Principal investigator: Viviane Pires Ribeiro
Article title: Self-Control at Work
Article authors: Supreet Kaur, Michael Kremer and Sendhil Mullainathan
Location of the intervention: India
Sample size: 102 workers
Sector: Job market
Type of Intervention: Effects of self-control among workers
Primary variable of interest: Self-control at work
Evaluation method: Other – Field research
Evaluation Context
The theory of the firm emphasizes a tension between workers and companies. This is because employers provide security to employees, and these, in turn, do not fully benefit from their own efforts, thus creating moral hazard. Workers with self-control problems do not work hard enough from the employer's point of view, directly harming the company's profits. As a result, both the company and the employee have an interest in overcoming these problems.
The workplace exists to organize the provision of effort by employees. The same characteristics that mitigate moral hazard—incentive contracts and work resources such as fixed hours—can also mitigate self-control problems. That is, the employer has the means and the motives to implicitly provide commitment mechanisms.
Intervention Details
To assess the empirical relevance and magnitude of temporal inconsistency, Kaur et al. (2015) studied an Indian data entry company in Mysore. Using the company's infrastructure – office space, software, and operational protocols – the authors designed a field experiment over 13 months with a sample of 102 workers.
At the data entry company, employees add, verify, and edit electronic data, working full-time and being paid based on the number of accurate fields entered each day. The workers analyzed used data entry software to type information into fields of scanned images. Following standard practice, workers were paid a piece rate of Rs. 0,03 for each field entered, plus a small daily rate of Rs. 15, which constituted about 8 percent of their compensation.
Methodology Details
Kaur et al. (2015) constructed a simple model to derive empirically testable predictions by distinguishing worker behavior over time. In firm models, employees are compensated with incentives when faced with extra risk. Thus, workers who are aware of their self-control problem value more pronounced incentives as a form of motivation for the future.
Employees were offered a daily choice between two types of incentive contracts. The first is a flat-rate piece-rate contract. The second is a dominated contract, which pays less than the other contract for low production levels, but pays the same amount for high production levels. Although the choice of contract is made daily, workers are paid weekly: on a specific day of the week, they receive their cumulative earnings from the previous seven days.
In addition to predicting the demand for dominated contracts, the model suggests that payday affects effort. As payday approaches, the source of the self-control problem diminishes: the rewards of work and the cost of work are closer in time. As a result, output should increase. The model also suggests an important role for heterogeneity. That is, workers with greater self-control problems show greater effects on payday and a greater desire for dominated contracts.
Results
The results found by Kaur et al. (2015) suggest the quantitative importance of self-control at work. Firstly, employees choose dominated contracts – which penalize low production but do not provide greater reward for high production – on average, 36% of the time. The use of these contracts increases production by the same amount as an 18% increase in the piece-rate wage.
Secondly, to test the impact of payday, workers were randomized into different payday groups, i.e., all were paid weekly, but with varying paydays. Worker output was 8% higher on paydays than at the beginning of the weekly pay cycle. An effect of this magnitude corresponds to a 24% increase in piece rate.
Thirdly, the authors found substantial heterogeneity in the extent of the effects of contracts and payday. Workers with above-average payday effects were 49% more likely to choose dominated contracts. Providing these workers with the option to choose a dominated contract increases their output by 9%.
Fourth, the option to choose dominated contracts has greater treatment effects when payday is far in the future. This is consistent with the fact that the self-control problem is smaller when it is close to payday, and the dominated contract therefore has less scope to affect effort. Evidence of learning was also found. As workers gain experience, the correlation between payday effects and the choice of dominated contracts increases. After 2 months of experience, workers with high payday effects are 20 percentage points (73%) more likely to select dominated contracts than workers with low effects.
Lessons in Public Policy
What is the importance of self-control at work? The results found by Kaur et al. (2015) indicate strong evidence that self-control problems distort employee effort to economically significant magnitudes, and these workers will require incentives to help them overcome these problems; such incentives can take a variety of forms. In the model developed by the authors, the dominated contract helped solve the self-control problem through high-powered incentives around a discrete threshold. Many companies, for example, provide bonuses to employees who meet targets or minimum production levels. In some cases, employers remove the worker's ability to choose certain dimensions of effort through rigid hours or, in more extreme cases, assembly lines that make it impossible to slow down. Therefore, self-control problems among workers can lead companies to adopt incentives or impose work rules to allow for monitoring of employee effort.
References: KAUR, Supreet; KREMER, Michael; MULLAINATHAN, Sendhil. Self-Control at Work. Journal of Political Economy, v. 123, no. 6, p. 1227-1277, 2015.
