Policy Implementation Entrepreneurs: A Review and Case Study of Constrained Programmatic Innovation

Policy Implementation Entrepreneurs:

A Review and Case Study of Constrained Programmatic Innovation

Abstract: This paper conceptualizes policy entrepreneurs in the implementation and administration of public programs, borrowing from the considerable research into entrepreneurship in policy adoption and diffusion.  The institutions and policy environment constrains the activity, but entrepreneurs are able to facilitate innovation without engaging contentious political forces.  A case study of a university-led partnership to improve middle school science education in traditional public schools through novel comprehensive curriculum reform provides evidence of the existence of policy implementation entrepreneurs.  By developing an analysis of characteristics of these leaders, it provides a means of understanding their role in this stage of the policy process.   

Key Words: STEM Education, Implementation, Education Policy, Program Evaluation, Policy Innovation

I. Introduction

In a society where entrepreneurship and innovation are aspirational goals that politicians, leaders, and cultural figures celebrate (Lindgren 2013), scholars of multiple disciplines have studied these concepts (Ireland & Webb 2007)- including how entrepreneurship drives adoption of policy changes (Mintrom & Norman 2009)- but have yet to focus attention on the purposeful diffusion of innovations in the implementation of public programs.  Entrepreneurial individuals and organizations seeking to build new programs or procedures and disseminate their results for the application in other similar arenas face a different set of risks and policy constraints in improving implementation and relevant outcomes.  This article uses a review of the literature and a case study to develop a frame to answer the question: what is the opportunity for policy entrepreneurship in the implementation phase of the policy process in innovation?  This is not a discussion of simply improving bureaucratic efficiency or fixing flaws of a program at the margin, but it is a theoretical and case study analysis of how implanting agents and organizations can create and expand new ideas within the context of existing systems and regulations.    

            

Entrepreneurs are actors- individuals or organizations- who devote resources with an element of risk with the goal of innovation.  In the private sector, there are financial costs and profit motives.  In the public and non-profit sector the goals are complex and innovation can involve a change, an improvement, to policies and programs.  The risks can be financial, but they can also involve political or reputational impacts.  As this paper will show, in policy implementation entrepreneurs are those actors changing or incubating changes in procedures, systems, and institutions within the political and legal structure and diffusing ideas for augmenting programs within existing structures.  These are not people who are simply more efficient or are simply bringing in best practices in the form of the New Public Management or other paradigm shifts, but they are active agents purposefully building towards outcomes in implementation.

            

Policy entrepreneurs in implementation can exist in multiple arenas of policymaking.  The “distributed” nature of education policy (Spillane, Harrison, & Diamond 2001; Spillane 2006), with street-level bureaucrats playing a determinant role in achieving desired outcomes (Lipsky 1980), makes it a useful testing ground for the ideas put forth in this article.  Ongoing reform efforts, particularly in science, technology, engineering, and mathematics (STEM) teaching and learning flow through multiple levels of governance and into American classrooms.  Policy entrepreneurs are seeking to design, test, implement, and disseminate best practices that serve client- in this case students and communities- to improve educational outcomes.  This article focuses on one example where a grant-based curriculum design and research program is implementing a comprehensive middle school STEM course within state and federal standards with a new paradigm of problem-based learning.

            

This paper will first introduce the concept of policy entrepreneurship and explain how this can meld with implementation research.  After developing the characteristics for analysis through the literature, it will explore the STEM education context before focusing on the case study.  Understanding policy entrepreneurship in implementation cannot only expand public policy scholarship, but it can help policymakers design opportunities to drive innovation at this stage of the policy process.              

           

One of the key purposes of policy scholarship is to identify change- particularly change that strays from incrementalism- in the public policy process.  To establish causality researchers seek to identify the causes and agents of these notable changes.  Policy entrepreneurs can serve as these change agents.  Kingdon’s (1984) multiple streams theory, for example, argues that it is policy entrepreneurs who identify an open policy window and connect the problem, politics, and policy streams shift the policy agenda and resultant legislation, regulations, or processes.  Mintrom (1997) built his theory of policy entrepreneurship around policy actors who innovate internally and disseminate policies in the diffusion process- based on Berry and Berry’s (2007) work on this pattern of policy change- spreading ideas for adaptation in other jurisdictions.  Policy implementation entrepreneurs are finding an open policy window within the constraints of implementation procedures and networking to diffuse innovations to other programs for reform.  The next section identifies the characteristics of policy entrepreneurs as they exist in all of the stages of the policy process.

II. Implementation Entrepreneurs and the Diffusion of Innovations

The policy literature is replete with studies and theoretical arguments on the purpose and characteristics of policy entrepreneurs.  Research into policy implementation is a vexing undertaking that scholar struggle to understand.  This section will attempt to merge the ideas about entrepreneurs with ideas about implementation to develop the underpinnings of empirical analysis into these policy actors.   

            a. Review of Policy Entrepreneurship

For all the theory related to overarching systems and institutions, public policy change and successful outcomes often boils down to the people within organizations.  Meier (2009) reflects on the public policy field:

           

            After spending a great deal of time studying the performance of    educational institutions, I am consistently struck by the basic fact that high performing schools and low-performing schools are often using the exact same programs in the same environment with relatively equal resources […].  The key difference is not the programs, but the people.  In short, every program in education policy has a massive Hawthorne effect linked to the organization itself.

Policy entrepreneurs are people and organizations and while not all policy and programmatic change has a positive impact, innovation is often a step forward for successful outcomes.

            

Policy entrepreneurs are not just politicians.  In fact, this class of actors can include lobbyists, policy analysts, and other key players in the policy process.  They can be bureaucrats (Roberts & King 1991), legislators (Mintrom 1997), executive branch politicians, and non-governmental organizations (Koski 2010).  Roberts and King (1991), through a series of interviews, establish categories of entrepreneurial activities:

·       Idea generation activities;

·       Problem framing activities;

·       Dissemination activities;

·       Strategic activities;

·       Demonstration project activities;

·       Activities cultivating bureaucratic insiders and advocates;

·       Collaborative activities with high-profile elite groups;

·       Activities enlisting support from elected officials;

·       Lobbying activities;

·       Activities attracting media attention and support;

·       Administrative and evaluative activities.

Mintrom and Norman’s (2009) more recent review identifies displaying social activity, defining problems, building teams, and leading by example as the role of entrepreneurs.  While this article analyzes these activities in the context of the policy process- particularly in policy change and adoption- these characteristics are also part of the implementation game.  The authors ask, “By what means can advocates of policy change come to have broad influence?”  Implementers can serve as these “advocates.”

            

Implicit in policy entrepreneurship is leadership; actors cannot innovate unless they can unify and find followers. Spillane (2006) states in the education policy arena that schools are complex organization and require a move away from heroic leaders.  He defines, “a distributed leadership perspective moves beyond the Superman and Wonderwoman view of school leadership.”  Fullan (2006) argues that education reform efforts require more significant action through transformational systems thinking that sustains contextual changes.  He notes the importance of networks and “co-dependent relationships” for leaders to achieve significant results in the world of education.  Vexing policy problems often require leaders that can work in these non-linear systems to achieve innovative reforms and education policy (Spillane, Resier, & Reimer 2002).  Leaders do not necessarily need to hold positions of power, as innovations can come from managers, mid-level employees, and those working on the front lines of policy interventions (Borins 2001; Damanpor & Schnier 2008; Spillane & Kim 2012). 

            

Leadership and entrepreneurship, however, are not synonymous as leaders may follow a successful course, while entrepreneurs risk resources, innovative, and diffuse policy and practice changes, disrupting the status quo in their venues.  Mack, Green, & Vedlitz (2008) delve into the attributes of policy entrepreneurs in a review of the research of how they work.  They attempt to quantify their role and empirically test entrepreneurship in a case study, concluding that “Those individuals with tighter linkages to their community are more likely to be the most active public entrepreneurs in other public entities and nonprofit organizations.”  This adds embeddedness and understanding of the institutional environment as characteristics that can turn leaders into innovators and entrepreneurs in this context.  It’s worth noting that not all attempts at innovation find success (Marsh & Edwards 2009) even if the research bias is for affirmative case studies. 

            

There is debate whether this drive from policy entrepreneurs from non-elected officials is consistent in a representative democracy, with roots in the generic public administration discussions of administrative discretion in governance and implementation.  Denhardt and Denhardt (2003), for example, express reservations with entrepreneurship as encouraged in the New Public Management because it shifts governmental actors away from a public service as “serving” “citizens” towards “steering” “clients.”    While there are dangers to the democratic process in having officials and administrators innovate without political controls, deLeon (1996) writes in her thorough examination of the ethics of public sector entrepreneurship, “Without prudent risk-taking and experimentation, without pushing the envelope of current practices, government will be unable to realize the dramatic increases in productivity necessary if it is to recover the public trust and confidence.”  The deLeon article focuses on government reinvention for the purpose of efficiency and effectiveness as the purpose of innovation and includes pilot programs as an opportunity to innovate without overstepping pluralistic values. 

            

As other studies and the empirical analysis of this paper show, pilot programs are often at the center of policy entrepreneurial activity.  Pilots are easy to start and difficult to replicate, but they are a relatively low-cost opportunity for experimentation in implementation (Durose 2009; Marsh & Edwards 2009; Petchy, Williams & Carter 2008; Dodgson & Scaggs 2012).  As per usual, scholars of diffusion must be wary of a selection bias.  Maranto and Wolf (2013) attempt to explain the lack of diffusion of innovations in “impossible jobs” such as teaching and policing and find that in these areas the role of the leaders with new ideas lose status as the quality of work reaches minimum politically desired levels.  It is the chance of failure, however, that can turn a manager or official into an entrepreneur.  

            

Before differentiating entrepreneurship in the policy science realm, it is also worth reflecting on entrepreneurship in multiple disciplines.  Ireland and Webb (2007) provide a thorough review on ten domains of academic pursuit through the dominant journals in that discipline where articles appear on entrepreneurship.  Their results show entrepreneurship as an expanding subject of academic research across disciplines since 1980, with management as the discipline most engaged in this area.  They note, “Similar to marketing, a significant amount of political science research examines the determinants and diffusion of policy innovation.”  Common themes they find in their broad bibliometric analysis include the individual-opportunity nexus, entrepreneurial risks, and identity construction.  All of these activities are part of the policy entrepreneurs’ roles as, like a marketing entrepreneur pushing a product into the market, they push their social, environmental, educational, other innovations in the policy sphere.  While this paper will not explore the psychology or sociology of entrepreneurship, it is clear that the instincts and challenges span a broad spectrum of human endeavor in improving processes and practices for a chosen objective.

            

Scholars need to further understand how policy entrepreneurs operate in their specific context.  Damanpour and Schneider (2008) use a regression analysis to explore the characteristics of public managers who facilitate the diffusion of innovation, getting into the personal attitudes of individuals in these positions.  While the large number of studies provides statistically significant data, their broad approach to these questions of entrepreneurship lacks specific contextual considerations of different policy environments. 

            

The works of James Spillane offer a window into leadership in schools- include entrepreneurial practices.  Affecting change beyond superficial alterations in complex organizations such as schools or myriad public and not-for-profit agencies requires willing participants whether the change arises from the top, middle, or bottom of the hierarchy.  Policy entrepreneurs are not pushing for incremental change, but rather what Spillane, Reiser, and Reimer (2002) consider a “level of change [that] represents loss for the implementing agent, in that it necessitates the discrediting of existing schemas or frameworks.”  This, for example, could mean changing from a traditional method of instruction to one pushing problem-based learning.  Their study, rooted in social psychology, notes, “What a policy means for implementing agents is constituted in the interaction of existing cognitive structures […] their situation, and the policy signals.  How implementing agents understand the policy’s messages about local behavior is defined in the interaction of these dimensions.”  (Spillane, Reiser, & Reimer 2002).  Particularly in the implementation phase, policy entrepreneurs need to consider the roles and interests of those involved in their innovation to enact meaningful change.

            

The context of the innovation and the constraints of the policy environment bear further discussion.  Regulatory pressures can narrow the scope and opportunity to alter formal and informal organizational routines (Spillane, Parise, & Scherer 2011).  It is, however, also worth noting that although there may be restrictions at the political or administrative levels, change can come from within the system and there are opportunities for bottom-up innovations that could diffuse across schools systems (Spillane, Halverson, & Diamond 2011; Spillane & Kim 2102).  Needless to say, there are ongoing tensions as policy entrepreneurs tackle the generalized resistance to change in the educational and other policy contexts.  The tasks of the policy entrepreneur are arduous.  The next section will describe the special challenge of bringing change in implementation in comparison to policy adoption. 

            b. From Adoption Phase Innovation to Implementation Stage Innovation

           

 The adoption of policy and putting a policy into action via implementation are separate phases of the policy cycle as widely presented in the scholarly and practitioner literature.  While the policy cycle model is not predictive and is an imperfect representation of policymaking procedures, there are fundamental differences in understanding the legitimating of laws, regulations, and programs and the activities of achieving outcomes through institutions, organizations, and policy actors (Peters 2013).  The core of the academic literature in policy diffusion relates to the adoption phase.  In education policy, for example, scholars have analyzed the development of charter school policies and have entrepreneurs have diffused these ideas and built coalitions for passage of relevant legislation at the state level (Mintrom 1997; Renzulli & Roscigno 2005).  Even at the city level, Marschall and Shah (2005) looked at “civic capacity” and the role of policy entrepreneurs in mobilization to put education reform ideas on the policy agenda. Actors in the implementation phase of the education process, however, also act entrepreneurial in developing and spreading stage-specific innovations. 

            

In general, there is less theoretical understanding of the implementation stage of the policy process compared to adoption.  Reflecting on the field, O’Toole (2000) notes that there has historically been a dearth, or at least a perception of a dearth, of research into implementation Robichau and Lynn Jr. (2009) argue that there is still a “missing link” as scholarly works, “do not conceptualize the distinction between policy outputs and policy outcomes, public policy theories tend to slight the administrative processes that constitute implementation.”  These authors, however, acknowledge the importance and ongoing growth of implementation research and its role in enhancing public policy as a field and occupation. 

            

One of the key debates about implementation, one that lies at the very core of the field stretching back to the writings of Wilson (1887), is the relationship between the political leadership and the administrative units that perform governmental services in American democratic society.  These issues of authority place limits on implementing agents that prevents them from acting in a manner incongruent with the political leadership.  There, however, exists flexibility and a level of autonomy for public management.  Hicklin and Godwin (2009) note, “Research has shown that in many cases, the implementation of the same policy instrument can result in vastly different outcomes across organizations.”  Policy implementation entrepreneurs do not necessarily become involved in the most contentious issues on the agenda.  In education, for example, they are not at the forefront of debates on choice, accountability, and salient controversial reform initiatives that plague political leaders in the educational policy subsystem (Ravitch 2010; Garelick 2012; Simon 2013).  They are, however, bringing reform within the political constraints through relevant policy tools. 

            

Salamon and Lund (1989) define, “A [policy] tool resembles an individual program in that it is a concrete mechanism for achieving a policy goal normally specified in legislation or manifested in identifiable organizations.”  Understanding the grant-based partnership is key to understanding the implementation of many innovative science and technology education programs, particularly those funded through the National Science Foundation (NSF).  Parties join a partnership to improve and expand on their mission and achieve their goals.  Parties to a partnership bring their own individual objectives and also form common objectives through joint activities.  Killion (2011) advises, “The sure way find and add value to each partner is to take adequate time to build the relationships with potential partners, assess potential partnerships, evaluate partnerships they enter, and avoid partnerships that might detract from their priorities and immediate needs.”

            

The NSF’s Math-Science Partnership Program (MPP) has been a useful area of analysis.  Scherer (2006) thoroughly reviews of the definitions and characteristics of partnerships.  In the introduction it makes the interesting point that, "For math and science education, the partnerships also may be functionally critical due to the dynamic nature of science, marked by the central notion of 'scientific progress,' and that changes in knowledge creates new demands." Foster et al. (2010), in an article highlighting successes of the MSP, note that although the partnerships between universities and schools are focused on improving K-12 instruction and outcomes, university faculty have also benefited in discovering innovations for their college-level instructional benefits, such as inquiry-based curriculum designs.   

            

With the NSF MPP serving as an incubator of entrepreneurship in science, technology, mathematics, and energy (STEM) education in grade school, the Center for Education Integrating Science, Mathematics and Computing at the Georgia Institute of Technology has developed the innovative Science Learning Integrating Design, Engineering, and Robotics (SLIDER) program as a pilot project for 8^th grade science classes in the State of Georgia.  The case study will show that, as with an adoption entrepreneur in public policy, this entrepreneurial unit that operates relatively independently of the university is fostering diffusion through an internally developed effort to change professional practice with a goal of disseminating and diffusing these ideas to the great STEM education community.  They are not looking to change politics or advocate for broad level policy adoptions, but are seeking to change how public education occurs and how teachers teach at the classroom level.  With NSF funding, they are affecting apolitical education reform.

III. Case Study of Entrepreneurship in a University-Led Partnership

At its core, the SLIDER project is about bringing problem-based learning through LEGO robotics into 8^th grade science classrooms.   The SLIDER program has evolved from the initial grant proposal to the NSF, but its focus remains the creation of a physical science curriculum for 8^th graders in Georgia using engineering principles through LEGO Robotics and researching its impacts and effectiveness in the classroom.  The goals of SLIDER outlined in its initial application are:

·       Goal #1—To develop a rigorous 8th grade physical science curriculum that uses engineering design and LEGO Robotics and Mechanics to teach physics-based concepts.

·       Goal #2—To implement the SLIDER curriculum in three schools, one rural, one urban, and one suburban, in a manner that enables the research staff to effectively conduct their educational research.

·       Goal #3—To conduct the SLIDER research project to determine student learning.

·       Goal #4—To train postdoctoral, graduate and undergraduate students in educational research and evaluation.

a.      The STEM Education Context

 A report of the National Governors Association (NGA) highlights opportunities to improve educational outcomes in states across the country.  To overcome STEM education failures it provides best practices of state programs that “Recruited and retained more qualified class-room teachers; Provided more rigorous preparation for STEM students; Used informal learning to expand math and science beyond the classroom; Enhanced the quality and supply of STEM teachers; [and] Established goals for postsecondary institutions to meet STEM job needs.”  (Thomasian 2011). There are a multitude of partnerships focused in the higher education sector on professional development and certification of teachers, with related research and evaluation into these programs (Rosenberg et al. 2009).  While these partnerships face similar challenges to grant-based programs and combine the resources of multiple actors through collaboration, they do not achieve unique objectives outside the traditional expansion and improvement of teacher training that is typically part of these institutions in the first place, with or without a formal partnership.  Unlike these partnerships and the major NGA recommendations, the SLIDER program is not looking to change the structure or personnel of modern education, but rather facilitate the enactment of a change in professional practice through problem based learning with minimal traditional instruction.

            

Problem based learning is a method of teaching students from kindergarten through graduate school that seeks to shift away from the traditional lecture and class work structures of American curricula.  Hmelo-Silver (2004) defines the method: “Problem-based learning (PBL) is part of this tradition of meaningful-experiential learning.  In PBL, students learn by solving problems and reflecting on their experiences.”  She lists the goals of PBL as allowing students to:

1)     construct an extensive and flexible knowledge base;

2)     develop effective problem-solving skills;

3)     develop self-directed, lifelong learning skills;

4)     become effective collaborators; and

5)     become intrinsically motivated to learn.  (Hmelo-Silver 2004).

Eisenkraft (2003) introduces a 7E model for inquiry learning in which the expectation is that teachers can achieve “transfer of learning” when they elicit prior understanding from the students, engage the students, and then have them explore, explain, elaborate, evaluate, and extend their knowledge.  Volkmann and Abell (2003) note that these efforts in STEM education can go beyond the “cookbook” laboratories in challenging the students to develop their courses of action in scientific problem-solving and learning.

Problem-based learning has many proponents, but it is not universally accepted as an effective improvement in scholastic education.  Kirschner, Sweller, and Clark (2006) argue that human cognitive architecture in not conducive to learning through “minimal guidance” instruction and declare, “Controlled experiments almost uniformly indicate that when dealing with novel information, learners should be explicitly shown what to do and how to do it.”  Schmidt et al. (2007), however, directly refute these notions of cognitive psychology and also write:

           

            [We] argue that problem-based learning is an instructional approach that cannot be equated with minimally guided instruction.  On the contrary, we contend that the elements of PBL allow for flexible adaptation of guidance, making this instructional approach more compatible with the manner in which our cognitive structures are organized than the direct guided approach […].  (Schmidt et al. 2007).

CEISMC, which also operates a full portfolio of other STEM education programs, recruited teachers in Georgia for the SLIDER pilot and has developed the curriculum, trained the teachers through professional development activities including an annual week-long summer institute, observed classes, and presented research results across the country.  While the program is ongoing, it has facilitated continued innovation and expansion of CEISMC activities even with the cited challenges of public school systems and strict and changing educational standards in the state.

b. Methodology

 The author of this paper as a member of the external evaluation team for SLIDER reviewed program applications, previous annual reports, publications, and website materials to understand the development of the project (See Appendix A).  The most important data collection this year involved semi-structured interviews with program staff (9 interviews in the fall, all conducted in person) and teachers (6 interviews in the winter- 4 conducted on site and 2 over the phone).  The themes of the protocol questions are below in Table 1.  In addition, the evaluation team attended project meetings, the 2012 Summer Institute, visited 2 of the 3 project sites, and met informally with program staff over the course of the year.

Table 1: Summary of Interview Protocols

Interview Subject	Key Question Themes
Program Staff	Individual Role on the Project and Activities to Date
	How SLIDER fits into Individual Portfolio of Projects
	Effectiveness and Relationships for the SLIDER team and the Overall Partnership
	Successes and Challenges of the Project Thus Far
	Ongoing Activities and Goals
	Personal Interest in the Project
	Work Product to Date
Teachers	Background Information
	Views on Problem-Based Learning and SLIDER
	Relationships with Georgia Tech and Program Partners
	Interactions with Other Teachers on SLIDER
	Perceptions of Students and Administration with regards to SLIDER
	Overall Progress to Date on SLIDER
	Challenges of SLIDER
	Robotics and Standards
	Expectations

The formative elements of the evaluation design focus on the inputs, partnership, and processes.  While outputs and outcomes informed the analysis, the ongoing research within the project will be informed by the program results in the outgoing year. The analysis followed established social science methods for program evaluation, referencing resources including Miles and Huberman’s (1984) work on qualitative data analysis and Nagel’s (2002) collected volume of guidance on policy evaluation.  Through qualitative coding of documents the evaluation team triangulated the data by source to establish the conflicts and consensus about program outcomes based on the criteria set out in the project guidelines.

c. Analyzing the Characteristics of this Case Study

CEISMC’s ongoing research has shown preliminary success in improved student performance, self esteem in science classes for students from a variety of demographic background, and interest in STEM with a causal relationship from SLIDER participation. A variety of challenges were identified by participants shaping the initial years of the project, however, including the following:

1.     There have been difficulties in establishing a stable set of partner institutions.  Due to difficulties with schools systems outside of the scope of SLIDER, the project partners stopped the project in two schools.  One staff member recalls, “We had one school that imploded early on [and] you cannot navigate around a school system that doesn't have a stable teaching force. “  One school relied heavily upon teachers drawn from the Teach for America program.  “[W]hen all of the teachers leave after one year and there's no guarantee that you're going to have stable teachers in place you can't […] have a 3-4 year implementation plan.”

2.     In the initial set of participating schools teachers had limited exposure to problem-based learning and instructional techniques.  This posed a significant challenge for both the curriculum development team and the research team to establish baselines for comparing progress towards installing and adapting the new curriculum to the specific context of schools.  There was also wide variance in the size of the classroom environments ranging from classes of 22 to classes of 40.  This meant that the curriculum design and problem based learning instruction would need to be robust enough to span the different class sizes.  The SLIDER project team also experienced wide variance in the receptiveness of teachers to problem-based learning approaches. 

3.     The initial project plan called for the use of Georgia Tech faculty and the student “fellows” to work with teachers, the project team, and the schools in the development and implementation of SLIDER.  However, over time the project team found it more effective to de-emphasize these elements of the project in favor of regular project staff who can maintain stable relationships with the teachers.

 The challenges associated with the start-up phase of SLIDER led the team to designate the first year of work as Year 0 to allow for progress at an appropriate pace.  Program staff considered this part of the learning experience in adapting a curriculum to the challenges of the modern classroom and has continued to run the program in light of these changes.  However, there have been important adjustments and tensions that arisen out of the change in the time scale for SLIDER.

To commence the activities, the SLIDER program developed, disseminated, and put into practice a “Launcher Unit” with two learning sets.  This was part of an iterative process to test the curriculum in the classroom to inform the Year 3 effort.  A conference paper provides an example of a difficulty and resultant change in the SLIDER curriculum. 

Halfway through the implementation of the DRK-12 Launcher unit it became clear, based on classroom observations and teacher feedback, that in a physical science class that is required to meet defined curricular standards, and where robotic building and programming are definitely not part of those standards, there is not enough time for students to realistically master LEGO NXT build and programming skills.  Teachers were under tremendous pressure to “get to the content” that would be tested on benchmark and national standardized tests —i.e. the science disciplinary core ideas.  It also became clear that the activities that took the longest and were done in groups were the most problematic because they took substantially longer in large, chaotic classes than they did in  pilot situations.  We therefore cut short the classroom implementation and the second half of the Learning Set, which taught LEGO programming skills, was not implemented with students in the school.  (Usselman et al. 2013)

            

The Launcher Unit was a basis for discussion at the Summer 2012 Summer Institute, with teachers and program staff observing the classes and examining student artifacts for lessons learned.  These lessons ranged from dealing with the competitive nature of students in the robot jousting to the impacts of a dirty classroom floor on the speed of the LEGO to pedagogic discussions relating to how the students experience the curriculum and understand the “real world” examples.  Fasse and Hendricks (2012) reports on the impact of the Launcher Unit in finding, “students had higher self-efficacy in science and had better attitudes toward science when they were involved in the SLIDER problem-based curriculum that utilizes engineering design as opposed to their traditional curriculum.”  In addition, the research team was able to test their observations and methods for exploring fidelity of implementation and other aspects of the program to prepare personnel for 2012-2013.

There are many influences that exist on teacher professional practice based on discussions with the program staff and teachers.  In addition to the innovative program- SLIDER,- teachers’ education and personal experience shape how they act in the classroom.  The community, in the form of administrators, parents, and students, as well as the state and district level standards, impact how teachers teach.

The teachers participating in SLIDER report strongly positive views about their experience in the project.  Not only do they acknowledge the expertise and diligent attention of the CEISMC staff, they also appreciate the additional mentorship and availability of opportunities and resources outside of SLIDER in dealing with issues in school and the professional development opportunities.  All of the teachers interviewed reported benefits from the use of problem-based learning.  While some see it as the best way to teach students, others believe it is part of the portfolio of educational methods.

Five of the six teachers who were working with the SLIDER curriculum this past school year have been with the project since its first summer institute and the consistency has helped to improve performance in implementation, build community between the teachers and with the program staff, and provide additional longitudinal data for the research.  Table 2 describes the experience of the 3 schools adopting the current curriculum based on teacher interview.  School A is the most recent addition and replaces one of the initial schools adopting SLIDER.

Table 2.  Teacher Experience in the 3 SLIDER Project Schools

School	A	B	C
Innovative Program	Adopted by 1 Teacher	Adopted by 3 Teachers	Adopted by 2 Teachers (All 8th Grade Science Classes)
Teacher Individual Experience and Preferences	Teacher has been incorporating PBL for 8 years.	Limited previous experience with PBL, now very much in favor of it.	Limited previous experience with PBL, view as part of the portfolio of teaching methods.
School Administration Priorities	Program only in one classroom, administration minimally aware of the program.	Program has created tension with the administration due to competing district-level priorities.	Administration is very favorable to the program and has publicized the activities.
Student, Parents, and Community	Stays “out of the spotlight.”	Minimal publicity for the program, not strongly aware.	Opportunity to inject additional costly resources into the school system.
Job Requirements and Standards	Consistent with the teacher’s previous practice, working on improving test preparation.	Serving to prepare teachers for upcoming changes to the standards, lean heavily on the partnership.	Difficult to run the entire curriculum and meet all testing requirements, lean heavily on the partnership.

For many of the teachers, the chief benefit of the program is being ahead of the curve on upcoming standards focusing on science skills and science writing.  While they believe that SLIDER matches the requirements it sets out to meet, what is most important is time for addressing everything on standardized tests.  Test scores are the most important metric for the administration’s continued acceptance of SLIDER in at least one of the schools.  Overall, the teachers were very candid in their responses, offering high praise for Georgia Tech.  Teachers made the following observations:

·       SLIDER teachers report that most administrators, non-SLIDER teachers, and even parents are unaware of the details of the SLIDER science curriculum. 

·       SLIDER teachers report mixed satisfaction levels with their school working environment, school administration and school districts ranging from negative to somewhat positive. 

·       SLIDER teachers are regularly in touch with other teachers in their eighth grade science cohorts about the curriculum.  However, there seems to minimal spillover of the SLIDER curriculum into other classrooms.

·       CEISMC serves as the communication hub on the SLIDER curriculum.  There is little communication amongst SLIDER teachers outside of the trainings and summer institute.

·       There is little publicity about the SLIDER curriculum beyond the occasional teacher sharing his or her experience or outside observation. 

·       The students generally enjoyed the curriculum, with the teachers believing that it benefits a variety of students with different learning types.

·       The SLIDER curriculum is useful for the English as a Second Language population. 

·       Students who have previously been successful in traditional science classes may find it challenging to adjust to the problem-based style, but those frustrations in the teachers’ view, have long-term benefits.

Many of the early struggles prior to 2012-2013 seem to have been settled amongst teachers.  Teachers report that they are more accepting of problem-based methods of instruction.  This was true whether the teacher employed problem-based learning as either as their primary method of instruction or as an important component of their instructional methods.  Teachers also reported a growing comfort level with having their classrooms observed.  Finally teachers reported that many of the early tensions that characterized the SLIDER project seemed to be settled during the 2012-2013 academic year. 

An important event for the 2012-2013 school year was the initiation of the AMP-IT-UP project at CEISMC (also funded by NSF).  In many ways this new project builds upon the successes of SLIDER.  There is an important opportunity to apply the findings of SLIDER in a new setting.  Some project personnel who are participating in both projects expressed enthusiasm for building on the lessons learned from SLIDER but also caution at becoming spread too thin between the two projects.

For both AMP-IT-UP and SLIDER, the research and dissemination function are a major part of CEISMC’s work.  While CEISMC operates as a relatively independent research unit of Georgia Tech with limited publicity, the organization takes on a significant publication and presentation component, as evidenced from SLIDER’s record in Appendix A.  Even the staff on the curriculum team seeks opportunities to share and influence other teachers and researchers as part of their entrepreneurial efforts.

IV. Discussion

The work of CEISMC is serial entrepreneurship in the implementation of STEM education programs.  There is a flow to program activities. NSF, on behalf of the federal government, supports innovation in policy in a similar manner to an entrepreneurship incubator at a university or in the local government in private sector entrepreneurship.  It is also worth noting that CEISMC is not a street-level bureaucratic agency, but rather a grant-based support ecosystem for the teachers in support of the students.  These mezzo-level policy actors exist throughout implementation subsystem affecting change by influencing implementing agents and diffusing best practices.

As innovators implementing a novel and nontraditional curriculum, however, these policy entrepreneurs face significant constraints in the design and execution of their program.  First and foremost, the pilot program cannot exist if it does not operate within the district, state, and federal minimums of education requirements.  Whether it’s the old state standards or the new Common Core Curriculum, as well as other local level requirements, CEISMC cannot get participation from teachers if the teachers will face criticism for trying something different that hampers critical test scores.  A decline in meeting objectives related to test scores would result in the elimination of the program from a particular site.  Additionally, the focus of the program is on students and teachers, with limited attention to parents, administration, and the greater community.  This under-the-radar operation provides flexibility and increased immunity from community pressures, but it also constrains the opportunities to celebrate program successes.  Even at Georgia Tech, CEISMC is a peripheral research unit that serves the university’s desire to influence K-12 STEM education but is not subject to the restrictions of an academic or other centrally managed unit.  

            As noted, the program has impacted students across the state, including 700 in the past school year alone, with measured successful outcomes through activities related to the first two goals listed above as outlined in the application.  On the external goals (#3 and #4), there is an active and engaged research program that is sharing these successes for potential adoption and adaption in other STEM educational contexts.  The content of this paper is limited as it only deals with one case that is still ongoing, but there is a clear opportunity to enhance this line of research and expand upon the role of policy entrepreneurs in the implementation process.  In a New Yorker article, Dr. Atul Gawande (2013) writes:

           

            In the era of the iPhone, Facebook, and Twitter, we’ve become enamored of ideas that spread as effortlessly as ether. We want frictionless, ‘turnkey’ solutions to the major difficulties of the world  […]. People and institutions can feel messy and anachronistic. They introduce, as the engineers put it,uncontrolled variability.  But technology and incentive programs are not enough. ‘Diffusion is essentially a social process through which people talking to people spread an innovation,’ wrote Everett Rogers […]. Every change requires effort, and the decision to make that effort is a social process.          

Policy implementation entrepreneurs deal with the “messy” and put forth the effort to make changes at this stage of the cycle a reality.

V. Conclusion

           

            Implementation entrepreneurs- actors who develop, execute, and disseminate innovations beyond increased efficiency but within the political context to improve outcomes- deserve increased attention in the policy literature.  CEISMC is an intriguing case of education reform, but it is not unique.  A better understanding of these actors could facilitate increased opportunities to enhance and incubate similar innovators and innovations.  With the recent failures of the healthcare reform implementation dominating the discourse, there is an opportunity to look beyond those failures and see how implementers can not only avoid bungling a program but rather act entrepreneurially to improve outcomes within- and, perhaps, beyond- the vision of the political policymakers.  Further research could better identify implementation entrepreneurs and explain their influences on the broader policy process. 

VI. Sources

Berry, F. S., & Berry, W. D. (2007). Innovation and Diffusion Models in Policy Research. In P. A. Sabatier (Ed.), Theories of the Policy Process (2nd Edition ed.). Boulder, CO: Westview Press.

Borins, S. (2002). Leadership and innovation in the public sector. Leadership & Organization Development Journal, 23(8), 467-476.

Damanpour, F., & Schneider, M. (2009). Characteristics of innovation and innovation adoption in public organizations: Assessing the role of managers. Journal of Public Administration Research and Theory, 19(3), 495-522.

deLeon, L. (1996). Ethics and entrepreneurship. Policy Studies Journal, 24(3), 495-510.

Denhardt, J. V., & Denhardt, R. B. (2003). The New Public Service. Armonk, NY: M.E. Sharpe.

Dodgson, M., & Staggs, J. (2012). Government policy, university strategy and the academic entrepreneur: the case of Queensland and Smart State Institutes. Cambridge journal of economics, 36(3), 567-585.

Durose, C. (2009). Front-line Workers and Local Knowledge: Neighbourhood Stories in Contemporary UK Local Governance. Public Administration, 87(1), 35-49.

Eisenkraft, A. (2003). Expanding the 5E model. Science Teacher, 70(6), 56-59.

Foster, K., Bergin, K., McKenna, A., Millard, D., Perez, L., Prival, J., et al. (2010). Partnerships for STEM education. Science, 329(5994), 906-907.

Fullan, M. (2006). The future of educational change: System thinkers in action. Journal of Educational Change, 7(3), 113-122.

Garelick, B. (2012, November 20). A New Kind of Problem: The Common Core Math Standards. The Atlantic.

Gawande, A. (2013, July 29). Slow Ideas. The New Yorker.

Hicklin, A., & Godwin, E. (2009). Agents of change: The role of public managers in public policy. Policy Studies Journal, 37(1), 13-20.

Hmelo-Silver, C. E. (2004). Problem-based learning: What and how do students learn? Educational Psychology Review, 16(3), 235-266.

Ireland, R. D., & Webb, J. W. (2007). A cross-disciplinary exploration of entrepreneurship research. Journal of Management, 33(6), 891-927.

Killion, J. (2011). The Perfect Partnership: What It Takes to Build and Sustain Relationships that Benefit Students. Journal of Staff Development, 32(1), 10-12.

Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational psychologist, 41(2), 75-86.

Kingdon, J. (1984). Agendas, alternatives and public policies. Little, Brown, Boston.

Koski, C. (2010). Greening America's Skylines: The Diffusion of Low-Salience Policies. Policy studies journal, 38(1), 93-117.

Lindgren, H. (2013, June 7). If you were trying to name the greatest invention in human history it would be (The Innovation Issue). New York Times Magazine.

Lipsky, M. (1980). Street-level bureaucracy: Dilemmas of the individual in public services. New York: Russell Sage Foundation Publications.

Mack, W., Green, D., & Vedlitz, A. (2008). Innovation and implementation in the public sector: An examination of public entrepreneurship. Review of policy research, 25(3), 233-252.

Maranto, R., & Wolf, P. J. (2013). Cops, Teachers, and the Art of the Impossible: Explaining the Lack of Diffusion of Innovations That Make Impossible Jobs Possible Public Administration Review, 73(1).

Marschall, M., & Shah, P. (2005). Keeping policy churn off the agenda: Urban education and civic capacity. Policy Studies Journal, 33(2), 161-180.

Marsh, I., & Edwards, L. (2009). Dilemmas of policy innovation in the public sector: a case study of the national innovation summit. Australian Journal of Public Administration, 68(4), 399-413.

Meier, K. J. (2009). Policy theory, policy theory everywhere: Ravings of a deranged policy scholar. Policy Studies Journal, 37(1), 5-11.

Miles, M. B., & Huberman, A. M. (1984). Qualitative data analysis: A sourcebook of new methods. SAGE PUBL. INC., P. O. BOX 5024, BEVERLY HILLS, CA 90210(USA), 1984, 264.

Mintrom, M. (1997). Policy entrepreneurs and the diffusion of innovation. American journal of political science, 738-770.

Mintrom, M., & Norman, P. (2009). Policy entrepreneurship and policy change. Policy Studies Journal, 37(4), 649-667.

Nagel (ed.), S. S. (2002). Handbook of Public Policy Evaluation. Thousand Oaks, CA: Sage Publications, Inc.

O'Toole Jr., L. J. (2000). Research on Policy Implementation: Assessment and Prospects. Journal of Public Administration Research and Theory, 10(2), 263-288.

Petchey, R., Williams, J., & Carter, Y. H. (2008). From Street-Level Bureaucrats to Street-Level Policy Entrepreneurs? Central Policy and Local Action in Lottery-Funded Community Cancer Care. Social Policy & Administration, 42(1), 59-76.

Peters, B.G.  (2013)  American Public Policy, 9^th ed.  Los Angeles: CQ Press. 

Ravitch, D. (2010, June 14). Why I changed my mind. The Nation.

Renzulli, L. A., & Roscigno, V. J. (2005). Charter school policy, implementation, and diffusion across the United States. Sociology of Education, 78(4), 344-366.

Roberts, N. C., & King, P. J. (1991). Policy entrepreneurs: Their activity structure and function in the policy process. Journal of Public Administration Research and Theory, 1(2), 147-175.

Robichau, R. W., & Lynn Jr, L. E. (2009). The implementation of public policy: Still the missing link. Policy Studies Journal, 37(1), 21-36.

Rosenberg, M. S., Brown, M., McCray, E. D., deBettencourt, L. U., Leko, M., & Long, S. (2009). Development and Sustainability of School-University Partnerships in Special Education Teacher Preparation: A Critical Review of the Literature (NCIPP Document No. RS-3). Gainesville, FL: National Center to Inform Policy and Practice in Special Education Professional Development, University of Florida.

Salamon, L. M., & Lund, M. S. (1989). The Tools Approach: Basic Analytics. In L. M. Salamon (Ed.), Beyond Privatization: The Tools of Government Action. Washington, DC: The Urban Institute.

Scherer, J. (2006). Partnership Implementation in the MSP Program. Washington, DC: National Science Foundation Math-Science Partnership Program Evaluation (MSP-PE)

Schmidt, H. G., Loyens, S. M. M., Van Gog, T., & Paas, F. (2007). Problem-based learning is compatible with human cognitive architecture: Commentary on Kirschner, Sweller, and Clark (2006). Educational psychologist, 42(2), 91-97.

Simon, S. (2013, November 8). Name-calling turns nasty in education world. Politico.

Spillane, J. P. (2006). Distributed Leadership. San Francisco: Jossey-Bass.

Spillane, J. P., Halverson, R., & Diamond, J. B. (2001). Investigating school leadership practice: A distributed perspective. Educational researcher, 30(3), 23-28.

Spillane, J. P., & Kim, C. M. (2012). An exploratory analysis of formal school leaders and positioning in instructional advice and information networks in elementary schools. American Journal of Education, 119(1), 73-102.

Spillane, J. P., Parise, L. M., & Sherer, J. Z. (2011). Organizational Routines as Coupling Mechanisms Policy, School Administration, and the Technical Core. American Educational Research Journal, 48(3), 586-619.

Spillane, James P., Brian J. Reiser, and Todd Reimer. (2002). "Policy implementation and cognition: Reframing and refocusing implementation research." Review of educational research 72.3. 387-431.

Thomasian, J. (2011). Building A Science, Technology, Engineering, and Mathematics Education Agenda. Washington, DC: National Governors Association Center for Best Practices.

Volkmann, M. J., & Abell, S. K. (2003). Rethinking Laboratories. Science Teacher, 70(6), 38-41.

Walker, J. L. (1969). The diffusion of innovations among the American states. The American Political Science Review, 63(3), 880-899.

Wilson, W. (1887). The study of administration. Political science quarterly, 2(2), 197-222.

VII.  Appendix A- Key SLIDER Documents

Category	Documents	Authors	Year
NSF Application	Application Packet	N/A	2009
	Response to NSF Questions	N/A	2009
Annual Reports	Year 3	N/A	2012
	Year 2	N/A	2011
	Year 1	N/A	2009
Curriculum Materials	Curriculum Binders	Curriculum Team	2012
	Summer Institute Training Documents	Curriculum Team	2012
SLIDER Website	ceismc.gatech.edu/slider	CEISMC	2013
SLIDER Presentations and Publications	Integrating K-12 Engineering and Science; Balancing Inquiry, Design, Standards and Classroom Realities	Usselman et al.	2013
	Innovating Science Curricula with Engineering: A Balancing Act	Usselman et al.	2013
	Supporting Teachers Adopting an Engineering-Based, PBL Middle School Science Curriculum	Grossman et al.	2013
	The impact of a problem-based learning launcher unit on eighth grade students' motivation and interest in science	Fasse & Hendricks	2012
	Implementing a problem-based learning curriculum in a university-school collaborative project for improving middle school science education: Lesson from year one on fidelity of implementation	Gane & Hendricks	2012
	Facilitating and Assessing Science Learning Within an Engineering Design-Focused Project-Based Curriculum	Grossman et al.	2012
	Some Lessons Learned: Researching and Evaluating STEM Teaching & Learning in K-12 Higher Education	Llewellyn & Usselman	2012
	STEM Education through Race to the Top and SLIDER (NSF) as Examples	Millman	2012
	Mathematical Academic Outreach to K-12: ALGEBRA3, SLIDER, and GIFT	Millman	2010
	The Research Rationale and Process for the SLIDER  Project at Georgia Tech	Llewellyn	2010