In his latest post on the Nesta blog, Geoff Mulgan describes why he's "skeptical" about how the idea of a theory of change is often presented, despite being "wholly supportive of the underlying notion." So while Mulgan highlights the importance of an organization having a "coherent account of why the things it does might have the effects it wants," he argues that the 'theory of change' framing can be misleading in two central ways:
"One is that they tend to be far too linear, assuming that inputs lead to outputs, and that outputs lead to outcomes. This sometimes happens. But anyone familiar with systems thinking will be dubious of linear explanations, especially where complex social phenomena like homelessness, poverty or isolation are concerned. I used to advocate mapping systems to see all the links between different causes, and then to be precise about the strength of knowledge about each of those links, and about who had the power to influence them. These types of system maps provide a very useful starting point for then thinking about how a particular organisation or intervention can work within a system, but there isn’t anything quite comparable in the theory of change literature.
The second related problem is that theories of change risk squeezing out the space for learning. The advantage of the kind of systems map I just described is that they are explicit about the limits of knowledge, and where, through action, you might want to generate better knowledge about what works and what affects what. Our own use of the phrase makes this explicit as we encourage organisations to move up the ladder of evidence, generating additional knowledge about what works and why."