I have seen that several studies using CPS data include many types of in-kind transfers in their analysis, including food stamps, lunch, housing and energy subsidies. However I can only find long data series for food stamps and energy subsidies, in the IPUMS database. Lunch and enrgy subsidies are only available between 2010 and 2013, as a part of the Supplemental Poverty Measure family.
Could you please tell me why this is the case and whether there is a way to recover these subsidies for a longer period of time?
The ASEC variable LUNCHSUB identifies households where children received a free or reduced price lunch and FRELUNCH indicates the number of children in a household that usually received a free or reduced price lunch. These variables are available from 1980-2014. Additionally, the Food Security Supplement asks whether children in the household received free or reduced price lunch during the past month, which is recorded in the variable FSLNCHFRC in various months from 1995-2010. Unfortunately, none of these variables either report the value of the lunch subsidy or distinguish between free and reduced price lunches; however, the Bureau of Labor Statistics paper “Supplemental Poverty Measure Thresholds: Imputing Noncash Benefits to the Consumer Expenditure Survey Using Current Population Survey” describes a probabilistic model for imputing school lunch values with CPS data.
As for housing subsidies, the RENTSUB variable indicates if a household was paying lower rent due to a government subsidy. This variable is available from 1976-2014, but does not report the value of the subsidy; however, the Census Bureau paper “Estimating the Value of Federal Housing Assistance for the Supplemental Poverty Measure” discusses several methods that have been used to estimate the value of housing subsidies with CPS data.
Thank you very much for your answer. I had seen those variables indicating wether subsidies were received or not but I am mostly interested in the value. Do you know if imputed values are available somewhere, perhaps applying the methodologies described in the papers you mentioned?
I am unaware of any publicly-available datasets with in-kind transfer values imputed for CPS respondents prior to 2010. You might consider contacting the authors of the papers linked to above, as they are likely more familiar with the available resources you need.
Sorry that I could not be of more direct assistance.