A UBI would increase discrimination towards women in the workplace
Robeyns ’01 (Ingrid Robeyns, PhD, Chair in Ethics of Institutions at the Ethics Institute of Utrecht University, “Will a Basic Income Do Justice to Women?”, Analyse & Kritik, 22, 5/1/2001, https://www.degruyter.com/document/doi/10.1515/auk-2001-0108/html?lang=en)
A last second order effect is an expected increase of statistical discrimination against women. Statistical discrimination is a form of indirect discrimination based on the fact that a person belongs to a group that has certain characteristics. These characteristics are used as proxies for the average productivity of that group. Women on average get one or two children and take maternity leave. They also work less hours on the labour market than men, and bear more responsibility for the household and the care of the children and the elderly. They have more career interruptions and are more absent from the workplace. For all these reasons, it is assumed by many employers that women are on average less productive than men. Furthermore, it is rational for an employer to think of an individual woman that she will share these characteristics of her group, as he has no information on her future commitments. Hence, an employer discriminates a woman (by not hiring her or giving her a lower wage) because the employer has no exact information on her productivity and therefore his perception of the average productivity of all women will count. A woman who does not want children, or wants to make a career or has a husband who takes half of the household responsibilities, will still bear the consequences of the fact that other women are more 'child and household oriented'-which is perceived as a good indicator for lower productivity. This implies that if a basic income is introduced, and a group of women decreases her labour supply, it will also be expected from other women that they will be less labour market attached. Hence, a basic income will probably worsen the statistical discrimination against women. One could argue that if statistical discrimination increases efficiency, this efficiency gain can be used for the worse off, for example women. If one follows this argument, statistical discrimination should not per definition be morally condemned. However, I strongly oppose to this argument. First, the little empirical research on discrimination that contains detailed and direct information on productivity (e.g. Wenneras/Wold 1997) shows that even when a woman is two times as productive as a man, employers might prefer to offer a job to him instead of her. I would therefore argue that the best assumption to start from is that statistical discrimination lowers aggregate productivity and efficiency. Nevertheless, this efficiency suboptimality can sustain because of three elements. First, it is rational for employers who have no information on individual productivity to use groups-based proxies. Second, perception biases (which can be formed by sexist, racist or other believes) can lead to a misfit between the observed proxies and real productivity. And finally, I see no reason why other considerations next to productivity (like a preference for employee homogeneity or preferences for certain types of people) would not interfere with the profit maximising strategies of employers hiring people. Furthermore, even if statistical discrimination would lead to efficiency gains, it should still be condemned on moral grounds. It violates the basic principle of equal concern and respect for all individuals. Therefore, I think it is a real challenge for theorists of justice, including basic income proposers, to find strategies to end statistical discrimination and integrate them in their theories.