FAYETTEVILLE, Ark. — A merit pay raise is like pornography: Most people can’t define it, but they know it when they see it. Not only can University of Arkansas researcher Nina Gupta define a merit pay raise, she has evidence that it may be different for men and women.

"Our earlier work showed that there is a threshold for merit raises of about seven percent," explained Gupta, professor of management in the Walton College of Business. "Calling a raise of less than that a merit increase is actually damaging to employee morale. It is not neutral — it is dysfunctional."

Gupta, along with associate professors Jason Shaw of the University of Kentucky and Atul Mitra and Susan Wurtz of the University of Northern Iowa, wanted to know if gender or personality had an impact on that threshold. The results of their studies are being presented this week at the 2002 Annual Meeting of the Academy of Management in Denver.

In addition to gender, the researchers looked at two personality traits — self-esteem and locus of control — to determine if there was a relationship with perceptions of pay raises. A person with an internal locus of control is more confident, task-oriented and realistic, while one with an external locus of control tends to be less confident and believes more in fate or luck.

The researchers conducted two studies, an experimental study with 192 participants and a field study with 166 participants. "We found that locus of control appeared to have no relationship to merit pay raise thresholds, but both gender and self-esteem seem to have an impact," Gupta said.

The experimental study involved hiring students to do basic coding, transferring data from a survey sheet to a scannable computer card. Recruited through fliers and ads in the student newspaper, the students were started at one of six pay levels and asked to work for one 3.5-hour time block.

Students were initially screened for personality characteristics and given a task to complete. After 30 minutes, they were given a questionnaire on their perceived effort, performance, speed and accuracy. Each person was "evaluated" by a researcher and told that the work was satisfactory and given a randomly assigned pay raise.

"It is important to note that we did not present the raises as percentages or absolute dollar amounts. We did not say, ‘you are getting a $2 raise’ or ‘you are getting a 10 percent raise.’ We just said, for example, ‘for the next hour you will be paid $6.80’," Gupta explained.

This ensured that the student workers would undergo the mental processes necessary to evaluate a noticeable difference. The students were debriefed after they had completed the simulation and all subjects were given the maximum amount possible for their participation.

The field study was conducted in a university hospital and included all types of employees, from nurses, physicians and laboratory technicians to housekeepers, staff associates and administrators. Participation was voluntary and participants were assured of confidentiality.

The study was conducted on all three shifts over a five-day period. Four months prior to a scheduled merit pay increase, 392 employees completed questionnaires during their lunch break. Four months after the pay increase, a second round of surveys was completed by 370 volunteers. Data was used from the 151 employees who participated in both surveys and had publicly available salary information.

"While results of the field study indicate that only self-esteem significantly predicts pay raise thresholds, the experimental simulation also suggests that gender is a significant predictor of perceptions of changes in pay," Gupta explained. "The results of the field study also suggest that high self-esteem employees require a bigger pay increase to consider the raise to be large or be pleased by it."