Count the People, Leave the Politics Aside

For a seemingly simple count of people, the census has always caused controversy. After a long-brewing feud, last Friday the White House and some members of Congress reached a tentative agreement about the census for the year 2000. Next year the census will be administered two different ways to two cities. In Columbia, South Carolina, residents will be subjected to the traditional census method, which is comprised of mailings and head counters who travel from house to house. At the same time in Sacramento, California, the new counting method, based on statistical sampling, that the Census takers have tried to implement since the 1990 census was taken will be tested. While this sounds quite straight forward, it actually has been quite contentious.

Why would a simple count cause so much controversy? It all stems from statistical sampling, the basis of the new method of census taking. Of course, statistical sampling does not sound that new to anyone who has had to sit through the QRR exam. For those of you who may have forgotten, through statistical sampling, census takers would directly count 90 percent of households and then extrapolate that data to look at the non-responders. As a result, the country could potentially save $900 million.

The 1990 census was the most expensive census ever taken, coming to a total cost of around $4 billion, and it missed more than 10 million people. Of the 290 million people in the country, only 61 percent answered the survey. While this seems like an obvious example of bureaucratic waste that must be revised, there has been strong opposition to the alternative statistical sampling method. The furor against statistical sampling arose when some Republicans released a report that said that through statistical sampling, they could lose as many as 30 seats in the House. Therefore, they claim that statistical sampling could be use to gain political leverage by the Democrats, and so they have tried to poke holes into the technique.

Ironically enough, Congress' own General Accounting Office uses statistical sampling, and so do several other national agencies. If other agencies see fit to use this device, surely it is accurate enough for the census.

It also seems fairly obvious that, given that the population is growing larger, statistical sampling will have to be adopted sooner or later. Why not now, then, so that we can save those several million dollars and do something more useful? The largest group of lobbyists for the census are business leaders who want to use the census as free market data. Although business interests are fine and good, there are a lot more useful things that could come of several million dollars for the rest of the country, and businesses could conduct their research with their own money.

That does not mean that the census should be eliminated, especially because as things stand now, it can't. The census is a large part of our national history, probably more so than in any nation's history. It was only through the census that the Constitutional Convention gained any ground. The 10-year census was written into law to ensure that the House of Representatives could function. And even then, when the census took 18 months and counted fewer than four million people, it was not accurate. Is it surprising that with a population 60 times that size that a few people are not counted?

Unfortunately, the vast majority of people who are missed by the census are at the lower income level. For that reason, the Republican party was upset at the idea of statistical sampling. Census takers believed that this new method would compensate for those missed, and since that part of the county tends to be mostly Democratic, the Republican party felt they would lose seats if districts were redrawn. However, also included in the people missed in the Census are the rural poor, who usually tend to vote Republican, and so most people believe that few drastic changes will be necessary.

But surely, if drastic changes are needed to redraw district lines to more accurately reflect the population, they should be made. When historians look back at the census of the millennium, we don't want them to take the political tenor of the country into their consideration of numbers.

Tanya Dutta's Column appears on alternate Tuesdays.