Executive summary

African Agricultural Trade Status Report

2017

Executive summary

To maximise the benefits of regional integration and look for new opportunities

to improve competitiveness, African policymakers, the private sector and

development partners need access to accurate and comprehensive data on intra

and inter-regional trade with respect to agricultural goods. It is in this context

that the ACP-EU Technical Centre for Agricultural and Rural Cooperation

(CTA) and the International Food Policy Research Institute (IFPRI)

commissioned the African Agricultural Trade Status Report, which examines

the current status, trends and outlook in African trade performance, making an

important contribution towards data and analysis of developments both at

regional and at continental levels.

The Report builds on the work by the African Growth and Development Policy

Modelling Consortium (AGRODEP) and the Regional Strategic Analysis and

Knowledge Support System (ReSAKSS) of CAADP and trade and also reflects

the CTA’s commitment to advancing knowledge and sharing of best practices

relating to agricultural trade.

In addition to accurate data to assist policy-makers to take informed decisions,

this collaboration aims at maximising the input from the highest African

analytical capacity on agricultural trade and strengthen an African pool of

expertise through AGRODEP.

Regional trade within Africa and between the various regions will offer the

biggest opportunities in the near future for the local private sector, SMEs and

producers and value chain actors. In this context, CTA and IFPRI believe that

an annual African trade report is needed and that for the next editions, a broader

range of partners would join this initiative.

Trade provides the potential for improving consumer welfare and producer incomes, boosting

overall economic growth, and reducing poverty. In Africa, increased and more diversified

agricultural trade on the global and regional levels could provide leverage for efforts to raise

productivity at all stages of the value chain, and facilitate the transformation of agriculture into a

high-productivity sector providing adequate incomes for producers and stimulating growth

throughout the economy. Increasing agricultural trade also has the potential to improve food

security and contribute to stabilizing local and regional food markets by making them less

vulnerable to shocks.

In addition to the benefits of global trade, intra-regional trade has increasingly been recognized as

a key element of efforts to increase food security and agricultural development in Africa. The 18th

African Union Summit in 2012 was organized under the theme of “Boosting Intra-African Trade.”

In 2014, African leaders committed to tripling intra-African trade in agricultural commodities and

services by 2025, as one of a limited number of commitments in the Malabo Declaration on

Accelerated Agricultural Growth and Transformation for Shared Prosperity and Improved

Livelihoods. The trade commitment included accelerating the establishment of a Continental Free

Trade Area and a continental Common External Tariff and taking measures to increase investments

in trade infrastructure and enhance Africa’s position in international trade negotiations.

Despite longstanding recognition of the benefits of trade and the importance of improving Africa’s

competitiveness, the continent is performing beneath its potential in global and regional

agricultural markets. Recent increases in exports have been offset by even larger growth in imports,

leading to a deterioration in Africa’s trade balance. Intra-regional trade in Africa is growing, but

remains significantly below the levels seen in other regions. These challenges result from a host

of factors, including historical trends and more recent developments inside and outside of Africa.

Action on many fronts is needed to remove constraints to improving the competitiveness of

Africa’s producers.

Highlights

The African Agricultural Trade Status Report (TSR) provides detailed descriptive assessments of

the current status and recent trends in Africa’s trade performance and competitiveness at the

continental and regional levels, as well as more in-depth investigations of the determinants of trade

performance and the relative importance of different drivers and constraints. The goal of the report

is to provide comprehensive and timely evidence and analysis on the status of African trade in

order to inform policy discussions on measures to enhance trade performance at the global and

regional level. In addition to the introductory and concluding chapters, the report is divided into

five chapters presenting findings on Africa’s trade performance and outlook.

Chapter two reviews trends and patterns in Africa’s global agricultural trade since 1998. The

chapter finds that although agricultural exports more than doubled between 1998 and 2013,

imports increased fivefold, leading to a growing trade deficit. The main drivers of this surge in

imports are rapid population growth and urbanisation, income changes due to economic growth,

and changes in dietary patterns. Among the major Regional Economic Communities (RECs), only

the SADC region has maintained a consistent trade surplus over the last decade.

The chapter finds that despite the increase in agricultural exports, the share of agricultural exports

in Africa’s total exports has declined by half over the period, due to more rapidly rising exports in

minerals and oil. Africa’s agricultural exports show signs of moderate diversification over the

period, while imports have remained fairly stable. The EU remains Africa’s top trading partner,

but both imports from and exports to the EU have dropped over the period, while trade with Asia

has doubled; Asia is likely to take the EU’s place as Africa’s top trading partner if these trends

continue. Recent efforts to pursue increased economic integration have resulted in significantly

increased intra-regional trade during the period, although the overall level of intra-regional trade

remains low.

Chapter three examines patterns in intra-regional trade at the continental level and among major

RECs, namely ECOWAS, ECCAS, COMESA, and SADC. The chapter finds that intra-African

agricultural trade has expanded significantly since 1998, increasing at about 12 percent per year in

value terms. However, the share of intra-African trade in total African trade is still very low

compared to other regions or continents. For example, 20 percent of Africa’s trade was intra-

regional in 2013, compared to around 40 percent among American countries, 63 percent among

Asian countries and 75 percent among European countries. Obstacles to better performance of

intra-regional trade in Africa include weak productive capacity and the lack of trade-related

infrastructure and services.

The largest increase in intra-REC trade in the past decade and a half took place in the ECCAS

region, while the slowest increase was in the SADC region. The chapter finds that ECOWAS

shows the highest regional trade integration, as measured by the ratio of intra-REC trade to the

REC’s trade with Africa; ECCAS shows the lowest. COMESA and SADC play larger roles as

destinations for and origins of African trade than do the other two RECs.

Chapter four reviews the changes in competitiveness of exports of different countries and different

agricultural products over the past three decades, and investigates the determinants of these

changes through econometric analysis. The chapter aims to shed light on the factors behind recent

improvements in trade performance in order to further accelerate gains and reduce trade deficits.

The chapter finds that most RECs saw their member countries maintain or increase their

competitiveness in global and regional markets, with the exception of ECCAS, whose member

countries tended to lose competitiveness. Improvements in the competitiveness of COMESA,

ECOWAS and SADC member countries took place primarily in intra-regional markets. The

majority of African export commodities gained competitiveness in global markets, with some

exceptions; however, the most competitive commodities account for a fairly small share of exports.

Africa’s top five most competitive commodities in global markets represent only 1.8 percent of

African exports to these markets, suggesting potential for expanding exports by leveraging

competitiveness gains among emerging export products. The chapter finds that determinants of

competitiveness improvements include the ease of doing business, institutional quality, the size of

the domestic market, and the quality of customs.

Chapter five examines the factors contributing to Africa’s improved agricultural export

performance, using a gravity model to assess the importance of different determinants of trade and

of the constraints to further improving exports. The study finds that supply side constraints,

including production capacity and the cost of trade, affect trade performance to a greater extent

than demand side constraints, which include trade policies and agricultural supports in importing

countries. This suggests a focus on removing domestic constraints to increased trade, including by

improving infrastructure and increasing agricultural productivity. For example, the study finds that

a 1 percent increase in land productivity increases trade flows to the global market by about 6

percent and to the African market by 7 percent. The chapter also finds that non-tariff barriers to

trade are increasing and present larger obstacles to exports than do tariffs. The chapter highlights

the potential of regional economic communities to promote the removal of barriers to trade at both

the regional and global levels, as well as the continued importance of global cooperation to

facilitate trade.

Chapter six focuses on the outlook for expanding intra-regional trade within West Africa, the

feature region of this report, and the potential effects of expanded trade on regional food markets.

The chapter finds that the distribution of production volatility among West African countries

suggests significant potential to lessen the impacts of domestic shocks through increased regional

trade, while patterns in agricultural production and trade show scope for increasing regional trade

levels. Analysis of a simulation model suggests that intra-regional trade will continue to increase

under current trends. Intra-regional trade growth can be accelerated through even modest

reductions in trading costs, modest increases in crop yields, or a reduction in trade barriers. In

particular, intra-regional trade in cereals during the 2008–2025 period is expected to increase by

23 percent over baseline trends following a 10 percent reduction in overall trading costs; by 36

percent following a removal of harassment costs; and by 33 percent following a 10 percent increase

in crop yields. The increased intra-regional trade resulting from these changes would reduce food

price volatility in regional markets.

The TSR chapters demonstrate undeniable improvements in Africa’s trade performance over the

past decade and a half, in both global and regional markets, as reflected by generally increasing

competitiveness for the majority of countries and commodities. However, progress has been

uneven, with some regions and countries consistently underperforming others. Challenges remain

in further enhancing Africa’s competitiveness on the global market and in increasing intra-regional

trade, which remains below its potential despite significant recent improvements. The findings of

chapter four point to the importance of the institutional and business environment in improving a

country’s export competitiveness, while chapter five also emphasizes the role of domestic factors

in increasing exports, including production capacity and trading costs. Chapter six focuses on the

West Africa region, demonstrating the role of potential domestic and regional policy actions to

increase intra-regional trade and enhance the stability of regional markets.

The chapters suggest a series of recommendations for policymakers, including efforts at the

country and regional level to increase agricultural productivity along the value chain, improve

market access, and improve the functioning of institutions; regional actions to enhance economic

integration; and continent-wide efforts to promote trade facilitation in international negotiations.

Policy actions such as these can influence the trends described in this report and accelerate

improvements in Africa’s trade performance, thereby increasing incomes and improving food

security across the continent.

Chapter 1. Introduction

Extracted from

African Agricultural Trade Status Report

2017

4

CHAPTER 1. INTRODUCTION

Trade provides the potential for improving consumer welfare and producer incomes, boosting

overall economic growth, and reducing poverty. In Africa, increased and more diversified

agricultural trade on the global and regional levels could provide leverage for efforts to raise

productivity at all stages of the value chain, and facilitate the transformation of agriculture into a

high-productivity sector providing adequate incomes for producers and stimulating growth

throughout the economy. Increasing agricultural trade also has the potential to improve food

security and contribute to stabilizing local and regional food markets by making them less

vulnerable to shocks.

In addition to the benefits of global trade, intra-regional trade has increasingly been recognized as

a key element of efforts to increase food security and agricultural development in Africa. The 18th

African Union Summit in 2012 was organized under the theme of “Boosting Intra-African Trade.”

In 2014, African leaders committed to tripling intra-African trade in agricultural commodities and

services by 2025, as one of a limited number of commitments in the Malabo Declaration on

Accelerated Agricultural Growth and Transformation for Shared Prosperity and Improved

Livelihoods. The trade commitment included accelerating the establishment of a Continental Free

Trade Area and a continental Common External Tariff and taking measures to increase investments

in trade infrastructure and enhance Africa’s position in international trade negotiations.

Despite longstanding recognition of the benefits of trade and the importance of improving Africa’s

competitiveness, the continent is performing beneath its potential in global and regional

agricultural markets. Recent increases in exports have been offset by even larger growth in imports,

leading to a deterioration in Africa’s trade balance. Intra-regional trade in Africa is growing, but

remains significantly below the levels seen in other regions. These challenges result from a host

of factors, including historical trends and more recent developments inside and outside of Africa.

Action on many fronts is needed to remove constraints to improving the competitiveness of

Africa’s producers.

In 2013, the Regional Strategic Analysis and Knowledge Support System (ReSAKSS), the official

monitoring and evaluation body of the CAADP, published its Annual Trends and Outlook Report

(ATOR) under the theme of “Promoting Agricultural Trade to Enhance Resilience in Africa.”

5

The report reviewed patterns in Africa’s global and regional agricultural trade and examined the

relationship between agricultural trade and the resilience of African countries and regions to

shocks, including food price volatility and weather shocks. The report detailed significant progress

made in improving Africa’s trade performance in recent years, as well as the remaining challenges

at the global and regional levels.

The current African Agricultural Trade Status Report (TSR) builds on the analysis presented in

the 2013 ATOR. The report provides detailed descriptive assessments of the current status and

recent trends in Africa’s trade performance and competitiveness at the continental and regional

levels, as well as more in-depth investigations of the determinants of trade performance and the

relative importance of different drivers and constraints. This report represents the first in a series

of annual publications examining current status, trends and outlook in African trade performance.

The goal of this and subsequent reports is to provide comprehensive and timely evidence and

analysis on the status of African trade in order to inform policy discussions on measures to enhance

trade performance at the global and regional level.

In addition to the introductory and concluding chapters, the report is divided into five chapters

presenting findings on Africa’s trade performance and outlook. Chapter two examines trends and

patterns in Africa’s global agricultural trade over the past decade and a half. The study assesses

trends in overall trade volumes and values and in trade of key agricultural commodities. The

chapter then analyzes the direction of agricultural exports and imports, changes in market shares,

and changes in the composition of Africa’s exports and imports, to provide a comprehensive

overview of Africa’s agricultural trade with the rest of the world.

Chapter three addresses regional trade, discussing patterns in trade among African countries at the

continental level and among its regional economic communities (RECs). The chapter reviews

intra-regional trade performance for the continent as a whole and for major RECs, before analyzing

trade direction, examining the role of individual RECs and countries in intra-regional trade, and

discussing the key commodities important in African intra-regional trade.

6

Chapter four presents a detailed analysis of the competitiveness of African agricultural exports in

global and regional markets. The chapter aims to shed light on the factors behind recent

improvements in trade performance in order to further accelerate gains and reduce trade deficits.

The study ranks countries and commodities according to their competiveness in export markets at

the global, continental, and REC levels. The chapter then performs econometric analysis of the

drivers of changes in competiveness at different levels and presents recommendations for further

improving competiveness.

Chapter five provides an in-depth examination of the determinants of African agricultural trade

performance. The chapter reviews broad categories of trade determinants, including production

capacity, cost of trade, trade policies, domestic agricultural supports, and global market shocks.

The chapter then develops a gravity model to assess the relative importance of determinants of

African trade and of different constraints to trade, and discusses how these constraints have

changed over time and vary across countries.

Chapter six focuses on the outlook for expanding intra-regional trade within West Africa, the focus

region of this issue, and the potential effects of expanded trade on regional food markets. The

chapter reviews recent trends in intra-regional trade and examines the possibilities for increased

regional trade to reduce food price volatility. The study then evaluates the scope for increasing

trade within the region. A simulation model is used to examine the effects of alternative policy

scenarios on regional trade and on the stability of regional food markets.

The final chapter concludes the report by reviewing findings from the preceding chapters. The

chapter synthesizes the results of previous analyses and summarizes policy implications for

addressing constraints to improved trade performance.

Chapter 2. Africa global trade patterns

Extracted from

African Agricultural Trade Status Report

2017

7

CHAPTER 2. AFRICA GLOBAL TRADE PATTERNS

Fousseini Traore IFPRI- Markets, Trade and Institutions Division, Regional Office for West and

Central Africa, Dakar, Senegal

Daniel Sakyi, Department of Economics, Kwame Nkrumah University of Science and

Technology (KNUST), Kumasi, Ghana

2.1 Introduction The trade performance of African countries has improved in recent years, though it is still below

expectations when compared to other regions of the world. This notwithstanding, and although the

region is currently considered as one of the fastest growing regions in the world, Africa’s trade

performance continues to be dominated by the agricultural sector. Overall, Africa’s

competitiveness has slightly improved and the trends in its exports have undergone major

diversification since 1998. This has become possible due to the region’s (i) participation in

multilateral and bilateral talks (WTO-DDA; EPAs, etc.), (ii) benefits received from preferential

trade agreements (AGOA, EBA, etc.), and (iii) deeper regional integration (FTAs, customs unions,

etc.). In addition, technological transfer from developed countries to the region has contributed

significantly to transformation of the agricultural sector and trade.

Although the agriculture sector still remains key with the potential to be an important player in

global food markets and continues to play a significant role in terms of value-added (NEPAD,

2015)1, the share of agricultural exports in total exports has declined since 1998. This has remained

so because the sector is still characterized by low productivity, which tends to pose a major setback

to Africa’s economic development and structural transformation. This presents critical challenges

for Africa given the continent’s rich natural resource endowments and its potential to transform

and export high valued agricultural products both within the continent and abroad. It is, therefore,

not surprising that the need to develop and transform the agricultural sector in Africa was heavily

discussed in the 2014 Malabo Declaration, as this was crucial to accelerate Africa’s development

campaign. Therefore, the commitment to boosting intra-African trade in agricultural commodities

and services (i.e. to triple, by the year 2025, intra-African trade in agricultural commodities and

services) is seen as key to growth because its expansion will trickle down to other sectors of the

region’s economy.

1 In fact, agriculture accounts for a significant portion of GDP in Africa (about 20% in 2015 (World Bank, 2015)),

and therefore presents considerable potential for supporting broader growth and the eradication of poverty and hunger.

8

In recent years the trends in international trade were largely driven by the sluggish economic

growth and the persisting economic and political turmoil in various parts of the world; from 2011

to 2014 world trade grew at a rate of less than 2 percent per year, due to generally lower economic

growth but also because trade has been much less responsive to output growth. This was

particularly the case for Africa (UNCTAD, 2015). Regarding agricultural products, while world

agricultural exports grew annually at 7% between 2010 and 2014, Africa’s exports grew at 5%,

highlighting more resistance for agricultural trade compared to trade in manufactures which grew

at 4% (WTO, 2015).

African agricultural export shares in global trade have increased steadily between 1998 and 2013,

with a diverging pattern among the main Regional Economic Communities (RECs). The ECCAS

and SADC regions registered a relative decline, while COMESA showed stability and ECOWAS

is characterised by huge short run volatility. However, the region’s imports still remain higher than

its exports in value terms, yielding a growing trade deficit. The main drivers of this surge in imports

are rapid population growth and urbanisation, income changes due to economic growth, and

changes in dietary patterns. Among the RECs, the SADC region is the only one registering a

consistent trade surplus over the last decade.

One noticeable feature is the direction of Africa’s trade to and from the European market that has

constantly showed a downward trend, while trade with regional partners and Asian countries keeps

rising. Africa also registered a decrease in the concentration of its exports over the last decade.

Another interesting feature is the relative decline of the share of agricultural exports in Africa’s

total exports, indicating that the main source of foreign earnings come now from non-agricultural

products. However, overall, despite the region’s attempt to integrate into the global market, there

is still some work to be done in increasing diversification, in furthering integration into global

value chains and in meeting international standards.

This chapter examines Africa’s global trade patterns from 1998 to 2013. Specifically, section II

highlights the trends of Africa’s agricultural trade both in values and in volumes with a focus on

the evolution of some key agricultural commodities. This is followed by a discussion of trends in

net agricultural exports in section III. Changes in market shares are presented under section IV;

this section also analyses in detail the direction of African’s exports and imports. Since the region’s

export and import composition changes over time, the composition of agricultural exports and

9

imports is also discussed under section V. We then examine under section VI the changes in unit

values of agricultural exports and imports. Finally, the last section concludes the chapter.

2.2 Trends in volumes and values of global agricultural trade (exports and imports)2

2.2.1 Global patterns

Fig. 2.1. Total agricultural trade, billion USD Fig. 2.2. Export shares in global agricultural

(nominal values) exports (nominal values)

Source: BACI Source: BACI

Globally, agricultural exports and imports have been increasing steadily since 1998 even though

imports have been generally higher than exports (Figure 2.1). After a fall in the nineties, Africa’s

exports have increased continuously over the last decade at 8% annually. Over the entire period

(1998–2013) exports more than doubled.

From 2008 to 2013 (the post crisis period), the annual growth rate of agricultural exports was 6.6%

which is much higher than total export growth (1.3%) due to sluggish economic growth in the

world (UNCTAD, 2015). Although the trend looks promising, exports still lagged behind imports.

The reasons behind this increase in exports include price booms of various commodities over the

last decade, the improvement in infrastructure in the continent (mostly transport and

telecommunication), economic growth, and more regional and global integration efforts.

2 Unless specified, all figures refer to aggregate continental trade, i.e. extra and intra Africa trade lumped together. The main source

of data is the BACI database built by CEPII. Based on UN COMTRADE, BACI has developed a procedure to reconcile exporter

and importer declarations using both mirror data and gravity modeling (Gaulier et al., 2010). This allows a significant increase in

the number of countries with available data. See the appendix for a complete description of the database.

0

20

40

60

80

100

120

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Exports Imports

0.00%

1.00%

2.00%

3.00%

4.00%

5.00%

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Africa SSA

10

While export growth has not been as high as expected, in contrast, the value of agricultural imports

has increased rapidly during the years since 1998. Over the entire period, imports have grown

fivefold. Specifically, there was a general rise in the value of agricultural imports from $19.07

billion in 1998 to approximately $68.28 billion in 2008 with a dip in 2009 ($60.61 billion). Total

trade in agricultural imports increased again between 2009 and 2011, peaking at approximately $

98.89 billion. However, since 2012, world agricultural imports have been slightly on the decline,

with the total value of world agricultural imports dropping to approximately $89.18 billion in 2013.

On the other hand, and as earlier indicated, exports have been rising over the period, with the 2013

value of approximately $63.85 billion being the highest for the period.

The higher imports may be attributed to both demand and supply factors. On the demand side, the

main elements to mention are the increasing income levels due to higher economic growth,

population growth and demographic changes, and changes in consumers’ dietary patterns

(Rakotoarisoa et al., 2011; Diao et al. 2008). The income effect due to economic growth is at play

in some countries like Ghana and Mozambique with consequences for dietary patterns. For

instance, with higher incomes, consumers demand more protein (such as meat, fish, milk, and

peanut). The other cause of increasing imports is population growth and rapid urbanization in

Africa with a concomitant increase of the population in rural areas. Africa is indeed the most

dynamic region in terms of demographics. Africa’s population has more than doubled in the last

30 years while the world’s population has grown by 60% with now two out of every five people

living in cities. The consequence of the rapid urbanization and population growth has been an

increase in the consumption of more diversified and richer animal products and in the consumption

of imported cereals (wheat, rice, and maize) rather than of the local cereals, roots, and tubers

generally consumed in rural areas (FAO, 2015). This trend will continue in the near future as

Africa’s population growth rate is twice the world average. On the supply side, the huge increase

in imports is mainly due to the poor performance in terms of competitiveness of African

agriculture, which has been unable to meet the requirements of the growing population. Low and

stagnating agricultural productivity, water constraints, the low use of fertilizers and low

mechanization are the key factors at play (FAO, 2015).

Export shares of Africa and SSA in global exports are given in Figure 2.2. The shares of Africa

and SSA’s exports in world exports have been fluctuating below 4% with a few exceptions, the

lowest share being 3.77% in 2008. The export shares of SSA countries in global exports have

11

experienced trends similar to those of Africa as a whole, with respect to the years of peaks and

troughs, meaning that North African countries do not account significantly for the region’s

agricultural exports. It is obvious from the trends given in Figure 2.2 that export shares of both

Africa and SSA in world agricultural trade are generally low. The contrasted evolution of Africa’s

share in global exports is reflected by the evolution of its competitiveness in world markets. Indeed

two third of the countries of the continent registered a loss in competitiveness while the remaining

ones managed to expand their exports in world markets faster than their competitors (Odjo and

Badiane, 2017).

The low share of Africa in world agricultural trade is to be contrasted with the facts that agriculture

products continue to contribute highly to GDP in most African countries and that agriculture

employs a large proportion of its workforce (WDI, 2015). The situation may however be explained

by the fact that compared to other countries or regions, agricultural production in Africa is largely

on a “peasant” scale (Bryceson, 2015; Collier and Dercon, 2014), making the overall share of

agricultural exports from Africa and SSA relatively lower. However the share of Africa’s

agricultural exports in world agricultural exports is slightly greater than the share of its

merchandise exports in global merchandise exports (Figure 2.2 versus Figure 2.3), showing the

relative specialization of Africa in agricultural products. Another interesting feature is the relative

decline of the share of agricultural exports in Africa’s total exports (Figure 2.4). Indeed the share

of agricultural products has been reduced by half since 1998, indicating a symmetric increase in

export earnings from other sources (mainly textiles, minerals and fossil oil). Agricultural exports

represent now 10% of Africa’s total exports.

12

Fig. 2.3. Share of Africa in world total trade3 Fig. 2.4. Share of agricultural exports in

(nominal values) total exports (nominal values)

Source: UNCTAD Source: BACI

Globally, the evolution of the market shares of the main RECs follow that of Africa as a whole

(Figure 2.5). The evolution in some groups is however more pronounced than for others. The

ECCAS group, which has the lowest share, is also characterized by a secular decline over the entire

period. This particular pattern of the SADC region is confirmed by its lack of competitiveness over

the last decade compared to its main competitors (Odjo and Badiane, 2017; see chapter 4). The

SADC region is also an example of a relative decline over the period after an increase of its market

share in the late nineties, with a decline in competitiveness. The ECOWAS region’s market share

is the most volatile one, with an improvement in the most recent years, while COMESA’s is

relatively stable over time. The divergent evolution of the market shares of the different RECs is

due to their differences in terms of specialization (commodities exported; see Annex 2) and to their

ability to respond to price booms and to compete with other exporters in global markets.

3 Goods and services

0

0.5

1

1.5

2

2.5

3

3.5

4

1998 2000 2002 2004 2006 2008 2010 2012 20140.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

16.00%

18.00%

20.00%

1998 2000 2002 2004 2006 2008 2010 2012

13

Figure 2.5. Exports shares of agricultural products by major RECs

Source: BACI

2.2.2. Evolution of some key exported commodities

This subsection focuses on some key commodities, particularly citrus, coffee, cocoa and cotton

(the main commodities exported in 1998) and fish and related products that are not part of the

WTO agreement on agriculture.

As evident in Figure 2.64, although citrus was the second most exported commodity in volume

terms after cocoa between 1998 and 2002, it outstripped the volume of cocoa exported from 2002

to 2013. Notwithstanding, cocoa remains the highest exported commodity in value (see Figure 2.7)

from 1998 to 2013, with the value of citrus, coffee and cotton all performing below that of cocoa

in the same period.

Globally, the price of cocoa and coffee in US$ per kilogram have grown continually since 2000

(see Figure 2.8). However, with the exception of the period 2001 to 2004, the coffee price grew

more rapidly than the cocoa price. Also cotton price (see Figure 2.9) maintained a relatively stable

growth rate between 2000 and 2009. By the year 2011, the price of cotton had more than doubled

from the price in 2000, though the highest price in 2011 did not last for the subsequent years.

4 Figure 2.6 illustrates the evolution of major agricultural exported commodities in millions of tons: citrus, coffee,

cocoa and cotton.

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

1.60%

1.80%

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

ECOWAS ECCAS COMESA SADC

14

What is interesting is the imperfect and even opposite correlation between the volume of exports

and world prices at the end of the period with the exception of cocoa (Figures 2.6 and 2.8). Indeed

despite the huge drop in the world prices of cotton and coffee, export volumes continue to rise

after 2011. This may be due to an imperfect transmission of international price shocks to local

producers’ prices (due to stabilization mechanisms at play, exchange rate movements between

USD and local currencies, etc.) but also to an income effect which pushes producers to supply

more when prices fall (i.e., negative supply elasticity; see Yotopoulos and Lau, 1974).

Fig. 2.6. Export volume of key commodities Fig. 2.7. Export value of key commodities

(Millions of tons) (Millions of USD)

Source: BACI Source: BACI

Fig. 2.8. Cocoa and coffee prices in US$/KG Fig. 2.9. Cotton (Cotlook A index cents/lb)

Source: World Bank Source: NCC

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Citrus Coffee Cocoa Cotton

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

1998 2000 2002 2004 2006 2008 2010 2012

Citrus Coffee Cocoa Cotton

0

1

2

3

4

5

6

7

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Cocoa Coffee

0

20

40

60

80

100

120

140

160

180

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

15

Fish and related products

Fish and related products represent a huge share of agricultural (extra-regional) exports for some

countries (such as Senegal) but are not part of the WTO agreement on agriculture. It is therefore

important to include them in the analysis. From 1998 to 2013, fish exports represented on average

15% of total agricultural exports.

Africa and SSA’s exports of fish and related products have doubled between 1998 and 2013,

increasing from $3.12 billion dollars and $2.29 billion dollars respectively to $7.17 billion and

$4.98 billion dollars (see Figure 2.10). For both Africa and SSA, exports of fish and related

products generally increased continuously from 1998-2008, fell between 2008 and 2010, and

increased again between 2010 and 2013.

Trends in the share of Africa and SSA in global fish trade have been similar for 1998-2013 (see

Figure 2.11). It is worth noting that Africa’s share in global fish exports is higher than its average

share in agricultural product exports, indicating a greater role and potential in that particular

market.

Fig. 2.10. Evolution of export value in USD millions Fig. 2.11. Share in global fish trade

Source: BACI Source: BACI

0

1

2

3

4

5

6

7

8

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Africa SSA

0

1

2

3

4

5

6

7

8

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Africa SSA

16

2.3 Trends in net agricultural exports

Since the early nineteen eighties, Africa’s agricultural exports have continued to lag behind its

imports. The agricultural trade deficit has therefore continued to dominate as the region recorded

a negative value in its net exports between 2001 and 2013 (see Figure 2.12). This pattern is also

confirmed by the normalized trade balance5 (see Figure 2.13). The main contributor to the trade

deficit is the America region (both North America and Latin America) with –US$4 billion in 2001,

–US$7 billion in 2005 and –US$18 billion in 2013. The EU and Asia regions recorded a surplus

of US$3.3 billion and US$0.9 billion respectively in 2001. Net agricultural exports to the global

market have worsened since, as Africa started recording deficits with both Asia and the EU in

addition to the America region. The lowest ever deficit recorded occurred in 2011 (US$39.7 billion

globally). In that same year, Africa recorded a negative value of US$8.3 million to Asia, US$1.6

million to the EU and US$25.3 billion to America. Although the deficit recorded in net agricultural

exports reduced somewhat, evidence for 2013 shows that net agricultural exports by African

countries have not been encouraging. Also, globally the deficit is mainly due to significant

increases in imports rather than a decrease in exports. The main import commodities causing the

deficit are sugar, maize, and wheat from the America region; wheat, milk and cream from the EU;

and rice, palm oil and wheat from Asia.

It appears that most of the RECs recorded a trade deficit over the period with the exception of the

SADC region which recorded a surplus over the entire period (see Annex 2). The trade deficit is

particularly important for North African countries, which are huge cereal importers. According to

recent studies, 23 countries in Africa are highly import dependent, with normalized trade balance

index values between -1 to -0.1 while 37 countries are net importers of food (FAO, 2015).

The growing agricultural trade deficit suggests that it is necessary that African countries take

relevant steps to improve export performance since the continent has the “agrarian” environment

to support agricultural exports. Agriculture on the continent must gradually be transformed from

being peasant-dominated to a more commercial type as doing so in addition to other measures

(such as improvement in technology and skills) will greatly improve agricultural exports.

5 The normalized balance is computed as a country's exports of agricultural products minus its imports of agricultural

products, normalized by dividing it by its agricultural trade (imports plus exports). The index varies between -1 and

1.

17

Figure 2.12. Evolution of net agricultural exports in US$ million (nominal values)

Source: BACI

Figure 2.13. Normalized trade balance

Source: BACI

2.4 Directions of agricultural exports and imports and changes in market shares

This section assesses the direction of Africa’s agricultural exports and imports as well as the

changes in Africa’s market shares in these regions. Africa as a region has been noted for its natural

resource abundance and a significant share of its exports are agricultural products, either semi

processed or in their raw state. Different types of exports are made by Africa to different regions

in the world. However, the most common agricultural export commodities are cash crops. In

particular, commodities such as cotton, cocoa, coffee, cassava, and sorghum are exported to other

parts of the world. The direction of these exports however depends on the demand for such

-30000

-25000

-20000

-15000

-10000

-5000

0

5000

10000

2001 2005 2013

World EU Asia America

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

18

products. In Figures 2.14 and 2.15, we present the direction of Africa’s agricultural exports from

1998 to 2013 to four regions: among African countries; Europe; Asia; and America. As shown in

Figure 2.15, thanks to free trade areas and improvement in local infrastructure, the rate at which

African countries export to each other has increased at a constant rate since 1998. This outcome is

however still low when compared to other regions outside Africa. The direction of exports among

African countries have averaged 15.70% between 1998 and 2012 in spite of the low take off rate

of 11% in 1998.

Exports to Europe have shown a downward trend since 1998, yet Europe remains the region’s

highest export destination. Consistently, Africa’s exports to the EU dropped from 62% of total

agricultural exports in 1998 to 37% in 2012. Some African countries started developing tropical

products for export to the EU market, to take advantage of the preferences granted by the EU (EBA

for instance), but EU standards and SPS dampen the level of agricultural exports (Otsuki and

Sewadeh, 2001; Kareem, 2014). It is also worth noting that the EU started negotiations with some

of Africa’s competitors such as Asia and Latin America, the risk being the erosion of preferences

for African countries for some commodities such as cocoa and bananas. Exports to Asia (and

Europe) are mostly agricultural products that are high-value and low-calorie in nature. Notable

among them are cotton, coffee, flowers, fruits, tea, tobacco and fish. As evident in Figure 2.15,

exports of agricultural products to Asia increased at a slower rate between 1998 and 2012 while

exports to America have been fairly low. Until 2012, the share of exports to America was below

9%. The highest export share to America since 1998 was recorded to be 9.69% which occurred in

2012. This reduced to 5.63% in 2013 (see Figure 2.14). Europe, on the other hand, received the

highest share of Africa’s exports (37.52%) in 2013 (see Figure 2.14) followed by Asia and Africa.

On the import side, as shown in Figure 2.16, the region imported 12.51% in 1999 from its own

area. This increased to 16% in 2003 and dropped to 12.37% in 2008. However these low figures

do not account for informal cross-border trade between African countries. This consists of flows

of local products and of import/re-export flows, sometimes in order to circumvent protectionist

policies put in place by some countries against imports from the international market (see the

Nigeria-Benin case, LARES, 2005; Golub, 2012). Since estimates of intra-regional trade volumes

are based on official statistics (customs declarations), the volume of trade is largely

underestimated. For instance, more than 50% of Benin’s trade in red meat, cattle and cereals was

informal in 2010 (ECNE, 2010). However some obstacles still remain for intra-African trade.

19

Among these are mentioned inadequate transport, storage and preservation infrastructure; tariffs,

non-tariff barriers and export bans; technical barriers; customs procedures; lack of harmonisation

of procedures and documents; lack of recognition of national certificates and standards; migratory

procedures; and roadside inspections (Levard and Benkhala, 2013; Rolland and Alpha, 2011).

Finally the share of intra-trade varies among commodities: cereals and live animals are the most

intra-exported while coffee, cocoa, and tea are mostly exported outside the continent.

The majority of Africa’s imports come from Europe. It is evident from Figure 2.16 that in 1998

42% of the region’s imports came from EU. Though the percentage of imports from the EU has

reduced since 1998, the EU still remains Africa’s largest origin of imports. Currently, imports from

America have been rising steadily; between 1998 and 2003, the share of imports from this region

averaged 26.62%. Moreover, the highest imports to Africa in 2011 came from America. Inside

America there is a sharp drop in imports from North America which benefited Latin America. The

share of imports from Asia has also increased from 11.30% in 1998 to 26.42% in 2012. This,

however, dropped in 2013 to 24.78%. The main feature here is the decline of Europe and the rise

of Asia over the period as Africa’s trade partner both for imports and exports.

Figure 2.14. Direction of agricultural exports and imports in 2013

Exports (nominal values) Imports (nominal values)

Source: BACI Source: BACI

20.14%

37.52%

31.71%

5.63% 5.00%

Africa EU Asia America Others

14.42%

27.98%

24.78%

24.29%

8.53%

Africa EU Asia America Others

20

Fig. 2.15. Directions of agricultural exports Fig. 2.16. Directions of agricultural imports

(nominal values) (nominal values)

Source: BACI Source: BACI

2.5 Changes in composition of agricultural exports and imports

The composition of agricultural exports and imports in Africa recorded mixed features over time.

It shows an increasing diversification of exports and a relative stability for imports, with slight

modifications from period to period.

It is widely recognized that African exports are highly concentrated (Kose and Riezman, 2001;

Songwe and Winkler, 2012). However, within the agricultural sector, Africa’s exports seems to

have started a gradual diversification as the composition and the shares of the region’s exports

changed over time. We report in Figures 2.17 and 2.18 the top ten exported products from Africa.

In 1998 the top 10 (HS4) products represented 57% of exports while in 2013 they represented

43%, indicating a decrease in the concentration of exports. However, 6 out of 10 products present

in 1998 were also present in 2013. By the end of the year 1998, cocoa beans were the region’s top

exported agricultural product. This is still the case in 2013 with 14% of total agricultural exports.

Coffee and cotton emerged as the second and third most exported products in that same year

(1998), amounting to US$2 billion and US$1.5 billion, respectively. Among others, sugar, tobacco,

tea, citrus fruits, grapes and apples were also among the top ten exported agricultural products in

1998. The region has since 1998 witnessed a drop in the export of cotton, citrus fruits and tobacco.

Conversely, cigars and cigarettes, oilseeds and frozen fish, which were absent from the list of top

exports in 1998, are now among the top ten products exported in 2013.

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

70.00%

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

Africa EU Asia

America Others

0.00%

5.00%

10.00%

15.00%

20.00%

25.00%

30.00%

35.00%

40.00%

45.00%

1998 2000 2002 2004 2006 2008 2010 2012

Africa EU Asia

America Others

21

Exports.

Fig. 2.17. Top ten products in 1998 Fig. 2.18. Top ten products in 2013

(in % of total agricultural export value) (in % of total agricultural export value)

Source: BACI Source: BACI

Unlike exports, Africa’s imports have remained quite stable in terms of composition and shares.

In 1998 the top 10 (HS4) products represented 52% of imports against 49% in 2013. As evident

from Figures 2.19 and 2.20, 8 out of the top 10 commodities imported in 1998 are also present in

2013. In Figure 2.19, the highest share of Africa’s agricultural imports is held by wheat and meslin

flour, which constituted about 16% of agricultural imports in 1998. Sugar was the second most

imported product, representing 8.28% of agricultural products imported by African countries. The

other products that were among the top ten imported products include maize, rice, wheat and

meslin flour, soya-bean oil, palm oil, sunflower-seed, and cigars and cigarettes. In 2013, wheat

and meslin continued to account for the highest share of agricultural imports. Rice is the second

most imported agricultural product followed by sugar, palm oil, and milk and cream. Meat and

edible offal of poultry, soya-bean oil and oil-cake and other solid residues are among the products

imported in 2013. The entry of meat and edible offal in the top 10 imported commodities highlights

the shift towards more protein-related products mentioned earlier.

0.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

0.00%2.00%4.00%6.00%8.00%

10.00%12.00%14.00%16.00%

22

Imports.

Fig. 2.19. Top ten imported products in 1998 Fig. 2.20. Top ten imported products in 2013

(in %of total agricultural imports) (in % of total agricultural imports)

Source: BACI Source: BACI

2.6 Changes in unit values of agricultural exports and imports

A plot of trends in the evolution of agricultural imports and exports unit values is given in Figure

2.21. It shows changes in unit values of agricultural imports and exports using 2000 as the base

year. The evolution of unit values is related to the so-called (deterioration of) terms of trade

literature which dates back to the Prebisch-Singer hypothesis (Prebisch, 1950; Singer, 1950) that

argues that the price of primary commodities declines relative to the price of manufactured goods

over the long run, causing the terms of trade to deteriorate for primary products exporting and

manufactured goods importing countries. However recent research regarding this topic has given

mixed results (Arezki et al. 2013).

From Figure 2.21, it can be seen that the unit value of both agricultural imports and exports have

generally increased for the 2000–2013 period with a mixed pattern. From 2000 to 2007, the

evolution of both indicators shows a significant increase, with imports rising faster than exports,

yielding a slight deterioration of the agricultural terms of trade. The period between 2008 and 2013

saw the evolution of the unit value of exports outstripping the unit value of Africa’s imports. This

improvement was mainly due to the huge increase in commodity prices in the late 2000s and is in

0.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

16.00%

18.00%

0.00%

2.00%

4.00%

6.00%

8.00%

10.00%

12.00%

14.00%

23

line with the evolution global terms of trade for Africa (UNCTAD, 2015) though more important

here than that of total trade6.

Figure 2.21. Evolution of exports and imports unit values (Base 100=2000)

Source: BACI

2.7 Conclusion

Africa has experienced a significant increase in both the value of its exports and imports over the

last decade, boosted by the increase in commodity prices in international markets. However, since

1998, Africa’s imports have increased more rapidly both in shares and in value terms than exports,

yielding a continuously deteriorating trade deficit. This growing trade deficit driven by imports is

mainly due to population and economic growth, change in dietary patterns, increasing income

levels and the lack of competiveness of the domestic sector. Among the main RECs of the

continent, the SADC region is the only one recording a surplus over the entire period.

Africa’s share of global trade in agriculture has been stable around 4%, though with some small

fluctuations for the last three years. The evolution of the market shares of the main RECs shows a

regular decline of the shares of the ECCAS and the SADC region, a relative stability of

COMESA’s share and a highly volatile pattern for ECOWAS. One of the main interesting features

is the secular decline of the share of agricultural exports in Africa’s total exports. The share of

agricultural exports in Africa’s total exports has been cut by half since 1998 to the benefit of

mineral and fossil oils.

6 This is due to mineral products that are not taken into account here.

0

50

100

150

200

Exports Imports

24

The composition of agricultural exports and imports in Africa recorded mixed features, showing

an increasing diversification for exports and a relative stability for imports. Indeed, within the

agricultural sector, Africa’s exports seem to have started a gradual diversification. Now the top ten

(HS4) exported products represent 43% of exports compared to 57% in 1998. However, most of

the products present in the top exported commodities in 1998 are still present, with a concentration

of cocoa beans, coffee and cotton. Unlike exports, Africa’s imports have remained quite stable in

terms of composition and shares, with the top ten (HS4) products still representing half the imports.

Imports remain dominated by cereals (wheat, rice, maize) and sugar, with a recent shift towards

more protein (meat and offal and fish).

In terms of directions of trade, Africa’s trade (both imports and exports) with the European market

has witnessed a continuous drop since 2000, though the EU still remains the first partner for the

continent. At the same time, Asia has emerged as a major partner for both imports and exports. If

recent trends were to continue, Asia will soon become Africa’s first trade partner. It is worth noting

that the ability to meet standards and SPS measures is still dampening Africa’s exports, in

particular to the EU and the US markets. There is also a risk of the erosion of preferences for some

African countries as the EU for instance has ongoing negotiations with some of Africa’s

competitors such as Asia and Latin America, the main sectors at risk being those of cocoa beans

and bananas.

African countries have also expanded their intra-trade over the last 10 years and become less

dependent on international markets. In particular, the share of agricultural imports and exports

among African countries more than doubled between 2000 and 2013. Recent improvement in intra-

trade is attributed to the effort of Africans to integrate into the regional and international market

(Bouet et al., 2013). Despite this improvement, intra-African trade is still low, hence should be

strengthened. Market fragmentation (lack of infrastructure; monetary, tax and trade fragmentation;

and red tape for traders) limits the development of the region’s trade potential. These barriers

should be tackled and given priority as they increase price instability within the region and

negatively affect food security (Badiane et al., 2014; NEPAD, 2005).

25

References

Arezki, R., Hadri, K., Loungani, P., & Rao, Y. (2013). Testing the Prebisch-Singer Hypothesis since

1650: Evidence from Panel Techniques that Allow for Multiple Breaks. IMF Working Paper 13/180.

Washington, DC: International Monetary Fund.

Badiane, O., Odjo, S., & Jemaneh, S. (2014). More resilient domestic food markets through regional

trade. In O. Badiane, T. Makombe & G. Bahiigwa (Eds.), Promoting Agricultural Trade to Enhance

Resilience in Africa. ReSAKSS Annual Trends and Outlook Report. Washington, DC: International Food

Policy Research Institute.

Bouet, A., Laborde, D., & Deason, L. (2013). Global trade patterns, competitiveness and growth outlook.

In O. Badiane, T. Makombe & G. Bahiigwa (Eds.), Promoting Agricultural Trade to Enhance Resilience

in Africa. ReSAKSS Annual Trends and Outlook Report. Washington, DC: International Food Policy

Research Institute.

Brulhart, M. (2008). An Account of Global Intra-industry Trade, 1962–2006. Background Paper, World

Bank 2009 World Development Report. Washington, DC: World Bank.

Bryceson, D. F. (2015). Reflections on the unravelling of the Tanzanian peasantry, 1975–2015. In M.

Stahl (Ed.), Looking Back, Looking Ahead: Land, Agriculture and Society in East Africa. Uppsala: The

Nordic Africa Institute.

Collier, P., & Dercon, S. (2014). African agriculture in 50 years: Smallholders in a rapidly changing

world? World Development, 63, 92-101.

FAO (Food and Agriculture Organization of the United Nations). (2003). Trade Reforms and Food

Security: Conceptualizing the Linkages. Rome: Food and Agriculture Organization of the United Nations.

FAO. (2015). The State of Agricultural Commodity Markets. Rome: Food and Agriculture Organization

of the United Nations.

Gaulier, G., & Zignago, S. (2010). BACI: International Trade Database at the Product-level: The 1994-

2007 Version. CEPII Working Paper 2010-23. Paris: CEPII.

Golub, S. S. (2012). Entrepôt trade and smuggling in West Africa: Benin, Togo and Nigeria. The World

Economy, 35(9), 1139-1161.

Kose, M. A., & Riezman, R. (2001). Trade shocks and macroeconomic fluctuations in Africa. Journal of

Development Economics, 65(1), 55–80.

LARES (2005). Le Trafic Illicite des Produits Pétroliers entre le Benin et le Nigeria: Vice ou Vertu pour

l'Economie Béninoise? Technical report. Cotonou: Laboratoire d'Analyse Régional et d'Expertise Sociale.

Levard, L., & Benkhala, A. (2013). How to Promote Intra-African Agricultural Trade? Analysis of

Possibilities and Impediments Regarding the Development of Intra-African Agricultural Trade. Nogent-

sur-Marne, France: Les Editions du GRET.

26

New Partnership for Africa's Development (NEPAD). (2013). Agriculture in Africa: Transformation and

Outlook. Addis Ababa: NEPAD/African Union.

Odjo, S., & Badiane, O. (2017). Competitiveness of African agricultural exports. In O. Badiane, S. Odjo

& J. Collins (Eds.), African Agricultural Trade Status Report.

Otsuki, T., Wilson, J., & Sewadeh, M. (2001). Saving two in a billion: Quantifying the trade effect of

European food safety standards on African exports. Food Policy, 26(5), 495–514.

Prebisch, R. (1950). The Economic development of Latin America and its principal problems. Economic

Bulletin for Latin America, 7, 1-12.

Rakotoarisoa, M. A., Iafrate, M., & Paschali, M. (2011). Why Has Africa Become a Net Food Importer?

Explaining Africa Agricultural and Food Trade Deficits. Rome: Food and Agriculture Organization of the

United Nations.

Rolland, J. P., & Alpha, A. (2011). Analyse de la Cohérence des Politiques Commerciales en Afrique de

l’Ouest. Document de Travail 114. Paris: Agence Francaise de Developpement.

Kareem, O. I. (2014). The European Union Sanitary and Phytosanitary Measures and Africa’s Exports.

EUI Working Paper RSCAS 2014/98. Fiesole, Italy: European University Institute.

Singer, H. (1950). The distribution of gains between investing and borrowing countries. American

Economic Review, Papers and Proceedings, 40, 473-485.

Songwe, V., & Winkler, D. (2012). Exports and Export Diversification in Sub-Saharan Africa: A Strategy

for Post-Crisis Growth. African Growth Initiative Working Paper 3. Washington, DC: The Brookings

Institution.

World Trade Organization (2015). International Trade Statistics 2015. Geneva: World Trade

Organization.

UNCTAD (United Nations Conference on Trade and Development). (2015). Key Statistics and Trends in

International Trade 2015: The Trade Slowdown. Geneva: UNCTAD, Division on International Trade In

Goods and Services, and Commodities.

UNCTAD (United Nations Conference on Trade and Development). (2014). United Nations Conference

on Trade and Development Database. www.unctadstat.unctad.org

World Bank. (2015). World Development Indicators Database.

http://databank.worldbank.org/data/reports.aspx?source=World%20Development%20Indicators

Yotopoulos, P., & Lau, L. (1974). On modelling the agrarian sector in developing countries: An

integrated approach of micro and macroeconomics. Journal of Development Economics, 1, 105-127.

27

Annex 2.1 The BACI global trade database

BACI stands for “Base pour le commerce international” and is the world trade database developed

by the CEPII7. The database is defined at a high level of product disaggregation and is the main

source used throughout this chapter. BACI is based on data from the UN COMTRADE database,

which is the world's largest database of trade statistics, maintained by the United Nations Statistics

Division (UNSD). COMTRADE is the main global source of trade statistics in goods, covering

more than 95% of world trade. BACI tries to improve UN COMTRADE by addressing the main

issues related to it: missing information for some African countries, reporting in different

nomenclatures, no distinction between zero trade flows and missing values in raw data, etc. To

address the issues, BACI has developed a procedure that reconciles exporter and importer

declarations using both mirror data and gravity modeling (see Gaulier et al., 2010). This procedure

allows for a significant increase in the number of countries with available data.

In its standard version, BACI provides export values and quantities at the HS 6-digit level. Data

are provided for over 200 countries since 1995. The database is updated every year. The

retreatment of data is particularly important for countries that do not report frequently to

COMTRADE (especially in Africa). Table A1 illustrates the data issue and the absence of

reporting for ECOWAS countries to UN COMTRADE from 1988 to 2010. In BACI all countries

are observed for imports and exports.

7 Centre d’Etudes Prospectives et d’Informations Internationals is a research center based in Paris and part of the

Prime Minister’s Office through the Centre d’Analyse Strategique, now “France Strategie.”

28

Table A.1: ECOWAS countries’ declaration to UN COMTRADE

1988 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Total

Benin Y Y Y Y Y Y Y Y Y Y Y Y Y 13

Burkina Faso

Y Y Y Y Y

Y Y Y Y 9

Cote d'Ivoire

Y

Y Y Y Y Y Y Y Y 9

Cape Verde Y Y Y Y Y Y Y Y Y Y Y Y Y 13

Ghana

Y Y Y Y Y Y Y Y 8

Guinea Y Y Y Y Y

Y Y Y Y Y

10

Gambia Y Y Y Y Y Y Y Y Y Y Y Y Y 13

Guinea-Bissau

Y Y Y

3

Liberia

0

Mali Y Y Y Y Y Y Y Y Y Y Y

Y 12

Niger Y Y Y Y Y Y Y Y Y Y Y Y Y 13

Nigeria

Y

Y Y Y

Y Y Y Y Y 9

Senegal Y Y Y Y Y Y Y Y Y Y Y Y Y 13

Sierra Leone

Togo Y Y Y Y Y Y Y Y

Y Y Y Y 12

NB of Countries declaring Imports 8 9 8 11 10 12 12 12 10 12 12 10 11

Note: Y stands for yes if the country declares that particular year.

29

Annex 2.2 Main descriptive statistics

Table A2.1: Africa’s top 15 exported products by destination in 20138

(nominal value in 1,000 USD and volume in tons).

World Africa America

HS4 Value Volume HS4 Value Volume HS4 Value Volume

1801 8949056 2588938 2402 1659452 34813.75 1801 933360 355881.7 5201 2590810 1517283 0303 919411.9 1424777 0901 224440.7 63762.26 0805 2535454 3700486 1701 669666.3 2604785 0805 187423.8 162056.8 2401 2417195 527845.9 0709 582662.9 131412.5 1803 182423.8 49903.43 1701 2257720 6833846 0902 513835.6 241690.7 1509 140348.4 37054.27 0901 2151131 1137948 2401 351878.4 123330.4 2204 119619.3 45068.34 1604 1948820 486239.3 1511 344980.4 2543051 0303 118032 57591.94 0303 1853421 1834613 1005 295261.2 1589105 2401 102585.4 25137.2 2402 1801219 46722.25 1101 285483 2860764 0802 100752.7 30831.81 1207 1472631 12451721 0901 278569.7 402602.5 1005 100256.6 293195.7 0801 1452097 1611323 2106 266778.5 128092.1 0801 90033.46 39679.83 0902 1347222 526269.2 1902 225433.6 1218597 1604 85488.46 14904.21 1803 1346488 391861.1 0102 215255.4 144427.8 0603 78142.34 30699.34 0603 1274794 266750.3 2202 207381.9 307924.4 1802 77175.89 23988.44 0307 1097386 246285.3 1604 196898.9 82363.39 1211 56772.54 27141.34

Source: BACI

Table A2.1 ctd.

Asia European Union

HS4 Value Volume HS4 Value Volume

1801 3999891 326122.8 1801 3576260 1738438 5201 2136118 1261332 1604 1582322 367771.1 0801 1264986 1440414 1701 1112135 3388362 0805 1214586 1984923 0901 1056904 468869.4 1207 1020944 650511.8 0805 954438.1 1170226 2401 1004399 151146 0603 908689.2 187579.4 0902 516089.2 173524 1803 882639.2 259285.2 0307 453306.5 100128.9 2401 767497.2 189279.5 0901 403178.8 132506.4 0806 710809.7 315833.6 0406 385945.1 100727.4 1804 635869.3 125528.2 1005 370671.5 1250419 0307 627574.8 128131.7 0104 369376.8 114676.2 0803 611995.4 630686.3 0713 336437.6 1635826 2204 599264.4 347734.6 0303 318154.1 137725.6 0304 571980.9 111401.3 5101 266543.5 46343.24 0702 540095.2 467447.5

8 See the list of products corresponding to the HS nomenclature in Table A2.4.

30

Source: BACI

Table A2.2: Africa’s top 15 imported products by origin in 2013

(nominal value in 1,000 USD and volume in tons).

World Africa America

HS4 Value Volume HS4 Value Volume HS4 Value Volume

1001 11315164 37956637 2402 1659452 34813.75 1701 4011909 9411226 1006 6192685 15621186 0303 919411.9 1424777 1001 3148162 10857641 1701 5789882 15825559 1701 669666.3 2604785 1005 2303404 8999137 1511 4536369 10423995 0709 582662.9 131412.5 2304 1835747 4138790 1005 3606254 14965351 0902 513835.6 241690.7 0207 1423035 1123708 0402 3365801 1062497 2401 351878.4 123330.4 1507 1006265 1293838 0303 3164988 2972965 1511 344980.4 2543051 1201 984894.7 1782988 0207 2295812 1755920 1005 295261.2 1589105 0202 738227.8 251871 2402 2256805 89823.44 1101 285483 2860764 0402 649771.9 170187.6 1507 2044662 2457679 0901 278569.7 402602.5 0713 387738.9 522599 2304 1926556 4629271 2106 266778.5 128092.1 1006 369890.1 1135149 1901 1749542 795508.4 1902 225433.6 1218597 0303 314510.4 242943.1 0202 1505473 570665.6 0102 215255.4 144427.8 0206 254577.8 198430.5 2106 1461689 577673.1 2202 207381.9 307924.4 2207 224787.8 228138.2 0902 1161753 438684.9 1604 196898.9 82363.39 2303 211352.9 484886.2

Source: BACI

Table A2.2. Ctd

Asia European Union

HS4 Value Volume HS4 Value Volume

1006 5568320 13508022 1001 4772036 14966210 1511 4142895 7450407 0402 1560448 453891.9 1001 1562951 5523065 1901 1272133 347804.6 1701 816755.5 3337692 0303 1161104 712226.3 1604 789623.4 284669.2 2106 829718.2 245938.9 0202 664980.4 215213.5 2208 818824.5 132224.1 0902 629274 192373.3 2403 805653.5 41764.71 0303 618133.8 496250.9 1507 784504.3 724614.6 2002 442291 434381.8 0207 764927 543898.6 0402 331543.1 183218.5 2204 528823.8 280842 0901 309772.6 138142.9 2202 469341.3 506755.1 1516 272551.5 305105.8 1107 462966.1 1037474 1512 241708.3 273454.3 2203 415659.4 421534 1905 238722 312852.2 0102 406498.5 104375 2009 232720.1 336060.3 2309 405577.4 687164

Source: BACI

31

Table A2.3: Exports, imports and trade balance for main RECS in nominal value (1,000 USD)

ECOWAS ECCAS COMESA

Exports Imports Trade balance Exports Imports Trade balance Exports Imports Trade balance

1998 6116465 3837574 2278891 985119 1316618 -331499 5919690 6675268 -755579 1999 5705731 4070148 1635583 914239.2 1138998 -224759 5953728 6225979 -272251 2000 4849950 3941394 908555.8 864932.7 1435258 -570325 6233086 6499117 -266031 2001 4959724 5063406 -103681 870867.2 1669209 -798342 6419539 7047405 -627867 2002 5691559 5443531 248028.1 769333.3 1892355 -1123022 6575509 7367812 -792304 2003 8174045 7172308 1001737 1034457 2352183 -1317726 7708798 8389307 -680509 2004 8390249 6861849 1528401 1103567 2679544 -1575977 8639757 9309681 -669924 2005 8182928 8082486 100442.2 1259674 3046911 -1787236 9907420 10646105 -738685 2006 8111680 9648551 -1536872 1250582 3733047 -2482466 10584645 12464647 -1880001 2007 10009034 13088053 -3079019 1427620 4784696 -3357075 12404233 15811640 -3407407 2008 12135190 14878796 -2743606 1590252 6346862 -4756611 14845553 24695229 -9849676 2009 13785804 14440253 -654449 1769146 5992694 -4223548 15491756 22310479 -6818723 2010 15283877 15294911 -11034.2 1800128 6405823 -4605695 16988548 28408191 -1.1E+07 2011 18861303 28161899 -9300596 1900651 8795311 -6894660 19639714 33633079 -1.4E+07 2012 19185691 20650589 -1464898 1860603 9031307 -7170704 18108289 32659977 -1.5E+07 2013 20289380 21339574 -1050194 1767716 9572699 -7804983 19923744 29564524 -9640780

Source: BACI

32

Table A2.3: ctd

SADC AMU

Exports Imports Trade balance Exports Imports Trade balance

1998 7316775 3996326 3320449 2253018 5898554 -3645536 1999 7414659 3550548 3864111 2603562 5080264 -2476702 2000 7674486 3686711 3987775 2664439 5519295 -2854856 2001 8231349 3772930 4458419 2695109 5702532 -3007422 2002 8705809 4728753 3977056 3084234 6698156 -3613922 2003 9624956 5483425 4141531 3564657 6679442 -3114786 2004 10467023 6865226 3601797 4242618 8502594 -4259976 2005 10838574 7175830 3662744 4837408 8735021 -3897613 2006 11324527 8807677 2516850 5359433 9470486 -4111052 2007 12726162 10976492 1749670 6482675 13694306 -7211631 2008 14353135 13310628 1042507 7558859 18944213 -1.1E+07 2009 14667621 12187492 2480129 6764896 14714422 -7949526 2010 15569389 13877386 1692003 6821328 17067732 -1E+07 2011 18192694 18090398 102296.3 7905469 22378653 -1.4E+07 2012 17702902 18748276 -1045374 7579879 22748748 -1.5E+07 2013 19622634 19302801 319833.3 8232886 24009848 -1.6E+07

Source: BACI

33

Table A2.4: list of products corresponding to the HS 4 nomenclature

HS4 Product Description

0102 Live bovine animals.

0104 Live sheep and goats.

0202 Meat of bovine animals, frozen.

0206 Edible offal of bovine animals, swine, sheep, goats, horses, asses, mules or hinnies, fresh, chilled or frozen.

0207 Meat and edible offal, of the poultry of heading No. 01.05, fresh, chilled or frozen.

0303 Fish, frozen, excluding fish fillets and other fish meat of heading No. 03.04.

0304 Fish fillets and other fish meat (whether or not minced), fresh, chilled or frozen.

0307 Molluscs, whether in shell or not, live, fresh, chilled, frozen, dried, salted or in brine; aquatic invertebrates other than crustaceans and molluscs, live, fresh, chilled, frozen, dried, salted or in brine; flours, meals and pellets of

0402 Milk and cream, concentrated or containing added sugar or other sweetening matter.

0406 Cheese and curd.

0603 Cut flowers and flower buds of a kind suitable for bouquets or for ornamental purposes, fresh, dried, dyed, bleached, impregnated or otherwise prepared.

0702 Tomatoes, fresh or chilled.

0709 Other vegetables, fresh or chilled.

0713 Dried leguminous vegetables, shelled, whether or not skinned or split.

0801 Coconuts, Brazil nuts and cashew nuts, fresh or dried, whether or not shelled or peeled.

0802 Other nuts, fresh or dried, whether or not shelled or peeled.

0803 Bananas, including plantains, fresh or dried.

0805 Citrus fruit, fresh or dried.

0806 Grapes, fresh or dried.

0901 Coffee, whether or not roasted or decaffeinated; coffee husks and skins; coffee substitutes containing coffee in any proportion.

0902 Tea, whether or not flavoured.

1001 Wheat and meslin.

1005 Maize (corn).

1006 Rice.

1101 Wheat or meslin flour.

1107 Malt, whether or not roasted.

1201 Soya beans, whether or not broken.

34

1207 Other oil seeds and oleaginous fruits, whether or not broken.

1211 Plants and parts of plants (including seeds and fruits), of a kind used primarily in perfumery, in pharmacy or for insecticidal, fungicidal or similar purposes, fresh or dried, whether or not cut, crushed or powdered.

1507 Soya-bean oil and its fractions, whether or not refined, but not chemically modified.

1509 Olive oil and its fractions, whether or not refined, but not chemically modified.

1511 Palm oil and its fractions, whether or not refined, but not chemically modified.

1512 Sunflower-seed, safflower or cotton-seed oil and fractions thereof, whether or not refined, but not chemically modified.

1516 Animal or vegetable fats and oils and their fractions, partly or wholly hydrogenated, inter-esterified, re-esterified or elaidinised, whether or not refined, but not further prepared.

1604 Prepared or preserved fish; caviar and caviar substitutes prepared from fish eggs.

1701 Cane or beet sugar and chemically pure sucrose, in solid form.

1801 Cocoa beans, whole or broken, raw or roasted.

1802 Cocoa shells, husks, skins and other cocoa waste.

1803 Cocoa paste, whether or not defatted.

1804 Cocoa butter, fat and oil.

1901 Malt extract; food preparations of flour, meal, starch or malt extract, not containing cocoa or containing less than 40% by weight of cocoa calculated on a totally defatted basis, not elsewhere specified or including; food preparations

1902 Pasta, whether or not cooked or stuffed (with meat or other substances) or otherwise prepared, such as spaghetti, macaroni, noodles, lasagne, gnocchi, ravioli, cannelloni; couscous, whether or not prepared.

1905 Bread, pastry, cakes, biscuits and other bakers' wares, whether or not containing cocoa; communion wafers, empty cachets of a kind suitable for pharmaceutical use, sealing wafers, rice paper and similar products.

2002 Tomatoes prepared or preserved otherwise than by vinegar or acetic acid.

2009 Fruit juices (including grape must) and vegetable juices, unfermented and not containing added spirit, whether or not containing added sugar or other sweetening matter.

2106 Food preparations not elsewhere specified or included.

2202 Waters, including mineral waters and aerated waters, containing added sugar or other sweetening matter or flavoured, and other non-alcoholic beverages, not including fruit or vegetable juices of heading No. 20.09.

35

2203 Beer made from malt.

2204 Wine of fresh grapes, including fortified wines; grape must other than that of heading No. 20.09.

2207 Undenatured ethyl alcohol of an alcoholic strength by volume of 80 % vol or higher; ethyl alcohol and other spirits, denatured, of any strength.

2208 Undenatured ethyl alcohol of an alcoholic strength by volume of less than 80 % vol; spirits, liqueurs and other spirituous beverages.

2303 Residues of starch manufacture and similar residues, beetpulp, bagasse and other waste of sugar manufacture, brewing or distilling dregs and waste, whether or not in the form of pellets.

2304 Oilcake and other solid residues, whether or not ground or in the form of pellets, resulting from the extraction of soyabean oil.

2309 Preparations of a kind used in animal feeding.

2401 Unmanufactured tobacco; tobacco refuse.

2402 Cigars, cheroots, cigarillos and cigarettes, of tobacco or of tobacco substitutes.

2403 Other manufactured tobacco and manufactured tobacco substitutes; homogenised or reconstituted tobacco; tobacco extracts and essences.

5101 Wool, not carded or combed.

5201 Cotton, not carded or combed.

Chapter 3. Regional trade patterns

Extracted from

African Agricultural Trade Status Report

2017

36

CHAPTER 3. REGIONAL TRADE PATTERNS

Anatole Goundan, International Food Policy Research Institute, West and Central Africa office,

Dakar, Senegal

Cheickh Sadibou Fall, Institut Sénégalais de Recherches Agricoles, Bureau d'Analyses Macro-

Economiques, Dakar, Senegal

3.1 Introduction

Deepening intra-regional trade among African countries, and especially Africa’s main RECs, is

essential for the continent’s resilience against international market shocks. Aware of that, African

leaders have positioned African economic integration as a central key in almost all continental

roundtables or political discussions. Important efforts have been made through several regional

trade agreements (RTA) such as the creation of free trade areas (FTA), customs unions (CU), and

economic and monetary unions. More recently, the 2012 African Union Summit mainly focused

on “Boosting Intra-African Trade.” Even if those agreements have generally and positively

impacted intra-African trade, the share of intra-regional trade in total African trade is still very low

compared to other regions or continents. For agricultural commodities, the view is similar (Figure

3.1). The share of trade in agricultural products among African countries that is intra-regional

varies between 13% and 20% over the period from 2000 to 2013, while its level is around 40%

among American countries, 63% among Asian countries and 75% among European countries.

Figure 3.1. Share of intra-regional agricultural trade value in total trade

Source: BACI and authors’ calculation, 2016.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

20

00

20

05

20

10

20

13

20

00

20

05

20

10

20

13

20

00

20

05

20

10

20

13

20

00

20

05

20

10

20

13

Africa America Asia Europe

Intra trade Extra trade

37

Many reasons could explain that low level of intra-regional trade in Africa. Obstacles to better

performance of intra-regional trade in Africa include weak productive capacity, the lack of trade-

related infrastructure and services, the limited role of the private sector in regional integration

initiatives, the low diversification of traded products, the small size of consumer markets, and the

quality of institutions.

This chapter focuses on the state of intra-African trade for agricultural commodities over recent

years. It will mainly (i) analyze the current performance of intra-African and intra-regional trade,

(ii) explore trade direction at the continental and REC levels, (iii) study the trading role of each

REC in African trade and each country’s individual share in the corresponding REC, (iv) examine

the main agricultural products traded among African countries, and finally (v) present the

evolution of import and export unit values.

3.2 A general perspective of regional agricultural trade and total trade

Over recent years (1998-2013), African exports have increased rapidly, with an annual growth of

12%. During the same period, trade exchange between African countries showed a significant

increase (16%), with an intra-African trade share growing from about 7% in 1998 to 13% in 2010.

The average intra-African trade share stood at 10%. In terms of agricultural trade, its share in total

trade has decreased over the years, passing from 18% in 1998 to about 9% in 2010. Total

agricultural trade has shown an annual growth of 8%. Agricultural trade between African countries

has experienced a significant growth rate of about 13% over the period, especially after the recent

food crisis, with an increase between 10 and 28% over the period from 2007 to 2012.

At the ECOWAS level, total exports have also considerably increased over the period, with an

annual growth of 14%. Trade within the region represents on average only 8% of total trade, but

has displayed a large increase between 1998 and 2013 of around 15%. Agricultural trade represents

about 15% of total exports, with an annual growth of 8%. Within the region, the agricultural trade

share stands at 18% on average, with on average 12% annual growth.

The total trade of ECCAS countries has displayed very high growth of more than 17% over 1998-

2013. However, this trade performance is not due to an increase in intra-regional trade, which

represents less than 2% of total ECCAS exports. Agricultural products represent only 4% of total

exports, with about 4% growth. The trade of these products inside the region represents 18% of

38

the total intra-regional trade. Over the period, intra-regional agricultural trade has grown

significantly, with an average growth rate of 16%.

For COMESA countries, total exports have shown significant growth over the period, with an

annual growth rate of 12%. Trade within the region, which represents on average only 6% of total

trade, has grown more rapidly than total trade (16% compared to 12%). Agricultural trade

represents about 17% of total exports, with an annual growth rate of 8%. Within the region, the

agricultural trade share stands at 33% on average, which is the highest share among the considered

RECs. The agricultural trade share grew by an average of 12% annually.

For SADC countries, total exports have shown rapid growth over the period, with an annual growth

rate of 16%, increasing from $11 billion in 1998 to $105 billion in 2013. Intra-SADC trade, which

represents on average only 4% of total trade, has grown rapidly, with a 19% annual growth rate.

Agricultural trade represents about 16% of total exports, with an annual growth of 7%. Within the

region, the agricultural trade share stands at 27% on average, which is the second highest share

among the considered RECs, with 17% average annual growth.

In terms of trade balance, Figure 3.2 depicts changes in the normalized trade balance over the

period 1998-2013 for agricultural and non-agricultural products for different regional economic

communities. This graph shows that the evolution of the trade balance depends immensely on the

product group and the region considered. Agricultural products tend to have a negative trade

balance, especially after the recent food crisis. Unlike agricultural products, non-agricultural

products have a positive trade balance for several RECs and years.

39

Figure 3.2. Evolution of the normalized trade balance by REC and product group

Source: BACI and authors’ calculation, 2016.

Note: (a) Total agricultural trade, (b) Total non-agricultural trade.

3.3 Trends in volumes and values of intra-African and intra-regional agricultural exports and imports

The evolution of agricultural trade in value and volume among African countries in general and

among some RECs (ECOWAS, ECCAS, COMESA and SADC) over the period from 1998 to

2013 is represented in Figures 3.3 and 3.49.

9 In the BACI trade dataset, intra-regional exports are set to exactly equate intra-regional imports. Therefore, we use

‘intra-regional trade’ to mean imports or exports. In terms of trends, imports or exports are equivalent.

-80

-60

-40

-20

0

20

40

1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

(a)

Africa ECOWAS ECCAS COMESA SADC

-40.00

-20.00

0.00

20.00

40.00

60.00

1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

(b)

Africa ECOWAS ECCAS COMESA SADC

40

The value of intra-African agricultural trade has grown rapidly over recent years, rising from $2.2

billion in 1998 to $12.8 billion in 2013 (Figure 3.3). The overall annual growth during this period

is around 12%. When the period is split into two sub-periods (before and after the international

crisis), an increase in the growth of agricultural products trade can be noted (13.62% between 2007

and 2013) compared the period before the crisis (11.47% between 1998 and 2006). The analysis

of intra-African trade in agricultural products in volume terms shows an overall growth of 15.84%,

which is greater than the nominal trade growth. Therefore, in general, growth in agricultural trade

between African countries over the selected periods was not driven by price increases.

Figure 3.3. Intra-regional agricultural trade over 1998-2013 by REC

Source: BACI and authors’ calculation, 2016.

Note: (a) trade value in billion US dollar, (b) trade volume in million metric tons.

Intra-ECOWAS agricultural trade shows an average growth of 12%, rising from $494 million in

1998 to $2.84 billion in 2013. Despite this apparent significant growth, agricultural trade between

ECOWAS countries was very erratic. In fact, seven negative growth-rates were noticed over the

considered period. The year 2006 saw the biggest decrease (-23.4%) and the largest increase was

reported in 2003 (95%). Over the two sub-periods, a big growth gap was noted. The sub-period

before 2007 showed an average growth of 5% while an intra-regional trade increase of 21% was

registered during the sub-period starting in 2007.

0

2

4

6

8

10

12

14

(a)

Africa ECOWAS ECCAS

COMESA SADC

0

5

10

15

20

25

30

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

(b)

Africa ECOWAS ECCAS

COMESA SADC

41

This could be the result of various initiatives during and after the international food crisis. As

examples of initiatives during the recent food crisis, Engel et al. (2013, page 20) mention the EU-

led Alliance Globale pour l’Initiative Résilience – Sahel (AGIR), the Comité permanent Inter-

état de Lutte contre la Sécheresse au Sahel (CILSS) initiative, the COMESA Alliance for

Commodity Trade, and the SADC Regional Indicative Strategic Development Plan, etc. In terms

of agricultural trade volume, overall growth of 11% is reported compared to 12% for nominal

trade. Trade increase between ECOWAS countries was then partly driven by commodity prices.

Figure 3.4. Average intra-regional trade growth (value and volume)

Source: BACI and authors’ calculation, 2016.

Note: (a) trade value, (b) trade volume.

Agricultural trade between ECCAS countries has shown the highest overall growth in value of

17%, with a nominal value which has increased from $14 million in 1998 to $147 million in 2013.

A significant change in intra-ECCAS trade can be noted over the two sub-periods. The first period

was characterised by an improving trade performance with an average annual growth of 27%, but

the growth rate of intra-exchange fell to 5% in the second period. Obviously, the 2007-2008 food

crisis has dampened the dynamic of agricultural trade inside the ECCAS zone. The volume of

agricultural trade between ECCAS countries showed the same dynamics as nominal trade.

Moreover, the average growth of trade volume was higher than that of trade value.

0%

5%

10%

15%

20%

25%

30%

Africa ECOWAS ECCAS COMESA SADC

(a)

1998-2006 2007-2013 Overall

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

Africa ECOWAS ECCAS COMESA SADC

(b)

1998-2006 2007-2013 Overall

42

In fact, the average trade volume (nominal trade value) growth was 38% (27%) over 1998-2006,

8% (5%) from 2007 to 2013, and 23% (17%) for the entire period. It could be concluded that on

average, trade flow of agricultural products was not driven by price increases.

Like other RECs, intra-regional agricultural trade in COMESA has displayed a significant increase

(14%) over 1998-2013, rising from $379 million in 1998 to $2.87 billion in 2013. Whereas the

first two RECs (ECOWAS and ECCAS) showed a major differences between our two sub-periods,

in COMESA, the growth gap between the two sub-periods is very thin (less than 3 percentage

points). Over the entire period (1998-2013), the volume of intra-regional agricultural trade has

shown a significant increase (22%).

The value of intra-regional trade of agricultural commodities in SADC has displayed the lowest

overall annual growth of 10%, with a nominal value which has increased from $871 million in

1998 to $3.82 billion in 2013. During the first sub-period, an 8% increase was reported, against

13% over the second sub-period. In value, intra-regional agricultural trade has increased after the

international food crisis. However, the volume trend is totally different over the two sub-periods.

A greater average increase was noted over the first sub-period (16%) compared to growth in the

second sub-period (13%). Therefore, the nominal intra-regional increase observed between the

sub-periods is essentially a price effect. Nevertheless, over the whole period (1998-2013), the

intra-regional trade volume increase (14%) is greater than its value increase (10%).

3.4 Direction of agricultural exports and imports in intra-African and intra-regional markets

The previous section presented trends in intra-African and intra-RECs trade over the period from

1998-2013. But, no mention was made of which country or REC leads in intra-regional trade.

Therefore, the target of this section is to shed light on that aspect. Before deepening the analysis

of intra-African and intra-RECs trade direction, Table 3.1 summarizes trading networks between

various African regions, by presenting the average trade flow (exports/imports) between them

over recent years (2010-2013). Exporting regions are in rows and importing ones are in columns.

Intra-regional trade is shown by the diagonal elements in bold.

43

Table 3.1. Value of intra- and inter-regional trade in agricultural products in Africa, 2010-2013

average (billion US dollars)

Regional market destinations

AFRICA ECOWAS ECCAS COMESA SADC SSA

Ex

po

rter

s

AFRICA 11.69 2.93 1.73 5.26 4.07 9.53

ECOWAS 2.40 1.91 0.13 0.06 0.09 2.13

ECCAS 0.30 0.01 0.16 0.15 0.08 0.27

COMESA 4.50 0.10 0.54 2.94 1.67 3.39

SADC 4.46 0.30 0.96 2.60 3.43 4.29

SSA 9.28 2.47 1.53 4.09 3.91 8.39

Source: BACI and authors’ calculation, 2016.

One interesting statistic is the ratio of intra-regional trade (ECOWAS, ECCAS, COMESA and

SADC) to the total trade of the REC with Africa as a whole. This statistic will show how one

REC’s trade with the continent is concentrated in that REC; it could be seen as an indicator of

regional trade integration. The results show that ECOWAS is the REC with the highest trade

integration with a ratio of 0.79, followed by SADC with 0.77, COMESA with 0.65 and ECCAS

with 0.52. Therefore, with the exception of ECCAS countries, each REC exchanges the principal

part of its trade with Africa inside its own bloc (UNCTAD, 2013). For example, ECOWAS’s intra-

regional agricultural trade represents, on average over 2010 and 2013, around 80% of its total trade

with Africa.

In terms of intra-African agricultural trade, Figure 3.5 represents the weight of individual RECs in

terms of origin and destination. As destinations or origins of intra-African trade, COMESA (42%

of exports and 34% of imports) and SADC (37% of exports and 42% of imports) are the main

regions, while ECCAS (14% of exports and 3% of imports) is last. One could note that COMESA

and SADC have opposite patterns. In fact, COMESA has gained trade share (exports and imports)

over the considered period while SADC countries have lost some. COMESA’s export share has

increased from 40% between 1998 and 2006 to 45% between 2007 and 2013, and the region’s

import share has risen from 32% between 1998 and 2006 to 37% between 2007 and 2013. In

contrast, SADC’s export share has decreased from 39% between 1998 and 2006 to 34% between

2007 and 2013, and the region’s import share has fallen from 46% between 1998 and 2006 to 38%

between 2007 and 2013.

44

Figure 3.5. Regional share in intra-African agricultural trade

Source: BACI and authors’ calculation, 2016. Note: (a) export value, (b) import value.

Inside any specific African REC, many efforts and political commitments exist to promote political

co-operation and economic integration. As seen previously, those commitments have increased

intra-regional trade. The objectives of the following subsections are to present the importance (in

terms of exports and imports) of different countries inside their regional bloc. Tables 3.2 to 3.5

present individual countries’ export and import shares in intra-regional trade (average shares for

1998-2006, 2007-2013 and 1998-2013).

Table 3.2. ECOWAS intra-regional trade share by country (%)

1998-2006 2007-2013 Overall

Exports Imports Exports Imports Exports Imports

Benin 6.3 5.5 5.9 3.9 6.0 4.5

Burkina Faso 14.8 7.7 4.2 10.2 7.9 9.3

Cape Verde 0.1 0.1 0.1 0.2 0.1 0.2

Côte d'Ivoire 25.0 15.3 26.8 12.5 26.2 13.5

Gambia 0.5 1.5 1.0 1.5 0.8 1.5

Ghana 3.7 10.3 11.1 8.9 8.5 9.3

Guinea 2.6 2.2 2.0 2.8 2.2 2.6

Guinea-Bissau 0.1 1.1 1.0 0.8 0.7 0.9

Liberia 0.1 0.4 0.1 0.7 0.1 0.6

Mali 17.7 8.4 6.0 9.7 10.1 9.3

Niger 10.9 8.5 17.9 5.8 15.5 6.7

Nigeria 3.0 14.8 6.9 27.6 5.5 23.1

Senegal 8.8 12.2 12.6 9.2 11.3 10.2

Sierra Leone 0.0 0.3 0.0 0.7 0.0 0.5

Togo 6.3 11.7 4.2 5.6 4.9 7.7

Source: BACI and authors’ calculation, 2016.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

(a)

ECOWAS ECCAS COMESA SADC

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

(b)

ECOWAS ECCAS COMESA SADC

45

Inside ECOWAS, Côte d’Ivoire remains the biggest exporter of agricultural products in the region

with about 26% of total intra-regional trade. Other important exporters to the region are Niger

(15.5%), Senegal (11.3%) and Mali (10.1%). In terms of destination, Nigeria is the main importer

of those commodities from the region with 23% of total trade, followed by Côte d’Ivoire (13.5%)

and Senegal (10.2%). Some countries have seen their exporting performance worsen over the two

sub-periods while others became more performant. For example, Burkina Faso’s export share has

fallen from 14.8% to 4.2%. In contrast, Ghana’s export share has increased from 3.7% to 11.1%.

Table 3.3. ECCAS intra-regional trade share by country (%)

1998-2006 2007-2013 Overall

Exports Imports Exports Imports Exports Imports

Angola 0.6 1.2 0.1 3.2 0.2 2.5

Burundi 2.0 0.8 2.2 3.9 2.2 3.5

Cameroon 50.5 20.8 41.5 11.7 42.7 14.4

Central African

Republic 1.6 10.9 0.4 8.6 0.8 9.2

Chad 4.1 11.6 0.1 8.6 1.3 9.7

Congo 16.9 18.7 11.7 18.7 13.1 18.5

Democratic Congo 0.5 5.2 4.9 21.0 3.4 15.9

Equatorial Guinea 0.1 6.5 0.0 7.1 0.0 7.0

Gabon 22.3 21.5 17.1 13.3 18.0 15.7

Rwanda 1.2 1.8 22.0 3.2 18.1 3.0

Sao Tome and Principe 0.2 0.9 0.1 0.6 0.1 0.7

Source: BACI and authors’ calculation, 2016.

For ECCAS countries, Cameroon controlled the export market inside this REC with around 43%

of the regional agricultural products market. Rwanda (18.1%), Gabon (18%) and Congo (13%) are

the other main exporters of agricultural products. In terms of destination, Congo (18.5%),

Democratic Republic of the Congo (DRC) (15.9%), Gabon (15.7%) and Cameroon (14.4%) are

the main markets for agricultural products. It is worth noting the impressive performance of

Rwanda, which has seen its export share rise from 1.2% over 1998-2006 to 18.1% between 2007

and 2013.

46

Table 3.4. COMESA intra-regional trade share by country (%)

1998-2006 2007-2013 Overall

Exports Imports Exports Imports Exports Imports

Burundi 0.4 1.4 0.4 1.6 0.4 1.6

Comoros 0.0 0.6 0.1 0.3 0.1 0.4

DRC 0.7 6.8 0.4 9.8 0.5 9.2

Djibouti 2.0 5.8 0.8 3.2 1.2 4.0

Egypt 5.6 22.6 21.1 14.3 17.0 16.6

Eritrea 0.0 0.8 0.1 1.1 0.1 1.0

Ethiopia 7.4 4.0 7.2 1.2 7.2 2.0

Kenya 28.0 13.2 21.1 11.6 22.9 12.2

Libya 0.0 0.2 0.1 10.2 0.1 8.3

Madagascar 1.3 2.5 0.7 2.5 0.8 2.5

Malawi 5.8 4.7 5.0 3.1 5.2 3.6

Mauritius 2.7 4.1 2.4 4.8 2.5 4.7

Rwanda 2.2 3.3 3.2 4.0 3.0 3.9

Seychelles 2.2 0.6 1.3 0.3 1.6 0.4

Sudan 6.4 11.9 2.6 16.6 3.5 13.7

Uganda 13.5 4.9 15.5 4.7 15.0 4.9

Zambia 11.9 6.8 15.5 3.0 14.6 4.0

Zimbabwe 9.9 5.6 2.3 7.6 4.3 7.2

Source: BACI and authors’ calculation, 2016.

Inside COMESA, Kenya (22.9%), Egypt (17%), Uganda (15%) and Zambia (14.6%) are the

leading exporters of agricultural products. In terms of imports, Egypt (16.6%), Sudan (13.7%) and

Kenya (12.2%) are the main markets for those products. Showing exceptional performance,

Egypt’s export share in the region has been multiplied by four, passing from 5.6% between 1998

and 2006 to 21.1% over 2007-2013.

Table 3.5. SADC intra-regional trade share by country (%)

1998-2006 2007-2013 Overall

Exports Imports Exports Imports Exports Imports

Angola 0.2 15.1 0.1 11.4 0.1 12.5

Democratic Congo 0.1 6.5 0.0 10.7 0.0 9.5

Madagascar 0.8 2.6 0.4 2.7 0.5 2.7

Malawi 4.0 8.1 5.1 5.5 4.7 6.3

Mauritius 1.5 7.7 2.3 6.5 2.0 6.9

Mozambique 4.8 13.3 5.0 13.6 4.9 13.5

SACU 59.9 18.3 57.0 12.6 57.8 14.3

Seychelles 1.2 0.9 1.3 0.6 1.2 0.7

Tanzania 2.1 3.9 3.8 2.9 3.3 3.2

Zambia 10.1 9.6 16.0 8.4 14.2 8.8

Zimbabwe 15.5 13.9 9.2 25.0 11.0 21.7

Source: BACI and authors’ calculation, 2016.

47

Within SADC, SACU countries, which are composed of Botswana, Lesotho, Namibia, Swaziland

and South Africa, constitute the major exporters with around 57% of intra-regional trade in

agricultural commodities. But in terms of imports, they are the second biggest market (14.3%)

behind Zimbabwe (21.7%). Mozambique is the third market for agricultural products in the region

with 13.5% of intra-regional trade.

3.5 Changes in export and import shares in intra-African and intra-regional agricultural markets

The bubble charts presented in the next subsections show primarily the changes in trade (imports

and exports) for each of the two sub-periods. The average trade in value for the sub-period is

represented on the X axis. The average trade in volume over the considered period is represented

on the Y axis. Each bubble corresponds to a country, and the bubble size shows the country’s

average GDP over the sub-period. This type of graph is chosen in order to capture whether the

observed changes in trade issue from a price effect or a volume effect. In addition, it provides an

idea of the size of the economies within the RECs.

3.5.1 ECOWAS

The changes in intra-ECOWAS agricultural imports are shown in Figure 3.6. It is found that in the

aggregate, the total value and volume of agricultural imports has doubled in the ECOWAS zone.

At the country level, we note that all countries have at least doubled the value of their agricultural

purchases from ECOWAS, except Togo, for which a 14% increase in the value of agricultural

imports from the ECOWAS zone is observed.

48

Figure 3.6. ECOWAS import changes

Source: BACI and authors’ calculation, 2016.

Note: Benin (BEN), Burkina Faso (BF), Cape Verde (CAPV), Côte d'Ivoire (CIV), Gambia (GAMB), Ghana

(GHA), Guinea (GUI), Guinea-Bissau (GUIB), Liberia (LIB), Mali (MAL), Niger (NIG), Nigeria (NIGA), Senegal

(SEN), Sierra Leone (SIER), Togo (TOG)

Over the two periods, the largest importers remain Nigeria and Côte d’Ivoire, which are the two

largest economies of the zone. Nigeria’s agricultural imports quadrupled in value and

approximately doubled in volume between the two periods. Other countries experiencing an

increase in imports in value and volume include Benin, Burkina Faso, Côte d’Ivoire, Guinea,

BEN BF

CAPV CIV

GAMB

GHA

GUI

GUIB

LIB

MAL

NIG

NIGA

SENSIERTOG

-50

0

50

100

150

200

-20 0 20 40 60 80 100 120

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

1998-2006

BEN BF

CAPV

CIVGAMB GHA

GUI

GUIB

LIBMAL

NIG

NIGA

SENSIER TOG

-200

0

200

400

600

800

1000

-100 0 100 200 300 400 500 600

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

2007-2013

49

Guinea-Bissau, Mali, Senegal and Sierra Leone. However, it should be noted that Senegal is the

country that buys the fewest agricultural products from ECOWAS in volume. This country is the

fourth largest economy in the zone after Nigeria, Côte d’Ivoire and Ghana. It is also in the top five

in import values in the two periods, as shown in Figure 3.6.

Figure 3.7. ECOWAS export changes

Source: BACI and authors’ calculation, 2016.

Note: Benin (BEN), Burkina Faso (BF), Cape Verde (CAPV), Côte d'Ivoire (CIV), Gambia (GAMB), Ghana

(GHA), Guinea (GUI), Guinea-Bissau (GUIB), Liberia (LIB), Mali (MAL), Niger (NIG), Nigeria (NIGA), Senegal

(SEN), Sierra Leone (SIER), Togo (TOG)

BEN

BF

CAPV

CIV

GAMB

GHA

GUIGUIBLIB

MALNIG

NIGASEN

SIER

TOG

-50

0

50

100

150

200

250

300

-50 0 50 100 150 200

Exp

ort

s vo

l (1

00

0 T

on

s)

Exports (million $)

1998-2006

BENBF

CAPV

CIV

GAMB

GHA

GUIGUIBLIB

MALNIGNIGA

SEN

SIER

TOG

-200

-100

0

100

200

300

400

500

600

700

800

-100 0 100 200 300 400 500

Exp

ort

s v

ol (

10

00

To

ns)

Exports (million $)

2007-2013

50

For the rest of the ECOWAS countries (Cape Verde, The Gambia, Ghana, Liberia, Niger, Togo),

an increase in the value of imports is noted, but the volumes remain almost unchanged. As a result,

the growth in value of imports recorded for these countries is due to the rising prices observed over

the 2007-2013 period. On the export side (Figure 3.7), it is noted that the total value of agricultural

exports has also doubled on aggregate. Aside from Burkina Faso, Mali and Sierra Leone, all other

countries have at least doubled the value of their average exports to the ECOWAS area. In volume

terms, it is also noted in the aggregate that intra-area agricultural sales have also doubled. However,

some countries such as Burkina Faso, Cape Verde, Mali, Niger and Sierra Leone have not

increased the volume of their agricultural shipments to ECOWAS destinations. At the country

level, Côte d’Ivoire remains in both periods the largest agricultural exporter in the area in value.

However, it is observed that during the second period Ghana has become the first supplier of

agricultural products for other ECOWAS countries before Côte d'Ivoire. Indeed, Ghana has

multiplied the volume of its agricultural exports to the region by 11. During the second period,

Niger is positioned as the second largest exporter of the zone in value with a quadrupling of the

value of its exports, but the volumes remain almost unchanged over the two periods. Niger has

taken advantage of the rising prices of livestock products during the 2007-2013 period. In contrast,

Mali and Burkina Faso, which were the main exporters behind Côte d'Ivoire in the first period, do

not benefit from the increasing agricultural prices. Instead they have experienced decreases in the

value of exports by 18% and 32%, respectively. As mentioned before, these two countries’ export

volumes have remained almost unchanged compared to the 1998-2006 period. Regarding Mali,

the political crisis that occurred in late 2011 could be an explanation for this decline.

3.5.2 ECCAS

Figure 3.8 illustrates the import changes in the ECCAS zone for the two periods. The total value

and volume of intra-ECCAS agricultural imports have tripled between the two periods. All the

countries in the zone, without exception, have at least doubled their imports in value. In terms of

volume, this upward trend in agricultural purchases from the area is observed except for Gabon

and Rwanda, where the level of import volumes remained stable over the two periods. Between

the two periods, the DRC is the country that has experienced the greatest growth in agricultural

purchases from its neighbours. This is due to rising prices in the second period. Actually, the DRC

is only the seventh importer in the area by volume over the period 2007-2013.

51

Figure 3.8. ECCAS import changes

Source: BACI and authors’ calculation, 2016.

Note: Angola (ANG), Burundi (BUR), Cameroon (CAM), Central African Republic (CAR), Chad (CHA), Congo

(CONG), Democratic Republic of the Congo (DRC), Equatorial Guinea (EGUI), Gabon (GAB), Rwanda (RWA),

Sao Tome and Principe (SAO)

ANG

BUR

CAM

CAR

CHA

CONGDRC

EGUI

GAB

RWA

SAO

-5

0

5

10

15

20

25

-2 0 2 4 6 8 10 12 14 16

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

1998-2006

ANG

BUR CAM

CAR

CHA

CONG

DRC

EGUI

GAB

RWA

SAO

-5

0

5

10

15

20

25

30

35

40

45

-5 0 5 10 15 20 25 30 35 40 45

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

2007-2016

52

In terms of exports, Cameroon remains the largest exporter of agricultural products in the ECCAS

area by doubling the value of its agricultural sales and the volume of its shipments to its neighbours

between the two periods. Two other major exporters of the zone, Congo and Gabon, also

experienced almost identical situations. However, Rwanda and the DRC are the countries that have

made the most progress in terms of exports. In fact, Rwanda has multiplied the value of its

agricultural exports in the area by 49 while the DRC has multiplied the value of its exports to its

neighbours in the area by 25. In volume, Rwanda and the DRC have multiplied the volume of

shipments by 25 and 31, respectively (Figure 3.9). Regarding Rwanda, which became the second

largest exporter of the area behind Cameroon, its performance is linked with the economic

performance recorded between 2000 and 2012 after the political crisis. In addition, Rwanda has

also intensified its commercial exchanges with neighbouring Kenya and DRC10.

Figure 3.9. ECCAS export changes

10 Rwanda is also part of COMESA with these two countries. We will discuss its performance further in the

COMESA subsection.

ANGBUR

CAM

CARCHA

CONG

DRCEGUI

GAB

RWASAO

-10

-5

0

5

10

15

20

25

30

35

40

-5 0 5 10 15 20 25 30 35

Exp

ort

s vo

l (1

00

0 T

on

s)

Exports (million $)

1998-2006

53

Source: BACI and authors’ calculation, 2016.

Note: Angola (ANG), Burundi (BUR), Cameroon (CAM), Central African Republic (CAR), Chad (CHA), Congo

(CONG), Democratic Republic of the Congo (DRC), Equatorial Guinea (EGUI), Gabon (GAB), Rwanda (RWA),

Sao Tome and Principe (SAO)

3.5.3 COMESA

Figure 3.10 shows the variations in terms of agricultural imports for the COMESA countries. In

the aggregate, trade has intensified in this area. Indeed, the value of imports was quadrupled while

traded volumes were doubled. In general, all countries in the region have at least doubled the value

of their purchases from their neighbours with the exception of Ethiopia for which the import values

remained almost unchanged over the two periods.

Regarding the volume variations, the trend remains the same, except for Ethiopia, Malawi and

Zambia. Regarding the latter, a highly significant decrease in the volume of agricultural products

imported from the area is observed. Indeed, the volume of imports in the second period is about

18 times lower compared to the first period. Despite this reduction, import values are found to

have doubled. Several elements of explanation could be advanced. First, import prices in this

country are very high. Second, given that Zambia is also a member of another REC, it may be that

this decline is offset by a sharp increase in quantities imported from the SADC area. Finally,

Zambia could have launched an agricultural self-sufficiency policy.

ANG

BUR

CAM

CARCHA

CONGDRC

EGUI

GAB

RWA

SAO

-10

0

10

20

30

40

50

60

70

-20 -10 0 10 20 30 40 50 60 70 80

Exp

ort

s vo

l (1

00

0 T

on

s)

Exports (million $)

2007-2013

54

Unlike Zambia, Madagascar has multiplied the volume of agricultural imports by 20, becoming

the largest importer in volume of the area before the largest economies of the region including

Egypt, Libya, Kenya, the DRC and Sudan. However, Libya has also stepped up its agricultural

orders from COMESA in the second period, 2007-2013. Indeed, they are multiplied by 225 with

respect to the value of the first period and by 280 for the quantities. Possible explanations include,

among others, the Libyan crisis that took place in 2011 and which has limited supplies to Libya

from Tunisia by land. Consequently, it appears that Libya buys more from COMESA.

In addition, three COMESA countries, Burundi, the DRC and Rwanda, are also members of the

ECCAS area. Regarding Rwanda, and despite the intensification of its exchanges in the ECCAS

zone, it should be noted that the values and volumes of its imports from the COMESA are

significantly higher than those from the ECCAS area. In other words, Rwanda purchases mainly

within COMESA. This is also true for Burundi and the DRC.

55

Figure 3.10. COMESA import changes

Source: BACI and authors’ calculation, 2016.

Note: Burundi (BUR), Comoros (COM), Democratic Republic of the Congo (DRC), Djibouti (DJI), Egypt (EGY),

Eritrea (ERI), Ethiopia (ETH), Kenya (KEN), Libyan Arab Jamahiriya (LIB), Madagascar (MAD), Malawi

(MALW), Mauritius (MAU), Rwanda (RWA), Seychelles (SEY), Sudan (SUD), Uganda (UGA), Zambia (ZAM),

Zimbabwe (ZIM)

BURCOM

DRCDJI

EGY

ERI ETH

KEN

LiBMAD MALWMAURWA

SEY

SUDUGA

ZAM

ZIM

-200

0

200

400

600

800

1000

1200

1400

-50 0 50 100 150 200

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

1998-2006

BURCOM

DRC

DJI

EGY

ERIETH

KEN

LiB

MAD

MALW

MAURWA

SEY

SUD

UGA

ZAM

ZIM

-200

-100

0

100

200

300

400

500

600

700

800

900

-100 0 100 200 300 400 500 600

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

2007-2013

56

Figure 3.11. COMESA export changes

Source: BACI and authors’ calculation, 2016.

Note: Burundi (BUR), Comoros (COM), Democratic Republic of the Congo (DRC), Djibouti (DJI), Egypt (EGY),

Eritrea (ERI), Ethiopia (ETH), Kenya (KEN), Libyan Arab Jamahiriya (LIB), Madagascar (MAD), Malawi

(MALW), Mauritius (MAU), Rwanda (RWA), Seychelles (SEY), Sudan (SUD), Uganda (UGA), Zambia (ZAM),

Zimbabwe (ZIM)

BURCOMDRCDJIEGY

ERIETH

KEN

LIBMAD

MALW

MAURWASEYSUD

UGAZAMZIM

-200

0

200

400

600

800

1000

1200

1400

-50 0 50 100 150 200 250

Exp

ort

s vo

l (1

00

0 T

on

s)

Export (million $)

1998-2006

BURCOMDRCDJI

EGY

ERI

ETH

KEN

LIB

MADMALW

MAURWASEY SUD

UGA

ZAM

ZIM

-200

0

200

400

600

800

1000

1200

1400

-100 0 100 200 300 400 500 600 700

Exp

ort

s vo

l (1

00

0 T

on

s)

Export (million $)

2007-2013

57

On the export side (Figure 3.11), it is found that the total value of intra-COMESA agricultural

exports has quadrupled, while volumes have doubled. At the country level, it is observed that all

countries in the region have at least doubled their agricultural sales (volume and value) in the area

over the two periods, with the exception of Djibouti, Malawi, Sudan and Zimbabwe. In Djibouti,

Malawi and Sudan, values have increased slightly, while they declined slightly for Zimbabwe.

Quantities shipped remained almost stable for Sudan. However, they have dropped more than half

for the other three countries. In contrast, Egypt is the country that has increased its agricultural

trade with its neighbours in the COMESA region the most, becoming the leading supplier of

agricultural products before Kenya, Uganda and Zambia. Concerning Rwanda, Burundi and the

DRC, these countries have at least tripled their trade in volume and value with other COMESA

countries. Compared to the ECCAS zone, it is noted that these countries sell more in the COMESA

region than in the ECCAS area.

3.5.4 SADC

Figure 3.12 shows the changes observed in imports within SADC. However, it should be noted

that in the database used, BACI, South Africa, Namibia, Botswana, Swaziland and Lesotho are

grouped within SACU. In fact, information is provided only for the SACU group, rather than for

the individual countries. On aggregate, it is found firstly that imports doubled in value and also

decreased approximately 20% in quantity. Malawi, Mozambique, Tanzania and Zambia are the

countries affected by the drop in traded quantities. Regarding Zambia, also a member of

COMESA, a sharp decline is also observed in the volume of its agricultural imports from its SADC

neighbours. Indeed, volumes were divided by 6. It seems that the trend for Zambia within

COMESA is also valid for SADC. This reinforces the hypothesis previously issued on the possible

implementation of a self-sufficiency policy to reduce imports, accompanied by a protectionist

policy. To a lesser extent, Malawi, also a member of COMESA, has also decreased its agricultural

purchases from SADC. Nevertheless, these two countries buy more within the SADC zone than

within the COMESA zone. Other countries concerned by the decline of imported quantities are

Tanzania and Mozambique. In contrast, the other countries of the zone have experienced an

increase in volumes purchased from neighbouring countries in SADC. Between the two periods,

Zimbabwe became the first buyer of agricultural products before Mozambique and SACU.

Furthermore, it is noted that Zimbabwe is a member of COMESA but buys more within SADC.

58

This observation is also true for the DRC, also a member of COMESA and ECCAS. For Angola,

also a member of ECCAS, the exchanges are also more intense in the SADC region. In general,

all countries that are at the same time members of SADC and another REC tend to import more

from the SADC area.

Figure 3.12. SADC import changes

ANGDRCMADMAL

MAUR

MOZ

SACU

SEYTAN

ZAM

ZIM

-500

0

500

1000

1500

2000

2500

3000

-50 0 50 100 150 200 250

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

1998-2006

59

Source: BACI and authors’ calculation, 2016.

Note: Angola (ANG), Democratic Republic of the Congo (DRC), Madagascar (MAD), Malawi (MALW), Mauritius

(MAU), Mozambique (MOZ), Southern African Customs Union (SACU), Seychelles (SEY), United Rep. of

Tanzania (TAN), Zambia (ZAM), Zimbabwe (ZIM)

Figure 3.13 shows the intra-SADC agricultural exports. It is observed in the aggregate that

exports values have increased and at the same time export volumes have decreased. In both

periods, SACU remains the top seller. Indeed, the value of exports from SACU exceeds the

aggregate exports of all other members of SADC. However, it should be noted that the quantities

exported by SACU have remained unchanged and are relatively low. SACU is the 10th exporter

in volume over the 11 countries.

ANG

DRC

MAD

MALMAUR

MOZSACU

SEYTAN

ZAM

ZIM

-200

0

200

400

600

800

1000

1200

1400

1600

-200 -100 0 100 200 300 400 500 600 700 800 900

Imp

ort

s vo

l (1

00

0 T

on

s)

Imports (million $)

2007-2013

60

Figure 3.13. SADC export changes

Source: BACI and authors’ calculation, 2016.

Note: Angola (ANG), Democratic Republic of the Congo (DRC), Madagascar (MAD), Malawi (MALW), Mauritius

(MAU), Mozambique (MOZ), Southern African Customs Union (SACU), Seychelles (SEY), United Rep. of

Tanzania (TAN), Zambia (ZAM), Zimbabwe (ZIM)

ANGDRCMADMALW

MAU

MOZ

SACUSEYTAN

ZAM ZIM

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

4000

-200 -100 0 100 200 300 400 500 600 700 800 900

Exp

ort

s vo

l (1

00

0 T

on

s)

Export (million $)

1998-2006

ANG

DRCMADMALW

MAU

MOZ

SACUSEY

TAN

ZAMZIM

-500

0

500

1000

1500

2000

2500

-500 0 500 1000 1500 2000

Exp

ort

s vo

l (1

00

0 T

on

s)

Exports (million $)

2007-2013

61

Products exported by this regional entity appear to be more expensive. Furthermore, concerning

the other SADC countries which are also member of COMESA (Zambia, Zimbabwe, Seychelles,

Malawi, Madagascar, and DRC), it is noted that the quantities shipped in the COMESA region are

greater. Only Madagascar exports more in value to COMESA than SADC.

In the next section, the changes in the composition of products traded between the different RECs

will be presented.

3.6 Changes in composition of intra-African and intra-regional agricultural exports and imports

Table 3.6 shows the trade variations in both periods by group of products. It is observed on

aggregate that the share of cereals in trade between African countries remained relatively stable.

Indeed, it was around 7% during both of the two periods. In addition, an increase in shares of dairy

products and other livestock products, fruits and processed food is observed in both periods. In

contrast, intra-African trade in coffee and oilseeds has slightly fallen.

Table 3.6. Changes in composition of intra-African trade (commodity groups)

Source: BACI and authors’ calculation, 2016.

1998-2006 2007-2013 1998-2006 2007-2013 1998-2006 2007-2013 1998-2006 2007-2013 1998-2006 2007-2013

Cereals 6,9 6,6 3,9 4,8 0,6 4,2 7,0 8,7 11,8 9,5

Coffee 10,4 7,4 0,4 1,5 0,9 0,5 27,4 17,0 2,8 2,2

Dairy products 2,8 3,5 3,3 2,9 1,9 3,7 1,5 4,4 3,7 3,3

Fish products 7,5 8,2 6,4 7,4 1,0 1,3 3,1 2,1 5,5 7,6

Fruits 2,5 3,3 2,7 2,4 0,1 0,2 1,2 1,1 2,8 2,8

Live cattle 2,8 3,0 10,5 8,8 1,3 3,5 1,6 3,7 1,3 1,0

Meats 0,8 0,8 0,7 1,6 0,2 0,2 0,6 0,2 1,6 1,4

Oilseeds 2,7 2,5 2,2 1,9 0,8 0,2 4,5 2,9 2,8 2,8

Processed Food 38,5 41,8 27,5 46,3 75,5 66,2 30,3 37,3 45,5 46,1

Others 25,0 22,8 42,4 22,5 17,6 19,8 22,9 22,5 22,3 23,2

Total 100 100 100 100 100 100 100 100 100 100

Africa ECOWAS ECCAS COMESA SADC

62

At the product level, Figure 3.14 shows the 10 most traded agricultural commodities in Africa.

Between the two periods, it is not noticed a major change in the composition of intra-African trade.

Indeed, only two products that were present in the first period are out of the top 10 most traded

goods between African countries in the second period. These products are cotton and food

preparations nes (not elsewhere specified). In contrast, vegetables and wheat flour are among the

10 most traded products in the second period but not the first. Also, it is observed that fishery

products become the most traded product between African countries in the second period. In the

next subsections, the changes observed in each REC will be presented.

Figure 3.14. Top 10 most traded commodities (Intra-Africa)

Source: BACI and authors’ calculation, 2016.

3.6.1 ECOWAS

Regarding ECOWAS trade by group of products (Table 3.6), trade increases in cereals, coffee,

fish products, dairy products, meat and processed food are noted. This latter group accounts for

almost the half of the trade of the second period, with an almost 20 percentage point increase

between the two periods.

0123456789

1998-2006

012345678

2007-2013

63

Figure 3.15. Top 10 most traded commodities in the ECOWAS zone

Source: BACI and authors’ calculation, 2016.

At the product level, it is found that cotton, which was the first traded product at the ECOWAS

level with a 25% share of trade between 1998 and 2006, is no longer part of the top 10 traded

products in the region. In contrast, trade in cigars and cheroots has intensified and the share of this

product quadrupled. To a lesser extent, exchanges of palm oil and frozen fish products have also

increased. In addition, it is noted that rice and pasta are among the 10 most traded food and

agricultural products in the ECOWAS region during the second period (Figure 3.15). For rice, it

is likely due to the rice self-sufficiency policies launched by many ECOWAS countries to cope

with the 2007-2008 food price crisis.

3.6.2 ECCAS

In the ECCAS zone, it is found during both of the two periods that processed foods account for

about 2/3 of the total trade share, despite a roughly 9-point decline in the trade of this group of

products between the two periods. In addition, cereals and fish products are the other most traded

groups (Table 3.6).

0

5

10

15

20

25

30

1998-2006

0

5

10

15

20

2007-2013

64

Figure 3.16. Top 10 most traded commodities in the ECCAS zone

Source: BACI and authors’ calculation, 2016.

At a more detailed level, sugar is still the most traded product, although its share has declined over

the second period. Generally, the composition of trade in the ECCAS zone does not change much,

even if a decreasing trend is noted for each product traded in the first period and still in the top 10

during the second period. For example, trade in cigars and cheroots halved between the two

periods. In terms of new products traded, it is found that wheat flour, sauces, milk and cream are

among the 10 most traded products in the ECCAS zone during the second period (Figure 3.16).

3.6.3 COMESA

It is found in both periods that the group of processed food products occupies the most important

position in intra-COMESA trade with over a third of the total trade share. Coffee trade has

decreased (-10 points), but represents a major product in intra-Community trade. As in the two

RECs presented above, an increase in cereal trade is noted. In addition, trade shares of dairy

products and live cattle have also increased. (Table 3.6).

0

5

10

15

20

25

1998-2006

0

5

10

15

20

2007-2013

65

Figure 3.17. Top 10 most traded commodities in the COMESA zone

Source: BACI and authors’ calculation, 2016.

Figure 3.17 gives an indication of the detail of the products traded. In general, the composition of

traded goods has not changed much. Only cotton, other oil seeds, and vegetables are no longer

among the most traded products. However, palm oil, dried leguminous vegetables and cigars and

cheroots are part of the 10 most traded products in the area during the second period.

3.6.4 SADC

As with the other RECs, processed food products are still the most important group, representing

nearly half of the trade over the two periods. In addition, the trade shares of fruits and oilseeds

have remained unchanged in both periods. Except for fish products, for which exchanges have

improved, it is found that all other group of products have experienced a drop in trade compared

to the first period (Table 3.6).

At the product level, the composition of trade is fairly stable. Sugar is still the most traded

commodity with an almost unchanged share in both periods. Maize and tobacco are the other two

most traded products, even if exchanges have fallen during the second period. However, a doubling

of the share of frozen fish products is found.

Furthermore, it is noted that oil trade has increased during the second period. Indeed, two types of

oil (cotton-seed oil and soya-bean oil) are now part of the top 10 most traded commodities, while

0

5

10

15

20

25

1998-2006

0

5

10

15

2007-2013

66

drinks (waters and beer made from malt) are no longer part of the 10 most traded commodities

(Figure 3.18).

Figure 3.18. Top 10 most traded commodities (Intra-SADC)

Source: BACI and authors’ calculation, 2016.

3.7 Changes in unit values of intra-African and intra-regional agricultural exports and imports

Trade unit values (TUV) are usually used as proxies for trade prices. They measure, for individual

commodity classes in a particular period, the total value of shipments divided by the corresponding

total quantity (IMF, 2009). To analyze the trends of this indicator for intra-African and intra-

regional trade, we use the Trade Unit Values dataset by Berthou and Emlinger (2011). This

database contains bilateral trade unit values at Harmonized-System 6-digit commodity categories.

In this database, 45 African countries are represented. Therefore, the following discussions are

related to the unit values (harmonic averages computed per year) of agricultural trade between

those 45 countries.

0

1

2

3

4

5

6

7

8

9

10

1998-2006

0123456789

10

2007-2013

67

Figure 3.19. UV changes for intra-African trade, $ per ton

Source: TUV Database and authors’ calculations, 2016.

Figure 3.19 gives the trends of intra-African agricultural trade unit values over the period 2000-

2013. The average unit values for intra-African agricultural trade have increased over the period,

with 3.54% growth for exports and 2.90% for imports. Export unit values have displayed slightly

greater growth over the period 2007-2013 (3.91%) compared to the period 1998-2006 (3.12%). In

contrast, import unit values have shown a slower increase during the post-crisis period (1.29%)

relative to the period before the crisis (4.81%).

Figure 3.20. UV changes for intra-ECOWAS trade, $ per ton

Source: TUV Database and authors’ calculations, 2016.

Export unit values for intra-ECOWAS agricultural trade have decreased over 1998-2013 (Figure

3.20), with a decrease of -4.67%. However, imports have become more expensive, with an overall

growth of 3.23%. Therefore, it is easier to export into the region than to import from the region.

0

500

1000

1500

2000

2500

3000

3500

2000 2002 2004 2006 2008 2010 2012 2014

Export UV Import UV

0

5000

10000

15000

20000

25000

2000 2002 2004 2006 2008 2010 2012 2014

Export UV Import UV

68

Since important progress in terms of economic integration has been made, one could attribute the

increase of import unit values to non-tariff measures, corruption, etc.

Figure 3.21. UV changes for intra-ECCAS trade, $ per ton

Source: TUV Database and authors’ calculations, 2016.

Inside ECCAS, a large gap is noticeable over the first sub-period compared to the second sub-

period (Figure 3.21). A 25.85% increase in export unit values and a 15.46% increase in import unit

values were reported over the 1998-2006 period, while export unit values (-4.83%) and import unit

values (-4.51%) have shown a decrease over the second sub-period. This may be interpreted as an

improvement in regional integration over the second period. It is worth noticing that trade unit

values in ECCAS are the highest among RECs.

Figure 3.22. UV changes for intra-COMESA trade, $ per ton

Source: TUV Database and authors’ calculations, 2016.

0

50000

100000

150000

200000

2000 2002 2004 2006 2008 2010 2012 2014

Export UV Import UV

0

5000

10000

15000

20000

25000

2000 2002 2004 2006 2008 2010 2012 2014

Export UV Import UV

69

Inside COMESA, trade unit values of agricultural products are more stable over the period (Figure

3.22). Export unit values showed a 4% increase while import unit values displayed 3.43% growth.

Over the two sub-periods, export unit values have registered a decrease in growth (5.89% over

1998-2006 and 4.09% over 2007-2013) but import unit values have shown increased growth

(1.77% over 1998-2006 and 3.5% over 2007-2013).

Figure 3.23. UV changes for intra-SADC trade

Source: TUV Database and authors’ calculations, 2016.

Export and import unit values for intra-SADC trade have shown steady growth over the period

considered (Figure 3.23). Exports displayed overall unit value growth of 7.5% and imports showed

a 5.7% increase.

Following the 2011 methodological note by OECD, we computed the export/import value index

for agricultural and non-agricultural products using the Fisher index (see Table A2 in the Annex

for the evolution of the export/import value index). Then, we derived the terms of trade for

different commodity groups as displayed in Figure 3.24. Before the recent food crisis, African

economies sold cheaper agricultural products but bought them more expensively from outside. On

the other hand, the terms of trade for non-agricultural products show that almost all RECs (with

the exception of ECCAS) have good prices for those products.

0

5000

10000

15000

20000

2000 2002 2004 2006 2008 2010 2012 2014

Intra SADC

Export UV Import UV

70

Figure 3.24. Evolution of the terms of trade by group of products

Source: TUV Database and authors’ calculations, 2016.

Note: (a) for agricultural products, (b) for non-agricultural products.

Conclusion

In this chapter, many indicators were discussed to measure the intensity of intra-regional trade

from 1998 to 2013 within African and within four RECs, including ECOWAS, ECCAS, COMESA

and SADC, using mainly the BACI database. The analysis of the current performance of intra-

African and intra-RECs trade showed that the value of intra-African agricultural trade has grown

rapidly over recent years, rising from $2.2 billion in 1998 to $12.8 billion in 2013.

50

60

70

80

90

100

110

120

130

1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

(a)

Africa ECOWAS ECCAS COMESA SADC

40

60

80

100

120

140

160

180

200

220

1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

(b)

Africa ECOWAS ECCAS COMESA SADC

71

The overall annual growth over this period is around 12%. Regarding the RECs, intra-regional

agricultural trade has in general displayed significant increases over the period. Intra-ECOWAS

agricultural trade shows an average growth rate of 12%, rising from $494 million in 1998 to

$2.84 billion in 2013. However, agricultural trade between ECOWAS countries was very erratic.

Trade increases between them were partly driven by commodity prices. Agricultural trade between

ECCAS countries has shown the highest overall growth of 17%, with a nominal value which has

increased from $14 million in 1998 to $147 million in 2013. Intra-regional agricultural trade in

COMESA has displayed a significant increase (14%) over 1998-2013, rising from $379 million in

1998 to $2.87 billion in 2013. In COMESA, unlike the other RECs, the growth gap between the

two sub-periods is very low (less than 3 percentage points). The volume of intra-regional

agricultural trade has also shown a significant increase (22%). Lastly, in the SADC area, the lowest

overall growth of 10% is observed, with a nominal value which has increased from $871 million

in 1998 to $3.82 billion in 2013.

The regional trade integration measures results showed that ECOWAS is the REC with the highest

trade integration with a ratio of 0.79, followed by SADC with 0.77, COMESA with 0.65 and

ECCAS with 0.52. Except for ECCAS countries, all the RECs exchange more inside their own

bloc. In terms of intra-African agricultural trade, as destinations or origins of intra-African trade,

COMESA and SADC are the leading regions before ECOWAS and ECCAS. However, it is noted

that COMESA and SADC have opposite patterns. In fact, COMESA has gained trade share

(exports and imports) over the considered period while SADC countries have lost some. Moreover,

it is also observed on aggregate that all the RECs have intensified agricultural exchanges within

their group. Regarding the main agricultural products traded between African countries, between

the two periods, no major changes are noted in the composition of intra-African trade.

72

References

Berthou, A., & Emlinger, C. (2011). The Trade Unit Values Database. CEPII Working Paper 2011-10.

Paris: CEPII.

Engel, J., Jouanjean, M., & Awal, A. (2013). The History, Impact and Political Economy of Barriers to

Food Trade in Sub-Saharan Africa: An Analytical Review. Overseas Development Institute Report.

London: Overseas Development Institute.

IMF. (2009). Export and Import Price Index Manual: Theory and Practice. Washington, DC:

International Monetary Fund.

OECD. (2011). Mexican Export and Import Unit Value Indices. STD/TBS/WPTGS(2011)4. Paris:

Organisation for Economic Co-operation and Development.

UNCTAD. (2013). Economic Development in Africa Report 2013: Intra-African Trade: Unlocking

Private Sector Dynamism. Geneva: United Nations.

73

Annex

Table A1: Regional agricultural trade share (%)

Africa ECOWAS ECCAS

Import Export Intra regional Import Export Intra regional Import Export Intra regional

1998 17.4 20.7 28.7 17.7 30.6 23.2 20.6 10.9 26.1

1999 16.5 18.1 23.0 18.3 24.0 20.1 22.3 8.8 30.0

2000 15.7 13.0 23.1 17.1 16.1 28.0 22.8 5.5 25.3

2001 15.8 14.4 22.7 17.9 17.4 18.6 19.0 6.3 22.4

2002 16.3 15.3 21.9 18.3 22.0 20.5 21.5 5.2 25.9

2003 15.8 14.8 23.1 19.2 22.4 26.5 21.4 5.7 25.1

2004 14.4 12.4 19.4 17.3 17.7 21.6 19.2 4.2 23.7

2005 13.0 10.1 15.8 16.4 13.0 15.4 17.9 3.1 29.3

2006 12.5 8.7 14.7 15.7 10.2 12.9 15.9 2.2 21.7

2007 13.7 9.0 14.5 16.4 11.5 12.7 16.3 2.0 6.2

2008 13.9 7.8 13.7 14.8 10.3 10.2 16.4 1.5 9.2

2009 14.4 11.7 17.9 16.1 16.7 16.8 16.4 2.8 6.6

2010 14.4 10.0 16.3 14.8 12.8 15.6 17.4 2.2 5.9

2011 16.7 9.8 18.4 15.8 12.0 17.6 22.9 2.0 28.8

2012 16.0 8.9 15.3 17.9 10.4 15.5 18.5 1.6 3.7

2013 15.1 9.9 15.9 16.4 10.8 14.1 19.4 1.5 3.5

1998-2006 15.3 14.2 21.4 17.6 19.3 20.8 20.0 5.8 25.5

2007-2013 14.9 9.6 16.0 16.0 12.1 14.7 18.2 1.9 9.1

Overall 15.1 12.2 19.0 16.9 16.1 18.1 19.2 4.1 18.3

74

Table A1: Regional agricultural trade share (%), contd.

COMESA SADC

Import Export Intra regional Import Export Intra regional

1998 20.1 29.0 37.3 18.5 34.6 33.7

1999 19.2 26.0 37.9 17.4 32.5 28.7

2000 19.3 20.6 40.3 17.8 25.8 36.4

2001 19.4 21.5 46.9 17.7 28.3 38.6

2002 20.3 22.6 36.0 21.0 26.4 41.1

2003 19.1 20.5 34.2 19.7 23.0 27.9

2004 17.5 17.1 36.3 17.6 17.2 29.6

2005 15.0 14.6 27.0 16.1 13.1 22.5

2006 14.4 12.0 26.2 15.4 10.8 19.2

2007 15.9 12.7 35.7 16.2 9.7 32.6

2008 17.1 11.0 27.2 15.1 6.6 24.3

2009 17.6 15.2 31.8 16.0 10.6 26.4

2010 18.6 14.3 33.5 16.5 8.3 24.9

2011 23.1 17.4 39.2 17.9 8.5 29.4

2012 20.4 14.0 34.0 16.8 8.1 29.9

2013 18.0 17.4 32.0 16.0 8.1 22.5

1998-

2006 18.3 20.4 35.8 17.9 23.5 30.8

2007-

2013 18.7 14.6 33.3 16.4 8.5 27.2

Overall 18.4 17.9 34.7 17.2 17.0 29.2

Source: BACI Database and authors’ calculations, 2016.

75

Table A2: Evolution of export/import value index and terms trade for agricultural products

Africa ECOWAS ECCAS

Import Export ToT Import Export ToT Import Export ToT

1998 0.897 0.765 85.284 0.832 0.775 93.192 0.863 0.827 95.799

1999 0.893 0.756 84.668 0.827 0.769 92.964 0.871 0.815 93.628

2000 0.890 0.738 82.936 0.818 0.754 92.186 0.875 0.777 88.823

2001 0.886 0.744 83.941 0.811 0.760 93.814 0.876 0.781 89.247

2002 0.883 0.751 85.074 0.820 0.778 94.902 0.874 0.766 87.574

2003 0.883 0.765 86.626 0.842 0.780 92.686 0.903 0.816 90.375

2004 0.885 0.772 87.226 0.849 0.773 91.105 0.896 0.808 90.181

2005 0.893 0.795 89.023 0.871 0.801 91.894 0.902 0.846 93.816

2006 0.902 0.814 90.198 0.867 0.813 93.845 0.916 0.861 94.040

2007 0.924 0.850 91.969 0.872 0.834 95.617 0.924 0.911 98.612

2008 0.943 0.897 95.125 0.886 0.867 97.862 0.925 0.978 105.815

2009 0.949 0.916 96.541 0.873 0.935 107.157 0.945 0.956 101.151

2010 0.966 0.981 101.494 0.863 0.956 110.800 0.937 0.978 104.353

2011 0.951 0.986 103.611 0.874 0.940 107.581 0.951 0.998 104.895

2012 0.977 0.980 100.299 0.883 0.969 109.636 0.949 1.076 113.330

2013 0.968 0.993 102.556 0.895 0.941 105.177 0.938 1.160 123.673

Source: BACI Database and authors’ calculations, 2016.

76

Table A2: Evolution of export/import value index and terms trade for agricultural products,

contd.

COMESA SADC

Import Export ToT Import Export ToT

1998 1.032 0.781 75.712 0.851 0.699 82.123

1999 1.032 0.766 74.257 0.852 0.686 80.554

2000 1.031 0.737 71.518 0.856 0.666 77.860

2001 1.030 0.753 73.066 0.859 0.680 79.090

2002 1.013 0.753 74.333 0.853 0.689 80.786

2003 0.990 0.780 78.833 0.861 0.714 82.947

2004 0.997 0.804 80.714 0.879 0.723 82.333

2005 0.988 0.849 86.021 0.868 0.751 86.445

2006 1.011 0.870 86.078 0.882 0.758 85.941

2007 1.070 0.959 89.626 0.895 0.774 86.499

2008 1.088 1.059 97.361 0.895 0.804 89.873

2009 1.081 1.064 98.400 0.900 0.817 90.763

2010 1.106 1.096 99.041 0.908 0.819 90.196

2011 1.085 1.098 101.185 0.920 0.856 93.077

2012 1.152 1.090 94.576 0.918 0.878 95.676

2013 1.120 1.130 100.919 0.911 0.898 98.516

Table A3: Evolution of export/import value index and terms trade for non-agricultural products

Africa ECOWAS ECCAS

Import Export ToT Import Export ToT Import Export ToT

1998 0.613 0.910 148.588 0.560 0.945 168.581 0.700 1.062 151.773

1999 0.639 0.906 141.847 0.625 0.982 157.127 0.676 1.088 160.818

2000 0.635 0.918 144.516 0.602 0.995 165.402 0.656 1.112 169.697

2001 0.638 0.903 141.567 0.594 0.972 163.639 0.703 1.111 158.055

2002 0.637 0.896 140.817 0.584 0.983 168.323 0.664 1.131 170.237

2003 0.647 0.915 141.355 0.584 0.986 168.773 0.676 1.133 167.686

2004 0.653 0.942 144.391 0.597 1.012 169.361 0.744 1.206 162.007

2005 0.669 0.983 146.887 0.643 1.006 156.432 0.728 1.224 168.181

2006 0.684 1.028 150.369 0.641 1.046 163.226 0.751 1.254 166.933

2007 0.752 1.042 138.501 0.713 1.046 146.832 1.292 1.281 99.164

2008 0.751 1.126 149.938 0.687 1.072 156.079 1.059 1.448 136.791

2009 0.736 1.094 148.593 0.631 1.066 168.901 1.079 1.374 127.303

2010 0.731 1.115 152.678 0.653 1.076 164.885 1.062 1.449 136.464

2011 0.709 1.110 156.467 0.638 1.143 179.102 0.770 1.425 185.211

2012 0.814 1.134 139.406 0.691 1.135 164.160 1.794 1.371 76.419

2013 0.796 1.145 143.755 0.706 1.195 169.375 1.549 1.461 94.336

77

Table A3: Evolution of export/import value index and terms trade for non-agricultural products,

contd.

COMESA SADC

Import Export ToT Import Export ToT

1998 0.674 1.109 164.543 0.627 0.995 158.754

1999 0.654 1.118 171.089 0.635 1.013 159.631

2000 0.649 1.112 171.312 0.638 1.042 163.264

2001 0.653 1.117 171.166 0.649 1.046 161.180

2002 0.653 1.104 169.045 0.646 1.059 163.798

2003 0.667 1.133 169.874 0.651 1.046 160.730

2004 0.654 1.132 173.238 0.695 1.042 150.046

2005 0.685 1.199 175.016 0.707 1.098 155.375

2006 0.705 1.270 180.277 0.713 1.142 160.177

2007 0.713 1.290 180.977 0.774 1.166 150.578

2008 0.756 1.367 180.725 0.731 1.184 161.873

2009 0.744 1.421 191.058 0.717 1.231 171.758

2010 0.749 1.539 205.578 0.737 1.429 193.913

2011 0.774 1.601 206.945 0.744 1.414 190.026

2012 0.787 1.539 195.531 0.751 1.347 179.381

2013 0.780 1.559 199.974 0.808 1.426 176.512

Source: BACI Database and authors’ calculations, 2016.

Chapter 4. Competitiveness of African agricultural exports

Extracted from

African Agricultural Trade Status Report

2017

78

CHAPTER 4. COMPETITIVENESS OF AFRICAN AGRICULTURAL EXPORTS

Sunday Pierre Odjo, International Food Policy Research Institute, West and Central Africa

office, Dakar, Senegal

Ousmane Badiane, International Food Policy Research Institute, Washington DC

4.1 Introduction African agricultural trade performance has been improving over the last decade. There have been

substantial gains in export value concomitantly with an increase in Africa’s share of world exports.

However, agricultural imports by African countries have increased faster and the continent is still

below the world market share it secured three decades ago. Thus, accelerating current export trends

and diversifying African export commodities and destination markets appear as a crucial policy

objective in an attempt to reduce foreign trade deficits across countries and help stabilize intra-

African food markets. To that end, a starting point is to understand how current advances in African

exports have been brought about. Of particular interest is understanding how changes in domestic

production and trading conditions have enabled improvement or degradation in African export

competitiveness in global as well as intra-African markets. This would provide more insight on

national and regional strategies that could help exploit untapped export potential and invest in

emerging markets and new export commodities.

The present chapter investigates the patterns and determinants of changes in export

competitiveness among African countries and products over the last three decades. It is based on

the measurement of changes in competitiveness through constant market share decomposition

analysis and the comparisons of derived competitive effects in alternative export destination

markets and across countries and commodity groups. In the next section we present the analytical

methods and data used for the derivation of country and commodity competitiveness changes.

Section 4.3 discusses the country and commodity rankings on their competitiveness in global

markets. Competitiveness rankings in global markets and intra-African markets are compared in

Section 4.4, while Section 4.5 deals with corresponding rankings in the markets of the regional

economic communities (RECs), including the Common Market for Eastern and Southern Africa

(COMESA), the Economic Community of Central African States (ECCAS), the Economic

Community of West African States (ECOWAS), and the Southern African Development

Community (SADC). Section 4.6 proposes an econometric model of the determinants of country

79

competitiveness changes in alternative agricultural export markets. Section 4.7 summarizes main

findings and derives some recommendations for policy actions.

4.2. Export share growth decomposition model and data

4.2.1 The model

Competitiveness has widely been explored through the Constant Market Share (CMS)

decomposition model as a means of assessing how countries compare to their competitors with

respect to their trade performance between time periods. Since its first application to trade analysis

by Tyszynski (1951), the CMS methodology has been refined and expanded through alternative

model formulations attempting to enrich its analytical features (Leamer and Stern, 1970;

Richardson, 1971) or to deal with some issues arising with its applications (Cheptea, Gaulier and

Zignago, 2005). The formulation used in this chapter was developed by Magee (1975). It explains

the growth in a country or region’s share of world markets by decomposing it into two major

growth sources, namely structural changes in market distribution and product composition and

competitiveness changes. The market share growth model starts with the following identity:

𝑆𝑡1

𝑚 = 𝑅𝑚 ∙ 𝑆𝑡0

𝑚 (1)

where 𝑆𝑡0

𝑚 and 𝑆𝑡1

𝑚 denote the shares of a given country or region 𝑚 in total world exports in the

beginning and end periods 𝑡0 and 𝑡1, respectively. 𝑅𝑚 represents a relative growth factor defined

as follows:

𝑅𝑚 =1+𝑔𝑚

1+𝑔𝑤 (2)

where 𝑔𝑚 and 𝑔𝑤 stand for the compound annual growth rate (between the beginning and end

periods) of total exports of country or region 𝑚 and of the world 𝑤, respectively. Equation (2)

expresses the growth of country or region 𝑚′𝑠 exports relative to the world’s exports and can be

rewritten as

𝑅𝑚 = ∑ (1+𝑔𝑖

𝑚

1+𝑔𝑤) (𝑋𝑖 𝑡0

𝑚

𝑋𝑡0𝑚 ) 𝑖 (3)

where

𝑋𝑡0𝑚 = ∑ 𝑋𝑖 𝑡0

𝑚𝑖

Expressing 𝑋𝑡0𝑚 for the different export products 𝑖 and destinations 𝑗 in (3), multiplying by

[(1 + 𝑔𝑖𝑤)𝑋𝑖 𝑡0

𝑚 (1 + 𝑔𝑖𝑤)𝑋𝑖 𝑡0

𝑚⁄ ] and by[(1 + 𝑔𝑖𝑤𝑗

) (1 + 𝑔𝑖𝑤𝑗

)⁄ ], and summing over 𝑖 and 𝑗 yields,

after rearranging and substituting the new expression for (3) in (1):

80

𝑆𝑡1

𝑚 = 𝑆𝑡0

𝑚 ∑(1+𝑔𝑖

𝑤) 𝑋𝑖 𝑡0𝑚

(1+𝑔𝑖𝑤) 𝑋𝑡0

𝑚𝑖 ∑(1+𝑔𝑖

𝑚𝑗)(1+𝑔𝑖

𝑤𝑗)𝑋𝑖 𝑡0

𝑚𝑗

(1+𝑔𝑖𝑤𝑗

)(1+𝑔𝑖𝑤)𝑋𝑖 𝑡0

𝑚𝑗 (4)

with

𝑋𝑡0

𝑚 = ∑ 𝑋𝑖 𝑡0

𝑚𝑖

𝑋𝑖 𝑡0

𝑚 = ∑ 𝑋𝑖 𝑡0

𝑚𝑗𝑗

where 𝑖 and 𝑗 are indices for export products and destinations, respectively.

Our objective in this chapter is to rank African countries and agricultural commodities on changes

in their competitiveness in different export destination markets, including global markets (as one

market entity), intra-African markets (as one market entity) and the regional markets of COMESA,

ECCAS, ECOWAS and SADC (taking each REC as one market entity). Therefore, the model is

applied in three different settings corresponding to different levels of exporters and products

aggregations as follows. In the first setting, 𝑚 represents Africa as a whole and the model

decomposes the growth in Africa’s share of world exports of each of 59 agricultural commodity

groups 𝑖. The second setting is a variant of the first where 𝑚 stands for each REC as an aggregate

exporter instead of Africa as a whole. Thus, the model explains the growth in the REC’s share of

world exports of each of 59 agricultural commodity groups. In the third setting, 𝑚 denotes each of

51 African countries and 𝑖 is an aggregate agricultural good. The model decomposes the growth

in a country’s share of world aggregate agricultural exports. In all three settings, calculations are

carried out for 𝑗 representing alternatively global markets, intra-African markets and the regional

markets of COMESA, ECCAS, ECOWAS and SADC. With exporters and products aggregated as

defined in the three settings, Eq. (4) simplifies to

In the case where 𝑗 represents global markets, Eq. (4) further simplifies to

𝑆𝑡1

𝑚 = 𝑆𝑡0

𝑚(1+𝑔𝑖

𝑚𝑗)

(1+𝑔𝑖𝑤𝑗

) (6)

𝑆𝑡1

𝑚 = 𝑆𝑡0

𝑚 ∑(1+𝑔𝑖

𝑚𝑗)

(1+𝑔𝑖𝑤𝑗

)𝑗

(1+𝑔𝑖𝑤𝑗

)

(1+𝑔𝑖𝑤)

𝑋𝑖 𝑡0

𝑚𝑗

𝑋𝑖 𝑡0𝑚 (5)

(𝑎) (𝑏) (𝑐)

81

From Eq. (1) it is clear that whether a country or region’s share in world exports increases or

diminishes during the considered time period depends upon whether the growth factor is greater

or less than unity. Given the reduced expression for 𝑅 in Eq. (5), the contribution of a

destination 𝑗 to the performance of a given country or region (in terms of the change in its export

share) can be decomposed into two components: a competitive effect and a market effect.

The competitive effect corresponds to the first expression (a) of the right hand side of Eq. (5). It

is a measure of the change in competitiveness experienced by country or region 𝑚 in exporting a

good 𝑖 to destination 𝑗. If it is greater (smaller) than 1.0 the competitive effect translates some gain

(loss) of competitiveness by the country or region compared to the group of its competitors in the

export destination considered.

The market effect corresponds to the product of the terms (b) and (c) in Eq. (5). It measures the

portion of the country or region’s export share growth which is due to faster or slower growth of

world exports of good 𝑖 to destination markets 𝑗 as compared to global markets. It reflects the

change in the importance of 𝑗 as a destination for the country’s exports attributable to the expansion

of markets 𝑗. For instance, in the case where 𝑗 denotes the regional markets of a REC, the market

effect translates the change in the importance of the community markets as a destination for its

members’ exports which is associated with the expansion of the regional markets. For an easier

interpretation, the market effect 𝑀𝑅𝐾 can be derived in value terms from the simplified expression

in Eq. (5) as follows:

𝑀𝑅𝐾 = [(1+𝑔𝑖

𝑤𝑗)

(1+𝑔𝑖𝑤)

𝑋𝑖 𝑡0

𝑚𝑗

𝑋𝑖 𝑡0𝑚 −

𝑋𝑖 𝑡0

𝑚𝑗

𝑋𝑖 𝑡0𝑚 ] 𝑋𝑖 𝑡1

𝑚 (7)

where 𝑋𝑖 𝑡1

𝑚 stands for the considered country or region’s total exports of good 𝑖 to world markets

in the end period. The value of 𝑀𝑅𝐾 measures the magnitude of the positive or negative impact

of the expansion of markets 𝑗 on the considered country or region’s export performance. As it

appears in Eq. (6), it is clear that no market effect can be derived in the case where global markets

are the destination under consideration.

4.2.2 Data and product and country coverage

The model is applied using data on the values of bilateral exports of agricultural products at the

HS4 aggregation level over the period 1998-2013.

82

The data are obtained from the BACI database for individual African countries, except for the

Southern African Customs Union (SACU) members, namely Botswana, Lesotho, Namibia, South

Africa and Swaziland, for which trade data are aggregated as SACU countries in the BACI

database.

For this analysis, bilateral export values are first aggregated so as to construct the variables of each

country’s total exports to world markets, to intra-African markets and to each REC’s regional

markets. These are then aggregated to construct the variables of Africa’s and each REC’s aggregate

exports to the different export destination markets under consideration. In addition, bilateral export

values are aggregated from the BACI database to construct the variables of the world’s total

exports of the different agricultural products to the different export destinations under analysis. In

order to reduce the number of HS4 product lines, the different variables are aggregated from HS4

to HS2 level, except for a few HS4 lines of interest which are kept as such.

The final dataset used for the CMS model comprises 59 commodity groups (hereafter also

designated as commodities or products) and 51 individual countries, including one SACU

countries aggregate. It includes all 11 ECCAS members and all 15 ECOWAS members. SADC

enters the dataset with 10 individual member countries while its other 5 members are aggregated

as one case (SACU countries). With Swaziland among the aggregated countries, COMESA is left

with 18 of its 19 members. The dataset also includes some countries that are not members of any

REC, including Algeria, Mauritania, Morocco, Tunisia, Saint Helena, Somalia, Western Sahara

and Tunisia.

In the present chapter only competitive effect values are reported and analyzed. Furthermore, the

results relative to the application of the model under the second setting – where the model

decomposes the export share growth for each REC as an aggregate exporter – are not presented in

this chapter. Thus, in the following development, the results that refer to the change in a REC’s

competitiveness reflect averages over the changes in competitiveness of its member countries.

Such averages reveal more meaningful differences between RECs than the results obtained from

modeling the RECs as aggregate exporting entities.

83

4.3. Competitiveness in global markets: country and commodity rankings

The values of the competitive effect derived from the share growth decomposition analysis for

individual African countries are presented in Table A4.1. They reflect the changes in

competitiveness experienced by African countries compared to their competitors as a group in

alternative agricultural export destination markets over the period 1998-2013. In Figure 4.1 the

values of competitive effect in global markets are sorted from lowest to highest, showing

corresponding countries from the least competitive to the most competitive.

As it appears on the figure, the coefficients of the competitive effect are smaller than 1.0 for 32

out of 51 countries under analysis, which means that those countries have underperformed the

group of their competitors in global markets. The least competitive among them include three

ECCAS members, namely Equatorial Guinea, Angola and Chad, for which estimates of the

competitive effect are not greater than 0.9. Between the 0.9 and 1.0 thresholds are the values of

the competitive effect estimated for all other ECCAS members, with only the exception of

Rwanda. Apart from Angola, almost two thirds of the other SADC members have revealed a

competitive effect within the 0.9 to 1.0 interval, the three exceptions being Tanzania, Mozambique

and Zambia. As many ECCAS and SADC members are also COMESA members, up to two thirds

of COMESA members are found among the countries that have underperformed the group of their

competitors. As for ECOWAS, half of its members are also found among underperforming

countries.

Figure 4.1 Change in country competitiveness in global agricultural export markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Equato

rial G

uin

ea

West

ern

Sahara

Angola

Chad

Sao T

om

e &

Princi

pe

Centr

al A

fric

an R

ep.

Zim

babw

eG

abon

Mali

D.R

. C

ongo

Madagasc

ar

Eritr

ea

Benin

Lib

yaG

uin

ea

Sudan

Mauritiu

sS

enegal

Congo

Côte

d'Iv

oire

Buru

ndi

Seyc

helle

sM

ala

wi

Com

oro

sC

am

ero

on

SA

CU

countr

ies

Gam

bia

Kenya

Maurita

nia

Sain

t H

ele

na

Togo

Moro

cco

Uganda

Tanza

nia

Nig

er

Tunis

iaM

oza

mbiq

ue

Burk

ina

Faso

Guin

ea B

issa

uS

ierr

a L

eone

Lib

eria

Zam

bia

Ghana

Rw

anda

Eth

iopia

Nig

eria

Egyp

tD

jibouti

Alg

eria

Som

alia

Cape V

erd

e

Change in

com

petit

iveness

84

However, for nineteen out of the 51 countries considered, the coefficients of the competitive effect

are greater than 1.0. These countries have succeeded in raising their levels of competitiveness by

expanding their exports to global markets faster than their competitors. The strongest increases in

competitiveness have been achieved by Cape Verde, Somalia, Algeria and Djibouti where

estimated values of the competitive effect are greater than 1.1. The other 15 countries have more

modestly outperformed their competitors, with competitive effect values between the 1.0 and 1.1

thresholds.

They include the other half of ECOWAS members, namely Niger, Burkina Faso, Guinea Bissau,

Sierra Leone, Liberia, Ghana and Nigeria. We can also see Tunisia among the outperforming

countries, as well as Tanzania, Mozambique and Zambia within SADC, and Uganda, Rwanda,

Ethiopia and Egypt within COMESA.

Changes in country competitiveness are plotted in Figure 4.2 against country shares in Africa’s

global agricultural exports as presented in Table A4.2. The figure shows that the most notable

changes in competitiveness have occurred among countries that contribute very small shares of

African global exports. Conversely, countries with higher export shares have not experienced a

remarkable change in competitiveness. Thus, Africa’s export performance has been improving

mostly among small exporters like Cape Verde, Somalia, Algeria and Djibouti while stagnating

among larger exporters like Côte d’Ivoire, Morocco and Kenya. It is worth noticing the

performance of Egypt and Ghana. Each represents at least 5% of Africa’s global agricultural

exports and has achieved an index of competitiveness change greater than 1.1.

85

Figure 4.2. Scatterplot of changes in country competitiveness against country shares in Africa’s

agricultural exports to global markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

In sum, ECCAS appears to be lagging behind in the fight to gain more competitiveness in global

agricultural export markets, but the proportions of underperforming countries within COMESA,

SADC and ECOWAS are also a concern. In order to get a clearer insight into the difference

between regional country groupings, average sizes of the competitive effect are plotted in Figure

4.3 and standard deviation values are shown on top of the bars. Within-group variations in

competitive effect values seem to be homogenous across groups, which justifies average effect

size comparisons. SADC and more notably ECCAS members appear to have on average lost

competitiveness, with ECCAS showing a bigger loss. In contrast, ECOWAS members have on

average raised competitiveness, while there has been no or little change for COMESA members

on average.

Equatorial Guinea

Western SaharaAngola

Zimbabwe

Côte d'Ivoire

SACU countriesKenya MoroccoTanzania

GhanaNigeria EgyptDjibouti

AlgeriaSomalia

Cape Verde

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0

Chan

ge

in c

ountr

y c

om

pet

itiv

enes

s

Country share in Africa's agricultural exports to the global markets (%)

86

Figure 4.3. Country-group average competitiveness change in global agricultural export markets

(1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries. Standard deviation

values are shown on top of the bars.

Table 4.1. Analysis of variance of country competitiveness changes in global agricultural export

markets (1998-2003)

Test Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.001 1 0.001 0.142 0.708 0.003

non-COMESA Within Groups 0.286 49 0.006

countries Total 0.287 50

ECCAS vs. Between Groups 0.06 1 0.060 12.919 0.001 0.209

non-ECCAS Within Groups 0.227 49 0.005

countries Total 0.287 50

ECOWAS vs. Between Groups 0.018 1 0.018 3.282 0.076 0.063

non-ECOWAS Within Groups 0.269 49 0.005

countries Total 0.287 50

SADC vs. Between Groups 0.006 1 0.006 1.009 0.32 0.02

non-SADC Within Groups 0.281 49 0.006 countries Total 0.287 50

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

An analysis of variance is carried out to statistically test the difference between each regional

country grouping and the rest of Africa as summarized in Table 4.1. The results confirm that

competitive effect sizes are on average significantly lower for ECCAS and higher for ECOWAS

compared to the rest of African countries. However, between-group variations account for very

little in the overall variations among countries. This means that the larger part of the variations in

0.057

0.078

0.069

0.051

0.076

0.92

0.94

0.96

0.98

1.00

1.02

1.04

COMESA ECCAS ECOWAS SADC AfricaAve

rag

e c

om

pe

titive

ne

ss

ch

an

ge

Regional Country groups

87

competitiveness change between countries is not related to regional factors but to domestic factors

like changes in total factor productivity and the competitiveness of most exported commodities by

individual countries. Indeed, as postulated by Hausman et al. (2005), what countries export matters

for their overall competitiveness.

Table A4.3 presents the values of the competitive effect calculated for agricultural commodities

through the decomposition of Africa’s commodity-specific export share growth in alternative

export destination markets between 1998 and 2013. They capture the magnitudes of changes in

competitiveness that Africa has achieved compared to the group of non-African competitors in the

different export destination markets over the period 1998-2013. In Figure 4.4 commodities are

sorted in increasing order of the changes in competitiveness as experienced in global markets. In

addition to the threshold of 1.0 that demarcates commodities in which Africa has lost some

competitiveness from those in which Africa has gained some, we will also consider the thresholds

of 0.95, 1.05 and 1.10 to help differentiate between lower and higher losses or gains.

Figure 4.4. Changes in commodity competitiveness in global agricultural export markets (1998-

2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

Gro

und

nut o

ilM

eat &

edib

le o

ffal

Org

anic

chem

icals

Poultry

Cotton, not card

ed o

r com

bed

Coffee

Cane s

ugar

Spic

es

Palm

oil

Fis

h &

sea foods

Hid

es &

skin

sO

ther

cere

als

Edib

le p

reps. of m

eat, fis

h &

cru

sta

ceans

Tea

Pre

ps. of vegs., fru

its &

nuts

Gum

s &

resin

sC

ocoa b

eans

Oth

er

anim

al pro

ducts

Gro

undnuts

Cotton, card

ed o

r com

bed

Edib

le fru

its &

nuts

Essential oils

& r

esin

oid

sS

ugar

confe

ctionery

Oliv

e o

ilO

ther

oils

eed

sO

ther

vegeta

ble

textile

fib

res

Mis

c. edib

le p

repara

tions

Ric

eF

urs

kin

sB

evera

ges, spirits &

vin

egar

Mill

ing industr

y p

roducts

Vegeta

ble

pla

itin

g m

ate

rials

Fin

ishin

g a

gents

for

textile

s &

paper

Sorg

hum

Maiz

eP

ota

toes

Tobacco &

substitu

tes

Tom

ato

es

Alb

um

inoid

al substa

nces

Resid

ues fro

m food industr

ies

Cocoa p

repara

tions

Medic

inal pla

nts

Wheat

Onio

ns &

substitu

tes

Oth

er

live tre

es &

pla

nts

Oth

er

edib

le v

egeta

ble

sO

ther

live a

nim

als

Oth

er

oils

& facts

Soybeans

Wool

Pre

ps. of cere

als

, flour,

sta

rch o

r m

ilkS

heep &

goats

Anim

al fa

tsR

oots

& tubers

Dairy, eggs &

honey

Silk

Cattle

Soybean o

ilR

ye, barley &

oats

Ch

an

ge i

n c

om

peti

tiven

ess

88

African exporters have lost competitiveness in global markets in the exports of 15 out of 59

commodities. Important food staples affected include groundnut oil, meat & edible offal, poultry,

palm oil, fish & sea foods, and some cereals11. However, the size of competitiveness loss is modest

as the corresponding estimates of the competitive effect are contained within the 0.95 to 1.0

interval.

For the majority of the commodities under analysis, Africa has experienced an increased

competitiveness in global markets by expanding its exports of these commodities faster than the

group of non-African competitors has done. Up to 44 out of 59 commodities considered show a

competitive effect value higher than 1.0. The strongest increase in competitiveness is acquired for

the following five commodity groups, for which competitive effect values are greater than 1.10,

including rye, barley & oats; soybean oil; cattle; silk; and dairy, eggs & honey. Many food staples

are found among the commodities for which competitiveness gains are higher than 1.05 though

smaller than 1.1, including roots & tubers, sheep & goats, other live animals, onions & substitutes,

and wheat. But a number of other staples are among commodities for which Africa has more

modestly outperformed the group of its competitors, including tomatoes, potatoes, maize,

sorghum, and rice, which show competitive effect values in the 1.0 to 1.05 interval.

African exporters have either lost competitiveness or modestly increased competitiveness for

major African traditional cash crops like coffee, cocoa beans, tea, cotton, groundnut oil, palm oil,

sugar cane, groundnuts and other oilseeds. In contrast, they have been able to improve their

competitiveness for new export commodities like wool, soybeans, soybean oil, live trees & plants,

and cocoa preparations. Figure 4.5 below helps assess the importance of the top ranked

commodities in terms of their share in the value of Africa’s total agricultural exports to global

markets compared to intra-African markets. For instance, it shows that the top 15 commodities

account for only 10% of Africa’s global agricultural exports and the top 40 commodities in the

ranking hardly reach the 50% share threshold. Conversely, the bottom 19 commodities in the

ranking represent up to 51.5% of African agricultural exports. This confirms our guess that

competitiveness gains in global markets are not occurring only for traditional African export

11 Within the commodity group comprising buckwheat, millet and canary seed.

89

commodities but also for emerging export products. It is indicative of the scope for further

expanding Africa’s global exports by exploiting increased commodity competitiveness.

Figure 4.5. Relative importance of the most competitive commodities in global and intra-African

markets

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

The same conclusions are demonstrated in Figure 4.6, which shows a scatter plot of changes in

commodity competitiveness against commodity shares in Africa’s global agricultural exports

(presented in Table A4.4). The figure indicates that changes in competitiveness generally have

been achieved for commodities that account for small shares of Africa’s global agricultural

exports. Conversely, little or no competitiveness change has been obtained in commodities that

represent higher export shares. Thus, African exporters have been improving their performance

mostly in minor export products like rye, barley & oats, soybean oil, and cattle, while their

performance has been stagnating in major export products like edible fruits & nuts, cocoa beans,

fish & sea foods, coffee, cotton, and cane sugar.

0

10

20

30

40

50

60

70

80

90

100

5 10 15 20 25 30 35 40 45 50 55 59

Cum

ulat

ive

aver

age

shar

e of

Afr

ican

agr

icul

tral

expo

rts to

the

diffe

rent

mar

kets

(%)

Number of top competitive commodities inthe different agricultural export markets

Global markets Intra-African markets

90

Figure 4.6. Scatterplot of changes in commodity competitiveness against commodity shares in

Africa’s agricultural exports to global markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

So far we have investigated how competitiveness has changed for countries and commodities in

global markets. We now turn to exploring changes in competitiveness in intra-African markets.

We will see how country and commodity rankings on competitiveness change in intra-African

markets compared to the above-described rankings related to global markets.

4.4. Competitiveness in intra-African markets: country and commodity rankings

Changes in competitiveness experienced by individual African countries in global and intra-

African agricultural markets are shown in Figure 4.7 below. They are measured by the coefficients

of the competitive effect derived through country-level share growth decomposition and

summarized in Table A4.1. In the figure, countries are sorted in increasing order of the changes in

competitiveness in intra-African markets. As it appears, competitive effect values are smaller than

1.0 for only 20 countries in this ranking compared to 32 countries in the ranking relative to global

markets (cf. Figure 4.1 above). This means that a smaller share of African countries have

underperformed the group of their competitors in intra-African markets compared to global

markets. Of those twenty, Saint Helena, Mali, Central Africa Republic and Chad have strongly

underperformed, with competitive effect values smaller than 0.9.

Groundnut oil

Meat and edible offal

Poultry Cotton, not carded or

combed.Coffee

Cane sugarSpices

Fish & sea foods

Edible preps. of meat, fish & crustaceansCocoa beans

Edible fruits and nutsTobacco and substitutesCocoa preps.

Other edible vegetables

Sheep & goats

Dairy, eggs and honeyCattle

Soybean oil

Rye, barley and oats

0.9

1.0

1.0

1.1

1.1

1.2

1.2

1.3

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

Chan

ge

in c

om

mo

dit

y c

om

pet

itiv

enes

s

Commodity share in Africa's Agricultural exports to global markets (%)

91

At the top edge of the ranking, twelve countries have strongly outperformed, with estimates of the

competitive effect greater than 1.1, among which the topmost 5 countries are Djibouti, Comoros,

Egypt and Algeria. It is worth recalling that only 4 countries have reached that level of increased

competitiveness in global markets. More interestingly, Figure 4.7 reveals that almost all

outperforming countries have in fact performed better in intra-African markets than in global

markets. And conversely, almost all underperforming countries have lost competitiveness more in

intra-African markets than in global markets.

Figure 4.7 Change in country competitiveness in intra-African agricultural export markets

compared to global markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

Table 4.2. Paired-sample T Tests for equality of country competitiveness changes in pairs of

African agricultural export destination markets

Paired Markets Paired Samples Correlation Mean paired

Differences t df

Sig.

(2-tailed) N Correlation Sig.

COMESA & global markets 48 0.417 0.003 0.002 0.086 47 0.932

ECCAS & global markets 46 0.631 0.000 -0.030 -2.183 45 0.034

ECOWAS & global markets 50 0.239 0.095 -0.009 -0.514 49 0.610

SADC & global markets 50 0.114 0.431 -0.025 -1.387 49 0.172

Intra-African & global markets 50 0.398 0.004 0.033 2.144 49 0.037

COMESA & intra-African markets 48 0.721 0.000 -0.024 -1.690 47 0.098

ECCAS & intra-African markets 46 0.479 0.001 -0.069 -4.069 45 0.000

ECOWAS & intra-African markets 50 0.487 0.000 -0.042 -2.532 49 0.015

SADC & intra-African markets 50 0.574 0.000 -0.058 -3.904 49 0.000

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Sa

int H

ele

na

Ma

liC

en

tra

l Afr

ica

n R

ep

.C

had

Sa

o T

om

e &

Prin

cipe

So

ma

liaB

enin

Zim

ba

bw

eN

ige

rM

ad

ag

asc

ar

To

go

Sie

rra

Le

on

eC

am

ero

on

Cô

te d

'Ivo

ire

Lib

yaLib

eria

SA

CU

co

un

trie

sK

en

yaG

ab

on

Bu

rkin

a F

aso

Ma

law

iS

ud

an

Gu

ine

aU

ga

nd

aG

am

bia

Ma

uritiu

sA

ng

ola

Tanza

nia

Se

ych

elle

sM

oza

mb

iqu

eM

au

rita

nia

Co

ng

oS

en

eg

al

Za

mb

iaD

.R. C

on

go

Eq

uato

rial G

uin

ea

Nig

eria

Ca

pe

Ve

rde

Mo

rocc

oG

uin

ea

Bis

sau

Gh

ana

Eritr

ea

Rw

an

da

Tu

nis

iaB

uru

nd

iE

thio

pia

Alg

eria

Eg

ypt

Co

mo

ros

Djib

ou

ti

Ch

na

ge

in c

om

pe

titiv

en

ess

Intra-African markets Global markets

92

Table 4.2 presents the results of paired-sample T tests of no difference between the competitive

effect values in global versus regional and intra-African markets. It can be read from the last row

of first panel of the table that changes in competitiveness in intra-African and global markets are

weakly and positively correlated. In other words, competitiveness changes are overall higher in

intra-African markets compared to global markets, but not consistently for all countries in the

sample. It also appears that there is a significant difference in the magnitude of competitiveness

changes between intra-African and global markets. On average competitiveness changes are higher

by 0.033 point in intra-African markets than in global markets.

It is of interest to see how the member countries of the different RECs have performed on average

in intra-African markets. Figure 4.8 reveals that COMESA members have generally achieved

higher gains in competitiveness than the rest of African countries in intra-African markets. Indeed,

we can see in Figure 4.7 that seven COMESA members have made it to the top 10 of the ranking,

namely Djibouti, Comoros, Egypt, Ethiopia, Burundi, Rwanda and Eritrea, and only Kenya is

found among the bottom 20 positions in the ranking. An analysis of variance of competitive effect

values in intra-African markets, summarized in Table 4.3, confirms that COMESA members have

on average performed significantly better than the rest of African countries. In contrast, there is no

perceptibly significant difference between ECCAS, ECOWAS and SADC members in terms of

changes in their competitiveness in intra-African markets. Part of the explanation may be found in

exploring differences in competitiveness gains achieved for particular export commodity groups.

Figure 4.8. Country-group average competitiveness change in intra-African agricultural export

markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries. Standard deviation

values are shown on top of the bars.

0.109

0.118 0.095 0.046

0.116

0.95

0.98

1.01

1.04

1.07

1.10

COMESA ECCAS ECOWAS SADC AfricaAve

rag

e c

om

pe

titive

ne

ss

ch

an

ge

Regional Country Groups

93

Table 4.3. Analysis of variance in country competitiveness changes in intra-African agricultural

export markets (1998-2013)

Test Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.075 1 0.075 6.196 0.016 0.114

non-COMESA Within Groups 0.579 48 0.012

countries Total 0.654 49

ECCAS vs. Between Groups 0.005 1 0.005 0.379 0.541 0.008

non-ECCAS Within Groups 0.649 48 0.014

countries Total 0.654 49

ECOWAS vs. Between Groups 0.011 1 0.011 0.806 0.374 0.017

non-ECOWAS Within Groups 0.643 48 0.013

countries Total 0.654 49

SADC vs. Between Groups 0.006 1 0.006 0.424 0.518 0.009

non-SADC Within Groups 0.648 48 0.014

countries Total 0.654 49 Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

Figure 4.9 below is constructed from Table A4.3 and represents the changes in competitiveness

that African countries have experienced in intra-African and global markets for individual

agricultural commodity groups under analysis. Commodities are sorted in increasing order of

changes in competiveness in intra-African markets.

For 29 out of 59 commodities under analysis, Africa has underperformed the group of its

competitors in intra-African markets. The corresponding number in the preceding ranking relative

to global markets is 15 out of 59 commodities. Furthermore, from Figure 4.9, it looks like Africa’s

performance in terms of commodity competitiveness gains is generally lower in intra-African

markets than in global markets, as it appears for the majority of commodities. The statistical

significance of these comparisons is verified in Table 4.4, which shows the results of a test for

equality of changes in commodity competitiveness in global markets compared to intra-African

and regional markets. The last row of the table shows that competitiveness changes in intra-African

and global markets are positively but weakly correlated. That is, changes in competitiveness tend

to be greater in global markets compared to intra-African markets, but not consistently across all

commodities. At the 10% significance level, competitiveness changes are indeed lower in intra-

African than in global markets. However, the difference is as small as 0.014 point on average.

94

Figure 4.9. Change in commodity competitiveness in intra-African agricultural export markets

compared to global markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

Many staple food products are among commodities for which Africa has underperformed,

including onions & substitutes, sheep & goats, meat & edible offal, poultry, sorghum, maize,

wheat, and other cereals.

We have seen above that Africa has strongly or weakly outperformed the group of its competitors

in global markets in exporting some of those staples, namely onions & substitutes, sheep & goats,

wheat, maize, and sorghum. Similarly to its competitiveness in global markets, Africa has

experienced positive changes in its competitiveness in intra-African markets for a number of other

important foodstuffs, including roots & tubers, cattle, other live animals, dairy, eggs & honey, rice,

potatoes, tomatoes, and fish & sea foods. In contrast and as in global markets, Africa has lost some

competitiveness in intra-African markets for its traditional cash crops like coffee, cocoa beans, tea,

cotton, groundnut oil, palm oil, groundnuts and other oilseeds.

Among the topmost ranked commodities we can see the same products that dominate the global

markets-related ranking, including rye, barley & oats (keeping the highest position), and soybean

oil. It also appears that African exporters have done better in intra-African markets than in global

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

Org

an

ic c

he

mic

als

So

yb

ea

ns

Gro

un

dn

ut o

ilS

ilkC

oco

a b

ea

ns

Fin

ish

ing

age

nts

fo

r te

xtile

s &

pap

er

On

ion

s &

su

bstitu

tes

Sh

eep

& g

oa

tsM

ea

t &

ed

ible

offa

lC

otto

n, no

t ca

rde

d o

r co

mb

ed

Co

ffe

eS

org

hu

mO

the

r o

ilse

ed

sE

ssen

tial o

ils &

re

sin

oid

sS

ug

ar

co

nfe

ctio

nery

Oth

er

cere

als

Co

co

a p

rep

ara

tio

ns

Ma

ize

Co

tto

n, ca

rde

d o

r co

mb

ed

Pa

lm o

ilB

eve

rag

es, sp

irits &

vin

eg

ar

Fu

rskin

sW

he

at

Oth

er

live

tre

es &

pla

nts

Po

ultry

Ed

ible

fru

its &

nu

tsG

rou

nd

nu

tsT

ea

Alb

um

ino

ida

l su

bsta

nce

sC

an

e s

uga

rM

isc. ed

ible

pre

pa

ratio

ns

Pre

ps. of ce

rea

ls, flou

r, s

tarc

h o

r m

ilkO

the

r a

nim

al p

rod

ucts

Me

dic

ina

l p

lan

tsF

ish &

se

a fo

od

sE

dib

le p

rep

s. o

f m

ea

t, fis

h &

cru

sta

ce

ans

To

ma

toe

sS

pic

es

Pre

ps. of ve

gs., fru

its &

nu

tsO

the

r liv

e a

nim

als

Po

tato

es

Ric

eR

esid

ue

s fro

m fo

od

in

dustr

ies

Da

iry, e

gg

s &

hon

ey

Mill

ing

in

du

str

y p

rod

ucts

To

ba

cco &

su

bstitu

tes

Wo

ol

Ca

ttle

An

ima

l fa

tsV

eg

eta

ble

pla

itin

g m

ate

ria

lsO

the

r o

ils &

fa

cts

Hid

es &

skin

sR

oo

ts &

tu

bers

Oth

er

ed

ible

ve

ge

tab

les

Oth

er

veg

eta

ble

te

xtile

fib

res

So

yb

ea

n o

ilG

um

s &

re

sin

sO

live

oil

Rye, b

arle

y &

oats

Ch

an

ge

in

co

mp

etitive

ne

ss

Intra-African markets Global markets

95

markets in exporting emerging export products like olive oil, soybean oil, gums & resins, other

(than cotton) vegetable textile fibers, hides & skins, and spices. Figure 4.5 above shows that the

top 15 commodities account for only 24.5% of intra-African agricultural exports and the top 25

commodities do not reach the 50% share threshold. However, the contributions of the same

numbers of the most competitive commodities in global markets to Africa’s global agricultural

exports are much smaller, as we have shown earlier with Figure 4.5. That is, more commodities

with relatively higher export value have gained increased competitiveness in intra-African markets

compared to global markets. This is in line with the faster growth of intra-African agricultural

trade in value terms over the period of this analysis.

Table 4.4. Paired-sample T Test for equality of commodity competitiveness changes in pairs of

African agricultural export destination markets

Paired markets

Paired Samples

Correlation Mean Paired

Differences t df

Sig.

(2-tailed) N Correlation Sig.

COMESA & global markets 59 0.475 0.000 -0.003 -0.306 58 0.761

ECCAS & global markets 59 0.430 0.001 -0.037 -4.238 58 0.000

ECOWAS & global markets 59 0.087 0.513 -0.020 -1.706 58 0.093

SADC & global markets 59 0.331 0.010 -0.015 -1.529 58 0.132

Intra-African & global markets 59 0.444 0.000 -0.014 -1.709 58 0.093

COMESA & intra-African markets 59 0.635 0.000 0.012 1.555 58 0.125

ECCAS & intra-African markets 59 0.377 0.003 -0.022 -2.246 58 0.029

ECOWAS & intra-African markets 59 0.294 0.024 -0.005 -0.484 58 0.630

SADC & intra-African markets 59 0.637 0.000 -0.001 -0.129 58 0.898

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

4.5. Competitiveness in regional markets: country and commodity rankings

In the preceding sections we have assessed and compared changes in country and commodity

competitiveness in global and intra-African agricultural export markets. We are now interested in

exploring the scope of Africa’s competitiveness gains or losses in each of four regional markets,

including COMESA, ECCAS, ECOWAS and SADC markets. To that end, four graphs analogous

to Figures 4.1 and 4.7 are constructed and pulled together in Figure A4.1. Each graph depicts the

ranking of African countries in increasing order of changes in their competitiveness in the

agricultural markets of a REC. They also help to see how competiveness changes in regional

markets compare to changes in global and intra-African markets.

96

Similarly, four other graphs equivalent to Figures 4.4 and 4.9 are assembled in Figure A4.2 and

show commodity rankings with respect to competitiveness changes in regional markets.

It can be seen from Figure A4.1 that 10 countries have underperformed in all four regional markets,

including Cameroon, Central African Republic, Kenya, Madagascar, Mali, Niger, Sao Tome &

Principe, Togo, Zimbabwe, and SACU countries as a group. Similarly, 9 other countries are found

that have outperformed in all regional markets, including Algeria, Egypt, Ethiopia, Malawi,

Mauritania, Morocco, Nigeria, Rwanda and Senegal. As a general pattern, country competitiveness

changes in regional markets tend to be lower than their performance in the broader intra-African

and global markets, in particular among the bottommost ranked countries.

The results from the test for equality presented in Table 4.2 above reveal that competitiveness

changes are significantly lower in ECCAS markets than in global markets by 0.03 point on

average. There are no significant differences between the other regional markets and global

markets as regards changes in country competitiveness. However, the test indicates that country

competitiveness changes are significantly lower in all regional markets than in the broader intra-

African markets, with differences ranging from 0.024 to 0.069 point on average.

Some of the findings conveyed by Figure A4.1 are summarized in Table 4.5 below. The table

presents two panels. The bottom row of the upper panel reveals that more than half of African

countries – 26-28 countries – have underperformed their competitors in ECCAS, ECOWAS and

SADC markets, with a revealed competitive effect value smaller than 1.0. Relatively fewer of them

– 19 countries – have similarly underperformed in COMESA markets. Indeed, COMESA markets

appear in the lower panel to be where at least half of African countries have outperformed their

competitors, with a revealed competitive effect value greater than 1.0.

The table provides a clearer insight into Africa’s performance in regional markets with a

breakdown of underperforming and outperforming countries by regional group membership. It

helps to apprehend for each REC how many of its members have underperformed or outperformed

their competitors in intra-regional versus extra-regional markets.

97

Table 4.5. Breakdown by REC membership of the numbers of underperforming and outperforming

countries in alternative agricultural export destination markets

Global markets

Intra-African markets

COMESA markets

ECCAS markets

ECOWAS markets

SADC markets

Number of underperforming countries (with competitive effect < 1.0)

COMESA members 12 4 4 8 11 6

ECCAS members 10 5 6 8 7 7

ECOWAS members 7 8 6 8 6 12

SADC members 8 3 4 8 8 5

Whole sample 32 20 19 26 27 28

Number of outperforming countries (with a competitive effect > 1.0)

COMESA members 6 14 14 8 7 12

ECCAS members 1 6 4 3 4 4

ECOWAS members 8 7 8 7 9 3

SADC members 3 8 7 3 3 6

Whole sample 19 30 29 20 23 22

Total number of countries in sample

Whole sample 51 50 48 46 50 50 Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

For instance, the first row of the upper panel of the table shows that for the COMESA region only

4 of its members have underperformed in their intra-regional markets compared to 11 members in

farther extra-regional markets located in the ECOWAS region. Conversely, we can read from the

first row of the lower panel of the table that for the COMESA region up to 14 of its members have

outperformed their competitors in their intra-regional markets compared to only 7 members in

extra-regional markets within ECOWAS. Similarly, the ECOWAS region also has a smaller

number of underperforming members in intra-regional markets than in remoter extra-regional

markets situated in the SADC region. The same is true for the SADC region which has fewer

underperforming members in intra-regional markets than in the remoter ECOWAS and ECCAS

markets. However, for the ECCAS region we see more underperforming and fewer outperforming

members in intra-regional than in extra-regional markets. This is surprising enough as one would

expect countries to be more performant in their region than in remoter regions.

Disparities between regional country groups as regards their competitiveness gains or losses in

intra-regional versus extra-regional markets are more clearly revealed in Figure 4.10 below.

COMESA members have achieved a positive average competitiveness change in intra-regional

98

markets and to a lesser extent in SADC markets, but a negative average change in the more distant

ECCAS and ECOWAS markets. ECOWAS members have also on average raised their

competitiveness in intra-regional markets and reduced their competitiveness in extra-regional

markets, with the biggest average reduction incurred in the remotest SADC markets. SADC

members have kept their average competitiveness level practically unchanged in intra-regional and

COMESA markets, but they have on average lost performance in ECOWAS markets and more

notably in ECCAS markets. The patterns are different for the ECCAS region, which has

underperformed in all regional markets and more remarkably in intra-regional markets.

Figure 4.10. Country-group average competitiveness change in regional agricultural export

markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

Furthermore, Figure 4.10 shows how group average competitiveness changes in regional markets

compare to corresponding Africa-wide average changes. The statistical significance of pairwise

comparisons has been tested through analysis of variance of country competitiveness changes in

regional markets. Major comparison test results are summarized in Tables A4.5-A4.8 in the

appendices. It appears that the COMESA region has raised its competitiveness in intra-regional

and SADC markets significantly more than the rest of Africa. The ECOWAS region has performed

significantly better than the rest of Africa only in SADC markets. And the ECCAS region has

0.88

0.90

0.92

0.94

0.96

0.98

1.00

1.02

1.04

1.06

1.08

1.10

COMESA ECCAS ECOWAS SADC Africa

Ave

rag

e c

om

petit

iveness

chang

e

Regional Country Groups

COMESA markets ECCAS markets ECOWAS markets SADC markets

99

undergone a significantly stronger loss of competitiveness than the rest of Africa in intra-regional

and COMESA markets. These patterns of disparities between regional groups of countries suggest

that differences in country competitiveness should be explained by other factors than trading

distance and costs. Differences in the competitiveness of most traded goods in individual countries

may contribute to the explanation.

As defined above, Figure A4.2 presents the rankings of commodities in increasing order of their

competitiveness change in the different regional markets. For some commodities, mostly among

those ranked towards the uppermost edge of the rankings, competitiveness changes are higher in

regional markets than in global and or intra-African markets.

However, for other commodities, mostly towards the lowermost edge of the rankings, the reverse

is true. In order to assess the consistency and significance of these comparisons, paired-sample T

tests of equality of competitiveness changes in regional markets compared with global and intra-

African markets are carried out and the results summarized in Table 4.4. The upper panel of the

table shows that commodity competitiveness changes in global markets are positively but weakly

correlated with changes in COMESA, as well as ECCAS and SADC markets. There is no

significant correlation between competitiveness changes in global and ECOWAS markets. On

average commodity competitiveness changes are lower by 0.037 point in ECCAS markets

compared to global markets at the 1% significance level, versus 0.02 point in ECOWAS markets

at the 10% significance level. In contrast, there is on average no significant difference in

competitiveness changes in global and COMESA or SADC markets.

Comparisons reported in the lower panel of the table reveal positive and weak correlations of

commodity competitiveness changes in intra-African and intra-regional markets, except for

COMESA and SADC, where competiveness changes are more strongly associated with changes

in intra-African markets. This means that changes in intra-African markets reflect changes in

COMESA and SADC significantly more than elsewhere in Africa. On average commodity

competitiveness changes are lower by 0.022 point in ECCAS markets than elsewhere in Africa at

the 5% significance level.

The distribution of commodities across different classes of competitiveness is summarized in

Table 4.6 below.

100

The loss of competitiveness by African countries has affected a greater number of commodities in

ECCAS markets compared to the other regional markets. For a total of 32 commodities,

competitive effect values are smaller than 1.0, including 26 with small competitiveness losses but

only 6 with high losses. Conversely, competitiveness gains achieved by African exporters have

benefited a greater number of commodities in COMESA markets compared to the other regional

markets. The benefit concerns up to 31 commodities with small gains and only 8 with high gains.

However, the number of commodities with increased competitiveness is still greater in global

markets than in regional markets. In other words, there is room for expanding Africa’s share of

total world agricultural exports by aligning competitiveness changes in regional markets with

improvements being made outside Africa.

Table 4.6. Number of commodity groups by class of competitiveness in alternative agricultural export

destination markets

Export destination markets

Competitiveness class Global

markets

Intra-African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

Competitive effect<=0.9 0 2 1 6 2 2

0.9<Competitive effect<=1.0 16 27 19 26 22 24

1.0<Competitive effect<=1.1 38 23 31 23 30 28

Competitive effect >1.1 5 7 8 4 5 5

Whole sample size 59 59 59 59 59 59

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

Among the commodities that have lost competitiveness in at least three regional markets we can

find cotton, wheat, sorghum, some oilseeds12, meat & edible offal, groundnut oil and tea. They all

have also been ranked among uncompetitive products in intra-African markets and, with the

exception of wheat and sorghum, in global markets. Therefore these commodities could be thought

of as the most uncompetitive commodities in African markets. Towards the topmost edge of the

rankings, many foodstuffs are found among the commodities that have gained competitiveness in

at least three regional markets, including rice, potatoes, onions & substitutes, fish and sea foods,

sheep & goats, other live animals13, and roots & tubers.

12 Not including soybeans and groundnuts 13 This group is a broad aggregate of live swine, horses, asses, mules and hinnies

101

They all have also shown a competitiveness gain in global markets, except for fish and sea foods,

as well as in intra-African markets, except for onions & substitutes and sheep & goats, as these

two commodity groups have lost competitiveness in ECOWAS markets. Therefore, ECOWAS

markets may be more stringent for African exports of onions & substitutes and sheep & goats, as

non-African markets may be for African exports of fish and sea foods.

In an attempt to assess how important the top ranked commodities are, Figure 4.11 shows the

cumulative share of Africa’s total agricultural exports to alternative markets that is contributed by

an increasing number of top competitive commodities in those markets. First of all, the figure

recalls the finding that the topmost competitive commodities in global and intra-African markets

account for small shares of African export baskets in these markets. The same is true as regards

regional markets. However, as we have already noted, the most competitive commodities represent

higher cumulative shares of export baskets in intra-African markets than in global markets. They

also account for higher shares of Africa’s exports to regional markets compared to global markets.

The top 5 and 10 commodities weigh more heavily in ECOWAS markets than in other intra-

African markets. For instance, the top 5 most competitive commodities in ECOWAS markets

account for 15.1% of Africa’s exports to that region while the corresponding shares as regards all

intra-African markets and global markets are 1.3% and 1.8%, respectively. Thus, the most

competitive products in the different markets are not among the most exported ones, which reveals

that competitiveness gains are happening among products that can be exploited for widening the

export bases of African countries.

102

Figure 4.11. Relative importance of the most competitive commodities in regional markets

compared to global and intra-African markets

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

The scope for expansion of intra- and extra-African exports by tapping into revealed

competitiveness gains appears in the fact that there is no single set of commodities gaining

competitiveness at the same pace in the different export destinations. This is demonstrated in

Figure 4.12 below which shows how dissimilar commodity rankings are in the different export

markets. The intuition behind the construction of the figure is that commodity rankings would be

considered to be similar if commodity ranks were approximately the same in the different rankings

(markets). In that case, all top K most competitive commodities in the different rankings would be

found in a unique set of K products as depicted by the 45 degree straight line.

The more the size of the set is greater than K the more dissimilar would be the different rankings.

The distance from the curved line to the straight line shows how dissimilar the rankings are. For

instance, the curved line shows that a set of 16 products encompasses all top 5 commodities in all

rankings. Similarly, the size of the set that includes all top 10 commodities in all rankings amounts

to 32. In other words, the most competitive commodities are not exactly the same in different

markets, which justifies the belief that there is scope for a diversified export expansion in the

different markets under analysis. Put differently, somewhat different baskets of non-traditional

export products are gaining competitiveness in the different markets and are good candidates for

export diversification and expansion.

0

10

20

30

40

50

60

70

80

90

100

5 10 15 20 25 30 35 40 45 50 55 59

Cumu

lative

aver

age s

hare

of A

frican

agric

ultur

al ex

ports

to th

e diffe

rent

marke

ts (%

)

Number of top competitive commodities inthe different agricultural export markets

COMESA markets

ECCAS markets

ECOWAS markets

SADC markets

Global markets

Intra-African markets

103

Figure 4.12. Dissimilarity of commodity rankings in the different export destination markets

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

4.6. Determinants of export competitiveness in global and regional markets

The preceding sections have highlighted considerable variations between African countries in

terms of changes in their competitiveness as compared to the group of their non-African

competitors in agricultural export markets. We have seen that the patterns of competitiveness

changes differ across export markets but also according to membership in the different RECs.

Trading distance and costs have appeared to affect the changes in competiveness experienced by

member countries of the different RECs in intra-regional markets as compared to extra-regional

markets. However, the larger part of differences between countries as regards their

competitiveness gains or losses seems to have to do more with country-specific production and

trade environments than with regional differences. Indeed, the analysis of commodity

competitiveness changes has suggested that differences in productivity gains and domestic market

conditions may play a large role in the differences of competitiveness gains or losses achieved by

African countries for the different commodities. This section is devoted to exploring the factors

behind the disparities among countries in terms of the changes in their competitiveness in the

different markets. Potential determinants considered include agricultural total factor productivity

changes from the USDA database, the World Bank’s Doing Business – Distance to Frontier (DB-

DTF) indicator, the World Economic Forum’s Global Competitiveness Index (GCI) and country

0

5

10

15

20

25

30

35

40

45

50

55

60

0 5 10 15 20 25 30 35 40 45 50 55 60

Siz

e of

the

set o

f top

K m

ost c

ompe

titiv

e co

mm

oditi

es in

all

expo

rt m

arke

tsun

der

anal

ysis

Value of K

104

attributes related to each of its 12 pillars, the International Logistics Performance Index and its

component indicators, and Transparency International’s Corruption Perceptions Index (CPI).

Tables 4.7 and 4.8 present the results of a linear regression analysis where the series of country

competitiveness changes in the different agricultural export destination markets are pooled

together as a single variable and regressed on the above-listed country-level indicators taken as

potential explanatory variables while controlling for REC membership and export destination

markets, as formally summarized in Eq. 8 below:

𝐶𝑂𝑀𝑃𝑚𝑟𝑗 = 𝛼 + ∑ 𝛽𝑟 ∙ 𝑅𝐸𝐶𝑟𝑅𝑟 + ∑ 𝛾𝑗 ∙ 𝑀𝐾𝑇𝑗

𝐽𝑗 + ∑ 𝜃𝑝 ∙𝑃

𝑝 𝐼𝑁𝐷𝑝 + 휀𝑚𝑟𝑗 (8)

where 𝐶𝑂𝑀𝑃𝑚𝑟𝑗 is the pooled variable standing for the change in competitiveness for country 𝑚,

which is a member of the Regional Economic Community 𝑟, in export markets 𝑗; 𝑅𝐸𝐶𝑟 represents

dummy variables for the different Regional Economic Communities and 𝑀𝐾𝑇𝑗 are dummy

variables for the different export destination markets; and 𝐼𝑁𝐷𝑝 stands for the different indicators

considered above as potential explanatory variables.

Table 4.7. Parameter estimates for the determinants of changes in country competitiveness

Coefficients Std. Error t Sig.

Constant 0.560 0.085 6.612 0.000

SADC region -0.062 0.016 -3.872 0.000

Intra-African markets 0.039 0.017 2.267 0.025

Doing Business - Distance to frontier a 0.003 0.001 2.242 0.026

Institutions (GCI 1st Pillar) b 0.043 0.018 2.316 0.022

Country market size (GCI 10th Pillar) b 0.048 0.011 4.182 0.000

LPI - Customs c 0.150 0.026 5.815 0.000

LPI - International shipments c -0.128 0.029 -4.396 0.000

Agricultural TFP growth estimates 1961-2012 -1.613 0.949 -1.701 0.091

a. Doing Business - Distance to frontier, maximum score between 2010 and 2016

b. Global Competitiveness Index, average attribute value between 2006 and 2015

c. International Logistics Performance Index (LPI 2014 score)

Source: Authors’ calculations.

105

Table 4.8. ANOVA and model summary

Sum of Squares df Mean Square F Sig.

Regression 0.769 8 0.096 12.321 0.000

Residual 1.381 177 0.008

Total 2.150 185

Number of observations 186

R Square 0.36

Adjusted R Square 0.33

Durbin-Watson 2.36 Source: Authors’ calculations.

The subset of explanatory variables that provide the best model fit are presented in Table 4.7. As

we have seen above, country competitiveness changes are higher in intra-African markets as

compared to global markets. They appear to be positively affected by the Doing Business –

Distance to Frontier score, the quality of institutions, country market size and the quality of

customs service. Surprisingly, the model reveals that changes in country competitiveness are

negatively associated with the ease of international shipments and changes in agricultural total

factor productivity. Table 4.8 shows that the model accounts for nearly two-fifth of the variation

in changes in competitiveness.

4.7. Conclusions

Changes in African agricultural export competitiveness have been explored in global, intra-

African, and regional markets over the period 1998-2013. Almost consistently in all export markets

under consideration, ECCAS members appear to have underperformed their competitors, while

SADC, COMESA and ECOWAS members have on average proved to have preserved their

competitiveness or outperformed the group of their competitors. In addition, changes in country

competitiveness are on average lower in ECCAS markets and generally higher in intra-African

markets than in global markets. The analysis has also shown that competitiveness gains have taken

place for the COMESA, ECOWAS and SADC members remarkably more in intra-regional than

in extra-regional markets. But for ECCAS, rare increases in country competitiveness have been

noted and they have happened in extra-regional markets and not in intra-regional markets.

However, it should be retained that while ECCAS is notably lagging behind, the proportions of

underperforming countries within COMESA, SADC and ECOWAS are also a concern.

106

Africa’s competitiveness analysis at the commodity level has revealed significant losses for some

important foodstuffs, though the majority of commodities have gained more competitiveness in

global markets. However, the levels of commodity competitiveness are lower in intra-African than

in global markets. They are even lower in regional markets, except in COMESA markets, where

the commodity competitiveness level is higher than in global and intra-African markets. In other

words, there is room for expanding Africa’s share of total world agricultural exports by aligning

competitiveness changes in regional markets with improvements being made outside Africa. The

top ranked commodities contribute a small share of intra-African agricultural export value and an

even smaller share of Africa’s global agricultural export value. This reflects the scope for

expanding African exports by exploiting increased competitiveness that arises among new and

emerging export products. The results show that the set of these candidate products for export

expansion varies remarkably across the different export destination markets, showing the scope

for product diversification for countries in conquering African and world markets.

Apart from REC membership, the Doing Business – Distance to Frontier score, the quality of

domestic institutions, country market size and the quality of customs service have been shown to

significantly contribute to the explanation of the variability in competitiveness changes.

107

References

Cheptea, A., Gaulier, G., & Zignago, S. (2005). World Trade Competitiveness: A Disaggregated View by

Shift-Share Analysis. CEPII Working Paper 2005-23. Paris: CEPII.

Hausman, R., Hwang, J., & Rodrik, D. (2005). What you export matters. Journal of Economic Growth,

12(1), 1–25.

Leamer, E., & Stern, R. (1970). Quantitative International Economics. Aldine.

Magee, S. (1975). Prices, income, and foreign trade. In P. Kenen (Ed.), International Trade and Finance:

Frontiers for Research. New York: Cambridge University Press.

Richardson, J. D. (1971a). Constant-market-shares analysis of export growth. Journal of International

Economics, 1(2), 227–239.

Richardson, J. D. (1971b). Some sensitivity tests for a “constant market shares analysis” of export

growth. Review of Economics and Statistics, 53, 300-304.

Tyszynski, H. (1951). World trade in manufactured commodities, 1899-1950. The Manchester School of

Economic and Social Studies, 19, 222–304.

108

4.9. Tables and figures

Table A4.1. Change in country competitiveness in alternative agricultural export destination markets, 1998-

2013

Global

markets

Intra-African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

Algeria 1.111 1.212 1.083 1.050 1.163 1.051

Angola 0.882 1.025 0.757 0.796 1.005 0.978

Benin 0.959 0.914 1.110 0.914 0.913 0.992

Burkina Faso 1.033 0.993 0.832 1.075 1.053 0.724

Burundi 0.976 1.183 1.089 1.037 0.900 1.071

Cameroon 0.984 0.966 0.841 0.971 0.964 0.865

Cape Verde 1.211 1.092 1.110 1.039 1.083 0.892

Central African Republic 0.903 0.818 0.715 0.706 0.948 0.859

Chad 0.900 0.859 0.958 0.650 1.067 0.931

Comoros 0.984 1.235 1.148 0.812 0.725 1.128

Congo 0.974 1.042 0.774 0.931 0.937 1.102

Côte d'Ivoire 0.976 0.971 1.032 0.976 0.999 0.895

Demo. Republic of Congo 0.939 1.071 1.087 1.027 0.972 0.911

Djibouti 1.104 1.236 1.178 1.095 0.940

Egypt 1.098 1.232 1.198 1.115 1.084 1.080

Equatorial Guinea 0.758 1.073 0.850 1.141 1.057

Eritrea 0.949 1.171 1.189 1.092 1.017

Ethiopia 1.071 1.203 1.110 1.107 1.057 1.103

Gabon 0.918 0.990 1.016 0.956 0.841 0.915

Gambia 0.986 1.022 0.991 0.879 1.040 0.849

Ghana 1.065 1.163 1.133 1.051 1.191 0.992

Guinea 0.966 1.011 1.010 0.772 1.066 0.837

Guinea Bissau 1.035 1.163 0.893 1.206 1.085

Kenya 0.987 0.976 0.980 0.939 0.952 0.997

Liberia 1.053 0.975 0.897 1.107 1.069 0.900

Libya 0.963 0.973 1.233 0.990 0.717 1.057

Madagascar 0.944 0.947 0.949 0.792 0.944 0.902

Malawi 0.984 1.004 1.061 1.032 1.003 1.013

Mali 0.931 0.805 0.703 0.859 0.779 0.717

Mauritania 0.995 1.030 1.073 1.033 1.012 1.177

Mauritius 0.971 1.024 1.020 0.758 0.967 1.055

Morocco 0.997 1.134 1.093 1.078 1.161 1.099

Mozambique 1.027 1.029 1.069 0.986 0.871 1.030

Niger 1.009 0.941 0.827 0.884 0.941 0.963

Nigeria 1.093 1.088 1.040 1.127 1.046 1.093

Rwanda 1.067 1.175 1.197 1.070 1.037 1.158

SACU countries 0.986 0.975 0.983 0.950 0.992 0.971

Saint Helena 0.995 0.731 0.719 0.841 0.822

109

Sao Tome & Principe 0.901 0.905 0.829 0.897 0.902 0.921

Senegal 0.971 1.044 1.099 1.019 1.074 1.029

Seychelles 0.982 1.027 1.084 0.966 0.889 1.032

Sierra Leone 1.045 0.963 1.060 1.135 0.920 0.734

Somalia 1.125 0.906 0.956 0.775 0.937

Sudan 0.968 1.008 0.996 1.016 0.877 0.743

Tanzania 1.004 1.027 1.025 1.125 0.965 1.056

Togo 0.995 0.950 0.807 0.934 0.937 0.871

Tunisia 1.022 1.176 1.044 1.047 1.063 0.930

Uganda 1.003 1.015 1.023 1.040 0.961 1.052

Western Sahara 0.853

Zambia 1.062 1.051 1.091 0.996 1.196 1.069

Zimbabwe 0.916 0.915 0.841 0.857 0.901 0.919

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

110

Table A4.2. Country shares in the value of Africa’s agricultural exports to alternative markets, 1998-2013

average (%)

Exporters

Global

markets

Intra-African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

Algeria 0.411 0.775 0.423 0.025 1.829 0.018

Angola 0.110 0.090 0.000 0.040 0.194 0.095

Benin 0.890 1.362 0.060 0.141 4.516 0.197

Burkina Faso 1.103 1.760 0.215 0.004 5.752 0.187

Burundi 0.178 0.097 0.167 0.177 0.014 0.078

Cameroon 2.399 1.098 0.203 5.239 0.333 0.288

Cape Verde 0.054 0.031 0.001 0.002 0.054 0.001

Central African Republic 0.078 0.076 0.098 0.193 0.040 0.030

Chad 0.261 0.108 0.023 0.429 0.048 0.027

Comoros 0.093 0.017 0.028 0.001 0.001 0.041

Congo 0.145 0.272 0.032 1.474 0.053 0.063

Côte d'Ivoire 12.225 7.124 0.227 2.476 17.027 1.301

Demo. Republic of Congo 0.149 0.160 0.293 0.260 0.027 0.040

Djibouti 0.137 0.276 0.552 0.000 0.006 0.010

Egypt 6.463 5.082 6.420 1.715 1.363 0.978

Equatorial Guinea 0.054 0.003 0.000 0.005 0.004 0.000

Eritrea 0.016 0.012 0.027 0.001 0.001 0.000

Ethiopia 2.894 2.887 3.490 0.045 0.057 0.227

Gabon 0.114 0.337 0.001 2.228 0.034 0.010

Gambia 0.124 0.137 0.003 0.012 0.481 0.032

Ghana 5.336 1.224 0.072 0.306 4.106 0.150

Guinea 0.344 0.610 0.008 0.030 1.416 0.006

Guinea Bissau 0.256 0.079 0.000 0.049 0.265 0.001

Kenya 5.974 7.380 13.475 3.592 0.573 4.468

Liberia 0.031 0.021 0.002 0.001 0.038 0.002

Libya 0.095 0.106 0.034 0.006 0.030 0.002

Madagascar 1.577 0.374 0.538 0.012 0.046 0.555

Malawi 2.030 2.331 2.854 0.335 0.154 3.982

Mali 1.125 3.068 0.286 0.005 10.757 0.340

Mauritania 1.557 2.712 0.057 3.888 8.192 0.026

Mauritius 1.889 0.841 1.347 0.070 0.591 1.540

Morocco 8.839 3.478 1.571 5.033 6.251 1.268

Mozambique 1.251 1.593 2.236 0.084 0.029 4.148

Niger 0.557 2.491 0.081 0.022 8.917 0.008

Nigeria 3.433 1.308 0.159 0.719 3.183 0.647

Rwanda 0.273 0.621 1.263 0.973 0.004 0.566

SACU countries 19.025 25.132 30.421 43.820 10.880 50.927

Saint Helena 0.024 0.005 0.006 0.000 0.001 0.005

Sao Tome & Principe 0.028 0.005 0.004 0.012 0.006 0.003

111

Senegal 1.774 2.417 0.062 2.858 6.608 0.050

Seychelles 0.885 0.441 0.875 0.001 0.055 1.086

Sierra Leone 0.091 0.017 0.002 0.001 0.027 0.007

Somalia 0.342 0.077 0.057 0.000 0.194 0.004

Sudan 1.437 1.098 2.187 0.003 0.013 0.090

Tanzania 2.882 2.521 4.754 5.161 0.144 2.487

Togo 0.819 1.163 0.025 0.262 3.695 0.044

Tunisia 3.112 4.430 7.082 0.796 1.664 0.115

Uganda 2.509 3.945 7.772 8.026 0.191 2.210

Western Sahara 0.006 0.004 0.000 0.000 0.015 0.000

Zambia 1.260 4.079 6.675 7.993 0.015 10.422

Zimbabwe 3.344 4.728 3.829 1.476 0.105 11.216

Africa 100 100 100 100 100 100

Source: Authors’ calculations using the BACI database.

112

Table A4.3. Change in commodity competitiveness in alternative agricultural export destination markets, 1998-

2013.

Global

markets

Intra-

African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

Cattle 1.130 1.058 1.129 0.996 1.000 0.980

Sheep & goats 1.092 0.949 1.051 1.010 0.999 1.036

Poultry 0.955 0.994 1.018 0.967 1.029 0.983

Other live animals 1.063 1.035 1.040 0.951 1.016 1.013

Meat & edible offal 0.951 0.949 0.991 0.918 1.009 0.918

Fish & sea foods 0.986 1.020 1.033 0.979 1.025 1.045

Dairy, eggs & honey 1.116 1.040 1.074 0.972 1.030 0.976

Other animal products 1.003 1.017 1.036 0.983 1.026 1.026

Roots & tubers 1.097 1.101 1.043 1.103 0.954 1.012

Other live trees & plants 1.055 0.990 1.031 0.997 0.942 1.004

Potatoes 1.034 1.035 0.967 1.015 1.066 1.002

Tomatoes 1.036 1.022 1.006 0.993 1.072 0.999

Onions & substitutes 1.054 0.949 1.021 1.030 0.905 1.019

Other edible vegetables 1.062 1.110 1.102 0.983 0.993 1.013

Edible fruits & nuts 1.009 0.996 0.980 1.005 1.016 1.006

Coffee 0.961 0.963 0.945 0.926 1.032 1.001

Tea 0.995 0.998 1.005 0.859 0.961 0.998

Spices 0.984 1.028 1.047 1.062 0.985 0.985

Wheat 1.050 0.990 0.934 0.997 1.177 0.933

Rye, barley & oats 1.216 1.243 1.140 1.045 0.846 1.382

Maize 1.031 0.987 0.991 1.035 0.971 1.033

Rice 1.019 1.037 1.042 1.017 1.023 1.071

Sorghum 1.030 0.967 0.950 0.798 0.968 1.007

Other cereals 0.993 0.985 0.976 1.006 1.020 0.974

Milling industry products 1.026 1.042 1.062 1.027 1.047 1.005

Soybeans 1.073 0.884 0.842 0.887 1.052 1.040

Groundnuts 1.005 0.998 1.089 0.992 1.016 1.014

Other oilseeds 1.014 0.967 0.954 0.997 1.034 0.975

Medicinal plants 1.044 1.019 1.016 0.946 0.992 0.998

Gums & resins 1.000 1.163 1.080 0.974 1.024 1.099

Vegetable plaiting materials 1.027 1.067 0.975 1.047 0.921 1.132

Animal fats 1.096 1.059 1.147 1.115 1.015 1.047

Soybean oil 1.148 1.138 1.147 1.162 1.246 1.068

Groundnut oil 0.943 0.935 1.004 0.935 0.949 0.992

Olive oil 1.013 1.173 1.205 1.073 1.250 1.164

Palm oil 0.985 0.988 1.066 0.925 0.921 1.026

Other oils & facts 1.066 1.071 1.063 1.080 1.033 1.041

Edible preps. of meat, fish & crustaceans 0.995 1.021 1.009 1.014 1.084 0.986

Cane sugar 0.963 1.002 1.001 1.009 0.982 0.977

113

Sugar confectionery 1.010 0.984 0.994 0.985 0.970 0.980

Cocoa beans 1.002 0.945 1.009 1.068 1.012 0.944

Cocoa preparations 1.039 0.987 0.991 1.012 1.039 0.982

Preps. of cereals, flour, starch or milk 1.087 1.016 1.041 0.997 1.011 1.011

Preps. of vegs., fruits & nuts 0.997 1.029 1.022 1.043 1.067 0.986

Misc. edible preparations 1.018 1.015 1.025 0.999 1.021 0.982

Beverages, spirits & vinegar 1.022 0.988 1.037 0.948 1.005 0.971

Residues from food industries 1.038 1.039 1.100 0.970 0.955 1.003

Tobacco & substitutes 1.035 1.042 1.044 1.029 1.110 0.997

Organic chemicals 0.952 0.859 0.901 0.898 0.821 0.873

Essential oils & resinoids 1.009 0.974 0.980 0.995 0.976 0.968

Albuminoidal substances 1.038 0.999 0.960 1.078 1.030 1.011

Finishing agents for textiles & paper 1.029 0.947 0.995 1.075 0.933 1.009

Hides & skins 0.993 1.088 0.963 1.030 0.920 1.235

Furskins 1.020 0.989 1.070 0.870 1.050 1.122

Silk 1.126 0.944 1.205 1.125 0.994 0.942

Wool 1.078 1.049 1.020 1.000 1.073 0.862

Cotton, not carded or combed 0.961 0.951 0.937 0.878 0.967 0.999

Cotton, carded or combed 1.009 0.988 0.997 0.907 0.911 1.012

Other vegetable textile fibres 1.015 1.130 1.149 0.905 1.140 1.023

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from commodity-level export share decomposition analysis for African countries as a group.

114

Table A4.4. Commodity shares in the value of Africa’s agricultural exports to alternative markets, 1998-2013

average (%)

Global

markets

Intra-

African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

Cattle 0.41 1.623 0.875 0.439 4.098 0.686

Sheep & goats 0.61 0.686 0.076 0.065 2.517 0.040

Poultry 0.02 0.120 0.104 0.043 0.033 0.229

Other live animals 0.29 0.477 0.708 0.081 0.454 0.225

Meat & edible offal 0.88 0.871 0.630 1.249 1.005 1.451

Fish & sea foods 11.66 7.599 3.512 11.800 15.716 5.486

Dairy, eggs & honey 1.18 3.171 3.520 3.693 2.804 3.675

Other animal products 0.37 0.228 0.196 0.035 0.530 0.200

Roots & tubers 0.04 0.015 0.021 0.006 0.023 0.006

Other live trees & plants 2.14 0.468 0.432 0.321 0.162 0.344

Potatoes 0.51 0.343 0.294 0.851 0.051 0.651

Tomatoes 0.87 0.107 0.103 0.058 0.102 0.087

Onions & substitutes 0.37 0.649 0.224 0.606 1.643 0.396

Other edible vegetables 3.35 2.800 2.616 1.769 1.461 1.793

Edible fruits & nuts 12.77 2.786 2.052 1.663 3.277 2.596

Coffee 4.66 3.852 2.377 0.584 0.509 0.832

Tea 2.68 5.216 10.621 1.014 0.563 1.775

Spices 1.01 0.532 0.584 0.138 0.162 0.563

Wheat 0.19 0.932 1.532 0.305 0.792 1.521

Rye, barley & oats 0.02 0.066 0.094 0.071 0.003 0.101

Maize 0.91 3.824 6.990 2.108 0.671 7.104

Rice 0.72 1.625 2.064 1.267 2.520 0.918

Sorghum 0.06 0.185 0.331 0.050 0.090 0.214

Other cereals 0.05 0.195 0.199 0.066 0.319 0.110

Milling industry products 0.74 4.008 6.087 8.829 2.953 5.924

Soybeans 0.07 0.225 0.380 0.445 0.011 0.351

Groundnuts 0.27 0.417 0.308 0.242 0.246 0.579

Other oilseeds 1.73 1.252 1.402 0.236 0.865 0.859

Medicinal plants 0.94 0.693 0.857 0.594 0.400 0.961

Gums & resins 0.67 0.376 0.280 0.813 0.385 0.180

Vegetable plaiting materials 0.22 0.849 1.010 0.015 0.009 0.077

Animal fats 0.11 0.102 0.146 0.025 0.098 0.157

Soybean oil 0.18 0.729 1.187 0.264 0.169 1.324

Groundnut oil 0.26 0.023 0.012 0.016 0.033 0.024

Olive oil 1.14 0.175 0.196 0.232 0.026 0.189

Palm oil 0.56 2.699 1.977 3.212 5.753 1.725

Other oils & facts 1.02 3.858 6.063 3.672 2.365 4.370

Edible preps. of meat, fish & crustaceans 3.56 1.889 1.081 4.429 2.896 1.755

115

Cane sugar 3.85 6.382 8.727 9.292 1.785 8.471

Sugar confectionery 0.62 1.691 1.474 2.496 1.595 2.008

Cocoa beans 12.18 0.570 0.012 0.010 0.342 0.416

Cocoa preparations 3.68 1.100 1.064 1.171 0.453 1.612

Preps. of cereals, flour, starch or milk 0.70 2.825 2.888 2.584 3.501 2.770

Preps. of vegs., fruits & nuts 2.39 2.069 2.674 1.614 1.244 2.458

Misc. edible preparations 1.62 5.366 3.301 5.065 8.795 4.087

Beverages, spirits & vinegar 3.11 5.578 3.964 16.045 4.270 9.001

Residues from food industries 1.05 2.319 2.314 0.509 0.948 2.835

Tobacco & substitutes 5.88 9.696 9.181 9.321 9.861 10.510

Organic chemicals 0.00 0.004 0.008 0.002 0.001 0.010

Essential oils & resinoids 0.27 0.083 0.097 0.059 0.031 0.174

Albuminoidal substances 0.03 0.094 0.142 0.090 0.052 0.157

Finishing agents for textiles & paper 0.00 0.018 0.034 0.004 0.006 0.038

Hides & skins 0.76 0.169 0.176 0.010 0.082 0.119

Furskins 0.02 0.002 0.001 0.001 0.001 0.004

Silk 0.00 0.002 0.003 0.001 0.000 0.006

Wool 0.47 0.037 0.074 0.001 0.003 0.053

Cotton, not carded or combed 5.69 5.971 2.366 0.270 11.073 5.398

Cotton, carded or combed 0.35 0.359 0.359 0.156 0.270 0.394

Other vegetable textile fibres 0.05 0.002 0.002 0.001 0.001 0.001

Agricultural exports 100 100 100 100 100 100

Source: Authors’ calculations using the BACI database.

116

Figure A4.1. Change in country competitiveness in regional exports markets compared to global and intra-

African markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3M

ali

Ce

ntr

al A

fric

an R

ep.

Sa

int H

ele

na

An

go

la

Co

ng

o

Togo

Nig

er

Sa

o T

om

e &

…

Bu

rkin

a F

aso

Zim

bab

we

Ca

me

roon

Lib

eria

Ma

da

ga

scar

So

ma

lia

Ch

ad

Ke

nya

SA

CU

co

un

trie

s

Ga

mb

ia

Su

da

n

Gu

inea

Ga

bo

n

Ma

uritiu

s

Ug

an

da

Tan

zan

ia

Cô

te d

'Ivo

ire

Nig

eria

Tun

isia

Sie

rra

Leo

ne

Ma

law

i

Mo

zam

biq

ue

Ma

urita

nia

Alg

eria

Se

yche

lles

D.R

. C

ong

o

Bu

run

di

Zam

bia

Mo

rocco

Se

ne

ga

l

Ca

pe

Ve

rde

Eth

iopia

Be

nin

Gh

an

a

Co

moro

s

Djib

ou

ti

Eritr

ea

Rw

and

a

Eg

ypt

Lib

ya

Ch

an

ge

in

co

mp

etitive

ne

ss

COMESA markets Global markets Intra-African markets

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Ch

ad

Ce

ntr

al A

fric

an R

ep

.

Ma

uritiu

s

Guin

ea

Ma

dag

asca

r

Ang

ola

Co

mo

ros

Equ

ato

rial G

uin

ea

Zim

ba

bw

e

Ma

li

Gam

bia

Nig

er

Guin

ea

Bis

sa

u

Sao

Tom

e &

…

Ben

in

Co

ng

o

To

go

Ken

ya

SA

CU

cou

ntr

ies

Gab

on

Seych

elle

s

Ca

me

roon

Cô

te d

'Ivoir

e

Mo

za

mbiq

ue

Lib

ya

Za

mb

ia

Sud

an

Sen

ega

l

D.R

. C

ong

o

Ma

law

i

Ma

urita

nia

Buru

nd

i

Ca

pe

Verd

e

Ug

an

da

Tu

nis

ia

Alg

eri

a

Gha

na

Rw

an

da

Burk

ina

Fa

so

Mo

rocco

Lib

eri

a

Eth

iop

ia

Egyp

t

Ta

nza

nia

Nig

eria

Sie

rra

Le

on

eCh

an

ge

in

co

mp

etitive

ne

ss

ECCAS markets Global markets Intra-African markets

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Lib

ya

Co

mo

ros

So

ma

lia

Ma

li

Sa

int H

ele

na

Ga

bo

n

Mo

zam

biq

ue

Su

da

n

Se

yche

lles

Bu

run

di

Zim

bab

we

Sa

o T

om

e &

…

Be

nin

Sie

rra

Leo

ne

Togo

Co

ng

o

Nig

er

Ma

da

ga

scar

Ce

ntr

al A

fric

an R

ep.

Ke

nya

Ug

an

da

Ca

mero

on

Tan

zan

ia

Ma

uritiu

s

D.R

. C

ong

o

SA

CU

co

un

trie

s

Cô

te d

'Ivo

ire

Ma

law

i

An

go

la

Ma

urita

nia

Rw

and

a

Ga

mb

ia

Nig

eria

Bu

rkin

a F

aso

Eth

iopia

Tun

isia

Gu

inea

Ch

ad

Lib

eria

Se

ne

ga

l

Ca

pe

Ve

rde

Eg

ypt

Eritr

ea

Djib

ou

ti

Eq

ua

toria

l G

uin

ea

Mo

rocco

Alg

eria

Gh

an

a

Zam

bia

Gu

inea

Bis

sa

uCh

an

ge

in

co

mp

etitive

ne

ss

ECOWAS markets Global markets Intra-African markets

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Ma

li

Bu

rkin

a F

aso

Sie

rra

Leo

ne

Su

da

n

Sa

int H

ele

na

Gu

inea

Ga

mb

ia

Ce

ntr

al A

fric

an R

ep.

Ca

me

roon

Togo

Ca

pe

Ve

rde

Cô

te d

'Ivo

ire

Lib

eria

Ma

da

ga

scar

D.R

. C

ong

o

Ga

bo

n

Zim

bab

we

Sa

o T

om

e &

…

Tun

isia

Ch

ad

So

ma

lia

Djib

ou

ti

Nig

er

SA

CU

co

un

trie

s

An

go

la

Be

nin

Gh

an

a

Ke

nya

Ma

law

i

Eritr

ea

Se

ne

ga

l

Mo

zam

biq

ue

Se

yche

lles

Alg

eria

Ug

an

da

Ma

uritiu

s

Tan

zan

ia

Lib

ya

Eq

ua

toria

l G

uin

ea

Zam

bia

Bu

rund

i

Eg

yp

t

Gu

ine

a B

issau

Nig

eria

Mo

rocco

Co

ng

o

Eth

iopia

Co

mo

ros

Rw

and

a

Ma

urita

niaC

han

ge

in

co

mp

etitive

ne

ss

SADC markets Global markets Intra-African markets

117

Figure A4.2a. Change in commodity competitiveness in regional exports markets compared to global and intra-

African markets (1998-2013)

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from commodity-level export share decomposition analysis for African countries as a group.

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

So

ybe

an

s

Org

an

ic c

hem

ica

ls

Wh

eat

Co

tto

n, n

ot ca

rded

or

co

mb

ed

Co

ffe

e

So

rgh

um

Oth

er

oils

eed

s

Alb

um

inoid

al su

bsta

nce

s

Hid

es &

skin

s

Po

tato

es

Ve

ge

tab

le p

laitin

g m

ate

rials

Oth

er

ce

reals

Esse

ntia

l o

ils &

re

sin

oid

s

Ed

ible

fru

its &

nuts

Co

coa

pre

pa

ratio

ns

Ma

ize

Me

at &

edib

le o

ffa

l

Su

ga

r co

nfe

ction

ery

Fin

ish

ing a

ge

nts

fo

r te

xtile

s &

pap

er

Co

tto

n, ca

rded

or

co

mb

ed

Ca

ne

su

ga

r

Gro

un

dn

ut o

il

Tea

Tom

ato

es

Co

coa

be

ans

Ed

ible

pre

ps. o

f m

eat, fis

h &

cru

sta

ce

ans

Me

dic

inal p

lants

Po

ultry

Wo

ol

On

ions &

su

bstitu

tes

Pre

ps. o

f ve

gs., fru

its &

nuts

Mis

c. e

dib

le p

repa

ratio

ns

Oth

er

live

tre

es &

pla

nts

Fis

h &

se

a fo

od

s

Oth

er

anim

al p

rod

ucts

Be

vera

ge

s, sp

irits &

vin

eg

ar

Oth

er

live

an

ima

ls

Pre

ps. o

f ce

rea

ls, flo

ur,

sta

rch

or

milk

Ric

e

Ro

ots

& tu

be

rs

Tob

acco

& s

ubstitu

tes

Sp

ice

s

Sh

ee

p &

goa

ts

Mill

ing in

du

str

y p

rodu

cts

Oth

er

oils

& fa

cts

Pa

lm o

il

Furs

kin

s

Da

iry, e

gg

s &

hon

ey

Gu

ms &

re

sin

s

Gro

un

dn

uts

Re

sid

ue

s fro

m fo

od

in

du

str

ies

Oth

er

edib

le v

ege

table

s

Ca

ttle

Rye

, b

arle

y &

oats

An

ima

l fa

ts

So

ybe

an

oil

Oth

er

ve

ge

table

te

xtile

fib

res

Oliv

e o

il

Silk

Ch

an

ge

in

co

mp

etitive

ne

ss

COMESA markets Global markets Intra-African markets

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

So

rgh

um

Tea

Furs

kin

s

Co

tto

n, n

ot ca

rded

or

co

mb

ed

So

ybe

an

s

Org

an

ic c

hem

ica

ls

Oth

er

ve

ge

table

te

xtile

fib

res

Co

tto

n, ca

rded

or

co

mb

ed

Me

at &

edib

le o

ffa

l

Pa

lm o

il

Co

ffe

e

Gro

un

dn

ut o

il

Me

dic

inal p

lants

Be

vera

ge

s, sp

irits &

vin

eg

ar

Oth

er

live

an

ima

ls

Po

ultry

Re

sid

ue

s fro

m fo

od

in

du

str

ies

Da

iry, e

gg

s &

hon

ey

Gu

ms &

re

sin

s

Fis

h &

se

a fo

od

s

Oth

er

anim

al p

rod

ucts

Oth

er

edib

le v

ege

table

s

Su

ga

r co

nfe

ction

ery

Gro

un

dn

uts

Tom

ato

es

Esse

ntia

l o

ils &

re

sin

oid

s

Ca

ttle

Oth

er

live

tre

es &

pla

nts

Oth

er

oils

eed

s

Pre

ps. o

f ce

rea

ls, flo

ur,

sta

rch

or

milk

Wh

eat

Mis

c. e

dib

le p

repa

ratio

ns

Wo

ol

Ed

ible

fru

its &

nuts

Oth

er

ce

reals

Ca

ne

su

ga

r

Sh

ee

p &

goa

ts

Co

coa

pre

pa

ratio

ns

Ed

ible

pre

ps. o

f m

eat, fis

h &

cru

sta

ce

ans

Po

tato

es

Ric

e

Mill

ing in

du

str

y p

rodu

cts

Tob

acco

& s

ubstitu

tes

Hid

es &

skin

s

On

ions &

su

bstitu

tes

Ma

ize

Pre

ps. o

f ve

gs., fru

its &

nuts

Rye

, b

arle

y &

oats

Ve

ge

tab

le p

laitin

g m

ate

rials

Sp

ice

s

Co

coa

be

ans

Oliv

e o

il

Fin

ish

ing a

ge

nts

fo

r te

xtile

s &

pap

er

Alb

um

inoid

al su

bsta

nce

s

Oth

er

oils

& fa

cts

Ro

ots

& tu

be

rs

An

ima

l fa

ts

Silk

So

ybe

an

oil

Ch

an

ge

in

co

mp

etitive

ne

ss

ECCAS markets Global markets Intra-African markets

118

Figure A4.2b. Change in commodity competitiveness in regional exports markets compared to global and

intra-African markets (1998-2013): commodity ranking

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from commodity-level export share decomposition analysis for African countries as a group.

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

Org

anic

chem

icals

Rye, barley &

oats

Onio

ns &

substitu

tes

Cotton, card

ed o

r com

bed

Hid

es &

skin

s

Palm

oil

Vegeta

ble

pla

itin

g m

ate

rials

Fin

ishin

g a

gents

for

textile

s &

paper

Oth

er

live tre

es &

pla

nts

Gro

undnut oil

Roots

& tubers

Resid

ues fro

m food industr

ies

Tea

Cotton, not card

ed o

r com

bed

Sorg

hum

Sugar

confe

ctionery

Maiz

e

Essential oils

& r

esin

oid

s

Cane s

ugar

Spic

es

Medic

inal pla

nts

Oth

er

edib

le v

egeta

ble

s

Silk

Sheep &

goats

Cattle

Bevera

ges, spirits &

vin

egar

Meat &

edib

le o

ffal

Pre

ps. of cere

als

, flour,

sta

rch o

r m

ilk

Cocoa b

eans

Anim

al fa

ts

Oth

er

live a

nim

als

Gro

undnuts

Edib

le fru

its &

nuts

Oth

er

cere

als

Mis

c. edib

le p

repara

tions

Ric

e

Gum

s &

resin

s

Fis

h &

sea foods

Oth

er

anim

al pro

ducts

Poultry

Alb

um

inoid

al substa

nces

Dairy, eggs &

honey

Coffee

Oth

er

oils

& facts

Oth

er

oils

eeds

Cocoa p

repara

tions

Mill

ing industr

y p

roducts

Furs

kin

s

Soybeans

Pota

toes

Pre

ps. of vegs., fru

its &

nuts

Tom

ato

es

Wool

Edib

le p

reps. of m

eat, fis

h &

cru

sta

ceans

Tobacco &

substitu

tes

Oth

er

vegeta

ble

textile

fib

res

Wheat

Soybean o

il

Oliv

e o

il

Chang

e in c

om

petitiveness

ECOWAS markets Global markets Intra-African markets

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

Wool

Org

anic

chem

icals

Meat &

edib

le o

ffal

Wheat

Silk

Cocoa b

eans

Essential oils

& r

esin

oid

s

Bevera

ges, spirits &

vin

egar

Oth

er

cere

als

Oth

er

oils

eeds

Dairy, eggs &

honey

Cane s

ugar

Sugar

confe

ctionery

Cattle

Cocoa p

repara

tions

Mis

c. edib

le p

repara

tions

Poultry

Spic

es

Edib

le p

reps. of m

eat, fis

h &

cru

sta

ceans

Pre

ps. of vegs., fru

its &

nuts

Gro

undnut oil

Tobacco &

substitu

tes

Tea

Medic

inal pla

nts

Cotton, not card

ed o

r com

bed

Tom

ato

es

Coffee

Pota

toes

Resid

ues fro

m food industr

ies

Oth

er

live tre

es &

pla

nts

Mill

ing industr

y p

roducts

Edib

le fru

its &

nuts

Sorg

hum

Fin

ishin

g a

gents

for

textile

s &

paper

Alb

um

inoid

al substa

nces

Pre

ps. of cere

als

, flour,

sta

rch o

r m

ilk

Cotton, card

ed o

r com

bed

Roots

& tubers

Oth

er

edib

le v

egeta

ble

s

Oth

er

live a

nim

als

Gro

undnuts

Onio

ns &

substitu

tes

Oth

er

vegeta

ble

textile

fib

res

Palm

oil

Oth

er

anim

al pro

ducts

Maiz

e

Sheep &

goats

Soybeans

Oth

er

oils

& facts

Fis

h &

sea foods

Anim

al fa

ts

Soybean o

il

Ric

e

Gum

s &

resin

s

Furs

kin

s

Vegeta

ble

pla

itin

g m

ate

rials

Oliv

e o

il

Hid

es &

skin

s

Rye, barley &

oats

Chang

e in c

om

petitiveness

SADC markets Global markets Intra-African markets

119

4.10. Statistical tests

The series of competitive effect values derived for all countries and all commodities and for

different destination markets are used to carry out two statistical comparison procedures. The first

one is an analysis of variance (ANOVA), which is used to test the hypothesis that the means of

competitiveness changes are equal across country groups. The second one is the paired-samples T

test of the hypothesis that competitiveness changes in two export destination markets are equal.

This is run both for country and commodity competitiveness changes. The results obtained from

these procedures are presented in Tables 4.1 - 4.4 above as well as Tables A4.5 – A4.8 below and

are discussed in sections 4.3 - 4.5.

Prior to running these procedures, the one-sample Kolmogorov-Smirnov test was first performed

to confirm the assumption of the normality of the distribution of competitiveness change indices

in each of the country groups under comparison. The same test was carried out the check the

assumption that for each pair of export markets the differences in competitiveness changes in those

markets follow a normal distribution. We also used the Levene's homogeneity-of-variance test to

check the assumption that country groups under comparison come from populations with equal

variances. In the large majority of comparisons, the Levene’s test confirmed an equality of

variances across groups, allowing us to perform an ANOVA procedure using the standard F

statistic. However, in the rare comparisons where variances are significantly different, a robust

ANOVA procedure using the Welch statistic was also performed to check whether we can trust

the p value associated with the standard ANOVA F statistic. The results of the Kolmogorov-

Smirnov test and the Levene's test are presented in Table A4.9 – A4.12.

Table A4.5. Analysis of variance of country competitiveness changes in COMESA agricultural export markets (1998-2013)

Country Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.187 1 0.187 11.970 0.001 0.206

non-COMESA Within Groups 0.720 46 0.016

countries Total 0.907 47

ECCAS vs. Between Groups 0.071 1 0.071 3.904 0.054 0.078

non-ECCAS Within Groups 0.836 46 0.018

countries Total 0.907 47

ECOWAS vs. Between Groups 0.014 1 0.014 0.697 0.408 0.015

non-ECOWAS Within Groups 0.893 46 0.019

countries Total 0.907 47

SADC vs. Between Groups 0.000 1 0.000 0.013 0.909 0.000

non-SADC Within Groups 0.907 46 0.020

countries Total 0.907 47

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

120

Table A4.6. Analysis of variance of country competitiveness changes in ECCAS agricultural export markets (1998-2013)

Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.003 1 0.003 0.182 0.672 0.004

non-COMESA Within Groups 0.629 44 0.014

countries Total 0.631 45

ECCAS vs. Between Groups 0.057 1 0.057 4.346 0.043 0.090

non-ECCAS Within Groups 0.574 44 0.013

countries Total 0.631 45

ECOWAS vs. Between Groups 0.006 1 0.006 0.389 0.536 0.009

non-ECOWAS Within Groups 0.626 44 0.014

countries Total 0.631 45

SADC vs. Between Groups 0.010 1 0.010 0.737 0.395 0.016

non-SADC Within Groups 0.621 44 0.014

countries Total 0.631 45

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from export share decomposition analysis for individual countries.

Table A4.7. Analysis of variance of country competitiveness changes in ECOWAS agricultural export markets (1998-2013)

Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.013 1 0.013 0.978 0.328 0.020

non-COMESA Within Groups 0.652 48 0.014

countries Total 0.665 49

ECCAS vs. Between Groups 0.002 1 0.002 0.164 0.687 0.003

non-ECCAS Within Groups 0.663 48 0.014

countries Total 0.665 49

ECOWAS vs. Between Groups 0.025 1 0.025 1.908 0.174 0.038

non-ECOWAS Within Groups 0.640 48 0.013

countries Total 0.665 49

SADC vs. Between Groups 0.003 1 0.003 0.186 0.668 0.004

non-SADC Within Groups 0.663 48 0.014

countries Total 0.665 49

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from export share decomposition analysis for individual countries.

121

Table A4.8. Analysis of variance of country competitiveness changes in SADC agricultural export markets (1998-2013)

Groups Sum of Squares df Mean Square F Sig. Eta Squared

COMESA vs. Between Groups 0.053 1 0.053 4.369 0.042 0.083

non-COMESA Within Groups 0.579 48 0.012

countries Total 0.632 49

ECCAS vs. Between Groups 0.001 1 0.001 0.077 0.782 0.002

non-ECCAS Within Groups 0.631 48 0.013

countries Total 0.632 49

ECOWAS vs. Between Groups 0.092 1 0.092 8.184 0.006 0.146

non-ECOWAS Within Groups 0.540 48 0.011

countries Total 0.632 49

SADC vs. Between Groups 0.008 1 0.008 0.612 0.438 0.013

non-SADC Within Groups 0.624 48 0.013

countries Total 0.632 49

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from export share decomposition analysis for individual countries.

122

Table A4.9. One-Sample Kolmogorov-Smirnov tests of normality of the distributions of competitiveness

changes for different country groups

Test groups

Export destination markets

Global

markets

Intra-African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

COMESA

countries

Kolmogorov-

Smirnov Z 1.039 0.793 0.506 0.756 0.536 0.695

Asymp. Sig.

(2-tailed) 0.231 0.555 0.960 0.617 0.937 0.720

Non-

COMESA

countries

Kolmogorov-

Smirnov Z 0.672 0.531 0.887 0.542 0.450 0.435

Asymp. Sig.

(2-tailed) 0.757 0.940 0.412 0.931 0.987 0.991

ECCAS

countries

Kolmogorov-

Smirnov Z 0.624 0.378 0.621 0.456 0.483 0.752

Asymp. Sig.

(2-tailed) 0.831 0.999 0.835 0.985 0.974 0.625

Non-ECCAS

countries

Kolmogorov-

Smirnov Z 0.892 0.970 0.837 0.664 0.568 0.744

Asymp. Sig.

(2-tailed) 0.404 0.303 0.486 0.770 0.904 0.638

ECOWAS

countries

Kolmogorov-

Smirnov Z 0.514 0.433 0.708 0.463 0.650 0.463

Asymp. Sig.

(2-tailed) 0.954 0.992 0.698 0.983 0.792 0.983

Non-

ECOWAS

countries

Kolmogorov-

Smirnov Z 0.775 0.752 0.752 0.751 0.421 0.775

Asymp. Sig.

(2-tailed) 0.585 0.623 0.624 0.626 0.994 0.586

SADC

countries

Kolmogorov-

Smirnov Z 0.414 0.888 0.729 0.620 0.883 0.576

Asymp. Sig.

(2-tailed) 0.995 0.410 0.663 0.836 0.416 0.894

Non-SADC

countries

Kolmogorov-

Smirnov Z 0.717 0.771 0.715 0.800 0.831 0.736

Asymp. Sig.

(2-tailed) 0.683 0.591 0.685 0.544 0.495 0.651

Note: The probability of the Z statistic is above 0.05, meaning that the normal distribution is a good fit for

competitiveness changes for the different country groups tested and across all export destinations. Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of the

competitive effect derived from export share decomposition analysis for individual countries.

123

Table A4.10. One-Sample Kolmogorov-Smirnov tests of normality of the distributions of differences in

country competitiveness changes in pairs of export markets

Pairs of markets N Kolmogorov-Smirnov Z Asymp. Sig. (2-tailed)

COMESA & global markets 48 0.973 0.300

ECCAS & global markets 46 0.796 0.551

ECOWAS & global markets 50 0.722 0.675

SADC & global markets 50 0.759 0.612

Intra-African & global markets 50 0.593 0.874

COMESA & intra-African markets 48 0.747 0.632

ECCAS & intra-African markets 46 0.899 0.394

ECOWAS & intra-African markets 50 0.824 0.505

SADC & intra-African markets 50 0.936 0.345

Note: The probability of the Z statistic is above 0.05, meaning that the normal distribution is a good fit the

differences of competitiveness changes in pairs of export destination markets.

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

Table A4.11. One-Sample Kolmogorov-Smirnov tests of normality of the distributions of differences in

commodity competitiveness changes in pairs of export markets

Pairs of markets N Kolmogorov-Smirnov Z Asymp. Sig. (2-tailed)

COMESA and global markets 59 0.626 0.828

ECCAS and global markets 59 1.023 0.246

ECOWAS and global markets 59 0.665 0.769

SADC and global markets 59 1.058 0.213

Intra-African and global markets 59 0.780 0.577

COMESA and intra-African markets 59 1.051 0.219

ECCAS and intra-African markets 59 0.747 0.631

ECOWAS and intra-African markets 59 1.073 0.200

SADC and intra-African markets 59 0.792 0.557

Note: The probability of the Z statistic is above 0.05, meaning that the normal distribution is a good fit the

differences of competitiveness changes in pairs of export destination markets.

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from commodity-level export share decomposition analysis for African countries as a

group.

124

Table A4.12. Levene's test for homogeneity-of-variance of country competitiveness changes for pairs of country

groups

Country groups

Export destination markets

Global

markets

Intra-African

markets

COMESA

markets

ECCAS

markets

ECOWAS

markets

SADC

markets

COMESA vs.

non-COMESA

countries

Levene

Statistic 0.834 0.543 4.551 0.201 0.000 0.897

Sig. 0.366 0.465 0.038* 0.656 0.994 0.348

ECCAS vs.

non-ECCAS

countries

Levene

Statistic 0.127 0.034 2.926 0.900 2.294 0.247

Sig. 0.723 0.854 0.094* 0.348 0.136 0.621

ECOWAS vs.

non-ECOWAS

countries

Levene

Statistic 0.044 1.042 0.060 0.019 0.069 0.655

Sig. 0.834 0.312 0.807 0.890 0.793 0.422

SADC vs.

non-SADC

countries

Levene

Statistic 1.370 9.432 1.710 0.006 4.206 6.343

Sig. 0.247 0.004* 0.198 0.939 0.046* 0.015*

Note: In the large majority of tests the significance value of the Levene statistic is above 0.10, which means that

we can assume an equality of variances for corresponding pairs of country-groups. The asterisk denotes a few

tests resulting in significance values below 0.10, meaning that the assumption of equal variances is violated for

corresponding pairs of groups.

Source: Authors’ calculations using the BACI database. Change in competitiveness is measured by the coefficient of

the competitive effect derived from export share decomposition analysis for individual countries.

Chapter 5. Determinants of African agricultural exports

Extracted from

African Agricultural Trade Status Report

2017

125

CHAPTER 5. DETERMINANTS OF AFRICAN AGRICULTURAL EXPORTS

Getaw Tadesse, International Food Policy Research Institute (IFPRI), Eastern and Southern

Africa Office, Addis Ababa, Ethiopia

Ousmane Badiane, International Food Policy Research Institute, Washington DC

5.1 Introduction

Trade is an important engine for economic growth, food security, reducing poverty and overall

development. However, it is a complex and sensitive subject for policymaking as it involves

negotiations, dialogues and agreements between partner countries residing in different socio-

political boundaries. It becomes more complicated when linked with agriculture, which is a sector

profoundly reliant on continuous social and ecological dynamism. Therefore, success in

agricultural trade heavily depends on the extent of understanding of the constraints facing

agriculture and its cross-broader trade.

Following the 1980s trade liberalizations, a series of studies have been conducted to document

agricultural trade trends, determinants and prospects both in Africa and elsewhere (Bouët, Bureau,

Decreux, & Jean, 2005; Bouët, Mishra, & Roy, 2008; Bureau, Jean, & Matthews, 2006; Croser &

Anderson, 2011; Moïsé, Delpeuch, Sorescu, Bottini, & Foch, 2013). These studies highlighted a

wide array of constraints that are crucially important for improving African agricultural trade.

More importantly they have indicated the importance of global trade policy actions and the need

to address the different trade constraints in a holistic manner. According to these studies,

agricultural trade determinants can be broadly classified into five major thematic areas, namely

production capacity, cost of trade, trade policies, domestic agricultural supports and global

market shocks. While production capacity and cost of trade are usually referred to as supply side

constraints, many trade policies (except export taxes) and agricultural supports in importing

countries are considered to be demand side constraints. Constraints related to global food, oil and

financial crises are taken as market level trade constraints. These constraints influence imports and

exports in different ways and to different extents both from the demand and supply sides.

Supply-side determinants limit the competitiveness of a country in global or regional markets by

increasing costs of production as well as costs of trading. These constraints include the nature and

extent of resource endowments, productivity (technology), quality of infrastructure and institutions

126

that facilitate trade, and domestic agricultural support services provided to smallholder producers

and traders in an exporting country. Demand side constraints usually emerge from trade protection

measures of importing countries. Africa exports more than 75 percent of its agricultural product

value outside of the continent. Many of its trade partners impose several trade protection measures

which directly or indirectly limit agricultural exports. This is particularly the case for certain

commodities such as tobacco, cotton, coffee, cocoa, and oilseeds, in which Africa has the

comparative advantage. Therefore, close monitoring of the extent and nature of these constraints

and their linkages with the flow of agricultural exports is required to guide effective evidence-

based trade policymaking in Africa.

The purpose of this chapter is to offer comprehensive and updated evidence to African agricultural

trade policy discussions through highlighting determinants that hinder the performance and

competitiveness of agricultural exports and underlining areas that should receive priority policy

attention at the continental, regional and national levels. Africa aspires to triple the current level

of regional agricultural trade by the year 2025, which requires a wide range of interventions in the

form of policies and investments. For these interventions to be effective and achieve the intended

targets, key areas of intervention have to be identified, prioritized and monitored regularly. In this

chapter, we attempt to review existing evidence, identify key determinants of trade in general, and

describe how these determinants are specifically important to African agricultural trade. In doing

so, we provide empirical evidence that shows the relative importance of trade constraints and

explains how the constraints are trending over time and varying across countries.

The chapter is structured as follows. The next section briefly reviews specific factors included in

each of the five major determinants of trade and their conceptual and empirical links with trade.

Following this section, the empirical assessment approach used to estimate the relative importance

of trade determinants is presented. This section explains the sources of data used, the variables

selected, and the overall results of gravity models estimated for global-Africa and intra-Africa

bilateral export trade. The subsequent section describes, discusses and tracks the major

determinants included in the gravity models. In this section, we discuss the significance of the

determinants, their magnitude and trends, and the conditions under which a factor becomes

detrimental. The last section summarizes major findings and draws conclusions that would help

policy dialogue and actions.

127

5.2 Review of trade determinants

The extent of agricultural exports has been constrained by many domestic and international factors

both from the demand and supply sides. Theoretical and empirical evidence suggests that these

factors can be broadly classified into five major thematic areas including production capacity, cost

of trade, trade policies, domestic agricultural supports and global market shocks. These

constraints influence imports and exports in different ways and at different magnitudes.

Production capacity refers to those factors that affect the production capacity of a country. These

factors include resource endowments and other technological and institutional factors that enhance

the productivity and comparative advantages of a country in global and regional markets. Both

classical and neoclassical theories have exhaustively explained the importance of comparative

advantage for improving performance of trade among countries. However, there has been strong

contention regarding the source of this production capacity and thereby the source of comparative

advantage. While the Ricardian hypothesis advocates the importance of technological (or

productivity) change as the major source of comparative advantage, the Heckscher-Ohlin

hypothesis argues for the importance of relative factor endowments as a prime source of trade.

According to the Ricardian theory, the relative efficiency of producing goods and services

determines the direction and magnitude of trade between two countries. In contrast, the Heckscher-

Ohlin factor endowment theory predicts that countries with an abundance of one or more of the

factors of production (land, labor and capital) will specialize in commodities that require much of

the abundant resources. However, empirical studies have confirmed that differences in

productivity (technology) and factor endowment explain a very small part of trade performance

variations over time and across countries (Bergstrand, 1990; Bernstein & Weinstein, 2002).

Moreover, recent evidence has suggested the importance of relative factor endowment over

productivity or technology to explain international trade (Amoroso, Chiquiar, & Ramos-Francia,

2011).

Cost of trade: factors that exacerbate costs of trade are very diverse. The two most important

factors that increase the cost of trade are poor infrastructure and institutional inefficiency related

to trade. Costs also include financial fees related to export and imports.

The role of infrastructure in enhancing trade has been widely discussed in policy circles and in the

literature (Bouët et al., 2008; Bougheas, Demetriades, & Mamuneas, 1999; Francois & Manchin,

128

2007; Moïsé et al., 2013). Empirical studies have generally confirmed positive and significant

effects of infrastructure quality in exporting countries on trade values. However, the relative

importance of infrastructural elements varies across studies. While road density has significant

positive effects on trade volumes of low income countries, the effect of mobile phone density has

been found to be less significant (Bouët et al., 2008).

Institutional efficiency refers to the ease of doing business in relation to agricultural imports and

exports. It includes procedures and delays in customs clearing, access to finance for traders, and

the strength of contractual enforcement. Although customs and administrative procedures are

essential for facilitating trade and implementing trade policies, they have the potential to restrict

trade, particularly in less developed countries where administrative systems are less automated,

capacitated and transparent. These procedures and requirements delay delivery and cause extra

costs related to storage costs and losses. Empirical studies have indicated that a 10 percent

reduction in the time spent to clear exports, the number of signatures required to clear exports, or

the number of documents needed to cross borders increases trade by 6 to 11 percent globally

(Wilson, 2007). Trade is more responsive to the number of documents than to the other metrics.

Trade policies include measures aimed at protecting trade through tariffs and non-tariff barriers.

The effect of tariffs on trade performance has been studied using economy-wide simulations (e.g.

Bouët, Bureau, et al., 2005), gravity equations (e.g. Bouët et al., 2008), and trade restrictiveness

indexes (e.g.Croser & Anderson, 2011). Although the magnitudes are different, all of the studies

indicated that the effect of import taxes on trade volumes is convincingly negative and significant.

Bilateral, regional and international trade agreements are also part of tariff policies that either

reduce tariffs through Free Trade Agreements (FTA) or facilitate cross border trade. The most

important of these agreements are trade preferences, particularly the non-reciprocal ones which

target opening markets to individual or sets of developing countries. This involves complete or

partial lifting of import tariffs and quotas for specified products. Preferences are usually designed

to offer commercial opportunities for poor countries. However, preferences are widely criticized

for not being utilized due to rules of origin, their focus on commodities for which developing

countries have little competitive advantage, and the presence of associated stringent standards

related to sanitary and phytosanitary requirements (Brenton, 2003; Panagariya, 2003; Topp, 2003).

Despite these critics, some recent studies have shown that preferences are still useful and beneficial

129

to less developed countries, particularly to countries in Africa south of the Sahara (Bouët,

Fontagné, & Jean, 2005; Bouët, Laborde, Dienesch, & Elliott, 2012; Wainio & Gehlhar, 2004).

Non-tariff measures include those trade barriers that limit the quantity and volume of imports

through a variety of technical and non-technical standards. UNCTAD classifies non-tariff trade

measures into sixteen broad categories, each of which constitutes several specific classifications.

The major ones are sanitary and phytosanitary (SPS) requirements, technical barriers to trade

(TBT) which include packing, labeling and standardizing, price controls (anti-dumping), licensing,

quantitative restrictions, export subsidies and export taxes. Non-tariff barriers constrain trade

through increasing the cost of inspection, certification and testing. This is particularly important

for developing countries which have poor quality assurance infrastructure and technological

capacity to conduct these processes and hence have to recruit third parties to access the services.

Domestic agricultural supports: Both developed and developing countries provide financial and

technical support to their agricultural producers for different reasons. However, the support

provided by industrial countries to protect their agricultural sectors has been considered to be the

most damaging for trade from developing countries. Supports in these countries take the form of

border measures (import tariffs, export subsidies) and domestic measures (production and input

subsidies). Domestic supports can be implemented through markets or through direct payments.

Both approaches have the potential to reduce the amount of imports from foreign countries. These

supports raise the price received by the producers of the supported country above the world price

so that they become artificially more competitive than imports from outside of the country.

Empirical studies assessing the link between domestic subsides and trade have revealed mixed

results depending on the type of support (coupled or decoupled) and commodity. Many have

argued that the removal of EU and US agricultural subsidies could have a significant effect on

world prices of some commodities such as cotton, tobacco and soybean (Bouët, Bureau, et al.,

2005; Bureau et al., 2006). However, the impact of domestic subsidies is lower than other cross-

border measures (Anderson & Martin, 2005; Hoekman, Ng, & Olarreaga, 2004).

Payments less related to the quantity produced (decoupled) have lesser impacts than payments

directly related to production (coupled); as a result many OECD countries are moving towards

payments which are less tied to the quantity of domestic production (Urban, Jensen, & Brockmeier,

2016).

130

Developing countries do also provide technical, financial and institutional support to smallholder

producers to boost productivity and improve market efficiency, thereby enhancing agricultural

exports. The extent of agricultural support provided to smallholders depends on the size, allocation

and efficiency of public agricultural expenditure. Agricultural public expenditure serves to

accumulate capital stock that would enhance the production as well as trading capacity of

smallholder producers (Benin, Mogues, & Fan, 2012). However, the actual effect on trade depends

on the focus and efficiency of public investments. Investments focused on export sectors would

likely improve trade more than those investments focused on domestic food production or food

security.

Global market shocks: Global food, financial and oil markets are increasingly interconnected

(Tadesse, Algieri, Kalkuhl, & Braun, 2014). Shocks to any of these markets would likely affect

the nature and extent of agricultural trade. The 2007/2008 food price crisis, for example, has caused

many countries to impose export barriers and relax import restrictions on food products, which has

further aggravated the problem of price spikes and adversely affected agricultural trade (Anderson,

2014; Anderson & Nelgen, 2012; Anderson & Thennakoon, 2015; Bouet & Laborde, 2012; Yu,

Tokgoz, Wailes, & Chavez, 2011). Similarly, the ongoing oil price crises may also affect the extent

of agricultural exports, particularly in those countries which are oil dependent. When the oil price

is declining, oil dependent countries would likely attempt to shift export dependence from oil to

agricultural products, for which prices are relatively stable.

5.3 Empirical assessment

5.3.1Data and methods

We used gravity-type econometric equations to examine the empirical and relative relevance of

the determinants listed above in the African context. The models are used to estimate the logarithm

of bilateral agricultural export values of African countries over a number of demand and supply

side factors. In addition to the four14 major thematic determinants explained above, scale variables

are included to control for the size of importing and exporting economies and income differences

between trading partners. Two to five specific variables were chosen to proxy each of the major

thematic determinants. Total GDP of both importing and exporting countries are used to proxy the

14 Variables to represent the fifth thematic determinant, global market shocks, are not considered due to their

invariability across countries. These variables can be captured in a time-series setting.

131

size of the economies of partnering countries. While GDP per capita in importing countries is used

to capture income effects, GDP per capita in exporting countries is used as a proxy for capital

endowment. Other assets such as farm machinery, irrigation facilities, etc., would have been a

good indicator of capital for agriculture, but the data on these variables suffers from a large number

of missing values. Quantity of land and labor are included to measure resource endowments; road

density, quality of port, index of trade infrastructural quality, index of customs clearing efficiency

and financial fees for exporting are used to measure costs of trade; frequency of non-tariff

measures, average ad valorem equivalent tariff rates and regional trade agreements are considered

to proxy external trade policy; and the ratio of the agricultural producer price index to the

manufacturing producer price index of importing countries and agricultural public expenditure of

exporting countries are used to measure the effect of domestic agricultural policy in importing and

exporting countries respectively. The list of determinants considered in the analysis and the metrics

used to estimate their magnitudes are described in Annex 1.

Data used in this analysis are obtained from different sources, mainly from World Bank World

Development Indicators (WDI), UN Comtrade, and World Integrated Trade Solution (WITS).

While data on income, resource endowments, infrastructure and efficiency of institutions are

gathered from World Bank WDI, UN Comtrade is used for trade data, and data on tariffs were

extracted from WITS. Other sources such as WTO, ReSAKSS, FAOSTAT, and OECD are used

for data on specific variables such as non-tariff barriers, public agricultural expenditure, producer

price indices and producer support estimates (PSE) respectively. The quality of trade data in Africa

has always been a big concern as sizable cross-border transactions are carried out informally and

unrecorded. However, the purpose of this chapter is not to show the size of trade, but rather to look

into the determinants of export flows. Thus, as long as the omitted trade transactions are random,

they will have little impact on our results. All export values are for agricultural products unless

and otherwise mentioned.

All the regressions are estimated using cross sectional data from 2013, which is the most recent

year for which adequate data are available for many of the determinants. However, one year lagged

values are used for some variables (productivity and public agricultural expenditure) which are

deemed to be endogenous to export values. Visualization of trade data over years indicates that

there were no extraordinary events in 2013 that could bias the results.

132

Two groups of models are estimated. The first group is used to estimate African agricultural

exports to the global market. In this models, only African countries are included as exporters ( i ).

In addition to African countries, countries from all continents which had frequent transactions with

Africa are included as importers ( j ). In general, a total of 49 exporters15 and 161 trade partners

are considered. The second group of models is used to estimate intra-African exports, with African

countries as both exporters and importers. We also estimated African exports to the rest of the

world for comparison purposes.

Of all possible pairwise transactions between 49 exporting countries and 161 importing countries,

about 58 percent have zero trade transactions. Excluding these transactions would likely cause

selection bias, while inclusion of them would cause censoring bias. Though previous studies have

excluded them and tried to control the selection bias using the Heckman approach, we choose to

include them in the analysis and address the censoring bias using a Tobit model approach. We

assume zero trade is an optimal outcome instead of a strategic choice of a country not to trade with

a specific partner.

Due to multiple data sources for different variables, the dataset is seriously affected by missing

values. To overcome the problem of missing values, several specifications are considered through

step-wise inclusion of explanatory variables, which have different sets of observations and

represent specific sets of determinants. A total of six specifications are estimated for African global

exports.

The first model estimates the effect of resource endowments together with scale variables. The

second model includes infrastructural and institutional variables in addition to the variables in

model one. The third model adds public agricultural expenditure and hence represents a domestic

trade model in which only domestic (supply side) constraints are included. The fourth model

includes international (demand side) variables such as non-tariff barriers, tariffs and regional trade

agreements. The fifth and sixth models are Tobit specifications without and with the agriculture-

to-manufacturing price ratio variable that represents domestic agricultural supports by OECD

15 Five southern Africa countries (Lesotho, South Africa, Botswana, Namibia and Swaziland) are treated as one

country as they have a common customs union called SACU. Trade data in many sources is reported for the five

countries together; for other variables we use the average or the sum of all or some of the countries, depending on the

variable.

133

countries. Since the price ratio is calculated only for OECD countries, the number of observations

is greatly reduced in the final specification.

5.3.2 Empirical Results

Table 5.1 shows results of the six specifications for African global agricultural exports. The

columns, denoted by the numbers 1 to 6, present the results of different specifications that could

help to test robustness under different numbers of observations and examine the predictive power

of additional variables. In general, many determinants show the theoretically expected signs,

except resource endowment variables. Variables related to infrastructure and institutional

efficiency are more significant than other domestic factors. These variables explain about 11

percent of the variation in agricultural export growth among African countries. Public expenditure

in agriculture appears to have positive and generally significant effect on trade. Trade policy

variables appear to be important determinants, next to the cost of trade, though there exists

significant variation between policy instruments. Non-tariff barriers and regional trade agreements

appear more important than tariffs. Resource endowment seems to be a less important factor for

African agricultural trade. The effect of producer price ratios which represent domestic agricultural

support in importing countries seems significant, but requires further explanation.

Table 5.2 shows results of intra-Africa trade determinants in comparison with African exports to

the rest of the world. In this case, we used the comprehensive models (four and five), as agriculture-

to-manufacturing price ratios are not available for most African countries. The results indicate that

many of the determinants are equally important for African exports either within Africa or outside

of Africa. The level of per capita income in importing countries is more relevant for intra-African

trade than for African exports to the rest of the world. Similarly, resource endowments and non-

tariff barriers are not as relevant for intra-African trade as they are for African trade with countries

in other regions. This is consistent with the facts that resource endowments within Africa are

closely similar and non-tariff barriers are not stringent as they are outside of Africa. We also learn

that public expenditures in agriculture are more relevant to reach markets outside of Africa than

markets within Africa.

Since the determinants for intra-African and global African exports are similar, in the subsequent

section we discuss why some variables are significant over the others, and track trends and

134

distributions of key determinants using the results of the global-Africa agricultural export

estimations. However, we briefly discuss the importance of a determinant for intra-Africa trade

whenever necessary.

135

Table 5.1. Response of African global agricultural export value to domestic and international

factors

Determinants

Logarithm of value of exports from i countries to j countries

OLS Tobit

(1) (2) (3) (4) (5) (6)

Importer’s GDP (billions of US$) 1.57*** 1.65*** 2.16*** 2.23*** 3.35*** 2.70***

Exporter’s GDP (billions of US$) 0.79*** 0.88*** 0.92*** 1.19*** 1.80*** 1.48***

Per capita GDP of exporters (US$) -

1.14*** -1.17*** -2.11*** -2.30*** -3.63***

-

2.67***

Per capita GDP of importers(US$) -

0.10*** -0.12*** -0.13*** 0.03 -0.04 -0.21

Arable land (millions of hectares) -

0.52*** -0.69*** -0.52*** -0.47*** -0.52***

-

0.91***

Agricultural labor (millions) -0.02 0.25*** -0.38** -0.43** -0.77*** 0.05

Road density (km per km2 of land) 0.01 -0.03 -0.02 0.03 0.37***

Quality of port 4.43*** 4.26*** 4.62*** 6.94*** 8.63***

Quality of transport infrastructure 1.80*** 1.17** 1.15** 0.82 1.47

Efficiency of customs clearing index 1.24*** 1.64*** 1.69*** 3.81*** 0.03

Export cost ($US per container) -0.05 -0.07 -0.01 -0.27 -0.13

PAE per agricultural GDP of exporter 0.12** 0.16** 0.46*** 0.28*

Incidence of importer’s non-tariff

barriers -0.32*** -0.39***

-

0.32***

Average tariff rate of importer -0.06 -0.18* -

0.46***

Being in a similar REC 3.52*** 5.39*** 5.24***

The ratio of agricultural PPI to

manufacturing PPI

-

5.96***

Constant 5.44*** -2.43* 3.30* 1.66 0.9 -1.44

Sigma (test for censoring) 4.32*** 3.21***

R-squared 0.30 0.41 0.41 0.49

N 6552 4836 4524 3113 3113 754

Note: All the determinants except REC are in logarithmic form and hence the coefficients are elasticities. i countries

refer to the 49 exporting African countries andjcountries include importing countries all over the world. PPI denotes

Producer Price Index and PAE denotes Public Agricultural Expenditure. The lagged value of PAE is used to control

for possible endogeneity.

136

Table 5.2. Determinants of intra-Africa agricultural exports

Determinants

Intra-Africa export African export to the rest of

the world

OLS Tobit OLS Tobit

Importer’s GDP (billions of US$) 1.91*** 2.75*** 2.31*** 3.48***

Exporter’s GDP (billions of US$) 0.32** 0.44* 1.22*** 1.84***

Per capita GDP of exporters (US$) -1.39** -1.89* -2.51*** -4.03***

Per capita GDP of importers(US$) 1.24*** 2.24*** 0.01 -0.06

Arable land (millions of hectares) -0.21 -0.1 -0.53*** -0.62***

Agricultural labor (millions) -0.43 -0.54 -0.43** -0.81***

Road density (km per km2 of land) -0.22 -0.37 0.03 0.12

Quality of port 4.46*** 6.83*** 4.68*** 7.05***

Quality of transport infrastructure 0.71 -0.45 1.26** 1.13

Efficiency of customs clearing index 2.39* 5.45** 1.51** 3.39***

Export cost ($US per container) -0.14 -0.63 0.02 -0.18

PAE per agricultural GDP of exporter 0.2 0.62** 0.14** 0.41***

Incidence of importer’s non-tariff barriers 0.2 0.24 -0.35*** -0.39***

Average tariff rate of importer 0.53*** 0.95*** -0.11 -0.32***

Being in a similar REC 3.55*** 5.68***

Constant -9.64* -20.95** 2.62 2.49

sigma 4.53*** 4.13***

R-squared 0.435 0.519

N 619 619 2494 2494

Note: All the determinants except REC are in logarithmic form and hence the coefficients are elasticities. i countries

refer to the 49 exporting African countries andjcountries include importing African countries for intra-African

trade and importing countries outside of Africa for export to the rest of the world. PAE denotes Public Agricultural

Expenditure. The lagged value of PAE is used to control for possible endogeneity.

137

5.4 Describing and tracking key determinants

5.4.1 Resource endowment and productivity

As this study exclusively considers agricultural products, we assume that agriculture is land and

labor intensive in the African context but less capital intensive compared to other sectors’ products,

expecting a negative effect of capital and positive effects of land and labor on agricultural exports.

However, all three resource endowment variables, labor, land and capital (represented by

exporters’ per capita income), show negative effects on agricultural exports (see Table 5.1).

According to this result, countries with higher per capita income are less likely to export

agricultural products than countries with lower per capita income. This is in line with the relative

resource endowment theory which predicts that a country specializes in an industry that requires

less of the scarcest resource in the country. Hence, while countries grow (accumulate capital), their

export portfolio shifts from agriculture (less capital intensive) to sectors which are more capital

intensive. Thus, capital endowment reduces exports of primary agricultural products.

The results also suggest that countries with scarce arable land and agricultural labor export more

than countries with abundant agricultural land and labor endowments. The negative effect of land

on agricultural exports is due to the exclusion of land productivity from the models. When land

and labor productivity are included in the model, the results become significantly different (Table

5.3). If productivity is controlled for, land positively affects the performance of agricultural exports

both to the world and African markets. The elasticity is greater for intra-African trade than for

global trade. The impact of labor has remained negative. Labor-abundant countries export less than

labor-scarce countries, keeping productivity constant. This could be due to the fact that African

agriculture is not labor intensive as we expected. Alternatively, in an area where labor is abundant

with low productivity, agricultural production may serve only for household subsistence without

any significant contribution to exports.

Similarly, while countries with high land productivity export at a higher rate than countries with

low land productivity, countries with high labor productivity export at a lower rate than countries

with low labor productivity. Labor productivity negatively affects trade, probably because

wherever the productivity of labor is high, the local market becomes more attractive to producers

than the export market. Increased agricultural labor productivity might be good for reducing

poverty, but it seems to negatively affect agricultural export performance in Africa. But the

138

negative effect may indicate the extent of economic transformation. Countries with higher labor

productivity are countries in which economic activity is shifting to the non-agricultural sector, and

hence the composition of their exports is shifting from agricultural to non-agricultural products.

All these imply that while availability of arable land and increased land productivity can positively

affect agricultural trade, having abundant labor alone does not necessarily lead to higher trade;

rather it may retard the continent’s global as well as intra-regional trade. Moreover, trade seems

more elastic for land productivity than land availability, implying that investment in land

productivity-enhancing technologies or institutions would help not only to increase farmers’

income but also to boost regional trade. A 1 percent increase in land productivity increases trade

flows by about 6 percent to the global market and 7 percent to the African market. Land

productivity has a stronger effect on intra-African trade than on global trade, which further

explains the importance of improving land productivity to triple intra-African trade. This is

because many African countries have similar resource endowments and closely similar trade

facilities, so their competitiveness in regional trade mainly depends on the extent of agricultural

productivity.

Table 5.3. African agricultural export response to land and labor endowments and productivity

(elasticity)

Endowment and productivity indicators Global trade Intra-African trade

(3) (7) (8) (9) (10)

Arable land (millions of hectares) -0.52*** 5.82*** 7.15***

Agricultural labor (millions) -0.38** -6.00*** -6.88***

Land productivity (US$ per ha) 6.24*** 0.56*** 7.21*** 0.35***

Labor productivity (US$ per person) -6.43*** -0.13 -7.40*** 0.00

R-squared 0.41 0.49 0.51 0.44 0.44

N 4524 3113 3435 3101 3397

Source: Authors’ estimation based on international sources

Note: Global trade denotes bilateral trade between African countries and selected countries globally, including other

African countries. Intra-African trade denotes trade among African countries only. Estimations include additional

variables for which results are not presented here.

139

5.4.2 Infrastructural quality and institutional efficiency

Variables addressing the quality of ports and transport, road density, efficiency of customs

clearing, and financial export costs have explained a significant part of the variation in agricultural

export performance among African countries (Table 5.1). However, there appear to be significant

differences among cost indicators in explaining trade flows. On one hand, road density and

financial export costs do not have statistically significant effects on export growth. On the other

hand, the quality of port infrastructure and the efficiency of customs clearing consistently and

positively affect trade performance.

Since the cost of trade affects not only export performance but also trade competitiveness, which

is defined as the ratio of a country’s exports to total African exports to the world or to the African

market, further analysis is made to shed light on how cost indicators affect the competiveness of a

country in global and regional markets.

Table 5.4 presents the effects of trade cost indicators on global and regional competiveness. From

these results, it is obvious that although road density and financial export costs have no effect on

export volumes, they do have significant effects on competiveness. This is particularly significant

when it comes to financial payments to clear exports. Financial export costs include all costs

exporters pay for documents, administrative fees for customs clearance and technical control,

customs brokers, terminal handling charges, and inland transport, and these costs are found to be

very crucial for trade competiveness. The lower these fees, the more likely a country becomes

competitive both in regional and global markets. Unfortunately, financial fees for exports are

increasing over time in Africa South of the Sahara (SSA) (Figure 5.1). Sixteen African countries

do not have their own ports. These countries incur higher per unit financial export costs than costal

countries. The cost gap between these groups of countries is widening over time. Lack of port

access may induce preferential fees for port services and increase inland transport costs, thereby

raising export costs. It also creates business insecurity.

140

Table 5.4. Effect of trade costs on agricultural trade competiveness in Africa (elasticity)

Cost indicators

Share of country i ’s supply in total African supply to

Global markets African markets

Road density (km per km2 of land) 0.002*** 0.003***

Quality of port 0.105*** 0.118***

Quality of transport infrastructure -0.003 0.000

Efficiency of customs clearing index -0.016*** -0.019**

Financial fees for export ($US per container) -0.004*** -0.006***

Source: Authors’ estimation based on international sources

Note: Estimations include additional variables for which results are not presented here.

Figure 5.1. Trends of average financial costs for export in SSA

Source: Authors’ calculation based on World Bank Development Indicators

Note: Land locked countries are those SSA countries which do not have their own ports. Costal countries are all SSA

countries which have their own port(s).

Although the effect of road density on export performance was insignificant in most specifications

(Table 5.1), it appears to have a strong and positive effect on competiveness (Table 5.4). This

100

02

00

03

00

04

00

0

US

$ p

er

co

nta

iner

2006 2008 2010 2012 2014

SSA Landlocked countries Costal countries

Figure 1. Trends of average financial costs for export in SSA

141

could be due to the fact that the African road networks are biased to connect local markets more

than regional markets (Gwilliam et al., 2008).

Even though domestic road networks have improved in many African countries over the past two

decades, they are not well-connected to the regional roads, and hence they failed to increase export

volumes but still contribute to the country’s competiveness. Unlike export volumes, which depend

primarily on external efficiency, competitiveness depends mainly on internal efficiency. A country

might be competitive compared to other producers but its export volumes may not grow at a faster

rate than others. This is exactly what the road density results demonstrate. Improved road density

improves a country’s internal competiveness to supply cheaper products to external markets, so

that the share of that country is higher than those of countries with lower road density. However,

since the roads do not adequately connect local markets with regional or global markets, their

effect on absolute export volumes remains insignificant. Despite the significance of road density,

Africa still remains poorly connected both internally and externally. According to the World Bank

Rural Accessibility Index, only 34 percent of the rural population in Africa South of the Sahara

lives within 2 kilometers of an all-weather road (Carruthers, Krishnamani, & Murray, 2010).

Port quality has remained important both for absolute export volumes (Table 5.1) and trade

competiveness (Table 5.4). However, Africa has the lowest port quality of all regions. Based on

the quality of port infrastructure, the World Bank classifies ports into 7 groups, 1 being extremely

underdeveloped and 7 being considered efficient by international standards. According to this

classification Africa South of the Sahara scores 3.65, which is 13 percent below the world average

and 29 percent below the average for high income countries. This indicates an urgent need for

African countries to invest in port infrastructure to improve both regional and global trade.

Other variables related to transport infrastructure and institutional efficiency are important for

export growth but not for competiveness (Table 5.4). The negative effect of institutional efficiency

on competiveness is very hard to explain. The institutional efficiency indicator is developed based

on the number of documents, number of signatures and number of days required to clear customs,

both for imports and exports. The mix of these requirements may explain how the institutional

efficiency index is related to trade competiveness.

142

Figure 5.2. Number of days and documents needed to clear exports

Source: Authors’ calculation based on World Bank World Development Indicators

Note: HIC refers to high income countries and LDCs to least developed countries according to the UN classification.

Values refer to the mean of an average country in the group.

Figure 5.2 shows the number of documents and number of days required for clearing exports across

different regions. In many instances, more requirements are imposed on imports than exports for

all indicators. SSA has the highest requirements for all indicators compared to other regions. On

average it takes more than 32 days to clear exports in Africa South of the Sahara as compared to

less than 10 for high-income countries and 27 days in all least developed countries. We observe

significant differences across regional economic communities, the worst being SADC member

states in which an average export takes close to 50 days. The same is true for the number of

documents required to clear exports. However, both indicators are declining over time (Figure

5.3). The number of documents has already declined from nine on average in 2006 to seven in

2010 and remained constant thereafter. It seems that countries’ progress in improving customs

clearing processes has stalled. The number of days continues to decline from 36 in 2006 to below

30 days in 2014, but the rate of decline remains very slow.

4.5

12.7

7

27.4

7.4

27.8

7.4

30.7

7.5

32.4

7.5

33.3

8.3

41.2

8.3

49.50

10

20

30

40

50

mea

n 2

00

6-2

01

4

HIC

LD

Cs

EC

OW

AS

Afr

ica

SS

A

CO

ME

SA

EC

CA

S

SA

DC

Figure 2. Number of days and documents needed to clear exports

Documents Days

143

Figure 5.3. Trends of export clearing efficiency in Africa South of the Sahara

Source: Authors’ calculation based on World Bank World Development Indicators

5.4.3 Public Agricultural Expenditure

The effect of domestic agricultural support in exporting countries could be an important

determinant of export growth in developing countries due to the fact that farmers and traders in

these countries are poor and less commercialized, and therefore less able to facilitate production

and trade by themselves. The support provided in these countries is different from the support

provided in high income countries. In developing countries support is given to facilitate provision

of agricultural extension, advisory, market access and financial services. Public agricultural

expenditure (PAE) is used as a proxy variable to measure the significance of government support

in promoting agricultural exports in Africa. The empirical results reveal that there exists a positive

and statistically significant association between PAE and export growth. On average a 10 percent

increase in public agricultural expenditure relative to agricultural GDP increases agricultural

exports in the following year by about 2 to 4 percent.

The correlation between public agricultural spending and export performance significantly varies

across countries. Figure 5.4 illustrates the correlation coefficients for selected African countries

calculated using time series data for the last ten years. Unexpectedly, public agricultural

expenditure has no or negative correlation with exports in many countries. While Ethiopia stands

30

32

34

36

Nu

mb

er

of da

ys

7.4

7.6

7.8

88

.2

Nu

mb

er

of do

cu

me

nts

2006 2008 2010 2012 2014

Number of documents Number of days

Figure 3. Trends of export clearing efficiency in Africa South of the Sahara

144

out as the country with the largest negative correlation, Rwanda takes the leading role as the most

successful country on the positive end.

Many factors could explain why countries experience a negative correlation. First, these countries

might have focused more on domestic food security and hence, public expenditure has little or no

relevance in promoting external trade. This is the case in Ethiopia, where a significant part of the

public budget is allocated to mega food security projects such as the Productive Safety Net

Program (PSNP) and extension personnel who primarily provide services for food crop production.

The country’s competitive commodities such as coffee, oilseeds, and hides and skins have been

receiving very little budget allocation, relative to their importance to exports. Second, these

countries’ investments in export commodities might be less efficient in facilitating trade and

production. Third, a decline in the terms of trade could explain part of the paradox, but empirically

this should have little contribution to the negative correlation.

On the other end of the graph (Figure 5.4), there are many countries which are able to utilize the

public budget to motivate agricultural exports. Rwanda is followed by Liberia, Ghana, and

Zimbabwe, in which expenditures and exports are strongly correlated, with coefficients above 0.8.

Policymakers aiming to achieve the Malabo target may consider having a preferential public

expenditure allocation towards commodities in which they have competitive advantage, and

should balance investments in domestic food self-sufficiency (non-tradables) and the export sector

(tradables).

145

Figure 5.4. Correlation between public agricultural expenditure and agricultural exports

Source: Authors’ estimation based UNCOMTRADE export data and ReSAKSS public expenditure data.

Note: Correlations are calculated between current export values and previous year’s public expenditure.

5.4.4 Regional trade agreements

Regional trade agreements remove or reduce tariffs and facilitate joint trade for member states of

Regional Economic Communities (RECs). These agreements create trade within the trade

agreement zone and divert imports from the rest of the world. Empirical results have shown that

the trade creation effect of African RECs such as COMESA, ECOWAS, SADC and ECCAS are

stronger than their trade diversion effects (Figure 5.5). The overall trade creation effect as captured

by the variable REC, which takes 1 if the importing and exporting countries are from the same

RECs and zero otherwise, has a positive and statistically and economically significant effect on

export growth. Being a member of any of the RECs increases a country’s export value by 3 to 5

percent. This effect captures not only the effect of free trade agreements but also the effect of trade

facilitations commonly targeted for cross-border trade. Countries within the same REC are

geographically closer to each other, and hence this variable may also capture proximity effects as

well. In any case, the trade creation effects of African RECs are convincingly large and significant.

-1-.5

0.5

1

Corre

latio

n co

effic

ent

Ethi

opia

Moz

ambi

que

Gui

nea

Equa

toria

l Gui

nea

Cong

o

Nige

ria

Buru

ndi

Mau

ritan

ia

DRC

Gui

nea-

Biss

au

Eritr

ea

Nige

r

Tuni

sia

Cent

ral A

frica

Mau

ritiu

s

Sene

gal

Burk

ina

Faso

Togo

Keny

a

Gam

bia

Cote

d'iv

ore

Ugan

da

Mal

awi

Zim

babw

e

Gha

na

Libe

ria

Rwan

da

Figure 4. Correlation between public agricultural expenditure and agricultural export

146

Figure 5.5. Trade creation and diversion effects of RECs in Africa

Note: The values under “REC” indicate the trade creation effects of all communities. REC is a dummy variable that

takes the value 1 if both importing and exporting countries are from the same REC and 0 otherwise. Effects denoted

by each of the RECs indicate the trade diversion effects. For example, the value under “COMESA” indicates the effect

of a variable that takes 1 if the importing country is a COMESA member and the exporting country is a non-member

and 0 otherwise, and hence measures the trade diversion effect of COMESA. The same holds for the other RECs. The

graph shows coefficients and 95 percent confidence intervals. If zero is included within the confidence interval, the

coefficient is interpreted as statistically insignificant.

The trade diversion effects of these RECs are not yet significant and uniform. The effects were

captured by including dummy variables for each REC that take the value of 1 if the importing

country is a member of a given REC and the exporting country is not, and zero otherwise. This

variable measures openness of member states to non-member states. As shown in Figure 5.5, the

variable representing ECOWAS has a significant and positive effect on exports, implying that

being a member of ECOWAS makes countries open to non-member states, signifying a positive

trade diversion effect. SADC has a protective effect, but it is only significant at 10 percent (90

percent confidence interval). COMESA and ECCAS have shown negative diversion effects, which

may imply import protecting effects to the detriment of non-member states, but the coefficients

are not statistically significant. The results are consistent with previous evidence (Makochekanwa,

2012). Since welfare depends on the extent of both trade diversion and trade creation, policymakers

REC

COMESA

ECOWAS

SADC

ECCAS

-1 0 1 2 3 4Coefficencts

Figure 5. Trade creation and diversion effects of regional economic comunities in Africa

147

should target increasing the diversion as well as the creation effects. Internal institutions and

efficiency may explain the differential effects of RECs on trade diversion.

5.4.5 Tariffs and Preferences

Despite declining trends in tariff rates imposed on agricultural products worldwide, tariffs are still

important determinants of trade. According to our estimation (Table 5.1), a 10 percent increase in

tariff rates reduces African agricultural exports by about 3 percent, which is closely similar to

previous studies (Bouët et al., 2008; Moïsé et al., 2013). Luckily, Africa, particularly SSA, is

increasingly receiving tariff preferences from importing countries. Figure 5.6 shows the average

tariff rates imposed by selected countries on agricultural products imported from the world as a

whole, least developed countries (LDCs), and SSA. Though India and Pakistan impose the largest

tariff rates on agricultural imports globally, they impose lower tariff rates for imports from SSA

than imports from the world. Other countries such as the US, Canada and Russia also impose lower

average duties on imports from SSA. As expected, SSA countries impose lower taxes on imports

from the region than imports from outside the region.

Figure 5.6. Tariff rates imposed by major African trade partners on agricultural imports

Source: Authors’ estimation based on WITS data. Note: Tariff rates are weighted averages based on amount of

imports. Each country or group of countries levies different rates for different countries for the same product. The

rates are averaged for three groups: for all countries, for LDCs and for SSA.

0 10 20 30Percent (weighted average)

Australia

Malaysia

US

Japan

Middle East

Russia

EU

Canada

China

Pakistan

SSA

Turkey

India

Figure 6. Tariff rates imposed by major African trade partners on agricultural imports

On all countries On LDCs On SSA

148

In some countries and regions, including the EU, China and the Middle East, agricultural products

from SSA are being taxed more than the world average. This could be due to the fact that

preferences, especially by the EU, are given for selected products and that preference rates are

exceeded by the tariff rates imposed on non-preferential products. In many countries, African

products are taxed at higher rates than the average for LDCs. This indicates that although several

preferences are enacted in the EU and the US, African products are still highly taxed compared to

other developing countries. Most importantly, SSA countries impose import tax on other SSA

countries at a higher rate than they impose on all LDCs. This implies that some African countries

are providing a lower tax rate for non-African countries than they impose on African countries.

Tariff rates applicable on imports of agricultural products from any part of the world are sharply

declining (Figure 5.7). Average tariff rates declined from above 12 percent in 2005 to close to 8

percent in 2014, which indicates a 3 percent annual rate of decline. Multilateral negotiations

through WTO and the increasing global food demand as demonstrated by the food price crisis in

2007/2008 might have contributed to this effect.

The decline is proportionally similar among the rates applicable to the whole world, SSA and

LDCs. Globally, African products are being taxed at lower rates than the world average since 2009

and the gap between these tax rates has widened since then.

149

Figure 5.7. Trends of tariff rates imposed on SSA, LDCs, and world exports

Source: Authors’ estimation based on WITS data

Despite clear evidence of preferences given to African products over the world average, there are

a wide range of debates regarding the benefits of these preferences in enhancing African trade.

One of the criticisms is that preferences are given on commodities or products on which Africa

has no comparative advantage. Through this criticism applies to comparisons of manufactured and

agricultural products, it can also be applicable among agricultural products. As shown in Figure

5.8, there exist significant variations in preference rates16 given to SSA by the world, the US and

the EU across different agricultural products. The US provides preferences for a wider range of

products than the EU and others. However, the US does not provide preferences for tobacco and

silk. In contrast, the EU provides the highest preference for tobacco. The US provides the highest

preference to dairy products followed by sugar and hides and skin. Though some African countries

could have comparative advantage in sugar and hides and skin, many countries may not have

global comparative advantage in dairy products (Badiane, Odijo, & Jemaneh, 2014). While

16 Defined as the difference between average tariff rates on imports from the world and imports from SSA.

68

10

12

14

%, ave

rag

e

2005 2010 2015

On all countries On LDCs On SSA countries

Figure 7 : Trends of tariff rates imposed on SSA LDCs and World exports

150

preference rates for cocoa are reasonably significant, preference rates for coffee and tea are

minimal, confirming that preferences are given irrespective of comparative advantage.

Figure 5.8. Rates of preference given to SSA exports for major products

Source: Authors’ estimation based on WITS data

Note: Values (rates of preferences) are calculated as average tariff rates imposed by all countries (world), the EU and

the US on world imports minus tariff rates imposed on SSA imports.

5.4.6 Non-tariff barriers (NTBs)

There is much empirical evidence, including the findings of this paper, that indicates that trade is

more responsive to non-tariff barriers than tariffs (Table 5.1). This shows the increasing

importance of non-tariff barriers following the declining trends of tariffs due to bilateral and

multilateral trade agreements and preferences. However, despite the growing understanding of the

significance of non-tariff barriers to trade, there are certain issues that are not yet clear. These

include 1) which type of non-tariff barriers cause significant impacts on trade; 2) which type of

non-tariff barriers are prevalent in agricultural trade; 3) how these measures are trending; and 4)

what strategic options African countries have to reduce the effect of NTBs on trade performance.

Figure 5.9 shows the prevalence of different NTBs across major African trade partners, which

import about 90 percent of African agricultural exports. Of all the countries, the US takes the lead

-10 -5 0 5 10 15

VEGETABLES

TOBACCO

SUGARS

SILK

OIL SEEDS

MEAT

LIVE TREES

LIVE ANIMALS

HIDES AND SKINS

FISH

DAIRY and EGGS

COTTON

COFFEE and TEA

COCOA

CEREALS

FRUIT AND NUTS

Figure 8. Rates of preferences given to SSA exports for major products

World EU US

151

in terms of the number of measures imposed on imports of agricultural products. During the past

four years, the US has imposed about 1,000 measures annually, which are counted across products

and types of NTBs. Close to 50 percent of these relate to SPS measures. SPS measures followed

by TBT are the dominant type of NTBs in many countries. Quantitative restrictions are widely

prevalent in the EU. Unlike many other measures, SPS requirements are politically and

environmentally acceptable as they relate to health, safety and hygiene. Unfortunately, these

requirements impact trade more than any other measures (Figure 5.10). A ten percent increase in

the number of products affected by SPS measures reduces trade by about 3 percent. This result is

consistent with a previous study which shows that SPS penalizes poor countries more strongly

than others (Disdier, Fontagne, & Mimouni, 2008). Export subsidies, which are prevalent in the

EU, the US and Turkey, are the next type of NTB which negatively and significantly affects

African agricultural trade. The involvement of state enterprises in imports and exports positively

affects African exports, probably due to the discretionary preference that these enterprises may

provide to African imports. The involvement of state enterprises in agricultural trade is most

prevalent in China and India and in some EU member states. The number of NTBs in general are

steadily increasing over time both in the US and the EU, which impose the largest number of trade-

reducing non-tariff barriers of all of Africa’s trading partners (Figure 5.11).

Figure 5.9. Frequency of non-tariff measures on agricultural products (average 2012-2015)

Source: Authors’ calculation based on WTO data Note: Frequency of non-tariff barriers is measured as the sum of all

types of measures for all HS6 classified products. For example, if 2 measures are imposed on one product, 3 measures

on 3 products, and zero on all other products, the frequency will be 2*1+3*3=11.

0

200

400

600

800

1,00

0

US

EU

Japa

n

Chi

na

Aust

ralia

Can

ada

Indi

a

Sing

apor

e

Turk

ey

Mid

dle

East

Rus

sia

Mal

aysi

a

SSA

Paki

stan

Figure 9. Frequency of non-tariff measures on agricultural products (mean 2012-2015)

SPS TBT Trade defence

Quantitative restriction Export subsidy State trading

152

The significant impact of NTBs on trade and their growth over time present significant challenges

to policymakers as to how to minimize the adverse effects of these measures. Because of domestic

public concerns, reducing their prevalence through international negotiation is not likely to be

possible. Rather, policymakers in Africa should focus on reducing the vulnerability of their trade

to these measures. The majority of the measures demand certification and labeling, which increase

the cost of trading. Efficient institutional and infrastructural arrangements are required to reduce

these costs. Establishing a certification and accreditation center for an individual country could be

costly and in some cases impossible. Therefore, regional cooperation should be an important area

of focus for African policymakers. Furthermore, there are areas in which individual countries can

facilitate exports by establishing export facilitation centers that would primarily assist exporters in

fulfilling the requirements imposed by importers.

Figure 5.10. Effects of non-tariff measures on export growth in Africa

Source: Authors’ calculation based on WTO data

Note: SPS refers to sanitary and phytosanitary measures and TBT refers to technical barriers to trade based on the

UNCTAD classification. The graph shows coefficients and confidence intervals. If zero is included within the

confidence interval, the coefficient is interpreted as statistically insignificant.

SPS

TBT

Trade_defense

Quantitative_restriction

Export_subsidy

State_trading

-.5 0 .5 1Elasticities

Figure 10. Effects of NTMs on export growth in Africa

153

Figure 5.11. Trends of non-tariff measures in US and EU

Source: Authors’ calculation based on WTO data

5.4.7 Domestic agricultural supports in OECD countries

The empirical link between domestic agricultural supports in OECD countries and the value of

agricultural exports in African countries is assessed using a ratio of agricultural and non-

agricultural producer prices. This price ratio may capture the effect of all border and domestic

supports including tariffs, export subsidies, and production and input subsidies. Since tariffs and

non-tariff barriers are included as explanatory variables, the price ratio should predict the effect of

domestic supports. As shown in Table 5.1, the effect of this price ratio is negative and statistically

significant. According to this estimation, a 1 percent increase in the price ratio reduces African

exports by about 5 percent. However, the implication of this elasticity depends on the actual

correlation of the price ratio with domestic support. Many economists argue that since most

payments to agricultural producers are made through direct payments, the impact of agricultural

subsidies on trade is very limited (Anderson & Martin, 2005; Croser & Anderson, 2011; Hoekman

et al., 2004). But if we compare producer prices of agricultural and manufacturing products, in

many cases we get a ratio greater than one, which implies that agriculture is treated preferentially

and that this treatment restricts imports from developing countries.

520

540

560

580

600

620

Nu

mb

er

of m

easure

s

2012 2013 2014 2015

USA EU

Figure 11. Trends of non - tarff measures in USA and EU

154

Generally we conclude that although the effect of domestic support might not be as large as cross

border measures such as tariffs and non-tariff barriers, it still plays a significant role.

It appears, however, that the rate of agricultural support in general is declining over time in many

OECD countries. Figure 5.12 shows trends in Producer Support Estimates (PSE) estimated by

OECD for selected countries and groups of countries. Of all countries considered, EU countries

provided the highest support throughout the last two decades. Emerging economies such as China

and Russia are also increasingly supporting their producers despite the instability and

unpredictability of their support. In these countries, support is said to be mainly through tariffs and

non-tariff barriers instead of subsidies.

Figure 5.12. Trends of Producer Support Estimates (PSE) in OECD countries

Source: Authors’ estimation based on OECD data

Both our empirical analysis and trends in the PSE suggest the importance of domestic support in

high income countries for the performance of African exports. However, African countries in

particular and developing countries in general have very few policy options to curb the adverse

effects of this domestic policy action in foreign countries.

01

02

03

04

0

Pe

rce

nt

2000 2005 2010 2015

EU Russia USA China OECD

Figure 12. Trends of Producer Support Estimates ( PSE ) in OECD countries

155

Although multilateral trade negotiations through the WTO are usually of limited effectiveness,

they remain the most likely avenue for developing countries to compel high income countries to

reduce or redesign their agricultural supports. Economic growth in many African and Asian

countries and the increasing threat of climate change may create leverage for developing countries

to organize themselves and enforce effective global policy actions through the WTO.

5.5 Conclusions

African countries are striving to expand market opportunities for domestic producers regionally as

well as globally. However, this effort is being impeded by emerging and evolving constraints.

Though many of the constraints seem conventional and traditional, the nature and extent of the

constraints are evolving dramatically following global and regional shocks and opportunities. This

chapter aims to closely monitor these evolutions and identify key determinants of trade

performance with the purpose of provoking discussions among policymakers and development

partners on how to help Africa achieve the targets set by the Malabo Declaration. To do so, existing

theoretical and empirical evidence is reviewed and comprehensive empirical assessments are made

to supplement existing evidence.

The review generally found that the existing evidence is not sufficiently comprehensive, updated

and focused on African context. Realistic and updated assessments are required to feed the

increasing policy momentum to improve African agriculture. We also learned that agricultural

trade determinants are diverse and complex, ranging from farm level supply side constraints to

global level demand side barriers. This calls for regular monitoring and prioritization of these

constraints for immediate policy and development actions.

The empirical analysis that aimed at identifying and tracking key determinants of trade indicated

that supply side constraints, which include production capacity and cost of trade, are more

important determinants than demand side global constraints. This gives the opportunity for African

policymakers to focus on domestic production and trade facilitation which can easily be influenced

through national and regional policies and investments. A lot can be achieved by simply focusing

on domestic factors instead of assuming that international factors are the culprits for low and, in

some countries, declining agricultural exports. This does not, however, rule out the importance of

cooperation, both regionally and globally.

156

Regional cooperation is key for enhancing trade through reducing trade barriers and increasing

productivity. The empirical analysis clearly confirmed that regional economic communities in

Africa are significantly contributing to the growth of agricultural exports. These regional units can

be further utilized to reduce regional as well as global barriers. One important function of regional

bodies could be joint trade facilitation initiatives that can help to fulfil the growing non-tariff trade

requirements of African trade partners.

Despite a growing tendency toward import tariff reductions partly due to preferential trade, non-

tariff barriers are significantly increasing and impacting African exports more than tariffs. This

trend demands not only regional cooperation but also global cooperation. Ensuring global

cooperation has always been a challenge for developing countries. However, there are growing

opportunities that can enhance the bargaining power of developing countries in general and

African countries in particular. These are the growing economic importance of the continent for

markets and investments and the global climate threat, in that Africa can play pivotal role in

mitigating the problem. However, global cooperation should not be viewed only as an instrument

to influence international trade policies; rather Africa should also seek this cooperation for

facilitating trade and enhancing domestic agricultural value addition.

157

References

Amoroso, N., Chiquiar, D., & Ramos-Francia, M. (2011). Technology and endowments as determinants

of comparative advantage: Evidence from Mexico. The North American Journal of Economics and

Finance, 22(2), 164-196.

Anderson, K. (2014). The intersection of trade policy, price volatility, and food security. Annual Review

of Resource Economics, 6(1), 513-532.

Anderson, K., & Martin, W. (2005). Agricultural trade reform and the Doha Developemnt Agenda. The

World Economy, 28(9), 1301-1327.

Anderson, K., & Nelgen, S. (2012). Agricultural trade distortions during the global financial crisis.

Oxford Review of Economic Policy, 28(2), 235-260.

Anderson, K., & Thennakoon, J. (2015). Food price spikes and poor, small economies: What role for

trade policies. African Journal of Agricultural and Resource Economics, 10(1), 16-31.

Badiane, O., Odjo, S., & Jemaneh, S. (2014). More resilient domestic food markets through regional

trade. In O. Badiane, T. Makombe & G. Bahiigwa (Eds.), Promoting Agricutral Trade to Enhance

Resilience in Africa. ReSAKSS Annual Trends and Outlook Report 2013. Washington DC: International

Food Policy Research Institute.

Benin, S., Mogues, T., & Fan, S. (2012). Agricultural growth and poverty reduction impacts of public

investments: Assessment concepts and techniques. In T. Mogues & S. Benin (Eds.), Public Expenditures

for Agricultural and Rural Development in Africa. UK: Routledge.

Bergstrand, J. H. (1990). The Heckscher-Ohlin-Samuelson Model, the Linder Hypothesis and the

determinants of bilateral intra-industry trade. The Economic Journal, 100(403), 1216-1229.

Bernstein, J., & Weinstein, D. (2002). Do endowments predict the location of production? Evidence from

national and international data. Journal of International Economics, 56, 55-76.

Bouët, A., Bureau, J.-C., Decreux, Y., & Jean, S. (2005). Multilateral agricultural trade liberalisation: The

contrasting fortunes of developing countries in the Doha Round. The World Economy, 28(9), 1329-1354.

Bouët, A., Fontagné, L., & Jean, S. (2005). Is Erosion of Tariff Preferences a Serious Concern? CEPII

Working Paper 2005-14. Paris: CEPII.

Bouet, A., & Laborde, D. (2012). Food crisis and export taxation: The cost of non-cooperative trade

policies. Review of World Economics, 148(1), 209-233.

Bouët, A., Laborde, D., Dienesch, E., & Elliott, K. (2012). The costs and benefits of duty-free, quota-free

market access for poor countries: Who and what matters. Journal of Globalization and Development,

3(1), 1-27.

Bouët, A., Mishra, S., & Roy, D. (2008). Does Africa Trade Less than It Should, and If So, Why? The

Role of Market Access and Domestic Factors. IFPRI Discussion Paper 770. Washington, DC:

International Food Policy Research Institute.

Bougheas, S., Demetriades, P. O., & Mamuneas, T. P. (1999). Infrastrucutre, transport costs and trade.

Journal of International Economics, 47(1), 169-189.

158

Brenton, P. (2003). Integrating the Least Developed Countries into the World Trade System: The Current

Impact of EU Preferences under Everything But Arms. World Bank Policy Research Working Paper

3018. Washington, DC: World Bank.

Bureau, J.-C., Jean, S., & Matthews, A. (2006). The Consequences of Agricultural Trade Liberalization

for Developing Countries. Paper presented at the International Association of Agricultural Economists

Conference, Gold Coast, Australia, Aug. 12-18, 2006.

Carruthers, R., Krishnamani, R. R., & Murray, S. (2010). Africa Infrastructure Country Diagnostic:

Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa. Washington, DC:

World Bank.

Croser, J. L., & Anderson, K. (2011). Agricultural distortions in Sub-Saharan Africa: Trade and welfare

indicators, 1961 to 2004. The World Bank Economic Review, 25(2), 250-277.

Disdier, A.-C., Fontagne, L., & Mimouni, M. (2008). The impact of regulations on agricultural trade:

Evidence from SPS and TBT agreements. American Journal of Agricultural Economics, 90(2), 336-350.

Francois, J., & Manchin, M. (2007). Institutions, Infrastructure and Trade. CEPR Discussion Paper 6068.

London: Centre for Economic Policy Research.

Gwilliam, K., Foster, V., Archondo-Callao, R., Briceño-Garmendia, C., Nogales, A., & Kavita Sethi.

(2008). Africa Infrastructure Country Diagnostic: Roads in Sub-Saharan Africa. Washington, DC: World

Bank.

Hoekman, B., Ng, F., & Olarreaga, M. (2004). Agricultural tariffs or subsidies: Which are more important

for developing economies? The World Bank Economic Review, 18(2), 175-204.

Makochekanwa, A. (2012). Impacts of Regional Trade Agreements on Trade on Agrifood Products:

Evidence from Eastern and Southern Africa. Paper presented at the African Economic Conference 2012,

Kigali, Rwanda, Oct. 30-Nov. 2, 2012.

Moïsé, E., Delpeuch, C., Sorescu, S., Bottini, N., & Foch, A. (2013). Estimating the Constraints to

Agricultural Trade of Developing Countries. OECD Trade Policy Paper 142. Paris: OECD Publishing.

Panagariya, A. (2003). Aid through Trade: An Effective Option? Mimeo.

Tadesse, G., Algieri, B., Kalkuhl, M., & Braun, J. v. (2014). Drivers and triggers of international food

price spikes and volatility. Food Policy, 47, 117-128.

Topp, A. (2003). Are Trade Preferences Useful in Advancing Economic Development? Australian

National University Working Paper 0503.

Urban, K., Jensen, H. G., & Brockmeier, M. (2016). How decoupled is the Single Farm Payment and does

it matter for international trade? Food Policy, 59, 126-138.

Wainio, J., & Gehlhar, M. (2004). MFN Tariff Cuts and U.S. Agrciutral Imports Under Nonresprocal

Trade Preference Programs. Paper presented at the 7th Annual Conference on Global Economic

Analysis, Washington, DC, June 17-19, 2004.

Wilson, N. (2007). Examining the Trade Effect of Certain Customs and Administrative

Procedures. OECD Trade Policy Paper 42. Paris: OECD Publishing.

159

Yu, T.-h., Tokgoz, S., Wailes, E., & Chavez, E. (2011). A quantitative analysis of trade policy responses

to higher world agricultural commodity prices. Food Policy, 36, 545-561.

160

Annex 1. List of determinants and their indicators used to estimate African agricultural export

performance

Determinants Indicators and definitions

Size and income

level

Total GDP and per capita GDP are used to control for the size of both importing and exporting

economies. GDP is measured as real values deflated by 2005 constant price in billions of US$.

Per capita GDP is measured in US$ per person. In both cases, the 2013 values are used.

Missing values are replaced by values of the previous year.

Resource

endowment &

productivity

Land and labor of the exporting countries are chosen to test the role of resource endowment

for trade. Land is measured as the total arable land in millions of hectares and labor is

measured as total agricultural labor in millions of persons. The productivity of these resources

are also included at a later stage of the analysis to test the relevance of endowment vs.

technology. Land productivity is measured as agricultural value added per hectare of land;

similarly labor productivity is estimated as the ratio of agricultural GDP to agricultural labor

force. All the data are obtained from the ReSAKSS database (www.resakss.org).

Infrastructural

quality:

Road density, quality of port and quality of trade transport infrastructure quality are used to

measure the effect of infrastructure on trade performance. Road density is obtained from

publicly available international sources17 and measured in terms of kilometer per square

kilometer. Indices of port and trade transport qualities are obtained from the World Bank

survey on ‘doing business”. The indices are represented by scalar cores that ranges from 1 to

7; 1 being extremely poor/inaccessible and 7 being very efficient/accessible. Since the survey

data is available in different years for different countries, the average of available data from

2010 to 2013 are used.

Institutional

efficiency

The World Bank Logistics Performance Index specific to the efficiency of customs clearance

process (1=low to 5=high) is used to proxy institutional efficiency related to trade. It

aggregates the respondents’ ranking of the efficiency of customs clearance processes (i.e.

speed, simplicity and predictability of formalities), on a rating ranging from 1 (very low) to 5

(very high). Scores are averaged across all respondents.

Financial cost of

exports

Both infrastructural quality and institutional efficiency used to proxy costs of trade do not

capture all costs involved in the export of import of commodities. The cost of export estimated

by the World Bank is used to control for unaccounted trade costs. The cost measures the fees

levied on a 20-foot container in U.S. dollars. All the fees associated with completing the

procedures to export or import the goods are included. These include costs for documents,

administrative fees for customs clearance and technical control, customs broker fees, terminal

handling charges and inland transport. The cost measure does not include tariffs or trade

taxes. The average cost from 2010 to 2013 of the exporting country is used.

17 http://www.nationmaster.com/country-info/stats/Transport/Road-density/Km-of-road-per-100-sq.-km-of-land-area

161

Public

agricultural

expenditure

This variable is included to examine the empirical link between public investment and trade

performance. While it is very relevant from a policy perspective, it may cause endogeneity

problems. It may also correlate with other explanatory variables. To avoid these problems, its

lagged value is used for the regression analysis. The nominal value is normalized by

agricultural GDP.

Regional trade

agreements

Regional trade agreement is included as a dummy variable that takes 1 if both trading

countries are members of the same regional economic community (COMESA, ECOWAS, SADC,

ECCAS) and 0 otherwise. At a later stage we also included dummies for each regional block to

measure trade diversion effects of each REC. In this case, for example, we include a dummy for

COMESA that takes 1 if the importing country is member of COMESA and 0 otherwise. Similar

dummies are used for the other RECs.

Tariff Aggregation is the primary concern for measuring the effect of tariffs on trade. The use of

tariff indices such as the trade restrictiveness index, ad valorum equivalent, trade reduction

index and nominal rate assistance is quite common to aggregate the different tariff lines.

These indices are preferred over averages because simple averages of tariff rates of the

different agricultural lines will include untraded products and the weighted average based on

imports will be endogenous to trade. However, an all-inclusive index for all the countries

considered in this study is not available. Thus, a mix of weighted and simple averages of ad

valorum rates from WITS (http://wits.worldbank.org/) is used to proxy the effect of tariffs on

trade. Weighted averages are used to aggregate tariff rates on products up to the H2 level

and rates imposed on different countries, and then simple averages are used to approximate a

tariff rate imposed by a country on global imports. Since only exports of African countries are

considered in this analysis, the weighted tariff rates of other countries are less likely to be

endogenous to trade, as the share imports from Africa is relatively small.

Non-tariff

measures

The total number of non-tariff measures (NTM) imposed by the importing country, which is the

sum of all measures reported to the WTO (http://i-tip.wto.org), is used to capture the effect of

non-tariff barriers on African trade. Measures are counted across products and types of

measures. Alternatively we use the frequency of six major types of NTM separately. Only

measures applicable to all WTO members are considered. Non-tariff measures imposed

bilaterally are not considered as they are mostly for non-African countries. Unfortunately, not

all countries reported to WTO, so this variable has many missing values.

Domestic

agricultural

supports

Data on the extent of domestic agricultural support specifically for production and input

subsides is not available for all countries. We used the ratio of the agricultural producer price

index (PPI) to the manufacturing producer price index for OECD countries as a proxy to

represent domestic agricultural support. The agricultural PPI is obtained from FAOSTAT and

the manufacturing PPI is collected from the OECD database (www.oecd.org).

Chapter 7. West Africa trade outlook: business as usual vs alternative options

Extracted from

African Agricultural Trade Status Report

2017

162

CHAPTER 7. WEST AFRICA TRADE OUTLOOK: BUSINESS AS USUAL vs ALTERNATIVE OPTIONS

Sunday Pierre Odjo, International Food Policy Research Institute, West and Central Africa

office, Dakar, Senegal

Ousmane Badiane, International Food Policy Research Institute, Washington DC 7.1 Introduction Recent studies have indicated that Africa as a whole and a number of individual countries have

exhibited relatively strong trade performance in the global market (Bouët et al. 2014) as well as in

continental and major regional markets (Badiane et al. 2014). The increased competitiveness has

generally translated into higher shares of regional markets as destinations for exports from African

countries and regions. Faster growth in demand in continental and regional markets compared to

the global market has also boosted the export performance of African countries. For instance,

during the second half of the last decade, Africa’s share of the global export market rose sharply,

in relative terms, for all goods and agricultural products in value terms, from 0.05 % to 0.21 % and

from 0.15 % to 0.34 %, respectively. This is in line with the stronger competitive position of

African exporters mentioned earlier. The increase in intra-African and intra-regional trade, and the

rising role of continental and regional markets as major destinations for agricultural exports by

African countries, suggest that cross-border trade flows will exert greater influence on the level

and stability of domestic food supplies. The more countries find ways to accelerate the pace of

intra-trade growth, the larger that influence is expected to be in the future. The current chapter

examines the future outlook for intra-regional trade expansion in West Africa and the implications

for the volatility of regional food markets. The chapter starts with an analysis of historical trends

in intra-regional trade of major staple food products as well as the positions of West African

individual countries in the regional market. This is followed by an exploration of the potential of

regional trade to contribute to stabilizing food markets, and by an assessment of the scope for

cross-border trade expansion. A regional trade simulation model is then developed and used to

simulate alternative scenarios to boost trade and reduce volatility in the regional market.

163

7.2. Long-term trends in intra-regional trade of staple food products

Over the last two decades, the cross-border trade of staple food products has followed an increasing

but unsteady trend. It appears from Figure 7.1 and Table 7.1 that fish and animal products—

including meat, dairy and eggs—are the most traded commodities between West African countries

in value terms. Intra-regional trade of these products has on average amounted to US$ 439.2

million in 2011-2013 from only US$ 165.7 million approximately a decade before. They are

followed by live animals and edible oils, the exchange of which has averaged US$ 95.7 million

and US$ 307.3 million, respectively, in 2011-2013. At this amount, the cross-border trade of

vegetable oils has grown fourfold compared to its average level in the early 2000s.

Intra-West Africa trade of cereals and vegetables has generally occurred in lower amounts. For

instance, the regional market of cereals and vegetables amounted on average to US$ 81.5 million

and US$ 28.5 million, respectively, in 2006-2010. The region then more than doubled the level of

its cereals trade in early 2000s. However, a remarkable contraction of the regional market of

cereals has occurred in 2011-2013. In contrast, a surge of trade in vegetables happened in 2011,

inflating the average market size to US$ 133.7 million for the period 2011-2013.

Oilseeds are the least traded product within West Africa in value terms. Cross-border exchange of

this commodity amounted to US$ 31.8 million on average in 2011-2013, reaching almost the

double of its value in the early 2000s. Other staple food crops including edible fruits & nuts and

live trees and plants like roots & tubers constitute a relatively larger regional market size. Their

regional trade reached on average the value of US$ 54.8 million in 2011-2013, more than doubling

the corresponding value in the early 2000s.

164

Figure 7.1. Trends in intra-regional exports of staple food products in West Africa, 1998-2013

Source: Author’s calculations based on HS4-level bilateral trade values from the BACI database, 1998-2013. Note:

West Africa is here extended to the ECOWAS/CILSS area, including 15 ECOWAS members and Chad and

Mauritania.

Table 7.1. Average value of intra-regional trade of staple food products in West Africa (million US

dollars)

2001-2005 2006-2010 2011-2013

Live animals 87.7 155.6 95.7

Fish & animal products 165.7 348.4 439.2

Vegetables 27.3 28.1 133.7

Cereals 30.1 81.5 64.5

Oilseeds 16.8 17.8 31.8

Edible oils 75.8 137.4 307.3

Other food crops 20.6 28.5 54.8

All staple food products 424.1 797.3 1127.0 Source: Author’s calculations based on HS4-level bilateral trade values from the BACI database, 1998-2013. Note:

West Africa is here extended to the ECOWAS/CILSS area, including 15 ECOWAS members and Chad and

Mauritania.

1

10

100

1000

199

8

199

9

200

0

200

1

200

2

200

3

200

4

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3

Mil

lio

n U

S d

oll

ars

in l

og s

cale

LIVE ANIMALS

FISH & ANIMAL

PRODUCTS

VEGETABLES

CEREALS

OILSEEDS

EDIBLE OILS

OTHER FOOD CROPS

165

In sum, the cross-border trade of major food products has been expanding among West African

countries. It is tempting to explore which countries are the major exporters versus importers in the

regional markets of the different commodity groups under analysis. Table 7.2 presents the net trade

positions of each country in the regional market of each commodity group. In each cell, a negative

(positive) number indicates for a net importing (net exporting) country its share in the total value

of net imports (net exports) of a commodity across all countries of the region. In the bottom line

of the table, the contributions of all countries add up to zero for each commodity since the regional

market clears in the sense that the sums of net imports and net exports of the commodity over all

countries are equal.

For instance, Table 7.2 shows that Nigeria is the biggest net importer of live animals, followed by

Côte-d’Ivoire and Senegal with 50.4%, 20.6% and 18.4% of the regional import market,

respectively. Thus, these 3 major importing countries account for 89.4% of the regional import

market, the remaining 10.6% being made up by net imports of Benin, Ghana, Guinea, Mauritania

and Togo. In contrast, Niger and Mali are the biggest net exporters of live animals, with 50.5%

and 43.2% of the regional export market, followed by Burkina Faso with 6.2%, while other

countries contribute negligible market shares. To help visualize major differences between

countries in terms of their regional market positions across the different commodity markets, the

results of Table 7.2 have been mapped into Figure 7.2.

166

Table 7.2. Contributions to values of net imports and net exports of staple food products among West

African countries, 1998-2013 (%)

Live

animals

Fish & animal

products

Live trees

& plants

Vege-tables

Edible fruits

& nuts Cereals Oilseeds Edible

oils

Benin -6.1 -1.4 -1.9 -6.4 -6.5 33.8 0.3 8.4

Burkina Faso 6.2 -4.1 -8.8 11.8 -3.7 -1.9 66.5 -7.7

Cape Verde 0.0 0.2 0.0 -0.1 -0.3 -1.1 0.0 0.0

Chad 0.1 0.4 -32.2 -0.1 -0.7 -0.8 0.0 0.0

Cote d'Ivoire -20.6 -54.6 52.6 -67.6 78.6 18.3 18.9 88.3

Gambia 0.0 -0.2 -0.3 -0.2 -0.7 -1.6 7.5 -0.2

Ghana -3.0 -7.2 -37.4 15.2 12.4 -1.9 -45.4 0.7

Guinea -0.4 8.8 -0.5 -0.1 1.8 -1.4 -1.4 -1.3

Guinea-Bissau 0.0 3.2 0.0 -0.1 0.1 -11.7 0.0 -0.4

Liberia 0.0 -0.4 -5.2 -0.4 -0.1 -0.8 -0.6 -0.4

Mali 43.2 -4.7 -1.4 -3.3 7.2 -21.7 5.1 -21.0

Mauritania -0.9 71.6 -1.3 -0.3 -7.3 -4.9 -3.7 -0.1

Niger 50.5 0.1 -1.2 71.2 -14.5 -30.0 1.7 -16.7

Nigeria -50.4 -25.4 47.4 -18.3 -13.4 -22.2 -11.9 -28.6

Senegal -18.4 15.8 -5.2 1.8 -52.7 39.5 -12.6 -23.6

Sierra Leone 0.0 -0.4 -0.9 -0.5 -0.1 -0.1 -1.5 0.0

Togo -0.3 -1.5 -3.6 -2.6 0.0 8.3 -22.9 2.6

Sum of contributions 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Source: Author’s calculations based on HS4-level bilateral trade values from the BACI database, 1998-2013. Note:

i) West Africa is here extended to the ECOWAS/CILSS area, including 15 ECOWAS members and Chad and

Mauritania; ii) Negative (positive) numbers indicate the shares of net importing (net exporting) countries in the sum

of net import (net export) values across all countries of the region.

What we have just said about country positions in the regional market of live animals appears more

clearly in Figure 7.2, where major net importers and net exporters are clustered at the top and the

bottom of the figure, and countries with modest market participations are spread in between.

Nigeria and Côte d’Ivoire are the biggest net importers of vegetables while Niger, Ghana, and

Burkina Faso are net exporters. In addition, Nigeria and Côte d’Ivoire are net importers of fish &

animal products while net exports are supplied by Mauritania, Senegal, Cape Verde and Guinea.

The regional oilseeds market is dominated by Ghana, Togo, Senegal and Nigeria as net importers

and by Burkina Faso, Côte d’Ivoire, Gambia and Benin as net exporters. Cereals are mostly net

imported by Niger, Mali, Nigeria and Guinea Bissau and net exported by Senegal, Benin and Côte

d’Ivoire. Edible fruits & nuts are particularly net imported in the regional market by Senegal,

Nigeria and Niger and net exported notably by Côte d’Ivoire and less considerably by Ghana.

167

The regional market of vegetable oils is dominated by Nigeria, Senegal, Mali and Niger as major

net importers and by Côte d’Ivoire as the only major net exporter. Finally, Ghana and Chad

dominate the market of live trees and plants as net importers while Côte d’Ivoire and Nigeria are

the biggest net exporters.

Figure 7.2. Distribution of net exports and net imports of staple food products among West African

countries, 1998-2013

LIVE ANIMALS

VEGE-TABLES

FISH & ANIMAL PRODUCTS

OIL-SEEDS CEREALS

EDIBLE FRUITS & NUTS

EDIBLE OILS

LIVE TREES & PLANTS

Nigeria

Côte d'Ivoire

Senegal

Ghana

Liberia

Sierra Leone

Gambia

Togo

Benin

Guinea

Cape Verde

Mauritania

Burkina Faso

Chad

Guinea-Bissau

Mali

Niger

LEGEND Country share in total net-imports value (%) Country share in total net-exports value (%)

]-100,-50] ]-50,-10] ]-10, 0] [0, 10[ [10, 50[ [50, 100[

Source: Author’s calculations, constructed from Table 7.2 above, based on HS4-level bilateral trade values from the

BACI database, 1998-2013. Note: West Africa is here extended to the ECOWAS/CILSS area, including 15 ECOWAS

members and Chad and Mauritania.

Before closing this section on historical trends in intra-regional trade, it is important to analyze

harassment practices that are perceived as bottlenecks to the free movement of goods and persons

across the region. Figure 7.3 summarizes survey data on checkpoints, bribes paid and delays along

major cross-border transport corridors in West Africa. The average numbers plotted are illustrative

of the importance of abnormal trade costs to traders that operate in the regional market.

168

Every 100 km at least 2 checkpoints are encountered and a minimum of CFAF 2000 are paid in

bribes across the surveyed corridors. More than 3 checkpoints are found along the corridor

connecting Bamako (Mali) and Ouagadougou (Burkina Faso) and average bribes exceed CFAF

6000.

Figure 7.3. Indicators of harassment practices along West African corridors, 2010-2012

0.0

1.0

2.0

3.0

4.0

Average number of checkpoints per 100 km

0

2000

4000

6000

8000

Fran

cs C

FA

Average bribe taken per 100 km

0102030405060

Min

ute

s

Delay per 100 km

Source: Authors’ calculations based on survey results by the Improved Road Transport Governance (IRTG)

Initiative.

169

The preceding analysis has demonstrated that cross-border trade of staple food products is

increasing. We now turn to exploring the potential for expanding the current level of intra-regional

trade.

7.3. Regional Potential for the Stabilization of Domestic Food Markets through Trade

Variability of domestic production is a major contributor to local food price instability in low

income countries. The causes of production variability are such that an entire region is less likely

to be affected than individual countries. Moreover, fluctuations in national production levels for

different countries tend to partially offset each other, so that such fluctuations are less than

perfectly correlated. Food production can be expected to be more stable at the regional level than

at the country level. In this case, expanding cross-border trade and allowing greater integration of

domestic food markets would reduce supply volatility and price instability in these markets.

Integrating regional markets through increased trade raises the capacity of domestic markets to

absorb local price risks by: (1) enlarging the area of production and consumption and thus

increasing the volume of demand and supply that can be adjusted to respond to and dampen the

effects of shocks; (2) providing incentives to invest in marketing services and expand capacities

and activities in the marketing sector, which raises the capacity of the private sector to respond to

future shocks; and (3) lowering the size of needed carryover stocks, thereby reducing the cost of

supplying markets during periods of shortage and hence decreasing the likely amplitude of price

variation.

A simple comparison of the variability of cereal production in individual countries against the

regional average is carried out to illustrate the potential for trade and local market stabilization

through greater market integration (Badiane, 1988). For that purpose, a trend-corrected coefficient

of variation is used as a measure of production variability at the country and regional levels.

Following Cuddy and Della Valle (1978), the trend-corrected coefficient of variation in cereal

production is calculated for each ECOWAS member country as follows:

𝑇𝐶𝑉𝑖 = 𝐶𝑉𝑖 ∙ √1 − 𝑅𝑖2

170

where 𝐶𝑉𝑖 is the coefficient of variation in the series of cereal production quantities in country 𝑖

from 1980 to 2010 and 𝑅𝑖2 is the adjusted coefficient of determination of the linear trend model

fitted to the series. Then an index of regional cereal production volatility 𝑇𝐶𝑉𝑟𝑒𝑔 is derived for the

ECOWAS region as a weighted average of the trend-corrected coefficients of variation of its

member countries with the formula (Koester, 1986):

𝑇𝐶𝑉𝑟𝑒𝑔2 = ∑ 𝑠𝑖

2 ∙ 𝑇𝐶𝑉𝑖2𝑛

𝑖 + 2 ∑ ∑ 𝑠𝑖 ∙ 𝑠𝑗 ∙ 𝑣𝑖𝑗 ∙ 𝑇𝐶𝑉𝑖 ∙ 𝑇𝐶𝑉𝑗𝑛𝑗

𝑛𝑖

where 𝑇𝐶𝑉𝑖 and 𝑇𝐶𝑉𝑗 are the trend-corrected coefficients of variation in cereal production in

countries 𝑖 and 𝑗, 𝑛 is the number of ECOWAS member countries, 𝑠𝑖 and 𝑠𝑗 are the shares of

countries 𝑖 and 𝑗 in the region’s overall cereal production, and 𝑣𝑖𝑗 is the coefficient of correlation

between the series of cereal production quantities in countries 𝑖 and 𝑗. Finally, the trend-corrected

coefficients of variation calculated at the country level are normalized by dividing them by the

regional coefficient.

In Figure 7.4, the bars represent the normalized coefficients of variation which indicate by how

much individual country production levels are more (normalized coefficient greater than 1) or less

(normalized coefficient less than 1) volatile than production in the ECOWAS region. The figure

shows that for almost all countries, national production volatility is considerably larger than

regional level volatility, with only the exception of Côte d’Ivoire. Gambia, Liberia, Mali, Niger

and Senegal show considerably higher volatility levels than the region. These countries would be

the biggest beneficiaries of increased regional trade in terms of greater stability of domestic

supplies. However, the likelihood that a given country will benefit from the trade stabilization

potential suggested by the difference between its volatility level and the regional average will be

greater the more the fluctuations of its production and that of the other countries in the region are

weakly correlated.

171

Figure 7.4. Cereal production instability in ECOWAS countries (1980-2010)

Source: Authors’ calculations based on FAOSTAT 2014 data for the period 1980–2010.

Therefore, we plot in Figure 7.5 the distribution of production correlation coefficients between

individual countries in the region. For each country, the lower segment of the bar shows the

percentage of correlation coefficients that are 0.65 or less, or the share of countries with production

fluctuations that we define as relatively weakly correlated with the country’s own production

movements. The top segment represents the share of countries with highly correlated production

fluctuations, with coefficients that are higher than 0.75. The middle segment is the share of

moderately correlated country productions with coefficients that are between 0.65 and 0.75.

Country production levels tend to fluctuate together as shown by the high share of coefficients that

are above 0.75 for the majority of countries. However, for some of them, including Guinea Bissau,

Liberia and Senegal, the share is smaller than 30%. The division of the region into two nearly

uniform sub-regions, sahelian and coastal, may be an explanation. In general, the patterns and

distribution of production fluctuations across countries in the region are such that increased trade

could be expected to contribute to stabilizing domestic agricultural and food markets. But that is

only one condition, the other being that there is actual potential to increase cross-border trade, a

question that is examined in the next section.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

Benin BurkinaFaso

Coted'Ivoire

Gambia Ghana Guinea GuineaBissau

Liberia Mali Niger Nigeria Senegal Togo

Norm

aliz

ed c

oeff

icie

nt

of variation

172

Figure 7.5. Distribution of production correlation coefficients between ECOWAS countries (1980-

2010)

Source: Authors’ calculations based on FAOSTAT 2014 data for the period 1980–2010.

7.4. The scope for specialization and regional trade expansion in agriculture

Despite the recent upward trends, the level of intra-African and intra-regional trade is still very

low compared with other regions. Intra-African markets accounted only for an average of 34 % of

the total agricultural exports from African countries between 2007 and 2011 (Badiane et al. 2014).

Among the three RECs, SADC had the highest share of intra-regional trade (42 %), and ECOWAS

the lowest (6 %). COMESA’s share of intra-regional trade was 20 %. Although SADC is doing

much better than the other two RECs, intra-regional exports still account for far less than half of

the value of the region’s total agricultural exports (Badiane et al. 2014). There may be a host of

factors behind the low levels of intra-regional trade. These factors may not only make trading with

extra-regional partners more attractive, but they may also raise the cost of supplying regional

markets from intra-regional sources. The exploitation of the stabilization potential of regional

trade, as pointed out above, would require measures to lower the barriers to and the bias against

transborder trade so as to stimulate the expansion of regional supply capacities and of trade flows

across borders. This suppose that there is sufficient scope for specialization in production and trade

within the sub-regions. Often, it is assumed that neighboring developing countries would exhibit

similar production and trading patterns because of the similarities in their resource bases, leaving

little room for future specialization.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Benin BurkinaFaso

Coted'Ivoire

Gambia Ghana Guinea GuineaBissau

Liberia Mali Niger Nigeria Senegal Togo

% s

har

e o

f co

rr. c

oef

fici

ents

Corr. coefficients < 0.65 Corr. coefficients between 0.65 and 0.75 Corr. coefficients > 0.75

173

There are, however, several factors that may lead to different specialization patterns among such

countries. These factors include (1) differences in historical technological investments and thus

the level and structure of accumulated production capacities and skills; (2) the economic distance

to, and opportunity to trade with, distant markets; and (3) differences in dietary patterns as well as

consumer preferences that affect the structure of local production. The different patterns of

specialization in Senegal compared with the rest of Sahelian West Africa and in Kenya compared

with other Eastern African countries illustrate the influence of these factors.

Consequently, we use a series of indicators to assess the actual degree of specialization in

agricultural production and trade, and whether there is real scope for transborder trade expansion

as a strategy to exploit the less-than-perfect correlation between national production levels to

reduce the vulnerability of domestic food markets to shocks. The first two indicators are the

production and export similarity indices, which measure and rank the relative importance of the

production and trading of individual agricultural products in every country. The two indices are

calculated for country pairs using the following formulas:

𝑆𝑄𝑖𝑗 = 100 ∑ 𝑀𝑖𝑛(𝑞𝑖𝑘, 𝑞𝑗𝑘)𝑘

𝑆𝐸𝑖𝑗 = 100 ∑ 𝑀𝑖𝑛(𝑒𝑖𝑘, 𝑒𝑗𝑘)𝑘

where 𝑆𝑄𝑖𝑗 and 𝑆𝐸𝑖𝑗 are the production and export similarity indices, respectively, 𝑞𝑖𝑘 and 𝑞𝑗𝑘 are

the shares of a product 𝑘 in the total agricultural production of countries 𝑖 and 𝑗, respectively, and

𝑒𝑖𝑘 and 𝑒𝑗𝑘 are the shares of a product 𝑘 in the total agricultural exports of countries 𝑖 and 𝑗,

respectively. The level of importance or position of each product is then compared for all relevant

pairs of countries within the region. 20 The indices have a maximum value of 100, which would

reflect complete similarity of production or trade patterns between the considered pair of countries.

The more the value of the indices tends towards zero, the greater the degree of specialization

between the two countries. Index values of around 50 and below are interpreted as indicating

patterns of specialization that are compatible with higher degrees of trade expansion possibilities.

Figures 7.6 and 7.7 present the results of the calculations covering 150 products in total. Each bar

represents the number of country pairs that fall within the corresponding range of index values.

The vast majority of country pairs fall within the 0-50 range.

174

A value of less than 60 is conventionally interpreted as compatible with higher trade exchange

between the considered pair of countries. The estimated index values therefore suggest that there

exists sufficient dissimilarity in current country production and trading patterns and hence scope

for trans-border trade expansion in the region.

Figure 7.6. Similarity of production patterns among ECOWAS countries (2007-2011)

Source: Authors’ calculations based on data from FAOSTAT, 2014.

Figure 7.7. Similarity of trading patterns among ECOWAS countries (2007-2011)

Source: Authors’ calculations based on data from FAOSTAT, 2014.

0

20

40

60

80

100

120

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100

Nu

mb

er

of

Co

un

try

Pai

rs

Production Similarity Index

0

100

200

300

400

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100

Nu

mb

er

of

Co

un

try

Pai

rs

Export Similarity Index

175

A third indicator, the revealed comparative advantage (RCA) index, is computed to further assess

the degree of trade specialization among countries within the region. It is calculated according to

the following formula (Balassa, 1965):

𝑅𝐶𝐴𝑖𝑗𝑘 =𝐸𝑖𝑗𝑘

∑ 𝐸𝑖𝑗𝑘𝑘

𝐸𝑤𝑗𝑘

∑ 𝐸𝑤𝑗𝑘𝑘⁄

where 𝐸𝑖𝑗𝑘 is the export value of an agricultural product 𝑘 from country 𝑖 to destination 𝑗 and

𝐸𝑤𝑗𝑘 = ∑ 𝐸𝑖𝑗𝑘 𝑖 is the world export value of the same product to the same destination.

The RCA index compares the share of a given product in a given country’s export basket with that

of the same product in total world exports. A value greater than 1 indicates that the considered

country performs better than the world average, and the higher the value is, the stronger the

performance of the country in exporting the considered product. Of the nearly 450 RCA indicators

estimated for various products exported by different ECOWAS countries, 73 percent have a value

higher than 1. Following Laursen (2000), the RCA index is normalized through the

formula 𝑁𝑅𝐶𝐴𝑖𝑗𝑘 = (𝑅𝐶𝐴𝑖𝑗𝑘 − 1) (𝑅𝐶𝐴𝑖𝑗𝑘 + 1)⁄ . Thus, the normalized RCA (NRCA) is

positive for RCA indicators that are greater than 1 and negative otherwise. For very high RCA

indicators, the normalized value tends towards 1. The 20 products with the highest normalized

RCA index values are presented in Table 7.3.

176

Table 7.3. List of the 20 products with highest normalized revealed comparative advantage index

values in ECOWAS countries, average 2007-2011

Commodity Country

Cashew nuts, with shell Guinea Bissau

Cake of Groundnuts Gambia

Groundnut oil Gambia

Cashew nuts, with shell Benin

Groundnuts Shelled Gambia

Cashew nuts, with shell Gambia

Groundnut oil Senegal

Copra Gambia

Cake of Groundnuts Senegal

Cake of Cottonseed Benin

Rubber Nat Dry Liberia

Cottonseed oil Togo

Cottonseed oil Benin

Sugar beet Gambia

Cashew nuts, with shell Cote d'Ivoire

Cotton Linter Benin

Cocoa beans Cote d'Ivoire

Cake of Groundnuts Togo

Cocoa Paste Cote d'Ivoire

Cocoa beans Ghana Source: Authors’ calculations based on FAOSTAT 2014

All the products listed in the table have normalized RCA values above 0.98. The rankings reflect

the degree of cross-country specialization within the ECOWAS region. For instance, a total of 12

products spread across 8 out of 15 member countries account for the highest 20 normalized RCA

indicator values for the region.

So far, the analysis has established the existence of dissimilar patterns of specialization in

production and trade of agricultural products among countries within ECOWAS. Two final

indicators, the Trade Overlap Indicator (TOI) and the Trade Expansion Indicator (TEI), are

calculated to examine the potential to expand trade within the region based on current trade

patterns. They measure how much of the same product a given country or region exports and

imports at the same time. The TOI measures the overall degree of overlapping trade flows for a

country or region as a whole, while the TEI measures the overlapping trade flows at the level of

individual products for a country or region.

177

The TOI and TEI are calculated as follows:

𝑇𝑂𝐼𝑖 = 2(∑ 𝑀𝑖𝑛(𝐸𝑖𝑘, 𝑀𝑖𝑘)𝑘 ) ∑ (𝐸𝑖𝑘 + 𝑀𝑖𝑘)𝑘⁄

𝑇𝐸𝐼𝑖𝑘 = 100 ∙ [𝑀𝑖𝑛(𝐸𝑖𝑘, 𝑀𝑖𝑘) 𝑀𝑎𝑥(𝐸𝑖𝑘, 𝑀𝑖𝑘)⁄ ]

where 𝐸𝑖𝑘 and 𝑀𝑖𝑘 denote the values of the exports and imports of an agricultural product 𝑘 by a

country 𝑖. The TOI varies between 0 and 1. It will be zero if the country only exports or imports

any individual products. It will be 1 in the unlikely situation in which the country both exports and

imports all traded products by an equal amount. As regards the TEI, it indicates the percentage of

the country’s exports (imports) of a product that are matched by the country’s imports (exports) of

the same product.

The results are presented in Figure 7.8 and Table 7.4. The Figure indicates that there is a

considerable degree of overlapping trade flows; 25 percent for Africa as whole and as much as 17

percent for the ECOWAS region. Normalized TOI values obtained by dividing country TOI values

by the TOI value for the region can be found in Badiane et al. (2014). In the vast majority of cases,

they are significantly less than 1. The overlapping regional trade flows must therefore be from

different importing and exporting countries. In other words, some countries are exporting

(importing) the same products that are being imported (exported) by other ECOWAS member

countries, but in both cases to and from countries outside the region. By redirecting such flows,

countries should be able to expand trans-border trade within the region.

The TEI indicates which products have the highest potential for increased trans-border trade based

on the degree of overlapping trade flows. Table 7.4 lists the 20 products with the highest TEI value

for the region. The lowest indicator value for any of the products is 0.41 and the average value is

0.56. RCA values for the same products, presented in Badiane et al. (2014), are all greater than 1,

except for fresh fruits. The fact that products with high TEI values also have high RCA indicator

values points to a real scope for trans-border trade expansion in the region.

178

Figure 7.8. Trade Overlap indicators, ECOWAS region, 2007-2011.

Source: Authors' calculations based on FAOSTAT 2014

Table 7.4. Trade Expansion Indicators, ECOWAS region, average 2007-2011

Commodity TEI value

Tobacco products 0.926

Fatty acids 0.763

Groundnuts, shelled 0.744

Hides, cattle, wet salted 0.681

Coffee, extracts 0.676

Fruit, fresh 0.62

Fruit, tropical fresh 0.592

Cigarettes 0.573

Tea, mate extracts 0.535

Oilseeds 0.524

Onions, dry 0.513

Oil, cottonseed 0.51

Pepper (piper spp.) 0.479

Margarine Short 0.456

Roots and tubers 0.454

Cereal preparations 0.439

Chickpeas 0.415

Vegetables fresh or dried Products 0.412

Fruit, prepared 0.412

Pineapple, canned 0.406 Source: Authors’ calculations based on FAOSTAT 2014. Note: Italics designate products with RCA < 1; six products

with high TEI but which are not being produced in the region are included, as they relate to re-export trade.

0

0.05

0.1

0.15

0.2

0.25

0.3

2007 2008 2009 2010 2011

Trad

e O

verl

ap I

nd

ex

Africa

ECOWAS

179

The findings above point to the existence of a real potential to expand intra-trade within ECOWAS

beyond current levels even with current production and trade patterns. The remainder of the

chapter therefore analyzes the outlook for intra-trade expansion and the expected impact on the

volatility of regional food markets over the next decade. This is done by simulating alternative

policy scenarios to boost intra-regional trade and comparing the effects on the level and volatility

of trade flows up to 2025 to historical trends and outcomes under a baseline scenario that would

continue these trends.

7.5. Regional trade simulation model

The preceding analysis presents evidence showing that ECOWAS countries could use increased

regional trade to enhance the resilience of domestic markets to supply shocks. The high cost of

moving goods across domestic and trans-border markets and outwardly biased trading

infrastructure are major determinants of the level and direction of trade among African countries.

A strategy to exploit the regional stabilization potential therefore has to include measures to lower

the general cost of trading and remove additional barriers to cross-border trade. This section

simulates the impact on regional trade flows of changes in that direction. Simulations of changes

are carried out using IFPRI’s regional Economy-wide Multimarket Model (EMM) described

below18. The original model is augmented in this study to account for intra- versus extra-regional

trade sources and destinations as well as informal versus formal trade costs in intra-regional trade

transactions. In its original version, the EMM solves for optimal levels of supply 𝑄𝑋𝑟 𝑐,

demand 𝑄𝐷𝑟 𝑐 and net trade (either import 𝑄𝑀𝑟 𝑐 or export 𝑄𝐸𝑟 𝑐) of different commodities 𝑐 for

individual member countries 𝑟 of the modelled region.

18 See Diao et al., 2007 and Nin-Pratt et al., 2010.

180

Supply and demand balance at the national level determines domestic output prices 𝑃𝑋𝑟 𝑐 as stated

by equation (1) while equation (2) connects domestic market prices 𝑃𝐷𝑟 𝑐 to domestic output

prices, taking into account an exogenous domestic marketing margin 𝑚𝑎𝑟𝑔𝐷𝑟 𝑐. The net trade of

a commodity in a country is determined through mixed complementarity relationships between

producer prices and potential export quantities, and between consumer prices and potential import

quantities. Accordingly, equation (3) ensures that a country will not export a commodity (𝑄𝐸𝑟,𝑐 =

0) as long as the producer price of that commodity is higher than its export parity price, where

𝑝𝑤𝑒𝑟 𝑐 is the country’s FOB price and 𝑚𝑎𝑟𝑔𝑊𝑟 𝑐 is an exogenous trade margin covering the cost

of moving the commodity from and to the border. If the domestic market balance constraint in

equation (1) requires that the country exports some excess supply of a commodity (𝑄𝐸𝑟,𝑐 > 0),

then the producer price will be equal to the export parity price of that commodity. Additionally,

equation (4) governs any country’s possibility to import a commodity, where 𝑝𝑤𝑚𝑟 𝑐 is its CIF

price. There will be no imports (𝑄𝑀𝑟,𝑐 = 0) as long as the import parity price of a commodity is

higher than the domestic consumer price. If the domestic market balance constraint requires that

the country imports some excess demand of a commodity (𝑄𝑀𝑟,𝑐 > 0), then the domestic

consumer price will be equal to the import parity price of that commodity.

𝑄𝑋𝑟 𝑐 + 𝑄𝑀𝑟 𝑐 − 𝑄𝐸𝑟 𝑐 = 𝑄𝐷𝑟 𝑐 (1)

𝑃𝑋𝑟 𝑐 ∙ (1 + 𝑚𝑎𝑟𝑔𝐷𝑟 𝑐) = 𝑃𝐷𝑟 𝑐 (2)

𝑃𝑋𝑟 𝑐 ≥ 𝑝𝑤𝑒𝑟 𝑐 ∙ (1 − 𝑚𝑎𝑟𝑔𝑊𝑟 𝑐) ⏊ 𝑄𝐸𝑟,𝑐 ≥ 0 (3)

𝑝𝑤𝑚𝑟 𝑐 ∙ (1 + 𝑚𝑎𝑟𝑔𝑊𝑟 𝑐) ≥ 𝑃𝐷𝑟 𝑐 ⏊ 𝑄𝑀𝑟,𝑐 ≥ 0 (4)

In the version used in this study, the net export of any commodity is modelled as an aggregate of

two output varieties differentiated according to their market outlets (regional and extra-regional)

while assuming an imperfect transformability between the two export varieties. Similarly, the net

import of any commodity is modelled as a composite of two varieties differentiated by their origins

(regional and extra-regional) while assuming an imperfect substitutability between the two import

varieties.

181

In order to implement export differentiation by destination, the mixed complementarity

relationship in equation (3) is replaced with two new equations which specify the price conditions

for export to be possible to both destinations. Equation (5) indicates that for export to extra-

regional market outlets to be possible (𝑄𝐸𝑍𝑟 𝑐 > 0), suppliers should be willing to accept for that

destination a price 𝑃𝐸𝑍𝑟 𝑐 that is not greater than the export parity price. Similarly, equation (6)

assures that export to within-region market outlets is possible (𝑄𝐸𝑅𝑟 𝑐 > 0) only if suppliers are

willing to receive for that destination a price 𝑃𝐸𝑅𝑟 𝑐 that is not more than the regional market

clearing price 𝑃𝑅𝑐 adjusted downward to account for exogenous regional trade margins 𝑚𝑎𝑟𝑔𝑅𝑟 𝑐

incurred in moving the commodity from the farm gate to the regional market. (See equation 17

below for the determination of 𝑃𝑅𝑐.)

𝑃𝐸𝑍𝑟 𝑐 ≥ 𝑝𝑤𝑒𝑟 𝑐 ∙ (1 − 𝑚𝑎𝑟𝑔𝑊𝑟 𝑐) ⏊ 𝑄𝐸𝑍𝑟 𝑐 ≥ 0 (5)

𝑃𝐸𝑅𝑟 𝑐 ≥ 𝑃𝑅𝑐 ∙ (1 − 𝑚𝑎𝑟𝑔𝑅𝑟 𝑐) ⏊ 𝑄𝐸𝑅𝑟 𝑐 ≥ 0 (6)

Subject to these price conditions, equations (7) – (10) determine the aggregate export quantity and

its optimal allocation to alternative destinations. Equation (7) indicates that the aggregate export

of a commodity by individual countries 𝑄𝐸𝑟 𝑐 is obtained through a constant elasticity of

transformation (CET) function of the quantity 𝑄𝐸𝑍𝑟 𝑐 sold on extra-regional market outlets and

the quantity 𝑄𝐸𝑅𝑟 𝑐 sold on intra-regional market outlets, where 𝜌𝑟 𝑐𝑒 , 𝛿𝑟 𝑐

𝑒 and 𝛼𝑟 𝑐𝑒 represent the

CET function exponent, share parameter and shift parameter, respectively. Equation (8) is the first-

order condition of the aggregate export revenue maximization problem, given the prices suppliers

can receive for the different export destinations and subject to the CET export aggregation

function. It says that an increase in the ratio of intra-regional to extra-regional destination prices

will increase the ratio of intra-regional to extra-regional export quantities, i.e. a shift toward the

export destination that offers the higher return. Equation (9) helps identify the optimal quantities

supplied to each destination; it states that aggregate export revenue at producer price of export

𝑃𝐸𝑟 𝑐 is the sum of export sales revenues from both intra-regional and extra-regional market outlets

at supplier prices, while equation (10) sets the producer price of export to be the same as the

domestic output price 𝑃𝑋𝑟 𝑐, which is determined through the supply and demand balance equation

(1) as earlier explained.

182

𝑄𝐸𝑟 𝑐 = 𝛼𝑟 𝑐𝑒 ∙ (𝛿𝑟 𝑐

𝑒 ∙ 𝑄𝐸𝑅𝑟 𝑐𝜌𝑟 𝑐

𝑒

+ (1 − 𝛿𝑟 𝑐𝑒 ) ∙ 𝑄𝐸𝑍𝑟 𝑐

𝜌𝑟 𝑐𝑒

)

1𝜌𝑟 𝑐

𝑒

(7)

𝑄𝐸𝑅𝑟 𝑐

𝑄𝐸𝑍𝑟 𝑐= (

𝑃𝐸𝑅𝑟 𝑐

𝑃𝐸𝑍𝑟 𝑐∙

1 − 𝛿𝑟 𝑐𝑒

𝛿𝑟 𝑐𝑒 )

1𝜌𝑟 𝑐

𝑒 −1 (8)

𝑃𝐸𝑟 𝑐 ∙ 𝑄𝐸𝑟 𝑐 = 𝑃𝐸𝑅𝑟 𝑐 ∙ 𝑄𝐸𝑅𝑟 𝑐 + 𝑃𝐸𝑍𝑟 𝑐 ∙ 𝑄𝐸𝑍𝑟 𝑐 (9)

𝑃𝐸𝑟 𝑐 = 𝑃𝑋𝑟 𝑐 (10)

Import differentiation by origin is implemented following the same treatment as described above

for export differentiation by destination. Equation (4) is replaced with equations (11) and (12).

Accordingly, import from the extra-regional origin will happen (𝑄𝑀𝑍𝑟,𝑐 > 0) only if domestic

consumers are willing to pay for the extra-regional variety at a price 𝑃𝑀𝑍𝑟 𝑐 that is not smaller

than the import parity price. Furthermore, import from the intra-regional origin is possible

(𝑄𝑀𝑅𝑟,𝑐 > 0) only if domestic consumers are willing to pay for the intra-regional variety at a price

𝑃𝑀𝑅𝑟 𝑐 that is not smaller than the regional market clearing price 𝑃𝑅𝑐 adjusted upward to account

for exogenous regional trade margins 𝑚𝑎𝑟𝑔𝑅𝑟 𝑐 incurred in moving the commodity from the

regional market to consumers.

𝑝𝑤𝑚𝑟 𝑐 ∙ (1 + 𝑚𝑎𝑟𝑔𝑊𝑟 𝑐) ≥ 𝑃𝑀𝑍𝑟 𝑐 ⏊ 𝑄𝑀𝑍𝑟,𝑐 ≥ 0 (11)

𝑃𝑅𝑟 ∙ (1 + 𝑚𝑎𝑟𝑔𝑅𝑟 𝑐) ≥ 𝑃𝑀𝑅𝑟 𝑐 ⏊ 𝑄𝑀𝑅𝑟 𝑐 ≥ 0 (12)

Under these price conditions, equation (13) represents aggregate import quantity 𝑄𝑀𝑟 𝑐 as a

composite of intra- and extra-regional import variety quantities 𝑄𝑀𝑅𝑟 𝑐 and 𝑄𝑀𝑍𝑟 𝑐, respectively,

using a constant elasticity of substitution (CES) function, with 𝜌𝑟 𝑐𝑚 , 𝛿𝑟 𝑐

𝑚 and 𝛼𝑟 𝑐𝑚 standing for the

CES function exponent, share parameter and shift parameter, respectively. The optimal mix of the

two varieties is defined by equation (14), which is the first-order condition of the aggregate import

cost minimization problem, subject to the CES aggregation equation (13) and given import prices

from both origins. An increase in the ratio of extra-regional to intra-regional import prices will

increase the ratio of intra-regional to extra-regional import quantities, i.e. a shift away from the

import origin that becomes more expensive. Equation (15) identifies the specific quantities

imported from each origin. It defines total import cost at the consumer price of imports 𝑃𝑀𝑟 𝑐 as

the sum of intra-regional and extra-regional import costs, while equation (16) sets the consumer

price of imports to be the same as the domestic market price 𝑃𝐷𝑟 𝑐, which is determined through

equations (1) and (2) as earlier explained.

183

𝑄𝑀𝑟 𝑐 = 𝛼𝑟 𝑐𝑚 ∙ (𝛿𝑟 𝑐

𝑚 ∙ 𝑄𝑀𝑅𝑟 𝑐−𝜌𝑟 𝑐

𝑚

+ (1 − 𝛿𝑟 𝑐𝑚 ) ∙ 𝑄𝑀𝑍𝑟 𝑐

−𝜌𝑟 𝑐𝑚

)−

1𝜌𝑟 𝑐

𝑚

(13)

𝑄𝑀𝑅𝑟 𝑐

𝑄𝑀𝑍𝑟 𝑐= (

𝑃𝑀𝑍𝑟 𝑐

𝑃𝑀𝑅𝑟 𝑐∙

𝛿𝑟 𝑐𝑚

1 − 𝛿𝑟 𝑐𝑚 )

11+𝜌𝑟 𝑐

𝑚

(14)

𝑃𝑀𝑟 𝑐 ∙ 𝑄𝑀𝑟 𝑐 = 𝑃𝑀𝑅𝑟 𝑐 ∙ 𝑄𝑀𝑅𝑟 𝑐 + 𝑃𝑀𝑍𝑟 𝑐 ∙ 𝑄𝑀𝑍𝑟 𝑐 (15)

𝑃𝑀𝑟 𝑐 = 𝑃𝐷𝑟 𝑐 (16)

Having determined export quantities and prices by destination and import quantities and prices by

origin, the regional market clearing price 𝑃𝑅𝑐 can now be solved. Equation (17) imposes the

regional market balance constraint by equating the sum of intra-regional export supplies to the sum

of intra-regional import demands, with 𝑞𝑑𝑠𝑡𝑘𝑐 standing for discrepancies existing in observed

aggregate intra-regional export and import quantity data in the model base year. Thus, 𝑃𝑅𝑐 is

determined as the price that ensures the regional market balance.

∑ 𝑄𝐸𝑅𝑟 𝑐

𝑟

= ∑ 𝑄𝑀𝑅𝑟 𝑐 + 𝑞𝑑𝑠𝑡𝑘𝑐 (17)

𝑟

Calibration is performed so as to replicate, for every member country within the region, the same

production, consumption and net trade data as observed for different agricultural subsectors and

two non-agricultural sub-sectors in 2007–2008. Baseline trend scenarios are then constructed such

that, until 2025, changes in crop yields, cultivated areas, outputs, and GDP reflect the same

observed changes. Although the model is calibrated to the state of national economies seven years

earlier, it reproduces closely the countries’ current growth performance.

Four different scenarios are simulated using the EMM. The first is the baseline scenario described

above which assumes a continuation of current trends up to 2025. It is used later as a reference to

evaluate the impact of changes under the remaining three scenarios. The latter scenarios introduce

the following three different sets of changes to examine their impacts on regional trade levels: a

reduction of 10 percent in the overall cost of trading across the economy; a removal of all

harassment costs, that is, a reduction of their tariff equivalent to zero; and an across the board 10

percent increase in yields. These changes are to take place between 2008, the base year, and 2025.

The change in cross-border exports is used as an indicator of the impact on intra-regional trade. In

the original data, there are large discrepancies between recorded regional exports and import

levels, the latter often being a multiple of the former. The more conservative export figures are

therefore the preferred indicator of intra-regional trade.

184

7.6. Intra-trade simulation results

The results are presented in Figures 7.9 and 7.10. The results of the baseline scenarios from 2008

to 2025 are shown in Figure 7.9. Assuming a continuation of current trends, intra-regional trade in

ECOWAS is expected to expand rapidly but with marked differences between crops. The

aggregate volume of intra-regional trade in staples would approach 3 million tons in the case if the

current rates of growth in yields, cultivated areas, population and income are sustained to 2025.

Cereals would see the smallest gains, while trade in roots and tubers as well as other food crops

would experience much faster growth. This is in line with the current structure of and trends in

commodity demand and trade. While the increase in demand for roots and tubers is being met

almost exclusively from local sources, the fast growing demand in cereals is heavily tilted towards

rice, which is supplied from outside of the region. The two leading cereals that are traded

regionally, maize and millet, therefore benefit less from the expansion of regional demand and

have historically seen slower growth in trade than roots and tubers.

Figure 7.9. Regional exports outlook, baseline

Source: Authors’ calculations.

0

500

1000

1500

2000

2500

3000

2008 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 2025

Tho

usa

nd

met

ric

to

ns

CEREALS ROOTS OTHER FOODS ALL FOOD CROPS OTHER CROPS

185

The graphs in Figure 7.10 show the cumulated changes in intra-regional export levels by 2025

compared to the baseline that would result from a reduction in total trading costs, removal of

harassment costs, and an increase in yields. The bars represent the proportional changes in percent

and the numbers on top of the bars indicate the corresponding absolute changes in 1000 metric

tons. The results invariably show considerable increases in intra-regional trade in cereals and roots

and tubers, the main food crops, in response to changes in trading costs and yields. Intra-

community trade levels in ECOWAS climb by between 10 and 35 percent for most products over

the entire period. The volume of cereal trade increases by a cumulative total of between 200,000

and 300,000 tons for individual products and that of overall trade in staples by between 1.5 and

4.0 million tons by 2025, compared to baseline trends. Cereals seem to respond better than other

products in general. It also appears that removal of harassment costs would have the strongest

impact on trade flows across the board. Table 7.5 below shows how individual countries are

affected in the different scenarios. Countries respond more significantly to the removal of

harassment costs than to the reduction of normal trade costs, except for Benin, Guinea Bissau,

Niger and Sierra Leone, which appear to be more responsive to increases in crop yields than to

reductions in normal trading costs or harassment costs.

Figure 7.10. The impact of changes in trade costs and yields on regional exports

Source: Authors’ calculations.

1801183

101

14641809

284

1154

2560

3999

5032

259

1290

995

2544

4706

0

5

10

15

20

25

30

35

40

CEREALS ROOTS & TUBERS OTHER FOOD CROPS

ALL FOOD CROPS OTHER CROPS

% o

f b

asel

ine

qu

anti

ty

10% reduction in trade costs Removal of harassment costs 10% increase in crop yields

186

Table 7.5. Country-level impact of changes in trade costs and yields on regional exports of staple food

crops

10% reduction in trade

costs

Removal of harassment

costs

10% increase in crop

yields

Benin 27.6 18.2 39.5

Burkina Faso 22.2 34.9 39.1

Chad 22.5 39.1 33.9

Côte d’Ivoire 8.9 17.7 14.2

Gambia 1.9 8.5 5.3

Ghana 5.7 24.1 15.5

Guinea 4.7 32.0 16.2

Guinea Bissau 51.1 37.1 91.5

Liberia 9.0 34.2 22.1

Mali 4.6 21.6 10.5

Mauritania 17.5 33.2 28.6

Niger 80.8 1.4 289.6

Nigeria 26.0 32.9 46.3

Senegal 10.6 32.6 25.3

Sierra Leone 93.4 40.3 117.6

Togo 6.6 32.1 21.1

Source: Authors’ calculations.

7.7. Regional market volatility under alternative policy scenarios

Under each scenario, model simulated quantities of intra-regional exports 𝑄𝐸𝑅𝑟 𝑐 are used to

estimate an index of future export volatility at the country and regional levels as follows. First, a

trend-corrected coefficient of variation 𝑇𝐶𝑉 is calculated for each country, using the same formula

as in section 7.3:

𝑇𝐶𝑉𝑖 = 𝐶𝑉𝑖 ∙ √(1 − 𝑅𝑖2) where 𝐶𝑉𝑖 is the coefficient of variation in the series of the intraregional

exports of staple food crops by a country 𝑖 from 2008 to 2025 and 𝑅𝑖2 is the adjusted coefficient of

determination of the linear trend model fitted to the series.

187

Then an index of regional volatility 𝑇𝐶𝑉𝑟𝑒𝑔 is derived for the ECOWAS region as a weighted

average of trend corrected coefficients of variation of its member countries with the formula

𝑇𝐶𝑉𝑅𝐸𝐶2 = ∑ 𝑠𝑖

2 ∙ 𝑇𝐶𝑉𝑖2

𝑛

𝑖

+ 2 ∑ ∑ 𝑠𝑖 ∙ 𝑠𝑗 ∙ 𝑣𝑖𝑗 ∙ 𝑇𝐶𝑉𝑖 ∙ 𝑇𝐶𝑉𝑗

𝑛

𝑗

𝑛

𝑖

where 𝑇𝐶𝑉𝑖 and 𝑇𝐶𝑉𝑗 are the trend-corrected coefficients of variation in the exports of staple food

crops in countries 𝑖 and 𝑗, 𝑛 is the number of ECOWAS member countries, 𝑠𝑖 and 𝑠𝑗 are the shares

of countries 𝑖 and 𝑗 in the region’s overall intra-regional exports of staple food crops, and 𝑣𝑖𝑗 is

the coefficient of correlation between the food crop exports of countries 𝑖 and 𝑗. Finally, the

coefficients of variation at the country level are normalized by dividing them by the regional

coefficient. The historical and simulated levels of volatility of cross-border trade in food staples in

the region under historical trends and each of the alternative scenarios are reported in Table 7.6.

Volatility levels under historical trends are calculated based on bilateral export volumes from the

TradeMaps database (1996-2012). In Table 7.7, simulated volatility levels under the various

scenarios are compared with the historical levels of volatility, with the difference expressed in

point changes. As can be seen from the figures in the two tables, regional cross-border trade

volatility decreases with a reduction of overall trading costs but rises under the removal of cross-

border trade barriers or with increases in yields. The magnitude of the changes are, however, rather

small across all three scenarios. The figures also show that under the continuation of current trends

of rising volumes of intra-regional trade, the volatility level in the region is expected to decline

compared to historical trends. A better comparison is therefore to contrast changes under the two

trade policy scenarios and the productivity scenario with expected volatility levels under the

baseline scenario. Furthermore, the direction and magnitude of changes in the level of intra-

regional trade volatility are determined by the combined effect of changes in the level of volatility

as well as the shares of cross-border exports by individual countries. Figure 7.11 below shows

changes in volatility levels (x-axis) and shares of exports (y-axis) by individual countries under

each of the trade and productivity scenarios compared to the baseline. The different dots indicate

the position of different countries under the three scenarios. The tilted distribution of country

positions to the left of the x-axis indicates that exports by most countries would experience a lower

level of volatility under regional policies that would reduce the overall cost of trading, eliminate

188

administrative and regulatory obstacles to trans-border trade, or raise yields of staple crops in

member countries.

Table 7.6. Volatility in cross-border exports of staple food products within ECOWAS

Historical trend (1996-2012)

Baseline trend (2008-2025)

10% reduction in trade costs (2008-2025)

Removal of harassment costs (2008-2025)

10% increase in crop yields (2008-2025)

Benin 1.753 0.703 0.629 0.660 0.618

Burkina Faso 1.269 1.566 1.353 1.643 1.539

Cape Verde 2.802

Cote d’Ivoire 0.285 0.657 0.531 0.631 0.591

Gambia 1.585 1.546 1.379 1.291

Ghana 2.145 0.214 0.191 0.135 0.126

Guinea 1.347 0.538 0.540 0.698 0.654

Guinea Bissau 2.101 2.188 2.156 2.020

Liberia 0.521 0.520 0.656 0.615

Mali 0.856 1.107 1.138 1.164 1.090

Niger 2.011 1.913 2.004 1.785 1.672

Senegal 0.926 0.029 0.048 0.166 0.155

Sierra Leone 2.741 3.407 2.667 2.499

Togo 0.863 1.492 1.574 1.641 1.538

ECOWAS 0.345 0.330 0.323 0.354 0.378

Source: Authors’ calculations from the TradeMaps database and EMM model simulation results.

Table 7.7. Change in trade volatility under alternative scenarios (2008-2025)

Baseline trend 10% reduction in trade costs

Removal of harassment costs

10% increase in crop yields

Point change compared to historical trend

Benin -1.050 -1.124 -1.093 -1.135

Burkina Faso 0.297 0.084 0.374 0.270

Cote d’Ivoire 0.372 0.246 0.346 0.307

Ghana -1.931 -1.954 -2.010 -2.019

Guinea -0.809 -0.807 -0.649 -0.693

Mali 0.251 0.282 0.307 0.234

Niger -0.098 -0.007 -0.226 -0.339

Senegal -0.897 -0.878 -0.760 -0.770

Togo 0.629 0.711 0.779 0.675

ECOWAS -0.015 -0.022 0.009 0.033 Source: Authors’ calculations from the TradeMaps database and EMM model simulation results.

189

Figure 7.11. Changes in country export shares and volatility compared to baseline trends

Source: Authors’ calculations from the TradeMaps database and EMM model simulation results.

The combined changes in export share and volatility for individual countries under each of the

scenarios are reported in Table 7.8. Changes in country production patterns resulting from the

simulated policy actions lead to changes in both the volatility as well as the level of exports and

hence the shares in regional trade for each country. The magnitude and direction of these changes

determine the contribution of individual countries to changes in the level of volatility in regional

food markets.

-5.0

-4.0

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

4.0

-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2

Change in export

share in % points

Change in export

volatility in points

10% reduction in trade cost Removal of harassment costs 10% increase in crop yields

190

Table 7.8. Change in volatility and share of staple exports under alternative scenarios, 2008-2025

Change in volatility compared to

baseline (points)

Change in share compared to baseline

(% points)

10%

reduction

in trade

cost

Removal of

harassment

costs

10%

increase in

crop yields

10%

reduction

in trade

cost

Removal of

harassment

costs

10%

increase in

crop yields

Benin -0.1 0.0 -0.1 2.8 -0.3 2.4

Burkina Faso -0.2 0.1 0.0 0.4 0.5 0.5

Cote d’Ivoire -0.1 0.0 -0.1 -0.4 0.4 -0.8

Gambia 0.0 -0.2 -0.3 0.0 0.0 -0.1

Ghana 0.0 -0.1 -0.1 -0.6 0.2 -0.7

Guinea 0.0 0.2 0.1 -0.1 0.1 -0.2

Guinea Bissau 0.1 0.1 -0.1 0.0 0.0 0.0

Liberia 0.0 0.1 0.1 0.0 0.0 0.0

Mali 0.0 0.1 0.0 -3.1 0.1 -4.5

Niger 0.1 -0.1 -0.2 1.1 -1.1 3.2

Senegal 0.0 0.1 0.1 0.0 0.0 0.0

Sierra Leone 0.7 -0.1 -0.2 0.1 0.0 0.0

Togo 0.1 0.1 0.0 0.0 0.0 0.0 Source: Authors’ calculations from the TradeMaps database and EMM model simulation results.

7.8. Conclusion

The current chapter has examined the potential to use increased intra-regional trade among West

African countries as a means to raise the resilience of domestic food markets to shocks. The

distribution and correlation of production volatility as well as the current patterns of specialization

in the production and trade of agricultural products among West African countries suggest that it

is indeed possible to increase cross-border trade to reduce the level of instability of local food

markets. The results of the baseline scenario indicate that continuation of recent trends would

sustain the expansion of intra-regional trade flows in the ECOWAS region. The findings also

reveal that it is possible to significantly boost the pace of regional trade expansion, which in turn

would contribute to creating more resilient domestic food markets, through a modest reduction in

the overall cost of trading, a similarly modest increase in crop yields, or the removal of barriers to

transborder trade. More importantly, the simulation results also suggest that such policy actions to

promote transborder trade would reduce volatility in regional markets and help lower the

vulnerability of domestic food markets to shocks.

191

7.9. References

Badiane, O., Makombe, T., & Bahiigwa, G. (2014). Promoting Agricultural Trade to Enhance Resilience

in Africa. ReSAKSS Annual Trends and Outlook Report. Washington, DC: International Food Policy

Research Institute.

Badiane, O. (1988). National Food Security and Regional Integration in West Africa. Kiel, Germany:

Wissenschaftsverlag Vauk.

Balassa, B. (1965). Trade liberalisation and “revealed” comparative advantage. The Manchester School,

33(2), 99–123.

Bouët, A, & Laborde Debucquet, D. (2015). Global trade patterns, competitiveness, and growth outlook.

In O. Badiane, T. Makombe, & G. Bahiigwa (Eds.), Promoting Agricultural Trade to Enhance Resilience

in Africa. ReSAKSS Annual Trends and Outlook Report. Washington, DC: International Food Policy

Research Institute.

Cuddy, J. D. A., & Della Valle, P. A. (1978). Measuring the instability of time series data. Oxford

Bulletin of Economics and Statistics, 40(1), 79-85.

Diao, X., Fekadu, B., Haggblade, S., Taffesse, A. S., Wamisho, K., & Yu, B. (2007). Agricultural Growth

Linkages in Ethiopia: Estimates using Fixed and Flexible Price Models. IFPRI Discussion Paper 00695.

Washington, DC: International Food Policy Research Institute.

Koester, U. (1986). Regional Cooperation to Improve Food Security in Southern and Eastern African

Countries. IFPRI Research Report 53. Washington, DC: International Food Policy Research Institute.

Laursen, K. (2000). Trade Specialisation, Technology and Economic Growth: Theory and Evidence from

Advanced Countries. Cheltenham: Edward Elgar.

Nin-Pratt, A., Johnson, B., Magalhaes, E., You, L., Diao, X., & Chamberlain, J. (2011). Yield Gaps and

Potential Agricultural Growth in West and Central Africa. Washington, DC: International Food Policy

Research Institute.

Chapter 8. Summary and conclusions

Extracted from

African Agricultural Trade Status Report

2017

192

CHAPTER 8. SUMMARY AND CONCLUSIONS

The African Agricultural Trade Status Report (TSR) has examined recent trends, current status,

and future outlook for African agricultural trade in global and regional markets. The report’s five

substantive chapters provide descriptive assessments of trade patterns as well as econometric

analyses of the drivers of observed trends. In this concluding chapter, we summarize the findings

of the preceding chapters and draw general conclusions and policy recommendations.

Chapter two reviews trends in Africa’s global agricultural trade since 1998. The chapter finds that

although exports have increased over the period, imports have increased more rapidly, leading to

a growing trade deficit. The increase in imports is due to demographic changes as well as the low

competitiveness of domestic producers. Despite the increase in agricultural exports, the share of

agricultural exports in Africa’s total exports has declined by half over the period, due to more

rapidly rising exports in minerals and oil. Africa’s agricultural exports show signs of moderate

diversification over the period, while imports have remained fairly stable. The EU remains Africa’s

top trading partner, but both imports from and exports to the EU have dropped over the period,

while trade with Asia has increased; Asia is likely to take the EU’s place as Africa’s top trading

partner if these trends continue. Recent efforts to pursue increased economic integration have

resulted in significantly increased intra-regional trade during the period, although the overall level

of intra-regional trade remains low.

Chapter three examines patterns in intra-regional trade at the continental level and among major

RECs, namely ECOWAS, ECCAS, COMESA, and SADC. The chapter finds that intra-African

trade has expanded significantly since 1998, increasing at about 12 percent per year. The largest

increase took place in the ECCAS region, while the slowest increase was in the SADC region. The

chapter finds that ECOWAS shows the highest regional trade integration, as measured by the ratio

of intra-REC trade to the REC’s trade with Africa; ECCAS shows the lowest. COMESA and

SADC play larger roles as destinations for and origins of African trade than do the other two RECs.

Chapter four reviews the changes in competitiveness of exports of different countries and different

agricultural products over the past three decades, and investigates the determinants of these

changes through econometric analysis.

193

The chapter finds that most RECs saw their member countries increase or maintain their

competitiveness in global and regional markets, with the exception of ECCAS, whose member

countries tended to lose competitiveness. Improvements in the competitiveness of COMESA,

ECOWAS and SADC member countries took place primarily in intra-regional markets. The

majority of African export commodities gained competitiveness in global markets, with some

exceptions; however, the most competitive commodities accounted for fairly small export shares,

suggesting potential for expanding exports by leveraging competitiveness gains. The chapter finds

that determinants of competitiveness improvements include the ease of doing business,

institutional quality, the size of the domestic market, and the quality of customs.

Chapter five examines the factors contributing to Africa’s improved agricultural export

performance, using a gravity model to assess the importance of different determinants of trade and

of the constraints to further improving exports. The study finds that supply side constraints,

including production capacity and the cost of trade, affect trade performance to a greater extent

than demand side constraints, which include trade policies and agricultural supports in importing

countries. This suggests a focus on removing domestic constraints to increased trade. The chapter

also finds that non-tariff barriers to trade are increasing and present larger obstacles to exports than

do tariffs. The chapter highlights the potential of regional economic communities to promote the

removal of barriers to trade at both the regional and global levels, as well as the continued

importance of global cooperation to facilitate trade.

Chapter six examines the potential of increased intra-regional trade in West Africa, the feature

region of this report, to stabilize domestic food markets in the region. The chapter finds that the

distribution of production volatility among West African countries suggests significant potential

to lessen the impacts of domestic shocks through increased regional trade, while patterns in

agricultural production and trade show scope for increasing regional trade levels. Analysis of a

simulation model shows that intra-regional trade is expected to increase under current trends. Intra-

regional trade growth can be accelerated through small reductions in trading costs, small increases

in crop yields, or a reduction in trade barriers. The increased intra-regional trade resulting from

these changes would reduce food price volatility in regional markets.

The TSR chapters demonstrate undeniable improvements in Africa’s trade performance over the

past decade and a half, in both global and regional markets, as reflected by generally increasing

194

competitiveness for the majority of countries and commodities. However, progress has been

uneven, with some regions and countries consistently underperforming others. Challenges remain

in further enhancing Africa’s competitiveness on the global market and in increasing intra-regional

trade, which remains below its potential despite significant recent improvements. The findings of

chapter four point to the importance of the institutional and business environment in improving a

country’s export competitiveness, while chapter five also emphasizes the role of domestic factors

in increasing exports, including production capacity and trading costs. Chapter six focuses on the

West Africa region, demonstrating the role of potential domestic and regional policy actions to

increase intra-regional trade and enhance the stability of regional markets.

The chapters suggest a series of recommendations for policymakers, including efforts at the

country and regional level to increase agricultural productivity along the value chain, improve

market access, and improve the functioning of institutions; regional actions to enhance economic

integration; and continent-wide efforts to promote trade facilitation in international negotiations.

Policy actions such as these can influence the trends described in this report and accelerate

improvements in Africa’s trade performance, thereby increasing incomes and improving food

security across the continent.