So if I use either MISH-4 or MISH=8 and I use several consecutive years of data there will be no double-counting?
And what about using the NBER ORG extracts? Do you know if a year’s data includes all the months from the CPS? What are the double counting issues in these extracts
a) within a single year
b) using consecutive years?
Yes, you are correct about using either MISH==4 or 8 to avoid double counting with IPUMS CPS data. Additionally, the NBER ORG files also include all outgoing rotation group observations, regardless of the fact if they were previously included in an ORG sample. So, you’ll need to make the same adjustments with NBER ORG data to avoid double-counting.