It turns out that such criteria, based on the current employment status of the candidates, are also not legal according to French law. Public hiring must be based on criteria that are directly relevant to the position. Factors such as sex or race are obviously illegal, but any other criterion irrelevant to the advertised position is also illegal (which is not the case for private hiring, which is more flexible). This is why you will never find the criteria mentioned in this section on any official document. This is also why the CNRS and other research institutions insist that all reports and any document produced by committee members remain confidential, an organized opacity on the entire hiring process.

It is by asking committee members about the reason for this organized opacity, which is clearly not a great help for candidates, that I discovered the legal-psychological concept of “not really illegal”, an amazing illustration of cognitive dissonance theory, certainly one of the most brilliant theories in psychology, developed by Festinger in the 1950s. It is worth giving some background information about this theory. It proposes that people have a drive to reduce cognitive elements that are dissonant. This can happen by different means, such as avoiding sources of information that are likely to be dissonant (e.g. speak with people with similar opinions), adding consonant elements, or changing one of the dissonant factors. Festinger first described the theory in an entertaining study of a small UFO cult who believed in imminent apocalypse, “When Prophecy Fails”. The group had predicted the apocalypse for December 20. They waited together on that day, but nothing happened. Subsequently, the group did not collapse. On the contrary, consonant elements were added to the current set of beliefs so as to deal with the new element. Two things happened: 1) they were now convinced that the apocalypse did not come because they had prayed so much, 2) they became more active in spreading their beliefs and thereby seeking social support from others. The theory was then presented in a more academic book, with a number of experiments to support it. The bottom line is that a consistent story is constructed a posteriori to justify judgments and beliefs so as to accommodate them with reality, and that the beliefs are strengthened by social support from a group of people sharing the same beliefs. The phenomenon is also well documented in philosophy of science (see for example Lakatos, Feyerabend and Kuhn).

So, committees know perfectly well why the hiring process must remain confidential and would tell you without any trace of embarrassment (in private!). It is to avoid lawsuits. They all have some cases in mind when candidates complained and could prove that they had suffered from what they and the justice considered as an unfair hiring process. To me and anyone else that is not in a committee, this means that confidentiality is imposed so as to hide illegal procedures, which is certainly not a good reason. But for most committee members I have talked with, this is not what it means. It is just to protect the hiring process from an endless series of lawsuits, which certainly would be a problem. They did not seem to see that a successful lawsuit means that the attacked procedure was illegal, and that perhaps the problem is rather the illegality of the procedure than the fact that there was a procedure.

I teased one committee member with a particular case which I am personally concerned about, which is that assistant professors are almost entirely denied access to senior research positions in CNRS. Indeed, those positions (“directeur de recherches” or DR) are officially open to all, but statistics show that more than 95% of successful applicants actually held a junior position in CNRS, with a very different statistics in all applicants.

This member, as other members I have talked with, recognized that it was indeed a problem for me and many other people. However, also like other members I have talked with, he found that I was being really extreme by complaining that it was illegal to use my current employment status as a criterion. Yet the legal texts on this subject are very simple and explicit, it is impossible to interpret them differently. In public hiring, there are so-called “external” positions and “internal” positions. Internal ones are reserved for promotion (there are a few in CNRS), external ones are open to anyone. Senior research positions are external, and advertised as such, with an explicit mention that no distinction will be made between internal and external candidates. I pushed the member a little bit with these arguments, but he could not admit that he was participating in an illegal procedure. This is where he came up with an amazing example of cognitive dissonance: the concept of “real illegality”. Alright, so the argument goes, perhaps using such criterion contradicts the law, but it is not really illegal. So here is what I understood of the concept of not being really illegal:

- something not really illegal contradicts the law;

- it is a problematic issue that “in an ideal world” (another interesting concept) should be avoided;

- it should remain confidential, for otherwise a legal suit would be successful;

- it is actually not considered illegal if it remains confidential;

- taking the not really illegal decision is not morally bad, but rather something that has to be done.

Thanks to the psychological concept of “real illegality”, committees seem to accommodate illegal substituted criteria quite well. In defense of these committees, I should add that, if what I heard is correct (likely, given the number of independent sources), this belief is actively supported by the direction of these institutions. However, committees are independent. They have no legal obligation to follow the orders or recommendation of the direction. That they nevertheless comply is a case of obedience to authority, another well known psychological bias illustrated by the Milgram experiment.