By Scott Kupor, Director, U.S. Office of Personnel Management
April 14, 2026

Named for the traditional practice of printing educational diplomas on parchment made from sheep hide, the “sheepskin effect” holds that individuals who, for example, complete a 4-year bachelor’s degree earn significantly more in the labor market than do individuals who complete 3.9 years of college, but fail to earn their degree. If the value of the education one received in college were the primary determinant of earnings, we wouldn’t expect much of a difference between these two individuals. But, in practice, the sheepskin effect demonstrates that the primary value of a college education is the degree itself, rather than the educational training one receives.

Why is that?

Well, there are lots of reasons – e.g., maybe people who stick out the final 0.1 years are more likely to persevere when times get tough – but I think the most compelling reason is the simplest: Hiring is hard and people are lazy!

Finding great people takes a lot of time – sourcing, interviewing, evaluating – so we take shortcuts. We look for proxies of skill that might ultimately predict success in the job instead of doing the hard work of truly evaluating the actual skills demonstrated by the applicant and determining their fitness for the task at hand. We use fancy monikers such as credentialing bias or signaling theory to make ourselves feel better, but the reality is it’s just sheer laziness. We’d rather rely on a university’s conferral of an academic degree as a shortcut for doing the hard work.

How many job descriptions have you seen that say: “bachelor’s degree required,” as if all bachelor’s degrees were created equal? A Russian Literature degree is not the same as a Computer Science degree – particularly if you are applying for an editing job at a Russian newspaper – and nor is a degree from Harvard University equivalent to one from the University of Phoenix Online (no offense to any Harvard grads).

And how many job descriptions quip: “a minimum of 5 years of work experience required,” as if all work experiences were created equal? Five years working on an oil rig may be more predictive of a worker’s future performance than is ten years working as a dexterous manipulator (aka, “handyman”). It depends on what experiences and skills you gleaned in each of those professions and how those skills relate to the job for which you are currently applying.

At OPM, we recognize that hiring for demonstrated merit matters and that proxies are just that – imperfect proxies for merit. So, we have begun a (too) lengthy process of re-writing all 604 federal occupational series – yes, there are really 604 – to ensure that any of these imperfect proxies for merit do not serve as gating mechanisms, and that skills-and competency-based hiring methods may be used to determine position qualifications. As a side note, we also aim to reduce the total number of series by at least ~25%!

While we are at it, we are also eliminating the long-standing government hiring practice of allowing candidates to provide their own self-attestation of their skills for the job. In its place, we are requiring hiring managers to use formal assessments – skills-based “tests” to determine one’s aptitude for the specific job skill requirements. For example, if you are applying for a software development job, you won’t just check a box stipulating that you are an expert Perl developer, but rather you will now complete a coding assessment to determine whether you have the requisite competencies for the role.

And today, we are rolling out the first of these new job classifications – the 2210 series of technical IT-related jobs in the federal government. For the first time, your fitness for the job will be determined via a formal assessment rather than based upon whether you have a bachelor’s degree or some minimum amount of work experience.

Why does this matter?

First, I hope it’s non-controversial that hiring decisions should be made on demonstrated merit. Relying on proxies for merit is unfair to candidates who may be well-qualified for the job based on their skills yet may have had the misfortune of being born into the wrong zip code and thus never had the chance to earn a sheepskin.

Second, many people complain that federal government salaries are too low to attract great people who may be otherwise attracted to private sector jobs. But this is only half-true.

Granted that the federal government does on average pay lower salaries than the private sector – and I suspect that won’t change anytime soon. However, an equally important comp-related issue in attracting people to federal roles is making sure they are leveled commensurate with their experience. Instead, because leveling in the government has been driven by degree requirements and minimum tenure requirements, we often cannot hire people at a compensation level appropriate to their actual skill set if they are lacking these credential proxies.

Ok – I am going to say it.

It reminds me of a quote I saw from an existing federal employee who objected to the employ of Edward Coristine (aka – well, you can Google it on your own time, at your own risk). The chief complaint of this employee was the fact that Ed was 19 years old and was a Northwestern University dropout. He told the reporter: “We never would have hired him in government because he didn’t have a college degree or enough experience.” Zero recognition of the fact that Ed was/is a world-class software developer. Instead, his lack of credentials was dispositive of his ability to do the job for which he was hired.

If we want the best talent in the federal government – which I think we do – we should not refuse to hire college dropouts, unless dropping out of college is somehow representative of their ability to work successfully on behalf of the American people. And, if they are in fact world-class engineers, then we should pay them at the level at which they are performing versus force-fitting them into a lower pay level because they have no prior work experience.

Once again, merit matters.