I attended the Intelligence Squared debate on America’s economic outlook. In that debate I heard something that stayed with me, but the significance of which I have only now come to appreciate. Asked if he agreed with the motion, “The GOP Tax Reform Bill Will Improve Our Outlook for Growth,” Moore, the architect of the bill, responded in the affirmative. Shaking both his arms vigorously, he declared:
I agree with these two gentlemen that capital is the name of the game. Whoever has the most capital wins.
Stephen Moore, a card-carrying free-market fundamentalist, has been associated with the Heritage Foundation and the Cato Institute. He has even written for the Weekly Standard and the National Review. But the “two gentlemen” who he said agreed with him, and who did not contest the claim despite ample opportunities, were Simon Johnson and Jason Furman. Johnson, an economics professor at MIT known for his hard-hitting work on inequality and political capture by Wall Street, received the Main Street Hero Award. Furman, the “wonkiest wonk” in the Obama White House who helped draft the Affordable Care Act, was Al Gore’s economic policy director and served in the Clinton administration. That these two card-carrying progressive liberal economists agree with Moore that “whoever has the most capital wins” is, as we shall see, highly significant.
We saw in the previous post that parental income is a poor predictor of kids income. This is because the controlling variable for an individual’s lifetime earnings is the prestige of the school they attended and parental income has very little effect on the kid’s odds of getting into a prestige school. The upper class, defined as high income families, is constantly being invaded by the smart kids of the hoi polloi. Parental income does have an economically significant effect on kids income after controlling for school prestige. But it is secondary and gets swamped by the influx from below. In other words, the upper class, defined as the top blah percent of income, does not have a viable mechanism to reproduce itself.
The problem is that we have an ill-defined notion of the upper class. We must define the upper class not in terms of income but rather in terms of wealth. For it is not high income but wealthy families who can reproduce their advantages generation after generation. Note that wealth is “twice” as concentrated as income in the sense that the Pareto decay parameter is roughly half as large. Figure 1 shows the share of the top 1 and top 10 percent for income and wealth in the United States. Note the wealth crash in 1966-1979. We are back to high classical levels of income inequality and closing the gap in terms of wealth inequality. Note also the dramatic peaks of top wealth shares during the late-twenties wealth boom.
Pfeffer and Killewald find that not only is parental wealth a strong predictor of kids wealth, grandparent wealth is also a significant predictor of grand kids wealth (even after controlling for parental wealth). Because wealth may “skip” generations, 2-generation dyadic correlations do not fully capture the inherited advantage. Using an inter-generational study, they find an elasticity of 0.44 for log net worth after controlling for age-effects (ie, doubling log parental wealth increases kids wealth by 44%) and 0.39 for net worth percentile.
In Figure 2, we replicate their baseline result (without controlling for age-effects). On the X axis we have self-reported parental wealth in 1989 and on the Y axis we have self-reported wealth of their kids in 2015. We find a robust elasticity of 0.46.
In Figure 3, we replicate their second result. On the X axis we have self-reported grandparents net worth in 1989 and on the Y axis we have self-reported wealth of their grandkids in 2015. Again, we find a robust elasticity of 0.28.
The results on wealth and income are in considerable tension with each other. Individuals not born in wealth but earning high labor incomes can save much of their high earnings, thereby accumulating wealth over their lifetimes. This is the source of entropy facing the upper class defined in terms of wealth. On the other hand, high rates on return on wealth allows the wealthy to pull further away from the pack. In this struggle for class reproduction, who has the upper hand?
In order to interrogate the struggle between entropy and the upper class, ie in order to ascertain the failure rate of wealth-holding families to reproduce themselves, we must read Piketty’s numbers in an entirely new light. Piketty and Zucman have shown that the controlling variable is the difference between the growth rate of income and the rate of return on wealth.
…our three inter-related sets of ratios—the wealth-income ratio, the concentration of wealth, and the share of inherited wealth—all tend to take higher steady-state values when the long-run growth rate is lower or when the net-of-tax rate of return is higher. In particular, a higher r-g tends to magnify steady-state wealth inequalities. We argue that these theoretical predictions are broadly consistent with both the time-series and the cross-country evidence.
It can be derived from accounting identities that the share of inherited wealth in the stock of total wealth is a function of the histories of the rate of wealth transmission (“rate”), the labor share, the savings rate and the difference between r and g (all of whom are time-varying) as follows.
This form of a function is called an integral representation. Intuitively, the numerator is inherited wealth, ie wealth transmitted to progeny, while the denominator is total net worth, ie the sum of inherited wealth and earned wealth. The share of inherited wealth is an increasing function of the rate of transmission and the “excess” rate of accumulation, r-g. And it is a decreasing function of the labor share and the savings rate. Inherited wealth is obtained by integrating/summing up the rate of transmission and the product of the labor share and the saving rate with exponential weights with exponential “parameter” r-g, which therefore emerges as the controlling variable.
Beyond r-g, the rate of wealth transmission emerges as a key social-demographic variable. It is a product of the ratio of the wealth of the dying (corrected for gifts already made to progeny) to average wealth and the mortality rate. Intuitively: Suppose that there are no gifts; only estates. The rate at which wealth is being passed on depends on the rate at which wealth-holders are dying and their wealth relative to the living.
The rate of wealth transmission is a function of social norms of gift/bequest and savings behavior. If people largely save for old age then they will exhaust most of their savings and the rate of transmission will be low. If the bequest motive is strong on the other hand, wealth will be transmitted at higher rates. Of course, we must correct for gifts made to progeny before death since there is substantial evidence that gifts are becoming an increasingly major conduit of wealth transmission.
At this point I am going suppress my desire to share all of their graphs. You can find them at the end of the paper on page 61. I will share just three more. First, the inheritance flow/rate of wealth transmission in France where the data allows quite precise measurements over the long run with the understanding that the “U” shaped curve is a common property of major Western countries.
And second, the share of inherited wealth in the stock of total wealth in France. We see that the share of inherited wealth is closing in on 70 percent compared to the antebellum peak of 90 percent in 1910. Piketty and Zucman report simulations that suggest that France and other advanced countries will be able to return to prewar levels towards the end of the 21st century.
And third, the share of inherited wealth in Europe and the United States. In the US, the wealth of the dead relative to the living has surpassed the 1930 peak of more than 60 percent. In Europe, the share fell steadily from the 1910 peak of more than 70 percent all the down to less than 40 percent in 1980, and is now at levels last prevailing in 1950, around 55 percent.
Time now to circle back to our question of who has the upper hand? The foregoing analysis suggests that the answer is both broadly political and historically-specific. None of these numbers is a coincidence. All the drivers identified by Piketty and Zucman—rates of return, capital/labor share, low savings rate, and perhaps even low growth rates—can be traced to the neoliberal counterrevolution. Above all, the sociopolitical transformation that is implicated by this analysis is a regime that maximizes the rate of return on capital—whoever has the most capital wins. This was not the case in the corporate-liberal synthesis of the postwar era; a regime geared to maximize economic output, not wealth.
What fueled this sociopolitical transformation that we call the neoliberal counterrevolution was the wealth crash of 1966-1979 that we noted earlier. It all started offshore. Burn has shown how the escape of capital from center country control punctured the quasi-public international financial order, whereupon speculative attacks forced the center countries to relinquish control of capital flows and the hard-currency exchange rates. Back home, the stagflation crisis provided the opening for a spectacular attack against Keynesianism, monetary subservience, and the unions. The counterrevolution was consummated in November 1979, Duménil and Lévy argued, when, with the dollar plunging, Carter appointed Paul Volcker to the Fed after panicked conversations with New York bankers over the weekend. Interest rates were deregulated with the passage of the DIDMCA in 1980 by a Democratic Congress and signed into law by a Democratic President; thus effectively ending financial repression. Henceforth, instead of counter-cyclical fiscal policy aimed at maximizing output and employment, we would have monetary management of the macroeconomy geared to fighting inflation and maximizing wealth; instead of “a re-industrialization of America” (as the title of a special edition of BusinessWeek demanded in 1980) and the revitalization of the great industrial corporations (that underwrote the Treaty of Detroit whereby labor shared in productivity growth), firms would be compelled to maximize shareholder value; instead of high taxes on capital and high incomes, we would have increasingly reactionary tax cuts; and above all, instead of financial repression we would have unfettered finance.
We have seen that income and wealth are implicated in three interrelated ways. First, in the definition of the upper class. We have seen how defining it in terms of the flow of income is no good since such a class does not have a viable mode of reproduction in modern America. We must instead define the upper class in terms of the stock of wealth since with that definition we have a concrete mode of class reproduction. Second, we have seen Piketty’s r-g, the difference between the growth rates of wealth and income, in a new light. It is the controlling variable for the inter-generational survival rate of the properly defined upper class. Third, income (GDP) and wealth (capital) map onto two very different modes of postwar political economy. In the postwar era, the entire system was geared to maximize the former. In the neoliberal era, the focus of the entire system has imperceptibly shifted to maximize the latter.
At issue also are the rigidities of the discourse on public policy, including the purely civil society discourse of economics. For why is there no disagreement, indeed violent disagreement, between the three gentlemen? There is no compelling macroeconomic reason why public policy—both fiscal and monetary—should be geared to maximize wealth/return/shareholder value instead of output and employment (or for that matter median income or blue collar wages). What we have instead is a political consensus forged and upheld by holders of capital.