the really hard problems

Harvard’s Division of Social Science hosted a symposium of “big thinkers” in social science, asking the scholars to reflect on the hardest problems of social science. The thinkers included people like Ann Swidler, Roland Fryer, and Nassim Taleb. The website contains video links to their remarks and a poll asking people to rank just how hard and important these problems really are.

Looking through the list of problems, there are a number of problems that are unarguably hard and important (e.g., Swidler’s question: How do societies create or re-build effective, powerful, and resilient institutions (for instance, governments)?). Some of the questions are hard but maybe not as important (e.g., Fowler’s question: What causes clustering in social networks; for example, how much is it the effect of shared environment (individuals in the cluster are connected by some shared context that isn’t obvious); homophily (the tendency of similar people to like and associate with each other); or — most interesting — influence?). Other problems seem more like general speculative questions but not fundamental problems (e.g., Carey’s question: How do we understand the human capacity to create and articulate knowledge?).  And others are straightforward research questions (e.g., King’s question: What is the relationship between democratization and international conflict?).

The symposium’s Facebook page lets readers offer their own problems. Some of my colleagues at Kellogg suggested a few additions to the list. Does organizational theory have any hard problems of its own? Even if the problems are hard, are they that important? I’d love to hear your thoughts.

Written by brayden king

April 24, 2010 at 5:19 pm

6 Responses

Subscribe to comments with RSS.

  1. So, here are some matters that I think are both hard and important in org theory (warning: my very biased angle):

    1. the micro-macro, aggregation/transformation problem: this one’s been around for a while, but I think there are still lots of seeming “emergent” effects and specifying their forms and nature etc remains a big (and hard) question.

    2. comparative orgs questions: I still think that much of our language and intuition is ‘tainted’ by (thanks to, among others, institutional theory) assumptions of similarity across organizations and even different types of collectives. I think we need better comparative theories, across different contexts, different types, and lifestages of organization. (We’ve talked about these categories before — but I’ll plug them once more.) I don’t know that this necessarily is a “hard” question, but one that is important and deserves attention.

    3. orgs-markets nexus. Quite a few questions still exist at the org-markets nexus: where do markets versus orgs begin? how do we, analytically, work in network forms into a theory of the markets-orgs nexus? What’s the relationship between new markets and new organizations? And, other small issues like reconciling Granovetter and Williamson, etc.

    4. the organizational actor. How do we conceptualize the organization itself, as an actor? This has big implications for strategy, identity, org-individual relations, agency-environment/structure debates, etc, etc.

    5. performativity stuff — yes, that’s right. I still think the limits (and possibilities?) of performativity deserve attention. Don’t know that this is necessarily an ‘organization’ theory question specifically — but an important question nonetheless.

    6. emergent forms of organization — at times our theories lag behind the phenomena, there are various, quite cool forms of organization that deserve additional theoretical attention: wisdom of crowds-type intuition, wikipedia/open source-etc type communal production, entrepreneurship and nascent collective action.

    7. org design — it seemingly died for a while, now there is increased attention — how do we design social interaction, efficient forms of organization? Organizationally, how do we ‘unleash’ (is that fuzzy enough?) human potential?

    8. human nature, orgs and markets: reciprocity, self-interest, norms, hierarchy, emergence?

    9. org theory versus social science. Sometimes our field is seen as exceedingly applied (we borrow theories from “true” disciplines), which bugs me — so, the hard question is to come up with additional, uniquely organizational, hard questions.

    That’s my (very) quick, biased two cents.



    April 24, 2010 at 7:29 pm

  2. Gary King’s problem was not the one you state. It was “Post-Treatment Bias in Big Social Science Questions” and he made the point that this inability to do anything about the post-treatment bias (if it exists) prevents a lot of important “hard questions” from being solved, the one you quote being one of them.


    Abhishek Nagaraj

    April 24, 2010 at 8:55 pm

  3. Abhishek,

    I just copied the problem from the poll, which listed it as King’s. Multiple problems on the poll belonged to King. fyi.



    April 24, 2010 at 9:31 pm

  4. Based on his paper here, King proposes that those other questions are instances of questions that are hard to answer because of post-treatment bias. You’re both right!



    April 24, 2010 at 10:22 pm

  5. Re King: I’m with Abhishek here, sort of. King was not saying that that particular question should command great attention from thousands of industrious scholars, but the general problem of how to bound the consequences of post-treatment bias deserves deeper consideration than big thinkers are really giving it. So … I would read it as a critique of the whole conference … “You all think you are going to get at these big questions by thinking deeply and writing eloquently, but you are probably going to go nowhere until you really get serious about formal models of causal processes.” Anyway, that’s how I read it, and I happen to agree with King.



    April 26, 2010 at 1:55 pm

  6. Am I reading those slides correctly in that “post-treatment bias” is basically a result of feedback effects? I was a bit too young to be there for the release of DSI, but when I did read it a few years ago I found it somewhat amusing how at a time when people like Douglass North were busy becoming Nobel Laureates, King and his cohorts maintained that the best way to deal with feedback effects is to not study cases were we suspect that it is present. Clearly, if a problem can’t be solved with a regression analysis, then it probably wasn’t worth solving in the first place!



    April 26, 2010 at 2:56 pm

Comments are closed.

%d bloggers like this: