Data vintages? birth place and previous residence, example Benin 1992

Dear IPUMS team / user community

Thank you for making this amazing resource available. I have been working with your data for several years and my question may be related to changes that may have occurred in your data over time [if it is not, as is more likely, related to my own carelessness].

I am trying to define variables for internal (within-country) migrants using (a) individuals’ locations at the time of the census and (b) variables related to birth place or previous residence. I will use Benin1992 as an example.

I originally accessed these data on 28th September 2018. I have re-downloaded them today (16th March 2022).

When assembling the migration data, I noticed several changes in birth places and previous residence variables for some (though not all) individuals between the values of these variables in the (to me) original (2018) vintage of the data as compared to today’s vintage. Has there been any data revision in the interim? If so, since I am interested in all African censuses, is there a systematic list of such
revisions available (apologies if such a list is an obvious place and I have overlooked it).

BEN_1992_discrepancies.csv (2.5 MB)

I attach a .csv file with all the discrepancies.

Many thanks in advance for your help.

Kind regards


It was in an obvious place:

" Errors affecting codes for BPLBJ2 in Benin 1979, 1992, and 2002 were corrected."

Is there a way to obtain previous version of the data?

Kind regards,


In the case of errors in the data, we would recommend using the most up-to-date versions of the data as they represent improvements. Please let me know if I have misunderstood your request.