The Census Bureau recently announced plans to eliminate the American Community Survey (ACS) three-year socioeconomic estimates for all areas over 20,000 persons. These estimates are often used to evaluate current and past socioeconomic conditions in medium sized communities and the loss of this data set would be a blow to doing important, timely analysis.
The Census Bureau collects data annually for all geographies, but because the ACS sample size is not large enough to provide single year data for all areas, the data are reported for one-year, three-year and five-year time periods. Areas of 65,000 persons or more receive the one-year data products as well as the three-year and five-year data products. Medium sized geographies, such as counties and places, between 20,000 and 64,999 persons receive estimates based on a three year period. All areas receive estimates based on the five year data collection.
Eight of Maryland’s 23 counties fall below the 65,000 person threshold and thus do not receive one-year estimates. Therefore, to compare data across all Maryland’s 24 jurisdictions the three-year or five-year ACS products must be used, with the three-year data being much more current.
Since 2007 the Maryland Department of Planning has used the three-year ACS data to analyze current economic and social conditions relating to changes in income, poverty level, educational attainment, transportation and many other subjects for Maryland’s 23 counties and Baltimore City. If the three-year ACS data did not exist, the much less timely five-year data would have had to be used.
In addition to the three-year ACS data being more timely than the five-year ACS data, the three-year data also allows for more frequent comparisons over time and allows for better analysis of major events such as the Great Recession. In order for multi-year ACS data to be used for comparisons over time, there cannot be any overlapping years between the time periods being compared.
For the Great Recession, which lasted from December, 2007 thru June, 2009, the three-year 2007-2009 ACS data (released in 2010) roughly corresponds to this period. The first post-recession data with no overlapping years that could be compared to the 2007-2009 data was 2010-2012, released in 2013. With the three-year ACS data for 2007-2009 and 2010-2012, the Maryland Department of Planning was able to show that the impact of the Great Recession was greatest on the rural counties on the Eastern Shore and in Western Maryland and in Baltimore City. (See, A Status Report on Maryland Jurisdictions from the American Community Survey.)
If only the five-year data were available, the coverage of the Great Recession period would have been much less precise (with possible beginning years reduced to 2006-2010, or 2007-2011, or 2008-2012) and there would be nothing to compare to for at least five years after the ending year of the initial data set used (2011-2015, released in 2016, 2012-2016 released in 2017 and 2013-2017 released in 2018.
The counties in Maryland that fall below the 20,000 population threshold and will now only receive the five-year ACS are: Caroline, Dorchester, Kent, Queen Anne’s, Somerset, Talbot and Worcester on the Eastern Shore and Garrett in Western Maryland.
The elimination of the three-year ACS data series would be a severe blow to important, timely data and the analysis which it supports for much of Maryland. Previous budget cuts to the Census Bureau resulted in the elimination of the U.S. Statistical Abstract and the Consolidated Federal Funds report. The elimination of the 3-year ACS would be a loss of another critical data product at a time when reliable, current data are needed more than ever for making wise decisions about the allocation of scarce resources.
