Data Sources

pydeflate currently supports three authoritative data sources: IMF, World Bank, and OECD DAC. This guide helps you choose the right source for your use case.

Source Comparison

IMF World Economic Outlook

Overview

The International Monetary Fund's World Economic Outlook (WEO) database provides economic indicators with estimates for future years.

Data URL: https://www.imf.org/en/Publications/WEO

Available Functions

from pydeflate import (
    imf_gdp_deflate,    # GDP deflator
    imf_cpi_deflate,    # CPI (period average)
    imf_cpi_e_deflate,  # CPI (end-of-period)
    imf_exchange        # Exchange rates
)

Strengths

• Future estimates: Includes ~5 year forecasts for GDP deflators and CPI
• Comprehensive: Covers ~190 countries with consistent methodology
• Frequent updates: Published twice yearly (April, October)
• Economic focus: Ideal for macroeconomic analysis

Limitations

• Estimates subject to revision
• Exchange rates are derived based on WEO data (instead of published directly as an indicator)

When to Use IMF

Use IMF data when you need:

• Future year estimates

  # Project 2024-2026 aid budgets to constant 2023 prices
  imf_gdp_deflate(df, base_year=2023, ...)
  

• Consistency with IMF publications

  # Ensure compatibility with IMF reports
  imf_cpi_deflate(df, base_year=2020, ...)
  

• End-of-period CPI for financial reporting

  # December year-end adjustments
  imf_cpi_e_deflate(df, base_year=2018, ...)
  

Data Characteristics

• GDP deflator: Broad measure of price changes in the economy
• CPI (period average): Average monthly CPI across the year
• CPI (end-of-period): December CPI value
• Exchange rates: Yearly average, derived from WEO data

World Bank

Overview

The World Bank provides a comprehensive collection of development indicators spanning decades.

Data URL: https://databank.worldbank.org/

Available Functions

from pydeflate import (
    wb_gdp_deflate,         # GDP deflator
    wb_gdp_linked_deflate,  # Linked GDP deflator
    wb_cpi_deflate,         # CPI
    wb_exchange,            # Market exchange rates
    wb_exchange_ppp         # PPP exchange rates
)

Strengths

• Deep history: Data from 1960 onwards
• Broad coverage: ~200 countries and territories
• Linked deflators: Includes estimates for some gaps in the series.
• PPP rates: Purchasing power parity comparisons
• Development focus: Tailored for development indicators

Limitations

• No future estimates
• Annual updates
• Some historical gaps for smaller countries

When to Use World Bank

Use World Bank data when you need:

• Long time series (20+ years)

  # Analyze trends from 1980-2020
  wb_gdp_linked_deflate(df, base_year=2010, ...)
  

• PPP comparisons for living standards

  # Compare GDP per capita across countries
  wb_exchange_ppp(df, source_currency="LCU", target_currency="USA", ...)
  

• Consistency with World Bank data

  # Match World Development Indicators
  wb_gdp_deflate(df, base_year=2015, ...)
  

• GDP series with some gaps filled

  # Handle breaks in deflator series
  wb_gdp_linked_deflate(df, base_year=2015, ...)
  

Data Characteristics

• GDP deflator: Annual percentage change, standard series
• Linked GDP deflator: Annual percentage change, with some gaps filled
• CPI: Consumer price index (period average)
• Exchange rates (market): From IMF IFS, yearly average
• Exchange rates (PPP): Purchasing power parity conversion factors

OECD Development Assistance Committee

Overview

The OECD DAC provides specialized deflators and exchange rates for Official Development Assistance (ODA), maintained specifically for aid flow analysis.

Data URL: https://www.oecd.org/dac/

Available Functions

from pydeflate import (
    oecd_dac_deflate,   # DAC deflator
    oecd_dac_exchange   # DAC exchange rates
)

Strengths

• ODA-specific: Designed for development assistance analysis
• DAC standard: Official methodology for aid reporting
• Consistency: Aligned with OECD.Stat and CRS database
• Donor focus: Optimized for donor country reporting

Limitations

• Limited to ODA context (not suitable for general economic analysis)
• Smaller country coverage (donors)
• Single deflator type (no separate CPI)

When to Use OECD DAC

Use DAC data when working with:

• Official Development Assistance

  # Convert ODA to constant prices
  oecd_dac_deflate(df, base_year=2018, ...)
  

• DAC reporting requirements

  # Ensure compliance with DAC standards
  oecd_dac_deflate(df, base_year=2020, ...)
  

• Creditor Reporting System (CRS) data

  # Match CRS deflation methodology
  oecd_dac_exchange(df, source_currency="LCU", target_currency="USA", ...)
  

DAC Codes

DAC uses its own country coding system. Enable with use_source_codes=True:

# DAC donor codes (examples):
# 4   = France
# 12  = United Kingdom
# 302 = United States
# 701 = Japan

data = {
    'dac_code': [4, 12, 302, 701],
    'year': [2015, 2016, 2017, 2018],
    'oda': [11400, 18700, 30900, 15500]
}

result = oecd_dac_deflate(
    data=pd.DataFrame(data),
    base_year=2018,
    id_column="dac_code",
    use_source_codes=True,  # Use DAC codes
    ...
)

Data Characteristics

• DAC deflator: Composite deflator for aid flows
• Exchange rates: Yearly average rates used in DAC reporting

Choosing the Right Source

Decision Tree

┌─ Need future estimates?
│  └─ YES → Use IMF
│
├─ Working with ODA data?
│  └─ YES → Use OECD DAC
│
├─ Need PPP exchange rates?
│  └─ YES → Use World Bank (wb_exchange_ppp)
│
├─ Time series > 20 years?
│  └─ YES → Use World Bank (wb_gdp_linked_deflate)
│
└─ General economic analysis?
   └─ Use IMF or World Bank

By Use Case

Comparing Sources

You can compare results across sources:

from pydeflate import imf_gdp_deflate, wb_gdp_deflate, set_pydeflate_path
import pandas as pd

set_pydeflate_path("./data")

# Test data
data = {
    'country': ['USA', 'GBR', 'FRA'],
    'year': [2015, 2016, 2017],
    'value': [1000, 1000, 1000]
}
df = pd.DataFrame(data)

# Apply both sources
imf_result = imf_gdp_deflate(
    df, base_year=2015, source_currency="USA", target_currency="USA",
    id_column="country", value_column="value", target_value_column="value_imf"
)

wb_result = wb_gdp_deflate(
    df, base_year=2015, source_currency="USA", target_currency="USA",
    id_column="country", value_column="value", target_value_column="value_wb"
)

# Merge and compare
comparison = df.merge(
    imf_result[['country', 'year', 'value_imf']],
    on=['country', 'year']
).merge(
    wb_result[['country', 'year', 'value_wb']],
    on=['country', 'year']
)

comparison['difference'] = comparison['value_imf'] - comparison['value_wb']
print(comparison)

Data Update Schedule

pydeflate caches data locally and checks for updates. If cached data is >50 days old, you'll see a warning to refresh.

Data Attribution

pydeflate relies on publicly available data from:

• IMF World Economic Outlook: https://www.imf.org/en/Publications/WEO
• World Bank DataBank: https://databank.worldbank.org/
• OECD DAC: https://www.oecd.org/dac/

This data is provided based on the terms and conditions set by the original sources.

Next Steps

• Deflation Guide - Apply deflators from each source
• Currency Exchange - Use exchange rates from each source
• Advanced Topics - Register custom data sources