My Two Census

Formerly the non-partisan watchdog of the 2010 US Census, and currently an opinion blog that covers all things political, media, foreign policy, globalization, and culture…but sometimes returning to its census/demographics roots.

Posts Tagged ‘Myanmar’

Language problems with the 2010 Census may lead to changes in reapportionment

Monday, May 17th, 2010

The following article from the Bellingham Herald is very well written and paints a vivid picture of the problems I have discussed about poor translation services and more:

By DANIEL C. VOCK – Stateline.org

WASHINGTON Upstate New York took in nearly 3,200 refugees during one recent year. That was nearly seven times as many as New York City did. The refugees, more than half of whom came from Myanmar, often need medical care and other social services, but the region does not have the same informational resources – such as translators and English-language classes – as New York City. To help them get those services, upstate hospital officials and other advocates want them recorded in the 2010 census and have helped spread the word to refugees.

It’s not an easy job, but it’s a potentially important one. The refugees from Myanmar who live in the county that includes Rochester, N.Y., speak six different dialects, making the task of finding a translator who understands medical terms even more difficult. When refugees do visit a doctor or the hospital in the Rochester General Medical Group, says Jim Sutton, who heads the group’s office of community medicine, their appointments last longer because of the language barrier and complications related to the fact that refugees often went years without any health care.

An accurate population count could highlight that need to government officials, Sutton says. “Politicians want to represent their constituencies. We have 8,000 refugees in our area. … If a representative saw that much of their population was voting members of their particular area, their ears may perk up a little bit when something comes before them regarding language.”

Quantcast

Quantcast


This is the kind of small but ultimately significant problem state and local officials are wrestling with all over the country.

Minnesota state demographer Tom Gillaspy knows how important the census count is for his state. He’s done the math himself. The once-a-decade tally is used for many things, but one of the most important is deciding how many seats each state gets in the U.S. House. According to Gillaspy’s latest projections, Minnesota could lose a seat by fewer than 1,000 people.

“It doesn’t get much closer than that,” muses Gillaspy, now involved in his fourth census for Minnesota. Miss just two college dorms – say, by counting them in June instead of April – and there goes the state’s eighth congressional seat.

“It is a huge operation to do a census. It is just an enormous, enormous thing. I don’t think people appreciate the precision which is required,” Gillaspy says. “It’s really at the core of everything that’s done in government and, to a large extent, in the private sector for an entire decade. So it better get done right.”

To the surprise of many, quite a few things are going pretty well this time. Across the country, 72 percent of residents have mailed in their census forms already. That’s roughly the same percentage that turned in their forms in 2000, which ended a three-decade slide in participation. That’s a good sign, according to experts, because the mail-in participation rate is a good indicator of how accurate the final count will be.

Experts credit several changes over the past decade for making it easier to educate residents about the census.

Perhaps most striking is the publicity blitz that promoted the mail-in portion of the census and continues now that 635,000 workers are going door-to-door to check with people who didn’t return their forms. The first big splash in the campaign was a much-maligned Super Bowl ad, but it was only the beginning. By the time the campaign is over, the U.S. Census Bureau plans to spend a record $133 million on advertising in 28 languages.

Behind the scenes, the federal government placed a greater emphasis on partnering with local organizations to get the message out. State and local governments have used a similar approach. Stacey Cumberbach, the head of New York City’s 2010 census office, says working with trusted leaders in different communities and across city government has helped the city boost its mail-in rates from 57 percent a decade ago to 60 percent this year.

Working with the city’s agency for public and subsidized housing helped get the message to one out of 12 New Yorkers, she says. Immigrants make up more than one-third of the city’s population, but that population in itself is very diverse. That’s why, Cumberbach says, it was so important for the city to rely on community leaders to promote the census.

In Minnesota, Gillaspy took advantage of a few other opportunities offered for the first time by the Census Bureau. In February, the state compared the numbers of addresses it had on its list for every block against the census’ count. Where there were big differences, the state asked the Census Bureau to double check its list of addresses.

Later this summer, Minnesota officials plan to compare state data for the capacity of group quarters – including prisons, nursing homes, halfway homes and dormitories – against the population count the census came up with in those facilities. If there’s a large difference, the Census Bureau will go back to recount the population there.

“It’s up to each individual state to volunteer to do this,” Gillaspy says. “I’m not aware that all states are doing this, but we certainly are.”

Gillaspy says Minnesota’s efforts during this cycle are more involved than they were a decade ago and far exceed the state outreach during the 1980 and 1990 headcounts. The Legislature approved funding for a three-year effort, and it can pay for itself by successfully counting even a relatively small number of people, he says.

Still, Kim Brace, the head of the consulting firm Election Data Services, is worried that some states have cut back on their outreach efforts to save money during this recession. He predicts, for example, that California will suffer because it couldn’t afford to better promote the census.

On the other hand, Brace says, technology has improved the amount of interim census data available to the public during the count.

“Ten years ago, we were lucky to have just to have an overall county-level count of the response rate at this time,” he says. “Now we’ve got it at the (census) tract level. That’s phenomenal.” Practically speaking, Brace says, that lets elected officials or community leaders check with the Census Bureau’s online maps to determine which areas are falling behind and respond immediately.

People who didn’t turn in their forms are less likely to answer the door when a Census worker comes knocking, explains New York City’s Cumberbach. And even if they do talk, she says, they may not provide accurate information.

In New York City, six people may share a one-bedroom apartment. Or a family of immigrants may include some people who are in the country legally and some who are not. “It’s almost like everyone has something in their home that they don’t want to share or that they’re nervous about,” Cumberbach says.

Neighborhoods with the lowest mail-in participation rates tend to have more blacks and more Hispanics than areas that turned in a bigger share of their forms, according to an analysis by the City University of New York. The 5 percent of neighborhoods with the lowest response rates were, on average, 54-percent minority. The rest of the country as a whole is 30-percent minority.

When it comes to states, many of those most in jeopardy of losing U.S. House seats – a number of them clustered around the Great Lakes – had some of the best response rates in the country. Wisconsin, Minnesota, Iowa, Indiana, Nebraska, Ohio, Pennsylvania and Virginia topped the charts.

This is especially important because the housing crisis has slowed the population growth of many Sun Belt states, and because many of those states also have below-average census response rates. Arizona, Texas, Nevada and Georgia all were expected to gain seats, but each had 70 percent or lower mail-in participation rates.

An inaccurate headcount can cost communities more than just political clout. A study by a census oversight board following the 2000 count said the country’s 58 largest counties would lose out on a combined $3.6 billion over the decade in funds distributed by population formula, more than $2,900 per person.

“Every person missed,” says Arturo Vargas, executive director of the National Association of Latino Elected and Appointed Officials, “is that much less federal resources for everything from schools and medical services to resources to pave the streets.”

MyTwoCensus Investigation: Conclusive Evidence That Burmese Translations For 2010 Census Are Wrong!

Monday, March 8th, 2010

UPDATE: The Census Bureau conducted business in early 2009 with an outside consulting firm to evaluate the accuracy of 2010 Census forms in four languages (Chinese, Russian, Korean, and Vietnamese). Additionally, here is some further evidence of problems from an external report (available in full HERE):

Errors were discovered in the Vietnamese-language materials, including the sample
Vietnamese Census Form.  The Bureau had been inconsistent in their word choice for “census,”
using both “điều tra” and “thống kê” interchangeably.  For the Vietnamese community, “điều
tra” or “government investigation” carries a negative connotation because it is associated with
the communist regime.  While the Bureau recently fixed the online form, it is uncertain whether
the corrections will appear in the printed census forms.

In February, after being tipped off about translation errors on the Census Bureau’s foreign language forms, MyTwoCensus set out to conduct an investigation into Diplomatic Language Services, the firm that was contracted to conduct all translations for the 2010 Census. Our Freedom Of Information Act request has not yet been answered, so we started to contact leading foreign language scholars to translate forms for us and judge the quality of translations.

One minority group that will suffer terribly because of poor translations is America’s Burmese community. Though there isn’t much reliable data on the Burmese-American community, a cursory read of the group’s Wikipedia entry reveals that “According to the 2000 U.S. Census, 16,720 persons of Burmese descent resided in the United States. That number is estimated to have risen to at least 50,000 today because of the large number of Burmese people seeking political asylum.”

Regarding the Burmese translation (available HERE), Julian Wheatley, who serves as the President of the Burma Studies Foundation and works in the Department of Foreign Languages & Literature at MIT told us,  “There are some usage problems, which probably arose because the original translator stuck too close to the English. More obvious, one paragraph has been repeated. Towards the end, well into the second page, you’ll see the phrase (2010 Census) in parentheses. Above it is a small three line paragraph, and the two longer paragraphs above that — you’ll see them — are identical (one in bold, one normal). Presumably that is not as intended.”

This investigation is ongoing. If you or anyone you know has noticed poor language translations on a 2010 Census form, we encourage you to contact us with specific information.