Can we manually merge multi-year ACS data by ourselves?


Dear IPUMS staff:

I am planning to use the 2009-2013 5 years ACS sample for online data analysis, but our existing muli-year dataset doesn’t have it. I am planning to merge it by myself (selecting ACS 2001-14 sample and use the datasetnumber filter to get the 09-13 sample). Would that be problematic?

Thanks for the help.


Follow Up to Answer to "Release of 2009-2013 5yr PUMS"

First, I want to alert you to the recently released 2013 and 2014 multi-year microdata samples. Since the 2013 multi-year files are not currently available through the online analysis system, using the microdata samples is the recommended method.

However, since the multi-year data files are simply a concatenation of the single-year ACS data files, you can approximately reconstruct the multi-year files yourself using the online analysis system. All respondents from the 2009, 2010, 2011, 2012, and 2013 1-year ACS files are combined to create the 5-year file. This means you can simply select these samples, as you described; however, there are two additional adjustments you will need to make. First, dollar values should be adjusted to constant 2013 dollars (see the IPUMS inflation adjustment page). Second, all weight values should be divided by five. While the Census Bureau does use more complex methods for generating multi-year weights (see Weighting Methodology from the 2012 3-year Accuracy Statement), simply dividing by five is the primary adjustment and provides extremely similar values to the Census Bureau calculated weights. You can verify this by trying to match the 2009-2013 PUMS Estimates for User Verification. If you are still uncertain about this combination method, I would recommend replicating your analysis using the combined 2010 to 2012 ACS samples and compare your results with the actual 2012 5-year ACS sample in the online analysis system.

Hope this helps.