Business Uses of Census Data
Census data are used by many in the private sector, by for-and not-for-profit organizations. Retail establishments and restaurants, banks and other financial institutions, media and advertising, insurance companies, utility companies, health care providers, and many other segments of the business world use census data. In the past, household-level data on consumers at the zip-code and census-tract levels have been classified by characteristics such as age, sex, and income. Increasingly, however, individual households are contacted by direct mail or some other type of direct media (e.g., newspaper inserts). This appendix contains some examples of uses of census data by various segments of the business community—evidence of the great use of census data by businesses is provided by the Division of Research and Statistical Services of the South Carolina Budget and Control Board (the state data center for census information in South Carolina), which estimated that 35 percent of the annual requests received for census data are from businesses.
Small-area data are important for many business applications. Some businesses use small-area data as a substitute for household-level data. More important, however, is the ability to aggregate the small-area census data to nonstandard geographic areas—for example for business trade areas. As long as these data are available, businesses can create aggregations of data into areas for which data have never been published. The smaller the level of geography for which data are available, the more creatively businesses can create aggregations and the more precisely they can define the geographic area.
Several key types of small-area census data (at tract and other geographic levels) are used for business purposes: age; education; employment; housing
unit age, tenure, heating fuel, type, value, and rent; income; occupation; persons in household; phone availability; race and ethnicity; commute to work; and vehicles per household. Products—such as maps showing the concentration of a specific racial or ethnic group by specific areas (e.g., county, census tract, or zip code) or maps showing moderate-, high-, or low-income areas—can be produced using census data. Data also can be used to create consumer profiles, which can help in targeting advertising to current and potential customers; finding new customers; and analyzing locations, selecting sites, and competing against other businesses in a market area. Both the maps and consumer profiles (which may also be linked to a map) are used by businesses to target their markets more effectively. As the use of geographic information systems has grown, the demand for small-area geographic data has also grown. And, in turn, the new-found congruence of accessible geographically referenced small-area data is promoting the use of small-area census data for business purposes even further.
For example, a retail corporation with plans to expand could analyze potential markets before selecting sites. A specific case study (Kintner et al., 1994) involved examining and assessing various markets for a corporation's planned expansion. Several potential markets were selected by the corporation for the expansion, and the corporation wanted to determine which of the potential markets would be the most successful. Although the company's staff would make the final decision about the exact location of the sites, consultants were hired to analyze the potential revenue for each market. First, the consultants developed a model for analyzing the potential markets. The model took into account a number of variables—such as population, number of firms employing 100 or more workers, number of vehicles entering the county, and size of the transient population—that could help predict the viability of a site in areas selected for analysis. Some data were from business sources, but census data provided an essential component for analysis. Information on existing markets was used in the model to help determine its accuracy. Then, the predicted revenues for each of the existing locations generated by the model were compared with the actual revenues of those markets, enabling the corporation to assess and identify the strengths and weaknesses of the model. Next, data were collected for the potential new markets. By adding the new data to the model, revenue estimates were created for the potential markets, and the markets were ranked based on their predicted revenue. Markets that were the most promising were selected for additional analyses and reviewed by the corporation's staff, who then were able to select the best markets for the corporation's expansion.
It is clear to the panel that businesses use small-area data creatively and effectively for a number of applications, and that small-area census data are important to those applications. However, it is difficult to foresee the effects of a loss of small-area census data. There could be a negative impact on efficiency and competitiveness—impacts that would be difficult to measure.
This appendix describes the business uses of census data for a variety of
industries, including retail and restaurant, banks and other financial institutions, media and advertising, insurance, utilities, health care, nonprofit, and others. The review is not exhaustive of all industries, nor comprehensive in the many ways that census data are used. Rather, the purpose is to highlight several common uses, for a variety of industries, to illustrate the specific ways census data are used to reach business decisions and to improve business marketing. The examples cited below are taken from Thomas and Kirchner (1991), a recent publication on desktop marketing that describes ways that demographic data are used by businesses.
RETAIL AND RESTAURANT
Retail and service businesses, such as restaurants, use data to decide where to locate their stores and how to effectively market their goods and services. A retail chain might use population, poverty, income, and labor-force data for a state and for a city or county to study the possibility of a retail outlet. For example, county-level population figures for women aged 16-34 years could be used to help determine the location for a maternity shop. Or a children's clothing retailer could use age data, income data, and retail statistics to select a location.
A fast-food restaurant chain was able to better target employee recruiting efforts and improve service by analyzing concentrations of the population with desirable employee traits/lifestyle characteristics (including longevity of employment). To accomplish this task, the restaurant chain identified the characteristics of its current base of employees and located areas with high concentrations of potential employees—a population whose characteristics were the same as the most successful current employees. In creating the profile of current employees, past and present employees with at least 6 months of service with the restaurant were categorized into 1 of 50 categories based on census block-group characteristics of their neighborhoods. The categories were charted according to the percentage of total current employees falling into the category. Using the data, the restaurant was able to identify categories of workers that were likely to become restaurant employees and determine areas where they lived. Recruiting efforts were targeted to those areas using mail and newspaper advertisements, among other techniques. The restaurant has found this ability to be useful in existing markets and new markets, and it has helped reduce turnover in the restaurants, resulting in improved customer service (Thomas and Kirchner, 1991:55-60).
For selecting restaurant sites, a general area, as well as specific sites for the restaurant can be evaluated. By looking at selected demographic data by specific levels of geography (e.g., counties and zip codes) the characteristics of the potential customers can be determined. Employment data at those same levels may also be evaluated. These analyses taken together can help the restaurateur
select the best site for a successful restaurant (Thomas and Kirchner, 1991:61-63).
BANKS AND OTHER FINANCIAL INSTITUTIONS
Like retailers and restaurateurs, banks and other financial institutions can select the best locations for branch offices by analyzing population, demographic, and economic data from the census. More importantly, however, banks and financial institutions require median household income and income distributions by census tracts to ensure compliance with federal mortgage lending guidelines regarding race, and for meeting other regulatory requirements, particularly the Community Reinvestment Act, Home Mortgage Disclosure Act, and the Federal Insurance Improvement Act of 1992.
For example, the Community Reinvestment Act mandates that financial institutions meet deposit and credit needs in the communities they serve. The federal agencies that supervise financial institutions are required to assess whether the financial institutions in an area are meeting the needs of the community. To assess its compliance with the mandates of the Community Reinvestment Act, a bank wanted to determine the ratio of its loans to its deposits. Using customer data and a software system that is able to link demographic and client information, the bank was able to determine the loan-to-deposit ratio for its service areas. Thus, the bank was able to assess itself whether it was complying with the Community Reinvestment Act before the regulatory agencies conducted their audits. If there were areas with a discrepancy between deposits and loans, the bank would be able to make corrections in those areas (Thomas and Kirchner, 1991:114-116).
Census data can be used by banks to develop locally focused marketing programs. For example, a bank can determine the potential success of a particular new service by looking at how and where to market the service. A demographic profile of service areas based on age, deposits, household income, and credit use can be created. By grouping and mapping the frequency of the four variables mentioned above, along with a consumer profile, areas where the service is likely to be used can be identified. Those areas then can be targeted for promotion and implementation of the new service (Thomas and Kirchner, 1991:93-97).
In trying to determine if acquiring a competing banking institution (Bank B) would be a feasible and profitable way to expand and diversify its services, Bank A wanted to assess the proximity of Bank B's branches to its existing branches, the comparability of existing customers of Bank A with Bank B, and the comparability of services offered by both banks. The population (current and future projected) of the areas surrounding branches was compared, and income estimates for Bank B's locations were analyzed by census tract level (Thomas and
Kirchner, 1991:102-108). Using these analyses, Bank A is able to make the best decision about acquiring the competing bank.
A bank can analyze the potential performance of new and existing markets by developing a profile for evaluating those markets. By combining demographic characteristics of data on national financial behavior with demographic data for a particular market, a profile of the bank's service area can be developed. Using the average state performance of branches as a benchmark, the bank can determine the amount of increased business for areas performing below the state average if those areas grow to the state average level. This can help the bank determine areas for increased market analysis and marketing efforts, while also pinpointing markets that are performing at or above the state average that need to be maintained and protected from competitors (Thomas and Kirchner, 1991:111-113).
MEDIA AND ADVERTISING
Newspapers use census data in stories to profile the demographics of blocks, neighborhoods, towns, cities, counties, states, and other geographic areas. Census data also provide demographic background for other stories of general and specific interest to the public, e.g., what are the socioeconomic characteristics of areas with the most lawlessness in the Los Angeles riots? What is the most ''middle class" tract in L.A. County? And what are commuter travel patterns in Orange County? Examples included in responses to the panel's survey of state data centers (see Appendix E) noted that all variables to the block-group level in various census geographic files can be used to describe the demographic and economic characteristics of places and areas. Also reported in a survey response was that the Los Angeles Times recently used 1990 census data in more than 300 news stories within one year.
The collection of consumer zip codes may be used to create a consumer profile for an area. For instance, a radio station might collect a caller's zip code and link it to demographic data to develop profiles of listener preferences (Thomas and Kirchner, 1991:34). In turn, the station can determine the potential success of a particular radio format for a given area and target marketing campaigns accordingly. Those profiles can also be linked with ratings information and used to optimize advertising revenue.
A cable television company analyzed purchase of pay-per-view events by census tract maps (Thomas and Kirchner, 1991:37) and created customer profiles by block-group level. Those customer profiles assisted the company in focussing its marketing efforts to specific customers. For example, pay-per-view sporting events can be marketed to the subscribers that are most likely to purchase the event, rather than to the entire customer base, thus increasing the advertising value.
In a case study, an insurance company wanted to determine if some of its offices had allowed policies to lapse more than others. The company first wanted to determine if sites with high lapse rates were located in areas with high-risk customers. To determine the different characteristics between lapsed customers and continuing customers, the company created a profile of current customers, as well as a profile of lapsed customers. Based on the profile, the company determined that the continuing customers were generally more affluent and more family-oriented. When the profile for continuing customers was compared to the profile for lapsed customers, the company found that lapsed customers "tended to be more downscale than average" (Thomas and Kirchner, 1991:119). Using the data, the company was able to estimate what the performance of various offices should be, based on their geographic locations. For example, some of the offices were located in areas where the population could be characterized as high-lapse customers. Those offices, it was determined, could expect lower overall performance (Thomas and Kirchner, 1991:117-121).
Utility companies use census data to target low-income areas or areas with special needs, as well as for market research. Most utility companies have special lower rates for poorer, elderly, or disabled customers. Census data help companies note special areas for individual contact and special services and rates. An electric or gas company can use customer records to determine their share of the market. Using customer address information, a utility company can determine areas where it might be desirable to increase customer volume through greater name recognition. Other companies are using census maps to plot the location of their utility lines so they can quickly reference the proximity of lines to population areas.
HEALTH CARE PROVIDERS
Health care providers use census data to determine the need for additional hospital services, physicians, urgent care facilities, or other type of medical services in an area. For example, a hospital used data to study population trends when looking into building an off-site facility in a rural area, so that better health care could be provided to residents in that area. Using characteristics such as race, age, sex, and income for the health service area, a provider can determine if there is a need for additional doctors or other health services in an area. By estimating the need for services in an area, the best site for a doctor's office can be determined (Thomas and Kirchner, 1991:130-136). A hospital's selection of urgent care center sites is aided by analyzing patient records (including address
and other information about use of the hospital's health care services). Information that may be used in the analysis includes physician's name, reason for hospital admission, distance from the hospital, and insurance coverage. Those variables can be geocoded and aggregated by physician or group practice and mapped to show concentration of use, need for services, type of services provided, etc. (Thomas and Kirchner, 1991:137-140). The same type of analyses can help determine a need for certain health services, e.g., mammography, obstetrical, or family practice services, in a particular area.
Nonprofit organizations, such as community advocacy groups and rural health clinics, need small-area census data to define and describe the needs of the population. Community-wide data does not always adequately describe a neighborhood or service area. For example, a rural health clinic would use census data for a special federal designation based on location and population served. Such clinics provide access to important medical services and establish reimbursement rates for services to ensure that clinics can remain in service. Community service organizations help small communities prepare grant applications for funding based on census data. Those same organizations also help evaluate and target community needs for fundraising campaigns. Local chambers of commerce rely on census data to keep businesses, attract new businesses, and assist established businesses market their products locally. A neighborhood housing service organization working to improve the quality of life of low-and moderate-income residents of communities would likely use census data to draft proposals and prepare applications for federally funded programs.
The examples above touch on only a few of the innumerable business uses of census data. Other uses and users include:
real estate appraisal companies that use census data to establish an inventory of existing real estate, the current and future demand for real estate, and the value of that real estate.
attorneys, who use census data, for example, to ensure equitable racial/ethnic representation on juries and to show the disparate impact of housing practices on particular classes of the population.
telecommunications firms, which might determine the percent of the population in a particular area that graduated from high school for marketing purposes.
In addition, there are firms that use small-area census data in developing
software systems that assist businesses in choosing sites, targeting populations, creating marketing strategies, etc. Those products are sold for marketing and demographic uses like those described above. Many of the private suppliers who develop business software systems for marketing and other purposes use block-group level data in creating the user-defined areas for their programs. Although it could be argued that block-level data are inaccurate, when the data are aggregated (even for small aggregations), the accuracy improves considerably. For these companies and business users, the need for small-area census data is focused on the ability to conveniently aggregate small-area data to user-defined areas, for which accuracy is likely to be acceptable, rather than on the use and accuracy of individual block-group data. The development, existence, and use of the products create a large market for census data and ultimately increases the overall value of the census.
As of 1991, other businesses that use census data in their products include:
CACI has several marketing information systems that are used by businesses in assessing sites and developing consumer profiles. One of CACI's database systems contains census geography, and telephone exchange and zip-code areas. Another of its database systems facilitates the creation of area profiles for targeting marketing efforts; this system classifies over 200,000 census block groups into 44 lifestyle categories. CACI also annually publishes a series of data on cities, counties, and zip-code areas (Thomas and Kirchner, 1991:208-209).
Claritas Corporation works mostly in the financial and media arenas. One of its marketing software systems allows for data manipulation and market analysis, mapping, and the preparation of reports. The system can be customized by users to fit their individual needs. A second Claritas system groups areas into 40 types for targeting consumers in those areas. The groups, which are based on census data and records of consumer purchases, can then be linked to other data available. Other products available from Claritas include Census Bureau demographic data, TIGER mapping files, and data on businesses (Thomas and Kirchner, 1991:210-211).
Donnelley Marketing Inc. has a database of over 85 million households and supplies population estimates based on census tract data. One of its database systems includes data on demographics, retail sales, and lifestyles. The system was initially used by retail business and real estate developers in site selection applications, but it is also currently used by financial service organizations, media, government agencies, and health care providers. Another database system classifies neighborhoods into 47 different lifestyles, and is available at the block-group, census-tract, and zip-code levels, and can be linked with other data, including a company's own customer files (Thomas and Kirchner, 1991:215-216).
National Decision Systems provides marketing services to a variety of business in the United States, including retail, restaurant, insurance, and advertising.
It provides services such as target marketing, market analysis, site evaluation, and direct marketing. One of its database systems merges business data with census data. The system classifies every U.S. household into one of 50 market categories. Those categories can then be used for a variety of marketing purposes. In addition, a business's information on its own customers can be incorporated into a specialized program designed by the client and National Decision Systems (Thomas and Kirchner, 1991:218-219).
In addition, there are several marketing and demographic information firms that produce marketing and demographic analysis programs specifically for use by the health care industry.
Kintner, H.J., T.W. Merrick, P.A. Morrison, and P.R. Voss 1994 Demographics: A Casebook for Business and Government. Boulder, Colo.: Westview Press.
Thomas, R.K., and R.J. Kirchner 1991 Desktop Marketing: Lessons from America's Best. Ithaca, N.Y.: American Demographics Books.