by Steve Machin and Pawel Bukowski
Do workers benefit from the firm’s success? This is a highly relevant macroeconomic question in our world of rising income inequalities (Piketty, 2014), falling labour share of income (Karabarbounis and Neiman 2014) and the rising market power of multinational companies (Autor et al. 2017).
In standard models of perfect competition, there should be no common connection between firms’ economic rents and wages. Yet, there is ample empirical evidence showing a positive firm-level correlation between wages and a measure of rents (e.g. profits), which has been interpreted as a sign that the labour markets are not perfectly competitive (Blanchflower et al. 1996; Manning 2011, Card et al. 2018).
A classic argument says that workers have bargaining power, and therefore they are able to claim a portion of a firm’s rents, a phenomenon known as rent-sharing. The recent work, however, document a significant fall in rent sharing since the 1980s in the UK and the US (Bell et al. 2018; Benmelech et al. 2018; Bell and Van Reenen 2011). This would explain the macroeconomic phenomenon of a lack of income being awarded to labour.
What remains unknown is whether the decline of rent sharing is universal or just specific to the Anglo-Saxon countries. Is it driven by the trade-induced shift in activity from manufacturing to the services sectors, by the progressing outsourcing, by the development of global value chains or maybe by the erosion of institutions protecting workers’ rights? This project for the Globalisation Hub of Rebuilding Macroeconomics is the first comprehensive attempt to look at cross-country, cross-industry and temporal patterns in rent sharing in order to shed light on these questions. We construct a consistent and representative panel of firms using data from the historical files of Bureau van Dijk’s Orbis, to estimate the extent to which companies are willing to share their profits with workers. We focus on ten manufacturing industries in twelve EU countries over two periods: 2000-2008 and 2009-2016.
Figure 1 shows a fall in rent sharing across almost all countries in our sample (Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Italy, Romania, Spain, Sweden and Great Britain) between 2000-08 and 2009-17.
To explain the measures, a one percent increase in profit per employee in GB between 2000 and 2008 led, on average, to a 0.005% increase in wages. While this elasticity might appear small, it implies that one standard deviation rise of profits per employee leads to 2.5% rise of wages, only due to rent sharing. After 2009, the rent-sharing fell to 0.0043% (or 1.4% of standard deviation rise). Among larger GB companies (the right hand panel) rent sharing fell from 0.0053% to 0.0035% (or from 2.5% to 1.2%).
Across all countries, the median decline in the rent sharing elasticity is 22% between the periods 2000-08 and 2009-17. In the sample of large companies, the median decline is double. Similar results emerge when we look at the evolution of rent sharing at industry level. Almost all manufacturing industries in the analysed countries, except “Electrical and optical equipment’, have experienced a decline of rent sharing after 2000. We also report, however, a substantial heterogeneity in the levels and changes of rent sharing across European countries and industries. Are there any systematic patterns in the cross-country and cross-industry levels and changes of rent sharing?
Figure 1: The Country-level rent-sharing elasticities
Notes: The scatter plots present the rent-sharing elasticities estimated at the country-level for two periods separately: 2000-2008 (x-axis) and 2009-2017 (y-axis). The left (right) panel shows the estimates based on a sample of firms, which average employment is larger than 20 (100). The red dashed line marks the 45-degrees line. Source: Orbis, own calculations.
In the second part of the project, we looked at the institutional influences of rent sharing by using country-level measures of unionization and labour market regulation from ICTWSS and OECD. We look at these two forms of labour market institutions, as they experienced considerable transformation in recent decades – arguably caused by international competition (e.g. Bertola and Lo Prete 2015). Since rent sharing has been interpreted as a measure of working bargaining power, one might expect to find a positive correlation between unionization and rent sharing, yet we find no evidence for such relationship. This paradoxical result is, nevertheless, in line with some previous studies (Manning 2011; Hildreth and Oswald 1997; Blanchflower et al. 1996), and we suspect that they are driven by a growing share of knowledge workers, who are likely to have higher bargaining power and work in less unionized sectors.
On the other hand, we find a strong and positive correlation between rent sharing and employment protection legislation (Figure 2), which covers all workers regardless of their union status. Labour market deregulation and the erosion of employment protection law might be thus an important factor contributing to the fall of rent sharing.
Figure 2: Rent sharing and employment protection
Notes: The scatter plots present the residualized rent-sharing elasticities and employment protection index. The left panel shows relationship between levels, the right panel between changes. The upper (lower) panel shows estimates based on a sample of firms, which average employment is larger than 20 (100). The red line marks the OLS fit line. Source: Orbis, OECD,, own calculations.
The third part of the project turns to the industry-level factors affecting rent sharing. Globalisation affects the level of economic rents and its volatility, on the one hand by enhancing competition and providing new risk-diversification strategies, but on the other hand by giving access to new markets, allowing potential dumping practices and increasing the risk of shocks propagated through trade network. Blanchard and Giavazzi (2003) argue that worker’s incentives to invest in their bargaining power should rise with the size of potential rents to capture. We find that indeed the size of economic rents, measured by profit margins, is positively correlated with rent sharing, and the relationship is strong and statistically significant. The average profit margins in our sample have declined since 2000, which might partially explain the fall of rent sharing over this period. A similar argument says that workers are also more likely to invest in their bargaining power when the volatility of rents is small and the employment is large. This is because workers are risk averse and there are economies of scale connected with the investment in bargaining power. Consistently with this logic, we document a negative correlation of rent sharing with the volatility of profit margins, and positive with the median employment size, however the relationships are not significant.
Exposure to global value chains increases the risk of outsourcing and breaks the connection between firm’s profitability and the local labour market conditions. Consistent with this, we find a negative correlation between the share of foreign value added in the output and our measures of rent sharing. A striking finding is that the more that firms’ profits arise from overseas, the less they share their rents with labour.
Finally, we explore whether the fall of rent sharing may contribute to growing income inequalities. There are two mechanisms to consider. First, the fall shifts income toward more concentrated capital. We show that industries, which have experienced a stronger decline of rent sharing, are also characterized by a stronger decline of labour share, encouraging calls for a bigger role for policies providing redistribution between capital and labour. Second, as rent sharing is a component of firm-specific pay premia, its decline will lower wage inequality, assuming that high-wage workers are employed in high-premia companies. Contrary to this, we show that industries with larger fall of rent sharing have also experienced widening of the between-firm wage gap. This might be explained by either by an unlikely scenario where low-wage workers work in high-premia companies, or that there is a third factor leading to a decline of rent sharing and a parallel increase in the sorting of workers across firms.