cross-sectional
A statistical comparison across different units (here, countries) at a point in time, as opposed to a time-series comparison of one unit over time.
As defined in Why Industrial Policy Is (Almost) Always a Bad Idea (with Scott Sumner)
Where it comes up · 2
- Gabriel Zucman on Inequality, Growth, and Distributional National Accounts2 claims
- Why Industrial Policy Is (Almost) Always a Bad Idea (with Scott Sumner)1 claim
Most important claims · 3 of 3Search all claims for “cross-sectional” →
- Because the analysis is cross-sectional—comparing the distribution year after year rather than following the same people—the finding of zero bottom-50% growth does not mean specific individuals failed to gain; an influx of low-skill immigrants, for example, could push down the average for the bottom 50% even if every prior resident became better off, making snapshot comparisons potentially misleading as a composition effect.Gabriel Zucman on Inequality, Growth, and Distributional National Accounts
- When income distributions are computed by age group (20-30, 30-40, 40-50), the bottom 50% shows essentially no growth within each age band, suggesting that even following people over the life cycle would reveal little income growth, because there is strong persistence in income rank and little mobility in or out of the bottom 50% as shown in Social Security data.Gabriel Zucman on Inequality, Growth, and Distributional National Accounts
- A cross-sectional comparison of countries supports neoliberalism better than U.S. time-series data does: while U.S. growth in the neoliberal era was slightly lower than the 1960s, countries that adopted more aggressive market reforms (China, India, Eastern Europe, Israel) typically did better than those that stayed with statist models.Why Industrial Policy Is (Almost) Always a Bad Idea (with Scott Sumner)