Evidence Use and Advocacy Coalitions: Intermediary Organizations and Philanthropies in Denver, Colorado

The increasing involvement of philanthropists in education policy has contributed to the emergence of a dynamic sector of intermediary organizations (IOs), entities that serve a number of functions in school reform, including advocacy, consultation, policy design, alternative teacher and leadership preparation, and research. In recent years, many IOs have converged into coalitions that are pushing for incentivist educational policies like “parent trigger” laws, charter schools, vouchers, and teacher merit pay or sanctions often tied to value added metrics of teacher performance. This article draws on data from a mixed-methods, multiyear study of research use and dissemination. In this article, we examine the role of foundations in a broader advocacy coalition in Denver, Colorado, a key site for various incentivist reforms, including teacher pay-for-performance and charter schools. We find that IOs and their affiliated networks broker the production and use of research evidence, often targeting government and education policymakers, journalists, and increasingly, influential bloggers and social media communities. This brokering function positions foundations as the “hub” of research production, promotion, and utilization.


Evidence Use and Advocacy Coalitions: Intermediary Organizations and
Philanthropies in Denver, Colorado The increasing involvement of philanthropists in education policy has contributed to the emergence of a dynamic sector of intermediary organizations (IOs), entities that serve a number of functions in school reform, including advocacy, consultation, policy design, alternative teacher and leadership preparation, and research.In recent years, many IOs have converged into coalitions that are pushing for incentivist educational policies like "parent trigger" laws, charter schools, vouchers, and teacher merit pay or sanctions often tied to value added metrics of teacher performance.Major philanthropies like the Broad, Gates, and Walton Family Foundations are examples of philanthropic organizations that are leading the funding of incentivist reforms."Incentivist" reforms are policies that incorporate the use of enticements to drive school change, alter instructional practices, or help to create overall school improvement.
These incentives can be offered to organizations or individuals and include charter schools, voucher plans, merit pay for teachers and school leaders, and pay-for-performance plans for students.These reforms are often highly contested by parents, public education advocates, and teachers unions.In addition, the research evidence on the efficacy of these reforms is similarly unsettled.IOs are engaged in efforts to produce, promote, and utilize research in order to persuade policy makers and the public that their preferred reform direction is the proper route to better educational outcomes.
The focus of these philanthropic-funded efforts have largely focused on remaking the governance and leadership of school districts through shifts in mayoral control, school board elections, or district leadership in order to secure the adoption or expansion of incentivist reforms.For example, in 2015, a memo emerged detailing a Broad Foundation-led initiative to invest $490 million dollars to convert more than half of the Los Angeles Unified School District's schools into charter schools (Watanabe & Blume, 2015).
Some researchers have termed these policy approaches to school governance and control "portfolio management models" (Hill, Campbell & Gross, 2012), while others have described these alterations of control to be a part of the global "new public management" (Tolofari, 2005).Other researchers regard the rise of incentivist reforms as a hallmark of neoliberal ideology, with its nonpartisan focus on the development of market approaches to the provision of public services and public governance, which has long been promoted through advocacy groups like the American Legislative Exchange Council (Anderson & Donchik, 2014).While the conceptual frames and terminology remain unsettled in the research literature for the urban district reforms underway, it is clear that national and local foundations, in addition to national, state, and local policy makers, are central in moving forward a reform agenda in which market approaches to schooling are embraced.In local contexts, national philanthropic investments are complimented by the investments of smaller, state and community-based philanthropies.
These investments are frequently concentrated on a variety of organizations-advocacy groups, charter management organizations, and testing organizations, for example.The result is a growing set of intermediary organizations that are poised to influence federal, state, and school district policies around the need for an expansion and sustaining of incentivist educational policies.Their aggregate influence also extends to the production, promotion, and utilization of research evidence on incentivst reforms.
Drawing on data from a multiyear study of research use and dissemination,1 this article examines the role of foundations in a broader advocacy coalition in Denver, Colorado, a key site for various incentivist reforms, including teacher pay-for-performance and charter schools.We have found in our research that IOs broker the production and use of research evidence, often targeting government and education policymakers, journalists, and increasingly, influential bloggers and social media communities (Lubienski, Scott, & DeBray, 2011).This sector has gained visibility and momentum at the national level and in local policy arenas, frequently serving as entrepreneurs and champions for the educational policy agendas promoted by philanthropists.However, little is known about how intermediaries and philanthropists work together to shape public policy and movements toward growing incentivist policies in public education.
A focus on foundations and research evidence becomes especially important as advocacy networks of IOs turn their attention from urban districts that they have regarded as experimental laboratories for seeding incentivist reforms to other districts and states.We are seeing evidence that IO advocacy networks are interested in expanding these reforms to other districts within and across states through the enactment of state and national legislation and regulations.Much of the advocacy activity within these networks includes sharing evidence of the reforms' effectiveness in school districts with policymakers, the public, and other IOs in order to encourage or curtail their adoption (e.g., Hassel, Brinson, Boast & Kingsland, 2012;McEachin, Welsh & Brewer, 2013).To date, however, there is not sufficient study of the political contexts in which research evidence is utilized through networks of IOs.
The idea of introducing incentives into public education has advocates and detractors from the IO sector, many of whom are active in funding incentivist reforms, and producing or promoting evidence to policymakers on their promise or pitfalls.In our typology of the IOs that are active in this incentivist educational policy terrain, we include foundations, charter school management organizations (CMOs), teachers unions, professional associations, advocacy organizations, news media, think tanks, and civil rights organizations (Scott, Lubienski, DeBray & Jabbar, 2014).We have studied these IOs and their efforts to produce and share research evidence on these reforms in large urban school districts and national policy coalitions.We have learned a great deal from this work, advancing existing theoretical and empirical understandings of the role that IOs and their advocacy coalitions play in defining, promoting, and disseminating research evidence for policymaking in several analyses.We have found that foundations play a particularly unique role in these large urban districts, serving as IOs, and also as financial hubs for other IOs, connecting them with funding, meetings and conferences, and research evidence dissemination strategies (Scott & Jabbar, 2014).
We begin with an introduction to our conceptual framing of the politics of research utilization within advocacy coalitions, and the unique role that foundations play in this context.Next, we describe our methodological approach.We then review our major findings from Denver, Colorado.Finally, we conclude with a brief discussion of the implications of our findings.

Conceptualizing Foundations and Intermediaries in the Politics of Research Utilization
To analyze the emerging role of new networks of intermediary organizations, and how foundations support them to channel the production and consumption of research, we draw from insights from several conceptual orientations.We started with the Advocacy Coalition Framework (ACF) (Sabatier & Jenkins-Smith, 1999).According to this theory, an advocacy coalition is made up of "actors from a wide variety of institutions that share policy core beliefs and coordinate their behavior in a variety of ways."(p.130).Three premises of the ACF are particularly salient for our study: 1) theories of the policy process should include an understanding of the role of technical information to policymakers, that is, the ways in which think tanks and analysts become important to the policy process; 2) the best unit of analysis is a policy subsystem, consisting of "those actors from a variety of public and private organizations who are actively concerned with a policy problem or issue, such as air pollution control, and who regularly seek to influence public policy in that domain" (p.119); and 3) in addition to administrative agencies, legislative committees, and interests groups, ACF considers the role of journalists, researchers, and policy analysts and all government actors in policy subsystems.While traditional research organizations such as universities are preoccupied with established marks of research quality such as peer-review, ACF suggests that coalitions will focus on core tenets when policy specifics are contestable.
Therefore, that the ACF led us to theorize that intermediary organizations will promote evidence that is in alignment with a coalition's agenda, while slighting or rejecting high quality research that could challenge their agenda or cause fissures in a coalition.If intermediary organizations were indeed acting within a coalition formed around an incentivist agenda, then we would expect to see a broader range of evidence promoted as "research," including reports that lack established hallmarks of research quality.We employ the ACF to understand how philanthropists and intermediary organizations work in shifting, ephemeral networks around specific issues such as school choice coalitions that differ on the question of charter schools versus vouchers.As a coalition, network members find common causes with other organizations that can move ideas forward through research production, policy circles, and the media.

Foundations as the Hub in IO Advocacy Networks
Insights from the ACF help us to identify the sundry IOs involved in incentivist reforms and to understand the advocacy dynamics in different policy contexts.Yet we are also clear on the need to build theory on the uniquely situated role of foundations in the IO networks and coalitions.Our research has found that foundations are often central in the ways intermediary organizational networks function, and that the role of foundations in the overall landscape of IOs is multifaceted.Foundations have provided the essential funding for IOs and their advocacy efforts, without which such organizations would not exist at their current scale and level of policy influence.Yet we find that foundations are also themselves playing an intermediary role because they too participate in research production and promotion, and they attempt to persuade policymakers to adopt their preferred incentivist reforms, such as teacher pay-for-performance and the expansion of high performing charter schools.
In other words, foundations play a multifaceted role in the local IO networks as central IOs and as funders of IOs.They thus occupy a unique space in the overall political landscape of educational advocacy and knowledge production and utilization for educational policy making.Indeed, while their advocacy role has been understudied, many venture philanthropies have been more explicit about their advocacy role (Reckhow, 2013).An example of this advocacy direction is the case of the Broad and Gates Foundation's $60 million "Ed in 2008" campaign to ensure that their preferred educational issues would take priority in the 2008 Presidential election (Herszenhorn, 2007).While a number of scholars are engaged in research on foundations and their role in educational policy (See, for example, (Ferris, Hentschke, & Harmssen, 2008;Gittleson, 2010& Hess, 2005), as yet, research has not fully attended to the role foundations play directly as IOs, often focusing instead on their policy influence via the IOs that receive their funding.We frame this relationship between foundations and intermediary organizations by drawing on the analogy of a hub and spoke structure, in which integral parts of a wheel work together to move an incentivist policy agenda forward.A "hub-and-spoke" relationship, while education IOs are not typically profit-generating vehicles, philanthropists see their investments in IOs as a way to realize more promising and effective educational interventions whose "profit" is understood to be a scaling up of reforms they favor; they often seek a return on their investment.Together, their investments have generated a reform movement propelled forward by the entrepreneurial organizations they fund.For example, philanthropist Eli Broad said in the Wall Street Journal that the aim of his philanthropic efforts is to seed promising educational "ventures."Broad asserted that he and other educational philanthropies were "not in the check-writing charity business.We're in the venture philanthropy business" (quoted in Riley, 2009).Certainly, other researchers have raised empirical and philosophical concerns about the significance and scope of foundation influence over educational policies, particularly around market-based reforms, for democratic processes and equity (Demarrais, 2006;Kumashiro, 2008;Saltman, 2012).
In our previous work, we examined the 2010 investments of the six foundations that contributed the most to public education and for which we could get complete financial reporting via Internal Revenue Service 990 forms.From these statements, we recorded all grants made to U.S. education, paying particular attention to grants related to data systems, assessments and testing, and research or information.Three key strategies toward information and data management emerged from the analysis: (1) the funding of data systems and assessment tools; (2) funding for education journalism and media; and (3) grants for conferences and workshops that allow district and state leaders, education reformers, IOs, and researchers to share ideas and evidence on data use (Scott & Jabbar, 2013).While not all of these investments went to IOs or exclusively supported incentivist reforms, in Denver, significant grants from the Gates, Broad, Dell, Kellogg, and Hewlett Foundations were awarded to IOs to help implement and design teacher merit pay and value-added assessments, expand or strengthen charter schools, or to IOs like the Aspen Institute or Center for American Progress to host the convening of policymakers and IO representatives to share evidence and strategic approaches to reforms utilizing incentives.
These and other foundations have helped to provide the financial backing for many charter management organizations (CMOs), to place system and district leaders in positions of authority, and to scale organizations to expand across multiple school districts (Wohlstetter, Smith, Farrell, Hentschke, & Hirman, 2011).Leaders of foundations actively promote the successes of their investments to policymakers in an effort to convince them that such reforms and reformers are also worth public support and alterations in public policy.Figure 1 provides a visual of the hub and spoke relationship we argue exists between foundations and IOs.

Figure 1. The hub and spoke relationship between foundations and intermediary networks
This conceptualization is helpful in capturing the ways in which foundations play multiple roles, acting as intermediaries themselves and as the funding support to other intermediaries.In addition to supporting local coalitions, they also help to create organizational coalitions at the national and state levels.To apply our understanding of the unique place of foundations in the overall landscape of IOs-with a particular focus on how foundations work in the context of research and evidence of the efficacy of incentivist reforms-we examined interview data and documents for themes and patterns that emerged from our diverse pool of respondents in Denver, CO.

Policy Context, Methods, and Data Sources
This research project combines interviews, event observations, and the analysis of documents produced to persuade policymakers and the public to embrace "evidence" of the legitimacy and effectiveness of specific policies.Documents included policy and research reports generated by IOs, strategic planning documents or mission statements, foundation spending reports, and annual reports by IOs and foundations.We report our findings on one city, Denver, from a larger multi-city study to examine the role of intermediary organizations in the policymaking process.We interviewed and collected documents from actors in policymaking positions, advocacy organizations, teachers unions, philanthropies, and community groups.

Methods
From 2011 to 2013, we conducted 40 interviews in Denver, CO.We interviewed key local and state policymakers, representatives from over 25 intermediary organizations, researchers, and journalists.Informants were selected initially to capture a broad range of viewpoints and perspectives.We then used a snowball sampling method (Bogdan & Biklen, 1992) to elicit further respondents.We used semi-structured interview protocols (Patton, 1990), asking respondents about their organization, their positions on the incentivist policies of interest, their decision-making process, and the sources and types of evidence they drew on in making decisions.Interviews lasted, on average, between 30 and 90 minutes, were recorded with permission, and were transcribed for further analysis.In addition, we observed key dissemination events, such as town hall meetings, panels organized by intermediary organizations, and other public forums.We took detailed field notes during these meetings.For analysis, we developed a "start list" of codes (Miles & Huberman, 1994) based on previous literature and theory, which we used to code all interviews.We developed new codes and categories iteratively and inductively.We also conducted ongoing analysis by writing memos on emerging themes related to research use and production and the role of intermediary organizations in policymaking.Coding and memoing allowed us to identify key themes related to research production, the politics of research use, and intermediary organizations.

Local Context
At the time of our research, the Denver Public Schools (DPS) was a diverse educational system, with 54% of the students Latino, 25% White, 16% Black, and 4% from other categories.Over 70% of students qualified for federal free and reduced price lunch.Forty of the 161 public schools in Denver were created as choice schools (i.e., charters and magnets), and the Janus and Piton Foundations have been active in promoting such policies.The school system launched the Denver Plan beginning in 2005, which accompanied a voter initiative approving salary increases for teachers that included merit pay, raising over $25 million from taxpayers.With support from the Janus Foundation's Education Alliance, the program has expanded.
The ProComp agreement with the Denver Classroom Teachers Association provides merit pay for teachers based on their willingness to teach in difficult to staff schools and their role in raising student achievement outcomes.Under Denver's ProComp performance pay program, teachers can build permanent salary increases, but can also lose them if student performance declines.Denver's program is known for its successful negotiation with the teachers' union.The raises are based on multiple measures, including commitment to teaching in hard-to-staff schools, knowledge and skill development, and student growth (Gonring et al., 2007;Wiley et al., 2010).Though teachers can lose raises previously earned, the system was designed to "advance the priorities of the school district, not an instrument to punish 'bad' teachers" (Gonring et al., 2007, p. 23).Philanthropists played a key role in the implementation of the plan.The first step, as Gonring, Teske, and Jupp (2007) write, was to draw teachers to the cause because having the teachers' union on board was key to passing the legislation.Local foundations also influenced the Board of Education's resolution to extend the length of Denver's pilot pay-for-performance program to four years (Wiley et al., 2010).
Since ProComp was implemented, incentivist reforms such as merit pay and charter schools continue to be a key focus.Superintendent Tom Boasberg's educational vision (on the DPS website), reads, in part, "In an age of competition for students and for talented teachers and school leaders, DPS must break from the monopoly model of the past century and push forward with decentralization reforms to give schools greater autonomy, coupled with accountability."In 2010, however, Colorado was unsuccessful in its application for the RttT program, despite Denver's alignment with many of the Obama Administration's priorities for educational reform.
In the Denver context, foundations invested in incentivist reforms through research advocacy and funding mandates.Denver is exemplary of the strong linkages between national and local advocacy coalition networks among foundations and IOs.Local foundations leveraged funding from the federal grant program, Building Charter School Quality, to support the Colorado League of Charter Schools.The Gates Foundation awarded $10 million to Denver School Teachers Association for teacher evaluations and also helped to fund the DPS system directly.While respondents viewed the Gates Foundation as the national "lead" in terms of philanthropy, they viewed the Donnell-Kay Foundation as the local lead.The Rose and Donnell-Kay Foundations took on leadership roles to advance the ProComp reform and also to promote charter schools.The Donnell-Kay Foundation also provided funding to DPS, and they also founded two reform organizations, Colorado Succeeds and Get Smart Schools.
The Rose Foundation acted as a figurehead champion for ProComp and invested specifically in gaining support of the DCTA.Informants from Denver Public Schools explained that the investment of time and resources by Rose Foundation and Donnell-Kay Foundation in ProComp was critical for teachers' participation and advancement of both ProComp reform and CMOs.In terms of evidence to support their cause, local foundation representatives expressed dissatisfaction with the research base in not being representative of "gold-standard" work, not very compelling, anecdotal, and thin.Once the needed IOs and foundations were in place, foundations helped to fund reports by local groups in order to continue to affect policy.
By 2014, a majority of the Denver school board was united in support of charters and ProComp.Because of the departure of opponents to these incentivist reforms due to term limits, adversarial politics on the board have been minimal, though advocacy groups and vocal parent and community advocates have raised questions about the board's endorsement of incentivist reforms (Gorski, 2015).Local philanthropies have catalyzed much of the recent policy movement, and intermediary groups have responded to funding possibilities by marshaling research evidence in support of the agendas the philanthropies are promoting (Scott et al., 2014).

Findings
We found that foundations play a key role as "lead investors" for such incentivist reforms.Many of our respondents indicated that there did not exist a robust research base that legitimated the rapid expansion of incentivist reforms, but that in light of a lack of "gold-standard" evidenceby this, respondents usually meant experimental design studies that showed causal, positive of achievement effects on standardized assessments-foundations leveraged personal contacts with researchers who had found positive effects and targeted their dissemination of those research findings to policymakers.We also found, like Reckhow (2013) and Saltman (2007), that foundations used funding to proliferate reforms.
Local foundations have catalyzed much of the recent policy movement in Denver, and intermediary groups have responded to funding possibilities and to meet the needs of marshaling research evidence in support of the agendas the philanthropies are promoting.Denver had been relatively free of the adversarial politics at the district level that characterize such reforms in other cities, and the local philanthropies saw themselves as building bridges across political constituencies, including reform organizations and teachers unions.Intermediary organizations are increasingly creating a more acrimonious policy climate as they embrace a politics of opposition to advance their agendas.Several smaller philanthropies have indicated ambivalence toward the incentivist reforms they fund, yet they lend support in their goal to be seen as "players" in the philanthropic and policy worlds, with direction from the policy agendas of larger, national foundations that fund reforms in their cities.Denver philanthropies have positioned themselves as visible champions of ProComp and charter management organizations.They have invested in these reforms in two ways: 1) advocating through selective dissemination and translation of research to a targeted audience of policymakers, often at meetings funded by local and national foundations; and 2) funding preferred agendas practices that lead to reform proliferation.Because the foundation support is quite concentrated and comes primarily from just a handful of organizations, the IO network in Denver is small.The Daniels, Piton, and Donnell-Kay Foundations provide most of the funding.Furthermore, the Denver Public Schools and Denver Classroom Teacher Association worked with the private sector actors to form a unique coalition in their convergence on ProComp1 and ProComp2, the teacher pay-for-performance programs implemented in Denver.

Lead "Investors" in Reform
Since 2005, Denver has had in place a merit-pay system for its teachers based on valueadded assessments known as ProComp.The Rose and Donnell-Kay Foundations have been supportive of ProComp, and this appears to have influenced The Denver Classroom Teacher's Association's (DCTA) favorable position on ProComp.As one union official explained, the Rose Foundation acted as a figurehead champion for ProComp and invested specifically in gaining support of the DCTA: ProComp didn't happen just because DCTA just as an organization woke up one day and said, "wow, why don't we just do things differently?"You have to have a champion.According to another union official, the investment of time and resources by Rose Foundation and Donnell-Kay Foundation in ProComp was critical for teachers' participation: So when we had that Joint Salary Task Force, I think what we did is we invited national experts to come and talk to us for one entire year.So we had resources to pay for a facilitator, resources for experts to come to Denver . . .most of it from the Rose [Foundation] but there were some other foundations, like Donnell-Kay Foundation.They were really into it.They allowed us to tap into different researchers.
The Rose and Donnell-Kay Foundations took on leadership roles to advance the ProComp reform and also to promote charter schools.While the Gates Foundation was viewed as the national "lead" in terms of philanthropy, Donnell-Kay was viewed as the local lead.Gates gives $10 million to Denver School Teachers Association for teacher evaluations and also helped to fund the DPS system directly.Donnell-Kay Foundation also provided funding to DPS, and they founded two reform organizations, Colorado Succeeds and Get Smart Schools.Local groups also helped to channel funding from the federal grant program, Building Charter School Quality, to support the Colorado League of Charter Schools.

Foundations' Research Use and Promotion
Despite a perceived lack of rigorous research on the reforms, and claims that the ProComp evaluations were contested on all sides because of how difficult it was to parse the effects of ProComp itself in isolation from other reforms, foundations helped to fund reports by local groups that promoted the reforms.Targeted dissemination of research to policymakers.Foundations are often a key link between particular intermediary organizations and national, state, local policymakers; in some cases, they explicitly seek to influence policy.In the Denver context, foundations invested in incentivist reforms through research advocacy and funding mandates.Often, foundations in Denver did not conduct their own research or fund research directly, but rather funded the dissemination of research findings.A representative from the Rose Foundation reported, for example, that their organization did not necessarily conduct its own research, but rather disseminated research evidence that it found compelling.He explained:

Lack of "
We don't do a whole lot of funding of research and policy reports.A lot of the times what we will do is step in and help with dissemination of it.If we feel like this report that just came out really needs to be shared with legislators or there needs to be luncheons where they're educated on the implications of the policy recommendations or of the research, then sometimes we step in and play that role.
A researcher at a local university very familiar with ProComp agreed with this characterization: "Rose Community and Broad nationally and a few others put in some money to make it happen.Like any researcher I would complain that they don't seem to put in enough money into evaluating in a really rigorous way what did happen, especially down the road.They want to move on to the next thing."A Rose Foundation representative also reported that research is presented in the form of policy reports in order to help "move" state representatives on policy issues deemed important.As an example, The Rose Foundation was working with The New Teacher Project to amend state licensure policy.According to the foundation: The state board is really difficult right now.It's kind of a difficult set of players and the board chair essentially have said that he wants to get rid of licensure altogether, he thinks it's completely useless and what he wants on the table is for it to be scrapped.So while they had hoped to make the changes through the state board, they're now recognizing what they're going to have to do is make the changes through the state legislature.So because of that, they came back to us and said, "What we now would like to do is provide a pretty robust policy report about the changes that we think need to be made." This representative of the Rose Foundation reflected on how policymakers' responses provoked the need to produce research on the desired changes in policy, often by working alongside advocacy groups and other reform organizations.
To disseminate this particular research, the foundation cultivated relationships with key policymakers and researchers whose scholarship found positive effects on the reforms it favored.The staff at the Rose Foundation and other local foundations had previously worked in state or federal government.According to a representative from the Donnell-Kay Foundation, ultimately, they wanted to influence policymakers: It's really about policymakers, from district to state level.So, at the state level, it's from governor on down and his cabinet, to legislators, to state board commissioner.Those are the people we're really trying to influence and educate.
And on the district side, it's superintendents, school boards, occasionally principals, and it's kind of just the-the grass tops, the policy elites, the people who actually read the paper-not only the front page, but probably the editorial section.I'm guessing in most communities, that's about a tenth of one percent of the people…Try and reach some of those people because they actually have the ability to make decisions and move things around.Not trying to appeal to broadly parents, citizens, teachers-just our target audience.
One key representative from the Colorado Education Association reported that foundations advocated for charters despite awareness of inequities: "They're aware of the inequity of the funding between these types of schools but yet they continue to advocate certain reform models knowing the inequity."This observation reflects foundations' opportunistic, alignment -oriented approach to advocacy.A Donnell Kay Foundation representative explained that local foundations cultivate relationships, drawing upon useful research, by asking themselves, "what are the right policies to have in place at a state level or a district level, what are the policies that are impeding getting to that point, [and] how do we create the right environment to get this work done?"Furthermore, their "opportunistic" activity involved taking advantage of key policy windows to push through reforms: I'd love to tell you it's like a super thoughtful process, that's part of a strategic plan-the reality is, we're very opportunistic…And those windows, they have to be open and they have to be open in alignment are political, innovation, tangibility, and an understanding of how the policy environment is in that.
Foundations were therefore explicit about their strategic role in shaping state and local policy, and were serving as an important hub that connected the work of IOs to policy makers.Relationships with state government were instrumental to enacting such 'right' policies and trumping 'impeding' policies.In this vein, the state created the Colorado Legacy Foundation.A Get Smart Schools representative posited that this foundation was specifically envisioned as a vehicle by which the 'right' policies could be advanced more efficiently: The Colorado Department of Education set up the Colorado Legacy Foundation basically as a way to get a 501(c)(3) that could get private money and could actually do stuff fast, not going through all the state stuff….They have a variety of different projects that just don't necessarily seem to fit together, but there are things that are a priority for the Department of Education that they can get outside money for…And so they got a huge grant from the Bill and Melinda Gates Foundation to basically figure out how to implement 191.
Foundations thus played a key role in advancing particular ideas, drawing upon evidence selectively to convince policymakers, and creating new intermediaries that could bypass traditional policymaking routes.

Funding Preferred Agendas to Proliferate Reforms
Apart from engaging in research dissemination and translation to key policymakers, foundations invested in their policy agendas through targeted funding.Funding influences the direction of grantee organizations' work, goals, and research initiatives.A Colorado League of Charters Schools representative expressed the explicit relationship with funding directives and organizations' missions: "However nice it would be that they would all just write a big check and say, 'use this however you choose,' they don't and it tends to be around specific projects or initiatives and efforts, so there's some fluctuation in terms of who is doing what and when."One specific initiative that was attractive to local foundations was charter school expansion.As such, foundations more readily funded CMOs and organizations that advocated on behalf of CMOs.Critics of the "CMOs only" approach expressed concern that local preferences for independent, autonomous charter schools were disadvantaged under this investment strategy, even as funders expressed overall support for the charter-school movement.Teacher-led innovation schools, for example, could not access support from national and local funders without making concessions.A representative from the Colorado Education Association argued, "Because they've pushed so many charters that a lot of our neighborhood schools just kind of feel ignored.All we do is focus on the new schools and we're not giving additional resources or support to the current schools."A teachers' union representative was keenly aware that organizations could garner funding by conceding to grantors' visions, but this came at the expensive of focus on localized innovation: If I am willing to become an innovation school, I have $500,000 right there from Walton.All I have to do is waive something.So I even met with a group that was submitting the proposal, and they were saying that they could really use the money.So I said: "Say you're going to waive the calendar year and then just change on the years, there you go.Bam, money.We have waivers." She goes on to add: "But just even the philanthropic community, we have not necessarily agreed with them.Because they've pushed so many charters that a lot of our neighborhood schools just kind of feel ignored.All we do is focus on the new schools and we're not giving additional resources or support to the current schools."Despite these concerns, local foundations seemed focused on this strategy and achieved some success in this regard, as many respondents credited local and national foundations for creating the CMO Get Smart Schools.The founding of Get Smart Schools is illustrative of local foundations' aim of innovation.A Get Smart Schools representative described how local foundations' monumental, collaborative investment was made in the vein of innovation: Get Smart was created by a number of the foundations in town, and it was actually a pretty interesting process.It was the largest foundations in Denver that contribute to public education.They had gone through an exercise of looking at the money that they had invested over, say, a ten year period in public education, trying to isolate what had impact and what hadn't.I think they concluded pretty quickly that the money that they gave directly to districts, there was really nothing that they could show for that money.The money that they had put into starting new schools, in some cases those schools hadn't succeeded.The times that the schools had succeeded, they were making such a dramatic impact that the return on investment was pretty significant, even if you take into account the schools that had failed.So they reached this conclusion that they wanted to have more of an organized new school strategy.
A Donnell-Kay representative echoed this belief in the idea that charter school expansion through more unified management achieves the promise of innovation: We're very supportive of charters by in large because they are a little bit better space to innovate.So we have funded individual charters and we quickly got out of that business because there was too many, and we started funding CMOs-Charter Management Organizations.Because we thought, and still believe that the potential there is greater to have a network of schools.We helped create innovation schools here in the state.
Increasingly, local foundations posited that the best space to innovate and grow was not simply in local charter schools, but rather in networks with the potential for scale.A represen tative from the Colorado League of Charter Schools, an outfit with ties to national charter advocacy groups, explained and critiqued how this point of view was influenced by the federal reform pro-CMO funding schema: I think as far as ESEA goes, we're concerned…that the real focus is not going to be on individual charter schools.You can see it the way the dollars being shifted now through the charter school program only on CMOs and these highperforming CMOs.Particularly those that have the lobbying muscle to get those dollars geared to them…I've heard that the KIPP lobbyist has really been pushing for more money to KIPP because if you have some peanut butter then why spread it around all the bread when you know it's working?Focus it on where it needs to be focused on.We don't like that at all.Every single CMO started with one school at one point and found their groove and found what works and kept expanding.
Foundation support of charter school expansion, by funding already high performing chartermanagement groups rather than individual or startup charters, or by selecting particular charters to invest in, via Get Smart Schools, for example, has the potential to significantly alter the charter school movement, shifting it from a focus on a broad range of diverse school forms to a more targeted expansion of larger charter networks.

Discussion and Conclusion
Our analysis suggests that foundations "outsource" the "work" of research production and advocacy to intermediary groups.In the Denver context, key foundations invested early in incentivist reforms through research advocacy and funding mandates.These foundations generally did not conduct their own research or fund research directly, but rather funded the dissemination of research findings.
Working in ideologically defined coalitions, while perhaps intending to preserve the credibility of philanthropies, those operating within interstices may instead lead to a more divisive rather than monolithic policy environment.As such, it is important to foreground the work of foundations in the creation of IO advocacy networks and to especially consider the role of foundations in producing, promoting, and disseminating research and other forms of evidence.Foundations are therefore not just funders-they are also investors and private policy makers.They view their financial support as an investment in realizing the adoption and implementation of incentivist reforms.In this sense, we argue that they are the "hub" that moves the "spokes" in a local IO coalition.Research production, use, and dissemination are key strategies that philanthropists use to fund, launch, and protect their investments in educational reform.In Denver, we see the importance of foundations for the production of evidence, the communication of evidence to policymakers, and the overall support to IOs to scale their organizations.Local and national foundations have also been critical in championing the school choice and merit-pay programs in Denver, but also in helping to produce and promote evidence of their effectiveness locally and to the national audience (although some critics claim they have not invested enough in high-quality research).Without this hub of funding and alignment around the importance of incentivist reforms, it is unlikely that such reforms would have moved forward at the size and scope that we witness in Denver and within the Colorado legislature.
Across our research sites, we have seen that foundations are central in the creation of advocacy networks supporting and resisting incentivist reforms.And we have found that research evidence-broadly defined-is moving though those networks.We have also found that supporters of incentivist reforms enjoy much more robust foundation support than do detractors, who largely receive funding from teachers unions, or other membership organizations.An example is Education Sector, a "boutique think tank" designed to influence public policy dialogue and policy formation.Education Sector primarily conducts secondary analysis of existing data sets, using these analyses to further particular narratives and arguments.It receives 100% of its funding from foundations, including the Gates, Lumina, Carnegie, Broad, and Joyce Foundations.
This divide raises important questions about the state of knowledge production in a democracy, and the question of the relationship between research funding and findings produced is being scrutinized in a variety of fields, including medicine, environmental science, and tax policies.Foundations have historically played an important, and often critical role in the funding, production, and dissemination of research findings in education, science, medicine, and other areas, and in education this support has led to important advancements in teaching, learning, and policy.Yet, we are also seeing evidence that in ideologically charged environments, advocacy-oriented foundations are informing and shaping public policies that require greater examination, especially since we have found that many IOs producing or disseminating research have relatively thin expertise in conducting rigorous qualitative and quantitative research.
Our respondents dependent upon funding have also indicated that their IOs follow the money.As a New Orleans think tank representative shared, "What we do and what we spend our time on is connected to the fundraising.If we raise $50,000 from someone with a particular interest on energy policy, then we're going to spend more time on energy policy."If it holds that the primary influencers of educational policy are largely producing knowledge that endorses the policy preferences of their funders, it suggests an even greater need for independent research to be generated so that the findings can be more thoroughly deliberated in the public square.In this regard, we see a continued need for government to invest in existing federal and state research agencies and research universities and academic researchers to provide additional research evidence.
Practice and Leadership in Education (www.ripplenetwork.ca)-a program of research, training and KMb aimed at learning more about how knowledge brokering (KB) can increa se research use and its impact in public service sectors by facilitating collaboration between multistakeholder networks.

Samantha Shewchuk
Queen's University s.shewchuk@queensu.caSamantha Shewchuk is an elementary school teacher and a doctoral student in the Faculty of Education at Queen's University.Her research explores knowledge mobilization at the intersection of the education and child welfare sectors; more specifically, how outcomes for abused children can be improved by increasing research use in these sectors.She is the program manager of Dr. Amanda Cooper's RIPPLE program, Research Informing Policy Practice and Leadership in Education (www.ripplenetwork.ca).

SPECIAL ISSUE Knowledge Mobilization
education policy analysis archives That champion wasn't me, the champion was Brad Jupp [of the Rose Foundation].… So I have a great admiration for him.He invested, I never saw someone invest so much time into something like this.He made us participate in this process; we found value in what he did.

gold standard" research.
Local foundation representatives expressed dissatisfaction with the research base in not being representative of "gold-standard" work.One representative from the Donnell-Kay Foundation explained: So most of what I see is either anecdotal or, get the words right…non-causal, like correlated.And so, we tend to look at it, but a lot of the things we want to do are not necessarily replications of what has been done, so I think that that research is extremely thin when you start taking these ideas and morphing them.When they found research to be "thin" or less than "gold-standard," local foundations leveraged existing relationships with researchers whom they knew to be supportive of value added programs and charter expansion.Interviewees cited Disrupting Class author Michael Horn, Clay Christenson of Harvard Business School, Rick Hess from American Enterprise Institute, Brian Hassel from Public Impact, and Michael Griffith of the Education Commission of the States as examples of researchers who "either will know or may have some really interesting ideas."The reliance on such researchers, many of who enjoy funding from national foundations supportive of these reform efforts aligns with two of the ACF's premises that research and technical information is important to policymakers and that researchers are important actors in policy subsystems.
I don't find a lot of the research very compelling in education-I think it's tough to do, tough to do really well, tough to do gold-standard…Double-blind.I mean, I just think that it's really expensive and it's really tough to construct.And it's rarely done.