Where do Theories of Change come From, and to Whom do They Belong?

In his most recent contribution to our dialogue about theories of change, Sean Stannard-Stockton says that "nonprofits should have sound business plans that address both operations as well as a Theory of Change for why their programs will affect social impact" and that "the designing of programs and researching of validity should be the domain of nonprofits," rather than of foundations.

A nonprofit organization must definitely craft its own business plan. But the fact is that neither nonprofits nor foundations have the capacity to design and test the theories of change that underlie most important social interventions. Research of this sort is typically beyond both their missions and budgets. Rather, it is carried by social scientists at universities, think tanks, or independent research organizations, and funded mainly by governments and occasionally (and inadequately) by foundations.

My last post cited the What Works Clearinghouse, which describes empirically-based theories of change in education. If you take a look at the studies summarized on its website, you'll get a sense of the special expertise and extraordinary costs involved. I also mentioned the studies that found that abstinence-only programs did not reduce teen pregnancy. These were funded through a congressional appropriation and carried out by Mathematica Policy Research, Inc. By the same token, MIT's Poverty Action Lab tests theories related to international development interventions in education, health, microcredit, and other areas.

Although nonprofits need to know what works, research of this sort is beyond their scope. Think about most direct services organizations that you know -- schools, teen pregnancy prevention centers, and the like -- and you'll realize that they are set up to implement theories of change, not to test them. To be sure, the ideas for new approaches may originate with social entrepreneurs as well as social scientists. But organizations often achieve social impact by replicating what works rather than by coming up with untested innovations.

This point was emphasized in a recent guest post by Jason Saul on change.org. "Most nonprofits are not in the business of proving theory; they are in the business of improving outcomes. For example, social science has proven that students who are more interested in school have higher attendance. A nonprofit doesn't need to re-prove that theory: it just needs to implement it effectively and measure the increase in student interest"

But what about a social entrepreneur who, like an idiot savant who can do complicated square roots in his head without explaining how, can design robust theories of change? Designing without testing is akin to Glendower's claim in Henry IV that he can "call spirits from the vasty deep." Hotspur replies: "Why, so can I, or so can any man. But will they come ...?"

Of course, many nonprofit leaders have strong intuitions born of their on-the-ground experience. The trouble with leaving evaluation to intuition, however, is that intuitions about social change are more often wrong than right. This is nicely captured in the New York Times obituary of Joan McCord, a criminologist "who marshaled mountains of evidence to question the effectiveness of social programs championed by both liberals and conservatives."

Who needs to know whether a theory of social change is valid and strong? Whoever is supporting interventions based on it -- the operating nonprofit planning to implement it, the funder planning to pay for implementation, or government in its role as service provider or funder.

In the end, like most knowledge, a strong empirically-based theory of change -- or good evidence that a theory of change doesn't work -- is a public good, equally useful for a nonprofit and its funders, but with no one having a special claim of ownership.