The Great Gatsby curve — the negative cross-country association between income inequality and intergenerational income mobility — is among the most cited stylized facts about inequality, but its causal status remains unresolved. A 2019 meta-analysis of published intergenerational income elasticity (IGE) estimates for the United States, Canada, and eight European countries found that, after controlling for study-design variables, country fixed effects, and paper fixed effects, the association of mobility with the Gini coefficient loses statistical significance while the association with the top 1 percent income share persists. That study explicitly framed its results as correlations, not causal effects, and identified the obstacles to causal inference: no exogenous variation in inequality, strong collinearity between the Gini coefficient and top income shares, the mechanical dependence of elasticity-based mobility measures on income dispersion, and unexplained heterogeneity across study designs. Later work (2021–2025) has expanded country coverage to 87 countries, corrected lifecycle bias in mobility estimation, formalized what alternative mobility measures identify, and shown that the sign of the inequality–mobility association depends on the mobility concept chosen — yet no study has identified the causal effect of inequality on intergenerational mobility. The open problem is to establish, with designs that support causal claims and are robust to the choice of mobility measure, whether and through which channels income inequality — overall dispersion and top-end concentration separately — reduces intergenerational mobility.

Background

Intergenerational mobility is the degree to which conditions at birth, chiefly parental economic status, determine a person's own adult economic status; higher mobility means adult outcomes depend less on origins. The most common summary statistic is the intergenerational income elasticity (IGE): the coefficient from an ordinary least squares regression of the logarithm of children's adult income on the logarithm of parents' income. A higher IGE means parental income matters more, i.e. less mobility. Two facts about the IGE matter here. First, it can be decomposed as the parent–child income correlation multiplied by the ratio of the standard deviation of children's log income to that of parents' log income; because of this dispersion ratio, the IGE is mechanically tied to how unequal incomes are, independently of any behavioral channel. Second, IGE estimates are sensitive to study-design choices — how many years of parental income are averaged, whether individual or family income is used, the age at which children's income is measured — with life-cycle bias (transitory and age-related income variation contaminating lifetime-income proxies) being the classic problem. Rank-based alternatives, such as the rank–rank slope (the slope from regressing the child's income percentile on the parents' income percentile), do not load on dispersion in the same way. On the inequality side, the two standard measures are the Gini coefficient, a summary of inequality over the whole income distribution, and the top 1 percent income share, which captures concentration at the very top; the two are highly correlated across countries and time. The 'Great Gatsby curve' is the observation, popularized by Corak (2013) and Alan Krueger's 2012 speech, that countries with higher income inequality tend to have lower intergenerational mobility (higher IGE). Within the United States, Chetty, Hendren, Kline, and Saez (2014) documented large geographic differences in mobility and showed they correlate with segregation, inequality, school quality, and family structure — but, like all of this evidence, these are associations. The source study (Amaral, Yen, and Wang-Goodman 2019) performed a meta-analysis: it pooled IGE estimates from previously published studies for ten countries, merged them with country-year Gini coefficients and top 1 percent income shares, and estimated meta-regressions (regressions of the published estimates on inequality measures plus controls) with country, paper, and methodological fixed effects (indicator variables absorbing unobserved country-, study-, and design-level differences). Its key findings: the bivariate Gatsby-curve association is strong for the Gini coefficient; but once methodological, country, and paper controls are included, the Gini association loses significance while the top 1 percent share association survives — suggesting top-end concentration may matter more for mobility than dispersion across the whole distribution. Joint models of the two inequality measures were uninformative because of strong collinearity (diagnosed with variance inflation factors, a standard regression diagnostic for collinearity). The authors emphasize that their design, which pools published estimates rather than exploiting exogenous variation or experiments, captures correlations, not causal effects. Subsequent research has reshaped the frontier in three ways: a global mobility database now covers 87 countries (Munoz and Van der Weide 2025); lifecycle-bias-corrected estimation shows mobility in Sweden and the United States remained largely stable despite rising inequality (Mello, Nybom, and Stuhler 2025); and the sign of the inequality–mobility association has been shown to depend on whether mobility is measured as persistence (IGE-type) or as progressivity of income growth (Mayoral, Ray, and Genicot 2025), so that measurement choice is now a first-order identification issue rather than a technical detail (see also Mitnik 2025 on what each mobility measure formally identifies, and Bouchard St-Amant, Marceau, and Garon 2024 on theoretical conditions for the curve to exist at all).

Problem Statement

The unresolved question left by the source study and its successors is whether higher income inequality causally reduces intergenerational income mobility — the causal status of the Great Gatsby curve — or whether the observed association is a statistical artifact of how mobility and inequality are measured and estimated. The open problem has three linked components. (1) Causal effect: estimate the effect of income inequality on intergenerational mobility with a design that supports causal claims, and determine whether the effect is zero or nonzero. (2) Separate components: determine whether top-end income concentration (top 1 percent or comparable top shares) affects mobility differently from, or independently of, overall inequality (Gini coefficient or comparable whole-distribution measures), overcoming the near-collinearity of the two measures. (3) Channels: identify through which channels any causal effect operates — in particular whether it works through the return to human capital versus disparities in the accumulation of human capital (the contrast the source study raises for future analysis), or through other mechanisms such as family investment or neighborhood effects. The question is settled by component (1): a causal estimate of the effect of income inequality on intergenerational mobility whose identification strategy and identifying assumptions are explicit, whose estimand is accompanied by a falsification, placebo, or sensitivity check, and which determines whether the observed association survives once the mechanical (dispersion) component of the mobility measure is separated from the behavioral component — the claim that the association is a measurement artifact and the claim that the causal effect is zero being distinct claims — all subject to the two robustness and harmonization conditions stated below. A zero effect is a precisely estimated null (via an equivalence or confidence-interval criterion, not a failure to reject), in which case no channel is operative; a nonzero effect entails both the separation of overall from top-end inequality and identification of the operative channels. Because elasticity-based mobility measures mechanically load on income dispersion (the intergenerational elasticity equals the parent–child correlation times the ratio of child-to-parent income dispersion, and progressivity-based measures — measures built on how income growth varies with a family's position in the initial income distribution — can even reverse the sign of the association), a causal conclusion about the effect is well defined only when it is robust to the choice of mobility measure (via measure-invariant formulations or an explicit decomposition of mechanical versus behavioral components), and because published estimates differ systematically by design (parental-income averaging windows, individual versus family income concepts, measurement age, lifecycle bias), it rests on mobility estimates whose design differences are harmonized or controlled. The source study additionally names policy evaluation — which public policies can break the inequality–mobility link — as future work; that applied question is downstream of and distinct from the causal inequality–mobility relationship itself.

References

  1. Amaral, Ernesto F. L., Shih-Keng Yen, and Sharron Xuanren Wang-Goodman (2019). 'A Meta-Analysis of the Association between Income Inequality and Intergenerational Mobility.' Socius 5, article 2378023119881282. DOI: 10.1177/2378023119881282. https://doi.org/10.1177/2378023119881282
  2. Corak, Miles (2013). 'Income Inequality, Equality of Opportunity, and Intergenerational Mobility.' Journal of Economic Perspectives 27(3): 79-102. DOI: 10.1257/jep.27.3.79. https://doi.org/10.1257/jep.27.3.79
  3. Chetty, Raj, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez (2014). 'Where Is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States.' The Quarterly Journal of Economics 129(4): 1553-1623. DOI: 10.1093/qje/qju022. https://doi.org/10.1093/qje/qju022
  4. Durlauf, Steven N., Andros Kourtellos, and Chih Ming Tan (2022). 'The Great Gatsby Curve.' Annual Review of Economics 14(1): 571-605. DOI: 10.1146/annurev-economics-082321-122703. https://doi.org/10.1146/annurev-economics-082321-122703
  5. Mayoral, Laura, Debraj Ray, and Garance Genicot (2025). 'The Great Gatsby Curve: Upward Mobility, Persistence and Inequality.' NBER Working Paper 34144 / SSRN. DOI: 10.3386/w34144 (NBER); DOI: 10.2139/ssrn.5395430 (SSRN). https://doi.org/10.3386/w34144
  6. Munoz, Ercio, and Roy Van der Weide (2025). 'Intergenerational Income Mobility around the World: A New Database.' World Bank Policy Research Working Paper 11166. DOI: 10.1596/1813-9450-11166. https://doi.org/10.1596/1813-9450-11166
  7. Mello, Ursula, Martin Nybom, and Jan Stuhler (2025). 'A Lifecycle Estimator of Intergenerational Income Mobility.' The Review of Economics and Statistics, pp. 1-44. DOI: 10.1162/rest_a_01585. https://doi.org/10.1162/rest_a_01585
  8. Nam, Jaehyun (2021). 'Does Economic Inequality Constrain Intergenerational Economic Mobility? The Association Between Income Inequality During Childhood and Intergenerational Income Persistence in the United States.' Social Indicators Research 154(2): 469-488. DOI: 10.1007/s11205-020-02579-2. https://doi.org/10.1007/s11205-020-02579-2
  9. Song, Lijie (2022). 'Examining the Relationship Between Intergenerational Upward Mobility and Inequality: Evidence from Panel Data.' Social Indicators Research 163(1): 1-27. DOI: 10.1007/s11205-022-02891-z. https://doi.org/10.1007/s11205-022-02891-z
  10. Schechtl, Manuel (2025). 'The association between childhood exposure to local wealth inequality and intergenerational income mobility in the United States.' Nature Communications 16, article 9164. DOI: 10.1038/s41467-025-64222-3. https://doi.org/10.1038/s41467-025-64222-3
  11. Bouchard St-Amant, Pier-André, Nicolas Marceau, and Jean-Denis Garon (2024). 'Uncovering Gatsby Curves.' The Journal of Economic Inequality 22(4): 833-864. DOI: 10.1007/s10888-023-09619-0. https://doi.org/10.1007/s10888-023-09619-0
  12. Mitnik, Pablo A. (2025). 'Inequality of Opportunity, Income Mobility, and the Interpretation of Intergenerational Elasticities, Correlations, and Rank-Rank Slopes.' Sociological Methods & Research 54(4): 1289-1338. DOI: 10.1177/00491241251352102. https://doi.org/10.1177/00491241251352102
  13. Bloise, Francesco, Paolo Brunori, and Patrizio Piraino (2021). 'Estimating intergenerational income mobility on sub-optimal data: a machine learning approach.' The Journal of Economic Inequality 19(4): 643-665. DOI: 10.1007/s10888-021-09495-6. https://doi.org/10.1007/s10888-021-09495-6