Starting Points: national census

Showing posts with label national census. Show all posts

Mar 3, 2010

A Write-In Campaign

Many Arab-Americans check 'white' on the U.S. Census. Why community leaders want to change that.

By Roqaya Ashmawey | Newsweek Web Exclusive

Mar 1, 2010

Ten years ago, when Sarah Kazem's dad filled out the U.S. Census form for the household, he racially identified his family of Egyptian descent as "white" when he answered the question on race. But this time around, Kazem, a 22-year-old Michigan resident, is going to make sure her dad marks "Some Other Race" and write in "Arab" instead.

"Why are we marking white when we're Arab?" she asks. But that is how the Census counts Arabs. After 20th-century Syrian immigrants won citizenship as "whites," Uncle Sam applied the label to all Arabs. To change that, a California-based group of Arab-American leaders formed the Arab Complete Count Committee and launched a campaign dubbed "Check It Right, You Ain’t White." The campaign—which has been circulated nationally through the Web—is an attempt to get people of Arab origin to write in their true ancestry.

"We're like an anomalous minority," says Omar Masry, 30, co-chair of the Orange County-based committee. "It seems we get counted or magnified whenever something bad happens, but where there's an opportunity for minorities, we feel marginalized."

If the campaign is successful, experts say the official Arab-American population could swell from an estimated 1.2 million people to more than 4 million—a leap that could help the group coalesce into a distinct and formidable political block.

The problem: many Arabs, fearful of terrorism-related witch hunts, are reluctant to fill out the census form at all, much less self-identify unnecessarily as Arabs. "There's fear of profiling, especially heightened after 9/11," says Masry, who is an Irvine, Calif., city planner of Lebanese and Saudi descent. "So what we try to do is educate Arabs that, by law, the information they provide can't be shared." (The Census Act prohibits the disclosure of individually identifiable information.)

"The laws that govern our confidentiality are very strict," says James Christy, regional director of the U.S. Census Bureau in Los Angeles. "We don't share information with any other agency; it's not allowed, and it's not done."

After being contacted by community leaders with concerns that Arab-Americans are fearful of filling out census forms, the bureau hired Arab-speaking specialists, such as Manar Fakhoury from Claremont, to work directly with the population, Christy says. "Throughout the years, Arabs have become distrustful of the government," Fakhoury said. "We understand they've been hurt in the past, but we reiterate that the census is safe."

Masry contends that an inaccurate count negatively affects Arab-Americans' political influence. "When policymakers and pundits in the political sphere see these artificially low numbers, it's easier for them to cast us as this minority-fringe element."

"We want to assimilate to the point where we are accepted and heard, but still want to hold onto our identities," says Ahmad Ullah, 25, a financial analyst who is primarily of Pakistani descent but also part Arab. "We want our political leaders to include us in dialogue concerning foreign policy in the Middle East."

Helen Samhan, executive director of the Arab American Institute Foundation in New York, says some Arab-Americans do not feel the need to create a separate identity. Being classified as white, she says, "is a reflection of acceptance into the majority culture, one that was not so easily offered to the pioneer immigrants a century ago."

But for other Arab-Americans, the racial categories designated by the Office of Management and Budget—which handles the classification standards of federal data on race and ethnicity—can be confusing or irrelevant."Those who have recently immigrated or who have come of age in the distinctly diversity-conscious America of the past several decades often relate more with American minorities and people of color," Samhan says. "The sting of racial profiling, discrimination, and cultural intolerance some have felt, especially since 9/11, has only added to the feeling of being distinct from the white majority."

While reluctant to identify as such on government forms, many Arab-Americans still strongly identify with their origins. "You don't stop referring to your culture just because you were born here or have been here for a long time," Kazem says. "Ethnicity and culture are strong in our families, so we'll always be American as well as identify with our mother cultures and pass them along to our children."

These issues have also become part of Arab-American comedy. "We're an underrepresented group," says comedian Dean Obeidallah, who is half Palestinian, half Italian.

"A lot of Arabs do not respond to the census, and they don't realize that it gives us more influence politically to have people know how many of us there are in this country." He raised the issue along with colleagues Aron Kader and Maysoon Zayid on their recent "Arabs Gone Wild" comedy tour, which had stops in Los Angeles and Anaheim, Calif., a city dubbed Little Arabia because of its high concentration of Arabs.

And if Arabs are still fearful of identifying their ethnicity, the activists are deploying humor too: "We [joke that Arabs] are already being tracked," Masry says. "Or you remind people that whether you're Arab or whether you're any other group, you've already lost your privacy or you've given it away by posting your life on Facebook."

Ashmawey is deputy editor of InFocus News, a Muslim newspaper in Los Angeles, and a part-time temp for the Census Bureau.

Jan 7, 2010

'Negro' Race Choice On Census Form Sparks Outrage - wcbstv.com

Question #9 Causes Uproar, Asks For Citizens To Pick 'Black,' 'African American' Or 'Negro' From Same Box

A fiery blast from the past is conjuring controversy in the new millennium. The word "negro" is now featured on an official U.S. document and now many are questioning if the Census Bureau is being insensitive.

Image via Wikipedia

It's a word that many African Americans associate with segregation, so imagine how shocked many were to see it on the 2010 U.S. census form.

"The fact that it's 2010 and they're still putting 'negro,' I am a little offended," said Secaucus resident Dawud Ingram.

Question #9 on the this year's census asks about your race. One of the boxes you can choose is "black," "African American," or "negro," all placed next to the same box. Ingram said it's not a word he uses to identify neither himself nor anybody else.

"African Americans haven't been going by the term 'negro' for decades now. It's really confusing," he said.

But census officials disagree, saying they found some older African Americans identify themselves that way and they're trying to be inclusive. In a statement, they said: "Results from the census in 2000 showed that a number of respondents provided a write-in response of 'negro' when answering the question on race."

In fact, Congress approved the form more than a year ago. Newark resident Jabbar Ali can't believe it.

"I thought it was something we left behind a long time ago – the word 'negro,'" said Ali.

Chanou Wilshire said the census form doesn't give her an option since it's got "African American," "black," and "negro" next to the same box.

"It's highly offensive," she told CBS 2.

But not everyone is offended.

"How you define yourself I guess is subjective. But for me, that on a form doesn't offend me at all," said Brooklyn resident Tiffany Campbell.

Others don't understand why the question of race has to come up on any form.

"I'm an American. What's wrong with just being an American?" asked Newark resident Derri Gowns.

Census bureau officials said they're preparing for the 2020 form, asking folks now in a questionnaire whether the word 'negro' should be removed.

Dec 26, 2009

Census Goes Off the Rez to Count American Indians

Image via Wikipedia

LOS ANGELES-- When United American Indian Involvement, Inc., opened its doors in 1974 on Los Angeles’ Skid Row, it was a modest shelter serving downtown’s American Indian community. Today it occupies a larger facility, has centers in Fresno and Bakersfield, and provides an array of social services.

Image via Wikipedia

UAII’s growth reflects the explosive surge in American Indians moving from tribal reservations to urban centers, like Los Angeles County. According to U.S. Census data, more than 90 percent of the state’s American Indian population now lives off tribal reservations. That demographic shift requires targeted strategies by the Census Bureau in getting an accurate count of American Indians in the upcoming 2010 census count.

“We’re looking at a huge challenge, due to the federal policies of relocation of our people over the years, and then just the outright economic conditions that were present and are still present, causing people to go off the reservation for work, to make a living,” said Tim Harjo, the American Indian partnership specialist for the U.S. Census Bureau in Los Angeles. “And then they get to these places [like Los Angeles], and it’s not easy. They find themselves in bad situations.”

Harjo’s job is to encourage participation in the 2010 Census by more than 300,000 American Indians in his region, which includes Southern California and Hawaii.

While Harjo and his colleagues have been outreaching directly to tribal governments across California, he says the tougher test in 2010 will be to count the American Indian majority who are increasingly leaving the reservations and moving into the urban areas. They are most likely not members of local or regional tribes but in fact may have migrated from elsewhere in the state or country.

According to 2000 Census data, more than 90 percent of California’s American Indian population now lives off tribal reservations. In fact, the bureau’s most recent population estimates show there are now more than 155,000 American Indian and Alaska Natives residing in Los Angeles county alone, making it the most populous county in the nation for that racial category.

Placed in historical context, the numbers are extraordinary.

When the federal government first recognized American Indians on the 1860 census, their official population nationwide stood at roughly 44,000. By 1950, they numbered 343,000, and in the 2000 census, the American Indian population was 4 million. It is a trend that the bureau expects to continue. By 2050, the nation’s native population could surpass 8.6 million, or 2 percent of the U.S. population.

“To some people those numbers may not seem like a lot, but it is huge for us,” said [explained] Harjo.

Which is why community organizations like UAII are paying close attention to the 2010 Census and becoming active participants in spreading the message of being counted to community members.

“We’ve already sent out fliers [encouraging census participation] to thousands of names on our mailing list,” said Jerimy Billy, associate director of UAII. “We took it upon ourselves to promote it using our own funds because we want to see an accurate count.”

Many UAII programs are funded through government contracts. And with an annual $400 billion in federal allocations based on 2010 Census numbers, UAII and other organizations have a vested interest in making sure their communities get a fair share of that funding.

“The census does play a part, because it allows us to have a number to work with,” said Billy. “Our funders want to get the most bang for their buck. They want to give where they think we can serve the most. And without that concrete data, it’s really hard to convince them we have a large population.”

Changing public perceptions of where American Indians reside, explained Billy, will be helpful for both funders and tribal governments alike in the years to come.

“Sometimes what happens is, people from the reservation come through LA, and they may need to access our health services, but the people from those other tribes or reservations may not really be aware of what we do,” he said. “They might say, ‘Well, urban organizations are serving non-Indians.’ No. For example, there is a huge Lakota population living in LA. Or we can go back to the Navajo tribe and say, ‘We’ve got a huge population here that we’re trying to serve.’”

Specifying tribal affiliation on the 2010 Census questionnaire will have huge ramifications for American Indian communities because there is a direct correlation between how many people identify with a particular tribe and how much money that tribe receives from the government for services.

Harjo points to tribal numbers from the 2000 census as proof that American Indians lost out on funding for their communities 10 years ago.

Of the 4.1 million American Indian and Alaskan Natives identified on the 2000 census, the tribal affiliations of over 1 million were “not specified.” While lack of education on how to fill out the form is part of the problem, Harjo says there are also structural problems with the census questionnaire that led to the missing information.

“Question number 9 [on the census questionnaire] is the race question, and that’s the biggest issue,” he said. “There are only 19 boxes to write in the name of your tribe. But many tribe names require more than 19 boxes. And then there are people who write-in their tribe, but it’s a traditional name that the data processors don’t recognize, so that doesn’t get counted. That’s a lot of money that was lost.”

In California, there are over 100 federally recognized tribal groups, as well as dozens of smaller tribes that are still unrecognized by the government.

Further complicating the census outreach efforts of organizations like UAII and census officials like Harjo, are deep-seated – and well-founded - feelings of distrust toward government.

“Confidentiality is always a big concern,” said Harjo, “but perhaps more so, there are quite simply a lot of people who are saying, ‘it doesn’t matter to me.’”

Between now and National Census Day on April 1, Harjo and his allies in the Los Angeles region will be putting their all into making sure the growing community of urban American Indians understand that it does matter.

“We don’t have the resources to get out and be everywhere,” said Billy. “Our staff is still not big enough to meet everybody’s needs. At least to capture that, and show there is a need… I think that’s the most important thing.”

Oct 14, 2009

Census 2010 to include outreach in more languages - San Jose Mercury News

Image via Wikipedia

By JULIANA BARBASSA Associated Press Writer

Posted: 10/14/2009 03:46:23 PM PDT

Updated: 10/14/2009 03:53:20 PM PDT

SAN FRANCISCO—With the launch of the 2010 census less than six months away, the U.S. Census Bureau is preparing its most targeted and far-reaching language outreach yet to more accurately count the country's immigrants.

Census Director Robert Groves briefed San Francisco elected officials and community groups Wednesday on ways the federal agency is working to connect with residents who speak languages other than English.

"We're come to realize that a census run out of Washington, D.C., on a one-size-fits-all model doesn't work very well," Groves said. "We're working to customize the experience, to use words, languages that make sense and make it easier for people to participate."

The campaign, expected to cost about $300 million, is designed to remedy that, Groves said.

Part of the outreach will occur through Census Bureau alliances with institutions such as churches and ethnic organizations to inform communities and overcome fears of working with the federal government.

A media campaign, including advertising in ethnic newspapers, will also help get the word out, Groves said.

Along with a 10-question form in English, postcards in five other languages—Vietnamese, Korean, Chinese, Russian and Spanish—will point users to a Web site in their language.

Assistance guides in scores of languages, designed to walk a respondent through the English forms, will be distributed through places such as barber shops and supermarkets,

or delivered door-to-door, Groves said. Over-the-phone help will be offered in up to 59 languages.

Bilingual forms in English and Spanish will be mailed to 13.5 million households in neighborhoods where at least one-fifth of households report speaking primarily Spanish.

Some advocacy groups wanted the Census Bureau to ask immigration authorities to stop workplace enforcement raids during the nationwide count, so immigrants feel less fearful of the government.

One federal agency cannot ask another to stop doing its work, Groves said, but the confidentiality of respondents will be protected.

"We assure them their participation will not lead to any harm, in any way," Groves said. "We take this oath very seriously. This is a very safe thing to do."

San Francisco is investing more than $800,000 in community outreach because undercounting comes at significant cost to the city.

In 2008, the city challenged the 2007 census estimate. New figures released by the U.S. Census Bureau recognized an additional 34,209 residents, giving San Francisco a total population of 799,185—and additional federal funds.

"Given how much federal funding is at stake, we have to invest in this," said David Chiu, a San Francisco supervisor.

On the state level, a PricewaterhouseCoopers census study estimated that more than half a million Californians were not counted in the last census, which meant a loss for the state of more than $1.5 billion in federal dollars.

Sep 22, 2009

Number of Foreign-Born U.S. Residents Drops - washingtonpost.com

Image by mikebaird via Flickr

Construction, Manufacturing Job Cuts and Enforcement Cited in Loss of Hispanic Immigrants

By Carol Morello and Dan Keating
Washington Post Staff Writers
Tuesday, September 22, 2009

The number of foreign-born people living in the United States declined last year, particularly among low-skilled immigrants from Mexico, according to a Census Bureau report released Tuesday.

The immigrant losses were particularly pronounced in California, Florida, Arizona and Michigan, all states where the recession hit early and hard. The metropolitan Washington area gained about 1,000 foreign-born residents, but a jump in the Asian population was offset by a significant drop in Mexicans and Salvadorans, the largest Hispanic immigrant group in the region.

The nationwide total of about 38 million foreign-born people decreased slightly, by just under 100,000. That brought down the share of the overall population that is foreign-born from 12.6 percent to 12.5 percent. Although the drop is relatively small, it was the first official decline in at least four years.

Demographers and other analysts said immigration is bound to pick up once the economy improves, although some said stricter enforcement of immigration laws played a role in the decline.

"This is clearly a downturn related to the economy in the U.S.," said demographer William Frey with the Brookings Institution. "What looks like negative immigration is something that, two or three years ago, you wouldn't have expected at all. It shows immigrants respond to the economy."

The statistics were part of the American Community Survey, an annual Census Bureau report that also includes data on household incomes and health insurance. The survey, conducted year-round, is based on a sample of about 3 million addresses.

The Washington area remains among the wealthiest places in the country. The median household income of $85,824 last year -- up from $83,200 in 2007 -- is second only to San Jose. Blacks, whites and Hispanics in the region are all on average the highest earners in the nation, while Asians here are the third highest, behind their counterparts in San Jose and Raleigh, N.C.

For the first time, the survey measured how many people do not have health insurance. Nationally, 15 percent are uninsured, but the figure varied widely among states. Texas had the most uninsured, at 24 percent, and Massachusetts had the least, at 4 percent.

The Washington area reflected big disparities among adults ages 18 to 64. In Prince George's County, for instance, 20 percent of adults in that age group have no insurance, compared with 10 percent in the District and 9 percent in Arlington County. About 16 percent of adults in Virginia have no insurance, as do 15 percent in Maryland.

The new statistics on foreign-born residents confirm findings by other researchers showing an ongoing drop-off in immigrants from Mexico, who comprise a third of all foreign-born residents and two-thirds of all Hispanic immigrants.

The Census found about 325,000 fewer immigrants from Mexico last year, a fall-off of 2.8 percent. Without that decline, there would have been a small increase in the overall number of immigrants.

Latinos decreased in all regions except the Northeast, where the population stayed flat. In the Washington area, the Mexico-born population dropped about 9,600, a net loss of 19 percent. Salvadorans were down about 10,700, a 7.4 percent drop. Together, they negated the addition of 16,500 Asians, a 4.4 percent increase.

A study this summer by the Pew Hispanic Center concluded that since 2006, there has been a sharp decline in new immigrants from Mexico, while the number who return home every year has stayed about the same.

Many Hispanic immigrants work in construction and manufacturing, and they have been particularly affected by the economic downturn, said Mark Lopez, associate director of the Pew Hispanic Center.

Some might have been deterred by stricter immigration enforcement, too, he noted. In a 2008 Pew survey, 10 percent of foreign-born Hispanics said they had been stopped by authorities and asked for their immigration status, as had 8 percent of Hispanics born in the United States.

"Many Hispanics worry that they themselves, or someone they know, may be deported," Lopez said.

But Steven Camarota of the Center for Immigration Studies, which favors fewer immigrants, said his research suggests almost 1 million immigrants left the country in the past year alone, most of them Hispanic.

"People continue to come, but significantly fewer are coming, and many more are going home," he said. "It appears the decline began before the economy went south. That strongly suggests increased enforcement played a significant role."

Others contend that the floundering economy is solely to blame and that the drop is temporary.

"We've had a lot of enforcement in play for years," said Michael Cassidy, head of the Commonwealth Institute for Fiscal Analysis, a think tank that analyzes the impact of policy on low-income Virginians. "I think that points to the economic reasons behind the shift, as opposed to the enforcement reasons. When there are no jobs, people aren't coming and they're not staying."

Aug 10, 2009

The 2010 Census Will Violate the Constitution

By JOHN S. BAKER AND ELLIOTT STONECIPHER

Next year’s census will determine the apportionment of House members and Electoral College votes for each state. To accomplish these vital constitutional purposes, the enumeration should count only citizens and persons who are legal, permanent residents. But it won’t.

Instead, the U.S. Census Bureau is set to count all persons physically present in the country—including large numbers who are here illegally. The result will unconstitutionally increase the number of representatives in some states and deprive some other states of their rightful political representation. Citizens of “loser” states should be outraged. Yet few are even aware of what’s going on.

In 1790, the first Census Act provided that the enumeration of that year would count “inhabitants” and “distinguish” various subgroups by age, sex, status as free persons, etc. Inhabitant was a term with a well-defined meaning that encompassed, as the Oxford English Dictionary expressed it, one who “is a bona fide member of a State, subject to all the requisitions of its laws, and entitled to all the privileges which they confer.”

Thus early census questionnaires generally asked a question that got at the issue of citizenship or permanent resident status, e.g., “what state or foreign country were you born in?” or whether an individual who said he was foreign-born was naturalized. Over the years, however, Congress and the Census Bureau have added inquiries that have little or nothing to do with census’s constitutional purpose.

By 1980 there were two census forms. The shorter form went to every person physically present in the country and was used to establish congressional apportionment. It had no question pertaining to an individual’s citizenship or legal status as a resident. The longer form gathered various kinds of socioeconomic information including citizenship status, but it went only to a sample of U.S. households. That pattern was repeated for the 1990 and 2000 censuses.

The 2010 census will use only the short form. The long form has been replaced by the Census Bureau’s ongoing American Community Survey. Dr. Elizabeth Grieco, chief of the Census Bureau’s Immigration Statistics Staff, told us in a recent interview that the 2010 census short form does not ask about citizenship because “Congress has not asked us to do that.”

Because the census (since at least 1980) has not distinguished citizens and permanent, legal residents from individuals here illegally, the basis for apportionment of House seats has been skewed. According to the Census Bureau’s latest American Community Survey data (2007), states with a significant net gain in population by inclusion of noncitizens include Arizona, California, Florida, Illinois, Nevada, New Jersey, New York and Texas. (There are tiny net gains for Hawaii and Massachusetts.)

This makes a real difference. Here’s why:

According to the latest American Community Survey, California has 5,622,422 noncitizens in its population of 36,264,467. Based on our round-number projection of a decade-end population in that state of 37,000,000 (including 5,750,000 noncitizens), California would have 57 members in the newly reapportioned U.S. House of Representatives.

However, with noncitizens not included for purposes of reapportionment, California would have 48 House seats (based on an estimated 308 million total population in 2010 with 283 million citizens, or 650,000 citizens per House seat). Using a similar projection, Texas would have 38 House members with noncitizens included. With only citizens counted, it would be entitled to 34 members.

Of course, other states lose out when noncitizens are counted for reapportionment. According to projections of the 2010 Census by Election Data Services, states certain to lose one seat in the 2010 reapportionment are Iowa, Louisiana, Massachusetts, Michigan, New Jersey, New York, Ohio and Pennsylvania; states likely (though not certain) to lose a seat are Illinois, Minnesota, Missouri, and Ohio could lose a second seat. But under a proper census enumeration that excluded illegal residents, some of the states projected to lose a representative—including our own state of Louisiana—would not do so.

The census has drifted far from its constitutional roots, and the 2010 enumeration will result in a malapportionment of Congress.

In the 1964 case of Wesberry v. Sanders, the Supreme Court said, “The House of Representatives, the [Constitutional] Convention agreed, was to represent the people as individuals and on a basis of complete equality for each voter.” It ruled that Georgia had violated the equal-vote principle because House districts within the state did not contain roughly the same number of voting citizens. Justice Hugo Black wrote in his majority opinion that “one man’s vote in a congressional election is to be worth as much as another’s.” The same principle is being violated now on a national basis because of our faulty census.

The Census Bureau can of course collect whatever data Congress authorizes. But Congress must not permit the bureau to unconstitutionally redefine who are “We the People of the United States.”

—Mr. Baker teaches constitutional law at Louisiana State University. Mr. Stonecipher is a Louisiana pollster and demographic analyst.

Jul 26, 2009

Now It’s a Census That Could Rip Iraq Apart

By ROD NORDLAND

BAGHDAD — When Iraqis were drafting their Constitution in 2005, the parties could not agree on who would control Kirkuk, the prized oil capital of the north. They couldn’t even agree on who lived in Kirkuk, which is claimed by the region’s Kurds, but also by its Turkmen minority and Sunni Arabs. For that matter, they couldn’t even agree on where Kirkuk was — in Tamim, Erbil, or Sulaimaniya Province.

So the Iraqis punted, inserting Article 140, a clause that called for a national census, followed by a referendum on the status of Kirkuk, all to be held by the end of 2007. What followed were a succession of delays, against a backdrop of sectarian violence and warnings that Kirkuk could blow apart the Shiite-Kurdish alliance that has governed Iraq since the Americans invaded.

Massoud Barzani, president of the Kurdish regional government, warned two years ago that if “Article 140 is not implemented, then there will be a real civil war.” He’s still waiting.

But so is the threat of civil war, which lurked quietly in the polling places this weekend as residents of Iraq’s Kurdish-dominated areas voted for their regional president and Parliament. Until the status of Kirkuk is clear, nobody really knows how much power those regional officials can wield within the national government, or even whether the Kurds will want to remain part of Iraq.

The problem with settling that is the Kirkuk referendum. There can’t be a referendum until Iraqis figure out who is eligible to vote in Kirkuk, which they can’t do until there’s a census. And any attempt to hold a census in this country may well end up, all by itself, provoking a civil war.

Even now, Sunnis don’t agree that they’re a minority of the nation, and that the Shiites are the majority, though it’s patently obvious. And in Kirkuk, everyone is in denial, one way or another.

Ethnically mixed and awash in oil, Kirkuk has always been something of a numbers game. There are 10 billion barrels of proven oil reserves — 6 percent of the world’s total and 40 percent of Iraq’s — all within commuting distance of downtown Kirkuk. Its fields, though half destroyed, still produce a million barrels of oil a day.

Both Turkmen and Kurds claim to be in the majority; the last reliable estimates, from a 1957 census, gave Turkmen a plurality in the city and Kurds a plurality in the surrounding district, with Arabs second in the countryside and third in the city. In the Saddam Hussein years, the Kurds declared Kirkuk part of their autonomous region of Kurdistan, but the dictator sent the army after the Kurdish guerrillas, known as pesh merga, and held onto the prize. He then set about Arabizing it, forcibly relocating families from the south while evicting Kurds and Turkmen alike.

After 2003, pesh merga troops quickly took control of Kirkuk as the Iraqi Army collapsed. Some local Arabs revolted, nurturing an insurgency that still festers. Others simply remained. Meanwhile, Turkmen appealed to powerful patrons in Turkey that they were undercounted and ignored by everyone, and Turkey came to their aid to make sure the Kurds didn’t get Kirkuk, which supplies much of Turkey’s oil. Only the presence of American troops has kept a lid on things; a brigade is still kept in Kirkuk.

And still there is no census. “The Iraqi government for the last three years, every year they say it will come this year,” says Mahmoud Othman, a Kurdish member of Parliament.

A date for a census is on the calendar — Oct. 24. But it is subject to ratification by Iraq’s cabinet, the Turkmen have announced that they will boycott it and Arabs in Kirkuk may well do the same.

One proposal for getting past this problem would be to hold a census everywhere but in Kirkuk. If that happened Kirkuk could end up, in effect, a disenfranchised province when the next general national elections are held in January.

Another suggestion is to hold a referendum on Kirkuk without a census, but that would invite a dispute about the validity of the results.

And then there’s the Lebanese solution, the one that so far seems likeliest: just do nothing. The last census in that sectarian hodge-podge of a country was in 1932; no one would dare hold one now, since the groups who would almost certainly lose representation — Maronite Catholics, Druze and Sunni Muslims — would simply go back to war rather than get counted out.

Already, the Kurdish regional government has been defying Baghdad and issuing contracts to develop its oil fields, including some in Kirkuk. The Iraqi government showed its displeasure by moving its 12th Division, some 9,500 troops, up to Kirkuk; there they have been provocatively patrolling into pesh merga-held areas and setting off a series of minor incidents recently.

“It’s very worrisome that these incidents continue to happen,” said Joost Hilterman, of the International Crisis Group. “Perhaps they will be contained, but the stakes are huge.”

For the moment, there are still plenty of American troops around to do the containing, but all American combat troops are due to pull out by next summer. That doesn’t leave a lot of time to broker an agreement, especially when no one is likely to really want it.

Abeer Mohammed contributed reporting.