Dynamic effects of affirmative action: Experimental evidence

Abstract

This paper studies the impact of Affirmative Action (AA) policies on workforce sorting in competitive payment schemes, workers’ productivity, and firms’ performance. Using a laboratory experiment, we analyse an environment where participants repeatedly choose between two firms with tournament-based pay or a firm offering piece-rate compensation. Experimental treatments vary based on whether participants can switch firms each period and whether one of the tournament-based firms implements AA policies. Our findings reveal that AA significantly boosts female participation in competitive settings without reducing male engagement or productivity, effectively closing the gender gap in tournament selection. However, AA firms face mild talent retention challenges. Importantly, AA induces strong sorting by gender and productivity, with highly productive women preferring AA firms and highly productive men opting for competitive environments in which AA is not in place. These findings suggest that while AA policies successfully attract more women to competitive workplaces, they may also contribute to gender-based clustering – where men and women self-select into different firms based on the presence of AA – while maintaining overall organizational efficiency.