david and goliath? small developing countries, large emerging … · 2021. 4. 16. · 576 small...

International Studies Quarterly (2019) 63 574ndash588

David and Goliath Small Developing Countries Large EmergingMarkets and South-South Preferential Trade Agreements

DA N I E L A DO N N O

University of Cyprus

AND

NI TA RU D R A

Georgetown University

Has the rise of large emerging economies influenced the foreign economic policies of smaller nations Many of the BRICSrsquo(Brazil Russia India China South Africa) dominance in export markets for low-skilled goods pose a particular challenge forldquosurplus-laborrdquo countries characterized by large populations of unskilled and underemployed labor We theorize the incentivesof firms and governments in surplus-labor countries to form South-South preferential trade agreements (SSPTAs) as a meansof diversifying and expanding trade relationships in the face of this challenge Of all the BRICS our findings show that Chinaposes the greatest challenge the countries forming the most South-South agreements are those whose exports have been mostdisplaced by China We verify this pattern using both systemic and country-specific measures of the China ldquoshockrdquo Importsfrom China in contrast have no significant effect on SSPTA formation Our account which helps resolve the dual puzzle ofdeclining trade with rich countries and the proliferation of SSPTAs in recent decades underlines the implications of Chinarsquosrise on the developing world

A surprising number of poor countries have witnessed asecular decline in trade with advanced industrialized coun-tries since the 1990s This trend is particularly pronouncedfor less developed countries (LDCs) with large pools of un-or under-employed labor which we refer to as surplus-laborcountries Figure 1 (top panel) shows that as a percent ofgross domestic product (GDP) exports to developed coun-tries have been steadily declining for surplus-labor LDCs es-pecially since the mid-1990s Likewise the gap between thisgroup and other nationsrsquo access to trade partnerships withrich countries (ie North South Preferential Trade Agree-ments NSPTAs) has been increasing over the same timeframe (Figure 1 bottom panel) This pattern raises ques-tions as to whether the international system set up to en-courage trade between rich and poor countries is under fireAre poor nations withdrawing from the liberal internationalorder

Research abounds on the sustainability of the multilat-eral trading system but mostly from the perspective of richcountries Those that do consider developing nations havefocused at length on the rise of large emerging markets orthe BRICSmdashBrazil Russia India China and South Africamdashand their simultaneous implications for great powers1 But

Daniela Donno is an associate professor of political science at the University ofCyprus Her research explores the international influences on democracy humanrights and economic policy including the formation of international institutions

Nita Rudra is a professor of government at Georgetown University Her re-search analyzes the international economic influences on fiscal policies povertydemocracy and the sources of antiglobalization sentiments

Authorsrsquo notes Our sincere thanks to Leonardo Baccini Stephen ChaudoinPeter Rosendorff Dennis Quinn and Stephen Weymouth as well as attendees atthe International Political Economy Society (IPES) meeting and the AmericanPolitical Science Association (APSA) annual meeting Yumna Naqvi Kristen Skill-man Shawna Metzger Zachary Karabatak Inu Manak Duncan Hobbs and RabeaKirmani provided excellent research assistance

1 Many speculate that rising powers are fundamentally changing the rules ofglobal trade opposing US hegemony and as a result increasing global political-economic instability Others counter this view maintaining that the BRICS possess

scholars have given relatively less consideration to how de-veloping nations are attending to fears of shifting globalpower constellations As the BRICS have assumed largershares of global export marketsmdashparticularly in labor-intensive low-skilled industries such as textile and apparelssmaller LDCs struggle to compete for access to coveted rich-country markets2 This can hamstring the growth of exportindustries that are critical for creating employment oppor-tunities and bolstering political stability Chinarsquos rise poses aparticular challenge due to its sheer size productivity anddominance of markets in low-skilled goods

Drawing from new new trade theory and theories ofcollective action we argue that governments of surplus-labor LDCs are responding to the growth of BRICS ex-ports by seeking new trade partnerships with other develop-ing economies By forging South-South preferential tradeagreements (SSPTAs) LDC governments seek to create al-ternate market opportunities for and thereby politically sa-tiate globally less-productive firms who are the losers in thisincreasingly competitive environment Both exporting andimport-competing firms may have reason to support SSP-TAs though exporting sectorsmdashtypically dominated by asmall number of larger firmsmdashhave more uniform incen-tives and are better able to overcome collective action costsfor lobbying

Like Baccini and Duumlr (2012) we focus on exporterinterests in supporting PTA formation but in our accountthe impetus stems not from exclusion from other PTAs but

little capacity or interest in challenging the global economic order Common toexisting accounts is a penchant for descriptive data and broad predictions aboutchallenges to rich countriesrsquo global authority

2 For examples complaints in Pakistan that the global textile business is domi-nated by India and China (Dawn News 2015) criticism in Africa that engagementwith China is ushering in a ldquonew form of imperialismrdquo (Economist 2015) backlashagainst China in Latin America where ldquodomestic-oriented business organizationsand unions have demonstrated and publicly criticized what is seen as unfair com-petition in domestic and global marketsrdquo (Peters 2015)

Donno Daniela and Nita Rudra (2019) David and Goliath Small Developing Countries Large Emerging Markets and South-South Preferential Trade Agreements InternationalStudies Quarterly doi 101093isqsqz041copy The Author(s) (2019) Published by Oxford University Press on behalf of the International Studies Association All rights reservedFor permissions please e-mail journalspermissionsoupcom

Dow

nloaded from httpsacadem

icoupcomisqarticle6335745532837 by guest on 16 April 2021

DA N I E L A DO N N O A N D NI TA RU D R A 575

Figure 1 Surplus-labor LDC exports to developed coun-tries (top) and NSPTA memberships over time amongsurplus-labor and non-surplus labor LDCs (bottom) NotesTop panel Included countries are those with levels of sur-plus labor above the mean for a majority of years in the sam-ple See the ldquoEvidencerdquo section for details on the coding ofsurplus labor Bottom panel NSPTA average excludes highsurplus-labor nations

from the displacement created by large emerging markets(see Figure 1) Although SSPTAs do not necessarily bringthe large-scale increases in trade associated with North-South agreements for firms in surplus-labor nations theycan nevertheless open up new markets in countries withsimilar consumer preferences and quality expectationsand they provide opportunities for learning-by-exporting asthey strive to join multinational supply chains (Blalock andGertler 2004 De Loecker 2013) For governments thenSSPTAs can bring political benefits by signaling that leadersare taking action to address economic conditions If ourintuitions are correct smaller LDCs are not withdrawingfrom the liberal international order rather in response tothe loss of rich-country markets to rising powers they arepursuing multilateralism of a different form

We conduct a range of empirical tests to assess this in-tuition Though built on firm-level foundations our the-ory generates country-level hypotheses about the conditionsunder which governments form more trade agreementsParticularly among surplus-labor countries we find Chinarsquosshare of world export markets is significantly associated withSSPTA formation We employ two measures of Chinarsquos ex-port ldquoshockrdquo to developing economies one systemic andone country-specific First using a systemic measure of eachBRICS countryrsquos share in global export markets our find-ings reveal that Chinarsquos rise is closely associated with therapid proliferation of SSPTAs and we confirm that this ef-fect surpasses that of any other BRICS state Employing thecountry-specific measure of the China shock we find thatsurplus-labor countries whose export markets have been mostdisplaced by China are those joining the most SSPTAs Anincrease in Chinese imports is not a significant predictor ofSSPTAs suggesting that pressure from local firms threat-ened by the flood of Chinese imports has less impact ontrade policy than the relatively smaller number of exportersdisplaced by China

A difference-in-difference analysis provides further evi-dence in support of our theory Chinarsquos membership inthe World Trade Organization (WTO)mdasha critical juncture

Figure 2 South-South (SS) PTA memberships over timeNote Figure depicts the average number of SSPTA member-ships per country by year The sample includes developingcountries only

in its global economic integrationmdashis associated with anincrease in SSPTA membership particularly among surplus-labor countries that do not border a developed coun-try (and thus lack a geographic advantage for access torich-country markets) In short using various measures weconsistently find that surplus-labor countries most adverselyaffected by Chinarsquos rise are those that are forming SSPTAsin the largest numbers

Our account helps explain the proliferation of SSPTAs inrecent decades Per country the average number of SSPTAmemberships has increased sharply since 1990 While 136SSPTAs were formed in the forty-year period between 1948and 1989 398 were formed in the two decades that followed(1990ndash2009) Figure 2 shows that most of these agreementshave been negotiated by non-BRIC LDCs with each otherNorth-South PTAs have also proliferated since 1990 (as seenin Figure 1) but importantly the bulk of these agreementsare bypassing high surplus-labor LDCs3 Most research fo-cusing on domestic political factors as the drivers of PTAformation fails to consider these distinctions between North-South and South-South agreements (Mansfield Milner andRosendorff 2002 Mansfield Milner and Pevehouse 2007Mansfield and Pevehouse 2013)4 Here we bring renewedattention to systemic economic forces that in combina-tion with domestic lsquopolitical economyrsquo conditions providea novel explanation of which countries are forming whichtypes of PTAs and when

Our analysis also sheds light on the consequences of ris-ing powers for global governance (eg Ikenberry 2008Kahler 2013 Armijo and Roberts 2014) We reject the im-plicit assumption that small nations are the pawns of powerpolitics (Hurrell 2006 Kahler 2013 Stephen 2014 ZengXiao and Breslin 2015) and theorize these countries asagents with a distinct role in shaping the evolution of in-ternational market expansion as well as the global institu-tional landscape Our findings in short suggest that therise of large emerging market economies is changing theliberal economic order but not necessarily in the direction

3 Thirty-three NSPTA agreements were successfully negotiated between 1948and 1989 and 116 from 1990 to 2009

4 But see Baccini (2011) on democratization and South-South agreementsMansfield and Milner (2012) identify how democracy and veto players shape in-centives for PTA formation but they stop short of exploring whether there maybe unique incentives that govern PTAs among developing nations

Dow



576 Small Developing Countries Large Emerging Markets and South-South Preferential Trade Agreements

of chaos5 Rather the encouragement of greater linkagesamong developing countries may be stability-enhancing onboth domestic and international levels

The Challenge of Rising Powers for DevelopingCountries

Since the second half of the twentieth century LDCs havebeen attempting to hasten the industrialization process byencouraging the development of manufacturing firms thatcan export Access to rich-country markets is particularlylucrative because they offer greater opportunities for pro-ductivity gains and export profits (Aw Roberts and Xu2011 Crinograve and Epifani 2012 Kang 2013)6 Yet this strat-egy has become increasingly challenging over the last threedecades as BRICS nations have been liberalizing and suc-cessfully meeting much of the global demand for labor-intensive manufactured goods Exports from small LDCs torich-country markets have been steadily declining since themid-1990s (Figure 1) As Gereffi and Sturgeon (2013 339)explain

[l]arge emerging economies clearly have more op-tions in terms of upgrading within GVCs [global valuechains] than small economies Large countriesclearly have more leverage in such arrangementsLarge countries with high potential for market growth(such as the BRICS) can also institute policies to driveFDI in technologymdashand capital-intensive sectors suchas electronics and motor vehicles

China has been especially successful and resilient inthis regard Among the BRICS China is unique becauseof its combination of size and productivity which is muchhigher than countries in similar income brackets (Rodrik2006) China has maintained a steady increase in its ex-ports to rich nations and has become the preeminant part-ner in the global supply chain for both labor-intensive andskill-intensive goods making North-South trade in man-ufactured goods less accessible for many poor develop-ing economies (Wood 1997 Kaplinsky 2006 Hanson andRobertson 2010)7 Appendix 1 illustrates Chinarsquos persistentdominance in a key labor-intensive sector textiles and cloth-ing8 The reality is that China is keeping its existing compar-ative advantage in low-skilled goods while continuing to ac-quire new ones that ldquostraddle the full span of technologiesand labor intensitiesrdquo 9 Liberalizing economies as diverseas Egypt Guatemala Vietnam Ghana and Bangladesh arefinding their still-nascent manufacturing industries facingincreasingly stiff competition from China (Tull 2006 Zafar2007 Alvarez and Claro 2009)

Among the BRICS China is the only developing econ-omy that ranks consistently as one of the top five trad-ing partners for almost every developed country over the

5 See Lieber (2014) Layne (2012) Gu Humphrey and Messner (2008)Christensen (2006) and Tull (2006)

6 This is especially true for firms located in comparative advantage industries(Bernard Jensen Redding et al 2007) or in industries that are vertically inte-grated with firms in developed countries (Manger 2009)

7 See Zylstra (2012) According to one business report the following questioninevitably rises in meetings and conferences around the world ldquowhatrsquos your Chinastrategyrdquo (Jhangiani and Stocking 2006)

8 The Financial Times (2008) notes that while many countries in Africa havewelcomed China ldquoChinese textile imports have caused more than 80 percent ofNigeriarsquos textile factories to shut down An estimated 250000 Nigerian workershave been laid off as a resultrdquo

9 See Yusuf and Nabeshia (2009) and Schellekens (2013 7)

last two decades10 Brandt Van Biesebroeck and Zhang(2012) find that Chinarsquos firm-level total factor productivitygrowth from 1999 to 2006 has been extremely high increas-ing from 29 percent to approximately 14 percent Policy-makers are duly concerned11 and many economists con-cur that its dominance threatens developing countriesrsquo ex-port prospects (eg Blazquez-Lidoy Rodriacuteguez and Santiso2006 Freund and Ozden 2006 Jenkins and Edwards 2006Jenkins Peters and Moreira 2008 Wood and Mayer 2011Montalbano and Nenci 2012) Accordingly we anticipatethat China of all the BRICS poses the greatest challengeto smaller LDCs aiming to penetrate global export marketsin labor-intenstive goods

We also anticipate that not all LDCs are affected byChina in the same way those hosting large pools of youngunder- or un-employed low-skilled (or unskilled) labor arelikely to be the most vulnerable We term these surplus-laborcountries Their economic landscape is dominated by lessglobally-competitive small and medium enterprises (SMEs)which account for up to 90 percent of businesses outsideagriculture (Stein Ardic and Hommes 2013) and whichtend to produce low-quality goods (World Trade Organi-zation 2016) As Chinarsquos exporters have met rising con-sumer demands to produce higher quality low-cost less-skilled goods in a time-sensitive fashion (WTO 2013a 12)the survival of both export-oriented and import-competing firmsin surplus-labor LDCs is potentially at stake

Challenges to Exporting Firms

Export-oriented firms in surplus-labor LDCs face high tradecosts with rich nations from the outset12 Poor infrastruc-ture (eg roads and communication networks) transporta-tion costs cumbersome border customs and port clear-ance procedures are perhaps the most prevalent examples(see Djankov Freund and Pham 2010 Nordas Pinali andGrosso 2006 Milner and McGowan 2013) Maskus Otsukiand Wilson (2005) estimate the substantial production coststhat firms in developing nations face in conforming to prod-uct and regulatory standards imposed by major importingcountries is approximately $425000 per firm or 47 per-cent of value added on average And this is despite (orsome argue because of) the steep decline in transportationcosts and improvements in information and communicationtechnology (Ernst and Kim 2002 Christopher Peck andTowill 2006 Aw et al 2011 WTO 2013b Hottman Reddingand Weinstein 2014) In effect high surplus-labor countriesare high trade-cost economies

This puts exporters at a disadvantage vis-agrave-vis economiessuch as China Only larger productive firms can incur thehigh entry costs to rich-country markets (Melitz 2003)13

While the number of such ldquosuperstarrdquo firms is rising inChina exporters based in low-income countries are in-creasingly becoming ldquolosersrdquo as they struggle to overcomepersistently high trade costs One Pakistani newspaper

10 See BACI 2010 Department of Foreign Affairs [Australia] (2013) IndustryCanada (2014) Comtrade (2014) and United States Census Bureau (2014)

11 See for example BBC (2005)12 The World Bank Trade Costs Database applies a ldquotrade-costsrdquo measure that

includes international shipping and logistics costs tariff and nontariff costs andcosts from different language culture and currencies The data show that BRICSrsquotrade costs with rich countries are on average 44 percent lower than that for otherdeveloping economies

13 There is ample evidence that firms that invest in new technologies laborproductivity and product quality have a higher probability of survival in exportmarkets (Verhoogen 2008 Lileeva and Trefler 2010 Aw Roberts and Winston2007)

Dow




summarizes how constraints such as aging machines and rawmaterial shoratages puts local textile firms at a disadvantagevis-aacute-vis rising powers

Pakistan is no longer in the race for a bigger sharein the global textile business Even our domestic mar-ket has been stolen from us by China and India Weare just struggling to survive and save our jobs (DawnNews 2015)

Likewise retailers in developed-country markets notehow Chinarsquos ascendance has come at a cost to other devel-oping nations

For quite some time there has been a whole raft ofAsian countries that New Zealand firms dealt withmdashthe Philippines Indonesia Malaysia Almost ev-ery industry is now focused tightly on China (NewZealand Herald 2005)

Challenges to Import-Competing Firms

Chinarsquos exports may also overwhelm import-competingfirms in smaller developing economies (Ademola Abiodunand Adewuyi 2009 Kaplinksy et al 2010 Herman 2011 DiGiovanni Levchenko and Zhang 2014 Golub et al 2017)Ademola et al (2009) calculate that China accounted foran extraordinary 934 percent of Africarsquos manufactured im-ports in 2007 This poses a challenge to local industries thatproduce labor-intensive goods such as textiles and footwearChinarsquos purported dumping in poor countries has incitedmuch local controversy and scholars contend that Chineseimports are correlated with significant job loss and loss ofmarket share particularly in the textile sector (KaplinskyMcCormick and Morris 2010 Nhlabatsi 2014) The prob-lem is further compounded by the fact that governmentsof surplus-labor countries are least likely to take successfulantidumping duty actions against China using dispute set-tlement litigation at the World Trade Organization (Bownand Hoekman 2005 Kim 2008)

South-South Trade Agreements

The losers in this shifting economic landscapemdashless globallycompetitive domestic firms and the workers they employmdashare demanding compensation from their political leaders 14

We expect the greatest pressures to come from large ex-porters (ie former ldquosuperstarrdquo firms that are forced to exitrich-country markets or are prevented from entering themin the first place) Such firms enjoy a collective action ad-vantage because of their size and small numbers and theypossess the financial and political capital to lobby for pol-icy solutions (Olson 1979 Hellman Jones Kaufmann et al2000)15 LDC governments are sensitive to such pressurenot only do they rely on the support of large businesses theyare potentially vulnerable to instability caused by growingunemployment (ILO 2016)16

Policy-makers thus have incentives to find alternative waysof increasing access to international markets Here we de-velop the core insight that forging trade agreements withother developing nationsmdashparticularly regional agreements

14 See for example Indo-Asian News Service (2006) Bangkok Post (2006)Business Recorder (2011)

15 In general the percentage of firms that export is small in some of the rich-est economies 10 percent of a nationrsquos firms account for more than 80 percentof exports (Bernard Jensen and Redding et al 2007)

16 The rising share of unemployed young people is of particular concern forleaders of small LDCs (ILO 2016) High youth unemployment is associated withpolitical instability and conflict (Urdal 2006 Collier 2000)

among neighborsmdashis an important tool that is both lesscostly than implementing structural reforms and can bringpolitical benefits by addressing the demands of exportingand (some) import-competing firms In effect SSPTAs canhelp compensate for the steady decline in trade (and tradepartnerships) with rich economies

These insights cut against the view that the economicbenefits from South-South trade are limited although theymay be lower than they would with rich countries17 Agrowing body of economics research suggests the bene-fits of reducing trade barriers among developing countries(Dollar 2005 World Bank 2002 Lindert and Williamson2003 Mohan Khorana and Choudhury 2012) Studies havefound that South-South trade agreements are associatedwith lower tariffs and increased trade albeit in some sec-tors and in some countries more than others (Greenawayand Milner 1990 Langhammer 1992 Baier and Bergstrand2007 Mayda and Steinberg 2009 Behar and Criville 2013Gray 2014 Shepherd 2016) Although some agreements fallshort in implementation this varies substantially across cases(Gray 2014) In sum though SSPTAs are not a substitute forNorth-South trade they represent a parallel strategy thatseeks to address the demands of political groups and setcountries on a path toward new trade partnerships We pro-ceed to outline how and why export-oriented firms import-competing firms and governments seek SSPTAs in responseto the China export threat or lsquoChina shockrsquo

Export Firms

Exporting firms in developing economies can benefit fromseeking new markets in other developing countries18 Re-call that productive Chinese firmsrsquo ability to overcome hightrade costs with rich countries stands in contrast with firmsbased in surplus-labor countries The latter face hurdles inovercoming trade costs in critical domains (eg regulatorystandards strong and well-functioning infrastructure com-petency and quality of logistics services) As Baccini andDuumlr (2012) argue exporters have incentives to mobilizewhen facing losses especially when foreign countries jointrade partnerships that exclude them Moreover researchshows that it is large productive exporting firms that benefitdisproportionately from preferential liberalization (BacciniPinto and Weymouth 2017 Baccini forthcoming) Buildingon this idea we maintain that exporters demand PTAs as aform of ldquocompensationrdquo for the loss of rich-country markets

In an effort to continue exporting intraregional tradein particular may become more appealing Negotiatinglower trade costs is easier with neighboring countries fac-tors such as language common history sharing a borderand participation in the same economic community canhelp firms strike successful agreements that can incorpo-rate a wide range of activities aimed at reducing tradecosts (OECDWTO 2015) Notably studies find geograph-ical proximity has the largest trade-improving effect relativeto all the other factors impacting trade costs (Arvis et al2013)

17 The Heckscher-Ohlin trade model predicts trade patterns on the basis ofcountriesrsquo factor endowments Larger welfare gains from trade occur when laborabundant (poor) nations export labor-intensive goods to capital-intensive (rich)countries

18 Firms generally seek markets where they face less interfirm rivalry (Koch2001 Johansson 1997 Porter 1986) As Johansson (1997 16) argues competitiverivalry leads to an emphasis on speed and delivery in new product developmentFirms in smaller LDCs have more difficulty meeting such demands when compet-ing for access to rich-country markets

Dow




For exporters (and potential exporters) the impact ofSSPTAs may ultimately be broader than many critics positFirst large low-skill manufacturersmdashsuch as those in theapparel and leather industrymdashare well positioned to learn-by-exporting thereby increasing productivity (Blalock andGertler 2004 De Loecker 2013) and innovation (Salomonand Shaver 2005) Second SSPTAs can provide opportuni-ties for firms to gain economies of scale by increasing theirsize or speed of operation as well as create bilateral or re-gional foundations for enhancing supply chain capabilitiesFor these reasons SSPTAs may ultimately be avenues formarket expansion for (now) globally less competitive firmsin surplus-labor countries potentially serving as a launch-pad for increases in trade with other countries

Import-Competing Firms

Support for SSPTAs among import-competing firms may notbe as strong or as uniform as among large exporters Import-competing firms face considerable mobilization challengesgiven that they are larger in numbers and more hetero-geneous in size and productivity Their incentives are alsomore mixed On one hand some import-competitors maywelcome the benefits of lower trade barriers with develop-ing economies that will help make them exporters But theleast productive import-competing firms may be threatenedby yet more competition from new partner countries andtherefore resist any steps toward trade liberalization evenwith other developing countries It is ultimately an empir-ical question whether the population of import-competingfirms have enough overlapping incentives to overcome col-lective action problems and form a coalition in support ofSSPTAs

Governments

The turn to South-South trade agreements in response torising economic powers is in many ways a new historical de-velopment To be sure in the sixties and seventies smallLDCs such as South Korea and Singapore faced similar polit-ical and economic challenges when other developing coun-tries were advancing in export markets for labor-intensivegoods The difference is that these nations had strong lead-ers and US market access that helped alter their compar-ative advantage through ldquocorrectrdquo pricing setting realisticexchange rate policies incentivizing key industries and in-vesting in human capital development (Lim 1983 Amsden1989) The surplus-labor economies of today do not havethe resources or state capacity to achieve such goals More-over in the present historical context the ability to use sub-sidies is limited and access to US markets far more challeng-ing than in the past

Thus facing the challenge of trade-oriented liberaliza-tion in a global marketplace dominated by low-skilledChinese goods LDC governments have limited policy op-tions One solution is to alleviate behind-the-border con-straints as the Asian Tigers continue to do for example byengaging in labor market reforms skill upgrading and tech-nological advancement But such solutions involve struc-tural changes that are financially and politically costly forpoor nations Even if governments have the capacity to un-dertake difficult reforms to facilitate entry to rich-countrymarkets uncertainty remains particularly given Chinese ex-portersrsquo advantages in terms of lower trade costs and gov-ernment policy instruments to promote their high pro-ductivity levels such as research and development directgrants (shangji bokuan) and tax incentives (jianmian shui)

(Hu Jefferson and Jinchang 2005) For the large number ofimport-competing firms protectionist policies (eg highertariffs quantitative restrictions) are desirable but politicalleaders may be rightly concerned that price-sensitive con-sumers in LDCs will punish them for higher prices Antago-nizing China may also have other negative repercussions fordeveloping countriesrsquo governments such as access to gen-erous amounts of Chinese foreign aid investment exportcredits and bank finance (Brautigam 2011)

South-South trade agreements represent a more feasibleand palatable policy option As outlined above politicallyinfluential exporters have incentives to support such agree-ments Just as critically for leaders SSPTAs may be polit-ically easier to negotiate than North-South agreements Itis easier for firms in developing nations to trade in lower-quality goods considered unacceptable by consumers inrich nations and below the World Trade Organizationrsquos(WTO) global benchmarks for health and safety (Hudsonand Jones 2003 Hallak 2006) In South-South agreementsless globally productive exporting firms can carry on largelywith business as usual securing access to neighboring mar-kets without requiring the costly structural changes typi-cally demanded by trade agreements with rich countriesSSPTA partners have to resolve fewer differences in do-mestic regulatory regimes labor investment intellectualproperty rights and government procurement as well asfeaturing significantly less (or zero) pressure to sign ontohigher human and labor rights standards (Chauffour andKleimann 2012 Osnago Roch and Ruta 2015) It is alsoconsiderably easier for developing nations to agree amongthemselves on sanitary and phytosanitary measures as wellas standards on product quality As post-1990 NSPTAs con-tain deep provisions in areas that only more productivedeveloping economiesmdashsuch as Chinamdashcan readily meet(see Baccini and Urpelainen 2012) SSPTAs became increas-ingly attractive In support of these points our examina-tion of the legalization of nontrade issues in PTAs revealsthat South-South agreements (as compared to North-Southagreements) contain significantly fewer standards related tohuman rights and environmental protection19

In closing we note that our claims center on govern-mentsrsquo political incentives to form SSPTAs Even if a PTAhas a limited or delayed effect on trade outcomes govern-ments may still gain favor from investors (Buumlthe and Milner2008) from powerful vested interests and from the domes-tic public for signing the agreement As Mansfield and Mil-ner maintain ldquoif PTAs are providing a visible reassurancemechanism for domestic publics they may be highly effec-tive even if they have little economic impactrdquo (2012 169)Herein lies the importance of a formal agreement ratherthan unilateral tariff reductions for it sends a clearer po-litical signal and addresses the problem of high policy un-certainty in developing countries (Hollyer and Rosendorff2012)

Case Example Summary and Hypotheses

To understand the interaction between firms andgovernmentsmdashand the primacy of export interestsmdashinnegotiations over SSPTAs consider the case of UgandaIn 2000 Uganda decided to join a Free Trade Area (FTA)agreement initiated by a subset of the Common Market for

19 Data from Lechner (2016) The mean values of the indexes for civil andpolitical rights were 29 in NSPTAs and 16 in SSPTAs for economic and socialrights 33 versus 2 and for environmental protection 35 versus 17

Dow




Eastern and Southern Africa (COMESA) nations20 Ugan-dan manufacturers were initially opposed to membershipin the FTA because they feared foreign goods would ldquofloodthe Ugandan market and thus push them out of businessrdquo(Africa News 2006) However companies like MukwanoGroup a Ugandan conglomerate that produces low-skilledmanufactures argued that joining the FTA would help themcompete with countries like China (East Africa Business Week2012) Ugandan president Yoweri Museveni has politicalincentives to support exporters such as the Mukwano familygiven that the government relies on them for ldquoattractingmore investorsrdquo for providing mass employment and even(allegedly) for campaign contributions21

After a series of discussions with the government theUgandan Manufacturers Association united in support forthe FTA (Africa News 2006 The Independent 2012) Two fac-tors seem to have played the largest role in their change ofposition (1) the chance to improve their competitivenessvis-agrave-vis global competitors such as China and (2) the re-gional FTA provided access to a wider market and cheaperintermediate goods (East African 2001 East African BusinessWeek 2012) They feared that by failing to join the FTAthey would have to ldquoreturn to the era of exporting unpro-cessed raw materials and importing finished productsrdquo (EastAfrican 2001) Similarly in Kenya local manufacturers lob-bied for regional trade agreements on the grounds that itwould give regional businesses the advantage of ldquoeconomiesof scale to take on fast growing economies such as Chinaand India whose foray into the region is a major challengerdquo(Africa News 2008) The Kenyan High Commission endorsedCOMESArsquos simplified ldquocertificate of originrdquo with the specificgoal of promoting ldquosmall scale cross-border tradersrdquo22

In sum this evidence from COMESA supports our claimthat firms and governments seek to use SSPTAs to encour-age manufacturing exports by opening up new marketsdeveloping regional supply chains and providing opportu-nities for more nascent firms to engage in learning by ex-porting In support of the idea that South-South agreementsseek to target manufacturing sectors in particular our ex-amination of all SSPTAs for which data are available revealsthat membersrsquo average tariff levels in low-skilled manufac-turing sectors are substantially lower than Most Favored Na-tion (MFN) levels for those sectors but interestingly tariffsfor agricultural and high-skilled manufacturing sectors donot exhibit such large differentials (Appendix 2)

We sum up our discussion by deriving testable hypothesesAlthough our argument is founded in part on insights aboutfirm-level preferences it yields hypotheses about govern-ment behavior that can be tested using country-level datasimilar to the approach employed by Helpman Melitz andRubinstein (2008)23

H1 Of all the BRICS the growing dominance of China in global ex-port markets is correlated with an increase in SSPTA membershipsmdashand regional SSPTAs in particularmdashamong developing countrieswith high levels of surplus labor

We have outlined two sources of pressure for SSPTAs as aform of compensation for Chinarsquos rise First exporting firmsare most likely to form coalitions demanding that surplus-

20 Nine of twenty COMESA member-states formed this FTA Djibouti EgyptKenya Madagascar Malawi Mauritius Sudan Zambia and Zimbabwe

21 See Uganda (2014) and Lule (2002)22 See Kenya in COMESA (2013)23 They develop a model of trade flows determined by firm-level productivity

tested using a gravity model with country-year as the unit of analysis

labor governments find immediate solutions An observableimplication of this causal mechanism (CM1) is as follows

CM1 Countries with high levels of surplus labor whose export mar-kets have been supplanted by China are joining more SSPTAs

The second source of pressure for SSPTAs may be thelarge number of import-competing firms if they can over-come collective action problems which are higher thanthose for export-oriented firms

CM2 Countries with high levels of surplus labor that have experi-enced a surge in Chinese imports are joining more SSPTAs

Evidence

If our core contention is correct we should observe thatsurplus-labor countries are turning to SSPTAs as China be-comes more globally dominant in export markets To testour hypotheses we construct a comprehensive dataset of135 developing countries from 1979 to 200924 The unit ofanalysis is country-year Data on SSPTA membership comefrom Duumlr Baccini and Elsigrsquos (2014) DESTA dataset whichincludes information on 535 SSPTAs and 150 NSPTAs Weestimate a series of ordinary least squares (OLS) modelspredicting the change in the number of (1) South-Southand (2) regional South-South PTAs in which country i is amember in year t25 We examine change as the dependentvariable in order to test our claim that countries are form-ingjoining new SSPTAs in response to Chinarsquos rise Ourmodel takes the following form

SSPTAit = β0 + β1 ChinaRisetminus1 + β2 Surplus Laboritminus1

+β3 Surplus LaborlowastChinaRiseitminus1

+β Zitminus1 + αi + εit

China rise and surplus labor are our primary variables ofinterest Zi tndash1 is a vector of controls and αi is country fixedeffects

26 All right-hand-side variables are lagged by one yearand all models include a linear time-trend variable27 Robuststandard errors are clustered on country

Our first measure of Chinarsquos rise is systemic the changein its percent share of world exports by year28 Chinarsquos ex-port share increased from 16 percent in 1978 to 98 percentin 2009 with the years of highest growth occurring afterChina joined the World Trade Organization (WTO) Thisvariable allows us to assess whether SSPTA formation is tem-porally correlated with Chinarsquos growing export dominanceThis would occur as governments respond to systemic chal-lenges even if Chinarsquos rise has not (or perhaps not yet) di-rectly affected their own trade patterns In order to assessthe effect of accumulated change in Chinarsquos export shareover time we also run models using the level of Chinarsquos ex-ports in the global economy and results are substantivelythe same (see Appendix 4 Models 1ndash2)

24 Some small countries are excluded due to a lack of economic data used toconstruct the measure of surplus labor Our sample begins in 1979 because this isthe first year in which data on Chinarsquos exports is reported

25 As a robustness check we run negative binomial models with the numberof SSPTA memberships as the dependent variable Results are substantively thesame See Appendix 3

26 Our model does not include a lagged dependent variable and has a rela-tively large T mitigating concerns about the bias induced by including countryfixed effects We nevertheless run the models with random effects and the find-ings are robust

27 This accounts for increases in the number of PTAs over time Here we donot include year fixed effects because the variable for Chinarsquos global export sharedoes not vary within years Our next set of tests feature country fixed effects

28 Data on exports are from the World Bank Development Indicators

Dow




Table 1 BRICSrsquo rise and SSPTA memberships

BRICS Brazil Russia India China South Africa

South-South PTAsChange in world export share (+) ns + ns + nsSurplus labor change in export share + (+) ns ns + nsSurplus labor level of export share ns ns ns (+) + -

Regional South-South PTAsChange in world export share (+) ns ns + + nsSurplus labor change in export share ns ns ns ns + nsSurplus labor level of export share ns ns ns (+) + ns

Notes (1) Dependent variables are the change in SSPTA and regional SSPTA memberships Control variablesare identical to those in Table 2 (2) () denotes significance at p lt 010 + denotes a positive effect significantat p lt 005mdashdenotes a negative effect significant at p lt 005

Our second key variable is the measure of surplus labor(as a percent of the total working age population)29 Wedemeaned it for each year and centered at zero which re-moves the temporal trend in the data and allows us to bettercompare cases across time High levels of surplus labor indi-cates that a high proportion of the population is either un-employed or working in the informal sector in practice itserves as a good proxy for the size of the disadvantaged less-skilled informal sector (Rudra 2002 2008) Note that thisis different from labor abundance as a factor endowmentThough the two are correlated a labor-abundant countrydoes not necessarily exhibit a high level of surplus laborSouth Korea Vietnam and Indonesia are examples In ro-bustness checks we substitute two measures of labor abun-dance for surplus labor the countryrsquos population (logged)and a variable for arable land per capita (the inverse of laborabundance) Results are broadly consistent with our mainfindings (Appendix 5)

Control variables capture additional factors that affectPTA membership level of democracy (measured as the coun-tryrsquos Polity score) GDP per capita and GDP growth Previousstudies have shown that democracy is associated with eco-nomic liberalization and PTA membership (cf Mansfieldet al 2002 Milner and Kubota 2005 Mansfield andMilner 2012) but questions remain as to whether this holdsfor South-South agreements (Baccini 2011) Next we in-clude a variable for the average number of SSPTA membershipsamong developing countries by year (calculated excludingcountry i) This captures diffusion effects as well as ldquosat-urationrdquo effects (ie as global SSPTA memberships growcountries may be less likely to join more) possibly becausethe marginal utility of adding an additional PTA decreasesWe also control for membership in other trade agreementsfirst a dummy for membership in at least one GeneralizedSystem of Preferences (GSP) agreement with a high-incomecountry Such agreements provide LDCs with (nonrecipro-cal) access to rich-country markets and thus may reduce de-mand for South-South PTAs Second a dummy for WTOmembership which we expect is associated with an increasedpropensity for membership in trade agreements in generalFinally because liberalized economies are more likely to joinfree trade agreements we include Sachs et al (1995) indi-cator for openness which measures the extent of trade liber-alization policies adopted by government30

29 Surplus labor is calculated as the working age population (minus studentsenrolled in secondary and postsecondary education) minus active labor-force par-ticipation (Rudra 2008)

30 Values for 1993ndash1999 are from Wacziarg and Welch (2008) values for post-1999 are interpolated (most countries that have liberalized had done so prior to1999)

To assess H1 we first compare results for Chinarsquos rise tothat of the other BRICS Table 1 summarizes the key resultsof a series of models estimating the effect of the BRICSrsquo riseboth as a group and as individuals It indicates where wefound a statistically significant effect for the change in theBRICSrsquo export share and for the interaction term betweenthis variable and surplus labor These results suggest Chinais indeed the only BRICS whose expansion in global exportmarkets is robustly associated with an increase in SSPTAmembership among developing countries The findings forthe other BRICS countries are inconsistent and largely notsignificant

Table 2 lends support to our first hypothesis as Chinarsquosglobal export share grows other developing countries butparticularly surplus-labor countries join more SSPTAs (Models1 and 2)31 Moreover Chinarsquos rise unlike other BRICS isassociated with an increase in regional SSPTA membershipamong surplus-labor countries (Models 3ndash4) Figure 3 dis-plays our key result in substantive terms graphing the ef-fect Chinarsquos export share (from its mean value to mean + 1SD) at different levels of surplus labor The y-axis shows thepredicted change in SSPTA memberships We see that theeffect of Chinarsquos rise on SSPTA membership is significantlydifferent from zero for countries at or above a value of ap-proximately 05 of surplus labor (which is scaled from 0 to1) This is near the median value for the sample In otherwords countries at roughly the top half of surplus labor inany given year exhibit a significant response to Chinarsquos grow-ing global export share

In contrast Chinarsquos export share is not significantly as-sociated with SSPTA membership at low levels of surplus la-bor Thus among developing countries it is those with largepools of underemployed (and presumably less globally com-petitive) labormdashsuch as Uganda Colombia Morocco andSenegalmdashthat are especially likely to join SSPTAs as Chinarsquosprominence in export markets has grown over time Devel-oping nations with smaller pools of surplus labor appear tobe less motivated to take such steps

We also find striking evidence that even as surplus-laborcountries have joined more South-South PTAs they joinfewer North-South PTAs We replicated our analyses usingchange in NSPTAs as the dependent variable (Appendices7 and 8) While Chinarsquos rise (measured as export share) is

31 The number of observations in Table 2 reflects the fact that data are miss-ing for indicators used to create the surplus-labor variable Missingness is drivenby countries experiencing long periods of civil war (eg Afghanistan and Soma-lia) or for which economic data are not reliably reported (eg North Korea andTaiwan) or which gained independence midway through our sample period (egpost-Soviet states) Otherwise we found no systematic patterns of missingness forcertain variables or time periods

Dow




Table 2 Chinarsquos rise and SSPTA memberships

SSPTA Regl SSPTA

1 2 3 4

Chinarsquos world export share 1875 minus4077 1453 minus3418(733) (1975) (650) (1751)

Surplus labor 057 012 021 minus015(027) (030) (024) (027)

Surp labor Chinarsquos export share 11335 9278(3494) (3098)

Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)

GDP growth minus001 minus001 minus001 minus001(000) (000) (000) (000)

Avg SSPTA membership in world minus038 minus040 minus023 minus024(005) (005) (005) (005)

GSP with high-income partner minus007 minus007 minus004 minus004(003) (003) (003) (003)

WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)

Constant minus321 minus355 minus200 minus228(076) (077) (068) (068)

R2 004 005 002 003N 2624 2624 2624 2624

Note Statistical significance levels p lt 01 p lt 005 p lt 001

Figure 3 Marginal effect of change in Chinarsquos export shareon SSPTA memberships

associated with an increase in NSPTA memberships in gen-eral this is not the case for high surplus-labor countrieswhere Chinarsquos rise is actually associated with relatively fewerNorth-South trade agreements32 This lends further supportto our claim that the China export shock is encouragingsurplus-labor countries to join more SSPTAs in part becausethey are relatively less attractive partners for NSPTAs

32 Conversely our results imply that low-surplus labor countries form moreNSPTAs We speculate this may be because low surplus countries are associatedwith higher human capital stronger regulations and private property rights andso are more globally competitive

Country-Specific Measures of China Export and Import Competition

We have shown that surplus-labor countries are respondingto a systemic shock in global export markets Next we con-sider whether particular countries whose exports have beendisplaced by China are responding by forming SSPTAs Webegan by identifying for each developing country i thatcountryrsquos top five trading partners in the baseline year of199033 We then recorded for each year the value of coun-try irsquos exports to these top five trading partners as well asChinarsquos exports to these same five countries Finally we cal-culated the ratio of Chinarsquos exports to country irsquos exports byyear

sum5j = 1 expchina j

sum5j = 1 expi j

Each value j represents country irsquos jth export partner in1990 A ratio greater than 1 indicates that China exportedmore to these ldquotop fiverdquo countries than did country i Weterm this ratio the ldquoexport threatrdquo from China Of interestto us is the percent change in this ratio where a positivechange indicates that Chinarsquos exports are increasing rela-tively more than country irsquos exports to its top five tradingpartners34 Note that this measure of international export

33 We selected 1990 because it provides a snapshot of trade relations beforethe most rapid period of Chinarsquos growth in the mid-1990s

34 We also run models with the level of the export-threat ratio rather thanthe change (Appendix 4) Results differ in that the level interacts negatively withsurplus labor in predicting SSPTA memberships This indicates that countries areresponding to changes (ie growing displacement in their export markets) butthat an accumulated high level of Chinese exports to a countryrsquos top trading part-ners is not associated with growth in SSPTA memberships

Dow




Figure 4 (a) Chinarsquos export threat in high and low surplus-labor countries (b) Chinarsquos import threat in high and lowsurplus-labor countries

market displacement is different from other recent mea-sures of the impact of the China shock on local labor mar-kets (eg David Dorn and Hanson 2013 Autor Dorn andHanson 2016)

In support of our insight that firms in surplus-labor coun-tries are those that have faced the most challenges in main-taining and expanding their exports Figure 4a reveals quitestarkly that the China export threat has been far largermdashandincreasing at a higher ratemdashfor countries above the samplemedian level of surplus labor

To assess whether China is threatening local import-competing firms we create a measure of the change inthe countryrsquos imports from China (as percent of GDP)35

Figure 4b graphs this variable over time for the set ofhigh (above median) and low (below median) surplus-laborcountries In contrast to Chinarsquos export threat there is nosubstantial difference in Chinarsquos import threat between highand low surplus-labor countries This is a first indication thatChinese imports should not be a predictor of the differentialrate of SSPTA membership between high and low surplus-labor countries

Table 3 replicates our previous analyses but featuresthe percent change in the China export threat ratio36

while Table 4 features our measure of the import threat Re-sults reveal indeed that changes in export patterns are themore important correlate of SSPTA membership whereaschanges in imports from China are not significant eitheralone or in interaction with surplus labor This supports theidea that Chinarsquos threat to exporting firms is a driver of SSP-TAs (CM1) more so than the large and diverse group ofimport-competing (less-productive) local firms (CM2) Thesignificant interaction term in Table 3 again indicates thatthe effect of export market displacement on SSPTA mem-bership is more pronounced among surplus-labor coun-tries Figure 5 illustrates this graphing the effect of Chinarsquosexport threat (Table 3 Model 2) on the predicted changein SSPTA memberships with 90 percent confidence inter-vals37 While the effect is not quite significant at the 95 per-cent level of confidence it is very nearly so (at p-values of006 to 008) at the bottom half and top third of surplus la-bor Also of note is that the export-threat ratio exhibits no

35 Trade data is taken from the International Monetary Fundrsquos (IMF) Direc-tion of Trade Statistics (DOTS) (International Monetary Fund 1999)

36 To smooth over year-on-year variation we use the two-year running averageof the percent change in the export-threat ratio The N in Table 3 drops comparedto Table 2 because the export threat is coded only after 1989

37 Figure 5 graphs the effect of a change in the China export threat variablefrom the fiftieth to seventy-fifth percentile

effect on countriesrsquo memberships in NSPTAs (Appendix 6and 7)

In sum using both systemic and country-specific mea-sures of Chinarsquos rise in export markets we find consistentresults this risemdashand resulting displacementmdashis associatedwith more SSPTA memberships but relatively fewer NSPTAmemberships for high surplus-labor countries

Difference-in-Differences

As an additional check we employ a differences-in-differences estimation strategy which allows us to comparethe effect of Chinarsquos rise in a ldquotreatmentrdquo group (countriesthat should be highly affected) and a ldquocontrolrdquo group (coun-tries more insulated from the China shock) Our controlgroup consists of high surplus-labor countries that share aborder with a developed country (Allee and Scalera 2012)38

These statesmdashsuch as Mexico (which borders the UnitedStates)mdashenjoy a cushion against the China export shockbecause their geographic location provides privileged andcheaper access to developed markets If our intuitions arecorrect we should observe a greater increase in SSPTAmembership as China rises among high surplus-labor na-tions that are not contiguous with a developed country (iethe treatment group)

The diff-in-diff setup provides another way for us to dis-cern the effect of Chinarsquos rise on SSPTAs as compared toother global temporal factors (which should affect the con-trol group just as much as the treatment group) Our di-chotomous measure of Chinarsquos rise is Chinarsquos entry into theWTO in 2001 This choice is informed by strong evidencethat Chinarsquos WTO accession had a major impact on its ex-port growth (Ching Hsiao Wan et al 2011)39 Drawingfrom the logic of our theory and earlier results we restrictthe sample to countries located above the sample-mean levelof surplus labor An important assumption of the diff-in-diffmodel is that of parallel trends between treatment and con-trol groups our inspection of the pretreatment data lendssupport to its validity We estimate the following model

SSPTAit = β0 + β1 ChinaWTOit + β2 NoBorderit+β3 ChinalowastNoBorderit + β Controlsitminus1 + εit

38 We code contiguity with the Correlates of War (Sarkees and Wayman 2010)as a land or sea border within 150 miles

39 We also replicate our analysis using 1995mdashwhich corresponds to an earlierincrease in Chinarsquos global export sharemdashas the cutoff point and results hold

Dow




Table 3 China export threat and SSPTA memberships

SSPTA Regl SSPTA

1 2 3 4

China export threat ratio ( change 2 yr avg) minus00003 minus0004 minus00003 minus0004(000) (000) (000) (000)

Surplus labor 003 minus014 minus011 minus026(055) (058) (042) (044)

Surplus labor China export threat ratio 001 001(000) (000)

Polity minus001 minus001 minus001 minus001(001) (001) (001) (001)

GDP per cap (log) minus030 minus033 minus011 minus013(026) (026) (021) (021)


Avg SSPTA membership in world 1338 1332 778 772(325) (321) (256) (252)

WTO 002 003 002 003(015) (015) (013) (013)


Openness minus008 minus009 minus009 minus010(016) (017) (012) (012)

Year fixed effectsConstant minus4355 minus4303 minus2559 minus2513

(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677


Here β1 represents the effect of Chinarsquos rise in the con-trol group and β3 in the treatment group40 Table 5 sum-marizes the key result surplus-labor countries that do notborder a rich country (the treatment group) join moreSouth-South agreements after Chinarsquos WTO entry Cruciallythe difference-in-difference is positive and significant forboth SSPTAs and (especially) regional SSPTAs When wereplicate the same analysis for countries below the meanlevel of surplus labor these results do not hold further in-dicating that the trend toward more SSPTAs is being drivenby countries with high levels of surplus labor

Validity Checks for Surplus Labor

As discussed surplus labor is a symptom of a large unskilledinformal sector behind the border constraints and poorinfrastructure It should therefore also be associated withhigh trade costs which make it difficult to compete withChina for access to developed country markets To checkwhether our measure of surplus labor is indeed capturingthe difficulty of North-South trade we employ the WorldBankrsquos (2018) Trade Cost dataset computing each countryrsquosaverage trade cost with all developed (Organisation for Eco-nomic Co-operation and Development OECD) countries41

As anticipated Figure 6 illustrates stark differences in tradecosts between high- and low-surplus-labor developing coun-

40 Control variables are Polity GDP per capita the global average number ofSSPTA memberships the openness index and a year trend

41 The data provide estimates of bilateral trade costs in agriculture and man-ufactured goods Symmetric bilateral trade costs are computed using the inversegravity framework which estimates trade costs for each country-pair using bilat-eral trade and gross national output When we run our figure and analyses ontrade costs in manufacturing goods only results are unchanged

tries Compared to all other LDCs the BRICSmdashand espe-cially Chinamdashhave lower trade costs Chinarsquos trade costs ac-tually declined between 1995 and 2010

We then investigate whether our variable for surplus la-bor is simply picking up the effects of being a low-incomecountry We first note that although surplus labor is neg-atively correlated with GDP per capita among developingcountries this correlation is not very high (Pearsonrsquos cor-relation coefficient of minus026) Indeed there are importantdifferences between the two indicators high surplus-laborcountries can be higher-income developing economies asis the case with some oil-rich states such as Saudi ArabiaOman and Venezuela When we substitute the surplus-laborvariable with GDP per capita in our models we find thatincome is not a significant predictor of SSPTA member-ships either alone or in interaction with our measures ofChinarsquos rise (Appendix 8) However higher-income devel-oping countries are forming more NSPTAs and this effectincreases with Chinarsquos global export share In sum whileeconomic development is a good predictor of North-Southagreements for South-South agreements it is surplus laborrather than income level that is explaining which countriesjoin as China rises

Conclusion

In an increasingly competitive global economy small de-veloping economies face the acute challenge of appeasingless globally competitive firms and restive populations ofunderemployed workers Their exports to rich-country mar-kets have been steadily declining since the 1990s AlthoughNorth-South PTAs are coveted by many LDC governmentstheir negotiation can be politically fraught on both sides

Dow




Table 4 China import threat and SSPTA memberships

SSPTA Regl SSPTA

Imports from China ( GDP) minus056 minus265 minus005 minus102(083) (338) (078) (318)

Surplus labor 092 089 036 035(045) (045) (042) (042)

Surplus labor imports from China 460 213(720) (678)






WTO 010 010 005 005(008) (008) (008) (008)

Openness 007 007 minus006 minus006(016) (016) (015) (015)

Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719


Figure 5 Effect of China export threat on SSPTA member-ship

Poorer countries are rarely able to secure such agreementscompared to large emerging markets such as China We ar-gue that SSPTAs are a more feasible option that can never-theless bring economic and political benefits and we showthat SSPTAs are a policy tool to which governments are in-creasingly turning in the wake of Chinarsquos rise Althoughwe anticipated that all BRICS might have a similarmdashalbeitsmallermdasheffect it is Chinarsquos growth in global export marketsthat correlates most closely with SSPTA formation More-over we find that those countries whose exports have beenmost displaced by China are those that have joined moreSouth-South agreements Our account therefore underlinesthe role of structural change in the global economy as medi-

ated by domestic political-economic conditions in explain-ing the rapid growth in SSPTAs in recent decades

That surplus-labor countries in particular are turning toSSPTAs further suggests that there is a political logic behindtheir proliferation High levels of surplus labor imply thatmore-productive firms are floundering and policy makershave failed to provide economic opportunities for marginal-ized segments of society Countries with high surplus laborthus face a dual problemmdashnamely the need to expand in-ternational market access and expand employmentmdashwhichgrows more intractable in the face of competition from ris-ing economic powers Politically influential exporting firmsin such countries have incentives to seek South-South tradecooperation as an alternative means to gain access to newmarkets

To be clear our account does not imply that LDCs arelocked in a zero-sum game with China nor that SSPTAs rep-resent an alternative to cooperation with China and otherBRICS As the experience of several African countries illus-trates nations can receive Chinese aid and investment whilestill pursuing membership in a growing number of SSPTAsThe latter may serve as a sort of insurance policy an attemptto cushion the negative effects of increased global competi-tion and to seek out more a more diversified trade portfo-lio as noted by Mansfield and Reinhardt (2003)42

Here we have established the relationship betweenChinarsquos rise and the formation of SSPTAs This is an im-portant finding apart from questions about the agreementsrsquoultimate effectiveness in developing countries PTAs canfunction as a political signal that greater opportunitiesmdashparticularly for the losers of liberalizationmdashare on the

42 This ldquoinsurance policyrdquo might also be beneficial when as Davis Fuchs andJohnson (2014) find developing economies face negative economic repercus-sions while trading with China and political relations go unexpectedly awry

Dow




Table 5 Differences in differences Chinarsquos rise and SSPTA memberships

Pre-China WTO entry Post-China WTO entry Diff-in-diff

Change in SSPTAsDiff (treatedmdashcontrol) minus008 016 024SE (007) (011) (014)

Change in Regional SSPTAsDiff(treatedmdashcontrol) 005 021 016SE (005) (0001) (003)


Figure 6 Trade costs with developed countries over time

horizon (Mansfield and Milner 2012) At the same timeour logic suggests that policy-makers have real incentives toensure that these trade agreements do ldquoworkrdquo toward pro-viding the learning and know-how to eventually access rich-country markets One next step in this research agendathen is to take the question of effectiveness head-on by ex-tending our analysis of tariffs (Appendix 2) as well as ex-amining whether SSPTAs increase trade in less-skilled labor-intensive sectors

In closing our analysis contributes to debates over theconsequences of the BRICSrsquo rise and China in particularMoving beyond a focus on great power politics and declin-ing US hegemony we take seriously the agency of smallerdeveloping countries as they seek new strategies to navigatean economic order that has thus far largely served the eco-nomic interests of the powerful As deputy Prime Ministerof Zimbabwe Arthur Mutambara explains

African countries must not deal with China as individ-uals because we are too small and we will be short-changed We must work on a framework that enablesus to negotiate as SADC [Southern African Develop-ment Community] with 250 million people COMESA(400 million) and Africa with a billion people Wemust understand that national interest lies within theregional interest That is how we survive under global-ization (All Africa 2012)

We conclude that rising powers are changing the shape ofthe world economy but not necessarily toward greater dis-order Rather their emergence is contributing to the turntoward regionalism and greater South-South cooperation(Baccini and Duumlr 2012) Here we have theorized the eco-nomic consequences of these developments for firms laborand governments in LDCs But the political consequences

are noteworthy as well as developing countries stand togain improved representation and clout in global-trade fo-rums (Mansfield and Reinhart 2003) as well as enjoy thepeace dividends of enhanced institutional ties with neigh-bors many of whom have a history of rivalry and conflict

Supplementary Information

Supplementary information is available at the InternationalStudies Quarterly data archive

References

ADEMOLA OYEJIDE T ABIODUN S BANKOLE AND ADEOLU O ADEWUYI 2009ldquoChina-Africa Trade Relations Insights from AERC Scoping StudiesrdquoEuropean Journal of Development Research 21 (4) 485ndash505

AFRICA NEWS 2006 ldquoUganda Manufacturers Bitter With Govt Over Delay inFree Trade Dealrdquo Africa News March 29 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem4JKC-D010-TX2J-N2F6-00000-00ampcontext=1516831

AFRICA NEWS 2008 ldquoKenya Local Traders Set to Benefit from Economiesof Scalerdquo Africa News October 22 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem4TRS-5N30-TX2J-N011-00000-00ampcontext=1516831

ALL AFRICA 2012 ldquoNew Approach to Trade Vitalrdquo All Africa Interview May 26httpssearch-proquest-comproxylibrarygeorgetowneducentraldocview10165819604C6058F884984291PQ1accountid=11091

ALLEE TODD L AND JAMIE E SCALERA 2012 ldquoThe Divergent Effects of Join-ing International Organizations Trade Gains and the Rigors of WTOAccessionrdquo International Organization 66 (2) 243ndash76

AacuteLVAREZ ROBERTO AND SEBASTIAacuteN CLARO 2009 ldquoDavid Versus Goliath The Im-pact of Chinese Competition on Developing Countriesrdquo World Develop-ment 37 (3) 560ndash71

AMSDEN ALICE H 1989 Asiarsquos Next Giant South Korea and Late IndustrializationOxford Oxford University Press

ARMIJO LESLIE E AND CYNTHIA ROBERTS 2014 ldquoThe Emerging Powers andGlobal Governance Why the BRICS Matterrdquo In Handbook of EmergingEconomies 503ndash20 New York Routledge

AUSTRALIA DEPARTMENT OF FOREIGN AFFAIRS AND TRADE 2013 InternationalAid and Transparency Initiative Dataset httpsdfatgovauabout-uscorporatetransparencyPagesiati-dataaspx

AUTOR DAVID H DAVID DORN AND GORDON H HANSON 2016 ldquoThe ChinaShock Learning from Labor-Market Adjustment to Large Changes inTraderdquo Annual Review of Economics 8 205ndash40

ARVIS JEAN FRANCcedilOIS YANN DUVAL BEN SHEPHERD AND CHORTHIP UTOKTHAM2013 ldquoTrade Costs and Development A New Data Setrdquo EconomicPremise 104 World Bank Washington DC httpsopenknowledgeworldbankorghandle1098617051

AW BEE Y MARK J ROBERTS AND DANIEL YI XU 2011 ldquoRampD Investment Ex-porting and Productivity Dynamicsrdquo American Economic Review 101 (4)1312ndash44

AW BEE Y MARK J ROBERTS AND TOR WINSTON 2007 ldquoExport Market Partic-ipation Investments in RampD and Worker Training and the Evolutionof Firm Productivityrdquo World Economy 30 (1) 83ndash104

BACCINI LEONARDO 2011 ldquoDemocratization and Trade Policy An EmpiricalAnalysis of Developing Countriesrdquo European Journal of International Re-lations 18 (3) 455ndash79

Dow




mdashmdashmdashmdash ldquoThe Economics and Politics of Preferential Trade AgreementsrdquoAnnual Review of Political Science (forthcoming)

BACCINI LEONARDO AND ANDREAS DUumlR 2012 ldquoThe New Regionalism and PolicyInterdependencerdquo British Journal of Political Science 42 (1) 57ndash79

BACCINI LEONARDO PABLO M PINTO AND STEPHEN WEYMOUTH 2017 ldquoThe Dis-tributional Consequences of Preferential Trade Liberalization Firm-Level Evidencerdquo International Organization 71 (2) 373ndash95

BACCINI LEONARDO AND JOHANNES URPELAINEN 2012 ldquoStrategic Side PaymentsPreferential Trading Agreements Economic Reform and ForeignAidrdquo Journal of Politics 74 (4) 932ndash49

BACI 2010 International Trade Database at the Product-Level 1994ndash2007Version CEPII Dataset httpwwwcepiifrCEPIIenpublicationswpabstractaspNoDoc=2726

BAIER SCOTT L AND JEFFREY H BERGSTRAND 2007 Do Free Trade AgreementsActually Increase Membersrsquo International Traderdquo Journal of Interna-tional Economics 71 (1) 72ndash95

BANGKOK POST 2006 ldquoRise of China and India Puts Pressure on Thai-landrdquo Bangkok Post August 7 httpssearch-proquest-comproxylibrarygeorgetowneducentraldocview464296697BF570AF05BF44312PQ1accountid=11091

BBC 2005 ldquoKenyan Leader Fears Job Cuts as China India increase textileexports to US EUrdquo Worldwide Monitoring May 26

BEHAR ALBERTO AND LAIA CIRERA-I-CRIVILLEacute 2013 ldquoDoes It Matter Who YouSign With Comparing the Impacts of North-South and South-SouthTrade Agreements on Bilateral Traderdquo Review of International Economics21 (4) 765ndash82

BERNARD ANDREW B J BRADFORD JENSEN STEPHEN J REDDING AND PETER KSCHOTT 2007 ldquoFirms in International Traderdquo Journal of Economic Per-spectives 21 (3) 105ndash30

BLALOCK GARRICK AND PAUL J GERTLER 2004 ldquoLearning from Exporting Re-visited in a Less-Developed Settingrdquo Journal of Development Economics 75(2) 397ndash416

BLAacuteZQUEZ-LIDOY JAVIER JORGE RODRIacuteGUEZ AND JAVIER SANTISO 2006 ldquoiquest Aacutengelo Demonio Los Efectos Del Comercio Chinordquo Revista de la CEPAL90 17ndash43

BOWN CHAD P AND BERNARD M HOEKMAN 2005 ldquoWTO Dispute Settlementand the Missing Developing Country Cases Engaging the Private Sec-torrdquo Journal of International Economic Law 8 (4) 861ndash90

BRANDT LOREN JOHANNES VAN BIESEBROECK AND YIFAN ZHANG 2012 ldquoCreativeAccounting or Creative Destruction Firm-Level Productivity Growthin Chinese Manufacturingrdquo Journal of Development Economics 97 (2)339ndash51

BRAumlUTIGAM DEBORAH 2011 ldquoAid lsquowith Chinese Characteristicsrsquo ChineseForeign Aid and Development Finance Meet the OECD-DAC AidRegimerdquo Journal of International Development 23 (5) 752ndash64

BUSINESS RECORDER 2011 ldquoTextile Clothing Exports Face CompetitiveThreatrdquo Business Recorder May 8 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem53V6-HYR1-JBTF-6325-00000-00ampcontext=1516831

BUTHE TIM AND HELEN V MILNER 2008 ldquoThe Politics of Foreign Direct Invest-ment in Developing Countries Increasing FDI Through InternationalTrade Agreementsrdquo American Journal of Political Science 52 (4) 741ndash62

CHAUFFOUR JEAN P AND DAVID KLEIMANN 2012 ldquoThe Challenge of Implement-ing Preferential Trade Agreements in Developing CountriesndashLessonsfor Rule Designrdquo SSRN Electronic Journal httpsecipeorgwp-contentuploads201412KleimannChauffour_pdf

CHING H STEVE CHENG HSIAO SHUI K WAN AND TONGSAN WANG 2011 ldquoEco-nomic Benefits of Globalization The Impact of Entry to the WTO onChinarsquos Growthrdquo Pacific Economic Review 16 (3) 285ndash301

CHRISTENSEN THOMAS J 2006 ldquoFostering Stability Or Creating a Monster TheRise of China and US Policy Towards East Asiardquo International Security 31(1) 81ndash126

CHRISTOPHER MARTIN HELEN PECK AND DENIS R TOWILL 2006 ldquoA Taxonomyfor Selecting Global Supply Chain Strategiesrdquo International Journal ofLogistics Management 17 (2) 277ndash87

COLLIER PAUL 2000 ldquoRebellion As a Quasi-Criminal Activityrdquo Journal of Con-flict Resolution 44 (6) 168ndash83

COMMON MARKET FOR EASTERN AND SOUTHERN AFRICA (COMESA) 2013 KeyIssues in Regional Integration Volume 2 httpwwwcomesaintattachmentsarticle1223Key20Issues20on20Regional20Intergrationpdf

COMTRADE 2014 UN Comtrade Database Dataset United Nationshttpscomtradeunorg

CRINOgrave ROSARIO AND PAOLO EPIFANI 2012 ldquoProductivity Quality and ExportBehaviourrdquo Economic Journal 122 (565) 1206ndash43

DAVID AUTOR H DAVID DORN AND GORDON H HANSON 2013 ldquoThe ChinaSyndrome Local Labor Market Effects of Import Competition in theUnited Statesrdquo American Economic Review 103 (6) 2121ndash68

DAVIS CHRISTINA L ANDREAS FUCHS AND KRISTINA JOHNSON 2014 ldquoState Con-trol and the Effects of Foreign Relations on Bilateral Traderdquo Universityof Heidelberg Department of Economics Discussion Paper Series (576)

DAWN NEWS 2015 ldquoCrisis Deepens in the Textile Sectorrdquo Accessed24 June 2019 httpswwwdawncomnews1182613crisis-deepens-in-the-textile-sector

DE LOECKER JAN 2013 ldquoDetecting Learning by Exportingrdquo American EconomicJournal Microeconomics 5 (3) 1ndash21

DI GIOVANNI JULIAN ANDREI A LEVCHENKO AND JING ZHANG 2014 ldquoTheGlobal Welfare Impact of China Trade Integration and TechnologicalChangerdquo American Economic Journal Macroeconomics 6 (3) 153ndash83

DJANKOV SIMEON CAROLINE FREUND AND CONG S PHAM 2010 ldquoTrading onTimerdquo Review of Economics and Statistics 92 (1) 166ndash173

DOLLAR DAVID 2005 ldquoGlobalization Inequality and Poverty Since 1980rdquo TheWorld Bank Research Observer 20 (2) 145ndash175

DUumlR ANDREAS LEONARDO BACCINI AND MANFRED ELSIG 2014 ldquoThe Design ofInternational Trade Agreements Introducing a New Datasetrdquo Reviewof International Organizations 9 (3) 353ndash75

EAST AFRICAN 2001 ldquoKenya Mukwano Therersquos Pride in Doingrdquo The EastAfrican December 3 httpsallafricacomstories200112110151html

EAST AFRICAN BUSINESS WEEK 2012 ldquoUganda Traders Nod COMESA FTAEntryrdquo East African Business Week December 5 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem4JKC-D010-TX2J-N2F6-00000-00ampcontext=1516831

ECONOMIST 2015 ldquoOne Among Manyrdquo Economist January 15 httpwwweconomistcomnewsmiddle-east-and-africa21639554-china-has-become-big-africa-now-backlash-one-among-many

ERNST DIETER AND LINSU KIM 2002 ldquoGlobal Production Networks Knowl-edge Diffusion and Local Capability Formationrdquo Research Policy 31 (8)1417ndash29

FINANCIAL TIMES 2008 ldquoConsumer goods Pros and Cons of Import-ingrdquo Financial Times January 23 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem4RND-3P80-TXDK-S0JV-00000-00ampcontext=1516831

FREUND CAROLINE AND CAGLAR OZDEN 2006 The Effect of Chinarsquos Exports onLatin American Trade with the World Photocopy Washington DC WorldBank

GEREFFI GARY AND TIMOTHY STURGEON 2013 ldquoGlobal Value chain-orientedIndustrial Policy The Role of Emerging Economiesrdquo In Global ValueChains in a Changing World ed Deborah K Elms and Patrick Low 329ndash260 Geneva World Trade Organization

GOLUB STEPHEN S JANET CEGLOWSKI AHMADOU A MBAYE AND VARUN PRASAD2017 ldquoCan Africa Compete with China in Manufacturing The Role ofRelative Unit Labour Costsrdquo The World Economy 41 (6) 1508ndash28

GRAY JULIA 2014 ldquoDomestic Capacity and the Implementation Gap in Re-gional Trade Agreementsrdquo Comparative Political Studies 47 (1) 55ndash84

GREENAWAY DAVID AND CHRIS MILNER 1990 ldquoSouth-South Trade Theory Evi-dence and Policyrdquo World Bank Research Observer 5 (1) 47ndash68

GU JING JOHN HUMPHREY AND DIRK MESSNER 2008 ldquoGlobal Governance andDeveloping Countries The Implication of the Rise of Chinardquo WorldDevelopment 36 (2) 274ndash92

HALLAK JUAN C 2006 ldquoProduct Quality and the Direction of Traderdquo Journalof International Economics 68 (1) 238ndash65

HANSON GORDON H AND RAYMOND ROBERTSON 2010 ldquoChina and the Manu-facturing Exports of Other Developing Countriesrdquo In Chinarsquos GrowingRole in World Trade ed Robert C Feenstra and Shang-Jin Wei 137ndash66Chicago University of Chicago Press

HELLMAN JOEL S GERAINT JONES DANIEL KAUFMANN AND MARK SCHANKER-MAN 2000 ldquoMeasuring Governance Corruption and State Capture- how Firms and Bureaucrats Shape the Business Environmentin Transition Economiesrdquo Policy Research Working Paper WPS2312 Washington DC World Bank httpdocumentsworldbankorgcurateden241911468765617541Measuring-governance-corruption-and-State-capture-how-firms-and-bureaucrats-shape-the-business-environment-in-transition-economies

Dow




HELPMAN ELHANAN MARC MELITZ AND YONA RUBINSTEIN 2008 ldquoEstimatingTrade Flows Trading Partners and Trading Volumesrdquo Quarterly Journalof Economics 123 (2) 441ndash87

HERMAN FANIE 2011 ldquoTextile Disputes and Two-Level Games The Case ofChina and South Africardquo Asian Politics amp Policy 3 (1) 115ndash30

HOLLYER JAMES R AND B PETER ROSENDORFF 2012 ldquoLeadership SurvivalRegime Type Policy Uncertainty and PTA Accessionrdquo InternationalStudies Quarterly 56 (4) 748ndash64

HOTTMAN COLIN STEPHEN J REDDING AND DAVID E WEINSTEIN 2014 ldquoWhatIs lsquoFirm Heterogeneityrsquo in Trade Models The Role of Quality ScopeMarkups and Costrdquo CEP Discussion Papers dp1294 Centre for EconomicPerformance LSE

HU ALBERT G Z GARY H JEFFERSON AND QIAN JINCHANG 2005 ldquoRampD andTechnology Transfer Firm-Level Evidence from Chinese IndustryrdquoReview of Economics and Statistics 87 (4) 780ndash86

HUDSON JOHN AND PHILIP JONES 2003 ldquoInternational Trade in lsquoQualityGoodsrsquo Signaling Problems for Developing Countriesrdquo Journal of In-ternational Development 15 (8) 999ndash1013

HURRELL ANDREW 2006 ldquoHegemony Liberalism and Global Order WhatSpace for Would-Be Great Powersrdquo International Affairs 82 (1)1ndash19

IKENBERRY G JOHN 2008 ldquoThe Rise of China and the Future ofthe West Can the Liberal System Surviverdquo Foreign Affairs Jan-uary 1 httpswwwforeignaffairscomarticlesasia2008-01-01rise-china-and-future-west

INDO-ASIAN NEWS SERVICE 2006 ldquoAsia Faces Up to the Challenge from Indiaand Chinardquo Indo-Asian News Service May 6 httpsadvance-lexis-comproxylibrarygeorgetowneduapidocumentcollection=newsampid=urncontentItem4JWT-3SW0-TWTC-833T-00000-00ampcontext=1516831

INDUSTRY CANADA 2014 Canadian Importers Database Dataset httpswwwicgccaeicsitecid-dicnsfenghome

INTERNATIONAL LABOUR ORGANIZATION 2016 World Employment and So-cial Outlook 2016 Trends for Youth Geneva International LabourOffice httpwwwiloorgwcmsp5groupspublicmdashdgreportsmdashdcommmdashpubldocumentspublicationwcms_513739pdf

INTERNATIONAL MONETARY FUND GENERAL STATISTICS DIVISION 1999 Direction ofTrade Statistics (No 1ndash4) International Monetary Fund WashingtonDC httpdataimforgDOT

JENKINS RHYS AND CHRIS EDWARDS 2006 ldquoThe Economic Impacts of China andIndia on sub-Saharan Africa Trends and Prospectsrdquo Journal of AsianEconomics 17 (2) 207ndash25

JENKINS RHYS ENRIQUE PETERS AND MAURICIO M MOREIRA 2008 ldquoThe Impactof China on Latin America and the Caribbeanrdquo World Development 36(2) 235ndash53

JHANGIANI AMITA AND CARL STOCKING 2006 ldquoChina Top Global Sup-ply Chain Partnerrdquo Trade Finance SeptemberndashOctober httpwwwtransactionservicescitibankcomtransactionserviceshomeabout_usarticlesarchive2006docs200610cbrpdf

JOHANSSON JOHNY 1997 Global Marketing Foreign Entry Local Marketing andGlobal Management ChicagoMcGraw-Hill

KAHLER MILES 2013 ldquoRising Powers and Global Governance NegotiatingChange in a Resilient Status Quordquo International Affairs 89 (3) 711ndash29

KANG KICHUN 2013 ldquoThe Choice of Export Destinations and Its Determi-nants Evidence from Korean Exportsrdquo Korean Economic Review 29 (1)139ndash60

KAPLINSKY RAPHAEL 2006 ldquoRevisiting the Revisited Terms of TradeWill China Make a Differencerdquo World Development 34 (6)981ndash95

KAPLINSKY RAPHAEL DOROTHY MCCORMICK AND MIKE MORRIS 2010 ldquoImpactsand Challenges of a Growing Relationship Between China and Sub-Saharan Africardquo In The Political Economy of Africa 389ndash409 ed VishnuPadayachee New York NY Routledge

KIM MOONHAWK 2008 ldquoCostly Procedures Divergent Effects of Legalizationin the GATTWTO Dispute Settlement Proceduresrdquo International Stud-ies Quarterly 52 (3) 657ndash86

KOCH ADAM J 2001 ldquoSelecting Overseas Markets and Entry Modes Two De-cision Processes Or Onerdquo Marketing Intelligence amp Planning 19 (1) 65ndash75

LANGHAMMER ROLF J 1992 ldquoThe Developing Countries and RegionalismrdquoJCMS Journal of Common Market Studies 30 (2) 211ndash32

LAYNE CHRISTOPHER 2012 ldquoThis Time Itrsquos Real The End of Unipolar-ity and the Pax Americanardquo International Studies Quarterly 56 (1)203ndash13

LECHNER LISA 2016 ldquoThe Domestic Battle Over Non-Trade Issues in Pref-erential Trade Agreementsrdquo Review of International Political Economy 23(5) 840ndash71

LIEBER ROBERT J 2014 ldquoThe Rise of the BRICS and American PrimacyrdquoInternational Politics 51 137ndash54

LILEEVA ALLA AND DANIEL TREFLER 2010 ldquoImproved Access to Foreign Mar-kets Raises plant-level Productivityhellip for Some Plantsrdquo The QuarterlyJournal of Economics 125 (3) 1051ndash99

LIM LINDA Y 1983 ldquoSingaporersquos Success The Myth of the Free Market Econ-omyrdquo Asian Survey 23 (6) 752ndash64

LINDERT PETER H AND JEFFREY G WILLIAMSON 2003 ldquoDoes Globalization Makethe World More Unequalrdquo In Globalization in historical perspective 227ndash76 ed Michael D Bordo Alan M Taylor and Jeffrey G WilliamsonChicago Ill University of Chicago Press

LULE KENNEDY 2002 ldquoUganda Mukwano Denies Funding Musevenirdquo Al-lAfrica December 12 allafricacomstories200212140085html

MANGER MARK S 2009 Investing in Protection The Politics of Preferential TradeAgreements between North and South Cambridge UK New York NY Cam-bridge University Press

MANSFIELD EDWARD D AND ERIC REINHARDT 2003 ldquoMultilateral Determinantsof Regionalism The Effects of GATTWTO on the Formation of Pref-erential Trading Arrangementsrdquo International Organization 57 (4) 829ndash62

MANSFIELD EDWARD D AND HELEN V MILNER 2012 Votes Vetoes and the PoliticalEconomy of International Trade Agreements Princeton Princeton Univer-sity Press

MANSFIELD EDWARD D AND JOHN C PEVEHOUSE 2013 ldquoThe Expansion of Pref-erential Trading Arrangementsrdquo International Studies Quarterly 57 (3)592ndash604

MANSFIELD EDWARD D HELEN V MILNER AND JOHN C PEVEHOUSE 2007ldquoVetoing Cooperation The Impact of Veto Players on PreferentialTrade Agreementsrdquo British Journal of Political Science 37 (3) 403ndash432

MANSFIELD EDWARD D HELEN V MILNER AND PETER B ROSENDORFF 2002 ldquoWhyDemocracies Cooperate More Electoral Control and InternationalTrade Agreementsrdquo International Organization 56 (3) 477ndash513

MASKUS KEITH E TSUNEHIRO OTSUKI AND JOHN S WILSON 2005 The Cost ofCompliance with Product Standards for Firms in Developing Coun-tries an Econometric Study Policy Research Working Paper Seriesno WPS 3590 Washington DC World Bank

MAYDA ANNA M AND CHAD STEINBERG 2009 ldquoDo South-South Trade Agree-ments Increase Trade Commodity-Level Evidence from COMESArdquoCanadian Journal of Economics 42 (4) 1361ndash89

MELITZ MARC J 2003 ldquoThe Impact of Trade on Intra-Industry Reallo-cations and Aggregate Industry Productivityrdquo Econometrica 71 (6)1695ndash725

MILNER CHRIS AND DANNY MCGOWAN 2013 ldquoTrade Costs and Trade Composi-tionrdquo Economic Inquiry 51 (3) 1886ndash902

MILNER HELEN V AND KEIKO KUBOTA 2005 ldquoWhy the Move to Free TradeDemocracy and Trade Policy in the Developing Countriesrdquo Interna-tional Organization 59 (1) 107ndash43

MOHAN SUSHIL SANGEETA KHORANA AND HOMAGNI CHOUDHURY 2012 ldquoBar-riers to Prosperity ndash Developing Countries and the Need for TradeLiberalizationrdquo IEA Discussion Paper No 44 httpsieaorgukwp-contentuploads201607Barrier20to20Trade20-20Mohanpdf

MONTALBANO PIERLUIGI AND SILVIA NENCI 2012 ldquoThe Trade Specialization ofChina India Brazil and South Africa A Threat to Whomrdquo Interna-tional Trade Journal 26 (5) 363ndash84

NEW ZEALAND HERALD 2005 ldquoChina Rapidly Turning into the Factory ofthe Worldrdquo New Zealand Herald July 29 httpswwwnzheraldconzbusinessnewsarticlecfmc_id=3ampobjectid=10337758

NHLABATSI ROSENA 2014 ldquoCheap Chinese Imports in Africa Implications andRemediesrdquo Consultancy Africa Intelligence March 6 httpswwwafdborgennews-and-eventsnew-afdb-study-takes-in-depth-look-at-china-africa-partnership-8377

NORDAS HILDEGUNN K ENRICO PINALI AND MASSIMO G GROSSO 2006 ldquoLogisticsand Time As a Trade Barrierrdquo OECD Trade Policy Working Paper No 35

OECDWTO 2015 Aid for Trade at a Glance 2015 Reducing Trade Costs forInclusive Sustainable Growth Paris OECD Publishing

OLSON RICHARD S 1979 ldquoEconomic Coercion in World Politics With a Focuson North-South Relationsrdquo World Politics 31 (4) 471ndash94

OSNAGO ALBERTO NADIA ROCHA AND MICHELE RUTA 2015 ldquoDeep Trade Agree-ments and Vertical FDI the Devil is in the Detailsrdquo Policy Research work-ing paper no WPS 7464 Washington DC World Bank Group

Dow




PETERS ENRIQUE D 2015 ldquoChinarsquos Evolving Role in Latin Amer-ica Can It Be a Win-Winrdquo Atlantic Council httppublicationsatlanticcouncilorgchinalatam

PORTER MICHAEL E 1986 Competition in Global Industries Boston HarvardBusiness School Press

RODRIK DANI 2006 ldquoWhatrsquos So Special about Chinarsquos Exportsrdquo China ampWorld Economy 14 (5) 1ndash19

RUDRA NITA 2002 ldquoGlobalization and the Decline of the Welfare State inLess Developed Countriesrdquo International Organization 56 (2) 411ndash45

mdashmdashmdash 2008 Globalization and the Race to the Bottom in Developing CountriesWho Really Gets Hurt Cambridge Cambridge University Press

SACHS JEFFREY D ANDREW WARNER ANDERS ASLUND AND STANLEY FISCHER 1995ldquoEconomic Reform and the Process of Global Integrationrdquo BrookingsPapers on Economic Activity (1) 1ndash118

SALOMON ROBERT M AND J MYLES SHAVER 2005 ldquoLearning By ExportingNew Insights from Examining Firm Innovationrdquo Journal of Economicsand Management Strategy 14 (2) 431ndash60

SARKEES MEREDITH R AND FRANK WAYMAN 2010 Resort to War 1816ndash2017Washington DC CQ Press

SCHELLEKENS PHILIP 2013 ldquoA Changing China Implications for DevelopingCountriesrdquo Economic Premise 114 Washington DC World Bank

SHEPHERD BEN 2016 ldquoTrade Costs and South-South Trade AgreementsBuilding Blocks Or Stumbling Blocksrdquo Developing Trade Consul-tants Working Paper DTC-2016-2 httpdeveloping-tradecompublicationstrade-costs-and-south-south-trade-agreements-building-blocks-or-stumbling-blocks

STEIN PEER OYA P ARDIC AND MARTIN HOMMES 2013 Closing the Credit Gap forFormal and Informal Micro Small and Medium Enterprises WashingtonDC International Finance Corporation

STEPHEN MATTHEW D 2014 ldquoRising Powers Global Capitalism and LiberalGlobal Governance A Historical Materialist Account of the BRICSChallengerdquo European Journal of International Relations 20 (4) 912ndash38

THE INDEPENDENT 2012 ldquoUganda Weighing COMESA Fortunesrdquo December1 httpsallafricacomstories201212030144html

TULL DENIS M 2006 ldquoChinarsquos Engagement in Africa Scope Significanceand Consequencesrdquo Journal of Modern African Studies 44 (3) 459ndash79

UGANDA 2014 ldquo President Museveni Commissions Acacia Mall Calls onUgandans to Emulate Mukwano Familyrdquo Accessed 1 Jan 2018 httpstatehousegougmedianews20140620president-museveni-commissions-acacia-mall-calls-ugandans-emulate-mukwano-fami

UNITED STATES CENSUS BUREAU 2014 International Trade DatandashSITC Datasethttpswwwcensusgovforeign-tradedataindexhtml

URDAL HENRIK 2006 ldquoA Clash of Generations Youth Bulges and PoliticalViolencerdquo International Studies Quarterly 50 (3) 607ndash30

VERHOOGEN ERIC A 2008 ldquoTrade Quality Upgrading and Wage Inequality inthe Mexican Manufacturing Sectorrdquo Quarterly Journal of Economics 123(2) 489ndash530

WACZIARG ROMAIN AND KAREN H WELCH 2008 ldquoTrade Liberalization andGrowth New Evidencerdquo World Bank Economic Review 22 (2) 187ndash231

WOOD ADRIAN 1997 ldquoOpenness and Wage Inequality in Developing Coun-tries The Latin American Challenge to East Asian ConventionalWisdomrdquo World Bank Economic Review 11 (1) 33ndash57

WOOD ADRIAN AND JOumlRG MAYER 2011 ldquoHas China Deindustrialised Other De-veloping Countriesrdquo Review of World Economics 147 (2) 325ndash50

WORLD BANK 2002 Globalization Growth and Poverty Facts Fears andan Agenda for Action Washington DC World Bank Publishinghttpdocumentsworldbankorgcurateden954071468778196576pdfmulti0pagepdf

WORLD BANK 2018 ESCAP-World Bank Trade Cost Database Datasethttpswwwunescaporgresourcesescap-world-bank-trade-cost-database

WORLD TRADE ORGANIZATION 2013a International Trade Statistics WorldTrade Organization httpswwwwtoorgenglishres_estatis_eits2013_eits2013_epdf

mdashmdashmdash 2013b World Trade Report 2013 Factors Shaping the Future of WorldTrade Geneva WTO Publishing httpswwwwtoorgenglishres_ebooksp_eworld_trade_report13_epdf

mdashmdashmdash 2016 World Trade Report 2016 Levelling the Trading Field forSMEs Geneva WTO Publishing httpswwwwtoorgenglishres_ebooksp_eworld_trade_report16_epdf

YUSUF SHAHID AND KAORU NABESHIMA 2009 Tiger Economies Under Threat AComparative Analysis of Malaysiarsquos Industrial Prospects and Policy OptionsWashington DC World Bank

ZAFAR ALI 2007 ldquoThe Growing Relationship Between China and Sub-Saharan Africa Macroeconomic Trade Investment and Aid LinksrdquoWorld Bank Research Observer 22 (1) 103ndash30

ZENG JINGHAN YUEFAN XIAO AND SHAUN BRESLIN 2015 ldquoSecuring Chinarsquos CoreInterests the State of the Debate in Chinardquo International Affairs 91 (2)245ndash66

ZYLSTRA KIRK D 2012 Lean Distribution Applying Lean Manufacturing to Distri-bution Logistics and Supply Chain Hoboken NJ John Wiley amp Sons

Dow




Figure 1 Surplus-labor LDC exports to developed coun-tries (top) and NSPTA memberships over time amongsurplus-labor and non-surplus labor LDCs (bottom) NotesTop panel Included countries are those with levels of sur-plus labor above the mean for a majority of years in the sam-ple See the ldquoEvidencerdquo section for details on the coding ofsurplus labor Bottom panel NSPTA average excludes highsurplus-labor nations

from the displacement created by large emerging markets(see Figure 1) Although SSPTAs do not necessarily bringthe large-scale increases in trade associated with North-South agreements for firms in surplus-labor nations theycan nevertheless open up new markets in countries withsimilar consumer preferences and quality expectationsand they provide opportunities for learning-by-exporting asthey strive to join multinational supply chains (Blalock andGertler 2004 De Loecker 2013) For governments thenSSPTAs can bring political benefits by signaling that leadersare taking action to address economic conditions If ourintuitions are correct smaller LDCs are not withdrawingfrom the liberal international order rather in response tothe loss of rich-country markets to rising powers they arepursuing multilateralism of a different form

We conduct a range of empirical tests to assess this in-tuition Though built on firm-level foundations our the-ory generates country-level hypotheses about the conditionsunder which governments form more trade agreementsParticularly among surplus-labor countries we find Chinarsquosshare of world export markets is significantly associated withSSPTA formation We employ two measures of Chinarsquos ex-port ldquoshockrdquo to developing economies one systemic andone country-specific First using a systemic measure of eachBRICS countryrsquos share in global export markets our find-ings reveal that Chinarsquos rise is closely associated with therapid proliferation of SSPTAs and we confirm that this ef-fect surpasses that of any other BRICS state Employing thecountry-specific measure of the China shock we find thatsurplus-labor countries whose export markets have been mostdisplaced by China are those joining the most SSPTAs Anincrease in Chinese imports is not a significant predictor ofSSPTAs suggesting that pressure from local firms threat-ened by the flood of Chinese imports has less impact ontrade policy than the relatively smaller number of exportersdisplaced by China

A difference-in-difference analysis provides further evi-dence in support of our theory Chinarsquos membership inthe World Trade Organization (WTO)mdasha critical juncture

Figure 2 South-South (SS) PTA memberships over timeNote Figure depicts the average number of SSPTA member-ships per country by year The sample includes developingcountries only

in its global economic integrationmdashis associated with anincrease in SSPTA membership particularly among surplus-labor countries that do not border a developed coun-try (and thus lack a geographic advantage for access torich-country markets) In short using various measures weconsistently find that surplus-labor countries most adverselyaffected by Chinarsquos rise are those that are forming SSPTAsin the largest numbers

Our account helps explain the proliferation of SSPTAs inrecent decades Per country the average number of SSPTAmemberships has increased sharply since 1990 While 136SSPTAs were formed in the forty-year period between 1948and 1989 398 were formed in the two decades that followed(1990ndash2009) Figure 2 shows that most of these agreementshave been negotiated by non-BRIC LDCs with each otherNorth-South PTAs have also proliferated since 1990 (as seenin Figure 1) but importantly the bulk of these agreementsare bypassing high surplus-labor LDCs3 Most research fo-cusing on domestic political factors as the drivers of PTAformation fails to consider these distinctions between North-South and South-South agreements (Mansfield Milner andRosendorff 2002 Mansfield Milner and Pevehouse 2007Mansfield and Pevehouse 2013)4 Here we bring renewedattention to systemic economic forces that in combina-tion with domestic lsquopolitical economyrsquo conditions providea novel explanation of which countries are forming whichtypes of PTAs and when

Our analysis also sheds light on the consequences of ris-ing powers for global governance (eg Ikenberry 2008Kahler 2013 Armijo and Roberts 2014) We reject the im-plicit assumption that small nations are the pawns of powerpolitics (Hurrell 2006 Kahler 2013 Stephen 2014 ZengXiao and Breslin 2015) and theorize these countries asagents with a distinct role in shaping the evolution of in-ternational market expansion as well as the global institu-tional landscape Our findings in short suggest that therise of large emerging market economies is changing theliberal economic order but not necessarily in the direction

3 Thirty-three NSPTA agreements were successfully negotiated between 1948and 1989 and 116 from 1990 to 2009

4 But see Baccini (2011) on democratization and South-South agreementsMansfield and Milner (2012) identify how democracy and veto players shape in-centives for PTA formation but they stop short of exploring whether there maybe unique incentives that govern PTAs among developing nations

Dow

























Dow



















Export Firms





Dow







Governments









Dow

















Evidence













Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow

























Dow



















Export Firms





Dow







Governments









Dow

















Evidence













Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow



















Export Firms





Dow







Governments









Dow

















Evidence













Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow







Governments









Dow

















Evidence













Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow

















Evidence













Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow


















Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow





SSPTA Regl SSPTA

1 2 3 4




Polity minus00003 00001 0002 0003(0004) (0004) (0004) (0004)

GDP per cap (log) 002 010 002 008(009) (009) (008) (008)




WTO 016 018 007 009(006) (006) (006) (006)

Openness 010 010 002 003(006) (006) (005) (005)

Year 010 010 006 006(001) (001) (001) (001)


R2 004 005 002 003N 2624 2624 2624 2624








sum5j = 1 expi j




Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow























Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow





SSPTA Regl SSPTA

1 2 3 4








WTO 002 003 002 003(015) (015) (013) (013)




(1131) (1112) (914) (896)R2 010 010 005 006N 1677 1677 1677 1677










Conclusion


Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow





SSPTA Regl SSPTA


Surplus labor 092 089 036 035(045) (045) (042) (042)







WTO 010 010 005 005(008) (008) (008) (008)


Year 009 009 009 009(004) (004) (004) (004)


R2 004 004 002 002N 1719 1719 1719 1719









Dow

















References















Dow



















































Dow





















































Dow


































Dow

















References















Dow



















































Dow





















































Dow


































Dow



















































Dow





















































Dow


































Dow





















































Dow


































Dow


































Dow



david and goliath? small developing countries, large emerging … · 2021. 4. 16. · 576 small...

Documents