Globalization in food and agriculture · 10.1 Globalization as an ongoing process Globalization refers to the ongoing process of rapid global economic integration facilitated by lower

265

The preceding chapter focused on the role of tradeand trade policies as driving factors for increasinglyintegrated markets. This chapter on globalizationwill expand the analysis by identifying the othermain factors that drive global economic integrationand analyse their main effects on food and agricul-ture. These are presented in three major sections.The first part provides a definition of the process ofglobalization, placed in its historical context.Emphasis is given to the importance of factors thatreduce transaction costs, notably on the impacts ofnew transportation and communication technolo-gies. The second section presents the main featuresof globalization in agriculture, and discusses whysome countries have been successful in integratingtheir food and agricultural economies into therapidly growing world markets, but also why othershave largely failed to do so. This includes factorssuch as openness to trade and capital flows, abilityto adopt external expertise and technologies, butalso the importance of factors relating to acountry’s geographic location or its infrastructureendowment. The third part presents the options,the potential and the limits that developing coun-tries are facing for future integration into globalfood and agriculture.

10.1 Globalization as an ongoingprocess

Globalization refers to the ongoing process of rapidglobal economic integration facilitated by lowertransaction costs and lower barriers to movementsin capital and goods. It has shown itself in agrowing interdependence of the world’s economies,rapidly rising trade flows, increases in capitalmovements and an increasing internationalizationof production, often organized within and betweenmultinational corporations. To a large extent, glob-alization has been brought about by a massivereduction in transaction costs, which in turn wasmade possible through more efficient transporta-tion and communication facilities as well as innova-tions in organizing complex logistical processes.Trade and capital flows have also been boosted by asystematic reduction in trade and investmentbarriers. This process has brought about massiveincome gains for those who participated. Broadly,the integration into a larger and more competitivemarket has increased the returns to investment forproducers and provided consumers with a greatervariety of products at lower prices.

This process of growing integration of theworld economy has also given rise to numerous

Globalization in food and agriculture

10CHAPTER

266

concerns. Most prominent is the concern about thegrowing marginalization of entire countries or soci-etal groups within countries. There are in factmany countries that have been left out of theoverall economic integration and growth process,in some cases, despite sincere efforts to open up toforeign trade and capital flows. Over the 1990s,rapidly integrating economies recorded a percapita income growth rate of more than 4 percentp.a. while the income available per person in lessintegrated countries shrank by 1 percent annually(World Bank, 2001e). The rapid growth in agricul-tural trade has given rise to concerns that diseasesand pests will be hard to control and contain at thelocal level. Moreover, there are socioculturalconcerns that globalization could destroy thecultural heritage (including dietary habits) as wellas traditional societal and social links that haveevolved over centuries. Finally, there is widespreadconcern about a growing economic, social andcultural dependence on a few dominant countriesor corporations, which are seen to have the poten-tial of disempowering entire societies.

While the term “globalization” has been coinedonly recently, its driving forces and its principalimpacts are of older date. Similar developments,albeit of a more limited scope, have characterizedglobal economic development in the past. Inparticular, innovations that reduce transactioncosts (better transportation and communicationtechnologies) have always had a strong acceleratingimpact on global integration. A look back alsosuggests that the process of economic integration isa non-continuous one. It is often a process of wavesthat occur when new technologies are widelyadopted around the world. Similarly, trade andinvestment liberalization is being negotiated andimplemented in rounds. The two developments,technological change and liberalization, can bemutually reinforcing and create particularly notice-able waves of globalization. The current wave ofglobalization is driven by major technologicalbreakthroughs in transportation and communica-tion technologies (notably the Internet, mobile tele-phone technology and just-in-time systems) intandem with various efforts to liberalize interna-tional trade and investment flows.

The first wave of globalization during the secondhalf of the nineteenth century. The first wave ofrapid global integration began in the second half ofthe nineteenth century and was brought about by acombination of breakthroughs in transportationand communication technologies. Trade betweencontinents was boosted by the shift from sail tosteamships, which resulted in a tremendous declinein transatlantic transportation costs as well as fasterand more reliable connections. Trade in agricul-tural commodities, typically bulky, perishable orboth, enjoyed a particular boost. Transatlantictrade in grains and oilseeds – previously circum-scribed by high transaction costs – shot up sharply.This brought new land into production, mostnotably in the Midwest United States and someparts of Australia.

Agricultural trade was further accelerated bythe advent of the railways, which resulted in afurther and sharp reduction of transportation costswithin continents. Lower transaction costs height-ened competition and brought about not only asignificant downward pressure on prices, but also agrowing convergence of commodity prices acrosscontinents. For example, in 1870, a bushel of wheatsold for 60 cents in Chicago but for double thatprice in London. The difference was largely aresult of high transportation costs from Chicago toLondon. With railroads and steamships,1 transportrates between Chicago and London fell to 10 centsa bushel between 1865 and 1900 and the pricedifferentials for wheat declined accordingly(Henderson, Handy and Neff, 1996).

The decline in transaction costs also had signif-icant impacts on the overall volume of interconti-nental trade, market shares and income. UnitedStates exports of grain and meat to Europe, forinstance, increased from US$68 million in 1870 toUS$226 million in 1880, which boosted farmers’incomes in the United States and the welfare ofconsumers in Europe. The new transportationfacilities also reduced costs for internal shipmentsand, together with cheaper food supplies fromabroad, increased food security at the local andregional level. For the first time in history, thisbrought about years of “lower agricultural produc-tion without famine” in Europe (Tilly, 1981).

1 Between 1850 and 1913, global overseas transportation capacity increased by more than 500 percent. At the same time, tankersand vessels with cooling facilities vastly expanded the range of products exchanged within and across countries and continents.

267

GLOBALIZATION IN FOOD AND AGRICULTURE

Lower transportation costs also affected labourmobility and labour costs. Sixty million peoplemigrated from Europe to North America andAustralia to farm the newly available land. As landwas abundant, it created an opportunity for manyimmigrants to earn an income that exceeded by farthe wages they used to earn in Europe. In Europeitself, in turn, it created a relative labour shortageand an upward pressure on wages both in absoluteterms and relative to land prices. Overall, immi-gration led to a narrowing of differences in wagesin all globalizing regions. “Emigration is estimatedto have raised Irish wages by 32 percent, Italian by 28 percent and Norwegian by 10 percent.Immigration is estimated to have lowered Argentinewages by 22 percent, Australian by 15 percent,Canadian by 16 percent and American by 8 percent”(Lindert and Williamson, 2001). Indeed, migrationwas probably more important than either trade orcapital movements.

The backlash after 1914. With the end of the FirstWorld War, trade policy went into reverse andmany nations stepped up border protection. Theincrease in tariffs was built on the notion thathigher protection would help rebuild the domesticindustries that had suffered or were destroyedduring the war. The process started in Europe.France, Germany, Spain, Italy, Yugoslavia, Hungary,Czechoslovakia, Bulgaria, Romania, Belgium andthe Netherlands all raised their import tariffs tolevels comparable to those before the war. Even theUnited Kingdom, a free trade nation, declared that“new industries since 1915 would need carefulnurturing and protection if foreign competitionwere not again to reduce Britain to a technologicalcolony”.

In June 1930, when the United States Congresspassed the Hawley-Smoot Tariff Act, the UnitedStates joined in the new protectionist wave.Agricultural tariffs were increased particularlysharply, both in absolute terms and relative toindustrial ones. In reaction to the sharp increase inUnited States tariffs, other countries enacted retal-iatory trade laws. The spiralling tariff increases puta brake on global trade and reversed much of theliberalization that resulted from the first wave ofglobalization. Between 1929 and 1933, UnitedStates imports fell by 30 percent and, more signifi-cantly, exports fell by almost 40 percent. The sharp

decline in trade aggravated the internal economicsituation, and the depression in the United Statesintensified and engulfed much of the rest of thethen economically integrated world.

The second wave of globalization, 1945-80. Theexperience gathered from the reversal to protec-tionist policies during the interwar period gave animpetus to a new wave of internationalism after theSecond World War. The new wave of trade liberal-ization was, however, more selective both in termsof countries participating and products included.By 1980, developed countries’ barriers to trade in manufactured goods had been substantiallyremoved, but barriers for developing countries'agricultural products had been lowered only forthose primary commodities that did not competewith agriculture in the developed countries. Bycontrast, most developing countries had erectedtrade barriers against imports from each other andfrom developed countries.

The resulting effect on trade flows was veryuneven. For developed countries, the second waveof globalization was a spectacular success. Freertrade between them greatly expanded theexchange of goods. For the first time, internationalspecialization within manufacturing becameimportant, allowing scale economies to be realized.This helped to drive up the incomes of the devel-oped countries relative to the rest of the world(World Bank, 2001e). For developing countries, itperpetuated the North-South pattern of trade, i.e.the exchange of manufactures for land-intensiveprimary commodities, and this impeded them inexploiting their comparative advantage in labour-intensive manufactures. In addition, as discussedbelow, many developing countries adopted a policyapproach that was not conducive to a greater inte-gration into the globalizing world economy.

The economic policy approach adopted in manydeveloping countries during the 1950s and 1960swas strongly influenced by the work of Raul Prebish.Under the assumption of balanced trade and pricestability, Prebish established the following relation-ship between the relative growth rates of aneconomy vis-à-vis its trade partners and the incomeelasticities for its exports and imports: gi / gw = ex /em, where gi and gw are the trend growth rates ofthe economy and the rest of the world, and ex andem the export and import income elasticities.

268

The policy message from this relationship wasstraightforward: if a country wants to grow morerapidly than the rest of the world, its export elas-ticity needs to be higher than its import elasticity.The actual situation in developing countries,however, was precisely the reverse. Typically, theyexported primary goods with low income elasticitiesand imported manufactured products with highincome elasticities. As a result, growth without abalance-of-payment constraint was assumed to beimpossible without a continuous depreciation of thereal exchange rate or the steady accumulation offoreign debt. This so-called “elasticity pessimism”was the main rationale behind the import substitu-tion policies of this period.

For much of the 1950s and 1960s, import-substituting industrialization (ISI) was seen as away out of this deadlock. ISI was based on the ideathat domestic investment and technological capa-bilities can be spurred on by providing homeproducers with (temporary) protection againstimports. Whether and to what extent ISI helpedor hindered development remains controversial.On the one hand, the so-called “‘consensus view”emphasizes that ISI policies were at the heart ofthe problems that many of their adopters encoun-tered in the subsequent decades when theyopened up their economies (for example, seeOECD, 2001c). On the other hand, there areclaims (Rodrik, 1997; Hausmann and Rodrik,2002) that ISI worked reasonably well, notably inraising domestic investment and productivity. Ithas been stressed that numerous economies inLatin America and the Near East recorded robustgrowth under ISI policy regimes.

There is, however, a broad consensus that ISIwas an ineffective response to weather theeconomic turbulence of the 1970s, which witnessedthe abandonment of the Bretton Woods system offixed exchange rates, two major oil shocks andother commodity boom-and-bust cycles. For agri-culture, ISI strategies meant higher input costs andtherefore negative effective protection, i.e. implicittaxation. In conjunction with explicit taxes onoutput and exports, ISI strategies created a consid-erable burden for agriculture in many developingcountries, put a brake on agricultural export growthand slowed their integration into global agricul-tural markets. On average, for the period from1960 to 1984, the bias against agriculture

depressed the domestic terms of trade for agricul-ture by 30 percent. In the extreme cases of Côted’Ivoire, Ghana and Zambia, the average biasagainst agriculture reached levels of 52, 49 and 60 percent, respectively (Schiff and Valdes, 1997).

The current wave of globalization. The last twodecades of the twentieth century marked the begin-ning of a new wave of globalization. Like the firstwave about a hundred years earlier, it was broughtabout by a combination of lower trade barriers andnumerous technological innovations that stronglyreduced transaction costs for movements not onlyof goods but also of people and capital. This isparticularly apparent from the substantial increasein international migration and capital movements,which were of less importance during the secondwave of globalization. Unlike its predecessors, thiswave of globalization included many more devel-oping countries, even though not all of them wereable to harness globalization to their benefit.Particularly countries in sub-Saharan Africa failedto participate, resulting in a further widening oftheir income gap with both integrating Asianeconomies and, even more so, the fully globalizedeconomies of the North.

Most countries in East Asia were able to reapsubstantial benefits by exploiting their compara-tive advantage of cheap and abundant labour.Some countries in Latin America and the NearEast/North Africa region were also able to inte-grate fast. A common feature of successful integrators is an above-average shift towardsexports of manufactures. Countries such asChina, Bangladesh and Sri Lanka already haveshares of manufactures in their exports that areabove the world average of 81 percent. Others,such as India, Turkey, Morocco and Indonesia areswiftly approaching the world average. Anotherimportant change in the exports of successfullyglobalizing developing countries has been theirsubstantial increase in exports of services. In the early 1980s, commercial services made up 17 percent of the exports of developed countries,but only 9 percent of the exports of developingcountries. During the third wave of globalization,the export share of services in the former group increased slightly, to 20 percent, but fordeveloping countries the share almost doubled to17 percent (World Bank, 2001e).

269


10.2 The main features of globalization and the correlates of success

10.2.1 Freer trade and outward-orientedpolicies

Chapter 9 has already dealt with the main devel-opments in global agricultural trade, its impor-tance within overall trade and the structuralchanges that have taken place over the past 40 years. It also provided an overview of likelytrade developments for the next 30 years and thetrade policy issues that are expected to arise fromthe projected shifts in trade flows. In this section,emphasis is placed on the potential role of trade fordevelopment and poverty alleviation.

The links between trade, development andpoverty have been subject to an extensive andheated debate. While proponents and opponentsagree on the central importance of freer trade forincreasing global welfare, there is considerabledisagreement as to whether and to what extentfreer trade can be harnessed by individual coun-tries as a means to promote development and fightpoverty. There is also considerable disagreement asto how the transition towards freer trade, i.e. thespeed, timing and sequencing of liberalizationmeasures, should evolve. Some of these issues willbe addressed in the following section.

The consensus view. Economists have been assertingfor a long time that trade liberalization is good foreconomic development, particularly in developingcountries. The benefits from openness are assumedto arise from the efficiency gains that flow fromsuperior resource-allocation decisions in more openmarkets (Bhagwati and Srinivasan, 1999). Theresult is an increase in economic growth. Morerecently there have also been numerous empiricalstudies that suggest that openness to trade andinvestment flows has had a positive effect not onlyon economic growth but also in helping to fightpoverty. Among the most influential empiricalstudies are those by Edwards (1998) and by theWorld Bank (Dollar and Kraay, 2000, 2001). Wolf(2000) summarizes much of this literature.

In view of its importance for the ongoing policydebate, the main conclusions of the World Bankstudy are summarized below. The first concerns the

link between growth and openness. Dollar andKraay examine this relationship using an econo-metric study covering a sample of 72 developingcountries. Avoiding some of the pitfalls of earlierstudies by using a single indicator of openness (theratio of trade to GDP), the authors arrive at anumber of important conclusions:■ Weighted for population, the per capita income

of the group of “globalizers” grew at 5 percenta year in the 1990s, compared with 1.4 percentfor the “non-globalizer” group.

■ Growth rates for the globalizers have beensteadily increasing since the mid-1970s, whilethose for the non-globalizers fell sharply inthe 1980s and recovered only marginally inthe 1990s.

■ Per capita income among the globalizers isrising more than twice as fast as in industrial-ized countries, while the non-globalizers arefalling further behind. On a population-weighted basis, countries that are open aregrowing 3.6 percent a year faster than others.On this basis, average income in a globalizingeconomy would double every 14 years,compared with 50 years in a non-globalizingeconomy: a growth gap that would haveprofound implications for poverty reduction.

The second conclusion concerns the relation-ship between economic growth and poverty reduc-tion. On the basis of an econometric exerciseanalysing economic growth in 80 countries over aperiod of four decades, it is argued that, on average,the income of the poor rises on a one-to-one basiswith overall growth. In other words, poor peoplecapture a share of any income increment thatreflects their existing share of income distribution.As the authors say: “It is almost always the case thatthe income of the poor rises during periods ofsignificant growth” (Dollar and Kraay, 2001).

On closer inspection, however, some of thenumbers look less impressive. One reason for thisis that averages have the effect of obscuring impor-tant differences between countries, especially whensamples are weighted for population (since thismeans that large countries such as China have adisproportionate influence). Using an unweightedaverage, the per capita growth rate for the global-izers in the 1990s falls to 1.5 percent. Moreover, tenof the 24 countries in the group have growth rates

270

of 1 percent or less. Further disaggregation revealsthat one-third of the “globalizing” countries havelower average growth rates for the 1990s than the“non-globalizing” group.

The critique of the consensus view. The basiccritique of the consensus view is that the linkbetween openness and growth is one of correlationbut not, or at least not necessarily, one of causation.Simply put, openness is essentially an economicoutcome, captured (in the case of the World Bankstudy) by the ratio of trade to GDP, but not aninput, i.e. a policy tool to arrive at higher growth.2

When focusing on the causal relationshipbetween trade policy, growth and poverty reduc-tion, the critics of the consensus view claim that itappears to be an upside-down version of reality(Rodrik, 2001 and Oxfam, 2002). In fact, theystress that some of the most successful globalizersare anything but radical liberalizers, while many ofthe most radical liberalizers have actually achievedvery little in terms of economic growth and povertyreduction. They claim that no country has everdeveloped simply by opening itself up to foreigntrade and investment and that practically all oftoday’s developed countries embarked on theirgrowth behind tariff barriers, and reduced protec-tion only subsequently (Rodrik, 2001).3

There are also many examples in agriculture,where appropriate domestic policy settings and thetiming and sequencing of liberalization steps haveproved to be more important than a complete andimmediate reduction of border protection. Some oftoday’s most successful agricultural exporters (e.g.China and Viet Nam) established their internationalcompetitiveness under protection and import substi-tution regimes and embarked subsequently on“policy reforms”.4 In many cases, success was builton a promotion of export-led growth combined witha domestic investment and institution-buildingstrategy to stimulate entrepreneurship and the will-ingness to assume risks. Another important factorhas been that mechanisms are put in place to ensure

that excess capacities are cut back, and to create exitpossibilities for non-performing sectors or actors,and that the opening-up process to internationalcompetition is phased in a determined manner (forexamples, see below).

Notwithstanding the importance of temporarytrade protection measures, the proponents of fastand full liberalization stress that no country hasdeveloped successfully by turning its back on inter-national trade and long-term capital flows. Veryfew countries have grown over long periods of timewithout experiencing an increase in the share offoreign trade in their national product. In practice,it is hard to imagine that a country can create andsustain growth if it remains shut off from the forcesof competition that help to innovate and upgradeits productivity. Moreover, it is equally hard toimagine that developing countries would notbenefit from imported capital goods that are likelyto be significantly cheaper than those manufac-tured at home. Policies that restrict imports ofcapital equipment raise the price of capital goods athome and thereby reduce real investment levels.Exports, in turn, are important since they permitthe purchase of imported capital equipment.

The agricultural sector in many developingcountries has been particularly adversely affected bythe inward-oriented industrial development strate-gies of the 1950s and 1960s. In some countries theanti-agriculture bias remained a policy featurethroughout the 1970s and 1980s (Schiff and Valdes,1997). Import substitution policies for manufac-tures restricted capital good imports for agriculture,raised input costs and resulted in often significantnegative effective rates of protection. This held backreal investment levels in agriculture and slowedexport performance in many developing countries.In some developing countries, industrial protectionand restrictions on capital good imports for agricul-ture were accompanied by direct taxation of agricultural exports, placing agriculture at a disad-vantage both relative to other sectors and vis-à-visdeveloped country competitors.

2 Dollar and Kraay acknowledge this possibility, when declaring that “we use decade-over-decade changes in the volume of trade as an imperfectproxy for changes in trade policy” (Dollar and Kraay, 2001).

3 Bussolo and Lecomte (1999) also stress that trade policy theory does not unambiguously suggest that protection has a negative impact on growthin developing countries. However, they emphasize that those countries that apply more open trade regimes, together with fiscal discipline and goodgovernance, have enjoyed higher growth rates than those implementing restrictive policies. An open and simple trade policy can foster someexternal discipline, helping to reduce distortions on domestic markets, and to narrow the scope for wrong or unbalanced policies in other areas,as well as for rent-seeking and corruption that do not normally favour the poor.

4 For much of the 1970s and 1980s, developing countries’ agriculture was heavily discriminated against (see Chapter 9). Wherever and as long asagriculture was taxed, either directly or through macroeconomic measures, development slowed and international integration suffered.

271


Openness and development in agriculture – somecountry examples. Viet Nam’s rapid economic andagricultural development over the 1990s is nowcommonly regarded as one of the most successfuldevelopment stories of the last decade. AnnualGDP growth has been consistently high throughoutthe 1990s, averaging 7.6 percent. Over the sameperiod, agricultural output has been growing atalmost 5 percent per year, far outstripping demandin local markets (Government of Viet Nam, 2001).Poverty has declined substantially and the numberof undernourished has dropped by 3 millionpeople (FAO, 2001a).

Export markets provided an important sourceof demand to sustain growth. Over the 1990s, thevalue of agricultural exports shot up by a factor of3.5 and, for a number of commodities such ascoffee and rice, Viet Nam emerged as a leadingexporter in world markets. By the end of the1990s, rice and coffee exports combined generatedabout US$2 billion in foreign exchange earnings(1997/99 average), accounting for nearly 20 percentof the country’s total merchandise exports.

The foundations for Viet Nam’s rapid integra-tion into the global market were laid in 1986 withthe introduction of Doi Moi, Viet Nam’s economicrenovation programme. At the heart of the reformwas a decollectivization process, through whichfarming families received most of the land. Intandem, farmers were allowed to increase sales tothe market and agricultural taxes were reduced.Agriculture also benefited from other fiscalreforms, the creation of a Treasury system, and thereform of the banking system, which created asecure deposit base and allowed fiscal operationsdeep into the country’s rural areas. These meas-ures had a profound effect on society, encouragingentrepreneurship and willingness to take risk.Finally, Doi Moi offered “return” options to workersin the new factories, thereby reducing risk forinternal migrants and further accelerating the fastdevelopment of rural areas.

There is no doubt that the success of the 1990swas also promoted by a growing openness in theglobal trading environment, in which Viet Nam’sexport performance benefited from decliningtariffs and non-tariff barriers. As in many coun-

tries, Viet Nam’s economy also enjoyed all otherbenefits of globalization, such as cheaper and fastertransportation and communications. However,while benefiting from improved market accessabroad, Viet Nam was slow in removing its ownborder protection or its trade-distorting subsidies.Particularly, agricultural import tariffs have beenraised repeatedly over the 1990s (see, for example,USDA, 1999c, 1999d and 2001b)5 and subsidieshave been provided with the aim of increasing agri-cultural production and exports. Fforde (2002)even maintains that the initial fast liberalizationprocess in the early 1990s did not allow the countryto build up enough expertise and competitivenessand put a brake on overall growth.

Policies also played an important role inmanaging the 2000/02 coffee crisis that severelyaffected large parts of Viet Nam's thriving agricultural sector. For example, a sizeable support programme was launched to help coffeegrowers regain international competitiveness. Theprogramme includes subsidies to upgrade coffeequality and to reduce production costs. It promotessmaller, less centralized processing factories andwarehouses suitable for the many different coffee-producing regions (USDA, 2001b) and supportsthe creation of overall and agricultural infrastruc-ture and the shift towards improved coffee vari-eties. But the new policy package also initiated arationalization process within Viet Nam’s coffeeeconomy. Changes in eligibility for the existing softloan programme are probably the most importantefforts in this context. Under the revised scheme,credit subsidies will not be offered to low-yieldproducers or inefficient operations, but only topotentially profitable farmers. In parallel, specialpreferences have been given to participatingfarmers to switch to arabica coffee or to improvetheir operation’s effectiveness (USDA, 2001b).

Overall there are probably three importantfeatures that have contributed to the success of thecoffee policy. First, policies play an active role inpromoting production, particularly production forexports. Second, support is not an open-endedgovernment commitment but is limited to kick-starting the process and helping the sector discoverwhere its comparative advantage lies. And third,

5 Viet Nam undertook, for instance, a far-reaching tariffication exercise in 1999 (effective December 1999), converting many non-tariff barriers intotariffs, in line with World Bank prescriptions. This process was in general accompanied by an increase in effective border protection, with tariffrates between 30 and 100 percent. Detailed tariff schedules are, for instance, available from USDA (1999c, 1999d).

272

once competitiveness is established, policies focuson the competitive producers and decline supportto non-competitive ones. In doing so, competitiveproducers are helped through the price troughswhile the non-competitive ones are encouraged toexit from the sector.6

A fairly similar set of reforms in China in thelate 1970s set the stage for the impressive economicperformance that has been the envy of poor coun-tries since. Per capita GDP (at current prices)increased by a factor of nine and the value ofexports by a factor of ten. Agricultural outputtripled, as did agricultural exports, while thenumber of undernourished declined by 76 millionpeople (from 1990/92 to 1997/99). In fact, Chinawas the single largest contributor to the reductionof undernourishment during the 1990s, accountingfor two-thirds of the progress made in fightinghunger (FAO, 2001a).

This rapid development process started withfairly simple initial reforms in the agriculturalsector. The communal farming system was loos-ened and the so-called household responsibilitysystem was introduced, allowing farmers to selltheir crops on the free market once they hadfulfilled their quota obligations to the state. Thegovernment remained actively involved in agricul-tural policy formulation and implementation. Theoverall process can be best described as one ofactive experimentation, in which productionexpanded rapidly under administrative pressure tofulfil production quotas, as well as under produc-tion incentives through input subsidies (water,fertilizer).7 In tandem, policies were put in placethat promoted the adaptation of new technologiesfrom abroad to the domestic production environ-ment (particularly the high-yielding varieties of thegreen revolution) which, over time, even enableddomestic researchers to take the lead in developing

new applications (hybrid rice, etc.).8 The impor-tance of adopting external knowledge and tech-nologies is discussed in Section 10.2.3 below.Finally, domestic policies also encouraged the exitfrom agriculture of unproductive farmers. Thesemeasures include the creation and promotion oftownship and village enterprises (TVEs) thathelped absorb the excess labour of rural areas or,more recently, massive investments in rural infra-structure to reduce transaction costs and increasecompetitiveness of farmers and food processors inChina’s hinterland.

Unlike Viet Nam and China, sub-Saharan Africalargely failed to take advantage of the growing tradeopportunities in global markets. Its share in globalexports, for instance, dropped from 3.1 percent inthe mid-1950s to 1.2 percent in 1990. This corre-sponded to an annual loss in export earnings ofabout US$65 billion. In trying to identify thecontributing factors to this decline, a World Bankstudy (Amjadi, Reinke and Yeats, 1996) found thattrade barriers abroad have not had a significantinfluence. On the contrary, once preferences weretaken into account, tariffs conveyed significantcompetitive advantages over competing goodsshipped from some other regions, and were even apositive factor for the location of commodityprocessing in Africa as opposed to some otherforeign locations.9 Similarly, non-tariff barriers(NTBs) of markets abroad did not account forAfrica's poor export performance. In fact, the shareof Africa's exports subject to NTBs (11 percent) isless than half the average for the group of devel-oping countries.

To draw general lessons from a few successstories is difficult. Nonetheless, there are a fewcommonalities that characterize successful global-izers. To begin with, all of them have bothoutward-oriented policies and domestic produc-

6 Viet Nam’s active policy engagement in promoting and disciplining production may also be regarded as a special case of policies that have beenpursued elsewhere in East Asia. “Where Korea differs from other developing countries in promoting big business, was the discipline the state exer-cised over these chaebols by penalizing poor performance and rewarding only good ones … The government as the controller of commercial bankswas in a powerful position to punish poorly managed firms by freezing bank credit. As a result only three of the largest 10 chaebols in 1965 –Samsung, Lucky-Goldstar and Ssangyong – remained on the same list 10 years later. Similarly, seven of the largest 10 in 1975 remained on thesame list in 1985” (Kim, in Nelson, 2000). The Korean Government was quick to shelve its plans for supporting particular firms or industries whennew information suggested that productivity would lag (Westphal, 1981, p. 34).

7 It should be noted that the system of incentives to increase or slow output, and even to leave or stay in agriculture, was accompanied by a rigidsystem of administrative measures that may not be at the disposal of policy-makers in market economies.

8 Agricultural policies were accompanied by non-agricultural policy measures that aimed to facilitate structural change and gradually to liberalize thenon-agricultural sector. The most important measures were the creation of township and village enterprises (TVEs), the extension of the “markettrack” into the urban and industrial sectors, and the creation of special economic zones to attract foreign investment.

9 The authors warn that Africa may experience some losses through the Uruguay Round erosion of these preferences, although such losses shouldnot be large.

273


tion incentives. Moreover, freer trade regimes areadopted after or in parallel with domestic policyreforms. The country examples also suggest thatopenness per se is unlikely to be a sufficient condi-tion for a successful integration into the globaleconomy. More important seems to be (i) thatfarmers can operate in the appropriate domesticincentive system; (ii) that the incentives arereduced where unproductive excess capacity iscreated and exit policies are in place; and (iii) thatadjustment and reallocation costs are minimized,e.g. through appropriate timing, sequencing andpacing of policy measures.

10.2.2 Freer movement of capital and theemergence of transnational companies

Alongside the expansion of trade flows, anotherfeature of globalization has been the rapid growthin international capital flows. Transnational corpo-rations (TNCs) have been the driving force behindthis rapid development and foreign direct invest-ment (FDI) is the main instrument through whichTNCs expand their reach beyond national bound-aries. Through FDI, TNCs affect production levelsand composition, production technologies, labourmarkets and standards, and eventually also tradeand consumption patterns. Through their controlover resources, access to markets and development

of new technologies, TNCs have the potential tointegrate countries into global markets.

Foreign direct investment: level, flows, and distribution. Between 1989/94 and 2000, annualglobal FDI inflows increased more than sixfold,from US$200 billion to US$1 270 billion (UN,2001c). The growth in FDI exceeded by far thegrowth in trade flows. Between 1991 and 1995 theaverage annual growth rate of FDI was 21 percentcompared with 9 percent for exports of goods andnon-factor services. Between 1996 and 1999, thedifference increased, with FDI growing at anaverage rate of 41 percent and exports growing at 2 percent. In 2000, total sales of foreign affiliatesamounted to US$16 trillion, compared with worldexports of goods and non-factor services of US$7 trillion. Developed countries absorbed themajor part (80 percent) of the FDI inflows but alsoaccounted for a similar proportion of outflows.

As an increasing number of countries integratedinto the global economy, FDI flows also becamemore evenly distributed and reached more countriesin a substantial manner (UN, 2001c). By 2000, morethan 50 countries (24 of which were developing) hadaccumulated an inward FDI stock of more thanUS$10 billion, compared with only 17 countries 15 years earlier (seven of them developing coun-tries). The picture for outward FDI is similar: the

FDI inflows FDI outflows1989/1994 2000 1989/1994 2000

Developed countries 137.1 1 005.2 203.2 1 046.3

EU 76.6 617.3 105.2 772.9

Japan 1.0 15.8 9.0 32.9

United States 42.5 8.2 49 139.3

Other 17.0 363.9 40.0 101.2

Developing countries 59.6 240.2 24.9 99.5

Africa 4 8.2 0.9 0.7

Latin America and the Caribbean 17.5 86.2 3.7 13.4

Asia 37.9 143.8 20.3 85.3

Other 0.2 2.0 0.0 0.1

Central and Eastern Europe 3.4 25.4 0.1 4.0

World 200.1 1 270.8 228.3 1 149.9

Source: UN (2001c).

Table 10.1 Regional distribution of FDI inflows and outflows (billion US$)

274

number of countries with stocks exceeding US$10 billion rose from ten to 33 (now including 12developing countries, compared with eight in 1985)over the same period. In terms of flows, the numberof countries receiving an annual average of morethan US$1 billion rose from 17 (six of which weredeveloping countries) in the mid-1980s to 51 (23 ofwhich were developing countries) at the end of the1990s. In the case of outflows, 33 countries (11 devel-oping countries) invested more than US$1 billion atthe end of the 1990s, compared with 13 countries(only one developing country) in the mid-1980s.

A closer look at the regional distribution,however, reveals that there is still a high concentra-tion of FDI flows within developing Asia (Table10.1). More than half of all FDI went to Asianeconomies, and within Asia, East and South Asiaaccounted for almost the entire inflow. At the otherend of the scale, FDI inflows to Africa have remainedminimal. While doubling in absolute terms, thecontinent’s share in total inflows to developing coun-tries fell by half, between 1989-94 and 2000, from6.8 to 3.4 percent (Table 10.1).

TNCs and FDI in food and agriculture. The basis forthe large TNCs that dominate today’s global foodeconomy was laid with the market concentrationprocess in developed countries. In the UnitedStates, for instance, four meat-packing firms havetraditionally controlled about two-thirds of the beefsupply, and by the mid-1990s over 80 percent of thebeef supply was controlled by four firms (OECD,2001d). High levels of firm concentration also char-acterize the retail food distribution system in otherOECD countries. For example, in Australia, over 75 percent of the retail food distribution system iscontrolled by three firms.

As the domestic markets for their productsbecame increasingly limited, these large foodprocessors extended their operations in two prin-cipal directions. First, they extended their reach“vertically” by taking over the principal operationsalong the food chain. The final result of this processis often a fully vertically integrated company withoperations that cover the entire food chain from the“farmgate to the dinner plate”. Second, theyexpanded horizontally, i.e. they extended their

reach by branching into foreign markets. Thecombined process of horizontal expansion acrosscountries and vertical integration within thecompany created the typical TNC in food and agri-culture. These TNCs are frequently referred to as“food chain complexes” or “food chain clusters”.

The three most advanced food chain clusters areCargill/Monsanto, ConAgra and Novartis/ADM.10

ConAgra, for example, one of the three largest flourmillers in North America, ranks fourth in cornmilling. It produces its own livestock feed and ranksthird in cattle feeding, second in slaughtering, thirdin pork processing and fourth in broiler produc-tion. United AgriProducts is part of ConAgra andsells agrochemicals and biotechnology products(seeds) around the world. The conglomerate alsoowns its own grain trading company (Peavey). Atthe retail level it widely distributed processed foodsthrough such brands as Armour, Swift and Hunt’s,and is second only to Philip Morris as a leading foodprocessor. The Novartis/ADM cluster also connectsthe different stages of food production fromgenes/seeds (Novartis and Land O’lakes) to graincollection (ADM) to processing across the globefrom Mexico, the Netherlands, France, China andthe United Kingdom. Alliances with IBP, the largestUnited States beef packer and second largest porkpacker, extend its influence down the food chain(Heffernan, 1999).

A more recent feature within the process ofvertical integration is that the food chaincomplexes have extended ownership and controlfrom the agricultural downstream sector (foodprocessing and marketing) into strategic parts ofthe upstream system. For instance, it is estimatedthat only three firms control over 80 percent of USmaize exports and 65 percent of US soybeanexports; only four firms control 60 percent ofdomestic grain handling and 25 percent ofcompound feed production (Hendrickson andHeffernan, 2002). While market concentration incertain parts of a country’s food system is a well-established feature in many countries, thesecomplexes have extended their influence acrosscountry borders and have created vertically inte-grated or coordinated production chains across theglobe (OECD, 2001d).

10 All of them are located in, and operate globally from, the United States. They control, through joint ventures and strategic alliances, important partsof the food industry that range from seeds to processed products such as meats, seafood and other foods (OECD, 2001d).

275


Table 10.2 shows the implantation of agrofoodTNC subsidiaries from different home regions ofthe parent company arranged by host region of thesubsidiary, i.e. how much and where TNCs havespread out their activities. It shows that most TNCsin the food industry operate from a westernEuropean or United States home base. Togetherthey account for about 84 percent of all TNCs thathave invested in markets abroad. Those from Asiaare largely found in Asia, although there are signif-icant numbers located in the EU and NorthAmerica, and those from Latin America arepredominantly in other Latin American countries.Europe and North America are both the home andthe hosts to the vast majority of TNC subsidiaries,their stage of development acting both as push andpull forces. TNCs from the EU and the UnitedStates have, to a significant extent, also establishedforeign affiliates in developing countries. In bothcases, Asia and Latin America are the most impor-tant destinations. TNCs from western Europe, forinstance, have nearly as many foreign affiliates inAsia or Latin America as they have in NorthAmerica. By contrast, Africa is home to very fewsubsidiaries, and those it has are almost entirelylocated within other African countries.

TNCs in food and agriculture: help or hurdle forrural development? The general view amongexperts – in developed and developing countriesalike – is that FDI is a powerful catalyst for overalleconomic development. A number of recent publi-cations (World Bank, 2001e and UN, 1999b) havedocumented the benefits that FDI can create fordevelopment. The 1999 issue of the WorldInvestment Report (UN, 1999b) identifies five majoradvantages that are carried along to the hostcountry alongside inflows of FDI: access tocapital,11 access to technology, access to markets,enhanced skills and management techniques andhelp to protect the environment

The UN report stresses that developing coun-tries have vastly benefited from the rapid increasein FDI inflows during the 1990s, particularlythrough added productivity growth. Several othersources underline and quantify the potential thatTNCs have in generating productivity gains (e.g.Sachs and Warner, 1995; Baily and Gersbach,1995).12 Baily and Gersbach (1995) stress that thepotential for productivity gains is particularly largewhere TNCs reinvest profits in the host countries,create forward and backward linkages with localfirms, upgrade the performance of a country’s firms

11 The catalyst for the East Asian financial crisis in 1996 was a huge outflow of funds, as commercial banks and institutional investors called in loans.The resulting losses were equivalent to more than 10 percent of GDP for some countries (based on data in IMF, 1999). By contrast, FDI remainedconstant throughout this period.

12 This is partly because of increased competition, partly a demonstration effect “…when companies based in one country set up operations inanother, they carry with them the production processes and productivity levels of their home country” (Baily and Gersbach, 1995, p. 309).

Host region

Africa Latin America North Asia Eastern and Western Totaland America central Europe Europe

Home region the Caribbean

Africa 58 0 0 1 3 2 64

Latin America 8 45 14 5 0 49 121

North America 52 390 1 295 234 114 818 2 903

Asia 9 37 103 587 1 90 827

Western Europe 84 233 312 268 104 1 948 2 949

Australasia 1 8 5 25 0 46 85

Total 212 713 1 729 1 120 222 2 953 6 949

Source: Agrodata (2000).

Table 10.2 Number of subsidiaries of the 100 largest TNCs by region (1996)

276

through the provision of superior expertise andtechnologies and hence boost growth (Box 10.1).Also to be remembered is that TNCs are the world’schief repository of economically useful skills andknowledge and that technology flows are increas-ingly important components of FDI (UN, 1999b).

Despite the vast potential of FDI for ruraldevelopment, there are a number of reasons tosuggest that simply opening up a country’s borderto FDI may not be the best way to reap the benefits.There are substantial differences in the “quality” ofFDI flows and governments may have to intervenein the process of channelling FDI. Furthermore,the complexity of the FDI package means thatgovernments face trade-offs between differentbenefits and objectives. For instance, they may haveto choose between investments that offer short- asopposed to long-term benefits; the former maylead to static gains, but not necessarily to dynamicones. Moreover, the level of FDI inflows to devel-oping countries can easily be overstated. TNCs canrepatriate much of the profits that they producefrom their investments in developing countries. Insub-Saharan Africa, for instance, an average of 75 percent of the profits have been repatriatedannually between 1991 and 1997 (IMF, 1999). Alsothe costs of attracting FDI through tax revenuesforegone in the host country can be substantial13–costs that need to be counted against the benefitsthese inflows bring.

There are also concerns that TNCs abuse theirmarket power, add downward pressure on rural

wages and disempower farmers through unfaircontractual arrangements. These concerns oftenarise from the notion that the linkages betweenfarmers and TNCs are based on contracts betweenunequal parties, one party consisting of a mass of unorganized small-scale farmers with littlebargaining power and few of the resources neededto raise productivity and compete commercially,and the other party being a powerful agribusiness,offering production and supply contracts which –in exchange for inputs and technical advice – allowit to exploit cheap labour and transfer most risks tothe primary producers. This imbalance in negoti-ating power has been described extensively for theinternational cocoa and coffee markets wheresmallholder farmers are at the starting-point of“buyer-driven supply chains” (e.g. Ponte, 2001; seealso Box 10.2). Such an imbalance in negotiatingpower can affect the distribution of benefits alongthe food chain (Talbot, 1997; see also Ponte, 2001;Gibbon, 2000; and Gereffi, 1994). For example, theshare of income retained by coffee producersdropped from 20 percent in the 1970s to 13 percentin the 1990s and is likely to have dropped furtherwith the dramatic price decline for green coffee in2001/02.

But it should also be noted that there have beenimportant domestic factors that have squeezed theprofit margins for producers. Some developingcountries have creamed off farmers’ profits throughexport taxes, export controls and mandatory sales.For instance, for much of the 1970s and early 1980s

13 For example, in the second half of the 1990s, the governments of Rio Grande do Sul and Bahia in Brazil gave General Motors and Ford financialpackages worth US$3 billion to locate factories in their states (Hanson, 2001).

Box 10.1 TNCs can be the source of major productivity gains

Baily and Gersbach (1995) carried out a comparison of labour productivity in Japan, Germany and the UnitedStates for a number of manufacturing sectors, including food and beer. The United States was most productivein both of these sectors, food productivity in Germany reaching 76 percent of the United States level while inJapan it was only 33 percent. For beer the comparable figures were Germany 44 percent, and Japan 69 percent.They relate relative productivity to a globalization index and find a significant positive relationship – high glob-alization leads to high relative productivity. The globalization index is a complex construct that takes intoaccount the extent of the exposure of a country’s firms in a particular industry to the productivity leader’s firms,through trade, production by the productivity leader’s subsidiaries in the country, or ownership. For food, theglobalization index is very low in Japan, but at a medium level in Germany, and rising.

The authors conclude that the entrance of foreign firms is the most significant impetus to productivityupgrading in an industry and that the indirect effects (on local firms and the supply chain) may be more significant than the direct effects.

277


Box 10.2 The global coffee chain: changing market structures and power

The 1990s saw a number of major changes in the structure of the international coffee market. The mainchanges include a growing market concentration of trading and roasting companies; a growing product differ-entiation in high- and low-quality brands; and a redistribution of the value added along the marketing andprocessing chain.

Growing market concentration. The general deregulation of the international coffee market that followed theend of the International Coffee Agreement (ICA) in 1989 opened the way for a growing consolidation of themarket. This process was particularly pronounced for roasting and trading, where a declining number of compa-nies control an increasing part of the market. In 1998, the two largest coffee traders (Neumann and Volcafé)accounted for 29 percent of the total market, and the top six companies controlled 50 percent (Figure 10.1).Amid growing market concentration, some smaller and specialized companies have emerged, focusing on tradein the speciality coffee market (high-quality and specific origins).

Concentration within the group of coffee traders was promoted by higher international price volatility, whichincreased after the end of the ICA (Gilbert, 1995). Mid-sized traders with unhedged positions suffered consid-erable losses or found themselves too small to compete with larger traders. As a result, they either went bank-rupt, merged with others, or were taken over by the majors. Within the exporting countries, the bureaucracy thatwas needed to monitor exports and ensure compliance with the quota restrictions of the ICA was no longerneeded. Coffee boards and other parastatals that regulated export sales have been dismantled, the capability ofproducing countries to control exports disappeared and their ability to build up stocks decreased. Despite verylow prices, current producer-held stocks are close to their lowest levels in 30 years.

Product differentiation. Despite the overall increase in market concentration there was a product differentia-tion into a system of first-line and second-line supply, subject to price premia and discounts. The differentiationwas created by roasters who declined shipments from countries that could not guarantee a reliable minimumamount of supply. In the case of arabica, this minimum is around 60 000 tonnes a year (Raikes and Gibbon,2000). Minimum supply requirements have created concerns that minor producers may become increasinglymarginalized in the future. In addition, product segmentation will further encourage international traders toengage in major producing markets such as Uganda in order to satisfy their major roaster clients (Ponte, 2001).

Redistribution of the value added. Increased consolidation in the coffee industry has also affected the distri-bution of total income generated along the coffee chain. Talbot (1997) estimates that in the 1970s an average of20 percent of total income was retained by producers, while the average proportion retained in consumingcountries was almost 53 percent. Between 1980/81 and 1988/89, producers still controlled almost 20 percentof total income, while 55 percent was retained in consuming countries. In the 1990s, the situation changeddramatically. Between 1989/90 and 1994/95, the proportion of total income gained by producers dropped to13 percent, and the proportion retained in consuming countries surged to 78 percent. The share of incomeretained by producers in the last three to four years is likely to have dropped further as a result of the currentsituation of oversupply and low prices for green coffee and the ability of roasters to maintain retail prices at relatively stable levels.

Figure 10.1 Market concentration in the coffee chain

Source: van Dijk et al., 1998

Philip Morris25%

Nestlé24%

Sara Lee7%

P&G7%

Tchibo6%

Others31%

Neumann16%

Volcafe13%

Cargill6%

Esteve6%

Aron5%Man

4%Dreyfus

3%

Mitsubishi3%

Others44%

Source: van Dijk et al., 1998

Coffee roasting and processing:Market shares of major TNCs (1998)

Green coffee trading:Market shares of major TNCs (1998)

278

cocoa farmers in Ghana were obliged to sell theircrop to the government for as little as a twentieth ofthe world price (The Economist, 2002a, p. 44).Likewise, profit margins have also been squeezedfor Ecuador’s banana growers, largely in theabsence of TNC activities. Ecuador’s governmenthas currently fixed the local price for bananas atUS$2.90 per box, whereas the export price is ashigh as US$17 per box. The low farmgate price“has squeezed farmers’ profits to almost nothing”(The Economist, 2002b, p. 54). Unlike CentralAmerica, where TNCs own almost all banana plan-tations, Ecuador’s banana economy is dominated bysome 6 000 small family farm producers.

Some experience with successful FDI in food andagriculture. The links created between TNCs anddomestic firms are crucial factors that determinewhether and to what extent a host country benefitsfrom FDI. In the food industry, these linkages areforged between the TNC and the farmers or thelocal procurement company. The potential forlinkage-intensive FDI is particularly substantial infood and agriculture. Linkage-intensive FDI isoften the result of the need to process perishableinputs such as milk or fruit and vegetables (UN,2001c). It can also be forced by logistical bottlenecksor by tariff barriers that make imported goods lesscompetitive. Moreover, TNCs may face restrictionson landownership in many developing countrieswhich can make it necessary for foreign affiliates torely on domestic producers and to engage in effortsto develop new and upgrade existing suppliers.

These linkages from the foreign affiliate to thenational farm sector can provide enormous bene-fits for farmers and their cooperatives and thushave considerable potential to stimulate ruraldevelopment. Field research conducted in India(UN, 2001c) provides a number of interestinginsights as to how these benefits are generated. Itreveals that the four leading TNCs (i.e. Pepsi FoodsLtd, GlaxoSmithKline Beecham Ltd, Nestlé IndiaLtd and Cadbury India Ltd) on average sourcedlocally 93 percent of their raw material (tomatoes,potatoes, basmati rice, groundnuts, cocoa, freshmilk, sugar, wheat flour, etc.) and 74 percent ofother inputs (such as plastic crates, glass bottles,refrigerators, ice chests, corrugated boxes, craftpaper, etc.). Through these linkages the TNCspromoted overall development by means of the

following methods. ■ Collaboration in product development. All four

TNCs are engaged in product developmentwith local research institutes or universities, todevelop hybrid varieties of crops and vegetablesand new agricultural implements, to alter crop-ping patterns and to raise productivity. Forexample, Pepsi Foods has evaluated more than215 varieties/hybrids of chili, probably thelargest scientific evaluation of chilies anywhere.Pepsi’s technology in chili cultivation has raisedits yield three times. In addition, Pepsi hasdeveloped 15 new agricultural implements tofacilitate planting and harvesting in India.

■ Technology transfer and training. New hybrid vari-eties, implements and practices have been trans-ferred to suppliers (primarily farmers) throughfarmer training camps. Pepsi provides itscontract farmers, free of charge, with variousagricultural implements and hybrid seeds/plantlets, as well as process expertise. CadburyIndia has a procurement and extension servicesteam that imparts training to potential andexisting suppliers in new techniques in planting,harvesting, quality control and post-transplanta-tion care of cocoa crops through technicalbulletins, video demonstrations, slides andcharts and live demonstrations on the use ofvarious agricultural implements.

■ Introduction of contract farming. Farmers arecontracted to plant the processors’ crops ontheir lands and to deliver to the processors, atpre-agreed prices and quantities of outputbased upon anticipated yields and contractedarea. Towards this end, a processor usuallyprovides the farmers with selected inputs suchas seeds/seedlings, information on agriculturalpractices, regular inspection of the crop andadvisory services on crops. Farmers have thechoice to leave some part of the output freefrom the contract arrangement to sell on theopen market (see also Box 10.3).

■ Financial assistance is provided to growers throughthe involvement of agricultural developmentbanks. For example, GlaxoSmithKlineBeechamacts as a guarantor, enabling its suppliers to takebank loans.

Technology transfer to local farmers has had apositive impact on farm productivity. Tomato

279


yields of local suppliers for Pepsi in Punjab, forexample, rose from 16 tonnes/ha in 1989 to 52 tonnes/ha by 1999. In general, foreign affiliatesmay have contributed to better farming practices(e.g. hybrid seeds, transportation innovation)resulting in increased incomes and yields(McKinsey &Company, 1997).

10.2.3 International flows of knowledgeand technological innovations

Probably the single most important transfer ofexternal technology to developing countries’ agri-culture took place during the green revolution.The literature documents where the new technolo-gies have been adopted, to what extent, and how

Box 10.3 Formalizing the linkages in agriculture: the importance of contract farming

An extensive study by FAO (FAO, 2001i) brings together numerous examples from many developing countriesthat confirm the generally positive influence of TNCs on the agriculture of these countries. But the FAO studyalso shows that policies play an important role in promoting the benefits that TNCs or the local processors canprovide for a country’s agriculture. Most important, it shows that the underlying contracts between farmers andthe company are crucial for success or failure. Numerous examples demonstrate how well-managed contractfarming works as an effective tool to link the small-farm sector to sources of extension, mechanization, seeds,fertilizer and credit, and to guaranteed and profitable markets for produce. When efficiently organized andmanaged, contract farming reduces risk and uncertainty for both parties. The principal benefits laid out in thestudy are the following.

Increased productivity. In northern India, Hindustan Lever, a food processor, issued contracts to 400 farmersto grow hybrid tomatoes for processing. A study of the project confirmed that production yields and farmers’incomes increased as a result of the use of hybrid seeds and the availability of an assured market. An analysisof the yields and incomes of the contracted farmers compared with farmers who grew tomatoes for the openmarket showed that yields of the farmers under contract were 64 percent higher than those outside the project.In Sri Lanka, a flourishing export trade in gherkins has been built on contracts between companies and morethan 15 000 growers with plots of around 0.5 ha each. On a much larger scale, more than 200 000 farmers inThailand grow sugar cane for the country's 46 mills under a government-sponsored system that assigns growers70 percent and millers 30 percent of total net revenue (FAO, 2001i).

Introduction of superior technologies. Small-scale farmers are frequently reluctant to adopt new technologiesbecause of the possible risks and costs involved. In contract farming, private agribusiness will usually offer tech-nology more effectively than government agricultural extension services, because it has a direct economicinterest in improving farmers' production. Indeed, most of the larger corporations prefer to provide their ownextension. In Kenya, for example, the South Nyanza Sugar Company (SONY) places strong emphasis on fieldextension services to its 1 800 contracted farmers, at a ratio of one field officer to 65 sugar-cane growers. Theextension staff’s prime responsibilities are focused on the managerial skills required when new techniques areintroduced to SONY’s farmers. These include transplanting, spacing, fertilizer application, cultivation andharvesting practices. Also, SONY promotes farmer training programmes and organizes field days to demonstratethe latest sugar-cane production methods to farmers.

Risks and problems. In addition, the FAO study emphasizes that contract farming can be a major tool for trans-ferring skills and providing access to credit – features that are particularly important for smallholders. But thestudy also underlines that certain risks and problems can be associated with contract farming. Considerableproblems can result if farmers perceive that the company is unwilling to share any of the risk, even if it is partlyresponsible for the losses. In Thailand, a company that contracted farmers to rear chickens charged a levy onfarmers' incomes in order to offset the possibility of a high chicken mortality rate. This was much resented bythe farmers, as they believed that the poor quality of the chicks supplied by the company was one cause of theproblem. Inefficient management can lead to overproduction, and in some cases processors may be tempted tomanipulate quality standards in order to reduce purchases. One of the biggest risks for farmers is debt causedby production problems, poor technical advice, significant changes in market conditions, or a processor’sfailure to honour contracts.

280

swiftly this has been the case. This section will focuson why some countries managed to adopt, exploitand further enhance the new technologies, whileothers failed to do so. It will try to identify the poli-cies that allowed some countries to embrace globaltechnological innovations, but also describe whyother countries are still struggling to adopt thetechnologies that developed country farmers havebeen using for many decades.

One of the most influential studies in thiscontext (Griliches, 1957) underlines the importanceof inventive adaptation. Griliches shows thatfarmers in Iowa and Illinois had long adopted high-yielding hybrid maize varieties suited to the CornBelt states, while farmers outside the Corn Belt (e.g.in Alabama) continued to grow inferior traditionalvarieties. This had little to do with the farmers’capabilities. Instead, differences in the agro-ecolog-ical conditions between the Corn Belt and Alabama,together with the sensitivity of hybrid maize to thesedifferences, resulted in the lack of adoption and thecontinuous technological distance between thesemaize-growing areas. As Griliches noted, “farmersoutside the Corn Belt could not reap the benefits ofthe new technologies until the adaptive researchhad taken place to make the technologies availableto the new environment”.

In general, the same holds for the transfer ofnew technologies to developing countries. Farmersin the Philippines got no direct benefit from manydecades of United States hybrid maize researchthat produced a tripling of United States maizeyields. They indirectly benefited from previoushybrid research in the United States only after theresearch capacity was created to adapt the hybridvarieties to local conditions in the Philippines. Inmany African countries, farmers are still cut offfrom the benefits of hybrid maize varieties, notbecause they are unwilling to import the tech-nology but simply because the technology has notbeen adapted to their local growing environ-ments.

Perhaps even more important, much of thesuccess of the green revolution was not or notprimarily based on the fact that new technologieswere made available to countries from outside.

While the new “foreign” technologies, i.e. high-yielding varieties, played an important role, thereis ample evidence that the superiority of these newvarieties was largely limited to the areas to whichthese new technologies were adapted. Evensonand Westphal (1994) documented how importantthe adaptation process to tropical environmentswas for the success of high-yielding rice varietiesoutside their subtropical homes.14 “It was in the1950s, after an Indica-Japonica crossing programmesponsored by the FAO and IRRI gave majorimpetus to rice improvement for tropical conditions, that the new technologies becameavailable where they were needed. By 1965, manynational rice breeding programs had been estab-lished in tropical conditions. India, for examplehad 23 programs in various locations. Around 200rice breeding programs existed in some 40 coun-tries by 1970. Most had, and have maintained, aclose association with IRRI, which has served as anodal point in the transfer of new germplasm.”

Evenson and Gollin (1994) quantified theimportance of adaptive research in the spread ofhigh-yielding rice varieties during the green revo-lution. They show that national research centresplayed a crucial role in the adoption and spread of the new “technology”. The International RiceResearch Institute (IRRI), as the central andexogenous provider of the new technology,accounted for only 17 percent of all new varietalreleases since 1965. IRRI played, however, acrucial role in generating the basic technology: itaccounted for 65 percent of all new releases ofparental varieties.

In the future, the very same factors will likelydetermine the extent to which the new agrobiotech-nologies will be adapted and diffused to the economicand agro-ecological environments of developingcountries, even though the issue is complicated by thefact that many of the new technologies are propri-etary ones. Countries that put in place the basic infra-structure that promotes the adaptation to localenvironments are likely to gain the most. Again, Asiancountries are likely to come first, followed by LatinAmerica, while there is a danger that African coun-tries will, once again, be left behind.

14 Evenson and Westphal also provide an extensive documentation of the events that kick-started the spread of high-yielding rice varieties in the1950s. Inter alia, they explain that: “The earliest rice improvement research activities were in Japan, where major gains were made in this centurythrough improving Japonica landraces suited to subtropical regions. It was not until World War II that concerted efforts were made to improve theIndica landraces. As of that time, rice producers in Japan, Korea, Taiwan and parts of mainland China had achieved a 50-year technological leadover the tropical rice producing areas”.

281


The factors that determine the success inreaping the benefits of new technologies havemuch in common with those that enable countriesto reap the benefits of open markets. The experi-ence of the green revolution suggests that the exis-tence of new productivity- enhancing technologiesalone is not a guarantor for a successful adaptationof these technologies. Likewise, opening up tointernational markets and reduction of bordermeasures will not, on their own, ensure that thepotential of freer trade can be fully exploited. Bothopenness to trade and to technological change areimportant, but what seem to be more important arethe policies and institutions that allow countries toexploit the opportunities offered by openness.These factors can help to acquire the often tacitknowledge that enables countries to adapt newtechnologies to the domestic market environment,help them to exploit the demand potential of largeinternational markets and employ trading rules totheir advantage. Taken alone, access to markets isunlikely to create an exportable surplus. If notlocally adapted, new technologies will not substan-tially increase productivity.

10.2.4 Greater mobility of people withinand across national borders: migration and urbanization

International migration. Massive movements oflabour have been a feature of all three waves ofglobalization. During the first great wave ofmodern globalization, from 1870 to 1910, about 10 percent of the world’s population relocatedpermanently (World Bank, 2001e). Internationalmigration was much more modest and geographi-cally limited during the second wave. The mainreason was that only a limited number of countrieswere involved in the second wave of globalization,and where intense migration pressures occurred,strict immigration controls helped to put a brakeon labour flows. These controls were somewhatrelaxed during the third wave of globalization andhad a powerful effect on transnational migration.By 1995, about 150 million people or 2.3 percentof the world’s population lived in foreign countries(Taylor, 2000). Roughly half of this stock ofmigrants was in the industrial countries and half inthe developing world. However, because the popu-lation of developing countries is about five times

greater than the population of the developed coun-tries, migrants account for a much larger share of the population in developed countries (about 6 percent) than in poor countries (about 1 percent).

The freer movement of people has always hadpowerful effects on wages in poor and rich coun-tries alike. Initially, different speeds of growthwithin and across countries promote inequality inwages and wealth, which in turn creates theeconomic pressure to migrate. Then migrationitself, in addition to increased trade and capitalflows, helps arrest or even reverse a growing wageinequality. The influx of low-wage labour putsdownward pressure on wages in immigrantregions, while raising wages in the emigrantnations. Moreover, wealth is also redistributedwhen and to the extent that emigrants send backremittances to their countries of origin. As alreadymentioned in Section 10.1, Lindert and Williamson(2001) conclude that migration was overall a moreimportant equalizing factor than either trade orcapital movements.

Globalization also affects international migra-tion in agriculture. During the first wave of globalization, migration was almost exclusivelydetermined by different speeds of agriculturaldevelopment. But even today many developedcountries turn to foreign-born migrants as animportant source of agricultural labour. Most ruralmigrants are attracted by higher wages in devel-oped countries’ fruit, vegetable and horticulturalsectors. In the United States, for example, an esti-mated 69 percent of the 1996 seasonal agriculturalworkforce was foreign-born (Mines, Gabbard andSteirman, 1997). In California, by far the nation'slargest agricultural producer, more than 90 percentof the seasonal agricultural workforce was foreign(Taylor, 2000). The majority (65 percent) of UnitedStates migrant farm workers originated fromhouseholds in rural Mexico. Despite the highconcentration of foreign-born workers in farm jobs,the vast majority of immigrants are employedoutside agriculture, most in low-skill service andmanufacturing jobs.

Agricultural migration is primarily a movementof low-skill labour from developing countries to thehigher wage environments of developed countries.As such, it is unlikely to be associated with many ofthe typical concerns that emigration would lead toa “brain drain” in developing countries and

282

deprive them of their most important capital forfuture development. In fact, empirical studiesshow that migrants seldom sever their ties withtheir source households after they migrate andfamily members who remain behind (oftenparents and siblings) reorganize both theirconsumption and production activities inresponse to the migrant’s departure. Migrants(often children) typically share part of their earn-ings with their household of origin through remit-tances (Taylor, 2000). Remittances or savingsaccumulated abroad can even create the basis forfuture investments in the rural economies of theirhome countries.

The impacts in the countries of immigrationare often more ambiguous. While an inflow ofunskilled workers from developing countriesbenefits the highly skilled workers in host coun-tries (their jobs are not threatened by these immi-grants, and the presence of immigrants will lowerprices for many things that the skilled workersconsume, including food, restaurant and hotelservices, and personal services), the same inflowwill reduce the real wages of unskilled workers.Such competition in the low-wage sector hasbrought about political tensions within many hostcountries and has often resulted in increasinglyrestrictive immigration rules.

As immigration rules tightened and the economicincentives to immigrate remained unabated, illegalimmigration and trafficking in human beingsincreased rapidly. The World Bank estimates thatthere is an annual inflow of about 300 000 illegalworkers to the United States alone (World Bank,2001e). Many more cross temporarily into theUnited States. In 1999 United States authoritiesapprehended 1.5 million illegal immigrants alongthe Mexican border. The great majority sent back toMexico attempt to cross again within 24 hours.Illegal migration into the EU soared in the 1990s,from an estimated 50 000 p.a. in 1993 to half amillion in 1999 (World Bank, 2001e).

Intranational migration and urbanization. Asdiscussed in the preceding section, internationalmigration has affected rural populations and agri-cultural labour forces in developed and developingcountries alike. However, despite its importance for certain regions or countries (notably NorthAmerica), international migration has become

negligible compared with migration within nationalboundaries.

A look at the United Nations population projec-tions by urban and rural areas reveals that a signif-icant proportion of the world's population growthexpected between 2000 and 2030 will be concen-trated in urban areas. Urban population was esti-mated at about 2.9 billion in 2000, and is projectedto be about 4.9 billion by 2030 (Figure 10.2). Most of the future increase will be in the cities ofdeveloping countries. The urban population indeveloping countries is projected to increase from 1.9 billion people in 2000 to about 3.9 billion peopleby 2030, thus accounting for almost the entireincrement in developing countries’ populationgrowth. But only a part of it is caused by increasedrural-urban migration. Also important will be thetransformation of rural settlements into urban areasand, most important, natural urban populationgrowth.

These shifts in population distribution areconsiderable. At the beginning of the 1960s, onlyabout 20 percent of the developing countries’population lived in urban areas. By 2000 theshare had risen to nearly 40 percent and isexpected to rise to 56 percent by 2030. The rural-urban population ratio declined from about 3:1 in the 1960s to almost 3:2 in 2000 and will be closeto 3:4 in 2030. Within the group of developingregions, urbanization will be most pronounced indeveloping Asia and sub-Saharan Africa. In otherdeveloping regions, notably Latin America,urbanization has already progressed to an extentthat leaves little room for further growth in urbanpopulations, at least relative to rural ones.

Numerous factors have promoted and willcontinue to promote urbanization in developingcountries. These factors are often classified aspush and pull factors, i.e. factors that either incitepeople to leave their rural homes or attract themto urban areas. Typical push factors include lowand declining profitability of agricultural produc-tion, lack of non-agricultural employment oppor-tunities and a general lack of services such asschools, medical treatment and entertainment.They have resulted from a general neglect of, oreven an outright bias against, agriculture.

On the pull side, expectations for better serv-ices, housing, higher wages and more reliablesources of food are the main factors that attract

283


migrants to urban areas. The typical drivers ofglobalization, notably better information facilities(television, etc.) have been instrumental increating these expectations. There is, however, alarge and widening gap between the expectationsand the realities of urban areas in developingcountries. Access to food, jobs and services isbecoming more limited and other amenities thatare often associated with “urbane” or “civilized”city life are entirely missing. This so-called prema-ture urbanization is associated with numerousexternalities. These include huge social costscaused by health and sanitation problems, urbanpoverty, crime, etc.

Despite these problems, it is widely acceptedthat urbanization is unstoppable, let alone reversible(The Economist, 2002c). Moreover, while it maynot even be economically desirable to stop orreverse urbanization, it can be very profitable toslow the trend and mitigate or avoid the externali-ties associated with premature urbanization. Themost important factor is a revival of rural areas indeveloping countries, which would amount to areversal of the internal and external policy biasagainst agriculture in developing countries (asdiscussed in Chapter 8).

10.2.5 How important are geographiclocation and infrastructure?

Geographic location. Economists have long notedthe crucial role of geographic location for economicdevelopment. Adam Smith, who is most remem-bered for his prescription of free market forces foreconomic development, emphasized that the phys-ical geography of a region can influence cruciallyits economic performance (Smith, 1976). Hecontended that the economies of coastal regions,with their easy access to sea trade, usually outper-form the economies of inland areas. Smith'srationale for the importance of geographic locationis that productivity gains depend on specialization,and that specialization depends on the size of themarket. The size of the market in turn depends onboth the openness of markets and the costs oftransport. Geography is a crucial factor in deter-mining transport costs.

Empirical studies based on geographic informa-tion systems (GIS) have drawn renewed attentionto the importance of the physical location foreconomic development. A frequent point of depar-ture for the GIS-based analyses of location-baseddevelopment questions is a map of income density,

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Bill

ion

peop

le

Rural Urban projected

projected

0

1

2

3

4

5

6

7

8

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Rur

al/u

rban

Latin Amererica & Caribbean

Near East Developed Countries

Sub-Saharan Africa Developing Asia

projected

Urbanization in developing countries will accelerate over the next 30 years

Rural-urban population ratios

Figure 10.2 Past and projected trends in urbanization of developing countries

284

a measure of how much GDP is produced within agiven area of land.15

Figure 10.3 provides such a map. The mapunderlines two principal geographic factors thataffect economic well-being. First, almost all high-income countries are in the mid- and high lati-tudes, while nearly all countries in the geographictropics are poor. Second, coastal economies havegenerally higher incomes than landlocked coun-tries and coastal, temperate, northern hemisphereeconomies have the highest economic densities inthe world. Indeed, outside Europe, there is not asingle high-income landlocked country, althoughthere are 29 non-European landlocked countries(Gallup, Sachs and Mellinger, 1999). Four areas –western Europe, Northeast Asia (coastal China,Japan and the Republic of Korea), and the easternand western seaboards of the United States andCanada – are the core economic zones of themodern world. These regions are the over-whelming providers of capital goods in globaltrade, the world’s financial centres, and the gener-ators of a large proportion of global production.

A look at the regions within the United States,western Europe and temperate-zone East Asia thatlie within 100 km of the coastline reveals that theseareas account for a mere 3 percent of the world’sinhabited land area, 13 percent of the world’spopulation and at least 32 percent of the world’sGDP measured at purchasing power parity. Ifcoastal China is excluded from the calculations(since it lags far behind the other economies in thisgroup), then the core coastal region has a mere 9 percent of the world’s population but produces atleast 30 percent of world GDP. According to WTOdata (1995), just 11 countries in North America,western Europe and East Asia, with 14 percent ofthe world’s population, account for 88 percent ofglobal exports of capital goods (machinery andtransport equipment).

Moreover, nearly all landlocked countries inthe world are poor, except for a few in westernand central Europe that are deeply integratedinto the regional European market and connectedby low-cost trade. Even mountainous Switzerland

has the vast bulk of its population in the low-eleva-tion cantons north of the Alps, and these popula-tion centres are easily accessible to the NorthAtlantic by land and river-based traffic. There are35 landlocked countries in the world with a popu-lation greater than 1 million, of which 29 areoutside western and central Europe. The differ-ence in the average GDP per capita is striking: thelandlocked countries outside western and centralEurope have an average income of aboutUS$1 771, compared with the non-Europeancoastal countries, which have an average incomeof US$5 567. The difference in economic densityis even greater, since the landlocked countriestend to be very sparsely populated16 (Gallup,Sachs and Mellinger, 1999).

The most important points that arise from theinspection of GIS-based information can besummarized as follows:■ Coastal regions, and regions linked to coasts

by ocean-navigable waterways, are stronglyfavoured in development relative to the hinter-lands.

■ Landlocked economies may be particularlydisadvantaged by their lack of access to the sea,even when they are no farther than the interiorparts of coastal economies.17

■ Location advantages are particularly importantfor successful economic integration of agricul-ture and the food industry. Many agriculturalcommodities are either bulky, perishable orboth, which leads to high transportation costsper unit value of product. High transportationcosts mean that countries with poor marketaccess conditions and inadequate infrastructuremight remain effectively insulated, even if alltrade barriers were removed.

Endowment with, and importance of, infrastruc-ture. Infrastructure can offset much of the disad-vantage that may arise from an unfavourablegeographic location. In fact, in many developedcountries and regions, access to infrastructureoffsets possible disadvantages caused by unfavourablelocations. The extensive and efficient transporta-

15 Ideally, such an income density map depicts information about population density and the spatial distribution of income. In the absence of thelatter, Figure 10.3 simply combines population density with average per capita income levels (1997/99) for a given country.

16 Fifty-nine people per km2 in landlocked countries compared with 207 people per km2 in coastal countries.17 Obvious reasons for this are: cross-border migration of labour is more difficult than internal migration; infrastructure development across national

borders is much more difficult to arrange than similar investments within a country; and coastal economies may have military or economic incen-tives to impose costs on interior landlocked economies.

285


tion and communication systems in landlockedparts of Europe or North America effectively linkthese regions to one another and integrate theminto world markets. A look across other regions ofthe world provides a more mixed picture. Whileparts of East Asia and Latin America hold relativelyhigh stocks of infrastructure, Africa and manycountries in sub-Saharan Africa in particular sufferfrom both unfavourable location and a lack ofinfrastructure.

The case of sub-Saharan Africa. Numerousstudies (e.g. Finger and Yeats, 1976; Amjadi, Reinkeand Yeats, 1996) have analysed and quantified theimportance of infrastructure as a factor for thesuccessful integration into international markets.Amjadi, Reinke and Yeats focus on sub-SaharanAfrica and how the region’s inadequate endow-ment with infrastructure weighs on its exportperformance. The study also documents the importance of insufficient infrastructure and relatedpolicies relative to other factors such as tariffs andnon-tariff barriers.

The importance of the region’s infrastructurerelative to its competitors is also underlined by acomparison of the barriers that sub-Saharan African

exporters face in markets abroad, relative to those itscompetitors faced when they embarked on export-oriented policies. For example, pre-Uruguay Roundtariffs facing African exports to the EU, Japan andthe United States averaged three-quarters of apercent (about 18 points lower than those the Asiannewly industrializing countries [NICs] faced whenthey began their sustained export-oriented industri-alization drive), and preferences give Africa an edgeover some competitors (Amjadi, Reinke and Yeats,1996). As long as transport is expensive, electricityunavailable or unreliable and access to phonesrestricted, the costs and risks of doing businessremain high and the possibilities of reaping thebenefits of globalization remain limited. But global-ization also offers new opportunities to leapfrogtraditional constraints. These new options will bediscussed in Section 10.3 of this chapter.

10.2.6 What are the impacts on foodconsumption patterns?

The effects of freer trade in agriculture, the oper-ations of TNCs in the global food sector as well asurbanization and migration become visible in

Figure 10.3 GDP density map of the world

286

changes in food consumption patterns. In general,these factors promote a convergence of foodconsumption patterns across different countriesand regions. The channels through which thesefactors operate are either direct or indirect via thelink of income growth in poorer countries. Risingincomes have an equalizing effect on foodconsumption patterns as they promote a shift ofpoorer consumers to higher-value food items,while higher-income segments are constrained bynatural consumption limits even for higher-valuefood items.

A number of concerns revolve around thegrowing convergence in food consumption patterns.Some analysts see convergence as an indicator of aloss of cultural identity that reflects in part thegrowing market power of transnationally operatingfood enterprises (“McDonaldization”). Moreover,there are concerns that a fast convergence in foodconsumption patterns may have unexpectedresource implications. A growing global conver-gence on, say, a typical United States diet is associ-ated with rapid growth in feedgrain needs and thuswith an extra burden on the available agriculturalresource base. Another concern associated with aconvergence in food consumption patterns is thatfood would have to travel over ever longer distancesand that the externalities associated with these extra“food kilometres” are not, or not fully, reflected inthe price of food. This section examines to whatextent food consumption patterns have alreadyconverged and what the projections to 2015 and2030 imply for future convergence.

Measuring convergence in consumption patterns.The comparison of food consumption patternswas undertaken on the basis of 29 primaryproduct groups.18 The need to compare diets ofsome 150 countries over a period of 70 years(1961 to 2030) strongly favoured the use of asingle indicator, namely the consumption simi-larity index (CSI). This index measures theoverlap in the diets of two countries by comparinghow many calories consumed in two given coun-tries originate in the same primary products. TheCSI is expressed as:

A CSIj,k of 1 means that the diets of country jand country k are identical, i.e. that consumersdraw the same number of calories from each of the29 food categories distinguished in this study,while a CSIj,k of 0 means that the diets of the twocountries are entirely different, i.e. that consumersin country j and k draw their calorie consumptionfrom completely different food categories. In prin-ciple, the CSI allows the food consumption patternof any country j to be compared with the one ofany other country k.19 It is important to stress thatthe CSI only captures the similarities in the struc-tures of the diet in terms of primary products asdefined in this study, but does not necessarilycapture similarities in the final processed productsthat are actually consumed. This means that itmeasures to what extent consumers in two coun-tries rely on a wheat-based or meat-based diet but not whether the wheat-based calories areconsumed in the form of noodles or bread, or thatmeat is consumed in the form of hamburgers ortraditional meat products. It is also important tonote that the CSI measures similarities in dietstructures, regardless of the absolute calorie intakelevels. This can result in surprisingly high similar-ities in diets that are indeed very similar as far as the overall structures are concerned but verydifferent regarding their respective levels ofcalorie intake (e.g. high shares of meat consump-tion in pastoral societies with low overall calorieconsumption generate a high similarity with meat-intensive OECD diets).

The CSI has been used to compare the foodconsumption patterns across countries and overyears. While CSI calculations have been under-taken for all combinations of countries, the resultsare only reported for the United States as a“comparator” country. All CSI coefficients arebased on a comparison of any given country’s dietwith the one of the United States. Convergenceover time is convergence towards the United Statesconsumption patterns. The convergence in food

18 The 28 food commodities given in Annex 1 of this study and a commodity, “other calories”, i.e. all calories consumed and not covered by the 28 commodities of this study.

19 Note that a CSI of 0.5 between country 1 and country 2 and a CSI of 0.5 between country 1 and country 3 do not mean that the overlap/simi-larity in the diets between country 2 and 3 has to be 0.5 as well.

∑ −−=i k

CALik

CAL

jCAL

ijCAL

kjCSI

2

11

,

287


consumption reported here may thus be regardedan “Americanization” rather than a globalization offood consumption patterns.

Evidence for convergence. A look at CSI develop-ments suggests that diets have indeed becomeincreasingly similar over time. The speed ofconvergence, however, differs markedly acrosscountries. The traditional OECD countries form acluster with consumption patterns that are veryclose to the United States diet. About 75 percent ofthe calories in many OECD countries originatefrom the same sources as in the United States.These countries are fully integrated into the globalfood economy and their food economies are tied toone another through effective and efficient trans-portation and communication infrastructures,similar food distribution systems, cold chains, etc.One of the striking features in the group of OECDcountries is the high similarity in consumptionpatterns within the cluster of English-speakingcountries (Australia, New Zealand and the UnitedKingdom) where 80 percent and more of all calo-ries stem from the same foodstuffs.

Many of these countries share not only the samelanguage, but have also a common food andcooking culture. The absence of language barriersand the similarity in food culture are importantparameters for an effective and low-cost operation

of transnational food enterprises. These similaritiesallow them to employ the same or similar adver-tising and marketing strategies and thus reapeconomies of scale in their market penetrationstrategies. Finally, there is also a geographicelement that plays a role in explaining similaritiesin diets. For instance, 85 percent of all caloriesconsumed in Canada stem from the same primarycommodities as in the United States. Similarly highvalues exist when diets of North African or westernEuropean countries are compared among eachother (not with the United States).

Outside the group of the well-integrated westerncountries, the similarity with the United Statesconsumption pattern is often considerably smaller.Again, there are a number of different groups ofcountries that exhibit different levels of overlap anddifferent dynamics in moving towards United Statesfood consumption patterns. Very dynamic changecan be observed within the group of East andSoutheast Asian countries. Japan's consumers areamong the most dynamic adapters of a UnitedStates-type food consumption pattern. Starting froman overlap of only 45 percent in 1961, similarity hadincreased to about 70 percent in 1999 and isexpected to reach a level of 75 percent in 2030(Figure 10.4).

Outside the group of OECD countries, a numberof different clusters can be identified across conti-

0.4

0.5

0.6

0.7

0.8

0.0

1

1961

1965

1975

1985

1995

2005

2015

Con

sum

ptio

n si

mila

rity

inde

x

Australia New Zealand Canada

United Kingdom Japan

Note: The dip for Japan in 1975 reflects a change in diet caused by the drastic price increases in world markets during the world food crisis in the mid-1970s.

Figure 10.4 Food consumption convergence in OECD countries

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1961

1965

1970

1975

1980

1985

1990

1995

2000

2015

2030

Algeria Niger Ghana Kenya

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1961

1965

1970

1975

1980

1985

1990

1995

2000

2015

2030

Korea Rep. Lao Indonesia Thailand

Large differences in the speed and extent of convergence within Africa

Globalisation determines the degree of consumption convergence in Asia

Con

sum

ptio

n si

mila

rity

inde

x

Con

sum

ptio

n si

mila

rity

inde

x

288

nents and regions. Within Africa, three major trendsin consumption patterns emerge. First there is thegroup of North African countries, where consump-tion patterns are characterized by a grain-rich dietand where often more than 70 percent of the calorieintake stems from cereals, notably from wheat.Within this group, food consumption similarityreaches levels of more than 90 percent. Compared toa United States diet, however, similarity has reacheda level of about 60 percent (Figure 10.5) and is, evenby 2030, not expected to exceed about 65 percent.This could seem surprising at first sight, given thegeographic vicinity to, and increasingly importanteconomic integration with OECD markets. However,other factors override the integrating forces of glob-alization/Americanization on food consumptionpatterns. These are rooted in (i) the traditional foodculture characterized by high consumption levels ofwheat-based staples (bread, couscous); and (ii) thenon-consumption of pork that limits the potential forshifts towards meat consumption.

In summary, the forces of globalization have hada significant impact on food consumption patternsand have resulted in a growing convergence ofconsumption patterns. Even though the relativeimportance of the various driving forces of globaliza-tion is difficult to gauge, openness to trade andinvestments, geographic location, income levels andgrowth and TNC activity are almost always associated

with a rapid convergence in food consumptionpatterns. Many of these factors are interrelated. Well-integrated countries also often enjoy higher incomegrowth that works, within the boundaries of incomeresponsiveness and overall calorie intake levels, as aforce for convergence. There are, however, factors ofa longer-term nature that put a cap on the conver-gence of food consumption patterns. They includecultural and religious constraints as well as deeplyrooted traditions in food consumption and prepara-tion. As a result they limit convergence, even in themost integrated OECD markets, to a level of about 80percent, a level that is not expected to be topped overthe next 30 years. Outside the OECD area, conver-gence levels off at about 60 percent overlap.

10.3 Some options to integrate developing countries better

Multimodal transportation systems. The smartcombination of various transportation modes(sea, air, rail and road) can help to overcome thefinancing constraints that many developing coun-tries face in building up traditional infrastructurecomponents. So-called multimodal transporta-tion systems have gained a considerablemomentum in integrating hitherto remote areasof Asia and Latin America (Box 10.4).

Figure 10.5 Food consumption convergence in Africa and Asia

289


The new transportation systems can be particularly efficient if combined with newcommunication tools. These new technologiesoffer considerable potential to overcome the loca-tion handicap that many remote areas in thedeveloping world face; they could provide newtrade opportunities for bulky or perishable goodsthat were previously excluded because of prohib-itively high transaction costs. This in turn couldhelp integrate agricultural producers andprovide a stimulus to rural areas, perhaps compa-rable to the rapid expansion of agriculturalproduction in the United States Midwest duringthe first wave of globalization.

Leapfrog traditional communication constraints: theInternet for trade facilitation. As during the firstwave of globalization, the availability of more effi-cient transportation systems has been accompaniedby the advent of more efficient communicationsystems. These new communication technologiesenabled shippers to tailor volumes and deliverydates of goods to the precise needs of importers.The Internet now allows even smaller-sized compa-nies to compete with their larger counterparts, whohad gained a competitive advantage through thededicated but more expensive electronic data

interchange (EDI) systems. For the low initial costof a personal computer, a modem and an Internetconnection, anyone can now access the Internet.More and more shippers and carriers choose to dobusiness through the Internet because of the loweradministrative costs involved in conducting trans-actions. This means significant savings becausecarriers and shippers depend less on third-partyvalue-added networks that are normally requiredto run EDI transactions.

There is also a growing expectation that theInternet will soon provide all the advantages thathad previously been restricted to expensive EDIsystems. Most important, Internet data transfer willbecome increasingly safe. Moreover, it is available24 hours a day, allowing business deals to be madeat the shipper’s and carrier’s convenience. It couldprovide niche markets for smaller carriers, enablingthem to capture a greater market for small packagedeliveries.

Economic agglomeration and special economic zones.Despite the possibilities of exploiting further effi-ciencies in the existing infrastructure, investments ina uniform expansion of infrastructure in every loca-tion may be neither an efficient nor an affordableoption for most developing countries, particularly in

Box 10.4 Multimodal transport offers new opportunities for developing countries

The advent of multimodal transportation systems adds to numerous new opportunities for developing countriesto integrate faster and more easily into the global economy. In China, for instance, a new service has been estab-lished connecting the country’s hitherto isolated hinterland with Europe. Multimodal Logistics, a Rotterdam-based company, is now offering rail transport for containers from Rotterdam to northwest China. The Marco PoloRail Express, as the service is commonly called, has two connecting points at Almaty and Druzhba. The transittime is between two and three weeks depending on its destination in China, which could be Alataw-Shankou,Jinghe, Wusu, Ürümchi, Turpan, Korla or Hami/Yumen (UNCTAD, 1999a).

There is also an effort to connect Singapore and Europe by rail. So far, India, the Islamic Republic of Iran,Pakistan and Turkey have been active in the project. Each participating country must bear the cost of completingthe link within its national boundary. The aim is to reduce the transit time for deliveries from Singapore toEurope and vice versa by two weeks. There is also a proposal to construct a tunnel that will link Taiwan Provinceof China and mainland China in order to accommodate the increase in goods traded and transported betweenthe two countries. The Taiwanese private sector has already responded by investing in new depots in Shanghaiand Shenzen in China.

There are also plans to create a land bridge that would connect Latin America’s Atlantic side with its Pacificcoast. These plans have grown out of the anticipated congestion of the Panama Canal, which is projected to faceserious capacity constraints within the next 15 years. The proposal is to integrate Bolivia’s eastern and Andeanrailways to form a link to and from the Chilean port of Arica and the Brazilian port of Santos. The increase intraffic from such a link is expected to be between 2 and 20 million tonnes with a cut in transport costs of aboutUS$16 per tonne (UNCTAD, 1999a).

290

areas where income and population densities arelow. As an alternative, a number of manufacturing/service agglomerations could be developed. Largeareas with low population densities (sub-SaharanAfrica and Central Asia) would still require severalsuch locations and a considerable labour mobility topopulate these places. Of particular interest from theperspective of globalization is the formation ofdynamic economic regions and export processingzones (EPZs), which often thrive in the open tradingenvironment (Scott, 1998). These zones can becomecentres of industry, producer services and urbanamenities. Their growth derives from trade, thecapacity to attract financing and skills, and the use ofagglomeration effects to create networking relation-

ships yielding the maximum of synergy. Recentresearch also suggests that advances in telecommuni-cations, by increasing the frequency of contactbetween people, can motivate greater face-to-faceinteraction and make it more desirable to live incities (Gaspar and Gleaser, 1999).

In most developing countries, a small number ofcities – with the capital city in the lead – generate halfor more of the country’s GDP. These interlockingmetropolitan areas comprising an economic regioncould become the principal “growth pole” for coun-tries (Simmie and Sennet, 1999). These regionscould be located in a single country as in Brazil,India and China or straddle two or three countries asthey do in Southeast Asia.

Box 10.5 The benefits and limits of economic agglomeration

China’s special economic zones are a well-known example of a successful formation of spatial agglomerations.These zones have attracted substantive but spatially limited public investments in infrastructure. They haveoffered free trade with otherwise protected countries and have thus been particularly successful in attractingforeign investment. Alongside foreign investment, skills, production techniques and management knowledgein private companies improved while domestic policy-makers gathered institutional knowledge and practicalpolicy experience. This hands-on experience then helped to attract additional foreign capital and is nowhelping to develop China’s hinterland.

Mauritius’ export processing zone (EPZ) provides another example of a successful spatial agglomerationstrategy. Operating under free-trade principles, it enabled an export boom in garments to European markets andan accompanying investment boom at home. The Mauritian EPZ was created as part of an overall developmentstrategy in the 1970s. Given the small size of the home market, it was not surprising that Mauritius wouldbenefit from an outward-oriented strategy. The challenge, however, was to smooth the adjustment process forthe existing domestic garment sector that had been long protected under the country’s ISI regimes. The EPZscheme also provided a way around political difficulties. The EPZ generated new opportunities of trade andemployment, without taking protection away from the import-substituting groups and from the male workerswho dominated the established industries. The segmentation of labour markets early on between male andfemale workers, with the latter predominantly employed in the EPZ, was particularly crucial, as it preventedthe expansion of the EPZ from driving wages up in the rest of the economy, thereby disadvantaging import-substituting industries. New employment and profit opportunities were created at the margin, while leaving oldopportunities undisturbed. This in turn paved the way for the more substantial liberalization that took place inthe mid-1980s and 1990s.

Where these linkages to the domestic sector are not developed, the success of the EPZ model is often lesssustainable. For example, during the 1980s the Dominican Republic was able to diversify out of its depend-ence on agricultural commodity exports by expanding its production of garments for the United States market.However, the country’s increasing share of the North American market owed less to domestic competitivenessthan to the arrival of United States subsidiaries and their subcontractors in the country’s EPZ. When wagesincreased, foreign investors relocated to lower-wage economies in Central America. Because the exportindustry never established domestic linkages or generated a national supply base, export growth did little toraise long-term capacity (Vincens, Martínez and Mortimore, 1998). Problems can also arise out of the exten-sive tax inducements granted in EPZs. EPZs typically offer tax concessions for five to ten years and, in somecases, as in Honduras, they are granted on a permanent basis (Agosin, Bloom and Gitli, 2000). The resultingrevenue losses for the national governments can be substantial. For Bangladesh, the revenue losses associatedwith tax concessions in the EPZ amount to around US$84 million p.a.

291


Most developed countries have vastly benefitedfrom the productivity gains that economic agglom-erations and clusters provide. Agglomeration gainswere probably most pervasive during the secondwave of globalization, when trade between devel-oped countries became determined not so much bycomparative advantage based on differences infactor endowments but by cost savings fromagglomeration and scale. Because such cost savingsare quite specific to each activity, although eachindividual industry became more and moreconcentrated geographically, the industry as awhole remained very widely dispersed to avoidcosts of congestion.

However, while agglomeration economies aregood news for those in the clusters, they are badnews for those left out. A region may be uncom-petitive simply because not enough firms havechosen to locate there. As a result “a ‘divided

world’ may emerge, in which a network of manu-facturing firms is clustered in some ‘high wage’region, while wages in the remaining regionsstay low” (Yussuf, 2001).

Intra-industry trade in agriculture. Classical tradetheory suggests that countries specialize in inter-national trade according to differentials inproduction costs for different goods. These costdifferentials can result either from efficiencydifferences in the use of production factors(Ricardian trade specialization) or from differen-tials in factor endowment (Heckscher/Ohlin).When developing countries specialized inproducing and exporting agricultural productsfor which they seem to have a classical “compara-tive advantage”, they were facing increasinglybinding constraints for their potential to grow(relative to world markets). This was not neces-

Box 10.6 Why and when is two-way trade in agriculture important for developing countries?

There is growing recognition that developing countries can reap important benefits from two-way trade (TWT)and there is evidence that TWT can provide an avenue for successful globalization strategies.

First, a shift towards TWT is typically associated with a shift in trade to processed products and thus higher-value goods. Within a given infrastructure, a shift in trade towards higher-value goods reduces the share of trans-action costs per unit of merchandise and thus helps overcome the geographic and infrastructure constraintsfaced by many developing countries. Second, TWT helps to cope with adjustments in factor markets arisingfrom the large swings in international commodity prices witnessed over the last decades. As trade is largely theexchange of similar and processed products, shocks result in a reallocation of production factors within anindustry, rather than between industries. The latter is typically a process that involves the discontinuation ofactivities in one sector and the loss of jobs, in order to move factors to other industries or sectors. This type ofresilience against international shocks can be an important argument for policy-makers in developing countriesto integrate their economies faster and more fully into international markets.

Third, TWT in food and agriculture offers economies of scale in the food industry. It enables domesticproducers and processors to sell products that are homogeneous with respect to factor requirements but hetero-geneous with respect to utilization and marketing to both domestic and foreign markets. Like increased naturalprotection, scale economies are particularly important when countries want or have to integrate their domesticeconomy into international markets and expose their domestic sectors to greater competition from abroad. Thisis a particular benefit for small developing countries, notably small islands, for which TWT could provide aninteresting avenue to reap economies of scale, diversify trade and escape the volatility of price swings that theywould face otherwise. Fourth, if intra-industry adjustments dominate, commitments towards freer trade arelikely to be more comprehensive and to last longer. Empirical studies (Caves, 1981) suggest that increasing IITfollowing trade liberalization keeps pressures from import competition low. As a consequence, politicians aremore likely to press ahead with the process of trade liberalization since the high political cost associated withresources shifting between industries is limited. Finally, IIT in food and agriculture enables those developingcountries that are scarcely endowed with productive natural resources (land, water, climate, etc.), to create andfoster trade opportunities independent of the ability to produce primary agricultural products. TWT in food andagriculture has boosted the food processing sector in many Asian economies and created a trade surplus forcountries that lack the climatic and agronomic conditions for a flourishing agricultural export economy.

292

sarily as a result of limits in domestic factorendowments or as a result of trade barriers inmarkets abroad (see Chapter 9), but simply becausethe export markets they produced for exhibitedlow and declining demand elasticities, while theirown import demand remained elastic.20 Thiscreated the so-called “elasticity pessimism” andthe economic rationale for the import-substitutingindustrialization (ISI) strategies pursued by manydeveloping countries in the 1950s and 1960s (seeSection 10.2.1 above).

Krugman (1986) showed how a country couldovercome the elasticity constraint by embarking onintra-industry trade (IIT) or two-way trade (TWT)in differentiated products. The principle is asfollows: if consumers have a certain taste forvariety, each new differentiated product creates aniche and the corresponding demand. If thenumber of products produced in a given country isrelated to the size of the economy, then the coun-tries with the fastest growth also tend to producemore products. Contrary to the traditional view,this mechanism does not need a price (exchangerate) or demand adjustment to equilibrate thetrade balance. Instead, the mechanism works

endogenously. The country with higher growthproduces more product variety, which in turngenerates its own export markets.

These microeconomic effects play a role in thegrowth process and interact with macroeconomicpolicies, notably with the exchange rate. Forexample, there is evidence from a comparisonbetween the Republic of Korea and TaiwanProvince of China that the latter was able togenerate more product variety than the Republicof Korea and rely less on a continuous competitivedevaluation to gain a market share (OliveiraMartins, 1992 and Feenstra, Yang and Hamilton,1999). More recent work has also shown a positiveand significant impact of product variety on rela-tive export intensity and growth (Funke andRuhwedel, 2001).

In practice, however, TWT in food and agricul-ture was largely limited to trade within developedcountries. In developing countries, trade patternsin food and agriculture remained biased towards atraditional (inter-industry) trade specialization,which broadly reflected two major factors. First, agreat number of developing countries experiencedlow GDP growth rates and failed to attain the GDP

20 Even for the development of a competitive food processing sector, most developing countries are simply too small to reach the necessary economiesof scale. Where developing countries have made significant inroads into food processing – for example, orange juice production in Brazil, cannedpineapples in Thailand and soluble coffee production in Colombia and Brazil – the scale required for efficient production means that upstream accessto raw materials and downstream access to markets must also be secured on a large scale. Many developing countries lack the raw materials, capitaland market access to make processing viable (UNCTAD, 2000).

Box 10.7 How has two-way trade been quantified?

To measure the order of magnitude and the development in IIT specialization, a modified Grubel-Lloyd (GL)index of two-way trade (TWT) has been computed. The modification of the GL index was necessary to accountfor the overall trade imbalance in food and agriculture that is characteristic of many developing countries. TheTWT index for n products (i) and a given country j is computed as follows:

where Xij, Mij are export and import values in current US$.

In general the TWT index measures the proportion of total trade (sum of values of imports and exports) thatis composed of trade in “similar products”. A value of the index close to one indicates that there is predomi-nantly IIT (i.e. in differentiated products), while a TWT value close to zero suggests that trade is primarily inter-industry trade, i.e. in different products. The computations reported are made at the highest possibledisaggregation level that is allowed by the data (the FAO trade database for food and agriculture comprises amaximum of 521 different products).

−.−= ∑∑∑=

==

n

in

i

ij

ij

n

i

ij

ij

j

M

M

X

XTWT

1

11

5.01

293


levels required to stimulate a greater diversificationof demand and eventually rising IIT. Second,growing TWT goes hand in hand with the devel-opment of an internationally competitive foodprocessing industry, a process for which mostdeveloping countries faced major constraints. Foodprocessing industries are well established in theindustrialized countries.

But indicators for trade diversification andspecialization (Table 10.3) also show that the difference in the levels of TWT between developedand developing countries was not always aspronounced as in the 1990s. Throughout the 1960sand 1970s, IIT in agriculture has been low whencompared to manufactures, in developed anddeveloping countries alike. However, while TWTremained at low levels in developing countries,many developed countries recorded a rapidgrowth in TWT trade over the 1980s and 1990s.For the United States and the EU, for instance, thelevel of TWT in agriculture increased rapidly since the early 1970s, while the TWT coefficientsremained largely unchanged throughout the devel-oping world. In sub-Saharan Africa, TWT accountsfor merely 16 percent of total agricultural tradeand this share has remained unchanged since

1970. The level is higher in the Near East/NorthAfrica and South Asia but there too IIT in food andagriculture stagnated.

A closer inspection of country-specific informa-tion suggests that the level of TWT in agricultureincreases rapidly when trade barriers are reducedand even more so when countries integrate theireconomy into a common economic market (Figure10.6). Such a change is accompanied not only bythe increasing amount of IIT, i.e. by the volume ofexports and imports of similar products, but alsoby a rapid increase in the number of productstraded. For countries that are firmly integratedinto a common market (the Netherlands, Belgium),TWT in food and agriculture has reached levelsthat are comparable to those attained in manufac-tures. Country-specific data also reveal that a risingTWT is often associated with a growing tradesurplus in food and agriculture and vice versa. Thissuggests that the ability to generate a trade surplusin food and agriculture does not so much dependon the ability of a country to produce a certainamount of raw material as on the country’s capacityto produce differentiated products and to cater forspecific markets, including market niches.

Region TWT (percentage) Number of products traded

1969/71 1984/86 1997/99 1969/71 1984/86 1997/99

Developed countries 25.2 30.2 34.6 252 297 279

EU 28.4 33.4 45.3 278 329 395

North America 31.0 33.8 44.8 280 315 390

Other countries 18.5 21.3 26.8 199 236 332

Developing countries 17.4 17.6 20.4 98 119 194

East Asia 13.9 17.5 16.0 101 136 186

Latin America and the Caribbean 14.0 15.3 18.1 122 139 234

Near East/North Africa 29.2 28.2 29.5 120 146 200

South Asia 17.7 17.6 13.5 96 167 227

Sub-Saharan Africa 16.4 15.7 16.8 77 79 142

Transition economies 26.3 19.0 28.9 118 121 269

Eastern Europe 25.2 19.3 28.0 117 119 278

Source: own calculations.

Table 10.3 Two-way trade in food and agriculture, by region

Spain

0

0.1

0.2

0.3

0.4

1961

1966

1971

1976

1981

1986

1991

1996

EU

Portugal

0

0.1

0.2

0.3

0.4

1961

1966

1971

1976

1981

1986

1991

1996

EU

Greece

0

0.1

0.2

0.3

0.4

1961

1966

1971

1976

1981

1986

1991

1996

EU

Sweden

0

0.1

0.2

0.3

0.4

0.5

1961

1966

1971

1976

1981

1986

1991

1996

TWT

EU

Finland

0

0.1

0.2

0.3

0.4

0.5

1961

1966

1971

1976

1981

1986

1991

1996

EU

Austria

0

0.1

0.2

0.3

0.4

0.5

0.6

1961

1966

1971

1976

1981

1986

1991

1996

EU

Ratio Ratio Ratio

RatioRatioRatio

TWT

TWT

TWT

TWT

TWT

294

10.4 Concluding remarks

Globalization – the growing integration of economiesand societies around the world – is a complexprocess that affects the world’s food and agricul-tural economy in numerous ways. Cheaper andfaster transportation, easier communication andthe development of the Internet are importantdrivers. Also important are a growing number ofinternational agreements that have codified andliberalized the flow of goods and capital. Thesefactors have resulted in a rapid expansion of tradeand FDI but also in the rise and growing influenceof transnational companies. The impacts of thesenew factors have been very positive overall, eventhough the benefits are distributed unevenly. Forexample, globalization has helped to fight povertyand undernourishment in China, Viet Nam andThailand, but has done little so far to integrate thepoorest in sub-Saharan Africa, to improve theirfood security, or to enable the region’s farmers tomake significant inroads into markets abroad.

This raises the question as to what factors deter-mine success or failure, integration or marginaliza-tion. Why have some countries been able to takeadvantage of the great development potential thatglobalization offers while others have failed to do so?

Some of the correlates of success or failure havebeen identified in this chapter. To begin with,openness to trade and capital flows, and the abilityto adopt and to adapt technological innovations areundoubtedly among the important factors forsuccess. Also geographic location and endowmentwith infrastructure can play a crucial role in deter-mining whether a country thrives or falls furtherbehind in an increasingly globalized economicenvironment. But probably most important are thedomestic incentive system and the companion poli-cies that facilitate the integration process.

A number of examples have been presented todocument success and failure in the process ofglobalization. No claim is being made that theseexamples are representative or comprehensive.Nonetheless, the examples suggest that a number

Figure 10.6 Two-way trade in food and agriculture, the effects of economic integration

295


of common features are associated with success orfailure in the process of global integration.

First, while openness to trade and investmentflows is an important contributor to a successfulglobal integration process, openness alone is not aguarantor for success. In many cases, openness hasemerged gradually alongside overall economic andagricultural development. At the same time, nocountry has recorded high growth in the long termon the back of infant-industry protection andimport substitution policies. Nowhere have insula-tion and protection spurred on agricultural growthand overall development in a sustainable manner.

Second, successful globalizers are masters inmanaging adjustment. They succeed in rational-izing excess capacities and create exit possibilitiesfor farmers and new employment opportunities atminimal cost. The creation of township and villageenterprises in China, the pruning of excesscapacity in Viet Nam’s coffee sector and creditrestrictions imposed on unproductive chaebols inthe Republic of Korea have been mentioned.Gradual adjustment is particularly important foragriculture, as a large part of the human and finan-cial capital of the sector is fairly immobile in theshort term. Managed transition provides an oppor-tunity to reallocate resources or gradually depre-ciate them. Active adjustment management alsohelps to mitigate adverse impacts for the poor.

While the process of structural change oftencreates greater opportunities for the poor in themedium term, it also means that they have to bearshort- to medium-term transitional costs that theyare ill positioned to absorb. This is particularly trueof trade reforms, where the adjustment costs oftencome upfront, while the benefits are seen only overa longer period of time. The policy measures tomanage adjustment entail an appropriate mix,sequencing and phasing of trade reforms; they alsoinclude measures that prepare farmers and proces-sors for international competition, e.g. throughtraining and technical assistance. Even where andwhen border protection is largely removed,farmers can vastly benefit from measures thatprotect them from excessive price swings (e.g.China and Viet Nam).

Successful integration is also a process oflearning and experimenting and there are meas-ures that promote this learning process. The two-track system in China’s agriculture (both in terms

of export orientation and in terms of free marketplus government control) seems to work. It hasallowed policy-makers to gather experience as towhat system works and what does not, as to whatfarmers can do best and what not. Two-tracksystems have also proved useful in opening up tointernational competition without facing the costsof massive and rapid adjustments for a wholeeconomy. Probably the most prominent examplesfor successful two-track systems are China’s specialeconomic zones and Mauritius’ EPZ. There arevarious channels through which these two-tracksystems facilitate transition towards freer environ-ments. They allow, for example, a country toprovide foreign investors with special conditionsthat may be difficult to guarantee for the wholeeconomy. They also help domestic companies toprepare for growing competition from abroad andallow policy-makers to adjust the domestic frame-work of competition policies to an environment offreer trade and capital flows.

Globalization has generated growth in FDI thatexceeded growth in trade flows. FDI inflows canplay a catalytic role for development. FDI providesnot only an important source of finance. Moreimportant, it is a carrier of technology, skills andmanagement techniques. But, as with trade, successrests not only on the degree of openness. As impor-tant as the quantity of inflows is the quality of FDI.High-quality FDI is characterized by low repatria-tion levels and intensive linkages to domesticfarmers. Experience from FDI in India’s foodindustry, in particular, has demonstrated its poten-tial for promoting agriculture and overall ruraldevelopment. TNCs provided farmers with betterseeds, enhanced technologies and more stableprices and thus boosted crop yields and farmincomes. The contracts that forge these linkages arecrucial for success. But there are also exampleswhere FDI largely failed to create linkages withlocal farmers and even instances where TNCs haveadded to the marginalization of whole farm popula-tions. Developments in the global coffee marketsillustrate this point. There is evidence for a growingconcentration in trading and processing of coffeeand there is also evidence that TNCs managed toreap a growing share in the total value created inthe coffee marketing chain. It is however less clearto what extent these developments reflect the abuseof market power and the absence of an appropriate

296

competition policy framework that can address thenew competition policies in a globalized market.

Globalization in agriculture has also beenbrought about by an internationalization ofproduction technologies. The green revolution wasthe single most important vehicle in this process.Again, while some countries have been phenome-nally successful in adopting these new technologies,others have largely failed to do so. As with tradeand investment flows, the correlates of success arenot merely openness to innovation. Numerousstudies suggest that it is more important to createthe appropriate domestic environment that allowslocal producers to employ the new technologiesgainfully. In short, adoption has to be accompaniedby adaptation to provide success. Similarly, successor failure in reaping the benefits of biotechnologieswill depend less on availability than on the capacityto adapt the new technologies to the agronomicand economic environments that prevail in aspecific location. Finally, geographic location andinfrastructure endowment play a crucial role insuccessfully tapping potential world markets.There is ample evidence that the lack of infrastruc-ture (not the existence of trade barriers) has beenthe crucial impediment that hindered sub-SaharanAfrica’s farmers from making significant inroadsinto OECD markets.

But there is also evidence that globalization canoffer new opportunities to leapfrog old obstaclesresulting from unfavourable locations or inade-

quate infrastructure. In general, new technologiesare cheaper and faster and can bring the mostremote areas to the heart of the markets. Theseinclude Internet-based business communicationsystems as well as multimodal transportationsystems. In sparsely populated regions, transactioncosts can be reduced by promoting economicagglomeration.

In summary, globalization offers a great poten-tial for farmers and the entire food sector of devel-oping countries. Many developing countries aresuccessfully tapping this potential, but not all ofthem are able to take full advantage of the newopportunities. The ability of a country to reap thebenefits of globalization depends on factors such asopenness to trade and capital flows, ability to adopttechnological innovations, and also geographiclocation or infrastructure endowment. The variousexamples suggest that openness and outward-oriented policies characterize many successful globalizers but per se they are not guarantors forsuccess. More important are the companion poli-cies on the domestic front that facilitate integrationinto global markets. These are policies that provideappropriate transition periods towards freer trade;help adapt new, external technologies to thedomestic environment; and provide competitionpolicy settings and design contracts that also allowsmall-scale agriculture to thrive within the opera-tions of TNCs.

Documents

Globalization in food and agriculture · 10.1 Globalization as an ongoing process Globalization refers to the ongoing process of rapid global economic integration facilitated by lower