I'm looking to classify persons as being in "near poverty" As I understand, this is usually defined as a having an income above 100% of the poverty level with some upper cutoff, for example 125%. I also understand that there is no uniform definition for this upper percentage and that values like 125% and 150% and maybe 200% are commonly used. To do this I think that you can use PUMS data and the variable POVPIP. Am I doing this correctly ? I need the results for a town (CSD) which is a subarea of the PUMA. I plan on using small area estimation (SAE) adjusting for some covariates from ACS tables for the town. Also I need age 60+ but this is just another wrinkle as AGEP is in the PUMS data. Does using the B17001* POVERTY STATUS IN THE PAST 12 MONTHS BY SEX BY AGE (with different tables for different races) make sense? This would give an adjustment for Sex Age and Race and throwing in other marginals using B17003-B17005 would adjust for Education, employment status and work experience. Does anyone out there have experience doing this? I would expect that this is a common analysis. This work is for the local council on ageing.
After thinking a bit I can adjust for many other ACS detailed tables such as Tenure alone (renter/owner) Education Attainment alone etc.
Dave Dorer
Hi Dave - Have you seen Table B17024 - Ratio of Income to Poverty by Age? As you suggested, the bins are 55-64 and 65-75Anything less than 200% FPL seems appropriate, unless you're in a high cost state, then I would consider a higher ratio