I am trying to compare puma population of different groups as compared to thier national average. Person weights give a good estimate of the national proportion of, say an ethnicity, but if I want to have get the proportion of population that is of a certain group by PUMA, I should have puma level weights. Does IPUMS ACS provides that?
With the use of PERWT, IPUMS-USA data can be considered representative of geographic areas of all available sizes. As a result, there is no need for a separate weight when performing analyses at the PUMA level. The standard errors of statistics based on smaller geographic areas will be larger than those based on national data, but this is due only to the reduced sample size.
Hope this helps.