The Marriage Earnings Premium Is Real — and the Explanation Is Not What It's Usually Made to Mean

The marriage earnings premium is one of the most replicated findings in labor economics: married men earn approximately 10–25% more than demographically comparable unmarried men. This finding has been used to support substantially different interpretive claims — all of which deserve evaluation against the actual evidence.

What the Data Shows

The premium is real, large, and has persisted across countries and time periods [1]. Married men out-earn unmarried men on multiple measures: hourly wage, annual income, career advancement rate, and employer performance evaluations.

The question is why.

The Candidate Explanations

Productivity channel. Proposed mechanism: marriage stabilizes life circumstances, reduces financial and emotional distraction, and frees up capacity for career focus. Some evidence supports this — studies tracking men before and after marriage show earnings increases that begin before the wedding, suggesting selection into impending marriage rather than causation by the event itself, with gains continuing after [1].

Selection effect. More productive, higher-earning men are more likely to be selected as marriage partners and more likely to marry. The premium may partly reflect that men who marry are already higher-earners — marriage being the outcome of traits that also drive earnings, not the cause of those earnings.

Employer discrimination. Studies show employers evaluate married male candidates more favorably than equivalent unmarried candidates, reading marriage as a signal of stability, dependability, and commitment. This is a direct wage premium mechanism that operates through third-party perception, not through any productivity effect on the worker.

> 📌 Killewald's 2013 Demography study — one of the most methodologically rigorous on this question — found that the marriage premium substantially reduced and in some specifications disappeared when controlling for selection effects and pre-marriage earnings trajectories, suggesting that much of the premium reflects what kind of man gets married rather than what marriage does to his earnings. [1]

What This Means and Doesn't Mean

The finding is not evidence that married life makes men more productive, or that unmarried men should marry for career reasons.

The finding is evidence that:

• Selection effects are large in life outcomes research
• Employer biases based on relationship status are real and financially significant
• Life stability — in whatever form it takes — may carry positive career effects that don't require marriage specifically

The popular interpretation runs like this: "marriage causes the earnings premium; therefore marriage is prescriptively good for career success." That's the classic move of confounding correlation with causation.

---