ARE THERE SHORTFALLS in MENA TRADE? To What Extent Are They Due to the Rise in EU Trade and Investment and Other Factors?

1
ARE THERE SHORTFALLS in MENA TRADE? To What
Extent Are They Due to the Rise in EU Trade and
Investment and Other Factors?

• RANIA S. MINIESY
• BRITISH UNIVERSITY IN EGYPT
• JEFFREY B. NUGENT
• UNIVERSITY OF SOUTHERN CALIFORNIA
• June 8, 9, 2006
• CNR Institute of Studies on Mediterranean
  Studies
• International Conference on Bridging the Gap
  The Role of Trade and FDI in the Mediterranean

2
INTRODUCTION

• Evidence Although still controversial, important
  evidence that trade growth is causal to income
  growth (Frankel Romer 1999).
• Growth has lagged in MENA at least until the
  exploding oil prices of the last two years.
• Could slow trade growth contribute to this?

3
Why has MENA trade not been thoroughly
investigated as a slow-growth determinant?

• 1) Other, possibly potentially more important,
  slow-growth determinants have been investigated.
• Role of institutions (Elbadawi 2005 , Kuran 2004)
• Human Capital Formation (UNDP 2003,2004,
  Pritchett, 1996)
• High fertility rates and rapid population growth.
• (Williamson and Yousef 2002).
• Inflexible labor market institutions
  (Pissarides and Vergazones-Varoudakis 2006,
  Campos, Hsiao and Nugent 2005).
• Capital accumulation

4
Why has MENA trade not been thoroughly
investigated as a slow-growth determinant?

• 2) The presence of numerous methodological
  difficulties that may have discouraged empirical
  investigations of both the determinants of trade
  and its effects on growth.
• a. Problems of measurement of trade policy
• b. Traditional measurement OPENNESS is an
  outcome variable (interdependence between trade
  and trade policy and economic growth.
• but c. Frankel Romer 1999s innovation.

5
Why has MENA trade not been thoroughly
investigated as a slow-growth determinant?

• 3) Effects of trade on growth have sometimes
  been found to be considerably weaker for MENA
  countries than for other countries.
• (Makdisi, Fattah and Limam 2005)
• 4) The most common measure of openness
  (XM)/GDP is often unusually large for MENA
  countries leading to conclusion that MENA
  countries do not trade too little

6
Purpose of this paper

• To use the gravity model to answer a series of
  questions
• 1) Do MENA countries trade more or less with
  themselves and with other countries than would be
  expected on the basis of gravity type
  explanations?
• MENA trades less than expected in both cases
• 2) What determines the extent of these trade
  shortfalls and the extent to which they might be
  reduced by the introduction of additional
  controls?

7
Purpose of this paper

• 3) Does EU-EE relationship compete with EU-MENA?
  
• EE has outperformed MENA in attracting FDI
• IFDI Performance Indices 1990-96
  1997-2004
• MENA 1.28 0.66
• EE 1.82 1.84
• 4) Can the shortfalls be reduced by introducing
  variables representing political shocks, (civil
  and international wars), trade policy
  considerations, and institutional considerations
  such as legal traditions and governance?

8
Organization of the paper

• Section 2 Gravity model and applications to the
  MENA region.
• Section 3 Identifies Data base used in this
  study, presents results with and without a dummy
  variable for MENA, and as additional explanatory
  variables are added.
• Section 4 Compares results across different
  specifications and estimation procedures
• Section 5 Compares MENA with East Asia and the
  MENA2 and MENA-World Trade Shortfalls across
  Specifications Summarizes and contains
  conclusions for policy and further research.

9
ALTERNATIVE MODELS FOR TRADE

• 1. Trade equations involving relative prices
  and other relative factors that reflect
  comparative advantage considerations.
• 2. CGE models.
• 3. Gravity models.

10
Advantages of Gravity Models

• Can be derived from underlying microeconomic
  foundations (Bergstrand, 1985).
• Less subject to simultaneous and omitted variable
  biases as compared to the other two models.
• Use much more reliable databases than CGE models.
• Results shown to be quite robust to different
  specifications, modeling assumptions and data
  sets and applicable to all kinds of countries and
  time periods (Leamer and Levinsohn, 1995)
• Qualifications (Anderson and Van Wincoop) Can be
  Mitigated by Including Broad range of size and
  other controls and Fixed effects

11
Gravity Models and MENA Trade

• Applications to MENA region are relatively rare.
• Ekholm et al. (1996)
• Low Potential for trade growth both within MENA
  and with EU. (May have discouraged further
  investigations in MENA).
• Limitations of Study
• Small sample of countries 13 DCs and 11 LDCs for
  one year
• Small number of variables included
• Findings Counterintuitive and inconsistent with
  most other studies. Hence, results could be
  misleading.
• Other Studies Al-Atrash and Yousef (2000)
• Miniesy, Nugent (2005), Miniesy, Nugent and
  Yousef (2004)

12
THE GRAVITY MODEL

• The standard gravity model estimated in the
  literature and followed with certain
  modifications here is
• Ln (Tijt) ß0 ß1 Ln Xijt ß2 Ln Yij ?1ij
  ?2t uijt
• Data Virtually all countries in the world for
  1970-2000 for data taken at three to five year
  intervals.
• We focus on aggregate trade flows.
• The loglinear equation is limited to the
  variables that are continuous. Many of the
  variables are either qualitative or dummy
  variables which are non-continuous.

13
Results Basic Gravity
gdp 1.021753 (0.02)
gdppc_time -0.0363614 (0.00)
LnGDPPCdiff -0.2578801 (0.07)
gini_final -0.0122259 (0.01)
gini_lngdpf 0.0030216 (0.00)
ll -0.2464937 (0.05)
14
Results More Basic Gravity
language 0.4676306 (0.06) cu 1.664587 (0.18)
regional 0.6934212 (0.18) OneFTA -0.0057971 (0.05)
colonizer 1.198327 (0.17) OneFTA90s -0.051591 (0.02)
colonial 1.057564 (0.51) m2gdp 0.0000944 (0.00)
15
Less Standard Gravity Variables
Common_L -0.1068 (0.08) war_intl -0.1479 (0.22)
French_L -0.1030 (0.06) war_intl_intensity 0.0791 (0.12)
German_L -1.668 (0.47) war_civil -0.0882 (0.03)
Socialist_L 2.7892 (0.30) war_civil_intensity -0.0731 (0.04)
oil1npn -0.0045 (0.00)
16
Impacts of MENA Interactions with Variables Named
Below
Variable Basic Model Basic Model gdppc distance areas border gov_ds
gdppc 0.32 0.32 0.35 0.32 0.32 0.32 0.34
distance -1.3 -1.3 -1.31 -1.4 -1.3 -1.3 -1.31
areas -0.12 -0.12 -0.12 -0.12 -0.14 -0.12 -0.12
border 0.63 0.63 0.39 0.64 0.62 0.7 0.68
gov_ds 0.09 0.09 0.09 0.09 0.09 0.09 0.09
MENA2 -2.38 -2.38 2.32 -5.71 -3.68 -2.33 -2.21
MENA2_World -1.13 -1.13 -0.46 -2.94 -2.25 -1.13 -1.15
MENA2_gdppc -0.3
MENA2_distance 0.44
MENA2_areas 0.1
MENA2_border -0.36
MENA2_gov_ds 0.18
MENA2_World_gdppc -0.04
MENA2_World_distance MENA2_World_distance MENA2_World_distance 0.21
MENA2_World_areas MENA2_World_areas 0.08
MENA2_World_border MENA2_World_border 0.2
MENA2_World_gov_ds MENA2_World_gov_ds -0.03
17
Effects of Adding MENA2, MENA2_W and EU-EE
Interactions with Year
Variable 1. Basic Model 2 MENA2 MENA 2-World 3. Col. 2 EU_EE
gdppc 0.32 0.3087 0.3085
distance -1.3 -1.4006 -1.4012
areas -0.12 -0.1480 -0.1483
border 0.63 0.6572 0.6568
gov_ds 0.09 0.1253 0.1253
MENA2 -2.38 -2.8553 -2.8578
MENA2_World -1.13 -1.1860 -1.1868
MENA2_1997 0.3974 0.3974
MENA2_2000 0.2231 0.2238
MENA2_W1997 -0.2670 -0.2670
MENA2_W2000 -0.3807 -0.3800
EU_EE -0.9443
EU_EE_only2000 0.2507
18
SUMMARY OF THE FINDINGS

• MENA countries trade less than their Asian
  counterparts and than would be predicted on basis
  of gravity model
• But less so for more distant countries
• 2) Why?
• Less trade-friendly policies
• Weaker governance characteristics
• Coincidence of EU-EE Trade expansion and MENA
  decline but these do not affect the MENA trade
  shortfalls.

19
SUMMARY OF THE FINDINGS

• 3) Having a common border has a less positive
  impact on trade among MENA countries, suggesting
  that customs procedures, cooperation on
  procedures, infrastructure, at borders within
  MENA may be more detrimental to trade than
  elsewhere.
• 4) Generally speaking, controlling for
  non-observed omitted factors through fixed
  effects increases size of the estimated
  shortfalls. Hence studies without fixed effects
  may underestimate the shortfalls

20
TRADE POLICY RECOMMENDATIONS

• Eliminate existing restrictions on current
  account transactions.
• Improve internal transport and communications
  infrastructure
• Simplify and improve customs procedures at the
  borders so that having common border with a
  trading partner would exert a stronger positive
  influence on trade of MENA than at present.
• Reduce the likelihood and strength of civil and
  international conflicts that exert rather sizable
  negative influences on trade.

21
TRADE POLICY RECOMMENDATIONS

• Increase financial depth measured by M2/GDP.
• Promote greater and better implemented trade
  pacts with both MENA and non-MENA countries
• Improve compliance with agreements with penalties
  for non-compliance.
• Improve governance in MENA countries and choose
  partners with better governance and at same
  levels of per capita income.
• EU should impose stricter time limits on MENA for
  homogenizing policies and institutions.