How to use IMF World Economic Outlook data for inflation research

The IMF World Economic Outlook (WEO) database is published twice a year — April and October — with an update in January and July. It contains actual and forecast data for macroeconomic variables across 190+ countries and country groups. For inflation research, the key series are:

• PCPIPCH: Inflation, average consumer prices — period average, percent change. This is the headline inflation series used in most cross-country comparisons.
• PCPIEPCH: Inflation, end of period consumer prices — useful when you need the December-over-December rate rather than the annual average.
• NGDP_D: GDP deflator — an alternative inflation measure covering the broader economy (not just consumer prices).

The WEO covers actuals (usually up to 1–2 years before the current date) and projections (up to five years ahead). Critically, it includes the IMF's own forecasts, which represent a consensus view from country desk teams with extensive local knowledge. The forecasts are widely followed as a benchmark for central bank and fiscal policy assessments. See our IMF WEO October 2026 analysis for the latest projections in context.

The WEO data is freely available. Three ways to get it:

1. WEO online database (web interface): Go to imf.org WEO database, select the edition (e.g., "October 2026"), then "By Countries (Country-level Data)" or "By Subject." Choose your countries, subjects (PCPIPCH for CPI inflation), and year range. Export as Excel or CSV.
2. Full WEO bulk download: On the same page, download the full dataset as a zipped tab-delimited text file. This is ~10MB and contains every series for every country — useful for bulk analysis. Load into Python (pandas) or R.
3. IMF Data API: The IMF provides a REST API at https://www.imf.org/external/datamapper/api/v1/ for programmatic access. Query syntax: /PCPIPCH/{country_code}?periods=2020:2027. Returns JSON. Useful for building dashboards. Our inflation dashboard guide includes a worked example.

Several interpretation pitfalls are common when working with WEO inflation data:

• Period-average vs end-of-period: PCPIPCH is the average of the 12 monthly readings for the year. This lags actual inflation trends — a country that had 12% inflation in January and 2% by December would report a ~7% annual average. PCPIEPCH (end-of-period) is more forward-looking but less commonly cited.
• Country classifications: The WEO groups countries as "Advanced Economies," "Emerging Market and Developing Economies," and sub-groups. Group aggregates are GDP-weighted averages — so the "Advanced Economies" inflation series is heavily influenced by the US and euro area.
• Revision history: The WEO revises historical data as national statistics offices update their series. If you are tracking IMF forecasts over time (e.g., did the April 2026 forecast turn out to be accurate?), you need the archived edition, not the current one. The IMF archives all WEO editions at the same URL.
• Comparison to World Bank WDI: The World Bank WDI also publishes annual CPI inflation (FP.CPI.TOTL.ZG) — the same series used by our inflation explorer. WEO and WDI use the same underlying national CPI data for actuals, but WEO extends further into the future with forecasts, while WDI provides more granular historical data and covers a broader set of supplementary indicators.

WEO data is particularly valuable for four types of analysis:

1. Tracking the global disinflation: Create a scatter plot of 2022 peak inflation vs 2026 inflation for advanced economies to identify which countries made the most and least progress. The WEO's regional groupings make this easy to segment.
2. Evaluating IMF forecasts vs outcomes: Download the October 2022 WEO projection for 2024 inflation, and compare to the October 2024 actuals. IMF inflation forecasts are systematically too low during inflation surges (as in 2021–22) — understanding this bias helps calibrate how to read current forecasts.
3. Building an inflation outlook model: Use WEO projections as a starting point (the IMF's "baseline") and then apply your own adjustments for country-specific factors. The inflation expectations literature suggests that professional forecast surveys (like Consensus Economics) and market-implied rates (breakevens) are more accurate than model-based forecasts at short horizons.
4. Combining with FRED for US detail: WEO gives you the cross-country picture; FRED gives you the US series at monthly frequency with full component detail. A common workflow is WEO for international context, FRED for US deep-dive.

When using WEO data in research, reports, or publications, use this citation format:

International Monetary Fund. World Economic Outlook Database, [Edition Month Year]. https://www.imf.org/en/Publications/WEO/weo-database/. Accessed [date].

Key caveats to note in any publication: (1) specify whether you are using actuals or projections; (2) note the edition used (different editions have different historical revisions); (3) for advanced economies, WEO CPI data is typically consistent with national statistics but may differ slightly from official national sources due to timing of revisions. Always cross-check critical data points against the primary national source (BLS, ONS, Eurostat, Rosstat, etc.). Our World Bank WDI guide covers similar methodology notes for the other major free database.