I want to calculate the family poverty rate using the CPS microdata. To do this, I have edited the data so that only the first member of each family is left in the data. Their poverty status (and that of their entire family’s) is indicated by OFFPOV. I take the weighted HWTSUPP crosstabs of OFFPOV to get the family poverty rate. The problem is that I am getting a family poverty rate that is quite different from the family poverty rate given by the Census Bureau. Any ideas on where I could be going wrong?
You need to take one additional step in order to match the Census figures: limit your sample to only primary families (FTYPE=1). This will remove single-person households and households where the householder lives exclusively with non-related persons. The Census Bureau does not include either of these living arrangements in its family poverty rate.
Hope this helps.