Lessons from the German census
When the results of the 2011 German census were announced recently, they included an embarrassing error – at least in the demographics world. It showed the German population was 1.5 million people short of what the government had expected. The news dealt a blow to Germany’s reputation for efficient record-keeping, and it’s also relevant to how the next U.S. Census is conducted.
The German census, which was the first since East and West Germany reunified in 1990, counted 80.2 million people. East Germany had last taken a census in 1981, and West Germany in 1987. Since then, Germany had updated its population estimates with data from lists of residents kept by each locality, which also contain basic demographic information. These “population registers” are supposed to be updated with births, marriages, deaths, or changes of address.
The major reason for the shortfall in Germany’s population count was that local registers were not updated when foreign-born residents left the country. The German statistical office had estimated that there were 7.3 million residents of Germany without German passports, but the census counted just 6.2 million.
Most Western European and Nordic countries keep such registers, and most also keep a central register with all the local-level information. Some have used this data in place of a house-to-house census.
In Germany, the West German government’s attempt to link census records to municipal population registers led to a lawsuit over privacy issues that forced officials to cancel plans for the 1983 census. History hangs over the privacy debate in Germany, where census data were used to target Jews and other minorities during the Nazi era.
The maintenance of population registers in Germany and other countries has been cited as a potential partial role model for the U.S., where the Census Bureau has considered for decades how government records (or those from commercial vendors) might help the traditional census-taking process. These records could include files from federal agencies, as well as state, tribal or local data.
The most obvious use of administrative records would be to count the growing number of U.S. residents who do not send back their census forms, and that has been the main focus of Census Bureau research on this topic. But in its 2011 interim report, the National Research Council’s Panel to Review the 2010 Census urged use of “records-based information to supplement and improve a wide variety of census operations,” such as the crucial task of updating the address list that is the backbone of a successful census.
As it has in Europe, use of these records could raise concerns in the U.S. about personal privacy. Information collected by the U.S. Census Bureau is protected by law, and is never supposed to be used to identify individual respondents, but some Americans do object to answering questions on privacy grounds. Adding other government records to the mix could provoke more objections.
Another obstacle to using administrative records is that they can be incomplete or incorrect. A Census Bureau research paper released in March reported on an experiment in which a sample of federal government records (including a postal change-of-address file) was employed to identify people who were counted more than once in the 2010 Census or counted in the wrong place. The results were “not very promising,” in part because the federal records were not current or complete. The German census results offer a reminder that this is a problem for other countries too.
D’Vera Cohn is a Senior Writer at the Pew Research Center's Social & Demographic Trends Project.