WASHINGTON -- As many of you know, the U.S. Constitution requires Congress to perform a census every 10 years. The data collected during the census, including the number of American residents and their household characteristics, is used to apportion seats in the House of Representatives, to redraw congressional and state legislative boundaries, and to allocate billions of dollars in federal assistance to state and local governments. Census data also provide information on population growth patterns and demographic information that are used by both the private sector and by federal, state, and local officials.

I chair the Senate subcommittee on Federal Financial Management, Government Information, Federal Services and International Security that is responsible for overseeing the census, and we have held a number of hearings over the past year looking at the Census Bureau's preparation for the upcoming 2010 census. Specifically, I have been trying to evaluate the Census Bureau's plans to ensure the accuracy of the census, scheduled to be conducted starting April 1, 2010.

With such a substantial reliance on the census data, accuracy is critical. Unfortunately, every census in the nation's history has failed to count all residents, resulting in an undercount of the general population. Looking back at the 2000 census, its budget was unprecedented – more money was spent on it than on any previous census. As a result of its hard work, the census bureau was able to reduce both the number of Americans who went uncounted.

Despite the Census Bureau's success, undercounting still remained an issue for many communities throughout this country. In 2000, the official census count was 281.4 million and the adjusted estimate was just over 284 million. The bureau reported a net error of just under one-half percent. That sounds pretty good. The truth is, though, that there were huge errors in 2000 that I don't think we can afford to repeat this time around. The bureau's own Accuracy and Cover Evaluation survey revealed that 6.4 million people were missed and 3.1 million were counted twice. In other words, the 2000 census produced a net undercount of 3.3 million people. 

The undercount would be less problematic if it were evenly distributed among all Americans, however studies show that undercounting tends to have a disproportionate impact on racial and ethnic minorities, children, and immigrants. In 2000, Asian-Americans were missed nearly twice as often as whites; African-Americans nearly three times as often; and Hispanics were missed four times as often.

In past censuses, the bureau reduced its net undercount by letting the people counted twice substitute for those missed. That may work for statistical purposes, but it is problematic. The people counted twice are not like the people missed. They are not the same race, they do not have the same income and they do not live in the same places. 

Children; urban residents, especially those in low-income central-city areas; people with limited English language skills; and racial and ethnic minorities all are much more likely to be missed in the census. These groups are less likely to have a regular address, may fear cooperating with government authorities or may face language or cultural barriers to complying with the census. By contrast, over counted groups are those who fill out are included on more than one census form. Studies have shown that these tend to be higher income, suburban households. For example, retirees with both a summer and winter residence may receive and fill out a census form for both addresses and college students may be counted at college as well as at their parents' home.

The importance of getting this right is striking when you think about the countless ways in which we depend on census data. For starters, the undercount affects the distribution of federal funds that are allocated on the basis of population. A study performed by Pricewaterhouse Coopers found that the undercount in 2000 would cost states $4.1 billion. This loss of federal funding taxes the resources of state and local governments and compromises the level of services provided to residents – the people we work for.

Undercount in the census also has serious political implications. If political representation is based on population, undercounted groups get less credit for their population. This skews the make-up of the House of Representatives and results in some communities being under-represented while others get more of a voice than they are due.

Reaching out to those Americans who are historically hard to count is even more important when you consider that for every 1 percent of the population that does not respond to the census, we're going to have to spend about $75 million to go door-to-door to get everyone counted.

It is vitally important, then, that we do the necessary hard work now so that we get an accurate, cost-effective count in 2010 that will serve us well in the next decade. 

As the Census Bureau begins its final preparations for 2010, we need to make sure that they are reaching out as aggressively as we can to historically undercounted groups.

And I want to urge everyone in Delaware and nationwide to participate in the census so we can get the best count possible.

For questions, comments, or problems enountered while using my website, please contact the Webmaster.