The 21st century global decline in productivity growth is not well understood. One possible contributor is a decline in economic dynamism. We explore the contribution of firm formation and employee movement to productivity using administrative data on the population of New Zealand construction...

Combining weekly productivity data with weekly productivity beliefs for a large sample of truckers over two years, we show that workers tend to systematically and persistently over-predict their productivity. If workers are overconfident about their own productivity at the current firm relative...

A large body of work has highlighted the importance of employment reallocation as a driver of aggregate productivity growth, but there is little direct evidence on the extent and nature of this process at the worker-firm level. We use an administrative matched employer-employee census for Chile...

Do recessions speed up or impede productivity-enhancing reallocation? To investigate this question, we use U.S. linked employer-employee data to examine how worker flows contribute to productivity growth over the business cycle. We find that in expansions high-productivity firms grow faster...

The 21st century global decline in productivity growth is not well understood. One possible contributor is a decline in economic dynamism. We explore the contribution of firm formation and employee movement to productivity using administrative data on the population of New Zealand construction...

Combining weekly productivity data with weekly productivity beliefs for a large sample of truckers over two years, we show that workers tend to systematically and persistently over-predict their productivity. If workers are overconfident about their own productivity at the current firm relative...

Churn, defined as replacing departing workers with new ones as workers move to more productive uses, is an important feature of labor dynamics. The majority of hiring and separation reflects churn rather than hiring for expansion or separation for contraction. Using the JOLTS data, we show that...