A new working paper from Harvard economist Rebecca Diamond links GLP-1 marriage rates and employment outcomes to the use of weight-loss medications among previously single and unemployed women, finding effects that persist well beyond the first year of treatment.
The paper, NBER Working Paper No. 35387, is titled ‘GLP-1–Induced Weight Loss and the Female Obesity Penalty’ and was issued in June 2026. Diamond, a Harvard undergraduate alumna and Ph.D. graduate, drew on data from the University of Southern California’s Understanding America Study, a nationally representative, probability-based internet panel of approximately 15,000 U.S. residents.
Among women who were single at the point they began GLP-1 treatment, the probability of being married or living with a partner rose 18 percentage points overall, and 29 percentage points after six or more quarters. Diamond attributes this to weight loss ‘dramatically affect[ing] first impressions of potential partners, leading to a large increase in marriage and cohabitation.’
Employment Gains Follow a Similar Pattern
The relationship findings sit alongside an equally pronounced employment effect. Among women who were not working when they began GLP-1 treatment, employment rose by 27 percentage points after six or more quarters, according to the NBER paper.
Crucially, neither the relationship nor the employment effects extended to women who were already partnered or employed at the study’s outset. Diamond argues this pattern helps rule out a general boost in ‘productivity or well-being’ as the explanation: if broader motivation were the driver, gains would be expected across all groups, not concentrated among those starting from a disadvantaged position.
She also reports that single women on GLP-1s experienced ‘no statistically significant change in depression, loneliness, or life satisfaction,’ which she reads as evidence that visible weight loss, rather than shifts in mood or motivation, is the operative mechanism. ‘GLP-1 weight loss changes outcomes on precisely the margins where visible body weight should affect first impressions,’ Diamond writes.
The Causation Question Around GLP-1 Marriage Rates
The paper is careful about causal claims, and readers should be too. The comparison group consists of women who wanted to take GLP-1 medications but did not; those who did begin treatment may have differed in ways the data cannot fully capture. Women who decided to start treatment may have been simultaneously reassessing other parts of their lives, making changes in relationship formation a consequence of a broader personal reorientation rather than weight loss alone. The possibility that a nascent romantic interest prompted both the decision to start medication and the subsequent partnership cannot be dismissed.
Diamond’s design attempts to address these concerns. The absence of change in depression or life satisfaction scores among the treated group is intended to close off mood-based explanations. The specificity of gains to those who began from a position of being single or unemployed, rather than those already established, is meant to distinguish a weight-related mechanism from a general confidence effect.
Whether that case fully holds will depend on peer review. The paper carries working-paper status and has not yet been published in a peer-reviewed journal.
A Growing Population of GLP-1 Users
The research arrives as GLP-1 use has expanded substantially. According to Gallup polling, 11% of U.S. adults currently take GLP-1 medications for weight loss, up from 3% in 2024, and 15% report having used them at some point. Over the same period, the U.S. adult obesity rate has fallen from a record high of 39.9% in 2022 to 36.4% in 2026.
The clinical picture behind those population shifts is well-documented. Research from the USC Schaeffer Center finds that GLP-1 medications, including semaglutide and tirzepatide, enable patients to lose 15% to 20% of body weight, a result that behavioural interventions have not achieved at scale. Young adults beginning treatment between the ages of 25 and 34 could, on those findings, gain nearly two years of life expectancy and spend close to six fewer years living with diabetes.
Diamond’s contribution is to ask what happens to the social and economic circumstances of women whose weight changes at that scale. As Harvard’s Faculty of Arts and Sciences has noted in covering the paper, the framing centres an ‘obesity penalty’ concept: that women who are visibly overweight face disadvantages in labour and relationship markets that GLP-1-induced weight loss may begin to reverse.
Whether the causal argument survives scrutiny, the data on employment and partnership formation among previously single and unemployed women will require an answer from any competing explanation. Peer review will be the next test.
