The Slowing in World Trade: a structural break, the cause of uncertainty


Sébastien Jean


13 February 2017

Available versions :



Jean Sébastien

Sébastien Jean

Director of the Centre d'études prospectives et d'informations internationales (CEPII, Co-author with I. Bensidoun and J. Couppey-Soubeyran of L'économie mondiale 2017, Editions La découverte

The Slowing in World Trade: a structural break, the cause of uncertainty

PDF | 736 koIn English

1. A Structural Break

As a matter of fact, the break is a clear one. Over the fifteen years prior to the crisis (1993-2007), the volume of world trade in goods and services increased by 7.2% on average per year, more than twice that of the volume of world GDP (3.1%[4]). Between 2012 and 2015 trade only grew at a pace of 3.3% per year on average, barely faster than GDP (2.6%). We should also point out that even the latter trend does not reflect the most recent developments: since the end of 2015, according to the monthly data released by the CPB (Netherlands), recognised for their quality, the volume of world trade has barely increased, even though the most recent figures point to a recovery, apparently linked to the recovery in industrial output.

Of course the slowdown in GDP growth and therefore in demand, mechanically plays a role in trade and this role appears all the more important since we are referring to the relatively dynamic period preceding the crisis. Although this led the IMF to highlight sluggish demand as the main reason behind the slowdown in trade, this conclusion is hardly convincing as we look at the situation with greater hindsight.[5] The now emerging break with the trend does not concern a slowdown in trade in the strict sense of the term as much as it does a decline in the ratio between the rate of growth in trade and that of GDP, often called trade to growth elasticity. For a time, sluggish investment was used as a major factor in explaining this, but again the argument does not withstand analysis as far as the recent period is concerned. Of course investment rates are historically low in many countries, particularly in the industrialised countries, but it has stopped declining which means that investment is not increasing less on average than the other elements that make up demand.

The increase in the share of services in our economies is another possible explanation, since services are traded less than goods. However, this does not seem enough to explain the slowdown, firstly because this secular trend has not experienced collapse concomitant to that of trade. The international trade in services has also slowed even though it has withstood better than the goods trade.

Finally, in explanation of the slowdown in world trade, two factors clearly stand out, even though their respective contribution to this remains uncertain. They are linked to China and international chains of value. An additional question concerns the previous and future roles of protectionist policies.

2. China's economic readjustment is slowing its foreign trade

The role of the Chinese economy should be highlighted because this country, which has become the world leader in the export of merchandise since 2009, has for the last decade been undertaking a major task of readjustment. Hence in the 2000s China reached an extraordinarily high level for a country of this size, since exports represented up to 35% of its GDP in 2007. This remarkable extraversion resulted from the development strategy pursued by the Chinese government, which mainly relied on opening, and in particular, on exports to facilitate the introduction of a market discipline and access to modern technologies, whilst simultaneously maintaining a strong dynamic to accumulate productive capital and production efficiencies. The attraction of foreign direct investment on the part of multinationals from wealthy countries and the assembly trade, in which they have gladly engaged, combining imported inputs with low cost Chinese labour, has been a key factor. In spite of its success this strategy was only ever a first stage. The long term sustainability of Chinese growth required readjustment in several ways. Basically the domestic market was to take over from external outlets that could not, in the long term, maintain their initial dynamic, given the already high market share achieved by Chinese export firms; consumption was to supersede high investment in relation to GDP; growth was to rely on services more than on industry; and exports more on national outlets than on the assembly of imported inputs.

This overall development started at the beginning of the 2000s and was reflected in trade, with a limit to the level of opening, since the export to GDP ratio declined after 2007 to return to 26% in 2015, whilst the trade surplus, which rose to 8% of GDP, returned to lie within the range of 2 to 4% of GDP. This slower pace especially concerned the assembly trade by foreign companies which only represented 33% of Chinese exports in 2015 against 46% in 2007. Beyond readjustment this trend was part of a strategy to boost the sector, whereby China progressively gained wider control of the components necessary for its exports of finished products, and to move upmarket, via the progressive increase in the quality and technology of its manufactures. Hence the unit value of Chinese manufactured exports has increased faster than that of its competitors in most sectors over the last ten years[6]. The rapid increase in Chinese salaries leaves no other choice. Recently China's external trade has not even shown any growth trend: according to the WTO in the third quarter of 2016 the volume of Chinese exports were lower than their level two years ago, whilst the volume of imports were lower than their level in the third quarter of 2012, down by 10% in comparison with their highest point in 2014![7] Even though the most recent figures show recovery, it is still too early to deem that this will last; the least we can say is that China is no longer the locomotive for world trade that it was for many years.

3. The dynamic of international chains of value is exhausted

The second major factor to explain the slowdown which is not disconnected from the first, but cannot just be reduced to that either, is the slowing in the extension of global value chains. With this we mean the fragmentation of the production process into a great number of tasks undertaken in different countries in order to take best advantage of differences in wages, the cost of capital, qualifications, technologies and the availability of inputs. The chain of productive tasks, the sources of value added, that is fragmented ever more finely between countries has led to increasing international trade. This process played a central role in the speeding up of world trade between the 1990's and 2000's. It is notably reflected in the fact that the value of exported manufactures between two countries no longer just corresponds to the displaced value added between them. This is the case with the value of imported inputs and those which will then be re-exported (towards the country of origin of the flow or towards a third country) after transformation or incorporation into another product. In all, only a fraction of the exported production is really exported value added, and this fraction diminishes as the international chains of value become fragmented. Hence this fraction is estimated to have decreased from 78% in 1990 to 68% in 2008 regarding all products, and from 59% to 46% in terms of manufactured goods.[8] Although the equivalent figure is not yet available for the most recent years, this downward trend has clearly stopped now and has even reversed slightly. In other words the international fragmentation of value chains has stopped extending and has even decreased slightly. This is shown by the direct fragmentation gauge developed by the OECD, but also by the decline in intermediary consumption in world trade.[9] This break in trend alone would explain two percentage points in the slowing of world trade in comparison with the pre-crisis period, ie around half of the overall slowdown that has been observed.[10] Hence the flows, in which the international fragmentation of value chains is the highest, are also those where growth has clearly been lower than expected, if we base ourselves on an analysis of pre-crisis factors.[11] The corollary of this observation is that the slowdown in international trade is not as sharp if we measure the flows of value added rather than gross output. However, it should be stressed that the break in question here is mainly a clear cut in the trend toward ever more finely fractioned international chains of value added, which does not necessarily mean the decline of fragmentation. Even though some multinationals have taken advantage of progress in robotics, decreasing price of machinery and the reduction of labour cost differentials to repatriate a share of the previously relocated production to rich countries, we cannot speak, at least for the time being, of a significant step backwards. The fashion surrounding the terms of relocation or reshoring should not hide the fact that statistics do not indicate any dramatic development in this direction.

This break in the international division of labour raises several issues, particularly that of understanding its determining factors, notably the share played by economic policies in this and their possible protectionist bias. The first element of explanation lies in the profits made in the international extension of international chains of production, which tend to decline as the process moves forward: the most profitable trade-offs have already been implemented. Also, although foreign direct investments have continued to rise in absolute terms, they have significantly declined in proportion to the gross formation of fixed capital in low or intermediate revenue countries, dropping from nearly 13% in 2007 to under 9% in 2014.[12] As for transport and distance coordination costs, the pace of their reduction has probably slowed and the profits to which they give rise are also yielding less. Finally economic policies cannot indefinitely be made to foster extraversion as was the case in most countries during the 2000s.[13]

Moreover the recent period has highlighted that the fine international division of manufacturing processes can also be a weak link, which seen as such can limit its deepening. The earthquake in Japan in 2011 or the floods in Thailand in the same year disrupted work in factories that were sometimes established on the other side of the world, particularly in the electronics industry. The economic and financial crisis also showed just how powerful and rapid the transmission of macro-economic and financial crises can be in this context.

4. Protectionism has not caused the slowdown but the latter means there is a threat of a return to protectionism

The return of protectionism is another possible explanation. Crises are periods that favour protectionist response, with governments and businesses trying to compensate for the decline in demand by a reduction of foreign shares on the national market. Aware of the dangers of the slippery slope inherent to trade wars, the G20 countries solemnly committed during the London Summit in 2009 to "reject protectionism".

This commitment has not prevented some protectionist reactions however and the WTO deemed in November 2016 that nearly 3,000 measures restricting trade had been introduced by the member States since 2008, 2,200 of which were said still to be in force. This observation is difficult to interpret however since a high number of measures that aim to facilitate trade were taken at the same time and comparable data is not available for the period prior to the crisis. The number of measures also says little of the effective impact of the restrictive measures, which the WTO deemed in 2014, in regard to those in force, covered less than 4% of the world's imports.[14] The supervisory work undertaken as part of the Global Trade Alert project[15] also suggests that a certain number of measures slip under the WTO's radar, but again data does not allow us to form a precise idea of developments in comparison with the pre-crisis period and quantitative studies have not revealed a significant link between the extension of these measures and trade slowdown.[16]

It remains that the slightest dynamic changes the politics of free-trade. Exports were generally seen as a primordial factor to boost growth prior to the crisis: in wealthy countries, by taking advantage of the rapid expansion of the emerging markets; in developing countries, by taking advantage of the technologies of multinationals from the most advanced countries. Since trade is now no longer synonymous to growth, governments and populations increasingly see a zero-sum game in which it is vital to protect oneself from foreign competition. Requests for protection have always existed but they are increasingly pressing. Tension regarding how the cake is to be divided up is all the stronger since the latter's growth has been stunted, leading to fear that not everyone will get a fair share.

This is all the more so since China's industrial and trade power are the source of fear. In the wealthy countries it is seen by many as a factor of deindustrialisation, even though it has contributed to this in a very limited way according to all of the analyses now available: in some poor countries, in India for example, it is deemed a threat in terms of its industrialisation strategies. The way China functions is still extremely centralised, which raises the question of whether the competition that it causes is fair to its partner countries, which respect stricter rules from the point of view State intervention, and notably regarding subsidies in support of selected sectors. The controversy over China's market economy status, linked to the December 2016 deadline of specific transitory measures included its accession protocol to the WTO in 2001, reflects these questions to a backdrop of massive overproduction in the steel sector.

Brexit plus the American presidential election are emblematic of this political response to globalisation. In particular, coming from a country that was the main architect and leader of the multi-lateral system as we know it, President Trump's protectionist rhetoric comprises a serious threat to the institutional context of trade exchange. Although the greatest uncertainty still remains about how he will implement it, the political importance he gives to the issue indicates that he could go a long way. His first decisions (abandonment of the transpacific partnership TPP) and his provocative declarations regarding Mexico, China and even Germany confirm this. The most spectacular of them are also the most unlikely. The threat to impose customs duties of 45% on Chinese and 35% on Mexican imports, for example is unrealistic given the prohibitive cost caused to the American economy and its businesses - an American study estimates potential job losses at nearly 5 million in the three years following the start of a major trade war like this.[17] Other levers of action seem more likely starting with the greater use of anti-dumping or similar measures, even though recent unfettered action by the USA in this area sets the bar high if we are to mark a difference. Beyond that, a very wide definition of trade defence against practices that are deemed unfair might serve as a pretext to use more or less targeted additional customs duties. The challenge made to free-trade agreements is the greatest threat and it has to be taken seriously. Projects to manipulate the taxation of companies to make it favourable to export and unfavourable to import have also been mooted, following the Republican Blueprint which advocates the introduction of an adjustment tax on the border that is highly unlikely to be in line with  the WTO's rules.[18] Finally it is possible - although for the time being it is unfounded - to label China officially as a manipulator of its exchange rates, even though we should note that the practical consequences, based on two laws dated 2008 and 2015 would be limited. And this list must not be considered exhaustive, given the disregard the new President seems to have for the USA's international commitments.

At this stage there have in the main only been declarations and threats which have not been acted upon. It is possible that the reality of political and economic constraints, the strength of counterbalances and the persuasiveness of the environment will bring Donald Trump round to a certain sobriety. It remains though that the risks are real for a multilateral trade system that has been weakened by economic sluggishness, political contestation and institutional paralysis.

Although until now protectionism does not seem to have played a leading role in the slowdown of world trade, its re-emergence might well speed up the trend, and even lead to a considerable decline. Nothing is certain as long as dynamics and chain reactions play a key role in trade developments, but uncertainty is weighing heavy and the consequences could be deep and long lasting.

[1] Bussière, M., Callegari, G., Ghironi, F., Sestieri, G. & Yamano, N. (2013), "Estimating Trade Elasticities: Demand Composition and the Trade Collapse of 2008-2009", American Economic Journal: Macroeconomics 5(3), 118-51.
[2] OMC (2013), Rapport sur le commerce mondial 2013. Facteurs déterminant l'avenir du commerce mondial.
[3] Jean S. (2016), " La croissance du commerce mondial en deçà des attentes de l'OMC. Comme prévu ! ", CEPII Blog, 29 September 2016,
[4] Source: base WEO, October 2016, IMF, calculation based on market exchange rates.
[5] Jean S. (2016), "Comments on IMF's "Global Trade: What's behind the Slowdown?" - or why there is more to trade slowdown than weak demand", CEPII Blog
[6] Lemoine, F. & Ünal, D. (2017), 'China's Foreign Trade: A "New Normal"', China & World Economy, to be published.
[7] These figures are not seasonally adjusted but they all concern the third quarter.
[8] A Portrait of Trade in Value Added over Four Decades, Robert C. Johnson, Guillermo Noguera, NBER Working Paper No. 22974. The figure mentioned refers to their measurement excluding the rest of the world for which their data is incomplete.
[9] Cf. Haugh, D.; Kopoin, A.; Rusticelli, E.; Turner, D. & Dutu, R. (2016), "Cardiac Arrest or Dizzy Spell: Why is World Trade So Weak and What can Policy Do About It?", OECD Economic Policy Papers 18, OECD.
[10] See the converging estimates on this point by Haugh et al. (2016), op. cit., et Timmer, M. P.; Los, B.; Stehrer, R. & de Vries, G. J. (2016), 'An Anatomy of the Global Trade Slowdown based on the WIOD 2016 Release', GGDC Research Memorandum 162, University Of Groningen, Groningen Growth And Development Centre.
[11] Crozet M., Emlinger C. and Jean S. (2015), "On the gravity of the trade slowdown", in The Global Trade Slowdown: A New Normal?, edited by B. Hoekman, VoxEU-CEPR. See also Jean, S. (2015), 'Le ralentissement du commerce mondial annonce un changement de tendance', La Lettre du CEPII 356.
[12] Hakobyan S. & Lederman D. (2016), "XXX", World Bank Working Paper 7777. The trend is more difficult to decipher in the developed countries where this ratio is more unstable.
[13] See for example Bureau J.-C., Guimbard H. & Jean S. (2016), "Competing Liberalizations: Tariffs and Trade in the 21st Century", CEPII Working Paper 2016- 12 , May 2016 , CEPII.
[14] WTO (2014), Rapport sur le commerce mondial 2014. Commerce et développement: tendances récentes et rôle de l'OMC, World Trade Organisation.
[16] See for example Crozet et al. (2015), op. cit.
[17] Noland M., Clyde Hufbauer G., Robinson S., and Moran T. (2016), "Assessing Trade Agendas in the US Presidential Campaign" (2016), PIIE Briefing 16-06, Peterson Institute of International Economics.
[18] Jean, S. (2017), " 'Destination-based cash-flow tax', la concurrence fiscale puissance quatre des républicains américains ", CEPII Blog, 16th January 2017,

Publishing Director : Pascale Joannin

The Slowing in World Trade: a structural break, the cause of uncertainty

PDF | 736 koIn English

To go further

Democracy and citizenship


Nathalie de Kaniv

17 June 2024

The philosophy and culture of the Enlightenment shaped the mental map of Europe, drawing a distinction between civilization and barbarism. The latter is now rising again, and contemporary...

Democracy and citizenship


Pascale Joannin

10 June 2024

The European elections were held from 6 to 9 June in the 27 Member States of the European Union. Turnout was higher than in 2019, demonstrating that people feel more concerned by European issues. No...

Democracy and citizenship


Maria Mateeva-Kazakova

4 June 2024

Impact on the polls to be held on 9 June 2024. The position adopted by Bulgaria towards Russia is turning into a structural divide. On the eve of the legislative and European elections of 9th June...

Democracy and citizenship


Emmanuel Sales

21 May 2024

Chancellor Olaf Scholz's recent visit to Beijing was the subject of much controversy. Since Xi Jinping's nationalist shift, most Western companies have turned their backs on China. Foreign investment...

The Letter

European news of the week

Unique in its genre, with its 200,000 subscribers and its editions in 6 languages ​​(French, English, German, Spanish, Polish and Ukrainian), it has brought to you, for 15 years, a summary of European news, more needed now than ever

Versions :