Review essay: not quite on the level.

Schansberg, D. Eric

The Spirit Level is a favorite book of one of my university colleagues. It is also on a topic of immense interest to me: income inequality. Yet I probably would not have taken notice of Wilkinson and Pickett's book (WP) without my colleague's repeated reference to their work. (1)

Wilkinson and Pickett are quite helpful in laying out certain income inequality data--as well as data on a wide and impressive array of correlated social problems and issues. They rely on comparisons within two samples: the twenty-three richest countries and the fifty states in America. They discuss some of the relevant theories and at least some of the relevant literature. (2) Then, they combine the data, theory, and literature review with some just-so stories that result in inferences somewhere between solid ground and an awkward narrative that fits various aspects of their worldview. (3)

At the end of the day, their data do not imply much about how to achieve their ends. Most of the time, WP are appropriately reluctant to make policy prescriptions. (4) Nevertheless, sometimes, they deviate a bit and get themselves into contradictory positions. Unfortunately, they are not forceful enough in insisting on humility among their readers--and therefore careless proponents of their work could easily end up being less cautious than would be wise.

In a review in Faith and Economics, A. M. C. Waterman summarizes the WP project well, though he is more generous than I concerning the statistical basis: Despite its flaws, "the overall effect of all the evidence is persuasive... [so let us] take it that the correlations do indeed tell us something true and important, and go on to consider some of the questions they raise." (5)

Income Inequality Measured

"There are lots of ways of measuring income inequality," WP write, "and they are all so closely related to each other that it doesn't usually make much difference which you use" (17). They then refer to the choice of quintiles and deciles for income measurement and the calculation of Gini coefficients.

Among those choices, they are correct; it does not make much difference. However there are other key data questions that are not even raised here: Are we looking at individual or household or family measures of income? Is income pre- or post-tax? Is income pre- or post-transfer (payments from government welfare programs)? Wilkinson and Pickett touch on these three questions only briefly (243)--and largely to dismiss them.

More broadly, what constitutes "income" in these measures? What about measures of compensation rather than wages? What was the impact of 1980s changes in tax laws that encouraged higher-income workers to receive more wages and less stock? What about much-more-equal measures of consumption (rather than income)? (6) To what extent is variance in income a function of differences in hours worked? (7) The immense failure to discuss all of these and WP's reliance solely on a narrow approach that favors their chosen narrative is, by itself, sufficient to reject WP's work out-of-hand as naive or biased. But let us assume the best and move along.

Wilkinson and Pickett focus on averages and aggregate measures. It would be interesting to know how much the numbers (and WP's inferences) would change with significant changes in the "super-rich"--for example, if one were to triple or halve the number and/or the incomes of those in the top .1 percent. (My guess is that the numbers would barely move, but the implications would be relevant either way.) Related to this, they criticize CEO pay (249-64), and they assume that CEO pay is a significant contributor to the underlying problem (such as it is). (8) This seems fanciful. In any case, it would be nice to see them address this somewhat--at least with static simulations.

Wilkinson and Pickett discuss social mobility as their final applied topic (chap. 12). In part, this is reasonable: It is arguably the most important consideration, so why not save it for last? On the other hand, it is curious to wait so long because the dynamics are so important--and so important to the foundational topic of true versus measured income inequality. More troubling, they only discuss a narrow slice of a complex and mixed literature on income and class mobility--a slice that favors their worldview. Again, this is a lack of scholarship or an unfortunate bias. (9)

Income Inequality versus Poverty

Wilkinson and Pickett are usually careful to avoid the common error of conflating income inequality with poverty. (10) But their passing references to the problems of poverty underline another point--poverty is at least as pressing as income inequality. That said, they rarely seem to recognize this--even implying that material poverty should be considered in strictly relative terms (15), rather than strictly absolute terms, or taking the more moderate position that material poverty is both relative and absolute.

This raises another question: In recent years, why has there been a greater focus on income inequality than poverty per se? One answer is that many "inequality warriors" are more concerned about income inequality, whether from a sense of morality; out of envy; or here, as with WP, from pragmatic concerns about its practical impact on society.

Yet many proponents of "equality" care about poverty too. This leads to another answer: In public debate, it is common (and easy) to use static analysis of simple statistics. (11) This approach works reasonably well for income inequality. However it works very poorly for the poverty rate--the favorite but simplistic metric in that realm. In short, a reliance on the poverty rate would indicate a massive failure for the War on Poverty because it remains roughly unchanged since the so-called war began. Although the poverty rate is deeply flawed (12)--and masks many of the problems of the poor--the public debate has become wedded to simple statistics and the flawed analysis that follows. Given that the poverty rate "doesn't help" the preferred narrative of advocates, it makes more sense for them to focus on inequality than poverty.

Social Implications of Income Inequality

This is the most impressive and useful section of the book: chapters 4-12 on community life; (13) mental health and drug use; health, obesity, and life expectancy; educational performance; babies born to teenagers; crime and punishment; and social mobility. Wilkinson and Pickett present a series of simple correlations--with aggregate statistics, mostly point-in-time. The approach is limited but becomes more persuasive as they apply it so broadly.

Arguably, the most powerful point in the book is that income inequality correlates with a number of troubling social indicators--for both the poor and the rich. It is one thing to say that inequality (or poverty) hurts the poor. It is much more provocative to say that it impacts the wealthy. In economic terms, WP are trying to make the case that income equality has positive externalities and thus should be fostered by some set of government policies. (14)

Even so WP either ignore or too easily dismiss the potential influence of many other contributing variables: religious belief and practice, family structure and stability, and globalization and other economic factors. They discuss Robert Putnam's concept of "social capital" at length (54-56, 79, 140) as another positive externality. Oddly, they give it little weight as a causal factor or a potential solution (208-9).

There are clearly other variables in play as well--limits in the data and other causal factors--given the significant dispersion along WP's set of univariate trend lines. In all of this it is strange to see WP be so sociologically selective, focusing almost exclusively on income inequality. They seem oddly passionate about looking for "a common underlying cause" (186), and for casting income inequality in that role--as if a single cause were likely to explain something so complex. (15)

Likewise WP implicitly hold genetics and culture constant--an odd and incoherent position. (16) They try to dismiss the former with a quick sentence (185): "This must be rejected because it is simply an expression of racial prejudice." If you do not reject scientific investigation of the topic, then you are a bigot? They extend the discussion briefly, to acknowledge that race correlates strongly for the results within the United States but walk away after that. In the rest of the book, correlation is king. Here it is dismissed as a joker.

Given this attempt to describe complex realities in simple terms, the authors are compelled to marshal many just-so stories to make the data fit the analysis. This narrow approach should be troubling to sociologists, economists, or anyone doing policy analysis.

In Coming Apart, Charles Murray provides a useful expansion on WP's discussion. (17) As with WP (44), but in far-greater depth, Murray discusses the impact on measured income inequality of "sorting" by class and education in marriage. As with WP (164), but in far-greater detail, Murray describes the cultural barriers increasingly implied by class differences. As with WP (chap. 9 and pp. 137-38), Murray talks about family structure and stability, but in a more comprehensive and coherent manner. For example, WP argue at some length that fathers and family structure matter--that there is more to life than material goods. Then, apparently conflicted, they argue that it seems "possible to safeguard children against most of the adverse effects of being brought up by lone parents" (187)--possible, but unlikely, given the bulk of the available research.

Policy Prescriptions

Even assuming away the above concerns, now what? In most cases, WP are noticeably and appropriately hesitant to make policy prescriptions. Other times they are mysteriously optimistic: "There are good reasons to feel confident that we can create a society in which the real quality of life... is far higher" (241). At the end of the day, their data do not have much to say about how to achieve the implied and desired ends. Hence, caution is the wise approach.

Wilkinson and Pickett are correct in pointing to the limits of economic growth--at least on average, in developed countries (5). This is a point made nicely in Charles Murray's indispensable book on public policy in broad terms, In Pursuit of Happiness and Good Government (Simon & Schuster, 1988). Murray uses Maslow's hierarchy to note that material needs are the first goal but not the only policy goal. In developed countries, the thresholds on material needs can easily be (and have been) met--while using policy to move beyond thresholds has relatively little value and creates a range of other problems.

At times WP deviate a bit from policy humility and get themselves into contradictory positions. Early on they trot out items from a common laundry list of policy prescriptions: (18) housing, health care, work/life balance, child support payments, and the "provision of high-quality early childhood education" (112). (19) Later, they argue that "attempts to deal with health and social problems through the provision of specialized services have proven expensive and, at best, only partially effective" (238). Then, as examples, they cite a wide range of such efforts--police, medical care, social work, drug rehab, and obesity--concluding that they are often "a form of window-dressing, a display of good intentions" (238). They leave unanswered or even unasked the question of how government actions could be better motivated, informed, and executed. (20)

Wilkinson and Pickett also ignore the challenges of large-scale federal policies in the United States, especially given the complexities of poverty and all sorts of diversity in a population. To sum up, one would expect federal policy to be more beneficial with smaller and more homogeneous populations and more damaging as population and diversity increase.

Exploring the many reasons cited by economists for middle class struggles and growing income inequality over the last forty years, WP unfortunately focus on Paul Krugman. (21) He argues that inequality has been driven by reduced unions, a lower real minimum wage, and changes in taxes and benefits (22)--a carefully chosen subset of the causes laid out by economists. Wilkinson and Pickett conclude by citing a straw man: "changes in income distribution are almost never attributable simply to market forces influencing wage rates"--so as to move to the claim that the key is "something much more like the changes in institutions, norms and the use of political power which Krugman describes in the U.S." (243).

Conclusion

Ultimately, income inequality and poverty raise ethical and practical questions, both in terms of current outcomes and potential (policy and nonpolicy) remedies. (23) Wilkinson and Pickett are helpful on the practical side of wrestling with the potential impact of income inequality. Nonetheless, one should recognize the limits and flaws in their analysis and should not forget about poverty or the practical and ethical limits of public policy.

At the end of the day, WP is worth a read for those who have invested heavily in this realm. For novices who want to jump into the topic, I would start with two more-helpful, less-statistical, and easier-to-read books: Tyler Cowen's Average Is Over (Dutton Books, 2013) and especially, Charles Murray's Coming Apart (Crown Books, 2013).

Notes

(1.) One problem exists: the subtitle is good, but the title of the book does not prompt its potential audience to think of its primary topic--income inequality.

(2.) I certainly do not know all of the amazingly broad literature well. However from the areas I do know, I would infer that they are selective in a biased sense.

(3.) For example, WP have an odd focus on "anthropogenic global warming" (AGW) throughout the book (but especially in chapter 15)--at least in terms of their efforts to connect it to their book's thesis. When authors go so far afield, it raises red flags about objectivity and worldview. That said, they have an innovative reference to "tradeable carbon quotas" (222)--the sort of approach that would be bothersome to most people concerned with AGW.

(4.) This contrasts with Thomas Piketty's more famous (or infamous) book, Capital in the Twenty-First Century (Cambridge: Belknap Press, 2014).

(5.) A. M. C. Waterman, "Inequality and Social Evil: Wilkinson and Pickett on the Spirit Level," Faith and Economics (Spring 2014): 38-49.

(6.) See David Cutler and Lawrence Katz, "Rising Inequality? Changes in the Distribution of Income and Consumption in the 1980s," AEA Papers and Proceedings (May 1992): 546-51; David S. Johnson, Timothy M. Smeeding, and Barbara Boyle Torrey "Economic Inequality Through the Prisms of Income and Consumption," Monthly Labor Review 128, no. 4 (April 2005): 11-24; and Bruce Meyer and James Sullivan, "Five Decades of Consumption and Income Poverty," NBER Working Paper #14827, March 2009.

(7.) Wilkinson and Pickett offer a brief mention of the aggregates (228) but do not discuss work differences between income classes.

(8.) Robert Reich provides the foreword for the book and provides a useful corrective here (xi). "CEO pay" turns out to be a complicated theoretical issue with a very mixed bag of empirical results. "The market for CEO's" is not the most competitive, but its flaws are often exaggerated. It is also really difficult to measure well. Not surprisingly, one can find empirical results all over the board.

(9.) For a useful set of essays on the related topics, see "Income Inequality in America: Fact and Fiction," Manhattan Institute, e21 Policy Brief, May 2014.

(10.) Even so, WP still occasionally stumble. See, for example, "most deprived neighborhoods" (195) and "denying adequate incomes" (264).

(11.) It is not clear whether this stems from its rhetorical usefulness or its lack of empirical knowledge and theoretical rigor.

(12.) See chapter 1, D. Eric Schansberg, Poor Policy (Boulder, CO: Westview Press, 1996).

(13.) Among many variables on "community life" in chapter 4, WP discuss "trust" and find significant differences between countries. This is not an area of expertise for me, but I have read one scholar who says that the international differences on this metric are not significant: Dan Ariely, The Honest Truth About Dishonesty: How We Lie to Everyone--Especially Ourselves (New York: Harper Perennial, 2013).

(14.) See Robert H. Frank, Falling Behind: How Rising Inequality Harms the Middle Class (Oakland: University of California Press, 2007).

(15.) Without any apparent sense of irony, WP (246) cite New Hampshire and Vermont as examples of respectively low and high tax burdens. The two states have similar cultures, populations, and results, so the differential tax burden seems unrelated to the outcomes.

(16.) See Thomas Sowell, Race and Culture (New York: Basic Books, 1995).

(17.) Charles Murray, Coming Apart: The State of White America, 1960-2010 (New York: Crown Forum, 2013).

(18.) Likewise, trade unions get a positive but passing reference--without any explanation of how private-sector unions could rebound in an increasingly competitive and globalized economy (245).

(19.) Wilkinson and Pickett are apparently unaware of the US government's expensive and largely failed attempts to do the latter through Head Start. See Health and Human Services, "Third Grade Follow-up to the Head Start Impact Study: Final Report," OPRE Report 2012-45, October 2012.

(20.) Wilkinson and Pickett want to avoid "concentrating power in the hands of the State" (255), but they say nothing about doing so with health care--or more importantly here, with K-12 education. One might reasonably hope to improve the K-12 education system by reducing the concentrated (monopoly) power "in the hands of the State" and empowering those in lower classes through enhanced school choice (tax credits, educational vouchers, charter schools, and so on). They do include an odd claim about public school spending--public spending as a percentage of total spending (not overall spending or spending inequality) is the key variable (161). On what theoretical basis would one choose that variable? This seems like the sort of variable choice that springs from data-mining or ideology.

(21.) Because WP are not Americans, they might reasonably rely on the popular and prolific Krugman as their primary guide to all things American. Giving the benefit of the doubt, we can excuse them for this instance of fallacy-of-authority.

(22.) Ironically, proponents of these changes typically focus on pre-tax/transfer measures of income inequality.

(23.) Waterman, "Inequality and Social Evil," 48-49, applies the "seven deadly sins" to this discussion. See also D. Eric Schansberg, Turn Neither to the Right nor to the Left: A Thinking Christian's Guide to Politics and Public Policy (Greenville, SC: Alertness Press, 2003) for a broad and thorough discussion of ethical and practical angles to social and economic policies.

D. Eric Schansberg

Professor of Economics Indiana University Southeast

(*) Kate Pickett and Richard Wilkinson, The Spirit Level: Why More Equal Societies Always Do Better (London: Allen Lane, 2009; New York: Bloomsbury, 2010).