Regionalization vs globalization: what is the future direction of trade? 

International trade is a key source of economic prosperity. That is why policymakers, economists and business leaders alike are concerned by recent developments. The COVID-19 crisis, as well as the Suez Canal blockage, add to a series of troubling events: rising protectionism and the China-US trade war to name just two. These developments have contributed to new discussions about the shortening of supply chains and building up resilience.

Some analysts and commentators see a wave of deglobalization forging ahead. The Economist predicted on 24 January 2019: “The new world will work differently. Slowbalisation will lead to deeper links within regional blocs. Supply chains in North America, Europe and Asia are sourcing more from closer to home. In Asia and Europe most trade is already intra-regional, and the share has risen since 2011.” And in the Financial Times, Martin Wolf wrote on 10 December 2020: “The plausible future is not that international exchange is going to die. But it is likely to become more regional and more virtual.”

Based on a careful analysis of merchandise trade data, we examined the extent to which data supports such projections. To do so, we established three indicators that reveal regionalization: the share of global trade between nations on the same continent; the share of global trade between nations featuring a common border; and the average trade-weighted geographic distance of global trade. All three signal a trend towards regionalization if they increased recently.

First, we considered historical trade data provided by Michel Fouquin and Jules Hugot, which spans back to the year 1825 (Figure 1: Tests for Regionalization in Historical Trade Data, 1815-2014). We observed that about a fifth of global trade occurs between neighbouring countries, roughly 60% among countries on the same continent. And the trade-weighted average geographic distance in global trade is about 5,000km but fluctuates over time – with declines during major disruptions. For all three indicators, there is no evidence of regionalization in the years prior to 2014.

These numbers provide a historical benchmark. But what does more recent data show? For this, we utilized the latest data provided by Keith Head and Thierry Mayer (Figure 2: Tests for Regionalization in Contemporary Trade Data, 1950-2019). Notice that the historical data covers a somewhat different set of countries compared to the contemporary data-set, which makes cross-data-set comparisons difficult and emphasizes within-data-set analyses.

Nevertheless, the contemporary trade data also provides no support for the regionalization hypothesis. The shares of global trade among neighbouring countries and those on the same continent is remarkably stable. And the average geographic distance that goods travel has increased in the past 20 years, especially because of China’s integration into the global economy. This rise compensated for the decrease in the 1990s that largely reflects the opening up of Eastern European countries.

Why then do many commentators argue regionalization is happening? Thus far, we might conclude that you could see regionalization everywhere, except in the data. However, the latest data that captures the year 2020 is concerning. While this data is not yet available for all nations, we can consider import statistics from the EU. Patterns in Figure 3 (Tests for Regionalization in Latest EU-28 Trade Data, 2016-2021) were remarkably stable until the COVID-19 crisis hit in February of 2020. Notably, the share of imports from countries on the same continent increased and the average geographic distance of imports fell significantly (this is not a Brexit effect as we look at the EU-28). Pandemic-related disruptions have clearly hit international supply chains.

One may argue that any emerging patterns in regionalization statistics are suppressed by long-run existing trade flows. We thus repeat the analysis focusing on “new trade”, i.e. product level flows between countries that never traded such goods beforehand on a commercial level (we define it as minimum $1 million). Here, we find that the average trade-weighted distance has continuously fallen over the past five years and in 2019 reached its lowest level since the financial crisis of 2008. Also, the share of new trade between bordering countries saw a significant uptick in 2019 to 18% from the 9-10% seen over the past two decades. The jury is still out on whether these recent indications of regionalization are the first signs of a systematic shift.

The COVID-19 economic crisis has put further strain on globalization. It remains to be seen if companies and policymakers permanently change their behaviour after the recovery. Growth in merchandise trade had slowed down to a crawl prior to the pandemic, according to data from the CPB World Trade Monitor. This slowdown in global trade growth appears to conflict with a contemporaneous trend: the number of trade agreements has surged in recent years.

According to the World Trade Organization, the cumulative number of regional trade agreements in force increased from less than 100 in the year 2000 to almost 500 today. Most of world trade now takes place between country pairs that have established a reciprocal trade agreement. Recently enacted trade agreements by the EU as well as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership suggest this number to increase further. Such agreements often facilitate trade between nations located in different geographic regions. Yet, the average geographic distance of free trade agreement partners has been remarkably stable at around 4,800km since the year 2000.

While the global financial crisis permanently reduced the speed of globalization, the world economy went into the COVID-19 crisis already with largely stagnant global trade volumes. The pattern of regionalization which now appears in the latest trade statistics enriches the debate and offers researchers new metrics to quantify the scope of regionalization.