Whether economic position passes down

The fading American dream: absolute mobility by cohort

For each United States birth cohort, the share of 30-year-olds earning more in real terms than their parents did at the same age. Chetty and co-authors measure this from linked tax records covering the full population. Averaged over the parent income distribution, absolute mobility fell from 91.5% for the 1940 cohort to 50.0% for the 1980 cohort. The decline is steepest for children of middle- and higher-income parents; for children with parents at the 25th percentile it fell more gently, from 93.0% to 53.6%. Because the estimates use population tax data rather than a sample, their sampling error is negligible; the readme documents the robustness checks behind them.

Source: Opportunity Insights, The Fading American Dream National baseline estimates (Online Data Table 1). Absolute mobility is the fraction of children with higher real income at age 30 than their parents. Cohorts 1940–1984; publicly available for use with citation. Methodology

Where you grow up: upward mobility by state

The Opportunity Atlas follows children into adulthood and asks, for those whose parents were at the 25th percentile of the national income distribution, what income rank they reach as adults. A rank of 50 would be the national middle; higher is more upward mobility. Rolled up from counties to states (weighted by the number of children), it ranges from 53.1% in North Dakota to 34.9% in District of Columbia, against a national child-weighted mean of 41.3%. The upper Midwest and Great Plains sit at the top; parts of the South and a few urban cores at the bottom.

The county detail is sharper than the state average, and the Atlas publishes a standard error with every estimate. Among counties with at least 5,000 children, the highest and lowest, with their published standard errors:

Source: Opportunity Insights, The Opportunity Atlas kfr_pooled_pooled_p25: mean adult household income rank of children with parents at the national 25th percentile, pooled over race and gender, with its published standard error kfr_pooled_pooled_p25_se. 3,137 counties carry the estimate (national range 15.4% to 68.8%, median standard error ±0.8 points). State figures are a child-weighted county rollup, not an Atlas-published number. Methodology

Across the world: education persistence by cohort

The World Bank GDIM measures how strongly a parent's schooling predicts a child's, economy by economy, for ten-year birth cohorts from 1940 to 1980. The measure is the parent-child education correlation: higher means more persistence, so less mobility. The rich-country trajectories drift down as schooling widened (the United Kingdom and Germany reach 0.28 and 0.31 for the 1980 cohort), while China's rises from 0.22 to 0.46 as its education system stratified, and India and Brazil stay high (0.55 and 0.52). Bangladesh appears in GDIM only for the 1980 cohort, at 0.58, and is carried on the Gatsby curve below.

Source: World Bank, Global Database on Intergenerational Mobility (GDIM 2023) Parent-child correlation of years of schooling, parent = the higher-educated of the two parents, children pooled by gender. Higher is less mobile. A gap is a cohort GDIM does not report for that economy; nothing is interpolated. Methodology

The Great Gatsby curve: inequality and stickiness

Each dot is one economy: its income inequality on the horizontal axis (the World Bank Gini, latest survey) against its intergenerational education persistence on the vertical (GDIM, 1980 cohort). More unequal economies tend to have stickier mobility, the relationship Alan Krueger named the Great Gatsby curve; across these 149 economies the correlation is +0.28 and the fitted line slopes up, though the scatter is wide, so this is a tendency, not a law. Low-inequality Denmark sits low and mobile; high-inequality Brazil and South Africa sit high on inequality. Bangladesh, highlighted, pairs a moderate consumption Gini of 31 with high education persistence of 0.58, above the fitted line: stickier than its inequality alone would predict.

Source: World Bank, Global Database on Intergenerational Mobility (GDIM 2023) | World Bank Poverty and Inequality Platform Sources stated per axis: the vertical axis is the GDIM parent-child education correlation (parent = max, 1980 cohort); the horizontal axis is the World Bank PIP Gini index, latest survey per economy (2005–2025), which mixes income- and consumption-based surveys. The mobility cohort (born 1980s) and the Gini (contemporary) are not the same era; the original Great Gatsby curve pairs child mobility with parental-generation inequality, a refinement noted here, not implemented. Correlation and slope are computed, not assumed. Methodology