The Great Leveler, page 53
Considering the counterfactual absence of all these powerful leveling forces, the most plausible outcome would seem to be peaceful leveling on a (much? how much?) smaller scale than that observed in actual history. But even this, my third and “intermediate” scenario, is probably too optimistic. If we assume that technological development in our counterfactual world mirrored that in real life, which seems reasonable in the long run, would not the many disequalizing pressures that harrow contemporary observers today—from resurgent sectoral income differentiation to intensifying globalization made possible by technological progress to computerization—have made themselves felt well before inequality had declined to anywhere near the levels achieved in our own world, and would societies that had not been shaped by the violent shocks of the world wars not have been less capable of withstanding them?
In this fourth and final scenario, inequality might indeed have declined to some extent during the second and third quarters of the twentieth century as social democracy and mass education curtailed wealth accumulation in elite circles but would since have rebounded, much in the way it has been doing in the real world, especially in Anglo-Saxon countries. In this case, perhaps the most plausible of my four counterfactual scenarios, inequality might very well have returned to levels that prevailed a century ago, putting us in a worse spot than the one in which we currently find ourselves (Fig. 14.1).
Figure 14.1Counterfactual inequality trends in the twentieth century
Fruitless though it might be to ponder at greater length the relative merits of these ideal-typical counterfactuals, they help us understand just how much would have had to have been different for substantial leveling to have occurred in the absence of violent shocks. First of all, we have to allow for the feasibility of gradual peaceful leveling under conditions of modernity, even though there is little empirical evidence to support this notion. Second, we have to posit an additional century of relatively peaceful conditions: any counterfactual shocks of comparable severity, regardless of their timing and specifics, would take us back to a close approximation of the real world and simply reinforce the preeminence of violent leveling. Third, we need to assume that the concentration of capital that existed in the early years of the twentieth century could somehow have been undone even without recourse to massive violent dislocations, which arguably entails an even greater stretch of the imagination. And fourth, we have to believe that any such leveling would not have been reversed by the disequalizing forces we have observed over the last generation. The first three conditions must apply for any significant nonviolent leveling to take place at all, and all four are required to approximate levels of inequality in the world we live in today. This is an exceedingly tall order, and it strongly suggests that without major violent shocks, developed countries would currently experience considerably higher levels of income and wealth inequality than they actually do. The only real question is how much higher.
Some might be tempted to dismiss this observation as irrelevant, not merely because it is impossible to verify but above all because that is not the world we actually live in. But this would be a mistake. The counterfactual of peaceful leveling under conditions of modernity is important for a very specific reason: if we cannot tell whether or how much inequality would have been reduced in the absence of the global violence of the Great Compression, how can we judge the prospects of leveling in the present or the future? For all the regional crises that compete for our attention, the world of relative peace and stability and economic integration outlined in my counterfactual is in fact the world that most of humanity inhabits today. How have these conditions shaped current inequalities, and what do they imply for the future of leveling?
1 In this and the next four paragraphs, I recap some of the basic points first made in the introduction (herein, pp. 5–9) and developed in Parts I–VI.
2 For early modern Europe, see herein, chapter 3, pp. 97–100. Milanovic 2016: 50 also rejects the notion of a link between inequality and economic growth in preindustrial societies.
3 See esp. herein, chapter 3, pp.164–173 and chapter 13, pp. 382–383, 387.
4 Quote: Milanovic 2016: 98. In 1790, Noah Webster considered Rome’s “vast inequality of fortunes” to be the principal cause for the fall of the Republic (“Miscellaneous remarks on divisions of property . . .,” http://press-pubs.uchicago.edu/founders/print_documents/v1ch15s44.html).
5 The clearest exposition of secular cycle theory can be found in Turchin and Nefedov 2009: 6–21. Cf. also 23–25 for more rapid and elite-centered cycles in polygynous societies, and see 303–314 for the results of existing case studies. Turchin 2016a applies an adapted version of this model to the United States. Motesharrei, Rivas, and Kalnay 2014 present a more abstract model of how elite overconsumption may precipitate the collapse of unequal societies.
6 Turchin and Nefedov 2009: 28–29 only briefly acknowledge exogenous factors. This can be a serious problem, most notably in the case of the Black Death in late medieval England, which defies endogenization: 35–80. For the societies mentioned in the text, see herein, chapter 3, pp. 94–101. Note that in a comprehensive survey Albertus 2015: 173–174 finds no connection between particular levels of land inequality and land reform or collective action leading to land reform.
7 I bypass here the debate about the causes behind the outbreak of global conflict in 1914 that has been given a boost by the recent centenary. Suffice it to note that in the most general sense, the world wars were endogenous to modern development in that they would not have been feasible without industrialization, and that mass mobilization was a corollary of the weapons technology available at the time: cf. Scheve and Stasavage 2016: 21–22. But this did not by itself determine the odds of actual war. Milanovic 2016: 94–97 proposes a more specific link between inequality and World War I that would allow the resultant leveling to be “‘endogenized’ in economic conditions predating the war” (94).
8 World War I: WWID. World War II: for putative bystanders, see herein, chapter 5, pp. 158–164. Switzerland: Dell, Piketty, and Saez 2007: 474; Roine and Waldenström 2015: 534–535, 545; and herein, chapter 5, pp. 158–159. For Argentina, see herein, chapter 5, p. 156.
9 For the disequalizing effects of globalization, see herein, chapter 15, pp. 413–414. British colonies in Africa tended to be quite unequal at the time of independence, even though inequality had in some cases already been declining in the postwar period: see Atkinson 2014b. For the importance of colonial assets for some European wealth elites, see Piketty 2014: 116–117 figs. 3.1–2, 148.
Part VII
INEQUALITY REDUX AND THE FUTURE OF LEVELING
Chapter 15
IN OUR TIME
RESURGENT INEQUALITY
The last generation to have lived through the Great Compression is rapidly fading. Ninety-five percent of Americans who served in World War II have passed away, and those who survive are mostly in their nineties. As with people, so with leveling. In developed countries, the massive decline in inequality that commenced in 1914 has long run its course. For about a generation, give or take a decade, income disparities have been growing in all countries for which we have reliable data (Table 15.1 and Fig. 15.1).1
In a sample of twenty-six countries, top income shares grew by half between 1980 and 2010, whereas market income inequality rose by 6.5 Gini points—an increase that could only partly be absorbed by an almost universal expansion of redistributive transfers. Statistically, 1983 was a major turning point, with downward trends in inequality reversed in Finland, France, Germany, Italy, Japan, and Switzerland, as well as modally for the entire sample. Anglo-Saxon economies had a head start, mostly in the 1970s: inequality began to rise in 1973 in the United Kingdom, in 1973 or 1976 in the United States, in 1977 in Ireland, in 1978 in Canada, and in 1981 in Australia. American wage dispersion already took off around 1970. Other metrics confirm this picture. Equivalized disposable household income Ginis and the ratios of top to bottom income shares generally increased from the 1970s or 1980s onward. Since the 1980s, the proportion of the population with middling incomes has been retreating relative to those in higher or lower income brackets in a number of OECD countries.2
Table 15.1Trends in top income shares and income inequality in select countries, 1980–2010
Country
Metric
1980
1990
2010
Lowest (year)
Australia
Top 1%
4.8
6.3
9.2
4.6 (1981)
Gini (m)
35.5
38.1
43.3
Gini (d)
26.9
30.3
33.3
Austria
Gini (m)
38.3 (1983)
44.0
42.3
Gini (d)
26.6 (1983)
28.4
27.4
Belgium
Gini (m)
33.0
30.7
33.1
Gini (d)
22.6
23.0
25.2
Canada
Top 1%
8.1
9.4
12.2
7.6 (1978)
Gini (m)
34.9
37.6
42.2
Gini (d)
28.2
23.0
25.2
Denmark
Top 1%
5.6
5.2
6.4
5.0 (1994)
Gini (m)
43.1
43.6
46.7
Gini (d)
25.5
25.8
25.3
Finland
Top 1%
4.3
4.6
7.5 (2009)
3.5 (1983)
Gini (m)
37.5
38.2
45.1
Gini (d)
21.7
21.0
25.6
France
Top 1%
7.6
8.0
8.1
7.0 (1983)
Gini (m)
36.4
42.6
46.1
Gini (d)
29.1
29.1
30.0
Germany
Top 1%
10.4
10.5 (1989)
13.4 (2008)
9.1 (1983)
Gini (m)
34.4
42.2
48.2
Gini (d)
25.1
26.3
28.6
Greece
Gini (m)
41.3 (1981)
38.6
43.2
Gini (d)
33.0 (1981)
32.7
33.3
Ireland
Top 1%
6.7
7.3
10.5 (2009)
5.6 (1977)
Gini (m)
41.3
42.6
45.2
Gini (d)
31.1
33.1
29.4
Italy
Top 1%
6.9
7.8
9.4 (2009)
6.3 (1983)
Gini (m)
37.0
39.7
47.2
Gini (d)
29.1
30.1
32.7
Japan
Top 1%
7.2
8.1
9.5
6.9 (1983)a
Gini (m)
28.3
31.3
36.3
Gini (d)
24.4
25.9
29.4
Korea
Top 1%
7.5
-
11.8
(6.9 (1995))
Luxembourg
Gini (m)
-
31.3
43.5
Gini (d)
-
24.0
26.9
Netherlands
Top 1%
5.9
5.6
6.5
5.3 (1998)
Gini (m)
33.8
38.0
39.3
Gini (d)
24.8
26.6
27.0
New Zealand
Top 1%
5.7
8.2
7.4b
5.4 (1988)
Gini (m)
29.7
36.0
35.5
Gini (d)
28.1
22.9
23.1
Norway
Top 1%
4.6
4.3
7.8
4.1 (1989)
Gini (m)
33.8
36.8
36.9
Gini (d)
23.5
22.9
23.1
Portugal
Top 1%
4.3
7.2
9.8 (2005)
4.0 (1981)
Gini (m)
33.9
45.1
50.5
Gini (d)
22.4
30.8
33.3
Singapore
Gini (m)
(41.3)
(43.7)
46.9
Gini (d)
(38.3)
(40.8)
43.3
South Africa
Top 1%
10.9
9.9
16.8
8.8 (1987)
Spain
Top 1%
7.5 (1981)
8.4
8.1b
7.5 (1981)c
Gini (m)
35.4
35.9
40.9
Gini (d)
31.8
30.2
33.3
Sweden
Top 1%
4.1
4.4
6.9
4.0 (1981)
Gini (m)
39.3
41.9
48.5
Gini (d)
20.0
21.4
25.8
Switzerland
Top 1%
8.4
8.6–9.2
10.6
8.4 (1983)
Gini (m)
46.3
39.7
40.7
Gini (d)
30.3
32.2
29.8
Taiwan
Top 1%
6.0
7.8
11.2
5.9 (1981)
Gini (m)
27.8
29.2
32.4
Gini (d)
26.3
27.2
29.6
UK
Top 1%
5.9–6.7d
9.8
12.6
5.7 (1973)
Gini (m)
37.0
44.4
47.4
Gini (d)
26.7
32.8
35.7
U.S.
Top 1%
8.2
13.0
17.5
7.7 (1973)
Top 1% (cg)
10.0
14.3
19.9
8.9 (1976)
Gini (m)
38.6
43.3
46.9
Gini (d)
30.4
34.2
37.3
Average
Top 1%e
6.7
7.8
10.0
6.1 (1983f)
Gini (m)
36.2
38.7
42.7
Gini (d)
28.0
28.1
29.8
Transfers
8.2
10.6
12.9
m = market income, d = disposable income, cg = including capital gains
a 6.4 in 1945.
b See note 3.
