Minimum Wages, Employment, and the “Harmony of Illusions”

Abstract

In contrast to Thomas Kuhn’s theory of scientific revolutions, the economic discipline appears to progress in a piecemeal, path dependent way continuously being dominated by the same paradigm. Instead of paradigmatic shifts, the history of economic thought is characterized by considerable paradigmatic resilience.

Drawing on the philosophy of science of Ludwik Fleck, this article demonstrates the basis of this resilience, as well as potential dangers to which it gives rise, and—while giving special consideration to research on the employment impact of minimum wages recently introduced in Germany—examines whether a necessary “thought style compulsion” may eventually turn into a “harmony of illusions.”

Keywords:

• minimum wage
• employment
• labor market
• scientific revolution
• paradigm

JEL Classification Codes:

• A11
• B41
• J30
• J40
• J42

Disclosure Statement

The author declares that there are no relevant financial or non-financial competing interests to report.

Notes

1 This “pre- or proto-idea” is similar to the pre-analytic vision in Schumpeter (1954, 41) and the heuristic or ontological dimension of Lakatos’s research programmes.

2 In the foreword to the English edition of Fleck’s Entstehung und Entwicklung einer wissenschaftlichen Tatsache [Genesis and Development of a Scientific Fact], Thomas S. Kuhn describes the “harmony of illusions” as “a damaging metaphor, for it reinforces the impression that, in the absence of social pressure, illusion might have been avoided” (Kuhn 1979, x). This incriminating impression appears to be justified if “absence of social pressure” does not simply refer to the “thought style compulsion” that necessarily characterizes every thought collective, but refers rather to the inadmissibility of different thought styles: i.e., of plurality.

3 Whereas Fleck emphasizes sociological factors in epistemic communities, the path dependency approach mentioned earlier focuses more on economic factors; see Yalcintas (2013).

4 Although there is no explicit reference to Fleck, Mirowski (2013, 354ff.) mentions denialism and agnotology as forms of thought style compulsion in a way that is very similar to him.

5 I focus here on the German-language literature—textbooks in this case—because the introduction of the minimum wage in Germany in 2015 represents the practical case that will be examined in what follows. But the American textbooks, which are particularly widely used internationally, could also be cited; see Dolar (2013).

6 Other studies arrive at similarly negative employment effects: Henzel and Engelhardt (2014) anticipate losses of between 470,000 and 1.45 million jobs; Arni et al. (2014) estimate a decline in employment of around 570,000 jobs. The different magnitudes are essentially due to different assumptions about the wage elasticity of labor demand.

7 Interestingly, most of the studies were conducted by precisely the research institutes that vigorously opposed the introduction of the minimum wage. This fact needs to be kept in mind when assessing the findings, as does the not unproblematic methodology of the incremental difference-in-difference approach, which has to replace the classical difference-in-difference approach that is normally used, because no regional comparisons (evolution of employment with and without minimum wage) can be undertaken in the case of a nationwide minimum wage. Time series analyses, in any case, do not show any sort of structural break since the introduction of the minimum wage.

8 These are jobs in which the employees earn no more than €450 per month and have reduced social security coverage.

9 A similar argument could be used to counter the theoretically conceivable claim that the minimum wage was raised to a level (w/P)_MW² considerably above the equilibrium level (w/P)* in figure 1, where the demand for labor also has a restrictive effect in the monopsonistic model—in contrast to the competitive model, however, the introduction of the minimum wage would not now have any macroeconomic effects on employment.

10 Although recent studies point to increasing labor market concentration in the United States (see Azar, Marinescu, and Steinbaum 2017), this basically only applies to rural, not to urban, labor markets. The proportion of workers who are in fact affected remains an open question.

11 These considerations (i.e., these thought style extensions) are by no means unknown in labor economics (see, for example, Reder 1969), but up to now they have not been applied to the minimum wage issue.

12 According to William Beveridge (1945, 18), there is full employment when the number of unemployed is equal to the number of vacancies. In our figure 2, for example, there would be (L*–L₁) frictional unemployment in equilibrium at a real wage (w/P)*, with the same number of vacancies at the same time.

13 In figure 1, the sum of classical and frictional unemployment (L₄–L₂) appears to be considerably higher under the conditions of a minimum wage than exclusively frictional unemployment (L*–L₁) in equilibrium without a minimum wage. But it is likewise obvious that the result could be different, if the position and slope of the curves, which can only be determined empirically, were represented differently.

14 The two-sector model comprises a sector that employs highly-skilled workers, and hence is unaffected by the introduction of the minimum wage, and a sector that employs low-skilled workers and hence is more affected by the introduction of the minimum wage.

15 One referee was not entirely convinced by the distinction between thought style extension and thought style supplementation because, as he understood it, they both boil down to altering some core assumptions. However, this would be a misinterpretation of my expositions as altering core assumptions would result in a thought style transformation which is exactly what is avoided. What can be altered only are auxiliary assumptions which either go beyond the standard model (though style extension) or allow for changes within the standard model (thought style supplements). The logic of adaptation of the model to empirical reality is different in both cases: while thought style extension changes the environment of the market actors (e.g., market structures or transaction costs) leaving it open whether the result is more in line with empirical evidence, thought style supplements change the actor’s behavior (e.g., quantity or price setting) leaving it open whether the argument is consistent with the underlying standard model.

16 In a recent issue of the Journal of Economic Perspectives, one of the world’s leading mainstream labor economists, Alan Manning (2021), addresses the “hard to find” and “elusive” employment effects of the minimum wage and comes to the conclusion finally that the search for other explanations is unproductive: In Manning’s view, minimum wage research should turn rather to determining the minimum wage level at which the expected employment effect is sure to arise. This represents yet a new pirouette in the harmony of illusions: The debate about a better explanation of the empirical anomaly is declared to be over and research is focused on a point that fits the customary theory better. The implicit assumption that the minimum wages hitherto observable around the world are not high enough to make the predicted effects demonstrable will irritate those mainstream German labor economists who cited the specific level of the German minimum wage as an argument for rejecting the findings of international minimum wage research. In any case, it does not occur to Manning either that a thought style transformation might be needed.

17 This refers, in particular, to the undeniable positive effects of the minimum wage: the higher income of the lowest income earners and increased job satisfaction (cf. Pusch and Rehm 2017a and 2017b, Gülal and Ayaita 2020).

Additional information

Notes on contributors

Arne Heise

Arne Heise is Professor of Economics at the Department of Socioeconomics at the University of Hamburg, Germany.