Does globalisation have a future? - Capital Economics
Emerging Markets Economics

Does globalisation have a future?

Emerging Markets Economics Focus
Written by Vicky Redwood
Cancel X

Globalisation has peaked, with any further major integration unlikely. In fact, a policy-driven period of de-globalisation is increasingly likely.

  • Globalisation has peaked, with any further major integration unlikely. In fact, a policy-driven period of de-globalisation is increasingly likely.
  • Flows of trade and foreign direct investment have flattened off as a share of nominal GDP over the past decade. This may partly be an after-effect of the global financial crisis. But even if that is the case, there are still several reasons why, trade wars apart, globalisation in its current form may have run its course.
  • First, all the major steps to integrate the global system have already been taken. Second, technological progress in advanced manufacturing techniques is making it cheaper for countries to produce domestically, instead of offshoring. This contrasts with many previous technological steps which, by reducing transport and communication costs, promoted globalisation. Third, complex supply chains have reached their limits. And fourth, China seems unlikely to open up its capital markets significantly.
  • Some pin their hopes on trade in services to drive the next wave of globalisation. And it is true that technology will help to make more services easier to trade. However, the liberalisation of services trade will remain inherently harder than that of goods, with countries continuing to find it hard to reach agreement on common standards. And the fastest growing areas of services are those which are unlikely ever to be traded, such as eating out or social care.
  • So old-fashioned globalisation has run its course. That said, flows of traded goods or capital will keep rising, just perhaps at a slower pace than the global economy. And even if the exchange of goods, capital and people grows at a slower pace, flows of ideas and information can keep growing rapidly. Moreover, to the extent that all this is partly driven by firms seeking more efficient ways of operating, this is a good thing.
  • More worrying is the growing risk of a malign form of de-globalisation driven by the deliberate erection of barriers by policymakers. This could reflect a worsening of the US-China trade war; current tensions reflect fundamental factors in the relationship between the US and China, which are unlikely to go away when President Trump leaves office.
  • Alternatively, the current trade war could be the start of a broader backlash against globalisation, as signalled by the rise of nationalist parties in many European countries.
    Such de-globalisation could take different forms. Perhaps the most immediate threat is that the current situation degenerates into a free-for-all, with a “beggar thy neighbour” rise in barriers, including tariffs, capital controls and migration limits.
  • Alternatively, we could see a move towards regionalisation, with the world splitting up into separate spheres of influence: a US-led bloc, a Chinese-led bloc and possibly a third European-led bloc. This might just result in a general trend of disengagement, but at the other extreme, it could lead to a technological iron curtain between blocs.

Does globalisation have a future?

This Focus is part of a series of reports about the end of globalisation and the impact of this on the global economy and financial markets. (See here for more.)

The trade war, as well as the spread of populist movements, has called into question the future of globalisation. In this Focus, we look at the recent trends in globalisation, where it was heading before the trade war came along, and what difference the trade war makes to its future.

Signs that globalisation has peaked

First of all, what is globalisation? Precise definitions vary, but fundamentally it refers to the increasing interconnectedness of national economies. In practice, this means the flow across borders of goods, services, labour, capital, knowledge and even values such as democracy.

Globalisation has been through a number of “waves” in the past, with the current one starting in the 1990s, spurred by the growth of low cost communications. (See Global Economics Focus, “What are the lessons from history about globalisation?” 8th October.) And in some respects, this wave is still going strong. The number of internet users remains on a sharp upward trend. (See Chart 1.) Cross-border data flows are rising rapidly. And the number of people visiting other countries is still rising rapidly; in 2017, the number of tourist trips to other countries rose by 7.3% compared to 2016.

Chart 1: % of World Population Using the Internet In the Past 3 Months

Source: UN’s International Telecommunication Union

In many other respects, though, the latest wave of globalisation seems to be reaching a natural limit. First, nominal exports of goods and services have flattened off as a share of global GDP over the past decade. (See Chart 2.) In real or volume terms, the share has kept rising (as the price of traded goods and services has fallen relative to the price to non-tradeables). But it has done so at a slower pace than seen during most of the 1990s and 2000s.

Chart 2: Exports of Goods & Services (As a % of GDP)

Source: WTO

Second, capital flows as a share of global GDP also appear to have flattened off. Interbank lending flows have, unsurprisingly, fallen since the financial crisis in the face of tighter regulation. But Chart 3 shows that flows of foreign direct investment (FDI) and short-term portfolio flows (of equities and bonds) have levelled off as a share of global GDP too.

Chart 3: Private Capital Inflows (As a % of Global GDP))

Source: WTO

And third, migration appears to have stalled, albeit only very recently. Up until 2015, the stock of international migrants (people moving abroad for at least a year) was rising relative to the world’s population. But it has since been flat, indicating that the flow of migrants in the past couple of years has been zero.

There are some misconceptions about these trends. One is that they reflect the recent trade war. But the charts above suggest that globalisation peaked about a decade ago. Another is that the slowdown in trade reflects the convergence of unit labour costs between China and the West. But that would not explain why global trade has slowed, given that production could just shift to other, cheaper countries.

It is also tempting to blame all of this on the prolonged after-effects of the global financial crisis. The impact at the time of the crisis is clearly visible in Charts 2 and 3; global trade and capital flows dipped sharply as a share of GDP, given that the crisis disproportionately hit highly-traded categories like investment and durable goods. But there may have been a longer-lasting effect, at least some of which might wear off over time – perhaps as regulation is rolled back, or credit conditions ease and trade finance increases. This would mean that the pace of globalisation has only temporarily slowed. However, we discuss later how some of the effects of the financial crisis are likely to be permanent.

Globalisation in current form has reached a limit

And in any case, there are several other reasons to think that globalisation in its current form has reached a limit. This is even before we factor in the effect of trade wars (which we return to later). That is not to say that trade and capital flows are about to grind to a halt, or even stop rising in absolute terms, but rather that they might not rise any further as a share of global GDP, or could rise less quickly than global GDP.

1. Fewer regions left to integrate

The first reason to think that globalisation in its current form has run its course is simply that the policy push towards more integration was running out even before the trade war began, given that all the major steps to integrate the global system have already been taken. One illustration of this is the fact that 93% of the world’s population now reside in countries which are members of the World Trade Organisation or WTO. (See Chart 4.) The share of world goods trade accounted for by WTO members is even higher at 98%.

Admittedly, this in itself does not tell us too much. After all, WTO members are still allowed to impose tariffs; rather, the WTO describes itself as a system of rules dedicated to “fair and undistorted competition”.

Chart 4: % of World Population in Countries Which are members of WTO

Sources: WTO, UN, Capital Economics

Nonetheless, it is true that restrictions on global trade have been dismantled over the past 40 years through several rounds of negotiations under the WTO and its predecessor the GATT. Major steps include the Uruguay round between 1986 and 1993 and the 2005 expiration of the multifibre agreement restricting textile imports to the US and EU. Average tariffs on goods are now very low, at just 2.6%. (See Chart 5.)

Chart 5: Average Global Tariffs (%)

Sources: World Bank, Coatsworth & Williamson (NBER, 1981)

Of course, trade barriers are not just about tariffs. The WTO’s Trade Facilitation Agreement – negotiated in 2013 and passed in 2017 – is the first multilateral trade agreement concluded since the establishment of the WTO in 1995 and focuses on reducing non-tariff barriers such as quotas. According to the WTO, fully implementing these commitments could reduce global trade costs by some 14%. Non-tariff barriers are especially important in services, which we discuss separately later.

Meanwhile, some pin their hopes on India or Africa to become the next China in driving globalisation further. However, an often overlooked point is that India and Africa are already integrated into the global economy. Their tariffs are low and Chart 6 shows that exports of goods and services already equate to 25% of GDP in Africa and 20% in India. So as far as trade is concerned, these areas are already at least as open as China and significantly more so than the US, where exports account for 12% of GDP.

Chart 6: Exports of Goods & Services (As a % of GDP, 2018)

Source: WTO

What’s more, many of the economies that have already opened up will be seeking to reduce their dependence on export-led growth. The obvious example is China. First, China’s sheer size and already high global export market share make it unlikely that double-digit export growth can be sustained. (See China Economics Focus, “The coming slowdown in China,” 14th May 2018.) Second, China is moving up the sophistication chain by producing some higher value-added parts and components itself, instead of importing them. This is reflected in the rise in the domestic value-added share in China’s exports. Meanwhile, production of low value-added goods is simply shifting to low-wage areas in China’s interior. Third, behind all this is a rebalancing of domestic demand; and we think that the decline in trade-intensive investment will be only partially offset by a rise in consumption and consumer imports. (See Global Economics Focus, “How will the rest of the world cope with 2% growth in China?” 2nd April.)

2. Robots to replace cheap labour

The second reason why further progress in globalisation looks limited is technological progress.

This may be surprising; after all the recent wave of globalisation has been driven by technology – with the internet helping the emergence of global online hubs like Amazon and driving global price transparency and competitiveness. But with a few exceptions (such as the fast-growing mobile payments systems in developing countries), the effects of this technology have already been felt.

It is possible that technology will be developed in a way we can’t currently imagine, which then promotes another wave of globalisation. But technology as it is evolving now does not seem likely to do this.

Admittedly, it could help further globalisation in some respects. For example, autonomous driving could reduce transport costs, while artificial intelligence (AI) systems could streamline logistics operations, both reducing the costs of trade as previous waves of technology did.

But more importantly, new advanced manufacturing techniques, such as 3D printing, will make labour costs less relevant to the location of manufacturing than before. Of course, the process of machines replacing human labour is nothing new; it has being going on for 200 years. But in the past, cheap labour was still needed to operate machinery; that will be less the case now. Chart 7 shows the high proportion of jobs that the OECD estimates is at risk of whole, or partial, automation.

Chart 7: Jobs At Risk of Automation

Source: OECD

Note that this does not just apply to the production of goods; it will apply in some services too. For example, advances in information processing are allowing computers to become much more advanced in tasks like pattern recognition and complex voice interaction – meaning, for example, that robots are already replacing call centre staff or radiography staff that identify diseases from scans. And technology could reduce trade flows in agriculture, too, by making it easier to grow crops outside their natural climates.

By the way, this does not mean that robots will lead to a reduction in the overall number of jobs. In fact, Chart 8 shows that when the number of robots started to rise around 2010, the share of manufacturing in world employment stopped falling. (The downward trend has subsequently resumed, but at no faster a pace than in the late 1990s/early 2000s). Indeed, in the past, the introduction of robots freed up time for more productive tasks, ultimately leading to a rise in employment. But the key point is that the jobs which can be replaced by robots will be disproportionately in tradeable sectors, whether manufacturing or call centres, while new jobs will be created in less-tradeable areas like robot maintenance, leisure or social care.

Chart 8: Robot Sales & World Manuf. Emp.

Sources: ILO, IFR

So far, evidence of any reshoring remains largely limited to anecdotes here and there (e.g. Adidas’s “speed factories”). However, productivity growth in the US has picked up and the downward trend in the share of manufacturing employment in total employment has flattened off in the past few years. (See Chart 9.) A similar trend in manufacturing employment has been seen in many other advanced economies. We might even see manufacturing’s share of the economy start to rise again in some of the developed countries where it has been falling.

Chart 9: US Manuf. Emp (As a % of Total Emp.)

Source: Refinitiv

Note, too, that the so-called import content of exports (i.e. the share of imported inputs in the overall exports of a country, also referred to as the ‘foreign value-added share of gross exports’) has on average been falling in OECD countries since 2012. (See Chart 10.)

Chart 10: Average Import Content of Exports of OECD countries (%)

Source: OECD. Weighted means weighted by each country’s GDP. Unweighted means a simple average across countries.

Admittedly, trade is not driven solely by where labour costs are lowest; traditional Ricardian trade theory, based on comparative advantage and specialisation, holds only up to a point. Geographical proximity, quality of infrastructure, historical precedent, clustering, skills, and economies of scale developed by early entrants into the market determine trade patterns too. Indeed, these factors explain why rich developed economies trade with each other – and this trade is unlikely to be significantly affected by advances in robotics.

But as Chart 11 shows, rich-rich country trade has unsurprisingly been taking up a decreasing share of global trade over time; it has fallen from over 80% in the early 1800s to 30% now. Interestingly, it is trade between poor countries that has really shot up over the past decade or so. This probably partly reflects rises in commodity prices (the chart is in nominal terms), which have boosted nominal flows of, for example, agricultural products from Latin America to China. However, some of these exports will have been the inputs and intermediate goods used to produce poor countries’ exports to rich ones. And it is such supply chains which we go on to discuss next.

Chart 11: Exports Between Rich & Poor countries (As a % of Global Exports)

Source: Our World in Data

3. Supply chains becoming less complex

The third reason why globalisation may be nearing its natural peak is that the complexity of supply chains seems to have peaked.

Digital control of supply chains has allowed industrial production to be split up into dozens or hundreds of stages, which are then allocated to producers around the globe according to efficiency and cost. For example, Boeing 787 aircraft are produced by over 500 different suppliers based in more than ten countries. This has resulted in more trade for any given value of final production, as the same good is often exported and re-exported several times at different stages of the production process.

However, that process is now largely complete. In fact, companies are waking up to the problems that complicated supply chains can bring. This includes a greater chance of something along the line going wrong; the difficulty of maintaining quality control; and a lack of traceability (especially important in food supply chains). What’s more, it lengthens the time it takes to get new products to consumers. This is crucial in sectors like fast fashion, especially now that technological progress in collating data means that companies can see more quickly what is selling quickly or is most searched for online.

More generally, there could be a shift in preferences away from buying or procuring things at the lowest cost. This extends to consumers seeking locally produced goods, partly for environmental reasons.

Accordingly, surveys suggest that many firms are now trying to shorten supply chains – both in terms of geographical distance and the number of links in the chain – in order to reduce risks, increase quality control and respond to consumer demand more quickly. For example, a 2017 survey of over 500 firms in 13 countries by the Economist Intelligence Unit showed that over the next five years, 15% thought that they would stay the same, 33% expected them to lengthen and become more complex, and 49% expected their supply chains to shorten and become simpler.

Chart 12 also suggests that the fragmentation of supply chains is easing off. The ratio of exports and imports of intermediate goods to that of final goods rose sharply in the late 1990s/2000s. But recently it has levelled off and, on OECD data, has been falling. Some of this reflects changes in the prices of commodities, a key intermediate good, but a more important factor is evolving supply chains.

Chart 12: Ratio of World Intermediate Trade to Final Trade in Goods & Services

Sources: BIS, OECD

4. Capital flows to be held back

Meanwhile, there are reasons to think that capital flows are unlikely to become significantly more globalised.

Admittedly, they could receive some boost from technological advances. For example, widespread adoption of blockchain technology could make cross-border financial transactions quicker and cheaper. Against this, though, three key factors are likely to hold back financial globalisation.

First, the financial crisis is likely to have at least some long-lasting legacy on the global operations of the banking sector. Since the crisis, banks have reduced their overseas exposures. This partly reflects the short-term pressures that banks were under to quickly shrink their balance sheets. However, it reflects some longer-lasting factors too, such as a reappraisal of overseas risks, extra regulation, and a recognition of the fact that foreign business was often less profitable than domestic business.

Second, capital flows might be held back deliberately by policymakers in recognition of the financial stability risks that they can bring. These risks have always been well-known for emerging markets. But more recently, awareness has grown among developed economies. For example, one of the factors leading to the global financial crisis was the large amounts of capital corresponding to current account deficits that flowed into foreign – primarily US – assets, driving down bond yields and interest rates. And a contraction in cross-border finance was a central feature of the 2012 euro-zone crisis. Indeed, capital controls were ultimately imposed in Greece and Cyprus. Even the IMF has, since the crisis, discussed when limits on capital flows might be justified. Controls might include inflow restrictions, minimum maturity requirements for inflows, taxes on foreign purchases of assets or unremunerated reserve requirements on inflows.

Third, a significant further globalisation of capital flows would require China to open up its capital markets. But that seems unlikely given the Chinese government’s desire to retain tight control over domestic financial conditions. China is likely to be especially nervous now when its banking system is looking shaky, making the risk of capital flight even more pertinent. Russia, too, may be reluctant to increase its dependence on the West; after it, it has taken deliberate steps to insulate itself in some areas because of the risks of sanctions.

5. Cultural barriers and immigration

Finally, language and cultural barriers will always present a natural limit to globalisation. And there are some signs that these limits are being reached.

Of course, by their nature, assessing these factors is more speculative than some of the other ones that we have discussed so far. What’s more, economic factors will continue to provide a strong impetus for migration – in particular, the big differences in the levels of incomes between emerging and developed economies. This is especially pertinent given the sharp rise in the African population that is forecast.

However, political factors might prevent this economic impetus from translating into actual migration flows. It is noticeable that there has been a backlash against migration in many countries, often reflecting concerns about the pressure on domestic infrastructure (e.g. schools, housing, doctors). And some companies have been forced to reverse outsourcing to, for example, Indian call centres, because customers prefer to speak to someone in their own country.

Meanwhile, technology has in some ways boosted cultural convergence – more tourists travelling to other countries, people being able to watch foreign films etc. But in the future, the influence of technology might be more nuanced, perhaps even spurring a divergence of cultural norms. For example, China has used technology to start social credit scoring, whereby surveillance will result in every citizen being given a social score based on their behavior, which will then determine things like what services they can and can’t access.

The future of trade in services

These are all reasons why globalisation has reached its limit. But what about the scope for trade in services to increase much further? Won’t that drive another wave of globalisation? Indeed, in contrast to trade in goods, trade in services as a share of global GDP is still rising in value terms. (See the solid lines in Chart 13.) That said, much of this reflects rising relative prices; in real terms, the underlying trend in services exports as a share of global GDP looks rather flatter. (See the dotted lines in Chart 13.)

Chart 13: Nominal Exports of Goods & Services (As a % of Global GDP)

Sources: WTO, Capital Economics

There appears to be scope for services trade to increase much further. Admittedly, services are already more important to global trade than Chart 13 suggests. Chart 13 shows gross exports. But services are an important intermediate input in the production of exported goods. In Chart 14, we have shown a measure of exports created by the Reserve Bank of Australia that adjusts for this. This measure – exports in gross value added terms – puts services exports much closer in size to goods exports. However, given the much bigger size of services sector output, there would seem to be scope for services exports to grow much further still.

Chart 14: Global Output & Exports (2016, $trn)

Source: WTO, RBA, Capital Economics

There are certainly some reasons to be optimistic about the outlook for services trade. First, there is scope for further growth in those services which are already growing strongly. Chart 15 shows that travel is the biggest sub-sector of global services exports (which includes the services provided to foreign visitors like accommodation and meals), while transport is the third biggest category. There is scope for these to grow faster as people get richer and spend a bigger share of their income on travel. The number of tourists from China, in particular, will rise rapidly. Recent estimates suggest that only 7% or so of Chinese residents have passports, compared with about 40% in the US and 75% in the UK. Moreover, Chinese citizens can currently only travel to 74 countries without a visa compared to 185 for citizens of the US and the UK. But that number is rising.

Chart 15: Global Services Exports ($bn)

Source: Unctad

Second, there is scope for other services which can in theory already be traded to be traded more as trade is liberalised – for example, education or business services (such as legal services).

And third, technological progress will in theory increase the number of services which are tradeable. For example, telerobotics (the control of robots from a distance) means that a doctor in China recently became the first to operate remotely, 30km away from the patient. In the future, surgeons might in theory be able to operate from a different country (creating an export for the medical services sector). Remote digital sensing technology means that aircraft maintenance can be done from afar (an export for aircraft services).

Improved AI translation services will overcome language barriers – which are far more important for services than for goods trade – further. There is a simple reason that allowed India to become a global leader in the outsourcing of business and financial services – the people speak English. English speaking abilities in other Asian countries are much lower. But this might be at least partially overcome with AI, increasing the global pool of labour that can be used to perform services overseas.

However, some significant hurdles will remain. First, the liberalisation of services trade will remain inherently harder than that of goods. Questions over sovereignty mean that it will not get any easier for countries to come to agreement on common standards, data localisation laws, common recognition of professional qualifications etc. Agreement is not impossible; for example, the recently concluded Trans-Pacific Partnership included the removal of restrictions on cross-border services. However, the EU is still a long way from completing the Single Market because trade in services is far from fully liberalised. Accordingly, even if technology makes it possible, for example, for surgeons to operate remotely, operations may still mainly be done within national borders.

Second, the sectors which technology will help to make tradeable will still be quite small. For example, even if surgeons can operate remotely, that is a tiny proportion of the medical services profession, with the vast bulk still needing to be based locally (think nurses, porters etc.). And in most cases, someone would need to stand ready locally to take over if the technology failed, meaning that it might not be commercially viable, even if technology in theory allowed it.

Third, technology will not just be trade-increasing; in some respects it will be trade-reducing. As we said earlier, technology will allow some services to be produced domestically more cheaply (e.g. AI used in call centres). And countries may not want to trade in some areas if increased technology leads to increased security concerns. For example, Western countries have recently expressed fears that allowing Chinese company Huawei to provide the technology for their 5G networks could allow China to create a “backdoor” into their data networks.

Fourth, there will still be large areas of services which cannot be traded easily. No amount of technology can alter the fact that many services cannot be stored and require face-to-face interaction in real time – for example, eating out, haircuts, social care. In Chart 16, we have shaded the different major components of services output in the US according to how tradeable they are. Around 40% are never likely to be traded very much.

Chart 16: Breakdown of US Services Output in 2018 According to Tradability

Sources: BEA, Capital Economics

What’s more, these are the areas that we expect to grow fastest as people get richer. Chart 17 shows the output of those private sector service sectors which we expect to remain relatively non-traded. (We have done this for the US given its availability of long-run data.) For the past 70 years, they have been taking a steadily rising share of US GDP and we expect that trend to continue.

Chart 17: US Output of Non-Tradeable Private Sector Services as a % of Total GDP

Sources: BEA, Capital Economics

So old-fashioned globalisation is over

The upshot is that many of the trends that have characterised the recent wave of globalisation – the increased exchange of goods, services, capital and people – are close to peaking as a share of the global economy. To use the rather awkward phrases that have been coined, slowbalisation, or perhaps even de-globalisation, is likely to be the way forward.

However, as we will discuss in more detail in one of our accompanying reports, the implications of this for the global economy need not be bad. If weaker trade is a result of firms finding a more efficient way of producing domestically with robots, or with a shorter supply chain, then productivity and GDP growth will, if anything, be stronger than before. And inflation will still come under strong downward pressure from technological progress. Reducing the reliance on global capital inflows could increase financial stability.

Moreover, a slowdown in the exchange of physical goods/capital/people does not have to mean that the exchange of ideas – the key to technological progress – will slow. In the past, technology was spread mainly through trade in goods, but now that process mainly happens through the flows of data and information. And those flows will continue. Take the production of cars, for example. Shared chassis designs and platforms are used by a range of vehicle manufacturers and models. So even if cars are produced locally, rather than traded, the entire production process still relies even more heavily on a sophisticated and global chain of design and ideas. Similarly, while robots might produce goods locally, the instructions and designs that are powering the robots can still be formed at a global level.

Could trade wars make things worse?

But could policymakers turn all this into a much more malign form of de-globalisation? After all, in the past it is policy which has been the cause of any actual rollback of globalisation.

And the current most likely cause of policy-driven rollback is the trade war between the US and China. As we have argued in other research, the trade war between the US and China is actually in and of itself not that big a deal. (See Global Economics Update, “Escalation of trade war to raise hit to global GDP to 0.5%”.) First, it has largely been confined to restrictions on goods trade. And second, US/China bilateral trade is small. The goods that the US and China buy from each other together sum to only 3% of all world goods trade. (See Chart 18.)

Chart 18: Breakdown of World Goods Trade (%)

Source: Refinitiv

But is the recent US-China trade war just the start of something more serious? That partly depends on its cause. It could just be a President Trump thing, which will therefore pass when he leaves office. However, as we will explain in one of our accompanying reports, we think that the trade war reflects far more fundamental factors in the relationship between the US and China, which are unlikely to go away, even when President Trump leaves office.

If this is true, then the trade war between the US and China might escalate further. After all, there seems to be little chance of either side backing down, with every period of calm followed by a renewed escalation of tensions. We already assume in our forecasts that tariffs will be imposed on all remaining US-Chinese imports and that China will extend administrative restrictions to more goods and blacklist more US firms. And the escalation might spread beyond trade barriers.

Could the trade war broaden out?

What’s more, there is a significant chance (perhaps 30%) that what has started off as a US-China centred wave of protectionism does not remain so.

After all, if the US succeeds in reducing its trade deficit with China then – unless the US changes its domestic policies to alter its saving and/or investment – that deficit will just get transferred to other countries instead. In that case, other countries might be drawn in to the US’s trade war. President Trump has already complained about Germany in particular in the past, especially regarding the amount of cars Germany sells to the US.

Indeed, although most of the US’s tariff increases have been directed at China, the steel and aluminium tariffs were introduced on a wider range of countries, including the EU. Meanwhile, the US’s bilateral trade deficits with Taiwan, Korea and Vietnam have increased markedly since the start of the trade war, so these countries might be next in line. What’s more, the US is now moving towards using trade as a way to achieve other aims (for example, recently threatening Mexico with tariffs unless it took action to reduce illegal migration to the US). Again, that might get more countries drawn in.

Admittedly, that would still leave the trade war as very much US-led. But what if the trade war does not just have its origins in the US. What if it is the start of a global backlash against globalisation? People might blame globalisation for increased inequality, financial instability, multinational tax avoidance and unwanted migration. Indeed, a key problem with globalisation is that it has bolstered the power of global markets and undermined the power of national governments (and there is no effective global government). These governments have limited control over globalisation, even though it affects a host of national issues, such as the distribution of income, labour rights, the environment and corporate behaviour including tax avoidance.

Indeed, past experience has shown that globalisation creates winners and losers and eventually the losers gain a political constituency and push back. For example, US farmers were pushing for higher tariffs against agricultural imports for most of the 19th and early 20th century.

How bad could the US-China trade war get?

In addition to tariffs on all trade, both countries could introduce more non-tariff barriers (e.g. slower tougher regulatory standards). This would bring more services trade into the fray – although services trade between the US and China is far smaller than goods trade. (See Chart 19.) (See China Economics Focus, “China’s next steps in the trade war,” 5th August.)

Chart 19: US-China Services & Goods Exports ($bn, Latest full-year)

Source: Refinitiv

Nearly 60% of China’s services exports to the US are in the telecomms/business services sectors. In contrast, nearly 60% of US services exports to China are travel and would be hit if limits were set on visitors and visas. (See Chart 20.) Of these travel exports, nearly half are in the education sector and so would be dented by a drop in foreign students.

Chart 20: Services Exports by Type (% of Total)

Sources: WTO, Capital Economics

Last summer, China told its citizens to beware of shootings, robberies and high costs for medical care in the US. The US retaliated by warning that China’s use of “exit bans” might lead to US citizens not being allowed to leave China. Further measures could include tightening rules for getting visas, setting limits on the number of visits or even banning them outright. Note that Chinese tourists to the US fell last year for the first time in 15 years.

The trade war might also extend into capital controls. Restrictions of portfolio capital flows (of bonds and equities) are already substantial, but to go further, the US could block Chinese companies from listing in the US. Or China could sell its holdings of US Treasuries (although we think it unlikely).

More important is FDI. There were already regulatory barriers on both sides and extra measures have been taken. The US stepped up investment screening by the Committee on Foreign Investment in the US, so it can block a wider array of foreign transactions deemed a threat to national security. And since 2017 China has implemented measures to limit “extensive overseas investments” by domestic groups. Both have hit Chinese FDI into the US. (See Chart 21.)

Chart 21: US China FDI ($bn)

Source: The US-China Investment Hub. 2019 = H1 annualised.

More generally, both countries could make it harder for foreign firms to operate domestically – for example, more ad hoc bans or making it harder for foreign firms to get business licenses. Conducting business would become less rules-based and more down to discretion and political controls.

China’s move to allow the renminbi to depreciate below the symbolic 7.00 level against the dollar in August could end up in a full-blown exchange rate war. The US could directly sell dollars. (See US Economics Update, “Currency intervention would fail without Fed’s full support,” 8th July.) Or it could undermine the value of the dollar by removing Fed Chairman Jerome Powell, stripping the US Fed of its independence and running an ultra-loose monetary policy. After all, Trump has been strident in his criticism of the Fed’s (in his view) overly tight policy.

Military conflict could even result. “Thucydides’s trap” is a term coined by Harvard professor Graham Allison to refer to the idea that rivalry between an established power and a rising one often ends in war.

And if there is a backlash against globalisation, it would be of little surprise that it has shown up in the US first; things usually do. In fact, there are already developments elsewhere that could be interpreted as a backlash against globalisation – in particular, the Brexit vote in UK and the rise in populist movements in Europe. (See Table 1.)

Table 1: Signs of Backlash Against Globalisation


Backlash against globalisation

Rest of Europe

(2012 onwards

Rise in support for populist nationalist parties in France, Italy, Germany, Netherlands, Sweden and Austria.

Russia (2014 onwards)

Trade and financial ties diminished since Western sanctions imposed in 2014. Government seems to be embracing isolation.

UK (June 2016)

UK (June 2016)

US (Nov. 2016)

Election of President Trump, with his protectionist agenda

Source: Capital Economics

Admittedly, these are not necessarily driven by discontent with globalisation. In the UK, the Brexit vote may have primarily been about migration – only one aspect of globalisation. Indeed, the Brexit party is in favour of free trade and no tariffs. Meanwhile, if the backlash has been against increased inequality, has that been driven by globalisation, or technology, or government policy – or the interplay of all three? And if there is a backlash, it is more of a Western phenomenon than a global one.

Indeed, survey evidence of attitudes towards globalisation paints a mixed picture, even within developed economies. Chart 22 shows how views have changed in the G7 countries over the past couple of decades. Sentiment towards globalisation has deteriorated sharply in Italy, and to a lesser extent in France, the US and Germany. But Japan, the UK and Canada now have a more favourable view. Indeed, this might suggest that the UK’s Brexit vote was more about the specific issue of migration or sovereignty, rather than globalisation more generally.

Chart 22: Opinions of Globalisation (Net % Balance, Change Between 2002 and 2018)

Source: Pew Research Centre. Exact question is: “What do you think about the growing trade and business ties between your country and other countries?” Opinion in each year is measured as the balance of people replying “a good thing” minus those replying “a bad thing”. The chart shows the change in this balance between 2002 and 2018.

However, even if there has not actually been much backlash against globalisation so far, that could change, perhaps in the face of a big migration push from Africa, or an economic downturn that pushed up unemployment. Although governments could in theory head off these concerns by helping those made worse off by globalisation, they might instead capitalise on voters’ concerns by pulling up the drawbridge. Indeed, the number of non-tariff trade barriers introduced by countries have been increasing since 2010, suggesting that, US-China trade war aside, this breakdown in co-operation has already begun. And note that the Doha round, the most recent programme of multilateral trade talks that began in 2001, ended up being canned in 2015.

How might things change?

So either a worsening of the US-China trade war, or a broader backlash to globalisation, could result in policy-driven de-globalisation.

There are various forms that such policy-driven de-globalisation could take. The most immediate threat might be the current situation degenerating into a free-for-all, with a “beggar thy neighbour” rise in barriers, including tariffs, capital controls and migration limits.

Alternatively, we could see a move towards regionalisation, with the world splitting up into separate spheres of influence: a US-led bloc, a Chinese-led bloc and possibly a third European-led bloc. Countries would have to choose which bloc to ally with. Note that in a potential taste of things to come, the US is already pressuring its allies not to participate in the China’s massive Belt and Road Initiative infrastructure programme.

In its mildest form, this regionalisation could simply manifest itself as disengagement from multilateral institutions. The rules-based international system and WTO may not survive or may be applied only voluntarily and/or in limited parts of the world. But at the other end of the spectrum, it could involve a separation of payments systems, different standards of regulation, technological divergence and even different IT networks. The flow of data and information would be the hardest thing to stop, but even here, steps could be taken, such as splitting the internet into two, with one half led by the Chinese (the so-called “splinternet”).

Some argue that globalisation has become so entrenched that it cannot be unwound as it was in the past, but on the whole we do not agree with this view. Capital controls would be fairly easy to impose. Such measures could include exchange controls (placing restrictions on the buying and selling of currency) or caps on the sale or purchase of different kinds of financial assets. Meanwhile, complicated supply chains could be unwound; reshoring activity is not impossible if countries have the time and money to invest in domestic production. And China already bans the sale of Western internet services to stem the flow of information; there is no Facebook or Twitter in China.

Protecting intellectual property would be tricky though. And even limits on legal migration would not stop illegal migration, although technology could help; instead of physical fences and walls, countries can install cameras and sensors (as the US has done along its Mexican borders, dispatching Border Patrol agents to suspected migrant crossings). In fact, it might be that policy would close off some areas of integration, but remain open to others. This is what has often happened in the past. For example, the Bretton Woods international system was organised on the basis of reductions in trade barriers, but controls to capital flows.


To sum up, the current wave of globalisation was nearing a natural limit even before trade wars struck. We think that globalisation as we know it has peaked and there is unlikely to be any further major integration. Flows of trade and capital are now unlikely to grow as quickly as the global economy. And to the extent that goods, people and capital do continue to cross borders, this may be more at a local/regional level, than a global one.

What comes next depends on policy. Globalisation may just stall over the next decade. After all, the current US-China trade war is, in isolation, not a major threat. But it may be a sign of more deep-seated problems. Accordingly, a period of policy-driven de-globalisation is increasingly likely.

Vicky Redwood, Senior Economic Adviser, +44 20 7808 4989,