GitHub

National Module

The national module has two aggregations available:

  • summary
  • detailed

The summary aggregation is loaded directly from the summary tables provided by the BEA. The detailed aggregation is an extrapolation of the summary tables using the detailed tables.

Build

Download Data

The following code will download the necessary data from the BEA website. It only needs to be run once.

sut_files = fetch_supply_use()

Subsequent runs can pull the file names from the directory.

data_path = "data/national"
sut_files = joinpath.(pwd(), data_path, readdir(data_path))

Summary Data

The following code will build the summary-level tables, calibrate, save them to a file, load a specific year, and run a model. Assumes the data has been downloaded.

using WiNDC, MPSGE

data_path = "data/national"
sut_files = joinpath.(pwd(), data_path, readdir(data_path))

summary_data = build_national_table(sut_files; aggregation = :summary)

SD,M = calibrate(summary_data; lower_bound = .5, upper_bound = 1.5)
save_table("summary_data.jld2", SD)

year = 2023
summary_year = load_table("summary_data.jld2", year)

# Initialize the model and solve a benchmark
M = national_mpsge(summary_year);
solve!(M, cumulative_iteration_limit=0)

# Set a counterfactual, in this case `Output_tax` to 10%
set_value!.(M[:Output_tax], .1)

solve!(M)

generate_report(M)

Detailed Data

using WiNDC, MPSGE

sut_files = fetch_supply_use()

data_path = "data/national"
sut_files = joinpath.(pwd(), data_path, readdir(data_path))

detail_data = build_national_table(sut_files)

SD,M = calibrate(detail_data; lower_bound = .5, upper_bound = 1.5)
save_table("detailed_data_partial.jld2", SD)


detail_year = load_table("detailed_data_partial.jld2", 2017)

M = national_mpsge(detail_year);

solve!(M, cumulative_iteration_limit=0)

set_value!.(M[:Output_tax], .1)

solve!(M)

generate_report(M)

Disaggregated Detailed Data

This assumes both the summary and detailed data have been built and saved.

using WiNDC, MPSGE

#sut_files = fetch_supply_use()

data_path = "data/national"
sut_files = joinpath.(pwd(), data_path, readdir(data_path))

summary_data = load_table("summary_data.jld2")
detailed_partial = load_table("detailed_data_partial.jld2")
summary_map = WiNDC.detailed_summary_map(sut_files[1])

detailed_yearly = WiNDC.national_disaggragate_summary_to_detailed(detailed_partial, summary_data, summary_map)

DD, _ = calibrate(detailed_yearly; lower_bound = .01, upper_bound = 100)

save_table("detailed_data.jld2", DD)


detail_year = load_table("detailed_data_partial.jld2", 2017)

M = national_mpsge(detail_year);

solve!(M, cumulative_iteration_limit=0)

set_value!.(M[:Output_tax], .1)

solve!(M)

generate_report(M)