To me, slogans like "making space for human agency" and "recognising women’s agency" are attempts to convey some useful messages from social science to the world of development practice. But they do that job badly.
By peppering our prose with "agency", we may make our writing seem more profound. However, this is spurious – an academic affectation, with questionable intellectual credentials, which gets in the way of clear, evidence-based discussion of an important set of real-world issues.
People do not have less agency when they have fewer choices
There was a time when social scientists genuinely needed to make an issue of agency.
That was when sociology and political science were significantly influenced by types of functionalist or structuralist theory that purported to be able to explain human behaviour without any reference to people’s intentions, motives or reasons.
That so-called structure/agency problem was put to bed definitively by Anthony Giddens and others at least thirty years ago. The then mainstream sociological and Marxist functionalisms have now all but disappeared. They have been replaced by a range of dominant perspectives – mostly 'new institutionalisms' of various kinds – that, whatever their other failings, do not have this particular weakness.
A core point in Giddens’ philosophical resolution of the structure/agency problem was that human beings have 'agency' simply by virtue of being human. This is not in conflict with the observation that human behaviour is typically quite structured. It is only in conflict with the notion that structures explain what people do in spite of (rather than by way of) their intentions, motives or reasons – for example, because the social order or the capitalist system has both ‘needs’ and the ability to secure them.
The Giddens position is consistent with ordinary language: agency is the quality that agents have. People do not have less agency when they have few choices (e.g. because they are poor, oppressed or enslaved); they just have fewer choices. Great leaders don’t have more agency than average leaders; they just use their agency differently.
Academic jargon infecting development practice
The notion that agency is something people have more or less of according to circumstances is contrary to both the Giddens solution and ordinary language.
Yet it has become the custom to use the word in exactly that way. In much academic and even practice-oriented writing, agency has become an empirical variable, something that varies inversely with the extent of structural constraint on behaviour – and something researchers and practitioners can appreciate or fail to notice depending on where and how they look.
I am aware that there is a back-story to the academic trend. This is about the way Giddens’ influence was blunted by a stream of less lucid sociological writing while the ink was still fresh on the page. I think it’s a pity that Giddens did almost nothing to defend his corner against this type of critique, then or since, unwisely counting on the sheer power of his argument. Someday, perhaps, someone will have the dedication to tell this story as it should be told. In the meantime, I submit, we should let ordinary language settle the matter.
Unfortunately, the nonsensical, either-or, language of structure and agency is now entrenched in many of the books and journal articles new generations of social scientists are made to read. As advisers to social development and governance programmes, young graduates have carried this into the world of practice.
As a result, the philosophical allure of ‘recognising agency’ is often hard to resist, pushing aside simpler, cleaner and more evidence-based thinking about development options.
Doing development differently
Why is this worth saying now? Because there are some currently rather exciting messages that social science needs to convey to development practice that will get blunted if they get mixed up with ‘agency’.
Often, appeals to "bring back agency" are substantively about recognising the historical roles of visionary, rebellious or entrepreneurial leaders. Or they are about the way the oppressed sometimes decide to exercise the little power they have in courageous, innovative or otherwise unusual ways.
Another valid concern sometimes conveyed in these terms is about complexity and uncertainty in human affairs: because of uncertainty, people may have more choices than they habitually admit, meaning that there is more room for change, and a greater role for ideas, than predicted by prevailing explanations.
These issues about uncertainty – and the implied scope for politically smarter, more adaptive or more entrepreneurial development work – are really important. They are at the heart of the lively debate now going on about doing development differently. The central ideas of this discussion very much need to be communicated to practical people, such as officials in development agencies who exaggerate the extent to which development or governance reforms have to be, and can be, planned, blueprint style.
We need better evidence on the circumstances in which innovative choices can be made by different sorts of individuals and organisations. We need to be capable of explaining clearly how smarter forms of development practice can help discover promising pathways of change. It will be easier to rise to this challenge if we are not encumbered with ‘agency’.
David Booth is a senior research fellow in the Politics and Governance Programme at the Overseas Development Institute (ODI), UK.