Rechecking the census

Sam Roberts writes:

The Census Bureau [reported] that though New York City’s population reached a record high of 8,175,133 in 2010, the gain of 2 percent, or 166,855 people, since 2000 fell about 200,000 short of what the bureau itself had estimated.

Public officials were incredulous that a city that lures tens of thousands of immigrants each year and where a forest of new buildings has sprouted could really have recorded such a puny increase.

How, they wondered, could Queens have grown by only one-tenth of 1 percent since 2000? How, even with a surge in foreclosures, could the number of vacant apartments have soared by nearly 60 percent in Queens and by 66 percent in Brooklyn?

That does seem a bit suspicious. So the newspaper did its own survey:

Now, a house-to-house New York Times survey of three representative square blocks where the Census Bureau said vacancies had increased and the population had declined since 2000 suggests that the city’s outrage is somewhat justified. In those blocks alone, census takers appear to have missed dozens of New Yorkers and to have overestimated the number of vacant apartments.

In Brooklyn, on a block near Ocean Parkway between Midwood and Gravesend, where the census said nearly half of the 148 homes were vacant, a resident said the only vacancies were in a new 33-unit apartment building that is partially occupied. . . .

On another block in Sheepshead Bay, Brooklyn, the number of vacancies on the block recorded by the Census Bureau far exceeds the number of unsold condominiums in a new apartment building. Superintendents of other nearby buildings say those had few vacant apartments when the census was conducted.

And on a block of two- and three-story homes in East Elmhurst, Queens, where the census recorded 26 fewer occupied apartments and 20 more vacant ones (defined as habitable but with no one living there) than a decade earlier, in 2000, a real estate agent said a one-family house had been illegally divided and had 14 residents — evidence of demand for housing.

Very impressive of the NYT to do their own survey rather than just reporting it as an amusing controversy. Roberts continues:

The magnitude of any undercount is uncertain. The Times survey did not replicate the methods the Census Bureau uses, including mailing questionnaires and making up to five visits to addresses that have not returned the forms. As a last resort, a census worker will consult with a landlord or neighbors and make a best guess about whether a home is occupied.

Often, though, owners of illegally divided houses are reluctant to disclose the number of tenants, who tend to include people who are in the country illegally and are leery of providing any information to the government. A visit from Times reporters may have proved less intimidating to landlords and residents.

City officials say as many as 80,000 residents appear to have been systematically overlooked in crowded immigrant neighborhoods like East Elmhurst and Jackson Heights in Queens and Sunset Park, Bay Ridge and Bensonhurst in Brooklyn.

Classrooms in those neighborhoods are overcrowded and “for rent” signs are rare. Some demographers say the number of vacancies was not all that anomalous, given some overbuilding before the recession and a surge in foreclosures. But of 500 houses or apartments on the three blocks surveyed by The Times, only four were in foreclosure or had been seized by the mortgage holder, according to an analysis conducted at The Times’s request by the Furman Center for Real Estate and Urban Policy at New York University.

What happens next?

Relying on earlier census surveys and evidence from the Postal Service and other sources, the city plans to formally ask the Census Bureau next month to review its findings. Census officials have acknowledged that a processing glitch is one possibility for any pattern of population declines and increased vacancies in specific neighborhoods.

3 thoughts on “Rechecking the census

  1. "A visit from Times reporters may have proved less intimidating to landlords and residents." Interesting, given that the Times reporters could share people's comments with anyone at all, while the Census Bureau is not allowed (by law) to share Census responses with other federal agencies or law enforcement… But I guess that's hard to convey to someone who starts out with an attitude of being afraid to share their info with the government. I'd be curious to read more about the psychology behind this.
    It just goes to show that statisticians can't just blindly trust the survey data we get. There are plenty of opportunities for underreporting, response errors, wrong sampling frame, etc. throughout the whole process!

  2. In the past, we've discovered that the NY Times definition of "representative" is not statistical. (remember representative wedding and engagement announcements on this blog?)

    In a large city, it is regrettable but inevitable that some local field work for the census will be shoddy. We'll have to wait for the follow-up to see how widespread the shoddiness is.

Comments are closed.