Who’s setting the agenda? Philanthropic donor influence in marine conservation

. We are in a period of unprecedented growth in conservation philanthropy. How will this influx of private funding affect conservation agendas? Inspired by a collaborative research co-design process, this paper addresses questions about how foundations influence conservation agendas in the places they work. We draw from a case study of the world’s largest philanthropic funder of marine conservation, the David and Lucile Packard Foundation and their 20 years of investment in marine conservation in Palau and Fiji. Conservation practitioners in both countries universally agreed that the Packard Foundation had a significant and positive influence on the agenda, which they attribute to both how the foundation worked and what they chose to fund. Specifically, our study reveals how the Packard Foundation shaped conservation agendas in Palau and Fiji in partnership with its grantees through a grant-making process characterized by relationship building, collaborative decision making, convening and promoting of collective action, flexibility, and long-term funding. Packard’s approach was often identified as unique, and contrasted with numerous other donors, including foundations and other types of donors, who use a more top-down approach. By describing a relative success story in how philanthropic foundations can work with conservation practitioners to co-design a shared conservation agenda, our work provides timely guidance for donors and practitioners.


INTRODUCTION
Philanthropic foundations are playing a growing role in funding environmental efforts (Bloomberg Philanthropies 2021, , Li et al. 2021. In light of the influx of new donors and their growing share of funding commitments for conservation, climate, and other environmental issues, it is urgent that we understand how philanthropic foundations may be shaping conservation agendas. Such commitments include, for example, the record $5 billion Protecting Our Planet Challenge that is funding efforts to protect 30% of the planet by 2030, and the $10 billion Bezos Earth Fund for climate change and nature protection. Yet empirical research on the influence of private philanthropy on conservation agendas is problematically thin. We advance scholarly and practical conversations on private donor influence in conservation agenda setting by using the results of a case study examining the David and Lucile Packard Foundation's (hereafter Packard) support over two decades for marine conservation in Palau and Fiji. Our research focuses on understanding both the extent to which Packard influenced the conservation agendas in Fiji and Palau and the processes through which that influence played out. We answer three specific questions: (1) Did Packard have a significant influence on the conservation agenda in each country? (2) How did Packard's grant-making approach affect the type and level of their influence on the conservation agendas? and (3) How did Packard-supported conservation (including individuals, organizations, projects, and ideas) influence priorities and agendas in Fiji and Palau, in some cases even after funding has ended? We begin by providing background on Packard's investments in Palau and Fiji during our study period and outlining our research methods. Next, we present our empirical results. We conclude by discussing implications for research and practice.
We use the phrase conservation agenda to broadly refer to the overall trajectory of marine conservation and its importance in relation to other issues. It also refers to the specific priorities that comprise that agenda, from the national level to the agendas of individual organizations in Fiji and Palau (Brulle 2000, Jenkins et al. 2017, Cunningham and Dreiling 2021. Conservation priorities refer to the objectives that are given importance and value, as well as what is being done in practice through research, action, and policy initiatives (Carwardine et al. 2008, Wilson et al. 2009, Pullin et al. 2013. Throughout this article, when we use the phrase conservation network we are referring to the individuals and organizations (including those in the government, non-governmental, and private sectors) based in-country in both Fiji and Palau and working on conservation in varying capacities.
We make two contributions with this work. First, we add to the small but growing body of empirical evidence that philanthropic foundations can have significant influence on conservation agendas (Brulle and Jenkins 2005, Barker 2008, Brulle 2014, Cunningham and Dreiling 2021. Our findings suggest that the growth in environmental philanthropy globally has the potential to translate into significant influence (positive and negative) on conservation agendas, and that diverse actors within the conservation field should be more engaged in observing and defining how this influence plays out in practice to ensure justice and equity, as well as accountability, in delivering on actions that https://www.ecologyandsociety.org/vol28/iss3/art2/ translate into meaningful outcomes for conservation and wellbeing. Second, we document a successful model of donor engagement in conservation agenda setting, as measured by largely positive perceptions among conservation practitioners about the level and type of donor influence and the processes through which that influence was exercised, the former largely being a result of the latter. Our case study reveals how Packard shaped conservation agendas in Palau and Fiji in partnership with its grantees through a grant-making process characterized by long-term relationships, collaborative decision making, convening and promoting of collective action with locally-based conservation networks, long-term investment, and flexibility. This approach and its resulting influence on the conservation agendas in Fiji and Palau were perceived as successful by conservation practitioners in the two countries, and also as increasingly rare in the broader field of ocean philanthropy. Because grantees are a primary conduit for channeling community viewpoints into foundation priority-setting and decision-making processes (Betsill et al. unpublished manuscript), meaningful collaboration with grantees in agenda setting may increase the potential for philanthropic-supported marine conservation that is more socially equitable and just by serving the priorities, interests, and needs of local stakeholders and communities (Bennett et al. 2020, 2021.

BACKGROUND: PHILANTHROPIC AGENDA SETTING AND PACKARD'S INVESTMENTS IN FIJI AND PALAU
Historically, philanthropic foundations have made significant contributions to environmental efforts (Barker 2008). In recent years, those contributions have increased substantially (California Environmental Associates 2019, Li et al. 2021), yet there has been little research on how philanthropic foundation support affects conservation priorities . To date, the area of research receiving the most attention concerns the ways foundations shape environmental movements (Brulle 2000, Faber and McCarthy 2005, Bartley 2007, Jenkins et al. 2017, Brulle et al. 2021) and embedded within that research are discussions about how foundations affect conservation agendas (Barker 2008, Tedesco 2015, Morena 2020. Some have argued that philanthropic foundations have the potential to significantly influence environmental agendas, and in so doing may preserve their privileged position in society (McCarthy 2004, Barker 2008, Mallin et al. 2019, Cunningham and Dreiling 2021. Others have argued that foundation influence on these agendas is complex and warrants further research Tang 2007, 2008). Although a handful of empirical studies have provided a macro-level analysis of the level and type of foundation influence on environmental agendas (Brulle and Jenkins 2005, Faber and McCarthy 2005, Brulle et al. 2021, Cunningham and Dreiling 2021, few studies examine the process through which this occurs. One notable exception is Delfin and Tang's work (2008) on how the donor's approach and role of grantees in the process can be important mitigating factors for foundation influence, a finding confirmed by our research.
The need for more detailed, empirical research on the processes by which foundations influence marine conservation agendas has been identified by practitioners, scholars, and donors who see a need to better understand, and in some cases reform, current practices in the context of a significant uptick in philanthropic support for marine conservation . From 2010 to 2019, foundation grants for ocean-related issues nearly doubled and were roughly equal to those from official development assistance in 2016 (California Environmental Associates 2019). Between 2010 and 2020, Packard was the largest philanthropic funder of ocean-related environmental work globally (California Environmental Associates 2019). Over a more than twenty-year period, between 1998 and 2020, Packard was a primary private funder of marine conservation in the two countries of focus for this paper (Palau and Fiji).
Packard's Western Pacific Program began work in Palau and Fiji in 1998, with their last grants ending by 2020. Between 1998 and 2017, Packard invested US$11,952,000 through 71 grants to 22 organizations for work in Fiji. In roughly the same period, Packard invested US$8,168,000 through 52 grants to 13 organizations for work in Palau. A majority of Packard's grants went to non-local organizations. In Palau 19 grants went to local organizations, whereas 34 went to non-local organizations. In Fiji, 27 grants went to local organizations and 44 to non-local organizations. Packard grouped their grants into four categories: site-based conservation and fisheries management; capacity building; public education and media; and policy reform, analysis, and applied science. In both countries the largest investments were made in site-based conservation and fisheries management (58% in Fiji and63% in Palau [Packard Foundation 2017, 2018]). Packard's investments in both site-based conservation (e.g., through marine-protected areas and locally-managed marine areas) and fisheries management reflect a recognition that "conservation" is indistinguishable from "sustainable use" in most Pacific Island worldviews, including Fiji and Palau (Govan and Jupiter 2013). In our paper, we thus use the phrase marine conservation to encompass objectives and tools associated with both biodiversity conservation and sustainable use, including area-based approaches and broader fisheries management.
In the early 2000s, Packard was the largest source of private foundation funding for marine conservation in Palau; there were few other major donors working in marine conservation there. There was a greater diversity of donors already working in Fiji when Packard entered into the scene. Packard pursued donor coordination there, particularly with the John. D. and Catherine T. MacArthur Foundation but also occasionally with the Marisla Foundation and Gordon and Betty Moore Foundation. Coordination applied to both strategy and funding (e.g., the locally-managed marine area movement). Although we recognize the importance of coordination for philanthropic agenda setting in Fiji, our focus remains on Packard.

METHODS
In order to understand how Packard affected the agenda in Fiji and Palau, we collected data via the following: semi-structured interviews with conservation practitioners and key Packard personnel, analysis of the entire portfolio of Packard-funded https://www.ecologyandsociety.org/vol28/iss3/art2/ projects, and knowledge co-production workshops. This paper is connected to a broader research project on philanthropic foundations in marine conservation called the Ocean Philanthropy Research Initiative (2018)(2019)(2020)(2021)(2022)(2023). The project was informed by a participatory research design process we conducted in 2018 with 81 practitioners in Fiji and Palau and 25 staff from six foundations working in marine conservation ). This paper addresses questions raised in the co-design process about how foundations influence marine conservation agendas. We selected Fiji and Palau because of the significance of Packard's support for marine conservation there, several of the authors' previous experience in those countries, and because of Packard's interest in understanding the impact of their exit there.
For our study we focus on three key questions. First, did Packard have a significant influence on marine conservation agendas in Fiji and Palau? Second, how did Packard's grant-making approach affect the type and level of their influence? Third, how did the Packard-supported conservation agenda (including individuals, organizations, projects, and ideas) influence other conservation priorities and agendas in Fiji and Palau, in some cases even after the funding ended? Our answers to these questions prioritize the perspectives and perceptions of conservation practitioners working in Fiji and Palau, many of whom were grantees of Packard.
Our study draws primarily on two sets of data collected as part of the larger project. The first dataset consisted of 82 semistructured interviews with marine conservation practitioners who worked in Fiji and Palau, as well as Packard employees who served as the primary points of contact for the two countries. Research respondents who worked in Fiji and Palau included Packard grantees (52), as well as other marine conservation practitioners (26) who were familiar with Packard's work in those countries. Respondents were primarily Fijian and Palauan nationals (60), but also included foreign nationals who lived and worked in country (16), and off island (2). They worked for local nongovernment organizations (NGOs), the private sector, large international NGOs with local field offices, government agencies, and research institutions (see Table 1). Thus, our results reflect perspectives of a professional conservation community that is primarily "local" in terms of their national origin and/or geographic location in-country. Given our focus on conservation professionals, results do not necessarily represent the views of broader stakeholder groups affected by foundation-supported initiatives. Our results must be read with this in mind, as interviewees' largely positive reflections on the success of Packard's agenda-setting role and processes are instructive but not necessarily representative. Interviews lasted from one to two hours and were conducted between March and September 2019. Interviews were recorded, transcribed, and entered into NVivo qualitative data analysis software. We used both open and axial coding processes (Saldaña 2021) to analyze responses to interview questions. The interview guide included such questions as: To what extent has Packard affected the focus of conservation here in Fiji/Palau? How and to what extent has Packard affected the selection of conservation priorities within your organization? What was it about the way Packard worked here that enabled them to influence the focus of conservation here? Our second dataset consisted of information on 34 Packardfunded initiatives in Fiji and Palau. We collected data through internal Packard documents and structured interviews conducted between May 2019 and February 2022 with 51 individuals who had worked on or had proximate familiarity with the initiatives. Although these data were collected primarily for a separate study on what happened to Packard-funded initiatives after the exit, participants gave open-ended explanations about Packard's influence on conservation agendas and priorities along with their responses to the survey questions. We used these data to validate and add nuance and depth to responses from the first dataset. Conservation practitioners are identified herein by a number and the country they worked in, or respondents are identified by a number and as being a donor. All interviewees from both datasets agreed to be listed in a separate appendix for this work, which can be found in Appendix 1.
Finally, we invited all respondents to attend one of two virtual knowledge co-production workshops held in June 2021. During these workshops, we shared preliminary results and asked participants to help refine our interpretation of the findings. The feedback and suggestions we received during these workshops validated and identified additional complexities within the findings. Workshop participants all indicated via a poll that they were "strongly" or "mostly" in agreement with the findings shared during the workshop.
We acknowledge that our own and interviewees' multifaceted relationships with Packard as both funder and research subject have the potential to affect our results and introduce real and/or perceived bias into the project. We are managing this with the support of an independent research advisory committee we have tasked with upholding the accountability of the project in pursuing rigorous research in a culturally appropriate way, with a multifaceted and balanced perspective on foundations. We have also offered confidentiality to interviewees and have continuously critically interrogated our own findings. Our confidence in our findings is further bolstered by three additional factors. First, interviews took place after Packard had announced their exit (and completed it for many initiatives and grantees in Palau); practitioners had little to gain or lose with respect to their relationship with Packard by sharing their candid viewpoints. Second, interviewees openly shared critiques of other donors, including some current donors, whom they contrasted with Packard. Below we share examples of this with quotes from diverse interviewees. Third, interviewees' descriptions and perceptions of Packard were highly specific and consistent across 1. Did Packard have a significant influence on the conservation agenda in each country?
Practitioners universally agreed that Packard significantly influenced the marine conservation agendas over a long period, even after they exited.
2. How did Packard's grant-making approach affect the type and level of their influence on the conservation agendas?
Practitioners emphasized that Packard's grant-making practices enabled them to successfully negotiate and pursue a shared conservation agenda that was perceived as well matched to local contexts and priorities. They highlighted the importance of Packard's commitment across five broad areas: Long-term relationships: Packard staff invested in relationships with practitioners over time. Authenticity, communication, listening, and trust characterized these relationships.
Collaborative decision making: grantees often (but not always) played a meaningful role in shaping funding decisions.
Convening and promotion of collective action: helped develop shared priorities and collaborative efforts on some conservation activities.
Long-term investment: enabled practitioners to design projects on realistic conservation timelines; experiment with more innovative ideas; and gave initiatives the time they needed to evolve and/or adapt when necessary.
Flexibility: created opportunities for practitioners to prioritize local conservation needs, align work with a changing local context, and be more innovative.
3. How did Packard-funded conservation influence priorities and agendas in Fiji and Palau, in some cases, even after funding has ended?
Practitioners underscored how Packard's investments influenced other conservation priorities at various organizational and temporal scales. We provide examples in two areas: Knowledge generation: Packard-supported research shaped conservation priorities by informing decision making or providing a basis for future conservation plans.
Capacity building: Capacity building empowered local practitioners to pursue their own conservation priorities by, for example, securing other funding and/or obtaining positions of decision making/power. different groups and countries, which we highlight in the paper with quotes from a wide range of respondents.

RESULTS
We first demonstrate that participants universally agreed that Packard had a significant influence on the conservation agenda in each country, and draw from our interviews to characterize that influence. Next, we discuss how this influence stemmed from Packard's grant-making approach. Finally, we discuss how Packard-supported initiatives, in turn, also influenced the conservation agenda. Table 2 summarizes our key findings.

Packard's influence on the conservation agenda
In response to our first question (whether Packard had a significant influence on the conservation agenda in each country), interviewees universally perceived that Packard had a significant influence on the marine conservation agendas in both Fiji and Palau that extended over a long period and continued even after their exit. They perceived this influence in the following ways: shifts in the conservation agenda over time, affecting the overall profile of marine conservation over other issues (see also: Lees and Siwatibau 2009); the emphasis on particular priorities (for example, area-based conservation via locally-managed marine areas, conservation areas, marine-protected areas, and/or ecosystem-based management [terminology and approach varied across Fiji and Palau]); increasing public awareness of specific conservation issues (e.g., sea turtle conservation); and building capacity among the network of locally-based conservation actors.
This influence may even have extended beyond Fiji and Palau through Packard's investments in the regional Locally-Managed Marine Area (LMMA) Network, which played a major role in sharing many concepts and the capacity for empowering communities. However, in light of Packard's exits from Fiji and Palau in 2020, the long-term durability of these priorities and activities will depend in part on whether and how governments and NGOs take up and continue conservation agendas supported initially by private philanthropy (Le Cornu et al. 2023).
Overall, respondents spoke quite positively about Packard's influence on the marine conservation agendas in both Fiji and Palau. One participant even referred to Packard as, "...the gold standard for foundations" (Palau 134). Positive statements reflect both the way respondents felt about Packard's overall effect on Fiji and Palau as well as how Packard worked, which we describe in detail in the following sections. These positive perspectives were often shared in contrast to other donors, both government aid and other philanthropic foundations.
The way in which Packard interacted with practitioners was often identified as unique, and contrasted with numerous other donors, including foundations and other types of donors. The quotations below are illustrative of how respondents in both Fiji and Palau emphatically differentiated Packard's more "bottom up" approach to agenda setting from other donors who wield their material resources and power to affect conservation agendas in a more "top down" manner: https://www.ecologyandsociety.org/vol28/iss3/art2/ Several interviewees felt there were some exceptions to this, and these were noted most often in relation to major shifts in Packard's strategy or program staff, or the impacts of engaging funding intermediaries who sub-grant or redistribute funds. In Palau, a few interviewees expressed concern that foundation funding (amid other political, social, and economic processes) may have played a role in the shift away from local and Indigenous models of conservation, which historically relied on strong customary marine tenure institutions (Carlisle and Gruby 2019 Though not specific to Packard or marine conservation, a 2009 study raised similar concerns broadly in Fiji, finding that 60% of the total conservation budget in Fiji was managed by international NGOs (Lees and Siwatibau 2009:22). The implication, they argue, was that "[in] both subtle and obvious ways this shifts the ownership of current conservation initiatives away from Fiji, from local people, and from local institutions that have the long-term responsibility for both the problem and solutions" (Lees and Siwatibau 2009:22). Packard's significant focus and investment in efforts to strengthen customary and Indigenous management of traditional fishing grounds through the Fiji Locally Managed Marine Area network may be an exception to this broader trend in Fiji, and could be one reason this concern did not arise with respect to Packard specifically in our interviews (Blackwatters et al. 2022).
Finally, we note that practitioners' perspectives about Packard varied depending on the individual Packard staff with whom they interacted, their organization's capacity, and their own background. The capacity and experience of practitioners and their organizations affected their interactions with Packard in their ability to navigate the power dynamics between donor and practitioners. Both donors and grantees reported that some grantees were better equipped to navigate interactions with donors to their advantage, and others simply found certain Packard staff easier to work with.

How Packard's grant-making approach enabled collaborative agenda setting
This section answers our second research question: how did Packard's grant-making approach affect the type and level of their influence on the conservation agendas? Respondents emphatically highlighted the importance of Packard's commitment across five broad areas: long-term relationships, collaborative decision making, the convening and promotion of collective action among the conservation network, long-term investments, and the flexibility of funding. Practitioners linked these grant-making practices to their ability to successfully negotiate and pursue a shared conservation agenda that was (mostly) well matched to local contexts and priorities. We discuss these in turn, below.

Long-term relationships
Packard staff based both in the Pacific region (Philippines and Papua New Guinea) and in the United States built trust over time by traveling once or twice a year to Fiji and Palau to get to know Ecology and Society 28(3): 2 https://www.ecologyandsociety.org/vol28/iss3/art2/ practitioners, making themselves accessible and listening to practitioner interests in order to understand local priorities. These relationships between practitioners and Packard staff enabled candid communication about conservation priorities and how best to achieve them, and, importantly, enabled practitioners to disagree openly with Packard staff without fear of losing a funder: Practitioners and Packard staff alike described how these relationships enabled them to shape initiatives together by acting as thought partners, listening to one another's perspectives, and engaging in mutual learning processes over time. As these practitioners recounted: Many practitioners noted that the personal qualities of Packard staff were an important part of being able to have this type of relationship. They emphasized how important it is for donors to hire staff who have a high level of cultural competence and are "a good fit" for the area they are working in. Respondents often described program officers as "friends," "mentors," and even "family" who genuinely cared about their work and their local relationships, as in this example: I think it's also people .... the people who were actually there, really cool people whom we interacted with oh man one of the best people I've ever met great mentors, people that you know, I can ask quality time, email anytime and ask question [ . . .] they're amazing people so to have a donor who's more than a donor who becomes a friend, a family member we consider really, really important. Someone who is not just you know, gonna give you money, but he's willing to hear your stories. To laugh with you and have a glass of beer and talk stories because they care. (Palau 136) It was also noted, however, that as a result of this dependence on rapport, relationships with Packard could fluctuate with changes in staff, especially program officers.
Collaborative decision making Packard was described as being more collaborative compared to many other donors, including other foundations (with a few exceptions) and other types of donors (e.g., multilateral and bilateral funding agencies and foreign governments). As one former grantee put it: "Packard was not pushing for their agenda. It wasn't like when you work with the U.S. Federal government, government grantors -they have agendas that they want to push and make happen in [a] short time" (Palau 61).
Practitioners explained that their relationships with Packard staff allowed for ongoing dialogue that informed Packard's funding decisions. Despite Packard ultimately holding the power to make final funding decisions, practitioners and Packard staff felt that grantees played a meaningful role in shaping those decisions: We ultimately set an agenda, but it was because we brought resources and then we had these partners that we trusted....there were a set of partners that dictated what we did. They understood our game, if you will, the process and they knew the place and we couldn't work without them, and they you know, masterfully worked with us on setting our priorities. (Donor 57) Respondents perceived that Packard's overall grant-making strategy for the Western Pacific Program and most of the initiatives they funded to implement that strategy were informed by the interests of practitioners working in Fiji and Palau. As these grantees explained: Although Packard's grant-making approach was predominantly described as "collaborative," there was variation in whose ideas ultimately informed funding decisions, and how. In some cases, Packard-funded initiatives were conceptualized entirely by practitioners working regionally. Packard would learn about potential interventions to fund through conversations with practitioners, and then use this information in decisions to support a particular initiative or organization. Respondents also described a number of funding decisions informed by the negotiated interests of practitioners and Packard staff. In those cases, sometimes Packard staff contributed input to the development of an initiative proposed by practitioners. Or conversely, an idea that Packard proposed would be further developed by practitioners. One practitioner summarized those efforts: "it was a real collaborative process" (Fiji Palau 116). A practitioner from Palau shared a similar experience:

I'm quite surprised that a funder is interested in
[The Foundation was] introducing this new program that they had just established, they were interested in...and then they asked us...do we think that could add value? And the answer was a resounding yes. The concept for the project came from Palauans...And Packard was so committed to ensuring that that was the approach... they would have somebody come in and help facilitate us through that discussion...so it was collaborative. (Palau 170) It should be noted that practitioners also mentioned a few examples where decisions were not made in a collaborative fashion, and these decisions tended to be viewed less positively. For example, Packard made a significant shift in their 2014-2020 Western Pacific Subprogram strategy, adding fisheries management to their longstanding focus on area-based conservation, alongside a shift in program officer. As described in the strategic plan (2013): "[I]n order to ensure sustainable management of nearshore marine resources in the Western Pacific, society (communities and governments) must effectively protect key coastal sites and establish and enforce nearshore fisheries management systems." As a result, several practitioners described narrowing their project's geographic focus or shifting gears, despite having misgivings about the new approach and its fit with local conservation needs. As one interviewee put it: "[W] e've just had to drop most of the things that we were doing" (Fiji 030). For those who were not interested in shifting their focus to fisheries management, Packard provided transition grants (up to US$100,000 for one year) to ease the impact of losing a donor (at least in Palau).

Convening and promotion of collective action
Packard encouraged the development of shared priorities and collaboration toward conservation goals among the local conservation networks in Fiji and Palau (Blackwatters et al. 2022). This was a primary purpose of Fiji's Locally-Managed Marine Area (FLMMA) network, which encouraged "collaboration among government ministries and departments, nongovernmental organizations, private or business sector, communities and individuals to better manage the 'i qoliqolis' of Fiji" (Govan and Meo 2011:6). In Palau, the Ecosystem Based Management (EBM) project funded grantees to explicitly work together on shared goals. Packard also encouraged practitioners working in various sectors to meet and develop shared conservation goals outside of funded grants.
The act of convening often led to insight on what direction to take initiatives or novel approaches to conservation challenges. Priorities were influenced when different practitioners met to share perspectives and discuss conservation needs. Packard encouraged and facilitated meetings among grantees and other practitioners (locally, regionally, and internationally), which often resulted in an exchange of ideas and sometimes collaborative conservation work. Occasionally, practitioners described this collaboration as having been "forced," resulting in friction among stakeholders or even competition for Packard funding for a particular issue. However, the majority of those who talked about the opportunities for convening and collaboration talked about these activities in positive terms.

Long-term investment
Interviewees reported that long-term investment in Fiji and Palau over 20 years enabled practitioners to identify, commit to, and pursue priorities and initiatives that require significant time and resources to implement. Although grant periods were similar to those of many other foundations (about an average of two years, although it varied from three months to four years), Packard maintained a norm of supporting many of its grantees consistently with back-to-back grants over time. This long-term commitment enabled practitioners to design projects that aligned with realistic conservation timelines. It also enabled experimentation with ideas that may have been more risky or innovative and gave those initiatives the time they needed to take shape, or to reflect and adapt if an approach did not work as anticipated. As practitioners explained, although this long-term engagement is not the norm in ocean philanthropy, it was a highlyvalued part of Packard's grant-making approach that enabled practitioners to conceptualize and commit to a different type of agenda than they could pursue with short-term funding: This long-term commitment also allowed practitioners to focus on project implementation, instead of time-consuming searches for funding, and empowered them to carry projects past an initial start-up phase. One practitioner described the advantage of these timelines as "security of guaranteed funding...compared to some funds that are committed annually or just three years and they're done" (Fiji 128). A Palauan practitioner who had worked with several local NGOs observed: Finally, long-term investments also contributed to the growth and increased capacity of locally-based conservation actors and NGOs who went on to have a major influence on the marine conservation agendas in both Fiji and Palau: If there was one factor that...I would say mattered most was consistent funding. In other words, people to be in a job for more than a year with some knowledge that they were going to have a job a year from now or a year after that. Because it gets people to start doing the types of professional development that they need to do. And by professional development, I don't mean, you know, some, you know, some retreat or something like that. It's investing in, they, not only do they themselves invest in their job, but their managers, the people around them, invest in them to do their job over a longer period. (Donor 15) Packard staff explained that long-term support and consistency in grant recipients and partners were intentional: "[W]e trust our partners and we follow them. And then you want to do that over time...you don't want to constantly shift between partners because you need to give time, things time to play out" (Donor 57). They felt that this generally met with positive results, but also acknowledged that it could sometimes be a liability if grantees perceived Packard as an "ATM." Several practitioner respondents also raised this challenge of striking a good balance between longterm commitment and encouraging sufficient critical selfappraisal among NGOs.

Flexibility
Practitioners described Packard as offering grantees flexibility in the design and direction of initiatives, enabling them to adapt to changing circumstances over time or in response to sudden unexpected disturbances, such as natural disasters. This flexibility was grounded in trusting relationships between Packard staff and practitioners and deep contextual understanding among program officers about the many reasons a project may not go as planned. Interviewees explained that this flexibility created opportunities for practitioners to prioritize local conservation needs, align their work with changing local context, and be more innovative: [Packard] gave so much flexibility they had so much understanding that at the community level, a lot of things can go wrong, and that was okay with them. But we learned from that, we adjust our plans and we move forward. And that to us, in the beginning, we were scared that we're gonna fail, we're gonna set ourselves to fail and you know, but to have that kind of a friend who was able to say, it's Okay it's okay, we make mistakes we, stand up, we move forward but we have to learn from those. And that meant a lot for us. (Palau 136) This flexibility was sometimes attributed to philanthropy as a category of donors more broadly and described by interviewees in contrast to other types of donors, such as the World Bank, or funding from the United Nations or the European Union: There's a lot more restriction around the types of things that [the World Bank] can be seen to be funding as well. Where...some of the philanthropies have a little bit more flexibility. (Fiji 31) Practitioners also talked about how the flexible funding enabled them to achieve other priorities than those that were explicitly outlined in a grant application.

The ongoing influence of the Packard-supported conservation agenda
Interviewees explained how, through its funded initiatives, Packard shaped other conservation priorities and agendas, sometimes long after grant periods ended. In the context of Packard's impending exit from Fiji, one individual from Fiji who worked at a large international NGO noted: "I think the Packard Foundation has actually helped set up the priorities for the next 20 years" (Fiji 05). Below we consider how Packard's agenda (and associated investments) for knowledge generation and capacity building influenced and continue to influence conservation priorities at various organizational and temporal scales in Fiji and Palau.

Knowledge generation
Initially in Fiji, Packard invested strongly in structured community knowledge generation (through implementation of the Learning Framework) and sharing through the regional LMMA Network and national networks. But in the second decade Packard made substantial investments in scientific research through their Western Pacific Program as well as their separate Conservation Science Program. They reported funding Ecology and Society 28(3): 2 https://www.ecologyandsociety.org/vol28/iss3/art2/ more research in Palau than in any other country in the Western Pacific Program. In both Fiji and Palau, Packard funding addressed a wide variety of topics, such as policy analysis, seagrass beds, coral bleaching, EBM, conservation area monitoring, fish spawning aggregations, and climate change indicators (Packard Foundation 2017, 2018. Practitioners described how priorities at the organizational, governmental, and community levels had been shaped by Packard-supported research, which informed decision making or provided a basis for future conservation plans: Packard funding actually provided us with funds to actually do socio-economic research at the community level, as well as some biological surveys...and these are very, very essential because they actually have us provide information on how much communities depend on the marine resources...and at the same time it helps to give us information on the status of their fisheries. Such information is very, very useful in informing decision making processes and management plans, not just at the community level, but also taking this to government level to push for changes in policy. (Fiji 05) Packard-supported research also "...inform [ed] partners about how to deliver conservation actions" (Fiji 85). In this way, research supported by Packard influenced what was being prioritized and done in practice. This practitioner describes, for example, how a specific research initiative informed efforts toward fisheries policy reform: So [Packard's] been able to support gathering the information needed in order to drive that as a priority... reaching that level of understanding, and then also the drive then to...say well this is a priority, this needs to happen. (Fiji 31) Other practitioners referenced the support Packard provided as having increased the amount of available information about local ecosystems to inform their choice of conservation priorities:

Capacity building
Capacity building is a second feature of Packard's funding portfolio that interviewees linked to Packard's ongoing influence on conservation agendas. Packard invested heavily in building the capacity of individuals and organizations working in Fiji and Palau. These investments, which came through the Western Pacific Program as well as via a separate Organizational Effectiveness Program, represented 15% of grants in Fiji (US$1,749,000) and 11% in Palau (US$887,000), and were primarily directed to Palauan and Fijian people and organizations. In an internal report, Packard estimated that as a direct result of their investment in Palau "there were at least two to three times more individuals-in villages, on NGO staffs, in government agencies-and organizations with the applicable skills, institutional infrastructure (laws, regulations, incentives), and financial resources needed to improve marine conservation and fisheries management in 2011 than there were in 2000" (Packard Foundation 2017). Capacity building took two forms. First, Packard directly funded initiatives designed to build capacity of individuals and local and regional conservation networks or to improve organizational effectiveness. These included, for example, grants to help local NGOs with organizational management skills, such as how to form a board of directors, or with conservation leadership training and the facilitation of regional learning networks, such as the regional locally managed marine area network. Second, capacity building not only happened through direct grants specifically for the purpose of capacity building or improving organization effectiveness but also grew out of the opportunities that Packard created, by funding projects and people to flourish, grow, and continue with their work in marine conservation in diverse ways, even after Packard funding ended. When asked how Packard influenced priorities at their organization, one respondent replied: Respondents noted, as above, that capacity building empowered local actors to pursue their conservation priorities (including securing funding from diverse sources) over the long term, even after Packard's funding had ended. As one practitioner working at a local NGO said, "It's not just me, it's the entire conservation sector and country of Palau that's benefiting from what I learned through Packard funding" (Palau 134).
Capacity building in local conservation networks also spilled over into the public sector, where former grantees/trainees in turn influenced government priorities. We heard many examples in both Fiji and Palau of individuals whose careers were supported by Packard funding and later went on to work for the government in a conservation context. One practitioner argued: In some instances, empowering certain locally-based actors to regrant funds was perceived to help align the conservation agenda to the local context. That worked well when those intermediaries were closely attuned to local conservation needs. However, there were other times when the flexibility afforded by Packard did not Ecology and Society 28(3): 2 https://www.ecologyandsociety.org/vol28/iss3/art2/ translate through the work of these intermediaries. For example, some respondents reported that when they received Packard funds through an intermediary, the stringent reporting requirements and lack of flexibility in other areas negatively affected their ability to carry out the work that was being funded.
Although Packard's capacity-building efforts were most often described in a positive light, there were a few cases in which the empowerment of certain practitioners (via grants to pursue particular priorities) was viewed less favorably. Sometimes conflicts emerged because individuals felt empowered to act in ways that others believed should have been approached differently. For example, one participant noted that "Packard... enabled organizations to push through some of their own agenda, and that sometimes is a good thing and sometimes a very dangerous thing" (Fiji 68). Additionally, whereas capacitybuilding efforts targeted local organizations and individuals, a majority of Packard's grants still went to non-local organizations. One respondent (Fiji 44) argued that "a lack of a Fiji marine resource management savvy civil society is a huge challenge at the moment" despite Packard's investments in capacity building, a point also raised more than 10 years ago by Lees and Siwatibau (2009).

DISCUSSION
Our paper presents an empirical analysis of one foundation's influence on marine conservation agendas in two Pacific Island countries. The case study is not broadly representative of how foundations work in marine conservation globally, and this is precisely why it is so important and instructive. Practitioners overwhelmingly described Packard's influence as positive: a success story in ocean philanthropy, in their view. We emphasize here how the entirety of Packard's grant-making approach was central to these positive perceptions of their influence on the conservation agenda: not only who and what they funded, but how they funded. However, our respondents echoed what much of the literature has also shown (McCarthy 2004, Ostrander 2007, Delfin and Tang 2008, Buteau et al. 2020: that such approaches have been uncommon in the broader field of philanthropy.

Lessons learned
Our research offers the donor and practitioner communities a number of lessons about how to think about and approach their roles in influencing conservation agendas. First, our study highlights the diversity of processes and moments for agenda setting throughout the grant-making lifecycle. Our analysis reveals agenda setting as an ongoing process of negotiation between a funder and its grantees that happens during the initiation of an idea for a grant (e.g., through building long-term relationships to generate initial funding ideas), the negotiation of what that grant looks like (e.g., through collaborative decision making), its adaptation in the face of unanticipated events (e.g., through flexibility), and legacy effects after the grant period ends (e.g., via knowledge generation and capacity building). Moreover, we show that agenda setting is not only about shaping specific conservation priorities, but also about influencing who gets to set those priorities (i.e., via local capacity building).
Second, our study highlights how critical grant-making processes are for determining how conservation priorities and agendas are influenced, and by whom. In Packard's case, these included hiring culturally competent staff who were able to build long-term relationships through which practitioners could have candid conversations about who, what, and where to fund; collaborative decision making about which of those ideas to pursue; convening conservation networks to advance shared agendas; long-term investment to enable practitioners to pursue priorities that required longer time horizons; flexibility to adapt an initiative to fit changing local contexts; and the ongoing work of the funded initiatives themselves. Considered together, these processes suggest that the type and mechanisms of foundation influence on conservation agendas are much more nuanced than suggested in the literature (Holmes 2012). Moreover, interviewees associated Packard's recognition and respect for practitioners with its high level of influence in Fiji and Palau. Importantly, Packard did not dictate the conservation agenda, but rather worked in partnership with grantees who themselves exercised significant agency and influence within the grant-making process.
Although Packard did many things right according to the practitioners we interviewed, there were still aspects of their approach that were viewed as too prescriptive, and these were viewed less favorably. We also recognize that our analysis is missing the perspectives of a critical population regarding Packard's agenda-setting role: the local communities, Indigenous rights-holders, and other stakeholders directly affected by the philanthropic-supported conservation agenda. Although Packard staff spent quite a bit of time in both countries, their primary engagement was with grantees, many of whom were non-local, and other professional conservation practitioners. For it to be deemed an unequivocal success story, we would need to more fully understand the perspectives of other affected stakeholders on the agenda and agenda-setting process, and the extent to which Packard grantees adequately engaged communities affected by Packard grant making and advocated for their priorities in collaborative agenda setting.

Implications
This work has implications for both research and practice. In summary, this paper reinforces an argument that others have begun to make: philanthropic foundations have the potential to significantly impact environmental agendas (McCarthy 2004, Barker 2008, Mallin et al. 2019, Cunningham and Dreiling 2021. Our work certainly adds evidence to this in the context of marine conservation. Some scholars have asserted that foundations play an excessively dominant role in setting the conservation agenda for environmental movements at local, national, and global scales (Holmes 2012, Jones 2012, Jenkins et al. 2017). The literature is critical on this point, and in certain circumstances rightly so, because domineering funders have significant potential to disrupt and override local institutions, culture, interests, and priorities, with poor outcomes for both people and nature (Bennett et al. 2020). Our results add nuance to this conversation: donor influence on conservation agendas can be a problematic or productive power, depending on how it is wielded. Donors can, and sometimes do, choose to engage with grantees as partners in co-creating conservation agendas that are at least in theory more contextually appropriate and responsive to local needs. To the extent that grantees effectively engage affected stakeholders and communities (and certainly not all do; Betsill et al. unpublished manuscript), that agenda has greater potential to be more equitable and just. Additional empirical research should more https://www.ecologyandsociety.org/vol28/iss3/art2/ deeply consider how equity, justice, and power dynamics shape interactions between donors, grantees, and affected stakeholders, and how these dynamics contribute to shaping conservation agendas.
In the meantime, this study can inform donor practice now. We are in a period of unprecedented growth in conservation philanthropy, with some of the largest commitments being made under the umbrella of the 30x30 global conservation agenda. As the global conservation agenda and associated funding travel to local contexts, donors must avoid overriding local conservation priorities and agendas. Packard's influence on the conservation agendas in Fiji and Palau was viewed positively because it happened in partnership with its grantees through a grant-making process that prioritized and legitimized local worldviews and practices. Positive perceptions and local legitimacy are, in turn, linked to more durable, effective, and equitable conservation (Bennett 2016). This more collaborative, long-term style of grant making deserves more attention from both scholars and the donor community, particularly as it is reported as being increasingly rare in practice, just as the field is expanding. As philanthropic foundations' funding for environmental initiatives continue to rise (Bloomberg Philanthropies 2021, Li et al. 2021), so too will their influence. We call upon the donor and practitioner communities to ensure that locally-based practitioners and associated communities play a significant role in shaping that influence.

Conflict of interest
Consistent with a knowledge co-production approach, we have embraced a multifaceted relationship with our funders who also serve as research participants and research subjects at various stages of the project. We have convened an external research advisory committee to help manage any conflicts of interest that arise from this arrangement.