This issue of U.S. Housing Market Conditions looks at the user cost of homeownership, a concept well established in academic literature but rarely applied in more policy-oriented discussions. The issue describes a method to estimate user costs and employs the rich data from the American Housing Survey (AHS) in combination with information from several other sources to develop household-specific user cost estimates. The method calculates user costs as a percent of house value, in dollars, and in terms of burden on the households (that is, user costs as a percentage of household income). Estimates of 1997 average user costs are shown for the Nation as a whole and for a variety of demographic, economic, and geographic dimensions (such as age, race, income, region, and location within cities). The sensitivity of the user cost measures is evaluated in relation to the assumption about the opportunity cost of capital.

The user cost of owning a home includes the usual out-of-pocket costs-mortgage interest, maintenance and repair costs, insurance costs, and real estate taxes-but also includes costs to the consumer that are not seen in monetary transactions, such as forgone earnings on the owner's equity in the house and depreciation. Appreciation in the value of the home, which offsets these other costs, is also included. In addition, user costs take into account the fact that mortgage interest and property taxes are deductible for Federal and some State income taxes. By adding costs that are not out of pocket and adjusting for tax offsets, the user cost concept provides an accurate measure of the real cost of owning a home and emphasizes the investment dimension of homeownership.

The AHS analyzes various out-of-pocket expenditures and relates these costs to income to measure the burden of providing shelter. Although very useful in the context of rental housing, this concept may not be useful for measuring burden for homeowners. The estimates developed in this issue provide a different perspective on the burden of homeownership.

Although researchers have used a variety of formulations of the user cost of owning a home (which tend to vary depending on their purposes), most have defined the user cost using the following general accounting format at right.

Mortgage interest and property tax rates are adjusted to reflect the income tax advantages enjoyed by homeowners. Because investment is a strong motive for owning a home, it is extremely important that the forgone earnings on the household's equity on investment be accounted for in the user cost calculation by including the equity rate of return. The equity rate of return is applied only to the equity proportion of the home's current value. The maintenance cost rate covers routine maintenance. The housing depreciation rate is the rate at which the house is expected to deteriorate over time. Last, the home appreciation rate is the rate at which housing prices increase.

The user cost formulation requires data on individual households, Federal and State income tax rates, local property tax rates, earnings on alternative investments, and home price appreciation. The following paragraphs discuss data sources employed to develop user cost estimates.

AHS Data

The AHS provides data at the household level for respondent estimates of housing values, household income, property taxes, mortgage interest payments, outstanding principal balances, and house maintenance costs. The AHS also provides the data on location, income, and age and race of homeowners used to categorized the results.

Several key factors used to generate user costs are computed from the AHS data for each household. The loan-to-value ratio is the ratio of the outstanding principal balance to the current value of the housing unit as estimated by the respondent. The property tax rate for a house is the ratio of property taxes to the house's current value. The mortgage interest rate is the average annual interest rate for the household's first two mortgages and all home equity loans (that is, all loans that apply to the household). The housing maintenance rate is the ratio of the annual maintenance cost to the current value of the housing unit.

Table 1 presents the 1997 average values of the above-mentioned variables for the national level of households with and without a current mortgage. Homeowners with mortgages have houses with higher average values than those without mortgages ($129,000 and $102,600, respectively), which is consistent with the fact that their annual average household income also is higher ($61,450 and 39,760, respectively). Average property tax rates are comparable for the two groups of households (approximately 1.4 percent). Homeowners with mortgages carry loans with outstanding balances that are approximately 53 percent of their homes' current value, on average, with an annual average rate of interest of 8.1 percent.

Marginal Income Tax Rates

Both Federal and State marginal income tax rates are used to compute homeowner user costs. Federal income tax rates are set by law; they are 15, 28, 31, 36, or 39.6 percent depending on income level. Federal income tax rates, income brackets, and deductions are available from the U.S. Internal Revenue Service. State income tax rates, income brackets, and deductions are unique to each State. The essential income tax information for all States is provided by the Federation of Tax Administrators (www.taxadmin.org/fta/rate/tax_stru.html). In order to identify appropriate Federal and State marginal income tax rates for the user cost estimates, a simulated tax return was compiled for each homeowning household in the AHS.

Housing Depreciation Rate

The housing depreciation rate is often an assumed fixed percentage of housing value. For this exercise, housing is assumed to have a useful life of 80 years and to depreciate at a fixed rate of 1.14 percent per year. This depreciation rate was derived by the U.S. Bureau of Economic Analysis using a Hulten-Wykoff geometric decay procedure.

Equity Rate of Return

The equity rate of return (or opportunity cost) usually is taken to be equal to the interest rate of a widely traded financial instrument-for example, the 10-year U.S. Treasury bond rate. However, the most appropriate financial instrument to measure the opportunity cost of capital is not known. Consequently, five values were chosen to test the "sensitivity" of the estimated user cost of owning a home to the choice of this parameter: 0, 2, 6, 10, and 15 percent. As a comparison, the current annual value of the 10-year U.S. Treasury bond rate is approximately 5.6 percent. Historically, the stock market has had a real rate of return of approximately 8 percent on investments. This is roughly the 10 percent used in the simulations. However, more recently, much higher rates of return (15 percent or more) have not been unusual.

House Price Appreciation

House price appreciation is defined here as the annual percentage changes in mean housing prices using a 3-year moving average. Data used to determine house price appreciation are annual values of Freddie Mac's Housing Price Index for Metropolitan Statistical Areas based on repeat sales. When combined with household information from the AHS, house appreciation rates are slightly higher on average for the owners without mortgages (4.4 percent) than for homeowners with mortgages (4.3 percent); see table 1.

Table 1:1997 Average Annual Homeowner User Cost Factors

User cost estimates of owning a home were compiled at the household level using the components described and presented in the previous section. Estimates of 1997 average annual mean user costs of owning a home are shown in tables 2 and 3. In addition, user costs are expressed in dollar values by multiplying user cost rate by the corresponding house value. The burden of homeowner user costs is determined as the ratio of household dollar-value user costs to the household's income.

As expected, the user costs of owning a home rise as the equity rate of return increases, whether the homeowner has a loan or not (see table 2). Generally, user costs are approximately 5 percent of the home's current value and are also fully within the range of estimates that have been published by other professional and academic researchers. However, user costs of homeownership can even be negative at very low equity rates of return (that is, near zero).

Table 2:1997 Average Annual Homeowner User Costs

At low or modest equity rates of return (0 to 6 percent), the user cost of homeownership is lower for households that have no mortgages than for those that have mortgages. However, at higher equity rates of return (10 to 15 percent), the user cost is higher for households without mortgages. The reason for this switch in trends is that homeowners without mortgages have more equity that is not earning high rates of return. At higher rates of return, this really matters. Another reason for the switch is that the assumed equity rate is higher than debt rates paid by households with mortgages. Put simply, the average cost of capital (equity and debt) is higher because there is no lower cost debt.

In dollar terms, housing costs range from negative values to over $12,000. Generally, they are in the $5,000 range and are similar to housing costs estimates based on a cash expenditure approach presented regularly in AHS publications. The cost burden estimates for owners range from negative 10 percent to nearly 70 percent of household income. However, they are generally in the 15 to 20 percent range and can be compared with the cost burdens of renters, which are approximately 30 percent. It should be noted that for owners with no mortgages, most of the user costs are the forgone earnings on their equity and do not represent out-of-pocket expenses.

Table 3 provides estimates of annual average user costs of homeownership that have been categorized by region, urban location, house value, age and race of householder, and household income. The equity rate of return is assumed to be 10 percent because it is similar to the historical rates of return from stock markets. Nationally, homeowner user costs are, on average, 2 percent higher for homeowners without mortgages than for those with mortgages. Interestingly, this difference holds regardless of how the data are categorized.

Table 3:1997 Average Annual Homeowner User Costs (Percent of Home Value) Where the Assumed Equity Rate of Return Is 10 Percent

The greatest variation in user costs appears across home values and household incomes. Families with the lowest house values have user costs of approximately 9 to 11 percent, while those with the highest home values have user costs that are between 6 and 8 percent. When comparing income, the variation is not as great but is still pronounced. In both cases, the difference could be driven by the income tax advantages, which are greater for higher income families. Specifically, user costs decline dramatically as house value or household income rises.

User costs are highest in the Northeast and the South and lowest in the Midwest and West. The difference between the highest and lowest user costs is about 2.2 percent. That user costs are lower in the Midwest and the West than the rest of the country may be due to their higher housing appreciation rates.

Among households with mortgages, user costs vary somewhat according to age of household head. This variance is likely the result of higher equity in older households and the fact that the assumed equity rate of return is greater that the mortgage rate. Finally, user costs appear to vary little (less than one-half percent) according to the race of the householder or according to the household's location (central city, suburb, or nonmetro).

This paper uses the richness of the data found in the AHS to estimate the user cost of homeownership. User costs are estimated by accumulating the various costs of owning a home at the household level. User costs of owning a home are shown in three ways. First, user costs are measured as a percentage of house value. Second, user costs are expressed in dollar value (in relation to the home's value). Last, the burden of user cost is measured by the dollar-valued user cost as a percentage of household income. Estimates of 1997 average user costs and their incidence are shown for the Nation as a whole and for demographic, economic, and geographic dimensions (age, race, income, region, and location within cities). The sensitivity of the user cost measures are evaluated in relation to the equity rate of return (or opportunity cost of capital).

The equity rate of return has been shown to largely determine user costs that the homeowner faces. Periods of high rates of return on equity (as recently experienced in the stock market) will impose high costs on current homeowners. It will also make raising homeownership rates more difficult. Also, during these periods, owners who have paid off their loans find the costs of owning their homes higher than if they still had a mortgage (due to the resulting tax advantages and the amount of equity tied up in the house).

Alternatively, user costs of owning a home can fall below zero during periods when real interest rates are extremely low or when house appreciation rates are high (as happened during the late 1970s and early 1980s, when both inflation and house appreciation rates were very high).

Finally, it is more advantageous for households (that is, user costs are lower) to reduce their outstanding mortgage principal balances and increase the equity in their homes when rates of return on alternative investments are low. On the other hand, households will want to increase their house-related debt when equity rates of return are high; for example, by getting a second mortgage or home equity loan.