Don’t overestimate shifts from globalization to regionalization

The COVID-19 pandemic and the conflict in Ukraine are widely seen as accelerating an ongoing shift towards more regionalized trade patterns. Yet data on actual changes in merchandise trade flows—at least up to 2021—do not provide robust evidence of rising regionalization. The growth in long-distance trade (which primarily takes place between different regions) has outpaced that of short-distance trade (which primarily occurs within regions).

Why do many expect trade flows to become more regionalized? Reports of a trend towards regionalization and companies nearshoring supply chains reach back as far as 10–15 years. More recently, resilience imperatives, geopolitical tensions, regional trade agreements, technologies such as automation and 3D printing, and environmental concerns have further fuelled interest in producing goods closer to the company’s customers.

Against this backdrop, a McKinsey & Company study of January 2019 reported that “value chains are becoming more regional and less global”, and The Economist stated that “as globalisation fades, the emerging pattern of cross-border commerce is more regional.” Although these publications picked up on a measurable shift in trade patterns, the changes they identified do not amount to an ongoing trend of rising trade regionalization.

The increase in trade regionalization that became apparent a few years ago was neither sustained nor robust to alternative ways of dividing up the world into regions. If we divide the world into seven regions using the World Trade Organization’s (WTO region classifications, we find that a rising share of trade took place within regions between 2012 and 2016, as shown in the figure above. However, that growing trend ended in 2016 and disappears entirely if we divide up the world into regions using alternative classification schemes (3 mega-regions, 6 continents, or 22 smaller regions, see maps below).

Since all methods of classifying countries into regions involve subjective judgment–even the appropriate number of regions to use is an arbitrary choice–we opt to track changes in global trade patterns using the average distance traversed by international trade flows. Regional trade normally takes place over shorter distances, so we would expect a shift towards more regionalized trade patterns to be accompanied by a decline in the average distance traversed by trade flows. Instead, we find the opposite pattern, illustrated in the figure below. Trade has tended to take place over longer rather than shorter distances.

It is especially surprising that trade flows took place, on average, over longer distances in 2020 and 2021, since many expected that the pandemic and the associated supply disruptions would have intensified reliance on nearby sources of imports. Even though some buyers did switch to closer suppliers, especially for time-sensitive products, the dominant pattern, as we report in the DHL Global Connectedness Index 2021 Update, was greater reliance on goods produced in Asia. Exports from Asia surged while import growth was more evenly distributed around the world. As a result, countries in Asia imported over shorter distances while countries located far away from Asia imported over longer distances.

These data do not support the idea of an ongoing trend towards more regionalized trade patterns. But are we likely to see such a trend emerge moving forward? There are indeed several reasons why trade flows might become more regionalized. Supply chain reconfiguration can take several years, and regional trade can contribute to resilience by cutting transit times and reducing cross-region interdependencies. Moreover, regional trade has long been especially attractive because of the stronger non-geographic links that typically connect proximate compared to distant countries. Countries in the same region are five times more likely to be linked by a trade agreement and nearly three times more likely to share a common official language1.

A transformational shift from global to regional trade, however, is unlikely for three main reasons. First, trade flows are already highly regionalized. Using the seven WTO regions discussed earlier, 54 per cent of trade already takes place within regions. That is about three times higher than one would expect in a hypothetical world, where distance and cross-country differences have no dampening effects on trade. The intra-regional share of trade according to this classification fell from 57 per cent in 2003 to 51 per cent in 2012. It would hardly be a major transformation if this share goes back up to 57 per cent or reaches two-thirds of total trade.

Second, many of the factors that have spurred the growth of long-distance trade continue to open up opportunities. Long-distance trade still contributes to specialization and scale economies, and provides access to inputs that may be scarce or even unavailable in a given region. Long-distance trade can also contribute to resilience, expanding the variety of sources countries can access to produce essential goods.

Third, the present conditions of high inflation and input shortages mean that companies in most industries will have to look near and far for the most efficient and reliable production and sourcing locations. Shipping costs are falling, easing the recent cost penalty for long-distance trade. Government budgets are increasingly stretched, implying that public funds will only be used for the relocation of the most strategic supply chains.

To conclude, the question whether trade flows will become more regionalized moving forward remains open, but we must keep in mind that trade flows are already highly regionalized and that there is no robust evidence yet of an ongoing trend towards increased regionalization. As a result, the most prudent path forward in business and public policy is to plan for a future in which both global and regional trade continue to be central features of the international environment.