We conduct a large-scale field experiment in the German labor market to investigate how information provision affects job seekers' employment prospects and labor market outcomes. Individuals assigned to the treatment group of our experiment received a brochure that informed them about job search strategies and the consequences of unemployment, and motivated them to actively look for new employment. We study the causal impact of the brochure by comparing labor market outcomes of treated and untreated job seekers in administrative data containing comprehensive information on individuals' employment status and earnings. The effects of our treatment tend to be positive, but concentrated among job seekers who are at risk of being unemployed for an extended period of time. Specifically, treatment effects in our overall sample are moderately positive on average but mostly insignificant. At the same time, we do observe pronounced and statistically significant effects for individuals who exhibit an increased risk of long-term unemployment. For this group, the brochure increases employment and earnings in the year after the intervention by roughly 4%. Given the low cost of the intervention, our findings indicate that targeted information provision can be a highly effective policy tool in the labor market.
jobsearch.pdf
The Regression Kink (RK) design is an increasingly popular empirical method for estimating causal effects of policies, such as the effect of unemployment benefits on unemployment duration. Using simulation studies based on data from existing RK designs, we empirically document that the statistical significance of RK estimators based on conventional standard errors can be spurious. In the simulations, false positives arise as a consequence of non-linearities in the underlying relationship between the outcome and the assignment variable, confirming concerns about the misspecification bias of discontinuity estimators pointed out by Calonico et al. (2014). As a complement to standard RK inference, we propose that researchers construct a distribution of placebo estimates in regions with and without a policy kink and use this distribution to gauge statistical significance. Under the assumption that the location of the kink point is random, this permutation test has exact size in finite samples for testing a sharp null hypothesis of no effect of the policy on the outcome. We implement simulation studies based on existing RK applications that estimate the effect of unemployment benefits on unemployment duration and show that our permutation test as well as inference procedures proposed by Calonico et al. (2014) improve upon the size of standard approaches, while having sufficient power to detect an effect of unemployment benefits on unemployment duration.
How does tax complexity affect people’s reaction to tax changes? To answer this question, we conduct an experiment in which subjects work for a piece rate and face taxes. One treatment features a simple, the other a complex tax system. The payoff-maximizing output level and the incentives around this optimum are, however, identical across treatments. We introduce the same sequence of additional taxes in both treatments. Subjects in the complex treatment underreact to new taxes; some ignore new taxes entirely. The underreaction is stronger for subjects with lower cognitive ability. Contrary to predictions from models of rational inattention, subjects are equally likely to ignore large or small incentive changes.
