Reinhold Kosfeld (p), Christian Dreger

Spatial econometric analysis of the wage curve are usually rationalized by monopsonistic competition within labour markets. External effects extracted from data of administrative regions insufficiently capture the competition effects. Moreover, issues of nonstationarity are usually ignored. Both aspects gain particular relevance when the existence of a wage curve is difficult to prove in high unemployment regimes. This issue is debated for East Germany in the literature. In this paper, we make use of labour market regions in establishing the existence of a wage curve for East Germany. Because of nonstationarity of spatial data a global panel cointegration approach is adopted. By specifying a spatial error correction model (SpECM), equilibrium adjustments are considered in both time and space. Within this framework, evidence for the existence of a long-run East German wage curve is found.

While the need of a proper handling of nonstationary variables is by now well understood in standard econometric panel analysis, aspects of spurious regression are usually ignored in spatial analysis. Global cointegration takes account of local and spatial adjustment processes between nonstationary variables. The modelling framework consists of a panel cointegrating relationship and a spatial error correction model (SpECM). The SpECM aims at merging the long- and short-run perspective in identifying co-movements of wages and unemployment within and between the cross-sections. The merit of the approach lies in the potential of revealing the consequences of spatial effects for the time series behaviour of the variables.

The results show that regional spillovers play a crucial role in establishing the equilibrium locus in East German labour markets. A highly significant cointegrating relationship is obtained when wage and unemployment spillovers from nearby regions are included. These corroborate the theory of monopsonistic competition. The existence of a regional Phillips curve is clearly rejected.

Merja Kauhanen (p), Terhi Maczulskij

Although job polarization of the labor market has been well documented in a burgeoning literature, we still know to much lesser extent about the implications of job polarization at the individual level.

Our study extends the empirical literature on the consequences of job polarization at the individual level in the following important policy-relevant ways. First, we provide new evidence on the mobility patterns of workers from declining middle-skilled routine occupations and the wage outcomes of mobility on which there is still scarce evidence. Second, in the analyses we distinguish routine cognitive and routine manual occupations, which makes it possible to compare the transition patterns and the wage outcomes between these two groups. Third, we extend our analysis to include regional aspect by taking into consideration whether within-country migration (to polarizing/non-polarizing regions) mitigates the negative effects of job market polarization.

In the analyses we use Finnish Longitudinal Employer-Employee Data (FLEED) for the period from 1995 to 2009. The data are based on various administrative registers that have been linked together using identification codes for individuals, firms and plants. We estimate multinomial (polytomous) logit regressions to study the transition patterns of routine cognitive and routine manual workers and wage growth regressions to study the impact of mobility.

Our results suggest that the consequences of job polarization and mobility patterns are not similar to all routine workers, but that there are distinct differences between routine manual and routine cognitive workers. We find that routine manual workers are more likely to end up in low-paying non-routine manual jobs or become unemployed, while routine cognitive workers are more likely to move upwards to non-routine cognitive jobs. Further, we find that worker migration to highly industrialized, urban regions mitigates the negative labor market effects of polarization.

Torben Dall Schmidt (p), Timo F. Mitze, Amjad Naveed

Identifying the labor market return to creativity is a revisited issue in the empirical literature. Although the link between the “creative class” and economic development has been identified in numerous publications, recently these empirical approaches have been criticized for lacking solid micro-foundations of the underlying effects taking place. Based on register data for Denmark, this paper attempts to trace out such returns to creativity at the micro level. To do so, we estimate standard wage and wage growth equations for Danish residents where we control for a variety of individual characteristics, which are typically found to confound with the observed labor market outcome. For instance, it is often discussed whether creativity and education are interchanged measured of the human capital implied. Based on the use of register data, we apply an identification strategy which keeps education levels constant for the individual and considers job changes from jobs that have been classified as non-creative to jobs classified as creative. We argue that wage increases from such job changes reflect the unobserved factors increasing human capital for given education levels. Results show that the return from creativity identified from this approach is in the range of a 3 percent increase in the wage level and 6 percent in the wage growth. We also find that the results vary in different regional contexts, e.g. when splitting the sample into job changes from non-creative to creative in large, medium-sized and small cities. We find the wage and wage growth effects are the strongest in large cities – likely reflecting positive urban externalities.

Anping Chen, Tianshi Dai (p), Mark Partridge

China is experiencing rapid urbanization with the steady emergence of large cities, leading to policy discussions of the role of large cities in the country’s development. There is a large related literature that documents the role of city size on productivity and wage premiums. While the consensus is that agglomeration plays an important role in economic development and large cities can act as engines of economic growth, there is relatively little empirical knowledge of the effects of agglomeration on inequality. This is an important gap in the literature: given that high-productivity firms and low-productivity firms may both sort to large cities, the continuing rapid growth of large cities may further increase Chinese income inequality, which is already at high levels.

In this study, we apply panel data from a micro firm-level survey and from city-level data to investigate whether there is causal relationship between agglomeration and establishment wage dispersion in China. We not only examine the short-run linkages between city size and wage dispersion, but also assess their long-run relationship to investigate whether the effects have temporal variation. Furthermore, we explore whether the effects of city size on wage dispersion differ between coastal and inland regions to assess development’s role in forging agglomeration links. In addition, we assess this relationship for different public and private ownership structures within these two regions followed by an investigation of whether it is high-wage or low-wage workers that benefit from agglomeration. Given potential endogeneity of city size, we employ an instrumental variable regression (IV) approach.

We find strong evidence that agglomeration has positive effects on wage dispersion in the short run and negative effects in the long run. The link between agglomeration and wage dispersion is much stronger in the inland region than in the coastal region and the differing coastal/interior region results are not due to differing private/public ownership distributions. Rather, these differences appear to be due to differing stages of development. We also find that positive short-term labor demand shocks tend to increase wages relatively uniformly in the upper-half of the distribution, while long-term agglomeration effects dampen average firm wages at the top and middle of the distribution and increase relative wages for firms at the bottom.