Methodology

This page explains how Global Economic Indicators collects, processes, and presents economic data. Understanding our methodology helps you interpret the charts and figures on the platform with appropriate context and confidence.

Data Collection

Our platform retrieves data from multiple institutional APIs at the moment a user visits a page. This approach ensures that the information displayed reflects the most recent data available from each source. The primary data retrieval flow works as follows:

1. The client-side application requests data from our Next.js API routes, which act as proxy endpoints.
2. These server-side routes forward the request to the upstream data provider (e.g., World Bank, FRED) and handle CORS, rate limiting, and error recovery.
3. The response is parsed, normalized into a consistent format, and returned to the client.
4. The client stores the response in a local cache with a 24-hour time-to-live to minimize redundant API calls.

For indicators that lack reliable free APIs, we maintain curated static datasets compiled from official institutional reports and publications. These fallback datasets are versioned and updated periodically. Examples include venture capital funding data (sourced from Crunchbase and Dealroom public reports), AI patent statistics (from WIPO annual reports), and digital economy adoption rates (from World Bank Findex and eMarketer publications).

API Sources and Indicator Codes

Each metric displayed on the platform maps to a specific indicator code from its source database. This ensures traceability and allows users to verify values against the original source. The table below summarizes key sources and their indicator categories.

Data Normalization

Different sources report data in different formats, units, and time granularities. We normalize all data into a consistent structure before display:

• Country naming: All sources use different country identifiers (ISO-2, ISO-3, full names). We standardize to readable names (e.g., “SouthKorea”, “USA”, “UK”) used consistently across the platform.
• Time aggregation: Monthly or quarterly data is aggregated to annual averages when displayed alongside annual indicators, ensuring consistent time-series comparisons.
• Unit consistency: Percentages, per-capita figures, and absolute values are clearly labeled. Currency values are reported in current U.S. dollars unless otherwise noted.
• Missing values: Where a country does not report a value for a given year, that data point is omitted from charts rather than interpolated or estimated.

Caching Strategy

To balance data freshness with performance, we implement a two-tier caching system. Server-side API routes cache responses from upstream providers for a configurable period, reducing the number of outbound API calls. On the client side, fetched data is stored in the browser's localStorage with a 24-hour time-to-live and a version identifier. When we release updates that change data structures or add new indicators, incrementing the cache version forces all clients to fetch fresh data on their next visit.

Composite Scores and Rankings

Some features, such as the Innovation Ranking Table and the Technology Heatmap, compute composite scores by normalizing individual metrics to a 0–100 scale and applying weighted averages. The normalization uses min-max scaling across all countries for the selected year. Weights are assigned based on the relative importance and data quality of each indicator. These weights are documented in the source code and can be reviewed in our technology indicators data file.

Limitations and Disclaimers