Calibrating Process and Outcome Accountability Systems in the Workplace to Meet Fairness and Efficiency Goals
(Original project title: What Companies Need to Do to Curb Bias in Employment Practices: A Multi-Method Exploration of the Perceived and Actual Efficacy of Process Versus Outcome Accountability)
Funded: November 2009 Completed: June, 2012
Philip E. Tetlock, Ph.D., University of Pennsylvania; Gregory Mitchell, Ph.D., J.D., University of Virginia; William T. Self, Ph.D., University of Missouri, and Barbara Mellers, Ph.D., University of Pennsylvania, J. Angus D. Hildreth, Ph.D., University of California, Berkeley
This project used an experimental design to explore the impact of process and outcome accountability on racial and gender biases in interview decisions. (Process accountability refers to holding managers accountability for following set procedures, whereas outcome accountability refers to holding managers accountability for meeting specified goals without regard to the process used to reach those goals.) The experiment examined the impact of process and outcome accountability manipulations on managerial interviewing decisions in a climate where discrimination against women and minorities was forbidden and managers were supposed to hire well-qualified applicants.
Where managers operated under outcome accountability constraints, fewer well-qualified applicants were selected for interviews and managers exhibited bias against white male candidates. Where managers operated under process accountability constraints, they avoided bias against women and African-Americans and made more meritocratic decisions. Managers reacted more negatively to outcome accountability than to process accountability.
Key Findings and Implications for Practice
• First, under outcome accountability, participants over-corrected in their biases more than under process accountability or no accountability, creating bias against white-male candidates. This bias became especially pronounced in the experimental condition in which traditionally disadvantaged minority groups suffered from human-capital deficits.
• Second, the pro-minority biases under outcome accountability resulted in the selection of less-qualified candidates across labor pools. These results are consistent with prior research suggesting that outcome accountability results in more cursory information processing and a shift in attention from job-relevant qualifications to job-irrelevant demographic characteristics.
• Third, participants reacted negatively to strong accountability instructions, reporting feeling less trusted in the outcome-accountability condition. These results suggest that strong accountability pressures can carry a social cost for organizations and lead to feelings of distrust among HR managers.
• For organizations seeking to debias personnel processes rather than achieve hiring targets, process accountability is likely to provide a more effective intervention.
• Process accountability resulted in better calibrated bias mitigation, reducing bias without prompting over-correction. This calibration had performance advantages with process accountability yielding a better-qualified pool of selected candidates.
• Additionally, unlike under outcome accountability, process accountability did not create a sense of distrust.
• If companies are operating in a labor pool in which there are large human-capital deficits in traditionally disadvantaged communities, if the companies' managers score high on indicators of egalitarian values (and display little or no baseline bias against traditionally disadvantaged groups), and if the companies embrace outcome accountability, our data suggest that these companies may pay a high price in workforce qualifications and potential efficiency – as well as a price in employee dissatisfaction.
Two hundred and ninety-seven undergraduate business students participated in a laboratory experiment in which they were asked to assume the role of a human-resources manager at a large company. They were informed that part of their job involved screening potential candidates for position openings at the company: they were to determine which candidates would be invited to interview for open positions. In the study task, participants were told they would review summaries of 75 applicants for an open position as an entry-level manager at a regional office. Participants were randomly assigned to a process-accountability, outcome-accountability, or no-accountability condition.
Download the full report.
View the full list of SHRM Foundation funded research.