Five states — New York, California, Texas, Arizona and Florida — are perilously close to losing out on congressional seats because of lackluster participation in the U.S. census.
The five were average or below average in mailing back 10-question census forms when compared to other states, trailing by as many as 5 percentage points, according to the final census mail-in tally released Wednesday.
Based on recent population trends, New York, California and Texas had been estimated to fall just above the cutoff for the last House seats when they are redistributed next year. Waiting behind them in hopes of picking up additional seats are Arizona and Florida, which are already expected to gain one seat apiece.
Responses from these states also raise a red flag because of their higher shares of residents who are Latinos. The Census Bureau has said one of its main concerns is whether tensions over immigration will discourage Latinos, and particularly illegal immigrants, from participating in the government count. That issue returned to the forefront after Arizona passed a tough immigration enforcement bill.
Latino residents represent a predominant share of the population growth in New York, California, Texas, Arizona and Florida, making up more than 50 percent of total growth since 2000. As a result, those states could face big losses if there isn’t full cooperation when the Census Bureau on Saturday begins knocking on the doors of those who did not respond by mail.
Of the five states on the cusp, the biggest potential losers are California and New York, which could have a net loss of one and two House seats, respectively. Texas may end up gaining just three House seats instead of four.
Arturo Vargas, executive director of the National Association of Latino Elected and Appointed Officials, said he was concerned about some skittish Latinos who may refuse to answer their doors, particularly given Arizona’s new immigration law.
“I’m incredibly disappointed with the Obama administration in their efforts to promote the census,” Vargas said, citing the government’s failure to halt immigration raids during the count as it did in 2000. “It may have the impact of shooting people in the foot if Arizona ends up losing out on a House seat.”
States such as Minnesota and Oregon are next in line to pick up seats. Minnesota had the nation’s second-highest mail response at 80 percent — a clear boost in its effort to avert the loss of a seat, even after Rep. Michele Bachmann, R-Minn., made clear her view that the 10-question census was an invasion of privacy.
North Carolina, which snatched a seat from Utah in 2000 when overseas missionaries were excluded from the count, also remains in play to gain a seat.
“It would be a bit ironic if Minnesota ends up a winner,” said Kimball Brace, president of Election Data Services, a Virginia-based firm that crunches political numbers. “With the immigration concern, that’s going to have an impact. Both New York and California are in the position of losing seats, but they haven’t done as much as they could in spending to improve on outreach.”
On Wednesday, Census Bureau director Robert Groves attributed the strong mail participation rate of 72 percent to the bureau’s advertising and outreach campaign, which helped overcome growing public apathy toward surveys as well as distrust of the government. But he said it remained uncertain how that will translate to “how the American public reacts when we knock on their door.”
“The census is not over,” said Groves, who noted the non-respondents were disproportionately low-income, lesser-educated or renters. “For those of you who haven’t been counted in the 2010 census, this is your moment.”
The midterm report comes as the Census Bureau prepares to begin door-to-door canvassing, the most costly and error-prone portion of the count.
In all, more than 600,000 workers will fan neighborhoods at rates of $10 to $25 an hour until mid-July to query people on the 10 census questions on race, gender and family relationships. It’s part of a government hiring spree the Commerce Department says could alter the unemployment rate by several tenths of a percentage point in April and May.
At training sessions this week, temporary census workers were instructed on the protocols of conducting interviews, such as how to tabulate answers on race (let people self-identify if they’re multiracial, but a label of “American” isn’t a sufficient response), where to ask questions (outside, since census workers should not ask to enter a person’s home) and carrying proper identification (government badges and a “U.S. Census Bureau” bag).
Census workers also are being told to be respectful if homeowners refuse to cooperate, to keep data confidential and to alert supervisors if there are signs of danger. In the last 12 years, there have been 21 work-related deaths involving census employees, including a dog attack on a 71-year-old worker in 2000.