How Will Bush Come to This Census?

  • Share
  • Read Later

Census Bureau director Kenneth Prewitt announces the results of the 2000 census

The good news for George W. Bush is that the country is going Republican. The bad news is that come January, he just might have another counting fight on his hands.

The Census Bureau announced Thursday that the population of the United States is up to 281,421,906, some 33 million more than in 1990. And with population figures determining the how the 435 House seats (and electoral college votes) are apportioned to each state, it will cheer Republicans to hear that most of America's population gains are in the states that November painted red.

Arizona, Georgia, Texas and Florida will each gain two congressional seats and thus two electoral-college votes; Colorado, Nevada and North Carolina and California (which went for Gore, if barely) will each gain one.

The losing states — although no state actually lost population, for the first census ever — are mostly blue. New York and Pennsylvania will lose two seats each, and Connecticut, Illinois, Michigan and Wisconsin each lose one, along with Bush winners Indiana, Mississippi, Ohio, and Oklahoma.

A new numbers game

And Bush's home state of Texas is now the nation's second-most populous, supplanting Democratic stronghold New York. Implications: If the election were held all over again — perish the thought — and Bush won exactly the same states, he would win the electoral college 278-260, rather than 271-267. And House Republicans would have a 16-seat majority instead of nine.

Of course, Gore 2004 isn't exactly slipping away. The populations on the fastest rise are urban and minority, in booming population centers like Atlanta, Miami and Las Vegas, and the nation's cities are classic Democratic stomping grounds. To assume Florida stays red after the 2000 election fiasco is an exercise in denial. And when the census gets its close-view, block-by-block count finished in March, those urban and minority head-counts will give Bush a chance to address the problems with his election in Florida — or irk those groups all over again.

Although census workers attempt to count every man, woman and child in America every ten years, a perfect count is unattainable, especially in the poor and urban areas that are traditionally Democratic. People get missed. And statistical sampling is how the Census Bureau makes up for it.

A model to avoid?

In addition to the nationwide head count covering the nation's 11 million "blocks," workers take another survey of 11,800 blocks, selected at random. When comparisons between the two counts of that sample turn up discrepancies, the bureau tallies them up, produces a statistical model, and applies that model to the rest of the country, theoretically producing a more accurate total count.

"Statistical adjustments" have been a bone of contention for Democrats and Republicans ever since 1990. The apportionment of House seats and electoral-college votes must by law be determined only by person-by-person counts, or so the Supreme Court has ruled. But redistricting — how and where congressional-district boundaries are redrawn every decade so that a state's representatives have equal numbers of constituents — can be determined using statistical adjustments. Or not.

And it's up to the president to decide. Republicans fight sampling on philosophical and ideological grounds, but mostly because they think it costs them political clout and federal money. Democrats have no problem with using a little science if it bumps up accuracy by accounting for people the census workers couldn't find — again, mostly heavily Democratic minorities in big cities who don't answer when the government comes knocking at their door.

Will Bush say no to sampling?

Starting to sound a little like Florida? Bush taking the hard Republican line on sampling would set off all the same vote-suppression firestorms and make Bush's picks of Colin, Condi and Rod Paige look like window dressing; giving the nod to sampling — for the first time in history — would be taken by GOPers as a betrayal.

Unsurprisingly, Bill Clinton has already found a way to walk this political tightrope: he left it up to the Census Bureau to decide in 2000 whether or not sampling would improve the accuracy of the final tally. With the differences expected to be marginal and the Republicans already happy with what the 2000 numbers can do for them, expect Bush to be similarly hands-off when decision time comes in February.

As Kenneth Blackwell, the Republican Ohio secretary of state who co-chaired the Census Monitoring Board, put it: "Is the juice worth the squeeze? No." After all, the whole thing — which is handled with much wrangling by state legislatures — is likely to be even more bitter in 2000 than it was in 1990, what with the national scales so easily tipped. It's one more partisan fight that a "uniter, not a divider" would want to stay as far away from as possible.