Gmos

Impediments to Intra-regional trade in Sub-Saharan Africa

Prepared for the Commonwealth Secretariat

Jodie Keane, Massimiliano Calì & Jane Kennan Overseas Development Institute

September 2010

Overseas Development Institute 111 Westminster Bridge Road London SE1 7JD, UK

Disclaimer: The views presented in this paper are those of the author(s) and do not necessarily represent the views of ODI or our partners.

Tel: +44 (0)20 7922 0300 Fax: +44 (0)20 7922 0399 www.odi.org.uk

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Contents

Contents Tables, figures & boxes List of acronyms Executive summary Introduction 1 1.1 1.2 1.3 1.4 2 2.1 2.2 3 3.1 3.2 3.3 4 Intra-regional trade in sub-Saharan Africa Theory and practice Newer contributions to theory Trends Potential for intra-regional trade – structural analysis The impact of non-tariff barriers on intra-regional trade Types and potential trade impacts of non-tariff barrier Quantitative assessment of the effects of NTBs on intra-regional SSA trade flows: SADC Addressing non-tariff barriers in SADC The SADC Trade Protocol and Rules on NTBs NTBs identified and measures in place in SADC Other strategies to address the trade effects of NTBs in SADC Conclusion References Appendix 1 – Major regional economic communities in Africa Appendix 2: GDP per capita for SADC and ECOWAS members Appendix 3: ECOWAS trade and revealed comparative advantages Appendix 4: SADC trade and revealed comparative advantages Appendix 5: Quantitative Assessment of NTBs on Trade, SADC Appendix 6: Non-tariff barriers in SADC i ii iv vi 1 2 2 3 3 12 18 18 21 27 27 28 32 36 38 41 42 43 47 53 56

The Review on Export Trade of Bangladesh with Saarc Countries

Introduction The policy of trade liberalization and free-market economy in the 1980s has created both challenges and opportunities for Bangladesh economy. The creation of the World Trade Organization has created new ways of enjoying the comparative advantage for Bangladesh. At the same time, the globalize trade scenario has opened up the structural limitations of Bangladesh economy, which needs ...

i

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Tables, figures & boxes

Tables

Table 1: Impacts of NTBs registered on TRAINs on Intra-Regional Trade in SADC Table 2: ECOWAS intra-regional exports (% share) Table 3: ECOWAS intra-regional imports (% share) Table 4: ECOWAS regional trade intensity indices Table 5: SADC intra-regional exports (% share) Table 6: SADC intra-regional imports (% share) Table 7: SADC regional trade intensity indices Table 8: RCA indicators for ECOWAS and rest of the world Table 9: Trade complementarities index in ECOWAS Table 10: RCA indicator for SADC and rest of the world Table 11: Trade complementarities index in SADC Table 12: Example of the recording of NTMs from UNCTAD TRAINS Table 13: Summary statistics of the main variables used in quantitative analysis for SADC Table 14: The impact of NTM on imports in Southern Africa, 2003–6 Table 15: The impact of NTMs on imports in Southern Africa by country Table 16: Types of NTB by SADC members between 21 January 2009 to 8 June 2010 Table 17: Reported NTBs imposed by SADC members, 21 January 2009 to 8 June 2010 Table A: GDP per capita for SADC members (current US$) Table B: GDP per capita for ECOWAS members (current US$) Table C: RCAs for ECOWAS Members compared to rest of the world Table D. Value of SADC exports for the period 2000–2008 (US$ millions) Table E: Value of SADC imports for the period 2000–2008 (US$ millions) Table F. SADC RCA compared to rest of the world– other top products Table G: RCAs for SADC members compared to rest of the world Table H: Results of GMM regressions according to sectoral aggregation (2) Table I: Sectoral aggregations Table J: NTBs identifiable in most recent TPRs by type and prevalence

ix 8 9 9 11 11 12 14 15 16 17 23 24 25 26 32 33 42 42 45 47 48 50 51 54 54 56

The Essay on Export Import Sitution of Bangladesh

It gives ours immense pleasure to submit an assignment on export import situation of BD. This assignment is submitted as a partial fulfillment as a part of our course. The preparation of the assignment has given ours and insightful experience and in-depth knowledge on “Export Import Situation of BD. We have given our best effort to make it worthy one and each aspect of the problem is considered ...

ii

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Table K: Number of NTBs reported to Tripartite Monitoring Mechanism, 21 January 2009 to 8 June 2010 59

Figures

Figure 1: Regional trade agreements in Africa Figure 2: World: intra-regional exports as a proportion of total exports, 1980–2008 (%) Figure 3: Composition of world exports by region, 2008 Figure 4: SSA RECs value of intra-regional exports (US$ million), 1980–2008 Figure 5: Intra-regional exports as a proportion of total exports in SSA RECs, 1980–2008 (%) Figure 6: Composition of exports from SSA RECs in 2008 Figure 7: GDP (US$ million) and intra-regional exports in SADC and ECOWAS Figure 8: Average number of full coverage NTMs per H6-sector, SADC vs other exporters Figure A: Value of top ten ECOWAS exports, 2000–8 (US$) Figure B: Value of top ten ECOWAS imports, 2000–8 (US$) Figure C: Total value of ECOWAS exports, 1980–2008 (US$ million) Figure D: Direction of ECOWAS exports, 1980–2008 (percentage by destination) Figure E: Value of top ten SADC exports for period 2000–2008 (US$) Figure F: Value of top ten SADC imports for the period 2000–8 (US$) Figure G. Total value of SADC exports, 1980–2008 (US$ million) Figure H. Direction of SADC exports for period 1980–2008 (% by destination) Figure I: Bound and applied tariffs among SADC countries

Boxes

4 5 5 6 6 7 7 24 43 43 44 44 47 48 49 49 54

Box 1: TradeMark Box 2: Rules of origin as provided for in Annex I of SADC Trade Protocol Box A: Estimating the Welfare Impact of NTBs Box B: Robustness Checks for Regression analysis

29 30 53 55

iii

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

List of acronyms

AEC AFTA ASEAN CEMAC CET CFA CIF CMA COMESA COMTRADE CU EAC ECCAS ECOWAS EU FE FOB FTA GATT GDP GMM HS IGAD IOC MMTZ NTB NTM QR RCA REC RoW RTA SACU SADC SITC SOE SPS African Economic Community ASEAN Free Trade Area Association of Southeast Asian Nations Communauté Économique des États d’Afrique Centrale Common External Tariff Communauté Financière Africaine Cost, Insurance and Freight Common Monetary Area Common Market for Eastern and Southern Africa (United Nations) Commodity Trade Statistics Customs Union East African Community Economic Community of Central African States Economic Community of West African States European Union fixed effect Free on Board Free Trade Area General Agreement on Tariffs and Trade Gross Domestic Product Generalised Methods of Moments Harmonised System (of trade classification) Intergovernmental Authority on Development Indian Ocean Commission Malawi, Mozambique, Tanzania and Zambia Non-tariff barrier Non-tariff measure Quantitative restriction Revealed Comparative Advantage Regional Economic Community Rest of the World Regional Trade Agreement Southern African Customs Union Southern African Development Community Standard International Trade Classification State-Owned Enterprise Sanitary and phytosanitary standards

Trade Network Essay Eurasian And African

In the trade networks between Africa and Eurasia from circa 300 C.E. to 1450 C.E., there were key continuities and changes. Although there was some continuity in Eurasian and African trade, it was mainly the key changes that led to advancements in technology, trade networks and involvement of other societies in trade. In 300 C.E., there was limited trade between Africa and Eurasia. The dominant ...

iv

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

SSA TBT TMEA TMSA TPR TRAINS UEMOA UMA UNCTAD VER WAEMU (also UEMOA) WAMZ WTO

Sub-Saharan Africa Technical barriers to trade TradeMark East Africa TradeMark Southern Africa Trade Policy Review (UNCTAD) Trade Analysis and Information System (see WAEMU) Arab Maghreb Union United Nations Conference on Trade and Development voluntary export restraint West African Economic and (Union Economique et Monétaire Ouest Africaine) West African Monetary Zone World Trade Organization Monetary Union

v

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Executive summary

Whilst there have been multilateral, regional and bilateral reductions in tariffs among countries and regions, including in sub-Saharan Africa (SSA), non-tariff barriers (NTBs) can act as important constraints on trade and limit the gains from increased market access through tariff reductions. As traditional barriers to trade such as tariffs decline there are concerns that NTBs may subsequently increase. There are a range of different types of NTBs, some of which may reflect perfectly legitimate public policy concerns. We define a NTB as an unnecessarily restrictive non-tariff measure (NTM) which affects trade in goods. The presence of NTBs can undermine the gains from trade liberalisation for new entrants and constrain diversification efforts, across products as well as markets. Intra-regional trade in SSA appears to be low and there are different views as to why this is the case; in addition to how and why increased intra-regional trade on the continent may be beneficial. Although it is often stated that NTBs are important constraints to trade in SSA, there have been limited attempts so far to systematically quantify their actual impact on trade. This analysis sets out to explore the extent to which intra-regional trade flows for African regional economic communities (RECs) are constrained by NTBs. A quantitative methodology for the assessment of the impact of reported NTBs on trade flows is developed and then applied to the Southern African Development Community (SADC).

The Dissertation on Shanghai Free Trade Zone and Its Impact on Chinese Economy

SOUTHWESTERN UNIVERSITY OF FINANCE AND ECONOMICS SCHOOL OF INTERNRATIONAL BUSINESS Shanghai pilot free trade zone and its impact on Chinese economy TABLE OF CONTENT INTRODUCTION.................................................................................................... - 2 BACKGROUND ..................................................................................................... - 3 ...

Based on the results of a quantitative assessment of the identified impacts of NTBs on intra-regional trade, we suggest policy measures and steps towards addressing them. Regional Integration: Theory and Practice One intended effect of a regional trade agreement (RTA) is, through the reduction and removal of tariffs, to enable more efficient producers in a region expand production (and reap economies of scale and scope) to the advantage of consumers and the detriment of less competitive producers. However, this potential may be limited unless other barriers to trade are also addressed, and harmonised. The process of fostering closer regional integration means developing new policy tools. This includes the development of regional NTMs to increase intra-regional trade flows, such as harmonised standards to facilitate trade, as well as rules of origin (RoO) which are required to avoid trade deflection. Some NTMs may reflect legitimate public health concerns, others more strategic regional or national developmental objectives. The challenge is to ensure that these types of NTMs are not unnecessarily trade restrictive and so become NTBs. The objectives of regional RoO include ensuring that members of RECs benefit from market access entitlements relative to competition from third party countries.

But if too restrictive or uncoordinated with emerging production networks and business strategies, these new rules may render regional trade liberalisation strategies ineffective. Regional standards or RoO may undermine potential dynamic gains if producers are not informed about them, unable to meet them or if they are implemented in an ad hoc, unpredictable and less harmonised way. Moreover, should the real intent of NTMs applied on intra-regional trade by individual members of RECs be to limit rather than better manage trade flows, for example, because of fears of disproportionate economic benefits accruing to other regional partners. Distinguishing between the intent and impact of NTMs on trade helps to determine the extent to which they may actually be NTBs. This information may be particularly helpful for those African RECs that seek deeper economic integration. There are many RTAs on the African continent, overlapping and complementing each other in some cases, but with conflicting objectives in others. For countries covered by more than one trade agreement, importers have a choice of regimes under which to import goods. For small and medium sized enterprises (SMEs), overlapping membership may pose particular difficulties through increasing the trade costs of exporting to different regional markets which may have varying standards-related entry requirements. Consequently, this could reduce the potential for benefits which result from scale and therefore constrain product and market diversification efforts. Although replacing overlapping membership with one allvi

The Essay on U.S. Trade Analysis with other Countries

Abstract Purpose- This paper presents the analysis of U.S. imports and exports by managing the trade balance. It also presents the leading U.S. imports and exports in terms of value along with the important partners. Design/methodology/approach- The author explains the balance of trade including the rise and fall of U.S. trade deficit using the analysis between different countries imports and ...

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

embracing REC may help in reducing the costs of trading for firms, it also makes the task of harmonizing rules and regulations greater for governments. Assessing the Impact of NTBs Information about the coverage and impact of NTBs is often hard to collect. This means that their impacts on trade and economic welfare are not easy to quantify or model (Fugazza and Maur, 2008).

Despite these measurement challenges, this analysis attempts to assess the welfare impacts of introducing or removing NTBs on intra-regional trade in SADC. Estimates of the potential welfare impacts of NTBs can help policy makers act to on them. Approaches to measuring NTBs range from frequency-type inventories to price comparison measures (tariff equivalents), to quantity impact measures based on estimation of trade flows. The United Nations Conference on Trade and Development (UNCTAD) has been collecting data on the incidence of NTBs for a large sample of countries in its Trade Analysis and Information System (TRAINS).

The Term Paper on Regional Economic Impacts Of Idaho State

The Regional Economic Impacts of Idaho State University, 1996 Dr. Richard Bowen, President of Idaho State University, requested the Center for Business Research to conduct a study of the impact of the University on Idaho, with particular emphasis on its regional impacts upon the Pocatello area. This publication reports the findings and results of that effort, which was begun in the fall of 1997. ...

Its database, which is country as well as commodity specific, uses a classification of over 100 trade measures, including those with a discretionary or variable component. Its main source of information includes GATT notifications, government publications and WTO Trade Policy Reviews (TPRs).1 The incidence of reported NTBs, either as a count or percentage of coverage of specific product lines (HS or Standard International Trade Classification (SITC)) include such measures as:

   

price control measures, such as multiple exchange rates, or foreign exchange allocation; finance control measures, such as anti-dumping or countervailing measures; quantity restrictions, such as non-automatic licensing, quotas; monopolistic measures;

 technical measures, such as regulations and customs procedures; and  miscellaneous including subsidies.

Despite the range of NTBs that may be reported, most of those listed fall in the category of ‘technical measures’. The database does not include any measures related to: corruption – which may include the use of roadblocks, export related measures, government procurement, intellectual property rights and other investment related measures. Each measure listed on the database is given an equal weight regardless of the intensity of the measure and its trade related impacts. This limits the ability to distinguish between different types of NTB reported in terms of the severity of trade impacts. Despite this, the TRAINS database is to the best of our knowledge the most systematic source of data on NTBs available across countries and products. Kee et al. (2008) use UNCTAD TRAINS data to compute indices of trade restrictiveness for NTBs across countries. However, as these indices are computed on the basis of the actual effects that these measures have on trade on one year for each country (and these years differ across countries), using them to estimate the effects of NTBs on trade would lead to biased estimations.2 We therefore take a different route and undertake a systematic evaluation of the actual impact of reported NTBs on the imports of selected SADC countries (HS 6-digit level) over a number of recent years. Because of data limitations, our analysis is limited to the SADC region and within it to Botswana, Namibia, South Africa and Swaziland; these countries therefore comprise our SADC sample.

1

2

However, as noted by Fugazza and Maur (2008), the categories used by TRAINS are not the same as those of the WTO. Although the database covers a wide range of NTBs, most listed fall within the Technical Measures category. In particular, the use of such indices would make the estimation endogenous.

vii

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

The idea is to perform a direct test by matching the NTBs recorded by UNCTAD, with corresponding import data from UNComtrade. Because the NTBs reported to UNCTAD TRAINS have not actually been assessed in terms of their trade restrictiveness, and the evidence on the extent to which they actually constrain trade is not available, we use the term NTM to refer to them, until such time as we have obtained the evidence to classify them as actual barriers to trade – NTBs. We create three different non-tariff measure (NTM) indicators: a ‘NTM partial coverage’ variable; a ‘NTM full coverage’ variable; and ‘NTM any coverage’ variable, which is a sum of the two.3 This is because of the nature of NTB reporting, sometimes it is not known whether a measure applies to each 6-digit sector or not. For instance a measure can be listed as applying to HS 02 (vegetable products) but with a ‘partial coverage indicator’, implying that the measure does not cover every 6-digit code within the 2-digit chapter. We then match the NTM data with bilateral import data (in current thousand US dollars) at the HS 6-digit sector level and create three different samples according to the trading partner considered: full; SADC only; and non-SADC. We test whether imports from the four reporting countries in a sector where one or more NTM is imposed at year t grow less (or decline more) than those in sectors which do not experience an increase in NTMs in t, as well as whether imports in a sector grow less (or decline more) in those periods when NTMs are not applied, relative to periods when NTMs are imposed. As our interest lies mainly in the impact of NTMs on intra-regional trade, we test for any differential impact of NTMs on imports from SADC countries vis-a-vis from other countries. In this way we examine whether NTMs are more or less of a constraint for intraregional flows relative to flows with the rest of the world. Results from Quantitative Analysis The results of the analysis and estimation of the NTM coefficients are presented in Table 1. We find that the NTM dummy is negative and highly significant for the imports from SADC countries (Column 2), while it is positive and significant for the imports from non-SADC countries (Column 3).

That is we find that the introduction of one or more NTMs in a sector significantly penalises imports from other SADC countries in that sector (intra-regional trade) to the benefit of non-SADC countries, whose exports increase. This confirms the hypothesis that the NTBs reported to TRAINS are indeed barriers to intra-regional trade for SADC countries. Moreover, to the extent that these barriers divert imports away from regional towards non-regional partners, their presence seems to stifle intra-regional trade. What drives this differential impact? There are two potential explanations: first, SADC countries could be on average much less able to adjust to the NTBs listed than other exporters to the Southern African region; second, SADC exports could be concentrated in product lines particularly susceptible to NTBs. Therefore the eventual difference in exposure by exporter would come from the sectoral composition of its exports to the SADC countries considered. Results from further analysis suggest that a little more than half of the impacts of the NTBs analysed on trade comes from the different abilities of exporters to adjust to the imposition of them.

3

The results presented here are based on the use of the ‘NTM full coverage’ variable, but they hold using the ‘NTM any coverage’ variable as well.

viii

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Table 1: Impacts of NTBs registered on TRAINs on Intra-Regional Trade in SADC

1 All NTM dummy coverage) Year Effects Importer-productexporter effects Observations Nr. of groups (full -0.055 (0.078) YES YES 428,208 207,800 2 SADC -2.099*** (0.268) YES YES 76,500 33,122 3 No SADC 0.236*** (0.081) YES YES 351,708 174,678

Robust standard errors in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%. Dependent variable is the percentage growth of imports over the previous year. Importing countries are Botswana, Namibia, South Africa and Swaziland. Endogenous variable in GMM estimations are the NTM variable and the NTM-sectoral interaction terms.

We also analysed the impact of NTBs separately by importing country, for Botswana, Namibia and South Africa (the data for Swaziland do not allow a separate estimation).

The analysis reveals that NTBs applied by South Africa have a more significant negative impact on imports from SADC than Botswana and Namibia. On the other hand in the case of Namibia and Botswana, whose imports from SADC are mainly from South Africa, the NTB coefficient is less significant, which is consistent with the hypothesis that traders in South Africa are better able to tackle these barriers than other SADC countries. Efforts to Address NTBs in SADC For those in favour of regional integration in SSA the results of the quantitative analysis undertaken should be of concern: they suggest that the imposition of NTBs by SADC countries is usually handled better by nonSADC than SADC countries, and within SADC by the economically larger members. In light of these results, initiatives aimed at tackling NTBs and their impacts on trade at the regional level become even more important. We supplemented the quantitative analysis undertaken with a qualitative assessment and synthesis of NTBs identified in most recent TPRs and other regional databases. Most of the NTBs mentioned as potentially problematic for importers and exporters in the recent WTO TPRs4 for SADC members fall within the following categories: price controls; quantity restrictions; and miscellaneous, such as other charges intended to protect local industry and/or encourage local processing. Although not explicitly discussed in TPRs as being problematic, a number of areas are mentioned as requiring closer regional cooperation and harmonisation of rules and regulations including: competition policy; protection of domestic industries; SPS/TBT; rules of origin; and customs procedures. We analysed recent notifications of NTBs to the NTB Monitoring Mechanism (established to support efforts to integrate the economies of the Common Market for Eastern and Southern Africa (COMESA), the East

4

We reviewed the most recent TPRs available for SADC members, most of which are dated from 2006 onwards. No reviews are available for Democratic Republic of Congo, the Seychelles or Zimbabwe.

ix

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

African Community (EAC) and SADC as part of the Tripartite Agreement).

At the current time the distribution of reported NTBs across sectors, and products, is not clear. However, most complaints registered relate to South Africa and its trade-related administrative NTBs (though it is not clear which types).

This category is also the most frequently reported barrier for all SADC members, followed by export and import licences and transit issues. NTBs imposed on importers from outside of SADC, and SSA, are not reported as much and may therefore be less problematic for importers and exporters, or less applied. These results appear to substantiate those of the quantitative analysis. Policy Implications Monitoring the nature, evolution and impacts of reported NTBs constitutes one step towards reducing negative impacts on trade and economic welfare. For policy makers, responding to private sector concerns about NTBs may entail policy shifts and the simplification of rules and regulations. There may be a need for increased awareness-raising among producers and traders about new rules and regulations. In addition to the provision of support to assist producers comply with NTMs. These steps may help avoid legitimate policy tools becoming barriers to trade. Once problematic NTBs have been identified, and analysed, it makes sense to start classifying them according to their intent and severity of trade impact, with a view to beginning to address them at the national and regional level. It might be easier to begin this approach on product-specific basis, first. For example, steps towards addressing NTBs may include: identifying regional priority sectors and products; monitoring reported NTBs on these sectors and products; analysing the intent and impact of reported NTBs; and reducing the negative impacts of them on trade either through their removal, harmonization of related policy, or provision of appropriate support to producers. In practice, however, progress on the reduction of NTBs is likely to take time, be costly and involve difficult trade-offs. Specific policy recommendations which arise from our analysis for SADC include:

 Improving the coding of the new tripartite NTB monitoring database. For example, by sector

and product, including indicators related to the severity of impact for importers and exporters – in terms of time, cost and related trade impacts.

 Linking the monitoring of NTMs to compliance and enforcement mechanisms. Once NTBs have

been reported and analysed they need to be acted upon which requires dialogue and collaboration at the regional level and between national institutions.

 Investing in conformity infrastructure. Limited testing infrastructure constrains the ability of

members to sign mutual recognition agreements with regional partners and instead the fallback position becomes the standards of the most dominant trading partner, which may themselves constitute NTBs for lesser developed regional partners. Results from our analysis suggest the economically smaller members of SADC find NTBs a particular challenge, which may be a result of scale and a limited ability to spread fixed costs over a large export basket, or because of limited conformity infrastructure within country.

 Harmonising infant industry protection. Article 21 of the SADC Protocol on the Protection of

Infant Industries permits the suspension of certain obligations of the Protocol, but it also specifies that terms and conditions should be imposed. However, at present there appears to be a policy void. Although SACU members are obliged to have a competition policy, commitments needs to be matched with enforcement. This is to say, infant industry across the SADC region could be much stronger if coordinated and monitored. This includes identifying regional priority products and harmonising related policy (for example, creating simple, consistent and predictable RoO) so as to spur regional growth dynamism. As efforts to foster closer regional integration in SADC gather pace it is important to note that the monitoring of NTBs does not end as efforts begin to reduce negative impacts, but instead needs to

x

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

continue as regional surveillance mechanisms, policy and related compliance and enforcement mechanisms further develop and strengthen. Whether or not efforts to further promote regional integration on the continent in the 21st century can lead to increased intra-regional, as well as international trade, arguably depends on a more strategic approach to ‘behind the border’ issues being adopted so that the potentially dynamic gains from deeper integration may be harnessed.

xi

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Introduction

Whilst there have been multilateral, regional and bilateral reductions in tariffs among countries and regions, including in sub-Saharan Africa (SSA), non-tariff barriers (NTBs) can act as important constraints on trade and limit the gains from increased market access through tariff reductions. As traditional barriers to trade such as tariffs decline there are concerns that NTBs may subsequently increase. We define a NTB as an unnecessarily restrictive non-tariff measure (NTM) which affects trade in goods. The presence of NTBs can undermine the gains from trade liberalisation for existing and new entrants, and can impede diversification efforts, across products as well as markets. Intra-regional trade in SSA appears to be low and there are different views as to why this is the case; in addition to how and why increased intra-regional trade on the continent may be beneficial. There are different types of determinants of intra-regional trade. These range from economic variables, such as differences in factor endowments and complementarities in trade structures, to policy variables such as tariffs and NTBs. Other aspects such as geographical location may serve as a natural non-tariff barrier to accessing particular markets, but like other market failures may be overcome through effective and targeted government intervention (Calì 2009).

NTBs are perceived to be important constraints to trade in SSA but there have been limited attempts so far to systematically quantify their actual impact on trade. Estimates as to the potential welfare impacts of NTBs can help to provide information to policy makers to act on them. How do NTBs impact on intraregional trade? Is there any differential impact of NTBs on regional partners vis-à-vis third parties? This information is particularly relevant for those regions seeking to foster deeper economic integration and enhance intra-regional trade flows. This paper sets out to explore the extent to which intra-regional trade flows on the continent are constrained by the imposition of NTBs. We develop a quantitative methodology in order to undertake this analysis which uses product level import and NTB data. This paper is based on three complementary building blocks and is organised as follows: Section One briefly reviews the theory and practice of regional integration, which suggests that similar production structures can impede intra-regional trade for low-income countries but may be conducive to intra-regional trade between high-income countries. We present and discuss recent trends in intra-regional trade flows in SSA, and in Economic Community of West African States (ECOWAS) and Southern African Development Community (SADC), in particular. We assess trade complementarities using specific indices, including revealed comparative advantages (RCAs), and discuss the implications of these findings for potential increases in intra-regional trade in sub-Saharan Africa. Section Two introduces the pathways through which NTBs may impact on trade, including intra-regional trade. We develop a methodology to quantify and analyse the impact of NTBs on intra-regional trade flows for SADC as a group and on selected members (Botswana, Namibia, South Africa and Swaziland).

Section Three complements the analysis with a qualitative assessment of NTBs based on the most recent World Trade Organisation (WTO) Trade Policy Reviews (TPRs) for SADC countries. We discuss the extent to which the effects of different types of NTBs on intra-regional trade are being addressed by existing policy measures. Finally, we conclude with a summary of the key findings and policy messages. Recommendations are provided, based on the results of our analysis and which draw on the experience of other regions in their efforts to harmonise and reduce NTBs so as to spur increases in intra-regional trade.

1

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

1

1.1

Intra-regional trade in sub-Saharan Africa

Theory and practice

Most countries now belong to at least one bilateral and regional trade agreement (RTA) which have continued to increase in number, and size, since the early 1990s; this includes in SSA. There are several stages in the regional economic integration process which range from the formation of a free trade area (FTA) to a customs union (CU) and the establishment of an economic and monetary union. Most of the agreements signed up to date constitute FTAs. However, an important step in the regional integration process is the formation of a CU, which not only eliminates tariffs and quotas on trade amongst member countries but also establishes a common external tariff (CET) which is applied to trade with non-members and third party countries. The literature on regional integration dates back to at least Viner (1950), who suggested that the effects of regional integration on trade can be either trade creating or trade diverting. Like any form of liberalisation, one intended effect of an RTA is to allow the more efficient producers in the region to expand production (and reap economies of scale) to the advantage of consumers and the detriment of less competitive producers. This is called trade creation. Trade diversion occurs when the removal of tariffs within the region leads to goods that were previously imported from outside (from the cheapest global source) being replaced by more expensive goods produced inside the region which can be sold for less because they no longer have to pay any import duty. Consumers still gain, although by less, but governments lose more in tariff revenue and the country as a whole is able to obtain fewer imports for a given value of exports. This implies that regional integration can lead to further trade, but that these flows may not always be welfare enhancing. However, there are other important roles that closer regional integration and cooperation can play. Regions can better support the provision of, for example, public goods; this includes ‘hard’ infrastructure like roads, energy and the physical networks required to support trade, as well as ‘soft’ infrastructure such as institutions, related to the governance of trade. Many competitiveness challenges are regional in nature. For example, a landlocked country is dependent on the appropriate infrastructure being available in transit countries for its trade flows. National development programmes will not normally consider activities with strong regional (or international) externalities as the benefits cannot be fully appropriated nationally. There have been various attempts to analyse the relationship between regional integration and trade in SSA. These studies suggest that the new wave of regionalism which began in the 1990s (with the creation of new blocs and the revamping of old blocs) has not led to further intra-regional trade.5 For example, te Velde (2006) argues that because intra-regional trade in Africa covers only a small percentage of total trade, any trade (and hence economic) effect of lower tariffs is likely to be small. Others have therefore argued that an alternative approach could be to foster deeper integration through the harmonisation of trade rules and standards and institutional co-operation (Gasiorek and Holmes, 2008).

New regionalism moves beyond static trade creation benefits and emphasises the potential dynamic trade and welfare gains from reductions in administration, transaction costs and the elimination and/or harmonisation of other types of NTBs. The reduction of NTBs and harmonisation of other non-tariff measures (NTMs) such as standards and customs clearings procedures constitute a deeper form of integration, with the potential for more dynamic gains in terms of trade creation if harnessed correctly.

5

See te Velde (2005) for an overview of the literature and the links between regional integration and poverty. 2

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

1.2

Newer contributions to theory

The literature on regional integration and trade has been informed by new developments within economic theory. The model developed by Venables (2003) includes some aspects of new trade theory such as the recognition of forces that may foster economic convergence as well as divergence, for example agglomeration effects. However, despite this, the model still relies on the classical theory of comparative advantage, which underpins the theory of the CU as developed by Viner (1950).6 Venables extends Viner’s model and shows how the creation of a CU between two low-income countries with a similar comparative advantage (and therefore factor endowments) may lead not only to trade diversion but also to greater economic divergence. This is because of traditional forces of trade diversion and creation working in a perfectly competitive environment. Venables (2003) therefore argues that a greater potential for welfare gains exists from trade between countries with vastly different factor endowments (similar to HeckscherOhlin theory).

This could take the form of the inclusion of a more developed country within South–South FTAs, leading to North–South–South arrangements.7 Others have argued that although the inclusion of countries with vastly different static comparative advantages may help to promote more economic convergence than divergence, the gains may still be disproportionate. Krugman (1993) and Puga and Venables (1995) show how relatively large welfare gains may accrue to the most developed country of a regional trade agreement, or hub, because of conditions of imperfect competition and the agglomeration effects of industries that choose to cluster closer together. They try to better include more dynamic aspects of trade-induced growth fostered through the process of regional integration, but similarly retain their classical roots by basing their analysis on comparative advantage as indicated by technological differences (similar to the Ricardian model).

The creation of an FTA or CU is likely to have both trade creation and diversion effects; the efficiency and welfare enhancing, or reducing, effects of which depend on what process – creation or diversion – dominates. There are many different pathways through which closer regional integration and the formation of an FTA can affect intra- and extra-regional trade. This study focuses on the policy environment and explores the extent to which NTBs may undermine regional integration efforts and constrain intra-regional trade. Before outlining these pathways, it first presents recent trends in intra-regional trade flows in SSA and within two selected regions in particular. It then analyses the structural characteristics of trade within these two regions and whether or not its composition suggests the potential for increases in intra-regional trade, based on traditional and newer contributions to theory.

1.3

Trends

The African Economic Community (AEC) – an integral part of the African Union – sets out the continental framework for economic integration. It recognises that the process of fostering economic integration at the continental, regional and sub-regional levels requires the rationalisation and harmonisation of Regional Economic Community (RECs).8 In addition to reductions in tariffs this includes due consideration of ‘soft’ infrastructure related to the governance of intra-regional trade, such as rules of origin, product standards and accreditation systems, the harmonisation of which may help to reduce costs for business and therefore facilitate increases in intra-regional trade.

6

7

Which is itself a neoclassical model since it assumes perfect competition in commodity and factor markets and factor mobility within country but not between countries; economies of scale are not considered. The European Commission also believes that regional integration between developed and developing countries can be beneficial; it has recently promoted North–South agreements, which are intended to build on Southern regional integration, termed South–South–North FTAs (European Commission, 2002).

See also Bilal (2007).

See UNCTAD (2009).

3

8

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

RTAs currently on the African continent are overlapping and complementing each other in some cases but at times have conflicting objectives. Figure 1 shows the memberships of these agreements in 2009. For countries covered by more than one agreement, importers have a choice of arrangement under which to import goods, which does not necessarily create difficulties. It may, however, introduce particular difficulties, as well as costs, for small and medium sized firms seeking to export. If product standards are more stringent in one market compared to another the complexity of exporting to different markets within a given region increases. This may reduce potential scale benefits and therefore constrain product and market diversification efforts. Although replacing overlapping membership with one all-embracing REC may help in reducing costs for firms, it also makes the task of harmonizing rules and regulations greater for governments.9

Figure 1: Regional trade agreements in Africa

Source: Adapted from UNCTAD, 2009a.

SSA has the lowest share of intra-regional exports as a proportion of total in the world (Figure 2) and this has been the case since the 1980s. There has been growth in intra-regional exports since the 1990s, from 6% in 1990 to 12% in 2000, compared to an increase of just 3% between 1980 and 1990. But this growth has not been sustained: intra-regional exports were 12% of total exports in 2008, the same level as in 2000. Despite this, it is important to point out that unlike some other regions such as developing America, SSA has not seen its intra-regional exports decrease. But nor has it seen as rapid increase in intra-regional exports as other regions such as developing Asia, which managed to almost to double its intra-regional exports over the period 1980–2008.

9

Appendix 1 presents the major RECs on the continent and their ultimate objectives, and includes UNCTAD’s (2009) summary of their current status. 4

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Figure 2: World: intra-regional exports as a proportion of total exports, 1980–2008 (%)

80 70 60 50 40 30 20 10 0

1980

1990

2000

1981

1982

1983

1984

1985

1986

1987

1988

1989

1991

1992

1993

1994

1995

1996

1997

1998

1999

2001

2002

2003

2004

2005

2006

2007

Sub-Saharan Africa Developing economies: Asia Developed economies: America

Developing economies: America Developed economies: Europe

Source: Adapted from UNCTAD 2009b. Note: Countries included within each REC for each year are not known.

Compared to other regions such as developing Asia and the Americas, SSA’s exports are highly concentrated and consist predominantly of fuels (Figure 3).

Despite this aggregate concentration in less sophisticated primary products, which are considered to embody limited potential to sustain trade dynamism, there are some clear differences in the recent performance of intra-regional trade flows within SSA’s RECs, and their composition.

Figure 3: Composition of world exports by region, 2008

100% 80% 60% 40% 20% 0% Sub-Saharan Africa Developing economies: America Developing Developed economies: Asia economies: Europe Developed economies: America

All food items (SITC 0 + 1 + 22 + 4) Agricultural raw materials (SITC 2 less 22, 27 and 28) Ores, metals, precious stones and non-monetary gold (SITC 27 + 28 + 68 + 667 + 971) Fuels (SITC 3) Manufactured goods (SITC 5 to 8 less 667 and 68) Iron and steel (SITC 67)

Source: Adapted from UNCTAD 2009b. Note: Countries included within each REC for each year are not known.

As shown in Figure 4, growth in intra-regional trade appears to have been faster for SADC and ECOWAS than for other RECs in Africa. The value of intra-regional exports was highest in ECOWAS – which also encompasses the West African Economic and Monetary Union (WAEMU) – during the period 1980 to 1990.

5

2008

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

However, since the first half of the 1990s intra-regional trade flows within SADC have accelerated, surpassing the value of those within ECOWAS. The increase in value of SADC intra-regional exports has been considerable, particularly from 2000 onwards.

Figure 4: SSA RECs value of intra-regional exports (US$ million), 1980–2008

20000 18000 16000 14000 12000 10000 8000 6000 4000 2000 0

1982

1985

1988

1991

1994

1980

1981

1983

1984

1986

1987

1989

1990

1992

1993

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

COMESA SADC

ECCAS UMA

ECOWAS WAEMU

Source: Adapted from UNCTAD 2009b. Note: Countries included within each REC for each year are not known

This recent take off in intra-regional exports within SADC is much more pronounced when presented as a share of total exports (Figure 5).

There appears to have been some tailing-off between 2000 and 2008, which suggests that the share of the region’s exports to the rest of the world (RoW) have simply growth faster in terms of value over that period, whereas in the period 1990 to 2000 intra-regional exports grew faster. Although lower in value, intra-regional exports for the sub-group of WAEMU as a proportion of total can be seen to have exceeded those of ECOWAS since 1980, and even though erratic, since the late 1990s their growth appears to have been sustained.

Figure 5: Intra-regional exports as a proportion of total exports in SSA RECs, 1980– 2008 (%)

20 18 16 14 12 10 8 6 4 2 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 COMESA ECCAS ECOWAS SADC UMA WAEMU

Source: Adapted from UNCTAD 2009b. Note: Countries included within each REC for each year are not known.

The sub-region of WAEMU has a more diversified export basket than ECOWAS; food items comprise a larger share of exports than fuels, and manufactured goods are more prominent (Figure 6).

However,

6

2008

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

compared to both these regions SADC has a much larger share of manufactured goods and lower share of primary products in its export basket.

Figure 6: Composition of exports from SSA RECs in 2008

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% COMESA ECCAS ECOWAS SADC UMA WAEMU Ores, metals, precious stones and non-monetary gold (SITC 27 + 28 + 68 + 667 + 971) Agricultural raw materials (SITC 2 less 22, 27 and 28) Manufactured goods (SITC 5 to 8 less 667 and 68) Fuels (SITC 3) Iron and steel (SITC 67)

Source: Adapted from UNCTAD 2009b. Note: Countries included within each REC for each year are not known.

ECOWAS exports’ protracted concentration in fuels, which are more subject to the vagaries of international markets and less conducive to dynamic gains, may be partly responsible for its more erratic as well as sluggish increases in intra-regional trade. In the following sub-sections we identify which countries within each REC account for the largest shares of intra-regional trade and analyse their structural characteristics in more detail. Suffice it to say here that in terms of the role of economic growth as a driver of intra-regional trade, and vice versa, there is no correlation between gross domestic product (GDP) and intra-regional exports in ECOWAS. In comparison, a stronger positive association is apparent in SADC (see Figure 7).

These results do not however, indicate causality.10

Figure 7: GDP (US$ million) and intra-regional exports in SADC and ECOWAS

500000 20 350000 300000 15 10 200000 100000 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 5 0 250000 200000 150000 100000 50000 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 GDP (current US$) Intra-regional exports, % of total exports 14

SADC

400000 300000

ECOWAS

12 10 8 6 4 2 0

GDP (current US$) Intra-regional exports, % of total

Source: World Development Indicators; UNCTAD 2009b.

10

The coefficient of determination for SADC is 0.38 which suggests that 38% of the variation in one variable may be explained by the variation in the other (compared to 0.5% for ECOWAS).

There is no correlation between GDP and intra-regional exports in ECOWAS (0.23), but a strong positive association in SADC (0.6).

It is beyond the scope of this report to explore the drivers of growth in more detail. See Tables A and B, Appendix 2 for per capita GDP data. 7

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

1.3.1

ECOWAS

ECOWAS seeks to promote regional political co-operation and economic integration, with the aim of eventually establishing an economic union among West African countries. The treaty which established ECOWAS was signed in Lagos, Nigeria, on 28 May 1975 and came into force on 20 June 1975. ECOWAS comprises most of West Africa and is designated as one of the five regional pillars of the AEC, together with Common Market for Eastern and Southern Africa (COMESA), the Economic Community of Central African States (ECCAS), the Intergovernmental Authority on Development (IGAD) and SADC; ECOWAS signed the Protocol on Relations between the AEC and RECs in February 1998. Although ECOWAS states as its objective the creation of a full economic union, it already includes one functioning CU – WAEMU, which became operational in 2000. The region is also divided by two currencies: the Communauté Financière Africaine (CFA) zone, which encompasses WAEMU, and the West African Monetary Zone (WAMZ).

Membership of ECOWAS has remained stable in recent years with only one new entrant (Cape Verde in 1977) and one withdrawal (Mauritania in 2002).

There are currently 15 members.11 Yang and Gupta (2007) point out that, despite increases, intra-ECOWAS trade has risen erratically relative to trade with RoW and shows no obvious trend over time, except perhaps within WAEMU – the largest subgroup within the region. They argue that this is due to the improved performance of the WAEMU CU – a model which the ECOWAS region is trying to emulate.12 But they also point out that empirical evidence on the effectiveness of regional trade agreements in promoting intra-African trade is limited. Moreover they recognise that although some increases in intra-regional trade flows have been recorded, these are not large enough to confirm that intra-ECOWAS trade has been beneficial to all members. As shown in Tables 2 and 3, Côte d’Ivoire, Nigeria, and Ghana account for most intra-regional trade – in terms of both exports and imports – and their shares have remained fairly constant in recent years.

Table 2: ECOWAS intra-regional exports (% share)

Country Benin Burkina Faso Cape Verde Côte d’Ivoire The Gambia Ghana Guinea Guinea-Bissau Liberia Mali Niger Nigeria Senegal Sierra Leone Togo 1980–84 0.7 2.3 0 34.5 1.4 3.2 0.2 0.1 1.1 4.5 6.1 31.4 10.2 0.1 4.1 1985–89 1 2.3 0 41 0.5 2.3 0.3 0.1 0.4 3.3 2.4 35.4 9.1 0.1 1.9 1990–94 0.9 1.5 0 38.9 0.6 7 0.4 0.1 0.1 0.6 3.4 39.1 5.1 0 2.4 1995–99 0.7 1.4 0 38.8 0.1 4.6 0.3 0 0.2 0.3 2.8 43.3 6.4 0 1.2 2000–4 1.5 1 0 35.4 0 3.1 0.5 0.2 0.3 0.2 2.5 41 8.9 0.2 5.2 2005–9 1.5 1 0 32.2 0 3.3 0.4 0.5 0.2 0.1 1.8 44.7 8.9 0.1 5.1

Source: IMF Direction of Trade Statistics Note: WAMZ members are presented in bold. All other countries except Cape Verde and Liberia are members of WAEMU.

11

12

Benin, Burkina Faso, Cape Verde, Côte d’Ivoire, The Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Niger, Nigeria, Senegal, Sierra Leone and Togo. Besides the establishment of an integrated internal market for goods and services free of tariffs and NTBs, the ECOWAS CU project also requires its member countries to agree on a CET to be applied to imports from outside the region. 8

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

There has been no major change in the level of intra-regional exports in ECOWAS over the period 19901999 compared to 2000-2009. However, in relation to imports, there is a significant increase in intraregional imports over the period 1990-1999 compared to 2000-2009.

Table 3: ECOWAS intra-regional imports (% share)

Country Benin Burkina Faso Cape Verde Côte d’Ivoire The Gambia Ghana Guinea Guinea-Bissau Liberia Mali Niger Nigeria Senegal Sierra Leone Togo 1980–84 2.6 8.5 0.1 15 0.6 25.8 0.6 0.4 1.7 12.1 10.8 4.6 9 4.1 3.9 1985–89 4 9.7 0.2 31.3 0.8 14 1.5 0.5 0.8 9.2 8.8 3 10.6 3 2.7 1990–94 1.4 6.8 0.5 27.7 1.5 17 5.5 0.4 0.9 9.6 4.7 7.2 7.3 6.8 2.8 1995–99 3.9 6.9 0.2 24.9 1 26.7 3.1 0.4 1 10.9 3.5 8.1 5.3 2.3 1.8 2000–4 5 6.8 0.2 23.6 1.9 20.8 3.2 0.9 1.3 9.7 3.9 9.5 9.6 1.9 1.6 2005–9 4.2 6.9 0.2 27 2 23.2 1.8 0.6 1.1 9.1 2.9 12.5 4.6 2 1.8

Source: IMF Direction of Trade Statistics Note: WAMZ members are presented in bold. All other countries except Cape Verde and Liberia are members of WAEMU.

Indicators on regional trade intensity can help to explore whether or not the value of trade between two countries, or in this case, the region with its members, is more or less than would be expected given their exports to the rest of the world. It is calculated as follows: TIij = (xij /Xi)/ (xwj/Xw).

That is, xij and xwj are the values of a country i’s exports to the region, and the rest of the worlds exports to the region, respectively while Xi is country i’s total exports and Xw world total exports. A value that is greater (less) than one indicates a trade flow that is larger (smaller) than expected based on the country i’s importance in world trade. In terms of regional trade intensity for ECOWAS (Table 4), the share of regional exports has increased to a greater extent for some of the economically smaller members such as Togo, Sierra Leone, Senegal, Benin and Guinea-Bissau. However, in most cases regional import intensity has declined which suggests that imports from third party countries have increased to a greater extent than those from the region.13

Table 4: ECOWAS regional trade intensity indices

Country Benin Burkina Faso Cape Verde Côte d’Ivoire The Gambia Ghana Guinea Guinea-Bissau Liberia Mali 1980 0.1 0.7 0.1 0.1 0.4 0.0 0.1 0.0 0.3 1985 0.0 0.2 0.0 0.1 0.2 0.1 0.0 0.2 0.0 0.1 Export intensity 1990 1995 2000 0.4 0.1 0.1 0.2 0.4 0.1 0.0 0.0 0.0 0.3 0.2 0.3 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2005 0.3 0.1 0.0 0.3 0.1 0.1 0.0 0.2 0.0 0.0 2009 0.4 0.2 0.0 0.3 0.1 0.1 0.0 0.5 0.0 0.1 1980 0.1 0.3 0.0 0.0 0.1 0.3 0.1 0.0 0.7 1985 -1.5 0.4 0.0 0.2 0.1 0.4 0.0 0.0 0.1 0.2 Import intensity 1990 1995 2000 -1.5 -2.7 -1.7 0.3 0.3 0.5 0.0 0.0 0.0 0.4 0.2 0.4 0.1 0.3 0.2 0.1 0.0 0.4 0.2 0.2 0.1 0.0 0.0 0.3 0.2 0.2 0.2 0.2 0.0 0.3 2005 -1.9 0.5 0.0 0.4 0.3 0.3 0.1 0.3 0.0 0.3 2009 -3.7 0.5 0.0 0.4 0.2 0.2 0.0 0.2 0.0 0.3

13

Regional import and export intensity figures are calculated based on value data. 9

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Country Niger Nigeria Senegal Sierra Leone Togo 1980 0.1 0.0 0.3 0.0 0.2 1985 0.2 0.0 0.2 0.0 0.1

Export intensity 1990 1995 2000 0.2 0.2 0.9 0.1 0.1 0.0 0.2 0.1 0.3 0.0 0.1 0.0 0.3 0.0 0.5

2005 0.4 0.0 0.5 0.0 2.1

2009 0.4 0.7 0.1 0.7

1980 0.2 0.1 -1.0 0.2

1985 0.2 0.0 0.1 0.2 0.1

Import intensity 1990 1995 2000 0.4 0.3 0.1 0.0 0.1 0.7 0.1 0.0 0.1 0.2 0.1 0.0 0.2 0.2 0.2

2005 0.2 0.0 0.2 0.1 0.1

2009 0.2 0.0 0.1 0.2 0.0

Source: IMF Direction of Trade Statistics Note: WAMZ members are presented in bold. All other countries except Cape Verde and Liberia are members of WAEMU.

1.3.2

SADC

SADC, with a membership of 15 countries,14 was originally conceived as a bulwark against apartheid South Africa and, by extension, the Southern African Customs Union (SACU).15 SACU is not a Common Market, nor an Economic Union, but within it operates a Common Monetary Area (CMA) which comprises Lesotho, Namibia, South Africa and Swaziland (Botswana is not a member).16 All members of SACU are also members of SADC. The 1992 SADC Treaty changed a loose organisation of member state into a legally binding arrangement to facilitate closer economic integration between members and formed part of the continent’s drive to harmonise RECs further to the establishment of the AEC in 1991. The SADC Trade Protocol was established in 1996, and since 2000 members have started to implement their commitments and an FTA was launched in 2008. Some SADC members also belong to other RECs, such as COMESA and the East African Community (EAC).

There are increasing efforts to harmonise membership across these RECs, and between them, institutionally. These efforts are being driven regionally, but with some external pressure applied – most recently brought to bear in light of Economic Partnership Agreement negotiations with the European Union (EU).

This section highlights trends in intra-regional trade flows within the SADC region across all members for which data are available. As shown in Tables 5 and 6, the main regional driver of trade within SADC, in terms of both exports and imports, is South Africa. Its regional importance is much more pronounced as a source of other members’ imports than as a destination for their exports. Angola is a lesser, but increasingly important regional partner in terms of its imports as well as its exports (which are concentrated in fuels); so too are the landlocked countries Zambia and Zimbabwe, whose shares of exports and imports increased steadily over the period 1980–2009.

14

15

16

Angola, Botswana, Democratic Republic of Congo (DRC), Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, Swaziland, South Africa, Tanzania, Zambia, and Zimbabwe. It was created in 1910 as a revenue sharing mechanism for the tariffs paid on imports which had to travel across the borders of the land-locked countries Botswana, Lesotho and Swaziland from ports in South Africa. The agreement of sovereign states was signed in 1969 and entered into force in 1970, replacing the 1910 arrangement between Botswana, Lesotho, South Africa, and Swaziland (Namibia became the fifth member in 1990).The agreement was most recently revised in 2002 in order to respond to the demands of its members related to revenue sharing mechanisms. Namibia officially joined in 1992, but had previously been a de facto member as a territory administered by South Africa, as noted in its most recent Trade Policy Review (2009).

Although Botswana is not a member of the CMA it maintains a crawling band system based on a basket of the South African rand and the IMF’s Special Drawing Rights. 10

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Table 5: SADC intra-regional exports (% share)

Country Angola Botswana DRC Lesotho Madagascar Malawi Mauritius Mozambique Namibia Seychelles Swaziland South Africa Tanzania Zambia Zimbabwe 1980–84 0.9 8.4 6.4 0.2 0.1 5.9 0.8 5.3 0.4 0.4 1.5 48.3 2.6 9.6 10.8 1985–89 1.1 15 5.9 0.3 0.7 6.8 1.5 9.9 0.2 0.4 2.7 37.9 2.8 12.5 5 1990–94 1.1 12.4 3.6 0.2 1.9 9.5 2 7.6 1.1 0.8 6 39.6 2.5 9.3 8.3 1995–99 3.6 2.8 4.2 0 3.2 8.6 6 13 0.8 0.8 2.1 23.4 4.6 11.1 18 2000–4 7 1 4.4 0 3.2 6.8 6.1 14.8 0.4 0.9 1.1 18.3 6.1 13.8 17.2 2005–9 6.6 1.6 9.1 0 2.2 4.4 3.4 12.1 0.4 0.8 0.8 25.2 5.1 14.4 14.7

Source: IMF Direction of Trade Statistics Note: Bold = member of COMESA; Italic = member of SACU; Underlined = member of EAC.

Despite recent growth in intra-regional exports for SADC, there has been no significant change in the level over the period 1990-1999 compared to 2000-2009. This is also the case for intra-regional imports, for the same period.

Table 6: SADC intra-regional imports (% share)

Country 1980–84 1985–89 1990–94 1995–99 2000–4 Angola 0 0 0 0.2 1.3 Botswana 4 8.1 3 1.1 1.2 DRC 0.2 0.4 1.2 0.2 0.7 Lesotho 0 0 0 0 0 Madagascar 0.1 0.5 1.3 0.8 0.8 Malawi 2.6 1.4 0.6 1.7 1.6 Mauritius 0.2 0.4 0.8 1.9 1.6 Mozambique 2.2 0.3 0.5 1.2 3.2 Namibia 0 0 0.2 0.3 0.3 Seychelles 0.1 0.1 0 0 0.4 Swaziland 0.2 0.1 0.1 0 0 South Africa 74.8 69.3 80 80.7 72.4 Tanzania 0.9 0.9 0.6 0.9 1.6 Zambia 5.4 4.2 2 3 5.2 Zimbabwe 9.3 14.4 9.8 7.9 9.9 Source: IMF Direction of Trade Statistics Note: Bold = member of COMESA; Italic = member of SACU; Underlined = member of EAC. 2005–9 12.3 1.1 1.3 0.1 0.3 1.6 1.4 3.2 0.3 0.4 0 58.9 1.8 8.9 8.5

In terms of regional trade intensity, as expected the landlocked countries of the region such as Zambia and Zimbabwe have a high degree of dependence on regional partners for both exports and imports. It may be the case that goods in transit may be recorded as if they are destined or originating from some member countries particularly South Africa (which has many major ports serving landlocked countries).

The region generally is more important as a destination for its members’ exports than as a source of their imports (Table 7).

However, contrary to trends apparent in ECOWAS, the levels of imports sourced from regional partners have held up in recent years compared to RoW (as indicated by regional import trade intensity).

Zimbabwe, Malawi and Mozambique have the highest regional export intensity. Zambia, Zimbabwe and Mozambique have the highest regional import intensity.

11

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Table 7: SADC regional trade intensity indices

Country Angola Botswana DRC Lesotho Madagascar Malawi Mauritius Mozambique Namibia Seychelles Swaziland South Africa Tanzania Zambia Zimbabwe 1980 -1.0 0.0 0.0 0.0 0.0 -1.0 0.2 0.0 0.0 0.0 -1.0 1985 0.0 0.0 0.0 0.1 0.0 0.2 0.0 0.0 0.0 0.1 0.0 Export intensity 1990 1995 2000 0.0 0.0 0.03 0.0 0.0 0.1 0.0 0.1 0.2 0.2 0.2 1.5 b 0.0 0.0 0.3 0.2 0.5 -1.9 0.4 0.2 0.5 0.0 0.02 0.02 0.0 0.1 0.5 0.2 0.1 0.1 0.0 1.5 b 0.3 2005 0.03 0.4 0.3 2.2 0.2 1.6 0.4 0.03 0.5 1.2 3.5 2009 0.02 1.2 0.2 2.1 0.3 1.6 a 0.3 0.04 0.5 1.4 3.0 1980 0.0 0.0 0.1 0.0 -1.0 0.0 0.0 0.0 0.0 1985 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Import intensity 1990 1995 2000 0.0 0.0 0.0 0.1 0.0 0.0 0.2 0.0 0.1 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.1 0.1 0.1 0.0 0.0 0.1 0.0 0.1 0.1 0.1 0.1 0.2 2005 0.0 0.1 0.0 0.3 0.0 0.1 0.0 0.1 0.1 1.0 0.3 2009 0.1 0.1 0.0 0.1 0.1 0.2 a 0.1 0.1 0.0 0.3 0.2

Source: IMF Direction of Trade Statistics Note: Bold = member of COMESA; Italic = member of SACU; Underlined = member of EAC; (a) 2008 data; (b) 2001 data

1.4

Potential for intra-regional trade – structural analysis

There are two main areas in which to look in order to try to explain variations in intra-regional trade flows. The first relates to factor endowments and the propensity to import and export as revealed by comparative advantages. Its calculation requires trade data only; the trade flows reported themselves reveal countries’ specialisation patterns and hence comparative advantage, though not the source of this advantage.17 The second focuses more on the policy environment and the extent to which trade flows are influenced by tariffs and NTBs. This section focuses on the first area. It explores the structural characteristics of trade flows for ECOWAS and SADC. It does this by making use of indices for identifying trade complementarities and comparative advantages. The following section focuses on the second area and explores the impact of policy barriers such as NTBs on intra-regional trade flows, in much more detail. The approach to assessing the potential for intra-regional trade first, and then by extension how this potential may be constrained by the imposition of NTBs, could be considered similar to the export diversification approach.18 As previously discussed, traditional and newer contributions to theory suggest that low income countries should seek to form RTAs with other countries whose export baskets are more diversified than their own in order to have more success in their efforts to maintain and benefit from such agreements. The literature is much less sanguine in relation to RTAs between low income economies which have very similar export bases, unless the potential for intra-industry trade exists.19

17

18

19

As de Benedictis and Tamberi (2001) argue, the greater independence of measurement from theory implies more freedom in the selection of the specific index of RCA to use in applied research on international specialisation issues, but it also requires greater awareness of the implications of that selection. See Yang and Gupta (2007).

Also Yeats (1998) who, with reference to Mercosur, argues that if members of an RTA have to rely heavily on third countries for a high share of their key imports, and as destinations for their major exports, this may reduce their commitment to the RTA. It is important to note, however, while the theoretical literature on regionalism is well developed, the empirical literature is still maturing (see Freund and Ornelas, 2010).

12

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Some have argued that because exports from countries in SSA tend to be highly concentrated in a few primary products, many of which are not important imports in other African countries, that this limits the potential gains from any RTA between them. Others argue that despite similar export baskets, regions such as ECOWAS, and indeed SADC, possess large enough regional markets for domestic producers to utilise and subsequently launch out from as a strong competitive force to other regions of the world. As in other parts of the developing world, the experience of West Africa with formal regional integration has been largely driven by the desire to overcome the constraint of small economic size, which can hamper countries’ ability to industrialise efficiently, by extending the logic of protected and state-led economic development to a larger number of countries (Aryeteey, 2002).

As argued by Foroutan (1993) simply saying that removing barriers to trade, including non-tariff barriers, will not have any effect if partners do not demand each other products misses the potential for trade based on economies of scale, product differentiation and intra-industry trade. Moreover, ignores the objectives of fostering closer economic integration: to alter existing patterns of production by taking advantage of a larger market, fostering development of new industries, relocating others and increasing the output of existing ones. We first explore the RCAs of each region and their individual members and attempt to reconcile with recent trends in intra-regional trade flows in ECOWAS and SADC. To capture the degree of trade specialisation of a country Balassa (1965) suggested the following index of RCA. RCA = (xij/Xi)/(xaj/Xa) xij = exports of product j from country i. Xi = total exports from country i. xa i= total exports of product j from the reference area (e.g. the world).

Xa = total exports from the reference area. A country, or region, reveals its comparative advantage in specific products when the RCA indicator scores higher than 1; this shows that its exports of a specific product are more than expected on the basis of comparison to the reference area, usually the rest of the world. Despite recent extensions to the RCA indicator, for example by Hausmann and Klinger (2006), this section focuses on the building blocks of approaches and simply applies the index developed by Balassa (1965) to trade flows within the selected regions and for their members.20

1.4.1

RCA and trade complementarity in ECOWAS

The first step in the structural analysis of trade flows from the ECOWAS region was to explore growth in and across product categories at the HS 2-digit level. The full results for the region are presented in Appendix 3, which shows how the composition of the region’s top exports and imports has remained steady in recent years, with no change in patterns discernible. The total value of exports from ECOWAS to RoW has increased in recent years, but this growth has mostly been accounted for by increases in the value of existing exports, such as mineral fuels and oils21 which are dominated by flows from Nigeria (Figure C,

20

21

Vollrath (1991) extended the RCA indicator and applied the same formula to imports. As summarised by Shirotori et al. (2010), he also amended the formula in two further ways; the first in order to avoid double counting, and the second so as to encompass both import and export aspects of comparative advantage. The indicator has also recently been extended further by the product proximity literature (Hausmann and Klinger, 2006).

It is beyond the scope of this study to review these two measures in more detail with respect to the selected regions and countries. See Figure 2. 13

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Appendix 3).

Levels of intra-regional trade, in comparison, have remained stable but low (Figure D, Appendix 3).

The results of region’s RCA for all products identified with an RCA of more than one in 2008 are presented in Table 8 for the period 2000–2008. In terms of the region’s RCA compared to rest of the world, it is highest for cocoa and mineral fuels – the region’s top exports. But cotton, a ‘top ten’ import also features.

Table 8: RCA indicators for ECOWAS and rest of the world

H S 01 08 13 14 18 27 33 41 46 52 71 Description Live animals Edible fruit and nuts Lac; gums, resins Vegetable plaiting materials Cocoa and cocoa preparations Mineral fuels, oils Essential oils & resin oils Raw hides and skins Manufactures of straw, other plaiting materials Cotton Precious/semi precious stones/metals 2000 1.3 1.4 0.2 3.1 21.2 7.7 0.5 0.1 1.2 3.8 1.5 2001 2.9 1.6 0.2 0.7 29.5 7.3 0.6 0.3 3.1 4.0 2.7 2002 2.0 1.5 0.3 0.1 36.3 7.2 0.7 0.2 1.2 4.8 1.7 2003 2.0 1.7 0.2 0.6 34.3 6.8 0.5 0.2 1.2 6.0 2.2 2004 6.3 6.0 0.6 4.5 108.3 1.1 2.0 0.3 3.6 18.3 3.0 2005 4.6 4.8 0.5 7.0 91.4 1.2 1.6 0.3 4.2 14.2 7.0 2006 1.0 1.4 7.8 1.6 21.7 5.8 0.4 0.5 0.6 2.6 1.7 2007 1.3 1.4 3.6 1.1 25.0 5.8 0.5 2.2 1.0 1.8 2.1 2008 1.3 1.2 2.0 1.6 19.7 5.3 1.1 3.7 1.0 1.4 1.8

Source: Derived from data obtained from UN COMTRADE database. Note: Based on available data.

The RCAs for each member of ECOWAS at the HS 2-digit level are identified in Table C, Appendix 3. The results for ECOWAS suggest that RCAs tend to be clustered amongst traditional commodity exports such as tobacco, salt, ores, slag, ash and precious metals. These products are generally considered to embody less of a potential to sustain trade dynamism in the form of technology, learning and other spillover effects. More non-traditional, or newer, products such as edible vegetables, fruit and fish also feature as common sources of advantage across members. Although such goods may be ‘newer’ than other more traditional commodities, they are generally considered to be less sophisticated compared to other types of light manufactures with less of a potential to sustain trade dynamism. But more sophisticated than other types of traditional commodities, for example, because they are supplied to customers ‘just-in-time’ and must comply with more stringent product and process standards. Trade complementarities The trade complementarity index is a measure of potential trade between two partners.22 It compares the export basket of country k to the import basket of country j and is constructed as follows: TCij = 100*(1- ∑ (|mik – xij| / 2))

i =1

Where xij is the share of good i in global exports of country j and mik is the share of good i in all imports of country k. Positive values up to 100 indicate the extent to which the exports of country i match the imports of country j; when values are less than zero and negative, export and import shares differ greatly, which suggests a limited potential for intra-regional trade, based on the logic of this index. Results for ECOWAS and its sub-regions WAMZ and WAEMU have recently been calculated by Goretti and Weisfeld (2008) and are summarised in Table 9.23 The complementarities between WAEMU countries as

22

See Shirotori and Molina (2009).

14

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

exporters and other ECOWAS countries as importers are as high as 29% – which exceeds the reference point of 25% considered to indicate a strong potential for enhancing trade. These results suggest that the trade complementarities between ECOWAS countries are substantial. Thus, even though ECOWAS members’ RCAs for exported goods tend to be clustered within similar products based on similar factor endowments, export baskets differ enough from other members’ imports to suggest the potential for intraregional trade according to logic which underpins the index.

Table 9: Trade complementarities index in ECOWAS

Source: Goretti and Weisfeld (2008).

Analysis of the trade complementarities of the sub-region of WAEMU by Goretti and Weisfeld (2008) also finds that with production patterns unchanged, the current level of intra-regional trade could be expanded marginally: from 11% of total trade to about 14%. They show how trade complementarities are highest for those economies with more diversified economic bases, such as Côte d’Ivoire and Senegal (also the economically larger members of ECOWAS; see Table B, Appendix 2).

But less for the economically smaller members of ECOWAS, such as Benin, Niger, and Burkina Faso, that have average export complementarities indices of less than 25%.

1.4.2

RCA and trade complementarities in SADC

Growth in and across product categories at the HS 2-digit level is presented for those members of SADC for which trade data are available in Appendix 4.24 It is clear that the export and import baskets of this region are more diversified than is the case for ECOWAS. Unlike ECOWAS, SADC is a large mineral fuels and oil importer; other important imports include general intermediate goods such as electronic machinery and plastics. Since 2000 the region’s exports have grown rapidly in terms of value (Figure G, Appendix 4).

This growth has occurred fairly evenly across the region’s top ten exports. Intra-regional trade flows have also increased

23

24

Goretti and Weisfeld (2008) use the trade complementarity formula devised by Michaely (1996).

This differs from that referred to by Shirotori and Molina (2009) in that it is not possible to obtain a figure of less than zero (zero means no products exported by country i are imported by country j).

In comparison, the inclusion of 1- by Shirotori and Molina indicates the use of absolute values; negative values therefore indicate the extent to which exports by country i do not match the imports of county j. Unfortunately, this analysis excludes Angola and DRC because no trade data have been reported from 2000 to date – the period under analysis. 15

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

over the period 2000–8 (Figure H, Appendix 4).

Exports to the world are dominated by South Africa, which in 2008 accounted for 74% of the region’s reported total exports in value terms (Table D, Appendix 4).25 The SADC region has an RCA compared to rest of the world in a large number of products. Table 10 below presents the top ten product groups in which the region has been identified as having an RCA, some of which, such as live trees and other plants, edible fruit and nuts, tobacco, aircraft and parts, differ from the region’s top ten exports (Table F, Appendix 4, presents other high-scoring products).

Table 10: RCA indicator for SADC and rest of the world

HS 06 08 24 26 36 71 74 75 86 88 Description Live tree & other plants; bulb, root Edible fruit and nuts Tobacco/manuf. tobacco substitutes Ores, slag and ash Explosives; pyrotechnic prod; matches Precious/semi precious stones/metals Copper and articles thereof Nickel and articles thereof Railway locomotives/rolling stock, parts Aircraft, spacecraft, and parts thereof 2000 1.9 4.5 8.4 8.2 7.7 8.6 3.7 3.5 10.7 2.8 2001 1.3 3.5 7.9 7.5 3.3 12.1 3.5 3.8 9.4 2.8 2002 2.2 3.9 5.0 9.2 4.2 7.8 3.7 4.5 10.3 3.5 2003 1.0 4.2 3.2 7.0 4.2 10.9 4. 1.5 8.2 4.2 2004 1.2 4.7 5.9 6.2 5.2 10.7 3.1 7.6 7.6 4.1 2005 1.4 4.4 5.4 6.0 3.8 11.9 3.7 2.5 6.6 4.6 2006 7.8 3.7 6.4 6.4 3.3 10.6 4.3 2.9 8.6 3.9 2007 2.4 3.9 5.0 7.0 4.0 10.0 4.2 5.3 8.0 4.1 2008 3.3 3.9 5.9 9.6 5.2 8.4 4.0 5.5 7.4 5.6 2009 6.9 5.3 4.8 10.7 4.6 5.8 6.9 4.0 8.8 5.6

Source: Derived from data obtained from UN COMTRADE database. Note: Based on available data.

The results for each member country are presented in Table G, Appendix 4. RCAs for SADC member countries tend to be clustered amongst similar products such as tobacco, salt, and ores, slag and ash. But articles of apparel feature, as do other types of light manufactured goods such as automotive parts. These products are generally considered to embody a greater potential for learning, technology and other spillover effects, if properly harnessed. Moreover, such products tend to be traded within production networks and as intermediate goods; they may therefore feature in countries’ imports and exports. Trade complementarities indices constructed for each member of SADC for which data permit are presented in Table 11. Despite the region having an RCA in more technologically sophisticated products compared to ECOWAS, the results of analysis of the complementarities of trade structures and therefore the potential for increased intra-regional trade in SADC appear less promising. The results suggest that only in the case of South Africa and, to a lesser extent Mozambique is the composition of exports conducive to increased intra-regional trade. This is because the products exported by these two countries are more similar to members’ imports, which suggest that the potential trade between these countries and the rest of the region is high. However, for other members this is not the case because their export and import baskets differ greatly from each other, which suggests limited potential for increased intra-regional trade according to the logic which underpins the index.

25

This excludes Lesotho and Swaziland. 16

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Table 11: Trade complementarities index in SADC

Mozambique South Africa Madagascar Zimbabwe -50 1 -7 -16 -4 2 -6 -17 -25 -31 -24 -26 Swaziland Seychelles Mauritius Botswana Tanzania Namibia

Lesotho

Botswana Lesotho Madagascar Malawi Mauritius Mozambique Namibia Seychelles Swaziland South Africa Tanzania Zambia Zimbabwe

-56 -50 -35 -61 -46 -53 -43 -52 -50 -59 -55 -53 -50 -60 -74 -59 -74 -64 -76 -61 -70 -77 -79 -77

-35 19 -41 -4 -30 -28 -19 -9 -45 -48 -54 -25

-61 -36 -59 -48 -57 -59 -66 -37 -72 -71 -78 -51

-46 -5 -41 -56 -54 -36 -45 -32 -52 -59 -62 -51

-53 15 -18 -9 16 6 16 22 6 7 3 11

-43 -18 -21 -21 9 -8 11 -8 -16 -13 -8 -12

-52 -66 -78 -79 -55 -76 -70 -69 -81 -82 -83 -70

-50 -34 -43 -2 -36 -49 -31 -48 -44 -52 -35 -39

-59 12 -78 22 32 43 40 37 29 46 39 36

-55 6 -17 -2 14 2 -4 -5 13 -14 -23 -6

-53 -36 -3 -37 32 -32 -25 -24 -30 -35 -31 -23

Source: Derived from data obtained from UN COMTRADE database. Note: Data for Lesotho are 2004; Swaziland, 2006; Malawi, Namibia and Seychelles, 2008, all others are 2009.

It is important to note the limitations of using such a high level of aggregation (HS 2-digit level) to assess RCAs and how these are similar or differ across economies in each region. In addition to the constraints analysis at such a high level of aggregation imposes on analysis of trade complementarities: it means that differences between production structures at a much lower level of aggregation are not captured, when analysis at such levels may be needed to better assess the potential or not for intra-industry trade. Both indicators assume that all other variables remain constant, such as demand in world markets, and related policy. Even though a country may have a high (low) revealed comparative advantage index compared to the rest of the world, this should not be automatically interpreted as being beneficial (weak) without more in-depth, country and industry specific analysis. As the earlier discussion on theory suggested, it may be more beneficial for countries to seek to trade against their comparative advantages rather than with it.

17

Zambia

Malawi

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

2

The impact of non-tariff barriers on intraregional trade

Section 1 presented and discussed recent trends and the structure of trade in SSA in general, and within ECOWAS and SADC, in particular. For these selected regions it explored what these patterns suggest in terms of the potential for intra-regional trade. This section sets out to identify the extent to which intraregional trade flows within these regions are constrained by NTBs. It first introduces the different types of NTBs. It makes a clear distinction between NTBs and NTMs, the former is defined as an unnecessarily restrictive non-tariff measure (NTM) which affects trade in goods; the latter may affect trade in goods, but may be reflective of legitimate public policy purposes. It argues that key to making this distinction is an understanding of both the intent of the measure and resultant impacts on trade. The following sub-section introduces our methodological approach to quantitatively assessing the impact of reported NTBs on intraregional trade for the selected regions.

2.1

Types and potential trade impacts of non-tariff barrier

Non-tariff barriers to trade may have similar effects to tariffs: they can increase domestic prices and impede trade to protect selected producers at the expense of other economic agents; they may also tax exports. Bhagwati (1965) has shown how both tariffs and NTBs can have equivalent effects when markets are competitive and therefore how the removal or reduction of NTBs can have similar effects to that of tariff reduction. Tariffs increase the costs for foreign suppliers while quotas and other types of NTBs serve to restrict the quantity of foreign-supplied goods in domestic markets; both may cause prices to increase in the domestic market. This in turn results in a decrease in economic welfare because of the distortion or wedge created between domestic and world market prices. It therefore follows that the removal or reduction of NTBs on imports could, as in the case of tariff liberalisation, increase imports and therefore impact welfare through effects on local producers, domestic consumers and government revenues. These effects can be summarised as follows:

 Increased imports may displace domestic producers by foreign suppliers, depending on the

assumed elasticity of substitution between imported and domestically produced goods;

 Consumers (and producers using imported inputs) may benefit from cheaper product prices;

and

 Governments may lose revenues for the product liberalised, e.g. revenues from quota auctions

or licences. Deardorff (1987) notes that NTBs are preferred by policy makers because their effects are are more certain, direct and predictable than the effects of tariffs – competitors cannot overcome them easily. NTBs are usually held steady whilst market conditions change and are therefore rigid control measures. 26 Quotas are more palatable for voters given that they are not directly associated with a price increase. Governments may choose to exercise considerable discretionary power through the allocation of quotas, with divergent outcomes on the welfare of consumers and producers. These discretionary powers are kept in check through membership of multilateral institutions such as the WTO as well as other regional and bilateral trade regimes.

26

The impact of a specific import duty compared to an ad valorem tariff can be vastly different if import prices fluctuate. It is also important to make the distinction between free on board (FOB) and cost, insurance and freight (CIF) values. 18

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

Based on the General Agreement on Trade and Tariffs (GATT), the definition of NTBs used by all WTO members includes export restraints, production and export subsidies, or measures with similar effect, not just import restraints. Other definitions of NTBs include any measure, public or private, that causes internationally traded goods and services, or resources devoted to the production of these goods and services, to be allocated in such a way as to reduce potential and real world income (Baldwin, 1970).

This is clearly an extremely broad definition and obscures the fact that some NTMs may be used for legitimate public policy objectives, for example, in relation to health and safety concerns (SPS) and other technical regulations (TBT).

Making the distinction between the intent and impact of a NTM is therefore crucial to determining the extent to which legitimate measures may serve as unnecessarily restrictive barriers to trade. That is, when a NTM serves to be a NTB.

2.1.1

Types of non-tariff barrier

Five types of policy barriers may be identified from the literature.27

 The first consists of quantitative restrictions (QRs) such as quotas and licences, which limit the

volume or value of goods allowed across borders.

 The second consists of non-tariff charges on imports, such as variable levies, border tax

adjustments or countervailing duties.

 The third includes government participation in trade, such as the use of subsidies,

procurement policies and competition policy.

 Fourth are customs procedures, clearance, and classification procedures.  The last category includes technical barriers to trade (TBT), such as packaging and labelling and

health and sanitary regulations. These barriers may be generic or product specific. For example, the former category could include NTMs that all goods must adhere to, such as customs, administrative entry and passage procedures, but which are implemented in a way that makes them unnecessarily trade restrictive. In comparison product specific barriers may result from the application of technical quality standards which on application tend to be more trade restrictive than facilitative. Even though measures may be generic, or product specific, in terms of their application, the consequent impact of them on trade flows, and as a result economic welfare, may differ markedly. For example, a direct quantitative restriction (QR) which covers whole industry will cause the import demand curve to become vertical at the permitted quantity. In comparison, an industrial standard which adds a fixed cost to each unit of a good imported may cause the demand curve to turn downward and become steeper.28 This means that in addition to the consideration of price and quantity effects, the responsiveness of demand and supply as indicated by product and income elasticities becomes important: the extent of responsiveness may differ across markets, producers and consumers, even if the reported NTB is the same in each. This is because NTBs may have much more uncertain as well as highly variable impacts on prices and quantities and therefore what is demanded as well as supplied. Additional uncertainty results from the very nature of the imposition of NTBs: how does one know when an import surge may result which warrants the use of an NTB?29 Because of these variable affects, some authors prefer to classify NTBs

27 28

The following categories are used in the UNCTAD TRAINS database. See Deardorff and Stern (1998), 29 This is known as the endogeneity problem, which we try to address in the empirical analysis we undertake. 19

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

according to their intent and impact.30 This means that once NTBs have been identified by traders and defined, the challenge is to measure their impact on trade so as to analyse the welfare impact of their removal, reduction or harmonisation.

2.1.2

Measures of non-tariff barriers

Approaches to measuring NTBs range from: frequency-type inventories, based on counts of observed NTMs in particular countries, sectors, and types of trade; price comparison measures (tariff equivalents); and quantity impact measures based on estimation of trade flows in the absence of the measure.31 Each approach has its own drawbacks.32 Here we briefly summarise the pros and cons of using different frequency type inventories of NTBs in the analysis of welfare impacts.33 Measures of frequency UNCTAD’s TRAINS database uses a classification of over 100 trade measures, including those with a discretionary or variable component. It contains NTBs reported for over 150 countries from 1988 to 2001. Since 2006 it has included data on anti-dumping measures. The incidence of reported NTBs, either as a count or as a percentage of coverage of specific product lines (HS or Standard International Trade Classification (SITC)), is categorised as follows:

 Price control measures, such as multiple exchange rates, or foreign exchange allocation;  Finance control measures, such as anti-dumping or countervailing measures, relating to credit

allocations;

 Quantity restrictions, such as non-automatic licensing, quotas;  Monopolistic measures;  Technical Measures, such as regulations and customs procedures; and  Miscellaneous, such as subsidies.

The database does not include any measures related to:

 Corruption;  Export related measures;  Government procurement;  Intellectual property rights; or  Other investment related measures.

This database which is country as well as commodity specific is one of the most detailed available. However, like most inventories, it is only as good as the data that are provided to it. The main source of information used in the database is taken from GATT notifications and other government publications, as

30 31 32 33

See Laird and Vossenaar (1991).

Deardorff and Stern (1998) also make reference to measures of equivalent nominal rates of assistance. See Bora et al. (2002) for a more detailed overview of these. Box 1 Appendix summarises some of the other approaches. 20

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

well as WTO TPRs. Although it is able to register a fairly wide range of NTBs, most of those listed fall within the Technical Measures category34. Other sources similar to UNCTAD TRAINS, which also take an ‘inventory’ or frequency approach to measuring the extent of NTBs, include the World Bank’s Trade Restrictiveness Index (TRI), which seeks to measure the effectiveness of protection. It is constructed based on UNCTAD TRAINS data and includes four types of NTBs: 1. Quantitative Restrictions (QRs); 2.Voluntary Export Restraints (VERs); 3.Enforcement of decreased prices; and 4.Tariff quotas. The difficulty with using frequency measures is that although they provide coverage on a range of restrictions, they are not able to capture their impacts or show differences in the intensity of their application. They c1an therefore only be used as a measure of the extent to which products (or countries) are subject to NTBs. They cannot be used to capture scale and/or growth effects. These limitations will obviously play out in our empirical analysis, which tries to quantify the impact of NTBs using TRAINS data. Bearing these caveats in mind, these data still allow us to analyse systematically the impact of NTBs on imports. More importantly, by comparing the effect of the same measure on different exporting countries we can largely avoid the problem of heterogeneity across measures and sectors.

2.1.3

Approaches to analysis of welfare impacts of NTBs

As the previous discussion has highlighted, NTBs may consist of several parameters, information about them is hard to collect and they are not straightforwardly quantifiable. This means that their economic impact is not easy to model (see Fugazza and Maur, 2008), and creates a number of methodological challenges for an empirical exercise which seeks to measure their trade impact. Despite these challenges, estimates as to the potential welfare impacts of NTBs can still help to provide information to policy makers to act to reduce or mitigate any negative impacts on producers or consumers that may result from their imposition. How do NTBs interact with intra-regional trade? Is there any differential impact of NTBs on regional partners vis-à-vis third parties? This information is particularly relevant for those regions seeking to foster deeper economic integration and enhance intra-regional trade flows. According to Deardorff and Stern (1998), the one method of empirical analysis applicable to any kind of NTB is time-series analysis of the periods in which the NTB is in place, combined with observation of changes in prices or quantities of imports at the time of implementation. However, they also note that unless the implementation of the NTB comes as a complete surprise to the public, it is likely to have effects – perhaps perverse ones – long before it is put formally in place. Moreover, if some other event happens to affect trade simultaneously with the NTB, then this approach may give misleading information unless the importance of that other event is correctly diagnosed. We try to take account of some of these methodological concerns in the empirical analysis which follows in the next section which seeks to quantitatively assess the extent to which reported NTBs to UNCTAD TRAINs actually impact intra-regional trade.

2.2

Quantitative assessment of the effects of NTBs on intra-regional SSA trade flows: SADC

While NTBs are widely perceived to be an important constraint to trade in SSA, there have been limited attempts so far to systematically quantify their actual trade impact. For example, Kee et al. (2008) use UNCTAD TRAINS data to compute indices of trade restrictiveness for NTBs across countries. However as these indices are computed on the basis of the actual effects that these measures have on trade on one year for each country (which differs across countries), using them to estimate the effects of NTBs on trade

34

In particular, the use of such indices would make the estimation endogenous.

21

Impediments to Intra-Regional Trade in Sub-Saharan Africa – Prepared for the Commonwealth Secretariat

would lead to biased estimations.35 We take a different route and undertake a systematic evaluation of the actual impact of NTBs on the imports of selected SADC countries (HS 6-digit level) over a number of recent years. However, data limitations mean that our analysis is limited to the SADC region. The idea is to perform a direct test by matching the NTBs recorded by UNCTAD, with corresponding import data. Because the NTBs reported to UNCTAD TRAINS have not actually been assessed in terms of their trade restrictiveness, and the evidence on the extent to which they actually constrain trade is not available, we use the term NTM to refer to them, until such time as we have the data to classify them as actual barriers to trade – NTBs. We test whether imports in a sector where one or more NTMs are imposed at year t perform worse than those in sectors which do not experience an increase in NTMs, as well as whether they perform worse than in those periods when the NTM was not applied. It is worth noting that NTMs in our dataset are applied on products from all sources. That is due to the nature of NTMs, whose objective for instance is to protect consumers from possible animal diseases, thus making the selective application of an NTM according to the source of imports uncommon. More formally, we employ the following expression:

I ijst  ijs  t  ntmist   ijst

(1)

Where ΔI is the percentage change in imports of country i from exporting country j in sector s at time t; α is importer-sector-exporter fixed effects and λ is time effects; and ntm is a variable that captures non tariff measures (see below).

The basic hypothesis we are testing is that β