Essays in empirical macroeconomics

Abstract

This thesis covers three topics in empirical macroeconomics. The first chapter, joint with David Dorn and Philipp Kircher, investigates the effect of national industry shocks on local employment. By providing a novel structural view on the Bartik framework, we show that the difference in national and regional employment growth trends can be attributed to within-region spillovers. We hypothesize that these spillovers are due to agglomeration economies and show that they can be quantified in a simple regression of regional employment change predictions versus actual regional employment changes. We find that the main driver of agglomeration effects are regions with particularly strong growth in employment which outperform their predictions, suggesting that the efficiency-equity trade-off in the distribution of regional growth lies between strong performers and the rest. The second chapter is an empirical study of the effect of increasing import competition from China in the US on regional air pollution. I find that increased exposure to Chinese import competition had significant effects on direct emissions and air quality in the US between 2000 and 2005. These results stand in contrast to current literature attributing the decline in US emissions almost exclusively to environmental policy changes. While I also find strong effects from the main environmental policy tool, its effects are not substantially larger than the import competition effect. The third chapter investigates pension bias and uncertainty in Germany. Using a representative panel I find that 61% of individuals approximate their pension claims, overestimating them by 18%-28% percent. A standard household finance model which includes a German style pension system and differentiates between rational agents and agents that are biased with regards to their pension claims is presented. After calibration I find that aggregate consumption of rational agents is 0.25% lower than for biased agents. This result is explained by much higher early life consumption of biased agents coupled with lower survival rates of all agents as they age. Specifically, biased agents may die before their sub-optimal savings behavior has repercussions.