Hello,
I understand not all variables from the MEPS HC are not currently included in IPUMS. I’m curious if the HMO flag (UPRHMO) under the Person Round File will be included and any feedback as to whether specific reasons hinder it from being included.
If I were to merge onto an already created IPUMS MEPS dataset, would you recommend merging/pooling using DUPERSID and PANEL?
I appreciate your feedback.
Best,
Adam
Dear Adam,
Currently, we only offer variables from the annual full-year consolidated MEPS-HC files. We have plans, contingent on continued funding, to add variables from other MEPS-HC files, including the PRPL file. When we get to the point where we are ready to add variables from other files, we can prioritize the UPRHMO variable. If you have other variables you’d like us to see add sooner rather than later, please let us know!
The short answer to your question about linking keys to merge: Yes, you can perform a many-to-one merge on DUPERSID and PANEL. In IPUMS MEPS, this information is saved in the variable MEPSID as a ten-digit string variable* or separately as DUID, PID, and PANEL, if you would rather go that route. If you do this, you will end up with a person-round-policyholder-establishment file (which is the structure of the PRPL file). The longer answer: Because the PRPL file has a very complex file structure, you have to make some decisions about what you want your merged file to look like. For example, if you would like to end up with a person-level file with UPRHMO added as another person-level variable (set of variables), that would change the data structure option you choose when downloading your IPUMS MEPS extract and require some work to reshape the file to get exactly what you want. You would also need to make some decisions about how you want to handle people who have more than one value of UPRHMO per round. If you get in touch with our user support team via ipums@umn.edu describing what kind of file you’d like to end up with, they can forward your email to me and I would be happy to walk you through steps to merge in the original MEPS-HC PRPL file and reshape to get what you want.
Hope this helps,
Julia
*If merging panels 9 and earlier from IPUMS MEPS to the original MEPS HC files, add a “0” to the beginning of the DUPERSID variable (e.g., gen mepsid = “0” + DUPERSID + PANEL, assuming DUPERSID and PANEL are string (not always the case in the original data!)). We left pad mepsid values in years where PANEL values are only one column wide to get a consistent variable width over time.
Hi Julia,
I appreciate your feedback. I have sent a request to ipums@umn.edu. I will stay tuned to hopefully see the PRPL file incorporated to this rich data source in the future.
Thanks!
-Adam