Is economie een empirische wetenschap? Je zou het niet zeggen als je naar neoklassieke arbeidsmarktmodellen kijkt. Daarin wordt verondersteld dat het loon wordt bepaald door de productiviteit van de laatst aangenomen werknemer. Afgezien van het feit dat die productiviteit moeilijk te meten is, zijn er meer problemen met deze veronderstelling.
by Dirk Damsma
Rather than providing a critical scrutiny of the way markets actually work, much of mainstream economic theory offers an ideological defense of their supposed superiority (as I argue here). This ideological slant is also apparent in the assumptions that underlie labor market models. Consider, for instance, the assumption that wages and salaries reflect a person’s marginal product. Or, put differently, that people’s wages are equal to the production value that is added by the last person hired. This assumption is unlikely to hold in reality.
First of all, a person’s marginal product can only be calculated in very specific circumstances. Take the pin factory of Adam Smith. In this example, Smith describes how one unskilled laborer could make at most twenty pins a day. With each extra worker the production process can be broken up into smaller parts, allowing each worker to specialize in only one or a few operations, such as heating, straightening or cutting the wire to size. This specialization process, Smith says, can lead a group of ten people to ‘make among them upwards of forty-eight thousand pins in a day’ (Smith 1776). This thus boosts their average productivity to four thousand eight hundred pins per worker per day. Each time the process is broken down into smaller parts, however, the productivity gains become smaller. In economists’ parlance: their marginal product decreases. If all workers are paid the marginal product of the last worker added, the productivity of all the non-marginal workers surpasses their wages and a profit can be made. So it only makes sense for a firm to equate wages to marginal productivity when the contribution to production of each extra person is less than the contribution of the previous person hired.
This logic breaks down when there is team production. If you have to move a couch, you need at least two people. If there is only one, his marginal product is zero. If you add a second, the job can be done. So in actual teamwork the concept of average productivity makes sense, but marginal productivity is a useless concept. In fully industrialized economies most production is genuine team production in this sense. A specific machine usually needs a specific technically determined amount of people to operate it. Not more, not less. When operated by the right amount of people, it churns out a technically determined amount of products per day. So in most cases it makes sense to speak of the marginal product of a team manning a machine, but not the marginal product of labor (or capital, for that matter). Moreover, the marginal product of each additional team is likely to be pretty constant, rather than fall. Under these circumstances, the marginal and the average product are the same. So there would be no profit left if the company paid each team its marginal product.
Most managers should have a low wage
All in all, remunerating workers according to their marginal productivity only makes business sense, if their marginal product falls as more workers are added. If wages are equal to employees’ marginal product, the amount of production foregone if a particular professional calls in sick or is fired should be exactly equal to the wage associated with their profession. This does not square with experience, however. When Ricardo Semler, the current CEO of Brazil-based Semco, took over from his dad, he fired almost all managers. If it were true that each professional is paid his or her marginal product, this should have resulted in falling productivity overall. But instead of falling, overall production and thus the average productivity of his employees actually rose quite dramatically (Wieners 2004). Apparently, then, the managers held them back from achieving their full potential. This implies that the marginal product of their managers was negative. So instead of getting wages many times those of the employees, they should have actually been charged for the privilege to manage them. In the Netherlands, Buurtzorg has illustrated the same principle (cf. De Beter 2018). Within the faculty of Economics and Business (where I work), you can see the same dynamic illustrated on a regular basis. If a timetable creator calls in sick for a week, all teachers and all students suffer. If a teacher calls in sick, only his or her students are inconvenienced. If the dean calls in sick, we probably wouldn’t even notice. Using this logic then, the dean should be paid less than teachers and teachers less than timetable creators.
In short, it is unlikely that wages reflect productivity. Instead, wages are negotiated between unequal partners. On the one hand, we have the employer that has the luxury of being able to choose between potential candidates and on the other the worker that desperately needs a job to gain a living. Framing these negotiations in terms of productivity only serves to defend the labor market as a fair place, which it is not.