By Per Skedinger (Photo: Karl Gabor)
The EU Commission has recently proposed that all workers in the Union should be covered by a “fair minimum wage”. Precise details have yet to be revealed, but the Commission has stated that the traditions of Member States in minimum wage setting, including collective bargaining, shall not be infringed upon (European Commission, 2020). The proposal has nevertheless been regarded by unions and employers in Sweden and the other Nordic countries as a threat to their system of collective bargaining and collectively agreed minimum wages.
The main motivation behind the EU initiative is the observation that the share of low-paid workers, as well as the share of “working poor”, has increased in the Union. But even if one accepts that minimum wages are effective instruments for addressing such problems – which is not self-evident in the case of the “working poor”, since most poor individuals do not work – there is still the question of whether these wages should be introduced and determined by the members states or at the EU level. The Commission is silent on this fundamental question.
The desirability of EU-wide minimum wages is closely related to how minimum wages will affect employment, in light of the Commission’s statement that wages for the low-paid could be increased without substantial loss of employment. To examine the reasonableness of resistance to the EU proposal among Nordic countries, I will assess what research has to say about the employment effects of minimum wages in labor markets in which these wages are collectively agreed.
Collectively agreed vs. statutory minimum wages
There is no statutory minimum wage in the Nordics and in some other European countries, such as Austria and Italy. In these countries, minimum wages are set in collective agreements between unions and employer organizations. Most other countries, in the EU and elsewhere, regulate minimum wages mainly by law.
Since the coverage of collective agreements is high in the Nordic countries – between 70 and 90 per cent of workers according to the OECD – the collectively agreed rates are functionally equivalent to statutory minimum wages in the sense that they cover almost the whole labor market. There are some important differences between the two systems, though. Firstly, absent statutory minimum wages, government involvement in wage formation is limited. Hence the collectively agreed system is to a large extent “self-regulatory”. Secondly, minimum wages determined in collective agreements differ across industries and often vary along several other dimensions, such as age, occupation and experience. This stands in contrast to national, statutory minimum wages, which usually apply uniformly to all workers (although in some countries they also differ by age and region). In collectively agreed systems, there is thus more potential for flexibility in adapting minimum wages to conditions in different labor markets.
A common measure of the importance of minimum wages relates to how far these cut into the wage distribution. The relative minimum wage, also known as the “bite”, is defined as the minimum wage divided by the median wage in the economy and is expressed as a percentage. It has been observed that relative minimum wages are higher on average in countries with collectively agreed rates than in statutory systems (Calmfors et al., 2018; European Commission, 2020). Flexibility in these countries has been achieved, one could say, at a higher level of minimum wage. This might have important implications in terms of employment effects, as the theories that I will discuss predict that detrimental employment effects are more likely with higher minimum wages.
One reason why most countries use statutory minimum wages, despite the associated disadvantages mentioned above, may be that the coverage of collective agreements is too low in these countries for collectively agreed rates to be an effective instrument for raising the lowest wages. The legal minimum wage then acts a substitute for collectively determined minimum wages. The German experience seems to support this interpretation, as it introduced a national minimum wage in 2015 after a long period of declining coverage of collective agreements. A causal relationship between weak collective bargaining and statutory minimum wages is not obvious, however. Aghion et al. (2011), for example, have argued that statutory minimum wages crowd out unionization and coverage of collective agreements, and especially so in France.
Economic theories of employment effects
There are various economic theories regarding the employment effects of minimum wages (see Cahuc et al., 2014, for a textbook treatment). A basic finding in this literature is that minimum wages need not decrease employment. According to the simplest model of a competitive labor market, a minimum wage set above the market-clearing level will cause employment to decrease, because it is not profitable for employers to hire workers to the same extent as before. But according to the monopsony model, in which there is one large employer or a cartel of employers, employment can increase in response to a legislated minimum wage. A monopsonist will find it profitable to cut employment to reduce the market wage. The introduction of a minimum wage above the monopsony wage will then raise employment, since there will be a larger labor supply. Such effects are not likely to be present with strong unions however, as in collectively agreed minimum wage systems, since the unions reduce the possibility for firms to exert monopsony power in wage setting.
Search and matching models seem to provide the most realistic theory of the relationship between minimum wages and employment. According to these models, it is both costly and time-consuming for workers to search for jobs and for employers to fill vacancies. With such frictions in the labor market, a minimum wage may be employment-enhancing if it increases job search and the propensity of employers to set up vacancies. If the minimum wage is set too high, however, the negative effect on employment from higher wage costs will dominate over the positive effect from the faster filling of more vacancies.
The relationship between employment and the minimum wage therefore takes the form of an inverted U, with higher employment when the minimum wage is increased from a low level and lower employment when it is increased from a high level. An implication of search and matching theories is that there is an optimal level of the minimum wage, which inter alia, increases in the productivity of the labour force and decreases in the recruitment costs for employers.
Two Swedish studies
There is a vast literature on the employment effects of statutory minimum wages (see Belman and Wolfson, 2014, for an extensive survey). But very few studies deal with collectively agreed minimum wage rates. The findings in the studies of statutory systems are quite mixed – some find negative effects, some no effects, and a few even positive effects. Many of the studies finding negative effects, especially for the UK and US, report small employment elasticities, close to –0.1. As mentioned above, collectively agreed minimum wages differ by industry, and are relatively high compared to statutory minimum wages. If minimum wages have any negative employment effects under collective bargaining, these effects should primarily be manifested in low-wage industries.
Swedish hospitality and retail are two such industries, examined by Skedinger (2006) and Skedinger (2015) respectively. The employment effects of minimum wages are identified in both studies by comparing outcomes for workers assumed to be affected by minimum wage changes to those for a control group of workers assumed to be unaffected, depending on their position in the wage distribution. With increasing minimum wages, a worker is defined as being in the affected group if she (it is mostly a she in these industries) had a wage in the previous year that was below the minimum wage in the current year. The control is defined as a worker with a wage in the previous year that was slightly above the current minimum wage.
The outcome under consideration in this context is the job separation rate, i.e., the probability that the worker is no longer employed by the firm in the current year. The control group is assumed to capture “structural” turnover in the industry, which is considerable in both hospitality and retail, and the difference is then attributable to “excess” turnover due to minimum wages. The estimated employment elasticities turn out to be around –0.5 in the hotels and restaurants industry and approximately –0.3 in retail. Since these elasticities are conditional on previous employment, they are not directly comparable to the (unconditional) elasticities typically estimated in the literature referred to earlier.
The findings are in line with those in most other studies for Sweden and other countries with collectively agreed minimum wages. Of 13 studies for Austria, Denmark, Finland, Italy, Norway and Sweden surveyed in Ek and Skedinger (2019) and Skedinger (2020), seven consistently indicate negative employment effects, four suggest mixed results, and two find no statistically significant effects.
Minimum wages in collectively agreed systems therefore seem to be associated with negative employment effects to a larger extent than in statutory systems (with the proviso that the number of studies is small in the first group). This is consistent with the theoretical prediction that detrimental employment effects are more likely when the minimum wage is high.
Weak case for EU regulation
One motive for supranational regulation could be a real or perceived race to the bottom, in which member states compete against each other by lowering minimum wage growth (or even minimum wage levels) in order to gain a competitive edge over other countries in the Union. But minimum wage levels in the new member states have approached those in the old member states (Skedinger, 2020). In 2000, purchasing-power adjusted minimum wages in the new member states were on average nearly 30 per cent of those in the old member states; the corresponding figure for 2018 had risen to two thirds. The development of minimum wages in the Union looks more similar to a race to the top.
Another problem with the “race to the bottom” argument is that it could be used for protectionist purposes, which does not sit well with the objectives behind the formation of the EU. Some old member states with far-reaching ambitions for income redistribution could support an EU-wide system for setting minimum wages because it helps protect their industries from competition from new member states.
Moreover, the theory and empirical work discussed above suggest that the employment effects of minimum wages are highly dependent on the characteristics of minimum wage systems as well as those of national labor markets. This gives member states an informational advantage in relation to the EU in setting minimum wages that are not harmful to employment. It is therefore hard to see what would be gained by regulating minimum wages at the EU level.
Aghion, Philippe, Yann Algan and Pierre Cahuc (2011), “Civil Society and the State: The Interplay between Cooperation and Minimum Wage Regulation”. Journal of the European Economic Association 9(1), 3–42.
Belman, Dale and Paul J. Wolfson (2014), What Does the Minimum Wage Do?. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research
Cahuc, Pierre, Stéphane Carcillo and André Zylberberg (2014), Labor Economics. Cambridge, MA, and London: MIT Press.
Calmfors, Lars, Petter Danielsson, Simon Ek, Ann-Sofie Kolm, Tuomas Pekkarinen and Per Skedinger (2018), Hur ska fler komma in på arbetsmarknaden?. Stockholm: Dialogos.
Ek, Simon and Per Skedinger (2019), “Wage Policies and the Integration of Immigrants”. In Calmfors, Lars and Nora Sánchez-Gassen (eds.), Integrating Immigrants into the Nordic Countries. Copenhagen: Nordic Council of Ministers.
European Commission (2020), “Consultation Document. First Phase Consultation of Social Partners under Article 154 TFEU on a Possible Action Addressing the Challenges Related to Fair Minimum Wages”. Brussels.
Skedinger, Per (2006), “Minimum Wages and Employment in Swedish Hotels and Restaurants”. Labour Economics 13(2), 259–290.
Skedinger, Per (2015), “Employment Effects of Union-Bargained Minimum Wages: Evidence from Sweden’s Retail Sector”. International Journal of Manpower 36(5), 694–710.
Skedinger, Per (2020), “Reglering av minimilöner på EU-nivå – rätt väg att gå?” Stockholm: SIEPS. Forthcoming.