Discrimination distorts talent allocation

Dominant (mainstream) economic theory is replete with studies and research on gender discrimination. The interest of economists in this topic stems from the fact that unequal treatment, all other things being equal, is not only aproblem for women, but is above all a problem for the economy, because it produces an inefficient allocation of human resources.

The problem arises when individuals with the same production characteristics have different probabilities of recruitment and promotion, or receive different wages, and these differences are systematically correlated with a non-economic characteristic of the individual such as the gender affiliation (Stiglitz 1973). In particular, the most recent literature points to the fact thatdiscrimination distorts the incentive structure and jams the talent allocation mechanism. Daron Acemoğlu, one of the three economists who have just received the Nobel Prize, states that discrimination represents a deviation from efficient and meritocratic market conditions, and emphasises the importance of a well-designed incentive structure in order to optimise talent allocation; inappropriate incentives may indeed distort occupational choices and hinder individual careers, but above all they reduce a country's development prospects.

In modern labour markets, where all agents are different from each other and all relevant information is private, the optimal match between individuals and jobs is hampered by the fact that competition for careers is unfair, becausethe evaluation of job performance is influenced by the bias of the evaluators. In particular, thetournament theory (Lazear 1998) defines unfair competition when the contenders, although of equal ability and applying themselves with equal effort, are not equally likely to win the tournament; in this case the match to the job position is suboptimal, the talent of the agents is not revealed and the social welfare is less than the potential.

In real labour markets, discrimination occurs when observable personal characteristics of individuals are not perfect predictors of productivity, and stereotypes of employers and evaluators find ample space in the gaps resulting from these information gaps. It is important to emphasise that biased behaviour is not intentional, conscious, and deliberately intended to harm. On the contrary, it is unconscious, automatic, out of the control of the decision-makers, and is often contradictory to their conscious intentions and values. Despite good intentions, however, stereotype conditioning is proven beyond reasonable doubt by thirty years of research in experimental economics that proves the existence of gender discrimination, i.e. unequal treatment of the female workforce. It is precisely experiments that make it possible to approximate the condition of ceteris paribus because, for example, when the same CV is alternately evaluated with a man's name or a woman's name, the differences observed in the responses highlight the disparity of treatment of people with the same productive characteristics but different genders (Bordalo et al. 2019).

The literature on stereotypes shows that talent and merit do not simply emerge in response to the intention of the decision maker and that goodwill is not enough to counteract the effects of prejudice. Since stereotypes arise from an information vacuum, the tools that can counteract them are: information, training, knowledge, disciplinary knowledge, research, data analysis, the scientific method, and so on. Hiding the gender affiliation of the subjects to be evaluated from the evaluators also seems to be a rather effective remedy because when selection is 'gender-blind', the results of the female component improve markedly. This is demonstrated, for example, by Goldin and Rouse's research (2000) on the recruitment of musicians in large symphony orchestras in the United States: the probability of a woman getting a place increases by 50% when the auditions are screened, i.e. when the evaluators have no information about the gender of the candidates.