Partly Cloudy, with a 30% Chance of Riots

Gary Gutting

In this morning’s New York Times, distinguished philosopher Gary Gutting raises a commonly discussed set of questions about social science research and theory and the ability to make accurate predictions. As is common in such arguments, physics is held up as the best “real” science because it gives us theories with clear predictions that always work. Experimental evidence has allowed for the clear elaboration of (what appear to be) invariant physical laws, at least in areas like classical mechanics (e.g. we can predict precisely where the moon will be 200 years, 3 days, and 7 hours from now). The social sciences are then held in contrast to this. As the argument goes, the social world is amazingly complex, making it hard to generate predictions, and it is often too difficult or too immoral to do controlled experiments on people, so social scientists could never test those predictions anyway. Gutting’s conclusion: “we need to develop a much better sense of the severely limited reliability of social scientific results.” And the implication he draws from this conclusion?

“Given the limited predictive success and the lack of consensus in social sciences, their conclusions can seldom be primary guides to setting policy.  At best, they can supplement the general knowledge, practical experience, good sense and critical intelligence that we can only hope our political leaders will have.”

I have to say, I accept that scientific findings (of all kinds) are limited, that predictions (of any kind) only work within specific scope conditions, and that breathless media accounts of new research findings often overstate claims. That said, I disagree with much of the argument Gutting makes in the piece. I think he oversells the ability of natural sciences to make perfect predictions, especially once we start moving away from Newtonian physics (consider how many times you use your umbrella, even when the field of meteorology claims a 0% chance of rain). I also think he undersells the usefulness of accumulated evidence and carefully tested theory in the social sciences for advancing knowledge of how the world works and for designing public policy to shape that world.

But whether or not Gutting is right, people are looking to social scientists for predictions so they can make informed choices–especially when it comes to collective behavior and social movements. For example, the Chicago police would love to be able to predict with great accuracy when and where acts of collective violence will occur during protests associated with the upcoming NATO meetings; and protestors would love to know when and how the police will act (and react) over the course of the events (here’s an NPR story about the already developing interactions between protestors and police in Chicago). And I’m sure lots of people would have liked to see clear predictions about the Arab Spring movements. And the Occupy movement. And the Tea Party movement. And so on.

Protestors in Chicago

(Nancy Stone, Chicago Tribune / May 11, 2012)

So that leaves me wondering. Can we ever hope to develop social movements research and theory to the point where we could make such predictions? And just as importantly, is that even a goal worth pursuing? To think about it more concretely: based on what you know from social science research and theory, what will happen in Chicago this weekend? How much confidence do you have in those predictions? And do you think it would be worthwhile to use those predictions to inform the “policies” adopted by the police, the protestors, or anyone else in the Windy City this weekend?

4 Comments

Filed under Daily Disruption

4 responses to “Partly Cloudy, with a 30% Chance of Riots

  1. willhmoore

    Though Gutting does not engage it, The Economist recently published a relevant piece , and though it is not specifically mentioned in the piece, Strategic Analysis Enterprises is one of the consulting firms involved in the ICEWS project. Back in 2008 I co-authored a piece with Steve Shellman that offers a different view from Gutting .

  2. willhmoore

    Chris Blattman has a relevant post from early May:
    http://chrisblattman.com/2012/05/05/predicting-conflict/

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s