Abstract:
Research productivity is not constant over the lifetime of a researcher but fluctuates substantially and often seems to follow a typical pattern. But despite stable aggregate patterns there is substantial variation of research output across individuals. Our paper aims at explaining systematic differences in lifecycle productivity patterns with differences in career incentives. We develop a theoretical model in which lifecycle research productivity is driven by a combination of incentives to invest in skills and/or to produce output. Both incentives depend on career characteristics which are set by within national university systems. From our model we derive testable hypotheses on variations in individual research productivity profiles within and across countries. We test our implications based on a unique data set which we collected for 112 (business) economists in the US and 189 in Germany. We find that promotion tournaments in the US as well as in Germany provide very effective incentives. In general this leads to elevated publication outputs in time periods preceding a major promotion and to reduced publication productivity afterwards. But we also find striking differences between US and German researchers. Skill acquisition is more important for German researchers in the screening period since the first promotion decision is strongly influenced by a qualification requirement, the so-called Habilitation. Also, German researchers lack a second major career step in comparison to US researchers, for whom a promotion to full professor is almost as important as the promotion to associate professor. Re-appointments in the German university system offer comparatively low gains and are thereby not attractive enough to induce a significant increase in research output. Therefore, incentives and publication productivity are highest early in the career of German researchers levelling off on a lower but decent level afterwards. For US researchers the situation is different. Their research output is not only significantly higher prior to their first appointment but also prior to a promotion to full professor, indicating that this promotion provides a second effective incentive to foster research output over a longer period of time. However, after promotion to full professor research output also levels off at a significantly lower level.