Generate family weights for summary statistics for families (not people or households)

USA
1

I am trying to generate summary statistics for families in the US – e.g., the number of families that earn below a very specific value (e.g., $12,345).

I understand that family is uniquely identified by the combination of variables ‘serial’ and ‘famunit,’ however, I am unclear on how apply weights to generate estimates for the number of families

Using ACS 2012 1-year estimates, I know from published data that there should be 76,509,262 families in the US. Is there sample STATA code that generates family weights so that I can replicate this value?

I have gotten this far – I can generate incomes by family, but I don’t have a weight to apply: collapse (sum) incwage, by (serial famunit)

2

While there is no specific weight generated for families, The Census Bureau uses Housing Unit Weights to generate family-level estimates. A little bit more info on the matter can be found in this forum response. Using the household weight you should be able to get close to published estimates, however, the Public Use file is a subset of the full sample and is subject to additional sampling error and further processing before being released so it is not likely that you will get the exact same figure.

I hope this helps.

1 Like