GPA Predicts GPA. It Doesn't Predict Much Else.

GPA cutoffs remain common in hiring, particularly for new graduates and professional services roles. The typical justification is that grades reflect intelligence, conscientiousness, and the ability to meet standards.

This is not entirely wrong. But the validity evidence for GPA as a job performance predictor is weak enough that using it as a primary screen deserves scrutiny.

What the research shows

The meta-analytic evidence on GPA and job performance is not encouraging. Roth et al. (1996) found a validity coefficient of .16 for the GPA-performance relationship, which is low by selection standards. More recent work has found similar results, with some studies finding coefficients near zero for roles more than a year after graduation.

Why does GPA predict so weakly?

• GPA reflects academic performance in a specific context: structured curricula, defined assignments, graded assessments. The skills required to perform well in that context overlap only partially with the skills required to perform in most jobs.
• GPA is highly field-dependent. A 3.8 in a rigorous engineering program involves different demands than a 3.8 in a less demanding curriculum.
• Grading varies across institutions and professors. GPA is not a standardized measure.
• GPA captures a four-year average that does not reflect trajectory, effort, or growth.

Why it still gets used

GPA is still used because it is cheap to collect, easy to compare across candidates, and feels like an objective standard. It reduces the volume of applications that need to be reviewed. For organizations hiring hundreds of new graduates, it functions more as an administrative filter than a predictive tool.

It also correlates with cognitive ability and conscientiousness, both of which do predict job performance. GPA works as a proxy for these constructs, but it is a weak proxy. A direct cognitive assessment or conscientiousness measure produces substantially higher validity than GPA.

The other reason: GPA is familiar. “Minimum 3.0 GPA” is easy to write into a job posting. It does not require a validation study. Organizations that have used it for twenty years have no strong incentive to question it until they run into legal trouble.

The adverse impact problem

GPA cutoffs produce significant adverse impact in most applicant pools. GPA correlates with socioeconomic status, which correlates with access to tutoring, campus resources, and the ability to study without working multiple jobs. Race correlates with socioeconomic status due to historical and structural factors.

Using GPA as a hard cutoff removes a disproportionate share of underrepresented candidates from the pool before the first interview. If the validity of the criterion is .16, that adverse impact is difficult to justify as business necessity.

What to use instead

The alternatives that outperform GPA on validity:

• Structured behavioral interviews: validity .50 to .60, measures the competencies that actually matter for the role
• Work samples and job simulations: high content validity, directly measures capability in the relevant domain
• Cognitive ability measures: validity .50 for general cognitive ability, though these also produce adverse impact and require careful implementation
• Structured skill assessments: appropriate for technical roles where specific knowledge is a job requirement

None of these are as cheap or simple to implement as a GPA filter. They are, however, substantially more predictive of who will perform well. For organizations serious about both quality of hire and legal defensibility, the tradeoff is worth making.