Volume 19 · Number 1 · November 2017 economic econsoc.mpifg.de sociology _the european electronic newsletter 19.1 Note from the editor Content 1 Note from the editor Inequality: Inequality: A Piketty et al. Moment in the Social Sciences A Piketty et al. Moment 7 How Did the Great Recession Affect Income Inequality in Spain? by Pierre Blavier in the Social Sciences 15 The Comparative Organizational Olivier Godechot Inequality Network: Toward an Economic Sociology of Inequality by Donald Tomaskovic-Devey et al. hat does the world Piketty, 2001 & 2003). Piketty and success of Piketty’s Saez’s 2001 NBER Working Paper 22 The Sociology and Economics of Capital in the 21st (published in 2003 in the Quarterly Wealth Inequality: Two Worlds Apart WCentury (2014) reveal? This book Journal of Economics) made the re- by Philipp Korom is not just one of the most aston- turn of inequality in the United 29 Avoiding and Protesting Taxes: ishing bestsellers in the social States visible to many through an Wealthy People and Tax Consent sciences in recent years, but it may extremely simple and striking by Camille Herlin-Giret also signal an important shift in U-shaped graph plotting the in- the way we consider inequality in come share of the top one percent, 38 Why No One Cares about economics and sociology, in the which peaked at 20% in 1928, Inequality (Except Us) social sciences and the public are- dropped after World War II, stabi- by Martin Schröder na, in political debate and day-to- lized at a low of 8% in the early 42 Book Reviews day conversations. 1970s, and grew rapidly in the Piketty warns the reader that 1980s and 1990s, reaching 14% in Editor the book should not be considered 1998. This single graph helped to Olivier Godechot, Sciences Po, a one-off one-man opus. It is the definitively bury Kuznets’ (1955) CNRS, and MaxPo, Paris sediment of a decade of collective optimistic argument of an inverted research involving many scholars, U-shaped evolution of inequality Book Reviews Editor including Anthony Atkinson, Em- under capitalism: low before the Lisa Suckert, Max Planck Institute manuel Saez, Gabriel Zucman, Industrial Revolution, very high for the Study of Societies, Cologne Gilles Postel-Vinay, and Facundo during it, and decreasing in the Editorial Board Alvaredo. It started at the turn of post-industrial era. Patrik Aspers, Uppsala University; Jens the millennium with publications Although this simple mes- Beckert, Max Planck Institute for the Study about the evolution of income ine- sage may seem revolutionary not of Societies, Cologne; Johan Heilbron, quality during the course of a cen- only in political terms, but also as Centre de Sociologie Européenne, Paris; tury in the United States and France a groundbreaking scientific con- Richard Swedberg, Cornell University, Ithaca (Piketty and Saez, 2001 & 2003; tribution, it was not that novel. The economic sociology_the european electronic newsletter Volume 19 · Number 1 · November 2017 Note from the editor: Inequality: A Piketty et al. Moment in the Social Sciences by Olivier Godechot 2 data (US tax data), the method (top income shares), being decomposable and not being fully in line with and the result (the return of income inequality) could basic axioms of welfare theory (Atkinson, 1970), which already be found in previous works (Kuznets and Jenks, led to alternative inequality coefficients (i.e. the Theil 1953; Feenberg and Poterba, 1993; Cutler and Katz, and Atkinson coefficients). Nevertheless, those coeffi- 1992). So how did Piketty et al.’s incremental innova- cients share with the Gini the fact that they overlook tions – e.g. greater historical depth, a more homogene- the heterogeneity in the evolution of inequality at dif- ous dataset, a richer description of top incomes, a de- ferent levels of the income distribution. composition of the respective contribution of wages In contrast, the Piketty et al. approach to ine- and property income – become a turning point in the quality contributed to fully tackling the heterogeneity social sciences? It is not merely because this research, of the inequality evolutions: the rhythm of the evolu- which may only have been incrementally innovative in tion of the top one percent share might differ from regard to our knowledge of the United States in 2001, that of the F95-99 or the F90-95. Without formulating led to the establishment of ignored facts when replicat- any explicit criticisms of other measures, it provided ed in other countries – China, India, Germany, and so the level of detail required to turn the quest for a per- on – and led to the development of a 10 000 000 unique comparative dataset on world Piketty & Saez (2001) income and wealth inequalities. It is Piketty (2001) also because it helped to shift the way 1 000 000 we view inequality in the distribution of affluence, whether in terms of in- 100 000 come, wages, or wealth. Baseline 10 000 The top one percent and 1 000 statistical measures of inequality 100 10 Asking whether inequality has in- creased or decreased seems like a fairly Decade of publication simple question that should have a Number of references listed in Google Scholar (August 25, 2017) 1 1901- 1911- 1921- 1931- 1941- 1951- 1961- 1971- 1981- 1991- 2001- 2011- simple and unequivocal answer. How- 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020* ever, it requires a statistical compari- Inequality AND (income OR wealth OR fortune OR capital OR wage OR salary OR pay) … AND "Lorenz curve" son of not just two distinct figures, but … AND ("Gini index" OR "Gini coefficient") two full distributions. Answers might … AND ("Theil index" OR "Theil coefficient") … AND ("Atkinson index" OR "Atkinson coefficient") differ depending on the area one focus- … AND ("D9/D1" OR "D9 to D1" OR decile ratio OR interdecile ratio) es on most (top, middle, or bottom) … AND share AND ("top 1 percent" OR "top one percent" OR "top 1%") and on the metric that one prioritizes … AND share AND ("top 0.1 percent" OR "top 0.1%") in order to summarize a full distribu- Figure 1. Statistical concepts used to approach inequality tion. This is the famous problem of Note and source: I collected the number of references published each decade that are listed on Lorenz curves comparisons. The “Gini Google Scholar. I used the keywords inequality AND (income OR wealth OR fortune OR capital OR wage OR salary OR pay) as the baseline search, to which I added various statistical concepts on the coefficient” (1921) is notoriously fa- measure of inequality. Hence, the Google Scholar search on the “top one percent share” concept mous for being an all-encompassing was inequality AND (income OR wealth OR fortune OR capital OR wage OR salary OR pay) AND share measure that solves this comparison AND (“top 1 percent” OR “top one percent” OR “top 1%”). puzzle. As shown in Figure 1, it is one Google Scholar is a publication database based on Google’s automated collecting of academic references in online scientific books and articles. It therefore comes with some limitations, including of the most used statistical measures of imprecise, undefined, and potentially inconsistent temporal coverage and some referencing inequality, and it has enjoyed rapid errors, especially dating errors, especially for the earliest decades. However, besides being free of growth, especially in the 1960s and charge and easy to use, Google Scholar has the great advantage of providing a comprehensive 1970s. The fact that it could be imple- list of publications, including books (Harzing and van der Val 2008). This graph has been included mainly to compare the evolution of the use of different measures of inequality. Absolute levels and mented easily on limited samples of evolutions where the number of references <10 should be treated with caution. the population contributed to its adop- * The number of references for the full decade 2011–2020 has been estimated by multiplying the tion. However, it was criticized for not actual number of publications for 2011–August 2017 (6.5 years) by 1.54 (=10/6.5). economic sociology_the european electronic newsletter Volume 19 · Number 1 · November 2017 Note from the editor: Inequality: A Piketty et al. Moment in the Social Sciences by Olivier Godechot 3 fect all-encompassing measure of inequality into an This equation has not only been adopted by new so- unnecessary illusion. cial protests keen on naming an enemy, but also by By focusing on the top of the affluence distribu- the social sciences beyond economics, especially soci- tion, the Piketty et al. approach also uncovered a skewed ology and political science, as a complement to and world that had largely been ignored, especially when one even a substitute for traditional ways of designating used the more traditional inter-decile ratios: the gap the highest group in the affluence hierarchy. Figure 2 separating the richest of the rich from the rest of the rich shows clearly that upper-percentile groups have be- is as large as the one separating the rich from the rest of come a challenge for class theories. Traditional no- the population. Hence, the Piketty et al. approach not tions such as “capitalists,” “bourgeoisie,” and “upper only popularized a “top one percent share” measure of classes” were clearly dominant at the beginning of the inequality – which had already been abundantly used by twentieth century for thinking the gap between the Kuznets and Jenks (1953) – but even more importantly, top and the bottom (Figure 2). Their popularity de- it promoted a “top 0.1% share” (and “top 0.01% share”) clined sharply after the 1970s, however, and fractiles measure of inequality that had been very rare before of affluence have been gaining ground in academic their work (Figure 1).