Why Excellent Performance Evaluations Generate Raises for Steve but not for Jennifer

tilt deskThat women are paid less than men at work is not news.  In the U.S., a gender pay gap is nearly ubiquitous.  Even in Hollywood.  We’ve figured out the gender pay gap is largely an artifact of women and men working in different jobs (and the jobs men work in simply pay more than the ones women work in). Yet when women and men work in similar jobs, men still tend to earn more than their female counterparts. What accounts for this difference?

A forthcoming article in the Academy of Management Journal by Aparna JoshiJooyeon Son, and Hyuntak Roh explores the way in which this job reward difference might be partly a result of gender differences in the way performance evaluation are tied to reward.  Their study is important, since very few scholars have paid much attention to performance evaluations. This is partly because it’s hard to get access to workers’ performance evaluations. And it’s partly because performance and performance metrics are often specific to particular jobs or firms, and thus make it difficult to draws comparisons across firms. What Joshi, Son and Roh have done, though, suggests that a focus on performance evaluation is exactly right; evaluations are directly linked to work reward.  In fact, their findings demonstrate how and where gender differences in pay can be found.  For this reason, the piece has obvious relevance for readers who seek not only to study gender differences at work. But it also warrants close attention from workers who may actually not know how their performance and evaluation connect (or in some cases, do not). Employers too will want to read this study, to calibrate their performance and reward systems more fairly.

The researchers performed a meta-analysis of studies conducted between 1985-2013 that reported on the relationship between performance evaluations and compensation among male and female workers. Their research encompassed a total of 190 effect sizes and 474,732 individuals from 142 studies.  So basically, if a study that had anything to do with the connection between an individual’s sex, performance evaluation, or work rewards and was written in the past roughly 30 years, they authors took a look at it.

Two major results leap off the page: First, gender differences in salary, bonuses, and promotions were fourteen times larger than gender differences in performance evaluation. And second, differences in performance evaluations could not explain  the gender gap in reward.  In other words, it was not that women systematically under-performed relative to men (in fact, they found no significant difference in the performance of women and men holding similar jobs).  What happened instead was that employers systematically under-rewarded women who performed relatively similarly to men.

The connection between gender and performance evaluation was context dependent.  Men tended to get higher performance evaluations than women in low-prestige, male-dominated occupations in industries with few female executives. Men also received higher evaluations when they held complex jobs in prestigious occupations within industries where women were severely under-represented among executives.   Women tended to get higher performance evaluations in jobs requiring routine tasks and when either their occupation was gender-balanced or their industry had a high representation of female executives.

Perhaps the most interesting connection is that between performance and reward. In prestigious occupations, women and men received similar performance evaluations, yet men were rewarded significantly higher than women.  The gender gap in the link between performance and reward grew in relation to the share of men in the occupation and the complexity of job being performed.  Women were rewarded at higher levels than men and received higher performance evaluations in only one setting: industries with a high proportion of female executives.

So, the “best” work setting for women–one in which their rewards are judged in much the same way as men’s—is in a routine job located in an occupation with a roughly equal share of women and men. Another “ideal” setting for women is in an industry with a high share of women in executive positions.  Try finding a gender-balanced occupation and one with routine job tasks in the U.S., though.  If “low skill” equates to routine, the choices are janitors and building cleaners and stock clerks and other filers. What about industries with a lot of female executives? According to the authors, there was one industry with more than 50% female executives: healthcare.

It does not look very promising for working women hoping to be rewarded the same as their similar-performing male colleagues unless employers do a few things.  Devise a better way to evaluate complex jobs in a way that does not emphasize the reward of stereotypically male traits (e.g., leadership, risk-taking, drive).  This would have to look something like removing the ambiguity of complex jobs. That could be done by asking people who do these complex jobs to carefully track what they do, compile those tasks over time and across individuals, and in this way form a baseline of what gets done on the job.

Here is an example of how job tasks might be compiled. It’s a composite of what faculty members actually do on an average  weekday.

academic tasks weekday averages

Source: https://thebluereview.org/faculty-time-allocation/ “The Long, Lonely Job of Homo academicus” by John Ziker.


Or this might mean including, as they authors suggest, a neutral accountability structure in the evaluation process (a role perhaps played by an HR manager) that could operate like a “checks and balance” system to ensure a match between performance and evaluation.  If a task is being rewarded for one group of workers in one way and in a different way for another, this manager would sound a warning and at minimum, require justification for the difference. Another approach would be: get women into management.  When they are there, by means we do not yet fully understand, the disconnect between women’s job performance and reward goes away.

The implications are very real. By ignoring the disconnect between women’s performance and reward, female workers stand to suffer from feelings of inequity and burnout. And, in the most severe cases, they leave paid work altogether.  That would be a shame –especially since it tends to be the case that women’s on-the-job-performance is just as good as men’s.

  1. grad student said:

    A valuable validation of theories presented in Castilla 2008 (AJS) and Castilla and Benard 2010 (ASQ)

  2. Pingback: Soji Sakura

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: