A Supply Side Rationale for Wage Floors: Evidence on Worker Collusion

Project Outcomes Statement

We implement and analyze the results from a field experiment to test whether communities of rural agricultural workers coordinate to improve their collective market wages, even in a context of daily informal spot contracts and no formal unions. We ask if workers adhere to a social norm during times of unemployment, refusing to work below the prevailing wage. In our empirical setting, workers are often bound by strong social and economic ties. For example, workers are dependent on each other for socialization, informal insurance, and job referrals. We argue that this creates the scope for communities to deter self-interested deviations.

In the experiment, we partner with local agricultural employers in seasonal periods of high unemployment to make job offers for one day of work. All jobs correspond to actual employment on the employer's land. The offers are made with cross-cutting, random variation in both the daily wage and the level observability to others in the community. First, the job is offered at a random wage level: at, or 10% below the prevailing wage. Second, we vary the extent to which the wage is publicly observable. In the "private" condition, the job is offered inside the worker's home and consequently is not directly observable to others. In the "public" condition, the employer offers the job outside on the street where neighbors, who are typically other workers, can overhear the offer. This allows us to obtain estimates of "private" versus "public" labor supply at different wage rates.

At the prevailing wage, 26% of workers offered the job take it. This take-up rate does not differ substantively with observability, suggesting that, when offers accord with the wage norm, observability in and of itself does not affect workers' decisions. In contrast, when a worker is offered a job below the prevailing wage, take-up depends crucially on whether his decision is observable to others. When a wage cut is offered in private, take-up remains a robust 18%. However, the willingness to accept falls by 13.6 percentage points (74%) when offers are publicly observable.

We collect two additional sets of data to examine whether accepting wage cuts has social consequences. First, we provide survey evidence on the nature of potential consequences. 90% of workers state that others would be angry with someone who accepted a wage cut. Moreover, respondents report that those who do not adhere to the norm may experience a decrease in referrals for future work.

Second, to construct a test for whether workers desire to sanction those who accept wage cuts, we build on our field experiment to design an anonymous costly punishment game. In the game, the decision-makers are randomly-chosen workers ("players") who were not offered jobs. Each player is paired with another (anonymous) worker who was offered a job with a local employer, as per our field experiment protocols. The player is told whether his paired worker: (i) accepted a job at the prevailing wage or at 10% below the prevailing wage, and (ii) lives in the player's own village or a village that is far away. The player can give up some of his participation payment to partially reduce the payment to his paired worker, thereby "sanctioning" the worker. The payments to the paired workers are unexpected (and again anonymous), so sanctioning results in a smaller windfall bonus. As expected, there is no sanctioning of workers who accept jobs at the prevailing wage. In contrast, 37% of players choose to pay to sanction those who accepted a wage cut. Surprisingly, sanctioning occurs even if the paired worker is from a distant village.

While our results are consistent with the prediction that a norm against accepting wage cuts could exacerbate unemployment in economic downturns, wage floors could still benefit workers overall. If the employers exert market power (monopsony), wage floors such those in our setting could increase both the overall surplus allocated to workers and could increase average employment levels. Even in the absence of monopsony power, we show in a simple back-of-the-envelope exercise that wage floors can allow workers to claim substantially more surplus from employers, with minor potential efficiency costs.

Our results on informal cooperation in markets are potentially more general than our specific empirical context. Our proposed mechanism hinges on two features: a clearly defined norm (with observable violations) and a mechanism to impose social sanctions (in order to prevent individual deviations). These features can arise naturally in many settings with repeated interpersonal interaction.