Labour market inequalities and the role of institutions.

• Author: Damiani, Mirella

1. Introduction

   A large body of studies has documented the changing wage structure and increasing inequalities that in the last decades have characterised many OECD countries (Katz and Autor, 1999, Atkinson, 2007, Machin, 2008). A short summary of the main facts gives the following picture: in the US a previous period of a monotonic surge of inequality, started in the mid 80s, was followed by a steep growth in the upper tail of wage distribution. Furthermore, recent studies have addressed new evidence signalling the 'changing nature of wage inequality' (Lemieux, 2008), such as a marked polarisation, no fully explainable by a supply-demand model (Autor, Katz and Kearney, 2008). European countries present a more heterogeneous scenario which calls for more attention on possible interplay of institutional settings and demand-supply forces. In addition, limited previous evidence on earning mobility seems to suggest considerable stability everywhere.

   The importance of these topics has been addressed by the workshop on 'Comparing Inequalities' organised by the Italian Association for Comparative Economic Studies (AISSEC), and held in Assisi in June 2010. One session of the workshop was devoted to "Labour Market Institutions and Wage Inequalities: a Comparative Perspective". This special issue, which includes a selection of papers that were originally presented at the workshop, offers contributions which can be helpful to obtain an enriched view of ongoing changes and a broader spectrum of plausible explanations.

   In this section, we offer a short appraisal of the large economic literature on wage inequality and institutions, with the main aim of clarifying how the papers collected in this symposium contribute to the related literature and in which directions they move.

   The basic facts at the origin of the debate have been changes in wage distribution which have revealed increasing inequalities from the mid '80s. Take the case of the US, the country which has been studied in much more detail and where the growth of inequality, at least in a first period starting from the 80s, was almost linear across all deciles: "Between 1979 and 1994, the ratio of the earnings of a male worker at the ninth decile compared with one at the median rose from 1.73 to 2.04. At the same time the earnings of that median male worker rose from 1.84 to 2.13 times the earnings of a worker at the first decile", as clearly synthesised by Blanchflower and Slaughter (1999, p.69).

   This empirical evidence has motivated a vast body of studies aimed at exploring these changes. In the 1990s, an early literature, based on an efficiency-driven perspective, has seen these inequalities as results from the interplay of supply and demand in the labour market for individual skills (Katz and Murphy, 1992). Indeed, one explanation obtaining a vast consensus assigned a central role to the skill based technological change (SBTC) which increased the relative demand for educated employees (Acemoglu, 1998). One broad conclusion reached by this literature was that the increased wage dispersion was simply the outcome of a "race between technological development and education" (the forces behind shifts in demand and supply of skilled employees) with technology winning (4). More precisely, "the surge of inequality evident in the 1980s reflected an ongoing, secular rise in the demand for skill that commenced decades earlier and perhaps accelerated during the 1980s with the onset of the computer revolution. When this secular demand shift met with an abrupt slowdown in the growth of the relative supply of college-equivalent workers during the 1980s ... wage differentials expanded rapidly" (Autor, Katz and Kearney, 2008, p. 300).

   The consensus reached by the SBTC hypothesis has remained central also in following years, even if some challenges have been advanced by a 'revisionist literature' which has signalled that explanations based on the SBTC view are difficult to reconcile with at least three related facts.

   First, international comparisons reveal that the growth in wage inequality is not an 'ubiquitous' phenomenon, even if technical changes have been pervasive. During the 1980s, many other industrialised countries have failed to record significant increase in wage inequality. Among OECD countries, in only five of them the bottom decile has fallen more than 10 per cent over the period 1980-2005 (Atkinson, 2007). Conversely, countries such as Germany, France, Japan, although affected by the same technological shocks represented by the computer revolution and ICT, have not experienced the same steep growth of wage inequality recorded in Anglo-Saxon economies. Germany, for instance, according to a recent study based on accurate micro-data, shows reverse patterns for the bottom tail of the distribution with respect to the US, even if shows similar changes with respect the US for the top coding (Dustmann et al. 2009). In addition, for Nordic countries, Gottschalk and Smeeding (1997) show that the phenomenon of widening wage distribution has been less evident, as found in Finland and Sweden.

   Evidence on OECD countries suggests that the magnitude and timing of the changes in the wage structure is far from homogenous. These different patterns of changes, like those observed in Continental Europe itself, call for a critical view of the SBTC hypothesis (Atkinson, 2007). A case in point, mentioned already above, is Germany, which for decades has been characterized by a stable structure of wages, but in recent years has shown noticeable changes and similar trends as those recorded in the US for the top part of the wage distribution.

   Second, the role of institutions is a significant determinant of the wage structure and contributes to explaining international differences. For example, in Anglo-Saxon countries, where union density is low and the wage-setting is decentralised, fall in demand lowers the earnings of less-skilled workers. By contrast, in Continental Europe where unions are strong and the wage-setting is more centralised, the same fall in demand increases unemployment but not wage inequality (Lemieux, 2008, p.23). Indeed, Koeninger, Leonardi and Nunziata (2004) find that a large set of labour institutions are associated with more compressed wage differentials. In addition, estimates of the impact of labour market institutions on inequality show that in the 1990s rising wage dispersion for workers in most OECD countries has been offset by reduction in unemployment, with ambiguous final effects on disposable income inequality (Burneaux, Padrini and Brandt, 2006). This would support the view that "European unemployment and US inequality are opposite sides of the same coin" (Machin, 2008, p.17).

   This macroeconomic framework, which represents a 'relatively new line of research' has been adopted by Checchi and Garcia-Penalosa (2008) and presented by Checchi in his invited lecture at the AISSEC workshop. The authors offer a unified framework in order to detect determinants not only of the wage dispersion, but also of the wage share, and of the personal distribution of income.

   Institutions play a significant role in explaining the growth of wage inequality also in a low...