, within higher education policy, is complicated by vagueness at the federal level, the use of the negotiated rulemaking process, and the existence of the triad. These problems underscore the need for each part of the triad to be strong. State governments, in particular, can be the purveyors of locally tailored regulation, enforcement, and quality control in postsecondary markets.
State governments behave similarly and use regulations and enforcement policies as symbols. While state-level regulations of for-profit colleges are nearly ubiquitous, enforcement is not. As such, the level of enforcement may symbolize a state government’s commitment to safeguarding students from predatory for-profit colleges. Moreover, regulation without enforcement undermines the effectiveness of such policies (Tombs, 2016). This study focuses on regulatory enforcement as a form of policy, and elucidates contributing factors.
As it is unclear what role the federal government will play in regulating the for-profit market in the coming years, it behooves higher education researchers to better understand state regulations, and whether or how these laws are enforced. Enforcement is not simply an on–off switch; the temporal nature is similar to a policy that must be renewed or reauthorized frequently. As part of the implementation process, enforcement is integral to meeting policy goals (Van Meter & Van Horn, 1975). Enforcement, as part of the policy process, can take two forms: active monitoring and reactive responses to specific complaints or issues (McCubbins & Schwartz, 1984). This study seeks to explore the latter type, which can be identified in the context of for-profit regulation as formal legal sanctions at the state level. We conceptualize enforcement as the filing of a lawsuit by a state attorney general (AG). This article seeks to understand what factors influence the enforcement of state regulations at a given point in time. We assess what internal and external variables contribute to state AGs bringing lawsuits against for-profit colleges. These lawsuits cover a range of issues including consumer protection, fraudulent filings, student aid abuses, and recruitment violations. We examine state political and economic factors, FPCU performance, and the potential for regulatory competition and diffusion among states. While scholars have examined the history (Angulo, 2016), framing (Natow, 2015; Serna, 2014), and impact (Darolia, 2013; Ward, 2017) of federal regulation, less is known about the roles states play in regulating the market. Public press outlets have covered the expansion of state consumer protection offices to include for-profit students and the large number of lawsuits brought against FPCUs by state and federal prosecutors, but there is little academic research. Like any other law, enforcement is circumstantial and may not be consistent. Of consequence, we explore here the potential factors contributing to enforcement, and seek to understand the relationship between internal and external forces.
To underscore the importance of regulating these schools, we couch this study in the current research about the behavior of FPCUs and outcomes of their students. We draw on the political science theories of policy innovation and diffusion to examine the enforcement of state regulations as a form of policy implementation. Given the commitment of the Trump administration to reducing federal oversight, state regulations are likely to play an important role in the coming years, thus heightening the urgency for a more complete understanding of these policies.
For-Profit Colleges: Regulation and Student Outcomes
For-profit colleges have a unique standing with the federal government and the Department of Education. First, FPCUs have a separate designation from public and private nonprofit colleges under the HEA. This designation allows the federal government to treat these institutions differently and pass FPCU-specific regulations. The most notable of these targeted regulations is the federally legislated “90/10” rule which states that no more than 90% of a for-profit college’s revenue can come from Title IV sources (e.g., Pell grants, work-study programs, and federally backed loans). Additional regulations have been applied to all institutions, but targeted at proprietary colleges. For example, in 2008 the proposed changes to rules governing CDRs were demonstrated to disproportionately push FPCUs over the 25% limit (Pelesh, 2010). Although the rules were eventually softened, this is a clear example of regulation targeted at the for-profit industry via the larger policy process. Moreover, federal legal action against exploitative FPCUs has increased significantly in recent years. From 2004 through 2014, the U.S. Attorney General filed 17 lawsuits against proprietary schools. This legal action was in response to reports of widespread abuse within the industry, and suggests that exploitative behavior has been seen across the country. Moreover, federal investigations since the Nunn Report have revealed widespread abuses within the for-profit industry (Kutz, 2011; McComis, 2011).
Similarly, many states have singled out FPCUs for regulation. While all states regulate higher education as part of the triad, the targeting of proprietary schools has been purposeful. Historically, states and local municipalities have sought to curb profiteering and exploitative behaviors. Although the modern incarnation of FPCUs have been scrutinized a great deal for their business practices, reports of abuse by FPCUs, specifically in their recruitment tactics, date back to the 1912 and 1918 reports on proprietary schools in Chicago and New York. Numerous states responded by passing laws specifically controlling and regulating these institutions to protect their citizens (Angulo, 2016). While the types of proprietary schools have changed over time, and the market for these programs has ebbed and flowed with economic fluctuations, states have consistently sought to prohibit abusive tactics as part of the postsecondary regulatory structure. As part of this attempt, states have targeted FPCUs by passing regulations and consumer protection laws that seek to safeguard students from abuse. For example, Pelesh (2010) notes that Texas uses a tiered system of regulation that disproportionately groups FPCUs together as “workforce-oriented institutions” (p. 99) while California and Florida use regional accreditation, which is less common among FPCUs, as qualifiers for certain aid programs.
The role states play in regulating FPCUs is important for two reasons: First, it creates a differentiated market of regulations across relatively small areas; second, it demonstrates the impact that state policy has in shaping educational options and opportunities for students. Given the mobility of students, differentiated regulatory environments increase uncertainty for students in procedures (e.g., state aid eligibility) and quality (i.e., if low-quality schools have been driven out of a specific state market). The ability of states to influence which schools operate within their jurisdiction gives state governments significant control over local postsecondary markets. In particular, the use of regulatory power can manage and dictate the quality of higher education options available to its residents.
The current debate over regulations in the for-profit market has become increasingly heated, especially given the reauthorization of the HEA (Posselt, Venegas, Ward, Hernandez, & DePaola, 2017). As the market has grown in size and import, scholars have dedicated resources to understanding FPCUs and their students. Although FPCUs’ market share has decreased in recent years, deleterious effects still plague their students. Research has shown that for-profit students receive a disproportionate amount of federal aid (College Board, 2014), and represent a disproportionate share of student loan defaults (Cellini & Darolia, 2017; Deming, Goldin, & Katz, 2012; Hillman, 2014; Looney & Yannelis, 2015; Mettler, 2014). Underlying causes of these findings may include the higher price of FPCUs compared with similar public offerings (Cellini, 2012; Knapp, Kelly-Reid, & Ginder, 2011), deceptive marketing practices (Denice, 2015), higher rates of unemployment upon graduating (Deming et al., 2012), and the relatively low wages of for-profit graduates as compared with the price premium these students pay (Cellini & Chaudhary, 2014; Deming et al., 2012; Denice, 2015; Lang & Weinstein, 2012). Given the large investment by the public in for-profit colleges, these findings indicate that the return on investment may not be worthwhile for students or taxpayers. From an equity perspective, the disproportionate FPCU enrollment of low-income and students of color means that a lack of regulation or enforcement may widen social inequality and reproduce systematic disadvantages.
At the federal level, the Obama administration pursued a strong regulatory environment and sought to curb abuses in the for-profit sector (Natow, 2015). The Trump administration, under Secretary of Education Betsy DeVos, has destabilized the future of for-profit regulation. Campaign promises have turned into executive orders mandating decreases in regulation across all departments. In June of 2017, DeVos suspended two regulations and called for further review (Kreighbaum, 2017). First, the Department will stop processing requests under the borrower defense to repayment rule, which helps students defrauded by FPCUs discharge their federal loans. Second, the Gainful Employment Rule, which dictates programs at FPCUs must produce graduates who are employed in their field and can repay their loans, has been suspended after years of the Obama administration battling in court to keep it alive. As of July 2017, AGs in 18 states and the District of Columbia have filed lawsuits against the federal Department of Education to reinstate these programs. Moreover, DeVos has appointed for-profit executives and lobbyists, who have vocally disapproved of current regulations on the market, to top Departmental positions. The steps the Secretary has taken thus far indicate a destabilization of the federal-level regulatory environment and uncertainty for measures intended to protect students. This ambiguity will likely result in state policy, and legal action, becoming significantly more important to FPCU regulation in the coming years.
Accompanying this wave of scholarship has been vigorous debate regarding the role of regulation in safeguarding taxpayer dollars and students. Specifically, the public and legislators have deliberated about the utility of the recently passed 2016 Gainful Employment Rule, as well as previous regulations like the 90/10 rule (Posselt et al., 2017). Evidence regarding the efficacy of such regulations is limited (Ward, 2017). Darolia (2013) finds that federal aid regulations designed to safeguard students from low-quality institutions are less effective at dissuading racial and ethnic minority students from attending schools with high default rates. However, there is some evidence that federal regulations on CDRs has limited the flow of student aid dollars to the worst-performing for-profits (Jaquette & Hillman, 2015).
While the body of evidence on the efficacy of federal regulation is limited, even less is known about the composition and effectiveness of state regulations on for-profit colleges. Our intent is to fill this gap, and do so while the federal government is on the precipice of deregulating the market. In addition to education scholars’ concerns over the quality of education received at for-profit colleges, deregulation at the federal level will likely increase the role that state governments play in shaping the for-profit market. We aim to understand factors that contribute to state enforcement of for-profit regulations. The widespread journalistic reports of abuse, the significant growth in regulations on FPCUs, and the recent use of legal action from the U.S. Attorney General all suggest bad-acting FPCUs are still embedded in the proprietary market. Research on FPCU student outcomes reinforces the need to eliminate the colleges that exploit vulnerable students. As the initial triad of regulation was designed, states played and continue to play an important role in monitoring and regulating their local markets. This study sheds light on the rigor with which state AGs have pursued regulatory enforcement to expunge local for-profit markets of low-quality schools.
Conceptual Framework
To understand factors related to the enforcement of state regulations of for-profit colleges, we call upon political science theories. The relationship between governments and colleges and universities can be characterized as one between a principal and an agent. 1 As such, governments design contracts, in the form of regulations, to guide the behaviors of colleges and bring them into alignment with the goals of the government. Given the profit motive of FPCUs, the principal–agent relationship is particularly fitting. Leaders of for-profits, as guided by their fiduciary responsibility, pursue profits over the public interest. For this reason, both federal and state governments seek to align FPCU goals with public interests. The principal uses regulatory enforcement, usually in the form of sanctions for bad behavior or rewards for good behavior, to induce desirable behaviors.
In this traditional description of the principal–agent relationship, the agent possesses more information about the inner workings and behaviors of the organization than the principal. Given the conflicting goals, this information asymmetry makes control more difficult for the principal to ascertain. As such, increased regulation will occur (Waterman & Meier, 1998). With the inauguration of Donald Trump came a realigning of goals. That is, the commitment to deregulation signals that the federal government no longer sees the goals of FPCUs as contentious with those of the government. State governments that see a misalignment between the for-profit sector and the public good will have to increasingly fill the regulatory void to maintain control of the industry.
At the federal level the principal–agent relationship has manifested itself in the 90/10 rule, bans on performance pay for FPCU recruiters, the Gainful Employment Rule, and CDR policies. These regulations, each with its specific clauses about Title IV revenue amounts or alumni employment or loan repayment rates, use objective measures to induce institutional behaviors. The enforcement of federal-level policies has come in the form of sanctions, restricted access to Title IV funds, and lawsuits brought by the U.S. Attorney General. In fact, during the years included in this study 17 lawsuits were filed by the Department of Justice. This reinforces the power of legal action as a policy mechanism as well as the nation-wide prevalence of abusive and self-interested actions of FPCUs.
In a similar fashion, states can use legal recourse to enforce their regulations on the proprietary market. Bringing a lawsuit against a for-profit college may actually be a state’s only form of regulatory action. Given that proprietary schools are highly reliant on federal student aid, not state aid (Ward, 2017), state governments cannot use access to a primary source of FPCU revenue as a policy lever. This leaves legal recourse as a remaining policy tool for state governments. Couched in this understanding of the principal–agent relationship is the policy-making process. In a federalist system of government, states have a great deal of autonomy to create their own policies and regulatory apparatuses. Understanding how state-level policies are created and implemented is useful for recognizing the differentiated opportunities that may exist for citizens in different states. In the case of FPCU regulation, the variability in policy implementation, and specifically the enforcement of policy, has the potential to generate drastically different levels of access and protection for students. As noted above, individual principal–agent relationships exist between state governments and FPCUs, and thus policy implementation varies between jurisdictions. These relationships and the implementation within each state are likely influenced by individual policy environments. Although these principal–agent relationships vary across state lines, they are likely to influence one another. Understanding the dyadic influence between states can provide valuable information regarding asymmetric implementation. Policy innovation and diffusion proves to be a useful lens for understanding individual policy environments and the relationship between states (Hearn, McLendon, & Linthicum, 2017). Below, we review the concepts of policy diffusion and innovation, and their applications to higher education research. We use this framework to examine factors that contribute to the enforcement of state regulations on FPCUs.
Theories of Policy Diffusion and Innovation
To understand these two concepts and how they work in tandem, a brief history of their development is useful. Policy diffusion (Walker, 1969) explains that state-level policies spread regionally and that states often adopt specific policies after their neighbors have done so. A long body of literature examines the influence some states have on policy adoption patterns of other states (see F. S. Berry & Berry, 2014). This idea has been applied to a multitude of policy areas, including higher education (e.g., Hearn & Griswold, 1994; Lacy & Tandberg, 2014; Li, 2017; McLendon, Heller, & Young, 2005; McLendon, Hearn, & Deaton, 2006).
Diffusion research has identified five specific explanations for policy adoption which are related to diffusion: learning, imitation, normative pressure, competition, and coercion. In the learning explanation, states derive ideas from other states. Learning often occurs by observing neighboring states, and is not limited to policy adoption. Implementation strategies have also been considered through a diffusion lens (Renzulli & Roscigno, 2005) and signal a way to think about patterns among regulatory enforcement. Policy makers may assume that neighboring states are similar geographically, economically, or culturally, and thus mimic policy ideas (Sponsler, 2010). Others have seen learning from geographic neighbors as driven by an inability to adequately assess and understand policy making in all other states, thus focusing attention on the most familiar, and closest, states (W. D. Berry & Baybeck, 2005).
F. S. Berry and Berry (2014) explain that imitation may come in the form of one state following the lead of another state that is perceived to be excellent. Often this perception is rooted in wealth or power. Normative pressures likely act similar to imitation in that wealthy and powerful states may influence the general norms and values related to postsecondary policies. However, it is also possible that nearby states are major influencers of behavior. Both imitation of leaders and normative pressure can be localized. That is a leader in the Northeast may be more influential for other Northeastern states, and may assert greater normative pressures than West Coast states that are considered ideal.
The final two factors are competition and coercion. F. S. Berry and Berry (2014) define competition as State A trying to secure an advantage or prevent State B from securing an advantage. Coercion, however, is when the adoption by State A increases the incentives of State B to adopt. The outcome of both mechanisms is the same: State B will seek to adopt the policy. However, how the pressure is exerted, and the connotation of this pressure is different. While a coercive policy attempts to force another state’s behavior for the benefit of the primary actor, a competitive policy exploits an opportunity to create an advantage for the primary actor.
Each component provides a logical mechanism for the geographic spread of regulatory enforcement. For example, if one state pursues enforcement as part of the implementation process, this may provide competitive or coercive pressures on neighboring states to pursue this aspect of implementation. This method has been used by numerous education scholars to understand postsecondary policy adoptions, but little evidence suggests that geographic diffusion is present (Sponsler, 2010). For example, diffusion was not found to be a significant factor in the spread of state merit aid programs (Doyle, 2006), dual enrollment programs (Mokher & McLendon, 2008), or prepaid tuition plans (Doyle, McLendon, & Hearn, 2010). More recent research has posited a hypothesis for a lack of findings supporting geographic diffusion: Perfect copies of specific policies are unlikely to spread (Lacy & Tandberg, 2014). The authors argue that “the deep parsing of larger policy categories [as previous work has done] may obscure geographic policy diffusion’s influence” (Lacy & Tandberg, 2014, p. 628), and thus look at postsecondary finance innovations more generally. We concur with the idea that because states differ on many attributes, exact regulations may not exist across states. It is more plausible that a state devises regulations based on the unique characteristics of its postsecondary market. However, regulations across states have a common underlying motivation: to safeguard students and keep predatory colleges out of the market. As such, we look at for-profit regulation more broadly, and specifically if enforcement is pursued.
A competing theory developed during a similar time and focused on internal determinants of the policy-making process. The notion of innovation in this process stemmed from the idea that in absence of similar policies in other jurisdictions of which a state can model its new policy, a state will need to innovate a new policy. This innovation process is seen as the result of internal factors (F. S. Berry & Berry, 2014).
Policy innovation is thought to be associated with three factors: available motivation, resources, and obstacles. Innovation is thought to be positively correlated with the first two and negatively correlated with the third (Mohr, 1969). Motivation has often been conceptualized as problem severity (F. S. Berry & Berry, 2014). As such, higher default rates may be associated with an increased likelihood of adopting for-profit regulation. F. S.Berry and Berry (2014) point out that resources are necessary to overcome obstacles because legislative bodies require funds to research new policies and explore their tenability in the state.
Internal aspects of policy innovation were found to be significant predictors in the aforementioned narrowly defined higher education studies (Doyle, 2006; Doyle et al., 2010; Mokher & McLendon, 2008). Not unsurprisingly, state-level factors appear to play an important role in the implementation of postsecondary policies (Perna & Finney, 2014). Given the importance of policy context in for-profit regulation broadly (Serna, 2014), internal state factors are expected to play a role in the regulatory action taken by each state.
Policy diffusion and innovation developed in parallel until F. S. Berry and Berry’s (1990) influential work on state lotteries married diffusion with policy innovation. In addition to uncovering evidence of a geographic spread of these policies, the authors contributed the important conceptual idea that both internal determinants and diffusion play a role in policy adoption. That is, states are influenced by internal factors, such as unemployment or political ideology, and external factors, such as neighboring states’ policy adoption status. Evidence of diffusion and innovation working in tandem has been demonstrated within higher education policies (Lacy & Tandberg, 2014; Li, 2017). Given the evidence of policy learning and the importance of context within implementation literature, we propose the application of ideas from both diffusion and innovation to regulatory enforcement. Given the individual principal–agent relationships that exist across states, variation between regulatory enforcement would be expected. However, just as policy adoption can be affected by geography and context, we expect implementation to take a similar functional form. We express this algebraically:
where regulatory enforcement (Re) is a function of a state’s economic conditions (EC), higher education structure (HES), geographic factors (GEO), legal authority (LA), FPCU characteristics (FPCU), and state political ideology (IDEO).
Theoretical Limitations
Of course, no theory of the policy process fully captures all aspects of a specific phenomenon. Given the limited research of FPCU regulation at the state level, this study is meant to be exploratory and propose one possible theoretical understanding. Here, we have created an ideal type, relying on previous policy research, and are testing it as a possible explanation. While policy diffusion and innovation can provide a useful lens for understanding regulatory enforcement, the topic is complex and there are likely other explanations that may complement our conceptualization.
Although our study incorporates objective measures of political leanings, theoretical understandings of the political process may provide useful ways of examining and understanding the enforcement process. Specific relationships that exist within each state government could be influential in the regulatory apparatus, and may complement the innovation and diffusion conceptualization presented here. As such, we do not propose that our framework is absolute in nature, but merely an attempt to explain differentiation between state behaviors. This study is meant to provide initial work on state regulatory enforcement that can spur future research that further develops a theoretical model of contributing factors.
Data
To understand how regulatory action manifests at the state level, we combine multiple data sources to measure internal and external factors in alignment with the policy diffusion and innovation framework. These data sources and variables are summarized in Table 1, and descriptive statistics are provided in Table 2. Given the focus on both the internal and external influences of policy making, a multitude of data sources are necessary to adequately measure diffusion and innovation. The variables included reflect previous diffusion and innovation studies in higher education (Hearn et al., 2017), and are intended to be a proxy for the elements of diffusion and innovation detailed above.
Variable Descriptions and Sources.
Note. FTE = full-time equivalent; FPCU = for-profit colleges and university.
Descriptive Statistics.
Note. FTE = full-time equivalent; FPCU = for-profit colleges and universities.
Outcome Variable
Policy diffusion and innovation studies seek to understand how policies develop and come into fruition. This study moves beyond policy making to policy enforcement. While state-level regulations of for-profit colleges are essentially ubiquitous (LexisNexis, 2016), enforcement may not be. While passing a policy can be symbolic, the enforcement or lack thereof is both symbolic and important for the efficacy of regulations. Evidence across many industries suggest that regulation without enforcement undermines the effectiveness as profit-seeking corporations will pursue private business interests (see, Lee & Baik, 2017; Tombs, 2016; van Rooij, 2010)
To assess the factors that influence state-level enforcement, we conduct an event history analysis using state lawsuits as the event of interest. Using cross-sectional data from 2007 through 2014, we merged data from multiple sources listed in Table 1. State lawsuits were catalogued from the National Consumer Law Center (2014) to indicate if each state filed a lawsuit in a given year. These lawsuits are intended to proxy enforcement of state regulations and serve as our outcome variable.
Of course, we acknowledge that this may be a limited conceptualization of enforcement. For example, each lawsuit differs in size and scope; the size of the institution being prosecuted varies, the type of degrees offered at the offending school, and some lawsuits may be part of larger multistate lawsuits. These caveats are all important and warrant future attention. However, given the exploratory nature of this study, we find that a broad conceptualization of enforcement allows for a preliminary investigation into the factors that contribute to this phenomenon.
Independent Variables
The independent variables in our model are meant to capture the concepts of diffusion and innovation. Diffusion has been broken into five underlying mechanisms: learning, imitation, normative pressure, competition, and coercion. As described above, these inherently have a geographic component, and as such we include a variable titled diffusion to identify if a neighboring state’s AG has filed a lawsuit against a for-profit.
Working in tandem with geographic diffusion are internal state factors. Policy innovation is thought to be associated with three factors: motivation, available resources, and obstacles. These factors are associated with multiple variables that help us approximate the construct. For example, states may be motivated to curb exploitative behaviors if FPCUs in the state have low loan repayment rates, if FPCUs comprise larger portions of postsecondary enrollments, or if the for-profit market is dominated by larger multicampus chains which thrive on a business model that exploits students (McMillan-Cottom, 2017). Additional factors related to motivation have been included from previous studies on higher education such as unemployment rate (Li, 2017) and state expenditures on higher education (Li, 2017; Tandberg & Hillman, 2014).
Financial resources also play an important role in policy innovation (F. S. Berry & Berry, 2014). As such, our model includes income per capita, to measure the wealth of a state. This measure has been used in previous research on higher education policy diffusion and innovation (Li, 2017; McLendon et al., 2005). In addition to financial resources, legislative resources are important. To account for this, we include the amount of government expenditures (e.g., legislative staff salaries, policy research resources, etc.) per state legislator.
Finally, we include proxies for potential obstacles that may prevent the enforcement of state regulations. Specifically, we include a binary variable for whether or not consumer protection responsibilities fall under the purview of the state AG in a given year. Given that FPCUs are private, profit-seeking companies, students are often considered and treated like consumers by institutions and the government. Just as with other industries, state governments often seek to protect consumers from fraud or exploitation. As such, it is possible that states that do not allocate these responsibilities to the AG limit the likelihood that the state will legally enforce regulations on FPCUs. We also include a measure of political ideology. As previous research has noted, Republican leaning state governments are less likely to actively seek to enforce regulations. We use a measure of ideological leaning (W. D. Berry, Ringquist, Fording, & Hanson, 1998) curated and updated by Richard Fording and made public on his website (Fording, 2015). The measure ranges from 0 to 100, where 0 is a conservative ideology and 100 is a liberal ideology.
Method
To assess the impact of diffusion and innovation, we conduct an event history analysis. Because lawsuits can, and do, happen on multiple occurrences, we calculate odds ratios of the likelihood of a state filing a lawsuit in a given year. Nested logistic regression is an appropriate approach for event history analysis with multiple opportunities for an event to occur, where annual observations are nested within states. We use a discrete-time approach as the data are measured on an annual basis. A continuous-time method, such as a Cox model, would be more appropriate if our data were measured in a more temporally precise manner (Chen & DesJardins, 2008). However, daily values of our control variables, such as loan repayment rates, is not feasible given that many are reported once each year. As such, our model measures the odds of a lawsuit being fired in a given state and year. The nested logit model is estimated by the following:
where we are examining the odds that state s will file a lawsuit in year t,
Six state-year observations had incomplete data for legislative expenditures. As such, we use Stata’s multiple imputation to generate missing values. The models with imputed values do not differ substantively from models ran excluding the six observations with missing values, thus, we present the imputed models here. To assess the robustness of our findings and isolate the potentially differential effect that diffusion and innovation variables have, we estimate our model in three phases. First, with only the geographic diffusion indicator; second, with only the innovation variables; and third, a combined model with all indicators. Finally, we cluster our standard errors at the state level as error terms can be serially correlated in panel data (Wooldridge, 2003).
Limitations
Although this study provides a new understanding of factors that contribute to regulatory enforcement, there are some limitations to our findings. One factor that could play an important role in enforcement is the amount of resources available for enforcement. While we include legislative expenditures, there may be an important link between capacity in AG offices and the ability to pursue enforcement. Future research should explore this relationship, including qualitative work that seeks to understand how different areas of regulatory enforcement are prioritized.
A second limitation stems from how diffusion occurs. Our study focuses on geographic diffusion, however, in 2017, diffusion could operate through different channels (Hearn et al., 2017). Ideas may spread between state leaders through technological means which bypass geographical locations, or through national conferences that purposefully seek to bring together leaders from disparate areas. Although future research should consider or focus on these other methods of diffusion, the findings will complement, rather than supplant, our findings on the importance of geography. Moreover, geographic diffusion uniquely captures market factors that result from proximal locations, whereas other forms of diffusion may better reflect organizational learning. Future work should give additional attention to the five aspects of geographic diffusion. Although, these mechanisms are difficult to operationalize in quantitative work, qualitative research could provide necessary nuance to these factors.
Findings
Table 3 shows the results of our models, where Model 1 is diffusion only, Model 2 is innovation only, and Model 3 is a combined model. While theoretically we believe that diffusion and innovation play roles in the likelihood of an AG filing a lawsuit against an FPCU, it is important to test the models separately. Using three different models, where the third is the combined model, allows us to estimate how sensitive diffusion and innovation are to each other. The change in magnitude and significance of an independent variable reflects how sensitive it is to the inclusion of other theoretically relevant factors. The coefficients are reported in odds ratios for ease of interpretation. In both individual models and the combined model, we find evidence that diffusion and innovation factors independently and jointly affect the likelihood of a state AG pursuing legal action against a for-profit college.
Summary of Findings.
Note. FTE = full-time equivalent; FPCUs = for-profit colleges and universities.
Model 1 shows a strong association between diffusion and the likelihood of a lawsuit, but is modeled without any additional controls. The odds that a state exposed to a neighbor that has filed a lawsuit are 5.09 times greater than a nonexposed state’s odds of filing a lawsuit in a given year. The geographic exposure represents a highly significant and powerful force in the likelihood that a state files a lawsuit against a for-profit college. However, theory and previous research suggest that diffusion does not act in a vacuum and, thus, this model likely suffers from omitted variable bias. Specifically, we are concerned with internal state factors that may influence the likelihood of a lawsuit. These factors are examined in Model 2.
Model 2 fits the relationship between innovation factors and the likelihood of a lawsuit. It demonstrates a strong link between the average 3-year repayment rate of for-profit students and legal action, and a weaker link between the composition of for-profit institutions within a state (i.e., the proportion of FPCUs that are single campus institutions) and a state lawsuit. Both factors show an odds ratio of less than one, indicating a negative association. States with a greater number of single-campus institutions are less likely to file a lawsuit against a for-profit. This is not entirely unsurprising given that many lawsuits are brought against large for-profit companies that typically have multiple campuses. The association between the composition of FPCUs and likelihood of regulatory enforcement diminishes in the full model suggesting that this factor may not be as important as initially thought. However, the differential effect between models indicates an area for additional research.
Additionally, states with higher loan repayment rates of for-profit students are less likely to file a lawsuit. With every 1 percentage point increase in repayment rate, the odds a state AG will file a lawsuit decreases by 6.7%, as indicated by Model 2. Given the significance of diffusion in Model 1 and our theoretical justification, we now turn to a fully fitted model that includes both internal factors and diffusion. This is Model 3.
In addition to in the combined model, the association between diffusion and loan repayment rates persists, although both become somewhat less influential; in fact, the impact of diffusion is cut in half. This suggests that internal factors capture some of the explanatory power that was attributed to diffusion in the limited Model 1. However, the effect size of diffusion in the full model is still quite sizable. Being exposed to a neighboring state that has filed a lawsuit is associated with a 265% increase in the odds of filing a lawsuit. That is, when State A’s neighbor has filed a lawsuit, the odds of State A filing a lawsuit, in a given year, is 2.65 greater than if State A had no neighbors that filed suits. Understanding this effect size is imperative, especially given the magnitude. We display the findings in odds ratios because they are more easily interpreted, however, odds ratios must be considered in relation to baseline odds. While a 265% increase seems very large, it comes on top of the baseline proportion of state-year observation that experience a lawsuit of 16% (see Table 2). That is, only 16% of state-year observations include the filing of a lawsuit, and a neighboring state’s previous filing of a lawsuit is associated with a 265% increase in the odds for a given state-year observation. Methodological research points out that extreme values of odds ratios are likely due to an overestimation; however, with baseline odds less than 20% this overestimation is likely small (Davies, Crombie, & Tavakoli, 1998). Consequently, we are cautious to accept the exact value associated with diffusion, but certainly accept the significant and large impact it has on increasing the likelihood of a given state pursuing legal action.
The link between repayment rates and the likelihood of regulatory enforcement is consistent and persistent across models. In the full model, an increase in one percentage point in the average repayment rate at FPCUs in a given state is associated with an approximately 6% decrease in the odds of a lawsuit being filed. That is, states with for-profit colleges who graduate students that are unable to repay their federal loans are more likely to seek state-level legal action against a for-profit college.
Models 2 and 3 include an expansive list of internal factors that literature and theory suggest may be associated with policy innovation. However, our findings do not demonstrate an association between these factors and increased likelihood of regulatory enforcement. Factors such as governmental ideology, legislative professionalism, and the consumer protection authority of the state AG do not appear to be connected to legal action. In addition to these factors being statistically insignificant, most of these factors are associated with small effect sizes strengthening the conclusion that there is no link between these variables and the likelihood of regulatory enforcement. This small effect size is persistent across Models 2 and 3 suggesting that the lack of influence of these factors in the full model is not attributable to the inclusion of diffusion.
Discussion
The findings have implications for students, policy makers, and those who do research on FPCUs. A better understanding of the factors that contribute to regulatory enforcement by state AGs contributes to our collective understanding of how and if students are being safeguarded as intended, and how different state markets for FPCUs function. The large effect of diffusion in our models reflects the five mechanisms of the diffusion process, and these mechanisms have important implications for how we understand state-level markets for education.
Diffusion research has identified five specific explanations for policy adoption which are related to diffusion: learning, imitation, normative pressure, competition, and coercion. While our study does not have a mechanism by which to approximate each individual facet of diffusion, we do show support for diffusion generally. The apparent diffusion of regulatory action across state boundaries may reflect learning, imitation, or normative pressures that result in a leader–follower diffusion action across state lines. These three mechanisms reflect ways in which state governments, as organizations, learn from neighboring governments.
The diffusion of regulatory actions through competition and coercion likely reflect market powers. A crackdown on exploitative and low-quality FPCUs in one state creates market pressure, either competitively or coercively, on neighboring states. While these two ideas are different, in the case of for-profit regulation they may be quite similar because of the mobility of students to freely cross state lines and the nimble characteristics of FPCUs to quickly open or close in new markets in order to secure profits (Deming et al., 2012; Tierney & Hentschke, 2007). Policy diffusion suggests that when a state enacts regulations that purportedly drive low-quality FPCUs out of the market, neighboring states will follow to maintain a similar level of quality among institutions. If a nearby market has a higher quality product, consumers will either gravitate toward that market or, at a minimum, refrain from engaging in their local, lower quality market. The adoption of a regulation in one state may spur the adoption of similar regulations in neighboring states in a coercive and competitive manner. This type of scenario may not hold for large states because markets are geographically spread out. However, for smaller states and residents living near state borders, difference in quality because of differential regulatory environments may give students the option to choose between lower and higher quality colleges. This puts pressure on states with looser regulatory environments to ensure better postsecondary options. While these aspects of diffusion are not directly operationalized in this exploratory study, we put this forward as an area that requires additional research. This competitive and coercive mechanism of diffusion indicates an area for further exploration, specifically in relation to geographic bounds.
The link between repayment rates and regulatory action likely reinforces the idea that market pressure plays an important role. As for-profit regulation seeks to weed out exploitative firms and protect students and taxpayers, states may be motivated to adopt regulations when residents are being harmed. One proxy for this may be low student loan repayment rates among for-profit students within a state. Low repayment rates may reflect an educational product that does not provide students with useful skills in the labor market. This metric can capture students who paid high tuition prices for training in a profession with low salaries. In both scenarios, low repayment rates indicate that market mechanisms, which theoretically prevent overborrowing or act as a quality control are failing. In turn, state governments may increase regulatory enforcement to correct market failures. The opposite scenario would logically fit as well: In states where students from for-profit colleges are performing well enough in the labor market to repay loans, states may see little reason to intervene legally. Considered next to the diffusion findings, this corroborates the notion that market pressures play an important role for predicting enforcement behaviors. That is, if the market appears to be failing, state governments may be more likely to use legal recourse to protect students and the state’s economic interests. In cases without market failure, regulatory enforcement may not be prioritized.
The significance of both diffusion and innovation factors reinforces previous research that policy develops from within and outside a state (F. S. Berry & Berry, 2014). While diffusion research that focuses on higher education policy has found limited evidence in studies of narrow policies, there is evidence that a broader conceptualization of policy innovation diffuses geographically (Lacy & Tandberg, 2014). The present study follows this notion and defines for-profit regulation broadly, rather than examining specific state rules. This captures a more comprehensive commitment by state governments to regulating the for-profit market, rather than specific types of regulation. As the nuance of regulations is differentiated across stateliness, this broader conceptualization more appropriately captures state behaviors that seek to safeguard students and rid the market of low-quality FPCUs.
In addition to furthering theoretical understandings of policy diffusion and innovation in higher education and how these theories may relate to the spread of enforcement as part of the implementation process, the findings from this study have implications for policy makers and students. Federal policy making and regulation has become increasingly politicized, and promises to deregulate the for-profit market have become part of the DeVos and Trump agendas (Posselt et al., 2017). The role of state policy makers in safeguarding students from predatory colleges will likely grow in importance over upcoming years. While regulation has grown in politicization at the federal level, our findings suggest that this same phenomenon may not be occurring at the state level. Ideological leanings of state governments show no significant association with the likelihood of a lawsuit. However, this finding should be interpreted cautiously as many factors likely contribute to regulatory enforcement. Political ideology may have played a limited role under the Obama administration which was actively trying to regulate the for-profit market. With a deregulation of this market under DeVos, states may be the only form of regulatory enforcement which may increase the importance of internal determinants, including political leanings. Additionally, political stability may play an important role and should be examined in future research. Although this study is exploratory and additional contextualization is necessary, we can be confident that the findings suggest that enforcement is not purely political in nature. This limited impact of politicization at the state level means that policy entrepreneurs may be able to spearhead regulatory movements within a state regardless of political ideology.
The inverse relationship between repayment rates and the likelihood of a lawsuit suggests that regulatory enforcement is related to the quality of education students are receiving. If lawsuits are more likely to be brought when graduates of FPCUs are left with debts they cannot pay, state governments may be actively seeking to protect citizens, regardless of party affiliation. In this way, state regulation may be an effective tool in movement to safeguard vulnerable students. It also suggests that at the state level regulatory enforcement is explicitly linked to ridding the market of low-performing or exploitative for-profits, thus, effectively seeking to help students.
The strong evidence of geographic diffusion also suggests that individual states may be able to influence broader regulatory enforcement. For example, if one specific state actively pursues a regulatory movement against FPCUs, neighboring states may need or want to follow this behavior. This raises the possibility that a state government that becomes strongly motivated to regulate the for-profit market can influence widespread legal enforcement. Table 4 provides a list of the number of lawsuits filed each year for each state. Although the coefficient on diffusion is large and highly significant, it is worth noting the dispersion of states that filed lawsuits. The three most active states – Massachusetts, Kentucky, and Florida – are not neighboring states. However, the significance of diffusion suggests that individual states, like these leaders, may be able to influence neighbors and thus broader patterns of regulatory action. In a period marked by federal withdrawal from regulation, individual states could potentially fill this gap.
State Lawsuits by Year.
Conclusion
This study moves beyond the creation of regulation to the enforcement of regulation. Enforcement can be seen as a policy decision in itself. As such, we examine factors that contribute to the enforcement of for-profit college regulations at the state level. FPCUs disproportionately leave students with higher debt loads, default rates, and unemployment rates. Federal and state governments have instituted regulations on this market to protect students from predatory FPCUs. Widespread exploitative behavior was present during the period studied, from 2009 through 2014, as indicated by the vast journalistic reports and the significant number of federal lawsuits targeting FPCUs across the country. However, state lawsuits were not as ubiquitous during this period.
Our findings support previous work that suggests policy is affected by internal state factors and geographic diffusion. The significance of neighboring states’ behaviors and the ability of FPCU students to repay student loans supports theories of policy diffusion and innovation, and suggests that market forces may play an important role in regulatory enforcement. These findings also suggest that entrepreneurial states may be able to create movements of FPCU regulatory action by influencing neighboring state. This could become important as the federal government has indicated an intention to deregulating for-profit colleges.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.