John Coglianese, Lucas W. Davis, Lutz Kilian, and James H. Stock,
"Anticipation, Tax Avoidance, and the Price Elasticity of Gasoline
Demand", Journal of Applied Econometrics, forthcoming.

This readme file describes the data and code used to perform the
empirical analysis in this paper. All files are ASCII files in DOS
format. They are zipped in the file cdks.zip. Some files will be
unzipped in the current directory, and others will be unzipped in three
subdirectories (see below). Unix/Linux users should use "unzip -a".

All datasets are in .CSV format, to facilitate use in all statistical
packages. The main analysis, including the creation of all figures, is
performed in StataCode.do, a do file for STATA.

First, unzip the file cdks.zip. Several of the more involved
datasets used in the analysis should be kept in separate directories,
which should happen automatically when cdks.zip is unzipped.
These subdirectories are:

  Pop Density from Census
  Tax Rates from Highway Statistics
  Unemployment Rates from BLS

In each case, the original data are included as a .CSV file, and DO1.do
(a STATA do file) cleans the data and produces a STATA file with the
appropriate data.  Details about the cleaning as well as about the
source for the data are documented in the do files themselves and in the
paper. Before running StataCode.do, you should run all three of these
ancillary do files. Before running each .do file, change the base
directory in the "global basedirectory" line of the .do file to the
directory where the code and data are located.

As explained in the main do file, the monthly gasoline prices by state
(in PET_CONS_PRIM_A_EPM0_P00_MGALPD_M.csv) were downloaded from the
Department of Energy Website,

http://tonto.eia.doe.gov/dnav/pet/pet_pri_allmg_d_nus_PTA_cpgal_m.htm.

Similarly, the monthly sales volumes by state (in
PET_PRI_ALLMG_A_EPM0_PTA_CPGAL_M.csv) were downloaded from

http://tonto.eia.doe.gov/dnav/pet/pet_cons_prim_dcu_SDE_m.htm.

These files have been saved as is in the main directory. The final two
.CSV files (cpi.csv and BLSprices.csv) were both downloaded from the
Bureau of Labor Statistics as indicated in the do file.

John Coglianese
Harvard University
John F. Kennedy School of Government
Cambridge, MA 02138
Phone: 339-368-1480

coglianese [AT] fas.harvard.edu
