This directory includes the Julia code and data needed to replicate the
Anderson and van Wincoop (AvW) model results and the associated figures in:

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 files in this directory are as follows:
  28-Feb-2025   2:28:00a          6,144   AvW-9-plot.jl
  28-Feb-2025   2:28:00a          9,075   AvW-deficit-US-plot.jl
  8-Dec-2024   12:52:00a          4,096   Fitted-value-Avw-USplot.jl
  8-Dec-2024   12:48:00a          7,168   fitted-value-avw.jl
  11-Jun-2025   9:55:00p          3,072   BRF-9fig.jl
  11-Jun-2025  10:37:00p          2,048   combined-graph-ARM-BRF.jl
  25-Dec-2025   7:59:28Am         6,295   9-graph-brf-2.csv
  11-Jun-2025  10:17:28A          3,285   US-uni-arm-2.csv
  11-Jun-2025  10:16:35A          3,271   US-uni-brf-2.csv

------------------------------------------------------------
AvW-9-plot.jl:
  Baseline AvW (Armington) model in Julia using MPSGE.jl. This script:
    • calibrates the model to a 9-region balanced trade matrix,
    • sets a base scenario with 60% US–China tariffs and 10% US tariffs on all
      other imports,
    • varies a coordinated "non-Chinese" retaliation tariff (0–100%) against US
      exports, and
    • generates a report DataFrame and a combined 3×3 (nine-panel) welfare plot.

AvW-deficit-US-plot.jl:
  AvW model extension that allows for trade imbalances and alternative benchmark
  calibrations. The script includes three trade matrices:
    (i) "tbl1" (balanced benchmark),
    (ii) "gtap" (GTAP benchmark with imbalances), and
    (iii) "ppml" (PPML fitted values / econometric calibration).
  It constructs the deficit terms (gamma and v_b), solves the model, and then runs
  a unilateral US tariff experiment (0–100%) with no retaliation. Output is stored
  in a report DataFrame and used for the appendix optimal-tariff figure comparing
  balanced vs. imbalanced vs. PPML calibrations.

Fitted-value-Avw-USplot.jl:
  AvW unilateral US tariff experiment (no retaliation) calibrated to the PPML fitted
  trade matrix. Loops over uniform US tariffs (0–100%), records welfare changes for
  each region and the world aggregate, and produces a plot focused on US and world
  welfare with the implied "optimal" tariff rate.

fitted-value-avw.jl:
  A working / exploratory script closely related to the PPML-calibrated AvW runs.
  It builds the AvW model (MPSGE.jl), sets the same base tariffs used in the paper,
  and runs the retaliation experiment. This file is useful for verifying results and
  for generating additional plots during development.

BRF-9fig.jl:
  Plotting script for the BRF (structural gravity / full model) retaliation experiment.
  It reads precomputed welfare results from 9-graph-brf-2.csv and produces a combined
  3×3 (nine-panel) welfare plot (one panel per region vs. world).

combined-graph-ARM-BRF.jl:
  Comparison plot that overlays the unilateral US welfare results from the ARM (AvW)
  and BRF model runs. It reads US-uni-arm.csv and US-uni-brf.csv and plots US and world
  welfare as a function of the US tariff rate, marking the welfare-maximizing tariff in
  each model.

9-graph-brf-2.csv:
  Precomputed welfare results for the BRF coordinated retaliation experiment used by
  BRF-9fig.jl.
  

US-uni-arm-2.csv:
  Precomputed welfare results for the unilateral (no-retaliation) US tariff experiment
  under the ARM/AvW model.
  

US-uni-brf-2.csv:
  Precomputed welfare results for the unilateral (no-retaliation) US tariff experiment
  under the BRF model.
  

------------------------------------------------------------
Notes for replication:
  1) Install required Julia packages:
       MPSGE, NamedArrays, DataFrames, CSV, Plots
  2) Some scripts contain absolute file paths (e.g., to /Users/suraiya/...).
     Update those paths (or place the CSVs in the same directory and use relative paths)
     before running BRF-9fig.jl and combined-graph-ARM-BRF.jl.
