Jobseeker Attribute Selection Invariance
Lowering application barriers could potentially draw in systematically lower-skilled or less-qualified candidates, diluting applicant quality and confounding returns to search interventions. Evaluating whether marginal applicants differ in baseline education, work experience, or employer-graded CV scores is essential for policy evaluation.
Picture this
Imagine a university dropping its application fee. If the additional students who apply have the exact same test scores, grades, and teacher recommendations as the students who applied previously, lowering the fee increases application volume without reducing the overall qualification level of the applicant pool.
What the evidence says
Treatment effects on applications did not differ significantly by employer-graded CV quality (interaction p = 0.186). Mean predicted interview probabilities based on observed jobseeker traits were identical at 6.3% for inframarginal and 6.3% for marginal applications (p = 0.874).
- Who
- N = 1,477 jobseekers with employer-graded CVs across 122,946 matches in Lahore, Pakistan.
- How
- Latent type compliance analysis and RCT regression models interacting treatment status with employer-assigned CV quality tiers (excellent, good, average/lower).
What to do
Expand application facilitation programs broadly without imposing strict pre-screening filters, as reducing initiation friction does not induce adverse selection in applicant qualification.
From the source
"Marginal and inframarginal applications come from jobseekers with similar observed characteristics... mean education and CV quality scores... are almost identical for the marginal and inframarginal applications."
Why_Don’t_Jobseekers_Search_More_Barriers_and_Returns_to_Search.pdf