Questions: Signaling Model

Question 1

In Spence's signaling model, the key assumption that makes education a credible signal is...

• A. Education is enjoyable for all workers equally
• B. The cost of acquiring education is lower for high-ability workers than for low-ability workers (single-crossing condition)
• C. Employers can perfectly observe worker ability before hiring
• D. All workers receive the same wage regardless of education

Answer: (B)

The signaling mechanism requires the single-crossing condition: education must be differentially costly across ability types. If acquiring a degree costs high-ability workers less (in time, effort, or psychic cost), then only high-ability workers will find it worthwhile to invest in the signal at the equilibrium education level. Low-ability workers face higher costs and choose not to signal. This cost differential is what makes education an informative (separating) signal — it reveals information precisely because it is not equally costly to acquire.

Question 2

In a separating equilibrium, both high-ability and low-ability workers are better off compared to a world without signaling.

• T. True
• F. False

Answer: False

In the pooling outcome (without signaling), all workers receive the average wage. In the separating equilibrium, high-ability workers earn a premium but incur signaling costs (tuition, time, effort). Low-ability workers earn a lower wage because they are now identified as low-ability. Whether high-ability workers are better off depends on whether the premium exceeds the signaling cost — they may or may not be. Low-ability workers are unambiguously worse off because they lose the pooling subsidy. From a social efficiency perspective, if education is pure signal, the resources spent on it are wasted — they produce no output.

Question 3

How can we empirically distinguish between the human capital and signaling explanations for the education premium?

Model answer: Key empirical tests include: (1) sheepskin effects — if large wage jumps occur at degree completion (12th grade, bachelor's degree) rather than smoothly per year, signaling is implicated (the credential matters more than the years); (2) whether the return to education is higher in sectors with more information asymmetry; (3) whether employer learning erodes the education premium over time (as employers observe actual productivity, the signal's value should diminish if it is pure signal); and (4) whether self-employed workers (who do not need to signal to employers) show similar education premiums.

Each test has limitations. Sheepskin effects are consistent with signaling but could also reflect curriculum structure. Employer learning evidence is mixed. Self-employment evidence is complicated by selection. The difficulty of cleanly distinguishing the two explanations is why the debate persists after 50 years. The empirical consensus leans toward a mixed model where education both builds skills and signals ability, with the mix varying by context.