1James K. Galbraith has written an ambitious book. It starts off with the observation that “in a deep sense inequality was at the heart of the financial crisis” (4) and then presents almost two decades of the author’s work on inequality in twelve chapters, before returning to the issue of the global crisis in a short concluding chapter. His work is based on a unique (and undeservedly undervalued) methodology and data source. In the University of Texas Inequality Project (UTIP) sectoral pay data and regional data are used to derive a single index of inequality, the Theil index. These data are on earnings, i.e. wages, but the assertion is that wage dispersion across sectors and between regions will, in practise, go a long way in measuring inequality. Most (but not all) of the work uses the UNIDO industrial sector database. The inequality statistic thus arrived at has two major shortcomings: it measures inequality in earnings (but not property incomes) and it is for industrial sectors only. For developing countries there is an additional issue as it only captures the formal sector of the economy. It has two big advantages: data availability and consistency. While a full household income census may be preferable, such data are typically only available for some countries for some years—and then there are various issues of consistency (household or personal income? Gross or net? Income or expenditures?). Chapters 2 and 3 discuss these data issues and contrast the UTIP Theil measure to the Deininger-Squire (DS) dataset, a widely used dataset on inequality that compiles available Gini coefficients and that exhibits several oddities, such as classifying Spain as Europe’s most equal economy, that are probably due to data issues. Galbraith has, with various co-authors, written numerous papers with these and similar data. These form the basis for the following chapters.
2Chapter 3 uses the dataset to investigate the Kuznets Curve. It overall finds support for a Kuznets Curve, i.e. decreasing inequality with increasing income, but also for a modified Kuznets Curve that has a second inflection point above which higher income comes with higher inequality. The chapter also finds that there are strong common (time) effects that operate in all countries. This standard estimation method in panel econometrics (time fixed effects are separate dummy variables for each time period) controls for shocks that happened in all countries in the same year. Chapter 4 uses the data to develop an estimate of household income inequality by testing the correlation with the DS data. Chapter 5 uses it to investigate the question whether democracies lead to less inequality. The chapter concludes that democracy itself does not have negative effects, but social democratic countries and communist (as well as islamist) regimes do have negative effects on inequality.
3Chapters 6 and 7 deal with the geography of inequality in the USA. Chapter 6 highlights that the nature of inequality has changed over the last decades. In the late 1990s half of the increase in inequality was due to the increase in five counties (Manhattan, where the Wall Street financial industry is located and some ‘Silicon Valley’ districts). Inequality in this period meant an inequality at the very top of the income distribution and it affected relatively few people (this of course is consistent with the Piketty and Saez findings). With the stock market crash inequality declined for a few years before it started rising again. Chapter 7 investigates inequality and voting behaviour, arguing that increased inequality benefits the Democrats but reduces voter turnout (which hurts Democrats).
4Chapters 8 and 9 deal with inequality and unemployment in Europe. These chapters contrast the (mainstream) Labour Market Flexibility Hypothesis, which argues that rigid labour markets are the cause of high unemployment in Europe and the (Keynesian) Macroeconomic Policy view, which argues that restrictive fiscal and monetary policies are to blame. Galbraith argues that both arguments accept that (national) labour markets are indeed inflexible in Europe (which he does not document in detail for the Keynesians), but that in fact, with European monetary and economic integration Europe should be regarded as one unit and then wage inequality is indeed larger in Europe than in the USA (which is documented in great detail). While inequality is greater in short distances in the USA, overall inequality is greater in Europe. Thus the Labour Market Flexibility Hypothesis is incorrect, because labour markets overall are more flexible in Europe. Galbraith thus equates labour market flexibility with wage inequality (and he spends little time justifying this notion). Furthermore he argues that, in fact, higher inequality leads to higher unemployment because “inequality creates an incentive to search, which yields unemployment” (168). People can now travel across Europe and search longer for a better paid job rather than accept a low paid job.
5Chapters 10, 11 and 12 analyse inequality in China, Argentina and Brazil and Cuba respectively. These chapters are more descriptive and highlight trends in inequality but do not offer a econometrically investigated explanation. Chapter 13, finally, in only six pages Galbraith returns to the theme of the book, the link between inequality and the world crisis. The chapter summarises some of the findings of the book and combines them with broad claims such as “global inequality [is] the product of forces affecting the global economy in common and systematic ways, forces impinging on individual countries ... forces that are largely financial in character” (289). But, indeed, the previous chapters had not demonstrated this. So what then is the link between inequality and the crisis? The chapter is brief here: The dotcom bubble came with rising inequality, which throttled demand once the bubble had burst. Demand was sustained under the Bush administration via the “‘ownership society’, overtly encouraging massive expansion of lending to weak credits, and relaxing regulatory standards” (293).
6Does the book achieve its aim? This is difficult to tell, because the introduction does not specify the aim of the book. But the title of the book (Inequality and Instability) and the introduction certainly suggest that the explanation of inequality and how it relates to the crisis are at the core. The reader looking for the latter will be disappointed. While there is a lot on inequality in the book, its links to the crisis are discussed in the passing in a few short paragraphs—“the link, of course, runs through debt”(3). There is no clarification of the channels; nor is there a discussion of the literature on the subject (e.g. Rajan 2010, Kumhof and Ranciere 2010, Lysandrou 2011, Stockhammer 2012). Rather Galbraith seems to posit that financial forces have driven the rise in inequality and they are at the root of the crisis, with a weaker feedback from inequality via the Bush administration’s encouragement of easy credit and subprime credit in particular: “The economics of inequality is, in large measure, an economics of instability; inequality is the barometer, in many ways, of the instabilities that global credit relationships create” (18). So the inequality and instability are linked, but there plainly is too little of analytical substance to justify the title of the book.
7This is not to downplay the scholarship and the contributions of the book in general. The Inequality measure that Galbraith has developed over many years is indeed an interesting one and is unjustly ignored in the literature. It allows for a consistent and relatively simple way of calculating inequality over longer time periods and for more countries than other available measures. Galbraith presents a rich set of analyses based on this methodology. There are several interesting questions such as how this measure correlates with the top 1% measures compiled by Atkinson et al (2011). The UTIP data are available online [http://utip.gov.utexas.edu] and awaiting further analysis.
8So how successful is Galbraith in explaining changes in inequality? Galbraith is prone to strong conclusions: he writes “evidence strongly suggests that global finance is a principal source of changing global patterns of pay inequality” (73). This is based on the finding that in Kuznets-type fixed effects estimation the time fixed effects show substantially increasing effects since the around 1980s. He interprets this as ‘global factors’ and highlights the financial factors among them. This is plausible, but hardly ‘strong evidence’ that global finance is the ‘principal source of changes’. The evidence is equally consistent with the interpretation that technological change associated with information and communications technology has been the driving force. An extension for future research could include financial variables explicitly (IMF 2007) to find statistically significant effects of financial globalisation on household Ginis).
9The book repeatedly makes strong claims that global, not national factors drive changes in inequality. For example Galbraith claims that “the model suggests that national policy choices respecting inequality were secondary” (70). This is, again, based on the finding of significant time effects with a clear pattern—but in a specification without controls for national policies. Moreover, a later chapter reports results where a set of dummy variables for political regimes are included in a similar specification. In this specification the time effects are not statistically significant any more. On the face of it, that would suggest that national effects dominate the global effects.
10With the compilation of the UTIP pay inequality data set James Galbraith has made an important contribution to the analysis of changes in income distribution. His book only makes a first step in putting this dataset to use. His findings are stimulating, but preliminary. The weakness of the book is that its title is overambitious. This is a valuable contribution to the analysis of inequality, but not to the relation between inequality and instability.