---

Changes in U.S. Family Finances from 2004 to 2007: Evidence from the Survey of Consumer Finances

(Errata paragraph added on March 6, 2009; see p. A56)

---

Brian K. Bucks, Arthur B. Kennickell, Traci L. Mach, and Kevin B. Moore, of the Board's Division of Research and Statistics, prepared this article with assistance from Gerhard Fries, Daniel J. Grodzicki, and Richard A. Windle.

The Federal Reserve Board's Survey of Consumer Finances for 2007 provides insights into changes in family income and net worth since the 2004 survey.¹ The survey shows that, over the 2004-07 period, the median value of real (inflation-adjusted) family income before taxes was little changed; median income had grown slightly in the preceding three-year period (figure 1). Across most demographic

groups, the pattern of change was mixed, but a few changes stand out: Income increased markedly for Hispanic or nonwhite families, while it declined substantially for families living in the Northeast or the Midwest and for families headed by a person who was retired or otherwise not working. In contrast to median income, mean income in the recent period climbed 8.5 percent, and the increases were spread broadly across demographic groups. The increases were most striking for families in the top 10 percent of the distribution of net worth and for families headed by a single parent, a person who was self-employed, or a person who was aged 65 to 74. Over the preceding three years, mean income had declined broadly. Differences in the rates of change in the median and mean signal a change in the distribution of income.

Unlike family income over the 2004-07 period, both median and mean net worth increased; the median rose 17.7 percent, and the mean rose 13.0 percent (figure 2). The increases were fairly broadly spread, but with a number of noteworthy exceptions, some of which entailed changes in medians and means within demographic groups that differed substantially, either in terms of relative magnitude or in the direction of change. Median and mean net worth for the lowest 25 percent of

the distribution of net worth plunged 36.8 percent and 43.8 percent, respectively; median net worth for the lowest 20 percent of the distribution of income fell 1.2 percent, but the mean rose 31.8 percent. Percentage increases in median and mean net worth were similar for white non-Hispanic families, while the increase in the median for nonwhite or Hispanic families was only about one-fifth of that for other families, and the increase in the mean was nearly three times the size of that for other families. Relative to other regions, both the Northeast and the Midwest saw sizable declines in median net worth. The clearest gains in both median and mean net worth were for high-net-worth families, high-income families, families headed by a person aged 65 or older, and families headed by a person who worked for someone else or who worked in a technical, sales, or service occupation. In the preceding three years, median net worth had increased only slightly (1.0 percent), while the mean had risen more strongly (6.0 percent); over that time, the data had shown a more complex pattern of mixed increases and decreases in wealth.

Unrealized capital gains were a particularly important factor in the increase in net worth over the 2004-07 period. The share of total assets attributable to unrealized capital gains from real estate, businesses, stocks, or mutual funds rose 5.1 percentage points, to 35.8 percent in 2007. Although the level of debt owed by families rose noticeably, debt as a percentage of assets was little changed. The largest percentage change in debt was in borrowing for residential real estate other than a primary residence.

With median and mean debt advancing faster than income, payments relative to income might be expected to increase substantially. In fact, total payments relative to total income barely increased, and the median of payments relative to income rose at a slower pace than it did between 2001 and 2004. Nonetheless, the share of families with high payments relative to their incomes increased notably.

This article reviews these and other changes in the financial condition of U.S. families between 2004 and 2007.² The discussion draws on data from the Federal Reserve Board's Survey of Consumer Finances (SCF) for those years; it also uses evidence from earlier years of the survey to place the 2004-07 changes in a broader context.

Return to menu

Economic Background

Families' finances are affected by both their own decisions and the state of the broader economy. Over the 2004-07 period, real gross domestic product (GDP) increased, on average, about 2.5 percent per year. However, toward the end of 2007, the pace of economic activity slowed noticeably. The unemployment rate stood at 5.5 percent in mid-2004, fell to 4.5 percent by late 2006, and then increased to 5.0 percent at the end of 2007. The rate of inflation, as measured by the consumer price index for all urban consumers (CPI-U-RS), increased somewhat over the period, from an annual average of 2.7 percent in 2004 to 2.9 percent in 2007; the increase was driven, in part, by the escalation of food and energy prices.

Developments in financial markets over the three-year period were varied. The major stock market indexes climbed over most of the period before beginning a decline in late 2007; from September 2004 to September 2007, the Wilshire 5000 index rose 41.7 percent. Interest rates on new consumer loans generally increased; for example, the interest rate on a new 30-year fixed-rate mortgage averaged 5.75 percent in September 2004, when about one-half of the interviews for the 2004 survey had been completed, and was 6.38 percent three years later. Yields also rose on liquid deposits, time deposits, and bonds; for example, the rate on a three-month certificate of deposit rose from an average of 1.86 percent in September 2004 to 5.46 percent in September 2007.

The national purchase-only LoanPerformance Home Price Index , produced by First American Core-Logic, increased more than 12.4 percent between September 2004 and September 2007. Price increases varied sharply across areas of the country. The largest increase in the index was a 49.9 percent rise for Hawaii. While most states saw an increase, the index declined 8.0 percent for Michigan and by smaller amounts for Ohio, Rhode Island, and Massachusetts. Homeownership rates were little changed over the period after a long and steady increase. Nonetheless, the number of homeowners rose with population growth, and subprime mortgages are generally thought to have played an important part in financing home purchases.

No major tax legislation was passed during the period, but other important institutional changes occurred. The BankruptcyAbuse Prevention and Consumer Protection Act of April 2005 altered the rules for liquidation of consumers' liabilities under bankruptcy. In particular, the new rules require that consumers with a certain level of income pay back at least part of their outstanding debts, whereas in the past the entire amount might have been liquidated. The law also mandated financial counseling for anyone declaring bankruptcy. Continuing innovation in financial markets over the period supported further proliferation of hedge funds and other sophisticated instruments for money management.

Several demographic shifts had important consequences for the structure of the population. The aging of the baby-boom population from 2004 to 2007 drove a 12.5 percent increase in the population aged 55 to 64. Overall population growth was about 2.9 percent, and, according to figures from the U.S. Census Bureau, 37.3 percent of that growth was due to net immigration. Also according to Census Bureau estimates, the number of households increased 2.3 percent--about the same pace as in the 2001-04 period--and the average number of persons per household rose slightly, from 2.59 people in 2004 to 2.61 in 2007.

Only a small fraction of the 2007 SCF interviews took place in 2008. Thus, the survey data are largely unaffected by the declines in economic activity in 2008, the fall in the market price of corporate equities, and the continued slide in house prices. Nonetheless, readers' views of the survey results may be colored by the knowledge that, in the first three quarters of 2008, a broad measure of the value of corporate equities declined more than one-third, and house prices overall declined approximately an additional 5 percent. At a few places in the article, an attempt is made to gauge the first-order effects of these changes on families' finances.

Return to menu

Income

The change in real before-tax family income between 2004 and 2007 diverged from the pattern seen in the preceding three-year period.³ While median income declined slightly over the more recent period, the mean rose 8.5 percent ( table 1).⁴ Over the preceding three-year period, the median had increased 1.7 percent, and the mean had declined 2.3 percent. The changes for both periods stand in much stronger contrast to a pattern of substantial increases in both the median and the mean dating to the early 1990s.

Underlying the recent change was a shift in the composition of income between 2004 and 2007 (table 2). The share of family income attributable to wages and salaries fell 5.2 percentage points over the period, which approximately balanced a 3.5 percentage point rise in the share of realized capital gains and a 2.7 percentage point increase in income from self-employment, a farm, or a business. These shifts were seen across all wealth groups except the group between the 75th and 90th percentiles. As may be seen across the years shown in the table, wage income tends to be a smaller factor for the highest wealth group.

Some patterns of income distribution hold generally across the years of SCF data shown in table 1.⁵ Across age classes, median and mean incomes show a life-cycle pattern, rising to a peak in the middle age groups and then declining for groups that are older and increasingly more likely to be retired. Couples tend to have higher incomes than single persons, in part because couples have more potential wage earners. Income also shows a strong positive association with education; in particular, incomes for families headed by a person who has a college degree are substantially higher than for those with any lesser amount of schooling. Incomes of white non-Hispanic families are substantially higher than those of other families.⁶ Families headed by a self-employed worker consistently have the highest median and mean incomes of all work-status groups. Families headed by a person in a managerial or professional occupation have higher incomes than families in the three remaining occupation categories. Income is also higher for homeowners than for other families, and it is progressively higher for groups with greater net worth.⁷ Across the four regions of the country as defined by the Census Bureau, the ordering of median incomes over time has varied, but the means generally show higher values for the Northeast and the West than for the Midwest and the South. Finally, families living in metropolitan statistical areas (MSAs), which are relatively urban areas, have higher median and mean incomes than those living in rural areas.⁸

Income by Demographic Category

Across the income distribution between 2004 and 2007, only the second quintile and the top decile experienced substantial percentage changes in median income; the medians for both groups rose approximately 2 percent, though the dollar amount of the increase for the second quintile was only about $600.⁹ For other groups, changes in the median varied in direction, and in all instances they were less than 1 percent in absolute value. Similarly, the direction of changes in mean income was mixed, and the only substantial increase in dollar terms occurred for the top decile of the income distribution; the mean for that group rose almost 20 percent, more than twice the rate of change in the overall mean. Median income measured in the survey had been relatively flat for all income groups since 2001 after an earlier period of growth before 1998. Over this longer period, the rise in the mean was greatest for the top decile of the income distribution despite a dip for this group between 2001 and 2004. For the rest of the distribution, the increase of the mean more closely resembled that of the median.

Substantial proportional gains or losses in median income occurred across all age groups in the recent three-year period. The median declined for the age groups between 45 and 64 and for the 75-or-more age group, while it rose for the rest. For the 75-or-more age group, the decline was 12.3 percent. Since 1998, the age groups between 55 and 74 experienced the largest proportional rises in the median. In contrast to the recent changes in the median, the mean rose for all groups but especially for the 45-to-54 age group (8.5 percent) and the 65-to-74 age group (40.9 percent); these groups had experienced a decline in the mean between 2001 and 2004.

By family structure, median incomes declined over the 2004-07 period for all groups except childless single families (those headed by a person who was neither married nor living with a partner); median income rose the most (5.1 percent) for childless families headed by a person aged 55 or older. The largest decline (4.5 percent) was for couples (families in which the family head was either married or living with a partner) with children. In contrast, mean income rose for all types of families except childless single families headed by a person aged 55 or older, for whom it fell 2.9 percent. Mean income rose the most (20.7 percent) for single families with children.

Across education groups, median incomes rose only for families headed by a person with less than a high school diploma and for families headed by a person with only some college education (who attended college but did not receive a degree); the increase of median income was relatively strong for the former group-4.2 percent--but that group still had the lowest median income of all education groups. Mean incomes rose substantially for all education groups after declines in the preceding three-year period. The increases were particularly pronounced for the groups with families headed by a person with only some college education (10.6 percent) or by a person with a college degree (11.4 percent).

In the 2004-07 period, the median income for white non-Hispanic families fell 4.6 percent, and the mean rose 9.4 percent. In contrast, the median for nonwhite or Hispanic families rose 12.5 percent, and the mean rose 8.7 percent. However, both the median and the mean values for nonwhites or Hispanics were substantially lower than the corresponding figures for non-Hispanic whites. Since 1998, the total gain in median income for nonwhite or Hispanic families was 23.9 percent, whereas it was 6.6 percent for other families; the gain in the mean over this period was larger and more similar for the two groups--27.0 percent for nonwhite or Hispanic families and 28.5 percent for other families.¹⁰

Median income rose from 2004 to 2007 for families headed by a person who was working for someone else (a rise of 4.6 percent) or was self-employed (a rise of 3.3 percent); the median fell for the retired group (7.8 percent) and the other-not-working group (9.7 percent).¹¹ In contrast, the mean over this period rose for all groups except the other-not-working group, for which it fell 13.7 percent. Of the increases in the mean, the largest proportional change was the 23.3 percent rise for the self-employed group--the group with the highest levels of median and mean income by far. Over the previous three years, median incomes had risen only for the retired and other-not-working groups, and the mean had risen only for the retired group.

Across occupation groups, median income rose moderately for families headed by a person working in a technical, sales, or service job (an increase of 7.5 percent), and it fell strongly for families headed by a person who was not working (a decline of 9.2 percent). For the other-occupation group, a group that predominantly comprises workers in traditional blue-collar occupations, the median was barely changed. In contrast, mean income rose for all groups, particularly for families headed by a person in a managerial or professional position (an increase of 10.8 percent) and for those headed by a person in a technical, sales, or service position (an increase of 16.0 percent), the groups with the highest mean incomes in 2007. Since 1998, the only substantial changes in the median were the increases for the managerial or professional group and for the technical, sales, or service group. The means for the groups showed a general pattern of increase over the period since 1998.

By region, median family incomes in the Northeast and the West converged from different directions to about the same value in 2007, and the medians in the Midwest and the South similarly converged. The median increased between 2004 and 2007 for families living in the South and the West, and it fell for others. The 8.1 percent decline for families in the Northeast offset only about one-half of a steep increase between 2001 and 2004. The rise for the West continued the only uninterrupted trend in the median across regions for the period shown. Declines in the median income in the Midwest since 2001 erased most of the substantial gains between 1998 and 2001. In 2007, mean income was highest in the Northeast, followed by the West. In 2001, the two had been closer, but growth flattened out for the West, while it continued for the Northeast. The mean incomes in the Midwest and the South have been comparable with one another since 1998, though the mean for the South increased strongly over the recent period while the mean for the Midwest fell back slightly since 2001.

In the recent three-year period, families in MSAs saw a 0.8 percent decline in median income, while those living in other areas saw a rise of 9.8 percent. Mean income has shown a general rise for both groups since 1998.

By housing status, median and mean incomes rose both for homeowners and for other families from 2004 to 2007. All the increases were modest except the 10.0 percent increase in the mean for homeowners. As noted later in this article, homeownership declined slightly in the recent three-year period after rising for a number of years. Thus, changes in the composition of the group are likely to be smaller than in earlier years. Nonetheless, such changes were sufficient to cause the change in the median for both groups to be positive at the same time that the change in the overall median was negative.

By percentile of net worth, median income rose more than 1 percent over the recent three-year period only for the lowest quartile, for which the median increased 4.4 percent; the median declined somewhat for the third quartile and for the group between the 75th and 90th percentiles.¹² The mean increased over the period for the lowest quartile (an increase of 6.2 percent), but it rose much more strongly (23.5 percent) for the top decile. Over the earlier years shown in the table, the most dramatic cumulative gains in the median were clearly for the top two groups. The mean rose at least somewhat for all groups, but the change was largest by far for the wealthiest 10 percent.

Income Variability

For a given family, income at a particular time may not be indicative of its "usual" income. Unemployment, a bonus, a capital loss or gain, or other factors may cause income to deviate temporarily from the usual amount. Although the SCF is a cross-sectional survey, it does provide some information on income variability. In 2007, 23.7 percent of families reported that their income for the preceding year was unusual--9.2 percent reported it was unusually high, and 14.5 percent reported it was unusually low (data not shown in the tables). For those reporting unusual income, the median deviation of actual income from the usual amount was negative 17.3 percent of the normal level. A larger fraction of families in 2004 reported that their income was unusual--8.7 percent reported it was unusually high, and 19.8 percent reported it was unusually low.

Although a family's income may vary, such variability may be a well-recognized part of its financial planning. In 2007, 31.4 percent of families reported that they did not have a good idea of what their income would be for the next year, and 27.2 percent reported that they do not even usually have a good idea of their next year's income. The figures for 2004 were similar.

Saving

Because saving out of current income is an important determinant of family net worth, the SCF asks respondents whether, over the preceding year, the family's spending was less than, more than, or about equal to its income. Though only qualitative, the answers are a useful indicator of whether families are saving. Asking instead for a specific dollar amount would require much more time from respondents and would likely lower the rate of response to the survey. Overall, from 2004 to 2007, the proportion of families that reported that they had saved in the preceding year was about unchanged at 56.5 percent, a bit higher than the level in 1998 but still lower than the 2001 level. The general pattern of changes across demographic groups in the recent three-year period is one of small shifts. The previous survey had shown a broad pattern of declines.

Estimates of the personal saving rate from the national income and product accounts (NIPA) show an annual saving rate of less than 1 percent over the 2004-07 period. However, the SCF and NIPA concepts of saving differ in some important ways. First, the underlying SCF question asks only whether the family's spending has been less than, more than, or about the same as its income over the past year. Thus, families may be saving, but those that are doing so may be saving a relatively small amount; those that are spending more than their incomes may be spending a relatively large amount. Second, the NIPA measure of saving relies on definitions of income and consumption that may not be the same as those that respondents had in mind when answering the survey questions. For example, the NIPA measure of personal income includes payments employers make to their employees' defined-benefit pension plans but not the payments made from such plans to families, whereas the SCF measure includes only the latter. The SCF measure also includes realized capital gains, whereas the NIPA measure excludes capital gains of all forms, realized and unrealized.

A separate question in the survey asks about families' more typical saving habits. In 2007, 6.0 percent of families reported that their spending usually exceeds their income; 16.1 percent reported that the two are usually about the same; 35.7 percent reported that they typically save income "left over" at the end of the year, income of one family member, or unusual additional income; and 42.2 percent reported that they save regularly (data not shown in the tables). The fact that these figures are not much changed over the last three surveys suggests that variations in economic conditions over this period have had little effect on the longer-run saving plans of families.

The SCF also collects information on families' most important motivations for saving (table 3).¹³ In 2007, the most frequently reported motive was retirement related (33.9 percent of families), and the next most frequently reported was liquidity related (32.0 percent of families), a response that is generally taken to be indicative of saving for precautionary reasons.¹⁴ At least since 1998, these have been the dominant reported reasons, but saving for retirement has increased in importance. The education-related motive also appears to be important but less so in the latest survey; in 2007, 8.4 percent of families reported it as their primary motive, down 3.2 percentage points from 2004. The importance of saving for purchases rose 2.3 percentage points in 2007 after falling since before the 1998 survey in its prevalence as a reported motive for saving.

The survey asks families to estimate the amount of savings they need for emergencies and other unexpected contingencies, a measure of desired savings for precautionary purposes.¹⁵ The desired amount increases with income, but the amount is a lower percentage of usual income for higher levels of such income than for lower levels (table 3.1).

Return to menu

Net Worth

From 2004 to 2007, inflation-adjusted net worth (wealth)--the difference between families' gross assets and their liabilities--rose strongly, both in terms of the median and the mean (table 4). The median rose 17.7 percent, and the mean rose 13.0 percent; the corresponding values for the period from 2001 to 2004 were 1.0 percent and 6.0 percent. Both the median and the mean have risen consistently over the period since 1998, but overall the mean has gained more--54.7 percent, compared with a 31.8 percent increase in the median.

Movements in the dollar value of families' net worth are, by definition, a result of changes in investment, valuation, and patterns of ownership of financial assets (tables 5, 6, and 7) and nonfinancial assets (tables 8, 9, and 10), as well as decisions about acquiring or paying down debt (tables 11 through 18). A variety of financial decisions underlie these changes. The box "Shopping for Financial Services" provides a discussion of the intensity of families' decisionmaking efforts and their sources of financial information.

After the end of 2007, house prices continued to decline, and equity prices fell sharply. Although the survey cannot provide direct results about the overall effects of these and other such changes, it can provide some indication of the implications for families' finances. For this purpose, the value of assets invested directly or indirectly in publicly traded equity, the value of privately held businesses, and the net value of nonresidential real estate are assumed to have fallen at the overall rate of the Wilshire 5000 index from the time of the interview until October 2008. In addition, the value of residential properties--both primary residences and other residential real estate--are assumed to have fallen in line with LoanPerformance Home Price Indexes from the time of the interview until October 2008.¹⁶ Changes are assumed to have affected all holders proportionately, and families are assumed to have made no changes in their holdings of these assets or any other assets or liabilities. Taken together, these assumptions imply large drops in median and mean net worth since the 2007 survey--17.8 percent and 22.7 percent, respectively. Relative to the values in the 2004 SCF, adjusted median net worth is 3.2 percent lower, and the adjusted mean is 12.7 percent lower.¹⁷

By age group, median and mean net worth show a "hump" pattern that generally peaks in the 55-to-64 age group. This pattern reflects both life-cycle saving behavior and a historical pattern of long-run growth in inflation-adjusted wages. The median and mean values of wealth rise in tandem with income, a relationship reflecting both income earned from assets and a higher likelihood of saving among higher-income families. Wealth shows strong differentials across groups defined in terms of family structure, education, racial or ethnic background, work status, occupation, housing status, and the urbanicity and region of residence; these differentials generally mirror those for income, but the wealth differences are larger.

Net Worth by Demographic Category

Analysis by demographic group for the 2004-07 period shows a pattern of gains of varying sizes in median and mean net worth for most groups. But a small number of groups experienced losses, and some had noticeably different shifts in their median and mean net worth.

Median net worth rose for all percentile groups of the distribution of net worth except for families in the lowest quartile. In that group, the median fell from $1,900 to $1,200; the mean fell from negative $1,600 in 2004 to negative $2,300 in 2007. For the rest of the distribution of net worth, the median and mean over the recent three-year period rose substantially for all other groups except the 75th-to-90th percentile group, which had seen relatively large gains over the preceding three years. Gains for the top wealth group were unbroken back to at least 1998.

Over the recent period, median net worth increased for all income groups above the 20th percentile and especially for families in the fourth quintile, for which the median rose 16.6 percent; the mean for this group was little changed. Families in the lowest income quintile had the largest proportional increase in the mean--31.8 percent--a rise due, in part, to an increase in the fraction of the group consisting of relatively wealthy families with incomes that are likely to have been temporarily low (data not shown in the tables). The mean rose for the other income groups, and it rose most for the highest decile group--an 18.8 percent gain. Over the preceding years shown, median net worth had increased for all groups except the second income quintile; the mean had risen for all income groups.

The survey shows some substantial movements of net worth by age group between 2004 and 2007. Median net worth rose most strongly--19.2 percent--for the 75-or-more age group, which had seen relatively modest change over the previous three-year period. The less-than-35 age group saw a large decline in the median--24.4 percent--over the more recent period; at the same time, median wealth fell 7.1 percent for the 55-to-64 age group. Mean wealth rose just more than 10 percent for families in the 45-to-54 and 75-or-more age groups, and it increased more than 30 percent for families in the less-than-35 and 65-to-74 age groups; mean wealth declined, however, for the 35-to-44 group and was about unchanged for the 55-to-64 group. Many of the changes observed contrast in size or direction with the changes in the preceding three-year period.

By family structure, single families with children had the largest increases from 2004 to 2007 in both median and mean net worth--13.9 percent and 45.3 percent, respectively--but these families had the second-lowest level of net worth (after younger single families without children). Median net worth increased for all family-structure groups except younger single families without children, and the mean increased for all except older single families without children.

From 2004 to 2007, median net worth increased for all education groups. The change was particularly large--46.9 percent--for the no-high-school-diploma group. At the same time, this group was the only one that did not see a rise in mean net worth; its mean declined 4.7 percent. The shifts for this group were the opposite of the pattern in the preceding three-year period, during which the median fell and the mean rose.

The data show gains from 2004 to 2007 in median and mean wealth for both categories of race or ethnicity. Gains in the median and the mean were roughly the same for white non-Hispanic families--10.3 percent and 12.2 percent, respectively. But for nonwhite or Hispanic families, the change in the median--2.2 percent--was far smaller than that in the mean--35.9 percent.¹⁸ In the preceding three-year period, both the median and the mean for nonwhites or Hispanics had risen more strongly than those for other families. Despite some continuing signs of convergence, in 2007, the median and mean of net worth for white non-Hispanic families remained much higher than those for nonwhite or Hispanic families. In contrast to the whole group of nonwhite or Hispanic families, the subgroup of African American families saw a 24.1 percent decline in their median net worth from 2004 ($22,400) to 2007 ($17,000), but their mean net worth rose 9.3 percent, from $121,500 to $132,800; over the 2001-04 period, the median for the group had shown virtually no change, while the mean had risen 36.4 percent (data not shown in the tables).

Among work-status groups, median and mean net worth rose from 2004 to 2007 for all families except those headed by persons who were not working for reasons other than retirement (the other-not-working group), which showed substantial declines in both measures. The group had the lowest levels of both median and mean net worth of all work-status groups. Although the dollar amounts of the changes in median and mean net worth for the self-employed group were far larger than those for the other groups over the period from 1998 to 2007, the percentage increase in the median for the self-employed group was below the rates for all other work-status categories except the retired group. The percentage increase in the mean for the self-employed group was just slightly higher than that for the working-for-someone-else group.

Median and mean net worth increased for all occupation groups in the recent three-year period, but they did so most markedly for families headed by a worker in a technical, sales, or service occupation or by a worker in a managerial or professional occupation. Over the period since 1998, the median for families in the residual other-occupation category barely rose, and the increase in the mean was the smallest of any occupation group. All other groups had greater than a 20 percent increase in their median and mean net worth over this period.

Between 2004 and 2007, median net worth fell for families living in the Northeast or the Midwest, while it rose strongly for those in the South or the West. Mean net worth for families in the Northeast or the Midwest also lagged behind that for families in the other regions. Over the longer period from 1998 to 2007, median and mean net worth moved up most strongly in the Northeast and the West; these regions ended the period with quite similar medians and means. The Midwest and the South also ended the period with fairly similar values, at levels considerably below those for the Northeast and the West.

By urbanicity of the place of residence, in the recent three-year period, median net worth increased by about the same proportion in MSA and non-MSA areas, but the mean advanced by a much larger proportion in non-MSA areas. However, over the longer period since 1998, median and mean wealth rose more rapidly for MSAs, and in 2007 both the median and mean net worth for families in MSAs remained substantially above that for families in non-MSAs.

By housing status, the percentage increases in median net worth between 2004 and 2007 were very similar for both groups, and the increase in the mean for non-homeowners (hereafter, renters) was somewhat higher. From 1998 to 2007--a time of rising house prices, on balance--the increase in median and mean net worth for homeowners far outstripped that for renters.

Return to menu

Assets

At 97.7 percent in 2007, the overall proportion of families with any asset was barely changed from 2004 (first half of tables 9.A and 9.B, last column). Overall, this figure has risen 0.9 percentage point since 1998 (data not shown in the tables). Across demographic groups, the pattern of changes in the recent three-year period is mostly one of small increases or decreases. Noticeable exceptions are declines for the following groups: the lowest quintile of the income distribution (2.4 percentage points); single families with children (1.2 percentage points); younger single families without children (1.7 percentage points); families headed by a person whose work status was retired (1.6 percentage points) or who was in the related retired-or-other-not-working category (1.2 percentage points); families headed by a person aged 75 or older (1.5 percent); and families living in the Northeast (3.3 percentage points).¹⁹ For many groups, the figure remained at or near 100 percent.

From 2004 to 2007, median assets for families having any assets rose 16.6 percent, from $189,900 to $221,500 (second half of tables 9.A and 9.B, last column), and the mean rose 13.1 percent, from $591,300 to $668,500 (memo line). These percentage changes closely resemble those for overall net worth, but examination of changes in median assets by demographic groups reveals differences. Because changes in ownership were generally small, these differences must largely represent variations in the amount of borrowing. Across net worth groups, the percentage changes in median assets and net worth were most similar for families in the top quartile of the distribution of net worth; for all except the lowest quartile of that distribution, the changes were more roughly similar; and for the lowest quartile of the distribution, the percentage decline in assets was much larger than that for net worth. For white non-Hispanic families, median assets rose 9.9 percent, while median net worth rose 10.3 percent; but for nonwhites or Hispanics, median assets rose 36.4 percent, and median net worth rose only 2.2 percent. For homeowners, median assets increased 8.1 percent, but median net worth increased 15.6 percent; for renters, median assets barely changed, but median net worth rose 15.9 percent. Percentage changes in the medians of assets and net worth were similar across region and urbanicity of the place of residence. Over the preceding three-year period, median assets had risen 9.8 percent and mean assets had risen 8.3 percent, compared with corresponding figures for net worth of 1.0 percent and 6.0 percent.

Financial Assets

Although the level of financial assets rose from 2004 to 2007, financial assets as a share of total assets fell 1.8 percentage points, to 33.9 percent (table 5, memo line); this movement continues a decline in this share from a level in 2001 (42.2 percent) that marked the high point observed in the survey since at least 1989. The relative shares of various financial assets also shifted. Declines in the percentage shares of transaction accounts, bonds, and "other managed assets" were mostly offset by increases in the shares of retirement accounts and pooled investment funds.²⁰ After declines in the previous two surveys, the share of assets attributable to publicly traded stocks held directly by families edged up.

Overall, the rate of ownership of any financial asset was virtually unchanged over the recent survey period, at 93.9 percent (first half of tables 6.A and 6.B, last column). However, the recent data show changes for some demographic groups. By income percentile groups, ownership fell for the first and third quintiles and rose or stayed the same for other income groups; by age, an increase appeared only for the 55-to-64 age group; by family structure, ownership increased for childless couples and childless single families headed by a person older than age 55 but declined for other single families; and by work status, ownership rose substantially for families headed by a person who was self-employed or neither working nor retired. Ownership increased for nonwhite or Hispanic families and decreased for white non-Hispanic families. The share of homeowners with financial assets rose, but the ownership rate fell for renters.

 

In contrast to the drop in the overall ratio of financial assets to total assets over the recent period, the median holding of financial assets for families having such assets rose 13.8 percent (second half of tables 6.A and 6.B, last column), while the mean rose 7.0 percent (memo line). The recent change in the median did not completely offset the decrease over the previous three-year period. The more detailed picture is one of increases in the medians over the recent period for most demographic groups, including substantial increases for the lowest two income quintiles and all age groups except the 55-to-64 and 75-or-more categories. Median holdings increased most markedly for single families with children and younger childless single families; for families in the 65-to-75 age group; for families living in the South or outside of MSAs; and for nonwhite or Hispanic families. Mean holdings of those with financial assets generally rose; among the scattered declines, the largest was a 52.0 percent drop for families in the other-not-working work-status group (means by groups are not shown in the tables).

Transaction Accounts and Certificates of Deposit

In 2007, 92.1 percent of families had some type of transaction account--a category comprising checking, savings, and money market deposit accounts; money market mutual funds; and call or cash accounts at brokerages. The increase of 0.8 percentage point in ownership since 2004 resumed the upward trend seen in earlier surveys after the ownership rate had remained essentially unchanged over the previous three-year period. Families that did not have any type of transaction account in 2007 were disproportionately likely to be in the bottom income quintile group, to be headed by a person younger than 35, to be nonwhite or Hispanic, to be headed by a person who was neither working nor retired, to be renters, or to have net worth in the bottom quartile. See box "Decisions about Checking Accounts" for a discussion of the reasons families do or do not have a checking account. Over the 2004-07 period, transaction account ownership rose noticeably--by 3 to 4 percentage points--for families in South, nonwhite or Hispanic families, and families headed by a person who did not graduate from high school or who was aged 55 to 64.

The slight overall expansion in ownership of transaction accounts in the recent three-year period is reflected in the small changes in the types of transaction accounts held by families. Ownership of checking and savings accounts inched up, while ownership of money market and call accounts slightly declined (table 6.1).

The savings account category includes a relatively small number of tax-preferred accounts such as medical or health savings accounts and Coverdell or 529 education accounts.²¹ For families with a savings account, ownership of any of these types of tax-preferred accounts increased, from 2.5 percent in 2004 to 3.8 percent in 2007. In both of these survey years, 529 plans accounted for about 80 percent of the number of tax-preferred savings accounts.

Median holdings in transaction accounts for those who had such accounts fell 2.4 percent from 2004 to 2007, while the mean fell 11.4 percent. Across demographic groups, the patterns of changes in the median are mainly a mixture of substantial increases and decreases. Median balances rose for the lowest and highest income groups and the lowest net worth quartile and fell or was unchanged for the middle income groups and all the other wealth groups; across age groups, the median increased substantially for the less-than-35 and the 65-to-74 age groups and fell or rose slightly for other families. By family structure, median balances increased sharply for single families with children and rose for childless single families headed by a person aged less than 55, but they fell for other families. Across work-status groups, median balances fell for all groups except the working-for-someone-else category. Holdings increased for households headed by a person in a technical, sales, or service occupation but decreased for the remaining three occupation groups. Median balances increased strongly for nonwhite or Hispanic families and fell somewhat for other families. By region, median holdings declined substantially for families in the Northeast and Midwest.

Certificates of deposit (CDs)--interest-bearing deposits with a set term--are traditionally viewed as a low-risk saving vehicle, and they are often used by persons who desire a safe haven from the volatility of financial markets. Over the 2004-07 period, the attractiveness of CDs increased as the interest rates on them rose. The resulting increase of 3.4 percentage points in ownership was the largest increase observed in the SCF since 1989. Over the recent period, ownership increased among almost all demographic groups. Increases in ownership were particularly strong for the top income group, the oldest age group, retired families, and the next-to-highest net worth group. The overall median value of holdings of CDs increased 21.2 percent over the three-year period, while the mean value decreased 7.6 percent. Consideration of changes in the median across demographic groups reveals substantial increases for the first and third income quintiles, the some-college education group, the other-not-working group, and the other occupation group. The overall decline in the mean suggests that balances on most new accounts tended to be moderate.

Savings Bonds and Other Bonds

Savings bonds are owned disproportionately by families in the highest 40 percent of the income distribution and by families in the top half of the distribution of net worth. Over the 2004-07 period, the ownership of savings bonds declined 2.7 percentage points, to 14.9 percent overall, and it fell for virtually all demographic groups. Median holdings fell 9.1 percent, but the mean rose 4.8 percent.

Other bond types tend to be very narrowly held, and the ownership rate fell to 1.6 percent in 2007, a drop of 0.2 percentage point from 2004.²² Although the ownership rate for such bonds fell only slightly, changes in the types of bonds held by families were somewhat larger and were driven mainly by a decline in the fraction of families owning bonds of multiple types. The proportion of families that owned government bills and bonds, mortgage-backed bonds, and corporate or foreign bonds fell in the recent period, while ownership of tax-exempt bonds was unchanged (table 6.2).

Ownership of any type of bond is concentrated among the highest tiers of the income and wealth distributions, and these groups saw little change in ownership from 2004 to 2007. The median value of bonds for families that had them rose 12.0 percent, while the mean fell 4.4 percent.

Publicly Traded Stock

The direct ownership of publicly traded stocks is more widespread than the direct ownership of bonds, but, as with bonds, it is also concentrated among high-income and high-wealth families. The share of families with any such stock holdings declined 2.8 percentage points from 2004 to 2007, to 17.9 percent, thereby continuing a decline observed over the previous three-year period. Across demographic groups, the recent decline was most marked for the highest decile of the income distribution, families headed by a person who was aged 55 to 74 or who was self-employed, families in the Northeast or the Midwest, and families in the top quartile of the net worth distribution.

The major stock price indexes increased about 30 percent over the 2004-07 period; at the same time, the median amount of directly held stock for families with such assets rose 3.0 percent, and the mean climbed 25.6 percent. The median value declined for many demographic groups but rose substantially for the two family-structure groups with children and for the self-employed. The mean amount of directly held stock increased across most demographic groups (data not shown in the tables).

The great majority of families with directly held stock owned stock in only a small number of companies. Over the three-year period, the share of families owning stock in only one company increased (table 6.3).

For 36.1 percent of stockowners in 2007, at least one of the companies in which they owned stock was one that employed, or had employed, the family head or that person's spouse or partner. Ownership of stock in a foreign company was less common; only 15.8 percent of stockholders had this type of stock (data not shown in the tables). The 2004 data show a similar pattern.

Pooled Investment Funds

Pooled investment funds are among the least commonly held of the specific financial assets shown in table 6.²³ As was the case for directly held bonds and stocks from 2004 to 2007, direct ownership of pooled investment funds fell--a decline of 3.6 percentage points, to 11.4 percent of families in 2007. Ownership of pooled investment funds declined for almost every demographic group over the three-year period. Both the overall change and the changes for demographic groups continue the pattern observed in the previous three-year period.

The survey also collects information on the different types of pooled investment funds owned by families. Ownership shifted over the recent period to stock funds from most other types of funds; the residual "other" category, which consists almost entirely of hedge funds and exchange-traded funds, decreased slightly (table 6.4).

Among families owning pooled investment funds, the value of holdings has continued an increase seen over the preceding decade; in the recent three-year period, the median holding rose 26.1 percent, and the mean rose 53.3 percent. Median and mean values increased across almost every demographic group, evidence that the decrease in ownership was concentrated among families with small account balances (data not shown in the tables).

Retirement Accounts

Ownership of tax-deferred retirement assets such as personally established individual retirement accounts (IRAs) or job-based 401(k) accounts tends to increase with families' income and net worth.²⁴ For several reasons, ownership is also more likely among families headed by a person less than 65 years of age than among the older groups. First, even though retirement accounts have been in existence for more than 25 years, they may not have become common until relatively late in the careers of many persons in the older groups. Second, beginning in the year that a person reaches age 591/2, funds held by that person in retirement accounts may be withdrawn without penalty, and some in the two oldest age groups may have already done so. Third, families may have used funds from retirement accounts accumulated from previous employment to purchase an annuity at retirement; annuities are treated in the SCF as a separate type of managed asset.

From 2004 to 2007, the fraction of families with retirement accounts rose 2.9 percentage points, to 52.6 percent; the increase offset most of the 3.0 percentage point decrease over the preceding three years. In the recent period, the fraction of families that had some type of account plan associated with a current or past job or that held an IRA or Keogh account increased, and the fraction that had at least one account of each type rose as well (table 6.5).

Over the 2004-07 period, ownership increased for nearly all groups. Substantial increases were reported for families in the 45-to-54 and 65-to-74 age groups, nonwhite or Hispanic families, families living in the South, and families in the technical, sales, or services occupation group.

In a continuation of the trend over the preceding decade, holdings in retirement accounts increased markedly in the 2004-07 period; for families having retirement accounts, the median rose 16.3 percent, and the mean rose 9.5 percent. Gains also appeared in the median holdings of most demographic groups over the recent period; some of the largest increases were for families in the middle of the income and wealth distributions, families in the high-school-diploma and some-college education groups, single families with children, nonwhite or Hispanic families, the self-employed work-status group, families in the South and West, and families residing in non-MSA areas.

Although tax-deferred retirement assets are clearly an important element in retirement planning, families may hold a variety of other assets that are intended, at least in part, to finance retirement. Such other assets might also be used for contingencies as necessary. Similarly, a need for liquidity might drive a family to liquidate or borrow against a tax-deferred retirement asset, even if it will be assessed a penalty for doing so.

Two common and often particularly important types of retirement plans are not included in the assets described in this section: Social Security (the federally funded Old-Age and Survivors' Insurance program, or OASI) and employer-sponsored defined-benefit plans. OASI is well described elsewhere, and it covers the great majority of the population.²⁵ The retirement income provided by defined-benefit plans is typically based on workers' salaries and years of work with an employer, a group of employers, or a union. Unfortunately, future income streams from OASI and defined-benefit plans cannot be translated directly into a current value because valuation depends critically on assumptions about future events and conditions--work decisions, earnings, inflation rates, discount rates, mortality, and so on--and no widely agreed-upon standards exist for making these assumptions.²⁶

However, the SCF does contain substantial information for family heads and their spouse or partner regarding any defined-benefit plans or other types of plans with some kind of account feature to which they have rights from a current or past job.²⁷ In 2007, 57.7 percent of families had rights to some type of plan other than OASI through the current or past work of either the family head or that person's spouse or partner, a level nearly the same as in 2004. For this group of families, the fraction with a standard defined-benefit plan with an annuity payout scheme declined over the recent period, while the fraction with a plan with at least some account feature and the fraction that had both types of plans increased (table 6.6).

In many pension plans with account features, contributions may be made by the employer, the worker, or both. In some cases, these contributions represent a substantial amount of saving, though workers may offset this saving by reducing their saving in other forms. An employer's contributions also represent additional income for the worker. In 2007, 87.1 percent of families with an account plan on a current job of either the family head or that person's spouse or partner had an employer that made contributions to the plan, a decline of 1.6 percentage points from 2004. In 2007, 91.4 percent of families with such plans made contributions themselves, an increase of 2.1 percentage points from 2004. The median annual contribution by employers who contributed to such accounts was $2,200 in 2007, and the median contribution by families that contributed was $2,500; both amounts fell slightly from 2004 levels (data not shown in the tables).

The eligibility of working heads of families to participate in any type of job-related pension rose from 54.8 percent in 2004 to 55.9 percent in 2007; it had declined 2.4 percentage points over the preceding three years (data not shown in the tables). Participation by eligible workers is usually voluntary. In 2007, 83.8 percent of family heads who were eligible to participate elected to do so, down slightly from 84.1 percent in 2004.²⁸ The choice to participate appears to be related strongly to income. In 2007, the fraction of eligible family heads declining to participate fell as income rose, and this general pattern was not substantially altered from 2004 (table 6.7).

Cash Value Life Insurance

Cash value life insurance combines an investment vehicle with insurance coverage in the form of a death benefit.²⁹ Some cash value life insurance policies offer a high degree of choice in the way the policy payments are invested. Investment returns on such policies are typically shielded from taxation until the money is withdrawn; if the funds remain untapped until the policyholder dies, the beneficiary of the policy may receive, tax-free, the death benefit or the cash value, whichever is greater. In contrast, term insurance, the other popular type of life insurance, offers only a death benefit. One attraction of cash value policies for some people is that they promote regular saving funded through the required policy premium.

Ownership of cash value life insurance is broadly spread across demographic groups, with a tendency toward increasing rates among families with higher levels of income and net worth and those with older family heads. Ownership of cash value policies over the 2004-07 period continued a declining trend, decreasing 1.2 percentage points, to 23.0 percent of families in 2007. The decline was shared by most demographic groups. Over the three-year period, ownership of any type of life insurance, cash value or term, also fell slightly--from 65.4 percent in 2004 to 64.9 percent in 2007 (data not shown in the tables). Of those families with some type of life insurance, the proportion with term policies was about unchanged, while the proportion with cash value policies fell; these changes are similar to trends in the earlier surveys.

After declining over the previous three-year period, the median value of cash value life insurance for families that had any such insurance rose 21.2 percent between 2004 and 2007, and the mean rose 23.7 percent. The median showed increases across most demographic groups, although it declined considerably for families in the other-not-working work-status category, renter families, and families in the second quartile of the wealth distribution.

Other Managed Assets

Ownership of other managed assets--personal annuities and trusts with an equity interest and managed investment accounts--is concentrated among families with higher levels of income and wealth and among families headed by a person who is aged 55 or older or who is retired.³⁰ Ownership of these assets declined 1.5 percentage points between 2004 and 2007 after a small increase over the previous three years. Ownership fell in the recent three-year period for almost every demographic group, with the largest declines for families in the Midwest and for the next-to-highest income and net worth groups. Across all families, the fraction with an annuity declined over the period, and the fraction with a trust or managed investment account inched up, while the fraction with both categories of managed assets was essentially unchanged (table 6.8).

Between 2004 and 2007, the median value of other managed assets for families that had such assets increased 41.7 percent, an increase that offset the decline in the preceding three-year period. Over the more recent period, the corresponding mean value increased 9.4 percent. Median holdings rose for many demographic groups; noticeable exceptions were families in the top two income deciles and families headed by a person who was working for someone else or who was working in a technical, sales, or service job or a job in the other-occupation category. The rise in the median value reflects substantial increases in annuities and modest increases in trusts or managed investment accounts. For families with an equity interest in an annuity, the median holding rose 23.1 percent, to $50,000 in 2007; for families with a trust or managed investment account as defined in this article, the median holding rose 9.1 percent, to $120,000 (data not shown in the tables).

As noted in the discussion of retirement accounts, some families use settlements from retirement accounts to purchase an annuity. In 2007, 30.4 percent of families with annuities had done so (data not shown in the tables). Of these families, 71.7 percent had an equity interest in their annuities.

Other Financial Assets

Ownership of other financial assets--a heterogeneous category including oil and gas leases, futures contracts, royalties, proceeds from lawsuits or estates in settlement, and loans made to others--fell 0.7 percentage point between 2004 and 2007, to 9.3 percent. Ownership of such assets tends to be more common among higher income and wealth groups, younger age groups, and families headed by a person who is self-employed. Ownership across demographic groups generally declined over this period, while the median holding for those who had such assets increased 36.4 percent, to $6,000.

Holdings may be grouped into four categories: cash, which includes money owed to families by other persons; future proceeds, which include amounts to be received from a lawsuit, estate, or other type of settlement; business items, which include deferred compensation, royalties, futures contracts, and derivatives; and other. The proportion of families holding various types of other financial assets remained fairly constant over the three-year period, with cash being by far the most frequently held component (table 6.9).

Some publicly traded companies offer stock options to their employees as a form of compensation.³¹ Although stock options, when executed, may represent an appreciable part of a family's net worth, the survey does not specifically ask for the value of these options.³² Instead, the survey asks whether the family head or that person's spouse or partner had been given stock options by an employer during the preceding year. In 2007, 8.3 percent of families reported having received stock options, a decline of 1.0 percentage point below the level in 2004; this decrease continues a downward trend since the peak of 11.4 percent recorded in the SCF in 2001 (data not shown in the tables).³³

Direct and Indirect Holdings of Publicly Traded Stocks

Families may hold stocks in publicly traded companies directly or indirectly, and information about each of these forms of ownership is collected separately in the SCF. When direct and indirect forms are combined, the 2007 data show a resumption of a trend of increasing stock ownership (table 7). Between 2004 and 2007, the fraction of families holding any such stock rose 0.9 percentage point, to 51.1 percent, a level still below the 2001 peak of 52.2 percent. Much like ownership of directly held stock, ownership of direct and indirect equity holdings is more common among higher-income groups and among families headed by a person aged 35 to 64. Over the recent three-year period, ownership increased for all income groups except the third quintile and top decile. Across age groups, ownership fell for families headed by a person younger than 45 or aged 55 to 64; ownership rose substantially for families headed by a person aged 45 to 54 or older than 65.

At the same time, the overall median value of direct and indirect stock holdings dropped 2.0 percent. Changes in the median value across demographic groups were mixed, with declines more common for groups that experienced increases in ownership, an indication that most new owners had small amounts. As a proportion of financial assets, holdings rose 2.0 percentage points overall, with substantial increases for the first and fourth income quintiles and the oldest age group.

As noted earlier in the discussion on net worth, the stock markets have undergone sizable declines since the data collection for the 2007 SCF was completed. To gauge the potential effect of these changes on the median amount of equity held by families, the equity values in the survey were deflated by the ratio of the average of the Wilshire 5000 index in October 2008 to the value of the index on the day of the interview, assuming a homogeneous rate of return for all equity holders and no changes in the portfolios of families since the time of the survey. Under this scenario, the median value of equity falls 35.7 percent, from the 2007 value of $35,000 to $22,500 (data not shown in the tables).

Among families that held equity, either directly or indirectly, in 2007, ownership through a tax-deferred retirement account was most common, followed by direct holdings of stocks, direct holdings of pooled investment funds, and managed investment accounts or an equity interest in a trust or annuity. Over the 2004-07 period, ownership of tax-deferred accounts rose, while ownership of all other types of equities fell; the fraction of equity owners with multiple types also declined (table 7.1).

The distribution of amounts of holdings over these types of equities shows a different pattern. Of the total amount of equity, 37.8 percent was held in tax-deferred retirement accounts, 33.6 percent as directly held stocks, 22.1 percent as directly held pooled investment funds, and 6.5 percent as other managed assets (data not shown in the tables).

Nonfinancial Assets

By definition, a rise in nonfinancial assets as a share of total assets must exactly offset the 1.8 percentage point drop in the share of financial assets from 2004 to 2007, which was discussed earlier in this article (table 5). The changes in these shares may have been driven by changes in portfolio choices, portfolio valuation, or both. The 2001 estimate of the value of nonfinancial assets as a share of total assets, at 57.8 percent, appears to be the low point since 1998 (table 8); the 2007 level of 66.1 percent is near the middle of the range over the past seven surveys (data not shown in the tables). Over the recent three-year period, the value of primary residences as a share of nonfinancial assets fell 2.2 percentage points, to 48.1 percent, still above its share before 2004. The share of equity in nonresidential property also declined. The largest offsetting increase was in the share of business equity, which rose 3.8 percentage points over the period to its highest recorded share of 29.7 percent in 2007.

In 2007, the level of ownership of nonfinancial assets was 92.0 percent of families, 0.5 percentage point lower than in 2004 (first half of tables table 9.A and table 9.B, next-to-last column). Across most of the demographic groups shown, the 2007 rate was 85 percent or more; exceptions were the lowest income and wealth groups, younger childless single families, families headed by a person who was neither working nor retired, renters, families headed by a person without a high school diploma, and families living in the Northeast. Over the 2004-07 period, ownership rose most for the 55-to-64 age group, families with children, nonwhite or Hispanic families, and families living in the South. Substantial declines in ownership were seen by the oldest age group, the lowest quintile of the income distribution, families without children, families headed by a retiree, and families living in the Northeast.