The Boston population debacle: Time for the Census Bureau to explain itself
The Boston Globe’s Michael Levenson and Yvonne Abraham have a follow up article today on yesterday’s report that Boston’s population actually increased by some 7,000 people between 2000 and 2005, rather than decreased by 30,000, as originally estimated by the United States Census Bureau
Here’s how the article explains how the revision came about:
After the bureau’s estimate for Boston was released, Menino ordered the Boston Redevelopment Authority to tabulate the number of housing units built and demolished since 2000. He said the city could more accurately measure housing that the census might have missed, such as new college dormitories and commercial buildings that were converted to housing.
BRA officials found that the Census Bureau had underestimated new housing production by more than 3,000 units, and overestimated demolitions by 1,800, said Alvaro Lima, the agency’s director of research.
Those numbers really raise about as many questions as they answer.
I assume more thnan 3,000 means less than 3,100, so let’s round it up to 3,100 for a total of 4,900 housing units that the Bureau’s originial estimate of Boston’s population had not taken into consideration. The Bureau extracted the 4,900 new and non-demolished old units into 37,000 people to add to the the population estimate.
That’s odd.
The 2000 Census counted 251,935 housing units in Boston. The 4,900 that the Census missed estimating in 2005 represent 1.9% of that total (though I don’t know what the current total is), but the population estimate was off by 5%, according to the revised numbers. On the other hand, the new population estimate is just a tad over 1% higher than the 2000 population count. But then again, the new population estimate squeezed 37,000 people out of about 4,900 housing units, which is about 7.6 people per unit, vastly higher than the 2.3 average in 2000.
I guess there could be a few different explanations:
1) There are factors that explain the estimate gap that aren’t mentioned in the Globe article.
2) Most of the previously unestimated population lives in group quarters (dorms, basically) and are students more often than immigrants.
3) The new estimate has little backing by evidence and is really just the product of a machine-politics big-city mayor twisting arms.
The Census Bureau should explain which new housing units it misses. It’d be a bit much to ask for a list of all 3,000 units, but surely it could reveal the two or three buildings with the biggest number of housing units that it overlooked.
It should also explain why it increased the estimate by 37,000 when the difference in housing units suggests an upward revision of less than 12,000 rather than 37,000. A revision of 12,000 would still have meant a population loss for Boston.
The United State Census Bureau: “Helping you make informed decisions.”
Looks like the Globe has some explaining to do itself. Yesterday it claimed that the Mayor hailed the news as a “psychic boost,” a rather preposterous idea, but today it writes the Mayor hailed it as a “psychological boost.” Is the Globe covering for the Mayor or did Levenson and Abraham screw up royally yesterday (I say royally because you’re really smearing a person if you falsely claim he hails something as a psychic boost)?
