Supported by an extended period of declining interest rates, the housing market in the Minneapolis-St. Paul Metropolitan Statistical Area, 1/ or MSA, recently experienced a period of strong, sustained growth in home building, homebuying and mortgage lending. 2/ The growth has been one factor behind the region's increasing homeownership rate, which jumped from 69 percent to over 72 percent between 1990 and 2000, as the stock of owner-occupied housing grew by 180,967 units, or 28 percent.
Nevertheless, as the other articles in this issue reveal, homeownership rates did not rise equally across the state and the region. Which leads to the following question: How did households in different income or racial and ethnic groups fare in their pursuit of homeownership, especially during the housing market boom that occurred from roughly 1997 to 2003? To complement this issue's features on homeownership and emerging markets, this article examines home lending trends in the Minneapolis-St. Paul MSA during the boom period. It uses data from Census 2000 and Home Mortgage Disclosure Act (HMDA) records for the 1996 and 2003 calendar years. Data from 1996 provide a baseline for understanding home purchase lending patterns before the recent market boom, while data from 2003 are the most recent available for analysis. 3/
Measuring the boom
Examining HMDA data for lending activity in the Minneapolis-St. Paul MSA reveals a period of robust home purchase lending. In particular, originated loans for home purchases increased dramatically from 54,545 to 80,519, or by 48 percent, between 1996 and 2003. (See the table.) In order to standardize this measurement of home purchase lending, the calculation measures all originated loans as a proportion of the total owner-occupied housing units for that year. Total units are calculated using data from U.S. Census 2000 and the Current Population Surveyfor 1996-2003. Using this standardized measurement reveals that the number of loans per 100 owner-occupied units appears to have grown consistently over this period, from 6.9 in 1996 to 9.0 in 2003. In addition to the number of home purchase loans originated, the volume of lending—in terms of both the average loan size and total value of loans—also increased during this period. In 1996, the average loan size was slightly more than $128,000, with approximately $7 billion in total annual loan volume.
Low- and moderate-income households see gains
Did minority and low- and moderate-income households see the same levels of growth? For low- and moderate-income borrowers, defined as making 80 percent or less of the area's median family income, the number of originated home purchase loans increased from 21,987 to 35,210, or by 60 percent, between 1996 and 2003. As a standardized rate, the number of originated loans increased by four percentage points—a slightly stronger rate of increase when compared to the overall population. This group's share of originated home purchase loans increased from 41 to 46 percent during this period.
In terms of the overall lending volume, low- and moderate-income households also increased their share, from 28 percent to 35 percent. Due to increased housing costs in the region, low- and moderate-income households increased the amount they borrowed. For example, comparing the dollar amount that represents 80 percent of area median income to the average loan size over time reveals that the ratio of income to loan size increased during the eight years, from 1.68 in 1996 to 2.41 in 2003.
Nonwhites increase their share
For nonwhite households, which include all groups but non-Hispanic whites, home purchase originations increased from 4,020 to 8,349, or by 107 percent, between 1996 and 2003. Using the standardized measurement for this group reveals that the number of loans per 100 owner-occupied units increased by almost three percentage points over the past eight years, once again reflecting the increased volume of home purchase lending to nonwhite borrowers. The number of home purchase loans originated for white households also increased, but at a more modest rate of 32 percent. More importantly, the total amount of loans for nonwhite households increased by 235 percent, and the share of all home purchase lending for nonwhite households increased from 6.3 percent to 9.8 percent during this period. The increased percentage is closer to the group's actual share of the metropolitan population, which was approximately 15 percent in 2000. 4/
Viewed at the census tract level, the increase in home purchase originations by nonwhites was generally concentrated in the core central cities. Gains also occurred in the suburbs, particularly those to the north and northwest of Minneapolis, such as Brooklyn Center and Brooklyn Park, and to the east of St. Paul in Maplewood, North St. Paul and Oakdale. To the southwest, the suburbs of Shakopee and Chaska also experienced an increase in home purchase originations by nonwhites.
The challenge ahead
In summary, the findings detailed above suggest that a number of positive trends for low- and moderate-income and nonwhite households developed during the recent housing market boom. During a short period of eight years, both groups increased their share of total loan applications and lending volume. While these gains are promising, the level of home purchase lending to nonwhite households remains low relative to their share of the region's population. For nonwhite households, whose household heads tend to be younger than the rest of the population, the challenge will be to maintain the strong gains made over the last eight years and to target those subgroups whose lending activity and rates of homeownership remain sluggish.
The Home Mortgage Disclosure Act (HMDA) was enacted in 1975 to target investments to needy neighborhoods, identify possible discriminatory lending patterns and help regulators determine whether financial institutions are meeting the mortgage lending needs of their service areas.
Under HMDA, mortgage loan data from a variety of financial institutions—such as banks, savings and loan associations, and credit unions—are collected and made public. Enforcement of the data collection process depends on an institution's asset size and other criteria. For a depository institution, regulators use assets, the locations of home and branch offices, and other factors to determine whether it must comply with the act. In 2003, HMDA covered depository institutions that had more than $32 million in assets and an office in a Metropolitan Statistical Area (MSA). The HMDA coverage threshold for nondepository institutions with an office or loan activity in an MSA was an asset size of $10 million or a record of making more than 100 home purchase loans during the previous calendar year.
Since the early 1990s, academics and community groups alike have used HMDA data to inform research and policy debates on a variety of lending issues. The data, released each year by the Federal Financial Institutions Examination Council, provide a useful annual snapshot of lending activity for a metropolitan region, especially with regard to home purchase and refinance. For each loan application, the lending institution collects information on characteristics of the loan, property and borrower. Information on the loan action taken by the financial institution is also gathered. Each loan is identified by census tract, county and MSA. Data on an applicant's race and gender and the gross annual income used in the lender's decision-making process are also collected.
Source: Federal Financial Institutions Examination Council, A Guide to HMDA Reporting. Getting it Right!May 2003. See also www.ffiec.gov/hmda.