Why Do Trade Numbers Differ? Unraveling the Mystery Behind COMTRADE and BACI

You’re deep into your research on global trade flows, tracking the rise of a particular product. You decide to cross-check the data between different platforms—only to find that the numbers don’t quite add up. What’s going on here? This is a puzzle that many trade data users encounter, and the answer lies in the way different datasets handle the complexities of global trade.

A PhD researcher from Norway recently faced this dilemma. While comparing Norway’s trade data with China’s figures, he noticed discrepancies that left him puzzled. “Could you please let me know your source of data for trade (imports and exports)?” he asked us.

To resolve this question, let’s dive into the core differences between two key datasets: COMTRADE and BACI.

At first glance, COMTRADE and BACI might seem like two sides of the same coin, but the way they handle data is different. Understanding these differences is key to making sense of the numbers.

1. Data Collection and Reporting:

• COMTRADE: Managed by the United Nations, COMTRADE is like a collection of puzzle pieces, each representing a country’s official trade data. Countries report data based on their own systems, which can result in mismatches. For example, in 2022, China reported $7.9 billion in imports from Norway, while Norway reported only $5.3 billion in exports to China—illustrating how varied practices lead to conflicting data.

• BACI: Developed by CEPII (Centre d’Études Prospectives et d’Informations Internationales), BACI reshapes data by harmonizing discrepancies. For Norway and China’s 2022 trade, BACI reports $5.64 billion, using historical patterns and cross-country comparisons. It employs “mirror statistics,” filling gaps in disaggregated data with partner country reports, reducing missing values, particularly from non-reporting countries or specific product flows.

2. Harmonization: The BACI Advantage:

• COMTRADE: While COMTRADE provides a raw, unfiltered view of each country’s reported trade data, it doesn’t reconcile discrepancies between what one country claims to trade and what its trading partner reports. This can lead to mismatched puzzle pieces—like in 2022, when China’s reported imports from Norway exceeded Norway’s reported exports by $2.6 billion.

• BACI: BACI acts as a meticulous puzzle master, aligning these mismatches by considering both sides of a trade flow. It adjusts the data to provide a balanced view, ensuring that the puzzle pieces fit together more accurately. BACI’s adjusted figure of $5.64 billion for Norway-China trade in 2022 exemplifies how it smooths out these discrepancies.

3. Long-Term Analysis:

• COMTRADE: The pieces of the COMTRADE puzzle can shift over time, making it challenging to see the big picture. Changes in reporting standards or practices can distort long-term trends.

• BACI: BACI offers a more stable and consistent dataset, making it easier to track long-term trends in global trade. This stability is crucial for anyone analyzing the evolution of trade relationships over time or studying the shifts in the economic complexity of countries.

4. Coverage and Detail

• COMTRADE gives you the raw pieces with all their imperfections, offering detailed data but with inconsistencies depending on the country.

• BACI smooths out the edges, providing a more coherent and reliable dataset, particularly valuable for historical analysis.

However, even BACI has its quirks—its harmonization algorithm can sometimes overemphasize exports from smaller nations, like small island countries. This is something to keep in mind when using the data—or when playing TRADLE.

At the OEC, we aim to provide the clearest view of global trade patterns, especially for analyzing trends and economic complexity over time. That’s why we rely on BACI. Here’s what makes BACI stand out:

1. Consistency Across Time and Borders:

• BACI’s harmonization process ensures that the data is consistent, making it easier to compare trade flows or economic complexity between countries and over time. This consistency is vital for anyone conducting in-depth research on global trade patterns.

2. Enhanced Reliability:

• BACI reduces the noise and potential errors in trade data by reconciling discrepancies. For example, if Norway reports exporting more to China than China reports importing, BACI’s algorithm adjusts the data to reflect a more accurate trade flow, making BACI a more reliable source for analysis.

3. Suitability for Historical Analysis:

• Historical trade data needs to be stable and comparable across different periods. BACI’s standardized approach allows researchers to analyze long-term trends without worrying about inconsistencies that can arise from raw data.

4. Global Coverage with a Balanced View:

• While BACI’s methodology is robust, it has its quirks, and the harmonization process can occasionally skew results. Despite this, BACI remains our choice for a coherent and consistent understanding of historical global trade.

Navigating the differences between datasets can be challenging, especially when discrepancies arise. As a PhD researcher from Norway discovered, these variations can be significant. By acknowledging each dataset’s strengths and limitations, researchers can gain a more accurate perspective on global trade flows. The OEC’s use of BACI underscores our commitment to providing reliable historical trade data, ensuring that users can trust the information for their critical work—while being mindful of its inherent limitations.