VATs are key to financing inequality-decreasing social spending.

The United States currently has one of the highest levels of inequality among the members of the Organization for Economic Cooperation and Development (OECD), as measured by the Gini coefficient before taxes and transfers. The U.S. Gini coefficient before taxes is 0.49, measured on a scale in which 100 is perfect in equality (one person has all the income) and 0 is perfect equality. Figure 1 shows that the U.S. Gini is higher than most of the other members of the OECD, and it is also higher than many developing countries (e.g., India).

This level of inequality would not matter if social mobility in the United States were high, because in that case every U.S. child could achieve the "American Dream." However, a recent OE pointed out: "Mobility in earnings across pairs of fathers and sons is particularly low in France, Italy, the United Kingdom and the United States, while mobility is higher in the Nordic countries, Australia and Canada" (OECD 2010, 183). The United States ranked third from the bottom among the twelve countries studied. The main reason was that it ranked first in the influence of parental background on student achievement in secondary education (OECD 2010, figure 5.3). In general, the OECD found that intergenerational social mobility tends to be lower in more unequal societies (OECD 2010, figure 5.10).

Numerous scholars have shown that social mobility in the United States is significantly lower now than it was in the period between 1945 and 1970, when inequality was also declining (Frank 2014). The combination of these trends is dangerous because it risks transforming the United States into a society where small elites capture most of the gains, a pattern in which growth cannot be sustained over time (Acemoglu and Robinson 2012; Zingales 2012). As Acemoglu and Robinson suggest, societies with "extractive" institutions tend to stagnate in comparison with societies with "inclusive" institutions. Moreover, as Zingales points out, the United States used to be more inclusive but is now becoming more extractive as established elites use their financial resources to lobby for and capture economic rents. A classic example is the tax treatment of the labor income of hedge fund managers as capital gains taxed at half the rate.

How does taxation enter into the picture? The level of inequality in the United States after taxes and transfers are taken into account is significantly lower, although it is still higher than in most OECD countries, and the trend is still for inequality to increase (Figure 2).

The striking fact visible in the data above is that the trend line is the same from 1984 to 2008—that is, the Gini coefficient moves more or less in parallel before taxes and transfers and after taxes and transfers are taken into account. This pattern is surprising because the United States relies heavily on individual income taxes, and those taxes are very progressive; and the top marginal tax rate and the tax rates capital gains and dividends changed numerous times during the relevant period, so if progressive income taxation reduces inequality, one would expect the before- and after-tax Ginis to diverge rather than march in parallel.

Compared to other countries, the United States relied heavily on individual income taxes throughout the relevant period, and those taxes are quite progressive. In 2001, the top 1 percent of the U.S. population by adjusted gross income (AGI) paid 33.89 percent of federal person income tax, and the top 5 percent paid the bottom 50 percent of the AGI distribution paid less thatn 4 percent of total income taxes collected). This is a significant increase from 1994, when the top 1 percent of taxpayers only paid 28.7 percent of federal personal income tax. In 2004, even after President George W. Bush's tax cuts, the top 1 percent still paid 32.3 percent of federal individual income taxes, and the top 5 percent paid 53.7 percent. In 2008, the top quintile paid 94.6 percent of federal individual income taxes, the highest percentage since 1979 (IRS 2013).

Given this level of progressivity, it is striking that the U.S. Gini after taxes increased steadily during the entire period from 1983 to 2008, parallel with the before-tax Gini. Moreover, if one examines the movements of the top individual tax rate in that period, one can see quite dramatic fluctuations: It went from 50 percent in 1983 down to 28 percent in 1986, then gradually up to 39.6 percent in 1993, then down to 35 percent in 2001, and back to 39.6 percent in 2012. The capital gains rate went up from 20 percent in 1983 to 28 percent in 1986, then down to 20 percent in 1997, then down again to 15 percent in 2003, and up to 20 percent in 2012. None of these changes seem to have made an impact on the after-tax Gini. A more detailed examination of the data, however, suggests the answer to this puzzle.

Figure 3 shows the Congressional Budget Office (CBO 2011) estimate of the contribution of taxes and transfers to the reduction in the Gini coefficient. Figure 3 indicates that (1) there has been some decline in the degree to which taxes and transfers reduced the Gini from 1979 to 2007, but the system as a whole is still quite progressive, reducing the Gini by about 17 percent; (2) both taxes and transfers contribute to the reduction, but transfers contribute more than taxes; (3) after 1986, despite numerous changes in the tax rates, the contribution of taxes to decreasing inequality is essentially flat, while all the fluctuation is due to changes in transfers (since the overall line tracks the transfers line).

These data suggest that the answer to increasing inequality is not more progressive income taxation. High rates of income taxation raise familiar problems such as increased tax avoidance, choosing leisure over labor, and potentially emigrating to low-tax jurisdictions (Avi Yonah 2012). While the U.S. top individual tax rate is lower than that in some OECD countries, those countries typically have lower rates on capital income, which is highly concentrated at the top of the income distribution. This comparison suggests that in a globalized world, the United States cannot raise the tax rate on the rich further. Certainly the 70 percent top marginal rate of 1980 and the 94 percent tax rate of the 1950s appear unlikely to return when the rich can easily move to other countries, give up their U.S. citizenship, and enjoy much lower rates.

Overall, the U.S. tax and transfer system has a very significant impact on inequality. As Figure 2 indicates, it reduces the U.S. Gini by about ten points, which is more than the difference between the pretax Gini of the United States and Sweden. So if it is not the progressive tax system that does the work, what does? The answer must be the much-maligned "entitlements": Social Security, Medicare, and Medicaid. These programs are very progressive because while only Medicaid is means tested, their benefits are more important to the poor, while the funding comes from progressive taxation of labor income (although in the case of Social Security, the funding could be made more progressive by eliminating the income cap).

Thus, one key to reducing inequality in the United States is to bolster the social safety net. We should at least maintain current benefits and possibly increase them. In particular, Social Security should be strengthened, since most baby boomers do not have nearly enough saved for retirement (and many have seen their savings decimated by the Great Recession). The Affordable Care Act has strengthened access to health care, but it will probably require more funding to support insurance for those with preexisting conditions and the poor if not enough young, healthy people sign up for the exchanges.

Moreover, strengthening the social safety net is important to sustaining growth. Open economies tend to have stronger safety nets, because the gains from having an open economy tend to impose risk on the people who lose from globalization; so in a democracy a strong safety net is a precondition to obtaining widespread political support for openness, which in turn produces growth (Avi-Yonah 2000). Thus, if we want to avoid the pattern that led to the end of the first era of globalization a hundred years ago, we need to maintain a strong safety net lest the U.S. public vote for protectionism, decreased immigration, and less tolerance for the "creative destruction" of technologically inspired growth.

How can the U.S. safety net be financially sustained in the long run? The answer cannot be raising the income tax. There are not enough rich people to support the safety net, and as stated above, the rich can adjust to a higher rate by avoiding taxes, working less, or mov ing. Nor would it be wise to increase the income tax on corporations (they can move even more easily than the rich) or on the middle class, since they already carry heavy burdens and cannot afford to bear the entire load (Warren and Tyagi 2007).

Another reason the social safety net cannot be financed by increased taxes on labor (either income or payroll taxes) is intergenerational equity. The benefits of Social Security and Medicare flow to the old, while income and payroll taxes are borne by the young. In addition to the dependency-ratio concerns raised above, it seems unfair (and will certainly appear so to many young voters) to transfer so much from the young to the old.

The only solution is to enact a broad federal consumption tax that is, a value-added tax (VAT). This proposal has been developed in detail elsewhere (Avi-Yonah 2010). For present purposes, a VAT has an important advantage: Unlike income and payroll taxes, consumption taxes do not discourage work, and because the old as well as the young consume, taxes are borne in significant part by the principal beneficiaries of the social safety net. In addition, VATs are used in over 150 countries and have a demonstrated capacity to raise revenues even with far weaker tax administrations than the IRS.

The VAT is regressive, and while there are ways to mitigate regressivity within the VAT (such as exempting basic living necessities), such exemptions subsidize well-off consumers and make the tax needlessly complicated. If VAT revenues are segregated and used to fund Social Security, Medicare, and Medicaid, the inherent progressivity of those programs mitigates the regressivity of the underlying tax.