This directory contains files used for

    ``Accounting for the Great Depression''
      American Economic Association Papers and Proceedings
      May 2002
      by V.V. Chari, Patrick Kehoe, and Ellen McGrattan

The appendix describes the computations done in the main codes:

   wedges.m   Matlab code that generates wedges
   aea.f90    Fortran code that computes equilibria
   aeasim.m   Fortran code that compares data with model

The file used to generate Figure 1 in the paper is wedges.m. 
Look at the last figure plotted after running it in Matlab.

I should note that some versions of Matlab may require some
editing of the code wedges.m because they give different 
names to the data being loaded than my version.  For example,
the code has "load nipa/605a" which loads in the file 605a
in the subdirectory nipa.  My version of matlab names it
X605a because the first character of the name is numeric.
Other versions use the naming convention: nipa_605a.

The specific input files for aea.f90 that generated the results 
in the paper are aea.inp.*, with *=a, taun,all, and taux.  Copy 
each to aea.inp (which is called by aea.f90) and compile aea.f90 
along with qgausl.f and uni.f.  The command I use is in file comp 
(where pgf90 is the compiler, -r8 and -fast are options for real*8
and optimization, a.out is the name of the executable, aea.f90 is 
the main routine, qgausl.f and uni.f are subroutines, and lapack 
and blas are necessary libraries.  Standard math Fortran libaries 
include lapack and blas routines.  If yours does not, the source 
code is free and can be obtained via ftp.  
(See, for example, http://www.netlib.org/lapack and
http://www.netlib.org/blas.)

To generate Figures 2-4 in the paper, run aeasim.m after compiling 
and running aea.f90. The output of aea.f90 is the input of aeasim.m. 

Data used in the project are located in subdirectories

   ./nipa
   ./labor

Files for my UM8313 class are located in 

   ./um8313

The students were given countries simulated from the Fortran
codes and asked to determine the sources of fluctuations -- 
during normal times and during times of depression.
The results of the experiments are summarized in countries.txt.
What the students discovered is intuition can get you only so
far.  What I found interesting about the particular experiments
that we did is this: the students oftentimes assumed that
changes in investment or capital tax rates were the source
of fluctuations -- even when they were fixed at a constant rate.

Ellen McGrattan
May 2002