When Performance Pay Backfires: Managing Incentives in Complex Work

Categories: Economics

Reading Time:

6–8 minutes

Performance-based pay has long been a default tool for motivating employees. In straightforward environments, it works well. Paying auto glass installers per windshield reliably increases output because the link between effort and results is clear.

In complex, high-skill roles such as teaching, healthcare, social work and public administration, the same logic often breaks down. Incentive programs tied to outcomes like test scores or patient metrics frequently produce disappointing or unintended results. Teachers may narrow instruction to test preparation. Clinicians may focus on measurable tasks at the expense of less visible but critical care. The issue is not motivation, but uncertainty.

Research¹ cowritten by Mike Kofoed, assistant professor in the Department of Economics at the University of Tennessee, Knoxville, Haslam College of Business, shows that output-based incentives fail in complex jobs because workers often cannot tell which specific actions will drive the outcomes on which pay depends. Multiple inputs influence results, and their impact varies across situations. A teacher knows both higher-order instruction and test prep affect scores but cannot predict which will matter more for a particular group of students in a given year. A physician understands that diagnosis, communication and coordination all affect outcomes, yet cannot precisely forecast which action will most influence a specific patient with multiple conditions.

When pay depends on uncertain outcomes, two predictable patterns emerge. First, effort shifts toward safer, lower-variance tasks. Workers gravitate to actions that feel more predictable, even if those tasks create less value. Second, overall effort can decline. If the connection between effort and earnings feels unclear, additional effort may seem futile. These responses are rational from the worker’s perspective but inefficient for the organization.

To study this dynamic, the researchers designed an experiment simulating a multi-task job under uncertainty. Participants chose between easy, low-value math problems and harder, higher-value ones, earning real money based on performance.

In a control setting, payoffs for each task were clearly displayed. Participants quickly learned which tasks generated the most earnings per minute and shifted effort efficiently.

In the uncertainty setting, payoffs varied randomly within known ranges, but the exact value was revealed only after each round. This mirrors many real jobs, where average relationships are known but case-specific outcomes are not. Under these conditions, participants consistently shifted effort toward the more predictable task, even when it was less profitable on average. As payoff uncertainty rose, effort became less responsive to actual value.

For managers, the implication is clear: Production uncertainty pushes employees away from the very tasks that generate the most value. Output-based incentives can unintentionally steer effort toward what feels safest rather than what matters most.

Several leadership principles follow:

Assess line-of-sight – output-based pay carries high risk is employees don’t clearly understand which actions drive the measured outcomes
Recognize uncertainty – when results depend on many interacting factors, tying pay heavily to outcomes shifts excessive risk onto workers
Use input-based incentives where appropriate – rewarding high-quality practices, processes or behaviors employees can control often produces better results than paying solely for outcomes
Blend input and output measures – combining process-based rewards with outcome goals maintains accountability while reducing unmanageable risk
Watch for behavioral drift – narrowing of effort, overemphasis on measurable tasks and avoidance of complex work signal incentive misalignment

Performance pay is not inherently flawed. In stable, well-understood production settings, it remains powerful. In complex, high-uncertainty roles, however, incentive design becomes a risk management decision. Compensation systems that account for uncertainty help ensure effort flows toward the organization’s highest-value work, not merely toward what feels safest under the pay scheme.

¹Kofoed, M. S., & Phipps, A. (2026). Unexpected costs of performance incentives with production uncertainty: Theory and evidence from a real‐effort laboratory experiment. Economic Inquiry, 64(1), 261-285. https://doi.org/10.1111/ecin.70021

Economics

Short Takes

How should states allocate treatment centers to reduce overdose fatalities and regional inequality?

A predictive optimization model shows that reallocating limited resources can simultaneously reduce deaths and narrow disparities: emphasizing total deaths favors dense counties, while emphasizing equity shifts resources toward underserved ones.

States can use quantitative optimization to balance efficiency and fairness when allocating public-health infrastructure.

Based on “Reducing Overdose Deaths and Mitigating County Disparities: Optimal Allocation of Substance-Use Treatment Centers,” by Matthew Baucum, Matthew Harris, Lawerence Kessler and Guanyi Lu, published in Manufacturing & Service Operations Management.

Do export-promotion programs shape firm behavior beyond increasing export volumes?

Export support boosts firm-level exports within existing markets, but it does not alter prices, quality, markups or marginal costs. This pattern shows that firms grow by shifting demand rather than by enhancing productivity or improving products.

Export-support programs are tools for market access and demand generation, not methods to automatically improve product quality or reduce costs. If the goal is to strengthen competitiveness through innovation, quality improvement or productivity growth, export support should accompany intentional investments in capabilities instead of being relied upon to achieve those results on its own.

Based on “Do Export Support Programs Affect Prices, Quality, Markups and Marginal Costs? Evidence From a Natural Policy Experiment” by Magnus T. Buus, Jakob R. Munch, Joel Rodrigue and Georg Schaur, published in the Review of Economics and Statistics.

How costly are border-processing delays for importers?

Very costly. Each extra day of customs processing in Peru imposes a burden equivalent to an 18 percent tariff. Adoption of the World Trade Organization Trade Facilitation Agreement measures significantly reduces these costs.

Streamlining administrative processes can generate welfare gains exceeding those from tariff liberalization, highlighting logistics reform as a robust trade policy tool.

Based on “Import Processing and Trade Costs,” by Jerónimo Carballo, Alejandro G. Graziano, Georg Schaur and Christian Volpe Martincus, publsihed in the Journal of International Economics.

Does uncertainty about competitors change how hard teams work in competitive settings?

Uncertainty about rivals’ costs or how much they want to win can boost effort. However, uncertainty about group size does not have the same impact. People may also put in more effort than standard self-interest models would suggest.

Information disclosure is a tool. Careful management of what participants know about competitors can increase effort, but it can also lead to unnecessary over-investment. Managers should match the contest design with the value of extra effort. 

Based on “Who Are We Up Against? Heterogeneous Group Contests with Incomplete Information,” by Vasudha Chopra, Hieu M. Nguyen and Christian A. Vossler, published in the Journal of Economic Behavior & Organization.

What happens to teen birth rates when high school access expands?

Mexico’s national education reform lowered teen birth rates by 2.8 percent in high-exposure municipalities. The decline stems from heightened aspirations and perceived opportunity.

Educational expansion is an effective social policy that improves both economic and health outcomes for future generations.

Based on “Improvements in Schooling Opportunities and Teen Births,” by Lucas N. Garcez, María Padilla-Romo, Cecilia Peluffo and Mayra Pineda-Torres, published in the Journal of Economic Behavior & Organization.

How can we model long-term memory in a rational optimization model when information decays?

Memories decay over time unless individuals expend costly effort to preserve them. This means there is a bias toward recent memories, and performance declines as the delay between learning and action increases.

Organizations should reduce the lag between information exposure and decision-making and design information systems that preserve older, relevant signals to counteract excessive recency bias.

Based on “Rational Memory With Decay,” by Nathaniel Neligh, published in the Journal of Economic Behavior & Organization.

Does increasing school-day length influence family stability and labor participation?

Yes. Extending the school day by 3.5 hours increased divorce rates in municipalities with greater exposure, driven by expanded female labor participation and shifting household dynamics.

Childcare and education reforms affect family structure. Social policy should anticipate how time allocation influences household decision-making.

Based on “Parents’ Effective Time Endowment and Divorce: Evidence from Extended School Days,” by María Padilla-Romo, Cecilia Peluffo and Mariana Viollaz, published in the Journal of Public Economics.

Limit information disclosure such as past winning bids to reduce profitable deviations and protect seller revenue in sequential auction design.

Do bidders have strong incentives to follow equilibrium strategies in sequential auctions?

In sequential Dutch and English auctions, a class of “mimic” deviations yields flat expected, implying weak incentives to adhere to equilibrium strictly. This fragility complicates learning and sustaining equilibrium play. 

Based on “Mimic Martingales in Sequential Auctions,” by Matt Van Essen and John Wooders, published in Economic Theory.

Contact

Michael Kofoed

Research Assistant Professor

[email protected]

Haslam College of Business - When Performance Pay Backfires: Managing Incentives in Complex Work