COMEX

Overview

COMEX (Comércio Exterior - Foreign Trade) dataset provides Brazil’s official international trade statistics extracted from Siscomex, the Integrated System of Foreign Trade maintained by the Brazilian government.

This dataset captures:

Export data: Brazilian goods leaving the country, disaggregated by municipality and product
Import data: Foreign goods entering Brazil, disaggregated by municipality and product
Monthly frequency: High-frequency trade data for detailed temporal analysis
Product classification: Detailed product codes and descriptions
Geographic coverage: Trade flows identified by Brazilian municipality
Long historical coverage: Available from 1989 onwards

COMEX is the primary official source for Brazil’s international trade statistics, widely used for trade policy analysis, business intelligence, academic research, and economic monitoring.

Data Source and Coverage

COMEX data comes from: - Official records from Siscomex (Brazil’s foreign trade system) - Mandatory declarations by exporters and importers - Updated monthly with current month data - Historical data from 1989 onwards

Important note on nomenclature: From 1989 to 1996, Brazil used a different system of product nomenclature (NBLC - Nomenclatura Brasileira de Mercadorias). All conversions to the current nomenclature system are available and the package handles this transparently.

For more information, visit the Brazilian Ministry of Productivity, Employment and Foreign Trade.

Available Datasets

1. export_mun (Exports by Municipality)

Export data disaggregated at the municipality level.

Coverage: All Brazilian municipalities engaged in international trade
Frequency: Monthly
Time period: 1989 onwards
Key variables: Export value (USD), quantity, product code, municipality, date
Use cases:
- Identify which municipalities are export hubs
- Analyze export diversification by region
- Track geographic shifts in export capacity
- Municipal-level trade policy impact assessment

2. import_mun (Imports by Municipality)

Import data disaggregated at the municipality level.

Coverage: All Brazilian municipalities receiving imports
Frequency: Monthly
Time period: 1989 onwards
Key variables: Import value (USD), quantity, product code, municipality, date
Use cases:
- Understand which regions import specific products
- Analyze import dependency patterns
- Track geographic consumption patterns
- Regional supply chain analysis

3. export_prod (Exports by Producer)

Export data organized by producer/exporter and product.

Coverage: All registered exporters in Brazil
Frequency: Monthly
Time period: 1989 onwards
Key variables: Export value (USD), quantity, product code, exporter code, date
Use cases:
- Firm-level export analysis
- Identify major exporters and their product mix
- Export concentration analysis
- Exporter persistence and dynamics

4. import_prod (Imports by Producer)

Import data organized by importer/distributor and product.

Coverage: All registered importers in Brazil
Frequency: Monthly
Time period: 1989 onwards
Key variables: Import value (USD), quantity, product code, importer code, date
Use cases:
- Firm-level import behavior
- Supply chain relationships
- Importer concentration analysis
- International sourcing patterns

Function Parameters

1. dataset

Selects which trade dataset to download.

dataset = "export_mun"   # exports by municipality
dataset = "import_mun"   # imports by municipality
dataset = "export_prod"  # exports by producer/exporter
dataset = "import_prod"  # imports by producer/importer

2. raw_data

Controls whether to download the original data or the processed/cleaned version.

TRUE: Returns raw data exactly as published by Siscomex
FALSE: Returns treated data with standardized formatting, English variable names, and cleaned values

raw_data = FALSE  # logical

3. time_period

Specifies which year(s) to download. Available from 1989 onwards.

time_period = 2020              # single year
time_period = c(2018, 2020)     # specific years
time_period = 2015:2020         # range of years

Note: Monthly data means each year can be quite large. Consider downloading specific years or ranges to manage file size.

4. language

Output language for variable names and documentation.

"pt": Portuguese
"eng": English

language = "eng"  # character string

Examples

# download treated exports data by municipality from 2020 to 2021
data <- load_br_trade(
  dataset = "export_mun",
  raw_data = FALSE,
  time_period = 2020:2021,
  language = "eng"
)

# download treated imports data by municipality from 2020 to 2021
data <- load_br_trade(
  dataset = "import_mun",
  raw_data = FALSE,
  time_period = 2020:2021,
  language = "eng"
)

Data Notes

Raw vs. Treated Data

Raw data (raw_data = TRUE): Original Siscomex format, potentially with inconsistencies and naming conventions from different time periods
Treated data (raw_data = FALSE): Standardized with English variable names, consistent units (USD for values), and cleaned formatting

Product Classification

1989-1996: Uses NBLC (Nomenclatura Brasileira de Mercadorias) - conversions are handled transparently
1997 onwards: Uses HS (Harmonized System) classification aligned with international standards
Product codes enable comparison with international trade databases

Data Characteristics

Monthly frequency: Data is reported monthly; aggregation to annual or quarterly is straightforward
Producer vs. Municipality:
- Municipality data groups trade by geographic origin/destination
- Producer data groups by firm/exporter-importer code
- Use municipality for regional analysis, producer for firm analysis
Missing data: Some small trade flows may not be reported
Currency: All values in USD

Nomenclature Conversion

When using data spanning 1989-1996 to 1997 onwards, be aware: - Product categories may differ between nomenclature systems - Conversions are available but not always 1:1 - Compare very old with recent data with caution