Scripts to reproduce analyses and tables should be run in the following order.

1. makedata.r
2. analysis.do (NB: this may take a long time to run)
3. descriptives.r
4. predictions.r

Two helper scripts, csv2tex.r and est2csv.do, should remain in the directory, but do not need to be called on their own.

Another file, ipedsgeo.r, can be used to create colfips2004.csv (see instructions in Data directory README file).

After the scripts are run, paper tables can be reproduced using tables.rnw in the ./tables subdirectory.