Inflation, Interest, and the Secular Rise in Wealth Inequality in the United States: Is the Fed Responsible?

Abstract

Two hallmarks of U.S. monetary policy since the 1981–1982 recession are declining interest rates and moderation in inflation, at least until recently. Coincident with these trends was a surge in U.S. wealth inequality, with the Gini coefficient up by 0.070 between 1983 and 2019. This article analyzes the connection between these two developments on the basis of the Survey of Consumer Finances. Contrary to expectations, the article finds that these two monetary effects reduced wealth inequality rather than increasing it. The effect is quite sizeable, with the Gini coefficient declining by 0.045 over these years. Asset price changes and debt devaluation also accounted for 72.6 percent of the advance of mean wealth and would have led to a 204.9 percent gain in median wealth compared to its actual rise of 23.4 percent. Moreover, they helped lower the racial wealth gap rather than enlarging it. These results are at odds with previous literature in which estimates range from a weak negative effect on inequality to neutral, small positive, and strong positive. In terms of methodology, this article differs from previous work by focusing on only the direct effects of interest rate changes and inflation on the household balance sheet.

Keywords:

• household wealth
• inequality
• racial inequality
• portfolio composition

JEL Classification Codes:

• D31
• H31
• J15

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1 The CPI-U-RS series begins in 1977.

2 Another problem is that they exclude pension accounts, like 401(k) plans in their empirical analysis because of a lack of comparable data. This is particularly a problem in the case of the United States where pension accounts made up 15.6 percent of total assets in 2007.

3 A related paper on the subject is Matthieu Gomez and and Emilien Gouin-Bonenfant (2020). They argue that while low rates decrease the average growth rate of existing fortunes, they increase the growth rate of new fortunes by making it cheaper to raise capital. Which effect dominates depends on the average equity issuance rate and leverage of individuals on the right tail of the distribution. They estimate this using new data on the trajectory of the fortunes of the top 100 individuals in the United States and conclude that the secular decline in interest rates played a key role in the recent increase of top wealth inequality.

4 In particular, Fed policy, like a change in the Federal funds rate, may affect stockholders’ expectation of future profits flows and, therefore, stock values. However, this is an indirect effect of Fed policy and is ignored in my analysis.

5 See, for example, Duca and Muellbauer (2021) for an extensive discussion of the factors that affect house and real estate prices.

6 It should be noted that homeowners with larger mortgages will benefit more from the mortgage interest rate reduction than those with smaller mortgages or no mortgages. In distributional terms, richer households will generally benefit more. It should also be noted that my approach differs from several papers cited above which include mortgage payments as a negative wealth entry. This article does not do that. Mortgage payments are an income flow and do not affect the balance sheet except for amortization.

7 This is analogous to the Campbell-Shiller CAPE ratio, the ratio of stock price per share deflated by the consumer price index to the average value over the past 10 years of corporate earnings per share deflated by the CPI (see Campbell and Shiller 1998). The Excess CAPE Yield (ECY) is likewise based in part on real long-term interest rates (see Shiller, Black, and Jivrav 2020).

8 Indeed, according to my calculations, the reduction in mortgage rates account for almost the whole increase in actual median home prices over years 1983-2019.

9 In the base scenario, it is assumed that the face value of liquid assets remains fixed over time. This assumption is later altered to include accrued interest on these assets (see Scenario IV of table 13).

10 On the other hand, the value of antiques and other “valuables” are included in the SCF in the category “other assets.”

11 My baseline estimates, as noted above, also exclude vehicles. Moreover, my calculations are based on the “public use” samples provided by the Federal Reserve Board, which are to some degree different from its internal files, so that my figures on mean and median net worth, as well as on wealth inequality, will in general be at a slight variance from the “standard” estimates provided by the Federal Reserve Board which include the value of vehicles in their statistics (see, for example, Kennickell and Woodburn 1999, and Bricker et al.,2016).

12 Though the standard SCF income measure includes realized capital gains as well as property income, these two components are excluded from S_T since they are already captured in the term r_iT.

13 Note that this time trend is rather different from that of the overall wealth Gini coefficient.

14 The residual group, American Indians and Asians, is excluded here because of its small sample size in most years.

15 Ratios of median wealth were even lower, at 0.07 in 1983 and 0.06 in 2007.

16 Differences in percentage gains over a period between the 30-year and 20-year Treasury bond are small because, as indicated in equation 3, the term C / (1 + i)^t becomes very small for t >20.

17 The only other long terms series on mortgage rates that I could find is for a 15-year mortgage. However, this series begins only in 1991.

Additional information

Notes on contributors

Edward N. Wolff

Edward N. Wolff is a professor at the New York University and a Research Associate at the National Bureau of Economic Research.