Top Incomes over the Twentieth Century: A Contrast Between Continental European and English-Speaking Countries

The central finding is that the decline in top income shares during the first half of the twentieth century was entirely due to the fall of top capital incomes, not a gradual reallocation of labor from low- to high-productivity sectors. This rejects Kuznets's inverse-U hypothesis that inequality naturally rises then falls with development. Piketty's analysis shows the decline took place during a very particular and politically chaotic period—1914 to 1945, especially both World Wars and the early 1930s—which raises serious doubts about any gradual, Kuznets-type explanation [citation:6]. The stability of wage inequality in countries like France further confirms that capital income, not wages, drove the dynamics.

The volume pioneered the systematic use of income tax records as the primary source for constructing long-run, homogeneous series of top income shares. Unlike household surveys, which only exist on a national basis after 1950 and in machine-readable form after the 1970s-1980s, tax data allows researchers to trace inequality back to the early twentieth century [citation:15]. Atkinson and Piketty applied the same raw data sources and methodology across all countries, creating annual series broken down by income source—labor versus capital—for ten OECD countries. This database has since become the foundation for the broader international research program on global inequality.

Piketty's analysis shows the 1914-45 drop in top income shares was entirely due to the fall of top capital incomes; top wage shares actually did not decline at all. The timing coincided with the two World Wars and the Great Depression, which destroyed capital and disrupted rentier incomes. Progressive taxation, introduced in most Western countries during this period, also played a role in the post-1945 stability of lower inequality [citation:2]. This contradicts the Kuznets hypothesis, which predicted a gradual decline due to sectoral reallocation from rural to urban sectors—a process that did occur but did not compress income inequality as Kuznets expected.

The volume documents a striking difference: in English-speaking countries like the UK, USA, and Australia, top income shares rose dramatically after 1970, while in continental European countries the increase was much smaller. Piketty attributes the US case primarily to the explosion of top labor incomes—executive compensation and wage inequality—while in other Western countries the smaller increase was due to the revival of capital incomes [citation:2]. This divergence points to institutional and political factors, such as tax policy and labor market regulation, rather than purely economic mechanisms.

Top Incomes over the Twentieth Century is the foundational empirical work that made Capital possible. Piketty himself describes his 2001 French study (the precursor to this volume) as 'somewhat monomaniacal,' relying almost entirely on French income tax tabulations, but it launched the international research program that led to Capital twelve years later [citation:13]. The 2007 volume extended the analysis to ten countries with the collaboration of Anthony Atkinson and Emmanuel Saez. Where Capital offered a grand synthesis across three centuries and twenty countries, Top Incomes provided the detailed methodological groundwork and country-by-country case studies that underpinned it.

The authors acknowledge significant limitations. Their long-run series are generally confined to top income and wealth shares, containing little information about the bottom segments of the distribution. Fully homogeneous cross-country data sets do not exist, and the database's primary value lies in understanding dynamics at the top rather than overall inequality [citation:15]. Additionally, the volume is descriptive and methodological in focus—it constructs the series and offers first-cut analysis—rather than providing a systematic statistical exploitation of the complete database for causal inference. The authors explicitly hope future researchers will use their data to explore causal mechanisms more rigorously.