A large number of mySidewalk datasets come from the American Community Survey (ACS). The ACS is an ongoing survey conducted by the U.S. Census Bureau, with nearly 3.5 million respondents participating in the survey every calendar year. The Census Bureau uses the ACS survey to annually publish its hallmark data product, the ACS 5-Year Estimates. On December 7, 2017, the Census published the 2012-2016 ACS 5-Year Estimates. 

What does this mean for mySidewalk Customers?

All mySidewalk datasets based on ACS 2011-2015 5-year estimates have been updated with the 2012-2016 estimates. Because the Census Bureau recommends against comparing 5-year estimates of overlapping years, mySidewalk will no longer curate the 2011-2015 ACS estimates. The update affects all saved assets and visualizations in mySidewalk that use ACS data, including reports, maps, and dashboards. If you have created any reports, maps, or dashboards using ACS 2011-2015 5-year estimates, be sure to update and check any written content associated with the data (the only piece that will auto-update is the data itself).

It's also worth mentioning that the updated ACS 5-Year Estimates will be incorporated in all time series datasets in early 2018. 

Understanding ACS 5-Year Estimates

The Census Bureau produces the ACS 5-Year Estimates by pooling together the results of the five previous survey years. Using pooled data frame, the Census Bureau generates estimates of thousands of economic and demographic characteristics for the entire United States. Through pooling multiple years of data, the Census Bureau trades precision with respect to time for improved estimation accuracy. Because they are generated with five years of survey data (as opposed to a single year), the five year estimates have several interpretative nuances.

  • ACS data are estimates. The Census Bureau collects data from a sample of the population in the United States rather than from the whole population. 
  • ACS 5-year estimates represent the average of characteristics over an interval of time (2012-2016) as opposed to a single point in time.
  • Five-year estimates cannot be compared with previously published five-year estimates consisting of overlapping years. For example, do not compare ACS 2012-2016 five year estimates with ACS 2010-2014 5-year estimates. However, ACS five-year estimates may be compared with Decennial Census 2000 and Decennial Census 2010.

Questions? Reach out  at any time via chat or to [email protected]

Did this answer your question?