Do federal managers play favorites in career advancement?

The Merit Systems Protection Board recently issued a report, titled Fair and Equitable Treatment:  Progress Made and Challenges Remaining (wow -- the folks at MSPB seem to be taking a leaf from the title writers at GAO!)  that is attracting attention for a survey of federal employees asking whether supervisors show favoritism in making promotion decisions. 

An article on the survey by Alyssa Rosenberg in Govexec.com has drawn 109 comments (!) online, as of this writing, and is listed as one of the Web site's most popular articles. MSPB conducted the survey in 2007, and found that 72 percent of federal employees said promotions were based on who they knew, while just 40 percent cited competence as a reason for advancement, and 36 percent believed the key factor to be hard work.
 
There are really good reasons to take these findings with a huge grain of salt in terms of how well they represent the actual behavior of supervisors. A really well-established finding in social psychology is that people have "positive illusions" (to use the phrase of Professor Shelley Taylor of UCLA, who has investigated the phenomeon most extensively) about themselves. In area after area, most people -- generally around 70 percent --  think they are above average, which of course is statistically impossible. (Technically, 70% could be above average if "average" means the mean, but it's a safe guess that most people are intuitively answering the question using the median as their benchmark -- the performance of the person in the fiftieth percentile of the population.) Most people think they are above-average drivers and above average in interpersonal skills. And a simliar 70 percent of employees believe they are above-average employees.

These (honest) perceptions create a built-in problem for any workplace decisions -- including promotion decisions (and pay for performance) -- that require managers to affirm that some employees are better than others. Most people turned down for promotion genuinely believe they are above-average and deserved to be promoted. So it's natural that when they aren't promoted, they blame something other than that they weren't as good as the person who got the job. The most obvious candidate is "favoritism."

An indication of how inaccurately many perceive the grounds for promotion is the finding from the survey that 31 percent of respondents thought promotion decisions were influenced by nepotism, which must be absurd given the tiny percentage of supervisor relatives who work in workplaces. Although I predict that this blog post will incite disagreement and maybe even insults, I still believe that nobody should conclude from these survey results that favoritism is a rampant problem in federal workplaces. Sure, one may assume that most supervisors like the person selected for promotion more than the ones they didn't choose, but in most (though obviously not all) cases, a supervisor has no real reason to like or dislike an employee other than job performance.
 
There are two caveats to this. First, don't take these thoughts as a ringing vote of endorsement for federal supervisors. As I suggested in a recent FCW column, the quality of first-line supervision is often a real problem in the government, and a target for reform. Surely there are supervisors who really do play favorites, without concern about workplace performance, although I am confident the problem is of nowhere near the magnitude this employee survey suggests. 

The second caveat is that positive illusions are a fact -- people genuinely do often think they are better than they are -- and therefore must be taken into account in designing workplace systems. This suggests, for example, that employee evaluations (and possible pay for performance) be based on objective performance measures to the extent possible, rather than subjective judgments about hard work or good attitude. It is probably also the reason why many local governments have, unlike the federal government, traditionally based promotions on objective factors such as length of service, which, however, I think is a cure worse than the disease.