Endnotes and citations are available in the PDF and Scribd versions.
President Barack Obama made the promotion of middle-class economic interests his highest priority at the start of his second term. In a July speech on the economy at Knox College in Illinois, the president said his top policy agenda is to “reverse the forces that battered the middle class for so long.” He argued that this requires not only fighting unemployment with better employment growth but also attacking the recent spectacular growth in income inequality. In making this case, the president pointed to the fact that “nearly all the income gains of the past 10 years have continued to flow to the top 1 percent … [but] the average American earns less than he or she did in 1999.”
This presidential attack on extreme inequality was particularly notable because it implied that the recent concentration of the proceeds of America’s productivity growth in a tiny number of ultra-rich households is inconsistent with the American Dream of upward mobility and economic growth. As President Obama put it, “growing inequality is not just morally wrong, it’s bad economics.” In short, the explosion of inequality that America has experienced in recent decades has been so extreme that it is inefficient, leading not just to a “battered middle class” but also to a future of low growth and reduced prosperity for all but the top 1 percent.
During his Knox College address, the president might have added that the dramatic post-1980 redistribution of national wealth to those at the very top has also had severe consequences for those whose standard of living falls far short of what anyone would consider to be “middle class.” The opportunity for upward mobility is at the heart of the American Dream, but there is compelling evidence that in the current era—which can appropriately be called the Age of Inequality—income mobility over individual careers has stagnated, and income mobility across generations is substantially lower in the post-1980 United States than in other high-income countries. Labor-market prospects for the vast majority of young American workers without elite college and graduate credentials have dramatically worsened since 1979.
That income inequality has become so high that it may now act as a drag on economic growth flies in the face of conventional economic wisdom, which underscores the importance of market incentives—big income payoffs—to extra effort and risk taking. This textbook economic thinking provided the intellectual foundations for the profound ideological shift toward free-market fundamentalism that began in the late 1970s. It became widely accepted by the end of that decade that the American welfare state—quite modest by international standards—was drastically undermining private incentives for growth. In short, America had become too egalitarian, and the best recipe for economic growth and future prosperity was more inequality. In this view, increasing incentives for work, investment, and risk taking could be achieved by deregulating labor, product, and financial markets; by shrinking the welfare state; and by legislating a much less progressive tax system. In sum, the conventional view was that America had to choose more efficiency and less equality. Reflecting this free-market vision, the United States, much like the United Kingdom, embarked on what Harvard economist Richard B. Freeman has called a great “laissez-faire experiment.”
In his influential 2012 book, Unintended Consequences: Why Everything You’ve Been Told About the Economy Is Wrong, Edward Conard, businessman and a visiting scholar at the American Enterprise Institute for Public Policy Research, argued that the recent rapid growth of American income inequality has been necessary to promote risky innovation and facilitate larger and “more liquid” financial markets. Indeed, in this view, soaring post-1980 inequality has been at the root of “what went right” in recent decades, as demonstrated by the (presumed) increasing superiority of U.S. economic performance over Europe and Japan. It all comes down to incentives for risk taking. Conard writes:
Europe and Japan lacked the economic incentives to take the risks necessary to transform their economies. … In the United States, more valuable on-the-job training, lower labor redeployment costs, and lower marginal tax rates increased payouts for successful risk taking. Higher payouts, in turn, increased risk taking. The outsized gains of successful risk takers diminished the status of other talented workers, which increased their motivation to take risks. Successful risk taking accelerated growth and the accumulation of equity. With more wealth in the hands of risk takers, US investors underwrote more risk. Larger, more liquid US financial markets allowed investors to further parse risk and sell risks they were reluctant to bear.
Not surprisingly, the editorial page of The Wall Street Journal applauded what it called Conard’s “bravado defense” of inequality and wealth creation and wrote approvingly of his conclusion that, “More equal societies work less, invest less, grow more slowly and ultimately leave everyone less well-off.”
This incentives-based case for high and rising inequality has been lent strong support by leading academic economists. Most recently and prominently, Harvard University’s N. Gregory Mankiw, in a 2012 paper published in the Journal of Economic Perspectives—“Defending the One Percent”—concludes, “The story of rising inequality, therefore, is not primarily about politics and rent-seeking but rather about supply and demand.” Mankiw, reflecting the mainstream view among economists, argues that the rise of extreme inequality is the outcome of three market developments: the increasing demand for skills in the information age, the failure of workers to develop the skills they need to succeed in the labor market, and the expanding responsibilities of CEOs as corporations have grown larger.
An alternative view is that the post-1980 U.S. trajectory from high to extreme inequality is less the result of politically neutral, technology-driven changes in labor demand that favored skilled workers and happened to correspond with slow growth in the supply of college-educated workers and much more the consequence of policy choices that reflected an ideological shift toward market solutions and away from moderate government regulation and redistribution. Facilitated by technological advances in information processing, communications, and transportation, this new laissez-faire policy regime of deregulation promoted the growth of the financial sector, the financialization of nonfinancial firms (in which the production of financial services and short-run returns to shareholders trumped longer-term investment in physical capital), and the offshoring of production to less-developed countries. Together with the dismantling of institutional protections for less-skilled workers—for example, a sharply declining minimum wage and hostility toward labor unions—these policy choices were at the root of the large-scale shift in political power in the 1980s away from middle-class economic interests and toward those of the top 1 percent, in what is an increasingly “rigged game.” In this vision, the rise of extreme inequality has undermined economic welfare and mobility for most American workers, as well as the prospects for economic mobility for their children, with potentially severe consequences for future economic growth and prosperity for the vast majority.
Does rapidly increasing inequality from already extremely high levels help promote economic performance and household welfare, or has the inequality of the post-1980 laissez-faire experiment gone much too far, if our yardsticks are national economic growth and the economic welfare of middle-class households?
This report compares the performance of the United States with other high-income countries on income inequality, economic growth, and the sharing of that growth with the vast majority of households. Because different inequality and growth indicators can make a big difference in these comparisons, our cross-country analysis makes use of three measures of income inequality and three measures of economic growth.
The report begins with a portrait of income inequality in the United States and in other affluent countries. Three common inequality indicators, each of which captures a quite different dimension of the income distribution, were used to create this portrait:
- Top and middle-class pre-tax income shares—the income shares accruing to those in the top 1 percent of income and to everyone falling in the 20th to the 79th percentiles of income
- The 90-10 ratio—the disposable after-tax income earned by households at the 90th percentile of the income distribution relative to the income of those at the 10th percentile
- The Gini index (or coefficient)—a summary measure of the overall dispersion of the income distribution that characterizes household disposable income inequality. An index of 0, for example, would indicate perfect equality, where the top and bottom 10 percent of the population would each receive 10 percent of the total income. In contrast, an index of 1 would indicate perfect inequality, where all income went to one household. Figure 5 reports that the Gini index ranges from 0.24 for Denmark to 0.38 for the United States in recent years.
On all three indicators, the United States ranked at or near the top of rich countries in 1980, and its relative inequality increased sharply over the following three decades:
- U.S. top and middle-class shares of total income tracked each other closely until the early 1980s, when the top income share exploded and the middle-class share began a long and steady decline. (see Figure 1)
- Compared to other rich countries, the United States shows by far the largest increase in the share of income taken by the top 1 percent (see Figure 2) and the largest decline in the middle-class share of total income—the middle 60 percent of households—since the mid-1980s. (see Figure 3) By 2007, the American middle class received the lowest income share in the rich world.
- At the start of the 2008 financial crisis, U.S. household income inequality—whether measured by the top 1 percent share, the 90-10 ratio or the Gini index—was by far the highest in the rich world, having risen substantially since 1980. (see Figures 2, 4, and 5)
- In the post-1980 Age of Inequality, higher income inequality—as measured by the Gini index—is closely associated with lower income mobility across generations. (see Figure 6)
Section 3 outlines alternative laissez-faire and political-economy explanations for the post-1980 explosion in income inequality. While the political-economy account is getting greater traction, the dominant story of the Age of Inequality, certainly among economists, has been that high and rising inequality just reflects competitive market pressures. Stagnant wages and skyrocketing top incomes are the consequences of new information technologies in the workplace that have driven up the demand for skilled workers faster than the educational system has increased their supply.
The political-economy explanation focuses on a radical ideological shift in favor of unregulated market solutions that appeared in the mid- to late 1970s. Political decisions led to institutional and policy reforms, here referred to as the “laissez-faire experiment,” which, together with technological advances in information processing, communications, and transportation, greatly empowered the finance sector, financialized the nonfinancial sector, and promoted the globalization of production. The result was squeezed worker wages and an appropriation of nearly the entire increase in national productivity by the top 1 percent, made up mainly of financiers and corporate executives. This political-economy story is outlined in Diagram 1.
Section 4 turns to alternative perspectives on the effects of high inequality on growth. Reflecting the alternative stories about the post-1980 surge in inequality outlined in Section 3, there are two general narratives. In the laissez-faire vision, what matters most for growth and prosperity is small government and strong market-based work and investment incentives, which imply strong economy-wide payoffs to high and rising inequality. Indeed, these incentives will promote the educational attainments that can ultimately at least moderate rising inequality.
In contrast, in the political-economy vision, above a certain moderate threshold, rising inequality can undermine social cohesion and the democratic process as financial elites increasingly dominate the political process. This, in turn, can squeeze wages as worker bargaining power declines. It also leads to inadequate investments in public goods, particularly those related to education, health, and the social safety net, as government budgets are also squeezed; a finance sector that is too large, wasteful, and destabilizing; and too little consumer income compensated for by too much household debt. In short, in this view, economic performance will be best in the long run if government plays an active regulatory, investment, and redistributive role, ensuring that middle-class households experience rising standards of living from market incomes (not debt). This model of shared growth is best promoted by a much more moderate and stable level of inequality than we have seen since the late 1970s. These two narratives are outlined in Diagram 2.
Sections 3 and 4 argue that soaring post-1980 inequality was largely a consequence of political decisions that reflected a strong pro-market ideological shift and that the consequences are not a good recipe for healthy, shared growth. Cross-country evidence is presented that suggests that it was political decisions regarding the regulatory and redistributive role of government—rather than demand-supply pressures—that best account for the exceptional character of the past three decades of American inequality:
- After the financial deregulation of the late 1970s and early 1980s, the United States has shown the fastest growth in the financial sector’s share of total compensation among the 15 rich countries for which there are data. (see Figure 7)
- Among 18 other rich countries, U.S. taxes and transfers had less effect on overall inequality than was the case for all other countries except South Korea. (see Figure 8)
- Government size is strongly inversely related to household disposable income inequality: Smaller government expenditure as a share of gross domestic product, or GDP, is associated with higher inequality, and the United States is an outlier on both metrics. (see Figure 9)
In Section 5, the report turns to evidence on U.S. economic performance and asks, “Has the extreme inequality of post-1980 America produced exceptional economic growth?” Attempting to link income inequality to economic growth is greatly complicated by the difficulty of measuring national output over time and especially across countries. Output growth is measured by the change in GDP, which is the sum of the net production in each sector.
For a variety of reasons, estimating the value of output is extremely difficult, particularly for finance, education, health, and government services, which together account—however measured—for an increasingly large share of the economy. There is also the question of what to measure output against. The possibilities include the total population (GDP per person); all adults (GDP per adult); all employees (GDP per employee); and all labor hours (GDP per hour, also known as standard labor productivity). Because the standard measures of growth are GDP per person and GDP per hour, this report focuses on these indicators. But because of the many serious questions that have been raised about the meaningfulness and consistency of the way GDP is measured over time, this report also compares countries with an alternative indicator—called “measureable productivity”—which is GDP per hour for those sectors with relatively well-measured output. (see “The Measurement of Economic Output” text box in Section 5).
The results of Section 5 show that the United States, while top ranked on all three measures of income inequality, is only a mediocre performer on output and productivity growth:
- Using three growth metrics—GDP per capita, GDP per hour, and measurable GDP per hour—to measure cumulative growth between 1980 and 2007 for six rich countries, the United States was never the top performer and was below or similar to Sweden on each. (see Figures 10a, 10b, and 10c)
- On the standard productivity metric, the United States scored below France, slightly higher than Germany, and substantially above Sweden and the United Kingdom on the level of productivity in 1994, 2000, and 2006. (see Figure 11a)
- On measurable productivity, however—which excludes those sectors for which value added is poorly measured—the United States had a lower score than France and Germany in 1994, and it was also below these two countries and Sweden in both 2000 and 2006. (see Figure 11b)
Section 6 addresses the relationship between inequality and growth. As Northwestern University economist Robert J. Gordon has emphasized, U.S. growth performance was impressive from the mid-1990s through 2004 but has since declined back to the lower levels of the 1980s and early 1990s. There is no reason to believe that the steady rise in income inequality since 1980 can help explain this slow-fast-slow pattern of productivity growth. More generally, the cross-country literature on the effects of inequality on growth in affluent countries is inconclusive. Consistent with this professional consensus, this section finds no evidence of a strong relationship—positive or negative—between levels of income inequality and standard measures of economic growth in advanced countries in recent decades. But there is an intriguing result for measureable productivity—my preferred indicator, since it excludes badly measured sectors, which can have big effects on cross-country results: If there is any relationship since the early 1990s, it appears to be that countries with lower inequality have had higher productivity-growth rates. I also show that the growth in the top 1 percent share grew much faster than standard productivity after (but not before) 1980 and that there is no statistical correspondence across affluent countries between top 1 percent income-share growth and productivity growth.
- There is no statistical association between disposable income inequality—the Gini coefficient for 2000—and standard indicators of output and productivity growth for 1994 to 2007. (see Figures 12a and 12b)
- But using measurable productivity, there is a negative relationship between income inequality and growth across high-income countries. If anything, lower inequality is associated with higher measurable productivity growth. (see Figure 12c)
- Using the conventional indicator, American productivity growth tracked the growth of the top 1 percent share of income almost perfectly from 1970 to 1986, but these indicators diverged sharply afterward, as the top income share rose much faster than productivity. (see Figure 13)
- The cross-country evidence does not suggest that growth in the share of income allocated to the top 1 percent is necessary for good productivity growth: Comparing 12 rich countries, only the United Kingdom showed higher growth than the United States in the 1 percent share between 1980 and 2007, but seven of the other 11 countries had higher cumulative productivity growth. (see Figure 14)
Section 7 asks why we care about economic growth in the first place. If the growth in a nation’s income is entirely captured by the top 1 percent—a good approximation for America in the Age of Inequality—the case for maintaining our commitment to the policies that defined the post-1980 laissez-faire experiment in order to maximize output must be made on grounds other than the well-being of the bottom 99 percent. The evidence summarized in this section indicates that America has shared substantially less of its less-than-stellar economic growth with nonsupervisory wage earners than have other rich countries.
An important consequence is that average annual growth in the median income of American households—the typical, middle-class family—has been substantially slower than that of other large, rich countries. At the same time, even this meager income growth has been dependent upon increasing household hours of work. In contrast to most other rich countries that have experienced a substantial decline in average work hours, American work hours—whether measured per worker or per adult—have increased, leading to less time for leisure or for working around the house (called by economists “household production”).
Compounding the pressure from stagnant incomes and rising hours of work, American working families pay far more out of pocket for essential education and health services than do their counterparts in other rich countries. The American public sector takes much less responsibility for early childhood education, for example, and this report concludes with an example that illustrates how poorly Americans are now performing on international tests of literacy and math proficiency, with potentially serious consequences for future economic competitiveness and prosperity.
- The United States has shared an exceptionally small part of its manufacturing productivity growth with nonsupervisory workers in the form of compensation since the early 1980s, consistent with the observed declines in the middle-class share of income over these decades. (see Figure 15)
- Compared to five other rich countries, U.S. compensation growth for manufacturing production workers between 1980 and 2007 was the slowest, and the gap with productivity growth was the highest. (see Figure 16)
- Compared to the same five other rich countries, American households had the slowest real median increase in incomes—0.4 percent—between the mid-1990s and mid-2000s, despite sharp productivity increases. (see Figure 17)
- But American adults paid for even this small increase in incomes with a sharp increase in annual work hours for the average working-age adult, from 1,213 hours in 1980 to 1,305 hours in 2002, while households in France, Germany, and Sweden combined faster household market income growth with declines in hours of market work. (see Figure 18)
- While the United States ranked fifth out of six rich nations in math proficiency in 2012 for those educated in the 1960s, its performance was at least close to
three other rich countries—the United Kingdom, Canada, and Germany. But Americans who attended primary and high schools in the 2000s were in last
place, well below the United Kingdom and far below Canada, Germany, France,
and Sweden. (see Figure 19)
The report concludes in Section 8 with a call for a return to shared growth. As President Obama put it, “When the rungs on the ladder of opportunity grow farther and farther apart, it undermines the very essence of America—that idea that if you work hard you can make it here.”
This legacy of America’s great laissez-faire experiment is undermining the future well-being of America’s young workers and, in turn, their children. It is time to return to policies explicitly aimed at reducing income inequality by raising wage levels; increasing the taxation of top incomes and the regulation of the financial sector; increasing job and health security; and above all, increasing public and private investments in skills, neighborhoods, and public infrastructure.
David R. Howell is professor of economics and public policy at The New School and a research associate at The New School’s Schwartz Center for Economic Policy Analysis and the University of Massachusetts Amherst’s Political Economy Research Institute.