Ulrich Zierahn

Racing With or Against the Machine? : Evidence from Europe

A fast-growing literature shows that technological change is replacing labor in routine tasks, raising concerns that labor is racing against the machine. This paper is the first to estimate the labor demand effects of routine-replacing technological change (RRTC) for Europe as a whole and at the level of 238 European regions. We develop and estimate a task framework of regional labor demand in tradable and non-tradable industries, building on Autor & Dorn (2013a) and Goos, Manning and Salomons (2014), and distinguish the main channels through which technological change affects labor demand. These channels include the direct substitution of capital for labor in task production, but also the compensating effects operating through product demand and local demand spillovers. Our results show that RRTC has on net led to positive labor demand effects across 27 European countries over 1999-2010, indicating that labor is racing with the machine. This is not due to limited scope for human-machine substitution, but rather because sizable substitution effects have been overcompensated by product demand and its associated spillovers. However, the size of the product demand spillover -- and therefore also RRTCs total labor demand effect-- depends critically on where the gains from the increased productivity of technological capital accrue.

Agglomeration, Congestion, and Regional Unemployment Disparities

Regional labor markets are characterized by huge disparities of unemployment rates.Models of the New Economic Geography explain how disparities of regional goods markets endogenously arise but usually assume full employment. This paper discusses regional unemployment disparities by introducing a wage curve based on efficiency wages into the New Economic Geography. The model shows how disparities of regional goods and labor markets endogenously arise through the interplay of increasing returns to scale, transport costs, congestion costs, and migration. In result, the agglomeration pattern might be catastrophic or smooth depending on congestion costs. The transition between both patterns is smooth.

The importance of spatial autocorrelation for regional employment growth in Germany

In analyzing the disparities in regional employment growth in Germany, in the recent empirical literature the so called shift-share-regression models are frequently applied. However, these models usually neglect spatial interdependencies, even though such interdependencies are likely to occur on a regional level. Therefore, this paper focuses on the importance of spatial dependencies using spatial autocorrelation in order to analyze regional employment growth. Spatial dependency in the form of spatial lag, spatial error, and cross regressive models are compared. The results indicate that the exogenous variables’ spatial lag sufficiently explains the spatial autocorrelation of regional employment growth.

Women and work: what role do social norms play?

Against the background of the current (economic) research which concentrates particularly on individual and structural factors, this paper examines if and to what extent social norms (in terms of attitudes towards gender roles and work commitment) can make a complementary statement in explaining women’s employment status and number of working hours. The impact is presumed to be enhanced through norms shared by people belonging to the same households, peer groups, and by residents of the same region. The analysis relies on a rich German dataset and employs a probit model with sample selection. The results highlight, among other things, the importance of the ‘relevant others’ (particularly partners living in the same household and peers sharing similar social and work characteristics, but not necessarily geographical proximity) in explaining women’s employment status.

Regional Unemployment and New Economic Geography

Regional labor markets are characterized by huge disparities. The literature on the wage curve argues that there exists a negative relationship between unemployment and wages. However, this literature cannot explain how disparities of these variables between regions endogenously arise. In contrast, the New Economic Geography analyzes how disparities of regional goods markets endogenously arise, but usually ignores unemployment. Therefore, this paper discusses regional unemployment disparities by introducing efficiency wages into the New Economic Geography. This model shows how disparities of regional goods and labor markets endogenously arise through the interplay of increasing returns to scale, transport costs and migration.

Field of Study, Qualification Mismatch, and Wages: Does Sorting Matter?

Graduates from Science, Technology, Engineering, and Mathematics (STEM) are usually found to have higher wages and a lower risk of overqualification. However, it is unclear whether we can interpret the effect of STEM subjects on overqualification and wages in a causal way, since individuals choosing these subjects might differ systematically in unobserved characteristics, such as ability. Using data on German male graduates we show that unobserved heterogeneity indeed matters for differences in the risk of overqualification and wages when STEM graduates are compared to the Business & Law group, while it plays only a minor role for the difference between STEM graduates and the Social Sciences & Humanities group.

The Importance of Spatial Autocorrelation for Regional Employment Growth in Germany

In analyzing the disparities of the regional developments in the volume of employment in Germany, in the recent empirical literature so called shift-share-regression-models are frequently applied. However, these models usually neglect spatial interdependencies, even though such interdependencies are likely to occur on a regional level. Therefore, this paper focuses on the importance of spatial dependencies using spatial autocorrelation in order to analyze regional employment development. Spatial dependency in the form of spatial lag, spatial error and cross regressive model are compared. The results indicate that the exogenous variables’ spatial lag sufficiently explains the spatial autocorrelation of regional employment growth.