How much redistribution policies can account for long-run changes in inequality? To answer this question, we quantify the extent of redistribution over time by the percentage reduction from pretax to post-tax inequalities, and decompose the changes in post-tax inequalities into different redistributive policies and changes in pretax inequalities. To estimate these redistributive statistics, we construct homogenous annual series of post-tax national income for France over the 1900-2018 period, and compare them with those recently constructed for the U.S. We obtain three major findings. First, redistribution has increased in both countries over the period, earlier in the U.S., later in France, to reach similar levels today. Second, the substantial long-run decline in post-tax inequality in France over the 1900-2018 period is due mostly to the fall in pretax inequality (accounting for three quarters of the total decline), and to a lesser extent to the direct redistributive role of taxes, transfers and other public spending (about one quarter). Third, the reason why overall inequality is much smaller in France than in the U.S. is entirely due to differences in pretax inequality. These findings suggest that policy discussions on inequality should, in the future, pay more attention to policies affecting pretax inequality and should not focus exclusively on "redistribution".