Executive summary African Agricultural Trade Status Report 2017
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.