Basis Report for Locked Out: Keys to Homeownership Elude Many Working Families with Children, Center for Housing Policy, March 2006

The views expressed in this paper are those of the author and do not necessarily reflect the views of the Center for Housing Policy, its Board or its sponsors.

HOMEOWNERSHIP AND CRITICAL HOUSING NEEDS AMONG WORKING FAMILIES WITH CHILDREN: 1978–2003 TRENDS AND THEIR POLICY IMPLICATIONS

Kathryn P. Nelson, Ph.D.

Summary
As U.S. homeownership rates soared to record highs after 2000, ownership among low- and moderate-income families with children continued to lag, failing even by 2003 to regain highs recorded in the late 1970s. Analysis of 1978–2003 trends among working families with children reveals that demographic shifts toward more single parents and minority households explain much of the lag. Over the 25-year period ownership rates rose significantly, albeit modestly, for non-Hispanic white working families with children, both couples and single parents, and also for minority working couples with children. Yet the incidence of critical housing needs and severe cost burdens also rose, particularly among minority renters with children. By 2003 working minority renters with children were appreciably more likely to experience cost burdens or critical housing problems than working owners with children.

Over the same period, both elderly and nonelderly households without children registered much more rapid gains in homeownership than families with children. Childless nonelderly households moved into ownership more quickly than families with children in both upper-income and working income groups, and in both high- and low-cost markets. These results imply that current policies to increase homeownership among low- and moderate-income families cannot be assumed to benefit either families with children or the low-income minorities who most need the benefits of homeownership.

Between 2001 and 2003, homeownership reached record highs in the United States. According to the American Housing Survey, by 2003 homeownership was at an all-time high of 68.3 percent, up from 65.2 percent in 1978. However, as earlier observed for 1999 (Nelson 2002), the ownership rate of 68.6% among all families with children in 2003 remained significantly below its 1978 high of 70.5%.

This report focuses on ownership and critical housing problems among working families with children—those earning no more than 120 % of the local median income but more than the full-time equivalent of the minimum wage.1 Between 2001 and 2003, ownership did rise for working families with children, to 59.6%. Despite this improvement, homeownership among working families with children remained lower in 2003 than its 1978 high of 62.5%.

Since 1988, both Republican and Democratic administrations have explicitly aimed to increase homeownership, often citing its demonstrated benefits for children (Retsinas and Belsky 2002; Haurin, Parcel and Haurin 2002). Yet this report demonstrates that the much heralded gains in ownership that have occurred over the past decade have been much higher among the elderly and childless nonelderly households than among families with children. This report first examines the characteristics and housing problems of working families with children, including the homeownership progress of minority working families with children. Two basic demographic shifts—relative increases in single parents and in minority households—are shown to underlie the national drop in ownership for this growing group, which by 2003 housed more than half of America’s children. Homeownership also became more expensive, causing rises in both average cost burdens and the shares of families experiencing severe cost burdens.

When the experiences of working families with children, both owners and renters, are compared with those of other households over the 1978-2003 period, it is clear that homeownership among working families with children also lagged in relation to most other American households. Increases in ownership have been much higher in almost all other groups, including the elderly. In contrast to the average drop among working families with children, it is particularly relevant that homeownership surged by 11 percentage points, or 27%, among childless nonelderly households in the same “working” income range.

In addition to comparing the housing experiences of working families with children with those of other household types, this study reviews the changing characteristics of working renters with children. The report concludes by exploring the implications of its findings for current programs and policies aimed at increasing homeownership among low- and moderate-income renters.

The experience of working families with children

Their growing importance. Between 2001 and 2003, working families with children grew both absolutely and relatively, continuing earlier trends (Table 1). By 2003, the 19.8 million working families with children comprised 18.7% of U.S. households. They included over half of families with children (52%) and housed over 54% of U.S. children. Over the 25 years between 1978 and 2003, the share of American children living in working families with children rose by 13 percentage points, whereas the shares of children living in families with marginal earnings or with upper incomes (above 120% of local median income) both declined.

Because more children lived in working families with children, the long-term decline in ownership rates among these families caused an even sharper drop in the share of children living in owner-occupied homes. As the table shows, among working families with children the share of children living in owned homes fell from 66% in 1978 to 60% in 2003, a percentage point drop twice as great as the net decline of 2.9 points in the ownership rate. Similarly, the share of all children living in owned homes fell by 4 percentage points, from 72 to 68%, double the 1.9 percentage point drop in ownership among families with children.

Shifts in family composition. The overall decline in ownership among working families with children over the past 25 years could statistically be attributed to a continuing shift toward families with only a single parent, families that have historically had lower ownership rates. But the minority share of working families with children also increased significantly between 1978 and 2003.2 As Table 2 details, the relative increase in minority heads (of 16 percentage points) was only slightly less than the 18-percentage point shift toward single parents. Whereas only 18% of working families with children had a single parent in 1978, in 2003 over a third (36%) of these families had a single parent. Similarly, while only 26% of working families with children had minority heads, by 2003 the minority share had risen to 42%. Over this quarter century, the share of working families with children headed by non-Hispanic white couples dropped by 24 percentage points as each of the other three groups gained. Both demographic shifts in household composition underlie the lag in homeownership rates.

Distinguishing working families with children by both family type and minority status reveals that ownership rates did rise between 1978 and 2003 for all 4 groups—both majority and minority couples and single parents. For each group except minority single parents with children, the rise in ownership was significant, with the increase greatest for white single parents (7 percentage points). The homeownership rate of 77.5% among white working couples with children was well above the national average. Among minority single parents, by contrast, barely one-third (32.8%) were owners and the increase over 25 years was minimal, and insignificant.

Shifts in household composition among working families with children, combined with lesser homeownership increase among minorities, changed the demographic composition of working renters with children appreciably. By 2003, whites comprised only two-fifths of this group, and only half of such renter families were headed by couples. Almost one-third (29%) of working renters with children were minority single parents.

Changes in homeownership among non-Hispanic whites and other minorities. When the race and ethnicity of working families with children are examined in more detail (Table 3B), homeownership was still slightly lower in 2003 than in 1978 for each minority group other than the relatively small group of Asians, American Indians and Pacific Islanders.3 Among this group, ownership rose by a statistically significant 10 percentage points between 1978 and 2003.4

As Table 3A’s second panel illustrates, long-term drops in ownership among working families with children occurred only among large families. Among white working families with 3 or more children, ownership fell from 79% in 1978 to 73% in 2003, with all of the decline occurring before 1991. Among minorities, ownership among large families dropped by 5 percentage points, from 53% to 48%. Although smaller white and minority working families consistently had lower ownership rates throughout the study period, their ownership dropped less during the 1980s, and rose above the 1978 levels by 2001.

Table 3B's fourth panel underscores one difficulty in achieving ownership that faces single parents. Between 1978 and 2003, ownership rose only among families with two earners, both white and minority. By contrast, ownership among families with only one earner fell rapidly between 1978 and 1991, and by 2003 ownership rates for these groups were still slightly below their 1978 values. Both the quarter century drop in ownership and the 1978–1991 drops were greatest among minority families with 3 or more earners.

Ownership by location. By 2003, ownership rates among white working families with children were higher than they had been in 1978 in each Census region. Among minorities, homeownership rose in the Northeast, but dropped by 7 percentage points in the Midwest, and also declined in the South. Surprisingly, ownership rose slightly among both whites and minorities in the West, despite its more expensive housing. There, both groups of working families registered especially large gains in ownership between 2001 and 2003.

The homeownership trends by metropolitan location in the last panels of Table 3A and Table 3B suggest that ownership rose most among whites outside of metropolitan areas, but fell most for minorities there. 5 In terms of concerns about the access of minorities to expanding employment opportunities in suburban counties, it is reassuring that between 1991 and 2003 ownership rates in suburbs rose almost as quickly for minority working families with children as for whites (11 vs. 12 percentage points).

Housing costs among working families with children. Over the 1978–2003 period ownership costs rose more quickly than income for both whites and minority working owner families with children (Table 4B).6 For both groups current-dollar income essentially doubled, but the increases in housing costs were appreciably higher for both minorities (264%) and non-Hispanic whites (229%). For both groups, payments for mortgage principal and interest rose the most. Among minorities, insurance payments and property taxes also rose more quickly than did median income. In relation to median income, mortgage payments rose by 5 percentage points for white working families with children, from 15% in 1978 to 20% in 2003. The increase was even greater for minorities, with the mortgage cost-income ratio rising by 7 percentage points from 15% in 1978 to 22% in 2003. Overall, housing outlays for white working owners with children rose from 21% of household income in 1978 to 24% in 2003. Among minorities, housing outlays in relation to income rose by 4 percentage points, from 22% in 1978 to 26% in 2003.

Reversing the long-term trend, between 2001 and 2003 housing cost burdens eased for working owners with children. For both white and minority working family owners, median incomes rose more than housing cash outlays over these two years. Ratios of median housing outlays to median income fell slightly, from 25% to 24% for whites, and from 27% to 26% for minorities.

In both the long and short runs, however, housing cost burdens rose more for working renters with children than for owners (Tables 4B and 4C). Between 1978 and 2003 housing outlays rose less for renters than for owners, but both white and minority renters had even slower growth in income. As housing costs outpaced incomes for white working renters with children, their median housing cost burden rose by 3 percentage points over the 25 years, to 27%. Minority working renters with children experienced both greater increases in housing costs and slower income growth than their white counterparts, so their housing cost burden soared by 8 percentage points, from 22% to 29%.

Nor did working renters with children benefit recently. Between 2001 and 2003, cost burdens rose an additional percentage point among whites, and remained at 29% for minorities.

Housing cost burdens and critical needs among owner and renter working families with children. Paralleling trends in median income and housing costs, the shares of working family owners with children with moderate or severe housing cost burdens rose appreciably between 1978 and 2001, but eased slightly thereafter (Tables 5A and 5B). Among white working owners with children, the share with moderate cost burdens (30–50% of income) rose by 8 percentage points between 1978 and 2001, but then fell by 4 percentage points between 2001 and 2003. The incidence of severe cost burdens (paying more than half of income for housing) more than tripled between 1978 and 2001, rising by 7 percentage points, before easing between 2001 and 2003. Over the entire 25-year period, critical needs more than doubled, rising from 4% in 1978 to 10% in 2003 among these owners.

Minority working owners with children experienced greater increases in cost burdens than did whites. The incidence of severe cost burdens rose by 7 percentage points to 11%, and moderate cost burdens increased by 13 percentage points, to 29%. By 2003 minority working owners with children were more likely than whites to have either moderate or severe housing cost burdens. They also more often had critical housing needs (12%).

White renters with children also had rising cost burdens during the study period, but as with owners the share with cost burdens eased slightly between 2001 and 2003. By 2003, the incidence of critical needs among working white renters was 10%, the same as that for working white owners with children. Among both groups, one-third had moderate or severe cost burdens.

Among working minorities with children who were renters, by contrast, all three indicators of housing problems reached record highs in 2003. As the second panel of Table 5B shows, the incidence of severe cost burdens among these renters rose surged from 1% in 1978 to 13% in 2003. Critical needs for this group more than doubled, from 8% in 1978 to 16% in 2003. Because the incidence of moderate cost burdens also rose steadily over the 25 years, by 2003 almost half (47%) of working minority renters had a severe or moderate housing cost burden, up from 20% in 1978. By 2003 working minority renters with children were appreciably more likely to experience cost burdens or critical housing problems than working owners with children.

Working families with children compared to other households

In contrast to the continued long-run ownership drop among working families with children, for almost all other households ownership rates recovered from their declines during the 1980s to record new highs by 2003. And ownership also rose markedly—by 3.7%—for upper-income families with children, reaching 90.8% in 2003. The lackluster changes in homeownership among non-Hispanic white and minority working families with children are particularly striking when compared to changes in ownership and household growth among other households, particularly childless households of similar income.

Ownership and household growth. Among all households without children, ownership rates rose by 6.3 percentage points, or 10%, from 61.8% in 1978 to 68.1% in 2003. As Table 6B details, among both white and minority childless households, ownership rates rose by 14%, much faster than those of families with children. Elderly households experienced the greatest long-run gains, with ownership up by 10 percentage points, or 14%, among whites, and by 9 percentage points, or 17%, among minorities. Yet such changes in ownership among the elderly are likely to reflect aging in place than first-time homeownership. According to the bottom panel of Table 6B, households with heads 65 or older comprised almost a third of the white childless households, but only one-fifth of minority households without children.

Among the nonelderly without children, ownership increased by 6 to 9 percentage points, or 14 to 17%, between 1978 and 2003. Among the subset of “working” nonelderly households without children (that is, nonelderly households with earnings and income in the same range as working families with children), ownership rates for non-Hispanic whites soared by 15 percentage points, or 35%, between 1978 and 2003. This increase was more than 7 times the 2 percentage point gain of white working families with children. But ownership rates for working minorities without children rose by only 4 percentage points, the same rise as that experienced by working minorities with children.

Between 1978 and 2003, the number of working nonelderly households without children more than doubled, rising from 9 to 20 million. As the bottom panel of Table 6B details, the increase was greatest (195%) among working nonelderly minorities without children, a growth rate even faster than the 168% of working minority families with children. Among both whites and minorities, numbers of elderly households rose less quickly. By 2003, there were as many working nonelderly households without children as the 20 million working families with children. The working nonelderly without children are thus effectively the largest group, and the most quickly growing one, to compete with working families with children for ownership opportunities.

Ownership trends compared in different types of housing markets. Rates of homeownership among low-income families with children have been found to be lower in high-cost markets (Nelson 2002; CHP 2004).7 To compare the experience of working families with children to that of the working childless nonelderly in comparably expensive markets, Table 7A and Table 7B examine ownership in markets differentiated by cost burdens for households in two income ranges: the “working” income range and “upper” incomes above 120% of area median. The results show that among non-Hispanic whites, gains in homeownershop among the working nonelderly were consistently greater than those of working families with children, with the disparity between the two groups ranging from 8 to 19 percentage points, much greater than the 4 percentage points shown by national averages. Among minorities, however, increases in homeownership were generally low among all household types in all markets.

Non-Hispanic whites. In each type of market white ownership rates rose appreciably more among the working childless nonelderly from 1978–2003 than among working families with children. The difference was greatest in markets with medium cost burdens: there ownership rose by 19 percentage points for the working nonelderly without children but not at all for working families with children. The difference in ownership gains between the two household types was also marked in the more costly markets: 12 percentage points. In the least costly markets the increase of the working nonelderly was 8 points higher than that of working families with children.

Because ownership increases were higher in each market type for both couples and single parents than for all white working families with children, some of the difference between working families with children and working childless households is due to the compositional shift toward single parents discussed earlier. But substantial differences remain, especially in the two more costly market types.

The gains made by the working white childless nonelderly are particularly notable when compared to the increases in ownership experienced by upper-income white childless nonelderly households in these markets. In each market type, the 25-year percentage point increases of the working white childless nonelderly were similar to or higher than those of upper-income childless nonelderly. And gains among both nonelderly childless white groups also consistently exceeded those of upper-income families with children. These differentials imply strongly that among non-Hispanic whites the working nonelderly benefited much more from efforts to raise homeownership over the past two decades than did families with children in the same income ranges.

The period-to-period changes for these four household-income groups reveal two other intriguing patterns. When rates in 1991 are compared to those in 1978, white working families with children were the only one of these four groups to experience a sharp drop in ownership during the 1980s; and that drop occurred mainly in the most expensive markets, for both couples and single parents. Similarly, recent changes between 2001 and 2003 suggest that working families with children are still disadvantaged vis-à-vis working childless nonelderly households in the more costly markets. There the 8 percentage point increase of the childless is more than twice those of either couples or single parents with children. In the other markets, recent homeownership gains among the working nonelderly without children were similar to those of working families with children.

Finally, the greater increases in homeownership by the white working nonelderly in each third of markets imply that the actual experience of working childless nonelderly exceeded that of working families with children more than national averages reveal. The excess of the 11 to 19 percentage point increases for working childless nonelderly in the three market types over the 7 percentage point average increase for all working childless nonelderly implies that these households were shifting into markets with higher cost burdens over the past 25 years. As they did, their demand for owned housing may well have bid up housing prices in those markets.

Minorities. Among minorities, there were no significant differences between working families with children and the working childless nonelderly in any of the three market groups.8 As expected, ownership gains were generally greater among upper-income minorities than among those with incomes in the “working” range. The most puzzling result in the table is the fact that ownership among working minorities rose most in markets with the highest cost burdens and fell most in the markets with the lowest cost burdens.

To sum up, the continuing evidence of faster transitions into ownership for working white childless nonelderly households raises questions about the design and effects of current programs intended to raise homeownership. The growing minority share of working renters combined with the evidence that homeownership has risen slowly, if at all, among minority working households with and without children is also disturbing in light of explicit efforts to raise minority homeownership.

Should ownership programs better aid low and moderate income working families with children?

The continuing lag in homeownership among non-Hispanic white working families with children compared to other nonelderly households in the same income range raises basic questions about the targeting of current ownership programs. Although few ownership programs are specifically directed at families with children, justifications for encouraging ownership repeatedly cite the advantages of ownership for children. Trends in ownership rates and critical needs suggest strongly that low-income renters with children, particularly minority single parents, have more pressing needs for homeownership than other working households.

In examining these trends, the experiences of households with “low” incomes (below 80% of area median income) are distinguished from those of households with “moderate” incomes (between 80% and 120%) because the main programs supporting ownership are aimed at one or the other of these income groups. Our focus on families with children makes this distinction particularly relevant. The largest federal programs assisting first-time ownership among households with incomes below 80% of AMI—HOME, CDBG, and Section 8—all adjust for household size in determining income eligibility. They thereby, to some extent, account for the fact that families with children, like other larger families, need larger incomes to cover basic expenditures for family members.

By contrast, the main federal and state programs facilitating ownership among households with incomes above 80% of AMI are implicitly biased against children because their income limits have little or no adjustment for household size. Households eligible for state ownership programs financed by mortgage revenue bonds must have incomes below the area median income if they have one or two persons, or incomes below 115% of the AMI for 3 persons or more. To meet the “low” income housing goals of the government-sponsored housing enterprises, households must have incomes below the area median family income, with no adjustment for household size.

Critical needs and ownership rates among working nonelderly households with and without children. When low-income working households are distinguished from those with moderate incomes (Table 8B), among both non-Hispanic whites and minorities it is households with low income that are more likely to have critical housing problems and less likely to be owners. Except for moderate-income white single parents, within each income range couples with children in 2003 have ownership rates some 20 percentage points above those of either single parents or households without children. Between 1978 and 2003, ownership rose most among working nonelderly childless whites in both income ranges, and it also rose by 16 percentage points for moderate-income single white parents with children. By contrast, long-term gains in homeownership were marginal for low-income minorities. Ownership among moderate-income minority couples and childless households rose appreciably, particularly between 2001 and 2003

The differentials in the table imply that low-income single parent renters with children have the most pressing needs for homeownership assistance, particularly among minorities:

• Within both race-ethnicity groups, low-income single-parent renters with children were most likely to have critical needs in 2003. This was especially true among minority renters (19%), who were also more likely to have severe problems than low-income owners.
  
  
• Critical needs grew faster for low-income single-parent renters with children between 1978-2003 (up 9 to 10 percentage points) than for any other group, although the rise among low-income minority renter couples with children was almost as great (8 percentage points). By contrast, critical needs fell among both majority and minority moderate-income renter couples and single parents with children, and also among minority renters without children.
  
  
• Minority low-income single-parent renters with children had lower ownership rates in 2003 (31%) than all other low-income groups, and essentially no increase in ownership rates between 1978 and 2003.
  

Policy implications. These results confirm and update earlier evidence that low-income families with children have worse housing problems and more need for assistance in moving into homeownership than their childless counterparts (Nelson and Khadduri 1992). As I have discussed in more detail elsewhere (Nelson 2002, 455ff), they suggest that HOME is the current program best targeted to aid low-income families with children become homeowners. HOME has the additional advantage of being usable for rental assistance, which may be more cost-effective than ownership in tight and expensive local housing markets.

Last November, a presidential tax advisory panel called for replacing the deduction for mortgage interest with a credit equal to 15 percent of mortgage interest paid. Such a credit would reduce homeownership costs for more low- and moderate-income families than does the current deduction. But as Capone (1995, 346) has shown, the current “differential in tax-bracket levels and standard deductions gives a homeowner tax benefit only to single individuals—rather than families—who buy moderately priced homes.” The results of this paper not only point to the desirability of replacing the mortgage interest deduction with a mortgage interest credit, but they further suggest that such a credit should be more explicitly targeted toward low-income families with children than is now the case.

Summary

In 2003, homeownership among working families with children was still lower than it had been 25 years earlier, even as homeownership rates rose to record highs among most other households. This report first shows that two basic demographic shifts—relative increases in single parents and in minority households—underlie the lag, and that ownership did rise among non-Hispanic white working couples and single parents with children and among minority working couples with children.

A wider comparison, however, reveals that ownership has increased much more quickly among non-Hispanic white working nonelderly childless households than among families with children. Moreover, ownership has increased least, and critical needs have risen most, among low-income minorities with children. These results raise questions about the direction and efficacy of efforts to increase homeownership among minorities over the past decade.

Notes:
1 Research by the Center for Housing Policy (2004) found that in 2001, ownership was almost 6 percentage points lower for this important group than it had been at its peak before 1980.
2 Procedures for identifying minorities on the American Housing Survey were changed in 2003, since respondents were allowed to identify as many categories as they wished. Despite the changed procedures, the number of working families with children who were classified as minorities was 13 million in both 2001 and 2003.
3 This group rose from 2% of working families with children in 1978 to 6% in 2003.
4 Among Hispanics and non-Hispanic blacks, ownership dropped by a insignificant 1 percentage point between 1978 and 2003.
5 These data on homeownership trends by metropolitan location must be interpreted with care, because the MSA definitions change over time as new counties were reclassified as metropolitan. The 1978 data are based on SMSAs as defined from 1970 decennial Census results; the 1991 data on MSAs as defined from 1980 decennial Census results, and the 2001 and 2003 data on MSAs as defined from 2000 decennial Census results. Although the geography included in these definitions is not constant over time, it does approximate the areas contemporaneously considered metropolitan.
6 Studies generally show rising cost burdens among owners over this time period (cf Nelson 2002).
7 To investigate relationships between homeownership and local housing affordability, observations were defined as being in high, medium, or low cost-burden geographic areas based on the median cash housing costs of owners in that area relative to the median household income of owners in 2001.
8 This is not surprising, because the 4 percentage point difference between them at the national level (Table 6B) was barely significant at the 90% confidence level.

---

References

Capone, Charles A., Jr. 1995. Taxation and Housing Tenure Choice: The Case for Moderate-Income Homeownership. Journal of Housing Economics 4:328-349.

Center for Housing Policy. 2004. Working Families with Children: A Closer Look at Homeownership Trends. Washington DC: Center for Housing Policy.

Haurin, Donald R., Toby L. Parcel, and R. Jean Haurin. 2002. Impact of Homeownership on Child Outcomes. Pp 427-446 in Retsinas and Belsky, 2002.

Nelson, Kathryn P. 2002. Housing Needs and Effective Policies in High-Tech Metropolitan Economies. Housing Policy Debate 13(2): 417-68.

Nelson, Kathryn P., and Jill Khadduri. 1992. To Whom Should Limited Housing Resources Be Directed? Housing Policy Debate 3(1): 1-55.

Retsinas, Nicolas P., and Eric Belsky (editors). 2002. Low-Income Homeownership: Examining the Unexamined Goal. Cambridge MA, Joint Center for Housing Studies.