Im­pact Se­ries: 03-22

Global Trade:
A Future in Doubt?

Michael Blanga-Gubbay
Senior Research Fellow at the Kühne Center for Sustainable Trade and Logistics at the University of Zurich

Mathilde Le Moigne
Senior Research Fellow at the Kühne Center for Sustainable Trade and Logistics at the University of Zurich

Outline

Talks of “deglobalization” or “slowbalization” have multiplied in the aftermath of the Great Trade Collapse of 2008/2009. The recent economic shock of the COVID-19 pandemic and the war in Ukraine have re-ignited fears of global value chain disruptions, and lead many in international trade to claim the end of globalization as we know it. In this Kühne Impact Series we examine these facts and find that while few statistics point towards a slowdown in global trade, looking at the broad picture we can still be cautiously optimistic. More concerns arise, instead, when looking at the policy landscape.

Full Article

Broad Trends: Is “deglobalization” visible in the data?

Claims of de­glob­al­iza­tion are large­ly sub­stan­ti­at­ed by the evo­lu­tion of the glob­al Trade-to-GDP ra­tio af­ter 2009 (Fig­ure 1). The Trade-to-GDP ra­tio is one of the most com­mon in­di­ca­tors for trade open­ness, or for the “vol­ume” of trade. It ex­press­es the rel­a­tive im­por­tance of in­ter­na­tion­al trade (the val­ue of im­ports and ex­ports) in the over­all econ­o­my (mea­sured by the GDP).

From the mid-1990s un­til 2008, glob­al trade was char­ac­ter­ized by a steady in­crease of the Trade-to-GDP ra­tio (from 16% in 1996 to 25% in 2008). The Great Re­ces­sion of 2008/2009 clear­ly in­ter­rupt­ed this trend, with a de­cline of the ra­tio of 5 per­cent­age points with­in a year. De­spite a sig­nif­i­cant re­cov­ery in 2010, the evo­lu­tion of this ra­tio in the af­ter­math of the re­ces­sion is char­ac­ter­ized by a steady de­cline. In 2019 (just be­fore the COVID-19 pan­dem­ic), glob­al trade rep­re­sents 21% of glob­al GDP, low­er than what had been at­tained in 2004.

As the Great Re­ces­sion has main­ly af­fect­ed de­vel­oped economies, one could won­der whether this de­glob­al­iza­tion trend af­fects every coun­try the same way. In Fig­ure 2 we present the evo­lu­tion over the same pe­ri­od of the Trade-to-GDP ra­tio of the five largest economies in the world: Chi­na, the EU (with and with­out in­tra-EU trade), In­dia, Japan, and the US. Three facts clear­ly emerge from this pic­ture. First, the Great Re­ces­sion af­fect­ed the Trade-to-GDP ra­tio of all the con­sid­ered economies, in­clud­ing Chi­na and In­dia, who were pre­sum­ably less con­nect­ed to the glob­al fi­nan­cial sys­tem: due to glob­al­iza­tion, coun­tries are much more in­ter­con­nect­ed and eco­nom­ic shocks spread quick­ly across bor­ders. Sec­ond, all se­lect­ed coun­tries with the ex­cep­tion of the EU show signs of a sim­i­lar de­glob­al­iza­tion or at least slow­bal­iza­tion: the Trade-to-GDP ra­tio of Chi­na, In­dia, and the US clear­ly fol­lows a down­ward slop­ing trend af­ter 2010, Japan’s ra­tio re­mains flat, and the Trade-to-GDP ra­tio of the EU is in­creas­ing when in­clud­ing in­tra-EU trade but flat oth­er­wise. Third and fi­nal­ly, Chi­na’s Trade-to-GDP ra­tio start­ed de­creas­ing be­fore the 2008/2009 eco­nom­ic cri­sis, and col­lapsed even fur­ther dur­ing the Great Re­ces­sion, de­spite a mild re­cov­ery in 2010. While Chi­na’s trade rep­re­sent­ed more than 60% of its GDP in 2006, this ra­tio falls to 28.5% in 2020. Part of this trend can be ex­plained by Chi­na’s ex­plic­it ef­fort to re­duce its open­ness to the world and to en­cour­age do­mes­tic con­sump­tion (in par­tic­u­lar in the 11th and 12th Five-Year Plans). A quick check shows that this how­ev­er doesn’t dri­ve the glob­al ag­gre­gate trend, as the evo­lu­tion of Trade-to-GDP glob­al­ly re­mains iden­ti­cal when ex­clud­ing Chi­na.

As a con­clu­sion, with Trade-to-GDP ra­tio as the sole sta­tis­tics to de­scribe glob­al­iza­tion trends, it ap­pears that in­deed, the Great Re­ces­sion marked the end of an era, and some ev­i­dence for a be­gin­ning de­glob­al­iza­tion in terms of “vol­umes”. Oth­er sta­tis­tics how­ev­er pro­vide a dif­fer­ent pic­ture that mit­i­gates this mes­sage. In fact, while Trade-to-GDP ra­tios pro­vide in­for­ma­tion on the “vol­ume” of glob­al­iza­tion, oth­er sta­tis­tics can pro­vide in­for­ma­tion on its “shape” and “out­reach”, re­spec­tive­ly, glob­al val­ue chain (GVC) par­tic­i­pa­tion in­dices, and the im­por­tance of dis­tance in bi­lat­er­al trade re­la­tion­ships.

The Great Recession marked the end of an era, and some evidence for a beginning deglobalization in terms of “volumes”.

Fig­ure 3 presents ev­i­dence on GVC par­tic­i­pa­tion. In­di­ca­tors of GVC par­tic­i­pa­tion track coun­tries’ en­gage­ment in GVCs and serve as prox­ies for the frag­men­ta­tion of the pro­duc­tion process, and are good in­di­ca­tors for the “shape” of trade. GVC par­tic­i­pa­tion is gen­er­al­ly com­put­ed as the share of GVC-re­lat­ed trade (trade flows of val­ue added that cross at least two na­tion­al bor­ders – fol­low­ing Borin and Manci­ni¹) over gross ex­ports. For ex­am­ple, a smart­phone as­sem­bled in Chi­na might in­clude de­sign el­e­ments from the Unit­ed States, high-tech com­po­nents from Japan, sil­i­cone chips from Sin­ga­pore, met­als from Bo­livia. Through­out this process, all coun­tries in­volved re­tain some val­ue and ben­e­fit from the ex­port of the fi­nal prod­uct.

Fo­cus­ing on the yel­low line, GVC par­tic­i­pa­tion is char­ac­ter­ized like the Trade-to-GDP ra­tio by a sharp de­cline dur­ing the Great Re­ces­sion, a strong re­cov­ery in 2010/2011, and a mild de­clin­ing trend af­ter­wards. The blue line ex­tends this trend us­ing data on in­ter­me­di­ate in­puts trade. In the medi­um run, GVC par­tic­i­pa­tion be­tween 2015 and 2020 ap­pears steady at 38% of gross ex­ports, sug­gest­ing that in­te­gra­tion of pro­duc­tion process­es re­cov­ered from the Great Re­ces­sion but did not ac­cel­er­ate any fur­ther – as was the case pri­or to 2009. The flat trend sus­tains the view that there is no dis­pro­por­tion­ate un­do­ing of GVCs de­spite a de­cline in “vol­umes”. This could po­ten­tial­ly sub­stan­ti­ate a claim for slow­bal­iza­tion but not nec­es­sar­i­ly for de­glob­al­iza­tion in terms of “shape”, and it could con­tribute to a de­cline in world trade growth and to low­er trade elas­tic­i­ty in the fu­ture.

If the “vol­ume” of glob­al­iza­tion has de­clined, but its “shape” ap­pears un­changed, this begs the ques­tion of whether the net­work – and more specif­i­cal­ly the “out­reach” of bi­lat­er­al trade re­la­tion­ships – has changed sub­stan­tial­ly af­ter 2009. In or­der to study this, we pro­pose a third sta­tis­tics to ex­plore glob­al­iza­tion over time: the dis­tance co­ef­fi­cient in a grav­i­ty equa­tion. Re­call that the grav­i­ty equa­tion links bi­lat­er­al trade vol­umes Xod be­tween ori­gin o and des­ti­na­tion d to the (eco­nom­ic) size of the re­spec­tive trad­ing part­ners, dis­tance, eth­ni­cal, his­tor­i­cal and in­sti­tu­tion­al links be­tween coun­tries (such as lan­guage, colo­nial de­pen­dence, or le­gal her­itage) and trade agree­ments. Con­di­tion­al on all oth­er fac­tors, the dis­tance co­ef­fi­cient rep­re­sents in any giv­en year how an in­crease of dis­tance be­tween two coun­tries by 1% af­fects the vol­ume of trade be­tween them. Tra­di­tion­al­ly it has been es­ti­mat­ed to be around –1% (Head and May­er ²), that is, a 1% in­crease in dis­tance be­tween two coun­tries de­creas­es their bi­lat­er­al trade vol­ume by 1%. Con­sid­ered over time, it in­forms us of how “long-dis­tance” trade is, and whether this dis­tance has in­creased or de­creased over time. We con­sid­er three dis­tinct sam­ples of analy­sis: trade be­tween the 10 largest economies in 2020 (in terms of cur­rent GDP, Fig­ure 4 a), be­tween the 50 largest (Fig­ure 4 b), and be­tween the 100 largest (Fig­ure 4 c).

Sev­er­al facts emerge from the com­par­i­son of the three fig­ures. First, the scale of the dis­tance co­ef­fi­cient re­veals that dis­tance has very lit­tle im­pact on bi­lat­er­al trade for the rich­est economies, but more so when rel­a­tive­ly small­er economies are con­sid­ered: on av­er­age the dis­tance co­ef­fi­cient is around –0.4 for the ten largest economies, –0.8 for the 50 largest, and –1 when we ex­tend the sam­ple to the 100 largest economies. Sec­ond, the ef­fect of the Great Re­ces­sion on the elas­tic­i­ty of trade vol­umes to dis­tance is some­what lagged, as we see the sharpest de­clines in three out of four fig­ures be­tween 2010 and 2011 (note that there is a rel­a­tive­ly small­er ef­fect of the Great Re­ces­sion on the dis­tance co­ef­fi­cient when a larg­er sam­ple is con­sid­ered). This is not nec­es­sar­i­ly sur­pris­ing, as trade re­la­tion­ships may take time to ad­just both on the in­ten­sive and on the ex­ten­sive mar­gin.

Distance has very little impact on bilateral trade for the richest economies, but more so when relatively smaller economies are considered.

Third, the evo­lu­tion of how “long-dis­tance” trade is dif­fers a lot de­pend­ing on the set of coun­tries one stud­ies. Not only does trade be­tween the 10 largest economies bare­ly re­spond to dis­tance, but this elas­tic­i­ty hasn’t changed at all dur­ing the pe­ri­od 1996–2020 ex­cept dur­ing the Great Re­ces­sion.³ There is a lit­tle and sig­nif­i­cant uptick of the co­ef­fi­cient in 2001/2002 that can be ex­plained by the en­try of Chi­na in the WTO, fol­lowed by a small down­ward trend af­ter 2005, but over­all, the pre-2009 trend is flat. When in­tro­duc­ing the next 40 largest economies in the sam­ple, the elas­tic­i­ty of trade to dis­tance in­creas­es (in ab­solute val­ue) and its time-trend changes. There is a sig­nif­i­cant de­crease in the (ab­solute) val­ue of the dis­tance co­ef­fi­cient be­tween 1996 (–0.83) and 2010 (–0.72), sug­gest­ing that as glob­al­iza­tion in­ten­si­fies, dis­tance be­comes less and less of a hur­dle to trade. As for oth­er sta­tis­tics, the Great Re­ces­sion (or slight­ly af­ter) marks a struc­tur­al break, and the co­ef­fi­cient re­mains rel­a­tive­ly flat af­ter­wards around –0.75 (based on the 95% con­fi­dence in­ter­val, the co­ef­fi­cients be­tween 2009 and 2020 are not sig­nif­i­cant­ly dif­fer­ent from each oth­er). A dif­fer­ent pat­tern emerges when con­sid­er­ing the 100 largest economies: the ef­fect of dis­tance on bi­lat­er­al trade vol­umes re­mains rel­a­tive­ly con­stant un­til 2014 (with a small but sig­nif­i­cant break in 2010/2011) around –1.05, and ex­hibits a sig­nif­i­cant and steep in­crease af­ter­wards. In 2020, a 1% in­crease in dis­tance be­tween two of the 100 largest economies led to a re­duc­tion of bi­lat­er­al trade vol­umes of “only” 0.88%. This im­plies that as we in­tro­duce small­er coun­tries in our sam­ple, with small­er bi­lat­er­al ex­changes, the “out­reach” of glob­al­iza­tion changes, and so should our in­ter­pre­ta­tion of whether there is a de­glob­al­iza­tion or not.

Tak­en to­geth­er, all three pic­tures point to­wards an in­crease in world trade in­te­gra­tion, also in the most re­cent decade, with no par­tic­u­lar sign of de­glob­al­iza­tion or slow­bal­iza­tion in terms of “out­reach”. Com­bined with the oth­er sta­tis­tics we pre­sent­ed, we ob­serve at best a sta­bi­liza­tion of trade in­te­gra­tion, which need not be in con­tra­dic­tion with a de­clin­ing Trade-to-GDP ra­tio.

Taken together, all three pictures point towards an increase in world trade integration, also in the most recent decade, with no particular sign of deglobalization or slowbalization in terms of “outreach”.

Potential Drivers: Is there a substantial change of paradigm?

To fur­ther ex­plore whether the Great Re­ces­sion marks the peak equi­lib­ri­um of glob­al­iza­tion or tru­ly a tip­ping point lead­ing to a slow but sure de­cline, we con­sid­er what could be the dri­ving forces be­hind a change of par­a­digm. Amongst the pos­si­ble can­di­dates, we ex­plore in turn la­bor costs, the in­creas­ing role of trade in ser­vices rel­a­tive to trade in goods, and the role of geopo­lit­i­cal un­cer­tain­ty and fear of GVCs dis­rup­tion in a trade “re­gion­al­iza­tion”.

One of the main pos­si­ble dri­ving forces for changes in the shape of glob­al val­ue chains in the re­cent pe­ri­od is the evo­lu­tion of the cost of la­bor across coun­tries. In par­tic­u­lar, it is some­times ar­gued – by ad­vo­cates of “reshoring” in par­tic­u­lar – that the cost of la­bor in Asian economies has steadi­ly in­creased over the past two decades, ren­der­ing the use of off­shored fa­cil­i­ties less at­trac­tive, as Chi­nese salaries lev­el with Amer­i­can ones. This would in turn fuel a de­glob­al­iza­tion trend. Fig­ure 5 shows that this isn’t ex­act­ly true. Chi­nese GDP per capi­ta has been mul­ti­plied by 15 be­tween 1996 and 2020, an in­crease that isn’t ob­served for many oth­er coun­tries. How­ev­er, GDP per capi­ta in Chi­na in 2020 is $10,500, where­as the US GDP per capi­ta is $64,000 and Ger­man GDP per capi­ta is $46,000. So the idea that the cost of la­bor in Chi­na is now sim­i­lar to the cost of la­bor in some of its largest trade part­ners does not hold in the data. The com­par­i­son with oth­er low-in­come economies how­ev­er sug­gest that there may be some ground for “friend-shoring”, that is the re­lo­ca­tion of off­shored fa­cil­i­ties clos­er to home in “friend” coun­tries with low­er la­bor costs. For the US for ex­am­ple, a nat­ur­al can­di­date is Mex­i­co, whose GDP per capi­ta in 2020 was $8,000, so in­deed low­er than Chi­na’s ap­prox­i­mate cost of la­bor. For the EU, one can con­sid­er for ex­am­ple Türkiye, whose GDP per capi­ta in 2020 was $8,500, as a sub­sti­tute for Chi­na’s tex­tile in­dus­try.

Fig­ure 6 how­ev­er shows that this re­ver­sal of la­bor cost rank­ings is very re­cent: in our ex­am­ple, Chi­na’s GDP per capi­ta be­came high­er than Mex­i­co’s or Türkiye’s first in 2017. In 2017, Chi­na still ranked 37 out of 50 in terms of GDP per capi­ta, and only reached rank 31 in 2020. As a re­sult, la­bor costs con­sid­er­a­tions are un­like­ly to have dri­ven any of the ob­served ag­gre­gat­ed trends, and over­all are more like­ly to pro­mote friend-shoring rather than reshoring.

Labor costs considerations are unlikely to have driven any of the observed aggregated trends, and overall are more likely to promote friend-shoring rather than reshoring.

An­oth­er po­ten­tial force at play in the per­ceived change of in­ter­na­tion­al trade par­a­digm is the ris­ing im­por­tance of trade in ser­vices, which be­came par­tic­u­lar­ly salient dur­ing the COVID-19 pan­dem­ic. Most sta­tis­tics about glob­al­iza­tion re­fer to or are dri­ven by trade in goods. How­ev­er, if ser­vices rep­re­sent an in­creas­ing share of eco­nom­ic ac­tiv­i­ty and by tran­si­tiv­i­ty of trade flows, the de­glob­al­iza­tion or slow­bal­iza­tion may only re­flect the de­cline of trade in goods rel­a­tive to trade in ser­vices, rather than a de­cline of in­ter­na­tion­al trade al­to­geth­er. Fig­ure 7 il­lus­trates the in­creas­ing share of trade in ser­vices in to­tal trade for a se­lect­ed group of coun­tries: Chi­na, the EU (incl. in­tra-EU trade), In­dia, Japan and the US.

There’s a clear in­creas­ing trend for all coun­tries ex­cept Chi­na, and the mag­ni­tudes we find are sig­nif­i­cant: in 2020, 35% of US trade was trade in ser­vices, and 16% of EU trade. The graph shows that the Great Re­ces­sion also had a short but sig­nif­i­cant im­pact on the share of trade in ser­vices for the US, Eu­rope and Japan, and a large im­pact in 2020 at the on­set of the COVID-19 pan­dem­ic. This pat­tern of re­sponse to shocks sug­gests that while trade in ser­vices may par­tial­ly ex­plain the de­cline in Trade-to-GDP ra­tio, it can­not be used to ex­plain a “tip­ping point” af­ter the Great Re­ces­sion.

The mul­ti­pli­ca­tion of trade crises in the re­cent decades (Great Re­ces­sion, Brex­it, COVID-19, Ukraine war) has shed light on the fragili­ty of glob­al val­ue chains and ex­ac­er­bat­ed the need to en­sure the re­silience of trade net­works. This has fu­eled dis­cus­sions on reshoring and friend-shoring, name­ly the repa­tri­a­tion of GVCs do­mes­ti­cal­ly or with­in spheres of geopo­lit­i­cal in­flu­ence. Us­ing our sta­tis­tics on GVC par­tic­i­pa­tion we can pro­vide some fac­tu­al ev­i­dence on that de­bate.

The multiplication of trade crises in the recent decades has shed light on the fragility of global value chains and exacerbated the need to ensure the resilience of trade networks.

Re­gion­al­iza­tion is no new con­cept. Fig­ure 8 il­lus­trates the strong re­liance upon in­tra-re­gion­al trade for most ge­o­graph­ic ar­eas over time. In that re­gard, Eu­rope does not ap­pear as an out­lier fa­vored by its cus­toms union: while the Eu­ro­pean share of in­trare­gion­al trade re­mains sta­ble around 70% be­tween 1995 and 2021, this share is at 50% for the Amer­i­c­as. An in­ter­est­ing dy­nam­ic can be ob­served for Asia: the share of in­tra-re­gion­al trade in­creas­es from 50 to 60% be­tween 1998 and 2021. From this pic­ture it ap­pears that any re­gion­al­iza­tion trend would have start­ed long be­fore the Great Re­ces­sion. Africa is a clear out­lier in this pic­ture, with a share of in­tra-re­gion­al trend av­er­ag­ing at 15%. It is the only re­gion for which a “re­gion­al­iza­tion” ef­fect seems to have picked up af­ter 2008.

Fig­ure 9 fur­ther il­lus­trates this point by de­com­pos­ing GVC par­tic­i­pa­tion of dif­fer­ent re­gions into the con­tri­bu­tion of re­gion­al and ex­tra-re­gion­al trade. The preva­lence of re­gion­al val­ue chains for most ar­eas, even pri­or to 2008 is co­her­ent with Fig­ure 8. For coun­tries in Eu­rope and Asia, and to a less­er ex­tent in North Amer­i­ca, GVC par­tic­i­pa­tion has risen large­ly on the back of stronger sup­ply link­ages with­in the re­gion it­self, while coun­tries in Latin Amer­i­ca have be­come in­te­grat­ed in GVCs by strength­en­ing link­ages with part­ners from oth­er re­gions.

Since 2011, re­gion­al GVC links have some­what weak­ened in Asia and Latin Amer­i­ca, while in oth­er re­gions they have re­mained broad­ly con­stant. Over­all, sup­ply chains re­main clus­tered at re­gion­al lev­el, with no clear in­creas­ing trend.

Tak­ing stock, it ap­pears that the ev­i­dence point­ing to­wards a de­glob­al­iza­tion or slow­bal­iza­tion is lim­it­ed. The de­cline of the Trade-to-GDP ra­tio since the Great Re­ces­sion is a clear in­di­ca­tion that “vol­umes” of trade are de­clin­ing, al­though ev­i­dence sug­gests that part of this can be ex­plained by a sub­sti­tu­tion of trade in goods by trade in ser­vices. How­ev­er, de­glob­al­iza­tion does not sim­ply hint at a de­cline in “vol­umes”, but also at a change of the “shape” of glob­al­iza­tion. In that re­spect, there is no ev­i­dence that the Great Re­ces­sion rep­re­sent­ed a tip­ping point for glob­al val­ue chain par­tic­i­pa­tion. If any­thing, our sta­tis­tics show that in that re­gard glob­al­iza­tion has in­ten­si­fied in the most re­cent decade, al­low­ing small economies (and pre­sum­ably poor­er ones) to in­te­grate the world trade sys­tem, even more so at the re­gion­al lev­el, and to sup­port more friend-shoring rather than reshoring.

There is no evidence that the Great Recession represented a tipping point for global value chain participation.

Policy Landscape: The role of multilateral and unilateral trade policies

An­oth­er fun­da­men­tal di­men­sion of analy­sis, for the trends in both glob­al­iza­tion and re­gion­al­iza­tion, is the role played by bi­lat­er­al and mul­ti­lat­er­al trade ne­go­ti­a­tions, as well as (uni­lat­er­al) pro­tec­tion­ist trade poli­cies. Re­cent decades have seen a pro­lif­er­a­tion of re­gion­al trade agree­ments (RTAs) which in­clude both free trade agree­ments and cus­toms unions (Fig­ure 10). As of Sep­tem­ber 2022, there are 353 of such agree­ments in force. The com­ing into be­ing of the World Trade Or­ga­ni­za­tion (WTO) in 1995 is con­sid­ered one of the main rea­sons for the sharp in­crease in the num­ber of RTAs. This process of po­lit­i­cal glob­al­iza­tion start­ed to slow down fol­low­ing the Great Re­ces­sion: the num­ber of no­ti­fi­ca­tions per year in­creased steadi­ly from the mid-1990 and peaked in 2009. It start­ed de­creas­ing slight­ly be­tween 2010 and 2015, and it de­creased even fur­ther be­tween 2016 and 2020. The only ex­cep­tion is 2021, as a di­rect con­se­quence of Brex­it: 86% of the RTAs en­tered into force in 2021 have the Unit­ed King­dom as one of the trad­ing part­ners.

While RTAs are reg­u­lat­ed un­der the WTO and com­pat­i­ble with the glob­al trad­ing sys­tem, there is a (grow­ing) view that they can be an ob­sta­cle to a more ef­fec­tive and fair glob­al trad­ing sys­tem. The large pro­lif­er­a­tion of RTAs is in fact of­ten con­sid­ered to be a sys­temic prob­lem for the WTO, un­der­min­ing its role as a glob­al ne­go­ti­a­tion fo­rum while fos­ter­ing in­stead bi­lat­er­al ne­go­ti­a­tions.

One of the main con­cerns is that bi­lat­er­al ne­go­ti­a­tions are in­her­ent­ly dis­crim­i­na­to­ry. RTAs could be de­signed to strength­en trade re­la­tion­ships with some coun­tries and pur­pose­ful­ly ex­clude or iso­late oth­er economies. In fact, these pref­er­en­tial trade ne­go­ti­a­tions are con­ceived to cir­cum­vent the Most Fa­vored Na­tion clause of the WTO, which is about nondis­crim­i­na­to­ry han­dling of trade, by al­low­ing two or more coun­tries to set tar­iffs to zero just among them­selves. A sec­ond con­cern is that the in­ter­ests of small de­vel­op­ing economies are less pro­tect­ed in bi­lat­er­al ne­go­ti­a­tions rather than in a mul­ti­lat­er­al set­ting. Trade agree­ments be­tween de­vel­oped and de­vel­op­ing coun­tries – the so-called North-South RTAs – are thought to dis­pro­por­tion­ate­ly rep­re­sent the eco­nom­ic in­ter­ests of the “north”, the part with the strongest bar­gain­ing pow­er in the ne­go­ti­a­tion.

The large proliferation of RTAs is considered to be a systemic problem for the WTO, undermining its role as a global negotiation forum while fostering bilateral
negotiations.

This view is some­how con­firmed by look­ing at the RTAs signed by the EU, the Unit­ed States and Chi­na (Fig­ure 11). The pref­er­en­tial trade ne­go­ti­a­tions signed by the three largest trad­ing economies in the world are de­signed to strength­en their ge­o­graph­i­cal in­flu­ence. The EU with the East-Eu­ro­pean bloc and the Mediter­ranean coun­tries; the US with North Amer­i­ca and the coun­tries of the Pa­cif­ic Rim; and Chi­na with South East Asia. In light of the above dis­cus­sion, these RTAs could be seen as the build­ing block for a fu­ture trend in both re­gion­al­iza­tion and friend-shoring. But the ex­ten­sive use of these in­stru­ments might cre­ate a sub-op­ti­mal out­come in terms of fair­ness and sus­tain­abil­i­ty com­pared to mul­ti­lat­er­al ne­go­ti­a­tions.

On top of bi­lat­er­al and mul­ti­lat­er­al trade lib­er­al­iza­tion poli­cies, the oth­er im­por­tant di­men­sion is rep­re­sent­ed by uni­lat­er­al pro­tec­tion­ist mea­sures. The most re­cent and most no­table ex­am­ples of such poli­cies are Brex­it, and the uni­lat­er­al tar­iffs set by Pres­i­dent Trump against Chi­na in Jan­u­ary 2018. As shown in Fig­ure 12, pri­or to the first pack­age of US tar­iffs on Chi­nese so­lar pan­els and wash­ing ma­chines, the US tar­iffs on Chi­nese ex­ports was around 3%, while the Chi­nese tar­iff on US ex­ports was around 7%, both in line with the tar­iff rates ap­plied to the rest of the world. Fol­low­ing a se­ries of tar­iff pack­ages and re­tal­ia­to­ry re­spons­es, tar­iffs in­creased to up to 20%.

One of the most sur­pris­ing facts is the stick­i­ness of these tar­iffs, and how com­pli­cat­ed the process of de-es­ca­la­tion is. Dur­ing the 2020 US pres­i­den­tial cam­paign, Biden ex­pressed his will­ing­ness to end the trade war with Chi­na. But since his vic­to­ry, no changes were made to the tar­iff sched­ule vis-à-vis Chi­na. De­spite the fact that many eco­nom­ic pa­pers showed how detri­men­tal these tar­iffs were to so­cial wel­fare, elim­i­nat­ing them does not seem a vi­able po­lit­i­cal ac­tion.

The trade war be­tween the US and Chi­na was also mo­ti­vat­ed by Pres­i­dent Trump as an at­tempt to “de­cou­ple” the two largest economies in the World. Four years af­ter the be­gin­ning of the trade war, and with­out a clear sign of a re­ver­sal in the im­posed tar­iffs, one could look at the ef­fect of these poli­cies on trade. Chad Bown, in a re­cent col­umn for The Pe­ter­son In­sti­tute for In­ter­na­tion­al Eco­nom­ics,⁴ has an­a­lyzed the bi­lat­er­al trade pat­terns be­tween the two coun­tries, show­ing that US im­ports from Chi­na have in fact de­clined in 2018, as a re­sult of the trade war (Fig­ure 13). The COVID-19 pan­dem­ic ex­ac­er­bat­ed this neg­a­tive im­pact – on im­ports from Chi­na as well as from the rest of the world. To­day, US im­ports from Chi­na have re­turned to the pre-trade war lev­els of June 2018, but re­main way be­low the pre-trade war trend. Chi­na is now the source of only 18% of to­tal US goods im­ports, down from 22% at the on­set of the trade war.

Yet, the two economies did not ful­ly de­cou­ple. Not sur­pris­ing­ly, the largest neg­a­tive im­pact on im­ports from Chi­na was on prod­ucts hit with the high­est US tar­iffs (Fig­ure 14 a). Si­mul­ta­ne­ous­ly, US im­ports of prod­ucts that did not face an in­crease in tar­iffs surged (Fig­ure 14 b). These were main­ly goods whose de­mand in­creased in re­sponse to the COVID-19 pan­dem­ic (e.g. lap­tops and com­put­er mon­i­tors, phones, video game con­soles, and toys).

When it comes to set­ting trade poli­cies, the trend to­wards de­glob­al­iza­tion and pro­tec­tion­ism is – at least on the po­lit­i­cal di­men­sion – def­i­nite­ly more clear. These de­ci­sions are not sim­ply po­lit­i­cal, but have a clear im­pact on trade and can lead to an at­tempt to de­glob­al­ize or to re­draw the pat­terns of trade to­wards reshoring and friend-shoring.

Concluding Remarks

Claims of de­glob­al­iza­tion have mul­ti­plied over the last decade, but the ev­i­dence is still mixed. The re­cent evo­lu­tion of the glob­al Trade-to-GDP ra­tio points to­wards a de­crease in glob­al­iza­tion in terms of “vol­umes”. Look­ing at GVC par­tic­i­pa­tion – the “shape” of glob­al­iza­tion – the pic­ture sub­stan­ti­ates at max­i­mum claims of slow­bal­iza­tion. In terms of “out­reach” we see that over time dis­tance is be­com­ing less of a hur­dle to trade, sup­port­ing views of con­tin­u­ing and in­creas­ing glob­al­iza­tion. When look­ing at these facts to­geth­er, and de­spite some signs of a slow­down, the main mes­sage can still be cau­tious­ly op­ti­mistic.

The pol­i­cy land­scape, in con­trast, gives a slight­ly more wor­ry­ing pic­ture. The pro­lif­er­a­tion of re­gion­al trade agree­ments can be read as an in­crease in glob­al­iza­tion. But these are of­ten used to sed­i­ment trends in re­gion­al­iza­tion and “friend-shoring” rather than fair and non-dis­crim­i­na­to­ry world­wide glob­al­iza­tion, which can be achieved only through mul­ti­lat­er­al ne­go­ti­a­tions. More­over, on the po­lit­i­cal are­na, we are ex­pe­ri­enc­ing in­creas­ing voic­es against glob­al­iza­tion. Re­cent uni­lat­er­al pro­tec­tion­ist mea­sures, like Brex­it or the es­ca­la­tion in the tar­iff war be­tween Chi­na and the US, are al­ready hav­ing clear reper­cus­sions on trade pat­terns. And the full im­pact of these mea­sures is yet to be seen.