American Community Survey Still Lagging?

Over the last few weeks the Census Bureau has released results from the 2009 American Community Survey (ACS) at the Tract level and the 2010 Census population counts at the State level. The 2010 Census counts were very favorable for D.C. as for the first time in six decades the District grew in population rather than declined. The 5.19% growth pushed D.C. back over 600,000.

The 2010 Census will ultimately reveal a lower level picture at Census Tract and Block Group for the ten basic questions on the short form including name, sex, age, date of birth, ethnicity, race, and homeownership status. But we are months away from that lower level of detail being published.

In the meantime we have the aforementioned 2009 ACS survey for Tract data. Using a Mapping America tool published by the NYTimes it is possible to view ACS results by tract in a visual geolocational form rather than mind numbing spreadsheets.

The Mount Vernon Triangle falls within Census Tract 47, Block Group 3. Since the ACS data does not drill down to Block Group we must settle for reviewing the entirety of Tract 47 which extends our Triangle borders eastward past New Jersey Ave to North Capitol.

Image from NYTimes ACS Mapping America tool; click to enlarge

The dot representation of race on that map suggests the Mount Vernon Triangle is still overwhelmingly black. Living here we know that not to be true. It should be interesting to see how these ACS estimates of race stack up against the 2010 Census estimates for the tract when they are published. Unfortunately the continued lag the ACS has in recognizing the changes in our community could continue to suppress the measure of our change in the areas of income and education. These attributes are not items on the 2010 Census and will be estimated purely by ACS going forward. While the Census could show that the Triangle has gained thousands in population the failure of the ACS to rapidly gather contextual information on that transforming population could steer the estimation techniques to suggest the growth is mostly more of the same as how the tract had been composed in the past (i.e. low income, hence the 15K median income in one of the charts below).

Image from NYTimes ACS Mapping America tool; click to enlarge

Image from NYTimes ACS Mapping America tool; click to enlarge

Image from NYTimes ACS Mapping America tool; click to enlarge

1 person likes this post.
SociBook Digg Facebook Google Yahoo Buzz StumbleUpon


  1. 1

    Lynda says

    Good points. I could potentially see that the initial poor quality population characteristics for the 2005-2009 growth in the Triangle could warp analysis in future years. The growth spike will eventually be characterized accurately by the ACS. However when analyzing the more accurate snapshot in 2012 or 2013 against substandard 2009 estimates it could give false amplification of the level of gentrification. It may look like a 1000+ low income households were replaced between 2009 and 2013 when the household growth was always affluent over that time but the estimation method wasn’t able to accurately capture it as such.