Missing a variable in data despite it being in the extract

Good morning,

I am downloading census data from different Latin American countries between the years 1990-2020s and have been consistently finding two variables missing from my dataset and the cleaning codes, despite them being selected in the data extract. The variables are AREAMOLLWGEO1 and AREAMOLLWGEO2, which represent the areas of certain geographical units in square meters.

Does anyone know why this might be happening? I checked on multiple samples (Brazil, Chile, Costa Rica) and the same problem persists. Could it be that the data is too large (with too many variables) and this one gets deleted? But why specifically those and not the ones at the bottom of the list?

I appreciate any leads.

From your description of your issue, it sounds like either the AREAMOLLWGEO1 and AREAMOLLWGEO2 variables are being dropped entirely from your data file or that the variables listed in your stats package have all values missing.

To attempt to replicate this issue, I created data extracts that I read into Stata with the IPUMS-provided .do file that contained all countries in the Americas for the 2010-round censuses and included the variables AREAMOLLWGEO1 and AREAMOLLWGEO2. Each country (including Brazil, Chile and Costa Rica) had valid numeric data for AREAMOLLWGEO1/O2 for each value of GEOLEV1 in the dataset. Additionally, I looked at data extracts associated with your user profile and resubmitted them. You may have received notification that some of your extracts were re-processed last week. Check to see if those files (you may need to resubmit them) contain the variables and data that you expect. Each extract that I examined had valid data for those two variables.

If these variables are still missing when you look at the recreated data files, please email us directly at ipums@umn.edu with the specific extract numbers that are causing issues and details about the stats package you are using to work with the data.