Conceptualizing responsible exits in conservation philanthropy

Conservation philanthropy has grown significantly in the past decade. As the number of philanthropic‐supported conservation initiatives increases, so too will the frequency of exits—the ending of funding relationships. A trend toward “strategic philanthropy,” where foundations fund time‐limited grants, is already contributing to near‐constant exits. We draw attention to exits as a critical and ubiquitous—yet understudied—part of conservation grantmaking processes that can have tremendous impacts on the people and places foundations invest in. This paper begins to address this research gap with the first empirical study of exit processes in the context of ocean conservation philanthropy. We draw on an analysis of interviews and a knowledge co‐production workshop with donors representing 36 foundations investing in ocean conservation globally to: (1) develop a conceptual framework that broadens and clarifies definitions of exits and provides a common language to characterize exits along varied dimensions, and (2) derive best practices for exiting responsibly. This paper provides timely guidance for environmental philanthropy broadly, including the need to think about exits early and often as an integral part of the grantmaking strategy. Responsible giving must include responsible exits.


| INTRODUCTION
The past decade has witnessed significant growth in philanthropic giving (Kolbert, 2018), with the environmental sector seeing some of the highest rates of growth in the field overall (Australian Environmental Grantmakers Network, 2022;Environmental Grantmaking Association, 2021). For example, in 2021, the Wyss Foundation and eight other philanthropic organizations pledged the "largest-ever private gift for conservation" (Mufson, 2021) consisting of $5 billion to support an international push to protect at least 30% of the planet's land and sea by 2030, known as the 30 by 30 target (Greenfield, 2021). That same year, the Bezos Earth Fund announced that they would commit $1 billion of their $10 billion climate pledge to conservation (Kulish, 2021), also with the goal of advancing the 30 by 30 target. With the unprecedented growth in "green giving," we argue that now is the time to think seriously not only about the beginning of new funding relationships, initiatives, and partnerships-but also their end. Responsible giving includes responsible exits.
Responsible exits warrant research attention for several key reasons. First, there is a timely opportunity to include exit planning in the early development of what may be the "new green era" of philanthropic giving (Greenfield, 2021). Second, foundations are increasingly moving away from multi-year grants as they turn to alternative funding modes like "strategic philanthropy," where they fund time-limited grants that lead to nearconstant exits (Behrens & Gordillo, 2019;Boris et al., 2016;Brest & Harvey, 2018;Buchanan & Patrizi, 2016;Buteau et al., 2009;Rogers, 2015). Third, exits were identified as a top research priority by the ocean philanthropy community, which expressed an interest in applied research that can inform donors about how to design "respectful exit processes" (Gruby et al., 2021, p. 7). Fourth, poorly planned and executed exits may present tremendous risks and implications for the people and places foundations invest in. The consequences can be broad and devastating for conservation organizations (e.g., wasted resources, weakened governance, loss of community trust) and can ultimately compromise conservation goals (Ruiz-Miranda et al., 2020). Grantees may fear that their organizations or initiatives will suffer financial setbacks during and after an exit, which could lead to reduced capacity and sustainability of conservation initiatives and raises broader issues of resource dependency (Davis & Cobb, 2010;Hillman et al., 2009;Le Cornu et al., In preparation;Pfeffer & Salancik, 1978). While scholars and practitioners tend to focus on the financial aspects of exits, there is also a recognition that more than money is at stake. Funders and grantees can develop close relationships which means that exits are also personal and, in some cases, long-term partnerships-sometimes friendships-may end along with funding (Kibbe, 2017;Perez, 2017). Kibbe (2017, p. 50) even compares exits to a "breakup." This is why exits are often characterized as risky, painful, confusing, and potentially damaging and destabilizing forces (Kibbe, 2017;Petrovich, 2011). But is this always the case? What can foundations do to manage exits in a responsible manner and reduce the potential for adverse impacts? Answers to even the most basic questions about foundation exits remain elusive: "given the frequency and ubiquity of foundation exits, the literature is painfully thin" (Kibbe, 2017, p. 50).
This paper begins to address this gap with the first empirical study of exit processes in the context of ocean conservation philanthropy. The growth in "green giving" is reflected in the oceans, where private foundations are injecting hundreds of millions of dollars each year to address ocean degradation and climate change (Berger et al., 2019;CEA, 2017;2018;Hamilton et al., 2021). Philanthropic grants for ocean conservation have more than doubled in the past decade (CEA, 2021), and foundations have been recognized as "increasingly important players" within ocean conservation (Wabnitz & Blasiak, 2019, p. 2). Gruby et al. (2021, p. 11) compare them to "keystone actors in ecosystems" due to their "profound and disproportionate influence on conservation agendas, research, organizations, networks, policy, and the local societies affected by these interventions." This paper draws on an empirical analysis of interviews and a knowledge co-production workshop with staff from 36 foundations working in ocean conservation globally to: (1) develop a conceptual framework that broadens and clarifies definitions of exits and provides a common language to characterize exits along varied dimensions, and (2) derive best practices for exiting responsibly. We emphasize there is no "one-size-fits-all" approach to exits. Instead, the typology and best practices can be used by funders and grantees as a starting point for conceptualizing, planning, and implementing exits in a more intentional way. While our empirical focus is ocean conservation philanthropy, there was nothing specific to this context in our findings that limit their broader applicability. The contributions of this paper will be useful for the conservation field generally and other philanthropic sectors beyond the environment.

| PHILANTHROPIC EXITS
With the increasing reliance on philanthropic foundations in conservation and the trends toward limited-life or spend-down foundations, where foundations themselves have a defined endpoint, and strategic philanthropy, where foundations increasingly fund time-limited strategic initiatives (Behrens & Gordillo, 2019;Boris et al., 2016;Loh & Buteau, 2017;Markham & Ditkoff, 2013;Petrovich, 2011;Wolcheck & Renz, 2009), there is a critical need and growing interest in better understanding and improving exit processes. In 2017, the Foundation Review published a special issue on "exit strategies" which included eight articles written by and for foundation staff. Every article identified philanthropic exits as a critical topic that deserves more research and attention. While there is significant practical knowledge and experience with exits, this topic has not been a priority for systematic research, learning, or discussion at the field level.
This paper addresses two key gaps in the limited literature on foundation exits. First, there is no shared definition or typology of philanthropic exits. When referring to exits, authors rarely define them explicitly. Rather, they implicitly characterize them in diverse ways, such as: closing a program or initiative, ending a funding relationship with specific organizations, or withdrawing from an entire country, region or field (Kibbe, 2017;Perez, 2017). Current conceptualizations lack clarity, which limits systematic learning, and are too narrow to capture the range of exit types and dimensions that matter for planning and managing exits responsibly. In some cases, this lack of conceptual clarity may mean a funder and/or grantee does not even recognize when they are going through an exit, or results in a failure to plan or pursue exits at all: "too often funders and grantees fail to acknowledge that exiting is a part of the investment life cycle. As a result, they also fail to discuss the realities of it and the importance and value of exiting" (Perez, 2017, p. 105). In addition, while the literature mentions drivers of exits (Herweijer & Kerkhoven, 2013;Petrovich, 2011), these discussions are usually quite general and descriptive and not considered in relation to diverse exit types. A comprehensive conceptual framework that captures the diversity of exit types and associated drivers, as we contribute here, can help foundations, grantees, and practitioners to better identify, discuss, plan for, and navigate exits.
Second, and relatedly, there is insufficient guidance for planning and conducting responsible exits. There are a handful of studies in the gray literature that consider exit best practices (Herweijer & Kerkhoven, 2013;Mackinnon & Jaffe, 2007;Petrovich, 2011) and, although growing, the peer-reviewed literature is still quite limited. The existing literature tends to focus on how funders can exit from specific grants or programs (Gienapp et al., 2017), with a few publications focusing on strategies for exiting from a field or on spend-down foundations specifically (Fleishman, 2017;Gardner et al., 2005;Halverstadt & Kerman, 2017;Loh & Buteau, 2017;Markham & Ditkoff, 2013;Ostrower, 2009Ostrower, , 2011. We identified only one publication focusing on how funders exited from a major place-based initiative (Cao Yu et al., 2017). In this literature, the authors focus on the importance of: (1) planning for and carefully executing the end of a funding relationship, (2) having early and continuous communication with grantees, and (3) developing strategies for ensuring sustainability after the exit.
There is a need for much finer-grained guidance to conduct responsible exits. For many funders, there is still a great deal of uncertainty and lack of planning and learning around exits (Beadnell et al., 2017;Giloth & Gewirtz, 2009), which can lead to irresponsible exit practices where foundations simply do not address or plan appropriately for the exit. Cao Yu et al. (2017, p. 64) explains that: "the [exit] process is too often treated as an afterthought, and funders rarely devote enough time to planning for and working through the tensions and issues that arise." A more comprehensive framework delineating best practices for exiting responsibly in various exit contexts could help funders, grantees and other practitioners think about exits as a priority on par with other parts of their grantmaking strategies.

| METHODS
Our study conceptualizes responsible exits in conservation philanthropy by examining the process of exiting as it is practiced by foundations funding ocean conservation. This study was conducted as part of a larger project on ocean philanthropy, where we used a knowledge coproduction approach (Lemos et al., 2018) to engage more than 100 individuals from local, regional, and international NGOs, foundations, national governments, universities, and intergovernmental organizations to develop the project's research themes . Responsible exits emerged as one of five research priorities . Our analysis is based on semi-structured interviews conducted in 2019 and 2020 with 47 individuals (primarily program officers from private foundations) with experience working-formerly or currently-with 36 ocean foundations (see Appendix, Table A1), and a virtual knowledge co-production workshop held in 2021 with 41 individuals (32 of whom also participated in the interviews) to share and to co-interpret our preliminary findings. Participants were initially selected based on an assessment of the foundations with the largest investments in oceans in 2018-2020 from the fundingtheocean. org website, then through snowball sampling. Additionally, some participants responded to a call for research participants from our funders sent to the Biodiversity Funders Group's Marine Conservation Program listserv. Our sample includes 10 of the 20 largest marine philanthropic funders globally in the period 2010(CEA, 2018. Each interview lasted between 1 and 1.5 hours and included prompts to encourage reflection on: (1) what defines an exit; (2) why a foundation decides to shift funding away from a geography, organization or issue; (3) lessons learned from a specific exit "story" or "experience"; and (4) advice for those embarking on an exit.
We conducted a systematic thematic analysis of interview transcripts and detailed notes from the workshops using QSR NVivo to identify general themes for the exit typology and best practices. Then, through an iterative process with co-authors and with feedback from workshop participants we nuanced the typology and discussed challenges and opportunities for conducting responsible exits. This process culminated in the development of: (1) a conceptual framework that identifies three different types of exits, their associated drivers and three other dimensions of difference, and (2) a set of best practices that can be implemented in various exit contexts. We present these findings below.

| Exit typology
Our typology captures the diverse ways that research participants define and talk about exits ( Figure 1). These results suggest that exits vary across a broader range of dimensions than is currently recognized in the limited literature on philanthropic exits. First, we differentiate between three exit types: embedded, organic, and forced. These types are defined primarily in terms of their distinct drivers. We then consider how exits may vary along three additional dimensions: what is ending or shifting, at what scale and to what degree. Overall, the typology highlights the complexity of an exit as a process relevant to all stages of grantmaking. In so doing, it reveals diverse opportunities for applying exit best practices from the moment a grant is initiated until well after it has concluded.

| Exit types
Embedded exits are defined explicitly -and often earlyin the grantmaking process and associated with a definitive and planned ending. These exits are most common in time-limited strategic grants with no expectations for continued funding, long-term initiatives with a set duration, and in some cases, limited-life foundations that operate from the beginning with the intention to spend down all their resources within a given period. What was particularly striking about embedded exits was that participants often did not think about them as exits. Some would even begin the conversation claiming they had little experience to share. Here is how one participant wrestled with the characterization: It wasn't really an exit per se, though is it? […] We had a few one-off grants, or two-year grants with organizations but then we hadn't actually supported them for a couple of years. By the time we made this move, we haven't been through that process of supporting [grantees] in a space for several years and then leaving. (RP29) In our view, embedded exits are a particularly important type of exit to reflect on and learn from in light of the growing interest in strategic philanthropy, where exits from time-limited grants are fundamental to the grantmaking philosophy and require near-constant attention and management from both funders and grantees.
Organic exits emerge in the course of the grantmaking cycle due to a specific driver or a combination of drivers both internal and external to the foundation. Examples of internal drivers may be that the board or program officer changes priorities based on new scientific discoveries, donor fatigue, results from a mid-way evaluation, personal interests or passion. One participant also made a "spread the love" argument: There is so much to be done in this field. And when we select a handful of places, lucky them. We owe it to the planet to spread the love is what [the funder] would say. And so, we can't just commit ourselves forever to [a place]. (RP7) Organic exits may also be driven by other factors. For example, a project, program or initiative may achieve its goals or financial sustainability earlier than expected-or the opposite, when insufficient progress is made. These exits may occur because a grantee or organization's performance is unsatisfactory, or a grantee or organization lacks engagement or communication with the funder. There are also instances of organic exits when a new funder enters a space and takes over some of the work.
Forced exits tend to be unexpected, unplanned, and abrupt and can be particularly disruptive for foundations, grantees and other practitioners. They occur for reasons outside of a foundation's direct control. For example, drivers may include sudden shifts in personnel at the foundation (e.g., death of a board member), serious financial setbacks or economic downturns, or mismanagement of funds by grantees. Sometimes a foundation can be asked to leave a geography, or a political context makes it too hard to achieve goals. All research participants who had direct experience with forced exits described them as a painful and difficult process for all parties involved. The following are a few examples, in their own words: Sometimes you have to exit because the economy tanked. I lived through the bubble bursting, we had to lay off half the foundation staff, and cut back on a lot of grantees. And it's tough. But sometimes your hand is forced. (RP35); So, sadly, our trustee passed away […] and the program was dismantled. So that [exit] process was really just cold. I mean, there was just no process. (RP38)

| Exit dimensions
We found that each exit type varies along three additional dimensions: what is ending or shifting, at what scale, and to what degree. The first dimension is related to what is ending or shifting. Foundations may exit from a particular geography (i.e., from a specific locality, country or region); a conservation issue (e.g., a shift from focusing on marine protected areas to small-scale fisheries); or a specific project, program, or initiative. The what that is ending may also include a funding relationship with a grantee or an organization (e.g., when conservation goals are achieved or, conversely, when goals are hard to achieve), or a relationship with a foundation staff member (e.g., when a F I G U R E 1 Conceptual framework for philanthropic exits. program officer quits), which can sometimes come with ideological and/or financial changes as new staff bring their own ideas, grantmaking styles, and funding priorities. A shift in staff or programmatic foci can be experienced as what some participants referred to as "mini exits" where the organization may stay engaged with the same geography and grantees. As this participant explains: We have not really had a hard exit, we've had transitions. So, the marine protected area work went from design to implementation. Many of the grantees have continued to engage but have shifted gears, and new grantees have come in, some have fallen out. But it wasn't like we left the space. (RP15) The second dimension refers to the scale of the exit and is related to time, geography, and investment levels. The temporal scale of the exit calls attention to the differences in ending long-term, medium-term or shortterm grants: ending a short-term strategic grant for a one-year project with a specific organization is qualitatively different than ending a 20-year investment including multi-year grants for multiple organizations. Second, the geographic scale of the exit could range from small to large, calling attention to the difference between exiting from a local, regional, national or global scale project, program or initiative-and implications for the types of grantees and communities who are likely to be affected. Third, the financial scale of the investment could range from small to large. We included this in acknowledgment that the relative size of the project, program or initiative that is ending will matter for the exit process. Again, we expect the impacts from ending a small investment versus a larger one to differ in ways that shape the stakes of an exit process and activities for managing it well.
The third dimension is the degree of the exit. Here we draw a distinction between complete or partial exits, with a focus on both the funding relationship and other relationships (e.g., personal and working). A complete withdrawal refers to the discontinuation of both funding and engagement with grantees and other relevant practitioners. A partial withdrawal may involve discontinuation of new large grants, scaling back of funding, continued funding through endowments, or a cessation of funding coupled with ongoing dialogue and working relationships with current or former grantees. Partial exits can also occur when a foundation or individual staff member has invested heavily in their relationships with grantees and wants to continue fostering these relationships even after an exit, maybe by providing connections to other funders or simply giving advice. This conceptual framework broadens and clarifies definitions of exits and provides a common language to characterize exits along varied dimensions. The exit types and dimensions co-occur and they should be considered together to characterize where a given exit "sits" within the conceptual framework. In identifying key distinguishing features, the typology may help reveal broader patterns of difference and similarity that can help funders, grantees and other practitioners deliberate, share lessons, and make informed decisions on how best to navigate specific exit types, a point we turn to in the next section.

| Best practices for exiting responsibly
We identified a total of 52 best practices, which include practices funders actually implemented as shared in their exit "story" or "experience", as well as more general lessons learned and advice they had for other funders. We organized these best practices into three categories, principles, administration and management, and sustainability, and 17 sub-categories (see Tables 1-3). Importantly, there is not a single "exit moment"; we identified best practices within all stages of the grantmaking process, with a majority appearing at the inception of a new grant. We emphasize that these best practices are a good starting point for reflection, discussion, planning and actions that may be further contextualized to fit particular organizations, exit types, and grantmaking contexts.

| Principles
The first category introduces overarching guiding principles and ethics. This category includes best practices that together reflect a "code of conduct" that guides the exit process, grounded in principles of respect, inclusivity, reflexivity, transparency and flexibility. First, research participants emphasized that being thoughtful and respectful of local contexts, especially when funding smaller, local, and/or Indigenous-led organizations, was key to conducting a responsible exit. This research participant illustrates the point that the principle of respect coupled with long-term relationship-building is crucial in settings where Indigenous communities have been disrupted over the years: If nothing else, have a more gentle withdrawal and transition because ultimately there are certain best practices in the field that we value. It's long-term, it's flexible and multi-year and it's sticking with things especially in the Indigenous world where things got so dismantled for so long. The infrastructure and capacity need to be rebuilt and it takes a long time. It's easy for us over the years to do this but other funders come in and give a year [of funding] and leave. (RP32) In addition, participants emphasized that the process should be inclusive of all players affected by the exit, including grantees, program officers and communities that are directly affected. For example, this participant explains that the foundation works with an Indigenousled council to help them make decisions, including the decision to exit: This new Indigenous group and council has actually preferred to focus on the Americas and the Pacific, so we're now in the process of exiting. (RP32) Being inclusive of all relevant players during the design and the implementation of the exit is one step forward in addressing the issue of power dynamics in decision-making during the exit.
Participants also indicated that reflexivity is important. They specifically mentioned intentionally learning from the exit process to improve it, whether it is learning from mistakes, getting feedback from grantees or asking other funders for advice about their exit experiences. This research participant reflected on exits as an opportunity to learn from affected communities: If we were really a lot more intentional in thinking about exits as an opportunity for […] putting the folks in the communities that have experienced the changes in the driver's seat to tell that story, it can be both a therapeutic moment and it also can be empowering.
[…] Some of the great examples I've heard about but never done is like using videography, you know, giving everybody a camera within that community. (RP3) Research participants also emphasized the importance of being transparent about decision-making processes, including their "messiness": T A B L E 1 Best practices for exiting responsibly: Principles Note: The table displays exit best practices pertaining to the "principles" category (n = 20) in (1) their respective sub-categories (on the left) and (2) the different phases of the grantmaking process (at the top) based on when research participants mentioned they were implemented. The most frequently mentioned best practices are in red. Note: The table displays exit best practices pertaining to the "Administration and Management" category (n = 24) in (1) their respective sub-categories (on the left) and (2) the different phases of the grantmaking process (at the top) based on when research participants mentioned they were implemented. The most frequently mentioned best practices are in red.

T A B L E 3 Best practices for exiting responsibly: Sustainability
Note: The table displays exit best practices pertaining to the 'Sustainability' category (n = 8) in (1) their respective sub-categories (on the left) and (2) the different phases of the grantmaking process (at the top) based on when research participants mentioned they were implemented. The most frequently mentioned best practices are in red.
If you're looking at exits, I think it's important to be honest about how some of the actual decision-making at foundations is done. It's messy, and even when it's intended to be really thoughtful, it can get really distorted and it doesn't sometimes turn out the way even that the leadership wants it to. (RP11) Finally, participants emphasized that it is also important to be flexible with the process and be ready for surprises and unplanned events. This is especially true for forced exits, when funders and grantees have to respond to unexpected circumstances. Some suggested that having a living exit strategy document or simply extending grants when needed are simple practices that could introduce some flexibility in the process while being guided by the original plan.

| Administration and management
The second category describes concrete, actionable activities that foundations can take to plan, manage, evaluate and coordinate the exit responsibly. It includes measures to address the exit context and its timing, communication, and financial and non-financial support. Most research participants emphasized the importance of planning the exit from the beginning of the grantmaking process, as in the following: Exits, to me, that's like life, birth, and death. It's going to happen and if we could just face the mortality upfront, we might live better and more fully […] and we can get out of that fantasy world on both sides. (RP32) When funders and grantees face and accept the reality that the relationship can eventually end, it enablesboth groups to better prepare for the exit, set clear and realistic expectations and end goals, a clear timeframe, and start building a "handover" plan.
Participants also identified best practices related to the way foundations manage and evaluate the exit. In this regard, participants said that ramping down gradually is important, especially for organizations that are already struggling and may further suffer from an abrupt exit. In addition, conducting mid-way evaluations was identified as a best practice because it allows funders to assess the progress of a certain project or the needs of a grantee and adapt depending on the situation (e.g., provide additional support or end the relationship sooner). Post-exit evaluations were also identified as important to assess the success of a project or a partnership with a grantee and its associated conservation outcomes. Participants highlighted the importance of using a clear set of metrics and criteria for these evaluations and having a clear reporting system that grantees can use to inform their work. Although considered necessary by many, these evaluations are not always implemented, and some participants questioned their utility: At the end of a strategy, we'll do some exit evaluations. Occasionally, there's a big tension in the foundation, whether or not we should be investing in those because oftentimes they're things that are for grantees [but] it's not clear if the grantees want them. They're sort of puff pieces. (RP3) The context and timing of the exit also emerged as key considerations. Some participants recommended postponing the exit if the context and/or the timing of the exit are unfavorable for the grantee, while others even questioned the premise of exiting altogether. During our knowledge co-production workshop, research participants engaged in a passionate discussion where several argued that exits are ultimately a value judgment and that the field needs to rethink the assumption and/or goal of an eventual exit entirely. Participants talked about the need to be more explicit about an exit's definition, methodology and conceptualization (i.e., is "exit" the wrong word?) and emphasized that the way exits are currently conceptualized reinforces a power dynamic where funders "call the shots," assuming that there is little opportunity for grantees to inform or shape decisions to exits. Other research participants questioned whether exits are necessary at all and argued that donors should prioritize long-term investing.
Having open and frequent communication with grantees was identified as particularly important for building relationships between funders and grantees where the exit can be co-constructed. A strong relationship can help foster transparent communication channels where funders and grantees can be honest and give each other feedback, and work through challenges together, as this participant explains: I think trust and relationship building is key so that you can have those clear honest channels for conversation and receive real feedback on what an exit looks like.
[…] Actually, this is good practice of philanthropy in general. (RP4) Participants emphasized that communication is particularly key when discussing an exit timeframe and expectations with grantees. Specifically, participants stressed the importance of giving ample notice before exiting and being crystal clear about when and why it is happening. As this research participant explains: Notice is key and giving a signal and then executing on that signal is really essential. I just recently had a conversation with a dear colleague who runs an NGO and she was lamenting the fact that one of her major funders kept saying maybe, maybe, maybe [about the exit], and she was essentially desperate. (RP17) Research participants also emphasized the importance of internal communication within the foundation-especially among program officers, program directors and board members, so that expectations and goals that are set at a higher level within the foundation are communicated properly to grantees and other practitioners, usually through program officers.
Participants identified best practices related to coordination to help ensure the void of their exit is filled by others (e.g., foundations, government, etc.). This includes working closely with governments to transition to other sources of funding after the exit or "showcasing" the work that has been done to a potential funding audience and gauging their interest in taking over and continuing the work after the exit. This is also true when a foundation leaves a space where another foundation is continuing to work. The exiting and remaining foundations should try their best to coordinate the exit. If multiple foundations exit around the same time, it can have devastating implications for grantees who would face a significant funding gap.
Finally, foundations can provide financial (e.g., setting up an endowment or providing exit grants) and nonfinancial support to grantees during the exit. Non-financial support can include providing grantees with resources beyond money, through networking or simply playing an advisory role if grantees need guidance on what to do next in terms of projects, partnerships, or funding sources. This research participant lists some of the non-financial support foundations can offer: We recognize that our primary goal is to offer financial resources, but what are the other things where we could maybe help? Is it media connections? Storytelling? Is it networking? Is it introductions to other funders or funding spaces? (RP28)

| Sustainability
The third category focuses on strategies that foundations can implement to ensure the long-term sustainability of the funded project, program, initiative, or organization after the exit. This category includes best practices related to building capacity and autonomy of grantees, and developing strong funder networks. First, participants underscored the value of including capacity-building strategies as part of the grantmaking process as they are intended to contribute to outcomes that endure beyond the exit. Foundations that invest in capacity-building may provide grantees with organizational, human, and infrastructure support to foster leadership and develop the skills and confidence to continue their work and/or pursue new priorities after an exit (Enrici et al., In review;Blackwatters et al., 2022). In addition, participants discussed the importance of avoiding resource dependency by investing in sustainable financing mechanisms and supporting grantees in diversifying their sources of funding to ensure they are financially stable once the exit comes, if it is something grantees see as beneficial. Participants emphasized the importance of being creative with these mechanisms whether they are endowments, government grants, revolving loan funds, community trusts or simply introducing new funders to grantees. This research participant defines exit strategies as a "pathway" to sustainability: An exit strategy is a deliberate pathway toward financial safeguards and transformational change that, once you leave, permanence is guaranteed because the structure is in place, like human capacity, but also financial flows. (RP9) Finally, participants stressed the importance of building a strong network of funders who can "take over" after an exit. New funders could be philanthropic foundations already working in the space or planning on entering the space, but also official development aid, businesses, or governments. Research participants saw the value of transitioning to other funders as an exit best practice to avoid letting down grantees, as this research participant explains: Get somebody else to take over. That's the best way to exit. You don't let down your grantees, you just transfer [them] to another funder. (RP35) 4.2.4 | Tailoring best practices to specific exit types?
During the co-production workshop, research participants suggested that not all the best practices can be implemented in every situation and that it is important to differentiate which best practices are most relevant for each type of exit. While we did not comprehensively investigate this link in the current work, a few patterns arose in our data.
Within each category, we observed that some best practices came up more frequently than others across the interviews (see Tables 1-3 for the 15 most frequently mentioned best practices in red). This may imply they are relevant across exit types, but additional research is needed to confirm. For example, research participants stressed the importance of having an exit strategy at the beginning of the grantmaking process and to have clear and continuous communication regarding the exit throughout the grantmaking process. Participants also mentioned mid-way evaluations as a key step to assess the progress of a grant and to make appropriate decisions for continuing the work or phasing it down. Capacitybuilding and creating sustainable financing mechanisms also emerged as transcending best practices that are necessary for grantees to attain a level of independence and to be sustainable long after an exit. Finally, ramping down slowly and giving ample notice were also consistently mentioned as important to help funders, and especially grantees, have a smoother transition. Still, we noticed that even these more commonly mentioned best practices came up relatively less frequently in interviews with participants who experienced embedded exits compared to those who experienced organic and forced exits. One reason could be that they are simply less relevant for embedded exits. Another is that these participants did not think about embedded exits as "exits" and thus had not yet honed a point of view on what constituted exit best practices in these situations-whether they include best practices captured in our study or others unique to embedded exits. We suspect that there is significant opportunity for foundations to better plan and manage embedded exits in particular. Further research is needed to add nuance to this discussion. In the meantime the typology and best practices can be used in combination to stimulate dialogue among funders about which best practices are most useful for their specific exit scenarios.

| Challenges to implementing best practices
Research participants identified several challenges to implementing responsible exits during the co-production workshop. These include internal constraints such as lack of capacity (e.g., lack of resources, knowledge or experience to execute the exit, or staff turnover) and hierarchical issues within the foundation (e.g., decisions to exit being made at the board level with program officers lacking control around when and why exiting). They also shared challenges related to managing relationships with grantees. For example, participants spoke of the tensions between grantees' expectations/needs for continued funding; the appropriate roles and responsibilities of grantees (i.e., should it be entirely the responsibility of funders to manage the exit responsibly?); and navigating relationships with grantees who have a hard time facing the reality of an exit and not being proactive in planning for it. In addition, they raised questions around their relationship with other funders working in the same space. For example, research participants were worried about the impacts of their exits on other funders because it can shift an unwanted funding burden to their colleagues, which can create disagreements and tensions. Future research and discussions should focus on how to overcome these challenges and others.

| DISCUSSION
This paper draws on an empirical study of exits in ocean conservation philanthropy to distill a conceptual framework and lessons learned that are relevant for research and practice in this specific context and beyond.
First, we advance the small body of literature on philanthropic exits by adding conceptual clarity around definitions of exits and nuance to discussions of exit best practices. The conceptual framework can provide a basis for more systematic research and learning about the different types of exits and associated best practices for managing them. A limitation of this study is that it solely reflects the views and experiences of donors. Future research must incorporate perspectives from grantees and other affected communities on what constitutes a responsible exit. There are many additional topics ripe for continued research, including: understanding the long-term impacts of different types of exits, grantees' experience with exits, impacts of exits on affected communities, and further refinement in best practices and how they may be tailored to particular types of exits. There is also a need to consider how broader systemic issues like resource dependency shape exit impacts. Additionally, there is a clear need for research on how funders overcome the challenges they identified in terms of implementing the best practices. More broadly, in light of the rapid growth in "green giving"-and associated exits-we hope this work can help elevate philanthropic exits as a topic deserving more scholarly attention.
Second, this study has potential to inform practice, not only for philanthropic foundations but also for grantees and other conservation practitioners such as NGOs, governments, other types of funders, and local communities. Foundations could use the typology to identify the type of exit they plan on conducting or are currently going through. The typology offers a shared language for foundation staff and boards to use internally and across organizations, to discuss, plan for and implement exits; share lessons learned; and deliberate on how to solve common challenges, or coordinate exit processes. Additionally, foundations could draw on the best practices to proactively create responsible exit strategies and embed them within grantmaking process as soon as they start investing. While some of the best practices will be more or less relevant to particular exit contexts, the broad categories transcend all types of exits and are a useful starting point for more contextualized discussions about how to design a responsible exit. In this sense, they may also provide a useful starting point for funders who may find the task of exit planning daunting.
This work may also help grantees and other ocean conservation practitioners to take a more proactive role in exit processes. This is especially relevant in the context of an increasing push for a more inclusive philanthropic field that: "aims to create broader communal involvement and benefit, to expand the circle of those who participate in shaping public decisions, and to center philanthropy in community, rather than in managerial efficiency and effectiveness" (The Faculty of the Lilly Family School, 2020). The typology and best practices provide a basis for initiating conversations about exit planning as early as possible-a conversation grantees can initiate if funders fail to do so. When grantees are working with long-term funding that is potentially renewable, the typology suggests they should be mindful of events (i.e., exit drivers) that might lead to forced or organic exits and better prepare for it. In addition, grantees and other practitioners could have more proactive conversations with funders about the type of practices they can expect or would like to see around exits-or push back against or question the rationale for exiting in the first place. For example, they could create a contextualized set of best practices they wish to see around exits and share it with the funders they are working with as a basis for a co-produced exit strategy (e.g., discuss a realistic timeline, explain the type of capacity or needs they will need to be sustainable after an exit, ask for non-financial support, ask for introductions to other funders or request a certain type of program evaluations). This would also be an opportunity for grantees to include affected communities, who are seldom engaged directly in exit processes. There are a number of emerging initiatives aimed at working toward a more inclusive philanthropic field, such as the "Trust-Based Philanthropy Project," "Center for Evaluation Innovation," "Fund for Shared Insight," and "Decolonizing Wealth Project." These resources offer additional tools that could be tailored in the context of exits (e.g., providing long-term flexible funding, offering support beyond funding, listening to the people at the heart of the work, etc.). In addition, grantees also develop their own exit strategies from conservation initiatives (Ruiz-Miranda et al., 2020;WWF, 2017). These exit experiences coupled with our framework offer opportunities for funders and grantees to learn in partnership and coordinate plans for exits in a more deliberative and holistic way.
Exits can be messy, scary, difficult and painful processes that can have significant risks for both funders and grantees, but they need not and should not be this way. With the rapid growth in green giving, there is a pressing need for applied research on conservation philanthropy that can inform practice. We call upon communities of research and practice to give exits a more prominent place in our work as responsible exits are an ethical imperative with tremendous opportunity to positively impact the trajectories of philanthropic-supported initiatives, grantees, practitioners and communities.

AUTHOR CONTRIBUTIONS
All authors conceptualized and designed the study. Elodie Le Cornu led writing and analysis with substantial guidance from Rebecca Gruby and Michele Betsill and critical input from all authors. As Principal Investigators, Rebecca Gruby, Michele Betsill and Xavier Basurto, secured funding and served as research advisors. Rebecca Gruby and Michele Betsill conducted all interviews. Elodie Le Cornu, Ashley Enrici and Jeffrey Blackwatters transcribed interviews and managed data. All authors organized and participated in the knowledge coproduction workshop.

CONFLICT OF INTEREST
Consistent with a knowledge co-production approach, we have embraced a multi-faceted 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.

ETHICS STATEMENT
The purpose of the interviews and the use of data were explained to all research participants. All research participants agreed that the content of the interviews could be used for scientific analysis and published in an anonymous form. Research participants listed in the appendices gave explicit permission to be acknowledged for their time.