Report No.:62992-AFR Facilitating Cross-Border Trade between the DRC and Neighbors in the Great Lakes Region of Africa: Improving Conditions for Poor Traders June, 2011 Poverty Reduction and Economic Management Africa Region World Bank Document Facilitating Cross-Border Trade between the DRC and Neighbors in the Great Lakes Region of Africa: Improving Conditions for Poor Traders Table of Contents Acknowledgement.............................................................................................................................i Executive Summary .......................................................................................................................... ii I. Introduction........................................................................................................................ 1 II. The Potential for Cross-Border trade in the Great Lakes Region ....................................... 4 III. Conditions Facing Cross-Border Traders in the Great Lakes Region .................................. 8 IV. Physical Conditions at the Border and Border Procedures .............................................. 11 V. Steps to Facilitate Cross-Border Trade in the Great Lakes Region ................................... 14 VI. Conclusions and moving forward ..................................................................................... 21 VII. References ........................................................................................................................ 24 Annex 1: Assessing the Impact of Roads and Borders on Agricultural Prices in the Great Lakes Region. ........................................................................................................................... 25 Annex 2: The survey of cross-border traders ............................................................................... 31 Figures Figure 1: Recorded Cross-Border Trade in the Great Lakes Region (2009, $000)…………………………6 Figure 2: GDP per Capita in the Great Lakes Region – 2007-2009 (US dollars).............................. 7 Figure 3: Reported frequency of risks by cross border traders ...................................................... 9 Tables Table 1: Recorded Cross-Border Trade in the Great Lakes Region (2009, $000)…………………………2 Table 2: Estimated Informal and Formal Trade in Food Staples in 2009 in East Africa (tons) ....... 3 Acknowledgement T his report has been prepared by a team consisting of Paul Brenton (TTL, AFTPM), Shiho Nagaki (AFTP3), Caroline Hossein (international consultant, traders survey) , Celestin Bashige, Jean-Baptiste Ntagoma (local consultants, DRC, traders survey), Alberto Portugal (DEGTI, price analysis), Nicholas Sabwa and Adrian Mukhebi (international consultants, cross-border trade), Michel Zarnowiecki (international consultant, customs). This report has benefited from seminars and presentations in Goma, Bukavu, Uvira, Kinshasa, Bujumbura and Kigali and to the Trading for Peace team in the COMESA secretariat. The seminars in the east of the DRC brought together traders and government officials. There have also been presentations in Geneva and Washington. The report has benefitted from the comments of the peer reviewers Luigi Giovine, Bernard Harborne, Andrew Roberts, Philip Schuler and Paolo Zacchia and the suggestions of Han Herderschee. This work was financed by the Multi-Donor Trust Fund for Trade and Development supported by the governments of United Kingdom, Finland, Sweden and Norway. i Executive Summary This report looks at the current situation and opportunities for cross-border trade between DRC and neighbors with a focus on agricultural products. It recognizes that much of current trade across these borders is informal and carried by individual traders, most of whom are poor women. The report shows that it is very important that informality not be equated with illegality. The large numbers of traders that cross the border every day do so through official border crossings. The aim of the report is to provide information on the current state and the potential for cross border trade between the DRC and Burundi, Rwanda and Uganda, to document the conditions and problems that traders face in crossing the borders and provide suggestions to policy makers on the priorities for trade facilitating measures at the border. Exploiting the potential for cross-border trade will be an important element of growth and poverty reduction in the region and a key mechanism for enhancing stability. Comparing prices for food products in a range of markets in the three countries, the report underlines the strong role that cross-border trade could play in reducing differences in price levels and price volatility of agricultural products. However, this potential is only partly being realized, particularly in exchanges across the DRC border. The effect of crossing the Burundi-Rwanda border on relative prices is equivalent to pushing the two markets an additional 174 km or 4.6 hours further apart. However, crossing the Burundi-DRC border is equivalent on average to pushing markets in each country 1824 km or 41 hours further apart whereas crossing the DRC-Rwanda border is equivalent to adding an extra 1549 km and an additional 35 hours. Crossing the DRC border is a risky business; traders face harassment, physical violence and financial losses. The report presents information from a survey of cross-border traders on the conditions that they face in crossing the border. The conclusions are stark: the livelihoods and activities of the primarily female traders are currently being undermined by high levels of harassment and physical violence at the border and the prevalence of unofficial payments and bribes. For many of the women, trading is the only source of income for their families and so despite these conditions they have no alternative but to continue to trade, they are stuck in a survival economy. Most lack representation and are poorly organized. Nevertheless, the vast majority wish to expand and grow their business. To be able to do this they need security and safety in crossing the border and transparent and predictable rules for border processing. The report suggests critical steps to facilitate cross-border trade in the Great Lakes region. The immediate priority is to improve the conditions at the border and the treatment of cross- ii borders traders. This will have a significant impact on the livelihoods of a substantial number of women and the households who depend on their income from trading. It will also lead to more cross-border trade and more widely available and cheaper food products. The second imperative is to provide and implement a policy framework by which these traders can become increasingly organized backed up by support services that improve access to information, facilitate access to credit and ensure better representation of traders’ interests. This will provide a route by which these activities can gradually become more formalized and distribution channels become more efficient. The key steps are: ï‚· Officials at all levels recognize the important role played by poor cross-border traders. Poor women cross-border traders are currently viewed negatively by policy makers reflecting the misconception that they are an important source of revenue loss. ï‚· A comprehensive border management policy is established and implemented. This must ensure that: (i) only the four agencies that are legally entitled to be at the border are present; much of the harassment arises from the large number of agencies (sometimes up to 15) that are operating at the border; (ii) that there is an increase in the transparency and predictability of the trade policy regime. Practical measures that could quickly be made are that all payments made to an official should be recorded and an official receipt provided; payments should be made only at a single, clearly designated building; there should be no payments in kind to officials. (iii) there are improvements to infrastructure at border crossings. Both the transport infrastructure and the poor quality and inappropriate design of buildings used by officials need to be rehabilitated, including the provision of basic facilities for officials such as electricity and water. Work is immediately required to minimize the risks to safety and security that arise from the dilapidated infrastructure and to provide an open and transparent area for officials and traders. (iv) strengthened cross-border coordination and cooperation. Key border agencies coordinate insufficiently with Burundi, Rwanda and Uganda. Normally, regular contact between Customs authorities of both countries is a key element of successful border management. There are opportunities for sharing of border facilities but decision is required at the highest level in the DRC for these to be taken up. ï‚· Increased professionalism of officials and greater gender awareness. A first step is to ensure that all officials are either in uniform or wear constantly a visible badge. Women are under-represented in official bodies at the border and so increasing the number of female border officials is an important step. Nevertheless, lessons show that what is needed are gender-aware officials who appreciate the issues confronting traders, poor women and men alike. Obvious policies such as only female officers should be allowed to body search women traders should be made mandatory and widely publicized. This must be supplemented by enhanced training of officials on customs procedures and on gender awareness. iii ï‚· Increase support for traders through traders associations. There is a need to increase the representation of traders through traders associations to ensure that their interests are properly represented in public policy and as a mechanism to provide information and training. ï‚· A strategy for integrating poor entrepreneurs into the formal economy. It is also necessary to start to address key concerns of these traders regarding their lack of access to finance, lack of information on prices and business opportunities and training in basic business practices. This report concludes that there would be very large economic pay-offs from interventions that reduce the size of the border and facilitate freer movement of goods and people across the border. In the short-term this pay-off would be in the form of greater security and increased incomes for poor women traders whose trading activities are crucial to the welfare of their households and hence to poverty reduction in communities along the border. In addition, interventions to remove the major barriers to cross-border trade between the DRC and Burundi and Rwanda would allow for trade to play a greater role in integrating food markets in the region. This would increase the returns to farmers and provide genuine alternative sources of income to those currently linked to activities that destabilize the region. There has been progress on this issue at both the regional and the national level. Seminars have recently taken place in Goma and Bukavu which have brought together representatives of the regional government, officials from border agencies and traders. The Governor of South Kivu has visited the border at Bukavu. In recent interviews, traders at Bukavu attest that this heightened interest from policy makers has been reflected in less harassment. In Kinshasa, the issues of cross-border trade and the conditions facing poor women traders were discussed at a workshop in April 2011. At this meeting agencies agreed a detailed and sequenced roadmap towards improving border procedures at two pilot crossings in the east (Goma and Bukuva). Implementation of the roadmap will streamline processes and achieve clear agreement between the different agencies as to who should do what, with the preparation of clear job descriptions. The Bank is implementing a pilot project to improve trade facilitation across the DRCs Eastern borders with a focus on the Petite Barriere at Goma. The project will seek strategic improvements in infrastructure to create a more transparent and open space and reduce security and safety risks, will provide training for customs officials and security officers for improved and professional behavior of officials at the border and support to empower and build the capacity of traders associations. The project will be accompanies by an extensive impact evaluation. The baseline for this will be collected in May/June and the project will start to make improvements at the border in July. iv I. Introduction Peace and relative stability since 2008 provides an opportunity for restoring the local economy in the Great Lakes region. Well known as a conflict zone since the early 1990s, the economy of the Great Lakes region in Africa and, in particular, the east of the Democratic Republic of Congo (DRC) has been devastated by conflict. This war is often cited as the world’s deadliest since World War II, killing 5.4 million people in DRC between 1998 and 2003. Though still struggling with instability and fragility, the region has begun to take steps forward to provide a base for economic recovery, particularly, through several peace agreements, reactivation of regional political dialogue, and the large scale of international aid including the world’s largest UN peacekeeping mission. At present most of the revival in the DRC has been in the informal sector reflecting continued uncertainty over security and lack of clear regulations and poor governance of state institutions.1 There is enormous potential for international trade to drive growth and poverty reduction in the region. At the forefront of these opportunities are exports to the global market of mineral exports from the DRC. But also, and relatedly, there is considerable scope for all countries and economies in the region to benefit from cross-border and regional trade in a range of goods and services, which can contribute to food security, to increasing employment throughout the region and the return of internally displaced persons (IDPs), as well as provide improved access and lower prices for critical inputs into the exports of other goods and services. Exploiting these opportunities for cross-border trade, while smaller in value terms than exports of minerals, is likely to have a much more direct impact on incomes and employment in the region and improve outcomes for many households. In addition, while an improvement in the security situation is necessary for output and trade to recover, greater economic opportunities and rising living standards will contribute to greater stability. For example, increasing returns to agricultural activities can provide genuine alternative occupations and sources of income to artisanal mining and thus facilitate the shift to a more organized and less disruptive mining sector. Harnessing the opportunities of regional and international markets will play a key role in delivering higher returns to farmers and contributing to food security. Cross-border trade will become increasingly important in linking food surplus areas to food deficit areas, especially as development is accompanied, and driven, by the increasing concentration of people and activities in towns and cities. Rising incomes will lead to increasing demands for a wider range of goods and services and enhance the scope for mutually beneficial cross-border exchanges 1 The DRC Country Economic Memorandum focuses on these issues and the political economy constraints that limit progress in addressing them. 1 between the DRC and the other countries of the Great Lakes region in agriculture, manufactures and services. Recorded trade flows between the DRC and neighbors in the Great Lakes are relatively small and are mainly towards the DRC. Table 1 presents the available official data on merchandise trade between the DRC, Burundi, Rwanda and Uganda in 2009.2 These recorded cross-border trade flows are very small relative to the total imports of each country. For example, Rwanda’s imports from the DRC amount to about one half of one percent of total Rwandan imports in 2009. It is also apparent that exports from Burundi, Rwanda and Uganda to the DRC currently exceed their imports from the DRC. Official data for the 1970s suggest that at that time the DRC was a net exporter to Rwanda and Burundi and a net importer from Uganda. In 2009 the key products in the recorded exports of Uganda and Rwanda to the DRC were related to construction – cement and iron and steel products. The main commodities recorded in these countries imports from the DRC were wood, vegetable oils, medicines and perfumes. Table 1: Recorded Cross-Border Trade in the Great Lakes Region (2009, $000) Importer Burundi DRC Rwanda Uganda Exporter Burundi 7037 3291 426 DRC 652 5975 4346 Rwanda 1738 11392 3071 Uganda 28515 78307 141853 World 337649 1109830 4246228 Source: WITS Currently it is informal cross-border trade that plays the main role in linking producers to markets in the Great Lakes region of Africa. Official statistics vastly understate the amount of trade that crosses borders in the region. For example, the number of trucks crossing the border is a tiny fraction of the tens of thousands of people who cross official borders in eastern DRC every day for commercial purposes. This informal trade is not illegal trade but is unorganized small scale trade which does not appear in the customs record. It has to be clear that this small scale informal trade is “officialâ€? in the sense that traders go through official border posts, pay a crossing fee to the immigration office, and if processed appropriately pay a duty on imports.3 Sabwa and Mukhebi (2011), in a background paper for this report, estimate that the volume of the DRC’s informal trade in food staples with countries in the Great Lakes region is almost 5 times greater than the volume of formal trade in these products (Table 2). For 2 Official data from the DRC have not been provided to the United Nations in recent years and so we use exports from each of the partner countries. 3 Even though there is no legal tariff on exports, traders are often asked to pay a duty. 2 the other countries in the region, informal trade in food staples, while not as dominant as in the DRC, is still an important aspect of cross-border trade. Cross-border exchange provides the main source of income for a large number of informal traders who are predominantly poor women carrying agricultural products and hence there is an important gender dimension to this issue. There is however, very little information on these traders and the conditions that they experience.4 There are key challenges relating to the illegal trade in minerals and timber. The issues regarding minerals have been well documented in reports under the ‘Trading for Peace’ initiative and international measures have been taken regarding schemes to certify minerals such as coltan and cassiterite. This report looks at the current situation and opportunities for cross-border trade between DRC and neighbors with a focus on agricultural products. The report brings together work that provides information on the current state and the potential for cross border trade between the DRC and Burundi, Rwanda and Uganda. Comparing Table 2: Estimated Informal and Formal Trade in Food Staples in 2009 prices for food products in a in East Africa (tons) range of markets in the three Congo DR Kenya Rwanda Uganda countries, the report underlines Formal 3276 22728 8286 205583 the strong role that cross-border trade could play in reducing Informal 16,078 9,116 2,177 114,879 differences in price levels and price volatility of agricultural Ratio of informal 4.91 0.40 0.26 0.56 to formal products. However, this potential is only partly being Source: Sabwa and Mukhebi (2011) realized, particularly in exchanges across the DRC border. This analysis is complemented by information from a survey of cross-border traders to ascertain the conditions that they face in crossing the border. The conclusions are stark: the livelihoods and activities of the primarily female traders are currently being undermined by high levels of harassment and physical violence at the border and the prevalence of unofficial payments and bribes. To improve conditions at the border requires improvements in infrastructure, simplification of trade procedures and improved border management by the authorities in DRC, training of officials and support to traders associations. Regional cooperation and regional economic integration can be key levers by which to reduce the policy and infrastructure constraints that constrain cross-border trade. The main 4 There is a report by International Alert (2010) that provides important information on the nature of goods carried and the conditions facing traders at the Goma-Gisenyi border crossing. Useful information has also been provided in Mthembu-Salter (undated), Coulibaly and Maburuki (2006) and Masinjila (2009). 3 actions to improve procedures and address governance issues have to be taken by countries themselves. The focus in this report is on the DRC. As a result of their own reforms and implementation of the EAC protocol on trade Burundi and especially Rwanda and Uganda, have progressed further in reforming policies at the border and strengthening their implementation. There is still however, much more to be done in these countries in assisting informal traders to prosper and integrate into to formal economy. Nevertheless, these individual policy reforms can be complemented by information sharing, harmonization of trade regimes and regulatory requirements and joint action at border posts and working together to identify and address key infrastructure priorities. These opportunities are yet to be realized and an important issue is the appropriate regional forum in which to address them. Burundi and Rwanda are members of CPEGL(Communauté Economique des Pays des Grands Lacs) and they joined the EAC in 2009 and started the process of integrating into the market in eastern Africa. The DRC is a member of COMESA, SADC, CEEAC and CPEGL, but does not have clear policy toward regional economic integration in terms of trade policy. The report proceeds by describing the current economic situation in the Great Lakes region and then provides a brief summary of a detailed analysis of relative prices of a number of basic food products in markets in Burundi, DRC and Rwanda and what this implies for the width of the borders in the region. The next section then summarizes the results from the survey of cross-border traders which is followed by a set of recommendations. The final section concludes and describes a number of steps that are being taken to address the constraints identified in this report and discusses ways in which these may be built upon. II. The Potential for Cross-Border trade in the Great Lakes Region The economies of eastern DRC, Burundi, Rwanda and western Uganda have been inextricably linked for centuries. However, the conflicts of recent years have taken a heavy toll on human life and on the regional economy. As the region emerges from conflict there is a huge potential for growth of the regional economy with massive reserves of natural resources. In addition to mining and forestry, the DRC has enormous agricultural potential, but political instability and insecurity have undermined agricultural production and food security. Food production in the DRC fell by between 30 and 40 per cent over the past 10 years (IPC (2009)). North and South Kivu in the east of the country are areas of highest agricultural potential and are a “potential breadbasketâ€? (Ulimwengu et al (2010)). Despite this potential, 4.5 million people in the DRC are living in food crisis conditions (8% of total population) 5. The situation is particularly acute in the east of the DRC where the conflict and continuing lack of security has 5 th The 4 Integrated Food Security Phase Classification (IPC) conducted by the Ministry of Agriculture with support from FAO 4 displaced around 14% of population or 1.3million people (HCR, 2010). These IDPs have less access to land (43.1%) for agricultural production activities than the local population (81.7%) while much of the land remains untended (WFP SK (2009)). Cultivation methods remain very basic and much of the current production is for subsistence rather than for the market, reflecting factors including the absence of access to both local and regional markets and lack of information on market opportunities. Returning the IDPs to their communities and bringing stability will encourage the recovery of agricultural production and increase opportunities for trade to the regional market. Cross-border trade is a key mechanism linking producers to markets and moderating price differences for key food products. While the instability and insecurity of the past 20 years has caused serious disruption to cross-border trade, substantial flows of people and goods continued to traverse the borders between these countries every day. Now that a degree of stability has returned, trade flows are again increasing but remain primarily informal. This report seeks to add light to the reasons why agricultural trade remains primarily informal. It shows that current movements are dominated by poor women, for whom trading activities are a critical source of household income, who often face harassment and extortion when crossing the border and have no support or framework for moving towards formal activities. Linking producers and consumers in the east of the DRC to large markets through international trade can play the key role in driving growth and income convergence. The east of the DRC is proximate to large and increasing concentrations of economic activity. The importance of exploiting economic links with east Africa is highlighted by Figure 1.6 The figure captures the location and intensity of economic activity through emissions of light at night. Given the location of North and South Kivu and the limited westwards links to the rest of the DRC, the closest markets are in neighboring countries. Hence, facilitating trade across borders will be an important mechanism to improve access of consumers to basic food products and to increase the returns to farmers. At present most of the food products are flowing into the DRC7, reflecting the remaining problems of insecurity in rural areas and movement of IDPs to the towns and cities, poor infrastructure in the DRC and the constrained level of agricultural production. Nevertheless, as normality returns and infrastructure weaknesses are addressed the 6 Measuring economic activity in regions where there is a large amount of informal activity is clearly very difficult. Recently, an alternative approach to measuring national, regional and sub-regional GDP/economic activity has emerged using the light emitted at night as measured by satellites. Henderson et al (2009) show that there is a strong correlation at the national level between night light and measured GDP which suggests that such information is useful for providing a picture of the geographic distribution of economic activity. 7 In the medium to longer term as stability is entrenched in the DRC and agricultural output is restored the potential is for flows of agricultural products out of the DRC towards the large and increasing agglomerations in Bujumbura, Kigali, Kampala and other cities in the Great Lakes countries but also to Nairobi and Dar-es-Salaam. This will raise issues of compliance with standards and non-tariff barriers that may restrict trade in these products which will necessitate increased regional cooperation. 5 potential for the east of the DRC to be a significant exporter of a range of agricultural products can be realized. Figure 1: The Distribution of Economic Activity in the Countries of the Great Lakes Region as shown by Night Light Data There is substantial scope for provinces in the East of the DRC to catch up the higher and fast growing incomes of neighboring countries. Incomes are growing strongly in East Africa (Figure 2) and currently GDP per capita in Rwanda and Uganda is more than twice that of the DRC. This contrasts with the situation in 1980 before the crisis when GDP per capita in the DRC was $530, more than double that of Rwanda ($224) and more than five times that of Uganda ($98).8 Income per head in North and South Kivu is a little higher than the average for the DRC but substantially below that of Rwanda, Uganda and Kenya. This suggests that there are good opportunities for incomes in the Kivu’s to catch up with those of the wealthier and currently faster growing neighbors in East Africa. 8 Source WDI 6 Figure 2: GDP per Capita in the Great Lakes Region – 2007-2009 (US dollars) 900 800 700 600 500 USD 400 300 200 100 0 North Kivu South Kivu DRC Burundi Rwanda Uganda Kenya 2007 172 215 160 125 396 388 720 2008 202 254 180 145 483 456 775 2009 195 211 160 160 522 490 738 Source: World Development Index, Provincial Macroeconomic Framework, World Bank 2011 Analysis of market level price data suggests that crossing the border into the DRC is a major hurdle for traders of food products. Annex 1 summarizes a study of the prices of key food products such as maize and cassava in different markets in the DRC, Burundi and Rwanda (full details are in the background report by Brenton and Portugal (2011)). The analysis seeks to explain differences in prices between markets and differences in the volatility of those prices. In competitive markets without major distortions, spatial arbitrage acts to ensure that the price difference between two markets will not exceed the full cost of shipping the commodity from the high price to the low priced market. The full cost will mainly comprise the costs of transportation. The activities of the arbitragers who make a profit from the price differences between markets has beneficial economic impacts in terms of moving food from areas of relative surplus to areas of deficit and reducing the price of food in the deficit area. When the markets are on different sides of a border then any additional costs of crossing the border constrain the role of arbitrage in reducing prices differences. Interventions that reduce the costs of shipping products between markets, such as better roads and removal of road blocks, will lead to smaller price differences. The analysis finds that after taking into account both product and market specific factors as well as proxies for the costs of transporting goods between markets that there is an additional effect that characterizes markets on different sides of a border. The variables that capture the effect of policy constraints at the border are positive and significant, although much larger for crossing the DRC border. The effect of crossing the Burundi-Rwanda border on relative prices is equivalent to pushing the two markets an additional 174 km or 4.6 hours further apart. However, crossing the Burundi-DRC border is equivalent on average to pushing 7 markets in each country 1824 km or 41 hours further apart whereas crossing the DRC-Rwanda border is equivalent to adding an extra 1549 km and an additional 35 hours. Crossing a border is also associated with an increase in the volatility of prices and again the impact is much greater when crossing the DRC border than the border between Burundi and Rwanda. This suggests that there are major barriers to cross-border trade between the DRC and Burundi and Rwanda and that there are substantial opportunities for trade to play a greater role in integrating food markets in the region. Hence, there would be very large economic pay-offs from interventions that reduce the size of the border and facilitate freer movement of goods and people across the border. It is the type and nature of constraints on cross-border trade to which we now turn. III. Conditions Facing Cross-Border Traders in the Great Lakes Region While cross-border trade is known to be important in the Great Lakes region, there is little information on the nature of this trade and the conditions under which it takes place. Cross-border trade is the key mechanism of supply of vital food products and other commodities to cities and towns on the borders of countries in the Great Lakes region such as of Goma, Bukavu. One of the few studies of such trade suggests that such trade directly and indirectly employs approximately 22,000 persons in the Goma/Gisenyi area while an estimated 500 to 1,000 cross-border traders are established in Bukavu, probably generating ten times as many derived jobs in the border region of South Kivu province.9 The monthly value or cross-border trader imports in Goma coming from Rwanda is estimated in excess of one million USD. Between 1,800 and 2,000 crossings are estimated to take place at Petite Barrière in Goma and 900 per day at the Ruzizi border post in Bukavu. These figures are high by any standard, and are likely to go up as the population of the corresponding cities increases – with a corresponding need for increased supplies of staple goods. To fill some of the large information gap on cross- border trade and trading conditions a survey of cross border traders was carried out in 2010 at 4 key border crossings in the region (details of which are summarized in Annex 2) Cross-border trade is dominated by women and provides an essential source of income to many households in the region. The survey identifies the following key features of cross-border trade: 1. The majority of traders are young women (85% of respondents were female and the average age of a trader is 32) and experienced traders, with 44% having been traders for more than five years. 2. Most of the officials who regulate the border are men (82%).10 9 International Alert report “La Traverséeâ€?, June 2010. 10 Official data from OFIDA for August 2010 shows that in North Kivu less than 20% of the 547 agents were women. 8 3. Most traders sell foodstuffs, particularly cereals, pulses, vegetables and fruits. The quantities involved are small and are typically carried by head. 4. Start-up capital is very small (less than $50) and is typically obtained from within the family. Very few traders have received loans from a financial institutional. The vast majority (95%) wishes to invest and grow their business but is constrained from doing so by the current environment at the border and lack of access to finance. 5. For almost two-thirds of the respondents, income from cross-border trade is the main source of income and most (77%) report that household income is heavily dependent on their trading activity. 6. Few are members of a traders association. Cross-border traders regularly have to pay bribes and suffer harassment. The responses from the detailed interviews and focus groups paint a dark picture of the conditions experienced by poor women cross border traders in the Great Lakes region. They face serious risks and losses each time they cross the border. Figure 3 summarizes the main findings. It is striking that payments of bribes is a regular occurrence for the majority of traders. Respondents at all four border posts repeated a catch phrase used by officials: “sans argent, on ne passe pasâ€? (no money, no passing). At the Goma-Gisenyi border 100% of respondents reported that they had to pay bribes to cross the border. A large number of traders also report having their goods confiscated and having to pay fines. Figure 3: Reported frequency of risks by cross border traders 9 An important feature of border crossings between the DRC and neighboring countries in the Great Lakes region is the large number and range of officials at the border. 11 This exacerbates the problem of poor governance with negative consequences for cross-border traders. This is compounded by a lack of transparency and awareness by both traders and officials of the rules and regulations that are supposed to govern cross-border movements of goods and people. A typical account of every day conditions is provided by an egg and sugar trader from Goma: “I buy my eggs in Rwanda; as soon as I cross to Congo I give one egg to every official who asks me. Some days I give away more than 30 eggs!â€? Reducing the number of agencies and officials at the border and increasing the transparency and predictability of the policy regime is crucial to provide an environment in which traders flourish and expand their business. A large number of traders report being subject to acts of violence, threats and sexual harassment. Traders are exposed to beatings, verbal insults, stripping, sexual harassment and even rape. Much of this abuse is unreported. While some borders are better organized, this gender-based violence, although more prevalent in the DRC, is being perpetrated on both sides of the border in Burundi, Rwanda and Uganda. Not only do officials harass traders but young unemployed men, called “maiboboâ€? (street youth and thugs) and “les viseursâ€? (watchers) are tolerated or even hired by state officials and given carte blanche to apply force as needed to extract money and goods from traders, particularly those that move by foot with goods strapped on their backs or carried by head. A typical occurrence is that women traders are often encircled by a group of men after they cross the border. Vulnerability to theft and physical abuse is rampant in these borderlands. Thus, cross-border traders face regular losses in the form of the almost mandatory payment of bribes and are regularly subject to harassment and physical abuse. This lack of economic and physical security and safety undermines the livelihoods of these traders and compounds their lack of access to finance, information and business knowledge. But it is clear that most traders perceive that under present conditions there is no alternative – they are currently stuck in a survival economy. Addressing these issues by providing security at the border and implementing a transparent and predictable regime for those crossing the border would facilitate trade, improve incomes, contribute to regional food security and in turn provide for greater stability in the region. 11 This is despite the decree No.036/2002, March 28, 2002 which stipulates that 4 agencies are legally entitled to be present at the border: Custom office (DGDA), Immigration office (DGM), Hygiene service (PNHF) and Quality control service (OCC). This was confirmed by the “Ordre Operationnelâ€? in December 2010 by the vice-prime minister in charge of security and interior. Nevertheless, the field survey for this report shows that the presence of officials from between 10 to 15 different agencies is not uncommon at the borders in the east of the DRC. 10 There is likely a vicious cycle of mistrust and disobedience between traders and officials at the border. A common perception of informal trade would be trade that seeks to avoid formal rules and institutions at the border. However, at many of the border crossings the formal institutions are at best very weak and the complexity and lack of transparency of rules and regulations governing cross-border trade provides a fertile environment for harassment and inappropriate behavior by officials and disobedience by traders. Hence, the source of informality at the border comes both from the traders and from the lack of formal institutions and from self-serving behavior by officials. The informality of trade can also be traced to the lack of recourse to formal institutions for traders to contest when they are subject to inappropriate or illegal actions by border officials. Informal traders, especially poor women food traders, lack representation and organization. As result, they are vulnerable to powerful officials who control the borders and are unable to contest the decisions and actions of officials. This is compounded by lack of knowledge and awareness of the relevant rules and regulations and the obligations with which traders must comply. For example, Congolese women taking products from Congo to sell in Burundi are often required to pay a tax by officials when they exit the country, yet there are no official export taxes. IV. Physical Conditions at the Border and Border Procedures The very poor quality and condition of the infrastructure at the border allows some of the worst types of harassment and violence. Most of the attention of the government and donors has been on improving conditions for formal trade and better infrastructure for trucks at the main border crossings. Informal cross-border traders have not been high on the agenda. The border crossings in Goma are a good example. Two crossings are regulated by customs: the “petite barrièreâ€?, for small informal traders, and the “grande barrièreâ€? for more formal traders. The differences between the two crossings are striking as is the stark contrast between the DRC side and the well-organized Rwandan side of the border. At the “grande barrièreâ€? the road is paved and buildings are in reasonable state and cross-border movements appear orderly. At the “petite barrièreâ€? (Goma side) it is chaos. Roads are very poor, and congested with vendors, and a number of storefronts tumble onto the busy roadside.12 Facilities for officials in the DRC are often in a bad condition. This is most apparent at the Petite Barrière in Goma where the different border agency huts/converted containers lack power and sanitation and are in the middle of a built-up area with large crowds of local inhabitants surrounding the border station. It is difficult to separate travelers from local 12 The situation has been transformed on the Rwandan side of the border where significant construction has delivered new customs and immigration buildings, improved roads and street lighting. 11 residents, and crowd control can be awkward. Adequate and well-designed facilities, which need not be lavish, contribute to a better control (and controlled) environment. For example, the chief Customs officer at Petite Barrière is in a back office with no view over the road and cross-border traffic and therefore cannot supervise the work of subordinate staff. On the other hand, the Chief Customs officer’s office at Ruzizi border station faces the access road, which would enable that officer to intervene immediately if necessary. A design that enables supervisors to see what is happening outside is an essential aspect of management. The role of Customs is insufficiently defined. The lack of consistent Customs intervention at the border facilitates smuggling, and generates additional corruption in other agencies. This has been further aggravated by the establishment of a special branch on the Congolese National Police, called the Border Police. Officially the role of that Border Police force is to (i) ensure law and order at border stations, and (ii) secure the borderline in the interval between two approved crossings13 with a view to curb illegal immigration and fight smuggling. However, conflicts with the mandate of Customs: according to the Customs code of February 20, 2011, Customs officials are allowed to carry weapons, for self-defense or to prevent smuggling (which is one of their major roles). Customs are therefore in a position to ensure their own security at border posts, or wherever else they operate. In addition, it appears that the Border Police at Petite Barrière are no longer on duty when the border closes at 6 PM and therefore do not serve any purpose in interdicting illegal crossings. It is a matter of concern that the establishment of this new force will likely result in an additional agency operating at the border, and reports from users suggest that Border Police officials are frequently involved in rent-seeking activities. Customs procedures are poorly applied. The normal way of assessing import duties is to calculate them on a Customs declaration where the importer describes the goods, places them against a tariff heading with a corresponding duty rate, and certifies the value. Understandably, the procedure can be simplified for cross border trade by individual traders, and the predictable nature of goods imported makes it easier to use generic classification and a flat duty rate, as well as a very simplified declaration format. It is often Customs officers who actually fill-in the form. This is broadly the approach used by Customs at the border stations in Kinshasa, Goma and Bukavu. A simplified import declaration (DSI) is filled by Customs, with a very general description of goods; Customs officials use a regularly updated reference price list, against which they check the value declared by the importers. However, this procedure has been seriously distorted: 13 This is occasionally known as the “greenâ€? border, 12 a. The DSI is only used as a weekly or monthly consolidated declaration. Therefore it does not itemize individual transactions, and is often not handed over to importers.14 b. The only trace of a specific transaction appears on the receipt which is supposed to be given to importers after payment. This receipt is not designed for assessment, and the way in which it is filled-in does not provide any record on the goods imported, or how the duties were calculated. c. As a result, there is no traceability, audit trail, or even way to verify what was imported, by whom, and how much was paid. There is a lack of transparency and of accountability of officials. It is difficult for Customs management to verify that field officials apply fair and consistent procedures to cross- border traders. This is made even more difficult by the fact that receipts are seldom issued to importers. A brief review of DSI and receipt books showed that on some occasions the counterfoil of DSIs remain attached to the book, and that originals of receipts were not detached from the receipt book; this implies that officials do not even bother to conceal the fact they are not issuing them to importers. Traders confirm that they are never issued with a receipt for any sort of payment they have to make. The overall impression is that there is a total lack of transparency, let alone visibility, in procedures and the way they are managed. While some managers acknowledge the fact, and explain it by weak training and absence of computerization15, others indicate that this situation is largely due to poor compliance, thus necessitating additional controls. However, simple observation and anecdotal evidence shows that there is an organized approach to ransoming of traders. There are too many agencies at the border with significant overlaps between responsibilities. Up to 19 (and possibly more) official or claiming to be official agencies operate at most border points. The tasks undertaken by these agencies are often superfluous or can be easily undertaken by the main border agencies. The nature of processing does not warrant specialized expertise in most cases: for example, the sanitary check entails verifying the international vaccination certificate, which any Immigration or Customs officer can do and does not require a separate agency at the border. The only justification for a specific sanitary control would be in the event of an epidemic, but the “visualâ€? health control which the Programme National d’Hygiène aux Frontières officials claim to be performing is not credible. The Office Congolais de Contrôle (OCC) performs quality and safety checks on some goods, but collects a fee on every import. However, there is no reconciliation between OCC checks and those performed by Customs, which could help in ensuring that everything imported is subject to Customs control. OCC and Customs also use two different valuation data bases to assess ad 14 An explanation provided was that the DSI form is sold for USD 8. Another option would be not to charge for that document. 15 This is not entirely true, as many border agencies in the world were honest long before computers reached them. 13 valorem duties. A multiplicity of agencies therefore is justified only if different agency processes are required for different categories of goods and persons but not if all are subject to the same treatment. Other countries have adopted a wide system of delegation, where specialized officials are only on standby should an unusual case occur and ordinary processes are checked by one official on behalf of all the other administrations. V. Steps to Facilitate Cross-Border Trade in the Great Lakes Region The immediate priority must be to improve the conditions at the border and the treatment of cross-borders traders. This will have a significant impact on the livelihoods of a substantial number of women and the households who depend on their income from trading. It will also lead to more cross-border trade and more widely available and cheaper food products. The second imperative is to provide and implement a policy framework by which these traders can become increasingly organized backed up by support services that improve access to information, facilitate access to credit and ensure better representation of traders’ interests. This will provide a route by which these activities can gradually become more formalized and distribution channels become more efficient. This will require: ï‚· Officials at all levels recognize the important role played by poor cross-border traders. Poor women cross-border traders are currently viewed negatively by policy makers reflecting the misconception that they are an important source of revenue loss. This is turn means that there has been little effort to improve the condition of the infrastructure at the border crossings they use and a lack of effort to bring discipline and order to the officials who regulate these borders. The vast majority of these traders are entrepreneurs who would like to invest and grow their business and emerge from their current survival mode if stability and security could be provided and policies at the border were transparent and applied in a predictable and fair manner. These traders therefore offer an opportunity rather than a constraint for growth and poverty reduction. ï‚· There has however, been significant progress on this issue at both the regional and the national level. Seminars have recently taken place in Goma and Bukavu which have brought together representatives of the regional government, officials from border agencies and traders. The Governor of South Kivu recently visited the border at Bukavu. In recent interviews, traders at Bukavu attest that this heightened interest from policy makers has been reflected in less harassment.16 In Kinshasa, the issues of cross-border trade and the conditions facing poor women traders were discussed at a workshop in April 2011. The workshop was opened on behalf of the Deputy Prime Minister and attended by the Governors of North and South Kivu as well as the Deputy Governor of South Kivu. Some 40 16 From transcript of interviews with a group of traders in Bukavu in March 2011. 14 officials including representatives of the agencies engaged in trade facilitation issues were represented. ï‚· A comprehensive border management policy must be established and implemented. This is an important pre-requisite. It should describe what is expected to be done at the border, by whom and how it will be done. Therefore, a detailed mapping of all processes should be carried out, and each step analyzed in terms of cost/effectiveness, and ways of streamlining it. This should then assemble into a logical framework for border operations. A first step in this regard has been achieved at the April workshop in Kinshasa where the agencies agreed a detailed and sequenced roadmap towards improving border procedures at two pilot crossings in the east (Goma and Bukuva). Implementation of the roadmap will streamline processes and achieve clear agreement between the different agencies as to who should do what, with the preparation of clear job descriptions. The border management policy must achieve the following: (i) There must be a reduction in the number of agencies at the border. Decree No.036/2002, March 28, 2002 clearly designated 4 service agencies at the border: Customs (DGDA), Immigration (DGM), Hygiene service (PNHF) and Quality control of goods (OCC). The government of the DRC through the Vice Prime Minister for security and interior has subsequently issued an operational order in December 2010 re-stating that only four agencies should be present at the border, taken in response to what was described as “persisting anarchy, disorder, and harassmentâ€? due to the multiplicity of State agencies at borders. It appears to have been largely ignored. While there has been some success in implementation, at many borders more than 10 different agencies remain.17 The main issue at this stage is how to obtain proper (and sustained) enforcement of the Operational Order through effective monitoring and sanctions when the order is disobeyed. (ii) An increase in the transparency and predictability of the trade policy regime. There is clearly a high degree of uncertainty and ambiguity about the legal rules that apply to cross-border trade and how they are applied in practice. Lack of transparency and consistency in the application of trade rules constrain trade across the borders. The application of simple and fair procedures and reasonable fees would support greater activity and a transition to more organized and formal modes of exchange. The results of the survey suggest a number of practical steps that could immediately be made: (i) all 17 For example, officials from the Ministry of Agriculture were witnessed at the Ruzizi 1 border post, thus contravening the Order. The local chief of agriculture produced a letter signed by the Ministry of Telecommunications on the letterhead paper of the Vice-Prime Minister stating that the Ministry, recognising the usefulness of agriculture at the border, fully recommended its presence. This letter was considered as sufficient authority for Ministry of Agriculture officials to operate and collect fees at the border. 15 payments made to an official should be recorded and an official receipt provided; (ii) payments should be made only at a single, clearly designated building; (iii) there should be no payments in kind to officials. Efforts are being made to simplify border procedures. The Government has convened a working group to prepare a simplified tariff for imports by small traders. It is currently being derived from Comesa guidelines, and the working group is expected to meet before the end of March 2011. However, the Comesa simplified trade regime is not being widely applied elsewhere and is not seen as a great success where it has been applied at the border between Zambia and Zimbabwe and Zambia and Malawi. The process has many bureaucratic requirements, which may be beyond the reach of the average border trader. For example, there is need for a certificate of origin, which for many small traders who are arbitraging basic food products across borders is unnecessary but costly and time consuming. Hence, there is a need for a more cooperative approach to simplifying trade procedures involving the countries of the Great lakes region and a more inclusive process that captures the views and potential impacts on small cross-border traders. For example, the procedures must reflect that many of the women traders have limited education. (iii) Improvements to infrastructure at border crossings. Both the transport infrastructure and the poor quality and inappropriate design of buildings used by officials need to be rehabilitated, including the provision of basic facilities for officials such as electricity and water. Work is immediately required to minimize the risks to safety and security that arise from the dilapidated infrastructure and to provide an open and transparent area for officials and traders. This should comprise providing floodlighting to the border crossing area, removing structures that provide opportunities for harassment and physical violence and providing a mechanism for senior officials to monitor the border crossing, for example, through the provision of closed circuit TV cameras. The border management strategy should provide for an assessment of the specific needs at each border station in terms of types and volumes of traffic, quantities of goods, local risks and potential for fraud or incidents, and categories of border commuters (e.g., women, disabled persons, nationals v/ foreigners, etc.). This assessment would include features such as parking space, holding areas, safe-keeping facilities, and specialized offices for control or search. A set of infrastructure modules designed by function (e.g., channeling, access control, examination) would then be defined. These modules would only be deployed where they are necessary and an inter- agency protocol for sharing assets would be drafted. Some cross-border consultation should take place when preparing the station design. 16 ï‚· Increased professionalism of officials and greater gender awareness. Women are under- represented in official bodies at the border and so increasing the number of female border officials is an important step. Nevertheless, lessons show that what is needed are gender- aware officials who appreciate the issues confronting traders, poor women and men alike. Information from the focus groups suggests that harassment may often come from women officials. More women at the borders may reduce sexual crimes, but that is not guaranteed if they do not have power. Lower-level female officials cannot stop higher placed males and their associates from carrying out gender-based violence. Obvious policies such as only female officers should be allowed to body search women traders should be made mandatory and widely publicized. There is therefore a need for enhanced training of officials. The purpose would be (i) to achieve a good understanding of the mechanics of border management and interaction between agencies. Inter-agency awareness should also be built into the training modules, with a view to promote (legitimate) cooperation between border officials. Regular information conferences should be held for the entire staff at the border, and possibly including border staff from the adjacent country.18 Additional specialized training needs would need to be identified, such as, for example, some basic training in Customs on respecting procedures and book-keeping. At the same time there is a need for (ii) serious training on human rights (and penalties for breaching them) and workshops to raise awareness of gender issues, especially at the border crossings used by informal traders. This can be supported by video’s that show the economic realities of poor women traders, the contribution of their trading activities to household income and the conditions and physical violence they encounter in their daily lives. A further approach could be role-play, if only to make officials aware of what some travelers may endure, which would also provide training in dissipating tensions (such as when a difficult control is taking place). A key issue is to ensure that all officials are either in uniform or wear constantly a visible badge (the latter being a cheaper alternative and an interim measure until proper uniforms have been issued). International experience shows that staff in reasonable uniforms, who are subject to quasi-military discipline, behave better. This was noticeable during border visits in the Eastern provinces, where DGM, Border Police, and to a large extent Customs uniformed personnel, showed more respect for their hierarchy (and consequently for the public) than other, non-uniformed, agencies. 18 A possibility would be to appoint, on a case-by-case basis, a local border coordinator, on a merit base and drawn from one of the approved agencies. It does not appear necessary to appoint one agency as the sole coordinator for all borders, as situations may differ significantly from one location to another. 17 This program of training and enhanced professionalism could provide a base for the introduction of performance assessments of officials in which treatment of cross-border traders is a key indicator. It is important to educate and not to shame the officials to change behavior, to define a clear code of practice and reward those officials who behave in manner consist with higher standards. Experience from elsewhere, such as customs reform in Senegal, suggests that non-financial rewards, such as widespread recognition within the institution and the wider community through employee of the month schemes, with such awards being kept on the employees file and being a factor in promotion decisions and eligibility for enhanced training can be as effective as direct financial rewards.19 This would need to be supported but an independent office of complaints which can record and act upon complaints when violations occur. More generally, as in all post-conflict countries there is a need to address the fundamental issues of governance in public administration and implement effective security sector reform (police, justice and military). The conditions that traders face at the border are in part a consequence of poor wages and often lack of payment and a culture of impunity due to the lack of functioning legal and justice system. ï‚· There is a need to strengthen cross-border coordination and cooperation. Key border agencies coordinate insufficiently with Burundi, Rwanda and Uganda. For example, while DGM maintains adequate contacts on the Eastern borders with their counterparts in Rwanda, there is very limited involvement by Customs. Normally, regular contacts between Customs authorities of both countries is a key element of successful border management, as it enables the establishment of close ties between agencies that do the same job along the same procedures, albeit from a different perspective. They help in collecting intelligence, detecting discrepancies, and preventing gross manipulations in reporting goods, but also in running daily operations.20 These ties are less formal than Mutual Assistance Agreements which are typically signed at the country-level, but provide for informal data exchange, often on an ad hoc basis. Border meetings normally take place on a monthly or even weekly basis. At Goma, there are apparently no such contacts between Customs. Although the Rwanda Revenue Authority (RRA) station manager at Petite Barrière visited his DRC counterpart, there has been no return visit, and Congolese officials say they do not have the time for it. Similar attempts to promote contacts by Ugandan customs with the DRC counterparts have not been successful Necessary country coordination is absent. Although it may seem anecdotal, the fact that on the Rwanda-DRC border at Goma Rwanda built a major facility for heavy traffic on one side while the government of DRC kept the same facility dedicated to pedestrian traffic, with no 19 See Cantens et al (2011). 20 For example, surges in traffic can be anticipated. 18 road worth mentioning connecting to the Congolese highways, it is revealing of some international planning deficiencies. The Rwanda facility cannot be used for the purpose for which it was designed, and it will take a while before the DRC side is ready. This appears to be a waste of resources, at least for the time being. Yet international cooperation gathers interest. This is particularly the case for the establishment of shared border facilities (see Box 1 for a brief discussion of the pros and cons of sharing border facilities). Officials of RRA and the Rwanda Immigration Service very much support the concept and have clearly indicated their willingness to share whenever possible their infrastructure with DRC officials. Cross-border traders too consider this would be a major step forward, in particular because DRC officials would work in contact with the Rwandese counterparts, and would be less inclined to indulge in corrupt practices on foreign territory. While provincial governors also favor the idea, the continuing legacy of the conflict means that decision at the highest level of government essential. ï‚· Increase training and support for traders through traders associations. There is a clear need to increase the representation of traders through traders associations to ensure that their interests are properly represented in public policy. An important development would be for more regular meetings between traders associations and border officials, moderated by an independent facilitator and with publication of minutes. At present meetings are ad hoc and typically centre on officials making traders aware of their responsibilities and they do not deal with the major problems and principles such that they are deemed to be rather ineffective. Associations of poor cross-border traders could be encouraged by reducing the cost of registering an association from $200 plus the cost of the legal services that are required to process the application to $50 with an expedited legal procedure. For associations of the poorest traders the current fees and costs are prohibitive. Officially registered associations will have more power to intervene at the border to support their members. The associations could be linked to sister organizations in other countries that have achieved progress in simplifying procedures for small cross-border traders and in improving the conditions faced by women traders. A further possibility is for the associations to benefit from the advice of a trained and experienced customs clearing agent who would provide training to the members on border procedures and their legal obligations, provide guidance and directly intervene with officials in some cases. 19 Box 1: The case for co-location of border facilities The original model adopted in Europe sixty years ago was simple and based on juxtaposition, with a Memorandum of Understanding between the adjacent administrations to deal with the simple extra- territorial matters that might arise (arrest, denial of entry, seizure). More difficult cases were to be deferred to a court decision, and/or international arbitration, but they were very few of these and they never prevented effective operations. Divergence in organisation, or even legal systems between the two countries never posed a major problem. On DRC borders, the benefits of a juxtaposed system would be the following: ï‚· It would eliminate the “parasiteâ€? agencies and individuals, as they would not be admitted into the adjacent country (when these facilities would be located on DRC’s neighbours’ territory), or even into a facility placed on Congolese territory (because such a new station would be designed in a way to prevent unauthorized access). ï‚· As a rule, officials behave better on foreign territory. This can be encouraged by a (moderate) “expatriation bonusâ€?, paid every time an official is on duty abroad. ï‚· Some of the major cross-border frauds (re-consolidation or splitting of consignments, switch in documents or number plates, etc.) which take place in the interval between the two border posts are de facto eliminated because both countries operate on the same platform. ï‚· It is easier to align to the adjacent country’s best practice. ï‚· Officials learn to work together, and this can be a useful prelude to more integration. A case against juxtaposed facilities is occasionally made on the following grounds: The country is not ready yet. This is not a valid argument, as all that is necessary is to assign a skeleton shift on the other side. There is no need for a full-fledged Customs house on foreign territory. In that sense, the juxtaposition essentially means pre-clearance, with more serious commercial clearance taking place in the country of destination. There is no computer system. This does not matter in the short term (most European countries had juxtaposed border facilities long before they introduced computer systems), and a local arrangement can provide for the use by one country of data declared in the other. Countries do not belong to the same economic group. Neither do Switzerland and Norway belong to the EU, which does not prevent Norwegian Customs to clear Swedish exports on behalf of Swedish Customs, or Swiss Customs to have shared facilities with all their EU neighbours. Political differences. Again in Europe these did not affect the operation of shared facilities between democratic countries and others that were less so. Co-location is simply a convenient way of performing border controls, and does not entail any qualitative judgment on political systems. Why should all facilities be on foreign territory? This is not the case, and different models can be envisaged. Given the current state of infrastructure in the DRC, it is suggested that a first pilot could be tried on Rwandese territory. The next pilot would then be on Congolese territory. That way, DRC would pay only for its portion of facilities on DRC territory. Although not the panacea, juxtaposition or co-location can provide a spectacular outcome, which can then be leveraged to achieve further streamlining. ï‚· Mobile technology could provide an innovative short term solution reducing harassments and could assist in increasing information on market prices. The survey suggests that more than half of traders have mobile phones and discussions in the focus groups show they are 20 able to send text messages using mobile technology, even though the illiteracy rate is high. It is possible then that a standard text message could be composed that could be sent to a hotline when a trader is being harassed and used to support an effective complaints mechanism. The use of mobile technology could also be used to provide more and better information on market prices to traders on both sides of the border. An opportunity arises from a Rwanda ICT project e-soko. This is part of the eRwanda Project, a World Bank funded project being implemented by the Rwanda Information Technology Authority (RITA) which intends to use mobile and other phone technologies to provide information on agricultural products from the Agricultural Market Pricing Information System. The possibility of integrating this project with a similar project in the east of the DRC could also be explored. ï‚· A strategy for integrating poor entrepreneurs into the formal economy. It is also necessary to start to address key concerns of these traders regarding their lack of access to finance, lack of information on prices and business opportunities and training in basic business practices. It would also be interesting to further investigate the potential beneficial impact of providing warehousing facilities close to the border and the development of cross-border markets building on the knowledge of the success of bazaars, for example, in central Asia (World Bank (2010)). Cross border markets could be an effective way to bring together traders from either side of the border in a more efficient way. VI. Conclusions and moving forward Trade and increasing integration with EAC neighbors has an important role to play in raising incomes and underpinning stability in the east of the DRC. At present trade is a critical mechanism to improve access to food and basic manufactured products for many households in border cities and towns in the east. But there is considerable agricultural potential in the east of the DRC that can be encouraged by linking farmers to large and growing markets in the EAC countries. At the same time providing greater security in the east of the DRC will create increasing opportunities for trade and economic activities and creating further stability. Continuous dialogue for regional cooperation needs to be strengthened through regional forums such as ICGLR (International Conference on the Great Lakes Region) and CPEGL. The borders in the east of the DRC are very thick. This reflects the continuing uncertainty over the security situation but also the difficulties, physical harassment and financial costs that traders face when crossing the border into and out of the DRC. A range of issues need to be addressed at the border to improve border crossing conditions including the application of clear and simple rules, increased professionalism of officials and zero tolerance of harassment and violence, better infrastructure and physical conditions for officials and traders at the border 21 and support to traders associations for better representation of the interests of poor small traders. Facilitating trade would contribute to food security but would also support the incomes of traders, most of whom are poor women, and the welfare of the households who depend on their trading activities. While the value of trade carried individually by these poor traders is very small, the contribution of their trading activities to the income of their household, many of which they head, is substantial. Given the large number of people crossing the border every day, the income earned from these activities is therefore a key element of the local economy and poverty reduction in communities around the border. It is very clear that the vast majority of these women traders wish to grow their business and escape from the survival economy in which they are currently confined. Initiatives are being implemented to improve conditions at the border. The government has validated the roadmap for reform of border procedures at the pilot border posts. This provides a prioritized set of interventions to improve border management procedures. The key now is to support and monitor implementation. The World Bank has commenced a small pilot project, financed by the Trade Facilitation Facility, with a prioritized set of interventions to improve physical conditions, provide training to officials and support traders associations with a focus on the Petite Barriere at Goma and Ruzizi in Bukavu. This project will fully involve the local provincial pilot committee in order to ensure sustainability of actions and monitor continuously progress of the project. This pilot needs to be scaled up at other border crossings throughout the region. Organizations such as International Alert are also supporting small traders. The Bank has organized the production of a video to highlight the conditions faced at the border and the importance of cross-border trade to the incomes of poor women traders and the welfare of their households. The video will contribute to the Bank’s advocacy on this issue and will be used as part of the training of officials to increase awareness of the situation of poor cross-border traders. Opportunities for cross-border coordination of policies and efficient use of facilities at the border are not being exploited. Finding an appropriate forum for the dialogue which can support coordination and cooperation is important. A COMESA based forum is the most obvious opportunity at present. Through the Trading for Peace program, COMESA has provided technical assistance in the form of COMESA trade information points which could provide valuable support to the coordination process, including at the border post in Goma. Both COMESA and the EAC are developing a simplified trade regime to improve conditions for small scale trade. The initial trials with the COMESA scheme suggest, however, that is not achieving the main objectives and is being reassessed. A process that revamps the COMEA scheme and harmonizes it with that of the EAC under the Tripartite process could be used to bring together officials and 22 traders from Burundi, DRC, Rwanda and Uganda (all of whom are COMESA members) to design simple documents and procedures for the quick and effective border processing of small traders. Further, if politicians in the DRC are bold, then implementing the COMESA free trade agreement would be a positive step forward towards deeper economic integration in Africa. This could be coordinated with the Tripartite process which is linking the EAC, COMESA and SADC into a Grand Free Trade Area. 23 VII. References Brenton, Paul and Alberto Portugal (2011) Road Infrastructure, Border Effects, and Agricultural Prices in the Great Lakes Region, background paper. Cantens, T., G. Raballand, N. Strychacz, and T. Tchouawou (2011) ‘Reforming African Customs: The Results of the Cameroonian Performance Contract Pilot’, Africa Trade Policy Note No. 13, World bank Washington, http://siteresources.worldbank.org/INTAFRREGTOPTRADE/Resources/13ReformingAfrica nCustomsREDESIGN2.pdf. Coulibaly, Kalamogo and Tembo Maburuki (2006) ‘Cross-border trade in Rwanda’, mimeo, World Bank. Engel, C., Rogers, J., 1996. How wide is the border? American Economic Review 86, 1112–1125. December. Hossein, Caroline Shenaz, Jean-Baptiste Ntagoma and Celestin Bashinge Bucekuderhwa (2010) ‘Risky Business? Poor Women Food Traders in the Borderlands of Congo, Uganda, Rwanda and Burundi’, background paper prepared for this report. International Alert (2010) “La Traversée: Petit commerce et amélioration des relations transfrontalières entre Goma (RD Congo) and Gisenyi (Rwanda),â€? report, June 2010. Integrated Food Security Phase Classification (IPC), July 2009 Masinjila, Masheti (2009) ‘Gender Dimensions of Cross-Border Trade in the East African Community – Kenya/Uganda and Rwanda/Burundi Border’, African Trade Policy Centre, Work in Progress No. 78. Mthembu-Salter, Gregory (undated) ‘Evaluating intra-African trade: The case of Cyangugu- Bukavu’, mimeo Seidman, Irving. 2006. Interviewing as Qualitative Research: A Guide for Researchers in Education and the Social Sciences. New York: Teachers College Press. Ulimwengu, John & Funes, Jose & Headey, Derek D. & You, Liang (2009). "Paving the Way for Development: The Impact of Road Infrastructure on Agricultural Production and Household Wealth in the Democratic Republic of Congo," 2009 Annual Meeting, July 26- 28, 2009, Milwaukee, Wisconsin 49292, Agricultural and Applied Economics Association. UNHCR (2010) Internal Displaced People Fact Sheet, Democratic Republic of Congo, May 2010 WFP (2009) Enquête sur les conditions socio-économiques des déplacés et population hôtes dans le Sud Kivu, WFP Sud Kivu World Bank (2010) Skeins of Silk: Borderless Bazaars and Border Trade in Central Asia, Washington DC SIDA (2009) Country Gender Profile, The Democratic Republic of Congo 24 Annex 1: Assessing the Impact of Roads and Borders on Agricultural Prices in the Great Lakes Region. To motivate investments and policy reforms that facilitate trade it is often useful to provide to policy makers with an assessment of the potential flows of cross-border trade if these barriers were to be removed. Commonly, the depth of market integration in a region is assessed by the volume of intra-regional trade. However, a substantial volume of intra-regional trade in the Great Lakes region is not recorded by customs offices, making such an assessment difficult. This is because a large amount of trade takes place informally. While there is a significant amount of illegal trade in minerals and timber, most of the activities involving agricultural products and household items are not illegal, the traders are typically very poor people passing through official borders but the amounts are not recorded. An alternative method to assess the extent of market integration consists of evaluating differences in prices of homogeneous products between markets on different sides of a border and the degree of co-evolution of those prices. In well-integrated and well-functioning markets, price differences between markets should be arbitraged away conditioned by the costs of transporting goods between markets. When the markets lie on either side of a border then arbitrage involves intraregional trade. High transport costs, border barriers, and imperfect markets for transport and other relevant services as well as for final goods limit the extent of movement between markets, resulting in large price dispersion and leading to potentially high welfare costs. In this analysis, we attempt to measure the impact of two key factors: roads and border effects, on the relative price and its volatility for a range of agricultural commodities between markets in Burundi, DR Congo and Rwanda.21 More specifically, we use a database collected by FAO on prices of agricultural goods in 32 towns in Rwanda, 15 towns in Burundi and 22 towns in DRC. Figures A1.1 and A2.2 show the markets and the road conditions in Burundi and Rwanda and the DRC respectively. Prices are monthly from May 2008 to September 2009. Because of data availability, we restrict our analysis to seven commodities that are produced, consumed and traded in the region: beef, fish, maize, maize flour, beans, palm oil, and cassava flour. We match commodity price data from these markets with a rich dataset on the road network in Burundi, Rwanda, and DRC that was compiled from GIS. This includes the length of the road separating any pair of towns in these countries, as well as a measure of road quality for each segment which is used to 21 The background paper by Brenton and Portugal (2011) provides a more detailed account of the approach and lists the towns/markets in the sample. 25 obtain the average number of hours required to travel from a town to any other town in the region. Estimates are carried out for both differences in relative prices and the volatility of prices. Figure A1 1: Markets and Road Conditions in Burundi and Rwanda Figure A1 2: Markets and Road Conditions in the DRC Clearly markets in the DRC are much more dispersed and this is taken into account in the statistical analysis. Nevertheless, we do implement a number of tests to assess the robustness of 26 the results. We find, for example, that the results are not sensitive to excluding markets that are more than 500km from the eastern borders of the DRC. Two types of estimates are carried out: i) Relative prices (in absolute value) The first estimates seek to look at the impact of road infrastructure and border effects on the level of relative prices of a commodity in two towns.  ln Pikt / Pjkt   β0  β1 ln RoadDistij  ï?¢ 2 BorderRWA  DRC , ij  ï?¢ 3 BorderRWA  BDI ,ij  ï?¢ 4 BorderDRC  BDI ,ij  δi , j  δk  δt  εikt (1) Where: Pikt is the price of commodity k in town i and month t expressed in dollars, RoadDistij is the road distance between town i and town j. BorderRWADRC is a dummy that equals one if one of the towns is in Rwanda and the other town in DRC and is zero elsewhere; the same definition applies to BorderRWA BDI ,ij and BorderDRC  BDI ,ij . Finally δi , j is a vector of dummies controlling for town-level fixed effects such as market size, δ k and δt control for fixed effects specific to commodities and months. The coefficient of road distance is expected to be positive, as prices of a commodity in more distant towns are expected to be more dispersed. ii) Volatility of relative prices. Analogously, the second set of estimates explains the effect of roads and borders- effects on the volatility of relative prices between two towns. ï?›  std.dev. ln Pik / Pjk ï??  β0  β1 ln RoadDistij  ï?¢ 2 BorderRWA  DRC , ij  ï?¢ 3 BorderRWA  BDI , ij  ï?¢ 4 BorderDRC  BDI , ij  δi , j  δk  εik (2) Again, longer distance between two towns should increase the price variability, whereas the border effect of Rwanda-Burundi should be smaller than the other two border effects, or even negative. Notice that estimates are in cross-section, whereas estimates on relative prices were in panel. 27 Results. Table A1 reports estimates of regressions with relative prices, with column 1 containing estimates of equation (1). All signs are as expected. Road distance has indeed a positive and significant effect on relative prices. The average inter-town distance in our sample being roughly 550 km, reducing a km would reduce relative prices by 0.016 % (=0.082/550). The border dummies are all positive and significant, although different in magnitude. The Burundi- Rwanda border effect (0.026) on relative prices is equivalent to additional 174 km (=550/0.082*0.272), whereas the Burundi-DRC and DRC-Rwanda border effects are respectively equivalent to 1824 km and 1549 km, respectively. Column 2 includes a single border dummy, instead of the three border dummies, and the coefficient of road distances does not change substantially. As Rwanda has more cities in the data (32) than Burundi (15) and DRC (22), we test the robustness of our results to reducing the number of Rwandan towns. Column 3 reports estimates of the baseline model on a reduced sample that keeps only 20 towns for Rwanda, and results do not change significantly. The detailed data on the road network allows estimating a measure of the average travel time between two towns in the sample. We replace road distance with a measure of the average number of traveling hours between two towns (column 4). Once again, estimates do not vary significantly. Estimates of the differential border and road distance effects across products are reported in Table A2. The coefficient of road distance is positive and significant across products, whereas border dummies are not significant for fish. Regarding border effects, the coefficient of the Rwanda-Burundi dummy is negative and significant for beef and haricot. Table A3 reports estimates on the volatility of relative prices. Coefficient estimates for equation 2 are reported in column 1 and column 2 with the full sample and with a reduced sample excluding some towns in Rwanda. In specifications 3 and 4, road distance is replaced by travelling time. The coefficient of road distance is significant and roughly equal to 0.0011 in all estimates. An additional kilometer of would increase price volatility by 0.00002, which is a much larger effect than the border effect for the US and Canada (0.00000057) found by Engel and Rogers (1996). The coefficients of dummies are positive except for the Burundi-Rwanda border, which again is evidence of a customs union effect. Finally, Table A4 reports estimates of the road and border effects on the volatility of relative prices across goods. 28 Table A 1: Benchmark Regressions (Relative Prices Dependent variable: ln Pikt / Pjkt   (1) (2) (3) (4) ln Road distance (km) 0.082*** 0.102*** 0.094*** [0.002] [0.002] [0.002] Ln Road distance (hrs) 0.081*** [0.001] border 0.175*** [0.002] border:BDI-DRC 0.272*** 0.260*** 0.245*** [0.003] [0.004] [0.004] border:BDI-RWA 0.026*** 0.016*** 0.027*** [0.003] [0.003] [0.003] Border:DRC-RWA 0.231*** 0.198*** 0.209*** [0.003] [0.003] [0.003] Observations 199512 199512 144948 199512 R-squared 0.37 0.35 0.34 0.37 Sample all all RWA select all Standard errors are clustered by town-pairs in brackets. * significant at 10%; ** significant at 5%; *** significant at 1%. All regressions control for town_i, town_j, product, and month-year, effects. Table A 2: Relative Prices by Product Dependant variable: ln Pit / Pjt .   (1) (2) (3) (4) (5) (6) (7) beef farinemais farinemanioc haricot huilepalme maisgraine poissonfrais lnRoad distance (km) 0.097*** 0.046*** 0.081*** 0.140*** 0.061*** 0.048*** 0.021** [0.009] [0.009] [0.010] [0.017] [0.007] [0.013] [0.009] border:BDI-DRC 0.278*** 0.285*** 0.433*** 0.437*** 0.014 0.195*** 0.046 [0.015] [0.016] [0.015] [0.024] [0.011] [0.018] [0.038] - border:BDI-RWA -0.078*** 0.313*** 0.025** 0.083*** 0.074*** 0.037*** -0.062 [0.011] [0.009] [0.010] [0.013] [0.010] [0.012] [0.052] border:DRC-RWA 0.516*** 0.024** 0.250*** 0.462*** 0.241*** 0.083*** 0.130*** [0.015] [0.010] [0.013] [0.022] [0.013] [0.015] [0.019] Observations 18052 37172 39438 39438 18185 36163 11064 R-squared 0.59 0.46 0.62 0.7 0.37 0.5 0.29 Standard errors are clustered by town-pairs in brackets * significant at 10%; ** significant at 5%; *** significant at 1%. All regressions control for town_i, town_j, and month-year, effects. 29 Table A 3: Benchmark Regressions (Volatility of Relative Prices) (1) (2) (3) (4) std_lratio std_lratio std_lratio std_lratio Ln Road distance (km) 0.011*** 0.012*** [0.002] [0.002] lnRoad distance (hrs) 0.011*** 0.011*** [0.002] [0.002] border:BDI-DRC 0.024*** 0.023*** 0.021*** 0.020*** [0.005] [0.006] [0.006] [0.006] border:BDI-RWA -0.016*** -0.022*** -0.017*** -0.021*** [0.004] [0.005] [0.004] [0.005] border:DRC-RWA 0.008** 0.004 0.004 0.001 [0.003] [0.004] [0.004] [0.004] Observations 15993 11021 15993 11021 R-squared 0.47 0.45 0.47 0.45 Sample All Select RWA All Select RWA Standard errors are clustered by town-pairs in brackets. * significant at 10%; ** significant at 5%; *** significant at 1%. Dependent variable is ï?› std.dev. ln P / Pjk  ik ï??. All regressions control for town_i, town_j, and product effects. Table A 4: Volatility of Relative Prices by Product Table A4 – Volatility of Relative Prices by Product (1) (2) (3) (4) (5) (6) (7) beef farinemais farinemanioc haricot huilepalme maisgraine poissonfrais Ln Road distance (km) 0.001 0.016*** 0.028*** -0.001 -0.001 0.011*** 0.013** [0.002] [0.004] [0.004] [0.003] [0.002] [0.003] [0.005] border:BDI- DRC 0.042*** 0.021*** 0.051*** 0.015*** 0.031*** 0.014* -0.065 [0.006] [0.007] [0.006] [0.005] [0.006] [0.008] [0.043] border:BDI- RWA -0.011** 0.015*** -0.009** 0.035*** -0.018*** -0.007 -0.362*** [0.005] [0.005] [0.004] [0.003] [0.006] [0.006] [0.045] border:DRC- RWA -0.034*** 0.050*** 0.089*** 0.054*** -0.054*** 0.038*** -0.102*** [0.004] [0.005] [0.006] [0.005] [0.003] [0.006] [0.009] Observations 2278 2415 2415 2415 2278 2414 1778 R-squared 0.75 0.86 0.9 0.89 0.8 0.81 0.68 Standard errors are clustered by town-pairs in brackets. * significant at 10%; ** significant at 5%; *** significant at 1%. Dependent variable is ï?› std.dev. ln P / Pjk  ik ï??. All regressions control for town_i and town_j effects. 30 Annex 2: The survey of cross-border traders A survey of cross-border traders and officials/other stakeholders was undertaken between mid- July and mid-September 2010 at 4 key border crossings in the Great Lakes region: at Uvira- Bujumbura (between DRC and Burundi); Bukavu-Cyangugu (between DRC and Rwanda); Goma- Gisenyi (DRC and Rwanda) and Kasindi-Mpondwe (DRC and Uganda), see Figure A2.1. A research team (ten persons in total) composed of both an international consultant and two lead local professors from the Catholic University of Bukavu (CUB) based in Bukavu, DR Congo as well as a team of local surveyors, mostly female students from CUB and local facilitators at each border check point.22 To be most effective, the approach was ethnographic; in that researchers had to observe and watch the communities before interviewing (Maxwell 2005). The team visited the borders to observe the border and see the many agencies operating at the border. Individual interviews were often conducted at people’s homes, or offices or in the markets. Visits to the nearby markets and residential slum communities near the border were helpful into getting a glimpse into the social and economic conditions of the cross border traders. Figure A2. 1: Key Border Crossings for the Survey 22 In Kasindi-Mpondwe border point, the local team was unable to carry out research in Uganda. They interviewed Ugandan elites and held focus groups with Ugandan traders on the DR Congo side of the border. 31 Data collection included interviews with two groups: traders and stakeholders (mainly officials). A total of 181 traders were interviewed, of which 100 had in-depth interviews and 81 persons participated in eight focus groups. A total of 58 officials/other stakeholders were interviewed, of which 54 completed structured interviews and four unstructured interviews at each location except in Cyangugu. The stakeholders were senior and lower level officials in customs, police, immigration, military and intelligence as well as a few from local non-profit organizations. The focus groups with traders had about ten main questions mapped onto key questions from the individual interviews. Each group had no more than eleven people and on average lasted one and half hours at a neutral location except in Gatumba where the session ran closer to two hours. Given the sensitivity of questions, mostly related to harassment by officials and gender issues, interview tools were standardized as much as possible but adapted to fit with the contextual realities. Interviews were designed so that people could tell a story and, at times, involved a dialogue so that trust could develop –but the focus in this project was to listen to people (Seidman 2006, 85-87). After each tool was piloted in July 2010, the interview tool for traders was modified to better respond to country context several times. There was an effort to keep the questions similar so that issues could be compared across the four border points, as appropriate. All individual interviews and FGs with traders were held at a neutral place with easy access for the subject. The pairing up of an international consultant with local consultants was important to this project. One clear limitation to this project was that the two main local consultants were from the DR Congo. During the election period in Rwanda (August 2010), the Congolese lead researchers were unable to complete interviews with Rwandan officials. To remedy this issue, local facilitators and surveyors from each border point were hired to work on this project to mitigate biases as much as possible. Moreover, the gender balance on the research team (six women, four men) ensured that there was a gender lens in the work. Full details of the survey, the questionnaire, the stakeholder and focus group questions are provided in the background paper by Hossein et al (2010). 32