Why statistical sampling rocks: … or rather, why it would rock in this situation.
From the passage of the Voting Rights Act in 1965 until the early 1990’s, state officials felt pressure from the Justice Department and the courts to help blacks and other minorities elect candidates of their choice. In practice, that meant that state legislators often tried to create districts in which blacks or other minorities accounted for more than 50 percent of the voting-age population. But in a series of landmark decisions starting with Shaw v. Reno in 1993, the Supreme Court struck down districting plans on the ground that state legislators had given too much weight to race as a factor in drawing the lines.
In March, the Census Bureau will provide detailed data to the states, showing total population and voting- age population, by race and by Hispanic origin. The bureau will decide late this month whether the data should be adjusted for a potential undercount or overcount.
Mark A. Packman, a lawyer who advises many state and local governments, said they “must use computerized census data and maps to take race into account in redistricting decisions,” and may be required to create black-majority districts where voting tends to follow racial lines. But, Mr. Packman said, “The very actions that state and local governments take to avoid liability under the Voting Rights Act expose them to liability under recent Supreme Court decisions.”
In Texas, State Senator Jeff Wentworth, a Republican who is chairman of the Senate Redistricting Committee, summarized the situation this way: “Race can be a factor, but it cannot be the dominant factor in drawing boundary lines. We still have to protect minority districts, and we will. But you can’t have very unusually shaped districts, as we did last time.”