The Creation of Performance Funding in Michigan : Partnerships , Promotion , and Points

While accountability in higher education has been a topic of debate for decades, in recent years the discussions have shifted to emphasize efficiency and economic measures of success. A prominent example of this accountability movement is the increase in popularity of performance funding policies. One of the most recent states to implement performance funding is Michigan, which began their performance funding policy in 2012. This study explored the creation and implementation of the state of Michigan’s performance funding policy. In particular, the decision making processes institutional administrators and state leaders engaged in while designing, promoting, and implementing the policy. Using a case study design and interviewing both higher education administrators and state leaders who were involved in the creation of the performance funding policy in Michigan yielded five large trends: 1) The importance of advocacy coalitions; 2) Securing support through a focus on higher education affordability; 3) Concerns with how to measure data and compare institutions; 4) Insufficient financial incentives; and 5) Limited impact on institutional decision making.

In response, states have focused on creating higher education policies with quantifiable and economic measures of success. One prominent example is the increase in popularity of performance funding policies. These policies create state specific metrics that institutions need to meet in order to receive financial support from the state. Performance funding policies allow state officials to emphasize metrics that they believe will better incentivize efficiency such as graduation rates, financial stewardship, credit hours earned, wages of graduates, or research grant dollars . The institutions that perform well on the metrics are financially rewarded through increases in their operating budget. One of the most recent states to implement performance funding, Michigan, is the focus of this study.

Study Purpose
The broad purpose of this study was to understand the creation and implementation of the state of Michigan's performance funding policy for the four-year sector and the decisions made during the process. In particular, I explored the decision making processes institutional administrators and state leaders engaged in while designing, promoting, and implementing the performance funding policy. This study contributes to the scholarly discourse by offering a robust analysis of the roles these stakeholders engaged in and their decision making processes during the design and implementation process of performance funding.
This research expands current literature in three notable ways. First, Michigan is a previously unstudied performance funding state. As performance funding spreads, it is important to expand the number of perspectives and states studied to understand trends and best practices that arise in and are applicable across multiple states. Second, this research was conducted two years post implementation and therefore gives a unique perspective on the implementation process. Much of the discussion regarding performance funding implementation takes a long term historical approach, asking participants to reflect back on experiences from many years ago. While beneficial and important, research of this type may be biased due to participants reflecting on the past through the lens of their current knowledge. This study strived to interview participants soon after the policy was implemented when the process was still fresh and unbiased in the participants mind. Finally, this study focused specifically on the decision making process state leaders and university administrators engaged in during policy creation and implementation. While numerous research studies have focused on who was involved in the policy process and how performance funding spreads (e. g. Dougherty & Natow, 2015;McLendon, Hearn, & Deaton, 2006), research exploring the decision making process of the individuals involved is lacking.
Given these limitations in the current literature, this study focused on the decision making processes institutional administrators and state leaders engaged in while creating and implementing the performance funding policy in the state of Michigan. To this end, I focused on the following research questions: 1. How was Michigan's performance funding policy created and implemented? 2. Who was involved in the policy creation and implementation process and how did these stakeholders interact?

Literature Review
While Michigan recently implemented their performance funding policy, discussions regarding accountability have occurred in higher education since the 1970s. Performance funding is a direct outgrowth of this focus on accountability in higher education. To situate this study in the current literature, this section provides an overview of accountability in higher education, a history of the spread of performance funding, and a summary of the impacts of performance funding.

Accountability in Higher Education
While accountability in higher education has been a topic of debate for decades, in recent years the discussions regarding accountability have shifted to emphasize efficiency and the demand for specific results and outcomes (Burke, 2005). Simply defined, efficiency is the measure of outputs (graduation rates, faculty productive, etc.) over inputs (cost of higher education). State leaders expecting greater efficiency from higher education pressure institutions to increase outputs, decrease costs, or both (Burke, 2005). Several factors have driven the emphasis on efficiency. First, the massification of higher education has shifted the college degree into a necessity for most Americans and brought the higher education sector into the spotlight (Burke, 2005). Second, economic competiveness and the belief that the United States is falling behind in the knowledge economy has increased the attention on higher education's ability to produce graduates with skills and abilities for today's labor market (Burke, 2005). Third, the recent economic downturn has strengthened this call for greater accountability due to the financial pressure placed on states and legislators' focus on the rising price of higher education (Powell, Gilleland, & Pearson, 2012). These three factors combined with the shifting social contract has led to a strong focus on accountability in higher education and a demand for quantified proof of the economic returns on the public investment in higher education.
Within a period of three years, reports by the National Governors Association (NGA), the National Conference of State Legislatures (NCSL), and the State Higher Education Executive Officers Association (SHEEO) all advocated for higher education policies that focused on state needs and economic competitiveness (NCSL, 2006;NGA, 2007;SHEEO, 2005). The renewed push for accountability was spurred on by the public's frustration with the quality of student learning, higher education's focus on research and graduate education, the expanding number of administrative staff on campus, low completion rates and limited employability of graduates (Burke, 2005). Further, legislators went beyond a push to compare institutions and began to advocate for the standardization of higher education, believing this would aid the comparability of institutions and enhance the measure of quality (Eaton, 2012).

History and Spread of Performance Funding
Performance funding is part of this growing trend in accountability. As of July 2015, performance funding policies existed in 32 states and five additional states were in the process of transitioning (NCSL, 2016). Performance funding is often thought of as occurring in two waves, frequently referred to as performance funding 1.0 and performance funding 2.0 (Dougherty & Natow, 2015). The adoption of performance funding 1.0 took place between 1979 (when Tennessee implemented the first policy) and 2000. In this version, performance funding is usually a bonus over regular state funding and involves small amounts of funding, usually between 1% and 6%. In performance funding 2.0, starting around 2007 and continuing to the present, the funding is embedded into the normal state funding for higher education. About two thirds of the programs created in the second wave are readoptions of previously discontinued policies (Dougherty & Natow, 2015). The amount of funding in 2.0 versions is also usually higher, with between five to 25% of all higher education funding coming from performance funding formulas. Of note, there is no national performance funding policy or formula and instead each state creates their own policy with great diversity across states and policies (NCSL, 2016).
Authors attribute the spread of performance funding and its fellow accountability measures to diverse factors, including: globalization, pressure to maximize productivity and efficiency, a shift towards marketization of higher education and governance, financial pressures from the great recession, changes in political leadership, and a growing frustration with the voluntary and self assessment of higher education (Dougherty & Natow, 2015;McLendon et al., 2006). Several studies have empirically explored the spread of performance funding. McLendon et al. (2006) conducted an event history analysis model to explore what sociopolitical factors influence a state's adoption of performance funding, performance based budgeting or performance reporting policies across all states. Of the ten areas they explored, they demonstrated that legislative party strength and higher education governance arrangements were the primary drivers. For performance funding specifically, a higher percentage of Republican state legislators and the absence of a consolidated governing board increased the probability of the state adopting a performance funding policy. The authors believe that Republicans favor performance funding over performance budgeting or performance reporting as it grants legislatures the strongest leverage for accountability (McLendon et al., 2006). Beyond Republican state legislators, governors who identify as Republican also seem to favor performance funding policies. Since 2007, nine states with Republican governors have implemented performance funding policies (Dougherty et al., 2014a). Dougherty et al. (2014b) examined how political forces influence the adoption of performance funding through a qualitative analysis in three states (Indiana, Ohio, and Tennessee). Through interviews with state legislators, higher education administrators, state governing board members, and business leaders, the authors determined that the main supporters of the adoption of performance funding policies were the governor, the state higher education coordination board and to a lesser extent the legislative leaders and business leaders (Dougherty et al., 2014b). This advocacy coalition arose from the shared beliefs that states needed to increase the number of residents with college degrees in order to meet economic demands, that the recession required higher education to become more efficient, and that a performance funding policy would allow the previous two goals to be achieved (Dougherty et al., 2014b).

Performance Funding Impacts
Dougherty and Reddy (2013) describe three levels of impacts of performance funding that institutions must progress through to see a positive effect of the policy. First, immediate impacts of performance funding are things such as institutions becoming aware of state priorities and their own performance relative to these priorities. Next, these immediate impacts must translate into intermediate institutional change such as modifications of institutional policies, programs, and practices that are designed to increase performance on state metrics. Finally, these intermediate impacts must result in changes to ultimate student outcomes such as increased graduation rates. It is these ultimate outcomes that performance funding policies seek to change.
In their meta analysis, Dougherty and Reddy (2013) discussed that several studies (mostly in the two-year sector) demonstrated that performance funding policies did make higher education administrators more aware of state priorities as well as led to a greater awareness of institutional performance in relation to the state priorities. Further, they found evidence that performance funding policies increase competition amongst the institutions. Some research on performance funding shows that higher education institutions change their academic and student service policies, programs and practices when performance funding policies are put in place. For instance, some colleges reported closing programs with low graduation rate or job placement success (Dougherty et al., 2014a). Institutions also reported changing course content and instruction as well as course sequences to try and increase course completion rates (Dougherty & Reddy, 2013). Further, some institutions reported increasing academic advising and job placement supports, and changing retention programs, registration procedures and student aid in order to increase retention and graduation rates (Dougherty & Reddy, 2013).
Many authors noted a concern that performance funding policies could create unintended outcomes, such as institutions favoring better prepared students and closing access to other students (Bell, 2005;Colbeck, 2002;Lahr et al., 2014;Naughton, 2004;Rutherford & Rabovsky, 2014). For instance, higher education administrators are concerned that performance funding policies would result in "creaming," where institutions choose to enroll more students who will perform well on the performance funding metrics (Dougherty et al., 2014b). Other higher education administrators suggested the possibility of weakened academic standards and restriction in admissions of underprepared students as a result of performance funding (Lahr et al., 2014).
As noted above, the call for accountability and efficiency in higher education has increased in recent years. Performance funding policies are a common outgrowth of this accountability culture. However, despite the increase in prevalence of performance funding policies, there is a dearth of research on the decision making process state leaders and institutional administrators engage in when in creating and implementing performance funding policies.

Theoretical Framework
Policy implementation research requires an understanding of the roles stakeholders have with each other and with the policy at hand. To answer the research questions, this study utilized Principal Agent Theory (PAT) to explore the relationships between institutional administrators and state leaders in the design and implementation of performance funding in the state of Michigan. While PAT began as an economic and political science framework, it is frequently used in higher education research to understand the relationships, beliefs, and power dynamics present between higher education institutions and governmental agencies and has frequently been used to understand these relationships within performance funding policies specifically (e.g. Hillman, Tandberg, & Fryar, 2014;Rutherford & Rabovsky, 2014). PAT is beneficial to higher education research as it provides a framework to understand the relationships and power dynamics that may be present between higher education institutions and state leaders and representatives.
Simply defined, PAT is where a principal enters into a relationship with an agent to perform a service that the principal cannot do alone (Lane, 2012). In the higher education context, the state (principal) enters into relationships with higher education institutions (agents) to educate and confer degrees on residents (service). PAT relationships may involve a formal written contract; however, in higher education, they often involve a more informal or implicit contract (Lane, 2012). This implicit contract underscores and guides the relationship between higher education administrators and state leaders and is pivotal to understanding the dynamics inherent in the creation and implementation of higher education policy. For this study, PAT helps highlight the significance of the relationship between higher education institutions and state legislators, the assumptions both parties hold regarding the relationship, and the competing priorities that may be placed on higher education institutions.
While the PAT contract may be implicit, the relationship between state legislators and higher education institutions is significant and strong. In public higher education, university charters or governmental relationships dictate an ongoing relationship between the principal and agent (Lane & Kivisto, 2008). The institution is intrinsically tied to the state through laws and regulations as well as institutional funding. Most public institutions rely on the funding they receive from state appropriations to continue operating (Mitchell, Palacios, & Leachman, 2014). On average, state and local appropriations makes up 53% of the revenue for public higher education institutions, (Mitchell et al., 2014). This relationship significantly influences how institutions and state leaders engage and interact with each other, as well as the power dynamics and abilities each party possesses. When exploring the creation of a policy, PAT helps provide an understanding of the unique relationship between the state and higher education institutions and the power dynamics and decision making abilities created.
PAT may also help explain the assumptions state leaders hold regarding higher education institutions, which may in turn influence the goals they have for performance funding. For instance, a frequent reason cited for implementing performance funding policies is the belief that institutions will pursue their own self interest or prestige over the needs of the state. In turn, legislators create performance funding metrics tied to specific state priorities to provide incentives for higher education institutions to pursue the priorities of the state as opposed to the institution's own self interests. PAT identifies this misalignment between the action of the agents and the preferences of the principal as shirking and attributes it to the agent pursuing their own self-interest (Lane, 2012). The belief that institutions engage in shirking gives context to the recent trend of state legislators adopting an accountability focused role with their higher education institutions. Additionally, higher education institutions are complex organizations that are often believed to make decisions and operate with limited transparency. In other words, the relationship between higher education institutions and state leaders is often marred by what PAT calls information asymmetry (Lane, 2012). Meaning state leaders do not have the same knowledge or specialized abilities as higher education institutions, which make it difficult for state leaders to accurately monitor the work or performance of the institutions. Utilizing PAT, and specifically the concepts of shirking and information asymmetry, can illuminate the assumptions state leaders may hold regarding the priorities of higher education institutions, which would influence how the legislators create and implement higher education policy.
Finally, PAT helps explain the competing priorities higher education institutions may face when making decisions. Public higher education institutions usually report to a board of trustees, but have multiple principals, such as the governor, state legislators, or higher education coordinating boards, that have authority or a controlling interest in the decision making processes of institutions (Lane, 2012). These multiple principals may have competing priorities, which the institution needs to consider and balance when making decisions. These competing priorities may lead to what appear to be inconsistencies between the desires of the state leaders and the actions of the institutional leaders. Understanding the relationship between the multiple principals and higher education institutions including the competing priorities, assumptions, and expectations the principals may place on institutions provides important context to policy research.

Case Study Context
In fiscal year (FY) 2011-12, as part of an effort to balance the state budget, the state of Michigan reduced university appropriations by 15%, or about $213 million. The following year, Michigan increased higher education appropriations for its four-year sector by 3%, or about $36.2 million, and restricted those funds to performance measures. Over the next three years the state continued to allocate additional money to higher education institutions via performance funding for a total increase of $132.7 million in FYs 2012FYs -13, 2013FYs -14, 2014FYs -15, and 2015. Each year, the increase in funding was a bonus over base funding amounts, but each year the performance funding increases were rolled into the base amounts for each institution for the subsequent year. Previous to the implementation of performance funding Michigan had "no permanent funding policy for allocating funding for operational support to the state's 15 public universities. Decisions about changes to university appropriations are made by the legislature on a year-to-year basis." (House Fiscal Agency, 2012).
Michigan's original performance funding formula included five metrics, some of which were funded in direct proportion to the value of the metric, while others were allocated based on a score in comparison to the institution's Carnegie peers. Eligibility for the funds were also tied to specific requirements including a tuition restraint provision, participation in reverse transfer agreements, and participation in the Michigan Transfer Network (HFA, 2012). Modifications to the calculation of the formula and the metrics used were made in the FY 2013-14 and FY 2014-15 budgeting process. The current formula includes the following six performance metrics: 1) Undergraduate degree completions in critical skill areas (direct proportion); 2) Six-year graduation rate (comparison score); 3) Total degree completions (comparison score); 4) Research and development expenditures (direct proportion); 5) Institutional support expenditures as a percentage of total core expenditures (comparison score); and 6) Pell Grant eligible students enrolled (comparison score). All but the Pell Grant metric were included in the original formula. For comparison score metrics, the original formula gave one point if an institution was above the national mean, two points if they were in the top 20%, and three points if they demonstrated improvement over the last three years (HFA, 2012). The current formula awards points in the following manner: three points if an institution is in the top 20% nationally, two points if it is above the national median, and two points for improving over a three-year period (HFA, 2014).
Michigan was chosen as the single state to examine for several reasons. First, to date, no study has examined Michigan's performance funding policy, whereas numerous studies have examined other common states such as Tennessee (Dougherty et al., 2014b;Dougherty & Natow, 2009;Hall, 2000;Lahr et al., 214;Sanford & Hunter, 2011), Ohio (Dougherty et al., 2014;Lahr et al., 214;O'Neal, 2007;Schaller, 2004), Florida (Bell, 2005;Dougherty & Natow, 2009) and Indiana (Dougherty et al., 2014;Lahr et al., 214;Umbricht, Fernandez, & Ortagus, 2015). Second, at the time of research, it was one of the most recent states to implement a performance funding policy. As noted above, most research on performance funding implementation occurs several years after the policy is implemented (e.g. Dougherty & Natow, 2015;Hall, 2000;McLendon, et al., 2006;O'Neal, 2007). The recent implementation of performance funding in Michigan allowed me to interview stakeholders less than two years after implementation as well as find a plethora of relevant documents related to the creation of the policy. Third, Michigan's lack of a higher education coordinated or consolidated governing board presents a unique policy creation and implementation process. Previous research has shown that state higher education governance structures influence performance funding policies (Dougherty & Natow, 2015;McLendon et al., 2006). As one of only two states without a higher education governing board, research on Michigan's performance funding policy allowed the exploration of how relationships and advocacy coalitions form without formal or legal authority connecting institutions to each other or the state government.
One key political context is also important to note in regards to Michigan during the implementation of performance funding. Similar to many other states that implemented performance funding policies, Michigan had a Republican controlled government and a recently elected Republican governor. However, in Michigan the governor made it a priority to implement state based performance assessments across various public entities through his Michigan Dashboard project. These dashboards were implemented to measure Michigan's performance in key areas including economic health, education, quality of life and public safety. While performance funding is not a part of this dashboard process, the promotion of performance assessment across public entities created an important opportunity for the performance funding policy creation process.

Methods
To understand the implementation of performance funding in Michigan, I utilized a case study research design. A case study approach allows for an in depth analysis of a complex topic, in this case, the performance funding policy in the state of Michigan (Yin, 2014). In particular, this study used a single case design to explore the research questions, with the state of Michigan as the unit of analysis. I interviewed higher education administrators and state leaders who were involved in the creation of the performance funding policy in Michigan. This research structure is frequently used in performance funding research (e.g. Dougherty & Natow, 2009 and ensures the voice of multiple stakeholders are represented. Three institutions out of 15 total in the state were selected through purposeful sampling. These institutions were chosen to ensure a variety of Carnegie classifications, total budget amount and percentage of budget that came from the state, level of involvement in creation of the policy and winner/loser status from the first year of the policy implementation. Within these institutions, purposeful sampling was again employed to interview senior administrators who either were directly involved in the policy creation process or whose role related to the implementation of performance funding (e.g. the governmental relations office, the finance office, or the president). Previous research has found that performance funding policies struggle to be disseminated across institutions, with senior level administrators familiar with the policy but lower level administrators and faculty not aware of the specifics of the policy (Burke, 2005;Dougherty et al., 2014a;Freeman, 2000;Hall, 2000;Schaller, 2004). Therefore, this study focused on senior level administrators and excluded faculty and lower level administrators. Additionally, participants were selected from the state and business sector who played a role in the creation or promotion of the performance funding policy. Publically available documents and snowball sampling were used to identify these participants. Specifically, participants were asked at the end of the interview to identify people they remember being involved in the policy design process.
In total, 14 participants were interviewed: nine administrators at three institutions and five state and business representatives. In order to protect confidentiality, the specific roles of the participants and institutions or state and business organizations interviewed will not be listed. Instead all results will simply identify participants as "university administrators" or "state leaders." University administrators included senior administrators in the areas of Finance, Government Relations, Academic Affairs, and Planning. The institutions included two research institutions (out of eight in the state) and one master's university (out of six in the state). State leaders were selected based on their involvement in the policy design and implementation process. Specifically, the leaders interviewed were selected based on recommendations from people familiar with the policy design process and were the names that repeatedly arose as those most involved and knowledgeable about the policy design and implementation process. These state leaders included state level education officials, legislative staff, members of the business community, and other stakeholders involved in the policy design process. It should also be noted that one further state leader was asked to participate, but declined as they felt that "my participation is too specific to not be identifiable." Semi structured interviews were conducted with each interview lasting between 45 and 90 minutes. The interview questions focused on two large themes: 1) How the performance funding policy was created and implemented and 2) What the decision making process entailed.
Additionally, data was collected from publically available records on state websites as well as the websites of the participating institutions and the business organizations. This data included meeting minutes, legislative documents, reports, and financial documents. Further, many participants provided documents (including notes, reports, presentations and memos) during and after the interview, which were also utilized in data analysis. This data helped provide supporting documentation to the statements and beliefs of the state leaders and university administrators.

Data Analysis
Thirteen of the 14 interviews were recorded and transcribed verbatim. One interview was not recorded and in this case, the notes from the interview were utilized in data analysis. Any quotes and ideas pulled from this interview were given to the participant to ensure validity. After transcription, all interviews were reviewed, analyzed and coded for themes using N'Vivo. All documents were reviewed, analyzed, and coded as well. Initial codes were developed based on previous literature and preliminary conversations related to Michigan's performance funding policy. Codes were added, removed, and grouped as necessary. Codes were analyzed for themes that were present across multiple data points and pattern coding (Miles & Huberman, 1994) was used to combine codes into a smaller number of themes and constructs. Multiple sources of evidence were used to triangulate the data, including the interviews, documents, and contextual data. Specifically, codes across interviews and documents were examined to confirm information and groupings. If contradictory or confusing data was found between the interviews or between an interview and documents or contextual data, the interview participants were contacted for a clarification conversation.

Results and Discussion
Coding the interviews and supporting documents led to the emergence five large themes: 1) The importance of advocacy coalitions; 2) Securing support through a focus on higher education affordability; 3) Concerns with how to measure data and compare institutions; 4) Insufficient financial incentives; and 5) Limited impact on institutional decision making. Each of these are presented below with data supporting the theme presented first, followed by interpretations and discussion.

Advocacy Coalitions
The most consistent theme that emerged regarding implementation was the importance of creating advocacy coalitions with business and industry as well as institutional peers. Many participants noted the key role the local business organization, Business Leaders for Michigan (BLM), played in promoting the policy and ensuring its legislative success. BLM is the state's business roundtable with membership comprising the CEOs of the largest businesses in the state. They strive to create strategies, policies and initiative that benefit the state's economy. Most of their initiatives are structured around their Building a New Michigan plan.
Many administrators mentioned knowing that BLM had the governor's ear and how important it was to ensure the institution's desires were heard by the organization. One university administrator stated, "We leveraged BLM to try to champion the ideas we came up with." Another mentioned putting together a proposal that had "broad base support from the institutions" that "they could work to sell to the BLM." Another university administrator believed that the policy took its particular shape "largely through BLM and their support of the model." Clear consensus emerged amongst the university administrators that the key to successful advocating your institution's desires for the performance funding policy was through BLM's support. Six out of nine administrators interviewed mentioned BLM's strategic role in the creation process. The other three university administrators did not mention BLM as a policy influencer and instead focused on institutional peers they felt had a significant role in policy formation.
The state representatives all identified BLM as having a large influence on the final policy that was created and the dominant role the organization played in getting the policy passed in the legislature. One state leader mentioned, "[BLM] were the ones behind Governor Snyder saying, 'here is legislation we need to change,' moving him forward. They were the real backers." Another stated, "We need BLM, the governor is going to listen to those CEOs." It was widely believed that without the support of BLM the policy would not have passed as easily or swiftly. Additionally, the specific structure and metrics used in the final policy was frequently attributed to BLM's support. Smith and Larimer (2013) believe that it is not unusual for one organization to have such power in the policy process. As was apparent to participants in Michigan, building networks of support around these organizations is seen as key to legislative success. Previous research has found that business leaders often play a role in the adoption of performance funding policies, though not in all states (Dougherty & Natow, 2015). The strong role BLM played in the creation of performance funding is not unusual given the priorities of the organization. BLM believes higher education is critical to meeting their goals of growing the state economy (BLM, 2011). As part of their turnaround plan for Michigan, BLM highlighted the need to increase funding for higher education institutions to a level that would move the state into the top ten for higher education appropriations (BLM, 2011). However, they tied this promotion of increasing higher education funding to the need for outcome based performance metrics.
Several university administrators also discussed the importance of reaching out to their peers to try to form advocacy coalitions and campaign for similar metrics. These administrators knew that forming constituency groups would allow them to more effectively advocate for metrics that would paint their institutions in a favorable light. One university administrator wanted to: Get universities to come to agreement on the formula. Not have to fight over it. If it is based on metrics we all come to agreement on then we can focus on… getting the legislature to put more money into higher education. Of important note, university administrators expressed the importance of working across institutional types, knowing that the coalition would be more effective if demonstrating support across institutional categories. When discussing which institutions their institution aligned themselves with, a university administrator said, We always want to be well aligned with [Institution] because they have a similar enrollment profile to us. Then we chose [Institution] for a lot of reasons, mostly because [of their geographic location], which is a very important district politically.
However others expressed that peer advocacy coalitions were not in their best interest due to their unique mission or values. They felt that the metrics they would hope to include in the policy would be unique or different from the ones promoted by other institutions. For instance, one university administrator noted, "We are focused more on access and wanted metrics that would include input adjusted measures, others would not be supportive of those metrics." According to Sabatier and Jenkins-Smith (1999) advocacy coalitions, consist of groups of people or organizations with shared beliefs that work together over a long period of time and coordinate their activity in response to a particular policy. The guiding principles that bring an advocacy coalition together are shared core beliefs even if there is some disagreement regarding the policy in discussion. The institutional advocacy coalition that formed in Michigan in response to the creation of performance funding support this belief. Many of the institutional peers that university administrators reached out to were long term alliances that had been utilized in the past during other policy discussions. Similarly, certain university administrators mentioned not participating due to distinct differences in the core beliefs of the institutions or disagreement regarding the core values of higher education.
While the university administrators clearly had motivation to unite and form coalitions, some state leaders were also aware of the need to work with institutions in the policy design process stating, "we need to move forward with something, we can't do this alone." Another state leader noted, We certainly made a concerted effort to involve all our universities. Were they all involved right from the beginning? No. It is difficult to get all 15 in a room from the beginning and come up with a set of metrics they all can agree to. So we started with a smaller group and then went from there. We took suggestions and made a couple changes. They all had input.
Some university administrators noted these dual involvement levels, expressing frustration at feeling that the policy "caught us off guard" and that there was "little communication or information ahead of time." This model of including the institutions in the design process fits what Birkland (2011) calls a "bottom up approach." This method of implementation values compromise between the goals of the policy creators (state legislators) and the norms and motivations of those implementing the policy (university administrators), with a focus on bargaining in order to maximize the likelihood of policy implementation and success. Given the dynamics at play between institutions and the state, this bottom up approach potentially helped relieve some of the tensions institutions may feel when state leaders try to create policy related to their operations.

Securing Support
This bottom up approach continued beyond policy formation with state leaders focusing on securing support from institutional administrators as well as the public. Both state leaders and university administrators commended the state for reaching out to institutions following the first year of the policy change. The state legislators invited institutions to provide feedback on the policy and what changes they would recommend. Participants noted the benefit of this process and the importance of having flexibility within the policy. One university administrator said, "We had a chance to respond directly as an institution about some of the things we wanted to see, so we had sort of a core set of principles that as an institution we thought was important." A state leader said, "It takes multiple years to sort of get it right…were the ones who were really performing the ones who were getting the funding?" Several changes were made after the first year based on the feedback received including: weighing the funding by the number of undergraduate students, reformatting how the rankings were determined, and adding new metrics. Rutherford and Rabovsky (2014) noted that a frequent reason accountability measures struggle is that policy makers fail to grant freedom for experimentation and innovation in the policy. Michigan allowing, and even encouraging innovation in the policy from both state legislators and higher education institutions helped ease the implementation process and increased the buy in from stakeholders such as higher education administrators. However, one university administrator expressed skepticism regarding the state reaching out, stating, "I'm not sure if the changes were actually in response to our feedback or because of public and political feedback." Another expressed concern with who the changes benefited, "They say they were trying to be more equitable, but who do these changes really benefit? Not [my institution]." Beyond creating a policy that higher education leaders would support, Michigan strived to create a policy that would have broad appeal and support in the public and to other stakeholders. State leaders expressed the importance of securing public support of the policy and structuring the promotion around college affordability and economic improvement. One state leader stated, "we try to make the case it's not just about giving money to institutions; it's about helping the economy" because residents see the economy as a higher priority than higher education funding and they wanted the policy to appeal to voter priorities. Beyond just words regarding the promotion of the policy, a state leader mentioned that making college affordable was a key part of the discussion in the actual creation and promotion of the policy. In particular, this was the reason the tuition restraint provision was included to qualify for the performance funding resources.
University administrators echoed this focus on promoting the policy as making college affordable with one administrator believing that the performance funding policy was created partially to help assuage the public concern with the rising price of higher education. Another university administrator believed that performance funding was, "inherently a political tool to show that [legislators] are taking charge of higher education." Even when creating the metrics, there was attention to selling the policy to the public with an emphasis on college affordability. One university administrator mentioned that when selecting metrics it was important to "pick ones that made sense to the public" and helped "demonstrate that something was being done about the cost of higher education." Jones and McBeth (2010) posit that narratives, symbols, and stories can be used to alter public perception and ultimately achieve policy change if these narratives are persuasive, fit current beliefs and come from a trusted source. By framing performance funding within the narrative of college affordability, proponents of the policy were able to appeal to residents who already were concerned about the price of higher education. Further, by broadening the focus of higher education funding to focus on economic improvements, policy makers are able to bolster support for higher education funding by showing a direct benefit to all residents.

Data Concerns
Previous research has found that higher education leaders hold significant reservations regarding the appropriateness of various metrics used in performance funding policies (Dougherty & Reddy, 2013;O'Neal, 2007;Sanford & Hunter, 2011). Almost all participants in this study (both university administrators and state leaders) echoed those findings and discussed concerns with how the data should be measured and analyzed in the policy creation process. Concerns revolved around the use of Integrated Postsecondary Education Data System (IPEDS) data, the way data was measured, and how to account for institutional differences.
One state leader expressed concern about using IPEDS data. "IPEDS data is always two years old…but they didn't want institutions to provide their own data, [they] wanted IPEDS for consistency." Another state leader expanded on the concern with IPEDS data saying, "Everyone was conscious of the fact that these measures aren't perfect but we had to settle on something so these seemed as good as anything." A university administrator had broader concerns with using only quantitative data stating, "A number in and of itself only tells part of the story." These concerns are not unique to Michigan. Previous research has shown administrators have great concern utilizing IPEDS data in performance funding formulas which, due to the nature of how IPEDS measures these metrics, offer only a limited slice of the students attending four year institutions and largely ignore transfer and part time students (O'Neal, 2007).
Participants expressed concern regarding the fact that different ways of measuring the data could result in different winners and losers in the policy. For instance, should the metric count the raw number of students graduating or use graduation percentages? Clearly institutions would be rewarded differently depending on how the results were tabulated. One university administrator shared, If you are us, if you are [Institution], you want things counted as much as possible in an absolute way, not in a relative way, because if you count things in an absolute way then our size, anything related to scale, we win.
A state leader mentioned, "Some institutions will do better if it is a percent, others will do better if flat rate. So obviously each is pushing what will be best for them." Another university administrator elaborated saying that funding the metrics in an absolute way, "completely eliminates the impact of the relevant part of the formula." They continued, If you score a 5 and you have 40,000 students, your adjusted score is 200,000. If you score 10 and you have 2,000 students you just beat my score, you doubled me, but you are only getting 20,000. I beat you by ten times because of my size even though I underperform you in every metric.
While university administrators expressed different concerns with the data, all the administrators interviewed and most of the state leaders expressed at least some concern regarding the data. Michigan addressed these concerns by creating a formula that measured some metrics through direct proportion and others through a comparison score. However, institutions still expressed some frustration with how the comparison score was calculated. In subsequent years, changes to the formula included changing how points were allocated for the comparison score metrics and weighing other metrics by undergraduate enrollment. Related, previous research has shown a concern with performance funding is how to account for institutional differences (Dougherty & Natow, 2015;Hall, 2000;Hillman et al., 2014;O'Neal;2007;Rutherford & Rabovsky, 2014). Administrators often feel that performance funding fails to understand the various missions of higher education institutions and their varying student demographics. In this study, university administrators also expressed concern regarding how the policy would compare open access regional institutions with selective research universities. One university administrator stated, [Institution] is going to graduate 93% of its students. Not because of what they are doing, but because their students are that dang smart…at the regional institutions, they're attracting a different student population so the student performance metrics need to change.
Others noted that the performance funding metrics didn't align with the goals and mission of their institution. "We are about access so we're looking more at what are the marketable degrees and they may encompass other degrees that are not included in the performance funding policy." State leaders seemed to understand this concern with one mentioning that it was important not to "compare institutions within Michigan unless they are in the same Carnegie classification…instead compare them to their national peers." Another state leader discussed the importance of recognizing "the diversity of institutions" when creating the policy and another noted, "we don't want to compare universities to each other." This recognition resulted in a large discussion during the implementation regarding how institutions should be measured: to their historical performance, to others in the state, or to their peers? Michigan decided to compare institutions to their national peers (based largely on Carnegie classification) and utilizing a point system. One state leader explained the reasoning behind the policy utilizing peer comparisons, stating, "Of course [Institution] and [Institution] are going to have better graduation rates, but if you are competing against your peers then it norms it…we were trying to be equitable." Another state leader said, "What was important…which I don't think many other states have done…we are going to compare how our schools do compared to their peer groups across the country." Another believed that when you compare institutions to their national peers, "you get a much more honest assessment of exactly how each school is performing." Many participants praised the fact that Michigan took the time to review other performance funding policies and worked with institutional partners to ensure that the metrics used and comparisons made would be fair for all institutions.

Insufficient Financial Incentives
A frequent concern in the literature is whether the limited amount of money tied to performance funding in many states can actually provide incentive for institutions to change their behavior (Dougherty & Hong, 2006;Dougherty & Reddy, 2013;Jones, 2012;Rutherford & Rabovsky, 2014;Sanford & Hunter, 2011;Shin, 2010). In many states, the portion of funding tied to performance metrics is less than 5% (Sanford & Hunter, 2011). Such small amounts may not be viewed as meaningful incentives to institutions, especially in relation to other possible revenues (Jones, 2012;Sanford & Hunter, 2011). This study supports that concern as all university administrators believed the small amount of funding currently in the policy was not enough to significantly alter their institution's behavior. University administrators referred to the funding amount as, "a tiny piece of our action" and "a small deal." Further, in order to be eligible for the increase in funds from the performance metrics, institutions had to agree to keep their tuition increase that year below a prescribed amount. While most institutions chose to meet the tuition restraint and qualify for the funds, participants stated this was mainly to be seen as "in compliance with the policy" rather than a change in behavior inspired by the policy. One university administrator believed it was "more of a PR sort of thing." However, no university administrator was able to name a particular percentage or dollar amount that would be a tipping point to promote institutional change. Most university administrators also voiced concern that increasing the funding amount could lead to financial volatility or unintended consequences and that state leaders needed to be mindful of those concerns when increasing the amount of funding tied to performance funding. One university administrator mentioned, "Obviously the bigger part, the larger a proportion of our appropriation this model represents, the more motivated anybody is going to be to work within it. I don't know how that gets accomplished effectively though." Another noted, If the funding levels were large it makes you think twice. You do look at these things and the impact they have on your institutions, but you also have to think about your institution and the betterment of your institution and students.
Likewise, the state leaders reported that they felt the amount of funding was small and that they knew "3% is not going to do a whole lot." Another believed that it is hard to measure the success of the policy "until the numbers start to get bigger." They understood the amount of money tied to performance funding should be increased, but were unclear on where the tipping point came in regards to incentivizing colleges to change behavior without creating issues of volatility in funding. They emphasized that it was "important for the incentives to mean something" without being life or death for an institution. Or that it needed to be enough to get schools to say, "OK, I really need to start paying attention to these metrics" without creating unintended consequences. Additionally, almost all participants (both university administrators and state leaders) expressed their support of the policy including only additional dollars, which ensured limited financial volatility for the institutions. One university leader stated, "You don't want to miss out on the money, because it matters, but you are not putting the whole entity at risk because it is only new money." Further, performance funding policies assume that institutions are intentionally not performing at optimal efficiency or have misaligned priorities and are simply waiting for financial motivation to change their performance (Rutherford & Rabovsky, 2014). However, the university administrators interviewed stated that the policy did not alter their behavior tremendously as the activities the performance funding policy promoted, such as increased retention rates, were values the institution already held and were striving to improve. One university administrator stated, "it's information that we had already been monitoring. So we look at our success rate, we look at retention rates. That's nothing new to us." Another said, "We care about graduation rates, not because somebody put it in a formula, but because we actually care about graduation rates." Many university administrators felt that the metrics were factors that were hard to change and that the performance funding policy did not recognize that it would take more than a short amount of time or a small increase in funding to see significant change on the metrics. "They don't seem to recognize the multi-year effort to influence these types of measures." When referencing work their institution was already doing for retention rates an administrator wondered, "I just don't understand how we would do things that differently if more of our appropriation was tied to performance funding."

Impacts
This study found that Michigan's performance funding policy did result in what Daugherty and Reddy (2013) termed immediate impacts. Administrators expressed becoming more aware of state priorities due to the policy and having greater awareness of institutional performance in relation to the expressed state priorities. However, the administrators believed that no intermediate institutional change occurred as a result of the policy. When discussing how performance funding influenced decision making at their institution, one administrator stated, "I think it is basically none." They elaborated, We want to be compliant…qualify for whatever incremental funding we would be exposed to, but in terms of leveraging all those metrics…there is not a whole lot more you can do to pick up an extra $300,000 that you are not already doing.
Another stated, "Have these initiatives changed our institutional behavior? The answer is not really." A final university administrator summarized it best by saying, "There's not a whole lot we are motivated by in terms of [performance funding] other than not wanting the headline to be [Institution] is not eligible for state funding." One university administrator even questioned the purpose of the policy asking, "Is it about rewarding those doing well or trying to drive institutional change?" A few mentioned that they felt the original goal of the policy was the tuition restraint provision and not actually focusing on changing institutional behavior in the realm of the metrics. This finding is different than previous research, which found that performance funding policies resulted in significant intermediate institutional change including academic and student service policies, programs and practices such as closing programs with low graduation rates, changing course content or sequences, increasing advising, changing retention programs and changes in financial aid (Dougherty et al., 2014a;Dougherty & Reddy, 2013). PAT highlights potential decision making processes that could play a role in this finding. The close relationship in the policy design process between university administrators and state leaders could imply an alignment of beliefs and values with limited concern on the part of state leaders regarding universities shirking due to their own self interests (Lane, 2012). Within such an environment, university administrators would not need to make significant changes to policies and programs to improve their performance on the state metrics as their policies and programs would already be in alignment with the state priorities and therefore the performance funding metrics. Conversely, the implementation process could imply that university administrators are indeed shirking and taking advantage of information asymmetry. When designing and promoting specific metrics, it is possible that university administrators placed their own self interests forward and recommended metrics that they knew would highlight their institution in a favorable light. Within this policy environment, university administrators would again not need to make significant changes to policy or practice as their institution would already be performing well on the performance funding metrics. Second, public higher education institutions have many stakeholders (or principals in the PAT framework) that they are accountable to including a board of trustees, state legislators, and their students. These stakeholders may have different priorities and therefore exert different pressures on university administrators (Lane, 2012). When these pressures are competing with the performance funding metrics, university administrators may determine that it is in their best interest to not make adjustments to university policies and practices in order to perform better on the performance funding metrics, and instead spend money and energy on different programs and policies that will meet the needs and desires of other stakeholders, especially given the limited financial resources provided by the performance funding policy.
While the policy did not seem to have the desired impacts during the course of this study, it did raise concerns regarding potential unintended outcomes. University administrators interviewed in this study noted the possibility of weakened academic standards and restriction in admissions of underprepared students as a result of performance funding. However, it was only mentioned as a possibility with no participants stating it had been an intentional decision at their institution. One university administrator ruminated, How would you engineer your class differently to achieve [higher graduation rates]? I know one way to do that but it leaves out a lot of the people that we espouse to be very important here in terms of access to higher education. So you can say we are going to make [performance funding] 10% of your appropriation, fine, no problem, but you are creating another problem elsewhere.
Another university administrator believed, "if graduation rates matter, you are going to be a little more careful on who you enroll." A third, discussing performance funding policies in other states, mentioned, "Some places it is all about retention rates, that's an easy metric to manipulate." While no participants noted changes to enrollment profiles, alterations to the performance funding policy in subsequent years did include the addition of a new metric that measures the number of Pell eligible students enrolled, which some may view as a precautionary step to promote access for low income students. This finding is similar to previous research that also noted a concern that performance funding policies could create unintended outcomes, in particular institutions favoring better prepared students and closing access to other students (Bell, 2005;Colbeck, 2002;Lahr et al., 2014;Naughton, 2004;Rutherford & Rabovsky, 2014).

Limitations
This study is limited in two key and important ways. First, this study was interested in the decision making process of those involved in the creation of Michigan's performance funding policy and therefore only participants who were involved in the creation process or university administrators who worked in areas related to the policy (e.g. finance, government relations) were selected to participate. While this helped answer the specific research questions of this study, it left the voice of significant areas of the university administration, most notably faculty, out of the research and results. Additionally, as this research was exploratory, only three institutions were included. However, the institutions selected are representative of the 15 institutions in the state in regards to geographic location, Carnegie classification, and reliance on the state for revenue.
Second, as noted above, the study intentionally occurred soon after policy implementation, with the focus on policy implementation and immediate outcomes. Therefore, the full effects of the policy, both intentional and unintentional, may not be seen. As previous research (e.g. Hillman et al., 2014) has demonstrated that institutional impacts may not be seen till many years post policy implementation, these results should be viewed simply as immediate outcomes with the understanding that future research may find more or different outcomes.

Implications and Conclusion
While the case study nature of this research prevents generalizations to other states, several implications and best practices for policy makers can be identified from this research. First, it can be extremely beneficial to include institutional representatives in the policy creation process. Numerous participants noted the benefit including these voices in the policy design process had on the eventual success and smoothness of implementing the policy. Of particular benefit was the opportunity for institutions to provide feedback after the first year of implementation and the corresponding adjustments that were made to the policy. The uniquely involved role of university administrators in the creation and implementation process led to the implementation of many of the metrics supported by the university administrators in the final policy. Allowing institutional leaders a voice in the policy design process can help limit shirking by aligning goals of the state with feasible actions by the institutions (Lane, 2012). Further, by allowing institutions to participate in discussions around metrics and performance, the state can become more aware of the processes involved in measuring the effectiveness of higher education, therefore hopefully decreasing the ability for higher education institutions to take advantage of information asymmetry.
Second, when designing metrics, great care should be taken in selecting what measures will be used and how institutions will be scored on those measures. The most contentious debates in creating Michigan's performance funding policy arose in how the state would measure institutional success in its performance funding model and what data would be best used to understand how the institutions were performing. In particular, significant discussion and consideration should occur in regards to how to score institutions on the performance metrics. In this realm, states could learn from Michigan's decision to compare institutions to their peers and give points based on an institution performance in comparison to those peers.
Third, in order to prevent unintended outcomes alluded to by university administrators in this study, states with performance funding policies should take care to protect the admissions of underserved students. In recent years, several states have implemented metrics that hope to counteract the desire of higher education institutions to engage in creaming. These metrics often include weighing underserved students differently in progression and completion metrics or a separate metric specifically focused on underserved student enrollment. States that have implemented such a weighted formula include Tennessee and Texas. Adding this extra weight to at risk students helps reduce the likelihood that institutions will simply reduce the enrollment of at risk students to improve their performance on the funding metrics.
Finally, it is important for policy creators to be aware that institutional change can be slow and the effects of a policy may not be seen in the first few years post implementation. Participants in this study believed that the performance funding policy did not alter their behavior for a variety of reasons. The slow nature of change at institutions could cause complications with performance funding policies that often push to see results and change quickly. Policies that could have financial implications for institutions (such as performance funding policies) should allow for a few years where institutions are aware of the changes they need to make and be given the chance to do so without financial repercussions. In fact, the CCA (2012) argues for a period of time where institutions can familiarize themselves with the metrics and how they would perform with limited financial impacts. Likewise, it is important to note that policy effects are often not seen till many years post implementation. Therefore policy makers should take great care to not make decisions based on information and results seen in the first few years post policy implementation. Readers are free to copy, display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for noncommercial purposes only, and no alteration or transformation is made in the work. More details of this Creative Commons license are available at http://creativecommons.org/licenses/by-nc-sa/3.0/. All other uses must be approved by the author (s)