Predictive social science

If you do an introductory sociology course* you'll at some point bump into the argument about social sciences not being sciences because they can't predict things will happen. This also comes up quite a lot in the evaluation literature, which I occasionally dive into like Tom Daley, as evaluations tend to be bloody useless at predicting what will happen if a policy intervention is rolled-out to a wider population c.f. the endless bloody debates about double-blind randomised controlled trials.

However, I just saw this tweet;

Fewer plans in poor areas: neighbourhood plans by deprivation decile. Reposting this after being at #PPPconf2015 pic.twitter.com/ODWzDx4cX0

— Alasdair Rae (@undertheraedar) September 10, 2015

And could proudly say, "we predicted that" - in this paper, I blogged about here. The paper came about from an ESRC seminar series and just some thoughts I had about our previous work on middle-class activism and how this would interact with the new neighbourhood planning system in England.

I don't have anything profound to say about the philosophy of social science and my ability to predict the future. But it does seem to me that social science can predict some things, the question is what things, and what are you predicting. We could never say exactly how many more neighbourhood plans there would be in affluent areas compared to deprived areas, but smashing together some theory, some findings from an evidence review and some statistical analysis of the British Social Attitudes Survey, we could roughly say "you'll probably get more neighbourhood plans in these sorts of places because of these reasons..." 

I should also add, I think this does say a lot for the realist school of evaluation from Ray Pawson's work on realist synthesis. It forces you to come up with predictive, useful, rules of thumb for whats works, in what contexts and why. This means we could easily apply the findings of our earlier study to the case of neighbourhood planning. 

In the meantime, PayPal me £20 and I'll predict your lottery numbers for you.

*I wistfully recall my A Level Sociology course and one of my friends proudly stating "a dog could walk off the street and get an A in A Level Sociology" to which another friend replied with "I got a B".