The threat of a "postcard census."

American Demographics, August, 1991 by Ken Hodges

THE SPOTLIGHT IS STILL ON 1990, but the Census Bureau is already looking for ways to improve the decennial census of 2000. Among the options being considered is a "postcard census" that would collect only the minimum information required by law. Such a move would improve coverage and lower cost, but it would also sacrifice the small-area demographic detail census users have come to expect.

The impact of a postcard census on business users could be severe. You might expect to hear howls of protest at the very idea of a minimal census, but so far there have been none. The reason is that not all businesses see the census in the same way.

Many businesses have mixed feelings toward the Census Bureau. They can't get enough census data, they can't get it soon enough, and they can't get it for small enough areas. That's why private data suppliers produce such ambitious small-area estimates. Also, the census isn't exactly what it would be if left to the preferences of business users.

If it were designed by business users, the census might look more like a large, annual survey of consumer behavior. Businesses would also want the bureau to permit them access to individual names and addresses, all at public expense.

In fact, businesses often use census data only when they lack a better alternative. Small-area demographic profiles are valuable to businesses not because of some inherent interest in small areas, but because they get marketers closer to target households. If something better were to come along, a loss of detail in census data would be of little concern to businesses.

Something better may have arrived in the form of database marketing. The term has become a buzzword, but generally it refers to using large consumer databases to market directly to individuals or households. The databases are generated in-house at the point of sale, or they are acquired from a list supplier. They can contain millions of household records, including names, addresses, and assorted demographic details. Unlike census data, consumer databases identify individual households that have target characteristics. They also have the advantage of being current.

As database marketing becomes more common, an increasing number of business users will declare themselves ready to leave the census behind. For their applications, even a bad list could be better than a good census.

There are important limitations to database marketing, however. The quality of demographic detail on consumer databases can vary widely from address to address. A data company's vested interest in selling data can make it difficult to evaluate a list's quality. Also, growing privacy concerns are threatening database marketing's progress.

The census will never land on the corporate junk heap, because some business users will never have list applications. Many users need accurate area data for sampling and site evaluation. Consumer databases are valuable for these applications only if they accurately reflect the characteristics of the area. As impressive as consumer databases have become, they will probably never be as accurate as the census in providing stand-alone demographics.

This is why decennial census data will be important to businesses for the foreseeable future. It is also why businesses should be concerned about the negative impact of a "postcard census."

--Ken Hodges, Donnelley Marketing Information Services, Stamford, Connecticut