(Illustration by iStock/Rudzhan Nagiev)
Not long ago, a clear sign a foundation took results seriously was an evaluation department. Today, it is a learning team. You can see the change in The Center for Evaluation Innovation’s 2023 benchmark report: “Learning” appeared in nearly two-thirds of surveyed evaluation staff titles, up from one in five in 2015, and overtook “evaluation” for the first time.
This is a good thing. As a corrective for very real failings of funder-led strategic philanthropy, this movement “from evaluation to learning” represents what Hilary Pearson describes as “a mindset shift for philanthropy’s language and practice.” Foundations have too often treated themselves as detached experts, imposed burdensome evaluation requirements, and measured what is easy to count rather than what matters. The learning turn rightly calls for greater humility about what funders can know and control, a more pragmatic focus on improving impact, and a deeper commitment to relationships with grantees and communities.
However, the learning turn also risks obscuring an obligation that funders cannot escape: the obligation to exercise judgment.
The Inescapability of Judgment
The learning turn changes what and how funders know, but it does not determine what they should do. And funders must do, fundamentally: they cannot escape deciding whose claims are credible, how to weigh different evidence, which values should inform a grant, and whether one use of resources is more compelling than another.
The learning turn does not deny that responsibility, of course. Its leading accounts describe learning as a basis for decisions and action, and some explicitly connect it to strategy development and implementation. Yet the field’s practical guidance tends to be strongest after a funder has already selected a broad direction: learning from work in progress, testing assumptions, and adapting future grantmaking strategy. That guidance generally has much less to say about how to compare competing needs, approaches, and funding opportunities before resources are committed. And where the guidance does address that earlier stage, it often emphasizes whose knowledge should shape the grant decision. Advocates call for “privileging community and practitioner voices,” demand that “philanthropy needs to trust the real experts—the people it supports,” and assert that “the people in the community know what they need.”
Such appeals respond to undeniable problems of funder overreach but leave important questions unanswered: What should funders do when stakeholders disagree, when their convictions conflict with other evidence, or when several legitimate needs compete for limited resources?
The reality of disagreement over how to address issues within communities illustrates the difficulty of turning learning into action. In his 1957 classic Community Conflict, James Coleman showed that local decisions, from school policy to public health, routinely split communities. More recently, Jeremy Levine’s ethnography of participatory planning in Boston found that “the community” had no fixed and consistent meaning. Residents and nonprofit leaders invoked the term to lend authority to their claims, even as redevelopment officials sometimes used it to justify decisions that those same residents opposed. These examples demonstrate how communities are not of one mind.
The veneration of practitioner knowledge encounters the same fundamental problem. Educators disagree over reading instruction, homelessness providers debate housing-first and treatment-first models, and addiction professionals differ in their acceptance of medication for opioid use disorder. Moreover, frontline experience cannot in and of itself establish whether a particular intervention deserves support. Efforts such as D.A.R.E., Scared Straight, and abstinence-only education show how practitioner conviction can endure even after rigorous evaluations cast doubt on outcomes.
These challenges do not mean that funders should not engage communities and practitioners, of course; their input can surface overlooked views and clarify stakes. But including these voices still leaves the hard work of deciding how to interpret conflicting views and which claims should steer the funder’s choices. The learning turn has offered little guidance on how to weigh what funders hear, especially when what they hear does not furnish a coherent strategic mandate.
That imbalance raises a practical question: How well has philanthropy equipped funders to make such judgments?
The Unfinished Work of Choosing Well
Strategic philanthropy, whatever its excesses, made important progress by encouraging funders to define their goals, articulate how their resources should advance them, and assess their results. Yet practice has often lagged rhetoric. In a 2009 report, the Center for Effective Philanthropy surveyed leaders of large private foundations, finding that most had board-approved strategic plans, yet fewer than half reported having a logic model connecting their resources to their goals, and just eight percent could describe the specific data behind their belief that they would achieve at least some of those goals. A 2018 study found the gap persisting: 65 percent of foundation CEOs said they understood what was working in their programs, yet 42 percent said their foundations were not investing enough in developing that understanding.
Put simply: by the time the learning turn arrived, the discipline of choosing well was still half-built. It has remained so.
The gap extends from high-level strategy to the selection of individual grantees. Strategic philanthropy has imported the language of due diligence from the investment world, but the practice typically centers on the prospective grantee’s leadership, governance, finances, and fit with funder priorities. A 2023 Bridgespan review of 26 diligence resources found a “wealth of tactical tools” for vetting organizations, but most presumed the funder had already chosen its areas of focus and intended impact. “Serious due diligence is one of the biggest missing ingredients in philanthropy today,” Herb Sandler observed in a 2015 article whose authors noted that few donors devote extensive time to researching social-change opportunities before major commitments.
If anything, the learning turn may be eroding the partial discipline that strategic philanthropy built. In the National Center for Family Philanthropy (NCFP) Trends 2025 survey, the share of family foundations that guide their giving with strategies for creating change, including theories of change and strategic plans, fell from 50 percent in 2020 to 31 percent in 2025, and the share guided by measurable goals dropped from 41 percent to 19 percent. Family foundations are also less likely to ask grantees to set and measure program goals than at any point in the past decade (21 percent, down from 34 percent in 2015), yet a third now describe themselves as active learning institutions, a 14-point jump in five years. That learning is gaining institutional ground is an undoubtedly positive development, but it is less encouraging that the practices for articulating intended change are ceding ground at the same time.
Nevertheless, working models of a more complete discipline do exist in different corners of the philanthropic sector. For example, GiveWell compares funding opportunities against a common cost-effectiveness benchmark, Robin Hood estimates the benefit-cost ratio of each grant in its anti-poverty portfolio, and Co-Impact publishes a handbook spelling out the evidence and the account of systems-level change it expects from candidates. While these approaches reflect particular missions and cannot simply be transplanted to other organizations, they demonstrate how different funders can make causal assumptions, expected outcomes, uncertainty, and tradeoffs explicit.
The unfinished work, then, is making impact diligence a norm across philanthropy. While conventional due diligence examines the prospective grantee, impact diligence examines the case for the grant itself. It focuses on whether the proposed work is likely to produce meaningful change, and whether it is a better use of limited resources than available alternatives. Such analysis should be rigorous enough to improve decision-making, flexible enough to accommodate different forms of social change, and proportionate to the stakes of each grant. To build that norm, philanthropy needs to move rigor upstream.
Moving Rigor Upstream
The learning turn has clarified what funders should do less of after making a grant: restrict expenditures unnecessarily, impose duplicative reporting, prescribe implementation, and demand fidelity to outdated plans. The remaining question is what funders should do before making a grant if they are to exercise less control afterward.
Moving rigor upstream means placing more of the funder’s analytical discipline on the deliberations that precede a grant, while the allocation decision is pending. In practice, moving rigor upstream has four implications for how funders approach grant decisions.
1. A consequential grant should rest on an explicit case detailing why and how the proposed work should produce meaningful change and why it would be a compelling use of limited resources given plausible alternatives. Making that case requires clarifying the causal pathway from funding to impact, reviewing the evidence base for anticipated outcomes, considering what would happen without the intervention or without the funder’s support, assessing costs, and identifying key uncertainties and potential tradeoffs.
2. The funder should shoulder the analytical burden. Trust-based philanthropy, a movement related to the learning turn and focused on shifting power toward grantees, already urges grantmakers to “do the homework” so that early vetting burdens funders rather than applicants. Much of that homework, however, concerns the organization: its leadership, finances, governance, and alignment with the funder’s priorities. As noted previously, impact diligence extends this scrutiny to the grant’s expected impact. A foundation considering several organizations working on the same problem, for example, could conduct its own review of the relevant research and use that evidence across its decisions rather than asking each applicant to make the evidence case from scratch. Grantees could then focus on what they uniquely know about local context and operational realities.
3. The written product of diligence should record the funder’s judgment without prescribing how the grantee must subsequently act. For a significant decision, a short internal memo might explain what the funder expects to change, why it believes the proposed approach can produce that change, what evidence supports the case, what alternatives it considered, and where uncertainty remains. Such a memo creates a record of the funder’s reasons for approving the grant without turning the funder’s initial assumptions into implementation requirements for the grantee. Once the grant is underway, the funder can revisit that record as new information emerges, enabling the funder to understand which assumptions held up, which did not, and what needs to change in future decisions.
4. The depth and form of diligence should match the nature of the grant decision. A small renewal to a long-trusted organization should require less analysis than a large commitment to a new program in an unfamiliar field. An evidence-rich direct-service intervention may lend itself to estimating costs and expected outcomes, while an advocacy or systems-change grant may require scenario analysis, contribution reasoning, and more explicit treatments of uncertainty.
Moving rigor upstream asks funders to take greater responsibility for the quality of their choices so that they can exercise less unnecessary control afterward. If the learning turn has taught philanthropy to listen more humbly, the next task is to choose more rigorously. Learning can enrich decisions, challenge them, and sometimes overturn them. But learning is not deciding.
Read more stories by Maoz (Michael) Brown.