Complexity is an important aspect to consider when designing workplace incentive schemes as it can affect worker effort and performance, according to new research from the Cornell SC Johnson College of Business.
David Huffman, the Ashley Family Professor in Behavioral Economics at the Samuel Curtis Johnson Graduate School of Management, and Collin Raymond, associate professor at the Johnson School, used data from warehouse workers coupled with online experiments to examine how workers respond to complex pay structures.
"In a surprising twist, we found that for the firm we worked with, complexity in workplace pay systems can lead to better outcomes," said Huffman. "When incentives are hard to understand, workers often ignore parts of them, and in some cases, this 'opacity' can lead to more effort, higher efficiency and even better pay. Complexity may not always be beneficial to firms and workers, though, so it is important to carefully consider this aspect of incentives."
The article, "Incentive Complexity, Bounded Rationality and Effort Provision," published Dec. 1 in the American Economic Review. Johannes Abeler, professor of economics at the University of Oxford, is a co-author.
Most companies, the researchers said, implement incentive systems such as bonuses or performance-based pay to motivate employees. However, they don't always work in employers' favor.
For example, incentives that compare current performance to benchmarks established by past outcomes can sometimes bring about what economists call the "ratchet effect": If workers realize that working harder today will raise expectations tomorrow, they may slow down to avoid tougher targets. This can defeat the purpose of the incentive scheme because it leads to low effort, and it might even prompt the employer to not offer a bonus pay scheme at all.
But these systems can become confusing, especially if there are many moving parts, like pay rates and targets that vary over time, and this confusion can keep workers from fully considering how working harder today affects future expectations. The team sought to explore whether workers understand dynamic incentives, and what happens if they don't.
To do so, the researchers partnered with a large warehouse company and ran two major field experiments. One assigned newly hired workers to different pay schemes. Some had incentives that adjusted based on their own performance (creating a ratchet effect), while others had fixed targets.
To the researchers' surprise, workers facing the ratchet effect didn't slow down. In fact, their effort barely changed - just a 0.1% drop, which wasn't significant. This suggested that either workers didn't realize their current performance would affect future expectations, or didn't know how to take advantage of it.
In a second experiment, called the "group trial," workers' performance affected not just their own future targets, but also those of their peers. Even with time to learn and social pressure, the effect was still small - a 1% drop in effort over 10 months.
To understand why workers weren't responding to these dynamic incentives, the researchers turned to online experiments. They re-created the incentive systems in a controlled setting and tested how workers responded when the incentives were framed in different ways.
They found that when incentives were explained clearly and simply, especially in terms of money rather than incentive scheme jargon, workers adjusted their effort. But when incentives were framed in abstract terms like "target rates" or included random elements, workers didn't react.
"Complexity made the incentives opaque," Raymond said. "Workers didn't see the full picture, and so they didn't change their behavior."
The study also found that workers with higher cognitive ability, measured using a test called the Cognitive Reflection Test (CRT), were more likely to understand and respond to complex incentives. Those with lower CRT scores were less likely to notice or mention the dynamic incentives, even when asked directly.
This suggests that the effectiveness of incentive systems may depend not just on their design, but on the employee. Using economic models, the research team showed that when workers ignore the ratchet effect, they end up working harder than they would if they fully understood the system. Further, they found that if firms are aware that workers don't fully understand the dynamics of their contracts and therefore won't exploit them, they are willing to offer more attractive incentive schemes, potentially benefiting both employer and employee.
The findings challenge traditional views of workplace incentives and the assumption that workers always respond rationally to pay structures.
"For policymakers, the study suggests the importance of considering complexity when regulating incentive systems, but also the need for a nuanced understanding of its effects," Huffman said. "In some cases, simplifying contracts might reduce productivity, while in others, it could prevent exploitation. Our study raises deeper questions about fairness, transparency and the role of cognitive ability in how workers respond to incentives. As companies continue to experiment with performance-based pay, understanding how workers perceive these systems could be just as important as the systems themselves."
Sarah Magnus-Sharpe is director of public relations and communications for the Cornell SC Johnson College of Business.
