How to handle locale-specific behavior?

[tbrown122387]

A student submission featured code that read in a data set. On his machine and on mine, it looked like

                      Series.Name    Series.Code   Country.Name Country.Code X2014..YR2014.
1 Income share held by second 20% SI.DST.02ND.20    Afghanistan          AFG             ..
2 Income share held by second 20% SI.DST.02ND.20        Albania          ALB             ..
3 Income share held by second 20% SI.DST.02ND.20        Algeria          DZA             ..
4 Income share held by second 20% SI.DST.02ND.20 American Samoa          ASM             ..
5 Income share held by second 20% SI.DST.02ND.20        Andorra          ADO             ..
6 Income share held by second 20% SI.DST.02ND.20         Angola          AGO             ..

However, when it runs on the server, it looks like this

                  X...Series.Name    Series.Code   Country.Name Country.Code
1 Income share held by second 20% SI.DST.02ND.20    Afghanistan          AFG
2 Income share held by second 20% SI.DST.02ND.20        Albania          ALB
3 Income share held by second 20% SI.DST.02ND.20        Algeria          DZA
4 Income share held by second 20% SI.DST.02ND.20 American Samoa          ASM
5 Income share held by second 20% SI.DST.02ND.20        Andorra          ADO
6 Income share held by second 20% SI.DST.02ND.20         Angola          AGO

All subsequent code that assumed column names therefore broke.

This stackoverflow post mentions that default column names are not always consistent across machines with different locales.

• what are other examples of inconsistencies?
• what is the best way to fix them so that the student doesn't have to worry about this sort of thing?