25 January 2011

Player-hating: Some clarifications on institutions

My fiercest critic [I’m sorry for the typos! I’m sorry! Please stop shouting at me!] just harangued me for my lazy and inadequate response to these comments on institutions.

So allow me to make amends.

Anonymous saidknowing that (persistent) social stuff matters isn't at all the same as knowing that institutions matter, and even if we define institutions to mean "persistent social stuff" that still doesn't say what needs fixing.”

Firstly, the terms we use are vague and confusing. What is an institution? What is governance? I like Paul Romer’s use of “rules.” It is easy to think concretely about what rules are, and why they might matter. Even if you think his prescriptions are a little crazy, he does a very good job of explaining the importance of institutions (rules), and you should really watch the TED talk if you haven’t already.

Secondly, you are absolutely right, we don’t know what needs fixing. Which is exactly my point – we know what good institutions are, but we don’t know how to make them.

Here is Dani Rodrik, discussing his canonical empirical paper Institutions Rule in the Finance & Development magazine:

Institutions are thus critical to the development process. But for each of the functions performed by institutions, there is an array of choices about their specific form. What type of legal regime should a country adopt—common law, civil law, or some hybrid? What is the right balance between competition and regulation in overcoming some of the standard market failures? What is the appropriate size of the public sector? How much discretion and how much flexibility should there be in arrangements for the conduct of fiscal, monetary, and exchange rate policies?

Unfortunately, economic analysis provides surprisingly little guidance in answering these questions.

Ed Carr questioned the power of institutions as an explanatory variable, “this is a relative measure - institutions are better than things like geography for predicting outcomes, but they still leave a hell of a lot of variance out there.”

This chart is from the same Rodrik Finance & Development article. There is a fair amount of variation, but in the world of real-world data, that is about as tight a correlation as you are ever going to find.


Finally Bottomupthinking askedMaybe the institutions operate so badly simply because they are not really understood by most of those who live in or with them (the wider populace)?”, which leads me to believe that we are again confusing terms? In any case, I think we agree on the difficulty of improving rules, governance, institutions, whatever you want to call it.

Lant Pritchett is doing some very interesting work on the related-but-not-quite-the-same question of state implementation capacity.

At their current pace of progress countries like Haiti or Afghanistan or Liberia would take hundreds (if not thousands) of years to reach the capability of a country like Singapore and decades to reach even a moderate capability country like India.

Pritchett is also one of the most vocal public advocates for migration as a development policy. It perhaps is no coincidence that he has also spent so much time studying development failure and “state capability traps.”


Ryan said...

My comments (mostly just a link to a working paper, sorry): http://ryancbriggs.net/post/2920347435/institutions-rule

My basic issue is that we don't have a coherent definition of institutions and usually the way that institutions are operationalized doesn't follow from the theory. Institutions as rules are basically just really hard to measure. My second point is that what most of the this literature taught us is that usually social stuff matters more (ceteris paribus) than geography. That is a very important finding, but I've yet to see a solid way of teasing out institutions from other examples of persistent social stuff.

The best example of the dangers of loose operationalizations of institutions is Glaeser et al's (2004) article "Do Institutions Cause Growth." They show that setter mortality can be used to instrument for human capital instead of institutions and will produce essentially the same findings as AJR and RST. The Glaeser paper also does a great job critiquing the concept of institutions.

Roving Bandit said...

Good points, I basically agree with you entirely. Perhaps I should have framed the initial discussion in terms of good governance (which is still pretty vaguely define, but perhaps slightly more concrete?).

Ryan said...

I agree that good governance is vague, but I think it is also thought to consist of more than rules, which in this case is a good thing. The more I think about this, the more that I think we should ditch most of these umbrella concepts. They give us an often misleading sense of generalizability, and we should probably just say exactly what we mean and see if we can build up from that. While I'm dreaming, I also want a pony.

On what is now a side note (and was the original point of your post), I agree completely with the pro-migration arguments.

Ranil Dissanayake said...

One of the biggest problems with Romer's ideas is that he thinks we know what rules work and which ones don't (and I think from this post you agree with him?)

The problem is anyone with an even basic knowledge of legal history and comparative law will be able to tell you that while it might be obvious ex post, it's rarely the case ex ante. We spend a very long time refining the content and application of rules through amendment, precedent etc., and as a result, what rules exist in different places look very different, both across space and within one place across time.

Rules and institutions (formal and informal) matter, but we assume we know much more about them than we really do.

I also do not share Pritchett's pessimism about improving institutions. Hong Kong built a civil service from scratch in less than 50 years. It's probably the most efficient in the world now. India has a way better set of institutions now (in most states, don't forget it's federal) than it did twenty years ago.

Karuna Herrmann said...

Thank you. I will try to shout less (it WAS a lazy response though).

HowMatters said...

Readers may be interested in the issue of ODI's Development Policy Review out this month, Aid, Institutions and Governance - What Have We Learned

Anonymous said...

If some one needs to be updated with most recent technologies then he must be
go to see this web page and be up to date everyday.

Feel free to surf to my page: www.pitch-putt.com

Post a Comment