Does Workplace Competition Increase Labor Supply? Evidence from a Field Experiment

incentives matter

"...by measuring the effects of workplace competition on labor supply and firm costs using a field experiment in which workers are assigned to competitive and non-competitive payment schemes. Our research design enables us to rule out variation due to productive technology because workers are all hired to perform the same job under identical conditions. The only difference is compensation, where workers in the competitive scheme compete for a bonus prize. We also shut the channel of worker self-selection into competition by assigning workers to schemes and by not informing them in advance of the possibility of a bonus prize. Our main finding is that work time is substantially higher among workers competing for a tournament prize bonus of $30 to the worker with the highest output in their pre-defined 4- person work group. In that treatment, 55% worked the full 40 minutes and only 15% worked less than 11 minutes. Workers in the $30 bonus group also invested more effort per minute of work than those who were not offered the chance to compete for a bonus. However, this intensive margin response had a small impact on overall performance relative to the extensive margin response. Paying a bonus increased total employer costs per work group. To recover these costs, we needed total work time to increase by at least 30%. The average induced increase in effort and performance was greater than 80%. Therefore, our costs per unit of input (work time, effort) or output (task performance) were significantly lower (over 30%) when we paid a bonus....Thus, our results clearly show a strong causal relationship between performance-based incentive pay and work hours....A majority of workers in the treatment without explicit financial incentives worked past the minimum time, but awarding a tournament prize increased work time and effort by over 80% and lowered costs of effort or output by over a third. Effort was similar with alternative (piece rate, low-prize tournament) bonuses. Men worked longer than women in the high-prize tournament, but for the same duration in other treatments.....This implies that the competitive nature of many high-status and high-pay careers, in which workers compete against their colleagues for bonuses and promotions, is likely to be part of the explanation for their long work hours. Workers in these jobs are not typically paid on an hourly basis or for fixed time shifts. Rather, they have flexibility about how long to work, and can voluntarily provide working hours well beyond the standard workweek in order improve their performance rank. Since workplaces typically hire workers of similar ability levels, even the highest ability workers will need to invest long hours to distinguish themselves from (or not fall behind) their equally capable colleagues. Furthermore, our results indicate that competitive compensation schemes can be cost effective and therefore profitable for firms. If they generalize, the induced increase in output that results from the longer working hours (even if only a subset of workers is actively vying for a promotion or bonus) is enough to justify the costs of providing performance-based incentive pay. The profitability of competitive pay is the direct result of the labor supply effects we estimate and can explain its widespread prevalence.... When tournament incentives are cost-effective, it is natural that firms will offer them, and competitive jobs will entail long work hours. Therefore, our findings highlight a fundamental challenge for policy efforts aimed at reducing work hours in order to enhance overall worker well-being or improve gender equality in elite occupations. These policies may be rendered ineffective with pay incentives that encourage workers to voluntarily supply additional labor and circumvent formal hours restrictions.Moreover, policies effective at reducing hours may entail substantial costs to employers and some employees. Our field experiment also revealed a significant gender difference in labor supply, but only in the high-prize tournament. The fact that men worked longer in that tournament hints at a second channel through which high-stakes workplace competition can contribute to gender gaps in labor markets. The first is based on our overall result that indicates that competition can increase the work hours needed to succeed in elite professional careers. This disproportionately deters women, who traditionally have more binding time constraints from greater household obligations. The second channel is through differential effects of high-stakes competition on labor supply, even absent any differences in time constraints...."
Amalia Miller, Ragan Petrie and Carmit Segal, "Does Workplace Competition Increase Labor Supply? Evidence from a Field Experiment," National Bureau of Economic Research, June 2019, https://www.nber.org/papers/w25948