Official PDF , 6 pages - World Bank Documents & Reports [PDF]

Public Disclosure Authorized Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized

86232 MARCH 2014 • Number 137

Global Value Chains in the Current Trade Slowdown Michael J. Ferrantino and Daria Taglioni

Real growth in global trade has decelerated significantly since its sharp recovery in 2010. Year-on-year growth in global real trade1 decelerated from 13.3 percent at the end of the first quarter of 2010, to 9.9, 3.1, and 0.5 percent at the end of the first quarters of 2011, 2012, and 2013, respectively, while picking back up to 3.9 percent in the year leading up to the fourth quarter of 2013.2 This aggregate deceleration in global trade includes absolute declines in real trade for many product categories and regions. In the wake of the Great Trade Collapse of 2008–9, understanding of the behavior of trade in slowdowns has improved. Among the many explanations offered for the Great Trade Collapse, including explanations related to uncertainty, trade financing, and new protectionist measures by governments, there has been a significant focus on whether the emergence of global value chains (GVCs) in international trade, and their behavior, are a contributing factor in trade slowdowns.3 Relationship between GVCs and Trade Decelerations GVCs involve trade in goods that have multiple production stages that take place in many different countries (that is, “production fragmentation” or “slicing up the value chain”) and in which multiple imports and exports of intermediate goods are necessary to produce a final good, which may also be exported. Since the emergence of the North American GVC in automobiles in the 1960s and the East Asian electronics GVC in the 1970s, the role of GVCs in international trade has become more important and has attracted increasing attention. There are several potential reasons why GVC trade may behave differently. First, there is a crude statistical argument for a relationship between GVC trade and slowdowns, arising from the fact that there are more trade flows in GVC trade than in nonGVC trade. If an exported good is produced entirely within the exporting country, and there is a reduction in demand, then one trade flow disappears. But if the exported good is produced with an imported input, there are two trade flows (the import of the input and the export of the final good), and

thus the same drop in demand for the final good causes two trade flows to disappear. This simple argument turns out to be fallacious (O’Rourke 2009): since there is twice as much trade in a GVC world, then a 1 percent drop in demand leads to a 1 percent drop in both intermediates and final trade, which is not unusual. However, global trade, in fact, consists of both GVC trade, say, complex goods such as automobiles, whose parts and components are produced in different countries and then assembled in one location, and non-GVC trade, such as exports of bananas that require no further processing. If there is a drop in global demand, and demand for complex goods such as automobiles drops more rapidly than demand for simple goods like bananas, then a compositional effect of decelerating demand for goods produced in GVCs (for example, complex consumer goods and capital goods) could cause trade to slow down more rapidly than it would otherwise. Since the world has gradually transitioned over the last several decades to a world in which GVC trade has become more important, there appears to have been a long-run increase in the elasticity of global trade with respect to global income (Escaith, Linden-

1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise

berg, and Miroudot 2010). This increase in elasticity reversed itself in the late 2000s. Whether this reversal is due to cyclical factors associated with the recent financial crisis, or to the possibility that the structural transformation associated with the increasing geographical fragmentation of production is now nearly finished, is not yet known. In any case, the fact that demand for the final good is ultimately spread out over demand for intermediates in many countries means that the transmission of demand shocks during a slowdown in any one region may be more rapid and greater in a world where GVCs are important. In addition, different parts of a GVC may behave differently in a downturn. In particular, reductions in final demand may lead to even larger reductions in intermediate demand, so that trade in intermediates declines more rapidly than trade in final goods, the so-called bullwhip effect (Altomonte et al. 2012; Zavacka 2012). When demand for final goods slows, exporters of such goods can continue for a while on inventory rather than ordering new intermediate goods for production, so intermediate goods trade declines more rapidly than final goods trade. On top of this, the effect of uncertainty is different in the upstream parts of the supply chain, especially for firms that are more peripheral in terms of decision making. Having less information about the drop in demand for final goods, the suppliers of inputs may start avoiding risk, cutting production and trade in intermediate goods even more rapidly than if they had the same information as final-goods producers. Figure 1. Trade in GVCs Has Decelerated More Rapidly Than Total Trade 40

Within these patterns, there is substantial regional and sectoral variation. While the above generalizations are useful, not every GVC produces a good whose final demand is unusually sensitive to declining income, and not every GVC exhibits bullwhip behavior at all times. For example, GVCs in more sophisticated value chains, especially those associated with large multinational lead firms making extensive use of high-tech supply chain management tools, may be more adept at managing fluctuations in demand. Three Representative GVCs: Apparel/ Footwear, Electronics, and Motor Vehicles and Related Parts A focus on trade in three important product groups characterized by GVC-type trade—apparel and footwear, electronics, and motor vehicles and parts—provides a simple picture of the role of GVCs in the current trade slowdown. The goods making up these GVCs, defined in Sturgeon and Memedovic (2011), can be divided into intermediate goods and final goods using BECs (Broad Economic Categories). Together, these goods make up about 20 percent of global trade. However, the GVCs for these goods include more trade than is included in the definitions. A significant amount of primary trade—in metals, chemicals, fossil fuels, electric power, and natural fibers produced in agriculture—is involved in providing inputs for the goods in this analysis. Moreover, GVCs exist for other complex goods, such as capital equipment, precision instruments, and chemical-based consumer goods such as cosmetics and personal care products. It is possible that the three GVCs analyzed here have had behavior similar to other GVCs, but the extent is uncertain. Overall GVC Trade in Current Slowdown

20

10

08 20 Q1 08 Q 20 2 08 Q 20 3 08 Q 20 4 09 20 Q1 09 Q 20 2 09 20 Q3 09 20 Q4 10 Q 20 1 10 20 Q2 10 Q 20 3 10 20 Q4 11 20 Q1 11 Q 20 2 11 20 Q3 11 Q 20 4 12 20 Q1 12 Q 20 2 12 20 Q3 12 20 Q4 13 Q1

0 20

year-on-year real growth growth for period ending at date shown (%)

30

-10

-20

All goods, average GVC, all

-30

-40

Figure 1 shows that in the current slowdown, trade in the three GVCs combined has declined more rapidly than global trade.4 In the Great Trade Collapse, the deceleration was also more rapid. However, during 2008–9, GVC trade collapsed more rapidly than global trade, and then boomed back. The current slowdown in GVC trade, relative to total trade, is more subtle. Arguably, GVC and non-GVC trade have grown at about the same rate since late 2011. However, in four of the last five rolling 12-month windows calculated on quarterly data, GVC trade has grown less rapidly than total trade. In the most

Source: Datastream, International Trade Commission (ITC), World Bank.

2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise

Figure 2. Trade in Motor Vehicles and Parts Has Led Decline in Current and Previous Slowdowns

30 20 10

08 20 Q1 08 Q 20 2 08 20 Q3 08 Q 20 4 09 20 Q1 09 20 Q2 09 20 Q3 09 20 Q4 10 20 Q1 10 20 Q2 10 20 Q3 10 20 Q4 11 20 Q1 11 20 Q2 11 20 Q3 11 Q 20 4 12 20 Q1 12 Q 20 2 12 20 Q3 12 20 Q4 13 Q1

0 20

year-on-year real growth growth for period ending at date shown (%)

40

-10 -20 -30

apparel/footwear electronics autos and parts

-40 -50 Source: ITC, World Bank.

Figure 3. The Europe-Centered GVC Contracted First, Then NAFTA and East Asia 60

40

EU27+ as importer East Asia as importer NAFTA as importer

30

ROW as importer

20 10

08 20 Q1 08 20 Q2 08 20 Q3 08 20 Q4 09 20 Q1 09 20 Q2 09 20 Q3 09 20 Q4 10 20 Q1 10 20 Q2 10 20 Q3 10 20 Q4 11 20 Q1 11 20 Q2 11 20 Q3 11 20 Q4 12 20 Q1 12 20 Q2 12 20 Q3 12 20 Q4 13 Q1

0 -10

20

year-on-year real growth growth for period ending at date shown (%)

50

-20 -30 -40

Sectoral patterns At the sectoral level, a decline in demand for motor vehicles and parts is leading the current slowdown, as it did during 2008–9. The bust-and-boom behavior of trade in motor vehicles and parts has been associated with the domestic stimulus programs in this sector pursued by many countries in the aftermath of the 2008–9 recession. In turn, the phasing out of these programs may be contributing to the disproportionate slowdown in motor vehicle trade. Regional patterns in GVCs GVCs often operate on a regional basis to optimize time and transport costs. To examine the different patterns, GVC trade can be divided into three regional supply chains: EU27+, which includes the 27 members of the European Union (EU27)5 as well as their GVC partners in Morocco, Tunisia, and Turkey; North American Free Trade Agreement (NAFTA) countries (Canada, Mexico, and the United States); and East Asia (including Japan, China, the Republic of Korea, and other large Asian traders). Each of the regions includes both developed and developing countries. Remaining trade is classified as “rest of the world.” For the three sectors analyzed here, over 96 percent of their trade involves one of the three big regional GVCs, either as an exporter or importer. On a regional basis, the slowdown in EU27 demand is most notable. EU27 imports of GVC goods (including intraregional imports) did not recover as dramatically as imports of other regions, and have been in absolute decline in real terms since 2011. In the most recent 12-month window ending in the first quarter of 2013, real growth in NAFTA and East Asian GVCs is also negative, while GVC imports by the rest of the world have been more resilient (figure 3).

Differences between intermedate and final goods trade Figure 4 shows one difference between the current trade slowdown and the Great Trade Collapse. Looking at the three sectors together, the strong bullwhip behavior (intermediate trade declining more than final goods

Source: ITC, World Bank. Note: EU27+ = EU 27 and Morocco, Tunisia, and Turkey; NAFTA = North American Free Trade Agreement; ROW = rest of world.

recent window for which data are available, that ending in the first quarter of 2013, total global merchandise trade increased by 2.0 percent, while GVC trade declined by 0.6 percent.

3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise

Figure 4. Within GVCs, Trade in Intermediate Goods Has Slowed More Rapidly Than Trade in Final Goods

2013M1

2012M9

2012M5

2012M1

2011M9

2011M5

2011M1

2010M9

2010M5

2010M1

2009M9

2009M5

2009M1

2008M9

2008M5

2008M1

year-on-year real growth growth for period ending at date shown (%)

recent years, the most notable slowdown in demand has been in Europe. While world economic activity grew by 4 percent in 2011 50 and 3.2 percent in 2012, with projections for 2013 at 3.3 percent, in the EU-27, it 40 only grew by 1.6 percent in 2011, and conworld world GVC, intermediate world world GVC, final tracted by 0.2 percent in 2012. The euro 30 area seems to have been driving the EU performance, with 1.4 percent growth in 2011 20 and 0.6 percent contraction in 2012, and a further contraction for 2013 estimated at 10 0.3 percent.6 The slowdown in Europe is having sig0 nificant spillover impacts on its trading partners in Africa and the Middle East. This -10 point is highlighted by the network representation of world trade (using a Minimal -20 Spanning Tree [MST]) shown in figure 5, which details overall bilateral trade for -30 2010. MSTs enable at-a-glance visualization of large and dense networks, such as world -40 trade. The lines in the network represent link weights and are proportional to trade -50 flows, while the size of the nodes reflects a Source: ITC, World Bank. country’s degree of centrality in global trade. Following the links of the network, it is possible to assess which countries’ exports are most likely to trade) that characterized 2008–9 does not appear in aggrebe affected by a contraction of activity and demand in the Eugate GVC trade in the latest slowdown. Trade in intermediate ropean Union. and final GVC goods are currently decelerating at a proporWorld trade appears polarized around three main blocks. tional rate. China, of course, is the most relevant trade partner for many Given the hypothesized origin of bullwhip behavior in the management of uncertainty and inventories, there are countries (higher degree of centrality), but so are the United some potential explanations for this difference. It could be States and Germany. More specifically concerning the Eurothat firms in GVCs have learned from 2008–9 and are able to pean Union, it appears that there are three distinct sub-blocks. adjust production in the supply chain more efficiently and One is centered around Germany and comprises mostly better anticipate the slowdown. Also, because the current countries from the wider European region. The second is censlowdown is more modest, there is less uncertainty in the systered around France and comprises a mix of Middle Eastern, tem overall for firms to react to, thus less bullwhip behavior. African, and European countries. The third sub-block centers However, breaking down the behavior of final versus inon Italy and comprises Mediterranean and North African termediate trade by sector and region, it appears that bullpartners. whip behavior is not entirely extinct. Over the last two years, The same representations can be carried out for indideclines in intermediates trade in excess of final goods trade vidual industries. Figures 6–8 provide MSTs of the three characterized the electronics GVC for most of 2011 and GVCs industries discussed here: electronics, motor vehicles 2012, but have recently moderated. Intra-EU trade in GVCs (including motorbikes), and apparel (including footwear). as a whole has exhibited bullwhip behavior. For the year endFor the electronics industry, the clustering broadly follows ing in the first quarter of 2013, real intra-EU GVC trade in the main regional blocks. The network centered around EU final goods increased by 1.3 percent, while intra-EU GVC countries comprises mostly European, Middle Eastern, and trade in intermediates decreased by 5.4 percent. North African countries. The European trade network links back to the rest of the world through the United States. MoGVCs’ Impact on EU Slowdown tor vehicle trade is also very regional in scope, but the number of countries clustered around Europe is higher than in Trade is likely to slow down most rapidly in countries exportthe case of electronics, for which Asian countries are most ing to regions where demand has slowed down the most. In 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise

Figure 5. World Trade Network, 2010

Figure 7. World Trade Network, 2010: Passenger Vehicles

NFK

MWI AFG

UZB LTU BLR

UKR

BRA

PRT

BFA PRK UGA

GIB

SEN

VGB

MAR

COM

LBN

FRA

VCT

CMR

IRL

MDG SVK GEO

TUR

COK

MKD

FLK BIH

SDN

MLT KAZ

HUN

AGO

CZE

POL

ROM

DNK

WLF

JPN CYM

BGR ESP

NRU WSM TON NFK FJI

BLX HRV

TCD GAB GIB

LAO

TUN

MAR

FRA

ASM

ARM

ZAF MHL

YUG

DEU

NLD

AUT

SPM CYP

TUR

CHE TKM

ISL

URY AZE

GIN

GRC

DMA

NPL SAU GNB BEN BLZ

DJI GNQ CIV

VNM

PRY TTO

CYM MYS SOM

SYR

SGP

PYF

FSM

LCA GUY

BLZ

PCN

ANT

CAN

USA

HTI

SUR AIA

EGY

ISR

GAB

KNA JAM

ESH

AND

DOM NIC

CRI

HND KNA BHS TCA SHN AGO IOT GTM LCA SLE NIC VEN MEX ISR BRA COL PAN COG CHL MSR KHM ABW VGB ARG ECU GUY ANT UMI SYC

KWT

GTM

TTO

TKL

DMA

VEN COL

SHN

ABW MDV

NGA

CXR

PAN

BHS

MSR

NIU

GRD

HND

MEX

BLR ATG UZB MNG KAZ MMR

SLV

TCA BGD

BRB LBY COK JAM CRI USA CAN DOM SLV

RUS

ECU

PER

HMD

Source: Comtrade.

BVT TMP CIV BRB

ATG

Figure 8. World Trade Network, 2010: Apparel and Footwear

Source: Comtrade.

MRT

AND

ASM TON

PAK

NZL

PER

BGD PAN ERI GHA BRB CIV PYF LAO IRQ

IOT

CHL

NFK CCK SPM

CAN ARG

IND

TKM

BTN

ISR

BRA MEX

AIA BEN

YEM

GRL LBY LBR

USA TCA FLK HMD

BIH

POL TUR GMB SVN HRV HUN

AND CAF YUG BDI LTU GNB DEU FRA PRT SVK CHE ESP ITA TUN BGR LVA DNK EGY AUT CYM SGS ROM SYR COG NLD RUS ZAF COM BLX CZE NER MLI KGZ TKL MKD ATF UMI UKR GBR SWE BLR ALB NOR GUM SHN MLT EST AZE IRL SLE MOZ TJK

CXR

AUS

BGR MKD HRV EGY

MNP

VEN

IRN

CHN

IRN DZA PRK ZMB ZAR DOM ARE WSM CUB MMR SAU SDN MHL SUR MUS BHR TKM SGP CHN UGA MNG TCD BFA VGB TUV MWI URY DJI GIN QAT AGO

IDN

NIC USA GRD TCA CCK HND SHN CYM BHR SLV ZAF TKL GTM CAN ZWE MEX ISR NGA SPM

CRI

KHM ASM VUT PLW GUYDMA NCL AFG GRD KIR TGO PNG ECU COL NIC JAM SLV ZWE LCA CRI FSM MNP VNM NRU JOR BRN SEN GTM MDG SLB TTO THA MDV TON IDN TMP KAZ UZB VCT HND LBN PHL ABW ISL GEO ANT PRY CYP ARM MDA MRT ESH CMR BOL KNA MSR NPL HTI CPV BMU GIB FJI RWA BHS GAB MAR BLZ STP NGA

JPN

GNQ

BVT

NIU

CAF

MNG

THA

BEN

DOM

SYC

FIN

PCN

IND

ZAR PHL TMP YEM SDN TZA MYS GNQ GAB SLE BOL URY VCT IRQ JPN LBR ARG PAK ABW PER MUS PYF CHL LCA PLW KOR FJI VNM DMA ANT SOM ATG PAN VUT PRK SUR RWA QAT CIV TTO UZB FSM BFA BRB KIR HTI COL GIN SLB ZMB LBY GUY ECU MLI BHS PRY NCL KAZ UGA AFG

MYS

CXR AUS

KEN

TZA

KOR PCN LKA ATG

OMN KWT

SOM

ETH

NFK

ISL VGB DZA TJK GEO NER OMN TUR ATF

COM BLZ BDI DJI TGO BTN KGZ KEN JAM NPL LBN WSM

GMB KWT SGP

MOZ

ITA

ROM ARM AIA BVT

TUN

BMU ALB ERI MDA UKR YUG BIH

ARE

MLT

FLK

Figure 6. World Trade Network, 2010: Electronics

COK

SLB TMP PER

TJK

ATF

SPM

NZL

IDN

KWT

SLE

CYP

NIU

MWI

UGA OMN MLI QAT

UKR

VEN

IRN

BOL

GUM

MDV

GEO

VUT

BGD ESH FSM DZA

PLW TUV

STP

ARE

JOR BHR

BLX

IOT

WLF

MOZ

JPN

ZAR

DNK

MDG

KOR

PRK

BRN

BFA

SWE

BHR

CHN

ZWE

EST

MNP CPV

HTI BMU PHL

CAF MKD

MUS

NOR TKL BTN

FIN

CUB

HUN

LBR

TZA

ETH

THA

CMR KGZ YEM

MLT

ZMB

ATF

BDI

IND LAO

AUS

LVA

LTU

ROM

ITA

BIH

POL

SVK SVN

MDA

AIA

KIR

MMR

SAU

SUR KEN

CZE

ALB VCT HMD RWA MRT

NRU

CCK

SOM

GRL

ERI ALB DZA GRC

NZL

PLW

THA

VUT

LBR

GBR

OMN

BEN TUV

TGO

TKM

FJI

WSM

AUS

TCD KHM VNM

SGP

CXR

KOR

ZAR

ETH

SWE

SVN YUG

NLD

ITA

GMB

PNG

BMU

IDN

AZE

GUM

SYR

YEM

MNG

MRT PHL IRN

PYF

DEU

NOR

STP

HRV

BGR

CHL

COG

CHN

BRN

SGS

AUT

ISL

LBY

RWA

GBR CHE

MUS

TON GNQ

IRQ PAK

KEN

NCL

SYC NER

CAF

TUN

ESP

MHL

TZA CUB

IRL

MNP ASM

TGO

NZL

NCL

PRT

GRL PCN

UMI

MYS

SLB

COM AND

EGY FLK SGS

IND

DJI

TJK CPV

NPL BTN

RUS

KGZ

LKA

ARG

PAK

IRQ

SDN AFG PNG GMB LKA NER GRD ERI

NGA

GHA

GIN

ZAF

ARM

FIN

URY

KIR

WLF

GNB

PRY

ZMB

MLI

MDA EST

BVT

ARE

BOL MOZ

GHA

LVA

QAT NIU

ZWE

LBN JOR

SEN

BDI

CCK

MWI

TCD

MMR

SGS

GBR

IRL

CYP HMD UMI

LAO

TUV BLX MDG

FRA

GRC LTU NLD

CHE

DEU

AZE

GRL

POL

HUN SVK

CZE

SVN

NRU

MHL

KHM

ESH

KNA

AUT

SWE

DNK NOR

BGD

MSR

BRA

ESP GIB

AGO

CUB IOT

BLR

RUS

LKA

SAU FIN

COK

SYR

EST LVA

PRT

MAR

CPV STP

COG

SEN MDV

GNB

SYC WLF PNG GHA ETH GUM CMR

BRN

JOR

Source: Comtrade.

GRC

Source: Comtrade.

central to the network. The most central EU country for motor vehicles is Germany. Finally for apparel, Italy, France, and Spain represent more central nodes in the global network. Dampened demand, which has concentrated in these countries, may explain the subdued growth of apparel GVCs shown in figure 2. Ongoing financial market friction in the European Union, particularly in the euro area, have also greatly increased uncertainty, leading to delays on investment and expenditure decisions, especially regarding durable goods. Increases in uncertainty can have a dampening effect on trade, as producers scale back investment decisions and inventory accumulation (as discussed above under the bullwhip effect) and consumers postpone purchases. Evidence shows that,

even taking traditional determinants of trade into account, exports to markets experiencing stock market volatility decline by more than expected (Taglioni and Zavacka 2012). The deterioration of financing conditions in the European Union also pose a problem. Inadequate financing constrains working capital, which affects firms’ investment decisions and limits funds, particularly for small and medium enterprises, which are the backbone of the EU economy. The prolonged financial crisis in the European Union likely has had a significantly negative impact on firms, which in turn has affected the exports of their foreign suppliers. Conclusion The recent slowdown in global merchandise (2010-early 2013) confirms some of the lessons learned during the Great Recession. Trade in complex products organized in GVCs has once again been more sensitive to global downturns than has trade in simple products, particularly for motor vehicles. For motor vehicles, the start-and-stop nature of incentives in

5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise

many countries may have added to the volatility. Conversely, trade in simpler products appears once again to be more resilient to global downturns. Although evidence is still accumulating, this suggests that the preference often expressed for countries to go “up the value chain” by specializing in more and more complex products may need to be qualified. While diversification is one way to manage risk associated with export portfolios, focusing on simpler products less associated with GVCs may be another way, and there is likely to be a tradeoff between these two strategies. Continued monitoring and research in this area is warranted. Acknowledgments The authors would like to thank Gianluca Santoni for the analysis of networks and accompanying charts, and Guillermo Arenas and Amir Fouad for additional research assistance. About the Authors Michael J. Ferrantino and Daria Taglioni are Lead Economist and Senior Economist, respectively, in the World Bank's International Trade Unit. Notes 1. Unless otherwise specified, the trade data used in this note represent the average of import and export data to smooth out the effect of short-run discrepancies in the two series. 2. World Bank’s Datastream, maintained by DEC. 3. Baldwin (2009) reviews the explanations for the Great Trade Collapse offered by economists in mid-2009, including the role of GVCs. 4. Due to lagged reporting of monthly figures at the product level, data for this analysis are only available through the first quarter of 2013. Nominal GVC trade has been deflated using

sector-specific import price deflators for the United States, taken as a rough proxy for the global market. 5. In June 2013, Croatia joined the European Union, becoming the 28th member. However, since this analysis covers the period up to mid-2013, the analysis includes only the EU-27. 6. See the International Monetary Fund’s World Economic Outlook (April 2013, http://www.imf.org/external/pubs/ft/ weo/2013/01/). References Altomonte, Carlo, Filippo Di Mauro, Gianmarco Ottaviano, Armando Rungi, and Vincent Vicard. 2012. “Global Value Chains During the Great Trade Collapse: A Bullwhip Effect?” European Central Bank Working Paper Series No. 1412. Baldwin, Richard, ed. 2009. The Great Trade Collapse: Causes, Consequences and Prospects. London: Center for Economic Policy Research and VoxEU.org. Escaith, Hubert, Nannette Lindenberg, and Sébastien Miroudot. 2010. “International Supply Chains and Trade Elasticity in Times of Global Crisis.” Economic Research and Statistics Division Staff Working Paper ERSD No. 2010-08, World Trade Organization, Geneva. O’Rourke, Kevin. 2009. “Collapsing Trade in a Barbie World.” http://www.irisheconomy.ie/index.php/2009/06/18/collapsingtrade-in-a-barbie-world/. Sturgeon, Timothy J., and Olga Memedovic. 2011. “Mapping Global Value Chains: Intermediate Goods Trade and Structural Change in the World Economy.” United Nations Industrial Development Organization, Development Policy and Strategic Research Branch Working Paper 05/2010, Vienna. Taglioni, Daria, and Veronika Zavacka. 2012. “Innocent Bystanders: How Foreign Uncertainty Shocks Harm Exporters.” World Bank Policy Research Working Paper No. WPS 6226, Washington, DC. Zavacka, Veronika. 2012. “The Bullwhip Effect and the Great Trade Collapse.” European Bank for Reconstruction and Development Working Paper No. 148.

The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise.

6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise