The 1990s saw increased lending to low-income families and young families seeking their first homes. Lenders introduced special programs aimed at first-time homebuyers and revamped their underwriting standards to work with the special circumstances of low-income families. Private mortgage insurers, the Federal Housing Administration (FHA), Fannie Mae, and Freddie Mac also played important roles in what some observers called “a revolution in affordable lending.”

Home Mortgage Disclosure Act (HMDA) data have documented the effects of new affordable lending efforts; for example, HMDA data show that the number of home-purchase loans originated during the 1990s for low-income borrowers and their neighborhoods grew at substantially higher rates than did the overall market. HMDA, however, does not provide separate information on first-time homebuyers—a segment of the housing market that certainly has benefited from the recent increases in affordable lending, but about which little is known. Insights on first-time homebuyers can be learned from the American Housing Survey (AHS), a national survey of the housing market funded by the U.S. Department of Housing and Urban Development and conducted by the U.S. Census Bureau. This issue of U.S. Housing Market Conditions will use AHS data to present a picture of the first-time homebuyer market by showing the trends and market shares for first-time homebuyers, as well as their characteristics.

The first section briefly describes AHS. After this, data on total home purchases and first-time homebuyers are reported for 1991–99, with separate analyses of minority and nonminority borrowers. The following section highlights the important role of FHA in the first-time homebuyer market. The final section examines differences in borrower and loan characteristics when first-time buyers are compared with repeat buyers.

AHS began in 1973, and the national sample¹ now collects data in each odd-numbered year. The most recent survey (1999) sampled 58,400 housing units, a sampling rate of approximately 1 per 2,000 units. AHS uses computer-assisted personal interviewing techniques to gather information on all aspects of housing: structure and unit characteristics, household composition and income, costs, equipment, mortgages, neighborhood characteristics, alterations and repairs, and recent movers. Data from AHS are published as tabular reports and as public-use microdata; each record of the data set corresponds to one housing unit. This article presents special tabulations of microdata from the 1993, 1995, 1997, and 1999 surveys.

It includes estimates of owner households who purchased units by year, race, and first-time buyer status. The subsample of owners who move within the periods between surveys is much smaller than the total AHS sample. For example, although the 1999 survey sampled 58,400 units, only 6,040 of these were purchased between 1997 and 1999; of these, 2,394 were purchased by first-time buyers. The previous survey years provide samples approximately the same size as the 1999 survey. Thus, our analysis is based on approximately 2,000 total and 850 first-time buyer observations per year. Although these sample sizes are adequate for generating statistically reliable totals, care must be taken when subdividing the sample further, such as by race or other household characteristics.

AHS surveys from 1993 through 1999 are used to estimate the number of buyers from 1991–99 by asking the occupants when they moved into their units. Because AHS interviews are conducted from July until November, the survey will miss households that move late in the year. To correct for this, estimates of movers are derived from the following survey (for example, 1997 movers are estimated using the data from the 1999 survey). Because this is not possible for the final year of a study, part-year data is ratio-adjusted to a 12-month basis.

Purchasers are also cross tabulated by race (determined by the race of the householder) and by first-time buyer status. Although AHS permits a finer division than simply “White” and “Minority,” data that use more race categories are somewhat inconsistent. In this study, “White” includes only non-Hispanic whites, and all Hispanic households are included in “Minority.” First-time buyers are identified by their answer to the survey question, “Has anyone in the household ever owned a home before?” Households who answer “no” are considered first-time buyers.

Analysis of AHS data shows that nearly 44 million home purchases were made in the 1990s, of which 8 million, or 19 percent, were made by minorities. (See table 1.) First-time homeowners comprised 40 percent of the total home purchases (17 million purchases). The overall trend in home-purchase volume was positive in the 1990s. In 1999, total home purchases of 5.6 million were higher than for any other year in the 1990s. (See table 2.) The rising trend in total home purchases was driven largely by disproportionate growth in the numbers of first-time homebuyers and minority homebuyers. There were 1.4 million first-time homebuyers in 1991, compared with 2.3 million in 1999. In 1991, first-time buyers represented 37 percent of total purchases, compared with 42 percent in 1999 and a 9-year average of 40 percent.²

Minority home purchases increased substantially throughout the 1990s. There were 640,000 minority homebuyers in 1991 compared with 1.3 million in 1999. Minority households are much more likely to be first-time buyers than are White households; first-time homebuyers accounted for 60 percent of all minority home purchases in the 1990s, compared with 35 percent of home purchases by Whites. In addition, the share of home purchases made by first-time, minority buyers increased during the 1990s. In 1991, 54 percent of minority purchases were by first-time buyers, compared with 66 percent in 1999. The increase in the share of first-time White buyers was more modest, increasing from 33 percent of purchases by White buyers in 1991 to 35 percent in 1999. As the above data suggest, growth of first-time purchases by minority buyers fueled most of the increase in first-time buyers during the 1990s. Total first-time home purchases increased from 1.4 million in 1991 to 2.3 million in 1999, with 53 percent of this increase due to minority buyers. The share of first-time home purchases accounted for by minorities grew from 24 percent to 36 percent. The minority share of total home purchases also increased during the decade, from 17 percent in 1991 to 22 percent in 1999.

The relative accuracy of AHS estimates is supported by comparing AHS totals to data presented in the bottom of table 2, which contains total estimates of home-purchase volume derived from totaling (1) new home purchases and (2) residential manufactured home placements from census data, and (3) existing home sales from NATIONAL ASSOCIATION OF REALTORS^® data. For the most part, AHS total home-purchase estimates are in the same ballpark as the alternative estimates. AHS data report 44 million home purchases during the decade, compared with 46 million estimated by the alternative data. However, it should be noted that for 2 years, 1993 and 1998, AHS data show decreased home-purchase activity relative to the prior year, whereas the alternative estimate (as well as HMDA data) shows an increase. Examining the trend differences evident in AHS data for these years requires further research.

The FHA market has long been recognized as a major source of funding for first-time homebuyers. FHA’s low downpayment requirements and liberal income qualification guidelines, combined with flexible credit underwriting, create a unique mortgage product for borrowers who have found it difficult to obtain mortgage financing in the private conventional market. An FHA-insured loan has often been the mechanism by which young, low-income, and minority families purchase their first home.

FHA expanded its traditional role of helping first-time homebuyers during the 1990s. The share of FHA-insured home-purchase loans going to first-time buyers increased from 62 percent in 1991 to 81 percent in 1999. In total, over this period, FHA helped 4.4 million first-time homebuyers realize their dreams of homeownership.

Table 3 combines program data from FHA with market data from AHS to examine FHA’s share of both the overall home-purchase market and the first-time homebuyer portion of that market. During the 1990s, FHA-insured loans comprised 14 percent of all home purchases, a percentage that remained approximately the same throughout the three subperiods (1991– 93, 1994–96, and 1997–99) reported in table 3.³ FHA’s share of the first-time homebuyer market was much larger than its share of the overall home-purchase market (25 percent versus 14 percent). FHA’s share of the first-time homebuyer market increased during the late 1990s, rising from an average of 23 percent in 1991–96 to 30 percent in 1997–99. It should be noted that AHS market data reported here include home purchases not financed with new mortgages, as well as those that are financed. The FHA share of the market would be slightly higher if only home purchases with mortgages were included in the market data.⁴ These comparisons still show the important role of FHA in funding first-time homebuyers.

As shown in the bottom portion of table 3, FHA’s role in the homebuying market is even larger when the analysis is restricted to minority homebuyers. During 1997–99, FHA-insured loans financed 27 percent of home purchases by African-American and Hispanic families, an increase from a 20-percent share during 1991–93. FHA’s share of the first-time market is even larger. From 1997–99, FHA accounted for 37 percent of first-time home purchases by African-American and Hispanic families, an increase from a 26-percent share during the 1991–93 period.

Table 4 compares first-time and repeat homebuyers in the late 1990s along a number of dimensions. First-time buyers are expected to be income constrained and somewhat limited in their purchase options, and the data bear this out. As expected, first-time homebuyers tend to be younger, have lower incomes, and buy less expensive houses than repeat buyers do. More than one-half of first-time buyers are under age 35, compared with less than one-quarter of repeat buyers. Thirty-nine percent of first-time buyers have incomes below 80 percent of the median, compared with 30 percent of repeat buyers. Fifty-four percent of first-time buyers purchase homes priced under $100,000, whereas only 37 percent of repeat buyers do so.

Consistent with the results reported in earlier, a higher proportion of first-time buyers are minority (32 percent), compared with 14 percent of repeat buyers. In addition, female-headed households represent a slightly higher proportion of first-time buyers than repeat buyers do: female-headed households comprise 22 percent of first-time buyers and 18 percent of repeat buyers. First-time buyers are somewhat more likely to purchase in the central city (28 percent) compared with repeat buyers (18 percent).⁵

Within these differences in distributions, first-time buyers exhibit a range of other characteristics, and even include high-income households, elderly households, and those who buy higher-priced homes. For example, 49 percent of first-time homebuyers have above median incomes.

When comparing characteristics of first-time buyers in the early part of the decade with those who purchased homes for the first time in the latter part, AHS data show that a growing share of first-time buyers were minority, were female-headed households, and purchased in central cities. In the early years, 23 percent of first-time buyers were minority compared with 32 percent at the end of the decade. Female-headed households increased slightly, from 19 percent of first-time buyers to 22 percent, while the share of first-time purchases in central cities grew from 26 to 28 percent. Finally, more first-time buyers (60 percent of the total) were under age 35 during the period from 1991–92, compared with only 55 percent from 1997–99. Examining the trend in terms of race shows that the overall trend is due to a shift among White first-time homebuyers. In the latter part of the decade, a larger share of minorities were under age 35 when they bought their first homes, compared with figures from the beginning of the decade. Despite the fact that the minority share of first-time homebuyers grew over the decade and that minorities still tend to purchase their first home at an older age than do Whites, minorities did not contribute to the overall trend of first-time buyers over age 35 during the 1990s. On the contrary, a growing representation of younger first-time minority homebuyers more than compensated for the age trend seen among White first-time homebuyers.

1. AHS also includes a metropolitan sample, but it was not used in this article.

2. First-time buyers were somewhat more prevalent in purchases of mobile homes than other housing; when mobile homes were excluded from the analysis, first-time homebuyers represented 39 percent of purchases during the decade rather than the 40 percent previously discussed.

3. In addition to examining FHA volume using FHA program data, FHA volume can also be estimated from AHS data, which includes an indicator for type of mortgage. Ignoring the small percentage of unknown loan types in AHS, approximately 14 percent of home purchases in AHS from 1997–99 involve FHA loans, which compares favorably to the 15-percent market share reported in table 3 using FHA program data. However, table 3 suggests that FHA accounts for 30 percent of purchases made by first-time buyers, whereas AHS data suggest that FHA’s share is only 20 percent of such purchases. The error appears to lie in the AHS data, because FHA program data show that 80 percent of FHA home-purchase loans are made to first-time buyers, whereas AHS data imply that only 60 percent of FHA home-purchase loans are made to first-time buyers. This discrepancy requires further study.

4. HMDA data show that FHA accounted for 19.5 percent of all home-purchase loans originated in metropolitan areas from 1997–99.

5. AHS data for all homebuyers in table 4 can be compared with HMDA data for home-purchase mortgages originated in metropolitan areas. From 1997–99, HMDA reports the following shares: 30.7 percent of home loans were for low-income borrowers (versus 33.2 percent of all homebuyers from AHS); 21.5 percent were for minorities (versus 21.6 percent from AHS); and 23.9 percent of home loans had a female as the primary borrower (versus 19.8 percent of homebuyers who were female-headed households from AHS). Given differences between the two databases—AHS includes both metropolitan and nonmetropolitan areas as well as homebuyers without mortgages, whereas HMDA data include only mortgages originated in metropolitan areas—the percentages from HMDA and AHS appear rather similar.