Print This Post Print This Post | Email This Post

Panic in Detroit: The Motor City and flaws in the U.S. Census

According to the most recent U.S. Census, Detroit lost a staggering 25 percent of its population between 2000 and 2010, dropping from 951,000 residents to 713,000. Detroit Mayor Dave Bing responded by saying Detroit will challenge the results. “Personally, I don’t believe that the number is accurate,” Bing said when the figures were released in March. “The Census has a history of undercounting residents in urban areas like Detroit.”

Detroit can begin checking the Census Bureau’s results on June 1st.  The effort is not a recount, but, according to former Detroit demographer Patricia Becker, “an extremely technical and painstaking process” done by city (not federal) workers to find uncounted residents. Moreover, no challenge can reverse the new reality Detroit faces:  rather than a major city in decline, Detroit is no longer a major city.

Detroit’s Mayor Bing is correct that the Census undercounts urban areas, especially those with large minority populations — and especially those as quickly changing as Detroit. Undercounts mean reduced public assistance, and Detroit  as well as the state of Michigan must respond to that.  But the population shift in Detroit shows a broader failure of state and federal authorities to react to Detroit’s gradual disappearance from the national map.

Why the population number matters

Except for post-Hurricane Katrina New Orleans, Detroit lost more of its population than any other large American city in the last ten years. No  American metropolis has declined faster in the last 30-40 years. In 1950, Detroit was the fourth most populous city in America with 1.8 million people. Today it has fewer residents than second-tier cities like Indianapolis and Columbus.

But population estimates made right before the census indicated a slightly less precipitous decline – for example, the Southeast Michigan Council of Governments put the Detroit population at 762,000. If, in fact, Detroit’s population has dropped below 750,000, it’s not just a a symbolic blow, it’s  an immediate practical problem. There are a number of Michigan laws that favor cities with at least 750,000 people. For example, cities with at least 750,000 can levy a higher income tax and utility tax. This tax money gave Detroit almost $300 million in extra revenue last year, especially critical for a city with a $150 million deficit.

Obviously, these “750,000 laws” were designed with Detroit in mind: the next most populous Michigan city is Grand Rapids with 188,000 people.  Democratic mayor of Detroit Dave Bing and Michigan’s Republican governor, Rick Snyder, together with the Republican-dominated state legislature, will probably struggle to find common ground but may manage to keep Detroit’s special status even with a smaller total population.  But the Census results also help determine  both how congressional districts are set up, and how much money Washington provides.

Nationally, about $450 billion, or $1,200 per person, is tied into Census figures. Population determines how much money Detroit will get for bedrock social service programs for the poor like Medicaid, temporary assistance to needy families, and food stamps. Money for highway planning and construction is also set by census figures as well as some education funding, such as money for college Pell Grants.  “The system is set up where you would almost be crazy not to put up a challenge because so many federal funding formulas are tied into population,” says Cynthia Taeuber, currently a private consultant in the Philadelphia area who worked at the Census Bureau for 30 years. “It makes sense to fight even if you have no particular basis for doing so.”

Not everyone counts

Given past census experience around the U.S., it’s a fair bet that Detroit’s population was undercounted, according to demographers and census experts “Historically, people who are non-white are undercounted,” says Andrew Reamer, a public policy at George Washington university. “It is a socioeconomic issue,” says Patricia Becker, Detroit’s former chief demographer who now runs a private consulting firm. “All big and poor cities are traditionally undercounted.” Socioeconomic factors are indirectly extrapolated as the Census records race, but not income or other measures of wealth.

The first problem is residents that don’t fill out Census forms mailed to them – 64 percent of questionnaires sent to Detroit addresses are mailed back, compared to 78 percent in Michigan and 74 percent nationally. Many people do not answer the door when Census takers knock. “Detroit always has one of the lowest participation rates in the country,” says Gregory Parrish of Data Driven Detroit. “It is a minority population very reluctant to respond to government intervention.”

The other long-term problem has been that Census takers often don’t know  where to knock. “Think about multi-unit buildings in cities,” Reamer explains. “A single family house is broken into four units with one mail basket up front. The Census Bureau doesn’t have a clue of how many units are in that building.”  Reamer says that minorities are significantly more likely to live in multiunit buildings.

These problems have been around for decades but they came to a head after studies of the 1990 Census showed that one out of every twenty black and Latino residents were missed. After that, the Census Bureau created a partnership program with about 140,000 church groups and neighborhood associations across the country. This led to aggressive community outreach and advertising – like “everybody counts” billboards in many cities across the country (not so much in Detroit, but more on that below).  Another change has been Census Bureau  collaboration with cities to update urban addresses. Called the Local Update on Community Address, or LUCA process, the U.S. Census develops a national address list three years prior to the census. State and city governments then review and make additions.

These changes have made a difference. In 1990, 1 out of 20 minorities were undercounted. In 2000, it was closer to 1 in 50 minorities that the census overlooked. Also, the Census Bureau notes on its Web site that of 39,000 jurisdictions across the country in 2000, count problems were found in just 1,180. However, urban undercounts are still a problem as evidenced by the fact that not just Detroit but New York City and, most likely, Chicago will also challenge the 2010 figures.

Moreover, a successfully conducted census depends on a competent and engaged city, which Detroit government has not been of late. City Hall was in shambles thanks to the corrupt regime of ex-Mayor Kwame Kilpatrick who is now in federal prison. “In 2009, Detroit was having a special mayoral election because of the Kilpatrick mess,” Becker says. “They didn’t have money and their focus wasn’t on the census.” Also, a huge spike in unemployment and home foreclosures created an almost unbelievable 80,000 housing vacancies – or 23 percent of all housing stock.  Becker called the census vacancy number “incredible.” It is up to the city to show that some of these units are still partially occupied.

Challenging the Numbers

Yet Becker says the cost of a challenge is “too much for Detroit to afford – the city is broke.” The challenge process is complex and requires real resources.   Detroit must make a “coverage challenge,” arguing that the census erroneously excluded certain city residences and therefore missed their inhabitants.  As  Becker explains, “The city has to get its own count of every block and there’s about 13,000 blocks in the city. Then they have to match that against the census figures and look for differences and try to explain differences.” Becker is agnostic about whether Detroit has 750,000 people, but their ability to find 750,000 people is “extremely unlikely here.”

Instead of focusing on a challenge today, Detroit could have better invested city resources in 2007 during the LUCA process, and most importantly in 2010, when census workers were on the ground. “The 2010 process was completely understaffed and under-resourced,” Parrish says. “The mayor’s office claims they were very responsive and tried and yet they insist there was an undercount. How do you reconcile those two things?” Repeated attempts to get in touch with the mayor’s office were unsuccessful – an office employee said that communications director Dan Lijana, “is not really good with returning phone calls.”  Still, Bing may be vindicated if the painstaking block-by-block work of a challenge yields 750,000 residents, instead of 713,000, given the state and federal money that will come with the new number.

The greater problem, though, is that cities in disrepair like Detroit – that literally need an accounting for – are bound to have trouble getting an accurate census tally, and the very people who are likely to be undercounted are the ones who need the most help in the form of social services the government provides.

The past twenty years of the Census Bureau show a federal agency in the continual process of self-evaluation and improvement. “The Census Bureau is odd in terms of most government agencies in that we measure our errors and report them in full detail,” Taeuber, the former census worker, says. The same does not seem to be true for failing cities like Detroit.  The Motor City prepared poorly for the 2010 Census, and it’s possible that its residents were undercounted as result.  If that leads to a further decline in the city’s health, it will be hard to say that the Census Bureau is to blame.

One Response to “Panic in Detroit: The Motor City and flaws in the U.S. Census”

  1. Michael Fay:

    The problem in doing a city-wide Census let alone a national count is not to assign blame afterwards but to prepare for follow up actions by both public and quasi-private interests at city, county, state, region, and Federal levels. It has been recognized that for decades that census undercounts in particular from economic subsistance level sections that are part(s) of American urban regions exist. The current under-budget situation(s) of the City of Detroit and Michigan have been known for some time. The set up of follow up plans and and additional counts with the assistance of the U.S. Census Bureau could have been designed and implemented prior to the 2010 Census with appropriated funding resources established. The expense factor is too large of a problem for one governmental unit to carry out follow up actions in particular when the Congress and the state of Michigan Legislature have majorities that are not in sympathy with the need to appropriate money to correct urban area undercounts like Detroit with a poverty level minority population. The U.S. Census Bureau and their counterparts in the fifty states for the 2020 Census and beyond will need to encourage local, county, and regional govermental levels to cooperate in planning and budgeting economic resources [quasi-insurance pool(s)?] to pay for follow up census counts including advocacy when requesting census corrections for known undercounts. Large percentage negative population census counts due to suspect undercounts for American cities like Detroit New Orleans in future decades’ census counts shall be the rule and not the exception.

    Michael Fay,
    College Station, Texas


    comment at 31. May 2011

Leave a Comment