Groups also use census
information as a way to be recognized, as a way for the rest of the
country to know they are here. If they are not counted, they become
statistically invisible. Arab Americans were telling me recently that
they now think there are three million of them in this country. They
want the rest of the country to know that, so they want us to find them
all. The same thing is true with Native Americans because some people
think that they are a shrinking population. They're saying, "No,
we're here. We're robust, and we want the country to know it."
Isn't the Census Bureau rethinking the decennial process in part because 10 years is a long time between counts?
Yes, the data are outdated rather quickly. We use vital statistics to update the raw counts. We know how many people live in the country and roughly how many people live in the different states. What gets dated very quickly is the information from the long form, answers to questions like: "How many veterans are there? Where are they living? What are peoples' language habits? How far do people drive to work?" All of those kinds of data, by the time you get to a new census, are 10 years out of date and not very useful.
We have an important innovation called the American Community Survey. If we can get Congressional money for it, the plan is to take the long-form questions and ask them of a large sample every year, a sample of about three million households. After five years, we would have talked to 15 million households -- approximately the number we now talk to when we're getting the long-form data. If you average the data across five years, you would effectively have long-form data every year after the first five years.
If we can implement this innovation, when we get around to the next decennial
census, we won't need to ask any long-form questions. We'll just do a
count for redistricting purposes. It's a very, very different kind of
census if you're basically counting and not trying to get all this other
information.
Obviously, the Decennial Census represents an enormous personnel challenge. Didn't Census 2000 hire a number of people who were on welfare?
The census has been well under way, in one way or the other, for over a year. Not the actual count, but all kinds of other work. In calendar year 1999, we hired and moved slightly over 4,000 people from welfare rolls to work. We're now over 6,000, and we haven't even done our big hiring yet. Our hope is that we will exceed employing and moving 10,000 welfare recipients to work. We've been very pleased with the quality of people who have applied, and we haven't had to give them much more training than we would give to anyone else.
We also have a big partnership with the Department of Labor. The Department of Labor is administering a contract to give some Welfare-to-Work people free training on census work. After they work for the Census Bureau, they'll get transitional training to move them into the permanent workforce. There's a major, and I think very healthy, Federal focus on using the census opportunity to employ welfare recipients because it is a big job producer for a short period of time. But, the real goal is to get Welfare-to-Work people into permanent employment.
