Monday, October 27, 2008

Conference at Nuremberg

On Friday and Saturday I was at a conference at the IAB (the research arm of the German Ministry of Labor - sort of like the U.S. Bureau of Labor Statistics). The program and the papers are here.

I particularly liked the paper presented by Margaretha Buurman of Erasmus University in Rotterdam. It formalizes a question closely related to one that Jim Heckman, Chris Taber and I wrote about informally in our paper on caseworker behavior at the Corpus Christi site of the Job Training Partnership Act (JTPA). That question has to do with the interaction between who you hire as a caseworker, the wage you have to pay, and what you can do in terms of selecting among applicants. One way to think about what outcome based performance standards, like those in JTPA its successor the Workforce Investment Act (WIA) do is that they push caseworkers away from doing what they would otherwise do, which is devoting resources to the hardest-to-serve among the applicants. This is probably efficient but the point here is that it reduces the non-pecuniary remuneration of altruistic caseworkers, which means that they will require higher pay. Of course, an alternative is to hire non-altruistic caseworkers devoted solely to efficiency.

The Buurman and Dur paper addresses the related question of the tradeoff between hiring altruistic caseworkers, who will work for less, and implementing sanctions on welfare recipients who do not take available jobs. The key is that altruistic workers do not like to sanction, so the local agency faces a tradeoff.

I think this type of public management question is understudied by economists both in terms of theory and in terms of empirical evidence.

The paper by my friends and Chicago graduate students colleagues (and fellow laborers in the Heckman vineyards) Gerald Marschke and Pascal Courty is also quite interesting. It shows that sites respond to changes in the payoffs to particular client characteristics implicit in the regression model that was used to adjust performance targets under JTPA.

The design and effects of performance management systems, a topic with broad implications in both the public and private sectors (it is a close cousin to the literature on performance based rewards for executives) is, in my view, similarly understudied.

The overview talks by Carolyn Heinrich (another Heckman student now at the LaFollette Institute at Wisconsin) and Burt Barnow are worth a look as well.