Transcript The 7th Annual Conference on Regional Integration in Africa (ACRIA7)

The 7th Annual Conference on Regional Integration in Africa (ACRIA7) on the
Topic
External Sector and Inclusive Development in West Africa
Jointly organized by
CREPOL WAIFEM WAMA & WAMI
Driving factors of intra-regional trade in agricultural goods:
The case of West African Economic and Monetary Union
Cotonou, July 8-9, 2016
Dr. Toussaint Houeninvo ,
African Development Bank
Regional office for Dakar, Senegal
Presentation Outline
1-Introduction and context
2- Intra-WAEMU trade in agricultural goods still low despite trade
liberalization measures since dthe 1990s
3-The research question
4-Theoretical and empirical foundations of the gravity model in the
analysis of trade in regional integration area
5- Presentation of the model, data sources and sampling
6- Model specification (fixed-effects or random-effects) and results
7- Conclusion and Policy recommendations
2
1 – Introduction and context
•
Since the 2008 financial crisis and the European debt in 2010, and their
impact on African exports, strengthening regional integration appears as
an alternative to build resilience to shocks and promote economic
development in West Africa.
•
In this context, and given the fact that agricultural value added in the
GDP represents between 20% and 50% of WAEMU Countries, boosting
Intra regional trade especially intra-WAEMU agricultural exports can
serve as a key factor for inclusive development in West Africa.
•
The hypothesis of the inclusive character of trade in agricultural goods
is strengthened by the fact that most of the population in West Africa live
in rural area and do not mainly benefit from the broad economic growth
at macro level.
•
In such a condition, Intra-regional trade in agricultural goods could
contribute to reduce, inequality and poverty
3
2 – Intra-WAEMU trade in agricultural goods still low despite
trade liberalization measures since the 1990s
• The structure of WAEMU tariff applied on external trade include the
Custom rate and two community tax namely Community Solidarity Levy
(1%) and the Statistical Tax (1%). .
•
Trade liberalization initiated in early 1990s with the structural adjustment
programs in different WAEMU Countries has been strengthened since
1996 by the Additional Act 96/04 of 16 May 1996 that has led today to the
removal of quotas and other quantitative restrictions and 0% tariff on
crude products originating from WAEMU including agricultural non
processed goods.
• Moreover, since January 2003, there is the lifting of the certificate of origin
requirement for raw products including agricultural intra-WAEMU exports
except seafood.
• Nevertheless non harmonized tariff rates charged by different member
4
countries could not promote the development
of intra-Community trade
2– Intra-WAEMU trade in agricultural goods still low despite
trade liberalization measures since the 1990s (cont..)
• Structure of tariff before the Common External tariff of WAEMU (1st January
2000)
Types of tariff
Maximal rate excluding Community Solidarity
Minimal rate
Tax (1%) and Statistical tax (1%)
Countries
Benin
2%
20%
Burkina Faso
9%
33%
Côte d’Ivoire
7.6%
35%
Guinea- Bissau
10%
105%
Mali
8%
30%
Niger
10%
30%
Senegal
25%
60%
Togo
6%
20%
5
2–Intra-WAEMU trade in agricultural goods still low despite
trade liberalization measures since the 1990s (cont..)
•
Structure of tariff after the Common External tariff of WAEMU (1st January
2000)
Community
Tariff
Tariff rate
Statistical tax (ST)
Categories
Solidarity Tax
Overall tariff
(CST)
Category 0
0%
1%
1%
2%
Category1
5%
1%
1%
7%
Category 2
10%
1%
1%
12%
Category 3
20%
1%
1%
22%
35%
1%
1%
37%
Category 5 (ECOWAS CET since
2015)
6
2 – Intra-WAEMU trade in agricultural goods still low despite
trade liberalization measures since the 1990s (cont..)
• In spite of all these measures one can notice that several years after the
creation of the UEMOA in 1994 and the adoption of a Common External
Tariff in 2000 (Customs Union), the intra-WAEMU trade is still low
estimated at 12% between 1994 and 2014.
• As compared to the expectations, at the time of the establishment of the
CET, WAEMU Commission was expecting to reach 25% of intraWAEMU trade at the end of 2005 (Coulibaly, Traore and Diarra 2015).
•
In terms of comparison to the other Regional integration zones, Intratrade is about 64% for European Union, 60% pour North American Free
Trade Area (NAFTA), 35% for ASEAN and 30% for MERCOSUR.
•
Intra-WAEMU agricultural exports represent only 4% of total WAEMU
exports
7
2 – Intra-WAEMU trade in agricultural goods still low despite
trade liberalization measures since the 1990s (cont..)
• In terms of share of intra-WAEMU agricultural exports in intra-WAEMU
exports they are estimated at 29.20% in 1996 and 32.3% in 2015, hardly
3% points increase over 21 years since the signing of the agreement
establishing the WAEMU zone
• According to a recent statement by the President of the
WAEMU this low level of intra-WAEMU trade is due to lack of
competitiveness due to cost factors, the supply, the
structural weakness of the physical infrastructure, lack of
complementarity between the economies and obstacles to the free
movement of goods in the countries
8
3 – The Research question
•
Beside value chain development and the level of processing, one of the
constraints to trade in West Africa is the quality of transport and the
related cost of transportation and logistics including Non-Tariff Barriers
that jeopardize competitiveness.
•
Most of the time, several goods imported from out of the continent
especially from Asia, Europe and America have been more competitive
than the intra WAEMU products.
•
Hence, the main research question is on the impact of distance
(transport / logistical cost) and the level of development on intraWAEMU agricultural exports.
•
The paper analyzes the determinants of intra-WAEMU trade in
agricultural products and therefore the variables on which policymakers
could act to promote intra-regional agricultural trade.
9
4 –Theoretical and empirical foundations of the gravity
model in the analysis of trade in regional integration area
• The underlining assumption is that the GDP, that is the economic mass,
acts as the attractor of trade between two trading partners and
therefore exert a positive effect on trade. In contrast, distance, a
measure of the cost of transport used by most studies, serves as
resistance factor and plays a negative role in trade.
• From an empirical point of view, This has been applied by Tinbergen
(1962); Linnemann (1966); Obstfeld and Rogoff (2000); Xubei (2001);
Jakab et al. (2001); Gbetnkom and Avom (2005); Coulibaly, Traore and
Diarra (2015)
10
5 – Presentation of the model, data sources and sampling
•
•
•
•
•
•
•
With a1, a2 >0 and a3<0
Where EXPAij is the value of Exports from country i to country j
B is an intercept
t is the period
Yit is the GDP of Country i during the period
Yjt is the GDP of Country j during the period
DISTij is the distance between country i and country j (i ≠ j). In practice it
is the distance between the major capitals and ports of the two countries
11
5 – Presentation of the model, data sources and sampling
(cont..)
• Taking the logarithm of equation1 lead to
•
Where a1 and a2 are >0 and a3 <0
But external trade is not influenced only by these two factors. Therefore the basic
model has been extended by adding other factors include the characteristics of
partner countries(control variables) following Luo Xubei (2001), Gbetnkom and
Avom (2005), Coulibaly Traore and Diarra (2015)
•
These control variables are Population, Foreign Direct Investment (FDI), Level
of Integration(Common external tariff),Political Stability and Absence of
Terrorism
12
5 – Presentation of the model, data sources and sampling
(cont..)
+
a Z  a PS  
5
it
6
it
it
Where
• Yit is the GDP of Country i during the period
• Yjt is the GDP of Country j during the period
• DISTij is the distance between country i and country j (i ≠ j). In practice it is
the distance between the major capitals and ports of the two countries
• POPit is the population of country I in the year t
• FDIit-2 is the Foreign Direct Investment of Country I in the year t-2
• Zit is the dummy variable for the Common External Tariff during the period
t (level of integration)
• PSit stands political stability and absence of terrorism (Governance
variable)
• ᵋ is the error term
13
5 – Presentation of the model, data sources and sampling
(cont..)
Expected signs of the variables
Variables
Intercept
GDPit
DISTij
POPit
FDIit
Zit
PSit
Coefficients
b0
a1
a2
a3
a4
a5
a6
Expected signs
+
+
-
-
+
+
+
Source of data
• WDI, UNCTAD, WAEMU Commission, www.Govindicators.org
• www.levoyageur.net
Sampling
• 126 observations of panel data covering 7 WAEMU member countries
over 1996-2013 yearly data
14
6 – Model specification (fixed-effects, random-effects or
mixed effects) and results
• Several types of test have been considered to that end including Fischer
test and Hausman test. Breusch and Pagan Lagrangian Multiplier test for
random-effects.
• While the Fischer global test measures the global significance of individual
effects (fixed-effects), Breusch Pagan Lagrangian Multiplier, the Hausman
test the presence of random-effects
• In the Fischer test using STATA software with which we regressed exports
on the other explanatory variables the “within” coefficient of determination
which gives the explanatory power of the fixed-effects model has a very
weak value of 0.19. This suggests that the assumption of the presence of
fixed-effects is not corroborated.
• The random-effects has been tested and the “between” coefficient of
determination which gives the explanatory power of the random-effects
15
model has a value of 0.69 corroborating
the random-effects assumptions
6 – Model specification (fixed-effects, random-effects or
mixed effects) and results
• Moreover in order to deepen the specification process we run the
Hausman test for which the probability is 54%. Since the probability is
greater than 10%, the Hausman test corroborates the choice in favor of a
random-effects model.
• This is consistent with the fact that the within estimator is unable to
estimate the marginal impact of variables that are invariant over time
(e.g distance in our case).
• Hence, estimating the model under the above assumption and after
performing the required econometric testing (Wooldgride autocorrelation
test, Breusch-Pagan heteroscedasticity) gives the following results
16
6 – Model specification (fixed-effects, random-effects or
mixed effects) and results(cont..)
Recapitulation of the results
Intercept
(1)
(2)
Log (EXPA)
Log (EXPA)
2.56
-
(0.31)
Log (GDP)
Log(DIST)
Log(FDI-2)
Log(POP)
1.01*
0.92**
(3.13)
(7.01)
-0.97*
-0.80**
(-3.02)
(-12.18)
0.27*
0.24**
(2.80)
(3.34)
-0.13
-
(-0.27)
PS
0.11
-
(1.13)
Z
0.15
-
(0.94)
Observations
111
111
Wald chi2 (6)
349.86
6196.70
Prob>chi2
0.0000
0.0000
t student in parenthesis;
** Coefficient significant at 1%
* Coefficient significant at 5%
17
6 – Model specification (fixed-effects, random-effects or
mixed effects) and results(cont..)
• As shown in table 2, in the panel random effects models, all the key
variables are significant and have the expected signs.
• The basic variables (the two factors of gravity) are significant at 1% level
with the expected signs.
• Regarding the four control variables meaning population (pop), foreign
direct investment (FDI-2), Political stability (PS) and the Common
External Tariff (Z), all of them have the expected signs but only the FDI
with 2 lag years is significant. Similar to the core variables of the model
the FDI is strongly significant at1%.
18
6 – Model specification (fixed-effects, random-effects or
mixed effects) and results(cont..)
• 1% increase in distance will lead to 0.8% decrease in intra-WAEMU
agricultural exports
• 1% increase of GDP leads to 0.9% increase in intra-WAEMU agricultural
exports
• 1% in the FDI (with 2 lag years) will yield 0.25% increase in intra WAEMU
agricultural exports.
19
7– Conclusion and policy recommendations
These results call for some policy recommendations including the
following:
• Macroeconomic and structural reforms that can lead to an overall
increase of GDP and the level of development of WAEMU which will in
turn favor intra-WAEMU agricultural exports.
• Construction/modernization of West African Corridors to facilitate
connection among West African Countries
• Reforms in agricultural sector including attracting FDI to modernize the
sector and promote agricultural transformation as it appears in an
agricultural transformation strategy for Africa under preparation at
African Development Bank
20
• Thank you
21