Essays on collective bargaining, wage inequality and firm dynamics

Abstract

In this thesis I study collective bargaining, wage inequality, firm dynamics and the ways in which they interact. In the first chapter I investigate the extent to which wages vary across different industries after controlling for detailed worker and job characteristics and how this is related to the wage setting institutions of a given country. In the second chapter I study the drivers of the growth in earnings and wage inequality in Italy between 1985 and 2018 and compare them to the USA. In the third chapter I build a large-firm search model with heterogeneous firms and endogenous firm entry in order to compare the aggregate implications of firm-level and sector-level wage setting. Chapter 1: Informal Coordination of Wage Bargaining and the Size of Sector Wage Premiums I use the Eurostat’s Structure of Earnings Survey which is a unique data set containing microdata harmonised across European countries in order to investigate the relationship between wage setting institutions and wage dispersion. First, I find that in countries where the main level for wage bargaining is the sector, the dispersion of wages across sectors after controlling for detailed worker and job characteristics is substantially smaller than in countries where wage bargaining occurs predominantly at the firm level. This is surprising given that sector-level bargaining implies equalising wages only for each worker type within industries. The result points towards strong informal coordination of wages across sectors achieved via pattern bargaining. Second, I find that the overall wage dispersion is larger in the countries with firm-level wage setting. As a result, the relative share of the overall wage inequality that can be attributed to the sector that a worker is employed in is not generally larger in the countries with firm-level wage setting. Third, I find that in countries with sector-level wage setting observable worker characteristics explain a larger fraction of the overall wage variance. This is likely because wages are not individually bargained, but are based on a collectively bargained formula that includes characteristics such as worker occupation, education, years of experience and tenure. Chapter 2: It’s the Sectors, not the Firms: Accounting for Earnings and Wage Inequality Trends in Italy 1985-2018 Using administrative data for the entire universe of private-sector employment in Italy for the period 1985-2018 we investigate the drivers of the growth in earnings and wage inequality and compare them with other countries, in particular the USA. First, we find that the majority of the increase in earnings inequality in Italy (62%) is due to an increase in the variance of average earnings between firms and only about 38% is due to increased variance within firms. This is very similar to the results found for the US (Song et al. (2019)). Second, we decompose the between-firm variance into the between-sector variance and the between-firms-within-sector variance. Whereas in the US, the contribution of the between-sector variance to the overall growth in earnings dispersion is minimal and the majority of the growth of inequality is a between-firm-within-sector phenomenon, in Italy the rising between-sector variance explains approx. 42% of the overall increase in earnings dispersion, with the between-firm-within-sector component playing only a small role. The most likely explanation for the different patterns of rising earnings inequality between Italy and the USA seems to be differences in wage-setting institutions. Wage bargaining in the US is at the firm level whereas in Italy over 90% of workers are covered by sector-level collective agreements that specify wage floors for each occupation. This does not necessarily mean that sector wage premiums became larger in Italy. It is much more likely that the sector-level negotiators simply allowed increases in the relative demand for high skilled workers driven by technological changes to be reflected in the minimum wages for different occupations. This in combination with the fact that occupational composition of the workforce differs hugely between the narrowly defined industries, but arguably much less within them can potentially explain why the growth of earnings (and wage) dispersion between sectors accounted for such a large share of the overall growth of inequality in Italy. Finally, we find that the pattern found in the USA, the UK and Brazil, that changes in the dispersion of average earnings between firms within the same narrowly defined industries explain the majority of the changes in earnings dispersion, is not universal. Chapter 3: The Aggregate Implications of Sector-Level vs Firm-Level Wage Setting in a Frictional Labour Market I compare a setting where a firm negotiates wages with each worker separately with a two-tier collective bargaining framework. In the latter case a sector-wide union and an employer organisation first bargain over the tariff wage that applies to all the homogeneous workers and then additional wage premiums are bargained collectively at firm-level. The model can vary the extent of centralisation of wage bargaining by adjusting the ability of workers to organise industrial actions at firm-level. The modelling framework is a search model with multi-worker firms that are heterogeneous in productivity. As a result of fixed costs of production there is a threshold firm productivity level and thus a firm-selection mechanism. Under firm-level bargaining (either individual or collective) there is wage dispersion across firms driven by rent sharing and the wage is an increasing function of the firm’s output per worker. Because of convex hiring costs firms only gradually grow towards their target size. Given that firm productivity is constant over the life of the firm and there are decreasing returns to scale, wage is declining in firm age. Firms with higher permanent productivity face higher wages along their entire growth path. My main finding is that reducing wage dispersion across firms while keeping average wage constant leads to a higher total value added. I provide two novel arguments in favour of sector-level bargaining. Firstly, centralised wage setting can reduce the young firm wage premium and thus encourage more firm entry. Secondly, it can weaken the link between firm size and wages and thus reduce the inefficiencies associated with the over-employment effect which has been identified by the existing literature.