Uses the fracking boom of the 2000s as a natural experiment — simulated production from shale geology interacted with year effects — to test whether improved male economic prospects reduce nonmarital fertility and increase marriage ("marriageable men" hypothesis, Wilson 1987). Analyzing a balanced panel of 2,044 Public Use Microdata Areas (PUMAs) from 1997–2012, the paper finds that fracking-driven male earnings increases raised both marital and nonmarital birth rates significantly but had no effect on marriage rates. A comparison to the Appalachian coal boom of the 1970s–80s reveals the opposite marriage response, consistent with evolving social norms rather than a stable structural relationship between male earnings and family formation.
| Coal boom 1970s–80s | Fracking boom 2000s | |
|---|---|---|
| Marital births (10% earnings ↑) | +7% (Black et al.) / +12.4% | +12.4% |
| Nonmarital births | −25.5% | +12.4% |
| Marriage rate | −9.6% never-married | No change |
"We find no evidence from the fracking context to support the proposition that as the economic prospects of less educated men improve, couples are more likely to marry before having children."
"The contrast of findings between the context of the coal boom and bust of the 1970s and 1980s and the fracking boom of the 2000s is consistent with the notion that social context matters."
The paper's contribution is to cleanly disentangle income effects on births from the marriage-market mechanism. The fracking instrument is well-suited: it is geographically determined by pre-existing geology, it generates large, plausibly exogenous earnings shocks for non-college men, and it operates through a channel (resources) rather than through norms. That this clean income shock raised both marital and nonmarital births without affecting marriage rates is strong evidence that the "marriageable men → marriage → children" pathway has weakened substantially since the 1970s. The social norms interpretation (supported by the heterogeneity by baseline nonmarital share and the coal-fracking contrast) is plausible but speculative — the paper cannot rule out that the fracking context differs from coal-mining for other structural reasons (e.g., migration, industry type, regional culture). For the wiki, this connects closely to the Autor-Dorn-Hanson-Wilson (2019) paper on manufacturing decline and marriage, but as a test of the reverse: what happens when male earnings improve. Together they suggest the marriage market responds asymmetrically to male earnings — negatively to declines, not positively to gains — consistent with a permanent norm shift rather than a cyclical one.