Deglobalization: Reasons and Effects

1 Introduction

From the end of World War II onwards, the volume of international trade in goods and services outgrew the value of global output as measured by gross domestic product (GDP). In particular, from the early 1990s to 2008, global goods trade grew by 2.4 % points per year more than GDP. In 2011, Harvard economist Dani Rodrik has famously used the term hyperglobalization to describe this process of rapid economic integration, implying that the process would not be sustainable.

2 Globalization, Hyperglobalization and Slowbalization

Indeed, at least since the global financial crisis of 2008/09 the world is in a different regime. Trade no longer outperforms GDP. From 2008 to 2018, there was no growth premium of goods trade over GDP anymore. Instead, the period saw sharp swings, driven by volatility of commodity prices and currencies. In 2019, the British magazine The Economist has used the term “slowbalization” to describe this situation.^[1] But importantly, services trade behaved differently. It tended to grow faster than GDP. The most recent trade data, however, suggest that the period of slowbalization might now come to an end with both services and goods trade expanding at a smaller rate than GDP.

Basic trade theory suggests that, absent any changes in trade costs and in the structure of consumers’ preferences, global trade should grow approximately at the average rate of growth of countries’ GDPs (corrected for changes in the dispersion of GDPs). Evidence shows that the period of hyperglobalization was marked by significant decreases in applied tariff rates and improved market access in many countries, brought about by the conclusion of the Uruguay Round, the creation of the World Trade Organization, and a boom in bilateral free trade agreements. Moreover, many countries, first and foremost China, engaged in internal reforms that allowed increasing segments of their economies to be active on world markets. There is also strong evidence that information and communication technologies have reduced trade costs significantly, e.g. through the digitization of logistics.

However, tailwinds turned into headwinds after the global financial crisis. According to the World Trade Organization, the number of new trade agreements concluded has been falling steadily from 21 in 2008 to 2 in 2022. So, policy barriers have probably not gone down; physical trade barriers – e.g. resulting from transportation of goods or from traders’ needs to communicate in order to strike deals – probably have continued to come down. In sum, it is likely that the speed at which trade costs have fallen has slowed down. As to consumer preferences, as per capita incomes of countries grow, their demand structure shifts from basic needs such as food and housing to manufactured goods and on to services. While manufactured goods are relatively easy to trade across borders, services are less readily tradeable. So, when the share of demand falling on services income goes up as countries grow, the ratio of trade over GDP falls. When the share of countries maturing into higher services shares in demand exceeds that of countries moving from basic needs into manufactured tradeables, this, too, can explain slowbalization – at least until technological and political breakthroughs make services more easily tradeable.

Deglobalization could be meaningfully defined as a situation in which international transactions grow at a slower rate than domestic ones, leading to a decline in the trade-GDP ratio. To the extent that such a development is driven by changes in the demand structure or in the dispersion of countries’ GDPs, this need not be worrying. It would be driven by market forces, not by government interference.

3 The Rise of Trade Impediments

The empirical evidence on discriminatory cross-border measures is very clear. Since 2009 the number of protectionist measures has grown substantially. This can be shown with the help of data assembled by the Global Trade Alert project at University of Sankt Gallen, Switzerland (Global Trade Alert 2023). Also, the latest World Trade Report of the WTO leaves no doubt that protectionism is on the rise. Export restrictions, a particularly dangerous type of measure, have made a massive return in the face of rising geopolitical tensions and shortages of critical goods and inputs.

When governments intervene with discriminatory measures at or behind the border, there is a risk that such policies reduce domestic welfare. The motivation behind such measures can be very diverse. In earlier centuries, governments imposed taxes on foreign goods – import duties – with the objective to raise money. This motivation is still of some relevance, e.g. at the EU-level, where tariff income is a substantial part of the overall revenue. Raising money by using tariffs is tempting to many governments because a fraction of the economic burden of those taxes is effectively borne by foreigners. Other political trade barriers, such as quantitative restrictions or discriminatory regulation, do not raise government revenue but they still protect domestic producers from foreign competition, thereby increasing profits while damaging consumer interests, where the latter effect outweighs the former. Protectionist measures can also take the form of export subsidies; then, they may increase the amount of trade between two countries above the “natural” level. While all the mentioned policies generally harm national welfare in competitive economies, they can be associated with higher or lower trade-over-GDP ratios than what an undisturbed equilibrium would bring about. In practice, however, the large majority of trade liberalization episodes led to more trade and, according to the results of standard trade models, to higher welfare.

Newer studies stress the importance of uncertainty over trade policy as an important and particularly harmful trade barrier. It operates like a non-tariff barrier, even before any possible or probable measures are even implemented. Because the establishment of trade relationships is almost always associated with some sort of sunk investment on either side of the partnership, uncertainty unambiguously reduces trade. So, it is well possible that one driver of slowbalization and potentially of deglobalizaton is uncertainty over trade policy rather than factual policy itself.

4 Industrial Policy and Internalization of Externalities as Trade Barriers

Next to the generation of government revenue, another empirically important motivation for protection that has appeared again and again in history is the pursuit of industrial policy objectives, the idea being that infant industries can only develop when they are, at least temporarily, sheltered from foreign competition. Current legislation, both in the EU and the US, goes into this direction. It strives to favor domestic inputs in electrical vehicles, semiconductors or other products deemed of strategic importance.

However, the return of industrial policy is not only driven by the wish to attract economic rents from abroad but also by the fear that foreign monopolies in key industries could be exploited by hostile foreign governments to extract political concessions. Whether in the form of de-coupling or its more benign sister called de-risking, such policies do reduce trade with the directly affected trade partner; they can also create new trade with alternative partners (trade diversion), but in sensibly parameterized simulation studies trade destruction always dominates trade diversion at the aggregate level so that the overall impact on global trade volumes is negative. Clearly, geopolitical tensions have a strong potential to lead to deglobalization. The international division of labor implies specializing production on certain sectors of activity, while requiring imports in others. If there is not enough trust, that trade partners will continue to cooperate, a strong reliance on trade might appear risky to policymakers who respond by gradually introducing disincentives for trade.

Trade barriers introduced for security reasons are likely to shrink GDP, but less than trade. Consequently, they would tend to yield deglobalization. Whether such policies also shrink national welfare depends on details. When they are well-targeted to internalize an international security externality that arises from uncoordinated corporate business decisions, they can end increasing welfare, in particular when they do indeed reduce the likelihood of supply disruptions.

A final motivation for border measures derives from governments’ wishes to internalize environmental externalities or to enforce human rights conventions by means of trade policies. The case of transboundary pollution is relatively straightforward. E.g., there is no doubt that CO₂-emissions associated with the international transportation of goods should be subject to a carbon price. Failing to do so amounts to a subsidization of international trade over domestic transactions, leading to hyperglobalization stricto sensu. Various simulation studies show that imposing a global CO₂-price would reduce the trade-to-GDP ratio. But if the tax on emissions is well calibrated it would increase welfare. In the case of local environmental pollution, the situation is more complicated. Trade barriers cannot be easily justified on welfare-theoretic grounds when there are no objectively measurable physical spill-overs between countries. Such policies can be seen as invasive and overreaching. Similarly, while one can defend restrictive trade policy measures targeting universally accepted human rights or social standards, the defense of “Western” values through trade policy is highly problematic.

Again, such measures tend to have in common that they depress cross-border trade. In many cases they also lower GDP as defined in national accounting. Nonetheless, under certain circumstances, they can increase welfare. But whenever externalities are used as a defense of restrictive trade barriers, one should take Amartya Sen’s warning contained in his famous 1970 paper “The Impossibility of a Paretian Liberal” very seriously. In this paper, Sen argued against the abuse of the externality argument to limit the degrees of freedom of other agents, in our context, foreign governments. The danger here is not only that such behavior would be unethical but that foreign governments could retaliate by using similar policies.

5 Outlook

It is likely that adjusting trade policy to reflect economic security concerns of countries and to achieve consistency with environmental objectives will tend to lower the trade-GDP ratio. However, in the long term, technological changes will continue bringing down trade costs, mostly in the area of services. When the geopolitical environment becomes more benign again, globalization may very well see a massive new boost.