The code in this directory is adapted from Balistreri, Boehringer, and 
Rutherford (2025), https://doi.org/10.21642/JGEA.100101AF.  The code is 
set for replication of the results presented in the paper:

Balistreri, Edward J., Suraiya Binte Ali, and Christine McDaniel (2025)
	"Tariff: the most beautiful word in the dictionary?," Working 
	paper, https://balistreri.createunl.com/Papers/Beautiful_Tariffs.pdf

The programs included here are open-source, but replication will require 
a recent version of the GAMS software installed with the solvers PATH an MPSGE.
The data are derived GTAP version 11 source data, which require a user license 
(https://www.gtap.agecon.purdue.edu/).  For replication only the compiled data
file "tradewar11c.gdx" is available upon request to edward.balistreri@unl.edu.

The model is DOS batch-file driven:
	run.bat executes all programs used to generate the ARM and BRF
		results in the paper.  The results will appear in the 
		constructed file "pivotreport.xlsx" which will include the
		tables and various displayed reports.  "pivotreport.xlsx"
		is copied from template.xlsx before the model output is 
		written to it.  Thus, the user must `refresh all' to view
		new reports.

Model code:
	model.gms includes the core model formulation from Balistreri, 
		Boehringer and Rutherford (2025).  The model is formulated
		in GAMS/MPSGE and includes a number of features and logical
		options that are not used in the Beautiful Tariff paper.  
		This renders the model as formulated less than transparent,
		even for the few people who take advantage of the MPSGE
		modeling language.

Accessible (algebraic) model code:
	run_clean.bat executes cleanmge.gms with the appropriate restart 
		files.
	cleanmge.gms replicates exactly the structure of model.gms as it is 
		executed in the Beautiful Tariffs paper but in a more
		accessible format.  There are two models in cleanmge.gms one
		is a simplified MPSGE model that does not include the unused
		features in model.gms.  The second is a model formulated in 
		GAMS algebra.  This plain algebra model is the source of the 
		equations documented in Appendix C of the paper.  Both of the
		models presented in cleanmge.gms are shown to represent the 
		non-linear system in model.gms (as executed for the present 
		paper) through a series of replication checks at the benchmark 
		and counterfactual equilibria.  The model.gms code is retained,
		however, for scenario analysis because the extensive reporting
		code (e.g., report.gms) is built around the nomenclature used 
		in model.gms.

Please direct any questions to edward.balistreri@unl.edu.