Winford H. Masanjala and Chris Papageorgiou, "Rough and Lonely Road to 
Prosperity: A Reexamination of the Sources of Growth in Africa Using 
Bayesian Model Averaging", Journal of Applied Econometrics, Vol. 23, No. 
5, 2008, pp. 671-682.

Results in this paper can be replicated using two datasets and three 
programs that can be found in the file mp1-data-programs.zip. All files 
are plain ASCII text files in DOS format. Unix users should use "unzip -a".


Documentation for programs
--------------------------

The programming language R with a BMA function was used. Both the R 
package and BMA function are publicly available from the world wide web. 
The program Global-BMA.R produces the "Global" results in Table 2. The 
program Interaction-BMA.R produces the "Africa" and "Non-Africa" 
posterior means in Table 2. The program Interaction-data.txt is used to 
produce the posterior variance for the interacted variables.


Documentation for data 
----------------------

The two datasets are formatted and ready to be used in R. The datasets 
used are subsets of the Sala-i-Martin (1997) dataset that is publicly 
available. In both datasets there are 104 rows (representing 104 
countries). Country names are presented in country-names.txt. The names 
of each variable are clearly marked in the first row of each data file. 
Definitions of these regressors can be found in the Appendix of the 
paper.

Global-data.txt

The Global-data.txt dataset should be used with the Global-BMA.R 
program. Column 1 is a row identifier (this is how data are formatted in 
R; do not remove this first column). The actual data starts from column 
2. Column 2 is the dependent variable (growth rate) and columns 3-26 are 
the 24 regressors used.

Interaction-data.txt

The Interaction-data.txt dataset should be used with the 
Interaction-BMA.R program. Column 1 is a row identifier (this is how 
data are formatted in R; do not remove this first column). The actual 
data starts from column 2. Column 2 is the dependent variable (growth 
rate); columns 3-27 are the 24 regressors in addition to a sub-Saharan 
Africa dummy; columns 28-51 are variables interacted with a sub-Saharan 
Africa dummy.


R programs and data
-------------------

To produce "Global" results in Table 2 of the published paper:

1a) Global-BMA.R 
1b) Global-data.txt

To produce "Africa" and "Non-Africa" results in Table 2 of the published
paper:

2a) Interaction-BMA.R
2b) Interaction-var-BMA.R
2b) Interaction-data.txt

For the country names corresponding to each row see:

3) Country-names.txt


Corresponding author contact information
----------------------------------------

Chris Papageorgiou
Research Department
International Monetary Fund
700 19th Street, NW
Washington DC, 20431
U.S.A.