82520 Women and Trade in Africa: Realizing the Potential Edited by: Paul Brenton Elisa Gamberoni Catherine Sear Women and Trade in Africa: Realizing the Potential Edited by: Paul Brenton Elisa Gamberoni Catherine Sear Foreword Women feature significantly in trade in Africa: they carry goods across borders, produce products, especially food, that can be exported, and own and manage trade-oriented firms. But these producers and traders often face specific constraints that undermine their economic activities. Women traders working in the informal sector are often subject to harassment and extortion at the border. Women are more readily denied access to key trader networks than men. Time-consuming trade procedures and documentary require- ments impinge more heavily on women, given the time they need for their household duties. And women working to produce exportable goods and services are typically less able than men to get the inputs and materials that would raise their productivity and allow them to compete better in overseas markets. This matters not just because of the adverse impact on the families and firms affected but because facilitating the role of women in trade will be essential if Africa is to achieve the enormous potential it has to trade more within its own borders and with the wider world. Achieving this trade poten- tial will contribute to better food security, vital job creation, and poverty reduction. This volume contains chapters that look at how women participate in trade in Africa, the constraints they face, and the consequence of those restrictions. It contributes to the rather small number of analytical work devoted to this issue and seeks to encourage researchers in Africa to explore the specific challenges faced by women in trade. The analysis leads to a set of key recommendations to help policy makers facilitate the participation of women in trade and in so doing assist Africa in achieving its trade potential. In particular, governments need to do more to: • recognize the role that women play in trade and ensure that officials at all levels understand the importance of that role. • ensure that rules and regulations governing trade are clear, predictable, and widely available at the border, which is critical for women traders working with very limited margins in the informal sector. • prioritize the simplification of trade documents and regulatory requirements, since this will benefit women in trade given the time and mobility constraints that arise from household responsibilities. • design interventions to develop trade in ways that ensure that women benefit. For example, programs that support improved access to information will miss women traders and entrepreneurs if the support is channeled through existing male-dominated trader networks. • help women address the risks that they face in their trade-related activities, given that they are typically more risk averse than men and respond to risk in different ways. Through its Gender and Trade Practices, the Africa region of the World Bank is seeking to improve the understanding of the barriers women trad- ers face and how those barriers can be effectively removed. By increasing analysis of these issues together with positive interventions—such as those described in chapter 2, which have supported traders’ associations and provided training to officials in the eastern part of the Democratic Republic of Congo—we can bring about the behavioral changes at all levels in society that will underpin inclusive trade in Africa. Marcelo M. Giugale Director Department of Economic Policy and Poverty Reduction Programs Africa THE WORLD BANK Acknowledgements We are grateful for the comments and suggestions of the peer reviewers: Ian Gillson (Senior Trade Economist, PRMTR) and Gozde Isik (Trade Economist, AFTPM). This work has been financed by the Multi-Donor Trust Fund for Trade and Development 2 supported by contributions from the governments of Sweden, the United Kingdom, the Netherlands and Switzerland.  vii Contents Chapter 1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Paul Brenton and Elisa Gamberoni Chapter 2 Barriers, Risks, and Productive Potential for Small-Scale Traders in the Great Lakes Region.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 María Elena García Mora and Sabrina Roshan Chapter 3 Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Louis Njie Ndumbe Chapter 4 Women Cross-Border Traders, Challenges, and Behavior Change Communications.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Susan D. Ityavyar Chapter 5 The Gender Dimension of Uganda’s Cotton Sector. . . . . . . . . . . . . . . 77 John Baffes and Laura Maratou-Kolias Chapter 6 Services Trade and Gender.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Antoine Coste and Nora Dihel Chapter 7 Gender in the Tourism Industry: The Case of Kenya.. . . . . . . . . . . . . 115 Michelle Christian, Elisa Gamberoni, and José Guilherme Reis Chapter 8 Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Thomas Bossuroy, Francisco Campos, Aidan Coville, Markus Goldstein, Gareth Roberts, and Sandra Sequeira Chapter 9 Trade and Gender in Tanzania: What Matters— Participation or Outcomes?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Josaphat Kweka and Mahjabeen Haji Figures Figure 2.1 Reported Cases of Corruption by Officials in DRC and Rwanda, by Type of Authority.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Figure 2.2 Number of Incidences of SGBV Reported by Traders in DRC and Rwanda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Figure 3.1 Key Border Crossings in the Southwest Region of Cameroon and the Mamfe-Ekok-Ikom Trade Corridor.. . . . . . . . . . . . . . . . . . . . . . . 47 Figure 4.1 Charter for Small-Scale Traders: Outreach Process. . . . . . . . . . . . . . . 72 Figure 5.1 Cotton Yields of Female Farmers as a Percentage of Yields of Male Farmers in Uganda by Region, 2009 and 2010.. . . . . . . . . . . . . . 83 Figure 5.2 Average Number of Pesticide Sprays by Men and Women Cotton Producers in Uganda by Region, 2009 and 2010. . . . . . . . . . . 87 Figure 6.1 Evolution of the Structure of Female Employment in Sub-Saharan Africa, 1992–2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Figure 6.2 Evolution of the Share of the Tertiary Sector in Female Employment in Selected World Regions, 1119, 2002, and 2012.. . . . 98 Figure 6.3 Female Employment, Ownership, and Leadership in Manufacturing and Services in Six World Regions. . . . . . . . . . . . . . . . 101 Figure 6.4 Female Employment, Ownership, and leadership in Sub-Saharan Africa and All Countries by Services Subsector. . . . . 103 Figure 6.5 Average Number of Employees in Manufacturing and Services Firms with Male and Female Top Managers in Sub-Saharan Africa and Elsewhere. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Figure 6.6 Proportion of Full-Time Female Employees in Manufacturing and Services Firms with Male and Female Top Managers in Sub-Saharan Africa and Elsewhere .. . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Figure 6.7 Proportion of Manufacturing and Services Firms Engaged in Exports with Male or Female Top Managers in Sub-Saharan Africa and Elsewhere. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Figure 6.8 Percentage of Firms Identifying Access to Finance as the Main Constraint, with Male or Female Top Managers in Sub-Saharan Africa and Elsewhere.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Figure 6.9 Proportion of Women in Top Positions of Professional Services Firms in Eastern and Southern Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Figure 6.10 Gender of Management and Export Status of Professional Services Firms in Eastern and Southern Africa.. . . . . . . . . . . . . . . . . . 108 Figure 6.11 Proportion of Women in Management of Professional Services Firms in Eastern and Southern Africa.. . . . . . . . . . . . . . . . . . 109 Figure 6.12 Proportion of Women in Top Positions of Professional Services Firms in Eastern ansd Southern Africa, by Sector.. . . . . . . . 110 Figure 6.13 Proportion of Female Fellows of the Chartered Institute of Management Accountants in Selected Economies. . . . . . . . . . . . . . . . 110 Figure A6.1 Regression 1 (Female Ownership). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure A6.2 Regression 2 (Female Top Managers). . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure A6.3 Regression 3 (Female Workers). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Figure 9.1 Exports of Goods and Services in Countries of the East African Community, 2000–2011.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure 9.2 Exports of Tanzania, 2006–09.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Figure 9.3 Historical Trend of World Exports by Destination, 1990–2009.. . . . 170 Figure 9.4 Active Exporters per Million Inhabitants in Rwanda, Tanzania, and Uganda, 2003–11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171  ix Figure 9.5 Contribution to Growth of Tanzania’s Manufacturing Exports, 2004–09. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Figure 9.6 Proportion of Female Ownership in Various Sectors of the Tanzania Economy, 2008.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Figure 9.7 Distribution of Profit of Firms by Gender Ownership in Tanzania, 2008. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Figure 9.8 Share of Domestic and Export Sales by Share of Female Ownership in Tanzania, 2008... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Tables Table 2.1 Household Indicators: Comparison of Traders’ Families in the Sample with Families in Selected Regions of the Democratic Republic of Congo, 2011.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Table 2.2 Main Products Carried by Small Women Traders in the Democratic Republic of Congo and Rwanda. . . . . . . . . . . . . . . . . . . . . . . 31 Table 4.1 Charter for Cross-Border Traders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Table 4.2 Outreach Materials for Communication for Behavioral Change. . . . 70 Table 5.1 Selected Demographic Characteristics of Surveyed Households in Uganda, 2009 and 2010 .. . . . . . . . . . . . . . . . . . . . . . . . . . 81 Table 5.2 Cotton Area, Production, and Yield in Surveyed Male- and Female-Headed Households in Uganda by Region, 2009 and 2010.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Table 5.3 Inputs Used in Surveyed Male- and Female-Headed Households in Uganda by Region, 2009 and 2010. . . . . . . . . . . . . . . . 86 Table 5.4 Cotton Income, Prices, and Marketing Characteristics of Surveyed Male and Female Cotton Producers in Uganda by Region, 2009 and 2010. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Table A.1 .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Table 8.1 Sample Characteristics of the Impact Evaluation Survey, the Finscope SME Survey, and the World Bank Enterprise Survey in Kwazulu-Natal Province, South Africa. . . . . . . . . . . . . . . . . 139 Table 8A.1 Summary Statistics on Entrepreneurs’ and Business Average Characteristics.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Table 8B.1 Predicting Whether a Firm Exports.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Table 8C.1 Predicting Whether a Firm Has More than Five Workers. . . . . . . . . . 156 Table 9A.1 Regression Results: Predictors of Export Sales (EXPORTERS).. . . . . 181 Table 9B.1 Regression Results: Predictors of Firm Profitability (Exporters). . . 183 Table 9C.1 Regression Results: Probability of Becoming an Exporter (Whole Sample).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 x Women and Trade in Africa: Realizing the Potential Boxes Box 1.1 Charter for Cross-Border Traders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Box 3.1 The Administrative Process of Exporting Eru: Documentation and Taxes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Box 3.2 Process for Obtaining Administrative Approval .. . . . . . . . . . . . . . . . . . 52 Box 4.1 Village Information Center Project in Uttar Pradesh, India .. . . . . . . . . 61 Box 4.2 Democratic Republic of Congo National Railway Company. . . . . . . . 64 Box 6.1 The Four Pillars of the World Economic Forum’s Global Gender Gap Index.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Box 7.1 European Buyers Promote Equitable Practices through Supplier Accountability.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Box 7.2 Starwood Hotels and the Mayan World Foundation Purchase Crafts from Female Artisans for Hotel Decoration.. . . . . . 124 Box 9.1 Some Key Gender Disparities in Tanzania. . . . . . . . . . . . . . . . . . . . . . . 173 Box 9.2 Gender Patterns of Fishing Activities in Tanzania. . . . . . . . . . . . . . . . 174 Table B9.2.1: Participation in Fishing Activities by Gender in Tanzania, 2004. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Table B9.2.2: Average Profit from Fishing Activities by Gender in Tanzania, 2004. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Introduction 1 1   Introduction Paul Brenton and Elisa Gamberoni A frican countries have enormous potential for trade with the global mar- ket and for more intensive trade among themselves. Regional trade in Africa can play a vital role in diversifying economies and reducing depen- dence on the export of a few mineral products, in delivering food and energy security, in generating jobs for the increasing numbers of young people, and in alleviating poverty and promoting a shared prosperity.1 New opportunities are appearing to provide food and manufactured goods to Africa’s growing cities, trade in services provides a route to more diversified exports and more competitive domestic services markets, and functioning regional markets have become a springboard to markets in other continents. Women play a key role in trade in Africa and will be essential to Africa’s success in exploiting its trade potential. Women make a major contribution to trade in most African countries through their involvement in the produc- tion of tradable goods as cross-border traders and as managers and owners of firms involved in trade. In many countries in Africa, the majority of small farmers are women, and they produce crops such as maize, cassava, cotton, and rice that have enormous potential for increased trade between African countries and with the global market.2 Women are also involved in providing services across borders, such as education, health, and professional services, including accountancy and legal services. Hundreds of thousands of women cross borders in Africa every day to deliver goods from areas where they are relatively cheap to areas in which they are in shorter supply. However, Africa’s trade potential is undermined by constraints that women face. The contribution of women to trade is much less than it could be because of various specific nontariff barriers that impinge particularly heavily on the trade activities of women and women-owned enterprises. These barriers often push women traders and producers into the informal economy where lack of access to finance, information, and networks jeop- ardize their capacity to grow and develop their business. Such conditions prevent women from taking full advantage of the opportunities created by trade and thus undermine the aspirations of countries in Africa to use trade Paul Brenton is the Africa Trade Practice Leader (AFTPM) at the World Bank. Elisa Gamberoni is an economist (AFTP1) at the World Bank. 2 Women and Trade in Africa: Realizing the Potential photo: Transaid as a driver of growth, employment, and poverty reduction. Without attention to these barriers, the objective of achieving an effective continental free trade area in Africa will not be realized. Yet, policy makers typically overlook women’s contribution to trade and the challenges they face. This neglect reflects, in part, the lack of data and information on women and trade in Africa and also the underrepresenta- tion of small traders and rural producers in trade and trade policy discus- sions. This volume brings together a series of chapters that look at the ways that women participate in trade in Africa, the constraints they face, and the impact of those constraints. It seeks to extend the rather small amount of analytical work that has been devoted to this issue and to encourage researchers, especially in Africa, to look more carefully at the specific chal- lenges women face. Women participate in trade in many ways. The chapters look at the conditions and challenges faced by three broad groups: informal cross- border traders; women who participate in the production of traded goods and services, ranging from rural farmers of cotton to professional activities such as legal and accountancy services; and women entrepreneurs with dominant ownership of exporting companies. While women in all these groups face challenges, they often differ and so broad programs to address gender inequalities may miss the specificities of the constraints that women face. The book highlights the importance of identifying and removing the conditions that prevent women from exploiting the full potential of trading Introduction 3 photo: Georgina Goodwin/World Bank activities. This introductory chapter provides a background to the volume and summarizes the contributions that are included. Here we derive a number of simple steps that policy makers in Africa can take to facilitate the participation of women in trade and in so doing assist Africa in achieving its trade potential. This in turn will drive poverty reduction and the generation of jobs in a more diverse economy. The key messages that emerge across the chapters of this book follow: 1. Recognition. Governments should recognize the role that women play in trade and ensure this is communicated to officials at all levels. A visit to the border from the highest levels of government and a discussion with women traders can provide a very clear signal of the importance that the country places on the activities of women traders, both to the traders and to the officials who regulate the border. Similarly, discussions in many African countries around exports and the country’s export strategy should reflect the critical role that women farmers play in producing exportable goods and that a substantial increase in agricultural exports would follow from addressing the constraints that limit the access of women to productive land, to better inputs, to knowledge and to key networks. 2. Openness and transparency. Government should ensure that the rules and regulations governing trade are clear, transparent and widely avail- able at the border. Clarity on the documents required, appropriate fees and taxes, and the rules governing competition for goods, services and 4 Women and Trade in Africa: Realizing the Potential people crossing borders is important for large firms. But predictable rules and regulations are even more important for small traders in developing countries working with very limited margins. The majority of these trad- ers in Africa are women working in the informal sector. 3. Simplify documents and regulatory requirements where possible. Given time and mobility constraints as a result of household responsibilities women are likely to benefit substantially from efforts to streamline the number and nature of documents that have to be submitted and regulations that have to be satisfied. Allowing documents to be obtained and submitted close to where women traders reside is of particular importance. 4. Design interventions to develop trade in ways that ensure that women benefit. Governments and donors are making concerted efforts to facil- itate trade, to increase productivity in export-oriented sectors, and to improve competitiveness. But these need to be better targeted to ensure that women who participate in trade are reached by these interventions and that it is not just men who benefit. For example, programs that support improved access to information for those who participate in existing trade networks will miss women traders and entrepreneurs who are often excluded from such networks. Similarly, vocational training in export-oriented activities must recognize the barriers that women face in attaining higher-level positions, such as safari guides in the tourism sector. 5. Help women address the risks they face in their trade-related activities given that they are typically more risk averse than men and respond to risk in different ways. This reflects the particular challenges that women face in participating in trade and managing enterprises. Because of these risks, small women traders appear more likely to operate in the infor- mal economy. Women farmers are less able to invest in the techniques and inputs that would raise output. Women entrepreneurs may be more inclined to spread risks across different activities rather than focus on the growth of a single business. This in turn makes it more difficult for women-owned firms to participate in trade, where size can be an impor- tant factor in allowing firms to enter overseas markets. While additional analysis is required, improving the access of women-owned firms to angel and venture capital investments could help female-owned firms spread risk across partners rather than across sectors, allowing them to grow and access foreign markets. Introduction 5 Women and Cross-Border Trade Poor women traders cross borders throughout Africa every day and make a major economic contribution to the continent. Estimates suggest that infor- mal cross-border trade contributes substantially to the economies of many African countries and is a source of income for about 43 percent of the total African population (Afrika and Ajumbo 2012). In southern Africa, informal cross-border trade is thought to amount to as much as 30–40 percent of total recorded formal trade between countries in the region, entailing some US$20 billion per year (SARDC 2008). This amounts to almost half of total development assistance to Sub-Saharan Africa as a whole. At one border post alone between Malawi and Zambia, informal trade may amount to almost US$3 million per month (Njiwa, et al. 2011). A range of studies throughout the continent confirms that the majority of informal cross-border traders are women.3 These traders play a key role in food security, bringing basic food products from areas where they are relatively cheap to areas where they are in short supply. The incomes they earn from these activities are critical to their households, often making the difference, for example, in whether children go to school or not. Small cross- border trade fosters shared economic growth photo: Salahaldeen Nadir / World Bank and interdepen- dence between populations and is especially important for those with a history of division and mistrust. According to surveys, informal cross-border trade in the Great Lakes region between the Democratic Republic of Congo (DRC) and Uganda and between south Sudan and Uganda is likely to be several times larger than officially recorded trade flows. The World Bank and other agencies, such as the United Nations Development Fund for Women (UNIFEM), as well as organizations such as International Alert,4 have been drawing attention to the conditions that women traders face. While the majority of traders are women, most of the officials who regulate the border are men. A lack of transparency and aware- ness by both traders and officials regarding the rules and regulations that are 6 Women and Trade in Africa: Realizing the Potential supposed to govern cross-border movements of goods and people provides fertile ground for the exploitation of the traders. The following comment made by the minister for East African Community affairs of Burundi makes clear the challenge: Women traders are the ones who face, every day, the problems of not knowing what their rights and obligations are in today’s EAC [East African Community]. Most of them haven’t a clue about their rights and obligations.5 Women traders often face poor conditions and harassment when cross- ing the border. Although most small traders are informal—that is, they are not formally registered as a business—many cross the border through official crossings and are processed by officials from customs, immigration, and other agencies. Indeed, in a recent survey of trade in the Great Lakes region, more than 80 percent of traders reported having to pay a bribe to cross the border. Worse, more than half had suffered from physical harassment and abuse, including beatings, verbal insults, stripping, sexual harassment, and even rape.6 Much of this abuse is unreported. Chapter 2, Barriers, Risks and Productive Potential for Small-Scale Traders in the Great Lakes Region, by María Elena Garcia Moria and Sabrina Roshan revisits traders at the borders of the DRC and Rwanda and reports the results of a more detailed survey of traders and officials. The chapter helps deepen an understanding of the socioeconomic background of both the traders and the officials at the border. This survey of more than 600 trad- ers confirms the importance of women in this trade, with 90 percent of the respondents being women. The households of these traders are larger than the national average, and cross-border trading activities provide the main source of family income for three out of four traders. The chapter looks at a number of indicators of the welfare of traders’ households. It finds that on measures such as the quality of the dwelling, access to electricity, type of cooking fuel used, and ownership of durable goods, the households of cross-border traders are just as well off as the aver- age urban household that is used as a comparator. Hence, trading activities are critical in enabling households in border areas to attain the levels of welfare enjoyed by the typical household elsewhere in the country. The survey confirms that food products are by far the most common items carried by small cross-border traders in the Great Lakes region. Also, most of the traders acquired the necessary start-up capital informally from their own savings or from friends and family. Very few traders have a bank account. Significantly, but perhaps surprisingly, almost 60 percent of the trad- ers report that they have registered with the government. Clearly, the traders Introduction 7 photo: Paul Brenton are not seeking to hide from or evade government agencies. The border offi- cials who were interviewed, however, typically see small traders as undertak- ing illegal activities. Many traders were unwilling to answer survey questions about sexual and gender-based violence (SGBV), perhaps due to lack of trust in the inter- viewer or fear of reprisals. Nevertheless, a number of traders reported having been subject to SGBV incidents during the previous 30-day period, and it is clear that traders do not perceive any mechanisms for disputing or seeking recourse from the actions and behavior of officials. The chapter also reports interesting results on the socioeconomic back- ground and conditions of the officials who regulate the border. The study finds that they are quite different from the traders in this regard and con- cludes that the officials and traders come from quite different social classes. The officials are older and more educated, with better-quality dwellings, and their monthly consumption expenditures are almost double those of the trad- ers. The interviews with the officials also suggest they have quite different perceptions from the traders about conditions and behavior at the border. Finally, the results presented in this chapter provide the baseline for an impact evaluation of interventions currently under way to provide train- ing and better information to traders on their rights and on the regulations governing trade and also to provide training to officials. By measuring the effects of reforms and training programs, the impact evaluation will pres- ent the government and development partners with evidence-based findings on the economic and social benefits of removing information barriers and reducing corruption. Women who are more aware of their rights may have a greater voice in the household and in society and, as a result, may be less 8 Women and Trade in Africa: Realizing the Potential photo: Africa Renewal susceptible to SGBV in the future. Mechanisms that enable traders to report violations of their rights and that ensure that authorities will be penalized for infractions will create transparent and effective social accountability at the border and empower individuals to control and curb violations. Such changes will have substantial positive impacts on the economic potential of trade at the border. In chapter 3, Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria, by Louis Njie Ndumbe looks at the participation of women, and the constraints they face, in the export value chain between Cameroon and Nigeria of eru, a high-value nonwood forest product. Women are heavily involved in eru trade as both harvesters and small-scale traders. A key feature that makes eru trading an attractive activity is that very little start-up capital is required relative to other activities and that those with little formal education can effectively participate. Based on information from focus groups of women traders and harvesters, the study identifies a number of gender-specific constraints. One group of challenges relates to attitudes in the home that limit the ability of women to make their own decisions about their trade activities, the reality of domestic violence as a result of time spent trading, and time- related problems arising from household responsibilities. Lack of women’s legal rights to own land limits their access to the credit that could help them develop their businesses and reduces their opportunities for the training that could improve their productivity and lead to greater returns. The available training tends to be concentrated in male-dominated activities, and women’s Introduction 9 household roles reduce the amount of time they have available to participate. Responsibilities in the home impede women’s access to contacts in the mar- ket, to role models, and to social interaction, thus narrowing their interaction with the “business culture” that serves as the main training ground for trade. Another set of constraints arises from the rules and regulations govern- ing trade and the way they are (or are not) implemented by officials. Traders report harassment from government officials, and at times from buyers, which cause delays and losses given the perishable nature of their products. Women may also encounter particular difficulties in acquiring administra- tive trade and transport documents, in bearing the costs of roadblocks and other barriers that raise transport costs, and in facing the market power of buyers in the Nigerian market, which pushes down returns to Cameroonian exporters. The chapter suggests key steps that can be taken to address the obstacles to women’s achieving their potential as small entrepreneurial traders in west- ern Cameroon. In general, the government needs to formulate and implement policies that promote gender equality and to initiate a process that gradually alleviates the conditions in the home that hamper the participation of women in trade. To support trade in eru directly, the chapter proposes removing road blocks and other physical barriers to trade along key export corridors, reform- ing and streamlining export permits, providing training and capacity building for women exporters, creating a market information system, and assisting organizations that help women traders them improve their bargaining power and their ability to generate economies of scale. These studies confirm that open access to information on trade rules and regulations is essential to small women traders; and while progress is being made, much more needs to be done. In response, the World Bank has photo: Josanna Lewin/The Advocacy Project 10 Women and Trade in Africa: Realizing the Potential proposed that a traders’ charter be clearly posted at borders (see box 1.1). This document enshrines a basic set of rights and obligations for traders and officials that the international community could support through training of officials and carefully defined programs to encourage behavioral change at the border. The traders’ charter would be a first step toward providing trans- parent, predictable, and safe procedures. The issue of how to change the behavior of traders, their families, officials, and senior members of the government is discussed in more detail in chapter 4, Women Cross-Border Traders, Challenges and Behavior Change Communications, by Susan Ityavyar. The chapter highlights the need for a change in attitude toward informal cross-border women traders both from the top down and from the bottom up. While women traders make a substantial contribution to the welfare of their families and the well being of their com- munities, there is a general lack of recognition of their role as small entre- preneurs or of the economic benefits they bring. The government often does little to recognize small-scale traders as genuine economic actors, perpetuat- ing the negative stereotypes of women traders as “petite hustlers” held by many border officials. Husbands of traders often do not fully value the eco- nomic contributions of the trading activities of their spouses. In consequence, women often face gender-based challenges, including sexual harassment and abuse, which create barriers to doing business effectively. The chapter looks at how behavioural change has been achieved in other development areas and how such lessons can inform strategies for women traders. The chapter discusses how important it is for women traders to have a clear understanding of their rights, for example, through clear publication of the Charter for Cross-Border Traders, but also the need for safe and effective means of seeking formal redress if they are not treated properly. The bor- der information centers established by some regional communities, such as COMESA (Common Market for Eastern and Southern Africa), could also house professionals with whom traders could discuss complaints about the behavior of officials at the border. Thus, in addition to providing information about trade opportunities and the relevant trade policies at the border, these infor- mation centers could also provide advice on mediation and judicial recourse. The chapter discusses the key role of information and communication technology—together with the media, such as the press—in increasing aware- ness and providing access to redress mechanisms for women traders. It looks at examples from elsewhere where information technology has provided a platform for reporting and documenting violence and harassment against women. The chapter concludes by identifying a number of steps that can be taken and issues that can be addressed that would improve monitoring at the border, including the use of cameras, help lines, and mediation services. Introduction 11 Box 1.1: Charter for Cross-Border Traders Basic rights and obligations for traders and officials at the border 1. All individuals shall be able to cross the border without verbal or physical abuse or harassment, including but not limited to sexual and gender-based violence. 2. Traders shall be processed at the border in an efficient and timely man- ner without discrimination. A receipt must be provided to the trader for any payment made and the payment properly recorded. 3. Only officials of the approved agencies shall be present at the border, and all border officials shall wear uniforms or identification badges that indicate their respective agency. 4. Physical checks of traders must be recorded with the reason and outcome provided. Female traders have the right to receive a physical check by female officials in a private but regulated and accountable environment. 5. All duties, fees, and taxes and the basis for their calculation shall be publicly available at the border. Any change to duties, fees, and taxes must be publicly announced at the border, with reasonable time allowed for traders to prepare, before their application. No unpub- lished fees or charges shall be demanded at the border. 6. Documentary requirements shall be clearly stated and publicly avail- able at the border. Any change in required documentation must be publicly announced at the border with reasonable time for traders to prepare before implementation. Simplified procedures should be applied to small traders. 7. Traders should be aware of their rights and obligations when cross- ing the border. Traders must present required documentation and pay appropriate duties at the border and obtain a receipt for any payments made to an official. Traders shall not attempt to bribe any official to avoid payment of duties or to obtain preferential treatment in any way, including avoiding queues. (Box 1.1: continued on the next page) 12 Wome n a n d Tra d e i n Af r i c a : R e a li z i ng t he Potenti a l (Box 1.1: continued from the previous page) With the support of the international community, governments should commit to the following: 8. By [agreed time], these basic rights and obligations governing cross- border movement of goods and people shall be clearly stated in the local language and visibly apparent at all border crossings. 9. By [agreed time] at every border post at least one agent shall have received gender awareness training; all senior officials at the border shall have received gender awareness training by [agreed time]; 50 per- cent of officials at any border post shall have received gender aware- ness training by [agreed time]. 10. At all border posts, traders shall have recourse to an independent and confidential mechanism for registering violation of any of these basic rights. Female traders must be able to register the violation of any basic rights with a female staff. 11. Strict disciplinary measures shall be taken against officials found to have violated the rights of a trader. 12. Support organizations of informal cross-border traders shall be allowed to disseminate information on these rights and obligations and to deliver advice and information to enhance the capacities of the traders. 13. Governments shall continue to improve the quality of infrastructure at all border crossings to provide an open and safe environment for traders, with attention to the specific needs of women traders, and appropriate facilities in which officials may undertake their work. 14. Governments shall improve the quality of data on small traders col- lected at all border posts, including the number passing through the border each day and the nature of the goods carried. Since the costs of satisfying rules and regulations governing trade may be particularly burdensome for women, more women traders may be pushed into the informal sector. Why is it that women traders dominate the informal sector and are much less visible as formal exporters? While female informal traders face many challenges in crossing borders, for many the alternative of being a formal trader is currently not realistic. Women in Africa continue to lack access to credit, markets, and formal employment opportunities. Findings from many surveys and interviews suggest that the majority of Introduction 13 female traders are breadwinners as a result of being widowed or divorced or because of the unemployment of their partner. Many indicated that they got involved in informal cross-border trade to supplement their husbands’ salaries, which were insufficient to cover living expenses.7 Informal cross-border trade provides a livelihood for millions of families in Africa, often in very difficult conditions; it reaches markets and consumers that are underserved, if they are served at all, by formal markets and plays a critical role in food security throughout the continent. Nonetheless, there are a number of sound economic and social reasons why a strategy for the pro- gressive formalization of this trade will bring benefits to both traders and the wider economy. Lesser and Moise-Leeman (2009) discuss how governments can benefit from an increasing shift toward formal trade through increased revenues, better governance of the border, improved compliance with health and safety standards, and greater control over cross-border movement of pests and diseases and more accurate data from the border to support better policy decisions. For traders, a move toward formality would entail a shift away from a survival economy toward one in which longer-term decisions can be made about scale and employment. Most important is that operating with a degree of formality should improve access to finance, information, and advice. Surveys confirm that the vast majority of women traders would like to grow and develop their business away from the informal sector. However, in addition to the lack of access to the capital necessary for reaching the scale to survive in the formal sector, the requirements for formal trading are likely to be too onerous. An often-overlooked factor is that women may be more con- strained than men in satisfying the documentary and other requirements that were typically designed with little heed to the realities that women traders face in Africa. Regulatory requirements for small businesses typically entail frequent trips to relevant government offices and either the use of influence or informal payments. For example, household commitments mean that for many women obtaining an export permit from a far-away city where bureau- crats are located is simply not feasible. Obtaining a certificate of origin to be granted the zero duties that should apply to goods traded within regional eco- nomic communities is often too difficult or too costly for small traders. There is some evidence that women can find it difficult to obtain required docu- ments from officials without the assistance or presence of their husbands or male relatives. Simplifying rules and regulations to reflect the constraints that women face will foster trade and regional integration. It will provide one step toward an environment in which small women traders can evolve into the formal sector. Facilitating a transition by which the small women entrepreneurs 14 Women and Trade in Africa: Realizing the Potential crossing Africa’s borders today can become medium and large firms tomor- row would be a major contribution to growth, job creation, and poverty reduction. Conclusions from a range of studies, including the PLAGE (Policy Leadership and Advocacy for Gender Equality) project in west Africa, suggest that governments should do the following: streamline the system of import and export registration and ensure that all traders are able to register with minimal time and cost; assess whether export and import permits are essen- tial for meeting public policy objectives and whether they can be removed or, if necessary, whether compliance could be made simple; and simplify procedures at the border and ensure that the relevant rules are well known by traders and fairly applied by officials. General trade facilitation measures aimed at reducing the complexity, time, and cost of formal trade procedures have been designed in Africa at the regional level (for example, the single customs document in COMESA) or bilaterally (such as the pilot one-stop border post between Zambia and Zimbabwe at Chirundu). However, the characteristics of most small-scale cross-border traders, including their limited resources and the very low vol- ume of their transactions, require targeted facilitation and support measures. COMESA has sought to address this issue through its simplified trading regime (STR). The STR is a pilot scheme designed to overcome problems associated with small traders who need to prove that goods have the origi- nating status that allows exemption from import duties under the COMESA customs union. It was developed for small consignments, currently defined as being below US$1,000. For these, a simplified customs document can be filled out by traders, which is accepted by customs instead of the full docu- ment that must be filled out by a customs clearance agent. In addition, for certain products the normal proof of origin required at a COMESA border post is relaxed, and a simplified certificate of origin can be used instead, although goods must still meet COMESA origin requirements. Ordinary COMESA origin certificates are usually issued in capital cities and must be stamped by the customs authorities. At the border, a trader must normally present the stamped certificate of origin together with an invoice and the customs document declaring the goods to the customs officer in the importing country. Under the STR, governments agree bilaterally on a list of products that do not require ordinary origin certificates for small consign- ments but can instead use the simplified certificate of origin. The lists of eli- gible products are supposed to be displayed at the relevant border posts, and the simplified certificates of origin can also be obtained there and are signed by the customs officer. If the goods do not appear on the common list, then a normal certificate of origin must be obtained and certified. Introduction 15 photo: Curt Carnemark / World Bank Beyond the simplified customs document and simplified certificate of origin, the STR makes no other changes to the border requirements for trade in small consignments. These remain the same and include travel documen- tation for immigration as well as licenses and certificates for agricultural produce. The latter must usually be obtained from the offices of ministries of agriculture and often involve export permits and phytosanitary certificates stating that the products are free from disease. In addition, the trader will have to pay other applicable taxes such as value-added tax, excise duty, and presumptive tax. Moreover, implementation of the STR on the ground by customs officials is often incomplete or perceived as arbitrary by traders.8 COMESA has also supported the establishment of trade information desks (TIDs) at each STR implementing border to (1) help small traders use the STR and deal with other formalities; (2) disseminate information on trade procedures and regional initiatives; and (3) act as a liaison between traders and border authorities in daily operations. TIDs have generally been found to provide useful services to traders where they exist, but there are still issues concerning the sustainability of their funding and staffing (Njiwa and Oldham 2011). While the STR is an important initiative, its use by small traders remains limited, for several reasons, with lessons for elsewhere in Africa: • The cost of satisfying formal trade procedures remains high, and the additional fees applied for certain procedures (including sanitary and phytosanitary certification and data entry into the computerized customs 16 Women and Trade in Africa: Realizing the Potential recording system) tend to be disproportionately high for small traders, given the very small margins they are working with. For instance, Malawi and Zambia still charge US$30 and US$10 respectively for each STR transaction processed in ASYCUDA (Automated System for Customs Data), despite a ministerial recommendation that the fee be pegged at US$1 (COMESA 2011; Njiwa 2012).9 • Despite some simplification, trade procedures remain complex for small traders. The number, complexity, length, and cost of trade procedures are still a major obstacle for actors who have limited capacities and frequently cross the border. The centralization of the issuance of export and import permits and SPS certification for agricultural products makes them costly for all traders but prohibitively so for the smaller ones, especially for women. Access to clear information on export and import procedures continues to be a challenge for small traders, especially if they are inexperienced or illiterate. In addition, opaque and arbitrary application of rules often means that small traders are not able to claim the advantages they are entitled to, even when they are aware of them. In this case, it is much easier and less risky to pay a bribe and pass through than to argue with officials. • Hostility toward small traders continues, and officials’ attitudes are an important factor in deciding where to cross the border. For example, Njiwa et al. (2011) observed that many more traders pass through the official border post at Mwami in Zambia than at Mchinji on the Malawian side, meaning that they use alternative routes to avoid the less trader-friendly Malawian border post. This chapter provides a number of guidelines for the reinvigoration of the COMESA STR and for simplifying trade regimes for small traders in other regions of Africa: further simplify trade procedures using a gender lens to identify those that affect women more heavily; deal with the issue of trans- parency of rules and regulations and relations between traders and officials by adopting the Charter for Cross-Border Traders and implementing strategies to change behavior at the borders; and provide a better foundation for these changes through support for traders’ associations and training of officials. Women as Producers of Tradable Agricultural Goods and Services Women are deeply involved in the production of agricultural products in Africa. The latest estimates of the International Labour Organization (ILO) show that in 2012 about 62.5 percent of women were employed in agriculture Introduction 17 compared to 61.8 of men (ILO 2012). Yet the potential for women farmers to contribute to the expansion of traditional agricultural exports appears to have been undermined by their limited access to key production inputs relative to male farmers (World Bank 2011). Evidence summarized by the 2012 World Development Report suggests that female farmers cannot seize the available trade opportunities because they lack critical inputs and are excluded from trade-related networks. For example, in Meru, Kenya, more than 90 percent of export contracts were issued to male household members, probably due to women’s weaker property rights to land (World Bank 2011, 270). Similarly, due to lower education levels, female producers were more constrained than their male counterparts in accessing international markets in Mozambique and Samoa (World Bank 2011, 271). This factor further limits overall produc- tion. FAO (2011), for example, estimates that closing the yield gap by provid- ing women with resources equal to those available to men would increase agricultural output in developing countries by 2.5 to 4 percent (FAO 2011). In chapter 5, The Gender Dimension of Uganda’s Cotton Sector, by John Baffes and Laura Maratou-Kolias offer a clear example of this potential by looking at the cotton industry in Uganda. Cotton is a key export product for the country, its third-largest export crop. Despite several reforms, production levels are less than a third of what they were before the 1970s, raising con- cerns that the sector still lags behind its potential. Part of this gap appears to be gender related; in 2009, male farmers produced 23 percent more cotton per unit of land than women did. The chapter seeks to explain the observed gender gap in cotton productivity. Although land characteris- tics such as soil quality, soil type, plot slope, land tenure system, and the way in which land was acquired have an important impact on yields, they do not explain the gender gap in productivity. Differences between men and women in their knowledge of farming practices appear to be an important factor in gender gaps in productivity. Social network–based training photo: Arne Hoel / The World Bank programs had a greater impact 18 Women and Trade in Africa: Realizing the Potential on yields for the poorest subsistence farmers, which include most female farmers, than the conventional training typically provided to larger and more productive farmers. The chapter also draws attention to differences between male and female producers in the inputs used in growing cotton in Uganda. Men are much more likely than women to use pesticides, although there is considerable variation in the difference across regions, reflecting historical and political factors that have favored certain regions over others. In addition to differences between male and female farmers in produc- tion inputs, including knowledge and use of pesticides, the chapter also reveals that men receive higher prices than female farmers. In 2009, male farmers received on average 7 percent higher prices than women, although the difference was much larger in some regions. This difference in prices when combined with lower yields and smaller areas under cultivation leads to much lower incomes for the households of female cotton farmers than for the households of male farmers. One reason for the lower prices is that women sell their cotton in fewer installments than men because of their greater requirement for cash to cover immediate needs and their lack of access to storage facilities. Male farmers are more likely to know their buyer in advance than female cotton farmers and to have arranged the sale before going to the market. Women are less likely to be aware of the price of cotton before selling their output. The chapter concludes by identifying two key policy interventions that would enhance the productivity of female cotton farmers in Uganda and hence improve their ability to contribute to the expansion of agricultural exports: increasing the flow of information on agricultural techniques that reaches women farmers and strengthening property rights so that women can undertake productive investments in their plots. While the majority of women in Sub-Saharan Africa are employed in the agricultural sector, the share of women employed in the service sector has dramatically increased in recent decades. This share rose by 7.6 percent- age points between 1992 and 2012 (see ILO 2012). Based on the experience observed in other regions, this share will likely increase further as service opportunities in Africa expand. In south Asia, Ghani and Kharas (2010) docu- ment that the service sector grew from less than 40 percent of gross domestic product (GDP) in 1980 to more than 50 percent of GDP in 2005. This increase appears to have especially benefited women. For example, women in India and Pakistan entered the service sector jobs at a much faster rate than those in manufacturing or agriculture (Ghani and Kharas 2010). Similarly, in Delhi and Mumbai, call centers now employ more than 1 million people, mostly women (World Bank 2011). Introduction 19 Antoine Coste and Nora Dihel, in chapter 6, Services Trade and Gender, argue that trade in services offers new opportunities for women in Africa. Women increasingly participate in services trade as employees in tourism and back-office processing, for example, and in professional services such as accountancy; and women are more likely to have ownership of service firms than of manufacturing companies. Liberalization of services trade can there- fore augment employment and income opportunities for women, but very few studies have examined the impact of trade in services on labor markets, businesses, and incomes through a gender lens. The chapter contributes to the discussion on whether services and services trade provide a ladder to more gender equality or whether these sectors continue to restrict women to low-wage, low-productivity jobs but in a context in which acute data and information gaps remain, especially in Africa. The chapter confirms that in Africa the share of female employment, female ownership, and female top managers is highest in the hotels and res- taurants sector and in the wholesale and retail trade sector, which suggests a concentration of women in low- and mid-skilled occupations in services, a feature that appears to characterize employment at the global level. The large proportion of women employed in the more tradable services sectors such as tourism (hotels and restaurants), wholesale and retail trade, or infor- mation technology services suggests that trade can have significant gender impacts; however, the outcomes achieved by women depend on the eco- nomic, social, and political conditions that shape their ability to participate in such activities. Analysis of data from the enterprise surveys undertaken throughout a large number of developing countries shows that, in general, firms with female management tend to be smaller than firms with male managers in both manufacturing and services, with services firms smaller than manu- facturing firms in both Africa and elsewhere. It is interesting that the share of female full-time employees is substantially higher in firms managed by women than in those managed by men—whether they are in manufacturing or in services, in Africa or elsewhere. The chapter also provides information from a detailed survey of the pro- viders of professional services in eastern and southern Africa and finds con- sistent evidence of gender imbalances at the managerial level of professional firms in all the examined African countries. Women represent only 6 percent of top senior positions and 12 percent of second-most-senior positions of professional services firms. Fewer firms with female top managers tend to engage in export activities than firms with male top managers. Nevertheless, these figures are broadly comparable to the statistics of women in top management in developed countries; women made up about 12.5 percent of 20 Women and Trade in Africa: Realizing the Potential the members of the corporate boards of FTSE-100 companies. Furthermore, the proportion of female fellows of the Chartered Institute of Management Accountants in countries such as Nigeria and Zambia is higher than in a number of countries in the Organisation for Economic Co-operation and Development. These findings suggest that the overall conditions that enable the participation of women in professional activities may be as favorable in Africa as in more developed countries. Hence, trade and regulatory reforms in professional services in Africa may have an important positive effect on women and important long-term development implications. In chapter 7, Gender in the Tourism Industry: The Case of Kenya, by Michelle Christian, Elisa Gamberoni, and José Guilherme Reis analyze the obstacles faced by women at various stages along the Kenyan tourism value chain. Because tourism is a key export sector in many African countries, it can be a significant source of new jobs. It also has important links to the rest of the economy and so can contribute to economic diversification. In 2011, the sector accounted for 12.5 percent of Kenya’s GDP (Kenya National Bureau of Statistics 2012) and supported 313,500 direct jobs. Its major markets are France, Germany, Italy, the United Kingdom, and the United States, with sup- ply mainly concentrated in two main product areas—safaris and beaches. The authors document that the more profitable segments and activities of the value chain are generally dominated by men. First, most of the tour opera- tors are male owned, particularly the most profitable tour operators. Second, in the excursion segment, men generally lead expeditions and are the safari drivers while women do not generally operate in camps, in line with a gender bias in the expected roles. Across ethnic-tribal groups, both male and female Maasai are at a disadvantage for employment at tourist camps due to their extreme lack of education. However, local hiring conditions, when built into land leasing contracts with the Maasai, provide jobs mainly for men. Third, in the accommodation sector, women appear more likely to be casual workers. The authors conclude by discussing promising interventions and pro- grams that can contribute to eliminating these disparities, such as programs that promote best labor practices and guidelines for tour operators. Brand sensitivity and the push toward responsible tourism create incentives for firms to follow these practices. In addition, the tourism sector is character- ized by a wide range of certifications, related in particular to social and envi- ronmental sustainability. These certifications could be expanded to include gender equity and be appropriately monitored. For example, the Better Work Program, sponsored by the ILO and the International Finance Corporation, seeks to improve compliance with international labor standards and national laws, while promoting business competitiveness in the apparel sector. A similar program could be applied to tourism. Dedicated programs that train Introduction 21 women in male-dominated occupations can help reduce the gender bias observed in certain jobs, such as those linked to the safari excursions. Finally, promoting new itineraries that support female commercial activities, such as artisanal markets, can also contribute to tourism diversification in Kenya. Women-Owned Firms and Trade The final section looks at gender gaps in export participation and volume among female- and male-owned firms. The trade literature has long empha- sized that exporters are more likely to be larger and more productive than firms that serve only the domestic market. Further studies suggest that there is a negative correlation between firm size and gender ownership (World Bank 2011). The question is whether, after controlling for size, female-owned firms face additional barriers. The final part of this book contains two chap- ters that analyze detailed data for South Africa and Tanzania to assess differ- ences between male- and female-owned exporting firms. In chapter 8, Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa by Thomas Bossuroy, Francisco Campos, Aidan Coville Markus Goldstein, Gareth Roberts, and Sandra Sequeira use informa- tion from a recent survey of more than 2,400 small and medium businesses in the Kwazulu-Natal Province of South Africa to examine the determinants of exporting and whether these differ between male- and female-owned enterprises. The data are also used to look at the determinants of the size of the enterprise, in a context in which size is typically seen as a determinant of whether a firm is an exporter. Interestingly, no significant difference is found between the likelihood of female- and male-owned firms being exporters after controlling for the key firm-level characteristics associated with exporting, such as size, access to finance, the owner’s skills and experience, and the sector of operation. Nevertheless, the analysis does identify a number of differences between the characteristics of male- and female-owned exporting firms, suggesting that women might face specific barriers and that interventions to promote exports will likely have different impacts on male- and female-owned firms. First, male- and female-owned businesses differ in their propensity to export according to the sector of business specialization: female-owned firms are much less likely to export than their male-owned counterparts when they operate across multiple sectors. It is postulated that female owners who oper- ate across many sectors may not have a strong business case in a given area of expertise and that they engage in business because they lack the skills to hold a salaried job. Some evidence also implies that this diversification strat- egy could be associated with risk aversion.10 Male-owned firms are associated 22 Women and Trade in Africa: Realizing the Potential photo: Josanna Lewin/The Advocacy Project more with industrial sectors, in particular agriculture and manufacturing, where firms in general are more likely to export. Women-owned businesses export in these sectors but are also more likely to be found in the general services sector. While networking activities such as finding new custom- ers online are strongly correlated with the propensity to export across the sample, they are more strongly correlated for women. In addition, the female- and male-owned businesses identify new customers in different ways. For female-owned firms, getting customers through referrals—not a particularly sophisticated way of obtaining clients—is negatively associated with the propensity to export. There is also some evidence that men are more likely to operate through traditional networks of contacts, while women need to find alternative mechanisms of reaching new markets, particularly in sectors where they do not traditionally operate. Finally, female exporters are also more educated than their male coun- terparts, with all of them having completed at least secondary education. The chapter identifies a number of potential interventions to address the barriers to the growth of women-owned small and medium enterprises in South Africa and their capacity to access foreign markets. These include: (1) exploring the potential of “angel” and venture capital investments to assist women-owned firms in spreading risk across partners rather than across sectors; (2) inject- ing new skills and offering more opportunity for joint decision making; (3) increasing potential networking possibilities through partnerships with more experienced and established entrepreneurs; and (4) supporting increased use of technology to circumvent entrenched networks and allow women-owned small and medium enterprises to reach out to new client bases and promote business growth and exports. The Internet and cell phones offer media Introduction 23 through which businesses can break into new markets without needing to rely on traditional networks. Testing these options through experimental pilot interventions will help shed light on the most important barriers faced by women-owned firms and contribute to finding sustainable solutions to pro- moting the development of small businesses in South Africa. In the final chapter, Trade and Gender in Tanzania: What matters — Participation or Outcomes by Josaphat Kweka and Mahjabeen Haji, conduct an analysis of gender differences in export participation among formal and larger businesses. The authors report the results of a statistical analysis of industrial survey data from Tanzania that provide information on the gender of the owner of the firm as well as other firm characteristics. After controlling for relevant characteristics, they find that firms owned by female entrepre- neurs have export values comparable to those of their male-owned coun- terparts. The analysis shows that for the full sample of firms there is little difference in profitability between male- and female-owned firms. However, for the subsample of firms that export, it is apparent that female-owned firms tend to be associated with less profitability than male-owned firms. Looking at the destination markets for exports, they find that female-led businesses prefer to export to neighboring countries unlike their male peers who export farther abroad. This finding might suggest that female-owned firms face more constraints in trading beyond the Tanzanian border than male-owned firms. It also shows the importance of regional trade in Africa, not only for informal cross-border traders but also for formal female-owned export firms. photo: Nathan Kone / World Bank 24 Women and Trade in Africa: Realizing the Potential Finally, the aim of this book is to make available new analysis on the participation of women in trade in Africa to a wide audience. It highlights the key role that women will play in achieving Africa’s potential in trade. In addi- tion to raising the profile of this public policy issue, we also hope that it will encourage more research and analysis over a wider range of African countries and so extend the knowledge base. Notes 1. World Bank (2012) discusses the opportunities for enhanced trade in Africa and the barriers that prevent that potential from being fully exploited. 2. World Bank (2012) shows that this potential arises from rapidly increasing demand, especially in cities, coupled with enormous potential for increasing production by raising yields, which are typically between one-third and a half of what farmers in other developing countries are achieving, and increasing areas under cultivation. 3. For example, Brenton and Isik (2012) find that 80 percent of cross-border traders in the Great Lakes region are women, while Njiwa, et al. (2011) report that 75 percent of informal cross-border traders between Malawi and Zambia are women, which is the same as an overall figure for Southern Africa that has been attributed to the United Nations Development Fund for Women. 4. Titeca, Kristof and Célestin Kimanuka. 2012. “Walking in the Dark: Informal Cross-border Trade in the Great Lakes Region.” International Alert, London. 5. Taken from http://www.trademarkea.com/ burundis-women-traders-hit-by-eac-propaganda-and-like-it/. 6. See Brenton and Isik (2012). Some of the challenges for traders due to poor governance are shown in the film Les Petites Barrieres, which is available at www.worldbank/afr/trade 7. Interestingly, SARDC (2008) suggests that men are less likely to be involved in informal cross-border trade because they are less able to with- stand the hassles and harassment from officials at the border. 8. For instance, some traders complain that the common list of products eligible to the STR is “disregarded” by Zimbabwean officials at Livingstone/ Victoria Falls and claim that they have been refused STR treatment for con- signments that should have qualified (see Coste 2013). 9. Coste (2013). Introduction 25 10. Croson and Gneezy (2009) review experimental evidence on differences between men’s and women’s preferences and found that women are more risk than men, with the exception of women at managerial levels. References Afrika, J.-G., and G. Ajumbo. 2012. “Informal Cross-Border Trade in Africa: Implications and Policy Recommendations.” African Development Bank Africa Economic Brief 3 (10): 1. Brenton, Paul, and Gozde Isik, eds. 2012. Defragmenting Africa: Deepening Regional Trade Integration in Goods and Services. Washington, DC: World Bank. Brown, Drusilla, Rajeev Dehejia, and Raymond Robertson. 2011. “Working Conditions and Factory Survival: Evidence from Better Factories Cambodia.” Better Work Discussion Paper 4, International Labour Organization, Geneva. ———. 2012. “Retrogression in Working Conditions: Evidence from Better Factories Cambodia.” Better Work Discussion Paper 6, International Labour Organization, Geneva. ———. Forthcoming. “Is There an Efficiency Case for International Labor Standards?” COMESA (2011) Report of the Meeting of Ministers on the Operations of the COMESA Simplified Trade Regime (STR) for Cross Border Traders. Document CS/IPPSD/REFORM/MIN-STR/I/7. Lusaka: COMESA Secretariat Coste, Antoine. 2013. “Informal Trade in Zambia.” Background paper pre- pared for World Bank Diagnostic Trade Integration Study on Zambia. (forthcoming). Croson Rachel, and Uri Gneezy. 2009. “Gender Differences in Preferences.” Journal of Economic Literature 47 (2): 448–74. FAO (Food and Agriculture Organization). 2011. “The State of Food and Agriculture: Women in Agriculture: Closing the Gender Gap for Development. Rome: Food and Agriculture Organization. Ghani, Ejaz and Homi Kharas. 2010. “The Service Revolution in South Asia: An Overview,” in: Ghani, Ejaz, ed., The Service Revolution in South Asia. (Oxford University Press). ILO (International Labour Office). 2012. Global Employment Trends for Women 2012. Geneva: ILO. Kenya National Bureau of Statistics. Kenya National Bureau of Statistics, 2012. Web (http://www.knbs.or.ke/) n.d. 26 Women and Trade in Africa: Realizing the Potential Lesser, C., and E. Moise-Leeman. 2009. “Informal Cross-Border Trade and Trade Facilitation Reform in Sub-Saharan Africa.” OECD Trade Policy Working Paper 86, Organisation for Economic Co-operation and Development, Paris. Njiwa, D. (2012), Informal Cross Border Trade: Challenges and Opportunities – A Case of COMESA and its STR Implementing Borders. Lusaka: COMESA Secretariat. Njiwa, D., T. Nthambi and J.Chirwa (2011), Reconnaissance Survey Report of Informal Cross-Border Trade at STR Implementing Borders of Zambia, Malawi and Zimbabwe. Unpublished report prepared by the CBT/ REFORM project team. Lusaka: COMESA Secretariat. Njiwa, D. and P. Oldham (2011) Analysis of the COMESA Simplified Trade Regime (STR) in Practice. Unpublished paper prepared by the CBT/ REFORM project team. Lusaka: COMESA Secretariat. SARDC (Southern Africa Research and Documentation Centre). 2008. “Optimizing Regional Integration in Southern Africa: Assessing Informal Cross Border Trade in SADC.” http://databases.sardc.net/books/ ICBTReport/index.php. Titeca, Kristof and Célestin Kimanuka. 2012. “Walking in the Dark: Informal Cross-border Trade in the Great Lakes Region.” International Alert, London. World Bank. 2011. World Development Report 2012: Gender Equality and Development. Washington, DC: World Bank. Barriers, Risks, and Productive Potential for Small-Scale Traders in the 27 Great Lakes Region 2  Barriers, Risks, and Productive Potential for Small-Scale Traders in the Great Lakes Region María Elena García Mora and Sabrina Roshan Introduction Evidence that crossing borders in Africa, especially for women, can be a costly and dangerous activity is increasing. This chapter reports the results of a World Bank Africa Region Gender Innovation Lab survey of traders that cross the border between the Democratic Republic of Congo (DRC) and Rwanda. It builds on previous work by Brenton et al.1 that obtained informa- tion about border-crossing conditions through a qualitative survey and focus group discussion at four regional border crossings: Uvira and Bujumbura (between DRC and Burundi), Kasindi and Mpondwe (between DRC and Uganda), Goma and Gisenyi (between DRC and Rwanda), and Bukavu and Cyangugu (between DRC and Rwanda). This study found that 85 percent of cross-border traders were female, with an average age of 32 (Brenton and Isik 2012). The majority traded food—cereals, pulses, fruits, and vegetables—in small quantities and trans- ported it mainly by head carrying. Start-up capital was less than US$50 and was obtained from informal sources of credit, particularly through family net- works. Cross-border trade served as the main source of income for 77 percent of the traders, and most of the traders were not part of formal trade networks or trader associations. The survey found that more than 85 percent of the traders reported having to pay a bribe to cross the border and that more than half reported being subjected to some form of physical harassment.2 Following on from this work, in August and September 2011 the World Bank’s Africa Region Gender Innovation Lab conducted a further survey in the towns of Goma and Bukavu on the border between DRC and Rwanda. Random sampling was used to collect information from 628 traders. 3 In addition, structured interviews were conducted with 66 border officials4 from the four main border agencies—customs, immigration, hygiene, and arms María Elena García Mora is social development specialist (LCSDE) at the World Bank. Sabrina Roshan is a consultant (AFTOS) at the World Bank. The authors thank Markus Goldstein (World Bank Africa Region Gender Practice Leader) and Kevin Croke (World Bank consultant AFTPM) for their invaluable comments and inputs. 28 Women and Trade in Africa: Realizing the Potential control. The survey was also followed up with focus group discussions with the traders in May 2012. The goal of the quantitative data collection exercise was to move beyond the basics of how much is traded, what is traded, and why it is traded to information about the socioeconomic background of the traders and officials and deeper information on the constraints the traders faced. This chapter discusses the main features of the survey results and seeks to improve our understanding of how and why sexual and gender-based violence (SGBV) and corruption take place across the border, what its affects are on small- scale trade for the victims, and what can be done about it. Thus, this follow-up quantitative study looks to broaden and deepen our understanding of these traders and the challenges they face at the border using a rigorous approach and, as a result, to support those in the government and develop- ment community in designing interventions with increasingly sizable and replicable impacts on the small, poor female traders. We now turn to the main findings of the data analysis. Socioeconomic Profile This section provides descriptive statistics and socioeconomic information about those sampled in the 2011 baseline survey of traders and border offi- cials. Ninety percent of small-scale traders were women, a finding consistent with the qualitative data collected by Brenton et al. (2011). Given that the project team was working closely with the Congolese government, we pur- posely oversampled Congolese traders, and thus 75 percent of traders in the sample lived in the DRC; 40 percent were from North Kivu Province and 35 percent from South Kivu Province. The remaining 25 percent of the sample photo: Arne Hoel / The World Bank Barriers, Risks, and Productive Potential for Small-Scale Traders in the 29 Great Lakes Region lived in Rwanda. The typical trader had been trading for 10 years and was 35 years old. The vast majority of traders lived with a spouse or partner, while 49 percent were married. Cross-border trade is the main source of family income for three out of every four of these traders. Based on just our sample, the livelihoods of approximately 3,400 people are dependent on trade.5 Women’s income from trading activities is of particular importance to households where the spouse in not employed and helps explain the high tolerance for the difficulties that they face in crossing the border. When traders were asked about their spouses’ main source of cash income, 29 percent had no regular cash-flow source, 20 percent had salaried employment, and 17 percent were generating earnings from nonfarm casual labor. Of the spouses who reported having an occupation in the previous 30 days, 18 were unpaid domestic workers, 15 percent were casual nonfarm workers, and 28 percent were in other indus- tries; 5 percent were unable to work due to illness. The households in the sample tended to be much larger than the national average: 7.4 members on average for the sample compared to the average household size in urban areas of 5.9 in DRC and 4.2 in Rwanda (USAID 2007; USAID 2010). 34 percent of the traders had finished primary school and had begun secondary school but not completed it. 10 percent had completed primary school. 27 percent had attended up to primary school but had not completed it, and 17 percent had had no educational instruction at all. Eight percent of traders were internally displaced people (of whom 87 percent live in DRC); and 1.6 percent of traders were refugees, of whom 70 percent reside in Rwanda. Indicators of Welfare of Traders’ Households We use the following measures as indicators of traders’ welfare: (1) quality of dwelling; (2) lighting source; (3) cooking fuel source; and (4) ownership of durable goods. To benchmark these measures, we compare data from trad- ers’ households with those from urban households in the same districts or provinces derived from the latest Demographic and Health Surveys (DHS) for DRC (2007) and Rwanda (2010). The DHS samples were restricted to urban households in North and South Kivu in DRC and to urban households in Western Province in Rwanda. Thus, this approach removes from the compari- son urban women living in the capital cities of Kigali and Kinshasa or other major cities outside the provinces of study. Table 2.1 summarizes the main findings. It suggests that the traders in our sample are doing just as well or perhaps slightly better than the average 30 Women and Trade in Africa: Realizing the Potential Table 2.1: Household Indicators: Comparison of Traders’ Families in the Sample with Families in Selected Regions of the Democratic Republic of Congo, 2011 percent South North South Rwanda North Kivu Rwanda Kivu Kivu DHS Kivi DHS DHS sample sample sample urban urban urban (%) (%) (%) (%) (%) (%) Dwelling quality: floor material Cement 30 33 44 27 29 43 Earth/sand 68 64 51 67 61 54 Access to electricity Yes 30 18 25 19 25 34 No 70 82 75 81 75 66 Source of cooking fuel Wood 14 6 9 75 29 34 Charcoal 80 90 84 18 71 60 Ownership of durable goods Mobile phone 63 54 61 50 44 46 TV 9 37 51 14 26 23 Radio 64 59 71 67 68 57 Bicycle 3 6 6 6 21 6 Source: World Bank Africa Region Gender Innovation Lab 2011–2012 survey. urban household in the same region. This finding suggests that the sample of cross-border traders, while quite poor, is still better off than the average urban household in the region. Trade Activities Data from the region demonstrate that women who own small businesses tend to use informal means (family, friends, and the like) to raise capital for their businesses. This has long-term impacts on the size, profitability, viability, and sustainability of their business activities. Our sample of trad- ers is representative of this norm among small, women-owned enterprises. Almost 36 percent of the traders attained the start-up capital for their busi- ness as a donation; 22 percent from their own savings; 13.5 percent from their spouse’s savings; and 27 percent from loans (of which 76 percent came from informal networks, including friends and family). Small, women-owned Barriers, Risks, and Productive Potential for Small-Scale Traders in the 31 Great Lakes Region businesses tend to be informal and to lack of access to formal bank accounts; the owners have no familiarity with formal tax applications or with raising relatively large-scale credit and make little use of other business development resources. Female traders face informational barriers related to business net- works and trade regulations. Only 20 percent of traders are part of a formal traders’ association. On average, Rwandans pay US$18 to join associations and Congolese traders pay US$8.6 The majority of traders trade in fruits and vegetables (table 2.2), which, according to the DRC tax code, when traded in small quantities should not be subject to import taxes. However, in DRC taxes at the border are administered in a haphazard and often opportunistic manner. Both traders and officials have little knowledge of regulations. In addition, the insecurity in the region leads to high price volatility for these goods. As a result, officials at the border often take advantage of this price and potential tariff variation, which increases traders’ vulnerability to offi- cials’ bad behavior at the border. Approximately 58 percent of traders are officially registered with the gov- ernment; 38 percent of traders in Rwanda, 70 percent in North Kivu, and 59 percent in South Kivu. Despite this significant percentage of official registra- tion, qualitative data from interviews with border officials reveal a common perception that these traders are not only conducting informal activities Table 2.2: Main Products Carried by Small Women Traders in the Democratic Republic of Congo and Rwanda percent DRC All Rwanda North Kivu South Kivu (%) (%) (%) (%) Fruits/vegetables 38.38 21.52 52.54 35.47 Fish 8.12 0.63 9.32 11.97 Goats 7.17 0.63 15.25 3.42 Chickens 3.98 1.27 5.08 4.70 Eggs 4.94 0.00 2.54 10.68 Flour 4.62 4.43 1.27 8.12 Plastics 4.46 0.63 3.39 8.12 Clothes/shoes 4.14 10.76 0.85 2. 99 Beans 3.50 10.13 1.69 0.85 Peanuts 3.18 1.27 2.54 5.13 Palm oil 1.91 4.43 2.12 0.00 Milk 1.75 1.27 2.12 1.71 Source: World Bank Africa Region Gender Innovation Lab 2011-2012 survey. 32 Women and Trade in Africa: Realizing the Potential but also participating in illegal business. This attitude can result in abusive behavior from officials at the border, with the traders lacking the power to counter with the information that would clarify their status as legally certi- fied small business owners. Furthermore, this attitude could also be prevent- ing the officials from seeing the potential and current benefits that this small cross-border trade brings to the region. Characteristics of Trading Activities Of the 628 traders in our sample, the majority, 87 percent, are importers. Of the exporters, most are from Rwanda. Most of exporters decide not to sell in the local market because of lower profit (26 percent) or too much competi- tion (23 percent); becaused they do not have market space (13 percent, all from Rwanda); because of low demand (5 percent); or because or they are trying to escape harassment at the local market (5 percent). In Rwanda, regulation prohibits street vendors from selling anything other than mobile phones or newspapers, which are permitted only in certain areas (Sommers 2012), which may help explain this dynamic. Traders normally purchase their goods in markets at the border (25 percent at the Gisenyi/Cyangugu border and 2.6 percent at the Goma/Bukavu border). While 21 per- cent buy their goods in a market village (19 percent near Gisenyi and 2 percent near Goma), 17 percent buy their goods in the city (8 percent in Gisenyi and 9 percent in Goma), 9 percent from wholesalers, less photo: Jonathan Ernst/World Bank Barriers, Risks, and Productive Potential for Small-Scale Traders in the 33 Great Lakes Region than 2 percent directly from a farmer group, and less than 1 percent directly from the farmers. Likewise, traders normally sell their goods in markets that are located directly across the border (59 percent across the DRC border in Goma or Bukavu), 6.5 percent in a market village near Goma, and 5 percent in a market village near Gisenyi; 4 percent sell goods door to door, 4 percent in the street, and only 3 percent at the border on the Rwandan side (Gisenyi/ Cyangugu). Less than 2 percent sell their goods in a shop or kiosk. The fact that these traders are not purchasing goods from farmers affects prices and indicates that trade happens through transactional middle men. Ninety-four percent of traders collect information about both purchase and sale prices by word of mouth, and 4 percent do so by calling suppliers. This suggests that greater use of mobile phones and short message service (or SMS) texting to share information about prices might increase efficiency and profits. One initial idea to improve conditions and reduce corruption was to subsidize use of customs brokers by traders. However, the baseline survey showed that only 13 percent of traders have used a customs broker and that among those that have not used one 82 percent believe it is not useful. These findings were influential in moving the project team away from this idea of a broker-based intervention and toward the training and information-based interventions. The average trader crosses the border 3.4 times a week, stays in the neighboring country for 3.5 hours, and requires 33 minutes to cross the bor- der. The majority of traders (55 percent) cross before 7 a.m., and up to 86 percent cross before 9 a.m. According to focus group discussions with traders, crossing the border during these windows offers advantages because the border officials are not always at their post, allowing traders to sometimes cross without paying fees, although it also reflects the early opening of markets. From the Rwandan side, most traders cross the border on foot or through head carriers7 (65 percent). Three percent use bicycles, 5 percent go in minibuses, 4.5 percent on moto scooters, and 4 percent in cars. From the DRC side, most traders cross by foot or through head carriers (68 percent), 6 percent go on bicycles, 6 percent in taxis, 5 percent on moto scooters, 4.6 percent in minibuses, and 1.6 percent in cars. Given the logistics in transporting the goods and the fact that few traders own bicycles, it is not surprising that most traders cross on foot. However, this leads to several vulnerabilities. The lack of storage and limited working capital require food traders to cross the border by foot several times a day, which has time-use impacts both inside and outside the household (that 34 Women and Trade in Africa: Realizing the Potential is, this time could be spent completing household chores, being produc- tive in other economic activities, caring for children, and so forth). While most traders cross the border only once a day on the day of their crossing, 57 traders (9 percent of all traders) cross more than once a day due to these constraints. More directly related to this study are the impacts that this mode of transportation has on increasing the vulnerability to sexual harassment and corruption. Trade Barriers The majority of traders refused to answer the set of questions related to avoid- ing difficulties, or holdups, at the border. This reluctance is most likely due to the sensitivity of the questions. Based on focus group discussions, there seems to be a general fear of reprisal from authorities. We assume that this fear also applies to the questions on avoiding difficulties: because traders may believe that authorities will have access to the information and because some of the avoidance activities are illicit, the traders may be wary of punishment. Second, there is likely a lack of trust in divulging this information to enumerators8 and to an international organization. As a result, the sample in this section is limited to 230 of the 628 traders. One-third of these traders believe that paying bribes, smuggling, and using connections with officials are the best ways for them to cross the border. Nine percent of these traders believe that knowledge of taxes and tariffs empowers traders to avoid difficulties with officials, while 12 percent think that using a customs broker would be helpful to them. Only 16 traders in the sample have reported difficulties in crossing the border to the authorities. Only 2.5 percent sought legal redress for harass- ment or unofficial fees, and none of the women who reported undergoing SGBV reached out to relevant service providers to receive medical treatment or counseling. It is unclear whether this lack of reporting is due to distrust of such ser- vices, lack of knowledge on who is responsible for the provision of relevant services, fear of repercussions, or the lack of outreach and effectiveness of the organizations involved. The mistrust may be rooted in past experiences; because most of the perpetrators of SGBV and corruption are officials at the border, the women traders are naturally less inclined to trust those to whom they must report their obstacles. Based on our conversations with other non- governmental organizations in the region, service delivery providers in North and South Kivu are completely overwhelmed by the demand for medical or psychosocial services. Traders may have felt that it would not prove benefi- cial to report to health and counseling experts or did not have the relevant information to reach out to these service providers. Barriers, Risks, and Productive Potential for Small-Scale Traders in the 35 Great Lakes Region Corruption To obtain information on corruption and its characteristics, we conducted focus group discussions with 100 of the 628 traders in Goma and Bukavu in May 2012. Examples of corruption that the traders provided during these dis- cussions included paying officials informally so that they could simply cross the border at all, paying officials to move ahead in long lines at the border, giving money to officials to cross the border after the official border closing, paying fees for unclear reasons or without justification, and paying officials to reduce their taxes. Of all 628 traders, 31 percent had, at some point, been involved in inci- dents in which they were forced to pay a bribe of some sort to cross the bor- der (whether through direct payment of cash or by giving up some of their goods); 130 of these traders were in North Kivu, 53 in South Kivu, and 19 in Rwanda. Some traders were victims of multiple illicit interactions: among all traders, there were 418 reports of individual bribery-related incidents, with 94 percent of these falling in DRC (see figure 2.1). Figure 2.1 Reported Cases of Corruption by Officials in DRC and Rwanda, by Type of Authority percent 1% 1% 1% 0% 3% 11% 25% 12% 18% 13% 15% DRC police OCC DGM RW customs DRC hygiene RW police DGDA Rwanda immigration DRC militaire RW military Source: World Bank Africa Region Gender Innovation Lab 2011-2012 survey. Note: DGM = Direction Générale de Migration (DRC); DGDA = Direction Générale des Douanes et Accises (DRC); OCC = Office Congolais de Controle; RW = Rwandan. 36 Women and Trade in Africa: Realizing the Potential Eighty-five out of 100 traders in the focus group discussions reported having had their goods confiscated by an authority at some point and could recall the specific authority responsible. These traders accounted for 146 specific incident reports. The largest number of the confisca- tions took place at the hands of the customs agency (DGDA) (30 reports). Twenty-five traders reported having had their goods destroyed at the border, mostly at the hands of the Congolese hygiene authority, fol- lowed by the Office Congolais de Controle. However, the lack of knowl- edge of regulations by traders and the lack of explanation by the border officials make it difficult to uncover whether these confiscations were legitimate for health and safety reasons or whether they reflect corrup- tion. This uncertainty reinforces the need for the more effective dissemina- tion of information on the rules and regulations governing cross-border trade and awareness-raising among both traders and officials. Sexual and Gender-Based Violence SGBV was measured by real incidents and perceptions of incidences wit- nessed while crossing the border during a 30-day period. To analyze the rates of physical and sexual violence, the team asked traders whether they had been insulted or spat on, groped, or sexually touched or had been subjected to rape or attempted rape while crossing the border. Insults and spitting are included to capture overall levels of aggression at the border. Within the 30-day period, 11 percent of all traders reported having been subjected to SGBV incidents that took place while crossing the border;9 and 5 percent had experienced some form of SGBV at the border. 10 Sixty-six percent of these latter incidents took place in DRC—again, with a stark contrast between North Kivu, where 52 percent of all incidents took place, and South Figure 2.2 Number of Incidences of SGBV Reported by Traders in DRC and Rwanda 197 62 23 19 12 7 Groping Sexual touching Attempted rape Rape Physical violence Insults/spitting Source: World Bank Africa Region Gender Innovation Lab 2011-2012 survey. Barriers, Risks, and Productive Potential for Small-Scale Traders in the 37 Great Lakes Region Kivu, where 13 percent of all incidents took place. While the reports of cor- ruption are strongly concentrated in the DRC, 34 percent of SGBV reports took place in Rwanda. Figure 2.2 shows that the most commonly occurring sexual and gender- based violation was sexual touching (23 reports), groping (19 reports), attempted rape (12 reports), and rape (7 reports). As for the proportion of individuals affected, 2.7 percent of traders reported having been groped, 3 percent had experienced sexual touching, 2 percent experienced attempted rape, and 1 percent had undergone rape in the past 30 days while crossing the border. According to the data collected on reports of incidents involving insults and spitting at the border, 35 percent were committed by the DRC police and 21 percent by the Rwandan police. Finally, it is worth remember- ing that fear of reprisal, shame, or social pressure may discourage the traders from reporting SGBV cases. Survey Results: Border Officials The study also interviewed 66 border officials from a range of government agencies on the Congolese side of the border. This section presents data on the living standards of border officials; their working conditions and prac- tices; their knowledge of tariffs, taxes, and procedures for cross-border trade; and the frequency with which they witness the types of corruption, harass- ment, and abuse that the traders were asked about (over the same 30-day reference period). One important caveat should be mentioned: although the research team had a listing of all relevant border officials, they could not take a random sample from this list. Government officials defined a subset of border officials from which the research team was permitted to draw a sample to interview. It is possible that the government officials chose border officials who would likely present a positive picture of conditions at the bor- der. “Negative” answers could therefore be interpreted as an (extreme) lower bound of the likely incidence of a given activity or occurrence. Border officials were on average 42 years old, 9 in 10 had completed secondary school, 9 in 10 were literate in French, and almost all reported that they were literate in Kiswahili. Just 5 percent lived in houses with dirt or sand floors, compared to 61 percent of traders, while half had electric- ity in their homes, compared to 20–30 percent of traders. Median monthly consumption measured as the sum of expenditures on food, transport, eating out, and mobile phones was US$334, almost twice as high as the US$170 in consumption reported by traders for the same categories. Clearly, border offi- cials either come from a socioeconomic class different from that of the cross- border traders or are extracting enough funds and goods from the traders to 38 Women and Trade in Africa: Realizing the Potential maintain this lifestyle; in either case, this difference further contributes to the power dynamics present in their interactions. The sample is drawn largely from the four main Congolese border agen- cies: the Direction Générale de Migration, Direction Générale des Douanes et Accises, Office Congolais de Côntrole, and the Programme National de l’Hygiène aux Frontières. Just two are from the (border) police, and none are from the army or other security services. The average tenure in the given agency was almost 10 years, and while border officials rotate as a matter of course, most officials had been at their current post (Goma or Bukavu) for 17 months and expected another 11 months in the same place before transfer. Three in 10 reported that they do not receive a regular salary, and just 15 percent reported being satisfied with their salary. The average salary reported was US$188, and the average allowances package was US$206 per month, suggesting average official compensation of US$394 per person. Seven in 10 had received some form of training. To assess their knowledge of rules and regulations, officials were pre- sented with a series of hypothetical cases and asked what fees and tariffs they would charge. Examples included a small-scale trader of fruits of vegeta- bles going from Rwanda to DRC or an exporter of goats from DRC to Rwanda. The results suggest substantial differences between officials and traders about conditions at the border. With regard to the example of the fruit and veg- etable trader, one in five suggested that a trader must pay for an immigration stamp (jeton). Just 10 percent replied that the trader must pay for a simplified import declaration, and only 3 percent suggested that the trader should pay import taxes. These figures are significantly at odds with the frequent pay- ments that traders report being forced to pay. Setting Up for the Impact Evaluation: Learning from a Potential Policy Response In response to the earlier study by Brenton et al. and following a request from the Congolese government, the World Bank has initiated a project to fund the training of border officials and small-scale traders on traders’ rights and obligations.11 Because of our initial focus on corruption and SGBV, the first training module informs traders about their rights and regulations on taxes and tariffs.12 The survey that is the focus of this chapter provides the baseline for a broader impact evaluation study. The evaluation will seek to measure the impacts of the training on rights and regulations related to tariffs, trade, and SGBV. Specifically, it will determine whether the treatment group of 324 Barriers, Risks, and Productive Potential for Small-Scale Traders in the 39 Great Lakes Region traders who were selected for training subsequently face fewer incidents of corruption and SGBV compared to the group that was not selected for train- ing. Follow-up data collection is planned for late 2013. Policy Implications and Future Work This study will allow for further investigation into current sexual crimes in North and South Kivu and Western Rwanda. Women’s limited voice and role in decision making are constraining the income-generating potential of small- scale, trade-based cross-border traders. Targeting this source of obstruction will support increased trade-related productivity for women, with positive long-term impacts inside and outside the household. Based on this prelimi- nary data collection, it is clear that SGBV for traders is closely linked to their understanding and perceptions of rights and violations. Increasing the transparency and predictability of taxes and fees applied at the border and rules and regulations governing trade is essential. The training program for traders and officials and the results of the impact evalua- tion should deepen our knowledge of what methods will increase awareness. Facilitating efficient trade and fair taxation at the border will lead to a credible revenue col- lection mechanism and positively affect the liveli- hoods of both traders and border officials. By measuring the effects of reforms and training programs, the impact evaluation will present the government with robust, evidence- based findings on the eco- nomic and social benefits from removing informa- tion barriers and reducing corruption. Women who have the information to determine their rights and obligations may have photo: Jonathan Ernst/World Bank a greater voice in the 40 Women and Trade in Africa: Realizing the Potential household and in society and, as a result, may be less susceptible to SGBV in the future. Mechanisms that provide traders with information on their rights, that open channels for reporting violations, and that consistently bring about penalties for infractions are integral to a transparent and effective social accountability system and will help ensure that individuals are empowered to control and curb violations on the border. Such a system will have substantial positive impacts on the economic potential for trade at the border. Finally, expanding and encouraging economic interdependence in the region are vital to providing stability (International Alert 2012). Collaborative economic activity by these small traders helps reduce tensions and build common interests along open and well-maintained borders. The regional inte- gration and collaboration that increased trade would promote will strengthen economic and diplomatic ties in the region and, as a result, support long- term stability. Notes 1. The survey collected information from 181 traders through in-depth inter- views with 100 of them and focus groups with 81. In addition, 58 interviews were conducted with stakeholders at the borders. 2. One issue that was identified was the large number of agencies operating at the border; as many as 17 different agencies could be present. Traders were not clear on who was truly authorized to be there and who could officially ask them for fees. Subsequently, a formal decree was issued by the Congolese government in November 2011 mandating a reduction in the number of agencies that were allowed to oversee crossings at the border to only four, including immigration (Direction Générale de Migration), customs (Direction Générale de Douanes et Accises), arms control (Office Congolais de Côntrole), and hygiene authorities (Programme National de l’Hygiène aux Frontiers). This limit on the number of agencies allowed at the border was seen as an important step toward curbing the potential for corruption and harassment. 3. The sampling of traders was challenging, given the mobility of the popula- tion as well as the informal nature of their businesses. To ensure the repre- sentativeness of the population our survey team proceeded with a listing of all small traders passing through the major border crossings between Goma, DRC, and Gisenyi, Rwanda (the Grand Barrière and the Petite Barrière) and between Bukavu, DRC, and Cyangugu, Rwanda (Ruzizi I and Ruzizi II). The listing of the universe of traders was compiled over a one-week period in which enumerators enlisted all small-scale traders at these crossings. This listing exercise allowed the survey team to compile the names of the main markets where these traders trade as a base for the second sampling strategy Barriers, Risks, and Productive Potential for Small-Scale Traders in the 41 Great Lakes Region that involved visiting the markets and randomly selecting the “nth” trader to be interviewed. Gender was taken into consideration in the selection as the sensitivity of the topic required female enumerators for female traders. 4. There are two important limitations that need to be underlined in the selection of the officials: first, the sample included only Congolese officials and thus our understanding of the Rwandan side of the border is incomplete; and, second, the need for authorization from government agencies reduced the selection of the officials interviewed and thus a random selection of offi- cials was not possible. 5. This was calculated by multiplying the number of traders (628) X (.75)— the number of traders that reported cross-border trade as their main source of income—by the average household size (7.4). 6. Further research is required to understand the benefits of membership for these traders in DRC and Rwanda in providing security against SGBV and corruption. 7. A group of—mainly—women wait for traders at the border crossings to charge a fee for carrying their goods on their head. 8. All the enumerators came from Bukavu. All female traders were inter- viewed by female enumerators and all male traders by male enumerators. 9. This refers to the entire process of crossing, from the first point of con- tact made in order to cross the border up until actually to exiting the border crossing. 10. This allows for repeated incidents to have been reported by the same trader. 11. The local implementing partner is International Alert. 12. When we asked traders what kinds of training would best enable them to improve their training activities, two-thirds (67 percent) opted for business training and business diversification, while 16 percent mentioned training on taxes and regulation and 9 percent on their rights vis-à-vis harassment. When asked specifically about a possible training on the regulations and fees that they should pay when crossing the border, 93 percent said that they would be interested. The responses to this question will serve as guidance for the follow-up training and other intervention. The team is currently exploring potential follow-up interventions that focus on entrepreneurship skills. References Brenton, Paul, and Gozde Isik, eds. 2012. Defragmenting Africa: Deepening Regional Trade Integration in Goods and Services. Washington, DC: World Bank. Brenton et al. 2011. Risky Business: Poor Women Cross-Border Traders in Great Lake Region in Africa. Washington D.C.: World Bank. International Alert. 2012. Ending the Deadlock: Towards a New Vision of Peace in Eastern DRC. London: International Alert. Sommers, Marc. 2012. Rwandan Youth and the Struggle for Adulthood. Studies in Security and International Affairs. Athens: University of Georgia Press. USAID (United States Agency for International Development). 2007. Democratic Republic of Congo: Demographics and Health Surveys 2007. Washington, DC: USAID, ———. 2010. Rwanda: Demographics and Health Surveys 2010. Washington, DC: USAID. Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 43 3  Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria Louis Njie Ndumbe Introduction Eru, a vegetable found in the forest, plays an important role in trade between countries in west and central Africa, especially between Cameroon and Nigeria. This chapter analyzes the environment for trade in eru as an example of a high-value nonwood forest product that has great potential both for export and for income-generating activities. Women are heavily involved in eru trade as both harvesters and small-scale traders. For many women, it is a key means for diversifying the income stream of their household and reducing their financial dependence on men. Hence, analyzing the eru value chain from a gender perspective is of particular importance. The discussion also focuses on the substantial constraints that women face relative to men in developing their eru business, constraints that limit their value added and the contribution of their work to household income. The chapter also explores a key challenge facing countries in the region—how to support trade in eru with its potential for poverty reduction, while also ensuring its sustainable cultivation. Current destructive harvesting techniques, extraction rates in excess of what can naturally be replenished, the rapid decline of the natural habitat of the plant, and the lack of effective management of forests, and of eru itself, compromise the long-term sustain- ability of this important natural resource. Gender imbalances in access to training and finance and in ownership of land undermine effective responses to this problem. Background Central Africa harbors the Congo basin forest, the second-largest forest reserve in the world after the Amazon forest. This forest has a huge diversity of biological resources, including two nonwood forest products that are widely traded in the region, Gnetumafricanum and bucholzianum, Louis Njie Ndumbe is an independent consultant and cofounder of Environmental Education, Conservation and Research, a nongovernmental organization in Cameroon. 44 Women and Trade in Africa: Realizing the Potential locally known as eru by the Anglophones and okok by the Francophones in Cameroon.1 Current estimates suggest that 40 percent of Cameroon’s population lives below the poverty line of US$1 per day. Over half the population lives in rural areas (INS 2007; ECAM 3). Sourced principally from Cameroon, eru is an important product for the country, given the volume of eru trade and its high economic value in an area of deep poverty. Even though exports of eru from Cameroon to Nigeria are not well recorded by customs officials at the border, evidence suggests substantial and increasing export flows. In 1992, exports of eru from Cameroon to Nigeria were estimated to be 428 tons (FAO 2009). By 2009, this trade measured more than 4,000 tons and represented about 78 percent of the overall volume of international trade in eru. That is, at the price of CFAF 1,000 per kilogram, Cameroon exported about CFAF 4 billion worth of eru to Nigeria or US$7.6 million.2 Overall, Cameroon’s exports of eru to the world are about CFAF 5.1 billion, equivalent to US$9.8 million, and represent 0.3 percent of total Cameroonian exports. Eru trade is a main source of income for individuals involved in the product’s value chain. Available estimates suggest that eru contributes on average 62 percent of a harvester’s annual income. Dependence on eru-based incomes increases for those further from the forest, providing on average 75 percent of a retailer’s (US$1,268) and 58 percent of an exporter’s annual incomes (US$7,000) (CIFOR 2009; Ingram et al. 2012). FAO (2009) reports that women trading eru, as harvesters and traders, make daily profit margins during the main season that vary between US$16 and US$160, depending on photo: Yosef Hadar / World Bank Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 45 the prevailing market conditions and whether the seller is a harvester or a trader.² Likewise, profits along the value chain increase the further away from the forests, with exporters enjoying the highest profit margins and harvesters making the least profits. Eru is also one of few products in which women are actively involved either as a harvester or as a small-scale trader. Women dominate the lowest segments of the value chain. Among a total of 759 harvesters interviewed in the southwest and littoral regions of Cameroon, almost 80 percent were female (Ndumbe, N. L., V. Ingram, and A. Awono. 2009.) Ndoye (1997) reported that about 94 percent of intermediaries operating in the humid forest zone of Cameroon were women. In addition, all the retailers identi- fied in the southwest region were women (Ndumbe, N. L., V. Ingram, and A. Awono. 2009.) Finally, of the estimated 265 exporters 58 percent were female (Ndumbe, N. L., V. Ingram, and A. Awono. 2009.) Profit margins for entrepreneurs working in this sector depend, among other things, on gender, family status, and age. For example, women are more restricted in their access to forest resources than men (Ingram, Ndumbe, and ElahEwane 2012). Eru of good quality—meaning large, thick, and dark green leaves—and quantity is generally found in closed-canopy forests and mostly on trees and tree canopies. Local customs, however, limit women’s access to trees (timber, commercial fruit trees, and commercial fuel wood) and other productive resources like bamboo, raffia, cola, and the like. Under communal tenure regimes, women traditionally obtain legal rights to family land, but they do not possess inheritance rights. Consequently, for crops like eru, women can exercise their right to cultivate in the forest. However, unlike male harvesting agents, women are restricted from climbing trees to collect eru. As a result, males have access to the best harvesting areas, collect the best quantity and quality, make larger profits from eru activity, and support more people from eru revenues (Ingram, Ndumbe, and ElahEwane 2012). A key challenge facing the production and trade of eru is overexploita- tion and the lack of governance to implement sustainable management of these natural resources. Central African countries have made commitments to sustainably managing their forest resources by ratifying a number of interna- tional legal instruments. The implementation of these legal tools, however, is often ineffective due to weak enforcement and inadequate consideration of subregional interests and realities (FAO 2009). Removing the constraints to eru trade and addressing the gender- specific barriers that are discussed in this chapter will be an important step toward a more organized and transparent market for eru. Likewise, it will provide a stronger base for implementing a regulatory environment that will ensure sustainable management of this important resource for the people of 46 Women and Trade in Africa: Realizing the Potential southwest Cameroon. An approach that simply restricts the collection of eru from the forests will have especially deleterious effects on the incomes of women. Women and Trade in Eru: Constraints along the Value Chain This chapter investigates the role of men and women along the eru value chain. The analysis is based on interviews, conducted through focus group discussions and complemented by some face-to-face interviews. Interviews were held with 60 women and 20 men involved in the eru value chain as harvesters, direct exporters, or intermediaries along the Bamenda-Mamfe- Ekok trade corridor leading directly to the Ikom market in Nigeria (see figure 3.1).3 This corridor is characterized by a high volume of eru trade. Around 40 percent of eru exporters who export to Nigeria through the southwest region of Cameroon use this route. Below is a summary of the key outcomes from the focus group discussions. The Benefits from Women’s Participation in Eru Trade The income from eru trade allows women to contribute financially to their household, especially to their children’s education. For example, one participant in the discussion group said, “Money from eru has helped me a lot. It helps us to afford food for the household and to pay for chil- dren’s education.” Taking part in eru trade is also seen as a source of empowerment for women. A response that reflects the overall impres- sion from the focus groups came from one participant: “I am happy to have [my] own business and [do] not receive orders from any boss[;] neither [do I] need to depend on any man for [my] livelihood.” Finally, 70 percent of the participants confirmed that they take part in eru har- vesting and trade because it contributes to diversifying the income stream of the household and hence reduces their vulnerability. A key feature that makes eru trading an attractive activity is that profits can be quickly realized relative to other seasonal products such as cocoa and egusi seeds.4 Trade in eru is an all-season activity. Furthermore, the product is found free in the forest. As a result, very little start-up capital is required relative to other activities. If capital is required, small amounts can be obtained from savings in community associations (njangi). Finally, cultivat- ing eru is an activity in which those with a low level of formal education can effectively participate. Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 47 Figure 3.1 Key Border Crossings in the Southwest Region of Cameroon and the Mamfe-Ekok-Ikom Trade Corridor MARKETS PRODUCTION SITES CITIES AND TOWNS NATIONAL CAPITAL MAIN ROADS NIGERIA PROVINCE BOUNDARIES NORD- INTERNATIONAL BOUNDARIES OUEST Mb u Takpe Bamenda Ikom Bache Euyumolock Ekok Mamfé Okoyong CAMEROON Okang Etoko Nchang Bafoussam Oran Kembong Ekenge OUEST CENTRE Mbam Mungo Ndor Calabar SUD-OUEST Mundemba 0 40 80 Kilometers Ekondo Titi Kumba 0 40 Miles uri Fiango Wo Bamusso Mbonge Sa’a Souza Bonamateke Gulf Buéa Buéa Nkapa Camp LITTORAL Obala of Tiko Douala Limbé ga Sana Evoudula G YAOUNDÉ ui EQ. GUINEA Nyong ne GSDPM a IBRD 40451 NOVEMBER 2013 Map Design Unit Gender-Specific Constraints While women are heavily involved in the eru value chain, they also face con- straints that are gender specific. Several key issues were raised in the focus group discussions. Domestic violence. Domestic violence can occur when women devote time to exporting and to intermediary business activities rather than to household responsibilities. Some female exporters and intermediaries are victims of verbal and physical abuse from their husbands. Women do not report such abuse because, they say, “the pride of the African woman is to be under the roof of a man and having children.” According to interviews, women say that the only solution appears to be to stop exporting or give 48 Women and Trade in Africa: Realizing the Potential their merchandise to colleagues, who sell eru on their behalf. Time-related constraints. Time-related constraints arise because of the social bias toward women performing household chores and having large families. About 40 percent of participants lamented their inability to spend more time on trading activities and said that they usually miss some of their planned trips to collect eru because of household responsibilities. More than 90 percent of exporters and intermediaries agreed that taking care of their business and household during pregnancy is very difficult. Furthermore, discussions revealed that women have very little or no knowledge about birth control. Discussants made it clear that they do not have the power to refuse their husbands; as one participant pointed out, “It is considered a pride of fruitfulness by the society for us women to have more children.” However, contrary to societal expectations, participants said they did not consider having many children a priority. Limited decision-making ability. Related to participation in eru activities and the allocation of the income generated from its trade is the issue of women’s limited ability to influence household decisions. While small- scale eru trade is dominated by women, decisions are made by women only after consultation with their husbands. Consequently, women tend to be dominated by the opinion of their husbands. Focus group discussions revealed several prevalent attitudes: • Husbands consider eru trading and harvesting a profession with low status, which they find suitable for women, including their wives. • Women have to consult with their husbands before investing or spending their profits. According to one discussant, “Our husbands are the heads of household. Because of the respect we have for them and in order to justify the time we spend doing business and the benefits of the business, we often make them ... aware of profits and losses and [of] how we may want to use the money. It is a lack of respect [for] the husband for a woman not to do this which may often result [in] problems with [the] husband.” Lack of ownership of land. The profit from eru is used mainly to support children’s education, health, and other household needs. If funds are available for savings, they are put into njangis, community financial associations based on mutual trust. Yet, according to participants in the study’s focus groups, funds from these associations are typically used by their husbands for cocoa production. Cocoa production is a male-dominated activity since it requires a longer-term investment in land that cannot generally be owned by women. A participant explained that culturally women do not own land because it is believed that “the girl child in a family Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 49 photo: Curt Carnemark / World Bank has not come to stay, but will one day become another man’s ‘property.’ And usually when the man dies and leaves the wife, one of the man’s relatives will ‘inherit’ the woman and she becomes his wife.” Men generally have exclusive rights to manage and administer property. As a result, because women lack the collateral to obtain access to credit, their capacity to grow their business is severely constrained. Limited education and access to training. According to the interviews, women involved in the eru value chain as exporters and harvesters want training on how to access credit and expand their business. In addition, some would like further education. They complain that available training activities are concentrated in male-dominated activities. Also, they say their role in the household limits the amount of time available for such activities. As one participant pointed out, “Our access to further education and training is inhibited by the obligation to take care of our children and to cook food for the household.” Of particular importance, given the challenge of sustainable harvest- ing and trade of eru, is that the women harvesters do not have access to knowledge on how to domesticate eru, despite expressing a strong interest in learning. This is complicated by the lack of female access to land upon which domesticated eru could be grown. Eru is becoming scarce. Distances traveled to get eru from the forest are increasing. In addition, the way in which eru is harvested affects its sustainability. It is reported that some males harvest eru unsustainably from the trees, along which the eru plant thrives, a practice that undermines the eru plant. According to women interviewed during focus group discussions, this practice is causing progressive exposure of the forest 50 Women and Trade in Africa: Realizing the Potential and deforestation. This example illustrates how traditional female knowledge of the resource and forest could be very useful in enhancing its protection and sustainability. In this context, Van Vliet (2010) shows that development agencies working around the Takamanda area have failed to take into full account gender differences in resource use and land tenure when designing projects to support sustainable agricultural activities. The ongoing shift in livelihood, especially that of men, from the sustainable harvesting of nontimber forest products to the production of cash crops, primarily cocoa, is leading to the conversion of closed-canopy forests into cocoa-forestry systems. With the transformation of the forests into cocoa-producing rather than eru-producing land, urgent action is needed to preserve the environment for eru and its financial rewards for women. Harassment from officials. Exporters and intermediaries also report harassment from government officials and at times from buyers, which causes delays and losses, given the perishable nature of their products. Officials sometimes accept promises from traders to pay “settlements” or bribes after sales; however, if the market is not favorable, this practice can lead to debt accumulation. Moreover, when debts are not paid, officials can confiscate products. This situation usually puts women in a vulnerable position whereby some may succumb to sexual blackmail, as focus group participants suggested, although no participant directly admitted to having been involved in this type of situation. Limited access to market contacts and trading networks. Responsibilities in the home limit the access of women to contacts in the market, to role models, and to social interaction. As a result, their contact with the “business culture,” which serves as the main training ground for trade, is constrained. Meanwhile, men, who are free of many household tasks, benefit from greater mobility and have better access to markets, credit, and business connections. Such access provides them with a forum for meeting, learning from, and gaining the support of others who trade in different products. These constraints on the interaction of women with other market players hinder the growth of eru activities and limits diversification into other trading activities. For example, bush mango trade requires a larger capital investment and access to “connectors” (that is, experienced traders), who match traders up with buyers in Nigeria. Trade- and Business-Related Constraints In addition to the explicit gender constraints discussed above, women, as well as men, involved in the eru trade face a range of trade- and Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 51 business-related constraints. However, in practice women often face greater difficulties in overcoming these constraints than men. Inability to acquire administrative trade and transport documents. Given that eru is a special forest product, sale and marketing are regulated. Box 3.1 outlines the documentation and taxes an exporter must provide the government. Despite the detailed list of documentation and taxes officially required by the government, there is no sense that any system is in place to monitor the documentation provided or the taxes paid, which underlines the argument that these administrative barriers to eru trade should be streamlined. For a given operator, administrative procedures for the marketing of eru are very cumbersome and begin at the central service of the ministry in charge of forestry, notably, by obtaining three main documents: • Approval document, which gives access to the forestry profession and includes a file-processing fee of CFAC 150, 000. Box 3.1: The Administrative Process of Exporting Eru: Documentation and Taxes To export eru, an individual should possess the following documents: an export authorization, a certificate of origin, a specification sheet, a phytos- anitary certificate, and proof of payment of the graduated surtax. Exporters are indeed obliged to pay two types of taxes: • Ordinary (or general) taxes applicable to all businessmen, payable at the taxation center with competent jurisdiction (full discharge tax, fees for market stalls, and others). • Eru taxes, which are based on the export volume of eru and include the following: ▼▼ Regeneration tax payable before exploitation: 25 percent as of January 31 during exploitation; 50 percent as of May 31 at the end of exploita- tion; and 25 percent as of October 31. ▼▼ Graduated surtax payable upon exportation of unprocessed eru: in the case where the quotas are exceeded (CFAF 10 per kilogram). ▼▼ Surtax upon exportation of fresh eru (CFAF 50 per kilogram) and dried Gnetum (CFAF 75 per kilogram). ▼▼ Phytosanitary tax (CFAF 6,500). The eru exporter should expect a payment of CFAF 650 per kilogram if the eru is transported by air and several other taxes and charges depending on the country of destination and the means of transport (FAO 2009). 52 Women and Trade in Africa: Realizing the Potential • Special permit to cultivate eru, which gives access to the resource. • Record of waybill receipts, which authorize the operator to transport the product. In addition, box 3.2 explains the two-step process for obtaining adminis- trative approval from Cameroon’s Ministry of Forestry and Wildlife. Since men have better access to credit and business connectors than women, they are in a better position to gain the support of others and to acquire cultivation and marketing permits from the government. For example, in 2009 only one out of six permit holders that exported through Idenau Port in the southwest region was a woman. The complicated nature of the administrative process for obtaining permits automatically disqualifies poor and uneducated women who do not meet the criteria but are the main actors involved in the value chain. Box 3.2: Process for Obtaining Administrative Approval The initial granting of approval is the first stage of the administrative procedure, and it is subject to the recommendation of the technical commit- tee, on the basis of a file submitted to the Ministry of Forestry and Wildlife against a receipt. The file comprises the following documents: • For individuals: A stamped application bearing the applicant’s full name, nationality, profession, and place of residence; curriculum vitae indicating professional experience and qualifications; a nonconviction certificate, not more than three months old; and the statistical and trade registration numbers. • For corporate bodies: A stamped application indicating the company’s name and address; a copy of the company’s articles of association; the statistical and trade registration numbers; the manager’s nonconviction certificate, not more than three months old; curriculum vitae of the man- ager or official in charge of forestry operations indicating his technical knowledge and professional experience; current and previous activities of the company; and an attestation of regular payment of contributions due to the national social insurance fund. • For individual and corporate bodies: The area of activity for which approval is sought; documentary evidence of the technical knowledge of the applicant, in the case of private individuals, or of the official in charge of operations, in the case of corporate bodies; a tax certificate; a state- ment of tax situation; and proof of file-processing fee payment. Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 53 To facilitate their export activity, some intermediaries and exporters purchase waybills from corporations rather than acquire the permits directly from the government because of the cumbersome and opaque administra- tive procedures. These corporations act as brokers and sell quotas at prices above the original regeneration tax. The waybill, which gives exporters the right to transport eru, must be presented at every checkpoint along the way. The average cost of a waybill for small-scale traders is CFAF 12 per kilogram of eru, representing 1.2 percent of the average price received per kilogram by exporters. It is important to note that it is illegal for large corporations to sell waybills to exporters; however, this practice is necessitated by the compli- cated and arduous process that exporters must go through to obtain such trading permits. According to traders, more powerful and influential permit holders have sought to exercise control over the trade in eru by “blocking” the access of smaller traders to permits through a process dogged by corruption. Again, this activity is likely to impinge particularly heavily on women, whose lack of contact with effective networks of traders is likely to limit their ability to confront corruption in the allocation of permits. To facilitate eru trade, the government should reform and streamline the administrative processes that limit female involvement in the market and create unnecessary barriers to trade. Traders should be able to obtain permits directly from the Ministry of Forestry and Wildlife with a simple, affordable, and transparent procedure. In addition to the administrative hurdles for eru exporters, the govern- ment also seems to have set up informal export quotas for unprocessed eru. Even though there is no official information on such informal quotas, these kinds of unsanctioned barriers to trade are another business-related con- straint traders face. High transport costs, roadblocks, and corruption. The poor state of the Mamfe-Ekok road and high transportation costs, compounded by hikes in the price of fuel, discourage road transport. Hence river transport via Satom and John Holt beaches, which are located on the Cross River, is the preferred means of transportation. Participants noted that the construction of the Mamfe-Ekok road should facilitate trade. However, informal payments to forestry guards, councils, customs officers, police, and quarantine and commerce officials along the route make road transport undesirable for exporters. Ultimately, these payments will undermine the value of the investment in the new road. Furthermore, even after presenting waybills, exporters must make informal payments to officials. For example, on average at the Satom and John Holt beaches, exporters still have to pay officials CFAF 1,000 per 200 kilogram bag of eru leaves transported. However, this is less than the average amount they would have to pay for using road transport— 54 Women and Trade in Africa: Realizing the Potential about CFAC 1,500 per 200 kilogram bag—where there are multiple roadblocks and consequently higher charges. Transport costs alone account for the highest proportion of exporters’ costs, about 23 percent. This, together with the five formal and informal types of taxes (forestry, council, quarantine, customs, and police), accounts for 59 percent of total costs and constitutes about 19 per- cent of the share of profit for exporters.5 The euphe- mistically named “police photo: Jonathan Ernst/World Bank tax,” used to cover bribes to police and gendarmes, usually en route, alone amounts to 14 percent of costs (Ndumbe, N. L., V. Ingram, and A. Awono. 2009.) For example, intermediaries traveling from Yaounde to Bamenda pay on average CFAF 5,000 but sometimes pay up to CFAF 15,000 for “settlements” during peak season (November-December) when roadblocks are prevalent. Market power of buyers in the Nigerian market. At the Ikom market in Nigeria, Cameroonian exporters have very little bargaining power. According to the focus group discussions, this is due to several causes. First, the Association of Afang Dealers in Nigeria (the Nigeria wholesalers) forbids Cameroonian exporters to sell directly to retailers. Second, Nigerian wholesalers have formed a cartel, thanks to their vibrant association. The association operates two major warehouses—the Akwa Ibom warehouse and the Ibo warehouse. Within these two major groups, there are minor “shades” or groups of at least five members. Each shade has its established customers in Nigeria and suppliers in Cameroon. Each member ensures the smooth running of the shade in the absence of the other members. At least one member of the shade always remains in the market to attend to suppliers and serve customers. Shade members then arrange the returns and expenditures among themselves based on financial records (Ndumbe, N. L., V. Ingram, and A. Awono. 2009.) For example, if a shade member owes money to a supplier Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 55 or if a customer owes money to a shade member, the money is paid to or by the member on duty for that day (whether he was involved in the previous transaction or not). In this situation, Cameroonian traders who do not have an association to represent them have little bargaining power, and the prices they receive are dictated by the Nigerian associations. Conclusions and Recommendations Trade in eru is an example of a natural product that provides significant contributions to the livelihoods of actors along the value chain in southwest Cameroon, who are predominantly poor women. Women are involved in this trade out of necessity, as it is the only major livelihood opportunity available to them and its revenues meet the basic needs of their households. Trade in eru also has a catalytic function for other economic activities, as profits from this activity are typically invested in other food and education. However, women face a range of societal-, business-, and trade-related constraints. Women suffer pervasive harassment, lack of access to credit and training, and restrictions on their mobility and capacity to exploit market opportunities. These factors explain their limited ability to take full advantage of trade and to increase control over their own lives and families. An increase in the involvement of women in economic activities outside the household would also mean an increase in autonomy within households, thus leading to greater domestic power and control. These factors also explain why women continue to concentrate their activities outside the household on eru. It is an activity that requires little capital and low skills. Given the potential of the eru value chain to promote household welfare, removing the constraints faced by female producers and traders of eru could have a great impact on economic growth and poverty reduction. In addition, access to more profitable export activities, such as bush mango and cocoa, which remain male dominated, would have a posi- tive effect on women’s economic and political status. Six key steps could be taken to address the obstacles faced by women in achieving their potential as small entrepreneurial traders in the west of Cameroon: 1. Formulate and implement policies that promote gender equality at the national level. The government should define laws and policies that guarantee equal rights for men and women in the exercise of control over assets such as land and in accessing services such as education and credit. This also requires giving women control over economic decisions that affect their lives and, on a political level, a greater sharing 56 Women and Trade in Africa: Realizing the Potential of reproductive responsibilities among women and men in society. The government should also ensure that institutions and officials that have a stake in the forestry and agricultural sectors are adequately trained to effectively implement gender equality under the law. The government should also work in collaboration with civil society to ensure that both women and men are educated on the rights of women. 2. Put in place a market information system and implement a policy frame- work to organize women traders to improve their bargaining power and economies of scale. Many informal exporters in Cameroon appear not to realize the full size of the Nigerian market, its potential to support their businesses, nor how prices are determined in that market. A catalogue or database of all actors in the chain could facilitate research and regulation of the sector while monitoring the resource. Although cultural attitudes toward organizations may challenge the level of association and collec- tive action possible, this sector would benefit from associations for the main actors and from strengthening the capacities of these associations. 3. Remove roadblocks and other barriers to trade along key corridors. The large potential Nigerian consumer market, with its proximity to the fer- tile production areas of Cameroon, offers a major opportunity for small Cameroonian exporters. However, poor infrastructure and costs create delays, allow corruption along the route, and act as significant barriers to efficient trade. 4. Reform and streamline the administrative permit system for exporters. The government of Cameroon requires unnecessarily complicated docu- mentation and administrative procedures for eru exporters. Simplifying documentary requirements for exporters and the inordinately long administrative procedures would contribute to a more sustainable trad- ing system for eru. 5. Training and capacity building for women. Education and skill building will broaden women’s range of choices and give them more influence within their households and in society. Building the human capital of women and girls will allow them to become better producers, more con- fident traders, and stronger citizens. 6. Benchmark progress in removing gender bias along the eru value chain. Baseline surveys and continuous monitoring and evaluation of sex-dis- aggregated data will greatly advance our understanding of gender differences in trade. Such studies will also demonstrate the positive implications of gender equality in supporting the formulation of appro- priate and gender-aware policies and programs in the forestry and Unshackling Women Traders: Cross-Border Trade of Eru from Cameroon to Nigeria 57 agricultural sectors. Making women’s voices heard at all levels of deci- sion making in these sectors is critical. Notes 1. In Nigeria, they are locally known as okasi by the Ibos and afang by the Akwa Iboms. Dealers in Cameroon and Nigeria locally refer to them as “salad.” The eru leaves are eaten raw or shredded and added to soups, stews, porridges, and fish and meat dishes. Both species of eru are highly nutritious and an important source of protein, essential amino acids, and minerals. 2. At the exchange rate CFA Franc/US$=524. 3. An interview guide (qualitative questionnaire) was used to conduct inter- views for each stakeholder group moderated by a female facilitator. 4. Egusi seeds are the fat- and protein-rich seeds of a cucurbitaceous plant used in preparing egusi soup, a kind of soup thickened with the ground egusi seeds cooked with water and oil and typically containing vegetables, season- ings, and meat, which is popular in many west and central African countries. 5. Profit margins were calculated using average prices and costs of exporters. Average figures are based on calculations from reported actual quantities, cost, and prices. References FAO (Food and Agriculture Organization). 2009. “The Legal Framework Governing the Exploitation and Marketing of Non-Wood Forest Products in Cameroon and Sub-Regional Guidelines on the Sustainable Management of NWFP of Plant Origin in Central Africa.” Produced by the Sub-working Group on Non-wood Forest Products; FAO-COMIFAC-SNV ICRAF-CIFOR. Rome: FAO. Ingram, Verina, Louis N. Ndumbe, and Marcus ElahEwane. 2012. “Small Scale, High Value: Gnetum africanumand buchholzianum Value Chains in Cameroon.” Small-Scale Forestry 11 (4): 539–56. Institue of National Statistics. 2007. Troisieme Enquete Camerounaise aupres des Ménages (ECAM 3). Ndoye, O. M. 1997. “The Markets of Non-Timber Forest Products in the Humid Forest Zone of Cameroon.” Rural Development Forestry Network Paper 22c, Overseas Development Institute, London. Ndumbe, N. L. 2009. “Market Baseline Situation for Gnetum Spp. in Nigerian Markets.” Center for International Forestry, Bogor, Indonesia. 58 Women and Trade in Africa: Realizing the Potential Ndumbe, N. L., V. Ingram, and A. Awono. 2009. “Eru Market Chain Baseline in the Southwest and Littoral Regions of (Gnetum spp.) Cameroon.” Center for International Forestry, Bogor, Indonesia. Van Vliet, N. 2010. “Participatory Vulnerability Assessment in the Context of Conservation and Development Projects: A Case Study of Local Communities in Southwest Cameroon.” Ecology and Society 15 (2): http://www.ecologyandsociety.org/vol15/iss2/art6/. Women Cross-Border Traders, Challenges, and Behavior Change Communications 59 4  Women Cross-Border Traders, Challenges, and Behavior Change Communications Susan D. Ityavyar This chapter examines the obstacles—formal and informal—that women face as cross-border traders and suggests reforms to promote social and behavioral change. Small-scale cross-border trade is categorized as legitimate cross- border trade carried out by commercial players (companies or individuals) in small quantities (Njiwa 2012). It constitutes a significant segment of the local economy and provides access to essential food staples and products for a large section of the African population, especially in rural areas. It is impor- tant to note the critical contributions that women have made to cross-border trade, especially in food security and access to agricultural goods. Women are often recognized as producers of food, but their role in cross-border trade is often overlooked. While women’s contribution to this sector is critical, their benefits from trade are sometimes marginal due to policy, institutional, cultural, economic, and regulatory issues, among others (Randriamaro 2008). Regardless of these challenges to trade, women’s cross-border trading activity is often the only source of income to their families and contributes signifi- cantly to economic development. In West Africa, women in cross-border trade support more than six dependents on average, and in Kenya, 80 percent of women cross-border traders rely on that trade as their sole source of income (USAID 2012a). Women often face discrimination at the border. Studies in East Africa have shown that female cross-border traders are forced to pay larger bribes than their male counterparts or must provide sexual favors to avoid detention by the border guards or confiscation of their goods (Higgins and Turner 2010). The cost of doing business at the border is very high across Africa. According to Brenton and Isik in their 2012 report, Defragmenting Africa, the cost of trading across borders is the highest in the Sub-Saharan Africa region—over twice that in East Asia or in OECD countries. This is also observed in the World Bank’s Logistics Performance Index, based on a world- wide survey of global freight forwarders and express carriers; it demonstrates that African countries lag significantly behind other regions in key areas such Susan D. Ityavyar is consultant (HRDDL) at the International Finance Corporation. 60 Women and Trade in Africa: Realizing the Potential photo: Arne Hoel / The World Bank as customs, infrastructure, competence in logistics, and timeliness of exports and imports (World Bank 2010).Women play a major role in cross-border trade. About 90 percent of small-scale traders across Sub-Saharan Africa are women (Brenton and Isik 2012). Yet they face more challenges when crossing the borders than their male peers: limited transport capacities, high trans- portation costs, roadblocks, corrupt officials, and harassment at the borders. They also have limited access to finance, time, and land and confront other market and social challenges such as unfavorable policies for women and skewed perceptions of female traders. Security, mobility, and service delivery constraints can significantly limit women’s access to markets, particularly markets across borders. Easing those constraints could increase opportunities for formal cross-border agricultural trade by women and has the potential to generate economic growth and promote food security as well as reduce poverty among vulnerable households (USAID 2012a). Change Mentality: From the Bottom Up and the Top Down There is need for a complete change in the attitudes toward women traders both from the bottom up and from the top down. Although women contrib- ute significantly to the economic well-being of their households, they are not generally recognized as legitimate business people and financial contribu- tors to their families and to the economy at large. At the top, the govern- ment often does little to recognize small-scale traders as legitimate entities and thereby perpetuates the negative stereotypes of women traders as “petit Women Cross-Border Traders, Challenges, and Behavior Change Communications 61 Box 4.1: Village Information Center Project in Uttar Pradesh, India The main objective of this project was to make key information readily accessible and to make people aware of government programs, entitle- ments, and important practices such as breastfeeding and hand washing (UNICEF 2011). Village Information Centers targeted the population mainly through interpersonal communication to reinforce best practices. These centers had about 15 to 20 members, depending on the population of the village. First, the project provided training to grass-roots functionaries to ensure that they had holistic knowledge on key issues and the desired skills for effective interpersonal communication and social mobilization. These functionaries developed a social map of their respective villages or hamlets with the help of local communities and identified the households on which they needed to focus their efforts. They also tracked behavioral change on a bi-monthly basis. Second, cultural mobilization campaigns used folk-art-like songs and street plays to raise awareness on social issues. Local nongovernmental organizations and community-based organizations were engaged to reach the communities effectively. These organizations assigned animators or facilitators to cover a cluster of 10 to 12 villages. Third, Village Information Center members also received training on the use of tailored communications and information materials and on plan- ning communications to effect behavioral change. After the training, they functioned as local resource persons for organizing community dialogues and disseminating key messages and information. A major lesson from this project is that a physical space is highly important for implementing suc- cessful Village Information Centers. In Uttar Pradesh, this initiative not only provided health information but also ensured sustainability and conve- nience in access to information. hustlers” in the eyes of low- and mid-level border officials. At the other end of the spectrum, husbands of traders often fail to fully value the economic contributions of their wives. As a result, women often face gender-based challenges, including sexual harassment and abuse, which create barriers to doing business effectively. Because women often face limited access to police and legal services, improvement of access to formal redress mechanisms at the border is a first step toward protecting female traders in their activities. Border information centers are one approach to addressing the problem. Such centers are a place where female traders could submit complaints or file reports pertaining to 62 Women and Trade in Africa: Realizing the Potential problems at the border. In the case of COMESA (the Common Market for Eastern and Southern Africa), trader information desks were put in place to help nonliterate traders in file forms and access information. These informa- tion centers could also enable traders to speak with professionals or anony- mously report harassment cases at the border. The Village Information Center initiative in Uttar Pradesh, explained in box 4.1, illustrates the potential of such centers to promote improved condi- tions. This success story shows that communication can bring about sustain- able behavioral change. The method used in Uttar Pradesh could be emulated and provide cross-border traders consistent access to information, mediation, and development services. COMESA, for example, maintains a physical space for trader information desks, which allows traders continuous access to infor- mation, mediation services, and feedback for sustained behavioral change. Examples of similar interventions in other contexts include the Democratic Republic of Congo National Railway Company explained in box. 4.2. Based on case studies in the World Development Report 2012, for such arrangements to work effectively, a significant percentage of employees at the information centers must be female because they understand the key problems women traders face better than men do and their presence is important for building a system of trust among the center, officials, and cross-border traders. Such cen- ters create a conducive and safe business environment for women by women. In addition, information centers can provide mediation, judiciary, and information services to female traders. Although there is little information on behavioral change specific to cross-border trade and the challenges faced by female traders, examples from other sectors, like the health sector, can be adapted to change behaviors among traders, government authorities, low- and mid-level border officials, and households. Approaches that encourage female cross-border traders to record challenges and concerns are critical to bettering their conditions. Strong legal and institutional reforms are needed to improve the expe- rience of women in cross-border trade. One key problem is inadequate implementation and enforcement of the law. In this regard, government institutions must be capable of carrying out the law and willing to do so and to work with supporting organizations that can effectively advocate for better conditions. Eliminating a culture of discrimination or violence against women within a border agency requires training and a zero-tolerance policy for viola- tions. In addition, women traders must be equipped with a rigorous under- standing of their rights and of legal processes as well as a safe and effective means of seeking redress if they are treated poorly. For example, the World Bank has just launched a charter that requires support from the international community, governments, cross-border traders, and officials for successful Women Cross-Border Traders, Challenges, and Behavior Change Communications 63 Table 4.1: Charter for Cross-Border Traders Target group Rights and obligations Traders 1. All individuals shall be able to cross the border without verbal or physical abuse and officials or harassment, including but not limited to sexual and gender-based violence. 2. Traders shall be processed at the border in an efficient and timely manner without discrimination. A receipt must be provided to the trader for any payment made and the payment properly recorded. 3. Only officials of the approved agencies shall be present at the border, and all border officials shall wear uniforms or identification badges that indicate their respective agency. 4. Physical checks of traders must be recorded with the reason and outcome provided. Female traders have the right to receive a physical check by female officials in a private but regulated and accountable environment. 5. All duties, fees, and taxes and the basis for their calculation shall be publicly available at the border. Any change to duties, fees, and taxes must be publicly announced at the border, with reasonable time allowed for traders to prepare, before their application. No unpublished fees or charges shall be demanded at the border. 6. Documentary requirements shall be clearly stated and publicly available at the border. Any change in required documentation must be publicly announced at the border with reasonable time for traders to prepare before implementation. Simplified procedures should be applied to small traders. 7. Traders should be aware of their rights and obligations when crossing the border. Traders must present required documentation and pay appropriate duties at the border and obtain a receipt for any payments made to an official. Traders shall not attempt to bribe any official to avoid payment of duties or to obtain preferential treatment in any way, including avoiding queues. Government 1. By [agreed time], these basic rights and obligations governing cross-border movement of goods and people shall be clearly stated in the local language and visibly apparent at all border crossings. 2. By [agreed time] at every border post at least one agent shall have received gender awareness training; all senior officials at the border shall have received gender awareness training by [agreed time]; 50 percent of officials at any border post shall have received gender awareness training by [agreed time]. 3. At all border posts, traders shall have recourse to an independent and confiden- tial mechanism for registering violation of any of these basic rights. Female trad- ers must be able to register the violation of any basic rights with a female staff. 4. Strict disciplinary measures shall be taken against officials found to have violated the rights of a trader. 5. Support organizations of informal cross-border traders shall be allowed to disseminate information on these rights and obligations and to deliver advice and information to enhance the capacities of the traders. 6. Governments shall continue to improve the quality of infrastructure at all border crossings to provide an open and safe environment for traders, with attention to the specific needs of women traders, and appropriate facilities in which officials may undertake their work. 7. Governments shall improve the quality of data on small traders collected at all border posts, including the number passing through the border each day and the nature of the goods carried 64 Women and Trade in Africa: Realizing the Potential Box 4.2: Democratic Republic of Congo National Railway Company In an intervention conducted by UNICEF, the Democratic Republic of Congo National Railway Company, in collaboration with the unions, set aside fund- ing to compensate surviving widows and orphans of employees who are due retiree benefits. The funds were directed to the beneficiaries and, in cases where they could not be immediately found, were kept in a dedicated fund for future claims. This approach was taken because widows and orphans tend to be less likely to either know about or assert their claim when their employed spouse or parent dies. UNICEF’s research revealed contradictions and gaps in the legal framework that prohibit women from access to justice and from judicial participation, including the training of community parale- gals and watchdog groups to assist women with regard to the law. Lessons learned from the Congo Railway example apply to female traders as well and illustrate the benefits of having frameworks in place that meet the specific needs of women and of liaising with various organizations, departments, and entities that have direct and indirect impact on traders. For example, liaising with buyer organizations, groups of border officials, and other stakeholders will promote the sustainability of behavioral changes among female cross-border traders. This framework draws on legal pro- cesses to change behavior among all target groups. In many cases, women are less outspoken about their challenges; therefore, if the system of border trade is reformed to provide easy access to judicial, information, advisory, and mediatory services, female traders will face far fewer difficulties. enactment. Known as the Charter for Cross-Border Traders (table 4.1), its main objective is to facilitate small-scale cross-border trade by improving the treatment of traders and the efficiency of border procedures. The clarification of the rights and obligations of all groups at the border is an essential step toward improving the process and enhancing partnership at the border. Several innovative campaigns that improve the judicial environment for women have been launched. Technology and telecommunications play a significant role in women’s access to mechanisms for redress. For example, “Take Back the Tech” is a campaign aimed at empowering people to use information technology systems to end violence against women. It provides an online platform that enables policy makers and campaigners to see the scale of gender-based violence and to create actionable strategies for elimi- nating the problem. The initiative developed an interactive map that enables users worldwide to share their stories anonymously—especially those from Women Cross-Border Traders, Challenges, and Behavior Change Communications 65 Africa, Latin America, and Asia. An anonymous reporting platform compa- rable to “Take Back the Tech” or Kenya’s Ushahidi could be set up for women to report harassment at the border so that everyone can better understand the magnitude of the challenges faced by female border traders and policy mak- ers and campaigners will have better information. The World Development Report 2012 on gender equality and development examines female experiences and challenges from which parallels can be drawn to the problems women face in cross-border trade (World Bank 2011). The report states the need for providing timely and effective assistance from police, legal, health, and social services to women. For example, Malaysia has one-stop crisis centers at government hospitals. In the case of cross-border trade, to make doing business at borders better for women traders, service providers need to target women explicitly, thereby bringing services closer to women, who face time and mobility constraints in accessing these services. Technology plays a key role in providing access to these services as illustrated by the “Take Back the Tech” campaign. Another example of facilitating access to services also comes from Malaysia, where community paralegals and mobile legal aid clinics were provided to help women handle emergencies. The same mechanism could be implemented to improve trade conditions for female cross-border traders, whereby women can access legal, health, mediation, and social services. These platforms may also partner with other service organizations like health centers to handle a variety of female development issues that not only improve the business environment for women but also provide sustainable venues for access to various types of assistance, including legal, health, busi- ness, and family planning. Overall, demand-driven communication is necessary for creating sustain- able behavioral change. Bringing services closer to individuals (demand) can be combined with increasing the awareness of service delivery organiza- tions, particularly management, about gender issues (supply). For example, the World Development Report 2012 includes the examples of Female-Headed Household Empowerment Program (PEKKA), a program in Indonesia that trains village paralegals, with a focus on domestic violence and family law. As a result of the rise in the number of women in the judicial system and police forces responsible for addressing domestic violence, women’s access to justice has been significantly improved. Another example is that of Tamil Nadu in India, which introduced 188 all-women police units to cover both rural and urban areas and to focus on crimes against women. These units increased women’s comfort in approaching the police, including ease of assessing judicial services and reporting crimes and concerns. These examples show that when legal actions are taken to implement a system that 66 Women and Trade in Africa: Realizing the Potential caters to the needs of targeted groups, sustainable strategies in communica- tions for behavioral change can be realized. When more women are part and parcel of judicial systems, the police, mediation personnel, and women trad- ers will be more comfortable in reporting concerns, needs, and suggestions. Also, such a system sends a clear message to all border traders, officials, and stakeholders that doing business at the borders is changing. It promotes an understanding of the consequences of disobeying border rules and regula- tions or of disregarding the Charter for Cross-Border Traders presented in table 4.1 as a means of implementing clear standards, requirements, and rules governing cross-border trade. Overcoming Stereotypes: Small-Scale Traders as Legitimate Business Women Stereotypes abound when it comes to female cross-border traders. They are often perceived as “petit traders” rather than as legitimate business women. Such an undervaluing of these female traders can be seen across the value chain. Strategies for changing behavior must be used to alter the views of female traders held by both officials and the public. For real change to occur, perceptions must change at the top and the bot- tom simultaneously. At the top, policies and frameworks designed specifically to eliminate the gender gaps in the cross-border trading system should be implemented. In households, the relative bargaining power between men and women is very important in the decision-making process within the family as well as in the way the household responds to policies. The balance of power in a household tends to equalize as more information is shared. Therefore, a constant flow of information should help the men in the household better understand the value of small-scale female traders. In addition, workshops and information panels for small-scale traders could be opened up to their husbands at times to increase their understanding of their wives’ profession. Charter for Cross-Border Traders Furthermore, elimination of gender-based trade obstacles requires a change in behavior and in society’s perception of female traders. These perceptions begin to form in the home, when husbands often do not treat their wives’ trading activities seriously; this derogatory attitude towards female cross bor- der trade is also manifested in the stereotypical views of border management and government officials. A cross border trade charter like that shown in table 4.1 could improve the treatment of traders and increase the efficiency of Women Cross-Border Traders, Challenges, and Behavior Change Communications 67 border procedures. It could also gradually change perceptions of small-scale female traders. The goal is to have the charter posted and disseminated at border cross- ings to increase understanding and support for its principles among the key stakeholders (that is, traders’ associations and the customs and officials of other border agencies). It is expected that the project will inform subsequent meetings of the African Union by providing examples of specific measures that can be taken to improve conditions at the border and to facilitate cross-border trade. The charter should promote a change in the behavior of small-scale traders and border officials because it will be posted in strategic locations and will guide and inform border officials and traders on laws, standards, and regulations. Trade Fairs Trade fairs could also serve to deliver messages through interpersonal con- tact, which is often more effective than written materials and handbooks. Following the example of successful child fairs in Uganda, trade fairs could disseminate information targeted at small traders, larger traders, border offi- cials, government officials, and households and help change the perceptions of all parties (Elmendorf, 2005). Child fairs—family-oriented events where children are vaccinated and provided with vitamins and de-worming medi- cines and where women are shown proper feeding techniques—offer a model for trade fairs to follow. At these events, messages can be delivered through photo: Curt Carnemark / The World Bank 68 Women and Trade in Africa: Realizing the Potential skits, plays, demonstrations, posters, and pictographs. A report on the project noted that child fairs proved to be a cost-effective way to reach people with both health services and messages aimed at changing behavior. Project managers spent between $500 and $600 per child fair and reached an average of 450 children per event, at a cost of about US$1.00–1.33 per child for a set of services that included inoculations, growth monitoring, and vitamin A supplements. The child health fairs in Uganda have proven to be a successful nonmainstream communications strategy. In cross-border trade, such fairs can be used as forums for marketing the new charter as well as an avenue for bringing traders, officials, and government representatives together in an informal setting. Cross-Border Traders’ Associations The establishment of cross-border traders’ associations is important in legiti- mizing small-scale traders and developing their businesses into small- and midlevel enterprises. Small-scale traders must form entities, coalitions, and associations before being represented by a formal body; such organizations will change perceptions and provide institutional backing for traders. Another very important initiative being implemented in the COMESA region is the formalization of trader groups into registered associations. For the past three years, the region has been mobilizing cross-border traders in all its member states to form associations recognized by governments so that small traders (some informal) can participate in policy dialogue. COMESA approached this initiative by encouraging governments to create a policy environment that allows traders to form associations while helping private sector groups mobi- lize themselves into associations. The Uganda Export Promotion Board recently launched a pilot project to develop a trade facilitation regime for Ugandan women involved in cross- border trade. The project targeted four of Uganda’s key borders and included a fast-track clearance process, a customs integrity action plan, trade capac- ity hubs for women at each border crossing, and capacity-building modules developed and approved by women involved in informal trade. Through training on export opportunities and capacity building for trade participa- tion, the project is expected to help informal women traders formalize their agribusinesses so that they can become small or medium export businesses by strengthening the quality of their products and forming market links with international buyers. Trader participation in national policy forums and decision-making panels is key. COMESA is also spearheading the inclusion of traders in policy forums in the region. The key interface is the institution of cross-border Women Cross-Border Traders, Challenges, and Behavior Change Communications 69 traders’ associations (CBTAs). Both the COMESA secretariat and its member states are encouraging the CBTAs to participate in various national and regional policy forums. To date, the existing national CBTAs in eight COMESA member states have been accepted as participants in the National Working Group Committee on Trade where a number of policy issues are discussed and considered for review and implementation (Njiwa 2012). As a means of changing household perceptions of female traders, one day of the forum may be opened up to spouses so that husbands especially can better understand the value that female traders add to their households and to the economy at large. Mass Media, Technology, and Communications for Behavioral Change The media play an important role in changing perceptions of women and behavior toward them. While human behavior is difficult to alter, the mass media, including print and broadcast outlets, provide platforms that inform and educate people as well as laying a foundation for behavioral change. Radio and television, which are powerful catalysts for behavioral change, are important channels for educational processes. Such mechanisms for behav- ioral change have been used predominantly for health-related issues. The National Demographic and Health Survey attested to the importance of the mass media in the Nigerian HIV/AIDS field; surveys in 1999 which evalu- ated independent samples of men and women found that the most frequently photo: Arne Hoel / The World Bank 70 Women and Trade in Africa: Realizing the Potential Table 4.2: Outreach Materials for Communication for Behavioral Change Target group Outreach materials and activities Government Policy notes, websites, posters, workshops, border charter principles, trade agreements Custom officials Posters, billboards, workshops, cross-border trader handbook for customs officials, brochures, information on banned import and export items, trade agreements, customs newsletters, briefings on border reforms Traders • Print media: Border charter handbook for traders, trade agree- ments, official list of trade costs and fees, trade newsletters • Billboards: “Border trade is changing, don’t get left behind.” • Brochures: New programs and procedures in customs at borders • Fliers/posters: Steps to excellence • Leaflets (women): Report harassment or foul play: zero tolerance • Workshops: All targets discuss and analyze changes and effects • Community rallies: Rallies to advertise and advocate border changes • Information kiosks: In market squares, villages, and public places • TV/radio: Broadcast the trader charter changes • Cell phones: Send e-messages to inform traders about the formal system • Community and trade Union leaders: Use these leaders to cham- pion messages to traders about changes and getting access to brochures, posters, and leaflets about the border transformation quoted source of AIDS information was the radio (72.1 percent for men and 45.6 percent of women), followed by television (39.8 percent of men and 46.6 percent of women), and newspapers (25.4 percent of men and 9.3 percent of women). As access to radios, televisions, internet, and mobile technology con- tinue to improve in developing nations, these media channels can be better utilized to change behaviors. Workshops that bring together government offi- cials, custom officers, and traders should be organized so that the three parties can discuss the new charter and identify the role of technology in building seamless communication channels among all parties. It is also important to target outreach activities toward each of the groups separately and finally address them as a whole, since they are all stakeholders in the process. For example, in the Booster Program for the Control of Malaria in Nigeria, work- shops were important and effective channels for providing information and learning platforms for prevention and cure of malaria, as well as for access- ing care for pregnant women and children younger than five. It was a highly successful way to educate the public and bring together key stakeholders in an interactive learning space. Table 4.2 provides information on targets, outreach materials, and activities used for communication for behavioral change. Women Cross-Border Traders, Challenges, and Behavior Change Communications 71 photo: Ami Vitale / World Bank Information, Communication, and Technology Information and communication technology (ICT) has the potential to empower women in cross-border trade by increasing their access to the mar- ket. The use of mobile technology in improving business efficiency can be seen across Africa. With access to information and the heightened awareness of traders’ rights and border processes, mobile technology is changing the face of business. Thirty-eight percent or 1.1 billion of the 2.9 billion women in low- and middle-income countries have mobile phones (USAID 2012a). By expanding the use of computers, the Internet, and cell phone infrastructure for trade, governments can limit corruption and harassment and improve efficiency in agricultural markets for both men and women. For example, when Ghana computerized customs services for cross-border taxes and regulations, reports of abuse and sexual harassment decreased, and there was an estimated 35 percent reduction in occurrences of corruption (USAID 2012b). In India and Kenya, phone systems have allowed rural women to obtain direct informa- tion on the price of food crops, enabling them to negotiate better prices with middlemen. The extent to which women are computer literate and able to use technology resources may vary greatly by region and culture. The innova- tion of new ICT services should be accompanied by skill-building workshops held at locations and times convenient to women. Mobile phones are significant instruments of behavioral change and can also increase efficiency and effective information exchange for cross-border trade. In the health sector, mHealth is viewed as a promising approach to 72 Women and Trade in Africa: Realizing the Potential Figure 4.1 Charter for Small-Scale Traders: Outreach Process Workshop Government Cross-border trade Low-mid level Small-scale groups, organizations border officials cross-border traders and unions Outreach Outreach Outreach • Pamphlets/flyers • Word of Mouth • Pamphlets/flyers • Information Sessions • Pamphlets/flyers • Information Sessions • Senior staff • Information Sessions • Posters • Posters • Posters • Billboards • Billboards • Billboards • Message boards fostering behavioral change; yet more evaluations of current interventions are needed to establish stronger evidence. Estimates indicate that half of all people living in remote areas of the world had access to a mobile phone by 2012 (Atun et al. 2006), and ownership of these devices is increasing, with the greatest growth primarily in low- and middle-income countries (Mechael et al. 2010). Proponents suggest that the advantages of using short message services to communicate with others include their low cost and their abil- ity to disseminate information quickly, ensure confidentiality, and confirm delivery (Atun et al. 2006). Another significant example of the role of mobile technology is the Groupe Spéciale Mobile Association Development Fund’s mWomen Program, which aims to reduce the gender gap in mobile phone ownership by 50 percent by 2014, as well as pair the expansion of mobile phone ownership by women with technical literacy programs. Mobile communication is a promis- ing technology for improving health care delivery and promoting behavioral change among vulnerable populations. Figure 4.1 illustrates the outreach process designed for communications aimed at behavioral change for the Charter for Cross-Border Traders as well as useful materials and instruments. Aside from those listed in figure 4.1, other strategies have been identified to promote behavioral change and trade efficiency at the borders. These include the use of nonmainstream communi- cations strategies such as e-marketing, mobile telecommunications, texting, Women Cross-Border Traders, Challenges, and Behavior Change Communications 73 social media platforms, and fairs. Advancements in technology have opened doors to nonconventional strategies for behavioral change. For example, technology can be used to expand the reach of services to women as well as to monitor efficiency at borders. Below are listed some proposed tools for monitoring border operations: • Hidden cameras. • Border information centers. • Help line (or call center) to provide advice and guidance to female traders. • Mediation centers where women traders can file workplace complaints. These can serve as a liaison among traders, officials, and governments in record keeping and data collection as well as mediating issues faced by traders. These activities can also be a function of the border information centers. • Mobile tracking, recording, and information software (for example, Jamobi Kenya for bookkeeping). • Peer reviews by border officials from other countries to identify areas that need improvement and to seek ways to provide better conditions for female cross-border traders. • Unannounced site visits from senior management to curb unruly behavior at the borders. Conclusion Sharing information about cross-border traders through information and communication technology platforms, such as mobile phones, along with targeted mass media outreach initiatives, will facilitate opportunities for female traders to conduct business at the border and to be better represented in the business community. Likewise, the Charter for Cross-Border Traders, a vehicle for educating the public on traders’ rights and obligations, can also play a part in transforming the perception of female traders as “petit hustlers” into a view of them as legitimate business women. This change in perception must occur at all levels of the value chain and among traders, trade officials, and the government alike. 74 Women and Trade in Africa: Realizing the Potential References African Union. 2009. “Declaration of the Africa Women’s Decade: Gender Equality and Women’s Empowerment: A Bottom-Up Approach. Addis Ababa: African Union. Atun, R. A., et al. 2006. “The Role of Mobile Phones in Increasing Accessibility and Efficiency in Healthcare.” Vodafone Policy Series 4, Newbury, United Kingdom. Brenton, Paul, and Gozde Isik. 2012. “Africa Loses Billions in Potential Trade Earnings, Falls Short of Vast Promise in Cross-Border Business.” News release, February 7, World Bank, http://go.worldbank.org/IGK9VWUS20. Brenton, Paul, and Gozde Isik, eds. 2012. Defragmenting Africa: Deepening Regional Trade Integration in Goods and Services. Washington, DC: World Bank, http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/AFRI CAEXT/0,,contentMDK:23092452~pagePK:146736~piPK:226340~theSit ePK:258644,00.html. Brown, Scott, Christine Cervenak, and David Fairman. “Alternate Dispute Resolution Practitioner’s Guide.” [n.d.] Washington, DC: USAID, http:// transition.usaid.gov/policy/ads/200/200sbe.pdf. Curtis, B. 2009. “The Chirundu Border Post, Detailed Monitoring of Transit times.” Sub-Saharan Africa Transport Policy Program Discussion Paper 10, World Bank, Washington, DC. Dada , Eniola. 2012. “Women’s Economic Empowerment Issues Paper; Special Report: Cross-Border Trade and Women in West Africa,” OECD Issues Paper 10, Organisation for Economic Co-operation and Development, Paris. Dada, Eniola. 2012. “Cross-border Trade and Women in West Africa,” West Africa Insight, June 2012: Cross-border Trade Issue, http://www. westafricainsight.org/articles/view/183. Dejene, Y. 2002. “Women’s Cross-Border Trade in West Africa.” WIDTECH Information Bulletin. Elmendorf, E., Cecilia Cabañero-Verzosa, Michele Lioy, and Kathryn LaRusso. 2005. “Behavior Change Communication for Better Health Outcomes in Africa: Experience and Lessons Learned from World Bank-Financed Health, Nutrition and Population Projects.” World Bank, http://siteresources.worldbank.org/EXTDEVCOMMENG/Resources/ behaviorchange.pdf. Gurman, T., S. E. Rubin, and A. A. Roess. 2011. “Effectiveness of mHealth Behavior Change Communication Interventions in Developing Countries: A Systematic Review of the Literature.” Journal of Health Communication Women Cross-Border Traders, Challenges, and Behavior Change Communications 75 17 Supplement 1:82–104, https://www.k4health.org/sites/default/files/10 810730%252E2011%252E649160.pdf Higgins, K. and Turner, L. 2010. “Integrating Poverty and Social Analysis into Aid for Trade Programs: Trade Facilitation and Trade-Related Infrastructure.” Brief 3, Department for International Development, London. Lewis, Trevor, Christina Synowiec, Gina Lagomarsino, and Julian Schweitzer. 2012. “E-Health in Low- and Middle-Income Countries: Findings from the Center for Health Market Innovations.” Bulletin of the World Health Organization 90:332–40, http://www.who.int/bulletin/ volumes/90/5/11-099820/en/. Mecheal, P., Hima Batavia, Nadi Kaonga, Sarah Searle, Ada Kwan, Adina Goldberger, Lin Fu, and James Ossman. 2010. “Barriers and Gaps Affecting mHealth in Low and Middle Income Countries.” Policy White Paper, Earth Institute, Columbia University, http://cghed.ei.columbia. edu/sitefiles/file/mHealthBarriersWhitePaperFINAL.pdf. Njie Ndumbe, L. 2013. “Unshackling Women Traders: Cross-border Trade of Eru from Cameroon to Nigeria.” Africa Trade Policy Note 10, World Bank, Washington, DC. Njiwa, D. 2012. “Informal Cross-border Trade: Challenges and Opportunities: A Case of COMESA and Its STR Implementing Borders, COMESA.” COMESA, http://www.cbtcomesa.com/downloads/file/201306171654160. IDEP_ICBT%20report%20_final%20draft.pdf. Oculi, O. 2011. “Cooperation and Integration in Africa: The Case of Informal Cross-Border Trade.” http://www.codesria.org/IMG/pdf/oculi.pdf. Randriamaro, Z. 2008. “Trade, Poverty and Women’s Economic Empowerment in Sub-Saharan Africa.” Division for the Advancement of Women, United Nations, New York, http://www.un.org/womenwatch/ daw/ws2009/documents/EC-WSRWD-2008-EP1-Randriamaro%20 23Feb09.pdf. Randriamaro, Z., and D. Budlender. 2008. “Findings of UNIFEM Baseline Studies of Women in Informal Cross Border Trade in Africa.” Regional Cross Country Report, UNIFEM, CODESRIA. UN WOMEN . 2011. “Unleashing the Potential of Women Informal Cross- Border Traders to Transform Intra African Trade.” http://www.unwomen. org/publications/unleashing-the-potential-of-women-informal-cross-bor- der-traders-to-transform-intra-african-trade/. UNICEF. 2011. “Behavior Change Communication Cells and Village Information Centres: Lessons Learned: Innovations, and Good Practices.” Briefing Paper 12, UNICEF, New Delhi, http://www.unicef.org/india/12_ BCC_Cell_and_Village_Information_Centre.pdf . 76 Women and Trade in Africa: Realizing the Potential United Nations Economic Commission for Africa. 2010. “Gender and Intra African Trade: The Case of West Africa (Assessing Regional Integration in West Africa ARIA IV),” United Nations Economic Commission for Africa, Addis Ababa, http://siteresources.worldbank.org/INTAFRREGINICOO/ Resources/1587517-1271810608103/UNECA-4th-Africa-RI-Assessment- May2010.pdf . USAID (United States Agency for International Development). 2000. “Women’s Business Links: A Preliminary Assessment of Women Cross- Border Traders in West Africa,” USAID, Washington, DC WIDTECH/ USAID, http://pdf.usaid.gov/pdf_docs/PNACL930.pdf. ———. 2012a. “Women in Cross Border Agricultural Trade.” Enabling Agricultural Trade Project (EAT), USAID, Washington, DC, http://eat- project.org/docs/EAT_PolicyBrief_WomenCrossBorderAgTrade_Oct2012_ FINAL457.pdf. ———. 2012b. “Behavior Change Communication Activities and Achievements: Lessons Learned, Best Practices, and Promising Approaches,” USAID, Washington, DC, http://transition.usaid.gov/in/ newsroom/pdfs/bcc_rpt.pdf. World Bank. 2010. “Connecting to Compete: Trade Logistics in the Global Economy – The Logistics Performance Index and Its Indicators 2010.” World Bank, Washington DC. World Bank. 2011. World Development Report 2012: Gender Equality and Development. Washington, DC: World Bank, http://siteresources.worldbank.org/INTWDR2012/Resour ces/7778105-1299699968583/7786210-1315936222006/Complete-Report. pdf. The Gender Dimension of Uganda’s Cotton Sector 77 5  The Gender Dimension of Uganda’s Cotton Sector John Baffes and Laura Maratou-Kolias Introduction This chapter looks at gender-based differences in productivity in Uganda’s cotton industry, using data from a study conducted in 2009 and 2010. Exploiting opportunities to remove these differences and augment produc- tivity would improve the competitiveness of the sector, raise output, and increase exports. These changes in turn would have positive impacts on the incomes of the households of women cotton growers in Uganda with spill- over effects into the wider rural economy. Hence, understanding the source of these differences in productivity between male and female cotton growers is important. Cotton is Uganda’s third-largest export crop after coffee and tea. During the past few years, cotton has accounted for about US$50 million, the high- est export level since the mid-1970s. Cotton is now the fifth-largest primary commodity export item after fish, gold, coffee, and tea. The industry consists of small landholders with an average farm size of about one acre. It is the main source of income for some 250,000 low-income households. Cotton in Uganda is a rain-fed crop with minimal use of purchased inputs. The average grower produces about 100 kilograms of cotton lint per year, equivalent to a gross income of about US$100 at current cotton prices. Introduced in the early 20th century, cotton became Uganda’s most important cash crop; it reached 60,000 tons during the early 1930s and sustained that level of production for four decades (valued at approximately US$300 million per year in today’s terms). The political and economic turmoil of the 1970s, however, drastically reduced output, which fell from 78,000 tons to 14,000 tons in just four years (1972–76), reaching a record low of 2,000 tons in 1987. In the late 1980s, Uganda embarked on an aggressive reform program that addressed both macro and sectoral issues, including agriculture (Collier and Reinikka 2001). John Baffes is a senior economist (DECOG) at the World Bank. Laura Maratou-Kolias is a consultant (PRMPR) at the World Bank. We would like to thank Elisa Gamberoni, Paul Brenton, Kenneth Leonard, and Katya Vasilaky for their helpful comments on earlier drafts. 78 Women and Trade in Africa: Realizing the Potential The reforms introduced in the cotton sector in 1993 coupled with high prices in the mid-1990s led to a considerable supply response, with cot- ton production reaching 20,000 tons in only two years. A well-functioning research program was put in place. Growers began receiving payments promptly while entrepreneurial activity increased enormously following the entrance of many private entities at all levels of primary processing, market- ing, and trade (Baffes 2001). Despite these achievements, there is a sense that the sector lags behind its full potential. This is a legitimate concern, especially considering that recent production levels are less than a third of what they were before the 1970s. What is of even more importance is that Uganda has made important policy reforms. In light of these concerns, various reports have identified constraints that impede further growth of the sector, including a Diagnostic Trade Integration Study (see Baffes 2009) and a multicountry cotton study (see Tschirley et al. 2010). One of the constraints highlighted by those studies was the limited knowledge about key aspects of the sector at a grower level, includ- ing its size, gender composition, and performance. As an illustration of that point, consider the wide variations in the estimates of the number of households involved in the Ugandan cotton sector. Gordon and Goodland (2000), for example, reported that there are approximately 300,000 to 400,000 cotton producers. FOODNET (2002) notes that cotton contrib- utes to the employment of 2.5 million people in Uganda, which is 10 percent of the population in the east, north, and west of the country. In addition, development of the cotton sector could benefit around 15 per- cent of the rural population and significantly reduce poverty. While some variation may be due to measurement issues (that is, whether a person or a household is counted), the figures at the high end of the range most likely reflect estimates made when the sector was much larger. In addi- tion, Baffes (2009), based on focus groups, reported the existence of a large productivity gap between male and female cotton growers, thus highlighting the importance of gender in understanding productivity. To help address the need for more information on the role of gender in productivity, this chapter summarizes key gender-related findings from a two-round cotton sector survey undertaken during 2009 and 2010. It finds that female-headed households have lower cotton productivity (as measured by output per acre) than their male counterparts as conjectured in the earlier study (Baffes 2009). The rest of the chapter proceeds as follows. The next section gives a brief description of the survey along with key demographic characteristics of the surveyed households. The third section discusses the significant results of the survey, including the male-female productivity gap, The Gender Dimension of Uganda’s Cotton Sector 79 input use, and marketing char- acteristics. The last section sum- marizes and discusses some policy implications. Survey and the Demographic Setting The survey was conducted in four out of the five cotton-growing regions of Uganda: the east, north, west, and West Nile regions. While some cotton is produced in the central region, production there has been declining, and thus it was not included in the survey. The photo: Arne Hoel / The World Bank first round of data collection was conducted after the 2009 cotton harvest (from February to April). The survey was followed by training experimental interventions during the summer of 2009 (from August to October). A second round of data collection occurred immediately after the second harvest of cotton (from February to April 2010). The timing of the surveys was chosen to make full use of information on cotton harvesting. Thus, the north and east regions were surveyed first (from January to March) followed by the west and West Nile (from March to April). The survey took place in 38 villages (14 villages in each of the east and north and five villages in each of the west and West Nile). About 14 cotton-grow- ing households were interviewed in each village, equally divided between female-headed households and male-headed households in 2009. Numerous sources were used for the design of the survey. They include the instruments used for a multicountry eastern African dataset for the Research on Poverty, Environment, and Agricultural Technologies Survey in 2003 and 2005 (Yamano and Kijima 2008). Another key source was the 2007 study by the International Food Policy Research Institute that included questionnaires for a number of farm-level studies in Burkina Faso, consump- tion and supply impacts of agricultural price policies in Senegal, and agri- cultural innovation and resource management in Ghana (IFPRI 2007). Other resources included Tibouti, Flori, and Juillet (1993), Deaton (1997), and Udry and Duflo (2004). The customization of the survey to local circumstances, institutional processes, and formatting issues reflected the work of Grosh and Glewwe (2000) and Grosh and Munoz (1996). Finally, the survey used Kreider’s (1995) insights on data collection on gender-time allocation; the 80 Women and Trade in Africa: Realizing the Potential gender division of labor, access to land, credit, and recruiting; and training agricultural extension agents. The survey included 491 households in 2009 and 460 households in 2010. Table 5.1 reports the allocation of surveyed households by year, gender, and region along with a number of other demo- graphic characteristics. The average age of the household head was 48 years (51 for females and 45 for males).The proportion of household heads that had finished primary school was 55 percent for males and 22 percent for females (all-region aver- age). But completion of primary school was higher for males in the north and West Nile regions and higher for females in the west and north. Similarly, the proportion of household heads with secondary school education was 5 percent for males and 1 percent for females (all regions). In terms of house- hold size, female-headed households were smaller than male-headed house- holds (six and seven members, respectively), a pattern that is consistent across all regions. On average, in all regions except the north, female-headed households had one less child than male-headed households, and that can adversely affect family labor. On average, female-headed households had fewer adult male members than male-headed households (1.2 versus 1.8), but almost the same number of adult female members (2). Key Findings Productivity The key motivation of the study was to measure the productivity gap between male and female cotton growers and identify its main causes. Table 5.2 reports average cotton area, production, and yield of the surveyed house- holds. Beginning with the yield (measured as kilograms of cotton per acre), males achieve much higher productivity than females. In 2009, males pro- duced 23 percent more cotton per unit of land than females (205 kilograms per acre versus 157 kilograms per acre). The 2010 gap was slightly higher, at 26 percent as seen in figure 5.1.1 Similar productivity gaps have been found elsewhere in the literature; the reasons for such gaps can be grouped roughly into four categories. The first category includes human capital and technology, less knowledge of cultivation practices, and limited access to and use of extension information (Vasilaky 2013; Croppenstedt, Goldstein, and Rosas 2013; Quisumbing and Pandolfelli 2010; Meinzen-Dick et al. 2011). The second category includes limited access to credit markets, which lowers the ability to purchase pesti- cides, mechanization, draft animals, and hired labor (explained in the next Table 5.1: Selected Demographic Characteristics of Surveyed Households in Uganda, 2009 and 2010 Total East North West W. Nile Male Female Male Female Male Female Male Female Male Female Number of households surveyed 2009 253 238 90 84 91 86 37 33 35 35 2010 233 225 80 76 84 86 36 27 33 36 Demographic characteristics Age of HH 45 51 46 51 43.6 49 48 51 42.5 58 a Primary education (%) 55 22 49 20 64 28 57 27 49 11 Secondary education (%)a 5 1 4 0 3 0 11 6 8 0 b Household size 6.8 6 7.7 6.8 6.7 5.8 6.6 6 5.5 4.7 Children (<15 years) 2.5 2.7 4.1 3.2 3.4 3 3 2 2.5 1.8 Adult malesc 1.8 1.2 2.0 1.5 1.6 1.1 2.0 1.4 1.5 1.0 Adult femalesc 1.6 2.1 1.6 2.2 1.6 1.7 1.7 2.8 1.4 2.0 Note: Age of HH = the average age of the household head. a. Percentage of the household heads that have completed primary and secondary education. b. Average number of persons living in the household. c. Average number of adult males and females living in the household. The Gender Dimension of Uganda’s Cotton Sector 81 82 Table 5.2: Cotton Area, Production, and Yield in Surveyed Male- and Female-Headed Households in Uganda by Region, 2009 and 2010 Total East North West W. Nile Male Female Male Female Male Female Male Female Male Female Area planted to cotton (acres) 2009 1.83 1.30 1.44 1.00 1.53 1.12 3.80 2.60 1.60 1.16 2010 1.50 1.00 1.17 0.86 1.14 0.80 2.50 2.31 2.00 0.80 2010 (1 plot) 1.62 0.85 0.96 0.70 1.00 0.64 2.57 3.80 2.80 0.58 2010 (GPS) 1.06 0.64 0.95 0.61 0.90 0.57 1.61 1.31 1.00 0.63 Area planted to all crops (acres) 2009 4.9 3.4 3 2.4 6 4 7.4 4.4 3.6 3.2 2010 4.8 3.2 4 2.7 5 3 5.1 4.5 5.1 3.1 Quantity harvested (kgs) Women and Trade in Africa: Realizing the Potential 2009 469 269 235 136 247 106 1532 1135 520 154 2010 351 197 110 46.5 184 96 790 751 796 161 Yield (kg/acre) 2009 205 157 150 127 153 113 357 353 318 145 2010 202 150 116 60 171 141 299 256 349 200 2010 (1 plot) 225 178 157 67 191 174 240 423 378 202 2010 (GPS) 335 309 123 70 208 216 513 2040 809 180 Note: 2010 (GPS) denotes the measure of area and yield obtained by the Global Positioning System device (it was used only in the 2010 survey). The Gender Dimension of Uganda’s Cotton Sector 83 Figure 5.1 Cotton Yields of Female Farmers as a Percentage of Yields of Male Farmers in Uganda by Region, 2009 and 2010 120 99 100 85 82 86 80 77 74 74 Percent 60 57 52 46 40 20 0 All East North West West Nile Region 2009 2010 section and discussed in Croppenstedt, Goldstein, and Rosas 2013). The third category includes less favorable land quality, size, and tenure arrange- ments (Zhang 2010; Ali, Deininger, and Goldstein 2011). The fourth category includes responsibilities other than farming, such as child care and labor for housework (Quisumbing 1995). A closer look at the numbers reveals that, on average, males harvested 469 kilograms of cotton in 2009 whereas females harvested 269 kilograms. The average difference for 2010 was 154 kilograms (351 for males and 197 for females). The harvest in 2010 was lower, mainly due to adverse climatic conditions in the east and the north. In the east and the north for both years, males had almost double the production of females. In the west in 2009, there was a smaller relative gender difference (1,532 kilograms for males and 1,135 for females) but large gender gap for both years in West Nile, with men harvesting almost four times more cotton than women. One difference between the two surveys was the use of a global posi- tioning system (GPS) device during the second round. Thus, for 2010 we have three measures of area allocated to cotton: (1) the area reported by the respondents for all their cotton plots; (2) a GPS-measured area of one selected plot used to grow cotton; and (3) the farmer’s estimate of the area of this same plot. These two measurements of the same plot, in method (2) and (3), show that many farmers misreported the plot size (especially for smaller plots). On average, growers overestimated their acreage of cotton by 53 per- cent (males) and 33 percent (females), implying that the misreporting applies 84 Women and Trade in Africa: Realizing the Potential to both genders. In conclusion, the gendered-based productivity gap reported in Baffes (2009) has been confirmed by both survey rounds. Based on the 2009 survey, Zhang (2010) examined the impact of selected land characteristics on yields, including soil quality, soil type, plot slope, land tenure system, and the way in which land was acquired. Zhang (2010) concluded that loam-type soil is associated with higher yields (almost 70 kilograms per acre, a statistically significant difference).2 In addition, farm- ers’ reported subjective measures of land quality were indeed accurate as medium land quality and poor land quality decreased the yields (by 25–130 kilograms per acre). She also examined the relationship between the type of land ownership and cotton yields. Land acquisition methods included land purchased, received as gift or inheritance, rented for fixed payments, bor- rowed, “just walked in,” or other. The customary tenure system is associated with lower cotton yields (by about 75 kilograms per acre) than other forms, which is not a surprising outcome since customary land—that is, land owned by indigenous communities—is regarded as common and therefore provides little incentive for investment. Land acquired by inheritance or gift, however, produced higher yields—80 kilograms more per acre. Despite the importance of these factors in explaining some of the differences in yields, these factors do not explain the gender gap in yield; after controlling for these factors, female-headed households still harvest about 45 kilograms less per acre. Meanwhile, the knowledge of farming practices does appear to have an important gender component. In a study that used some of the survey data reported here, Vasilaky (2013) compared the impact of a year-long social network–based agricultural training program to that of a conventional cooperative education extension program for cotton. She found that the social network–based training program had a significant impact on yields for the poorest subsistence farmers, which included most female farmers, whereas the conventional training favored larger and more productive farmers. Participation in either the conventional or the social network–based training program increased both the probability of remaining a cotton farmer between 2009 and 2010 and the information acquired by the farmer. This was measured by a quiz in the 2010 questionnaire. For the average female Ugandan farmer who produces between 100 and 200 kilograms per acre, the social-based network program increased yields by more than 70 kilograms per acre, with the effect declining for the highest-yielding cotton growers. The impact of the conventional extension program on inputs and yields is being further evaluated. The Gender Dimension of Uganda’s Cotton Sector 85 Inputs Used There are notable differences in the inputs used by male and by female cot- ton producers. In all regions of Uganda and in both 2009 and 2010, a higher proportion of men than women used pesticides, except in the west in 2010, as shown in table 5.3.3 However, the variation across regions was considerable. Moreover, political, historical, economic, religious, and tribal issues divide the northern and southern regions of Uganda. In our data, the west and West Nile belong to southern Uganda, and the east and north belong to northern Uganda. These two areas have differed politically and economically since the precolonial era, and the northern area is underdeveloped. The long-standing (north-south) regional inequalities diminished during the civil insecurity in the north in the 1990s (Moncrieffe 2004). In the west and West Nile, there are fewer observable inequalities between men and women. Based on field observations in the survey, it is apparent that in the west women run orga- nized and important groups in villages and can be some of the wealthiest and most productive farmers. Not only was the proportion of males who sprayed higher than that of females, but the number of sprays was higher as well. This breakdown is illustrated in figure 5.2. In the north in 2009, men sprayed an average of 0.4 times, while women sprayed 0.1 times on average. The number of sprays in the north increased to 0.7 for males and 0.5 for females in 2010, but the gender gap remained. The reverse trend was observed in the east. The gender gap remained in the east for both years, but in 2010 the number of sprays was lower than in 2009 for both genders. The west had the highest number of pesticide sprays. In the west, there was not a gender difference, and over time the number of sprays decreased from 3.3 to 2.6 for males and for 3.0 to 2.7 for females. The number of pesticide sprays in West Nile for males remained constant over time (1.2 times) with a reduction for females (from 0.9 to 0.5). Similarly, the proportion of women that rented land for cotton grow- ing was slightly higher than that of men in 2009 (19 percent and 18 percent, respectively). In 2010, the number of households that rented land fell to around 14 percent. In 2009, the proportion of rented land was highest in the west, followed by West Nile and the east. Very few cotton plots were on rented land in the north. There were small differences in the gender breakout of labor used in cotton production. Both males and females hired labor for cotton plots 66 percent of the time in 2009 and 50 percent of the time in 2010. The propor- tions were the highest in the west followed by the east, the north, and West 86 Table 5.3: Inputs Used in Surveyed Male- and Female-Headed Households in Uganda by Region, 2009 and 2010 Total East North West W. Nile Male Female Male Female Male Female Male Female Male Female Used pesticides for cotton (%)a 2009 54 45 75 71 16 5 100 97 46 34 2010 40 28 32 16 29 22 80 85 42 22 Number of pesticide sprays 2009 1.8 1.2 2.7 1.9 0.4 0.6 3.3 3 1.2 0.9 2010 1.1 0.75 0.93 0.51 0.7 0.45 2.6 2.7 1.1 0.5 Rented land (%) 2009 18 19 15 17 6 5 49 52 25 26 2010 14 14 10 11 4 3 41 53 16 16 Hired labor (%) Women and Trade in Africa: Realizing the Potential 2009 66 63 63 58 62 84 94 77 60 66 2010 49 35 44 40 41 60 46 84 84 49 Hired draft animals (%) 2009 26 49 57 14 15 0 0 0 4 0 2010 16 33 47 8 10 0 0 0 0 4 Rented equipment (%) 2009 8 9 0 0 1 1 41 56 9 6 2010 11 6 2 0 1 1 56 36 0 0 a. The percentage of growers that sprayed at least once. The Gender Dimension of Uganda’s Cotton Sector 87 Figure 5.2 Average Number of Pesticide Sprays by Men and Women Cotton Producers in Uganda by Region, 2009 and 2010 2009 2010 3.50 3.00 3.00 2.50 2.50 2.00 Number Number 2.00 1.50 1.50 1.00 1.00 0.50 0.50 0.00 0.00 All East North West West Nile All East North West West Nile Region Region Male-headed households Female-headed households Nile. There were no gender differences in any region except for West Nile, where males had a higher proportion by 10 percentage points. As for using rented equipment to cultivate cotton, the local tradition in the west is opposed to using draft animals for humane reasons, and there- fore farmers used equipment more frequently than farmers in the other regions—41 percent of males and 56 percent of females in 2009. In 2010 in the west, the gender difference was reversed, but the magnitude was similar. West Nile followed with 6–8 percent of farmers renting equipment in 2009, but none in 2010. Almost no farmer rented equipment for cotton in the east and the north. Income and Marketing In addition to productivity and input use, the gender-based differences extend to marketing characteristics and the way in which income from cotton was spent. Table 5.4 shows the gender- and region-based differences in cotton prices and marketing. As expected, more land and higher productivity by men mean more income from cotton. On average, males earned U Sh 277,000 from cotton sales in 2009, whereas females earned only U Sh 148,000. A similar gender gap was observed in 2010. There was considerable variation across regions. In both years, farmers in the west had the highest income from cotton, followed by West Nile, east, and north (the same rank for har- vest and yields is shown in table 5.2). Income from cotton in 2010 was greater than in 2009, mainly due to higher cotton prices for all regions as harvest and yield were smaller in 2010 88 Table 5.4: Cotton Income, Prices, and Marketing Characteristics of Surveyed Male and Female Cotton Producers in Uganda by Region, 2009 and 2010 Total East North West W. Nile Male Female Male Female Male Female Male Female Male Female 2009 277 148 132 59 129 49 950 676 314 96 2010 304 163 76 34 136 76 724 626 748 146 Price per kg (U Sh) 2009 568 533 580 538 523 462 609 610 611 617 2010 806 779 780 725 756 734 874 849 900 865 Number of sale installments 2009 1.8 1.6 1.5 1.3 1.6 1.7 3.1 2 1.5 1.6 2010 1.6 1.4 1.5 1.2 1.8 1.6 1.6 1.6 1.8 1.7 Cost of transport to place of sale (thousand U Sh) Women and Trade in Africa: Realizing the Potential 2009 7 6 11 2 1 2 14 18 14 2 2010 6 4 3 2 9 1 7 10 12 5 Advance knowledge of price (%) 2009 79 62 58 49 80 56 95 80 89 78 2010 81 73 72 57 80 65 98 95 82 97 Knew the buyer prior to sale (%) 2009 82 68 84 64 75 61 95 94 67 64 2010 90 75 93 55 86 78 97 98 86 75 Arranged the sale in advance (%) 2009 44 27 30 16 41 31 69 48 33 13 2010 41 28 40 22 40 38 54 32 30 10 Started spending the money from the given installment (days after sale) 2009 6.4 5.3 5.5 5.8 5.5 6.7 9.2 2.4 5 4.2 2010 3.8 4.9 4.7 6.1 3.7 4.9 3.2 4.5 3 3.6 Total days to finish spending the money from the given installment 2009 12.5 10.5 10.4 10.6 10 6.1 14.3 11.9 23.4 20.3 2010 13 13 11 12.6 10.7 10.5 13.9 10.3 22.2 18.9 The Gender Dimension of Uganda’s Cotton Sector 89 90 Women and Trade in Africa: Realizing the Potential for all groups—except for males in West Nile. In 2009 in the east and the north, men had double the income of women, which equates to a U Sh 70,000 difference. An increase in income for both genders took place in 2010, but the gap remained the same. Cotton income reached U Sh 950,000 for males and U Sh 676,000 for females in the west (it increased to U Sh 724,000 for males and only U Sh 626,000 for females in 2010). The largest gender gap is observed in West Nile, almost U Sh 200,000 and more than U Sh 600,000 in 2010. While most of the difference in income from cotton is accounted for by plot size and yield, some of the discrepancy is a result of the lower price received by women. For example, the average price of cotton for 2009 was U Sh 568 per kilogram for males and U Sh 533 per kilogram for females. Most significant is that the gender-based price gap was consistent across the two years of the survey and the regions. The smallest gender gap occurred in the west and West Nile in 2009. In general, farmers in the west and West Nile obtained higher prices than farmers in the east and the north. The difference was about U Sh 100 per kilogram in both years. In addition to income and prices, marketing characteristics differ according to both gender and region. The third panel of table 5.4 reports the number of cotton sale installments made by men and women. Reasons for selling cotton in more than one installment include expectation for future higher prices, storage availability, immediate need for money, and access to a buyer at the time of harvest. For women, the need for cash drives them to sell cotton in fewer installments than men. Men sold their cotton harvest in 1.8 installments in 2009 whereas women sold it in 1.6. In 2010, the num- ber of installments fell to about 1.6 installments for both genders, except in West Nile, where the number of installments increased for both genders. Comparatively, there was no gender gap in the east and north, and the num- ber of installments did not change from 2009 to 2010 (about 1.5 in the east and 1.7 in the north). In the west in 2009, males sold their cotton harvest in three installments, but females sold their harvest in two. There are also some differences in the way in which cotton is brought to market. The modes of transport were mainly manual labor, followed by a personal bicycle and borrowed bicycle for both men and women in 2009 and 2010. The average cost of transportation is reported in the fourth panel of table 5.4. Males spent U Sh 7,000 to transport their cotton to the place of sale in 2009, whereas females spent around U Sh 6,000. Transportation expenses decreased in 2010 for both genders (U Sh 6,000 for males and U Sh 4,000 for females). Again, consistent with other characteristics, there was considerable variation across regions, but the patterns were not consistent. For example, in the east and in West Nile males spent more on transportation than females in The Gender Dimension of Uganda’s Cotton Sector 91 both years. But females in the north and west spent more on transportation than males in 2009 (it was reversed in 2010 in the north). As for market prices, men were more knowledgeable about the cotton price before the sale than females were, with proportions in 2009 averag- ing 79 percent and 62 percent, respectively (increasing to 81 percent and 73 percent for 2010). There were regional differences as well with farmers in the east the least knowledgeable about cotton prices before the sale; 60 percent of males and 50 percent of the females knew the price in advance in 2009. The percentages rose to 72 percent and 57 percent, respectively, in 2010. There was a gender difference in the north as well. In the west and West Nile in both years, most famers knew the price in advance, although the gender gap was 5–10 percentage points. Likewise, more male growers than female growers knew the buyer before the sale, 82 percent of males as opposed to 68 percent of females in 2009; the proportions went up by 10 percent in 2010, but the gender gap remained. Farmers in West Nile were the least knowledgeable about the buyer (about 67 percent for both genders). Farmers in the west were the most knowledge- able about the buyer (almost 100 percent), followed by the east, north, and West Nile. There was a gender difference of about 10 percentage points in all regions except for the west, where both genders were knowledgeable about the buyer. Due to men’s strong connections in the marketplace, the survey shows that 44 percent more male than female farmers arranged the sale of cotton in advance. Across regions, farmers in the west had the highest propor- tion, followed by the north, West Nile, and east. The gender gap was more photo: Yosef Hadar / World Bank 92 Women and Trade in Africa: Realizing the Potential pronounced in the east and in West Nile, where the proportion of females arranging the sale was half that of males. The survey also examined the number of days it took growers to start spending the proceeds from cotton sales. That number varies considerably across years, gender, and regions. Males began spending their income six days after the sale in 2009, while females began spending the cotton income in five days. In contrast, in 2010 the average start date of spending was four days for males and five days for females. Male farmers in the west waited nine days before starting to spend their cotton income. In all regions in 2010, farmers started spending their income earlier than in 2009. Men also held onto their income two days longer than female farmers. This is explained in the lower panel of table 5.4, which reports the number of days it took farmers to spend the income from each installment of the sale. These results show that there is not much gender variation in how long it takes for the income to be spent. However, there is a large difference between West Nile (more than 20 days) and all other regions (about 10 days). Overall, the results confirm findings from earlier studies that because cotton sales provide the first cash income of the year, it is spent quickly. Summary This chapter reported preliminary findings from a household survey of Uganda’s cotton sector. The survey, which included 491 male and female cotton growers, was conducted during 2009 and 2010. Key results from the analysis show that female cotton growers achieve lower yields than their male counterparts and that they receive slightly lower prices than males. Women have smaller cotton plots with worse soil quality and less secure land tenure arrangements, making them less able to make productive invest- ments in their land. Moreover, fewer women used pesticides, and those who did applied them less frequently than men. However, women sell their cotton more quickly than men and spend their proceeds faster, indicating that they need money more urgently than men do. Regarding marketing, females are less knowledgeable about the price ahead of the sale, and they are less likely to know the buyer and arrange the sale in advance. Two areas in which policy intervention is likely to increase the welfare of female cotton growers are enhancing the dissemination channels so that information reaches females in a more efficient manner and strengthening property rights so that women can undertake productive investments in their plots. Addressing the constraints faced by female cotton growers, in addition to improving their welfare, will have important implications for trade, both domestic and international. Indeed, earlier research (see Collier and Reinikka The Gender Dimension of Uganda’s Cotton Sector 93 2001; Deininger and Okidi 200) reported that following the removal of inter- ventions of the traditional exports (including coffee and cotton) during the reform process of the early 1990s, nontraditional exports increased as well, in turn leading to the shift from subsistence production to market-based activi- ties, especially in the north. Furthermore, according to Larson and Deininger (2001), growing cotton enhances the likelihood of marketing other crops, again leading to increased local trade and potential growth of exports. Notes 1. The analysis of the survey data is still underway. The explanation of the yield gap discussed here is based on some preliminary analysis by Zhang (2010) and Vasilaky (2013). 2. Loam soil is composed of sand, silt, and clay in relatively even concentra- tion (about 40-40-20 percent concentration, respectively). It contains more nutrients and moisture and drains better than other types of soil. 3. The survey included two experimental interventions with extension training and free pesticides for cotton given to some cotton growers in the east and the north. The nature and implications of the interventions are discussed in a forthcoming publication. References Ali, Daniel Ayalew, Klaus Deininger and Markus Goldstein. 2011. “Environmental and Gender Impacts of Land Tenure Regularization in Africa: Plot Evidence from Rwanda.” Policy Research Working Paper 5765, World Bank, Washington, DC. Baffes, John. 2001. “Policy Reform Experience in Cotton Markets.” In Commodity Market Reforms: Lessons from Two Decades, 165–90, edited by T. Akiyama, J. Baffes, D. Larson, and P. Varangis. Regional and Sectoral Studies. Washington, DC: World Bank. ———. 2009. “The ‘Full’ Potential of Uganda’s Cotton Industry.” Development Policy Review 27: 67–85. Collier, Paul, and Ritva Reinikka. 2001. “Reconstruction and Liberalization: An Overview.” In Uganda’s Recovery: The Role of Farms, Firms, and Government, 15-48 edited by Ritva Reinikka and Paul Collier. Regional and Sectoral Studies. Washington, DC: World Bank. Croppenstedt, Andre, Markus Goldstein, and Nina Rosas. 2013. “Gender and Agriculture: Inefficiencies, Segregation, and Low Productivity Traps.” Policy Research Working Paper 6370, World Bank, Washington, DC. 94 Women and Trade in Africa: Realizing the Potential Deaton, Angus. 1997. The Analysis of Household Surveys. Baltimore, MD: Johns Hopkins University Press, World Bank, and Living Standards and Measurement Study. Deininger, Klaus and John Okidi. 2001.”Rural Households: Incomes, Productivity, and Nonfarm Enterprises,” In Uganda’s Recovery: The Role of Farms, Firms, and Government, edited by Ritva Reinikka and Paul Collier. Regional and Sectoral Studies. Washington, DC: World Bank. FOODNET. 2002. “Transaction Costs Analysis: Final Report.” Prepared for the Plan for the Modernization of Agriculture, Natural Resources Institute and the International Institute of Tropical Agriculture, Chatham, UnitedKingdom. Gordon, Ann and, Andrew Goodland. 2000. “Production Credit for African Smallholders: Conditions for Private Provision.” Savings and Development 24: 54–83. Grosh, M., and Glewwe, P., eds. 2000. “Designing Household Survey Questionnaires for Developing Countries: Lessons from 15 Years of the Living Standards Measurement Study.” World Bank, http://www.world- bank.org/lsms/manage/designbook.html. Grosh, Margaret E., and Juan Munoz. 1996. “A Manual for Planning and Implementing the LSMS Survey.” Living Standards Measurement Study Working Paper 126, Washington, DC, World Bank, http://www.world- bank.org/LSMS/research/wp/a121_140.html#wp126. IFPRI (International Food Policy Research Institute). 2007. “Assessing the Impact of the National Agricultural Advisory Services (NAADS) in the Uganda Rural Livelihoods.” Discussion Paper 00724, , Washington, DC, IFPRI. Kreider, Henderson Helen, ed. 1995. Gender and Agricultural Development: Surveying the Field. Tucson: University of Arizona Press. Larson, Donald, and Klaus Deininger. 2001. “Crop Markets and Household Participation.” In Uganda’s Recovery: The Role of Farms, Firms, and Government, edited by Ritva Reinikka and Paul Collier. Regional and Sectoral Studies. Washington, DC: World Bank. Meinzen-Dick, Ruth, Agnes Quisumbing, Julia Behrman, Patricia Biermayr- Jenzano, Vicki Wilde, Marco Noordeloos, Catherine Ragasa, and Nienke Beintema. 2011. Engendering Agricultural Research, Development and Extension. IFPRI: Washington, DC. Moncrieffe, Joy. 2004. “Uganda’s Political Economy.” Report prepared for DFID (U.K. Department for International Development) Uganda, http:// www.gsdrc.org/docs/open/DOC44.pdf. Quisumbing, Agnes. 1995. “Gender Differences in Agricultural Productivity: A Survey on Empirical Knowledge,” FCND Discussion Paper, International The Gender Dimension of Uganda’s Cotton Sector 95 Food Policy Research Institute, Washington, DC, http://www.ifpri.org/ sites/default/files/pubs/ divs/fcnd/dp/papers/dp05.pdf. Quisumbing, Agnes, and Lauren Pandolfelli. 2010. “Promising Approaches to Address the Needs of Poor Female Farmers: Resources, Constraints, and Interventions.” World Development 38: 581–92. Tibouti, Abdelmajid, Yves-Antoine Flori, and Anne Juillet. 1993. “Household Surveys of Health Care Demand and Health Expenditures: A Practical Guide.” Bamako Initiative Technical Report Series 17. Tschirley, David L., Colin Poulton, Nicholas Gergely, Patrick Labaste, John Baffes, Duncan Boughton, and Gerald Estur. 2010. “Institutional Diversity and Performance in African Cotton Sectors.” Development Policy Review 28: 295–323. Udry, Christopher, and Ester Duflo. 2004. “Intrahousehold Resource Allocation in Côte d’Ivoire: Social Norms, Separate Accounts and Consumption Choices.” NBER Working Paper 10498, National Bureau of Economic Research, Cambridge, MA. Vasilaky, Katya. 2013. “Female Social Networks and Farmer Training: Can Randomized Information Exchange Improve Outcomes?” American Journal of Agricultural Economics 95: 376–83. Yamano, T., and Y. Kijima. 2008. “The Impacts of Soil Fertility and Market Access on Household Income: Evidence from Rural Uganda.” Working Paper, National Graduate Institute for Policy Studies and Foundation for Advanced Studies on International Development, Japan. Zhang, Xiaojie. 2010. “Gender Differences in Cotton Productivity.” Honors thesis, University of Maryland, Department of Agricultural and Resource Economics. Services Trade and Gender 97 6  Services Trade and Gender Antoine Coste and Nora Dihel Introduction Services offer new opportunities for women in Africa, given the rising share of tertiary female employment in developing economies. Trade in services can play a key role in leveraging these opportunities to higher levels. While trade liberalization of services has the potential to augment employment and income opportunities for women, studies that examine the impact of trade in services on labor markets, businesses, income, and access through a gender lens are largely absent. Available evidence shows that trade in ser- vices—mainly through modes 1 and 41—can provide gender-specific benefits through the increased participation of women in exports of services such as back-office processing and call centers and the increased mobility of women to provide services such as education, health, or professional services abroad (Puri 2004). Some studies also show that imports of services can increase access to energy, telecommunications, education, and health, especially for the least-favored segments of the population, including women (UNCTAD 2011). However, the increase in female employment in services that comes with trade expansion is not always matched by an equal reduction in pov- erty; liberalization may entail risks with respect to access to essential services and can generate unemployment. The specific challenges faced by women need to be understood so that trade and other policies can be designed and implemented to maximize opportunities for all. At the global level, not only did the share of females employed in services sectors (47 percent) surpass the share of males employed in tertiary activities (41 percent) in 2012, but it also exceeded the share of females employed in manufacturing and registered dynamic growth rates compared to the stagnant or declining shares in manu- facturing and agriculture (ILO 2012). Growth rates of female employment in services are equally dynamic in Sub-Saharan Africa, with women moving out of agriculture into services sectors (figure 6.1). Despite such positive devel- opments that resulted in an increase in women employed in services sectors from 25 percent in 1992 to about 30 percent in 2012, Sub-Saharan Africa’s Nora Dihel is a senior trade economist (AFTPM) at the World Bank. Antoine Coste is a consultant (AFTP4) at the World Bank. 98 Women and Trade in Africa: Realizing the Potential Figure 6.1 Evolution of the Structure of Female Employment in Sub- Saharan Africa, 1992–2012 80 70 60 50 Percent 40 30 20 10 0 1992 2002 2012 Year Agriculture Industry Services Source: ILO 2012. Figure 6.2 Evolution of the Share of the Tertiary Sector in Female Employment in Selected World Regions, 1119, 2002, and 2012 80 60 Percent 40 20 0 EU a a ia rib a & ia ic st n Pa st S sa d ld ric ric So ea cif As CI ie ope As Ea nd & a or Ca eric ia -E Af Af b & W st h e h om el ut rth n dl e Ea As ut Am ra v op id So on De No ha M ur tin Sa rE La b- he ec Su Ot Region 1992 2002 2012 Source: ILO 2012. Services Trade and Gender 99 performance in female services employment remains slightly below that of most developing and developed regions (figure 6.2). Furthermore, while detailed data on female shares of major occupational groups in services sectors are scarce for Sub-Saharan Africa, ILO (2012) shows that in developing countries including Africa women’s employment in services is most heavily concentrated in low- and mid-skilled occupations such as “clerks and services workers” and “shop and market sales work- ers.” This seems to suggest that women do not have the same opportunities to access the full range of services occupations as men, who are overrepre- sented in “craft and related trade workers” and “managerial and legislative” occupations. In addition, Staritz and Reis (2013) and ILO (2012) show that even when women work in the same occupation as men, they have fewer responsibilities, lower pay, and lower status for reasons not attributable to their skills or experience. Current analysis and research remain divided on whether services and services trade provide a ladder to more gender equality or whether these sectors continue to restrict women to low-wage, low-productivity jobs. The ambiguity on the impact of increased female services employment is accentu- ated in the case of Africa, given the acute data and information gaps. Against this background, this chapter describes the results of a cursory review of available evidence on women’s services employment and participation in services trade in Africa. In addition to anecdotal evidence and data from existing databases such as the World Bank Enterprise Surveys, the chapter also employs information from new business surveys on professional services conducted in 17 Sub-Saharan African countries. The main objective of this chapter is to summarize briefly the current knowledge of female services employment and participation in services trade in Africa and provide sugges- tions for the future research and policy agenda on this topic. What Do We Know about Female Services Employment and Participation in Services Trade in Africa? Data collected through the World Bank’s Enterprise Surveys confirm the ILO’s findings on female services employment in Africa: the proportion of women employed by services firms exceeds female employment in manufacturing, with shares comparable to but slightly lower than those observed in East Asia, Latin America, and Eastern Europe. By contrast, Africa’s share of female services employment is significantly higher than that recorded in the Middle East and North Africa and South Asia regions (figure 6.3a). A similar trend is observable for most regions for female participation in the ownership of 100 Women and Trade in Africa: Realizing the Potential services firms compared to manufacturing and the proportion of female top managers of services firms (figure 6.3b and 6.3c). A possible explanation for this trend is that it is easier for women to participate in the ownership of services firm given the predominance of small firms in services sec- tors in all regions. The results of regres- photo: Eric MIller / World Bank sions analyzing female ownership and manage- ment of services firms on the four pillars of the World Economic Forum’s “Global Gender Gap Index” (see box 6.1) reveal that the “economic partici- pation and opportunity,” the “educational attainment,” and the “health and survival” indexes are significant explanatory factors for the engagement of women in services (see annex). These results confirm that services opportu- nities in the form of increased employment and participation in ownership and management depend on the reduction of preexisting gender imbalances at the macro- and microlevels, the most critical factors related to education and training for women and inequalities in the distribution of income. These factors may further translate into a differential effect of liberalized services on men and women, given their different economic and social roles shaped by economic, sociocultural, and political factors. Regarding employment in services subsectors, it is worth noting that “hotels and restaurants” and “wholesale and retail trade” perform best in female employment, female participation in ownership, and female top man- agers (see figure 6.4). The results on female employment by services subsec- tor confirm the ILO’s findings on the concentration of women in low- and mid-skills occupations. While the large proportion of women employed in highly tradable services sectors such as tourism (hotels and restaurants), wholesale and retail trade, or information technology services suggests that trade can have a significant gender impact, the effect depends on the economic, social, and political conditions that shape the ability of women to participate in such activities. For example, a study of the tourism sector in Kenya showed that out of a total of 31 Category A KATO (Kenya Association of Tour Operators) operators2 only two were female owned and led. Access to international Services Trade and Gender 101 Figure 6.3 Female Employment, Ownership, and Leadership in Manufacturing and Services in Six World Regions 6.3a: % of female full-time employees 43 35 35 36 35 36 29 28 Percent 20 14 8 3 SSA EAP ECA LAC MENA SAR Region 6.3b: % of firms Manufacturing Services with female participation in ownership 53 53 45 39 37 37 32 Percent 24 19 16 11 6 SSA EAP ECA LAC MENA SAR Region Manufacturing Services 6.3c: % of firms with a female top manager 31 29 24 Percent 18 17 16 15 13 8 5 2 0 SSA EAP ECA LAC MENA SAR Region Manufacturing Services Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Note: SSA = Sub-Saharan Africa, EAP = East Asia and the Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia. 102 Women and Trade in Africa: Realizing the Potential Box 6.1: The Four Pillars of the World Economic Forum’s Global Gender Gap Index • Economic participation and opportunity that capture the difference in labor force participation rates gap, a combination of the ratio of esti- mated female-to-male earned income, and a qualitative variable calcu- lated through the World Economic Forum’s Executive Opinion Survey regarding wage equality for similar work, the ratio of women to men among legislators, senior officials, and managers and the ratio of women to men among technical and professional workers. • Educational attainment that is measured as the ratios of women to men in primary, secondary, and tertiary education. • Health and survival that combines the sex ratio at birth and the gap between women’s and men’s healthy life expectancy, calculated by the World Health Organization. • Political empowerment that combines the ratio of women to men in minister-level positions, the ratio of women to men in parliamentary posi- tions, and the ratio of women to men in years in executive office. Source: WEF 2013. markets and the ability to comply with the standards required by foreign tour operators are limited for female entrepreneurs. The main challenges relate to access to financial resources, care and financial responsibilities, gender expectations, and difficulties of breaking into the network of large national operators (Staritz and Reis 2013). Furthermore, the Enterprise Survey data reveal that, in general, firms with female management tend to be smaller than firms with male managers in both manufacturing and services, with ser- vices firms smaller than manufacturing firms in Africa and elsewhere (figure 6.5). Interestingly, the share of female full-time employees is substantially higher in firms managed by women, whether they are in manufacturing or in services, in Africa or elsewhere (figure 6.6). Globally, firms managed by women tend to engage less in export activi- ties than firms managed by men. In addition, the export performance of African services firms managed by women is below that observed in similar firms elsewhere in the world (figure 6.7). Moreover, African services firms with a female manager appear to be more constrained by access to finance than those with a male manager, a pattern not observed globally (figure 6.8). In a next section, we supplement this information with a case study of a specific service sector requiring skilled labor—professional services. Services Trade and Gender 103 Figure 6.4 Female Employment, Ownership, and leadership in Sub-Saharan Africa and All Countries by Services Subsector 6.4a: % of female full-time employees 50 47 44 42 35 31 31 29 Percent 27 22 23 20 21 16 Construction Maintenance Wholesale & Hotels & Transport Computer & Other & repair of retail trade restaurants IT-related services motor vehicles services Subsector All 6.4b: % of firms with countries female SSA participation in ownership 47 41 41 37 34 34 34 30 31 30 28 Percent 24 24 20 Construction Maintenance Wholesale & Hotels & Transport Computer & Other & repair of retail trade restaurants IT-related services motor vehicles services Subsector All countries SSA 6.4c: % of firms with a female top manager 33 27 25 Percent 18 13 14 12 12 10 11 10 11 8 5 Construction Maintenance Wholesale & Hotels & Transport Computer & Other & repair of retail trade restaurants IT-related services motor vehicles services Subsector All Source: Author’s calculations based on World countries Bank SSA raw data, using for each country Enterprise Surveys the latest available year. 104 Women and Trade in Africa: Realizing the Potential Figure 6.5 Average Number of Employees in Manufacturing and Services Firms with Male and Female Top Managers in Sub-Saharan Africa and Elsewhere 6.5a: In regions other than Sub-Saharan Africa Number of employees 86.5 40.0 45.2 24.1 Manufacturing Services Male top manager Female top manager 6.5b: In Sub-Saharan Africa 54.7 Number of employees 33.1 32.7 16.2 Manufacturing Services Male top manager Female top manager Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Gender Imbalances at the Managerial Level in Professional Services in Africa Professional services3 in Sub-Saharan Africa are a small but growing and significant sector that is important to many industries, including mining, manufacturing, and services in general. They can aid growth through such benefits as higher productivity, lower transaction costs, and better produc- tion processes. For example, on the basis of a survey of more than 500 firms, the study found that east African firms that use professional services are 10–45 percent more productive than firms that do not use these services. Professional services are also an important avenue for export diversification and higher exports. Services Trade and Gender 105 Figure 6.6 Proportion of Full-Time Female Employees in Manufacturing and Services Firms with Male and Female Top Managers in Sub-Saharan Africa and Elsewhere 6.6a: Full-time female employees in regions other than Sub-Saharan Africa % of employees 52 57 29 33 Manufacturing Services Sector Male top manager Female top manager 6.6b: Full-time female employees in Sub-Saharan Africa % of employees 47 42 21 25 Manufacturing Services Sector Male top manager Female top manager Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. African policy makers have started to recognize the economic importance of professional services and have initiated steps to remove the restrictions that currently prevent trade in professional services and to address the regulatory issues that allow for effective competition in an integrated regional market. For example, the development of mutual recognition agreements of profes- sional qualifications to facilitate the movement of professionals supplying services in Sub-Saharan Africa ranks high on the agenda of regional group- ings, and the Common Market for Eastern and Southern Africa (COMESA) has requested guidance from the World Bank on regional integration in profes- sional services. Despite some progress on such reforms, gender-related issues remain largely unaddressed. Most senior staff are male, with very few women holding managerial positions in professional services firms. Gender imbal- ances are acute in all professional services, particularly at the senior level in all examined African countries.4 For example, women represent only 6 percent 106 Women and Trade in Africa: Realizing the Potential Figure 6.7 Proportion of Manufacturing and Services Firms Engaged in Exports with Male or Female Top Managers in Sub-Saharan Africa and Elsewhere 6.7a: % of male and female managers in regions other than Sub-Saharan Africa 23% % of firms 11% 9% 9% Manufacturing Services Sector Male top manager Female top manager 6.7b: % of male and female managers in Sub-Saharan Africa 17% 14% % of firms 9% 5% Manufacturing Services Sector Male top manager Female top manager Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. of top senior positions and 12 percent of the next most senior positions in professional services firms in eastern and southern Africa. Furthermore, fewer firms with female top managers tend to engage in export activities than firms with male top managers (figure 6.10). Comoros, Madagascar, Rwanda, Swaziland, and Zambia are among the best performers in reaching gender balance in management positions of professional firms, with more than 15 percent women at the top management level, while female managers seem totally absent in the Arab Republic of Egypt, Ethiopia, and the Seychelles (figure 6.11). A similarly heterogeneous picture emerges at the sectoral level: while women seem to be better rep- resented in management positions of accounting and legal firms, they are almost nonexistent in the management of engineering and architectural firms in COMESA (figure 6.12). Services Trade and Gender 107 Figure 6.8 Percentage of Firms Identifying Access to Finance as the Main Constraint, with Male or Female Top Managers in Sub-Saharan Africa and Elsewhere 6.8a: % identifying access to finance as the main constraint in regions other than Sub-Saharan Africa 31 23 % of firms 15 13 Manufacturing Services Sector Male top manager Female top manager 6.8b: % identifying access to finance as the main constraint in Sub-Saharan Africa 40 % of firms 32 34 26 Manufacturing Services Sector Male top manager Female top manager Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. These figures are broadly comparable with the statistics of women at the top management level in developed countries. For instance, women made up about 12.5 percent of the members of the corporate boards of FTSE 100 companies.5 In addition, evidence from a survey carried out by the Chartered Institute of Management Accountants (CIMA) reveals that the proportion of female CIMA fellows in countries such as Nigeria and Zambia is higher than in several countries in the Organisation for Economic Co-operation and Development (OECD) (figure 6.13). Such findings seem to indicate that the overall conditions that enable the participation of women in professional activities are sometimes more favorable than in more developed countries; thus trade and regulatory reforms in professional services in Africa may have an important positive effect on women. 108 Women and Trade in Africa: Realizing the Potential Figure 6.9 Proportion of Women in Top Positions of Professional Services Firms in Eastern and Southern Africa 12% % of women 6% Most senior employee Second most senior employee Position Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Figure 6.10 Gender of Management and Export Status of Professional Services Firms in Eastern and Southern Africa Percent 16% 11% Male top managers Female top managers Gender Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Note: Sample size = 1,182. Notwithstanding the positive results highlighted in this chapter, improv- ing the gender balance seems to receive less attention in Africa than in many developed and other developing countries. Few Sub-Saharan African coun- tries have policies in place that address gender imbalances in education, in labor force participation, or in boardrooms. The experience of countries—for example, Tanzania’s National Education Policy that aims at reducing gender imbalances in higher education admissions or the experience of selected OECD countries in achieving greater gender diversity in education and employment, including at the managerial level—could serve as possible mod- els for African countries in addressing their gender disparities. Services Trade and Gender 109 Figure 6.11 Proportion of Women in Management of Professional Services Firms in Eastern and Southern Africa 40 30 Percent 20 10 0 Egypt, Arab Rep. (83) Congo, Dem. Rep. (60) Swaziland (40) Madagascar (61) Comoros (40) Zambia (60) Rwanda (35) Djibouti (30) Burundi (35) Zimbabwe (37) Mauritius (61) Kenya (100) Uganda (100) Malawi (60) Ethiopia (100) Seychelles (40) Eritrea (40) Republic of South Sudan (63) Most senior employee Second most senior employee Average most senior emloyee Average second most senior employee Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Note: Sample size in parentheses for each country. photo: Trevor Samson / World Bank 110 Women and Trade in Africa: Realizing the Potential Figure 6.12 Proportion of Women in Top Positions of Professional Services Firms in Eastern ansd Southern Africa, by Sector 16 15 Percent 10 10 10 7 2 2 Accountancy (271) Legal (268) Architecture (217) Engineering (289) Sector Most senior employee Second most senior employee Source: Author’s calculations based on World Bank Enterprise Surveys raw data, using for each country the latest available year. Note: Sample size in parentheses for each country. Figure 6.13 Proportion of Female Fellows of the Chartered Institute of Management Accountants in Selected Economies 21 Percent 18 17 13 13 12 9 9 9 7 6 6 4 4 Sri Lanka Singapore Malaysia Zambia Australia China USA France UK Nigeria Canada Hong Kong Switzerland South Africa Economy Source: http://www.cimaglobal.com/women (fellows are members with “considerable leadership experience”). Conclusion This chapter illustrates the acute need to fill the huge information gaps related to female participation in services activities and transactions in Africa. As female services employment continues to grow, systematic and sustainable data collection will become even more important. Better data and analysis will lead to a deeper understanding of what prevents women from participating in higher-skilled services jobs and of how gender dynamics Services Trade and Gender 111 shape the trade performance of services firms. Studies such as Staritz and Reis (2013), UNCTAD (2011), and Randriamaro (2006) make a strong case for the incorporation of gender perspectives into overall trade policy design and implementation. Particular emphasis needs to be placed on the specific chal- lenges and opportunities created by trade policy and to design and implement trade and macropolicies that maximize those opportunities and facilitate the successful integration of women into technologically advanced (services) sec- tors of the economy (UNCTAD 2009). Annex: Explaining Female Participation in Services This annex shows the results of regressions of (1) the proportion of firms with female ownership; (2) the proportion of firms with female top managers; and (3) the proportion of full-time female employees. The analysis is based on the four pillars of the WEF Global Gender Gap Index: economic participation and opportunity, educational attainment, health and survival, and political empow- erment. The results show that the Economic participation & opportunity, the Educational Attainment, and the Health & survival indices are significant for the three regressions. See table A.1.Figure A6.1: Regression 1 (Female Ownership) Table A.1 (1) (2) (3) female_own_serv female_mng_serv female_workers_serv econ_score 49.99*** 23.46** 28.46*** (3.58) (2.50) (3.25) educ_score 32.15* 29.83*** 37.73*** (1.82) (2.95) (4.46) health_score 248.8* 176.1** 184.2** (1.90) (2.13) (2.10) political_score 12.50 -8.444 -10.31 (0.77) (-0.79) (-0.94) _cons -269.4** -194.3** -195.7** (-2.16) (-2.44) (-2.31) N 80 65 67 R2 0.307 0.358 0.481 Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01. 112 Women and Trade in Africa: Realizing the Potential Figure A6.1 Regression 1 (Female Ownership) 40 30 ZWE C NI e( female_own_ e( female_own_ ZWE PHL PHL GUY NI C VNM 20 TUR BRA 20 AGO BWA VNM PRY FJI BRA GUY PRYMDG BHS GHA MDGGTM KGZBWA GEO serv | X ) serv | X ) KGZ ROM GEOMDA BL RETH MNG AGO BOL ROM TUR GTM ZMB BOL POL FJI CRIKENUKR ARM TTOUZB UGA LVA MOZ TZA 10 BHS BLRPOL MDA ZMB KEN UKR IDCRI SL N V HND 0 SLV N ID HND ARG NAM NPL HUNSVN NGATJMWI K JABRB M GMB ETH GHA UZB MNG ARG HUNNAM ARM MRT PAK BLZ VEN DOM LKA EST BGR COL MKD PER HRV RUS LTU KAZ BFA 0 YEM UGA NPL TT LVA BRB O SVN ZSUR BL YEM CHL SVK CMR ECU ZAF AZE PAN NGATJ K PAK MWI JAMBGR ESTVEN PER CHL DOM MKD –20 SUR BGD MUS MEX CZE AL B –10 TZAMRT LK ALT COL U KAZ HRV MEX MUS SVK URY MOZ GMB CMR BGD ECU ZAF RUSCZE PAN AZE –40 –20 BFA AL URY B –0.4 –0.2 0.0 0.2 –0.3 –0.2 –0.1 0.0 0.1 e(econ_score | X) e(educ_score | X) coef = 49.988186, coef = 32.151771, se = 13.961186, t = 3.58 se = 17.710453, t = 1.82 40 40 e( female_own_ e( female_own_ C NI ZWE C NI 20 ZWE VNM PHL TUR GUY BRA PHL PRY AGO MDG GTM 20 VNM GUY serv | X ) serv | X ) KGZ ROMFJI TUR PRY BRA AGO BWA ZMB BOL BHS MDA POL BLR UKR ETHV YEM SL GTMMDG FJI BWA KGZ CRI GHA KEN N ID HND MNG ARG ROM GEO BOL ZMB POL 0 GEO NPL NAM SVN UZ UGA TTO LV NGA A B HUN MOZ BRBTJK ZMRT BL BHS UKR KEN GHA SL MDA ID V BL N R ETH HND ARG CRI MWI SURVEN JA BGR M MNG NAMUGATT ARM PAK TZA PER MKD DOMBFA CMR ECU ZAF CHL GMB LTU LKEST A HRVCOL KAZ MEX SVK RUSMUS 0 YEM NGA ARM UZ HUN TJ B PAK SVN KSUR BRB MWILVAO NPL MOZ –20 BGD PANCZE BL ZMRTDOM JA KAZ M COL VEN GMBBGR EST TZ A PER MKD LTU HRVCHL BGD LK A URY MUS MEX SVK CMR RUSBFA ECU ZAF AZ E B AL CZEAZE PAN –40 –20 ALB URY –0.04 –0.03 –0.02 –0.01 0.00 0.01 –0.2 –0.1 0.0 0.1 0.2 0.3 e(health_score | X) e(political_score | X) coef = 248.75835, coef = 12.503412, se = 131.24442, t = 1.9 se = 16.242729, t = 0.77 Figure A6.2 Regression 2 (Female Top Managers) 20 NIC 20 NIC e(female_mng_ e(female_mng_ PHL MNG POL VEN VEN POL LV A IDN 10 IDN ROM ZWEVNM UKR 10 PHL LV ZWETUR UKR A Z SUR BL BGRMWI BHS MNG ROM BOLGEO serv | X) serv | X) TUR BOL BLZ HND GEO JAETH M BRB VNM BHS HND BGR SL VARM SUR SL NPL ARM BL PAN EST R AGO TTO KAZBFA TCD MWI NPL PRY PAN JAM EST BLR 0 V PRY URY CRI 0 BRB TTOURY CRI SVK GUY KAZ ECU GTM SVK GUY ECU CMRPERKGZ RUS MKD MDA TJK BWA AGOYEMCMR KGZ PER MKD RUS HUN SVN ETH HUN MDA BWA MEX YEM GTM BGD EA LTU LSO LK ARG CZ BGD LS TJK LKAOSVN CZE ARG BRA DOM –10 UZB –10 BFA LTU AZE MEX BRA DOM AZE UZB CHL COL TCD COL –20 CHL –20 –0.4 –0.2 0.0 0.2 –0.6 –0.4 –0.2 0.0 0.2 e(econ_score | X) e(educ_score | X) coef = 23.455017, coef = 29.831782, se = 9.3949442, t = 2.5 se = 10.103127, t = 2.95 20 NIC 20 NIC e(female_mng_ e(female_mng_ POLVEN VEN POL 10 PHL VNM IDLV NTUR AUKR MNG 10 BLZ TUR MNG UKR IDNZW LVA PHL BHS SUR BGRROM Z BL BHS SUR VNM E serv | X) serv | X) ZWE MWI BOL HND SL V ROM SLVGEO BOL BGR HND NPL 0 GEO ET PAN CRI H JA BLR EST BRB PRY AGO TTOKAZ GUY M TCD YEM URY SVK 0 YEM ARM JAM PRY KAZBRBBL EST TCPAN R D MWI ETH TTO AGO NPL ECU CMR BFA KGZ RUS URY SVK CMR BFA KGZ GUY MDATJK GTM HUN GTM BWARUS TJK MKDPER CRI MKD LSO CZEMEX HUN SVNMDA ECU BGD –10 ARM PER SVNARG LKAB LT UZ U BRA –10 BRACZEMEX DOM BGD BWA DOM UZB AZELTULSO ARG LKA CHL COL COL CHL –20 AZE –20 –0.04 –0.03 –0.02 –0.01 0.00 0.01 –0.2 –0.1 0.0 0.1 0.2 0.3 e(health_score | X) e(political_score | X) coef = 176.11906, coef = –8.4435029, se = 82.633347, t = 2.13 se = 10.731081, t = –0.79 Services Trade and Gender 113 Figure A6.3 Regression 3 (Female Workers) e( female_workers_ 20 20 e(female_workers_ ROM ROM BLZ NGA AGO NGA UKR SUR MEX 10 BLZ UKR ETH RMBRB BL 10 AGO BLR HNDARM GUY BWA JA MNG JAPER M BRB DOM RUS POL NIC FJI serv | X) MEX HND IARMBWA LVABHS TTO LVBGR ABOL KGZTUR VEN ECU SL V serv | X GUYML PERBENTTO RUS BGR 0 MLI CMR EST MNG BHS LTUURY SVK COL GTM SUR DOM POL C ESTKGZ VNM LTPHL LSO U ETH VNM NPL ZWE PHL ID CRI MDACHL KAZ AZEN MUS 0 NI BOL ECU CMR VEN NPL URY ZWE KAZ MDA COLBFA UZB BEN OMKD UZB LS HUN PRY CZE TUR SL FJI GTM V ID SVK CRI N HUN AZE MKDTJK MWI TCD –10 MWI YEM SVN BRA CHL PRYCZE BFA TJK –10 MUS SVN –20 MDG ARG BRA MDG TCD –20 YEM ARG –30 –0.4 –0.2 0.0 0.2 –0.6 –0.4 –0.2 0.0 0.2 e(econ_score | X) e(educ_score | X) coef = 28.46269, coef = 37.732997, se = 8.7477643, t = 3.25 se = 8.4589823, t = 4.46 20 ROM 20 e(female_workers_ e(female_workers_ ROM AGO BLZ BLZ NGA AGO MEX UKR NGA 10 SUR ETHBL GUY R JAM 10 UKR SUR MEX BL R GUY ARM serv | X) serv | X) PER HND BRB MLI MNG JAM ML BRB BWA I ETH HND LV TTORUS BEN A POL C BGR RUS DOM BEN PER NIC 0 ARM BWA DOM NI BOL ECU CMRTUR KGZ VENV GTM FJI BHS EST SL URY LTU 0 GTM TUR FJI BHSMNG CMR SL URY BOL LV POL BGR KGZ VEN EST V VNM A TTO ECU NPL NPL VNM PHLKAZ COL SVK MDA KAZ COL SVK LTU ZWE BFA PHL N CHL CRI BFA ID UZB LSHUN OMUS MUS UZAZE B HUN MDA IDN MWI MKD MKD PRYYEM TCD CZE YEM CZ EPRY TJK CRI –10 ZWE MWI TJK BRA –10 BRA TCD LSO CHL SVN AZE ARGMDG MDG SVN –20 –20 ARG –0.04 –0.03 –0.02 –0.01 0.00 0.01 –0.2 –0.1 0.0 0.1 0.2 e(health_score | X) e(political_score | X) coef = 184.1877, coef = –10.307619, se = 87.744564, t = 2.1 se = 10.954984, t = –0.94 References CIMA (Chartered Institute of Management Accountants). 2010. Breaking Glass: Strategies for Tomorrow’s Leaders. London: CIMA. ILO (International Labour Organization). 2012. Global Employment Trends for Women 2012. Geneva: ILO. Puri, L. 2004. “Trade in Services, Gender and Development: A Tale of Two Modes.” In Trade and Gender –Opportunities and Challenges for Developing Countries. United Nations Conference on Trade and Development, Geneva, http://www.unctad.org/en/docs/ edm20042_en.pdf. Randriamaro, Z. 2006. Overview Report on Gender and Trade. Brighton, UK: Institute of Development Studies. Staritz, Cornelia, and José Guilherme Reis, eds. 2013. Global Value Chains, Economic Upgrading, and Gender: Case Studies of the Horticulture, Tourism, and Call Center Industries. Washington, DC: World Bank. UNCTAD (United Nations Conference on Trade and Development). 2011. “Gender Equality and Trade Policy.” Resource Paper, United Nations Network on Women and Gender Equality, Geneva. ———. 2009. “Mainstreaming Gender in Trade Policy.” UNCTAD Secretariat Note, TD/B/CI/EM.2/2/Rev1, UNCTAD, Geneva. 114 Women and Trade in Africa: Realizing the Potential WEF (World Economic Forum). 2012. Global Gender Gap Report 2012. Geneva: World Economic Forum. World Bank. 2012. Survey of Professional Services Providers in Eastern and Southern Africa. Washington, DC: World Bank. World Bank. Enterprise Surveys (database), World Bank, Washington, http:// www.enterprisesurveys.org/Data. (Endnotes) Notes 1. Trade in services takes place through four modes of supply: (1) mode 1 is cross-border supply, or services supplied from the territory of one country into the territory of another country; (2) mode 2 is consumption abroad, or services supplied in the territory of one country to the consumers of another country; (3) mode 3 is commercial presence, or services supplied through any type of business or professional establishment of one country in the territory of another (for example, foreign direct investment); and (4) mode 4 is tem- porary presence of natural persons, or services supplied by nationals of one country in the territory of another. This last mode includes both independent service suppliers and employees of the services supplier of another country. 2. Kenya Association of Tour Operators (KATO) Category A firms have finan- cial capital, marketing ability, and access to tourists. 3. The examined professional services include accounting, architectural, engi- neering, and legal services. 4. The findings are based on the firm surveys covering over 1,000 providers of professional services in 17 COMESA countries conducted by the World Bank in 2012. 5. FTSE 100 Women on Board, a report published by Lord Davies of Abersoch to increase the number of women on boards of listed companies. Gender in the Tourism Industry: The Case of Kenya 115 7  Gender in the Tourism Industry: The Case of Kenya Michelle Christian, Elisa Gamberoni, and José Guilherme Reis Introduction This chapter provides a qualitative analysis of the impact of gender differ- ences in the Kenyan tourism value chain. Kenya is one of the best-established international tourist destinations in Africa. According to the most recent figures released by the Kenya National Bureau of Statistics (KNBS 2012), the sector generated over US$1 billion (K Sh 97.9 billion) for 2011, contributing 12.5 percent of gross domestic product. In 2011, the sector supported 313,500 direct jobs. Incorporating a gender dimension into analyzing the competitive- ness of the economy or a sector is critical, because gender inequality affects the extent of female participation in the economy, as well as the quality of that participation, with consequences for productivity and growth. Global tourism has consistently risen over the past 20 years, account- ing for about 6 percent of world trade and 30 percent of the world’s exports of commercial services (UNWTO 2012). Globally, tourism ranks fourth after fuels, chemicals, and food. For developing countries, global tourism is perceived as a way to earn foreign exchange, promote economic growth, and generate employment for large numbers of low-skilled workers. In 2007, for example, tourists spent US$295 billion in developing countries, three times the level of aid provided by the international donor community (Mitchell and Ashley 2010). Tourism in developing countries accounts for up to 40 per- cent of tourism arrivals and 30 percent of tourism receipts (UNCTAD 2010). Among the least-developed countries, 23 are dependent on tourism as a top- three-industry source of foreign exchange, while for 7 countries tourism is their largest revenue earner (UNCTAD 2010). In the case of tourism in Kenya, increased gender equality could help bring greater product variety by fostering more varied tourism experiences, Michelle Christian is a postdoctoral scholar for the Social Science Research Institute of Duke University. Elisa Gamberoni is an economist (AFTP1) at the World Bank. José Guilherme Reis is a country sector coordinator (ECSPF) at the World Bank. This chapter is based on Michelle Christian, “Global Value Chains, Economic Upgrading and Gender in the Tourism Industry,” in Global Value Chains, Economic Upgrading, and Gender: Case Studies of the Horticulture, Tourism, and Call Center Industries, chapter 3, edited by Cornelia Staritz and José Guilherme Reis (Washington, DC: World Bank, 2013). 116 Women and Trade in Africa: Realizing the Potential photo: © Curt Carnemark / World Bank including female-led tourism cooperatives working in traditional dance and food activities. Increased gender equality could also boost the productivity of the tourism sector by leading to improvement of employment relations and lower labor turnover. Small improvements, such as training to combat sexual harassment and enforcement of current policies, would raise worker morale and productivity. Furthermore, programs that empower women to take on male-dominated occupations, like guiding tours, and that provide women with professional support to be successful are ways to change the perception that certain tourism products are only for men. Based on desk and Kenyan field research, the chapter provides snap- shots of the gender distribution of employment and key constraints faced by women in three of the main segments of the Kenyan tourism value chain: the tour operator, the excursion package, and the accommodation provider. The final section discusses approaches adopted by the private and public sectors to improve the situation of women in the tourism sector. Many countries in Africa have enormous unexploited potential in the tourism sector. A number of these countries are at a relatively early stage in the development of the sector, for example, Sierra Leone. Constraints on the effective participation of women in the sector limit the extent to which this potential can be realized. There are therefore important lessons that these countries can draw from the experiences and policies of countries such as Kenya, which have a longer history of developing their tourism sector and the policies that support the inclusive and broad-based growth of the sector. Gender in the Tourism Industry: The Case of Kenya 117 The Tourism Value Chain in Kenya: Occupational Segregation and Limited Female Participation Although Kenya is a well-established tourist destination, its tourism market is concentrated mainly in two main product areas—safaris and beaches— and the subsections of these types of tourism are fragmented. Global tour operators play dominant roles, and a weak regulatory environment points to human-wildlife conflict management and broader development issues.1 The Kenyan tourism value chain is influenced by the outbound markets from which tourists originate and the arrangement of the global firms that are located in these markets. The largest source market is Europe, specifically, France, Germany, Italy, and the United Kingdom. The United States is Kenya’s second-largest market. These markets exhibit different distribution channels (travel agents, tour operators, or independent travelers) and types of tourist products (for example, all-inclusive beach holidays, multicountry destination safaris, or multiregion packages in Kenya). At the local level, the national distribution sector is populated by approximately 2,000 Kenyan tour operators and travel agents, mostly based out of Nairobi and Malindi, north of Mombasa (UNCTAD 2008). All firms sell some form of a safari package that includes transportation, accommoda- tion, guides, and food as well as beach tourism packages. Many packages include beach and safari combinations. Most excursion activities involve safari experiences in or around national parks and reserves. City tours, such as township tourism, are growing, along with activities connected to the photo: © Curt Carnemark / World Bank 118 Women and Trade in Africa: Realizing the Potential beach: fishing, snorkeling, and bay cruises. Most safari activities are included as part of the tour operators’ services, and beach activities are included in beach hotel holiday packages. Finally, Kenya’s tourist accommodation sector is made up of five-star luxury offerings, one- and two-star alternatives, and a range of products, including safari lodges and camps, guest houses in and around Nairobi serving as transitional lodging before and after a safari, and condos and villas for rent along the Mombasa coastline.2 The accommodation subsector has a range of ownership types, including foreign (13.7 percent), national (63 percent), and joint ventures (22.6 percent), and it contains fran- chise, management, or owner-managed operations (UNCTAD 2008). Overall, as with the rest of the formal economy, tourism jobs and busi- ness ownership are dominated by men who make up two-thirds of tour- ism employment. Although women are involved in all subsectors and are relatively evenly spread throughout job classifications, they are outnumbered and excluded in other ways, for example, from key positions considered male preserves and from access to tourists. Tour Operators Most of the tour operators are male owned, particularly the most profitable ones. Based on figures from the Kenya Association of Tour Operators (KATO), the gross annual turnover for their members is between US$140,000 and US$1.7 million (for tour operators in Category A). KATO Category A firms have financial capital, marketing ability, and access to tourists, and usually Photo: © Curt Carnemark / World Bank Gender in the Tourism Industry: The Case of Kenya 119 they find some way to harness political and community support in gaining access to local tourism destinations, for example, and an ability to establish exclusive camps in safari destinations. Yet, out of a total of 31 Category A KATO tour operators, only two are female owned and led. Most national tour operators also have distinct gender compositions within their businesses. Young women hold jobs primarily as ticketing agents, client representatives, tour consultants, and office receptionists, and men still outnumber women in the offices. Some offices have women in man- agement and higher support positions. Many of the smaller operators that are African-Kenyan owned and run are family businesses in which the father, brother, or husband takes on the highest role. Other family members are support staff. The family-based ownership structure can also limit the ability of women to find places as interns in other firms, since women from outside the family have trouble getting internships in these family-owned businesses. One of the fundamental positions for tour operators is as driver or guide. Men are perceived to have the appropriate characteristics for these roles. They can operate off-road vehicles and fix and maintain them, and they have the physical strength to lead trekking excursions. Men are also deemed more appropriate than women to lead expeditions. As a result, there are remarkably few female drivers or guides. None of the 11 gold-certified guides with the Kenya Professional Safari Guides Association3 are female, and only six female certified-silver guides exist, all of whom are of European or American background. Many tour operation camps have an almost entirely male staff. Both overt and unintentional gender discrimination perpetuates this. The remote locations of these camps are a detriment to women who have responsibilities at home and are also expected to stay close to their families. The staff hous- ing accommodations are not designed to meet the needs of women for pri- vacy from male staff. Women who do work at camps may also have to face gossip and consider their “reputation” as a result of working at such a place. In addition, for ethnic-tribal groups, opportunities for employment with tour operators pose a challenge to Maasi women in particular. Both male and female Maasai are at a disadvantage for employment at tourist camps due to their extreme lack of education, where the average may be only a year- and-a-half of schooling. Local hiring conditions, when built into land leasing contracts with Maasai, provide jobs mainly for men. Maasai women are more likely to work as artisan jewelry makers and cultural performers, but even then, they must share their income with their husbands, the Massai commu- nity, and the tour drivers and guides. Land policies overtly disadvantage women through the exemptions that exist for customary laws, which impede their access to title deeds and thus 120 Women and Trade in Africa: Realizing the Potential to collateral for loans. Customary laws protect the various traditional tribal customs and are based on patriarchal traditions and beliefs on distinct differ- ences between men and women as they pertain to roles, rights, and expecta- tions (Ellis et al. 2007; Wamai 2009). Without access to land, for example, Maasai women do not receive direct fees from tourism development on Maasai land. They are less likely to benefit from fees from conservation and leasing agreements because becoming private deed holders is difficult. They are also excluded from the negotiation and information process of the elder Maasai leadership who work with the private tour companies (Homewood, Kristjanson, and Chenevix Trench 2009). Excursion Tourism A strong gender bias also exists in the marketing of safaris—the key product of Kenyan tourism—which prevents female access to this segment of the sector. To evoke a particular visceral reaction and a consumer’s desire to buy their product, foreign and national tour operators use racially charged imagery and narratives based on gender stereotypes. The imagery of the “white hunter,” for example, is used in east Africa and particularly in Kenya to market safari products. Given the dominant focus on the safari tour, few tour operators mar- ket other tourism products. For example, little attention is given to Kenya’s potential for a vibrant cultural tourism experience. This focus would favor the female-led tourism cooperatives, which include artisan handicrafts and traditional dance and food. Accommodations Men outnumber women as hotel employees, while women are more likely to be casual workers because of the gender-bias built into some hiring mechanisms, such as maternity leave or the location of hotels far from fam- ily homes in the up-country. Given the nature of their contracts, women’s employment is thus less protected than men’s. For example, the hotel work- ers’ union has years of collective-bargaining agreements with the Kenya Association of Hotelkeepers and Caterers. However, maintaining a strati- fied labor force of permanent, contract, and casual workers to save costs is permitted by the labor code. A large number of casual workers do not have access to unions nor the support to lodge complaints related to their work. The provisions against sexual harassment, and to some extent the strengthened labor code for mothers, have not helped female hotel work- ers. The Employment Relations Act of 2007 outlines what constitutes sexual Gender in the Tourism Industry: The Case of Kenya 121 photo: Georgina Goodwin/World Bank harassment but leaves out proper measures for recourse and enforcement. Similarly, the strengthened maternity leave clause, without affirmative gender hiring strategies, can discourage employers from hiring women, which one hotel manager saw as a growing trend. These factors help maintain a male- dominated workforce and provide an environment for gender discrimination. Sex tourism is another key issue in this value chain that can be inadver- tently supported by hotels and thereby affect hotel workers, commercial sex workers, and children. A hotel’s response to sex tourism and child sex tour- ism have gendered outcomes and can be used to separate them from their competitors. Sex tourism in Kenya is not solely a female domain (Kibicho 2009), nor is it limited to the adult population. ECPAT (2008) calls Kenya an emerging child sex tourism location with potentially 15,000–30,000 child victims. While it is impossible for hotels to monitor the intentions of all their guests, they can introduce proactive policies that discourage exploitation and monitor ways in which employees may be inadvertently supporting it. Both men and women are affected, but women are more susceptible to this abuse. Best Practices for Mitigating Gender Constraints in the Kenyan Tourism Sector The constraints women face in the tourism sector are deeply rooted and require a national strategy that analyzes and eliminates gaps in economic opportunities, endowments, and agency for women. This section shares promising interventions that have contributed to a comprehensive national 122 Women and Trade in Africa: Realizing the Potential strategy for more gender-equitable participation in the tourism sector and its positive effects on the economy and society of Kenya. A starting point for many countries is the incorporation of gender issues in national tourism policies. Plans may include a monitoring or certification process that documents how tourism businesses are successfully meeting gender-equitable targets. Kenya’s 2006 National Tourism Policy (ROK 2006) states that female tourism initiatives will be encouraged through educational programs to stimulate tourism awareness as a development option in commu- nities; to foster community-based tourism projects; to ensure the quality of employment and female rights at work; and to promote women with “dig- nity” and “respect” in marketing initiatives. However, these kinds of initia- tives face challenges in their implementation, as outlined in the first section of the chapter. Efforts to educate the public are not only the responsibility of the government but also the responsibility of the private sector; it too must Box 7.1: European Buyers Promote Equitable Practices through Supplier Accountability In Europe, the Tour Operator Initiative (TOI) “consists of a commitment by tour operators to be accountable for the social, economic, and environmen- tal consequences of the supply chain” (TOI 2004, 3). Several multilateral organizations are involved in its sponsorship, including the United Nations World Tourism Organization; the United Nations Educational, Scientific, and Cultural Organization; the United Nations Environment Program; the World Wildlife Fund; and Conservation International, as well as prominent tour operators, such as TUI AG and Thomas Cook. The tour operators involved in this effort promote best practices and guidelines for tour operators to follow in integrating sustainability throughout their supply chain. Their definition of sustainability is comprehensive and includes economic, social, cultural, and environmental aspects. Brand sensitivity and the push toward responsible tourism create incentives for leading global firms to ensure that their suppliers and destination locations follow best labor practices and promote gender-sensitive policies. TOI highlights best practices by member operators and through its cooperation with destination communities. It has brought diverse stakeholders together to create and begin practices such as creating stronger links with local communities, building capacity, and imple- menting environmental protections. Destinations where TOI has worked include Side, Turkey, and Salvador de Bahia, Brazil. Source: TOI 2004. Gender in the Tourism Industry: The Case of Kenya 123 ensure appropriate resources and provide awareness of equitable gender poli- cies in new employee induction training, as well as in ongoing training. Global tour operators and global hotels and brands can also encour- age their own suppliers to implement gender parity in employment, market access, and anti-sexual-harassment policies. Box 7.1 describes the Tour Operator Initiative in Europe to improve the social, cultural, and environ- mental aspects of European tourism supply chains. At a local level, pressure could be applied through leading industry associations, such as KATO, to implement similar practices. Certifications accompanied by appropriate training for compliance in this area can also help implement gender-equitable practices. The tourism sector is characterized by a wide range of certifications, particularly those related to social and environmental sustainability. These certifications could be expanded to include gender equity and appropriately monitored. In a report commissioned by the Netherlands Development Organization, the Center for Responsible Travel argues, for example, that “responsible tourism” has done slightly better than the tourism industry as a whole during the economic recession. Several global trends support this finding, based on increased Internet bookings, the growth of online travel communities, and the “increas- ing consumer demand for travel that offers authenticity, connections with nature, environmental stewardship, and personal growth” (CREST 2009, 8). Leading firms can be encouraged to adopt third-party monitoring of suppliers in destination locations and training for compliance. An example is the Better Work Program, sponsored by the International Labour Organization and the International Finance Corporation, which aims at improving compliance with international labor standards and national laws, while promoting business competitiveness in the apparel sector. A similar program could be applied to tourism. As discussed in the previous section, the views on gender roles in key segments of the value chain, as in the case of safaris, prevent women from accessing certain occupations. Women can achieve more professional mobil- ity through dedicated programs that train them in male-dominated occupa- tions. One example in Kenya is the Karen Blixen Camp, which takes a strong gender-sensitive position by promoting gender equality in hiring, promotion, work practices, and working conditions for all employees.4 While still a long way from gender balance among employees, often due to a lack of female candidates, the company has women working in traditional male roles both as safari drivers and as spotter guides. Accommodation providers can also play a role by promoting certain tour operators’ itineraries that support women’s commercial activities and at the same time contribute to the tourism diversification of Kenya. Similarly, tour 124 Women and Trade in Africa: Realizing the Potential Box 7.2: Starwood Hotels and the Mayan World Foundation Purchase Crafts from Female Artisans for Hotel Decoration Through a joint venture in Mexico, Starwood Hotels and the Mayan World Foundation developed a program to support local craftsmanship in areas around the hotels. They set up seventeen microenterprises focused on the revival of traditional jewelry making and other handcrafts such as weav- ing and embroidery. The organization offered women training in traditional Mayan techniques and provided important access to markets for their goods, not only through sales in hotel gift shops and visits to artisan work- shops but also through sales to the hotels themselves for room décor and guest amenities. In 2007, at least 180 women were directly involved with the program, and efforts were under way to help them establish coopera- tives and explore new markets. The organization was awarded Travel and Leisure’s Global Vision 2010 prize for artisan revival. Source: Ashley et al. 2007; Travel and Leisure 2010 operators and hotels can include products that support women’s community groups among their offerings. For example, the Agape Women’s Group, with support from the Netherland’s Development Organization and the Tanzania Tourist Board, has created cultural tourism products to add to traditional safa- ris and have established links with safari lodging. Hotels should be encour- aged to source their local gift shops at least partially with products directly from women’s artisan collectives (see, for example, box 7.2). They could also devote specific days of the week or month to “artist days” with an emphasis on female artists. Hotels can work directly with national nongovernmental organizations, such as Ecotourism Kenya, in identifying collectives and in training female artisans to meet whatever standards the hotel may have. Reducing the burden of working in the home for women can free up time to participate in commercial activities. Encouraging firms to develop flexible and innovative practices that ease these constraints for their female employees increases productivity and positively affects the firm’s bottom line. One example is the Equilibre program established by the IBIS Hotel Chain, part of the Accor Group, which revises work pattern practices that ham- pered female employees’ ability to balance their productive and reproductive responsibilities. Central to the program’s success was its multiskills training, allowing employees to qualify for more than one job. The hotel saw working conditions improve with a reduction in attrition and absenteeism. Gender in the Tourism Industry: The Case of Kenya 125 Conclusion By increasing the annual supply and diversity of tourism, firms can retain their workforce year-round (thus maximizing the use of fixed-capital invest- ments) and have more incentive to train their employees, since the firms can capture more fully the returns on training. As a result, the number of temporary work contracts, mainly for women, could potentially be reduced and more secure employment and incomes be provided. Kenya’s tourism sector can eliminate some of the barriers women face by adapting work schedules to accommodate the needs of the female workforce, by training women in typically male-dominated jobs, and by creating and applying gender-equitable policies and communications. Actors along the tourism value chains—distribution, excursion activities, and accommoda- tions—can follow examples from the practices outlined in this chapter to sup- port gender parity in job distribution and greater tourism product and market diversification to promote the competitiveness of tourism goods and services in Kenya. Notes 1. Broadly speaking, Kenya is labeled as “overburdened with institutions and regulations” within the tourism sector, with 44 different legislative instru- ments being used (World Bank 2010, 33). Tourism is somewhat unique: poli- cies on land use, environmental protection, and indigenous group rights take on importance in addition to the traditional industrial policy base. 2. The Kenya Tourist Board reported in 2006 a total of 174 hotels, including city hotels, vacation hotels, and lodges. Of those, only 17 were five star. More recent bed occupancy figures from the Kenya National Bureau of Statistics indicate most bed occupancies for 2009 were located in coastal beach areas (3,011,400) and Nairobi (1,164,100), although Kenya as a whole only had a 37 percent occupancy rate that year due to political unrest. In 2011 the occu- pancy rate rose to 45 percent. The central and Maasailand zones, which are dominated by game reserves, are also building their bed capacity. 3. The Kenya Professional Safari Guides Association is a volunteer regulating organization that works in collaboration with tourism-related ministries and county councils to set the standards for safari driver guides. 4. See http://karenblixencamp.com/. 126 Women and Trade in Africa: Realizing the Potential References Akama, John S. 2004. “Neocolonialism, Dependency and External Control of Africa’s Tourism Industry: A Case Study of Wildlife Safari Tourism in Kenya.” In Tourism and Postcolonialism Contested Discourses, Identities, and Representations, 130–152, edited by C. Michael Hall and Hazel Tucker. London, England: Routledge. Ashley, Caroline, Peter De Brine, Amy Lehr, and Hannah Wilde. 2007. “The Role of the Tourism Sector in Expanding Economic Opportunity.” CSRI Report 23, Harvard University and the Overseas Development Institute, Boston. Christian, Michelle. 2013. “Global Value Chains, Economic Upgrading and Gender in the Tourism Industry.” In Global Value Chains, Economic Upgrading, and Gender Case Studies of the Horticulture, Tourism and Call Center Industries, 43–70, edited by Cornelia Staritz and José Guilherme Reis. Washington, D.C.: World Bank. CREST (Center for Responsible Travel). 2009. “The Market for Responsible Tourism Products with a Special Focus on Latin America and Nepal.” Netherlands Development Organisation, The Hague. ECPAT (End Child Prostitution, Child Pornography and Trafficking of Children for Sexual Purposes). 2008. “Combating Child Sex Tourism Questions and Answers.” ECPAT International, Bangkok. Ellis, Amanda, Jozefina Cutura, Nouma Dione, Ian Gillson, Clare Manuel, and Judy Thongori. 2007. Gender and Economic Growth in Kenya: Unleashing the Power of Women. Washington, DC: International Finance Corporation. Homewood, Katherine, Patti Kristjanson, and Pippa Chenevix Trench, eds. 2009. Staying Maasai? Livelihoods, Conservation and Development in East African Rangelands. New York: Springer. Kibicho, Wanjohi. 2009. Sex Tourism in Africa: Kenya’s Booming Industry. Hampshire, England: Ashgate. KNBS (Kenya National Bureau of Statistics). 2012. Kenya Facts and Figures 2012. Nairobi: KNBS. Mitchell, Jonathan, and Caroline Ashley. 2010. Tourism and Poverty Reduction: Pathways to Prosperity. London: Overseas Development Institute. ROK (Republic of Kenya). 2006. “National Tourism Policy, Final Draft.” Ministry of Tourism and Wildlife, Republic of Kenya. SNV (The Netherlands Development Organisation), Lao National Tourism Administration, Mekong Tourism Development Project, Lao Women’s Union, and Gender Resource Information and Development Center. 2005. Gender in the Tourism Industry: The Case of Kenya 127 “Gender Assessment and Strategic Action Plan of the Tourism Sector in Lao PDR.” Vientiane: SNV.  TOI (Tour Operators’ Initiative ). 2004. “Supply Chain Engagement for Tour Operators: Three Steps toward Sustainability.” Madrid, Spain: Tour Operators’ Initiative for Sustainable Tourism Development and the Center for Environmental Leadership in Business (CELB).Travel and Leisure. Travel and Leisure. 2010. “Global Vision Awards 2010: Artisan Revival, The Haciendas, Yucatatán, Mexico,” November, http://www.travelandleisure. com/articles/global-vision-awards-2010/3. UNCTAD (United Nations Conference on Trade and Development). 2008. FDI and Tourism: The Development Dimension. Geneva: UNCTAD. . 2010. The Contribution of Tourism to Trade and Development Note by the UNCTAD Secretariat. Geneva: UNCTAD. UNWTO (United Nations World Tourism Organization). 2011. Global Report on Women in Tourism 2010: Preliminary Findings. Geneva: UNWTO. . 2012. UNWTO Tourism Highlights 2012 Edition. Madrid: UNWTO. Wamai, Njoki E. 2009. “Women and Poverty (Land Rights and Ownership).” Policy brief on the Kenyan situation prepared for the African Women Rights Observatory, Addis Ababa, Ethiopia. World Bank. 2010. Kenya’s Tourism: Polishing the Jewel. Washington, DC: World Bank. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 129 8  Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa Thomas Bossuroy, Francisco Campos, Aidan Coville, Markus Goldstein, Gareth Roberts, and Sandra Sequeira Introduction Exporting is commonly regarded as one of the main drivers of firm growth. Competing in international markets is often associated with higher levels of productivity and better business practices. Consequently, understanding how to support the development of businesses so that they are able to access and be competitive in export markets is a policy priority for stimulating private sector–led growth. The trade literature highlights barriers to entry for exporting, requiring that firms reach a certain size and quality to compete in foreign markets. These constraints may be particularly binding for female-owned busi- nesses, which are often younger and smaller than their male-owned com- petitors. The available evidence shows that female-owned businesses tend to be in sectors with low barriers to entry, have limited potential to grow and cater to foreign markets, and face constraints in accessing the factors of production that would allow them to reach sectors with export potential (Hallward-Driemeier 2011). In developing countries, most of the analysis on constraints to the growth of women-owned enterprises has been focused on microenterprises, where female-owned businesses tend to be concentrated or, alternatively, on regional comparisons using small within-country samples Thomas Bossuroy is a young professional (AFTSW) at the World Bank. Francisco Campos is an economist (AFTFW) at the World Bank. Gareth Roberts is a consultant (AFTFW) at the World Bank. Aidan Coville is an economist (DECIE) at the World Bank. Markus Goldstein is the Africa Gender Practice Leader (AFTPM) at the World Bank. Sandra Sequeira is a lecturer in development economics in the Department of International Development at the London School of Economics, a research affiliate at The Suntory and Toyota International Centres for Economics and Related Disciplines (STICERD), and the lead academic for Mozambique at the International Growth Center (IGC). The authors are thankful for the support of the Knowledge-for-Change Program, the Gender Action Plan, and the Bank-Netherlands Partnership Program Trust Funds, which funded this work. We would also like to thank Elena Bardasi, Ana Margarida Fernandes, Neil Rankin, and Volker Schoer for their comments. All views expressed in this chapter are those of the authors and do not necessarily reflect those of the institutions they represent. 130 Women and Trade in Africa: Realizing the Potential photo: Trevor Samson / World Bank of women-led small and medium enterprises (SMEs). However, there are a number of female entrepreneurs—especially in middle-income countries with large income disparities like South Africa—whose businesses have grown beyond survivalist microenterprises and become more fully fledged micro, small, or medium enterprises (MSMEs), competitive in both domestic and international markets. MSMEs represent more than 95 percent of the total number of firms and employ more than 50 percent of the workforce in South Africa (World Bank 2007). SME growth is often seen as an important driver of job creation in South Africa and more generally (for South Africa, see DTI 1995; National Planning Commission 2012). The National Export Development Programme (DTI 2013) is a recent example of the importance that the South African government is placing on promoting exports by creating an enabling environ- ment for businesses—especially Black, African, and female-owned busi- nesses—to access international markets. Hence, understanding what sets the women-owned exporting businesses apart from the majority of female-owned microenterprises is critical. Few datasets offer data rich enough to conduct gender-disaggregated analysis of potential determinants of access to markets. While the Enterprise Surveys of the World Bank Group use cross-country analysis to study the nature of higher-end firm dynamics, limited sample sizes and geographic heterogeneity often constrain detailed within-country gender-disaggregated analysis. Alternatively, larger detailed household surveys or firm-level impact evaluations often focus on microenterprises at the lower end of the economic Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 131 spectrum, offering little insight into issues facing SMEs, such as barriers to export markets. This chapter uses data from the 2012 survey of businesses that employ up to 25 workers in the Kwazulu-Natal Province (KZN) of South Africa. This dataset of more than 2,400 businesses comes from the baseline survey of an impact evaluation on increasing the market access of emerging SMEs. We use these data to identify the determinants of firms’ propensity to export and determine if these vary according to the gender of the business owner. Then, given the importance of gender in explaining exporting, we explore the determinants of firm size and differences in these between male- and female- owned businesses.1 Strikingly, we find no difference in the likelihood of exporting between female- and male-owned firms in our sample. This is the case both before and after controlling for factors commonly identified as exporting enablers: firm size, access to markets, access to finance, owner’s skills and experience, length of establishment, and sector. We discuss how that similarity may be related to this particular sample and driven by the self-selection of businesses as potential beneficiaries of the networking marketplace being evaluated, as well as the initial sample frame of firms made up of MSMEs with a relatively high degree of sophistication. In the following section, we use other repre- sentative datasets from KZN to show that (1) women are disproportionately more likely to own microenterprises than SMEs and (2) there is a posi- tive relationship between firm size and the likelihood of exporting, which strengthens the argument that the equalization of export propensity across gender is specific to the group of firms targeted in this analysis. This finding likely reflects the homogeneity of business-exporting potential (regardless of gender) for population subgroups but not necessarily across the full spectrum of MSMEs operating in the country. We do, however, identify in our sample the differences in the character- istics of men and women that export. These include gender differences in the importance of sectors, risk-preferences, access to networks, and skills. Although being in a manufacturing or agricultural sector is strongly asso- ciated with exporting potential for male-owned businesses, female-owned businesses access international markets mostly through the services sector. According to our study, female-owned businesses are significantly less likely to export when they operate in multiple sectors simultaneously—a finding that can be interpreted as a proxy for “opportunistic” rather than strategic firm behavior or as high risk aversion. In addition, networking activities such as finding new customers online are significantly correlated with propensity to export across the sample but more strongly for women. Female exporters are also more educated than their male counterparts, with all of them having 132 Women and Trade in Africa: Realizing the Potential completed at least secondary education. Finally, our evidence suggests that business partnerships seem to play an important role in the firm size of female-owned businesses, a relevant finding given the significance of firm size for exporting potential.2 The results presented here should be treated with some caution since we do not address the possibility of either (1) reverse causality, in which exporting and firm size determine the covariates of interest rather than the other way around, or (2) omitted variable bias, in which other factors cor- related with both exporting and firm size and our covariates may be driving the relationship. However, the goal of this chapter is to help identify ways for female- owned businesses to access international markets and offer insights into future areas of research. Based on our analysis, we identify three main areas of focus for policy aimed at promoting women’s entrepreneurship: (1) risk preferences associated with women-led businesses operating in multiple sectors, with limited investments and limited expectations for growth in each of them; (2) skills development in targeted areas based on evidence of the importance of education and management processes; and (3) links that address underlying constraints in building business networks. The chapter begins with a brief overview of the literature on gender gaps in entrepreneurship and its relationship to trade. This is followed by a descrip- tion of the dataset used for the analysis and a comparison with other datasets used in the region. The next two sections present an overview of the charac- teristics of female-owned enterprises and regression analysis of the potential factors influencing the propensity to export and grow. The final section offers policy recommendations linked to a more rigorous research agenda. Literature Review Female-owned enterprises are likely to be smaller than male-owned busi- nesses.3 There is also evidence that smaller female-owned firms are less productive than male-owned firms with the same number of employees (see, for example, Bruhn 2009). In their work on the constraints to productivity growth, Berniell and Sanchez (2011) identify the importance of time con- straints related to household responsibilities in determining women’s prefer- ences on the decision to start and grow a business. Sabarwal, Terrell, and Bardasi (2009) analyze the differences between male- and female-owned firms in Eastern Europe, Central Asia, Latin America, and Sub-Saharan Africa. They find that while female-owned firms generally have fewer employees, differences in productivity, firm growth, and return on investments are less pronounced. Furthermore, the authors find no Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 133 evidence of discrimination in access to formal finance and argue that much of the gap in firm size can be explained by the specific characteristics of the sectors in which the female-owned business operate. Bruhn (2009) shows that female-owned firms in Latin America are smaller in sales, costs, and physical capital. While they have lower profits, this is often related to the quantity of labor and capital inputs that female- owned firms employ. The author finds no evidence that differences between male- and female-owned firms are due to differential access to finance or to regulatory burdens. The study does, however, suggest that there is less demand for such finance by women. Klapper and Parker (2011) support this finding, discovering no evidence of explicit legal discrimination. In contrast, they argue that the sector selec- tion itself may be driven by barriers to accessing finance since women are generally “better represented in labor intensive sectors… than [in] capital intensive manufacturing industries.” The authors suggest that part of the rea- son for this difference is that women are generally less likely to have physical and “reputational” collateral. In South Africa, the Financial Services Regulation Task Group (2007) sees the main financing challenge facing SMEs is their inadequate collateral and a limited formal credit history. This deficiency, consequently, may make it more difficult for newly established SMEs, particularly those owned by Africans, to finance any expansion. In contrast, FinMark Trust (2010) argues that there is considerable heterogeneity across subregions with regard to the self-reported problems that SME owners in South Africa face. Major self-iden- tified constraints include access to finance and suppliers or the infrastructure required to support operations. In the KZN Province studied in this chapter, access to finance and cash flow problems were identified as the most impor- tant self-reported constraints to growth by FinMark Trust. Another aspect of the relationship between gender and entrepreneurship is the importance of social capital. Baughn, Chua, and Neupert (2006) see a positive relationship between “normative support” for women’s entrepre- neurship and the ratio of female- to male-owned firms. They find that this “normative support” is “embedded in overall attitudes about entrepreneur- ship and gender equality.” Fletschner and Carter (2008) find evidence to sup- port their contention that such social factors may limit the ability of women to accumulate entrepreneurial capital, and Macours and Vakis (2009) find that social interactions have an effect on aspirations. Indeed, the Department of Trade and Industry argues that “narrowing the gap between the growth in women[’s] entrepreneurship and the contextual reality is contingent on skills training and tertiary education, removal of hidden and subtle gender dis- crimination, change in existing prejudices and stereotypes regarding the role 134 Women and Trade in Africa: Realizing the Potential of women in a male-dominated economy, demand for socio-economic rights, and policy advocacy” (DTI, 2005, page 10). Similarly, De Bruin, Brush, and Welter recognize the effect that self-per- ceptions may have; they argue that women may be more inclined to perceive incorrectly that “they may not have the right opportunities and know-how to start or grow their own businesses” (2007, page 330). IFC’s Access to Finance for Women Entrepreneurs (IFC 2006) suggests that in South Africa female entrepreneurs may be less confident about approaching financial institutions for loans. Several of these obstacles to female entrepreneurship are also likely to constrain participation in international trade. And yet, while there is an exten- sive literature on access to foreign markets more generally, the gender dimen- sion has remained relatively unexplored. The first such general barrier is the productivity of the firm. According to Wagner (2007), the available evidence suggests that “exporters are more productive than non-exporters” and that “more productive firms self-select into export markets, while exporting does not necessarily improve productivity” (2007, page 66). Thus, if female-owned firms are less productive, fewer are likely to participate in export markets. There is also evidence of a number of underlying constraints to increased productivity and the decision to export. Berman and Héricourt, for example, show that access to finance has a significant effect on the decision to enter export markets and that in countries with low levels of financial development there is “a disconnection between firms’ productive and export status” (2010, page 206). For many of the other barriers to exporting, any association to gender may be less obvious. For example, Freund and Rocha find that transit delays are the most significant barriers to exporting from Africa, although these are not due to geography but are instead the result of institutional features “such as border delays, road quality, fleet class and competition and secu- ¸ eker (2011) confirms the findings of both Berman and rity” (2011, page 361). S Héricourt (2010) and Freund and Rocha (2011) that bottlenecks such as these, and regulatory quality, have an effect on exporting. Kneller and Pisu (2011) highlight the importance of nontariff barriers, including language differences, imperfect information, and institutional quality. However, Kneller and Pisu present evidence to show that “the higher the export experience of firms, the lower the trade costs they face” and that these barriers are not related to productivity and size, which are generally associated with entry into exporting (2007, page ii). In other words, while the productivity and size of female-owned enterprises may not necessarily inhibit export growth, there is a virtuous cycle associated with experience in exporting. Consequently, if there are any female-specific barriers to exporting, Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 135 these may inhibit the learning required to reduce trade costs. Furthermore, Edwards, Rankin, and Schöer point out that in South Africa there is evidence of a “robust size-exporting relationship” and argue that this may be due to the high fixed costs associated with “making contact with, or travelling to meet foreign buyers, organizing bank accounts or export permits, or investing in new capital equipment…to produce goods of the required standard” (2008, page 16). They argue that these costs elevate the importance of expectations about future market conditions, especially for smaller exporting firms. Rankin (2013) finds evidence that many existing exporters in South Africa export only a small proportion of their output and that again exporting is associ- ated with productivity and growth; he suggests that increasing the export volume requires existing firms to export more, while new exporters first need to increase their export potential—that is, to grow and become more produc- tive. Gumede (2004) focuses on the extent to which SMEs in South Africa are likely to become exporters, also finding that most SMEs that export do so only sporadically. As a result, SMEs most likely need to partner with larger firms to overcome many of the barriers to exporting. Consequently, if female-owned firms are more risk averse (perhaps for social reasons) or if they have smaller networks of contacts, those tendencies reduce the likelihood of exporting. photo: Trevor Samson / World Bank 136 Women and Trade in Africa: Realizing the Potential Data: Source and Sample In the absence of firm-level census data in the majority of developing coun- tries (and particularly in Africa), the recent studies on gender and entrepre- neurship have used four main sources of data. The first source is household surveys, including the Living Standards Measurement Surveys. The main issues with these surveys for enterprise-level analysis are the short modules on business characteristics and the oversampling of household-level small- scale enterprises at the expense of larger firms. Other studies have relied on targeted household surveys, including the FinScope SME survey focused on access to finance (FinMark Trust 2010). Due to their design, these surveys are likely to have a high concentration of micro survival-type firms. A third source of data comes from the World Bank Enterprise Surveys of larger firms, which use a pool of data from different countries.4 These data are con- strained by the small sample sizes of within-country women-owned enter- prises, limiting the possibility for country-level or regional analysis. Finally, the constraints to female entrepreneurship are studied at length in impact evaluations of financial and private sector development.5 These studies are mostly focused on microenterprises. For example, samples of businesses in Fafchamps et al. (forthcoming) and Dupas and Robinson (2013) have typi- cally no paid workers. These constraints on the availability of data have hence limited the pos- sibility of studying the relationship between gender and the development of more sophisticated firms, such as those participating in international trade or business-to-business relations. This chapter contributes to the literature by analyzing a large dataset (2,422 enterprises) of businesses with up to 25 employees in KZN, South Africa.6 The dataset is rich in gender-disaggregated business char- acteristics of firms with workers, a group of businesses overlooked in the country-level literature due to lack of available data. The data were collected as the baseline survey for the impact evaluation of a networking market- place connecting MSMEs with larger enterprises through an online- and cell phone–based platform. Throughout this chapter, we define female-owned businesses as those in which the business owner responding to the survey is a woman.7 In over 95 percent of the cases, this respondent is the main source of decision making in the business, although in one quarter of the cases, this owner needs to con- sult a business partner or a manager. In two-thirds of the cases, the business partner consulted is also female. The setting for our analysis is a middle-income country where income inequality8 and firm-level heterogeneity are high. Given the limited quality Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 137 of firm-level registration data, the sampling frame was based on lists of businesses from various sources, including a major partner commercial bank, the local chamber of commerce, government programs,9 and a private database of firms. Using these different sources ensured that the businesses sampled had been previously screened for quality and size by these credible organizations. The survey was conducted over the phone with the main business owner and elicited information on the entrepreneurs’ profile, business character- istics, performance, access to technology, access to markets (including cor- porate and government contracts, as well as export markets), and access to finance. To achieve completion of the 2,788 interviews, the survey firm con- tacted over 20,000 companies.10 During the interviews, respondents were first asked two questions to screen whether they would qualify for the networking marketplace impact evaluation study: (1) whether the firm sells products or provides services to other businesses (a necessary condition for them to benefit from the business-to-business marketplace); and (2) the number of workers the business employs (given the focus of the intervention on small business development, the sample was restricted to firms with a maximum of 25 employees). The survey is thus not representative of the firm population in South Africa or of KZN. It captures details of a specific group of businesses that are large in relative terms—that is, the average number of workers including business owners working in the business in this sample is 7.1—and that have a certain degree of sophistication and knowledge of best business prac- tices. Given that the datasets used included business lists from government departments, banks, and chambers of commerce, the sample overrepresents formal firms. We can draw some comparisons between our pool of businesses and the samples of two other major business surveys in South Africa: the Finscope SME survey (FinMark Trust 2010) and the World Bank Enterprise Survey (World Bank 2007). The Finscope SME Survey was a firm-level national sur- vey that identified businesses by sampling households and asking whether or not household members operated any business. This survey (892 businesses in KZN) comprises mostly microentrepreneurs, including the self-employed, while the Enterprise Survey (127 businesses in KZN) is weighted more heav- ily toward medium and large businesses. The KZN impact evaluation survey presented here studies firms ranging in size between these two, with a focus on MSMEs. Table 8.1 provides an overview of the entrepreneurs’ and businesses’ average characteristics. Female-owned firms make up approximately one- third of our sample. This compares with 60 percent of the firms in the 138 Women and Trade in Africa: Realizing the Potential KZN Finscope SME Survey and just 13 percent in the Enterprise Survey.11 Compared to the age (10 years old on average) and size (7 workers) of busi- nesses in our sample, the firms in the Finscope SME Survey are on average younger (seven years old) and smaller (1 worker), while the businesses in the Enterprise Survey are older (19 years) and larger (67 employees). Our sample also finds itself in an intermediate position in the race distribution of ownership. Fourteen percent of our businesses are white-owned, which com- pares with 6 percent in the Finscope survey and 33 percent in the Enterprise Survey. The propensity to export follows the same pattern. Five percent of businesses in our sample export, whereas just 1 percent of the Finscope busi- nesses do and 16 percent in the Enterprise Survey. A comparison with other data sources shows that our sample excludes the bottom and top ends of businesses in South Africa. It is composed of businesses established enough to have entered databases of the government or commercial banks’ SME departments and sophisticated enough to be inter- ested in an online networking platform. It excludes the bottom end of the distribution, which is larger for women-owned businesses.12 But the sample is also limited to businesses with no more than 25 employees and includes only businesses that expressed an interest in broadening their customer base. Those with a large enough pool of existing customers are excluded. The comparison across different data sources helps us identify three things: (1) that gender is not strongly linked to exporting probability within the data sources except in the Enterprise Survey; (2) that the likelihood of exporting appears to be associated with firm size and type (observing that exporting ranges from 1 percent in microenterprises in the Finscope SME Survey to 16 percent in the larger firms’ Enterprise Survey); and (3) that female-led businesses are much more likely to be represented in surveys covering smaller firms, reflecting a strong relationship between gender and firm size. These observations suggest that a nationally representative survey would likely find strong gender differences in exporting as a whole, since women dominate the lower end of firms’ distribution identified in the Finscope SME survey. The analysis in this chapter should then be interpreted accordingly as a “within-group” analysis. Gender and Export: Some Descriptive Facts Although our sample comprises a subset of MSMEs with relatively high access to business tools, technology, infrastructure, and capital, there is still significant heterogeneity within it. We first focus on the overall gender differ- ences within the full sample. We summarize here the full results presented in annex A. Table 8.1: Sample Characteristics of the Impact Evaluation Survey, the Finscope SME Survey, and the World Bank Enterprise Survey in Kwazulu-Natal Province, South Africa   Impact Evaluation Survey in KZN, 2012   Finscope SME Survey in KZN, 2010 WBG Enterprise Survey in KZN, 2007 Total Male Female Total Male Female Total Male Female Gender 68.2 31.8 40.0 60.0 87.4 12.6 Exports (%) 5.5 5.5 5.5 0.8 0.8 0.8 15.8 14.4 25.0 Employees (number) 7.1 7.3 6.5 1.0 1.0 0.9 66.8 70.4 41.4 African (%) 63.7 59.9 72.0 71.2 64.2 75.9 22.1 19.8 37.5 Colored (%) 2.2 2.1 2.2 2.9 2.8 3.0 8.7 7.2 18.8 Indian or Asian (%) 19.7 23.3 11.8 19.7 25.2 16.1 36.2 37.8 25.0 White (%) 14.5 14.7 14.0 6.2 7.8 5.1 33.1 35.1 18.8 Firm age (years) 9.5 10.1 8.1 7.2 7.0 7.2 19.2 18.4 24.9 N 2,422 1,652 770   892 357 535   127 111 16 Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 139 140 Women and Trade in Africa: Realizing the Potential Female business owners are more likely to be African than male entre- preneurs. Female-owned firms have fewer employees than male-owned firms and are less likely to be registered for value-added taxes (VAT).13 They are also slightly less likely to keep financial records and have a bank account. This suggests that they may be less “formal” than their male counterparts. Significantly fewer female-owned firms have access to email or the Internet, and they are also less likely to have submitted proposals for corporate tenders or won contracts of more than three months with their clients. Men appear to have better access to business networks (referrals or online), which may be related to the bigger average firm size. There are also differences in the sectors of operations, although magnitudes are small. Male-owned businesses are more often found in construction, transport, trade, and services (relative to other sectors), while women are more represented in hotels and restau- rants. Men make major decisions alone more frequently and consult partners less often. Yet it is interesting to see that the share of businesses that export is exactly the same across gender groups, 5.5 percent. A second order of comparison focuses on exporting businesses versus nonexporting businesses. Four key differences stand out. First, the average size of exporting businesses is much larger than that of nonexporters, and many more exporters employ more than five workers. Second, exporting sec- tors include a high proportion of firms in manufacturing, while construction and hotels make up a much smaller share of exporters than their share of the full sample. Third, white-owned businesses, which represent only 14 percent of the sample, constitute 50 percent of the exporters. Although African-owned businesses form the vast majority of the sample, they represent less than a third of exporters. Finally, exporting firms are older, more formal, and rely more heavily on online marketing for business opportunities. We then compare the characteristics of female-owned and male-owned businesses when restricting only to exporters (columns 4, 5, and 6 of Table 8A.1). In this comparison, we find nonsignificant gender differences in sectors for firms that export (although this may be a reflection of a smaller sample size),14 when there were gender differences in the full sample of busi- nesses for the sector.15 The gender differences on race of exporters also equal- ize (for exporters largely dominated by white business owners irrespective of gender). Similarly, other factors associated with larger, more sophisticated businesses also equalize—VAT registration and whether the business has a bank account—when all exporting businesses are considered. The statistically significant differences across gender are the size of the business (with male-owned exporting businesses 16 percentage points more likely to have six employees or more) and networking capacity (with male- owned exporting businesses having 11 percentage points more referrals). Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 141 Noticeably, while a similar proportion of men and women in the overall sample do not have a high school certificate (15 percent), men in this group are still able to become exporters, while all female exporters in our sample have at least a high school degree. Male exporters are also more likely to take out loans from a bank while female firms re relatively more likely to borrow informally. Thus, while some factors related to exporting are gender neutral, it is clear that this is not the case across all measures. These differ- ent attributes have implications for policy, indicating that some interventions designed to promote exports are likely to affect male- and female-owned firms differently. Determinants of Export and Firm Size In this section, we analyze the main factors that constrain access to export markets (detailed analysis is presented in annex B) and growth in firms’ size (see annex C).16 The latter has been identified in the literature as being strongly correlated with exporting potential and is thus used as a proxy for propensity to access export markets. Accessing Export Markets Male- and female-owned firms combined. A set of factors associated with exporting are what we can call the firm’s core structure. These factors include the business location and its proximity to main economic centers,17 the internal practices of decision making,18 and compliance with regulations.19 The results also confirm the importance of the size of the business— measured by the number of people working in the business20—for accessing international markets. Firms with access to online customers are more likely to be exporting, which suggests that networking practices may matter for accessing foreign markets. Of course, the reverse may also be true: firms that export have to use email to stay in touch with buyers.21 Finally, firms that export are less likely to receive bank loans (although significant, the size of the coefficient is small at a 1.5 percentage point differ- ence). This could reflect several different factors: (1) businesses are choosing to enter export markets that require limited capital investment to participate; (2) businesses that export are also more likely to have high liquidity (or other financing options) and therefore have less need for bank financing; or (3) exporting firms are less successful at receiving bank loans than domes- tic market-oriented businesses even though they are in need of financing. We consider (1) and partially (2) by examining whether export businesses 142 Women and Trade in Africa: Realizing the Potential are more likely to be interested in a loan and find no significant relation- ship between the two. We assess the possibility of (2) by verifying whether businesses may not need credit (are not interested in a loan) or already have access to credit through other (informal) sources and find that neither seems to be the case. Finally, we indirectly explore (3)—the possibility that export- ing firms are less likely to be successful in their loan application—by using as the base case for the regression the businesses that applied and were unsuc- cessful or did not believe that they would be successful if they did apply. Given that exporting firms are not significantly more likely to be interested in obtaining a loan, access to capital is unlikely to be a potential constraint for exporters. In addition, the point estimate on the relationship between access to credit and exports is small in economic terms. Combining this with the fact that South Africa has a highly developed financial sector and a number of programs aimed at assisting potential exporters,22 access to credit does not seem to have a strong influence on exporting opportunities. Differences across male and female-owned firms. Across gender, we identify a consistent relationship between firm size and exporting potential; however, this result, although providing a similar coefficient size, is not statistically significant for female-owned businesses, most likely because of the smaller sample size of these businesses. Race and outreach to online customers are highly correlated with export- ing across gender. For women, regulations remain strongly associated with exporting status, while access to finance (accessing a bank loan) is negatively associated with exporting (relative to women who have been turned down for, or expect to be turned down for, a loan), which may reflect the limited importance of this constraint. Human capital seems to matter more for women. Female entrepreneurs that have not completed secondary school are very unlikely to export: none of the 125 female business owners with less than a high-school education in our sample export.23 Having more than a secondary school education does not seem as important as reaching that level. For male-owned businesses, education does not seem to be associated with exporting. Operating in more than one sector at the same time is negatively and significantly correlated with exporting for women-owned businesses, but this tendency is not associated with exporting for male-owned businesses. A busi- ness operating in multiple sectors is sometimes related to the lack of a strong business case in a given area of expertise. The business often starts small, seeking opportunistic operations in other industries to maintain an income stream.24 For women, who are traditionally seen as less likely to intend to grow their businesses,25 trying different activities can be a solution, but not one that allows the business to achieve the sophistication and growth needed Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 143 to compete in international markets. Furthermore, in South Africa, anec- dotal evidence suggests that a number of businesses apply for tenders across a range of sectors without a clear comparative advantage and are hence involved in multiple operations. In our sample, we notice that 67 percent of the female-owned enterprises with multiple businesses have applied for tenders in the past, compared to only 44 percent of those that are specializing in just one sector.26 This diversification strategy could be associated with risk aversion (as firms hedge the risks of receiving contracts) or lack of skills, in which case, mitigating interventions could include mentoring or training schemes or risk sharing with equity investors.27 The other important difference is that for women-owned firms getting customers through referrals—not seen as a particularly sophisticated way of obtaining clients—is negatively associated with access to export markets; however, female business owners are twice as likely to access customers online compared to male business owners. It is plausible that this differ- ence—although not statistically significant—reflects the fact that men are more likely to operate through traditional networks of contacts, while women need to find alternative mechanisms of reaching new markets, particularly in sectors where they have not traditionally operated. While this argument cannot be fully tested given data and design constraints, women appear to obtain customers through referrals primarily in services, in contrast to men who do so in construction and manufacturing, as well as in services. This finding suggests perhaps that women do not rely so heavily on referrals when they operate in male-dominated sectors.28 There are also differences between male and female participation in certain sectors, as becomes evident from both the descriptive statistics in annex A and the regression results in annex B. Men seem to be associated with industrial sectors, in particular in agriculture and manufacturing where exports are high. Women-owned businesses export in these sectors but are also likely to export in the general services sector (annex A shows 52 percent for female exporters versus 36 percent for female nonexporters), which is not the case for male-owned firms.29, 30 Becoming an SME Male- and female-owned firms combined. Following the analysis of the main factors affecting the propensity of firms to export, we now estimate the relationship between these critical drivers of exporting with the size of the business. The literature has identified a minimum business size associated with access to export markets. This analysis allows for a more nuanced review of exporting, identifying constraints on the development of 144 Women and Trade in Africa: Realizing the Potential the firms that are in local or international markets. For instance, business growth in sectors prone to export can reflect a firm’s preparedness to trade internationally. Physical networks and business partnerships may be more important for growth in local markets than in international ones when these networks are entrenched and one group has a monopoly, as is commonly perceived to be the case in South Africa. In this analysis, we examine the main drivers of becoming an SME. In South Africa, an SME is commonly defined as a firm with more than five workers (DTI 2003). The official definition incorporates turnover and capital measures and differs by sector, but for policy dialogue the defined number of workers is critical for targeting programs. It is also used in survey analysis both in South Africa and in other countries in the region (FinMark Trust 2010; World Bank 2007). In annex C, we estimate the probability of becoming an SME to help us understand what factors are important for growing the business out of micro- enterprise status. We conduct a similar analysis disaggregated by gender. As noted above, the sample is made up of high-end MSMEs. Given this charac- teristic, it seems to be a logical and obtainable goal for microenterprises in the sample to grow into SMEs. Forty-eight percent of the enterprises surveyed have more than five people working in the business.31 The analysis identifies some factors that, while not important in explaining the propensity to export, are positively correlated with firm size (being an SME). These include number of years in operation, being in partnership (more than one business owner), keeping financial records, partner- ships with other firms to supply products or services, and access to tenders. Other core aspects of the business structure, including whether the busi- photo: John Hogg/World Bank ness owner consults others Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 145 or the firm complies with regulations and obtains customers online, remain important explanatory variables for becoming an SME. Interestingly, there is a significant and negative relationship between white-owned enterprises (relative to African) and this measure of size. This factor could be driven by affirmative action regulation in South Africa that affects white-owned businesses once they grow to have annual turnover in excess of R 5 million (US$600,000). However, if affirmative action was an important determinant of strategic growth control by white businesses, we would expect to find that the race-size relationship could at least be partially accounted for when con- trolling for the business’s compliance with the Broad-Based Black Economic Empowerment (B-BBEE) status.32 However, the differential remains signifi- cant when controlling for having a B-BBEE status, implying that other factors are likely to be driving this relationship. Differences across male- and female-owned firms. When we compare male- and female-owned firms, a few differences stand out as explanations for larger firm sizes. For female-owned firms, the impact of race is different, with both white and Indian and Asian firms more likely to be smaller than African-owned firms. This is significantly less likely to be the case for male- owned enterprises. The effect of location is also different, with male-owned firms more likely to be smaller in Durban than elsewhere. In addition, female-owned hotels and restaurants are more likely to be larger than their male-owned counterparts. Finally, female-owned enterprises that had discussed partnerships with other firms are significantly more likely to be an SME than their male counterparts. The importance of engaging with others is also apparent when we look only within the group of women-owned firms. Female entrepreneurs who consult others in making decisions are significantly more likely to be larger than firms in which the female owners make decisions alone. Fifty-five percent of the people consulted in decision making by women-owned SMEs were other business owners33 (two-thirds of these are all women), 34 percent are the business owners’ spouses, and 11 percent are managers. Relative to female-headed microenterprises, the largest difference is in the proportion of firms consulting with managers (11 percent for female-led SMEs versus 7 per- cent for women-owned microenterprises). These findings are also in line with the targets of the consultations conducted by male-owned SMEs (57 percent other business owners, 36 percent the spouse, and 7 percent the manager). This gender disaggregated analysis of the factors associated with size is helpful in providing more robustness to the conclusions. Factors that are cor- related with both size and exports can be first areas of focus for testing poli- cies. Since, conditional on sector of operations, larger firms are more likely to export (see for example Rankin, 2013), understanding that these factors, for 146 Women and Trade in Africa: Realizing the Potential instance, consulting others in the decision-making, education level, and use of alternative mechanisms of reaching customers are associated with export- ers and larger firms is supportive of its potential significance. Discussion and Policy Recommendations This chapter discusses the relationship between the gender of the firm owner and the propensity of a firm to export for a particular set of small businesses with high potential for becoming exporters. Based on a dataset of over 2,400 small-to-medium firms operating in Kwazulu-Natal in South Africa, we observe the absence of a strong correlation between the gender of the entrepreneur and whether or not a firm exports within this group of firms. When comparing our sample to two other comprehensive surveys, we provide evidence to suggest that specific sections of the firm population have been excluded from this analysis and that women are more likely to operate microenterprises in this province of South Africa. Our study indicates that gender gaps in exporting likely do exist but not in firm types. While we do not find a difference by gender in whether a firm is export- ing, our analysis does suggest that the constraints and paths to exporting differ in male- and female-owned enterprises. The analysis of the size of the business identifies gender differences and highlights specific factors that seem relevant for determining firm size but not the propensity to export. Given the correlation between size and exports, as demonstrated in the literature, consistency of the findings for the two outcomes is important for ensuring the robustness of results. From the evidence, we identify the following potential barriers to the growth of women’s enterprises and access to foreign markets: • Limited internal and external partnership opportunities: Discussing potential partnerships with other firms is much more important for female-owned firms than for male owners. The firms of women entrepreneurs who make decisions on their own are likely to be smaller than female-owned firms that make decisions with others. Taken together, these facts suggest that the growth path for female-owned firms is strongly reliant on accessing other decisions makers. This finding may reflect a difference in female leadership style (more inclusive decision making) as much as a potential constraint (lack of partners to make decisions with). Either way, these internal and external networks seem important for the growth of female-owned firms. Women also need to seek alternative ways of reaching customers and markets. Constrained by limited access to networks of contacts, perhaps for historical or social reasons, women that are more successful seek alternatives like online Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 147 marketing, rather than relying on traditional mechanisms of building business relationships. • Lack of basic skills: We find that high school education is strongly associated with female-owned firms’ propensity to grow and export, although this constraint does not seem to be as binding for male-owned firms. Providing access to education can help create a path for women to grow their businesses and increase their exporting potential. • Risk aversion: Our analysis suggests that women diversify into multiple, small-scale enterprises despite not being particularly constrained in accessing finance, at least relative to men.34 Business diversification may support firm growth locally but appears to inhibit export potential. In South Africa, working across multiple sectors is commonly seen as a business strategy used to keep tender opportunities open, reflecting an opportunistic approach to contracting. This may be a rational choice by businesses that lack networks or expertise in a particular field but can also signal a lack of confidence or commitment to a specialized strategy. For male-owned businesses, where network constraints seem to be less important, operating in multiple sectors does not affect their success in exporting. For women, however, such diversification becomes a negative factor. It is not clear whether this is a quality issue (diversified businesses are not able to produce at a high enough quality to compete internationally) or a signaling issue (other businesses perceive diversification as a proxy for poor quality when information is asymmetric). As such, policies addressing the underlying risk aversion leading to multiple sector operations, as well as those addressing skills and information asymmetries, may help unpack the latent drivers of exporting potential for female-owned businesses. In South Africa, the framework provided in the National Export Development Programme identifies the importance of skills and networks in the government’s plan to promote export growth in South Africa. Beyond skills development, the plan proposes to create “export villages” that can foster photo: John Hogg / World Bank business partnerships and 148 Women and Trade in Africa: Realizing the Potential market links through collective action. These are envisaged as including both physical and virtual villages, harnessing the power of information technology to reduce information barriers associated with collective action opportunities. However, little is known about the importance of the constraints that these interventions implicitly plan to overcome, how these ideas may differentially affect female- and male-owned businesses, and what targeted interventions may further expand export access. The key findings from this chapter may help fill in this gap. The following policy options provide potential avenues for addressing the identified constraints to growth and exporting: skills, business networks, and risk aversion, leading to excessive business diversification. Given the nonexperimental nature of our analysis and the particular sample of firms, relationships identified here should be interpreted as potential opportunities for future, more rigorous, experimental research to measure the effectiveness of the proposed approaches in practice: • Explore the potential of angel and venture capital investments: More important than overcoming the debt risks associated with growing a start-up company by bringing in business partners with a stake in the business, equity investments may help (1) reduce the risk of specializing for female-led businesses by spreading risk across partners rather than across sectors; (2) inject new skills and offer more opportunity for joint decision making; and (3) increase potential networking possibilities through partnerships with more experienced and established entrepreneurs. Based on the identified constraints, this approach can help promote both firm growth and exporting opportunities for a specific group of female-owned businesses with the strongest potential. From a policy perspective, this approach may require developing the equity market thus, creating opportunities for a growing number of women to access this critical support. • Use technology to circumvent entrenched networks: The use of online and ICT-related marketing more generally to reach out to new client bases is an opportunity for promoting business growth and exports. The Internet, along with cell phones, offers a medium through which businesses can circumvent traditional networks and break into new markets. Technology offers businesses a platform for reaching out to both domestic and international markets at relatively low cost and for overcoming potential information asymmetries. Building skills focused on leveraging these instruments and other managerial tools seems important avenues to explore as well. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 149 Testing these options through pilot interventions will help shed light on the most important barriers women-owned firms face and contribute to find- ing sustainable solutions to promote the development of small businesses in South Africa. Annex A: Average Characteristics of Entrepreneurs and Businesses in Survey Sample Table 8A.1 provides descriptive statistics of the main covariates of interest, broken down by business gender and export status. The first three columns provide the mean values for men, the mean values for women, and the dif- ference between the two, with t-tests used to assess significance for the entire sample of 2,422 businesses. The next three columns restrict this comparison to only businesses that export (132). The last two columns present the mean differences between exporting and nonexporting businesses for men and women, respectively, with t-tests detecting significance levels. The covariates presented here are then used in annexes B and C for the regression analysis presented in the chapter. The dependent variables of interest in the chapter are whether the business is engaged in exporting and whether the business has more than five employees (and could therefore be defined as a small or a medium rather than as a microenterprise). The covariates of interest are broken into three broad groups: (1) owner characteristics; (2) firm charac- teristics; and (3) sector of operation. The following is a summary of variable descriptions in the cases in which they are not necessarily self-explanatory: • Owner race is defined, per South African census definitions, as being either Black/African, Colored, Indian/Asian, or White, representing the main racial groups in the country. • Owner takes decisions: the owner is asked whether he or she usually makes business decisions alone or jointly (consults others in decisions). • Matric refers to high school graduation in South Africa, with postmatric referring to any education beyond high school • Firm age is the number of years the business has been in operation. • Online customers refers to whether the business finds its customers through the Internet. Similarly, referral customers come from business referrals, and tender customers indicates that the firm applies for, and gets, clients through tender opportunities. • Sector refers to the main sector in which the business operates. When businesses operate in multiple sectors, more than one sector = 1, and the main sector is recorded for this analysis. 150 Table 8A.1: Summary Statistics on Entrepreneurs’ and Business Average Characteristics Nonexporters vs.   Full samplea   Exportersb exporters Male Female Difference Male Female Difference Male Female Sample (%) 68.2 31.8 68.2 31.8 0.0 0.0 Exports 5.5 5.5 0.0 More than 5 workers 50.1 43.8 6.4*** 73.3 57.1 16.2* -24.6* -14.2*** Number of workers 7.3 6.5 0.8*** 10.5 8.8 1.7 -3.4*** -2.4*** Owner characteristics African 59.9 72.0 -12.1*** 24.4 33.3 -8.9 37.5*** 40.9*** Coloured 2.1 2.2 -0.1 4.4 0.0 4.4 -2.5 2.3 Indian or Asian 23.3 11.8 11.5*** 22.2 14.3 7.9 1.2 -2.6 White 14.7 14.0 0.7 48.9 52.4 -3.5 -36.2*** -40.6*** Women and Trade in Africa: Realizing the Potential Owner age 42.5 42.4 0.1 47.0 45.1 1.9 -4.8*** -2.9*** Owner takes decisions 75.5 71.0 4.5** 64.4 54.8 9.7 11.7*** 17.2** Consults others in decisions 19.6 24.0 -4.4** 25.6 35.7 -10.2 -6.3 -12.4*** Less than matricc 15.4 15.7 -0.3 11.1 0.0 11.1** 4.6 16.6*** c Matric 35.2 33.3 2.0 30.0 40.5 -10.5 5.5 -7.7 Postmatricd 49.3 51.0 -1.7 58.9 59.5 -0.6 -10.1*** -9.0 Firm characteristics Firm age 10.1 8.1 2.1*** 15.2 12.0 3.3 -5.4*** -4.1*** Registered for VAT 59.3 46.1 13.2*** 86.7 81.0 5.7 -29.0*** -36.9*** Has bank account 94.0 91.9 2.1*** 97.8 97.6 0.2 -4.0 -6.0 Has received loan from bank 22.2 21.0 1.2 28.9 14.3 14.6*** -7.1 7.1 Borrows informally only 20.9 26.9 -5.9*** 6.7 23.8 -17.1*** 15.1*** 3.3 Online customers 20.0 17.1 2.9*** 53.3 57.1 -3.8 -35.2*** -42.3*** Referral customers 93.3 90.1 3.2*** 96.7 85.7 11.0** -3.6 4.7 Tender customers 43.2 45.2 -2.0 53.3 42.9 10.5 -10.7** 2.5 Keeps financial records 89.9 87.7 2.2*** 97.8 97.6 0.2 -8.3*** -10.5** Sector Agriculture 3.6 3.5 0.1 5.6 4.8 0.8 -2.1 -1.3 Construction 22.6 16.4 6.2*** 6.7 7.1 -0.5 16.8*** 9.8*** Manufacturing 11.6 10.8 0.8 20.0 21.4 -1.4 -8.9*** -11.3** Transport and public utilities 6.2 4.0 2.2** 8.9 4.8 4.1 -2.8 -0.8 Trade 10.1 7.9 2.2*** 13.3 7.1 6.2 -3.4 0.8 Hotels and restaurants 3.9 20.7 -16.8*** 0.0 2.4 -2.4 4.1** 19.3*** Services 41.7 36.8 4.9** 44.4 52.4 -7.9 -3.0 -16.5*** Other 0.4 0.0 0.4 1.1 0.0 1.1 -0.8 0.0 More than one sector 28.5 39.4 -10.8***   18.9 9.5 9.4   10.2** 31.6 Note: *** p<0.01, ** p<0.05, * p<0.1. a. N = 2,422. b. N = 132. c. Matric refers to high school graduation in South Africa. d. Postmatric refers to any education beyond high school. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 151 152 Women and Trade in Africa: Realizing the Potential Annex B: Explaining the Propensity of Firms to Export We present the results from two sets of probit regressions in which marginal effects are reported.35 First, we combine men and women and look at the relationship between the gender of the firm owner and propensity to export and firm size. In the second regression (the columns that fall under (2)), we interact all the explanatory factors with gender to see if there are differ- ences between male- and female-owned firms for a given factor’s explanatory power. The final column tells us if the differences between the coefficients for male- and female-owned firms are statistically significant. Table 8B.1, column (1), presents the results on what is associated with whether or not a firm exports for male- and female-owned firms combined. The gender dummy coefficient is nonsignificant and close to zero. Given the large sample size, this is a strong result. The racial divide—and associ- ated with it the different average quality of education, depth of networks of contacts, and family background—seems to be highly correlated with exporting. Measuring the unconditional differences between men and women and introducing a gender dummy in the propensity-to-export regression are very crude ways of assessing the existence of a gender gap in exporting. There may be strong gender differences in the pathways to exporting that affect women’s likelihood to do so successfully. Analyzing this requires estimating coefficients separately for male- and female-owned businesses, which we do in the second specification in table 8B.1 by running a pooled regression with each of the explanatory variables interacted with gender dummies and run- ning coefficient equality tests. Table 8B.1: Predicting Whether a Firm Exports Gender-disaggregated results Total sample (propensity to export) (2) (1)  Female Male Test of male and female Marginal Marginal Standard Marginal Standard coefficientsa effect Standard error effect error Effect error Female owner 0.008 0.007 Colored 0.021 0.028 0.045 0.041 . Indian or Asian -0.002 0.008 0.003 0.02 -0.006 0.008 0.63 White 0.049*** 0.017 0.054** 0.039 0.046** 0.023 0.90 Owner characteristics Consults others in decisions 0.018** 0.01 0.034* 0.027 0.012 0.012 0.43 Others take decisions 0.029* 0.022 0.054 0.058 0.029 0.026 0.67 Owner age 0.000 0.002 -0.003 0.003 0.001 0.002 0.36 Owner age squared (/100) 0.001 0.002 0.004 0.003 0.000 0.002 0.29 Post matric 0.003 0.006 0.01 0.013 -0.001 0.007 0.44 Firm characteristics Firm age 0.000 0.000 0.000 0.001 0.000 0.000 0.92 More than one owner -0.003 0.006 -0.003 0.011 -0.003 0.007 0.99 Keeps financial records 0.015 0.008 0.041 0.048 0.015 0.016 0.66 Durban 0.013** 0.006 -0.001 0.011 0.022** 0.010 0.12 (Table 8B.1: continued on the next page) Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 153 154 (Table 8B.1: continued from the previous page) Gender-disaggregated results Total sample (propensity to export) (2) (1)  Female Male Test of male and female Marginal Marginal Standard Marginal Standard coefficientsa effect Standard error effect error Effect error Registered for VAT 0.017** 0.007 0.037** 0.024 0.013 0.009 0.37 Has access to Internet 0.002 0.008 -0.009 0.011 0.008 0.009 0.26 Number of workers 0.001** 0.001 0.001 0.001 0.001** 0.001 0.89 Sector Agriculture 0.022 0.022 -0.007 0.018 0.040* 0.033 0.24 Construction -0.013 0.007 -0.008 0.013 -0.014 0.007 0.72 Manufacturing 0.023** 0.012 0.023 0.025 0.019* 0.014 0.89 Transport and public utilities 0.014 0.016 0.007 0.029 0.020 0.020 0.73 Women and Trade in Africa: Realizing the Potential Trade 0.005 0.010 -0.001 0.018 0.013 0.014 0.59 Hotels & restaurants -0.027** 0.006 -0.020 0.007 . Other 0.052 0.122 0.053 0.101 . More than one sector -0.010 0.007 -0.022*** 0.006 0.000 0.008 0.03** Market access Online customers 0.037*** 0.011 0.064*** 0.034 0.025*** 0.012 0.24 Referral customers -0.013 0.014 -0.029** 0.010 0.013 0.015 0.02** Tender customers 0.007 0.006 -0.003 0.010 0.014** 0.008 0.21 Gender-disaggregated results Total sample (propensity to export) (2) (1)  Female Male Test of male and female Marginal Marginal Standard Marginal Standard coefficientsa effect Standard error effect error Effect error Borrows informally only -0.012 0.007 0.003 0.017 -0.020** 0.006 0.09*** Not interested in loan -0.011 0.007 -0.011 0.010 -0.011 0.007 0.96 Number 2,422 2,341 R squared 0.218       0.243     Note: This table shows a Probit estimation on access to export markets, with robust standard errors. All specifications control for source of sampling frame used. For race, the omitted category is African. For decision making, the omitted category is takes decision by him/herself. For sector, the omitted category is services, while for access to finance the omitted category is owners who were denied credit or expected they would be rejected. Equality prob>2 Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 155 156 Annex C: Predicting Whether a Firm Has More than Five Workers Table 8C.1: Predicting Whether a Firm Has More than Five Workers Total sample (propensity to Gender-disaggregated results export) (2) (1)  Female Male Test of male and female Marginal Standard Marginal Standard Marginal Standard coefficientsa effect error effect error Effect error Female owner -0.034 0.025 Colored 0.066 0.068 0.151 0.141 0.019 0.094 0.45 Indian or Asian -0.063* 0.034 -0.172** 0.066 -0.041 0.038 0.09*** White -0.117*** 0.04 -0.269*** 0.057 -0.054 0.048 0.01* Owner characteristics Consults others in decisions 0.084** 0.031 0.145*** 0.052 0.054 0.039 0.17 Women and Trade in Africa: Realizing the Potential Others take decisions 0.037 0.059 0.097 0.106 0.021 0.072 0.56 Owner age -0.001 0.007 0.006 0.013 -0.004 0.008 0.51 Owner age squared (/100) 0.001 0.008 -0.007 0.014 0.004 0.009 0.54 Postmatric -0.017 0.023 -0.051 0.042 -0.003 0.028 0.34 Firm characteristics Firm age 0.007*** 0.002 0.011** 0.004 0.005*** 0.002 0.22 More than one owner 0.093*** 0.027 0.138*** 0.047 0.071** 0.033 0.25 Keeps financial records 0.089** 0.036 0.046 0.062 0.126*** 0.045 0.30 Durban -0.033 0.025 0.027 0.043 -0.057** 0.029 0.09*** Total sample (propensity to Gender-disaggregated results export) (2) (1)  Female Male Test of male and female Marginal Standard Marginal Standard Marginal Standard coefficientsa effect error effect error Effect error Registered for VAT 0.155*** 0.024 0.177*** 0.042 0.148*** 0.029 0.55 Has access to Internet 0.070*** 0.026 0.052 0.047 0.079** 0.032 0.63 Sector Agriculture 0.204*** 0.056 0.275** 0.093 0.176** 0.068 0.41 Construction 0.087*** 0.031 0.147** 0.059 0.080** 0.036 0.33 Manufacturing 0.056 0.036 0.111 0.067 0.044 0.043 0.41 Transport and public utilities 0.042 0.048 0.028 0.105 0.053 0.056 0.83 Trade 0 0.039 0.06 0.078 -0.017 0.046 0.39 Hotels and restaurants 0.029 0.043 0.112* 0.057 -0.095 0.069 0.02** Other -0.097 0.211 -0.114 0.213 . More than one sector 0.003 0.025 -0.031 0.043 0.016 0.03 0.37 Market access Online customers 0.147*** 0.03 0.175*** 0.056 0.150*** 0.035 0.70 Referral customers 0.02 0.04 0.009 0.068 0.028 0.052 0.82 Tender customers 0.064*** 0.022 0.063 0.041 0.061** 0.027 0.97 (Table 8C.1: continued on the next page) Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 157 158 (Table 8C.1: continued from the previous page) Total sample (propensity to Gender-disaggregated results export) (2) (1)  Female Male Test of male and female Marginal Standard Marginal Standard Marginal Standard coefficientsa effect error effect error Effect error Borrows informally only -0.003 0.034 0 0.061 -0.008 0.042 0.91 Not interested in loan 0.002 0.031 0.039 0.059 -0.011 0.037 0.46 Number 2,422 2,341 R squared 0.218         0.243     Note: This table shows a Probit estimation on being a small or medium enterprise (more than five employees), with robust standard errors. All specifications control for source of sampling, sample frame used. For race, the omitted category is African. For decision making, the omitted category is takes decision by him/herself. For sector, the omitted category is services; for access to finance, the omitted category is owners who were denied credit or expected they would be rejected. equality prob>2. Women and Trade in Africa: Realizing the Potential Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 159 Notes 1. We find a positive and significant relationship between exporting and firm size, in line with current evidence that, accounting for selection issues, firm size is an important determinant of exporting potential (for example, Rankin, Söderbom, and Teal 2006). Our coefficient estimates on the relationship between firm size and exporting are in line with those of the South African Enterprise Survey. 2. Conditional on the owners making decisions on their own (controlling for number of owners and other potential confounding factors), women-owned firms are smaller. In addition, female-owned firms that discuss business partnerships with other businesses are larger than other female-owned enter- prises. A similar relationship is not found for male-owned firms. 3. See Klapper and Parker (2011) for a review of the gender and entrepre- neurship literature and World Bank (2011) for an overview of the constraints women face in accessing economic opportunities. 4. See, for instance Sabarwal, Terrell, and Bardasi (2009). Bruhn (2009) uses a combination of enterprise surveys and household surveys, acknowledging the latter are limited in information about larger firms. 5. See, for example, Mckenzie and Woodruff (2012) for review of business training related impact evaluations. 6. The KZN Province is the second-largest province in South Africa with over 10 million people. Its capital is Pietermaritzburg, and the major city is Durban, a major port city. 7. This also allows us to use personal characteristics of the entrepreneur (education, age, and the like) in the analysis. 8. The Gini Index was 63.1 as of 2009 (http://data.worldbank.org/indicator/ SI.POV.GINI). 9. These include Small Enterprise Development Agency (SEDA), Black Business Supplier Development Programme (BBDSP), South African Woman Entrepreneurs Network (SAWEN), and The Critical Infrastructure Programme (CIP) (http://www.dti.gov.za/). 10. 2,788 interviews were completed, but nonresponses to particular ques- tions led to using 2,422 interviews for this analysis. Marketing research com- panies in South Africa indicate hit rates of 10–20 percent as the norm. 11. “Female ownership” in the Enterprise Survey means that women are included among the owners, which is a broader definition than that used for our analysis. 160 Women and Trade in Africa: Realizing the Potential 12. The KZN Quarterly Labour Force Survey, which is representative of the labor force, shows that only 12 percent of the total number of women-owned firms in the labor survey has more than four employees. That compares with 56 percent in our sample. 13. Over 95 percent of these enterprises are registered at CIPRO, the National Department of Trade and Industry’s business registration database. Firms with annual turnover greater than R 1 million need to register for VAT, which can help explain the difference between the proportion of VAT-registered male- and female-owned firms. 14. The sample of businesses that export totals 132, while the full sample including nonexporters and exporters, totals 2,422 businesses. 15. Combining the two, there are also sector differences between (wo)men- owned exporters and (wo)men-owned nonexporters. For instance, 52 percent of women exporters are in services when that is the case for less than 36 percent of the nonexporters. In the opposite direction, only 2 percent of the women exporters are in hotels and restaurants, while almost 22 percent of the women nonexporters are in that secto. 16. While we also considered the possibility of conducting an analysis of the gender gaps in productivity as a means of understanding the potential dif- ferences in accessing export markets, the quality of the financial data in our sample—the high nonresponse rate—does not allow for rigorous estimations. 17. Our dataset is not detailed enough to identify which markets exporters are serving. This would help understand if businesses located in Durban are more likely to export to more advanced markets, given the port located in the city. This would also help in identifying the quality and sustainability of exporting. 18. In our sample, this is measured by the likelihood of the decision making being concentrated solely or not on the business owner. Seventy-five percent of the nonexporters make the decisions by themselves, while that is the case for 62 percent of the exporters. 19. Over 95 percent of firms are registered; thus we are measuring here a degree of compliance on top of being registered. It may also reflect the size of the business because only firms with turnover greater than R1 million need to register for VAT. 20. We control for being in partnership in the regression. 21. We attempt to control for a firm’s level of sophistication and technology adoption in the regression by including an indicator of whether the firm has access to the Internet. We have other proxies for IT access, including access to email and computers that generate the same results. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 161 22. Including the Department of Trade and Industry’s Exporting Marketing and Investment Assistance programs, the Export Credit Insurance Corporation, and the Credit Guarantee Insurance Corporation (SouthAfrica. info 2013). There are also a large number of programs specifically aimed at assisting female entrepreneurs; see, for example, Deloitte Special Investment Team (2012) for a comprehensive list of the funding options available to female entrepreneurs in Kwazulu-Natal or the Department of Trade and Industry (2010) and the Small Enterprise Development Agency (2013) for a list of national programs. Furthermore, the Financial Sector Charter (National Treasury 2003) specifically requires that the financial sector “commits itself to fostering new, and developing existing BEE [Broad-Based Black Economic Empowerment] accredited companies.” 23. Given that not completing high school (matric) perfectly predicts failure to export for women and the observations in that status are dropped in the regression, the reported results in the table look only at the dummy on com- pleting a degree postmatric to examine the additional difference of this status, keeping the observations in the regression. 24. In our sample, for instance, there are women entrepreneurs operating at the same time in construction and catering, or laundry services and events organization. 25. Banerjee (2010) hypothesizes that firm-level activity may be a way of “buying” a job (especially when the household is not dependent solely on this income). 26. A similar difference is identified for male-owned enterprises, but the pro- portion of those with multiple sectors of operations is smaller for men. 27. The Southern African Venture Capital and Private Equity Association (2009) shows that private equity and venture capital (PE/VC) investments are associated with increased exporting but also that the reach of such invest- ments is limited when it comes to African women. Only 5 percent of the businesses that have access to PE/VC investments had some African female ownership. Lamprecht and Swart (2010) point out that the venture capital market in South Africa is geared toward “high tech, high growth potential businesses,” and that while there are a number of (predominantly publicly funded) instruments offering finance and equity that are focused on SME sector development, these are not regarded as VC-type investments. And while most of the publicly funded financing options are debt instruments (for example, the products offered by the Small Enterprise Funding Agency), the National Empowerment Fund (2012) appears to be the only government- supported program that provides equity products to assist emerging African- owned businesses at different stages of their development. 162 Women and Trade in Africa: Realizing the Potential 28. For services, 92 percent of women-owned firms get customers from refer- rals. For manufacturing and construction (traditionally male owned), that figure falls to 83 percent. For male-owned businesses, the difference between the two types of sectors is from 95 percent in services to 92 percent in con- struction and manufacturing. 29. Annex B regressions have services as the base category. For that reason, this result is not reflected in the table. 30. The industries in which there are few (if any) exporters include hotel and restaurants and construction. 31. Although the references here may be divergent, we take the approach of including in the headcount business owners working in their own business. A firm with three business owners that work in the business and no employ- ees has a labor force more comparable—albeit with different skills—to a firm that has one business owner and two employees, than to a firm with one business owner and no employees. A similar situation happens with unpaid employees or apprentices and family members working in the business. 32. This is the affirmative action policy enacted in South Africa with the aim of distributing opportunities more equitably. The B-BBEEE status gives an indication of a business’s compliance with the regulation. The status provides a picture of the participation of previously disadvantaged groups in a firm’s management, ownership, employee structure, preferential procurement, and the like. 33. There is a possibility, though, that this result reflects the fact that busi- nesses with more than one owner are more likely to have people from the opposite sex, which may increase the likelihood of classifying a business as female owned when men also participate (and may have actually been the original founders). However, if we define women owned as only the firms for which all owners are women, the results are still robust, and as discussed, even in the general case, over two-thirds of business owners consulted by women are other female business owners. Unfortunately, our dataset does not include an indicator on who started the business, which would be another way of looking at this possibility. 34. Small firms with lines of credit are relatively high (22 percent%) com- pared to other emerging economies such as Mexico (11 percent) or Nigeria (3 percent), for instance (World Bank 2013). 35. The results are robust to a linear probability model. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 163 References Anjinho, N., and N. Rankin. 2008. “Exporting and Size: Comparisons between Developing Countries.” University of the Witwatersrand, mimeo. Banerjee, A. 2010. “Making Sense of Microfinance.” Presentation at the Innovations for Poverty Action conference “Microfinance Impact and Innovation,” New York, Oct. 21–23. Baughn, C. C., B. L. Chua, and K. E. Neupert. 2006. “The Normative Context for Women’s Participation in Entrepreneurship: A Multicountry Study.” Entrepreneurship Theory and Practice 30 (5): 687–708. Beneria, L. 2001. “Changing Employment Patterns and the Informalization of Jobs: General Trends and Gender Dimensions,” International Labour Office, Geneva. Berman, N., and J. Héricourt. 2010. “Financial Factors and the Margins of Trade: Evidence from Cross-Country Firm-Level Data.” Journal of Development Economics 93 (2): 206–17. Berniell, M., and C. Sánchez-Páramo. 2011. “Overview of Time Use Data Used for the Analysis of Gender Differences in Time Use Patterns.” Background paper for the World Development Report 2012, World Bank, Washington, DC. Berry, A., M. von Blottnitz, R. Cassim, A. Kesper, B. Rajaratnam, and D. E. van Seventer. 2002. “The Economics of SMMES in South Africa.” Trade & Industrial Policy Strategies., unpublished paper. Bruhn, M. 2009. “Female-Owned Firms in Latin America.” Research Working Tapers 1 (1): 1–28. De Bruin, A., C. G. Brush, and F. Welter. 2007. “Advancing a Framework for Coherent Research on Women’s Entrepreneurship.” Entrepreneurship Theory and Practice 31 (3): 323–39. Deloitte Specialist Investment Team. 2012. “Women Entrepreneurs in Business: How to….” Presentation presented at the KZN Women’s Economic Empowerment Conference, Durban, August 16–17. DTI (Department of Trade and Industry). 1995. “National Strategy for the Development and Promotion of Small Businesses in South Africa,” DTI, Pretoria. ———. 2003. “National Small Business Amendment Act,” http://www.info. gov.za/view/DownloadFileAction?id=68002. ———. 2005. “South African Women Entrepreneurs: A Burgeoning Force in Our Economy: A Special Report,” DTI, Pretoria. ———. 2010. “National Directory of Small Business Support Programmes,” DTI, Pretoria. ———. 2013. “National Export Development Programme.” DTI,, Pretoria. 164 Women and Trade in Africa: Realizing the Potential Dupas, P. and Robinson, J. 2013. “Savings Constraints and Microenterprise Development: Evidence from a Field Experiment in Kenya.” AEJ: Applied Economics, 5 (1), pp.163-92, January 2013. Edwards, L., N. A. Rankin, and V. Schöer. 2008. “South African Exporting Firms: What Do We Know and What Should We Know?” Ideas, October, http://ideas.repec.org/p/pra/mprapa/16906.html. Fafchamps, M. McKenzie, D., Quinn., and Woodruff, C. “Microenterprise Growth and the Flypaper Effect: Evidence from a Random Experiment in Ghana.” Journal of Development Economics, forthcoming. Financial Services Regulation Task Group. 2007. “SMEs’ Access to Finance in South Africa: A Supply-Side Regulatory Review.” Internal Task Group Report. FinMark Trust. 2010. “Finscope Small Business Survey 2010,” Johannesburg, South Africa. Fletschner, D., and M. R. Carter. 2008. “Constructing and Reconstructing Gender: Reference Group Effects and Women’s Demand for Entrepreneurial Capital.” Journal of Socio-Economics 37 (2): 672–93. Freund, C., and N. Rocha. 2011. “What Constrains Africa’s Exports?” World Bank Economic Review 25 (3): 361–86. Gumede, V. 2004. “Export Propensities and Intensities of Small and Medium Manufacturing Enterprises in South Africa.” Small Business Economics 22 (5): 379–89. Hallward-Driemeier, M. 2011. “Enterprising Women: Expanding Opportunities for Women in Sub-Saharan Africa.” Washington DC,: World Bank. International Finance Corporation (IFC). 2006. Access to Finance for Women Entrepreneurs in South Africa. Washington, DC: IFC. Klapper, L. F., and S. C. Parker. 2011. “Gender and the Business Environment for New Firm Creation.” World Bank Research Observer 26 (2): 237–57. Kneller, R., and Pisu, M. 2007. “Gender and the Business Environment for New Firm Creation.” World Bank Research Observer 26 (2); 237-57. Lamprecht, S., and E. Swart. 2010. “2010 SAVCA Venture Solutions VC Survey,” Southern African Venture Capital Association (SAVCA), Illovo, South Africa. Macours, K., and R. Vakis. 2009. “Changing Households’ Investments and Aspirations through Social Interactions: Evidence from a Randomized Transfer Program.” Policy Research working paper ; WPS 5137, World Bank, Washington, DC. McKenzie, D., and C. Woodruff. 2012. “What Are We Learning from Business Training and Entrepreneurship Evaluations around the Developing World?” Policy Research Working Paper 6202, World Bank, Washington, DC. Shape Up and Ship Out? Gender Constraints to Growth and Exporting in South Africa 165 National Empowerment Fund. 2012. Annual Report. National Planning Commission. 2012. “National Development Plan 2030: Our Future: Make It Work,” Presidency, Republic of South Africa, Pretoria. National Treasury. 2003. “Financial Sector Charter,” National Treasury of the Republic of South Africa, Pretoria. Rankin, N. 2013. “Exporting and Export Dynamics among South African Firms,” Occasional Paper 149, Economic Diplomacy Programme, South African Institute of International Affairs, Johannesburg. Rankin, N., M. Söderbom, and F. Teal. 2006. “Exporting from Manufacturing Firms in Sub-Saharan Africa.” Journal of African Economies 15 (4): 671–87. Sabarwal, S., K. Terrell, and E. Bardasi. 2009. “How Do Female Entrepreneurs Perform? Evidence from Three Developing Regions.” Paper presented at the World Bank conference “Female Entrepreneurship: Constraints and Opportunities,” World Bank, Washington, DC, June 2–3. ¸ eker, M. 2011. “Trade Policies, Investment Climate, and Exports across S Countries.” Policy Research Working Paper 5654, World Bank, Washington, DC. ———. 2012. “Rigidities in Employment Protection and Exporting.” World Development 40 (2): 238–50. Small Enterprise Development Agency (SEDA). 2013. “Women-Owned Enterprise Development Information,” SEDA, Pretoria. SouthAfrica.Info. 2013. “Incentives for Exporters,” www.southafrica.info. Southern African Venture Capital and Private Equity Association (SAVCA). 2009. “The Economic Impact of Venture Capital and Private Equity in South Africa,” SAVCA, Illovo. Wagner, J. 2007. “Exports and Productivity: A Survey of the Evidence from Firm-Level Data.” World Economy 30 (1): 60–82. World Bank. 2007. Enterprise Survey: South Africa (database), World Bank, Washington, DC, http://www.enterprisesurveys.org/Data/ ExploreEconomies/2007/south-africa. ———. 2011. World Development Report 2012: Gender and Development. Washington, DC: World Bank. ———. 2013. The Little Data Book on Financial Development. Washington, DC: World Bank. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 167 9  Trade and Gender in Tanzania: What Matters—Participation or Outcomes? Josaphat Kweka and Mahjabeen Haji Introduction This chapter looks at the impact of gender inequality on economic growth as reflected in trade performance in Tanzania. It provides a statistical analysis of the participation of, and outcomes for, women in export activities. Unlike most of the previous studies, which have focused analysis on small firms to explain the lower participation of women in exporting, this study analyzes large firms. We look at the participation of women in export activities in an entrepreneurial context, rather than as wage workers, and assess how female ownership is associated with the determinants of firms’ export performance. The analysis is based primarily on the Annual Survey of Industrial Production (ASIP 2008) in Tanzania, which contains firm-level data disaggregated by gender on 729 industrial establishments, 110 of which are exporters. To provide some context, the chapter first examines the key features and trends in trade policy performance in Tanzania. It then focuses on gender in trade policy by reviewing female participation in trade activities in Tanzania. We find that while trade policy documents include gender-differentiated regulations, in practice, there is little evidence of any programs that lever- age regulations to address gender inequalities related to trade activities. In addition, the literature review finds that despite a plethora of opportunities for women’s increased participation in trade activities, barriers to female participation persist in labor-intensive export-oriented activities, such as the cut flowers and textiles trade. The chapter then reports the results of a statistical analysis of the survey data and finds that, controlling for observable characteristics, the export values of firms owned by female entrepreneurs are comparable to those of their male-owned counterparts. The analysis shows that for the full sample of firms, there is little difference in profitability between male- and Josaphat Kweka is a senior economist (AFTP5) at the World Bank. Mahjabeen Haji is a consultant (AFTP5) at the World Bank. We would like to acknowledge helpful comments and guidance from Paul Brenton, Elisa Gamberoni, and Catherine Sear. 168 Women and Trade in Africa: Realizing the Potential female-owned firms. However, for the subsample of firms that export, it is apparent that female-owned firms tend to be associated with less profitabil- ity than male-owned firms. Finally, looking at the destination markets for exports, we find that female-led businesses prefer to export to neighboring countries, while their male peers export farther abroad. This finding might suggest that female-owned firms face more constraints in trading beyond the Tanzanian border than male-owned firms. It also shows the importance of regional trade in Africa not only for informal cross-border traders but also for formal female-owned export firms. Trade Policy and Gender in Tanzania While the trade policy regime in Tanzania has evolved significantly toward support for market reforms and integration in the global economy, it has been less effective in achieving the high levels of competitiveness and diversifica- tion that would ensure broad-based growth and transformation.1 The policy reinforces the observation that liberalization alone does not automatically produce export dynamism or rapid growth of income. Trade policy reform can help boost the productivity of firms and labor, but complementary policies focused on macroeconomic stability, favorable investment climate, infra- structure investments, and investment in human capital are essential preconditions for trade to lead directly to pov- erty reduction and economic growth. With the rapid rise in exports over the past decade, Tanzania has grown substan- tially more open. The ratio of trade to gross domestic prod- uct (GDP) has risen from 13.5 percent to over 30 percent dur- ing the past decade, the high- est in the region (see figure 9.1a).2 Although imports grew faster than exports (hence the growing deficit in the cur- rent account balance), the significant increase in trade photo: Arne Hoel / World Bank openness is the result of high Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 169 Figure 9.1 Exports of Goods and Services in Countries of the East African Community, 2000–2011 9.1a: As % of GDP 9.1b: In current dollar value 35 700 30 600 25 500 U.S. dollars Percent 20 400 15 300 10 200 5 100 0 0 02 06 10 02 06 10 00 04 08 00 04 08 20 20 20 20 20 20 20 20 20 20 20 20 Year Year Burundi Kenya Tanzania Uganda Source: World Development Indicators. Note: Current dollar value, 2000 = 100. growth rates for both imports and exports, with dollar exports having multi- plied more than fivefold over the same period (see figure 9.1b). Furthermore, the trade structure has been shifting toward South-South trade, comparable to what happened in Asian economies in the 1980s and 1990s. Nonetheless, Tanzania’s export prospect also benefits from deepening regional integration in east and southern Africa, partly due to its geographical advantage as a coastal economy. Nevertheless, the country has been able to diversify to some extent into manufactures (figure 9.2) and discover new destinations, moving away from traditional European countries to Asian and regional markets: some 55–60 percent of all manufacturing exports in 2009 and 2010 were sold to the East African Community and the Southern African Development Community trad- ing blocs (figure 9.3). The current challenge emerging for Tanzania is how to sustain this momentum, given the persistence of supply-side constraints (infrastructure, human capital, and low productivity), and how to ensure that trade trans- lates into shared growth and gains in poverty reduction (Kweka and Booth 2005). One way for exports to increase is to reach out to more destinations. In an empirical study, Newfarmer and Brenton (2007) have shown that in contrast to successful exporting countries, most African countries concentrate their existing products in a few destinations. In that context, Tanzania needs to take advantage of its proximity to several national borders by lowering 170 Women and Trade in Africa: Realizing the Potential Figure 9.2 Exports of Tanzania, 2006–09 US$ millions 40 30 U.S. dollars 20 10 0 Manufactures exports Agricultural raw materials exports Ores and metals exports (% of merchandise exports) (% of merchandise exports) (% of merchandise exports) 2006 2007 2008 2009 2010 2011 Source: IMF Direction of Trade Figure 9.3 Historical Trend of World Exports by Destination, 1990–2009 US$ millions 700 2,500 World Exports - U.S. dollars 600 2000 Regions - U.S. dollars 500 1500 400 300 1000 200 500 100 0 0 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 20 20 20 19 19 19 20 20 20 20 20 20 20 20 20 Year World ASIA EU N. AMERICA SSA EAC SADC Source: UN COMTRADE. Note: EU = European Union; N. America = North America; SSA = Sub-Saharan Africa; EAC = East African Community; SADC = Southern African Development Community. transport costs and removing nontariff barriers, so that intra-industry trade can be further exploited through increased specialization and scale. According to the World Bank Exporter Dynamic Dataset, the number of Tanzanian exporters has been increasing—from 1,190 in 2003 to 1,807 in 2009—which represents growth of 7.7 percent on average per year. This rate Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 171 Figure 9.4 Active Exporters per Million Inhabitants in Rwanda, Tanzania, and Uganda, 2003–11 120 100 Millions of inhabitants 80 60 40 20 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year Tanzania Uganda Rwanda Source: Cadot, Regolo, and Yoshino 2013. is higher than that of Kenya (-3.9 percent) and South Africa (0.3 percent) and higher than the average of all countries in the data set (2.7 percent) over the same period. However, if compared to the size of the country’s popula- tion, the number of exporters in Tanzania is smaller than that in Rwanda and Uganda, and the growth is rather stagnant (see figure 9.4). This growth in the number of exporters has been associated with the diversification into manufactures (see figure 9.5). However, exports remain concentrated, dominated by a few exporters, in a limited number of products and markets per exporter. For instance, between 2003 and 2009, the top 1 percent of Tanzanian exporters (by size) accounted for 61 percent of total exports, and the top 25 percent, for 99 percent. On average, a Tanzanian exporter sold only four products between 2003 and 2009, a rather limited number in comparison to exporters in Kenya (7 products) and South Africa (13 products).3 The number of destinations per exporter, on average, is 2.7 higher than in Kenya and Uganda but smaller than in South Africa (Cadot et al. 2011). Finally, Tanzanian exporting firms are characterized by high mortal- ity rates. That is, even though the absolute number of Tanzanian exporters is growing, the exit rate is also high, suggesting a low survival rate among exporting firms. Based on the Exporter Dynamics Database (see Yoshino 2013), on average, the exit rate increased from 35 percent in 2003 to 53 percent in 2009, leading to a declining survival rate over time. Consequently, and consistent with the findings from Cadot et al. (2011), the survival rate of 172 Women and Trade in Africa: Realizing the Potential Figure 9.5 Contribution to Growth of Tanzania’s Manufacturing Exports, 2004–09 percent Others, 0.1% New exporters, existing products, 44.8% Old exporters, new products, 20.0% 20.0% 44.8% 13.2% Old exporters, existing products, 13.2% 21.9% New exporters, new products, 21.9% Source: IMF Direction of Trade. Tanzanian exporters declined from 45 percent in 2003 to barely 29 percent in 2008.4 In addition to the apparently weak competitiveness and the effects of the global financial crisis, the drastic decline in the survival rate is largely explained by the low level of export entrepreneurship. Gender in the National Trade Policy Tanzania has committed itself to the global policy objective of promoting gender equality and the empowerment of women, as shown by the ratifica- tion of various international and regional declarations and the establishment of a policy and institutional framework.5 Subsequently, Tanzania prepared a National Gender Policy in 1999 that was revised in 2002. This national policy aims to provide guidelines for developing gender-sensitive plans and strate- gies in all sectors and institutions, while emphasizing equal opportunity for men and women. In the specific case of trade policy, gender is reflected as one of the critical aspects and is recognized as a cross-cutting issue. However, the policy does not identify specific programs or interventions that reflect that emphasis. The ongoing review of the National Trade Policy aims to create an environment that fosters female participation in export activities by support- ing micro and small enterprises. However, the policy does not include strate- gies to execute this objective. There remains a disparity in gender equality despite the best intentions of policy makers (box 9.1). One explanation is that Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 173 Box 9.1: Some Key Gender Disparities in Tanzania • Sixty-six percent of unpaid family helpers are women. • Male-headed households are more likely to receive credit than female- headed households (86.7 percent versus 13.2 percent). • Women are less likely than men to be self-employed. • Women earn less than men even with comparable educational levels (39.5 percent of rural women are illiterate, compared to 25.3 percent of rural men). • Men spend a total of 5 percent of their day on unpaid care work (less than an hour and a half ), compared to 14 percent for women (3.5 hours per day). Source: TGNP 2009. the various policies are not supported by corresponding resource allocation (TGNP 2010). A study funded by the International Development Research Centre (IRDC) demonstrated the need for strategies that promote gender equality in the trade sector. It analyzed four export-intensive sectors (fishing, textiles, cut flowers, and coffee), showing that the employment, asset ownership, and dis- tribution of benefits in these sec- tors are less favorable for women (see the case of fishing in box 9.2) (IDRC 2004). Further, Kessy (2011) found that, in the case of cut flowers, female participation is prominent in casual rather than in managerial tasks. Empirical Evidence This section focuses on the link between trade and gender in Tanzania by looking at the participation of women and outcomes related to female ownership among exporters. photo: Trevor Samson / World Bank Several studies have examined 174 Women and Trade in Africa: Realizing the Potential Box 9.2: Gender Patterns of Fishing Activities in Tanzania The IDRC study showed the gender patterns of employment, asset ownership, and distribution of benefits. It concluded that the patterns are less favorable for women. For instance, female profits were far less than male profits due to lack of capital, a constraint that limits women to low-value activities along the value chains where earnings and profits are small. Table B9.2.1 Participation in Fishing Activities by Gender in Tanzania, 2004 Fishing category Total sample Men Women Fishermen 28 25 3 Retailers 61 49 12 Both 6 6 — Others (selling shells) 1 — 1 Total 96 80 16 Source: IDRC 2004 Table B9.2.2 Average Profit from Fishing Activities by Gender in Tanzania, 2004 Amount sold Profit (T Shs) Profit per unit (T Shs) Type of fish Male Female Male Female Male Female Dried Nile 7,000 1,000 3,500,000 50,000 500 50 perch (pieces) Fresh Nile 15 0.5 9,000,000 300,000 600,000 600,000 perch (tons) Tilapia (tons) 4 1.1 1,980,000 324,000 495,000 294,545 Sardines (tons) 13 0 6,500,000 0 500,000 0 Source: IDRC 2004. female entrepreneurship in smaller enterprises (see Mckenzie et al. 2012 and Valdivia 2012 for a review of the literature), including studies conducted for Tanzania by Kweka and Fox (2011). However, few have explored the link between firm ownership and export performance from a gender perspec- tive. The Annual Survey of Industrial Production 2008 dataset is particularly conducive to this type of analysis because, while there is limited information on individual characteristics such as age and educational levels, it allows us to disaggregate firm ownership by gender, which we then analyze in the Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 175 context of firm performance. The main indicators employed for the purposes of this analysis are the “share of female ownership” and a dummy variable indicating whether the firm has any female owners, since a majority of the industrial establishments in the ASIP 2008 have multiple active owners and there are data on the gender of each one. Sector Density The concentration of female ownership is unevenly spread across sectors. Female owners are underrepresented in industrial establishments, with the highest average share of female ownership reaching only 33 percent in the wood and paper manufacturing sector for exporters (figure 9.6). We do not find any export firms with female owners in the mining sector, electronics manufacturing sector, and services industry sector. Among these sectors, there is a stark gap in female ownership—as exporters and nonexporters—in the mining sector, which has the highest average productivity and export volumes. Firm Productivity and Profitability The profit distributions for firms with female owners and those with no female participation in figure 9.7a suggest little difference between the Figure 9.6 Proportion of Female Ownership in Various Sectors of the Tanzania Economy, 2008 35 30 25 20 Percent 15 10 5 0 o s t ts r g r s s eu ls/ pm / od er en he to ge/ pe al cc al ice ui cts in uc pr m tic ac ca et at rry ba pa su rv eq u ra m le m i d ar em a d/ Se ve on ro n s/ qu no oo be cp ph Ch rc ile g/ s/ W d/ he xt ni in al Te o tro Ot in et Fo M ec M El All (Mean share of female owners) Exporters (Mean share of female owners) Source: ASIP, 2008. 176 Women and Trade in Africa: Realizing the Potential Figure 9.7 Distribution of Profit of Firms by Gender Ownership in Tanzania, 2008 9.7a Profit distribution for all firms .2 .15 Density .1 .05 0 8 10 12 14 lnprofit 9.7b Profit distribution for exporters .4 .3 Density .2 .1 0 8 10 12 14 lnprofit Has female owner No female owner Note: kernel = epanechnikov, bandwidth = 0.7150 Source: ASIP, 2008. profitability achieved by male- and female-owned firms in the full sample of firms. However, when the sample is limited to exporters (figure 9.7b), the profit distribution of male-owned firms is higher and skewed to the right of the profit curve related to firms that are partly or fully female owned. The figure shows female-owned firms as a whole are associated with less profit- able activities. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 17 7 This result is interesting, because earlier studies have noted that gender gaps in profits tend to decrease as firm size increases, including for Tanzania. For example, Gamberoni and Haji (2013) used the National Panel Survey (NPS 2008/2010) and the Investment Climate Assessment (ICA 2006) databases for Tanzania and found that earnings gaps between male- and female-owned firms disappear in larger businesses. However, in this data- set, we find that profitability gaps are apparent when we limit the dataset to exporters from the ASIP 2008. However, there are differences in the way that female-owned firms are classified in the two datasets, and the analysis in this chapter is based on exporters of industrial and manufacturing goods, while the ICA 2006 includes all firms in manufacturing, retail, and other sectors. Export Volumes and Domestic Sales In the sample, we find that the volume of exports by the firm decreases as the proportion of female ownership increases. One of the factors that could explain the decreasing export volumes by share of female ownership could be the degree of concentration of domestic sales. Indeed, firms with a greater share of female ownership are more likely to have a lower share of export sales and a higher share of domestic sales (figure 9.8). It is interest- ing that the gap widens once the share of female ownership goes beyond one-third. Here we also note that while the overall volume of exports declines with share of female ownership, the importance of exports to neighboring countries increases while that of exports to countries farther away declines. photo: Arne Hoel / World Bank 178 Women and Trade in Africa: Realizing the Potential Figure 9.8 Share of Domestic and Export Sales by Share of Female Ownership in Tanzania, 2008. percent 90 100 Share of sales/month (%) 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Share of female ownership (%) Share export sales Share local sales Source: ASIP, 2008. Hence, exporting firms with high levels of female ownership tend to focus on nearby markets, suggesting perhaps greater difficulties in accessing the global market than those encountered by male-owned firms. To delve further into the possible determinants of export performance for female-owned export firms, we used regression analysis. Specifically, we looked at the factors determining two main indicators of export and firm per- formance: export sales relative to domestic sales and firm profitability (details are provided in annex A). Since the ASIP 2008 database includes both export- ers and nonexporters, we also look at the factors that affect the probability of becoming an exporter and what traits, if any, are significantly different between exporters and nonexporters (annex B). Consistent with the descriptive statistics presented above, the statisti- cal analysis shows that export firms with female owners are more likely to have a lower ratio of export sales to domestic sales than firms with only male owners. While this result does not have the strongest statistical significance (it is significant only at the 10 percent level), it suggests that, after control- ling for other determinants of exporting, firms with female owners are likely to have export sales that are 1.7 percentage points lower relative to domestic sales than export firms with no female owners. Moreover, all industries were negatively associated with the share of export sales; however, the base group Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 179 for this comparison is the mining industry, which reflects the fact that the mining industry accounts for a large share of export sales.6 Interestingly, export firms with plans to invest in the company by expanding production capacity or upgrading equipment are significantly more likely to have a larger share of export sales. This factor could stand as a proxy for firm performance, as firms that are doing well are more likely to want to invest further and those that have a larger portion of export sales tend perform better than their counterparts. However, one should note that the share of export sales may not neces- sarily be the best indicator of firm performance, as further regressions on the predictors of profit yield contrasting results. Notably, the share of female ownership is significant and negatively related to firm profitability, until industry controls are added. This finding suggests that the sector is a better predictor of firm profitability; however, this relationship is potentially endog- enous, if as a result of gender-related constraints females cluster in sectors that are less productive or profitable. Among the expected results, we find that worker productivity, firm size, and investment plans are all significant and positive predictors of firm profitability. In an additional set of analyses, we look further at the characteristics associated with becoming an exporter. While we find that smaller firms are less likely to be exporters and those firms with greater assets are more likely to be exporters, the analyses based on the ASIP 2008 may not capture additional factors that influence the likelihood of becoming an exporter. For instance, many unobservable characteristics (such as individual ability) may also help explain firm performance; but because we were not able to control for them in the analysis, the results may contain some omitted-variable bias (see, for instance, Gamberoni and Haji 2013). Conclusion and Policy Implications This chapter provides tentative empirical evidence from Tanzania on the links between trade and gender by focusing on both participation and outcomes for women-owned exporting enterprises. The findings show that profitabil- ity decreases as the share of female ownership increases among exporters. However, women-owned firms perform relatively better in cross-border exports than in exporting abroad, suggesting that female-owned firms face more constraints in trading beyond the Tanzanian border than their male counterparts. While it is important to increase female participation in trade activities, the findings underscore the need to focus more on policies for addressing the productivity of female-owned enterprises. Policies that address education and training opportunities for women appear to have significant 180 Women and Trade in Africa: Realizing the Potential potential for narrowing the gender gap in trade outcomes such as export earnings. Furthermore, given the gaps in data on gender indicators, it is important to focus on including gender dimensions in data collection. The findings also lend support to the efforts to promote cross-border trade and regional integration, since they have the potential to improve gender equality. Interestingly, despite the participation of large (formal) enterprises, cross-border trade provides the main source of income for a large number of informal traders who are predominantly poor women and hence represents an important gender dimension to the outcomes on livelihood (Brenton et al. 2011). Focusing on the latter, various studies on Tanzania (Mbise 2012; Meena Ruth 1992) document the challenges and advocate for deliberate measures to address the constraints faced by women in cross-bor- der activities. Such measures could include, for instance, simplifying border trading requirements and increasing awareness of the regulations, as well as adding safeguards to protect female traders. Annex A: Predictors of Export Sales In the first regression specification, we use a fractional logit model (for bounded outcomes in the dependent variable) to look specifically at whether the share of female ownership in export firms is significantly associated with trading at the border, given ownership, firm, and industry characteristics. (1) !ℎ!"#  !"  !"#$%&  !"#$%! = ∝  +  !! !"#  !"#  !"#$%"  !"#$%! +  !! !"#$%&'()('*! +  !! !"#$  !"#$! +  !! !"#"$%&'! +   !! !"#ℎ!"#"$%! + !! !"#$%&'! +  !!  !"#$%&'$"&  !"#$%! +  !! !"#$%&  !"#$%&! +    !! !"#$%&'(! +   !! The explanatory variables used include female ownership among export- ers (1 if any of the active owners are female, and 0 otherwise); productivity (as measured by value added per employee); firm size (measured by num- ber of employees); manufacturing capacity; type of technology employed (manual, semi-automatic, computerized); network reach (measured by an index of membership associations that the firm belongs to); investment plans (whether the firm has plans to invest in expansion or upgrades); source of credit (the firm’s main source of finance for investments and operating capital: options include government funding, bank credit, personal sources, foreign sources, and other means); and industry (sector that the firm belongs to) to control for industry-specific effects. Results are presented in table 9A.1. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 181 Table 9A.1: Regression Results: Predictors of Export Sales (EXPORTERS) (1) (2) (3) (4) share_ share_ share_ share_ Variables exportsales exportsales exportsales exportsales anyfemown 0.0524 -0.426 -0.576 -1.712* lnvalueadd -0.179 -0.0444 -0.323 firmsize_Small -1.652 -1.810 0.984 firmsize_Medium 0.536 0.572 2.768** lncapacity 0.124 0.0788 0.0462 tech_manual -0.00596 -0.00213 0.00124 tech_semiAutomatic -0.026*** -0.0160* -0.0230*** tech_computerized -2.066 -0.0381* -0.0743** network 1.626*** -1.409 0.0941 investment_plan 1.581*** 2.589*** source_govt -7.048*** -5.174*** source_bankcredit 0.182 0.504 source_personal 0.214 -0.348 source_oth -2.649** 0.862 source_for 3.432*** 2.112*** industry_Food -8.044*** industry_Textiles -8.030*** industry_Wood -8.924*** industry_Chemicals -14.44*** industry_Metals -12.46*** industry_Electronics -12.41*** industry_Consumer -9.439*** Constant -0.863*** 0.134 -1.153 8.967*** Observations 57 41 41 41 N 57 41 41 41 df_m 1 10 13 19 II -30 -15.96 -14.74 -10.09 Note: Robust standard errors are in parentheses. *** p<0.01, **p<0.05, ***p<0.1 182 Women and Trade in Africa: Realizing the Potential Annex B: Predictors of Firm Profitability In the second specification, we look at the extent to which gender and proportion of border trade are associated with profits while controlling for ownership, firm-, and sector-specific characteristics. (2) ln !"#$%&' ! =  ∝  +  !! !"#$%"  !"#$%  !ℎ!"#! +  !! !ℎ!"#  !"  !"#$%#  !"#$% + +    !! !ℎ!"#  !"  !"#$%&  !"#$% +  !! !"#$%&'()('*! +  !! !"#$  !"#$! +  !! !"#"$%&'! +   !! !"#ℎ!"#"$%! + !! !"#$%&'! +  !!  !"#$%&'$"&  !"#$%! +  !!" !"#$%&  !"#$%&! +   !!! !"##$%#&! + !!" !ℎ!""#$%#&! +  !!" !"#$%&'(! +   !! This specification aims to examine the strength of association between export firm profitability and share of border trade from a gender-differenti- ated angle. Explanatory variables include share of female ownership, share of border trade (as compared to trade beyond country borders), manufacturing capacity; type of technology employed; network; investment plans; source of credit; barriers faced by the firm (these include options such as problems with supply capacity, stringent sanitary and phytosanitary rules, low compli- ance with market standards, inability to meet quality standards, inadequate capacity to exploit preferential trade agreements, limited promotion, low skills, and other barriers); perception of challenges (measured by an index of challenges perceived by the firm); and type of industry. Results are presented in table 9B.1. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 183 Table 9B.1: Regression Results: Predictors of Firm Profitability (Exporters) (1) (2) (3) (4) Variables lnprofit lnprofit lnprofit lnprofit sharefemown -0.0172** -0.0132*** -0.0203*** -0.0169 fraction_border -0.276 -0.423** -0.891** share_exportsales -0.470 -0.463 -0.250 lnvalueadd 0.863*** 0.792*** 0.935*** firmsize_Medium 0.832*** 0.874** 1.264* firmsize_Large 1.662*** 1.416*** 2.294** lncapacity 0.0279 0.0127 0.0186 tech_manual -0.00194 -0.00483 -0.00293 tech_semiAutomatic -0.00349 -0.00634** -0.00587* tech_computerized -0.00535 -0.0120*** -0.00635 network 0.877 0.873* -0.363 investment_plan 0.530** 0.595*** 0.398* source_govt 2.726*** 4.083*** source_bankcredit 0.376*** 0.362* source_personal -0.260 0.184 source_oth -0.444 -0.593 source_for -1.732*** -1.640*** barrier_SupplyCapacity -0.0833 0.226 barrier_SPSrules -0.402 -0.265 barrier_CannotMeetStd -0.178 -0.900 barrier_LowCapacity 0.0830 0.539 barrier_LowSkills -0.238 0.232 barrier_Other -0.519 -0.346 challenge_index 0.113 -0.795 industry_Food 0.413 industry_Textiles -0.295 industry_Chemicals 0.853 industry_Metals 0.781 industry_Electronics 0.326 industry_Consumer 0.634 industry_Services -0.346 Constant 11.75*** 2.350** 3.768*** 1.646 Observations 57 39 39 39 R-squared 0.084 0.881 0.976 0.988 N 57 39 39 39 df_m 1 12 22 28 F 6.968 31.72 rss 106.7 7.385 1.499 0.715 II -98.75 -22.89 8.205 22.65 Note: Robust standard errors are in parentheses. *** p<0.01, **p<0.05, ***p<0.1 184 Women and Trade in Africa: Realizing the Potential Annex C: Probability of Becoming an Exporter In the third specification, we use a probit regression framework to analyze the probability of being an exporter conditional on owner, firm, and industry characteristics. The sample includes all firms in the database, and the set of control variables in this specification include owner and employee character- istics, firm characteristics, and industry characteristics. The regression is specified in the following way: (3) !"#$%&'%! = 1  if  !! !"#$%"  !"#$%  !ℎ!"#! +  !! !"#$%&'()('*! +  !! !"#$  !"#$! +   !! !""#$"! +    !! !"#"$%&'! +   !! !"#ℎ!"#"$%! + !! !"#$%&'! + !! !"#$%&'$"&  !"#$%! +!! !"#$%&  !"#$%& +   !!" !"#$%&'(   + !! >0 !"#$%&'%! = 0  otherwise Where Exporter takes the value 1 if the firm is identified as an exporter in the ASIP (2008) and takes the value 0 otherwise. Controls include busi- ness-owner characteristics (such as share of female owners); firm character- istics (such as productivity), manufacturing capacity (as a proxy for size), and type of technology used (proxy for level of innovation); network reach (measured by source of finance and membership in business associations); and sectors of specialization (to control for sector-specific effects). Table 9C.1 shows the marginal effects calculated at the mean of the continu- ous variables on exporting activities. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 185 Table 9C.1: Regression Results: Probability of Becoming an Exporter (Whole Sample) (1) (2) (3) (4) Variables exporter exporter exporter exporter sharefemown 0.000329 0.00107 0.00110 0.00121 lnvalueadd 0.0306* 0.0297* 0.0209 firmsize_Small -0.239*** -0.240*** -0.278*** firmsize_Large 0.0173 0.0180 -0.0134 lnassets 0.0175 0.0172 0.0218** lncapacity -0.00210 -0.00140 -0.00107 tech_manual -0.000299 0.00040 0.000158 tech_semiAutomatic -0.000246 0.000311 0.000213 tech_computerized -0.00101 -0.00111 -0.000827 network 0.0471 0.0525 0.0581 investment_plan -0.0279 -0.0248 -0.0273 source_govt 0.0348 0.136 source_bankcredit 0.00464 -0.180 source_personal -0.000127 0.00716 source_oth 0.0303 0.00912 source_for -0.0553 -0.0586 industry_Food -0.107 industry_Textiles -0.0759 industry_Wood -0.0729 industry_Metals -0.0843 industry_Electronics -0.0403 industry_Consumer -0.0715 industry_Services -0.105* Observations 410 220 220 216 N 410 220 220 216 df_m 1 11 16 23 F rss II -165.2 -71.73 -71.33 -67.29 Note: Robust standard errors are in parentheses. *** p<0.01, **p<0.05, ***p<0.1 186 Women and Trade in Africa: Realizing the Potential Notes 1. The current goal of Tanzania’s National Trade Policy is to facilitate smooth integration into the Multilateral Trading System by transitioning the economy toward competitive export-led growth. 2. Based on data from the Africa Development Indicators (World Bank 2013), the share of exports and services to gross domestic product increased from 12.6 percent in 1990 to 26.1 percent in 2009–10; in other countries in the region, that share rose 25.7 percent to 27.5 percent in Kenya, 7.2 percent to 23.8 percent in Uganda), 5.6 percent to 10.9 percent in Rwanda, and 24.2 percent to 27.3 percent in South Africa. However, the extent of openness in Tanzania is below the average of 31.2 percent for the Sub-Saharan Africa region, and in some export-intensive economies—such as Mauritius (over 50 percent), Botswana (33 percent), and Ghana (about 30 percent)—openness is quite high. 3. This is based on an analysis at the six-digit level in the Harmonized Commodity Description and Coding System. 4. Survival rate t = number of survivors t/number of entrants t. “Survivors” are those that did not export in t-1 but exported in t and t+1. 5. For instance, following the Beijing Platform for Action (UN 1995) that required all United Nations member states to generate and disseminate gender-disaggregated data in policy, the government of Tanzania through the Ministry of Finance published the Tanzania Gender Indicators booklet in 2010 to increase understanding and awareness of gender-disaggregated data in policy planning, programming, monitoring, and evaluation. 6. We use the mining sector as a comparator, which, even though it has a large discrepancy in female ownership (as exporters and nonexporters), it is the sector with the highest average productivity and export volumes, which is important because we are also looking at the overall sector correlation with the share of export sales. References Brenton, Paul, Celestin Bashinge Bucekuderhwa, Caroline Hossein, Shiho Nagaki, and Jean Baptiste Ntagoma. 2011. “Risky Business: Poor Women Cross-Border Traders in the Great Lakes Region of Africa.” Africa Trade Policy Note 11, World Bank, Washington, DC. Cadot, Olivier, Leonardo Iacovone, Denisse Pierola, and Ferdinand Rauch. 2011. “Success and Failure of African Exporters.” Policy Research Working Paper 5657, World Bank, Washington, DC. Trade and Gender in Tanzania: What Matters—Participation or Outcomes? 187 Cadot ,Olivier, Julie Regolo, and Yutaka Yoshino. 2013. “Analysis of Tanzania Export Survival Diversification” Paper prepared as part of the Technical Background Report on Export for the 2014 World Bank Tanzania Country Economic Memorandum (mimeo). Gamberoni, Elisa, and Mahjabeen Haji. 2013. “Individual Characteristics and Business Dynamics: An Analysis of Male and Female Entrepreneurs in Tanzania.” Unpublished paper. World Bank, Washington, DC. IDRC (International Development Research Centre). 2004. “Synthesis of Sector Studies: Kenya, Tanzania and Uganda.” Research Project on Gender and International Trade in East Africa, March 2004, Economic and Social Research Foundation, Dar es Salaam. Kessy, F. 2011, “Gender Dimensions in the Cut Flower Sector in Tanzania: Implications for Economic and Social Policy.” Mimeo, Dar es Salaam. Kilic, T. Amparo Palacio-Lopez, and Markus Goldstein. 2013. “Caught in a Productivity Trap: A Distributional Perspective on Gender Differences in Malawian Agriculture.” Policy Research Working Paper 6381,World Bank, Washington, DC. Kweka, J., and D. Booth. 2005.“Precedence for Trade and Poverty Linkage in Tanzania.” Tanzania Trade and Poverty Programme, ESRF/ODI, London. Kweka, J., and Louise Fox. 2011. “The Household Enterprise Sector in Tanzania: Why It Matters and Who Cares.” Policy Research Working Paper 5882, World Bank, Washington, DC. Mbise, N. 2012. “Final Report on the Assessment of Women Traders Engaged in Cross Border to Support Formation of Networks in Tanzania.” United National Entry for Gender Equality and the Empowerment of Women, UNWOMEN, Dar es Salaam, Tanzania. McKenzie, David & Woodruff, Christopher. 2012. “What are we learning from business training and entrepreneurship evaluations around the develop- ing world ?,” Policy Research Working Paper Series 6202, The World Bank. Ruth Meena. 1992. (ed), Gender in Southern Africa: Conceptual and Theoretical Issues, Harare, SAPES Books. Brenton, Paul & Newfarmer, Richard. 2007. “Watching more than the Discovery channel: export cycles and diversification in development,” Policy Research Working Paper Series 4302, The World Bank. TGNP (Tanzania Gender Networking Programme). 2010. “Employment, Livelihoods and Trade: What Plans Are There for the Poor and Marginalized?” TGNP Policy Brief, Gender Resource Center, Dar es Salaam. 188 Women and Trade in Africa: Realizing the Potential TGNP (Tanzania Gender Networking Programme). 2009. “Women Participation in Positions of Power and Influence in Tanzania”, TGNP, Dar es Salaam. Tran-Nguyen, Anh-Nga, and AmericoZambeti, eds. 2004. “Trade and Gender: Opportunities and Challenges for Developing Countries.” UN Inter- Agency on Women and Gender Equality Task Force on Gender and Trade, United Nations, New York. United Nations, 1995 “World Summit on Beijing Platform for Action” New York. URT (United Republic of Tanzania). 2003. “National Trade Policy: Trade Policy for a Competitive Economy and Export-Led Growth.” Dar es Salaam, Ministry of Industry and Trade. URT (United Republic of Tanzania). Annual Survey of Industrial Production, 2008. National Bureau of Statistics. Ministry of Finance. Dar es Salaam. URT (United Republic of Tanzania). Tanzania National Panel Survey Report -. Wave 1 (2008-2009), Wave 2 (2010–2011). National Bureau of Statistics. Ministry of Finance. Dar es Salaam. Valdivia, Martin. 2012. “Training or technical assistance for female entrepre- neurship? Evidence from a field experiment in Peru”, Mimeo. GRADE. World Bank. 2006. Enterprise Surveys. “Investment Climate Assessment: Enterprise Performance and Growth In Tanzania.” Washington DC. World Bank. 2011. “Gender Equality and Development: World Development Report 2012. Washington, DC: World Bank. World Bank. 2013. The Little Data Book on Africa. Washington, DC: World Bank. World Development Indicators (database). World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators. Yoshino, Yutaka. 2013. “Uncovering Drivers for Growth and Diversification of Tanzania’s Exports and Exporters.” Background Report for the 2014 World Bank Tanzania Country Economic Memorandum, World Bank, Washington, DC. Front cover photo credits clockwise: Arne Hoel for the World Bank; Trevor Samson for the World Bank; Arne Hoel for the World Bank; James Martone for the World Bank; Josanna Lewin for the Advocacy Project.