I am interested in performing a national rural analysis with 2012-2015 ACS PUMS data. The ACS PUMS documentation recommends using data from the Missouri Census Data Center (MCDC) <http://mcdc.missouri.edu/websas/geocorr14.html> to identify other geographies. The national level data downloaded from the MCDC includes state, county 2010, CBSA 2010, CBSA type, urban-rural portion. These data are then merged with 4 years of ACS data by state and puma.
Tabulations of rural areas, using CBSA type variable or the urban-rural portion, results in over 2 million missing observations. Has anyone else encountered this problem? I realize that the PUMA represents areas with populations of 100,000 or more people and a 1:1 match between PUMA:rural areas may be limited- I just didn't expect so much missing data.
While, iPUMS is a great resource, I would like to work directly with ACS data from the Census Bureau. Do any of you have other recommendations for identifying rural areas with the ACS PUMS?
Regards,
Shondelle Frederick