Every two years, Intermediair Magazine and Nyenrode Business University (note: despite the term “university” Nyenrode is not a scientific institution, but a business school) perform a study of Dutch salaries, which they call the National Salary Survey (NSO; Nationaal Salaris Onderzoek). Their 2019 report shows (among other things) a 6% discrepancy between the salaries of men and women up to 35 years of age, to the detriment of women. Naturally, journalists, in particular those with feminist sympathies, grab this “fact” to once more discuss the “wage gap” and how bad women have it in the western world.
I was rather surprised by the report’s findings, as the Dutch Bureau for Statistics (CBS; Centraal Bureau voor Statistiek) has reported since 2014 at least (I have not looked further into the past) that the salaries of women up to 36 years of age are 8% higher on average than those of men. In De Volkskrant, two articles in the same issue bring up both the 6% less and 8% more figures, without noting the discrepancy. I tried to ignore the discussions on the wage gap this time around, as I am getting rather tired of them, but since the 6% figure is now coming up in more articles I feel obliged to say something (despite the fact that few people read this, but sometimes I need to rant a bit).
The first question is: which of these figures is correct? The answer to that is clear: the 8% more for women figure of CBS is correct, as CBS writes its reports based on objective studies for the government — a government, I might say, which has rather feminist leanings as far as social justice is concerned. Note that, according to CBS, this figure is explained for the most part by the higher level of education that women tend to have.
The second question is: where does the difference between the results of CBS and the NSO originate? That is also clear: CBS has access to salary data of all Dutch citizens and bases its study on millions of data points. The NSO bases its results on data which are derived from a website, where they asked people to self-report on their salaries. They came up with close to 44,000 data points, which sounds like a lot, but is nothing compared to the millions of data points of CBS. Moreover, self-reports are notoriously unreliable, and salaries are something which men typically like to boast about and women tend to be humble about. So we may conclude that the NSO’s findings are unreliable and should be thrown into the garbage.
The third question is: why the hell does the NSO exist at all, when CBS produces a report on Dutch salaries every two years, based on objective data of every working person in the country? Producing the NSO is just a waste of time and money. Moreover, I cannot imagine that the leaders of the NSO research do not know about the CBS figures, and they should therefore have realized that their data must be skewed far too much to draw any conclusions from. Had they been responsible scientists, with the knowledge that their results are worthless, they would have declined to produce the report. So the only answer that I can come up with for this question is that the research leaders of the NSO are just interested in the attention that their report gets, and have no qualms about closing their eyes to the truth.
Note that CBS has examined the variance between salaries of all Dutch citizens using factor analyses, and has concluded that gender is in no way an explanation for observed salary differences. Therefore, any article or report which tries to pose salary differences between men and women as a gendered issue are fundamentally wrong. You may expect journalists and activists to erroneously state something along the lines of “but the fact that women earn less than men means that gender discrimination is at work, right?” but a scientist who bases their studies on statistics should understand their statistics better. Thus the NSO is no more than a pseudo-scientific manifesto and does not deserve the attention that it gets.