Barrientos Et Al 2003

World Development Vol. 31, No. 9, pp. 1511–1526, 2003� 2003 Elsevier Ltd. All rights reserved

Printed in Great Britain0305-750X/$ - see front matter

-750X(03)00110-4
doi:10.1016/S0305www.elsevier.com/locate/worlddev
A Gendered Value Chain Approach to Codes

of Conduct in African Horticulture

STEPHANIE BARRIENTOSInstitute of Development Studies, Brighton, UK

CATHERINE DOLANUniversity of East Anglia, Norwich, UK

and

ANNE TALLONTIRE *

University of Greenwich, Chatham, UK

Summary. — Codes of conduct covering the employment conditions of Southern producersexporting to European markets mushroomed throughout the 1990s, especially in the horticulturesector linking UK and European supermarkets with export firms in Africa. The majority ofemployment in this sector is ‘‘informal,’’ a significant proportion of which is female. This paperexplores the gender sensitivity of codes currently applied in the African export horticulture sectorfrom an analytical perspective that combines global value chain and gendered economyapproaches. Through an analysis of these two approaches, it develops a ‘‘gender pyramid,’’ whichprovides a framework for mapping and assessing the gender content of codes of conduct. Thepyramid is applied to codes that cover employment conditions in three commodity groups andcountries exporting to European markets: South African fruit, Kenyan flowers and Zambianvegetables and flowers. It concludes that the gender sensitivity of codes needs to be greatlyenhanced if they are to adequately address employment conditions relevant to informal andespecially women workers.� 2003 Elsevier Ltd. All rights reserved.

Key words — Africa, gender, codes of conduct, employment, export horticulture

*This paper is based on a research project funded by

the UK Department for International Development

(ESCOR/SSR). DFID supports policies, programs and

projects to promote international development. DFID

provided funds for this study as part of that objective

but the views and opinions expressed are those of the

authors alone. Final revision accepted: 25 March 2003.

1. INTRODUCTION

Codes of conduct 1 covering the employmentconditions of Southern producers exporting toEuropean markets mushroomed throughoutthe 1990s, especially in the horticulture 2 sectorlinking UK and European supermarkets withexport firms in Africa. Much employment inthis sector is flexible and informal (i.e., tem-porary, seasonal, casual, migrant, contract), ofwhich a significant proportion is female. Buthow gender sensitive are these codes of con-duct? Are they able to address the specificproblems linked to informal and feminizedemployment within the sector? This paper ad-dresses these questions through a gender map-ping of the codes currently being applied inAfrican horticulture.

151

Codes have become prevalent in the UK foodretail sector, where most supermarkets are nowimplementing their own codes to cover theirsupply chains. But these are not the only codesthat suppliers of horticultural products face.Various actors, such as importers, exportersand local trade associations have also devel-oped their own codes. Externally, independent

1

WORLD DEVELOPMENT1512

codes have been established through organiza-tions such as Social Accountability Interna-tional (SAI) in the US and the Ethical TradingInitiative (ETI) in the United Kingdom. As aresult, suppliers are faced with a plethora ofcodes, some of which are similar, but amongwhich there can be considerable variability.This variability is particularly marked in thecase of gender issues. Some codes integrate anumber of international conventions relating togender discrimination and inequality, yet othercodes make no mention of gender at all. Evenwhere codes address gender issues, their cov-erage and sensitivity can often be limited. Manycompanies adopt codes to reduce their risks ofnegative exposure to poor employment prac-tices within their supply chains. Yet if codes failto address the poor working conditions andunequal treatment faced by certain groups suchas women, these risks of exposure will persistand the overall effectiveness of codes will besignificantly reduced (Barrientos, 2000).This paper explores the gender sensitivity of

codes currently applied in the African exporthorticulture sector from an analytical per-spective that combines global value chain(GVC) and gendered economy approaches.Through an analysis of these two approaches,it develops a ‘‘gender pyramid,’’ which pro-vides a framework for mapping and assessingthe gender content of codes of conduct. Thepyramid is applied to codes that cover em-ployment conditions in three commoditygroups and countries exporting to Europeanand particularly UK markets: South Africanfruit, Kenyan flowers and Zambian vegetablesand flowers. The paper is organized as follows.Section 2 examines the GVC approach, andhow it provides an analytical framework forunderstanding the development of codes ofconduct. It considers the specific context ofhorticulture in South Africa, Kenya andZambia, and the gendered employment rela-tions that operate at the production end of thechain. Section 3 explores the concept of a‘‘gender economy,’’ and how this can be inte-grated into the GVC approach to analyze em-ployment codes within this sector. It alsodevelops the ‘‘gender pyramid’’ as a frameworkfor analyzing the gender content of codes ofconduct. Section 4 maps the plethora of codesthat currently exist in African horticulture, andassesses their gender content. A final sectionconcludes the paper by assessing the limitationsand potentials of these codes, and suggests av-enues for further research. 3

2. A VALUE CHAIN APPROACH TOCODES OF CONDUCT

While a number of analytic frameworks havebeen put forth to examine the crossnationalactivities of firms, 4 the GVC approach is par-ticularly useful in analyzing the role that stan-dards and private sector codes of conduct playin the governance of international trade. Thisapproach, 5 developed by Gereffi and Kor-zeniewicz (1994) explores how the linkagesbetween the production, distribution and con-sumption of products are globally intercon-nected along value chains that embody anetwork of activities and actors (Kaplinsky,2000; Sturgeon, 2001). Gereffi (1994) identifiedfour main dimensions of GVCs. These include:an input–output structure or the value-addedsequence in the production and consumption ofa product; a territorial configuration or thegeographical concentration and/or dispersionof production and marketing; a governancestructure or the power relations that determinehow financial, material and human resourcesare distributed within the chain; and an insti-tutional framework that identifies how local,national, and international contexts influenceactivities within chains (Gereffi, 1995). Thelatter two dimensions offer significant insightsinto codes of conduct.The concept of governance has attracted

considerable attention, largely because gover-nance structures determine the prospects offirms in developing countries to engage inglobal trade and how the benefits of participa-tion are distributed along the chain (Gereffi,Humphrey, Kaplinsky, & Sturgeon, 2001;Gibbon, 2000; Humphrey & Schmitz, 2001).Gereffi (1995) originally distinguished betweentwo types of governance structures: producer-driven and buyer-driven. Producer-drivenchains are typical of capital- and technology-intensive industries where transnational manu-facturing firms ‘‘drive’’ the chain, controllingthe core technologies and production facilities,often through vertical integration. In contrast,in buyer-driven chains, large retailers or brand-name companies make the key decisions aboutthe structure and activities of actors in thechain without actually owning any manufac-turing facilities themselves. While the producer/buyer-driven dichotomy has recently beenqualified to capture the dynamic nature ofGVCs, 6 the typology is nevertheless helpful inteasing out the way that power is exercisedwithin chains. The UK–Africa horticultural

A GENDERED VALUE CHAIN APPROACH 1513

value chain, which is characteristic of a buyer-driven commodity chain, is a case in point.Powerful lead firms (supermarkets) governsupply networks that span several Africancountries, defining not only what is to be pro-duced but also how and under what conditionsit is to be produced (Dolan & Humphrey,2000). These supermarkets do not own anyproduction or processing facilities themselvesbut rather steer the process from conception topoint of sale through intensive supply chainrelationships.The key factors driving buyer-driven gover-

nance in horticulture (and the proliferation ofcodes) relate to Gereffi�s fourth dimension: thesocial and economic context in which the chainoperates. For example, the changing nature ofconsumption patterns in northern countries hasincreased the importance of branding andproduct differentiation and shifted the focusaway from price-based competition towardquality, innovation and value-added as the keyperformance criteria for suppliers. Underlyingthis trend is the increasing salience of cre-dence 7 factors among consumers who are notonly concerned about quality and safety, butalso about the social and environmental con-ditions under which products are produced(Reardon et al., 2001). The more complexconsumer and regulatory environment faced byretailers has obliged them to manage theirsupply chains more closely, both to avert neg-ative publicity and to differentiate their prod-ucts from their competitors. One way thatretailers have achieved this is by codifying theknowledge required to meet quality specifica-tions in standards and grading systems, or byadopting the standards defined by other privateand/or public bodies. Codes of conduct are anextension of this process and one way thatglobal buyers endeavor to reduce risk and‘‘govern’’ their supply chains.A GVC approach, therefore, can further an

analysis of codes by focusing on how their ap-plication is shaped by the nature of power re-lations between agents within the chain, as wellas the global context that influences the way thechain operates. This is particularly relevant inhorticulture, where codes are being introducedon a multiple basis by different actors (super-markets, importers/exporters and trade associ-ations) at key points along the GVC, each ofwhich have distinct motivations for theiradoption. But, the institutional context of thehorticultural value chain operating within thedifferent countries varies according to the spe-

cific country context. These have potentiallyfar-reaching implications for firms and workersat the base of the chain. 8

(a) The export horticulture value chain

In many SSA countries, diversification intoexport horticulture has become a promisingstrategy for generating increased employmentand foreign trade. Over the last two decades,SSA�s horticultural exports have doubled, out-stripping the region�s exports for coffee ($1.84billion) and cotton ($1.52 billion), and for allother individual commodities other than cocoaby the end of the decade (Thoen, Jaffee, &Dolan, in press). The growth rates of totalhorticulture commodities have been impressivein all three countries. In Kenya, horticulture isthe fastest-growing sector of the economy,generating over US$270 million and accountingfor 22% of all agricultural exports in 2000. Thisperformance is largely attributable to cutflowers, which surpassed coffee as the country�ssecond largest source of foreign exchange inagriculture, bringing in US$118 million in theyear 2000 (Gachanga, 2002). In South Africatotal fruit exports accounted for 30% of allagricultural export trade in 1999, when thetotal value of deciduous exports alone stood atUS$700 million (Deciduous Fruit ProducersTrust, direct communication, 2000). In Zambiahorticultural products have led the growth inagricultural exports over the past decade. In thelatter part of the 1990s year-on-year growth ofhorticultural agricultural exports exceeded 40%on average and amounted to $63 million in1999 ($20 million for vegetables and $43 millionfor flowers; Giovanucci, Sterns, Eustrom, &Haantuba, 2001).There are important differences in the pro-

duction processes in Kenya, South Africa andZambia, and the way in which their valuechains operate. South Africa has approximately2,000 deciduous fruit farms, most of which arecommercial farms producing directly for export(De Klerk, n.d.). Kenya has a larger number offlower farms, approximately 500, but 75% oftotal exports are supplied by two dozen large-ormedium-scale flower operations (Thoen et al.,in press). Zambia has far fewer producers, with25 companies in the export horticulture indus-try. 9 Notwithstanding this diversity all threecountries rely heavily on European marketsfor their exports, and are therefore very de-pendent on any changes in the European mar-ket, including the trend towards ethical trade.


In 1998–99, for example, 66% of all SouthAfrican fruit, 94% of Kenyan flowers, and100% of Zambian export-quality vegetablesand flowers were supplied to the EuropeanUnion.Levels of employment in the three sub-sec-

tors we focused on varied between the coun-tries. 10 South African deciduous fruit employsan estimated 283,000 workers, in the Kenyancut flower industry estimated employment is40,000–70,000, and in Zambian flowers andvegetables, estimated employment is 8,000 (seeTable 1). 11 While the pattern of employmentvaries somewhat by sector and country, womencomprise between 50% and 75% of total em-ployment in all three countries. Women areconcentrated in the segments of the productionprocess that hold the most significance for thequality of the final product such as picking andpacking, and value-added processing activities.Companies perceive women as more ‘‘pro-ductive,’’ citing women�s ostensibly ‘‘nimblefingers’’ and capacity to perform tedious anddelicate work as essential to fulfilling thequality imperatives of overseas buyers. Femaleemployment is characterized by highly gen-dered and informal employment relations. Inall three countries, women form the core of thetemporary, seasonal and casual work force,while men tend to be concentrated in the fewerpermanent jobs (Barrientos, McClenaghan, &Orton, 2000; Dolan & Tewari, 2001). Womentend to be crowded into a narrow range ofseasonal occupations characterized by longhours and few opportunities for meetingdomestic responsibilities (due to insufficientchildcare, social provision and maternity leave).Informal female employment is accompaniedby job insecurity, risk and lack of employmentor social protection, often with the poorestconditions of employment amongst horticul-tural workers (Barrientos et al., 2000).The prevalence of informal employment in

horticulture can be partially explained by thenature of the industry, in which large numbersof workers are required for planting, picking

Table 1. Estimates of employment in export hor

Total employmen

South African deciduous fruit 283,000

Kenyan flowers 40,000–70,000

Zambian horticulture 8,000 (veg. and flow

Sources: De Klerk (n.d.), Kritzinger and Vorster (1995, 1

Blowfield, Malins, and Dolan (1998) and KFC (2002).

and packing at particular points of the year(Dolan & Humphrey, 2000). In cases such asKenya, however, where export vegetable pro-duction is now year-round, the majority ofworkers remain employed on temporary con-tracts. The informality of horticultural em-ployment needs to be situated within thebroader flexible production strategies adoptedby African firms to maintain competitiveness ina context of globalization. The global trendtoward flexible labor allows employers to varyemployment levels on a highly fluid basis toaccommodate demand instability, therebydriving labor costs down and avoiding many ofthe non-wage costs of employment (Standing,1989, 1999). Within this, informal employmentrelations allow producers to shift the risks ofproduction (from adverse conditions or marketfluctuations) onto workers in the sector. It alsoreflects pressures intrinsic to the value chain,particularly its buyer-driven nature where retailconcentration has created a more fiercelycompetitive landscape for developing countrysuppliers. The retailers� adoption of just-in-timeproduction methods passes the costs of demandinstability and inventory control upstream toproducers. This has made elasticity of labor acompetitive asset.Traditionally the division between formal

and informal forms of working has been de-noted by distinct formal and informal sectors.This rigid classification is untenable in a globaleconomy where the boundary between formaland informal is blurred, and informal workersare increasingly employed within ‘‘formal’’sectors. This process has been accelerated in aclimate of deregulation, with increasing num-bers of both men and women now employed in‘‘informal’’ types of employment, as the shareof secure, permanent, full-time jobs declinesthroughout the world (ILO, 2002; Lund &Srinivas, 2000; Standing, 1999). Today, infor-mal work arrangements are becoming the normin many export sectors, including horticulture,with a ‘‘continuum’’ emerging between formaland informal work. Toward the informal end of

ticulture in South Africa, Kenya and Zambia

t Temporary or seasonal (%) Female (%)

65–75 53

65 75

ers) 60–74 (veg. only) 65 (veg. only)

996), NZTT (1999) and calculated from farm records;


the continuum, workers lack security of em-ployment, have few employment rights, receiveinadequate employment benefits or social pro-tection, lack trade union organization and beara high level of risk and vulnerability withinemployment (Barrientos & Ware Barrientos,2002).While the global expansion in informal work

partially explains women�s dominance in hor-ticulture, the employment strategies of horti-cultural firms are not simply a product offlexible production methods. They also reflectthe fact that labor markets themselves embodyand transmit gender inequalities, and are asElson suggested, ‘‘bearers of gender’’ (Elson,1995, 1999). The next section examines this,and the gendering of the value chain as aframework for a gender analysis of the codes ofconduct themselves.

3. A GENDERED VALUE CHAINAPPROACH TO CODES OF CONDUCT

Gender analysis of GVCs is still at an earlystage (Barrientos et al., 2001), but in the con-text of codes of conduct it is the intersectionbetween value chains and employment at theproduction end of the chain that is key. Thisemployment is embedded within the institu-tional context in which value chains operate,and takes place within labor markets that arethemselves gendered institutions, which reifyand reflect socially constructed gender divisionsof labor (Rai, 2002). Similarly, codes haveevolved as a reflection of national and in-ternational regulations and institutions thatreinforce traditional gendered patterns of em-ployment. Hence codes themselves need to beanalyzed in the context of GVCs and employ-ment patterns that are themselves embedded in,and structured by a gendered economy.The idea that the economy is a gendered

structure emerged from feminist analysis of theeconomy that has become central to recentgender analyses of economic growth and trade(C�agatay, Elson, & Grown, 1995; Grown,Elson, & C�agatay, 2000; Whitehead, 2001). Incontrast to a standard economic analysis thatviews the economy as ‘‘gender neutral,’’ yetprovides only a partial examination of eco-nomic activity that is linked to the market, agender economy approach argues for the in-separability of the reproductive and productivespheres. The understanding of economic activ-ity is therefore extended to include not only

market-oriented activities but also the ‘‘repro-ductive economy’’ (unpaid domestic work andchildcare) that underpins productive marketbased activity, and which is largely undertakenby women (Elson, 1999; Folbre, 1994; White-head, 2001). As feminists have long argued,nonmonetized caring activities such as child-care, cooking and housework are indispensableto the functioning of the ‘‘productive econ-omy’’ as they both maintain and reproduce thelabor force. The reproductive economy alsostrongly differentiates the options of men andwomen to participate in market activity andconditions their subsequent experience of thatemployment.Labor market institutions themselves are

embedded within, and constructed by a gen-dered economy. They reflect the socially de-rived gender division of labor, and are situatedat the intersection between productive (paid)and reproductive (unpaid) work (Elson, 1999;Humphries & Rubery, 1984). Men are con-centrated in the former and are more likely tohold permanent jobs with higher wages. Incontrast, women are concentrated in the ‘‘twi-light’’ zone between the two forms of work,increasingly engaged in ‘‘informal’’ types ofemployment, moving ‘‘flexibly’’ between theproductive and reproductive economy as re-quired by dictates of work. Labor marketregulations and norms have traditionally re-inforced this gender division of labor as theyare predicated on a model of male permanentemployment. This is reflected in trade unionpractices 12 as well as labor law, both of whichare founded on the notion that women aresecondary earners who can rely upon theearnings of men to buffer them against the riskof economic insecurity (Elson, 1999).Codes of conduct are a form of employment

regulation that are drawn up and implementedwithin value chains that reflect the genderednature of labor markets and economic activity.They are thus introduced in the context of thegendered forms of employment that operatewithin the value chain. The remainder of thispaper conceptualizes the gender sensitivity ofemployment codes based on a ‘‘gender pyra-mid.’’ The gender pyramid in Figure 1 dividesthe key issues relating to employment into threeinter-linked levels: formal employment issues(segment A); employment-related issues (seg-ment B); and the wider socioeconomic contextthat affects an individual�s ability to accessparticular types of employment (segment C).This analytical framework enables us to

Reproductive Work

C

I. FormalEmployment

A

II. Informal Employment

B

Figure 1. Gendered employment and the gender pyramid of codes. (A) Regulation and provision of formal conditions ofemployment. (B) Regulation and provision of employment related benefits. (C) Non-employment related benefits and

social provision supporting reproductive work.


examine the variable coverage of gender issuesbetween the various codes and the scope oftheir gender sensitivity.One of the main vehicles for achieving gender

equity in employment is through labor regula-tions. There is a growing consensus that a goodlabor code should be based on standards set bythe International Labor Organization (Diller,1999; Ferguson, 1998; Lee, 1997; Seyfang,1999). The ILO Core Conventions 13 establishthe standard for labor rights worldwide, ex-tending labor protections across nationalboundaries and along GVCs. The advantagesof using ILO Conventions in codes of conductare that they have been internationally negoti-ated and agreed to through a tripartite processinvolving governments, employers and tradeunions. National governments that are mem-bers of the ILO have a formal obligation toadopt the Core Conventions, which shouldenhance the symbiosis of private and publicregulation. 14 ILO Conventions also providesome basis for commonality across the plethoraof codes that exist. Some codes refer to nationallabor legislation as the basic standard of em-ployment protection. Where national legisla-tion is weak, if the code covers ILO coreconventions and wider issues, it can supplementlegislation by providing a floor for employmentconditions. Where national legislation is pro-gressive, it can enhance the coverage of a code,but there is often wide disparity in the extentand depth of gender issues addressed throughnational legislation (which generally only cov-ers segment A).

The implementation of ILO conventionsthrough national labor legislation takes placewithin the context of long-standing institutionalforms of employment that are constructedaround male norms, thereby reinforcing labormarkets as bearers of gender (Elson, 1995,1999). The primary focus of these forms ofregulation is permanent full-time employment,where the worker (usually male) is separatedfrom reproductive activity (usually female)(Elson & Gideon, 1999). Labor regulation isnormally less relevant to informal employment,usually undertaken by women, who balanceand move ‘‘flexibly’’ between productive andreproductive work, and whose employmentneeds can take different forms to permanentworkers. Both national and international laborregulations also strongly depend on employeesbeing represented in a collective bargainingagreement (Ladbury & Gibbons, 2000). Yet inagriculture generally (including horticulture)formal union membership is very low (for ex-ample at most 8% in South Africa) and therepresentation of women�s interests by tradi-tional rural unions is also often weak.In relation to our analytical framework, ILO

conventions and national legislation form thefoundation of segment A of the gender pyra-mid, which covers all issues of employmentregulation that relate to formal employment.These issues include hours and conditions ofwork, wages, contracts of employment, andemployment benefits (e.g., holiday leave, sickleave, and social insurance). Segment A alsoincludes gender issues that relate to formal


employment, such as sex discrimination andunequal pay. The core of workers that aremost likely to benefit most from this type ofemployment regulation are those who are infull-time, permanent employment. Part-time,temporary and seasonal workers might alsoenjoy these benefits, and in so far as they doexperience a formalization of their employ-ment. But the informal nature of most tempo-rary employment typically restricts access bythese workers to the coverage of employmentconditions in segment A (a contract of em-ployment is at most only temporary with lim-ited employment benefits). But many workersin informal employment do not have access tominimal employment rights even where theyare stipulated by legislation (e.g., many receiveno contracts of employment or employmentbenefits). While both men and women in formalemployment can have access to the types ofregulation covered in segment A, given thatmen form the majority of employees in thissituation, there is clearly a gender bias in for-mal employment coverage. This reflects theextent to which labor regulations (be they na-tional, international or private sector codes) are‘‘bearers of gender’’ as discussed above.Segment B of the gender pyramid covers the

types of employment provision and regulationthat relate to and facilitate employment, butextend beyond the minimal formal contractnormally provided. In particular, segment Bincorporates issues that facilitate the combiningof paid productive employment with repro-ductive labor. Such issues include childcareprovision, maternity and paternity leave, re-productive rights, as well as social provisionsuch as housing, transport to and from work,health provision linked to work, and sport fa-cilities. These can affect those that are both informal and informal employment, but will be ofmost benefit to women who are primarily re-sponsible for combining paid work with re-productive labor. Segment B can also includeemployment related benefits that are particu-larly relevant to those in informal employment,such as assurance of seasonal re-employmentand the extension of employment benefits (suchas health and pension insurance) beyond peri-ods of formal employment. Overall, the benefitsand regulations covered in segment B are par-ticularly beneficial to women, who are concen-trated in informal employment, and thereforehave the potential to reverse the male bias ofmuch formal employment regulation and pro-vision. The further codes cover issues in seg-

ment B of the gender pyramid the greater theirgender sensitivity, and the deeper their cover-age of issues that are predominant in informalforms of employment.Segment C of the gender pyramid covers re-

productive work (including domestic responsi-bilities, care of children and elderly relativeswithin the home). This is usually unpaid work,and given the gendered social relations in so-ciety is predominantly undertaken by women.In contrast to segment B, segment C encom-passes issues that are exogenous to the work-place, which no codes of conduct cover. Wehave included segment C to help illuminate thegender issues faced in segment B, which is at theintersection between formal employment andreproductive work, and provide the gendereconomy context to the employment pyramid.The gender division of labor, with womenbearing primary responsibility for reproductivework, conditions their particular employmentneeds. It also constrains their ability to accessformal paid employment, increasing women�sconcentration within informal work, with thehigher level of job insecurity and risk this car-ries.Codes of conduct designed solely around

ILO core conventions and relevant nationallabor legislation (segment A) may cover genderissues related to formal employment. But theyare less likely to be sensitive to the genderedneeds of workers combining productive withreproductive work, and to the needs of those ininformal forms of employment, with womenforming the majority of both these groups inAfrican horticulture. The employment relatedissues relevant to these groups are included insegment B of the gender pyramid, and thegreater the coverage of the issues in segment B,the greater the gender sensitivity of codes. Thegender pyramid reflects the gender economy,and the embedded nature of labor market in-stitutions that contextualize the gender valuechain within horticulture. In the next section wewill use this framework to assess the gendersensitivity of codes of conduct currently beingapplied in African horticulture (South Africa,Kenya and Zambia).

4. GENDER MAPPING OF CODES INAFRICAN HORTICULTURE

In mapping codes, it is important to identifywhich agents within or connected to the GVChave introduced them and why, as this is likely


to influence their form of operation within thechain as well as their potential for gender sen-sitivity. In African horticulture, codes havebeen introduced from three different origins: bydominant buyers such as supermarkets, im-porters and/or individual exporters within thechain (company codes); sectoral trade associa-tions linked to the horticultural value chain(sectoral codes); and independent bodies com-prising companies and a range of civil societyorganizations (independent codes). 15

In the horticultural sector company codesinitially emerged from the more demandingregulatory environment faced by retailers dur-ing the 1990s. In response to food safety andpesticides legislation, several UK supermarketsdeveloped their own codes to govern foodsafety, hygiene, and quality assurance through-out their supply chains. More recently super-markets have expanded the content of thesetechnical codes to include social and environ-mental criteria, or developed separate companycodes to address specific areas (e.g., humanrights and worker welfare, environmental pro-tection, animal welfare, integrated crop man-agement, etc.) (Blowfield, 1999). These codesare applied to all horticulture producers inSouth Africa, Kenya and Zambia that supplythe main UK multiples. Similarly, importerssupplying fresh produce to the United King-dom have also developed their own codes ofconduct, largely in response to the needs of thesupermarkets.Sectoral codes, developed by industry-wide

organizations and/or trade associations, wereoriginally focused on the environmental aspectsof production. Some sectoral codes have theirorigin in the North, and are being adopted byAfrican suppliers either voluntarily or as a re-quirement to supply certain buyers. These in-clude the EUREPGAP protocol, developed bya network of European retailers to ensure bestpractice in the production of fresh produce andMPS, which covers the production of flowers.Other sectoral codes have been establishedthrough consortia of trade associations andproducers in Africa, who moved early to in-troduce their own standards to promote ethicalproduction. These include the following codes:Kenya Flower Council (KFC), Fresh ProduceExporters Association of Kenya (FPEAK),Zambia Export Growers Association (ZEGA),and Horticultural Promotion Council (HPC),Zimbabwe. In both Kenya and Zambia thehorticultural export associations have alsoparticipated in the COLEACP regional har-

monized framework, which provides guidingprinciples for local codes covering productsafety, environmental and social responsibility.In addition, consortia of retailers, trade

unions, nongovermental organization (NGOs)and companies have developed independentsocial codes such as the UK-based ETI BaseCode, and the US based social managementstandard, SA8000. The ETI Base Code hasbeen particularly significant in the sourcing ofAfrican produce, as seven of United Kingdom�slargest supermarkets are ETI members 16 andhave agreed to apply the ETI Base Code to alltheir fresh produce suppliers. In South Africasupermarket codes are the main standards thatgrowers face, as there is currently no othersectoral code that covers the deciduous fruitsector. 17 In Zambia, growers face supermarketand importer codes, as well as the ZEGA codediscussed below. In the Kenyan flower sector,the application of supermarket codes is onlynow being established as most of the major UKsupermarkets previously accepted the KFCcode of practice as the standard.The GVC and the employment within it is

gendered, as discussed above in relation to the‘‘gendered economy.’’ The following mappingof the multiple codes applying to African hor-ticultural producers supplying UK supermar-kets highlights that these codes are alsogendered in a way that reflects the genderedvalue chain. We explore this by firstly examin-ing the independent, social codes that arecommon to all three countries, followed by ananalysis of company-specific codes, and sec-toral codes, some of which are specific to indi-vidual countries.

(a) Independent social codes: SA8000 and ETIbase code

Two of the most significant independent so-cial codes in the horticulture sector are SA8000and the ETI Base Code (see Barrientos et al.,2001). While the ETI Base Code is not in itself adefined auditable standard, the ETI has estab-lished a number of pilot projects with the aimof experimenting with different multi-stake-holder approaches to monitoring. The princi-ples embodied in SA8000 (established by SAI)and the ETI Base Code are essentially the same,and they share several common elements (seeTable 2); both are based on ILO conventionscovering minimum labor standards, are inde-pendent and freestanding social codes, andboth were originally developed in consultation

Table 2. Corporate codes of conduct––summary of contenta

Code Audited

code

S/holder

participation

design//

implement

Linked

production

standards

Envi-

ron-

ment

National

legislation

Free-

dom of

associa-

tion

Collective

bargaining

Forced

labor

Child

labor

Discrimina-

tion

International independent

ETI Base N Y//Y N N Y Y Y Y Y Y

SA8000 Y Y//Y N N Y Y Y Y Y Y

Company

Summary UK importers Y N Y Y/N Y Y Y Y/N Y Y

Summary UK supermarkets Y Y//N Y Y/N Y Y Y/N Y/N Y Y

Sectoral

COLEACP N N Y Y Y Y Y Y Y Y

EUREPGAP Y N//N Y Y Y N N N N N

MPS Y Y//Y Y Y N Y Y Y Y Y

KFC Y N//N Y Y Y Y Y Y Y Y

FPEAK Y N//N Y Y Y Y Y N Y Y

ZEGA N N//N Y Y Y Y Y Y Y Y

Safety/

hygiene

Living

wages

Regular

employment

Abuse Work

hours

Work

contract

Accommoda-

tion quality

Leave)holi-day+ special

Grievance

system

Manage-

ment system


ETI Base Y Y Y Y Y Y Y N N N

SA8000 Y Y N Y Y N Y N Y Y

Company

Summary UK importers Y Y/N Y/N Y/N Y Y Y Y/N Y/N N

Summary UK supermarkets Y Y/N Y/N Y/N Y Y/N Y Y/N N Y/N

Sectoral

COLEACP Y Y Y Y Y Y Y Y Y

EUREPGAP Y NL NL N NL N Y N N N

MPS Y Y N N Y Y Y Y N Y

KFC Y Y N N Y Y Y Y Y Y

FPEAK Y Y N N Y Y Y Y Y Y

ZEGA Y Y N Y Y Y Y Y Y Y

aY ¼ Yes, N ¼ No, Y/N ¼ mixed, NL ¼ national legislation.

AGENDERED

VALUECHAIN

APPROACH

1519


with multiple stakeholders (including NGOs,trade unions and private companies). Finallythey both make reference to national legislationin the country of application and cover a rangeof issues, as summarized in Table 2.Given that SA8000 and the ETI Base codes

incorporate the core ILO Conventions, theyclearly address segment A of the gender pyra-mid in terms of establishing basic employmentrights and entitlements. These codes also in-clude other issues that affect all workers, suchas safety and hygiene, living wages, abuse, andhours of work. The ETI is stronger in terms ofits coverage of insecure employment thanSA8000 in that it includes specific reference toregular employment and work contracts,though the wording is vague. In terms of gen-der, both include sections on discrimination butwhere ETI has a reference to nondiscriminationin compensation, SA8000 has a more generalclause on equal pay. But, as can be seen fromTable 3, both the ETI and SA8000 codes areweaker once you move beyond specific em-ployment to employment-related issues. Nei-ther includes coverage of reproductive rights,maternity or paternity leave, protection forpregnant women 18 or childcare. These are allimportant employment related issues forwomen workers in particular, affecting theiraccess to, and experience of employment. Thus,the coverage of the ETI and SA8000 codes islargely confined to segment A of the genderpyramid, and whilst their coverage is good atthis level, they are limited in their scope ex-tending to segment B. Both SA8000 and ETIincorporate the principle of stakeholder par-ticipation by NGOs, companies, and tradeunions in their organizations and the develop-ment of codes.

(b) Company codes

The ETI has had an important influence onthe evolution of company codes applied tohorticulture sectors in several African coun-tries. Some UK supermarkets have recentlybegun to use the ETI Base Code directly intheir supplier auditing, which means thatoverall their coverage of gender issues has im-proved. This is significant, as our analysis ofthe supermarkets� own codes revealed that theytended to be weaker than the ETI in terms ofthe coverage of gender issues, with the excep-tion of discrimination. At the level of segmentA of the gender pyramid, only one UK super-market code includes equal pay, while one

other includes sexual harassment and abuse. Interms of employment related issues (segment Bof the gender pyramid) supermarket codes areweak. While one supermarket code includesmaternity and paternity leave (in accordancewith national laws), the other codes provide fewprotections. Importer codes tend to be based onthe supermarket codes, but because importersfrequently supply more than one supermarket,they (and their growers in Africa) are forced tocomply with several company codes that em-body different criteria. As a result many UKimporters acting for different supermarketshave developed codes that address their clients�combined requirements. This has led to severalof the importer codes being at least as good as,or better than the individual supermarket codes(see Tables 2 and 3).

(c) Sectoral codes

The plethora of codes and criteria thatgrowers face is made clear when sectoral codesare combined with international and super-market codes. The EUREPGAP protocol pro-vides a unified standard for growers amidst thevariety of supermarket and importer codes, andhas been widely adopted by suppliers of Afri-can fresh produce. EUREPGAP is largely atechnical code, however, into which a smallnumber of social provisions have been in-cluded. In general the code makes recourse tonational law as a guiding principle on socialissues. In terms of segment A of the genderpyramid, none of the ILO Conventions arementioned, nor are any issues such as wages,abuse, working hours or discrimination (seeTable 3). As such EUREPGAP is much weakerthan both the ETI and SA8000 codes, and mostof the UK supermarket codes. If suppliers wereonly required to comply with the EUREPGAPcode, they would not address any of the gender-specific issues raised in either segments A or Bof the gender pyramid, except where they werecovered by prevailing national legislation.MPS has become the most important code to

gain entry to the Dutch flower auctions, theworld�s largest market outlet for cut flowers.MPS originated as an environmental standard,and has only recently added a social chapter.Because it is based on the ILO Conventionsand the Universal Declaration of HumanRights, growers who adopt the full SocialChapter address segment A of the gender pyr-amid including discrimination, equal pay, col-lective bargaining, forced labor and child labor.

Table 3. Corporate codes of conduct––gender issuesa

Discrimina-

tion

Abuse Sexual

harass-

ment

Confidential

complaints

Reproduc-

tive rights

Mater-

nity leave

Paternity

leave

Child-

care

Pregnant

women

protection

Equal

pay

Equal pay/

work of

equal value


ETI Base Y Y Y N N N N N N Y N

SA8000 Y Y Y N Y N N N N (Y) N

Company

Summary

importers

Y Y/N Y/N N Y/N Y/N Y/N N N Y/N Y/N

Summary

supermarkets

Y Y/N Y/N N N Y/N Y/N N N Y/N N

Sectoral

COLEACP Y Y Y Y Y Y N Y Y Y N

EUREPGAP N N N N N N N N N N N

MPS Y N Y N N Y N N Y Y N

KFC Y N N N N Y N N Y Y N

FPEAK Y N N N N Y N N Y Y N

ZEGA Y Y Y Y Y Y N Y N Y N

aY ¼ Yes, N ¼ No, Y/N ¼ mixed.

AGENDERED

VALUECHAIN

APPROACH

1521


But, as we move down the gender pyramid toemployment-related issues, the coverage of theMPS code is more variable. While the codespecifically addresses sexual harassment, ma-ternity leave, work contracts and hours, it doesnot cover other employment related issues suchas confidential complaints, grievance proce-dures, regular employment, childcare, paternityleave and reproductive rights (see Table 3).As a guiding framework, COLEACP pro-

vides the most progressive coverage of socialissues of all the codes in this study. It makesreference to compliance with all the core ILOConventions, and covers key direct employ-ment issues such as equal pay, safety and hy-giene, work hours, contracts and discrimination(see Table 3). Its coverage of segment B of thegender pyramid is also relatively comprehen-sive, incorporating provisions for housing,workplace childcare, and maternity leave.While it does not cover paternity leave, it is oneof the few codes to address reproductive rights.Within the framework set by COLEACP, thereare three sectoral codes that apply nationallywithin Kenya and Zambia: KFC, FPEAK andZEGA. All three codes have their origins inproduction and quality management systems,and were developed to enhance the marketingpotential of horticultural exports in the Euro-pean Union. The degree to which these codesaddress gender-specific criteria varies betweenthe codes and across the types of issues. Seg-ment A of the gender pyramid is well covered.All three codes stipulate compliance with thecontent of various ILO Conventions such asforced labor, child labor, equal pay, anti-dis-crimination, collective bargaining, and freedomof association as well as adherence to nationallegislation. All three codes also address manyof the gender specific criteria in segment B ofthe gender pyramid (see Table 3), however theZEGA code is much more comprehensive. Allinclude issues such as maternity leave, no use ofwomen in pesticide-related tasks, and housing.There are however, several issues that are onlycovered by the ZEGA code, which are partic-ularly important for women in the horticulturesector. These include housing, sexual harass-ment, confidential complaints, childcare, sexualabuse, and reproductive rights. There are alsoseveral issues that none of the codes cover, in-cluding regular employment 19 and paternityleave.A common element, therefore, to most sec-

toral codes is that they include social provisionas an extension of existing management, pro-

duction and environmental standards. Thecomprehensiveness of their social provision,and especially their gender provision, variesgreatly from EUREPGAP that has minimalcoverage to COLEACP that is much morecomprehensive. At a national level, theFPEAK, KFC, and ZEGA codes are relativelythorough on segment A of the gender pyramidyet more variable in their inclusion of the em-ployment-related issues found in segment B.None of these national sectoral codes extend tothe issues beyond the workplace found in seg-ment C, nor are they as comprehensive asCOLEACP, which only provides a frameworkfor these national codes. While Kenya andZambia both have relatively good sectoralcodes, they are linked to national legislation,which lags behind South Africa. Therefore inKenya and Zambia the provisions of the codeswill often be of primary importance in pro-tecting labor conditions. South Africa, on theother hand, is still only at an early stage indeveloping a sectoral code, but where companycodes require compliance with national legis-lation, this sets a strong standard. 20

5. GENDER LIMITATIONS ANDPOTENTIALS OF CODES––CONCLUDING REMARKS

Codes of conduct have come about as a re-sult of civil society pressure to improve poorworking conditions often found within buyer-driven GVCs, such as horticulture. They formpart of the governance of GVCs, and help toavoid the risk of adverse publicity to thedominant buyer, or supermarkets in the case offresh fruit, vegetables and flowers. But theirgovernance of value chains also allows domi-nant buyers to off-set many of the risks ofproduction and distribution onto producers bysetting strict conditions, such as meeting highproduction standards, accepting falling com-petitive market prices, and working to tight‘‘just-in-time’’ production schedules. Producersin turn cope with the risks and volatility ofsupply through flexible production and em-ployment methods. An important elementwithin this is the use of informal employment,where many of the risks are born by workerswho have no job security, formal employmentor social protection, or labor organization.Codes of conduct are meant to improve thepoor working conditions experienced by suchworkers. Yet they form part of a governance


system within GVCs that itself encourages theuse of informal workers to meet ‘‘just-in-time’’production requirements, when employmentconditions for informal workers are oftenpoorest. This anomaly itself is likely to act as alimitation on the effectiveness of codes of con-duct in improving the position of workers.Employment within GVCs is further shaped

by the social and institutional context withinwhich they operate. This environment is notgender neutral, but is shaped by the gendereconomy, predicated on a division of laborbetween productive and reproductive work.Women�s primary responsibility for reproduc-tive work constrains their access to formal paidemployment, increasing their concentrationwithin informal work. It also conditions theirspecific gender needs as workers, juggling theirreproductive roles in the context of the inse-curity and lack of protection provided by in-formal work. They face intensified gender risks,such as lack of reproductive rights, maternityleave and childcare, and do not have adequatejob security or employment protection to copewith those risks.Combining a GVC and gendered economy

approach has helped us to examine the genderlimitations of codes in the context of the genderpyramid. Codes have been formulated in thecontext of labor market regulation and insti-tutions that are embedded within the gendereconomy. As such, while many codes address

the employment needs of formal workers, theyoften fail to address the more complex genderneeds of informal workers, where the condi-tions of employment are often worst. If codesare to address the employment issues faced bythe majority of informal workers, their gendersensitivity will need to be enhanced to incor-porate broader employment-related issues thatare of concern to women workers in particular.We have found that in general codes that havebeen developed on the basis of a multi-stake-holder approach are more likely to addressgender issues, but that often this remains con-strained to formal employment. It is beyond thescope of this paper to examine the importantrole civil society organizations could play in theimplementation of codes in developing coun-tries, which is the subject of ongoing researchby the authors in African horticulture. 21 Butdespite the gender limitations of codes, the factthat they have come about through civil societypressure raises the possibility that NGOs andlabor organizations could continue to influencecode development, potentially resulting incodes with enhanced gender sensitivity in thefuture. While such organizations are not im-mune from reinforcing adverse gender normsand practices, they also embody the potential toidentify and represent the needs and interests ofworkers (including women and those in ‘‘in-formal’’ work arrangements) in the process ofcode development.

NOTES

1. Codes of conduct have been defined as ‘‘a set of

ethical principles and standards that attempt to guide a

firm�s environmental and social performance’’ (Utting,

2000, p. 4).

2. In this paper, horticulture encompasses fresh fruits,

vegetables and cut-flowers.

3. The main aim here is to explore the content of codes

in African horticulture using the analytical framework

provided by a gendered value chain approach. It is

beyond the scope of this paper to address the imple-

mentation of codes. For a more detailed empirical study

of codes within horticulture in sub-Saharan Africa, see

Barrientos, Dolan, and Tallontire (2001).

4. These include Porter�s theory of value chains (1990),

Malsot�s (1980) theory of ‘‘fili�eeres,’’ Ruigrok and van

Tulder (1995) concept of the industrial complex, Law�s(1999) notion of Actor Networks and Ernst�s (2000)

notion of the global production network. See Gereffi,

Humphrey, and Sturgeon (2002) for a review of such

approaches.

5. Gereffi�s (1994) concept of global commodity chains

is based on the work of Hopkins and Wallerstein (1986),

is distinguished from Porter�s (1990) concept of ‘‘value

chains’’ through its embodiment of an explicitly inter-

national dimension.

6. Humphrey and Schmitz (2001) have shown that

different parts of the same chain can be governed in

different ways. Recent work has extended governance

types to include ‘‘international trader-driven’’ chains

(Gibbon, 2000), ‘‘€ıınfomediary-driven’’ chains where

internet-based ‘‘informediaries’’ such as America On-

line (AOL) or yahoo! control e-commerce (Gereffi,

2001) and typology of governance types (market,

modular, relational, captive and network) (Gereffi

et al., 2002).


7. The quality and safety characteristics that constitute

credence attributes include the following: (a) food safety;

(b) healthier, more nutritional foods; (c) authenticity; (d)

production processes that promote a safe environment

and sustainable agriculture; (e) ‘‘fair trade’’ attributes

(e.g., working conditions) (Reardon, Codron, Busch,

Bingen, & Harris, 2001).

8. Convention theory also provides a framework of

analysis for understanding standards in the governance

of international trade. Based on the work of Salais and

Th�eevenot (1986), Boltanski and Th�eevenot (1989), Syl-

vander (1995), Valceschini (1993), and Sauvee (1998),

convention theory suggests that in situations where

product quality cannot be discerned through price or

observation, quality conventions are necessary to convey

information to consumers. While convention theory is

useful in identifying the role that standards and regula-

tion play in value chain governance, it has tended to

focus on the local or national levels of the chain (Raikes,

Friis Jensen, & Ponte, 2000). In contrast, the GVC

approach is better equipped to capture economic

dynamics along the full length of global chains.

9. Only two companies are responsible for nearly all

the exports of vegetables to Europe, as well as a large

proportion of the flowers exported. Other companies in

the industry export only flowers.

10. This paper focuses on wage employment. Fewer

than 2% of Kenya�s smallholders are directly engaged as

outgrowers in export horticulture production (Bawden,

Aust Sterns, Harris, & Berdegue, 2002). Smallholders

are not used in the production of South African fruit or

Zambian flowers, and produce only a very small volume

of Zambian vegetables.

11. Exact employment figures are difficult to obtain for

all three countries given a lack of official statistics,

therefore the figures given here are estimates based on

interviews.

12. While more unions are broadening their objectives

to represent workers located outside standard forms of

work, this has historically not been the case.

13. The ILO core labor standards cover freedom of

association and collective bargaining, forced labor, child

labor, discrimination and equal remuneration (ILO,

2001).

14. Implementation by national governments often

remains weak.

15. The main codes covered here apart from individual

company codes are: ETI (Ethical Trading Initiative),

SA8000 (SAI), COLEACP (Liaison committee Europe–

Africa–Caribbean–Pacific), EUREPGAP (Euro-Retailer

Produce Working Group protocol for Good Agricul-

tural Practice), KFC (Kenyan Flower Council), MPS

(Milieu Project Sierteelt), FPEAK (Fresh Produce Ex-

porters Association of Kenya), ZEGA (Zambia Export

Growers Association).

16. Supermarket members of the ETI are ASDA, The

Co-Op, J Sainsbury, Marks and Spencer, Safeway,

Somerfield, and Tesco. They are applying codes to all

their ‘‘own brand’’ products, including fresh produce,

which is counted as ‘‘own brand.’’

17. This reflects the fact that since deregulation of the

sector in South Africa, no single body has been in a

position to introduce such a code. However the Decid-

uous Fruit Producers Trust is considering such a move

using EUREPGAP, and there is a multi-stakeholder

initiative by AGRI-WESTCAPE (the main agricultural

union representing farmers).

18. The ETI base code does not refer to protection of

pregnant women, and SA8000 only refers to dismissal of

pregnant women as an aspect of discrimination within

its guidance notes.

19. Few codes applied to agricultural sectors cover

regular employment due to the widespread reliance on

casual, temporary and seasonal work.

20. Although formal legislation introduced since 1994

is good in South Africa, enforcement still remains very

weak. In this situation codes of conduct can act as an

additional means of enforcement.

21. See http://www.ids.ac.uk/ids/global/gendeth.html

for further details.

REFERENCES

Barrientos, S. (2000). Globalization and ethical trade:assessing the implications for development. Journalof International Development, 12, 559–570.

Barrientos, S., Dolan, C., & Tallontire, A. (2001).Gender and ethical trade: a mapping of the issuesin African horticulture. Natural Resources Institute

http://www.ids.ac.uk/ids/global/gendeth.html


Report No. 2624. Chatham Maritime: NaturalResources Institute.

Barrientos, S., McClenaghan, S., & Orton, L. (2000).Ethical trade and South African deciduous fruitexports––addressing gender sensitivity. EuropeanJournal of Development Research, 12(1), 140–158.

Barrientos, A., & Ware Barrientos, S. (2002). Extend-ing social protection to informal workers in thehorticulture global value chain. Social ProtectionDiscussion Paper No. 0216. Washington, DC:World Bank.

Bawden, R., Aust Sterns, P., Harris, S., & Berdegue, J.(2002). Increasing rural household incomes in Kenyathrough horticulture. In Partnerships for Food In-dustry Development––Fruits and Vegetables. Lansing,MI: Michigan State University.

Blowfield, M. (1999). Ethical trade: a review of devel-opments and issues. Third World Quarterly, 20(4),753–770.

Blowfield, M. E., Malins, A., & Dolan, C. (1998). KenyaFlower Council support to enhancement of socialand environmental practices: report of the designmission. Mimeo. Chatham Maritime: Natural Re-sources Institute.

Boltanski, L., & Th�eevenot, L. (Eds.). (1989). Justesse etjustice dans le travail. Paris: CEE/PUF.

C�agatay, N., Elson, D., & Grown, C. (1995). Specialissue: gender and macroeconomics. World Develop-ment, 23(11), 1825–2017.

De Klerk, M. (no date). Deciduous fruit industry study.Cape Town: Commission of Inquiry into the Provi-sion of Rural Financial Services.

Diller, J. (1999). A social conscience in the globalmarketplace? Labor dimensions of codes of conduct,social labeling and investor initiatives. InternationalLabor Review, 138(2), 99–129.

Dolan, C., & Humphrey, J. (2000). Governance andtrade in fresh vegetables: the impact of UK super-markets on the African horticulture industry. Journalof Development Studies, 37(2), 147–176.

Dolan, C., & Tewari, M. (2001). From what we wear towhat we eat: upgrading in global value chains. IDSBulletin, 32(3), 94–104.

Elson, D. (1995). Male bias in the development process.Manchester, UK: Manchester University Press.

Elson, D. (1999). Labor markets as gendered institu-tions: equality, efficiency and empowerment issues.World Development, 27(3), 611–627.

Elson, D., & Gideon, J. (1999). The internationalcovenant on economic, social and cultural rights andthe empowerment of women. NY: UNIFEM.

Ernst, D. (2000). Global production networks and thechanging geography of innovation systems: implica-tions for developing countries. Working Paper No. 9.East-West Centre: Honolulu.

Ferguson, C. (1998). A review of UK company codes ofconduct. London: Department for International De-velopment (DFID).

Folbre, N. (1994). Who pays for the kids? Gender and thestructures of constraint. London: Routledge.

Gachanga, S. P. (2002). Horticultural industry in Kenya.Paper Prepared for the Globalisation, Productionand Poverty Workshop, Nairobi, Kenya.

Gereffi, G. (1994). The organization of buyer-drivenglobal commodity chains: How US retailers shapeoverseas production networks. In G. Gereffi, & M.Korzeniewicz (Eds.), Commodity chains and globalcapitalism (pp. 95–123). Westport, CT: Praeger.

Gereffi, G. (1995). Global production systems and thirdworld development. In B. Stallings (Ed.), Globalchange, regional response: The new internationalcontext of development (pp. 100–142). Cambridge,MA: Cambridge University Press.

Gereffi, G. (2001). Beyond the producer-driven/buyer-driven dichotomy: the evolution of global valuechains in the Internet era. IDS Bulletin, 32(3), 30–40.

Gereffi, G., Humphrey, J., Kaplinsky, R., & Sturgeon,T. (2001). Globalization, value chains and develop-ment. IDS Bulletin, 32(3), 1–9.

Gereffi, G., Humphrey, J., & Sturgeon, T. (2002). Thegovernance of global value chains: an analyticframework. Review of International Political Econ-omy (under review).

Gereffi, G., & Korzeniewicz, M. (Eds.). (1994). Com-modity chains and global capitalism. Westport, CT:Praeger.

Gibbon, P. (2000). Back to the basics through delocali-sation: the Mauritian garment industry at the end ofthe twentieth century. Working Paper 00.7. Copen-hagen: Centre for Development Research.

Giovanucci, D. P., Sterns, P. A., Eustrom, M., &Haantuba, H. (2001). The impact of improved gradesand standards for agricultural products in Zambia.Phase One Assessment and Recommendations forUSAID. Lansing, MI: Institute for Food and Agri-cultural Standards.

Grown, C., Elson, D., & C�agatay, N. (2000). Specialissue: growth, trade, finance, and gender inequality.World Development, 28(7), 1145–1390.

Hopkins, T., & Wallerstein, I. (1986). Commoditychains in the world economy prior to 1800. Review,10, 157–170.

Humphrey, J., & Schmitz, H. (2001). Governance inglobal value chains. IDS Bulletin, 32(3), 19–30.

Humphries, J., & Rubery, J. (1984). The reconstitutionof the supply side of the labor market: the relativeautonomy of social reproduction. Cambridge Journalof Economics, 8(4), 331–346.

International Labour Organisation (ILO) (2001). Fun-damental ILO conventions. Available: http://www.ilo.org/public/english/standards/norm/whatare/fun-dam/index.html

International Labour Organisation (ILO) (2002). Decentwork and the informal economy. InternationalLabour Conference, 90th Session, Geneva: ILO.

Kaplinsky, R. (2000). Spreading the gains from globali-sation, what can be learned from value-chain ana-lysis. IDS Working Paper No. 110, Brighton: IDS.

Kenya Flower Council (KFC) (2002). A growingresponsibility. Available: http://www.kenyaflow-ers.co.ke/page4.html

Kritzinger, A., & Vorster, J. (1995). The labor situationin the South African deciduous fruit industry.Research Report, Stellenbosch, SA: University ofStellenbosch.

http://www.ilo.org/public/english/standards/norm/whatare/fundam/index.html



http://www.kenyaflowers.co.ke/page4.html

http://www.kenyaflowers.co.ke/page4.html


Kritzinger, A., & Vorster, J. (1996). Women farmworkers on South African deciduous fruit farms:gender relations and the structuring of work. Journalof Rural Studies, 12(4), 339–351.

Ladbury, S., & Gibbons, S. (2000). Core labour stan-dards. Key issues and a proposal for a strategy.London: Department for International Development(DFID).

Law, J. (1999). After ANT: complexity, naming andtopology. In J. Law, & J. Hassard (Eds.), Actor-network theory and after (pp. 1–14). Oxford: Black-well.

Lee, E. (1997). Globalisation and labor standards: areview of issues. International Labor Review, 136(2),173–189.

Lund, F., & Srinivas, S. (2000). Learning from experi-ence: A gendered approach to social protection forworkers in the informal economy. Geneva: WIEGOand ILO.

Malsot, J. (1980). Fili�eere et effets de domination dans lesyst�eeme productif. Annales des Mines, 1, 29–40.

NZTT, NRDC/ZEGA Training Trust (1999). Informa-tion leaflet. Lusaka: NZTT.

Porter, M. (1990). The competitive advantage of nations.London: Macmillan.

Rai, S. (2002). Gender and the political economy ofdevelopment. Cambridge: Polity.

Raikes, P., Friis Jensen, M., & Ponte, S. (2000). Globalcommodity chain analysis and the French fili�eereapproach: comparison and critique. Economy andSociety, 29(3), 390–417.

Reardon, T., Codron, J. M., Busch, L., Bingen, J., &Harris, C. (2001). Global change in agrifood gradesand standards: agribusiness strategic responses indeveloping countries. International Food and Agri-business Management Review, 2(3).

Ruigrok, W., & van Tulder, R. (1995). The logic ofinternational restructuring. New York: Routledge.

Salais, R., & Th�eevenot, L. (Eds.). (1986). Le travail,march�ee, r�eegles, conventions. Paris: INSEE-Economica.

Sauvee, L. (1998). Toward an institutional analysis ofvertical coordination in agribusiness. In J. Royer, &R. Rogers (Eds.), The industrialization of agriculture(pp. 27–71). Aldershot, England: Ashgate.

Seyfang, G. (1999). Private sector self-regulation forsocial responsibility: mapping codes of conduct.Ethical trading and globalisation: self-regulationand workers� experience, Working Paper 1. OverseasDevelopment Group, Norwich, UK: University ofEast Anglia.

Standing, G. (1989). Global feminisation through flexiblelabor. Working Paper no. 31. Labor Market Analysisand Empowerment Planning, Geneva: ILO.

Standing, G. (1999). Global feminization through flex-ible labor: a theme revisited. World Development,27(3), 583–602.

Sturgeon, T. (2001). How do we define value chains andproduction networks? IDS Bulletin, 32(3), 9–18.

Sylvander, B. (1995). Convention de qualit�ee, concur-rence et coop�eeration: Cas du label rouge dans lafili�eere volaille. In G. Allaire, & R. Boyer (Eds.), Lagrande transformation (pp. 73–96). Paris: INRA-Economica.

Thoen, R., Jaffee, S., & Dolan, C. (in press). Equatorialrose: the Kenyan–European cut flower supply chain.In Kopiki, R., (Ed.), Supply chain development inemerging markets: Case studies of supportive publicpolicy. Cambridge, MA: MIT Press.

Utting, P. (2000). Business responsibility for sustainabledevelopment. Occasional Paper No. 2, Geneva:UNRISD.

Valceschini, E. (1993). Conventions �eeconomiques etmutation de l��eeconomie contractuelle dans le secteurdes l�eegumes transform�ees. Economie Rurale (Novem-bre–Decembre), 19–26.

Whitehead, A. (2001). Trade, trade liberalisationand rural poverty in low-income Africa: agendered account. Background Paper for theUNCTAD 2001 Least Developed Countries Re-port, Mimeo.

Barrientos Et Al 2003

Documents

Transcript of Barrientos Et Al 2003