A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Africa’s Development And The Global Trading System: Challenges And Options

Working paper by OSABUOHIEN, Evans S. / Covenant University, Nigeria and EGWAKHE, Johnson A. /Adventist University of Central Africa, Rwanda, 2011

Download original document (English)

This paper empirically explores development in Africa in relation to the global trading system using panel data techniques. It examines the economic development of African economies in relation to the countries’ regional grouping and also assesses Random Effects (RE) and Generalised Method of Moments (GMM) estimates. The results establish, among other things, that sub-regions with higher level of domestic investment had higher values in economic development indicators.

AFRICA’S DEVELOPMENT AND THE GLOBAL TRADING SYSTEM:
CHALLENGES AND OPTIONS






By


Dr. Evans S. OSABUOHIEN,
Department of Economics and Development Studies,


Covenant University, Ota, Ogun State, Nigeria
E-mail: stephen.osabuohien@covenantuniversity@edu.ng and pecos4eva@gmail.com


Tel: +2348028858727


&


Dr. Johnson A. EGWAKHE
MBA Coordinator, Adventist University of Central Africa,


Masoro, Kigali, Rwanda.






PAPER SUBMITTED FOR FLACSO-WTO CHAIR AWARD,
AYACUCHO 551, 1426 CABA, ARGENTINA


SEPTEMBER, 2011





1


AFRICA’S DEVELOPMENT AND THE GLOBAL TRADING SYSTEM:
CHALLENGES AND OPTIONS




Abstract
This paper empirically explored development in Africa in relation to the global trading
system (trade and tariff) using panel data technique. It examined the economic
development of African economies in relation to the countries’ regional grouping and also
assessed Random Effects (RE) and Generalised Method of Moments (GMM) estimates.
The results established, among other things, that sub-regions with higher level of
domestic investment had higher values in economic development indicators. It was also
found that domestic investment and labour played a more crucial role in Africa’s economic
development process. The challenge noted was the fact that increased trade integration
do not translate to increased economic development in Africa. Thus the option is to
improve domestic investment and enhance labour productivity, which are more crucial for
economic development than trade and tariff.


Key words: Economic development, Export, Tariff, Trade.
JEL Codes: F11, F31.









2


1 Introduction
The impact of the global trading system trade has generated intensive debate among
academic commentators, but the impact on developing countries especially within Africa
has helped to fuel the contention. Some scholars advocated for trade liberalisation as a
prerequisite for economic growth (Edwards 1997; David and Scott 2005). Stiglitz (2002)
cautioned against drastic trade openness. Along these perceptions disparities, the African
commodities oriented countries are in dilemma as World Trade Organization (WTO)’s
tariff crusade makes the continent an economic-lake; importing and consuming a variety
of products without significant improvement in exports.


Many African countries have instituted strategic vision aiming at 2020 as the
developmental hallmark especially with a view of improving the welfare of their citizens
and attaining meaningful development. However, some countries anchored their
developmental strategy on the expansion of primary product exports, which may not yield
much result. This is because the price of primary products at the international market is
easily affected by unfavourable terms of trade. The occurrence of recent global economic
crisis that has resulted in decrease in commodity prices lends credence to this stance.
This, among other reasons, may be instrumental in development ‘lukewarmness’
experienced in many African countries even when most of them have had 50 years of
political independence. For instance, while many countries of the world have recorded
steady improvements with regards to human development in the past years, many African
countries suffered human development reversals from which they are yet to recover
(Human Development Report -HDR 2009).


Tariff and trade may or may not favour countries at the same rate due to difference in
economic structures. This should be understood as a system error, that economic
development is both interrelated and interactive. The advantage experienced by the
developed countries within the international trading system, which has made
improvements in the welfare of their citizens create an element of hierarchy within the
global system. Hence, instead of comparative advantage, it is somewhat selective-
advantage while there could be possibilities of complementary trade-off. The current trade
exhibits characteristics of have-nots for Africa despite participation in trade. Hence, the




3


promise of world trade benefits for Africa remains unfulfilled. For instance, the total trade
for services for Africa in 2007 was 3 per cent compared to that of Asia of 24 per cent and
average contribution of Sub-Saharan Africa (SSA) to world total export in 2008 was
modicum of 0.05 per cent compared to the global average of 0.64 per cent (World Bank
Group 2010).


The above is crucial given the targets in Millennium Developmental Goals (MDGs). Some
African countries have made some efforts in that regard but much success has not been
recorded in many respects especially in economic and human development. Africa had
some signs of impressive economic performance as the real per capita GDP in the first
half of 1970s where Africa’s 2.7 per cent growth rate (1970 to 1974) was similar to that of
Latin America and Caribbean and even higher than South Asia. However, Africa has had
some disappointing records in development in recent times. For instance, Africa had the
lowest value of real per capita GDP for the period 2005 to 2008 compared to other regions
of the world (World Bank Group 2010). On a similar note, human development index (HDI)
for Africa was in the very low ebb compared to other regions of the world. The average
HDI for Sub-Saharan Africa was 0.389, which was far lower than world average of 0.624
in 2010 (HDR 2010). This paper was motivated with a view to examining the extent to
which trade and tariff issues are relevant in explaining economic development in Africa.


In articulating this, 50 African countries were selected, which covered good representation
of the five sub-regions, namely: Central, East, North, Southern and West Africa. Data
sourced from HDR, World development Indicators (WDI), and World Trade Indicators
(WTI) were analysed using both descriptive and econometric techniques. The rest of the
paper is structured as follows: some stylised facts; theoretical underpinning and literature
review; empirical model formulation; presentation of econometric results and analyses;
and conclusion.




2 Stylised Facts on Development, Tariff and Trade


Table 1 gives some stylised facts on trade and development nexus across some regions
of the world including Africa from 1970 to 2008 using some period averages. This is with a




4


view to further situate the nexus on development, trade and tariff in Africa within the global
context. The information on tariff was for the period 1995 to 2008 while that of HDI was
1980 to 2010; periods that data were available. Trade outcome indicators using total
trade integration ratio (trade as % of GDP) and export integration ratio (export as % of
GDP) are shown in Part A and B of Table 1, while growth rates of real per capita GDP and
population are in Part C and D, respectively. Segments E and F report data on tariff and
HDI in Africa in comparison to other regions and global average.




From Table 1, Africa’s trade in the early 1970s was relatively at par with some other
regions presented and but it was even above South Asia using both indicators. There
were little fluctuations along the line; however, during the period 2000 to 2008, Africa
seems to have integrated very much in trade as its value of 38.28 per cent was very close
to that of Latin American and the Caribbean with the value of 38.48 per cent. It was even
above that of South Asia that has the value of 29.33 per cent. More so, the value of
Africa’s real per capita income for the period 1970 to 1974 was far above that of South
Asia (almost five times) and a little lower than that of Latin America and the Caribbean.
The irony is that towards the end of the period (2005 to 2008) Africa had the lowest real
per capita income amongst all the regions. The above is converse to its population growth
rate that has the highest value all through the period.




With respect to tariff, segment E of Table 1 indicates that the average applied tariff had
consistently reduced in Africa. For instance, the average for 1995-1994 was 22.44 per
cent which significantly reduced to 12.79 per cent for the period 2005 to 2008. In addition,
this paper presents another indicator of development, which incorporates the human
aspect of development (HDI). The HDI is composite measure of human development
entailing measures of living a long and healthy life (life expectancy), level of education
(adult literacy rate, enrolment rate and years of schooling) and decent living standard
(gross national income per capita at purchasing power parity). As can be seen segment F
of Table 1, values HDI for Africa was lower than those of other regions as well as the
global average all through the period 1980 to 2010.




5


Table 1 Development, Trade and Tariff Indicators across the World


Period ⇒ 1970-74 1975-79
1980-


84
1985-


79
1990-


94
1995-


99
2000-


04 2005-08 1970-08
Regions ⇓ (A) Trade as % GDP
East Asia & Pacific 64.51 70.23 77.04 78.99 87.80 94.27 115.23 138.52 90.17
Latin America &
Caribbean 59.16 81.31 80.65 78.17 85.01 85.60 85.34 86.75 79.59
South Asia 24.18 77.85 80.76 50.54 58.23 65.96 64.04 48.27 59.71
Africa* 58.72 69.84 69.50 65.64 65.54 72.33 77.47 83.36 69.50


(B) Export as %GDP
East Asia & Pacific 36.09 34.32 34.06 33.64 38.27 43.40 55.84 69.07 42.48
Latin America &
Caribbean 28.38 37.27 35.49 36.42 39.11 38.52 38.24 38.48 36.53
South Asia 10.26 33.71 33.50 22.64 26.15 31.81 31.38 29.33 27.71
Africa 27.32 29.08 28.22 28.02 27.01 30.61 35.16 38.28 30.12


(C) Real Per Capita GDP Growth (%)
East Asia & Pacific 5.37 3.90 2.17 2.15 3.51 2.06 2.13 4.35 3.57
Latin America &
Caribbean 3.31 2.71 -0.04 1.81 1.55 2.07 1.24 3.61 2.13
South Asia 0.54 2.40 3.75 2.89 2.64 2.82 3.84 3.75 2.85
Africa 2.66 0.91 -1.01 1.18 -1.93 2.83 2.36 2.75 0.67


(D) Population Growth (%)
East Asia & Pacific 2.37 2.03 2.21 2.02 2.00 1.76 1.47 1.46 1.88
Latin America &
Caribbean 1.82 1.72 1.76 1.49 1.53 1.41 1.30 1.17 1.55
South Asia 2.45 2.48 2.38 2.33 2.18 1.89 1.72 1.57 2.13
Africa 2.48 2.77 2.90 2.81 2.33 2.61 2.37 2.26 2.64


(E) Average Applied Tariff on All Goods (%)
1995-99 2000-04 2005-08 1995-08


East Asia & Pacific 15.96 9.80 9.83 12.01
Latin America &
Caribbean 13.64 11.67 8.22 11.39
South Asia 29.76 18.53 14.33 21.34
Africa 22.44 14.93 12.79 17.00
Global Average 11.93 10.24 9.35 10.59


(F) Human Development Index (HDI). Values range from 0 to 1; the higher, the
better.


1980 1990 1995 2000 2005 2009 2010
East Asia & Pacific 0.383 0.446 0.519 0.559 0.600 0.636 0.643
Latin America &
Caribbean 0.573 0.614 0.640 0.660 0.681 0.699 0.704
South Asia 0.315 0.387 0.415 0.440 0.481 0.510 0.516
Africa 0.293 0.354 0.358 0.315 0.366 0.384 0.389
Global Average 0.455 0.526 0.554 0.570 0.589 0.619 0.624


Note: *Africa largely denotes Sub-Saharan Africa.
Source: Human Development Reports (2010); World Development Indicators (2009),


World Trade Indicators (2010).




6


The summary from the stylised facts above concludes: a) Africa is becoming increasingly
more integrated in trade; b) Africa is fast becoming the least region with respect to
economic development; c) Africa has the highest population growth; and 4) Africa’s
average applied tariff has reduced consistently and remarkably. The above observations
can help to infer that Africa with high population growth experienced less economic
development and more importantly the increasing Africa’s trade integration do not reflect
significantly in development process of the continent. This discourse brings some issues
to limelight: the possible nexus between trade and economic development.




3 Theoretical Underpinning and Literature Review


There is a general assertion that cross-countries trade is imperative for economic growth
and to some extent economic development. This conclusion has been fostered by some
empirical and theoretical studies (e.g. Dollar 1992; Sachs and Warner 1995; Edwards
1997; David and Scott 2005). Also, Grossman and Helpman (1995) presume that the
world integration has an influence on the entrepreneurs which directly impact the social
fabrics of countries economic system. Hence, it is conventionally accepted that trade
openness is a vital component of economic growth and development (Winters 2004;
Mackay and Winters 2004).


Past studies offer some insights into the relationship between the trade, other factors and
economic growth and development. However, the studies have divergent conclusions
(Ackah and Morrissey 2010). The thrust of Solow (1956) argument was that market-
centred trade liberalisation will accelerate the dynamic of economic growth and
development. With respect to individual productivity pay-off, the aggregate market
interactions were to trigger growth, which is in accordance with the neoclassical theory of
trade and growth (Bhagwati 1988).


The progress in trade is becoming even more important in the analysis of economic
growth as well as development. Thus, it is necessary to examine theoretical and empirical
evidences towards substantiating the claims of WTO that the removal/reduction of tariff
influences economic growth. Some authors such as Berg and Krueger (2003) and Mackay




7


and Winters (2004) give reasons for trade liberalisation, and its propensity to promote
economic growth. These cross-countries empirical studies conclude that the liberalisation
of world trade has impacted significantly the economic growth of countries.


Mackay and Winters (2004) observed that the importation of capital goods and
technological goods create knowledge spill-over which increases international
competition. Through competition, trade is believed to enhance growth and concomitantly
leads to variety of goods availability at cheaper prices. The modern trade theory
developed by Helpman and Krugman (1985) and the new growth theory by Grossman and
Helpman (1991) illustrate that the benefits from trade is a fundamental argument for free
trade which makes it instrumental for economic growth. Although these studies were
Western-based, some economists believe that the argument for freer trade provides
significant incentives for developing countries (Srinivasan 2000 and Stiglitz, 2002).


In a similar manner, some empirical studies have related trade and trade issues to wealth
accumulation (Levine and Renelt 1992; Taylor 1998). In another perspective, Tilat (2002)
concludes that trade has no significant association with long-term economic growth and
suggested that short-run effects out-weigh the perceived benefits of trade liberalisation.
However, Mackay and Winters (2004) found that in the short run, trade liberalisation
harms poor actors in the economy and even in the long run, successful open countries
may create a return to below the poverty line, which means an escalation in poverty
density and a punctured economic growth.


The traditional theory of trade as illustrated by Stolper-Samulson reveals that an increase
in the relative price of a commodity results a corresponding increase in the real-return to
factors utilised in producing that commodity (Dixit and Norman 1980). However, some of
the literature did not examine the possibility of ‘Goliath-David trade’ to plummet economic
growth. Unfortunately for most Africa countries, the expected benefits of international
trade have not been sufficiently experienced; hence it is not difficult to link trade openness
with a countries’ economic less performance along e.g. primary extraction/commodities.




8


To investigate the relationships between trade openness through tariff removal to
economic growth within Africa, the effects on total factor productivity is imperative. Studies
show that reduction in trade barriers were followed by significant increases in total factor
productivity (Winters 2004). This resulted from the increase in import competition
according to Ferriera and Rossi (2001) with the study in Brazil, Jonsson and Subramanian
(2001) in South Africa, and Kraary (1997) obtain inconclusive results for China, while Aw,
Chung, and Roberts (1999) discovered little evidence for Latin America and Asia.
However, the significance of these studies resonate the debate about whether agricultural
commodities and primary extractions export for the poor countries in Africa is the option
for tariff reduction.


Freer trade by absolute definition involves greater interdependence among countries, and
Tilat (2002) linked it to the phenomenon of globalisation. Although reforms have been
uneven, there is clear evidence that protection of import substitutes with tariffs and non-
tariff barriers within Sub-Saharan Africa has declined significantly (Nash 1993). However,
Africa’s share in global exports reduced from 4.5 per cent in 1977 to 2.0 per cent in 1997,
and also, Africa’s share of total developing country exports dropped from 15.5 per cent in
1981 to 9.2 per cent in 1997 after many countries implemented the Structural Adjustment
Program (African Development Bank 2008).


Nevertheless, the study of Agama (2001) in Africa which utilised a database to investigate
the connection between trade openness and economic growth for 40 countries in Africa is
subjective. Agama argues that between 1980 and 1999, the more open countries in Africa
experienced higher economic growth rates than those that remained closed. Hence,
concludes that although trade liberalisation and economic integration increases economic
growth for African countries, increases in government consumption expenditure retards
the growth. Most studies believe that a significant relationship exists between exports,
measure of trade integration, and economic growth (Khalifa Al-Youssif 1997; Agama
2001) and cross-country study tends to confirm the importance of exports for developing
countries (Ngoc, Phuong Anh, and Nga 2003). The doubts exist pertaining to the
importance of trade. For example, Clarke and Kirkpatrick (1992) utilised data for 80
developing countries (1981 to 1988) to estimate the impact of trade policy reform on the




9


economic performance and conclude that trade reform does not affect profoundly the
economic performance.


Theoretically, the profound implication of international trade on development especially
along economic growth, income distribution, poverty, and employment are impressive
(Krugman 1983; Bhagwati 2004). This is anchored on the economic theory that opines a
completely liberalised global market constitutes the most efficient path to foster growth,
because a particular country that specialises in producing the goods and services in which
it has a comparative advantage gains from trade. Nonetheless, trans-national corporations
have become instruments of eroding nations’ comparative advantage since they dominate
the global marketplace and create a non-flattened relation of power and information.
Further, the problem is that free trade based on comparative advantage is not actually and
equally free. For example, agricultural subsidies and other designed trade barriers
common to the USA and some European countries can hinder Africa poor countries from
entry and participating in these vital markets despite the comparative advantage concept.


The debate about a positive empirical association between trade and economic growth
especially within the Africa domain remains far from being over. In spite of the recent
movement towards trade reforms for most Africa countries, there remain some major
controversies regarding certain aspects of trade and the message of WTO. The effects of
trade tariff removal/reduction and economic growth appear to be direct and imperative for
some selected Africa countries. To contribute to the academic debate and to recommend
some policies for Africa leaders, this paper examines the relationship between trade, tariff
reduction and economic growth and development among selected Africa countries (1995-
2008).


4 Empirical Model and Estimation Technique


The model for this paper, as informed from literature and the theoretical framework and is
discussed in the section. It draws insights from endogenous growth theory (also referred
to as Neo-classical theory) that has labour and capital as basic explanatory factors for
growth and also allows the incorporation of other variables of interest. The model




10


assumes a relationship between indicators of development of selected African countries
and capital, labour with the inclusion of other explanatory variables especially trade and
tariff. This is based on growth literature (e.g. Agama 2001; Stiglitz 2002; Winters 2004)
that have established the influence of trade on economic growth in some countries.


It is commonly purported that the more countries are open to trade, the better their level of
economic development. Thus, a functional relationship between economic development
(usually proxied by per capita income) and trade can be related. It has been argued that
real per capita income covers mainly the economic growth. To handle this, this paper
employs HDI, which incorporates other aspect of economic development especially
human component. The key explanatory variables of interest that can exert influence on
development are trade and tariff. This is represented by the functional relationship below:
EDevtJit = f(Labit, Investit TradeKit,Trdgrotit, Aptit,Uit)-------------------------------------1
The above equation is expressed in explicit form as:


2543210 −−−−−−−++++++= itititit
K


ititit
J


eAptTrdgrotTradeInvestLabEDevt αααααα


Where:
EDevtJ: Economic development in the selected countries. J=1 and 2. This represents two
equations: the real per capita GDP, rpgdp (economic aspect of development) and human
development index, HDI (the human aspect of development). rpgdp is measured at 1990
constant prices in United States dollars (USD), while HDI is taken as reported in HDR with
values ranging from 0 to 1; the higher, the better.
Lab: Labour force measured in million persons.
Invest: Domestic investment proxied by gross fixed capital formation measured in million
USD at 1990 constant prices.
TradeK: Trade integration, with K =1 and 2. This is measured by total trade
openness/integration, trdint (defined as ratio of total trade to GDP) and export integration,
expint (defined as ratio of export to GDP. These are the key measures of trade integration
with the third as import integration (import/GDP). However, it has been noted that the first
two measures are expected to positively influence growth and development but the impact
of import integration is ambiguous (Leyaro and Morrissey 2010). Hence, this paper used
trdint and expint.




11


Trdgrot: Real growth in trade, which shows rate of growth in trade over a given period.
The inclusion of this variable is necessitated with a view to examining the influence of real
trade growth over the studied period as it is possible to have trade integration without real
trade growth.
Apt: Average applied tariff.
eit: Error terms that captures other factors influencing the dependent variables that
are not included in the model. They are assumed to be identically and independently
distributed (iid) with zero mean and constant variance N(0, σ2).
it: The countries and time dimensions.


)50( −=iiα : Parameters to be estimated, which show the constant and the rate of
change in the dependent variable induced by the respective chosen explanatory variables.
Their apriori expectation is such that )4,,0( −=iiα >0. This means that the explanatory
variables are expected have positive influence on the indicators of development. Tariff can
be positive or negative depending on the economies.


To estimate the formulated model, the paper used a panel data regression technique.
Panel data regression technique is a relevant method of longitudinal data analysis
because it allows for a number of regression analyses in both units and time dimensions.
It also gives room for data analysis especially when the data are from various sources and
the time series are quite short for separate time series analysis (Baum 2006). In panel
data analysis, there are usually choices to be made from three possibilities: pooled
Ordinary Least Squares (OLS) regression, Fixed Effects (FE), and the Random Effects
(RE) models. However, there are some issues such as omitted variables, unobserved
heterogeneity, measurement errors and endogeneity biases (Baum 2006; Leyaro and
Morrissey 2010). To address them, the Generalised Method of Moments (GMM) estimator
was employed. GMM procedure allows freedom in specifying the lag structure for the
instruments.


The data used were sourced from World Development Indicators (WDI); World Trade
Indicators (WTI); United Nations Statistics; and Human Development Reports. The period
covered was 1995 to 2008 based on availability of relevant data, while STATA 10.1




12


program was used in the estimation process. The results from OLS, FE, RE and GMM are
reported and analysed in the next section.




5 Empirical Results and Analyses


The number of countries in Africa selected was 50 drawn from the five sub-regions in the
continent, namely: Central, East, North, Southern, and West Africa. The selected
countries represent about 87.72 per cent in number and over 95 per cent in economic size
with respect to GDP and population. Thus, this will give a good representation. The list of
countries selected arranged according to their sub-regions is reported in Table A1 in the
appendix.




5.1 Descriptive Analysis
To have first-hand information of the key issues, the paper plotted the major variables of
interest as shown in Figure 1.


As can be observed in Figure 1, indicators of trade, namely trade and export integration
ratios increased remarkably throughout the period, though they experienced little
fluctuation during the period 2000 to 2003. On the other hand, indicators of development,
namely real per capita GDP and HDI maintain a somewhat minimal increase over the
period. However, average applied tariff decreased markedly and consistently over the
same period. The above finding implies that Africa has experienced some form of
increased trade integration and declined tariff rate but the level of development has not
considerably improved. This denotes the challenge of Africa’s trade inability to translate to
development.




13




Note: Mean values were used for the graph; logarithm of rpgdp was taken to get the rate.
Source: Authors’ computation.




5.2 Summary of Statistics
The paper reports the summary of statistics of the variables used in the estimation
process with a view to making comparison across the five sub-regions in Africa, namely:
Central, East, North, Southern, and West; and make discussion on them. This is reported
in Table 2.


Values in Table 2 show that the mean of real per capita GDP for the 50 sampled African
countries was USD 1308.19; while across the sub-regions, it was highest in North Africa,
which was about twice the means of other regions, while the least was in West Africa. A
related observation is seen for HDI that was 0.50 for Africa, which was even lower for
Central Africa with the value of 0.48. On the other hand, both measures of trade
integration indicate that Southern Africa was more integrated in trade than the rest of
Africa, which was followed by North with the least being Central. Real growth in trade was


0
20


40
60


80
10


0


0
5


10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85


va
ria


bl
e


s


1995 19961997 1998 19992000 20012002 20032004 20052006 20072008
year


lmeanrpgdp meanhdi
meantrdint meantrdgrot
meanexpint meanapt


Figure 1 Trend in selected variables 1995 to 2008




14


higher in Central Africa distantly followed by North, Southern, East, and West Africa,
respectively.


Table 2 Summary Statistics of Selected Variables
Variables Statistics Africa Central East North South West
Rpgdp Mean 1308.19 1314.62 1360.29 3030.91 1535.92 552.87
Std. Dev. 1815.84 1782.50 2260.02 2654.32 1298.36 322.12
Hdi Mean 0.50 0.48 0.51 0.69 0.53 0.44
Std. Dev. 0.13 0.13 0.13 0.08 0.11 0.25
Trdint Mean 78.96 71.46 74.55 72.98 105.82 72.80
Std. Dev. 44.59 50.75 48.03 26.24 47.87 33.66
Trdgrot Mean 6.94 9.99 6.13 6.58 6.52 6.11
Std. Dev. 13.06 18.11 11.64 6.47 10.24 13.21
Expint Mean 34.87 33.35 29.69 38.49 49.39 29.98
Std. Dev. 22.89 27.38 23.92 17.57 26.24 12.44
Apt Mean 14.78 17.17 15.50 14.05 11.57 15.36
Std. Dev. 8.08 2.76 8.20 7.45 6.43 10.63
Invest Mean 2680.00 904.00 1220.00 11500.00 4760.00 694.00




Std. Dev. 6260.00 926.00 1580.00 6950.00 11100.00 800.00
Lab Mean 6.58 4.51 9.21 10.00 4.82 5.81




Std. Dev. 8.30 5.77 9.22 7.56 4.92 9.81
Observations(N)




700 126 168 70 126 210
Note: Only mean and standard deviation were reported for brevity sake.


Source: Authors’ computation.


Conversely, the mean values of applied tariff indicate that Central African sub-region has
the highest tariff rate followed by East, West, North and South, in that order. Table 2
points out that the average domestic investment was highest in North African sub-region
with value that was more than 10 times above those of West and Central African sub-
regions. In terms of labour force, the highest was in North Africa followed by West. The
inference that can be made from the discussion on the summarized statistics is that the
regions with higher level of domestic investment, as well as labour force seem to have
higher values in development indicators. This does not hold for tariff and trade given the
fact that sub-regions had higher values in trade integration and real growth in trade do not
necessarily reflect better indicators of development.




5.3 Correlation Test of Variables
Taking a step further before presenting the estimation result, the paper reports the
correlation matrix to examine possible problem of collinearrity among the explanatory




15


variables. The variables except real growth in trade were used in their logarithmic forms
given the assertion that this process helps to mitigate the issue of heteroscedasticity and
also brings the variables to a more comparative form (Olokoyo, Osabuohien and Salami
2009).




Table 3 Coefficient of Correlation among the Variables




lnrpgdp Lnhdi lntrdint Trdgrot lnexpint lnapt lninvest Lnlab
lnrpgdp 1.0000




Lnhdi 0.8348 1.0000


lntrdint 0.4243 0.4948 1.0000


Trdgrot 0.0266 0.0032 0.0696 1.0000
lnexpint 0.5359 0.5425 0.8308 0.0590 1.0000
Lnapt -0.1047 -0.2526 -0.0878 -0.0827 -0.0590 1.0000
lninvest 0.4425 0.4614 0.0004 0.0740 0.1743 -0.2358 1.0000
Lnlab 0.2875 0.2625 0.1849 0.0512 0.1937 0.0250 0.1877 1.0000


Source: Authors’ computation.


The values in Table 3 show that the two measures of economic development, namely:
real per capita income and HDI exhibit strong correlation between them. This is not
unexpected as they give interpretation on level of economic development. Since both of
them are used differently as dependent variables, it does not pose any challenge. This
observation is also similar to correlation between trade and export integration ratios, which
necessitate their usage differently in estimation process. In addition, their separate use
helps to ascertain, which of the two measures are more relevant for Africa’s economic
development process.


Surprisingly, the coefficient of correlation between trade and tariff was very minuscule (far
less than 0.1), which implies that they can be combined together without problem of
multicollinearity as the issue of multicollinearity becomes crucial when the coefficient of
correlation becomes high, say above 0.5 (Baum 2006). This is quite surprising given the
fact that tariff should have been expected to influence trade; the reason for this especially
in Africa is sufficient for another research. In summary, the correlation test has shown
that there is no challenge of multicollinearity among the explanatory variables and as such




16


the estimated results can be relied upon for useful deductions. Other tests before the
estimation would have been panel unit root and panel co-integration tests. However, given
the fact that GMM was among the estimators engaged which uses difference in the
variable, these pre-test are not always essential (Leyaro and Morrissey 2010). Hence, the
paper reports and analyses the estimated results in the next sub-section.




5.4 Presentation of Estimated Results and Analyses
Tables 4a and 4b present results from OLS, FE and RE and as well as GMM for the two
indicators of economic development used as dependent variables, namely: real per capita
income (Lnrpgdp) and human development index (Lnhdi), respectively.


The test statistics in the estimations in the last segments of Tables 4a and 4b, namely:
coefficient of determination (Adj. R2/R2), F-Statistics (F-stat), Wald Statistics (Wald-stat)
which were significant at 1 per cent denote that all the coefficients are jointly significant.
For instance, the values of Adj.R2/R2 in Table 4a were 0.8633, 0.5234 and 0.4459 for
OLS, FE and FE, respectively using the equation with lntrdint, while those in Table 4b
were 0.7672, 0.5660 and 0.5005. The values of F-stat for FE estimate was 38.00 and
Wald-stat for RE and GMM estimates were 72.39 and 85.25 in Table 4a, respectively
using the equation with lntrdint . Their counterparts in Table 4b in were 42.00, 85.74 and
87.18. The above discussion underscores the models good-fit. However, this does not
address the issue of endogeneity and measurement errors. The paper also appraised the
results from FE and RE using Hausman test, which indicates that the estimates from RE
were more efficient than those of FE.


Furthermore, the system GMM, which helps to address endogeneity and measurement
challenges are reported. To evaluate whether models were correctly specified and
whether instruments were valid, The Hansen/Sargan J statistics and the test for first and
second order serial correlation of the residual in differenced equation were carried out
[(AR(1) and AR(2)]. If the model is correctly specified, the variables in the instrument set
should be uncorrelated with the idiosyncratic component of the error term.





17




Table 4a Estimated Results with Real GDP per Capita
Dependent Variable ⇒ Lnrpgdp


Estimators⇒
/Variables


OLS FE RE GMM (Syst)
GMM
(Syst) OLS FE RE


GMM
(Syst)


GMM
(Syst)


Lnrpgdp(-1)
0.6475a
(0.000)


0.6587a
(0.000)


0.5208a
(0.000)


0.5403a
(0.000)


Trdgrot
0.0090a
(0.001)


0.0019b
(0.033)


0.0031a
(0.004)


0.0015a
(0.000)


0.0014a
(0.000)


Lntrdint
0.1825
(0.103)


0.0666
(0.167)


0.0134
(0.235)


0.0126
(0.477)


0.0139
(0.474)


Lnapt
0.0423
(0.423)


-0.0448
(0.180)


-0.0617
(0.240)


-0.0114
(0.580)


-0.0123
(0.589)


0.1101
(0.054)


-0.0357
(0.170)


-0.0432
(0.142)


-0.0035
(0.864)


-0.0306
(0.195)


Lninvest
0.6654a
(0.000)


0.1378a
(0.000)


0.3404a
(0.000)


0.0642a
(0.000)


0.0664a
(0.000)


0.6507a
(0.000)


0.1563a
(0.000)


0.3184a
(0.000)


0.0886a
(0.000)


0.0914a
(0.000)


Lnlab
0.7036a
(0.000)


0.4446a
(0.000)


0.4554a
(0.000)


0.1380c
(0.072)


0.1060c
(0.054)


0.6660a
(0.000)


0.3026c
(0.060)


0.4418c
(0.000)


0.1377c
(0.073)


0.1465b
(0.046)


Lnexpint
-0.0011
(0.983)


0.0197
(0.582)


0.1095
(0.202)


0.0137
(0.414)


0.0002
(0.489)


Constant
4.0223a
(0.000)


2.6174a
(0.000)


6.4644a
(0.000)


3.0932b
(0.014)


2.5447
(0.162)


2.7717a
(0.000)


1.2076
(0.525)


6.3720a
(0.000)


3.4012a
(0.006)


3.4734
(0.136)


Adj. R2/(R2) 0.8633 0.5234 0.4459 0.8421 0.4916 0.4399
F-stat


38.00a
(0.000)


42.300
(0.000)


Wald-stat.


72.39a
(0.000)


85.25a
(0.000)


96.74a
(0.000)


48.44a
(0.000)


73.30a
(0.000)


47.71a
(0.000)


Hausman
test


13.22a
(0.000)


19.86a
(0.000)


Hansen J


0.5722 0.4807
AR(1)




0.001 0.002
AR(2)




0.367 0.4212
Time effect




No Yes No Yes
Note: OLS - Ordinary Least Squares; FE- Fixed Effects; RE- Random Effects; GMM- Generalised
Method of Moments. R2 for OLS is adjusted but for FE and RE it is the overall. The probability
values are in parenthesis. Superscripts a,b and c denote significant at 1, 5 and 10 per cent,
respectively.


Source: Authors’ computation.




More so, for the instruments to be valid, the probability values for Sargan test and the
AR(2) tests should both be greater than 0.05. The AR(1) test is asymptotically distributed
as a standard normal under the null of no first-order serial correlation. The GMM estimator
requires that there is first-order serial correlation, AR(1) but no second-order serial
correlation, AR(2) in the residuals (Arellano and Bond 1991; Leyaro and Morrissey 2010).
The tests statistics show that the estimates are reliable. Hence, this paper focus
discussions on RE and GMM estimates.




18


In Table 4a, using real per capita GDP as indicator of economic development, the results
show that the key determinants of economic development in the selected African
countries within the studied period include: past level of economic development, real
growth in trade, domestic investment and labour force. This is given based on the fact
their coefficients were significant at the usual levels. Surprisingly, trade and export
integration ratios were found not to be significant in influencing economic development,
though they had the expected positive association with export integration (expint) being a
little greater than trade integration (trdint). In a similar fashion, average tariff had negative
sign; however it was not significant at 10 per cent.






Table 4b Estimated Results with HDI
Dependent Variable ⇒ Lnhdi


Estimators⇒
/Variables


OLS FE RE GMM (Syst)
GMM
(Syst) OLS FE RE


GMM
(Syst)


GMM
(Syst)


Lnhdi(-1)
0.0167
(0.886)


0.0898
(0.462)


0.0497
(0.662)


0.0659
(0.580)


Trdgrot
0.0031a
(0.002)


0.0005b
(0.030)


0.0010b
(0.043)


0.0002a
(0.007)


0.0001c
(0.074)


Lntrdint
0.0490
(0.135)


0.0414
(0.130)


0.0744
(0.210)


0.0373
(0.282)


0.0469
(0.197)


Lnapt
-0.0095
(0.647)


-0.0027
(0.837)


-0.0101
(0.451)


0.0116
(0.724)


0.0186
(0.637)


-0.0033
(0.759)


-0.0072
(0.293)


-0.0100
(0.143)


0.0283
(0.373)


0.0309
(0.407)


Lninvest
0.1660a
(0.000)


0.0652a
(0.000)


0.1334a
(0.000)


0.0544b
(0.015)


0.0368b
(0.014)


0.0861a
(0.000)


0.0329a
(0.000)


0.0636a
(0.000)


0.0719a
(0.003)


0.0550b
(0.040)


Lnlab
0.1304a
(0.000)


0.2896a
(0.000)


0.1034a
(0.000)


0.5246a
(0.000)


0.6260b
(0.017)


0.0726a
(0.000)


0.1283b
(0.030)


0.0506a
(0.000)


0.4575a
(0.000)


0.5541b
(0.017)


Lnexpint
0.0103
(0.286)


0.0355
(0.120)


0.0465
(0.201)


0.0585
(0.160)


0.0568
(0.191)


Constant
-2.3740a
(0.000)


-6.6354a
(0.000)


-2.2162
(0.000)


-1.0074a
(0.000)


-1.1101a
(0.070)


-0.2222a
(0.000)


-


2.2291a
(0.000)


-0.1869
(0.120)


-


1.4619a
(0.000)


-


1.4488c
(0.091)


Adj. R2/(R2) 0.7672 0.5660 0.5005 0.7626 0.5735 0.5256
P. Value 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000


F-stat
42.00


(0.000) 54.4500
Wald-stat.




85.74
(0.000)


87.18
(0.000)


90.70
(0.000)


96.04
(0.000)


91.36
(0.000)


96.84
(0.000)


Hausman
Test


26.14
(0.001)


40.98
(0.000)


Hansen J


0.6071 0.5787
AR(1)




0.000 0.001
AR(2)




0.3976 0.4851
Time effect No Yes No Yes


Note and Source: Same in Table 4a




19


Table 4b reports the result using the second indicator of economic development (human
development index). The results appear a somehow similar to the previous one as the
major determinants were observed to be domestic investment, labour force and real trade
growth. However, previous level of HDI was positive but not significant. Both trade and
export integration ratios were not significant but real trade growth was significant. The
interpretation for this may mean that in the long-run, real rate of Africa’s growth in trade
has the possibility of impacting on economic development unlike mere trade integration.
This implies that unguided openness to trade does not lead to economic development as
it has been noted that the argument for protection of infant industry in Africa is far from
been over. This is in line with the submission made by Ackah and Morrissey (2010).


The major policy recommendation emanating from the results and findings in this paper is
that tariff reduction argument does not favour African countries, while trade integration
may be good but it is not significantly sufficient in influencing economic development.
Another implication for policy is that efforts to promote the level of domestic investment
and labour productivity are very germane for economic development in Africa. Such
policies should encompass enhancement of internal mechanism including vibrant financial
institutions, political stability, functional telecommunication and transport facilities, inter
alia. This is because the above efforts will help boost the level of capacity utilisation as
well as capital formation, which are essential for economic development.




6 Conclusion
The debate on economic development, trade and tariff is not yet over especially in Africa
with varying opinions in literature. This motivated the present paper, which examined the
nexus between economic development, trade and tariff for the period 1995 to 2008. To
achieve the objective of the study, data sourced from human development reports, world
development indicators, and world trade indicators were analysed using both descriptive
and econometric techniques.


The results obtained from the empirical analyses established that regions in Africa that
had higher level of domestic investment experienced higher indicators of economic




20


development. The paper equally found that domestic investment and labour played more
significant contribution to economic development in Africa than both trade and tariff. The
challenge in this regards is the fact that increased trade integration in Africa do not
significantly lead to enhancement of the economic development process. Thus, improving
domestic investment and enhancing labour productivity will promote economic
development more than trade and tariff. Hence, the paper cautioned against swift trade
integration and unguided tariff reduction since they did not exert much impact on
economic development over the period studied.


The implication of the above findings is that there is a somewhat challenge of Africa’s
trade integration measures though having the potentials, which do not significantly result
in enhancement of economic development in the continent. Another implication from the
result is that domestic investment, labour, real growth in trade are important factors for
Africa’s economic development. Thus, the choice before African countries is to enhance
domestic investment and harnessing of their labour force in order to improve their level of
economic development. This policy recommendation can be engendered through
promotion of functional and technical education, which will help to adequately develop and
utilise the abundant labour force in most African economies. Another measure will be the
pursuance of vibrant and resilient financial sector that will be focused to play active role
for meaningful domestic resource mobilization, which is fundamental in stimulating
domestic capital.


The submission of the paper is crucial given the event of global financial crisis where
commodities prices and global demand of primary products, which are traded by most
African countries, have nose-dived. Thus, reliance of domestic investment and labour
force, which are not so subject to external shocks, will be a better policy choice more than
clamour for mere trade integration and tariff reduction.







21


Acknowledgements
The authors appreciate The Council for Development of Social Science Research in Africa
(CODESRIA) for sponsoring the full cost of participation at CODESRIA Guy Mhone 2010
Conference on “The Renaissance and Revival of African Economies”, held 20-21
December 2010 at Dar es Salaam, Tanzania as well as helpful comments from
participants at the conference where the first draft of this paper was presented. The
assistance of Dr (Mrs) Dauda R.O.S. of Dept. of Economics and Development Studies,
Covenant University during the revision process is acknowledged. The first author
acknowledges grant for PhD Thesis Writing from CODESRIA (Ref:SGRT.141/T08) and
PhD Fellowship awarded by Swedish Institute (Ref:00350/2009) as well as useful
assistance from faculty in Department of Economics, Lund University, Sweden during his
visit as guest PhD candidate. The gesture of Covenant University Management for
providing tuition-free post-graduate studies is also appreciated by the first author. The
second author appreciates Adventist University of Central Africa for time granted him and
to the University of Eastern Africa, Baraton for the books and on-line journals provided.







22


REFERENCES
Ackah, C. and Morrissey, O., 2010, ‘Who Gains from Trade Liberalization in Developing
Countries? Evidence from Macro-Micro Analysis’, paper presented at International
Workshop on the Determinants and Effects of Trade and Foreign Direct Investment in
SSA, Accra, Ghana, 26-28 October.


African Development Bank, 2008, Selected Statistics on African Countries, Vol. XXVII,
Tunis: African Development Bank.


Agama, L. A., 2001, ‘Assessing the Desirability of a Free-Trade Area in Southern Africa’,
International Economic Review, July/August: 1-6.


Arellano, M. and Bond, S. R., 1991, ‘Some Tests of Specification for Panel Data: Monte
Carlo Evidence and an Application to Employment Equations’, Review of Economic
Studies, Vol.58, No.2, pp.277-297.


Aw, B.Y., Chung, S. and Roberts, M.J., 1999, ‘Productivity and Turnover in the Export
Market: Micro Evidence from Taiwan and South Korea’, NBER, February.


Baum, C.F., 2006, An Introduction to Modern Econometrics Using Stata, Texas: Stata
Press.


Berg, A. and Krueger, A., 2003, ‘Trade, Growth, and Poverty’, IMF Working Paper,
WP/03/30.


Bhagwati, J., 1988, ‘Export-Promotion Trade Strategy: Issues and Evidence’, The World
Bank Research Observer, Vol.1, pp.27-57.


Bhagwati, J. N. ,2004, In Defense of Globalisation, New York: Oxford University Press.


Clarke, R. and Kirkpatrick, C., 1992, ‘Trade Policy Reform and Economic Performance in
Developing Countries: Assessing the Empirical Evidence’, in R. Adhikari, C. K Kirkpartick




23


and J. Weiss,eds., Industrial and Trade Policy Reform in Developing Countries,
Manchester: Manchester University Press.


David, M. and Schott, A., 2005, Macroeconomics: Understanding the Wealth of Nations, A
Handbook, New York: John Welly & Sons.


Dollar, D., 1992, ‘Outward-Oriented Developing Economies Really Do Grow More Rapidly:
Evidence from 95 LDCs, 1976-1985’, Economic Development and Cultural Change,
Vol.40, No.3, pp.523-544.


Dixit, A. and Norman, V., 1980, Theory of International Trade, Cambridge: University
Press.


Edwards, S., 1997, ‘Openness, Productivity and Growth: What Do We Really Know?’,
NBER Working Papers, No. 5978.


Ferriera, P. C. and Rossi, L., 2001, ‘New Evidence on Trade Liberalisation and
Productivity Growth’, Ensaios Economicos da, EPGE 433.


Grossman, G. M. and Helpman, E., 1991, Innovation and Growth in the Global Economy,
Cambridge: MIT Press.


Grossman, G.M. and Helpman, E., 1995, ‘Trade Wars and Trade Talks’, Journal of
Political Economy, Vol.103, pp.675-708.


Human Development Reports, Various Issues, New York: United Nations Development
Programme.


Jonsson, G. and Subramanian, A., 2001, ‘Dynamic Gains from Trade: Evidence from
South Africa’, IMF Staff Papers, Vol.48, No.1, pp.187-224.




24


Khalifa Al-Youssif, Y., 1997, ‘Exports and Growth: Some Empirical Evidence from the
Arab Gulf States’, Applied Economics, Vol. 29, pp.693-697.


Kraay, A., 1997, ‘Exports and Economic Performance: Evidence from a Panel of Chinese
Enterprises’, Mimeo, Washington DC: The World Bank.


Krugman, P., 1983, ‘New Theories of Trade among Industrial Countries’, American
Economic Review, Vol.73, No.2, pp.343-347.


Levine, R and Renelt, D., 1992, ‘A Sensitivity Analysis of Cross-Country Growth
Regressions’, American Economic Review, Vol.82, No.4, pp.942-963.


Leyaro, V. and Morrissey, O., 2010, ‘Trade and Growth: Is Sub-Saharan Africa Different?’,
CREDIT Research Paper, No.10/04.


Nash, J., 1993, ‘Trade Policy Reform Implementation in Sub-Saharan Africa: How Much
Heat and How Much Light?’, Mimeograph, Washington DC :The World Bank.


Ngoc, P.M., Phuong Anh, N. T. and Nga, P.T., 2003, ‘Exports and Long-Run Growth in
Vietnam, 1976-2001’, ASEAN Economic Bulletin, Vol. 20, pp.1-25.


Olokoyo, F.O., Osabuohien, E.S. and Salami, A.O., 2009, ‘An Econometric Analysis of
Foreign Reserves and Some Macroeconomic Variables in Nigeria (1970-2007)’, African
Development Review, Vol.20, Issue 3, pp.454-475.


Sachs, J. D. and Warner, A M., 1995, ‘Economic Convergence and Economic Policies’,
Brookings Papers in Economic Activity, Vol.1, pp.1-95.


Solow, R. M., 1956, ‘A Contribution to the Theory of Economic Growth’, Quarterly Journal
of Economics, Vol.70, pp.65-94.




25


Srinivasan, T.N., 2000, ‘Developing Countries in the World Trading System: from GATT,
1947, to the Third Ministerial Meeting of WTO, 1999’, The World Economy, Vol.23,
pp.437-454.


Stiglitz, J. E., 2002, Globalisation and its Discontents, New York: Norton and Company.


Taylor, A.M., 1998, ‘On the Costs of Inward-Looking Development: Price Distortions,
Growth and Divergence in Latin America’, The Journal of Economic History, Vol.58, No.1,
pp.1-28.


Tilat, A., 2002, ‘Impact of Globalisation and Liberalisation on Growth, Employment and
Poverty: A Case Study of Pakistan’, UNU-WIDER Research Paper.


Winters, L. A., 2004, ‘Trade Liberalisation and Economic Performance: An Overview’,
Economic Journal, Vol.114, No.2, pp.F4-21.


Winters, L. and Mackay, A., 2004, ‘Trade liberalisation and Poverty: The Evidence So
Far’, Journal of Economic Literature, Vol.42, No.1, pp.72-115.


World Bank Group, 2010, World Trade Indicators 2009/10.
(http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/TRADE/0,,contentMDK:2242195
0~pagePK:148956~piPK:216618~theSitePK:239071,00.html). 12 February 2010).





26


Appendix


Table A1 List of Countries used in the estimation


Central East North Southern West
Burundi Comoros Algeria Angola Benin
Cameroon Djibouti Egypt Botswana Burkina Faso
Central Africa
Rep. Eritrea Libya Lesotho Cape Verde
Chad Ethiopia Morocco Mozambique Cote d’Ivoire
Congo,DR Kenya Tunisia Namibia Gambia
Congo, Republic Madagascar




South Africa Ghana
Equatorial Guinea Malawi




Swaziland Guinea
Gabon Mauritius




Zambia Guinea Bissau
Rwanda Seychelles




Zimbabwe Mali


Sudan


Mauritania


Tanzania


Niger


Uganda


Nigeria




Senegal




Sierra Leone




Togo
Source: WTO (2009) International Trade Indicators





Login