The last financial and economic crisis aggravated macroeconomic distortions such as unemployment, especially for the young and of the long-term type, increasing social and income disparities, and also hit the macroeconomic distribution of income. We are interested in the role of institutions in explaining this phenomenon. Every advanced economy has a set of established policies and institutions designed to reduce those disparities in order to sustain household incomes. There is extensive research on the effectiveness of institutions (such as, e.g., employment protection legislation or the wage bargaining process) on wage determination and employment. Some of these studies focus on attributing the decreasing path of the labor income share to globalization, labor-saving technological progress, or competition, but find little or no effect of labor market institutions. At the beginning of the new century and after the depression of 2008-2009, data reveal that the labor shares of European countries declined more than those of Anglo-Saxon countries. Labor share dynamics depend on factors moving the growth rates of real wages and labor productivity. Vergeer and Kleinknecht (hereafter VK, 2010) document that many European countries and Japan had substantially larger real wage growth and labor productivity growth than Anglo-Saxon countries, at least until 2005, but that real wage growth followed labor productivity growth more closely in the latter than in the former. For example, labor productivity growth declined from an average rate of 2.8% in 1963-1973 to 0.7% in 2003- 2010 in the UK, and wage growth was 3.1% and 0.6% in the corresponding periods (UK Office for National Statistics, 2011). Scholars attribute the main causes of this different behavior to the reduction of union representation, accruing bargaining power of employers, inducing (downward) wage flexibility. In Europe, however, unions may have a role in increasing productivity by providing agency services, which boost both the supply of and demand for firm-specific human capital (Young and Zuleta, 2011, and the literature mentioned therein). More stringent employment protection legislation and more generous unemployment benefits counteract the flexibility of European labor markets (less wage flexibility and higher hiring/firing costs). Furthermore, a lower demand for unskilled labor, due to the ICT revolution of the 1990s, deregulation and more competition in product markets, and lower barriers to entry pushed productivity growth up in Europe at the end of the 1990s. Vergeer and Kleinknecht (2014) argue that the higher productivity of European workers is due to labor-capital substitution induced by labor-saving technological change, the slow innovative process of ‘creative destruction’, longer job tenure, training and firm loyalty.

Furthermore, they argue that lower employment protection forces workers to work more hours to reach the same GDP growth rate (VK, 2014, p.382), as happened in Anglo-Saxon countries, thus resulting in jobs with lower productivity. As labor productivity behaves differently in each country, real wage changes are not the only drivers of change in the LS (Bentolila and Saint-Paul, 2003). This paper follows the empirical approach of these authors and Young and Zuleta (2011) to estimate the impact of social policy expenditures and labor market institutions on the labor share, controlling for factors such as globalization, (labor-saving or skillbiased) technological progress, competition, physical and human capital accumulation, and labor force participation. Finally, it provides an empirical analysis of whether the impact of each policy weakened after the beginning of the crisis. The paper is structured as follows. Section 2 reviews the empirical literature on the main determinants of the labor income share. Section 3 provides a description of the diverse labor market institutions and government social policies characterizing the countries in our sample and discusses the expected impact of such institutions on the labor share. Section 4 describes the data, while Section 5 defines the econometric framework and discusses the results. Finally, Section 6 concludes.