The labour share of income: heterogeneous causes for parallel movements?

Abstract

Die Dis cus si on Pape rs die nen einer mög lichst schnel len Ver brei tung von neue ren For schungs arbei ten des ZEW. Die Bei trä ge lie gen in allei ni ger Ver ant wor tung der Auto ren und stel len nicht not wen di ger wei se die Mei nung des ZEW dar.Dis cus si on Papers are inten ded to make results of ZEW research prompt ly avai la ble to other eco no mists in order to encou ra ge dis cus si on and sug gesti ons for revi si ons. The aut hors are sole ly respon si ble for the con tents which do not neces sa ri ly repre sent the opi ni on of the ZEW.Download this ZEW Discussion Paper from our ftp server:ftp://ftp.zew.de/pub/zew-docs/dp/dp10024.pdf Non-technical summaryDeclining labour shares in a large number of countries, particularly in continental Europe, have renewed the academic and political debate about the factors which explain these downward trends. While the share accruing to labour in the division of national income was long being seen as constant over time, this stylised fact has recently been challenged. Several explanations for temporal and persistent movements of the labour share have been brought forward in the debate: effects of structural and technological change, influences of globalisation and increased product market integration, and the importance of institutional settings, often with a focus on wage bargaining structures.The aim of this paper is to take a structured empirical approach at assessing the relative importance of those factors across countries. In particular, we focus on proper dynamic model specification and test the validity of the homogeneity assumption of slope coefficients frequently implied in previous studies. We employ fixed effect estimators as well as pooled mean group and mean group estimators, the latter two in a dynamic heterogeneous panel framework.In a sample of OECD countries, we find negative effects for two explanatory variables of the labour share: the capital output ratio and trade openness. Furthermore, we are not able to reject the homogeneity assumption on the slope coefficients. This first finding lends important support to the standard theory on labour share movements. However, as far as other explanatory variables often found in the literature go, the picture is more mixed. Total factor productivity, in particular, is found to exert heterogeneous effects across countries and no clear support for the pooling assumption of slope coefficients in a linear dynamic model is found.In order to add more detail to our analysis and to address the role of institutional arrangements, in particular with respect to the bargaining process, we split the sample into two groups of countries characterised by differently strong unions. We find important differences in the coefficient values and levels of significance. For more marketoriented countries with lower union density, we see the labour share being driven down by variables capturing technological change and shifts in the relative usage of factors of production. For countries with strong unions, however, we find ...

The Decline of the U.S. Labor Share Across Sectors

“…These findings appear to hold for other countries. Hogrefe and Kappler (2013) and Rada and Kiefer (2016) show similar results on the basis of econometric analysis for OECD countries. Unlike Elsby et al (2013), the latter authors also find that union density remains a fairly robust correlate of the labor share even when including an index of globalization.…”

The Long‐Run Dynamics of the Labor Share in Japan

“…The estimator is part of the panel time‐series literature which emphasizes: (i) possible non‐stationarity of the processes; (ii) cross‐section dependence, that is, the possible correlation in the disturbances across sectors; and (iii) slope, not just group time‐invariant, parameter heterogeneity (Eberhardt, 2013). Like other mean group approaches (Pesaran, 2006; see Hogrefe and Kappler, 2013, for a specific application to the analysis of the labor share), the AMG estimator first estimates N group‐specific ordinary least‐squares regressions and then averages the estimated coefficients across groups. Cross‐sectional dependence is controlled for by the inclusion of a common dynamic effect, which in the AMG is obtained in the first step estimation of a pooled regression model augmented with year dummies, obtained by first difference ordinary least squares.…”

The relation between labor productivity and wages in Brazil:

“…Finally, overarching these microeconomic dynamics, technology-biased innovation and investment have been identifi ed by many authors as a cause of declining labor shares of income (in other words, as a cause of divergence between labor productivity and wages) across industries and countries (Hogrefe;Kappler, 2012;Bentolila;Saint Paul, 1999). In this analysis, capital-augmenting technical progress, such as the widespread adoption of computers from the 1990s onwards, may generate factor biases between capital and labor, with the degree of bias determined by activityspecifi c elasticities of substitution between factors (Karabarbounis;Neiman, 2013;Findlay;Jones, 1999;Feenstra;Hanson, 1999).…”