I used IPUMS data to study residential neighborhood, and now I’m working on a project that examines how living in segregated/integrated neighborhood affects health outcomes. But I cannot find geoid from IPUMS health surveys. Any ideas of how to link data from IPUMS to IPUMS health surveys?
I assume you are using IPUMS USA or IPUMS NHGIS to study residential neighborhoods. IPUMS USA and IPUMS NHGIS both provide data from the U.S. Decennial Census and the American Community Survey. IPUMS NHGIS provides aggregated data tables by geographic area, while IPUMS USA provides microdata (person- and household-level data). IPUMS Health Surveys, on the other hand, provides data from the National Health Interview Survey (NHIS), which is conducted by the National Center for Health Statistics (NCHS) at the CDC, and data from the Medical Expenditure Panel Survey (MEPS), which is conducted by the Agency for Healthcare Research and Quality (AHRQ). The NHIS, the MEPS, and the Census/ACS are all distinct surveys. The Census and ACS cannot be linked to the NHIS or MEPS. Geographic levels in IPUMS Health Surveys are much broader than those available in IPUMS USA and IPUMS NHGIS. In the NHIS, the smallest geographic level is U.S. region (REGION). The variable METRO, which reports whether a respondent lives in a metropolitan statistical area, is available in NHIS samples from 1963-1996. In the MEPS, no information on geography of respondents is available.
Some data on health insurance are available in some IPUMS USA samples (list of health insurance variables) and in IPUMS NHGIS in some years (select “Health Insurance” in topics to view the available tables), but no variables that directly measure health. The ACS also includes a series of questions on disability.
HI Isabell,
You’re right, I use IPUMS USA to study residential neighborhood. Now, I’m in a process of checking whether I can study individual’s differences of living in integrated/segregated neighborhood and their health outcomes. I think IPUMS health surveys are also individual-level, however, it does not provide geographical identifiers like PUMA id. I notice that they have region, metro id, correct?
What would you recommend a better way for my research question? Is there a way that I can use IPUMS data for it?
Thanks!
Shuang
Thanks for the information!
None of the IPUMS data collections provide data on health outcomes at the neighborhood level. You are correct that the NHIS and MEPS do not provide detailed geographic identifiers like PUMA ID. In the NHIS, the smallest geographic level is U.S. region (REGION). The variable METRO, which reports whether a respondent lives in a metropolitan statistical area, is available in NHIS samples from 1963-1996. METRO does not specify which metro area a respondent lives in—only whether or not they live in a metro area. In the MEPS, no information on geography of respondents is available.
You can use NHIS data from IPUMS Health Surveys to study health outcomes at the U.S. region level. Some data on health insurance are available in some IPUMS USA samples (list of health insurance variables) and in IPUMS NHGIS in some years (select “Health Insurance” in topics to view the available tables), but no variables that directly measure health. The ACS also includes a series of questions on disability.
The restricted versions of both NHIS and MEPS data include more geographic identifiers. The NCHS provides information on applying for access to restricted NHIS data, and the AHRQ provides information on applying for access to restricted MEPS data.
If none of these options is suitable for your purposes, I would recommend seeking out data sources outside of IPUMS, such as CDC PLACES.