Artistic labor markets are puzzling. In most advanced countries, census data provide quite similar pictures about artists' earnings : artists actually appear to get lower returns from their educational investments than is the case in other comparable occupations, to have more variable income across time and to experience huge income inequalities. Nevertheless artists are not deterred from entering such an occupation in growing numbers, nor is there as much withdrawal from artistic careers as would be expected.

Given that these labor market disparities persist across time, any type of usual labor market disequilibrium is unlikely to be the cause. I’ll review several alternative explanations: risk-loving, miscalculation due to overestimation of one’s chances of success, psychic rewards, impact of work contingency. Then I’ll turn to the issue of talent : art worlds have developed an insuperable engine to rank artists by quality level and market value, to select and signal the best works out of an ocean of products through endless competitive comparisons, to let the whirl of fads and fashions promote or eliminate aspiring superstars, to celebrate skyrocketing and ephemeral fame. I’ll ask under what conditions inequalities of such a magnitude may be occupationally legitimate.