A minimum wage is certain to emerge as a cornerstone policy of the next German government. The debate is under way, in the coalition negotiations between the CDU/CSU and SPD, but also in the media and academia, about its level, coverage and the mode by which it is to be implemented and determined over time.
A view often heard is that a minimum wage is „a good thing“ but it must not be „too high“. Indeed. To be too high is not a good thing. At some point when wages are raised from existing levels a price will be paid in terms of lost employment. But this begs the question: what is too high? Meanwhile conservatives and some business representatives are insisting that Tarifautonomie – the principle of collective bargaining between workers/unions and employers – should not be encroached upon. Social partners should negotiate regionally and/or sectorally differentiated minima.
A European comparison
An interesting new study by the European Foundation for Living and Working Conditions (a draft is available here) sheds, by way of a European comparison, some interesting light on these issues. There are a lot of interesting findings, but I will focus here on two of key relevance to the current German debate: the extent of wage inequality at the bottom of the labour market and the inadequacy of existing collectively agreed provisions.
The study examines, using detailed data on national wage structures, what the mechanical effects would be of introducing a minimum wage norm across Europe, according to which no worker is to be paid less than 60% of the national median. A key finding (see pp. 23-34 of the study) is that, of all the countries in Europe, Germany would be the one whose wage structure would be most profoundly affected. This is for two reasons. First, the share of workers below the 60% of the median line is the highest in the entire EU, at 24.5%: virtually a quarter of German workers earn less than 60% of the national median. In Finland the figure is 6.8%, in culturally similar Austria it is 14.6%, and even in the „neoliberal“ UK (where there is, however, a statutory minimum wage), it is substantially lower (18.9%). But that is not all. The distribution of wages in this segment of the working population also needs to be factored in. The authors of the report have calculated the average “distance” between the existing wages of low-paid workers and the 60%-of-median threshold. This, too, is (much) greater in Germany than in other countries.
The two measures – share of low-paid workers and wage distance – are combined in the following figure (Fig. 5 in the report, p. 34): the further the country is away from the origin, the greater the impact of a wage norm. Given that Germany comes out top (or rather bottom) on both measures, it is not surprising that this is also true of the combined measure. But the extent of the gap to the cloud of countries below (or rather above) is simply breath-taking. In European comparison Germany is a crass outsider regarding the bottom of its national labour market.
The second key finding to which I want to draw attention relates to the relative level of existing national minimum wages. (The discussion is based on Table 2, p. 23.) As is well known, most EU countries (21 out of 28) have statutorily based systems. Germany is one of the exceptions. The Eurofound study draws on work by Kampelmann, Garnero and Rycx (here) to produce an estimate of collectively agreed minimum wages in those countries lacking a statutory floor. It consists of a sort of weighted average of the collectively agreed minima prevailing in the country concerned. The collectively bargained figure for Germany is 8.8% below the 60%-of-median threshold, which is rather a small gap compared with other countries. However, almost one in six German workers (16.7%) earns less than this collectively agreed wage floor. This is the second-highest figure in the EU (after Italy, which has in some respects a similar system). The coverage of the (average) collectively agreed minimum in Germany, in short, is extremely poor. For instance in Austria “only” 7.3% of workers earn below the collectively agreed minimum, even though that floor is just 4.3% below the 60%-0f-median norm; in both cases less than half the percentages in neighbouring Germany.
Conclusions for the German minimum wage debate
The Eurofound study does not tell us what is the “right” level for a statutory minimum wage in Germany. But the two findings discussed offer two clear lessons.
Conservative and business-oriented commentators are concerned about the “bite” of the minimum wage proposals under discussion, and particularly the EUR8.50 per hour championed by the SPD: too many workers will be affected and the rise in labour costs will be excessive. What we can clearly see from this study is that such concerns are the mirror image of the crass exceptionalism of the German labour market: nowhere in Europe are the relative disparities at the bottom of the labour market anything like as severe. If Germany wants to move towards the European mainstream – and there appears to be a social consensus on this – then a large “bite” of any new measure is inevitable and desirable. That does not necessarily mean that EUR8.50 is right. But large numbers of workers with substantially higher hourly wages will have to happen if Germany is to leave its wage-inequality Sonderweg.
Secondly, those insisting on the sanctity of collective agreements to set wages and rejecting the “politicisation” of wage setting need to look at the figures – or face the charge of hypocrisy. Roughly one in six German workers, the second highest share in Europe, earn (hourly) wages below the minima set in collective agreements. This does not necessarily mean that a statutory national minimum wage is right. But little short of a revolution in German collective bargaining structures would be needed if coverage were to be brought in line with European norms without substantial state involvement.
 The median wage is the wage earned by the person exactly in the middle of the wage distribution. It is used in such studies rather than the average (mean) because it is not affected by small numbers of high earners and, in particular, it does not mechanically adjust when wages at the bottom of the distribution are raised, as occurs with the average.
 Note that this is not a working hour effect, as account has been taken of this factor in the calculations.