Kelman: The quest for objective research

I recently attended a conference in Vaxholm, Sweden, a beautiful island in the Stockholm archipelago. The theme was academic research on privatization. About 20 scholars — a few Americans, but mostly Swedes and other Europeans — all economists (except me), presented papers, and many of them were about contracting. The conference was sponsored by the Research Institute of Industrial Economics, an industry-supported group, so I was a little concerned before coming that the papers would be driven by a contract-out-everything agenda. I was pleasantly surprised. The research was of high academic quality, the empirical papers — those analyzing actual data — were very impressive. The theoretical ones — filled with math formulas — were interesting, though sometimes abstract and somewhat unrealistic in their assumptions.What impressed me most was the nonideological tone of the research. The papers proceeded neither from the assumption that contracting out is either always good nor that it is always bad. Instead, the empirical papers tried to look dispassionately at the data, and the theoretical papers presented both upsides and downsides of outsourcing.My favorite paper was by a young Swedish economist named Erik Lindqvist who examined the effects of the choice to contract out treatment services for at-risk teenagers or keep the services in-house. The theoretical predictions the paper tested were that private for-profit providers would have an incentive to increase the length of treatment (because the contracts were time-and-materials based, and therefore providers made more money the longer the treatment lasted), while public providers would have an incentive to get rid of hard-to-treat children because they created trouble but provided no “profit” to the government. The data backed both predictions. The average treatment time by for-profit providers was far longer — almost twice as long, controlling for other factors — than by public providers. But public providers were more likely to stop treating troublesome kids.Other papers also went where the data led. A careful analysis of the effects of Sweden’s radical school choice/voucher system — which has led to the establishment of a large for-profit education sector — showed that, controlling for other factors, pupil achievement was slightly better on some dimensions and not better on others because of the greater competition for-profit schools provided in a school district.The results, though, were far less dramatic than earlier studies on the same topic, which showed larger impacts but were based on fewer years of data. Given the importance of contracting as a way governments deliver services, the quantity and quality of academic research on the subject are disappointingly weak. This conference shows the way scholars should be going to investigate this subject, on behalf of those who need to make decisions on the proper scope for and management of government contracts. ()