NPR: Debate Over the 2010 Census

Yesterday, I listened to NPR's All Things Considered broadcast on the 2010 Census debate, which can be found here. Having briefly considered working part-time for the Census after receiving an e-mail from my college saying they were hiring, this particular topic peaked my interest. I never realized how much debate there is over the census, but now that I think about it, I can see why.

Jacki Lyden,the host, states that more than 100,000 people started working for the Census Bureau this week. There will be 1.4 million temporary jobs with the Census Bureau over the course of conducting the 2010 census. However, there is still widespread debate over the best methods to use in collecting the census data. The 1990 Census missed so many people that the Clinton Administration tried to implement the use of statistical sampling in the 2000 census. This sparked tremendous political debate between Democrats and Repulicans. Democrats want to employ a method of statistical sampling in which a model could be used to account for the people who are missing from the actual raw data. Republicans disagree with this practice and believe that physically counting each citizen is the only way to conduct the census and that no adjustments should take place after that. The battle wound up in the Supreme Court and eventually President George W. Bush precluded statistical sampling as an option.

The stakes are high in this particular debate. Census data is not just there to provide something for all of us political science majors to use in our reports, it has actual consequences in the political realm! How much federal money goes to your state, the number of representatives your state gets in the House of Representatives, and how many people vote on your state's behalf in the electoral college all depend on the census data.

Representing the Republican point of view, Newt Gingrich spoke on the show. Mr. Gingrich states that all Republicans want is for every American to be counted, the way the Constitution intended. Their concerns with statistical sampling are that it will not accurately represent the population and that it will distort the figures. His solution is an outsourcing of the census to private companies to avoid workers who have an agenda and to enhance accuracy. When asked if he thought President Obama would try to reverse the restriction President Bush made against statistical sampling he said that he thinks President Obama will try to change the rules and adjust the count after it is taken.

Representing the Democratic point of view was Robert Shapiro, Supervisor of the 2000 Census. Mr. Shapiro stated that the census is the largest nonmilitary operation the Federal government carries out. He explained that statistical sampling happens after all of the census data is collected. Then 1,000,000 homes are selected at random. The occupants are re-interviewed and the data from all one million homes is used as a sample. The sample is checked against the entire population to see if the population is adequately represented. If it is not, then based on a statistical model, they make adjustments. He emphasizes that statistical sampling is not in place of an enumeration, it is in addition to it, and actually makes the census data more reliable. Mr. Shapiro also states that the Census data does not have the kind of impact on Congressional results that the Republicans think there is.

After listening to an argument from both sides, I have to agree with the Republicans on this one. Though I have used statistical sampling and relied on it in many of my classes, I do not think that I would chance the amount of federal aid, how many representatives we have and how many electoral votes we have on a statistical model that could always be flawed. I do not see why we cannot just have a good, old-fashioned head count. I realize that some people avoid participating in the census out of fear or just apathy and that this is why so many people go unaccounted for. Perhaps then, we can determine some kind of incentive to motivate people to participate in the census and to provide accurate information. If we can convince people that the census is a good thing and that no one is going to get in trouble because of it, then we would not have the need for statistical sampling. Maybe we should be focusing our efforts there.