Printer Friendly

The role of homeownership and home price appreciation in the accumulation and distribution of household sector wealth.

*Thomas M. Holloway is Senior Economist, Mortgage Bankers Association of America, Washington DC. This article is adapted from one presented at the 32nd Annual Meeting of the National Association of Business Economists, September 23-27, 1990, Washington, DC. Helpful comments from Lyle E. Gramley and Richard W. Peach are gratefully acknowledged.

1 See footnotes and references at end of text.

This article reviews trends in the levels of household sector assets, liabilities and net worth since 1960. The sources of changes in net worth are examined, especially capital gains. After a review of the distribution of household wealth, the article assesses the likelihood of a decline in home prices in the years ahead.

WEALTH ACCUMULATION plays a key role in planning for retirement, determining consumer saving and spending, providing bequests to descendants, and influencing national capital formation and economic growth.(1) The purpose of this article is to explore the role of homeownership in the accumulation of household sector wealth.(2)


Household sector balance sheets showing assets, liabilities, and net worth are regularly prepared by the Federal Reserve and permit an analysis of the relative importance and growth of various balance sheet entries.(3) On the asset side, owner-occupied housing, land, and consumer durables are the most important tangible assets. Checking and saving deposits, credit market instruments, corporate equities, life insurance and pension reserves, and equity in noncorporate business are the most important financial assets. On the liability side, home mortgages and installment consumer credit are the major household sector liabilities.

Not surprisingly, throughout the 1960-89 period, nominal balance sheet measures tended to reflect general patterns of inflation. The inflation rate based on the implicit price deflator for personal consumption expenditures - averaged 2.7 percent during 1960-70, 7.3 percent during 1970-80, and 4.6 percent during 1980-89; over the entire 1960-89 period, the inflation rate was 4.8 percent. The nominal growth rates in assets, liabilities, and net worth also were lowest during the 1960-70 subperiod and highest during the 1970-80 subperiod. During the most recent subperiod, growth rates of assets and liabilities - like the inflation rate - were similar to the entire 1960-89 period average.

Table 1 illustrates the effects of adjusting the balance sheet measures for inflation. In 1982 dollars, assets increased at a compound annual rate of 3.4 percent from 1960 to 1989, liabilities increased 4.9 percent, and net worth increased 3.2 percent. Because liabilities grew smoothly, most of the variation in net worth around its trend reflected movements in assets.

The table also shows selected components of the major balance sheet entries. Owner-occupied real estate consists of both owner-occupied structures and land. Therefore, it reflects the assets involved in a typical home purchase. Other tangible assets consist mainly of consumer durables. Financial assets consist of the instruments and other financial claims noted earlier. Home mortgages consist of mortgages on owner-occupied real estate. Other liabilities consist mainly of installment consumer credit, other consumer credit, and bank and other loans.

The role of homeownership in the growth of balance sheet aggregates is highlighted in the owner-occupied real estate, home mortgage, and home equity categories. The value of owner-occupied real estate increased more rapidly than other asset categories from 1960 to 1989. Growth was particularly rapid in the 1970-80 subperiod, when the rate of increase was more than double the rates of the other two asset categories. During the same period, home equity also increased at a much more rapid rate than the increase in net worth from other sources. In short, home equity increased at a more rapid rate than both the inflation rate and net worth arising from a collection of assets and liabilities excluding those directly connected with homeownership.

Part of the strong gains almost certainly reflected the coming of age of baby boomers and the extraordinarily low real after-tax mortgage interest rates during the period, both of which contributed to a surge in housing demand and higher prices. However, during the 1980s, these factors virtually reversed course. Huge federal budget deficits contributed to very high real mortgage interest rates, the baby boomers had already formed households, leading to a sharp reduction in new household formations, and home price appreciation slowed markedly. As a result, home equity increases fell short of the inflation rate, so that home equity in constant 1982 dollars declined.


Changes in household sector net worth result from changes in total assets minus changes in total liabilities. In turn, these changes can be subdivided into component categories - such as those shown in Table 1 - and further subdivided into changes due to capital gains or losses and changes due to net acquisitions.(4) Capital gains arise from changes in prices of existing financial and tangible assets from one year to another. Net acquisitions reflect new purchases net of sales and depreciation.

Given this decomposition, changes in net worth can be stated as the sum of: (1) the change in net worth due to capital gains on owner-occupied real estate; (2) net acquisitions of owner-occupied real estate minus the change in mortgage debt outstanding; (3) the change in net worth due to capital gains on all assets excluding owner-occupied real estate; and (4) net acquisitions of assets excluding owner-occupied real estate, minus the change in liabilities excluding home mortgages.

It is the first two items that highlight the role of homeownership. Table 2 shows all of the sources of change in household sector net worth. Figure 1 zeroes in on changes in home equity. It is clear from the table and chart that capital gains account for most of the variation in home equity. These capital gains generally reflect movements in the inflation rate.

A statistical analysis suggests that, on average, capital gains on owner-occupied housing approximately matched inflation during the 1960-89 period. Because owner-occupied housing is a leveraged asset and home mortgages do not automatically increase with inflation, the analysis further implies that increases in home equity rise more rapidly than the inflation rate. For this reason, homeownership has often been viewed as an effective hedge against inflation. Home equity as a share of net worth increased during periods of accelerating inflation, such as 1972-74 and 1976-79, and, generally, decreased as a share during periods of decelerating inflation, such as 1974-75 and 1980-86.

Net acquisitions of owner-occupied real estate minus the change in home mortgages account for the remainder of the changes in home equity. This term was negative every year since 1961. The reason is that net acquisitions were less than the increase in home mortgages outstanding. In every year, net acquisitions alone showed increases because gross investment exceeded depreciation. However, the change in home mortgage debt typically exceeds net acquisitions because transactions involving existing owner-occupied real estate do not affect net acquisitions of the household sector as a whole, but do affect home mortgages.(5)

These trends clearly highlight the fact that homeownership plays an important role in accounting for growth in aggregate household sector net worth. However, the significance of homeownership becomes much more pronounced when considering the distribution of wealth.


One of the most noteworthy aspects of homeownership is that it is very broadly distributed, even among low income households. Figure 2 shows the homeownership rate by income quintile and illustrates several important points about homeownership.

First, the overall homeownership rate is about 60 to 65 percent - an important indicator of the widespread distribution of this type of asset. Second, even in the lowest two income quintiles, nearly one-half of all households owned homes. Third, homeownership rates increased with income levels in almost every year. Fourth, over time, homeownership rates have generally declined in the lowest income quintile, but generally increased in the highest income quintile.

Additional detailed data on the distribution of household wealth are available only at irregular intervals and differ in quality and coverage. Two of the most comprehensive studies were conducted for the Board of Governors of the Federal Reserve. The first, the 1962 Survey of Financial Characteristics of Consumers (SFCC), collected detailed data on household balance sheets and provides an extremely comprehensive inventory of assets (Projector and Weiss 1966). The second, the 1983 Survey of Consumer Finances (SCF), is the most comprehensive study since the SFCC (Avery, Ellienhausen, Canner, and Gustafson 1984a, 1984b; Avery and Elliehausen 1986).(6) Both the SCF and SFCC place special emphasis on sampling the very wealthy. This special emphasis is essential because the very wealthy control such a large percentage of total wealth. There are several other recent studies of wealth, but the SCF is widely regarded as the best current profile of the distribution of wealth.

Table 3 provides a detailed look at the distribution of owner-occupied housing compared to other types of assets using Gini coefficients based on the 1962 SFCC and 1983 SFC studies.(7) Gini coefficients range in value from 0 to 1.0 and measure the degree of inequality in a distribution. A Gini coefficient of 0 implies complete equality (i.e., all households own the same amount of wealth); a Gini coefficient of 1.0 implies complete inequality (i.e., a single household owns all wealth). Gini coefficients between these extremes reflect differing degrees of wealth concentration. Gini coefficients are shown for all households and reflect the distribution of assets among all households whether they own the asset or not.

The table indicates striking differences in the degree of concentration among different types of assets and liabilities. In 1983, the least concentrated item was vehicles with a Gini coefficient for all households of .51. The most concentrated item was trust fund equity with a Gini coefficient of .998. Owner-occupied housing was more equitably distributed than any other item except for vehicles. The Gini coefficient of .65 was significantly lower than the coefficient of .77 for total assets or the coefficient of .79 for net worth. Other common assets, such as savings deposits and corporate stock, were also significantly more concentrated than owner-occupied housing. The upshot is that owner-occupied housing is distributed much more equitably than most other assets.

However, the table also suggests that the degree of wealth concentration of total assets, net worth, and owner-occupied housing increased from 1962 to 1983. The Gini coefficients increased sharply for assets and modestly for net worth. In 1962, the Gini coefficient for owner-occupied housing was .63, but increased to .65 by 1983. Thus, the increase in the concentration of wealth included increased concentration in owner-occupied housing, but housing still remained significantly less concentrated than most other types of assets.


Undoubtedly, most equity gains arise through home price appreciation. For this reason, the outlook for home prices is of critical importance in determining the amount and distribution of household sector wealth. Further, from the point of view of mortgage lenders, nothing is more important than the amount of equity in a home in gauging the security of a loan. Quite simply, if equity is substantial, borrowers will not default. (More specifically, even if a borrower becomes severely delinquent, the property will be worth substantially more than the loan amount, and the borrower will sell the property to avoid foreclosure, thereby retaining some of the equity.)

Consequently, it is not surprising that several recent attention-grabbing articles proclaiming the collapse of home prices have generated a great deal of interest among homeowners, lenders, the media, and housing economists. In the August 22, 1988 issue of Barron's, the investment strategy team of Comstock Partners Inc. proclaimed: "We think real estate is heading for a fall." Soon after, a more substantial study by Harvard economists N. Greg Mankiw and David N. Weil (1988) was released that stated: "Real housing prices will fall substantially." The softening of prices in some northeastern cities in the late 1980s following explosive run-ups in the mid-1980s fed media interest in the topic, and both Newsweek and Time ran articles in late 1989 discussing the possibility of a significant weakening in home prices.

While all of this was going on, the consensus among housing economists appeared to be that the claims were greatly exaggerated. Still, the Mankiw and Weil study, in particular, merited serious reflection.

Based on their analysis, Mankiw and Weil make the startling prediction that real housing prices will fall 47 percent by the year 2007, or about 3 percent per year.(8) They summarize their section on the price forecast with the following comment: "Even if the fall in housing prices is only one-half what our equation predicts, it will likely be one of the major economic events of the next two decades." There is little doubt about that.

If prices fall as sharply as Mankiw and Weil envision, there will be a comparable decline in real owner-occupied real estate values on the asset side of household sector balance sheets and further significant shifts in the composition of wealth. Given that home equity is much more equitably distributed than other types of wealth, it is also likely that sharp declines in home prices would lead to a less equitable distribution of wealth in the U.S.

Will these sharp price declines come about? Most housing analysts appear skeptical. To begin with, the Mankiw-Weil results seem to suffer from a number of significant econometric problems. The basic Mankiw-Weil equation explains real home prices as a function of a demographic variable (representing housing demand) and a time trend. The demographic variable reflects movements of the baby boomers through the age-distribution of the population, and attendant shifts in new household formations (i.e., fewer new formations in the 1990s as the baby boomers age). This accounts for part of the reason their equation predicts such a large decline in real home prices, but another critical factor is that the time trend is relatively large and negative. In a recent paper examining the Mankiw-Weil study, Patric Hendershott and Richard Peach (1990) demonstrated that dropping the time trend (the presence of which is not defended by Mankiw and Weil in the first place) causes the equation to lose most of its explanatory power. Hendershott and Peach also note that the time trend alone would suggest about a 6 1/2 percent decline per year even if there were no change in demand. Clearly this is an implausible result. Further, the Mankiw-Weil equation appears to break down in recent subperiod fits (e.g., for 1970-87, the demand variable is not significant) and performs very poorly within sample forecasting tests. Finally, tests of the equation using regional data reveal enormous variation in the coefficients and highlight the critical role of omitted variables. In short, the Hendershott-Peach critique suggests that the Mankiw-Weil equation, is at best, not robust, and most likely misspecified. It is unlikely that Mankiw and Weil's use of a time trend and correction for serial correlation adequately address their specification problems.

One significant reason housing analysts doubted the Mankiw-Weil results so strongly in the first place is that likely trends in many other factors that influence home prices are ignored by Mankiw and Weil and do not point toward a collapse in prices. Let's review several of these factors:

1. Real incomes will continue to grow in the decade ahead. Consequently, many buyers will be able to afford significantly higher prices, and that will exert upward pressure on prices.

2. Although household formations will slow because of demographic shifts, the same process will move many existing households into age groups that have higher homeownership rates than their younger cohorts. This factor could significantly influence the homeowner-renter decision in favor of homeownership and could exert upward pressure on housing prices.

3. People must live somewhere, and, no matter what is happening to the age distribution of the population, the fact remains that the population itself continues to grow. An important, and perhaps growing component, is the large number of immigrants who enter the U.S. both legally and illegally. In either case, they contribute to population growth and are likely to contribute to upward pressure on housing prices.

4. Houses are not built for free. The components underlying construction costs (land costs, lumber and other materials costs, wages, regulatory expenses, etc.) affect the supply side of the market and play an important role in determining housing prices. With these costs likely to continue to rise in the 1990s, housing prices are also likely to rise accordingly.

5. Affordability problems kept many potential buyers out of the market in the 1980s. Any easing of real interest rates, softening of prices, or other factors affecting affordability would likely soon exert upward pressure on housing prices as this pool of would-be home buyers entered the market.

6. Homeowners probably would not stand for a sharp and sustained drop in prices. There is evidence that homeowners are often willing to keep their homes on the market for longer periods of time, rather than accept a price below what they perceive to be fair. If that attitude is pervasive, it puts a floor of prices from which to build.

7. Homeownership provides one of the few potential tax shelters for most households. The tax advantages of homeownership will help sustain demand and keep prices from falling. Further, if marginal tax rates are increased in federal budget accords, the after-tax return to housing will increase and induce additional demand.


The findings in this article lead to several fundamental conclusions. First, homeownership has been an important contributor to the growth of aggregate household wealth in the United States since 1960. Second, capital gains play a very important role in the accumulation of household wealth in general, and in shifts in portfolio composition in particular. During periods of relatively high inflation and high inflation expectations, capital gains accounted for the increase in the relative importance of homeownership. Third, homeownership plays a crucial- role compared with other types of assets in making the distribution of wealth in the United States more equitable. Home equity is more equally distributed than most other major components of household net worth. Fourth, if home prices decline sharply in real terms, there will be significant shifts in the composition of household sector portfolios. It is likely that such shifts would contribute to a less equitable distribution of wealth in the U.S. Fifth, a collapse in real home prices does not seem likely.

These conclusions point toward a continued important role for homeownership in household sector balance sheets in the 1990s. Wealth gains in the future through homeownership are almost certain to be less than those enjoyed in the 1970s, when real housing prices soared, but relatively stable real prices in the 1990s will still provide a useful vehicle for real wealth accumulation through the leverage and tax advantages provided by owner-occupied housing and through the gradual retirement of mortgage debt. Further, the likely rise in homeownership rates in the 1990s accompanying the further aging of the baby boomers suggests an even broader distribution of this type of asset - an asset that remains one of the nation's most equitably distributed. In short, homeownership in the 1990s, as in earlier decades, will be fundamental to the accumulation and equitable distribution of wealth in the U. S.


1 For a discussion of some of the relationships among wealth, saving, and capital formation, and how they are measured, see Holloway (1989).

2 For an expanded version of this paper, see Holloway (1990).

3 Annual balance sheets are prepared in connection with the Flow of Funds Accounts (FFA); see Board of Governors (1990). For comments on limitations of the FFA, and corresponding limitations in the balance sheets, see Holloway (1990) and de Leeuw (1984).

4 The Federal Reserve makes estimates of capital gains for most categories of assets, but not for liabilities. Currency and deposits are not revalued because they are not subject to capital gains or losses. Credit market instruments are not revalued either even though they are subject to market fluctuations. This may be problematic at times; estimates by others suggest that differences between par and market values of credit market instruments are sometimes fairly large. See de Leeuw and Holloway (1983, 1985).

5 For a detailed example illustrating this point, see Holloway (1990).

6 The 1983 SCF was updated in 1986, but the 1986 sample was a considerably smaller and there were fewer characteristics examined. The fundamental findings of the update support the findings discussed here.

7 Gini coefficients are derived from Lorenz diagrams. For a simple and concise discussion of these measures, see Suits (1977).

8 A summary of Mankiw and Weils' methodology is reported in Holloway (1990).


Avery, Robert B.; Elliehausen, Gregory E.; Canner, Glenn B.; and Gustafson, Thomas A. 1984a. Survey of Consumer Finances, 1983." Federal Reserve Bulletin 70 (September): 679-92.

Avery, Robert B.; Elliehausen, Gregory E.; Canner, Glenn, B.; and Gustafson, Thomas A. 1984b. "Survey of Consumer Finances, 1983: A Second Report." Federal Reserve Bulletin 70 (December): 857-68.

Avery, Robert B. and Elliehausen, Gregory E. 1986. "Financial Characteristics of High-Income Families." Federal Reserve Bulletin 72 (March): 163-77. Board of Governors of the Federal Reserve System, 1990. Balance Sheets for the U. S. Economy 1945-89 (April).

de Leeuw, Frank. 1984. Conflicting Measures of Private Saving." Survey of Current Business 64 (November): 17-33,

de Leeuw, Frank and Holloway, Thomas M. 1983. "Cyclical Adjustment of the Federal Budget and Federal Debt." Survey of Current Business 63 (December): 25-40.

de Leeuw, Frank and Holloway, Thomas M. 1985. "The Measurement and Significance of the Cyclically Adjusted Budget." Journal of Money, Credit, and Banking 17 (May 1985): 233-42.

Hendershott, Patric H. and Peach, Richard. 1990. "Is the Mankiw and Weil Forecast Believable?" Paper presented at the Mid-year Meeting of the American Real Estate & Urban Economics Association. Washington, D.C. (May).

Holloway, Thomas M. 1989. "Present NIPA Saving Measures: Their Characteristics and Limitations." In Robert E. Lipsey and Helen Stone Tice, eds. The Measurement of Saving, Investment, and Wealth. NBER Studies in Income and Wealth, Volume 52. Chicago: University of Chicago Press, pp. 21-83.

Holloway, Thomas M. 1990. "The Role of Homeownership and Home Price Appreciation in the Accumulation and Distribution of Household Sector Wealth." Paper presented at the 32nd Annual Meeting of the National Association of Business Economists. Washington, D.C. (September).

Mankiw, N. Gregory and Weil, David N. 1988. "The Baby Boom, the Baby Bust, and the Housing Market." NBER Working Paper No. 2794 December).

Suits, Daniel B. 1977. "Measurement of Tax Progressivity." American Economic Review 67 (September): 747-53.

Table 3

Concentration of Household Wealth by Selected Types of Assets and Liabilities
Type of Gini Coefficient for
Asset or All Households
Liability 1962 1983
All assets .71 .77
housing .63 .65
Other real estate .96 .94
Vehicles .61 .51
Demand deposits
 and currency .81 .82
Time and savings
 deposits .84 .85
Corporate stock .98 .98
Trust fund equity .999 .998
All Liabilities .75 .81
Mortgage debt .80 .84
Other debt .82 89
Net worth .77 .79

Sources: Edward N. Wolff and Marcia Marley, Lodg-term Trends in U.S. Wealth Inequality: Methodological Issues and Results," paper presented at the NBER Conference on the Measurement of Saving, Investment, and Wealth, Baltimore, Maryland, March 27-28, 1987. Wolff add Marley computations are based on data in Dorothy S. Projector and Gertrude S. Weiss, Survey of Financial Characteristics of Consumers (Washington, D.C.: Board of Governors of the Federal Reserve, 1966) and unpublished data from the 1983 Survey of Consumer Finances. Additional computations on the Gini coefficient for all households were made by the author. Tabular Data Omitted Figuration Omitted.
COPYRIGHT 1991 The National Association for Business Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Holloway, Thomas M.
Publication:Business Economics
Date:Apr 1, 1991
Previous Article:Some financial perspectives on comparative costs of capital.
Next Article:Promoting the U.S. flag merchant marine.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters