I'm not going to defend Lant Pritchett. But there is a serious issue here. In Ghana--where 60% of the urban population has no access to the sewer system, where 70% of energy still comes from burning wood and charcoal (and where rain acidity as a result at times reaches Black Forest levels), where 40% of people drink contaminated water, and where 15% of people suffer from waterborne diseases--should taxicabs have to have catalytic converters installed?
I finally got around to reading Andrei Shleifer's rant against the public sector in the Fall 98 J of Econ Perspectives. For those of you who haven't looked at it (and I can understand why), Shleifer argues that contract theory (post Williamson and Hart) supports the idea that almost any good or service can be provided more efficiently by the private sector. He even concludes with a brief, superficial plug for educational vouchers. He does not appear to recognize that his entire approach begs the question by assuming that all workers, whether publicly or privately employed, are opportunistic and self-interested, and that only competition and/or incentive-compatible contracts can light a fire under them. This is equivalent to denying the existence of public service motives. Clearly, without an ethos of public service, government becomes stultifying and tyrannical. A rational research agenda would look for the circumstances under which public service flourishes or withers.
Well, as the co-editor of the JEP who agreed with associate editor Oliver Hart's decision that Andrei Shleifer would be an interesting person to write about the public sector (and as an ex-college-roommate of Andrei Shleifer's as well), let me say three things:
First, Andrei's major goal--at least as I understood the paper that I edited, and if it didn't come through in the published version then I messed up--was to try to break the link between the case for social democracy (in the sense of substantial redistribution of income and wealth and substantial government funding of education, infrastructure development and so forth) and the case for the public provision (in the sense of the government itself employing a lot of people and managing a lot of production). As Andrei put it to me the first time we talked about the paper, we don't have the government try to raise nutritional standards for the poor by hiring people to grow wheat on state farms which is then processed in state mills, baked in state bakeries, and then distributed through state warehouses--instead we have food stamps. And he (and I) suspect that food stamps are more effective than the state-run agricultural and nutritional center will be.