Family finances in the U.S.: recent evidence from the Survey of Consumer Finances.

Using data that have just become available from the 1995 Survey of Consumer Finances (SCF) along with data from the 1989 and 1992 versions of the survey, this article details recent changes in the income, net worth, assets, and liabilities of U.S. families. It also presents information on family savings, unrealized capital gains, debt payments, and institutional providers of credit. Its primary focus is on changes in family finances between the 1992 and 1995 surveys; however, selected 1989 survey data provide a broader context within which to interpret the more recent changes.(1)

Although the data from the 1995 survey are still preliminary, some findings appear particularly noteworthy. First, between the 1992 and 1995 surveys, both median family income and median family net worth rose in constant dollars. The former, however, remained below the level measured in 1989, whereas the latter returned to the 1989 level. Second, the ownership and the amount of holdings of publicly traded stock by families expanded greatly over this period. In 1995, more than 41 percent of families had direct or indirect stock-holdings, compared with about 37 percent in 1992, and these assets accounted for about 40 percent of their total financial assets, compared with a little more than 34 percent in 1992. Finally, the survey provided little evidence of a serious rise in debt payment problems between 1992 and 1995, even though both the share of families with debt and the median amount of their debt rose.

Economic and Financial Backdrop

Between 1992 and 1995, changes in family finances were influenced by various macroeconomic events and long-term trends. Overall, the period was one of continuing economic expansion. In September 1992, as the U.S. economy was beginning to recover from the 1990-91 recession, the civilian unemployment rate was 7.6 percent.(2) By 1995, the fourth year of the expansion, the unemployment rate had fallen to 5.6 percent. Inflation was subdued over the period, with consumer prices rising at an average annual rate of 2.7 percent.

Conditions for families were also changing in the financial sector. Interest rates on deposits, which were very low in 1992, rose on net over the period; for example, the average rate on three-month certificates of deposit increased from 3.1 percent in 1992 to 5.8 percent in 1995. In contrast, interest rates on newly originated conventional mortgages declined from 9.7 percent in 1992 to 7.4 percent in early 1994 and then edged up to 7.6 percent in 1995. At the same time, the Standard and Poor's index of 500 stock prices increased at an annualized rate of 10.8 percent, and home prices rose at a moderate pace.

Several institutional trends were important for family finances over the period covered by the surveys. Two such trends may have accelerated the growth of holdings of financial assets, particularly stocks. First, the variety of mutual funds available to families continued to expand, as did the number of no-load funds. Second, employers increasingly offered tax-deferred saving plans as a way for workers to accumulate savings for retirement. Often such employer-provided plans offer an option that allows participants to invest in corporate equities. On the liability side of the balance sheet, growth in credit card holding was driven by increased marketing, relaxation of credit standards, lower credit card interest rates, and usage-based incentive plans (for example, cash rebate programs).

A key demographic trend was the aging of the large post-World War II cohort. From 1992 to 1995, the proportion of families headed by persons between 45 and 54 years of age rose from 16.2 percent to 17.8 percent. The financial decisions of families with heads in this group are likely to be influenced by the cost of college education for their children and the need to save for their own retirement.

Family Income

The SCF requests information on families' total cash income, before taxes, for the calendar year preceding the year of the interview (see box "The Survey of Consumer Finances"). Median and mean family income for 1994 (derived from the 1995 survey) had moved above the depressed levels of 1991 (from the 1992 survey) but had not yet moved back up to the pre-recession levels of 1988 (from the 1989 survey).(3) In addition, the percentage of families with incomes of more than $50,000 declined over this six-year period.

Some cross-sectional patterns hold generally across the three SCF surveys. Income is successively higher for each age-of-head group through 45-54, and then it declines (table 1). Education and net worth are positively associated with income in each of the surveys.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Although only limited external information on assets and liabilities is available for comparison with SCF data at the level of demographic groups, data on household income for the preceding year are available from the Current Population Survey (CPS) of the Bureau of the Census. Like the SCF, the CPS shows an overall pattern of decline in median and mean household income between the 1989 and 1992 survey periods. In contrast to the SCF, in the CPS mean income rose between the 1992 and 1995 surveys to a level above that in the 1989 survey. Median income in the 1995 CPS declined further to a level slightly below the 1992 level.

Two changes in the CPS over the period considered here make comparing recent estimates from the two surveys difficult. First, the CPS dramatically increased the size of the maximum income values recorded in the survey. This change alone could cause some income statistics like the mean to increase even in the absence of real change. Second, the redesign of the CPS in 1994 substantially altered the survey questions and introduced computer-assisted data collection. According to some estimates, the redesign may have induced complex changes in measured income.(4)

Changes in Income

by Demographic Categories

Between the 1989 and 1995 SCFs, median and mean income declined for families headed by persons in the 35 to 54 age groups. Median and mean income also declined for all educational groups, with the largest declines for the groups with at least some college education.(5) Other data in the survey suggest that one explanation for the incomplete recovery of the income of the group with a college degree may be that the fraction of the group not working in 1995 remained above the level of 1989.

Between the 1989 and 1995 surveys, median income rose somewhat for nonwhite and Hispanic families but fell for other families. In contrast to findings in other surveys, the proportion of families in the 1995 SCF that identified themselves principally as either nonwhite or Hispanic declined. This decline may reflect sampling error or other technical factors or may be simply a result of a change in the way respondents answer this question.(6)

After holding steady between the 1989 and 1992 surveys, the median income of families headed by machine operators and laborers (a group commonly referred to as unskilled blue-collar workers) rose in the 1995 survey, and that of families with self-employed heads fell. The increase in median income of machine operators and laborers was accompanied by an increase in the share of families with heads in such occupations, likely reflecting cyclical recovery. In contrast, the share of families with self-employed heads fell over the six-year period.

Gains in median and mean income between the 1992 and 1995 surveys were largely concentrated in the groups with net worth between $25,000 and $100,000. However, the only group for whom median and mean income were higher in 1995 than in 1989 was the group with net worth between $25,000 and $50,000.

Family Saving

Saving out of current income is an important determinant of changes in family wealth. Therefore, since 1992, the survey has asked respondents whether, over the preceding year, their family spent less than their income, more than their income, or about as much as their income. Despite the simplicity of this question, it appears to be a good indicator of families' saving behavior.(7) Deriving a reliable measure of the dollar amount of spending or saving would require substantial additional time from respondents and might adversely affect response rates.

Overall, the proportion of families reporting that they saved in the preceding year fell from 57 percent in 1992 to 55 percent in 1995 (a change that is consistent with the fact that saving rates generally decrease in expansions and increase in times of recession). Over this period, the aggregate personal saving rate as reported in the national income and product accounts fell from 5.9 percent to 4.7 percent. According to the SCF data, the share of families spending less than their income declined for most demographic groups. However, there were some interesting exceptions. A somewhat higher proportion of families with heads aged 75 and more were savers in 1995 than in 1992; this finding moderates the traditional tendency for the share of savers to decline with age. For both the group with less education than a high-school diploma and the group of machine operators and laborers, the proportion of savers rose, possibly as a result of the cyclical improvement in their incomes. Self-employed workers also showed an increased likelihood of saving in 1995. By net worth category, only the $10,000 to $24,999 group became more likely to save, altering only slightly the generally positive association between wealth and saving.

As in the past, the most common reason for saving given in 1995 was to increase liquidity, a category that includes a variety of precautionary motives (table 2).(8) However, the data show some strong shifts in the motives for saving between 1989 and 1995. Families became progressively more likely to report saving for retirement -- the increase perhaps reflecting the rising share of baby boom families in the population as well as the perceived uncertainty of future retirement benefits. The frequency of reporting of saving for education also continued to rise, a trend that is likely related to demographic shifts and continuing increases in the cost of college education.

2. Proportion of families citing selected reasons as most important for saving, 1989, 1992, and 1995

Percent

     Reason       1989   1992   1995

Education          8.1    9.0   10.9
For the family     3.1    2.7    2.6
Buying own home    4.9    4.0    5.4
Purchases          7.9    5.2    7.7
Retirement        18.4   19.3   23.5
Liquidity         34.2   34.0   33.1
Investments        8.0    7.6    4.3
Other reasons      6.7    6.2    5.8

Note: figures sum to more than 100 percent because some families cited more than one reason as most important.

Net Worth

After falling between 1989 and 192, both median and mean net worth rose. By 1995, median net worth had returned to virtually the same level as in 1989, but mean net worth had not fully recovered (table 3).(9) Median net worth rose 6.8 percent between 1992 and 1995. From 1992 to 1995, mean net worth rose 2.7 percent, but it was still down 5.0 percent from 1989.(10) Generally, a rise in median net worth that is larger than a rise in the mean suggests relatively less growth for wealthy families than for families in the middle of the wealth distribution.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Changes in Net Worth

by Demographic Categories

When these changes in the overall distribution of net worth are disaggregated by demographic categories, the picture becomes more complex. Between 1992 and 1995, median net worth rose for groups with incomes of less than $25,000. These groups also had higher median net worth in 1995 than in 1989, as did the group with incomes of $100,000 or more. The patterns of increases at the top and bottom of the income distribution are similar for mean net worth. For families in the $25,000-49,999 income group, however, both median and mean net worth in 1995 were below the 1989 level, and the mean was below the 1992 level.

Changes in median and mean net worth differed across age-of-head groups. From 1992 to 1995, median net worth increased for families with heads less than age 55, while the mean for each of these groups held steady or declined. In contrast, median net worth hardly changed for families with heads aged 65 to 74, while the mean increased.

The data within each year show net worth rising with the level of education of the family head, but over 1992-95, both median and mean net worth moved up markedly only for the groups with education at the level of a high-school diploma or less. Over 189-95, median and mean net worth rose only for the group with a high-school diploma.

The 1989 and 1992 surveys showed an appreciable increase in the median and mean net worth of nonwhite and Hispanic families, although their levels of net worth remained substantially below those of other families. The 1995 survey indicated that the median net worth of nonwhite and Hispanic families had changed little since 1992, whereas mean net worth had fallen below the level of 1989. For other families, mean and median net worth had risen from the depressed level of 1992 but still remained below the levels of 1989.

Over the six-year period, the changes in net worth by occupation classes show some similarity to the changes in family income. Median and mean net worth were higher in 1995 than in 1989 for families with heads who were in technical, sales, and clerical jobs; machine operators and laborers; service workers; and retired persons. In contrast, median and mean net worth were lower in 1995 than in 1989 for those in professional and managerial jobs, precision production jobs, and self-employment. For "others not working" -- including unemployed workers, students, and homemakers -- median net worth rose, and mean net worth declined. Changes for this group are likely dominated by cyclical variation in the composition of the group.

Median net worth moved down consistently for homeowners over the six-year period, whereas it moved in an opposite direction for renters. In 1995, mean net worth for both groups remained below the level of 1989. The results for homeowners do not appear to be driven by shifts in the level of home values, which generally rose over the period. Possible explanations could be the influx of new homeowners, an increase in the proportion of homeowners with mortgages, and a rise in the amount of mortgage debt owed.

Assets

According to the SCF, the share of financial assets in families' total asset holdings has risen steadily, from 28 percent in 1989 to 31 percent in 1992 to 34 percent in 1995 (table 4). By definition, the share of nonfinancial assets -- mainly vehicles, real estate, and businesses -- fell correspondingly.

4. Distribution of amount of financial assets of all families, by type of asset, 1989, 1992, and 1995

Percent

         Financial
           asset                  1989    1992    1995

Transaction accounts              19.7    17.7    13.5
Certificates of deposit           10.4     8.2     5.5
Savings bonds                      1.6     8.2     1.4
Bonds                             11.0     8.5     5.5
Stocks                            14.6    16.6    18.0
Mutual funds (excluding
    money market funds)            5.0     7.7    13.2
Retirement accounts               18.8    24.4    25.1
Cash value of life insurance       6.2     6.3     7.9
Other managed assets               6.6     5.5     5.7
Other                              6.0     3.9     3.2
    Total                          100     100     100

Memo
Financial assets as a
    percentage of total assets    27.9    30.5    34.1

Note: In this and following tables, figures may not sum to totals because of rounding.

Financial Assets

Substantial shifts in the composition of financial assets from 1992 to 1995 generally continued trends observed from 1989 to 1992. The share of financial assets held in transaction accounts and certificates of deposit, the traditional saving vehicles, fell sharply, from 30 percent in 1989 to 26 percent in 1992 to only 19 percent in 1995. At the same time, the share of tax-deferred retirement accounts, publicly traded stocks, and mutual funds in financial assets rose strongly, from 38 percent in 1989 to 49 percent in 1992 to 56 percent in 1995.

Although the proportion of families having at least some financial assets rose only slightly, ownership increased considerably for families earning less than $10,000 a year, for nonwhite and Hispanic families, and for families headed by precision production workers or machine operators and laborers (table 5). Overall, median financial assets rose a bit; this gain was shared by most demographic groups except families with heads aged 65 and older, for whom the median fell.(11)

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Transaction Accounts and Certificates of Deposit

From 1992 to 1995 the proportion of families having transaction accounts -- checking, savings, and money market deposit accounts; money market mutual funds; and call accounts at brokerages -- held steady, whereas the median holdings in such accounts fell a little. In 1995, families that had no transaction account were clustered mostly in lower-income and younger age groups. However, ownership rates for lower-income groups were substantially higher in 1995 than in 1989. The discussion in the box "Families without a Checking Account" gives some background on the reasons survey respondents gave for not having accounts.

For most demographic groups, the percentage of families having transaction accounts changed little between 1992 and 1995. However, ownership of such accounts increased substantially for families headed by machine operators and laborers, largely because of cyclical improvements for the group, and declined substantially for families with self-employed heads, largely because of a shift in the composition of the group. While median account balances declined slightly for most groups, there were three notable exceptions: the median declined markedly for families with incomes of $100,000 or more and rose for families with heads aged 75 or more and for nonwhite and Hispanic families.

Ownership of certificates of deposit (CDs), another type of deposit traditionally considered a safe investment, declined broadly, continuing a trend observed since the 1989 survey. The median balances of holders of CDs fell more than 40 percent over 1989-95. Modest increases were registered for only a few demographic groups. The decreased popularity of CDs likely reflects the continuing low interest rates in a period when returns to other assets, particularly equities, were rising sharply.

Savings Bonds and Other Bonds

Overall, the percentage of families owning savings bonds and the median amount of their holdings rose slightly between 1992 and 1995. Increases in ownership were marked for families with heads aged 65 to 74 and for families with heads working as machine operators and laborers or in self-employed occupations.

In contrast, direct ownership of other types of bonds -- that is, excluding bonds held either through mutual funds or as a part of a retirement account -- fell from 4 percent of families in 1992 to 3 percent in 1995; the median holding also fell sharply, from $32,600 to $26,200. The declines in direct ownership and in median holdings were shared by most demographic groups. As with CDs, the fall in bond holdings is part of a longer trend that probably reflects a response to lower interest rates.

Publicly Traded Stocks

The percentage of families having direct ownership of publicly traded stocks -- that is, owning stocks other than those held either through mutual funds or as a part of a retirement account -- held steady from 1989 to 1992 and then fell for most demographic groups. The exceptional groups were families with incomes of $10,000 to $24,999 and those with heads aged 65 and older.

Moreover, for families that continued to hold stocks, the median value of holdings in 1995 was down a little from the 1992 level but identical to the corresponding 1989 figure. The decline in direct stockholding is part of a broader shift in the ways that families own stocks (see section "Holdings of All Types of Publicly Traded Stocks").

Mutual Funds

Continuing a very strong trend seen between 1989 and 1992, the 1995 results show expansion in the ownership of mutual funds of all types (not including money market funds and funds held as a part of a retirement account). The median value of holdings also continued upward. These changes are not surprising given the run-up in the stock market, the surge in the number of mutual funds available, and the intense marketing of funds.

Changes at a more disaggregated level are mixed. The ownership rate moved up for non-Hispanic whites but remained unchanged for other families. Ownership rates rose most strikingly at the top of the income distribution and for families with heads aged less than 55. Median holdings show a different picture, however, with older families and families at both the top and bottom of the income distribution showing the largest increases.

Retirement Accounts

The survey questions on retirement accounts cover Keogh accounts, individual retirement accounts; and employer-sponsored plans from which loans or withdrawals can be made, such as 401(k) accounts. Over 1989-95, the proportion of families owning these assets rose strongly, and the share of families' financial assets accounted for by retirement assets also rose. These assets complicate straightforward interpretation of families, portfolios because they may comprise holdings of stocks, bonds, mutual funds, real estate, limited partnerships, or virtually any other type of asset.

The percentage of families with retirement accounts grew in almost every demographic group between 1992 and 1995. Although some of these changes are not large, they continue a strong trend noted between 1989 and 1992. Median holdings of the demographic groups shown in the table moved in ways that have no obvious systematic interpretation. However, the median grew strikingly for families with incomes of $100,000 or more.

Other types of retirement plans, particularly defined-benefit plans, provide an annuity income in retirement based on workers, wages and years of service. Because these annuity benefits are difficult to value and frequently depend on complex assumptions about future work decisions, they are excluded from the asset figures reported here.(12)

In general, coverage by any type of employer-sponsored pension plan remained fairly constant over 1989-95: Around 40 percent of all families had coverage from a current job. However, the type of coverage has shifted considerably. The percentage of families participating in a 401(k)-type plan increased dramatically over the period, with 19 percent of families covered under such a plan in 1989 and 27 percent in 1995. At the same time, coverage by defined-benefit plans declined from 28 percent in 1989 to 19 percent in 1995. The shift toward 401(k)-type plans places a more obvious demand on workers to plan for their retirement.

Participation in 401(k)-type plans is voluntary. According to the 1995 SCF, slightly more than one-fourth of family heads who were eligible to participate in such a plan failed to do so in 1995. The data indicate that this choice is related strongly to income: Heads of families with incomes of less than $25,000 were less likely to participate than others. Interestingly, among the group of workers who chose not to participate in these plans, almost half were covered by a defined-benefit plan. Among participants in 401(k)-type plans, less than one-fourth were covered by a defined-benefit plan.

Cash Value Life Insurance

Cash value life insurance combines an investment vehicle with insurance coverage in the form of a death benefit. In recent years, policies normally referred to as "universal life" have added more complex investment elements to the traditional policy. The proportion of families that held cash value insurance declined between 1992 and 1995, continuing the downward movement between 1989 and 1992. This decline may reflect increased competition from the pure insurance coverage offered by term insurance as well as the attraction of other types of investments. For those holding cash value insurance, however, the median cash value rose overall and for most of the demographic groups shown -- a trend that did not appear in the earlier data. The decrease in use together with the increase in median value suggests that families who have such insurance may be using it more intensively as an investment vehicle or may have held it for sufficient time to have accummulated large cash value.

Remaining Financial Assets

Ownership of other managed assets (personal annuities, trusts, and managed investment accounts) and other financial assets (a heterogeneous category including such items as oil and gas leases, futures contracts, royalties, proceeds from lawsuits or estates in settlement, and loans made to others) showed no important movements from 1992 to 1995, just as they had shown no systematic changes from 1989, to 1992. The overall median of other managed assets moved up, but the pattern of changes over demographic groups suggests no clear explanation for this finding. A strong upward shift in the share of such assets invested in stock suggests that the increase in the median may be due to asset appreciation. The median holding of other financial assets was little changed over the period.

Holdings of All Types of Publicly Traded Stocks

As noted earlier, families may hold stock many different ways -- through direct ownership, through a mutual fund, through a retirement account, or through a trust or other type of managed investment account. Data from the three SCFs between 1989 and 1995 show that when all these types of ownership are combined, stock holding rose dramatically (table 6). The proportion of families owning stock through any means rose from 32 percent in 1989 to 37 percent in 1992 to 41 percent in 1995. The median equity holding also moved up, as did the share of stock in total financial assets. Moreover, all income and age groups saw a substantial increase in stocks as a share of financial assets, the shares doubled for families with incomes of less than $25,000.

6. Direct and indirect stock ownership, by selected characteristics of families, 1989, 1992, and 1995

Percent except as noted

                             Families having direct
    Family                 or indirect stock holdings
 characteristic
                               1989   1992   1995

All families                   31.7   37.2   41.1

Income (1995 dollars)
Less than 10,000                2.3    6.9    6.0
10,000-24,999                  13.1   19.4   25.3
25,000-49,999                  33.1   41.6   47.7
50,000-99,999                  54.0   64.1   66.7
100,000 and more               79.7   79.1   83.9

Age of head (years)
Less than 35                   23.3   28.3   38.5
35-44                          40.5   42.2   46.7
45-54                          40.2   47.3   49.3
55-64                          34.2   44.8   41.4
65-74                          26.1   31.9   34.0
75 and more                    24.7   28.1   28.1

                        Median value, among families
                                with holdings
    Family               (thousands of 1995 dollars)
 characteristic
                             1989   1992   1995

All families                 10.4   11.5   13.5

Income (1995 dollars)
Less than 10,000             35.0    5.9    4.0
10,000-24,999                 9.2    4.3    5.0
25,000-49,999                 5.5    7.6    8.0
50,000-99,999                10.4   14.6   21.3
100,000 and more             47.9   74.6   90.8

Age of head (years)
Less than 35                  3.7    3.8    5.4
35-44                         6.3    8.1    9.0
45-54                        12.3   14.4   24.0
55-64                        18.6   25.3   20.0
65-74                        25.8   21.7   25.0
75 and more                  28.2   27.1   28.1

    Family                 Stock holdings' share
 characteristic         of group's financial assets

                             1989   1992   1995

All families                 26.3   34.4   40.4

Income (1995 dollars)
Less than 10,000             10.0   15.2   21.1
10,000-24,999                10.3   18.6   21.6
25,000-49,999                20.3   25.4   33.0
50,000-99,999                25.6   35.1   39.9
100,000 and more             31.4   40.2   47.6

Age of head (years)
Less than 35                 25.4   25.6   32.4
35-44                        25.6   30.8   41.4
45-54                        29.9   39.4   44.2
55-64                        28.4   37.3   45.3
65-74                        26.2   34.4   34.3
75 and more                  20.7   28.6   39.5

Note. See note to table 1 and note 1, table 3.

Nonfinancial artificial Assets

The counterpart to the strong growth in financial assets of families from 1989 to 1995 was the drop in the share of their assets held in nonfinancial forms (table 7). The composition of family nonfinancial assets changed as well, most notably in the decline in the share of such assets allocated to investment real estate and the rise in the shares of business assets and vehicles. Nonetheless, the primary residence remained the largest single part of families' nonfinancial assets.

7. Distribution of value of nonfinancial assets of all families, by type of asset, 1989, 1992, and 1995

Percent

  Nonfinancial asset           1989   1992   1995

Vehicles                        5.4    5.5    6.9
Primary residence              44.1   45.7   45.9
Investment real estate         21.4   21.4   16.6
Business assets                27.0   25.7   29.0
Other nonfinancial assets       2.1    1.6    2.1
  Total                        100    100    100

Memo
Nonfinancial assets
  as a share of total assets   72.1   69.5   65.9

A few general patterns in ownership of the types of nonfinancial assets are considered here (table 8). Ownership rises with income, it also tends to rise with the age of family head and then may decline in the older age groups. Similar patterns are observed in the median holdings of these assets over income and age groups. Overall, the percentage of families in various groups owning some type of nonfinancial asset changed little between 1992 and 1995. At the same time, the median value of the holdings rose for almost all groups, with the principal exceptions of families at the top and bottom ends of the income distribution.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Vehicle,

Ownership of vehicle@, automobiles, motorcycles, vans, trucks, sport utility vehicles, motor homes, recreational vehicles, airplanes, and boats) fell somewhat from 1992 to 1995, but these items remained the most widely held nonfinancial assets. The decline in ownership was spread over most of the demographic groups, although ownership rose for families with incomes of less than $10,000 and those headed by machine operators and laborers and by service workers. In contrast, the median value of vehicles owned rose from $7,400 to $10,000 over the three-year period, and it rose for every group.

A part of the decrease in the percentage of families owning vehicles may be attributed to an increase in the percentage leasing vehicles, which rose from 3 percent in 1992 to about 5 percent in 1995. Most of the increase was concentrated among families with incomes of $25,000 or more. Among the group with incomes of $100,000 or more, the proportion of families leasing a vehicle surged from 10 percent in 1992 to 17 percent in 1995. Although the share of families leasing vehicles is fairly small, leased vehicles represented 35 percent of all new vehicles acquired by families in 1995, up from 22 percent in 1992.(13)

Primary Residence

Between 1992 and 1995, homeownership moved up slightly, continuing a trend from 1989-92. The median home value showed a similar pattern. Over demographic groups, changes in ownership rates were mixed. For owners, the median house value generally rose for families with incomes of less than $100,000 and fell somewhat for higher-income families. This difference may partly reflect low rates of price appreciation for more expensive houses.

Investment Real Estate

Overall, ownership of investment real estate (vacation homes, rental units, commercial property, farm land, undeveloped land, and all other types of real estate except primary residences and property owned through a business) fell nearly 2 percentage points from 1992 to 1995. The 1992 figure in turn was nearly 1 percentage point lower than the 1989 level.

The decline in ownership was widespread. All income groups except the lowest were less likely to have investment real estate in 1995 than in 1992. By age groups, the decrease was concentrated among families with heads aged less than 65, and the change was especially striking for the 55 to 64 age group. The changes in median holdings showed no clear pattern between 1992 and 1995; however, increases for the groups with heads between the ages of 55 and 74 seem to continue a trend apparent since 1989.

Business Assets

The fraction of families owning business assets (sole proprietorships, limited partnerships, other types of partnerships, subchapter S corporations, other types of corporations that are not publicly traded, and other types of private businesses) fell slightly between 1989 and 1995. Ownership declined appreciably for families having incomes of $100,000 and more and for nonwhite and Hispanic families. In contrast, from 1989 to 1992. ownership of businesses rose for non-white and Hispanic families.

The median value of business assets fell sharply over the entire six-year period. The median declined between 1992 and 1992 for all age groups and for all income groups except those at each end of the income distribution. Median holdings declined markedly for families headed by self-employed workers, but the composition of this group may have changed between 1992 and 1995 through growth in the portion of the group consisting of "small-scale" entrepreneurs.

Other Nonfinancial Assets

For the remaining nonfinancial assets (a broad category of tangible assets including artwork, jewelry, precious metals, and antiques), ownership rates rose a bit between 1992 and 1995, while the median amounts of such assets rose more strongly. The median value of these assets rose particularly among older and lower-income families.

Unrealized Capital Gains

Along with saving out of current income, changes in the values of assets such as businesses, real estate, and stocks are a determinant of family wealth. Until these assets are sold, and thereby appear as part of family income, any gains are unrealized. To measure this part of wealth, the survey obtains information about changes in value from the time of purchase of a primary residence, investment real estate, businesses, publicly traded stock, and mutual funds. Despite the, large increase in stock prices from 1989 to 1995, median and mean unrealized capital gains fell overall in the 1992 and 1995 surveys (table 9). The declines were apparently driven by substantial declines of unrealized gains in housing and businesses.

9. Family unrealized capital gains, by selected characteristics of families, 1989, 1992, and 1995.

Thousands of 1995 dollars

                              1989              1992
      Family
  characteristics        Median     Mean    Median    Mean

All families              11.7      86.9      8.1     76.8

Income (1995 dollars)   [dagger]     9.7   [dagger]   14.6
10,000-24,999               .6      33.4      1.2     29.2
25,000-49,999             12.3      52.1      6.7     86.8
100,000 and more         132.7     552.2    138.2    497.4

Age of head (years)
Less than 35            [dagger]    26.4   [dagger]   16.1
35-44                     13.1      66.2      5.4     58.7
45-54                     41.1     144.8     19.5    112.4
55-64                     36.2     133.2     31.5    142.3
65-74                     31.4     132.6     32.5    122.7
75 and more               19.2      98.1     27.1     71.9

                             1995
      Family
  characteristics        Median    Mean

All families               6.1     68.8

Income (1995 dollars)   [dagger]   18.9
10,000-24,999               .2     26.2
25,000-49,999             27.0     63.1
100,000 and more          75.7    574.9

Age of head (years)
Less than 35            [dagger]   12.0
35-44                      4.5     39.8
45-54                     18.8     96.0
55-64                     30.0    133.5
65-74                     30.1    112.6
75 and more               33.0     95.4

Note. See note to table 1 and note 1, table 3. [dagger] Less than $50.

Liabilities

Family debt and family assets both rose strongly from 1989 to 1995. As a result, family debt as a proportion of assets (the leverage ratio) held fairly steady at about 16 percent over the period (table 10). Home mortgages and home equity borrowing as a share of total family debt grew strongly, however, probably because of wider home ownership and a continuing shift of debt into these tax-deductible forms. The share of credit card debt also grew strongly between 1992 and 1995, but it remained a small part of total family debt. Offsetting these increases was a strong decline in the share of borrowing for investment real estate. The share of installment borrowing declined between 1989 and 1992 and then rose slightly between 1992 and 1995.(14)

10. Distribution of amount of debt of all families, by type of debt, 1989, 1992, and 1995

Percent

      Type of debt                      1989   1992    1995

Home mortgage and home equity
   loans and lines of credit            56.6   64.2    68.2
Installment loans                       13.9   10.3    11.2
Other lines of credit                    1.0     .7      .4
Credit card balance                      2.3    2.8     3.7
Investment real estate mortgages        24.5   19.8    14.4
Other debt                               1.     2.2     2.2
     Total                             100    100     100

Memo
Debt as a percentage of total assets    15.9   16.3    16.0

Families' Holdings of Debts

Overall, the proportion of families with debt rose slightly between 1989 and 1992 and then more substantially by 1995 (table 11). Following a similar pattern, the median amount of debt outstanding rose 15 percent from 1992 to 1995 after having been flat over the previous three years. The increases between 1992 and 1995 in both the prevalence of borrowing and the median amount of debt owed would normally be expected in a period of economic expansion. The increases were spread widely over demographic groups, with the salient exceptions of families in the highest income group and families with self-employed heads.

The prevalence of debt tends to increase with family income, but the sizes of the increases are fairly small as the level of income rises above $25,000. The median amount of debt owed shows much larger increases with income, likely because of borrowing associated with the acquisition of nonfinancial assets. By age group, the proportion of families borrowing varies only a little for the groups with heads younger than 65, but it falls off quickly after that. The drop-off in median borrowing in these older groups is even sharper. The age pattern is largely explained by the paying off of mortgages on primary residences.

Home Mortgages and Home Equity Borrowing

For homeowners, mortgages serve at least two purposes: a means of paying for the home and a substitute for other borrowing. Since the Tax Reform Act of 1986, which phased out the deductibility of non-mortgage debt, loans secured by home equity (traditional mortgages, home equity loans, and home equity lines of credit secured by a primary residence) have increasingly served as a source of tax-preferred funds. Declining interest rates between 1992 and 1995 gave families yet another incentive to refinance existing mortgages and obtain funds for other purposes at the same time.

The proportion of families borrowing through mortgage loans in 1995 was up slightly from the 1989 level, but the median amount outstanding rose about 30 percent over the six-year period. Over the same period, the median value of a primary residence rose only 4.8 percent; the much larger rise in the size of mortgage debt suggests that families were using more of their home equity for purchases or investments other than the purchase of their primary residence. Both the prevalence of mortgage borrowing and the median amount owed rose for most demographic groups.

Not included in the debt figures shown for mortgages are families that have a home equity line of credit that was not being used at the time of the survey. Home equity lines of credit became a less important source of credit between 1992 and 1995; both the proportion of families with such a credit line and the proportion using the line declined.

Installment Borrowing

The share of families using installment borrowing was lower in 1995 than in 1989. Such a decline during a period of strong growth in borrowing probably reflects a substitution to other types of financing, particularly mortgages, credit card debt, or automobile leasing.

The median amount of installment debt rose 22 percent between 1992 and 1995, and much of the increase was associated with borrowing for automobiles. The median amount of installment debt rose for most groups, with the exceptions of families with incomes of $100,000 or more and families with self-employed heads.

Borrowing on Other Lines of Credit

The use of personal lines of credit other than home equity lines declined from 1989 to 1995 for almost every demographic group. On the demand side, the decline may reflect a strong increase in the use of credit cards or a rise in mortgage refinancing. On the supply side, many lenders stopped offering unsecured lines. Changes in the dollar amount of this type of credit are difficult to evaluate because the instrument is so narrowly used.

Credit Card Borrowing

In 1989, 40 percent of families had an outstanding balance on some type of credit card -- bank-type cards (such as Visa, Mastercard, Discover, and Optima), store and gasoline company cards, travel and entertainment cards (such as American Express and Diners Club), and other credit cards -- after paying their most recent bill. The proportion of families with such debt rose over the six years, to 48 percent in 1995. This strong growth likely reflects, among other things, the decline in some credit card interest rates and the intensive marketing of cards by issuers in recent years.

From 1992 to 1995, the use of some type of credit card for borrowing and the median amount borrowed rose for almost all demographic groups. Exceptions are the groups of families with heads aged 65 or over. However, the declines for those families do not entirely reverse the increases observed for these groups between 1989 and 1992.

Much of the growth in credit card debt over the past several years has been in bank-type cards. Between 1992 and 1995, continuing a trend from 1989, the share of families having such cards rose (from 62 percent to 48 percent), as did the fraction of cardholders running a balance on them (from 53 per cent to 56 percent). The median balance for those with positive balances rose almost 40 percent, from $1,100 in 1992 to $1,500 in 1995. At the same time, the median total credit limit on such cards jumped from $5,400 per cardholding family in 1992 to $9,000 in 1995. As a consequence, the median proportion of the credit limit used by those with balances fell from 28 percent in 1992 to 23 percent in 1995.

Investment Real Estate Debt

Reflecting in part the decline in the proportion of families having investment real estate, the percentage of families borrowing to finance such assets declined from 1992 to 1995 for most groups, whereas the median amount owed by families that had such loans moved up 8 percent. This divergence in prevalence and the amount owed suggests that the families that dropped out of the market were those with smaller holdings.

Other Debt

Other borrowing (loans on insurance policies, loans against pension accounts, borrowings on a margin account, and other unclassified loans) rose slightly overall between 1992 and 1995. The median amount of such debt fell 26 percent, however. This movement results largely from decreases in the amounts borrowed against employer-sponsored pension accounts and against the cash value of life insurance. The use of margin accounts and other loans was little changed.

Reasons for Borrowing

The SCF provides detailed information on the reasons that families borrow money (table 12).(15) One subtle problem with the use of these data is that, even though money is borrowed for a particular purpose, it may only offset the use of other funds. For example, a family may have sufficient assets to purchase a home without using a mortgage, but it may choose not to do so in order to invest those other funds in another asset. Thus, trends in the data can only be suggestive of the underlying use of funds by families.

The survey shows that the proportion of total borrowing directly attributable to home purchase has risen dramatically -- from 53 percent in 1989 to 65 percent in 1995 -- a rise nearly equaled by the size of the decline in borrowing for investment real estate. Borrowing for other investment also declined over the period. The share of credit attributable to vehicle borrowing shows a cyclical pattern. Other reasons for borrowing show little change over the period.

In 1995 the SCF for the first time gathered information on the use of the funds families obtained from refinancing their mortgages. Approximately 10 percent of all families with mortgages in 1995 had refinanced their current mortgage at some time. Among this group, the major uses reported for the funds were home improvements or repairs (42 percent of refinancers), payment of bills or bill consolidation (21 percent), investments (12 percent), education (5 percent), and vehicle purchases (4 percent).

12. Distribution of amount of debt of all families,

by purpose of debt, 1989, 1992, and 1995

Percent

        Purpose of debt           1989      1992    1995

Home purchase                     53.0      59.9    65.2.
Home improvement                   2.0       3.3     1.9
Investment, excluding real estate  2.6       1.6      .9
Vehicles                           8.6       6.2     7.1
Goods and services                 4.8       5.0     5.2
Investment real estate            25.8      20.5    15.3
Education                          1.9       2.5     2.5
Unclassifiable loans against
    pension accounts                .1        .2      .2
Other unclassifiable loans         1.1       2.0     1.8

Total                            100       100     100

Choice of Lenders

As one might expect in light of the continuing restructuring of the financial services industry, large shifts occurred in consumers, borrowing patterns (table 13). With the ongoing contraction of the sayings and loan industry, lending at savings and loans and savings banks as a share of total debt held by families declined between 1989 and 1995. Partly offsetting this decline was a surge in lending by finance companies that was particularly strong from 1992 to 1995. The market shares of banks and credit unions grew steadily from 1989 to 1995.(16)

13. Distribution of amount of debt of all families, by type

of lending institution, 1989, 1992, and 1995

Percent

      Type of institution            1989     1992     1995

Commercial bank                      29.4     33.2      35.1
Savings and loan or savings bank     23.4     16.6      11.3
Credit union                          3.3      4.0       4.2
Finance or loan company               9.6     13.5      21.0
Brokerage                             2.9      3.3       1.9
Real estate lender                   13.3     14.2      12.9
Individual lender                     6.8      4.5       4.4
Other nonfinancial                    1.3      1.5        .7
Government                            2.0      1.9       1.7
Credit and store cards                2.3      3.0       3.7
Loans against pension accounts         .1       .2        .2
Other unclassifiable loans            5.4      4.2       2.9

Total                               100      100       100

Debt Burden

The ability of families to service their loans is determined by the size of the payments they are required to make and the amount of income or assets they have available to make those payments. As noted above, the surveys measured before-tax family income for the preceding calendar year. Between the 1992 and 1995 surveys, median and mean income moved up, and the amount borrowed rose. By the end of the period, most interest rates on newly originated loans were lower than they were at the start. Thus, it is not obvious how the ability of families to service their loans out of their current income has changed or, indeed, if it has.

On net, the survey indicates that the overall ratio of the total debt payments of all families to total family income -- a common measure of "debt burden" -- changed only slightly between 1989 and 1995 (table 14).(17) When this measure is computed separately for different income groups, families with incomes below $10,000 show an increase in debt burden between 1992 and 1995, and families with incomes of at least $100,000 show a large decrease. The ratio did not move appreciably for other families.

This aggregate measure may not adequately describe the typical debt burden of families with debt. A better indication of typical debt burden is the median ratio of debt payments to income computed for debtors alone. For all families with debt, this measure moved up by I percentage point from 1992 to 1995. For debtors with incomes below $50,000 the ratio rose, and for those with incomes of $100,000 and more it fell.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Two indicators of potential financial distress are the share of families with debt who have payments exceeding 40 percent of their income and the share who were late with their payments by sixty days or more at least once in the preceding year. The 1992-95 period saw little change in the proportion of highly indebted families (that is, those with payments exceeding 40 percent of their income), but the proportion of debtors who were late payers rose nearly 1 percentage point.(18)

Summary

Data from the 1995 SCF provide a detailed view of recent changes in the income and net worth of families. Although median and mean family income reported in 1995 had risen above the cyclically depressed levels of the 1992 survey, neither measure had reached the level reported in 1989. Median net worth reported in 1995 also had returned to a level nearly the same as that in 1989. However, mean net worth remained below its 1989 level.

Among families' assets, the relative importance of corporate equities and of retirement accounts rose steadily from 1989 to 1995. During this time, the home ownership rate rose about 1 percentage point, while median home values increased slightly. The median value of business assets declined sharply over the six-year period. Unrealized capital gains also fell over the six-year period, mostly because of substantial declines in the unrealized gains in homes and businesses.

Although the ratio of debts to assets changed only slightly over this period, two shifts in borrowing are particularly interesting. First, mortgage debt grew strongly as a share of family debt, likely reflecting the increase in the home ownership rate as well as refinancing related to declining mortgage interest rates between 1992 and 1995. Second, both the proportion of families using credit cards for borrowing and the median balances outstanding on cards rose markedly from 1989 to 1995.

Appendix: Survey Procedures And

Statistical Measures

Since 1989, the questionnaires for the SCF have changed only slightly. Generally, changes have been introduced to gather additional information needed to understand other data in the survey -- for example, the 1995 SCF introduced a question on uses of funds for mortgages that were taken out after the time a primary residence was purchased. Also, the major aspects of the sample design have been fixed over this time. Thus, the information obtained by the survey is comparable over 1989-95.

The survey is intended to provide an adequate descriptive basis for the analysis of family assets and liabilities. To address this requirement, the SCF combines two types of samples. First, a standard multi-stage area-probability design is selected to provide good coverage of characteristics, such as home ownership, that are broadly distributed in the population. Second, a special list sample is included to over-sample wealthy families, who hold a disproportionately large share of such assets as non-corporate businesses and tax-exempt bonds. This list sample is drawn from a sample of tax records made available for this purpose under strict rules governing confidentiality, the rights of potential respondents to refuse participation in the survey, and the types of information from the interview that can be made generally available. Of the 3,906 completed interviews in the 1992 SCF, 2,456 families were from the area-probability sample and 1,450 were from the list sample; the comparable figures for the 4,299 interviews completed in 1995 are 2,780 families from the area-probability sample and 1,519 from the list sample.(19)

A very important factor in the ability to conduct surveys is the generosity of the public in giving their time for an interview. In the 1995 SCF, the average interview required -O minutes. However, for some particularly complicated cases, the amount of time needed was substantially more than two hours.(20)

Data for the 1992 and 1995 surveys were collected by the National Opinion Research Center at the University of Chicago (NORC) between the months of June and December in each of the two years. The great majority of interviews were conducted in person, although interviewers were allowed to conduct telephone interviews if that was a better arrangement for the respondent. In the 1995 survey, one important change was the introduction of laptop computers for use in administering the questionnaire. This change increased the length of the interview somewhat, and it may also have had some effects on the quality of information collected.(21) Nonetheless, the effects of the change in the mode of questionnaire administration appear to be fairly small.

Errors may be introduced into survey results at many stages. Sampling error, the variability expected to occur in estimates based on a sample instead of a census, is a particularly important source of error. Such error may be reduced either by increasing the size of the sample or by designing the sample to reduce important types of variability; the latter course has been chosen for the SCF. Estimation of sampling error in the SCF is described further below.

Interviewers may introduce errors, though SCF interviewers are given lengthy project-specific training to minimize this problem. In addition, computer control of the 1995 survey greatly reduced technical errors made by interviewers. Respondents may introduce error by understanding a question in a sense different from that intended by the survey designers. For the SCF, extensive pretesting and other review of questions tend to reduce this source of error.

Nonresponse -- either complete nonresponse to the survey or nonresponse to selected items within the survey -- may be another important source of error. As noted in more detail below, the SCF uses weighting adjustments to compensate for complete non-response. To deal with missing information on individual items, the SCF uses statistical methods to impute missing data.(22)

Response rates differ markedly in the two parts of the SCF sample. In both 1992 and 1995, about 70 percent of families selected for the area-probability sample actually completed interviews. The overall response rate in the list sample was about 34 percent. Detailed analysis of the data suggests that the tendency to refuse participation in an interview is highly correlated with wealth. The response rates for both samples are low by the standards of other major government surveys. However, unlike other surveys, which almost certainly also have differential non-response by wealthy families, the SCF sample frame provides a basis for adjusting for nonresponse by such families. To provide a measure of the frequency with which families similar to the sample families could be expected to be found in the population of all families, analysis weights are computed for each case to account for both the systematic properties of the design and for nonresponse. A major part of research by SCF staff is devoted to adjustments for nonresponse through the analysis weights for the survey.(23)

For this article, the weights were further adjusted to decrease the possibility that the results could be overly affected by a small number of observations. Such influential observations were detected using a graphical technique to inspect the underlying data. Most of the cases found were holders of an unusual asset or liability or were members of demographic groups for which such holdings were rare. Trimming the weights of such cases is likely to make the key findings in the article more reliable.

To estimate the standard errors due to sampling that are reported in the main part of this article, a replication technique was used. Replication methods draw samples from the set of actual respondents in a way that incorporates the important dimensions of the original sample design. In the SCF, a bootstrap procedure was used to select 1,000 sample replicates, and a new weight was computed for all the cases in each of the selected replicates.(24)

Comparable standard errors are not available for the 1989 survey. Weights for the 1989, SCF were computed using an earlier version of the methodology applied in 1992 and 1995. Although estimates using these 1989 weights are comparable to estimates from the later surveys, the difference in the weight construction is a source of variability in comparisons. An ongoing project computes weights for the 1989 SCF using exactly the same methodology as that applied in the later surveys. Corresponding bootstrap replicates and their weights will also be computed. A set of tables for the 1989 data reporting the major detailed estimates presented in this article for 1992 and 1995 will be published in a later issue of the Federal Reserve Bulletin.

Generally, the survey data correspond well to external estimates, when such information is available. Comparisons of SCF data with aggregate data from the Federal Reserve flow of funds accounts suggest that when proper adjustments are made to achieve conceptual comparability, these aggregate estimates and the SCF estimates for 1989 and 1992 are very close.(25) In general, only medians from the SCF can be compared with those of other surveys because of the special design of the SCF sample.

The definition of "family" used throughout this article differs from that typically used in other government studies. In the SCF, a household unit is divided into a "primary economic unit" (PEU) -- the family -- and everyone else in the household. The PEU is intended to be the economically dominant single individual or pair of individuals (who may be married or living as partners) and all other persons who are financially dependent on that person or those persons. In other government studies, for example, those of the Bureau of the Census, a single individual is not considered a family. As noted in the main text, the Census definition of household is closer to the SCF definition of family. The term "head" used in this article is an artifact of the organization of the data and implies no judgment about the structure of families. In a PEU containing only a single economically dominant individual, the head is taken to be that individual. In other PEUs, the head is taken to be the male in the core couple of the PEU or the older person in a same-sex couple.

The data used in this article from the 1989 and 1992 SCFs are derived from the final versions of those surveys. Results reported in this article may differ in some details from earlier results based on the preliminary data from those surveys. The 1995 data used here represent the best estimates at the current advanced stage of data processing. Data from the 1995 SCF, in a form designed to protect the privacy of respondents, should be available in the first half of 1997 from the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161, [phone].

(1.) for a detailed discussion of the results from the 1989 and 1992 SCF, see Arthur B. Kennickell and Martha Starr-McCluer, "Changes in Family Finances from 1989 to 1992," Federal Reserve Bulletin, vol. 80 (October 1994), pp. 861-82.

(2.) All other aggregate statistics cited in this section are for September unless otherwise noted. This month was the midpoint of the periods during which interviews were conducted.

(3.) If the PCE deflator is used instead of the CPI to adjust income to 1995 dollars, the overall decline since 1989 is a little less pronounced. Using this alternative index, median (mean) income is $31,400 ($49,200) in the 1989 survey, $28,800 ($43,100) in the 1992 survey, and $30,700 ($44,100) in the 1995 survey.

(4.) See Paul Ryscavage, "A Surge in Growing Income Inequality?" Monthly Labor Review, vol. 118 (August 1995), pp. 51-61, for a discussion of the effects of various factors on the measured income distribution. A general description of the CPS redesign may be found in Anne E. Polivka, "Data Watch: The Redesigned Current Population Survey," Journal of Economic Perspectives, vol. 10 (Summer 1996), pp. 169-80.

(5.) Published data from the CPS show roughly similar changes over the six-year period for the groups without a college degree. For families with heads having at least a college degree, the CPS data indicate that median income declined slightly between 1989 and 1995 but, in contrast to the SCF estimates, that the mean income of this group rose over the period. In addition to the changes in the CPS noted earlier, a change between 1989 and 1992 in the education data collected in the CPS further complicates the comparison of income for the group with a college degree.

(6.) Data from the CPS, for example, indicate that the fraction of households that identified themselves as nonwhite or Hispanic increased from 1989 to 1995. The CPS gathers this information through a two-part question, asking first about racial identification and second about Hispanic origin. In the SCF, respondents are asked to choose the group with which they most identify from a list that includes "Hispanic" as a choice.

(7.) See Arthur B. Kennickell, Saving and Permanent Income, Finance and Economics Discussion Series 95-41 (Board of Governors of the Federal Reserve System, 1995), for a more detailed discussion of this variable.

(8.) All families were asked to report a saving motive regardless of whether they were currently saving.

(9.) The asset values reported in this article do not include any adjustments for future tax liabilities. For example, a family that sold its stock would be required to pay taxes on any increase in the value of the stock.

(10.) As with family income, the choice of the CPI or the PCE deflator to adjust the data makes little difference to the overall conclusions. If the data are adjusted to 1995 dollars using the PCE deflator, median (mean) net worth is $55,800 ($213,300) in 1989 and $52,200 ($197,609) in 1992.

(11.) In discussing the dollar holdings of detailed components of net worth of the demographic groups considered in this article, we present only the median amounts held. In general, the median is a better indicator of typical holdings of an item than the mean, especially when ownership of the item is concentrated among a relatively small number of families.

(12.) For a calculation of net worth that includes the value of annuity pension benefits, see Arthur B. Kennickell and Annika E. Sunden, "Pensions, Social Security, and the Distribution of Wealth" (Board of Governors of the Federal Reserve System, 1996).

(13.) For additional evidence on vehicle leasing. see Ana Aizcorbe and Martha Starr-McCluer, Vehicle Ownership, Vehicle Acquisitions, and the Growth of Auto Leasing, Finance and Economics Discussion Series 96-35 (Board of Governors of the Federal Reserve System, 1996).

(14.) The term "installment borrowing" in this article describes consumer loans that generally have fixed payments and a fixed term, such as a standard automobile loan. This usage contrasts with alternative usages that include most types of nonmortgage borrowing by households.

(15.) For all but two of the types of loans covered by the 1995 SCF, respondents were queried about the use of the funds borrowed. It was deemed infeasible to ask the purposes for funds borrowed with credit cards. For purposes of the analysis here, credit card debt is included in the category "goods and services." The second exception was first mortgages that were taken out when a property was obtained. In this case, it was assumed that the funds were used for the purchase of the home. In the earlier surveys, questions were also not asked about the use of funds obtained from any other type of first mortgage and from borrowing against a pension account.

(16.) For evidence on similar changes in borrowing patterns for small businesses. see Rebel A. Cole, John D. Wolken, and R. Louise Woodburn. "Bank and Nonbank Competition for Small Business Credit: Evidence from the 1987 and 1993 National Surveys of Small Business Finances." Federal Reserve Bulletin, vol. 82 (November 1996), pp. 983-95.

(17.) When calculated using macroeconomic data, this ratio declined from 16 percent at the time of the 1992 survey to a low of 15.3 percent in 1993. It had moved back up to 16.6 percent by the time of the 1995 survey. The ratio calculated from macroeconomic data uses an estimate of aggregate debt service payments. In contrast, the ratio calculated from SCF data uses information on debt service payments obtained directly from respondents. See Glenn B. Canner, Arthur B. Kennickell, and Charles A. Luckett, "Household Sector Borrowing and the Burden of Debt," Federal Reserve Bulletin, vol. 81 (April 1995), pp. 323-38. In the case of credit cards, the payment is estimated using an average market rate on credit cards.

(18.) The periodicity of the SCF obscures some important fluctuations in the frequency of payments problems. Industry data available on a monthly basis show a sharp drop in delinquency rates on consumer debt from 1992 through 1994. By the time of the 1995 survey, rates had increased to levels somewhat higher than those in 1992, and they have continued to rise since then.

The measure of late payments in the SCF differs conceptually from the aggregate delinquency rate in some important respects. Whereas the delinquency rate records late payments on each loan in a given period, the survey asks respondent., whether they have been late or behind in any of their payments during the past year. Thus a person with three delinquent loans would be counted three times in the aggregate data but only once in the SCF.

(19.) The 1992 SCF represents 95.9 million families; the 1995 SCF represents 99.0 million families.

(20.) The role of interviewers in this effort is normally not sufficiently recognized. Without the dedication and perseverance of the project field staff, the survey would not have been possible.

(21.) For more information on the effects of computer-assisted interviewing in the 1995 SCF, see Arthur B. Kennickell, "Using Range Techniques with CAPI in the 1995 Survey of Consumer Finances" (Board of Governors of the Federal Reserve System, 1996).

(22.) See Arthur B. Kennickell, "Imputation of the 1989 Survey of Consumer Finances: Stochastic Relaxation and Multiple Imputation" (Board of Governors of the Federal Reserve System, 1991).

(23.) For a description of the weighting methodology, see Arthur B. Kennickell, Douglas A. McManus, and R. Louise Woodburn, "Weighting Design for the 1992 Survey of Consumer Finances" (Board of Governors of the Federal Reserve System, 1996). The weighting design for the 1995 survey is identical. A review of the 1995 weights will be available later in the year.

(24.) See Kennickell, McManus, and Woodburn, "Weighting Design," for details.

(25.) For the details of this comparison, see Rochelle Antoniewicz, "A Comparison of the Household Sector from the Flow of Funds Accounts and the Survey of Consumer Finances," Review of Income and Wealth (forthcoming).

RELATED ARTICLE: The Survey of Consumer Finances

The Survey of Consumer Finances (SCF) is a triennial survey sponsored by the Federal Reserve with the cooperation of the Department of the Treasury. It is designed to provide detailed information on U.S. families' balance sheets and their use of financial services, as well as on their pension rights, labor force participation, and demographic characteristics at the time of the interview. The survey also collects information on total family income, before taxes, for the calendar year preceding the survey. The term "family" as it is used here is more comparable to the U.S. Bureau of the Census definition of "household" than to Census's use of "family," which excludes single people. The appendix to this article provides a more detailed description of the SCF, including a more complete definition of this term.

The underlying statistical methodology of the surveys has been largely unchanged since 1989, and the questionnaires have been modified only slightly, mostly to reflect changes in the availability of financial services or in the financial behavior of families. Thus, the data since that time are comparable.

The need to measure financial characteristics imposes special requirements on the survey design. The survey must provide reliable information both on items that are broadly distributed in the population --- for example, vehicle ownership -- and on items that are highly concentrated in a relatively small part of the population -- for example, investment real estate. To address this problem, the SCF employs a dual-frame sample design that includes a standard geographically based random sample and a special oversample of relatively wealthy families. Weights are used to combine information from the two samples for estimates of statistics for the full population.

This article draws principally upon the final data from the 1992 SCF and preliminary data from the 1995 SCF. Any differences between figures for 1992 reported here and those published earlier in the Federal Reserve Bulletin are attributable to additional statistical processing of the data. To provide a larger context, some information is also included from the final version of the 1989 SCF. The SCFs for 1992 and 1995 were conducted by the National Opinion Research Center at the University of Chicago (NORC) between July and December of each survey year. For the 1992 survey, 3,906 families were interviewed, and for the 1995 survey, 4,299 were interviewed.

All dollar figures in this article are adjusted to 1995 dollars using the consumer price index (CPI) for all urban consumers. Concerns about how accurately the CPI represents inflation for families have been discussed in the literature. If, as generally supposed, the index overstates the true degree of inflation, upward adjustments to past dollar amounts will appear too large, and trends in positive dollar amounts will be biased downward. An index sometimes proposed as an alternative to the CPI is the implicit price deflator for personal consumption expenditure (PCE), which is reported as a part of the national income and product accounts. Over 1989-95, price changes measured by the CPI and the PCE deflator differed by a relatively small amount.

To provide a measure of the significance of the developments discussed in this article, standard errors due to sampling are given for selected estimates for the 1992 and 1995 data. Space limits have precluded reporting such figures for all estimates. No standard errors are given for the 1989 estimates for technical reasons mentioned in the appendix. Although we do not directly address the statistical significance of the results, the article highlights findings that are significant or are interesting in a broader context.

RELATED ARTICLE: Families without a Checking Account

The portion of families without any type of transaction account (checking, savings, and money market deposit accounts; money market mutual funds; and call accounts at brokerages) was about 13 percent in both 1992 and 1995 -- down from 15 percent in 1989.

More narrowly, the portion of families without a checking account fell continuously over the six-year period, from 18.9 percent in 1989 to 16.5 percent in 1992 to 15.1 percent in 1995. Among these families, 85.0 percent had incomes of less than $25,000 percent had incomes of less than $10,000), 59.5 percent had heads under age 45 (36.9 percent had heads under 35), and 54.0 percent were nonwhite or Hispanic. The survey asked all such families the reason they did not have an account, and several interesting patterns appear in the responses (table below).

From 1989 to 1995, a declining portion of such families reported that they did not write enough checks to make an account worthwhile, although this reason remained the most commonly reported. A sharp increase in the expressed dislike of banks appears in 1995. The proportion reporting that they had insufficient money for an account remained approximately unchanged. At the same time, the proportion reporting that they could not manage or balance a checking account almost doubled. The importance of service charges as a reason declined in 1995, and reporting of minimum balance requirements as a reason was unchanged from 1992.

Reasons reported by families without a checking account for not having one, 1989, 1992, and 1995

Percent

      Response category            1989    1992    1995

Do not write enough checks
   to make it worthwhile           34.3    30.4    27.1
Minimum balance is too high         7.6     8.6     8.6
Do not like dealing with banks     15.0    15.3    22.8
Service charges are too high        8.4    11.2     7.9
No bank has convenient hours
   or location                      1.2      .9     1.2
Do not have enough money           21.8    20.9    20.5
Cannot manage or balance
   a checking account               4.6     6.4     8.6
Other                               7.1     5.7     3.4
   Total                          100     100     100