Innovating and Responding to Customer Needs:
The Residential Finance Survey

by Mary B. Schwartz and Robert J. Schneider, Jr.

Fact-finding is one of America's oldest activities. In a country of 118 million housing units and 275 million people, providing timely, relevant, and quality national data is often a monumental task. At the U.S. Census Bureau, we conduct the Residential Finance Survey (RFS) the year after each decennial census.

You may have already participated in this survey at one time or another, considering that we’ve been doing it since 1950. The reason we conduct the RFS is to collect and produce data about the acquisition and financing of non-farm, privately-owned residential properties. Our survey data are used by several sources. For instance, researchers and policymakers use the data to design policies that ensure a sound residential finance and delivery system. Real estate and mortgage finance businesses use the data to gauge and improve their competitiveness. "Many large banks used the 1991 RFS data to determine the degree of their future involvement with home equity loans," according to Peter J. Fronczek, Chief, Financial and Market Characteristics Branch (F&MCB) of the Housing and Household Economic Statistics (HHES) Division, U.S. Census Bureau.

We have designed the Residential Finance Survey as a two-stage survey. In the first phase, we ask owners of sampled residential properties how they acquired their properties, whether they have mortgages on their properties, and how they would describe the properties. If the properties are mortgaged, we ask the property owners to identify their mortgage lender(s). In the "lender phase," we follow up with the lenders to verify technical information, which the owners are not likely--or even required--to have.

Industry Changes Necessitate Changes to Lender Survey

During the 1980s, the financial services industry landscape changed. Large multinational lenders replaced smaller lending institutions. As a result, in the last (1991) RFS, larger lenders were inundated by many eight-page paper questionnaires, which they were required to complete in a short time. Given the changes in the industry, we began actively investigating how some emerging technologies could be applied to reduce the burden on large lenders and to increase the efficiency of the survey. Fronczek, who is directing the 2001 RFS, identified one of the primary objectives for the survey as "making the task of responding as simple and practical as possible for lenders."

To achieve this, the Financial and Market Characteristics Branch of the Housing and Household Economic Statistics (HHES) Division, which coordinates the survey, initiated a series of meetings with members of our Computer Assisted Survey Research Office. They explained that diskette Computer Self-Administered Questionnaires (CSAQs) were already being used for many company surveys. A CSAQ diskette enables a company to easily report similar information on each of its establishments. For example, a company could report information on payroll, revenue, and number of employees for each of its establishments in one CSAQ by importing the data from their databases rather than keying it. Companies would also be spared the burden of sorting through questions to determine their applicability, because the CSAQ would automatically skip questions that did not apply. In addition, automatic calculations remove the burden of lender calculations. Furthermore, computer programs resolve data discrepancies during lender reporting, which eliminates the need for telephone follow-up.

This piqued our interest in using a CSAQ to make it easier for participating mortgage lenders to report relevant information common to each of their mortgages, such as:

• beginning and outstanding principal balances
• interest rates
• principal and interest payments
• insurance payments
• real estate taxes

We quickly decided against the option to administer a Web-based CSAQ, because other survey results showed that many people are still understandably concerned about the security risk of sending sensitive data over the Internet. Although our RFS web site cannot be used for actual reporting, we update it quarterly to provide current detailed information about the RFS.

What May Be Economical for Some May Be Uneconomical for Us

We reviewed this technology and--much to our dismay--determined that the development costs of a CSAQ for the RFS were too high. If the RFS were an ongoing survey, the development costs could be amortized over time. However, because the RFS is taken only once every ten years, the fixed costs were excessive. "None of this really mattered, because we didn’t have enough time to properly develop and test a CSAQ before the RFS goes to the field. So, the CSAQ turned out to be an unrealistic option for us all the way around," Fronczek added.

Alternatives We Are Investigating

We have not, however, entirely abandoned the option for lenders to report electronically in the 2001 RFS. In a presurvey, we will ask large lending institutions if they prefer to "provide a computer file of mortgage information in a format suggested by the Census Bureau," assuming that any of the properties on which they hold mortgage(s) fall into the RFS property sample. To meet the anticipated demand for an electronic survey, we are consulting with staff in various areas throughout the Bureau to learn from their experiences. For example, we are talking to staff working on Finance, Insurance, and Real Estate (FIRE) industry surveys to determine if their methods in the five-year census of companies in FIRE industries can be applied to our RFS lender survey. "We are interested in, and aggressively investigating, similar survey methodologies that other Bureau divisions have successfully employed," said Fronczek.

Commitment to Our Customers

Census Bureau employees proudly help to carry out the Bureau’s mission by being innovative in their work and responsive to our customers. "We constantly strive to improve our survey methods and techniques to ensure the accuracy and reliability of our data. Examining the potential for electronic RFS questionnaires is another significant way in which we anticipate and attempt to satisfy your needs," Fronczek added. Through continuous communication with you, our customers, we are able to achieve these goals.

About the Authors
Robert J. Schneider, Jr. is a Marketing Specialist, at the U.S. Census Bureau in Suitland, MD. You may reach him at (301) 457-4081 or Robert.J.Schneider.Jr@ccmail.census.gov
Mary B. Schwartz is in the Housing and Household Economic Statistics (HHES) Division of the Census Bureau. You may reach her at (301) 457-3190 or mary.b.schwartz@ccmail.census.gov.