Heterogeneity in Macroeconomics: Empirical Evidence

Prosjektet gav svar på flere økonomiske spørsmål: Framoverskuende pengepolitikk er ikke et like effektivt politisk virkemiddel som tidligere antatt så lenge heterogenitet er tatt høyde for. Finanspolitikk er ikke mer effektivt i en situasjon med resesjon eller likviditetsfelle sammenliknet med normale tider så lenge heterogenitet er tatt høyde for. En finanspolitisk stimuli kan være effektiv på kort sikt, men på mellomlang sikt vil velferdseffektene alltid være negative. Vi utviklet en metode som gjør oss i stand til å måle viktigheten av medarbeidernes ferdigheter for en arbeiders lønn. I kontrast til tidligere forskning som lider av metodiske svakheter fant vi at medarbeidernes produktivitet, og ikke bare arbeiderens egen produktivitet, betyr noe. Rollen av slike friksjoner og sortering av effektene som driver lønnsspredningen estimeres til å være relativt lave. Metoden vi utviklet gjennom prosjektperioden tillater oss å adressere mange klassiske spørsmål i arbeidsøkonomi uten å involvere den sterke antakelsen som er pålagt i eksisterende litteratur.

The project has allowed PI to devote the significant time to the research topics and achieve progress during the project period. The project has pushed further the research collaborations with UPenn, UToronto, USaarland and Statistics Denmark. The developed methodology will be implemented in real-world data that will allow to deliver answers to many important questions.

The proposal builds on a method developed in my previous work (Hagedorn, Law, Manovskii (2012,2014)) to provide substantive empirical answers to a number of classic but yet unanswered economic questions. The method can be used to identify unobserved worker and firm productivities and the consequences for output and productivity from moving any worker to any firm in the economy. The empirical implementation relies only on the routinely available matched employer-employee data sets, which are available (and will be used) for Denmark, Germany and Norway. The theoretical basis for the method is the classic assortative matching model of Becker (1973) with a frictional labor market augmented to allow for on-the-job search and stochastic match quality. Most fundamentally, the method allows to compute the optimal (e.g., output maximizing) assignment of individual workers to individual firms in the economy. The proposed method will be used to measure the extent to which sorting on unobservables can account for wage differences across groups of employers (large or small, exporters and no-exporters, belonging to different industries, located in different geographic regions, etc.) which were found to be present after controlling for observed worker and firms characteristics.The proposed research will also provide decompositions of wages into components due to workers, firms, and the assortative matching between them and will estimate the role of search frictions and sorting in driving the observed wage dispersion. The method will be used to study the extent to which mismatch between workers and firms varies over the business cycle and the role of this misallocation in accounting for the large income differences across countries. It will also measure wage discrimination and the effectiveness of policies, such as unemployment insurance, in facilitating better matching of workers to firms.