Economie%20Mediterranee

1
International Workshop Bridging the gap the
role of trade and FDI in the Mediterranean Napoli
, Castel dellOvo 8-9 Giugno, 2006 FDI
Potential and Shortfalls in South Mediterranean
Countries Determinants and Diversion Effects A.
Ferragina (aferragina_at_unisa.it) F. Pastore
(fpastore_at_unina.it)
2
Outline

• FDI in the MED-10 vis-a-vis Central and Eastern
  Europe overview of performance and potential.
• Determinants of FDI with the help of the gravity
  model the role of institutional and policy
  variables.
• Simulation analysis of the magnitude of normal
  value of FDI expected on the basis of the
  explanatory variables.
• Diversion of FDI from MED to CEECs?

3
Introduction

• A worrying stylised fact Southern Mediterranean
  countries (MED so far) receive little FDI from
  most other regions in the world.
• FDI flows into some of them have tended to grow
  slowly over the 90s and to decline after 2000,
  while they have been booming in Central and
  Eastern Europe (CEECs).

4
World FDI outflows to CEEC10 and MED10

• Actual dollar values of FDI outflows from 1994 to
  2004 from the World to CEEC10 and MED10 million
  USD and as percentage of total FDI outflows
  (Unctad, 2005).

5
EU investment are crucial

• EU is the main provider of FDI to the South
  Mediterranean.
• In 2004 the EU provided on average more than 70
  of the FDI to MED10.
• The European presence is not equally distributed
  in all the MED but for some countries it is
  really striking Turkey received more than 75 ,
  Morocco 73 (more than 95 in 2001), Tunisia 65.
• According to the MIPO database from ANIMA, 59
  of the investment projects are coming from
  European investors, essentially France, Spain,
  the United Kingdom, and Germany.

6
Things are not going well

• EU FDI a slow trend of growth up to 1997
• an upsurge between 1997 and 1999 and a good
  performance up to 2002 overall from 5 to 15
  billions and from less than 1 to almost 14 of
  total EU FDI.
• However, from 2002 to 2004 a strong decline has
  eroded previous results a fall from almost 14 to
  4 per cent, corresponding to a decrease of more
  than 10 billions USD (from 15 to 5 billions ).

7
EU FDI outflows to CEECs and MED10

• actual dollar values of EU FDI outflows to
  CEEC10 and MED10 from 1994 to 2004 as percentage
  of total EU FDI outflows (Eurostat, Economics and
  Finance Statistics) this cursory evidence might
  suggest investment diversion from MED to CEECs

8
Source Eurostat, Economics and Finance
Statistics.
9
Perfomance and Potential Index

• two indicators for ranking countries with respect
  to FDI (UNCTAD, 2004)
• 1) a performance index which relates the share
  of the FDI flows to a country to its share in the
  world GDP. An index above (below) one indicates
  that the couuntry attract more (less) FDI as a
  percentage of its economic dimension
• 2) a potential index, which is calculated as
  simple average of 12 structural variables which
  corresponds to the main FDI determinants
  identified by theoretical and empirical models
  (market size, degree of openness,
  infrastructures, technologies, qualified labour
  force at low cost, natural resources endowment,
  regulatory framework, busineess climate and
  country risk which influence the degree of
  confidence of investors).

10
Performance of MED in 2004 ranking over 140
countries (Unctad, 2005)

• UNCTAD ranking of countries according to the
  inward FDI performance index MED lag behind at
  the international level Lybia is at the bottom
  of the list (116), followed by Turkey (111),
  Egypt (108), Algeria (95), Lebanon (90), Israel
  (83). Better Tunisia (67), Morocco (65), Jordan
  (48) and Syria (39).
• CEECs position is far better for most countries
  Bulgaria 12, Estonia 16, Slovak R. 25, Czech R.
  28, Croatia 33, Romania 35, Poland and Albania
  42, Hungary 46, Latvia 47, Lithuania 59, Slovenia
  60, Macedonia 72.

11
Potential of MED in 2003 ranking over 140
countries (Unctad, 2005)

• UNCTAD ranking of countries according to the
  inward FDI potential index in 2003 USA (1),
  China (12), Slovenia (28), while Egypt 75 Morocco
  87, Syria 95!
• The MED (excluding Israel) all from 60 downward.

12
Evolution of Performance

• Only in few MED changes occurred over the last
  decade through economic reforms and the adoption
  of more friendly policy towards investors have
  translated into significant improvement.
• This is suggested by the evolution of the
  performance index over the period 1993-95 and
  2000-2002.
• Jordan, Algeria and Lebanon have realised an
  important increase in the index of development of
  FDI.
• But Tunisia, Egypt, Morocco, Syria, Turkey show a
  negative evolution.

13
Evolution of Performance
14
Evolution of Potential

• Most of the countries with upsetting FDI
  performance (Turkey, Tunisia, Morocco, Egypt,
  Syria) have had a negative or static dynamic also
  as far as the potential index is concerned
• Only Israel shows a big increase in potential
  from 0,20 to almost 0,40

15
Evolution of Potential
16
Perfomance and Potential

• Comparing the Potential Index value with the
  Performance Index value gives an indication of
  how each country performs against its potential.
• Countries in the world can be divided into the
  following four categories
• front-runners (countries with high FDI potential
  and performance)
• above potential (countries with low FDI potential
  but strong FDI performance)
• below potential (countries with high FDI
  potential but low FDI performance)
• under-performers (countries with both low FDI
  potential and performance

17
Perfomance and Potential

• In the figure below, it is shown the position of
  the 10 Mediterranean partners with respect to
  both these two indicators in the period
  2000-2002.
• On the horizontal axis, it is portrayed the
  Performance index and on the vertical axis the
  Potential index.
• For comparison, to measure the distance of
  Mediterranean economies from international
  partners, there are also countries with high and
  average potentials, such as USA, China, Poland,
  Venezuela.
• Striking the potential value of China and Poland,
  quite close competitors for MED.

18
Perfomance and Potential Index

• Four partitions of the figure
• MED with high (low) performance those which on
  the horizontal axis intercept value of the inward
  FDI performance index above (below) one
• countries with high (low) potential those with a
  potential index above (below) the average for the
  Mediterranean area (South and North) as a whole

19
Perfomance and Potential Index

• 1) MED with high performance combined with high
  potential (frontrunners) only Israel, close to
  Spain, France, Slovenia, Portugal and Cyprus.
• 2) MED with high performance but low potential
  (countries above their potential) Morocco,
  Tunisia, close to Albania, Croazia, Macedonia.
• 3) MED with worrying position both as
  performance and as potential (underperformers),
  above all Egypt, Turkey, Syria, but also Jordan,
  Algeria, Lebanon.

20
(No Transcript)
21
(No Transcript)
22
Why so little FDI?

• Does the current situation with both low
  performance and low potential and with some
  countries even showing performance above their
  potential suggests that MED have reached an
  equilibrium although of low level?
• What role is played by distortions of an
  institutional and policy type that economic
  agents are submitted to in MED?

23
Determinants of FDI

• pick up the determinants of FDI flows to a large
  sample of host economies
• a gravity model is estimated to this purpose with
  panel data techniques based on aggregate
  country-level data on bilateral FDI flows of
  fourteen European countries and two non EU
  countries (USA and Japan) into a large sample of
  developed and developing partners (84), using
  among the relevant explanatory variables also
  institutional and policy factors for the years
  1994-2004

24
Simulation analysis

• Use these estimates to perform forecasts for FDI
  flows to both the single CEECs and Southern
  Mediterranean countries, subsequently comparing
  these estimated flows to actual flows
• For CEECs, the stock adjustment might have
  already taken place. Hence, actual and expected
  flows should not be strongly misaligned.
• On the contrary, we would expect that actual
  capital flows to MED are much below the expected
  flows because the stock adjustment process still
  has to take place.
• However, the current situation although not
  corresponding to an optimum allocation of
  resources, might also correspond to an
  equilibrium taking into account the distortions
  of various types that economic agents are
  submitted to in MED (actual flows not below
  expected).

25
Diversion of FDI from MED to CEECs?

• Here we use the gravity model to assess whether
  changes in FDI flows to CEECs which are
  economically integrating appear to be associated
  with negative changes in FDI flows to MED.
• Our methodological approach is based upon that of
  Sapir (1997) who sought to identify whether a
  domino effect had characterised the impact of
  European integration upon bilateral trade flows.
• We experiment by including interaction of
  regional dummies with dummy variables for
  particular sub-periods 1994-1998 (transition
  period), 1998-2004 (pre-accession period) and for
  years, checking how regional dummies change.

26
Choice of the model underlying theoretical
reasons

• Why firms produce abroad and face additional
  costs instead of simply servicing the markets via
  exports?
• Dunning (1977, 1981), the OLI framework,
  considers FDI as determined by Ownership,
  Location and Internalisation advantages which the
  MNC holds over the foreign producer.
• The so-called New Theory of FDI takes
  inspiration from the OLI approach and refers
  mainly to the Ownership and Location advantage to
  introduce MNCs in general equilibrium models,
  where they arise endogenously.
• The early literature (Helpman 1984, Helpman and
  Krugman 1985) was mainly able to explain
  vertical FDI, i.e. investment that takes place
  in order to take advantage of differences in
  relative factor endowments (hence in factor
  prices) across countries.

27
Other theoretical explanations

• There are also horizontal FDI similar types of
  production activities, owned by MNCs, taking
  place in different countries. This phenomenon is
  better clarified if multinational activity is not
  driven by factor endowments differences, but
  rather by the trade-off between proximity and
  concentration (Brainard 1993 Markusen and
  Venables , 1995).
• The proximity advantage stems from firm-level
  economies of scale, whereby RD activity (or any
  other type of knowledge capital) is
  transferable to affiliates and allows MNCs to be
  closer to the foreign market.
• The concentration advantage derives from
  traditional plant-level economies of scale,
  which make it more profitable to concentrate
  production in one location and then export.
• Whenever the former outweigh the latter, foreign
  investment will take place, and this will be more
  likely the higher are intangible assets relative
  to fixed costs of opening up an affiliate and the
  higher are transport costs, which are assumed to
  be positive and an increasing function of
  geographical distance in this model.

28
Gravity model

• When we get to the empirical analysis, to compare
  attractiveness across countries and explain the
  geographic distribution of FDI we need a model
  that can pick up all these common determinants.
• To synthesise the two approaches discussed above,
  i.e. Helpman and Krugmans treatment of vertical
  FDI and Brainards of horizontal one, we will
  include in the model the following main
  variables
• a measure of the economic space between the two
  countries, given by the sum of the two GDPs and
  by the two countrys populations to catch the
  market-seeking aspect of FDI
• the relative factor endowments, an index of
  countries similarity in size measured by their
  relative GDP.
• additional variables, such as distance, a common
  language, a common border, or preferential trade
  agreements, that may reduce the costs
  (transaction and transportation costs) of
  locating abroad and which can be introduced via
  dummy variables.
• This type of gravity model approach has been
  already applied to studies of FDI as a means of
  picking up the common determinants of FDI flows
  across countries (Eaton and Tamura, 1996 Brenton
  and Di Mauro, 1999).

29
Enlarged gravity model

• We apply this type of model but we propose a
  broad version
• factors traditionally considered in gravity
  models such as proximity and market size should
  make countries attractive locations for FDI and
  should play a decisive role, but
• moving from the consideration that the success of
  FDI attractiveness of CEECs was mainly due to the
  prospects of EU membership and to the fact that
  most CEECs have succeeded in attaining both
  institutional and political stability
• .we attempt to explain FDI shortfalls of MED
  with a gravity model enlarged to include policy
  and institutional factors

30
Specification

• ln (Bilat FDIijt)
• traditional gravity ß0 ß1 ln SUMGDPijt ß2 ln
  POPit ß3 ln POPjt ß4 lnDiffGDPPCijt ß5 ln
  Distij ß6ln Areasij ß7 LLij ß8
• Borderij ß9 Langij ß10 Colonialij
• policy and instit. ß11 Regionalijt ß12 ln
  (IMP/GDP)jt ß 13 ln (M2/GDP)jt ß 14 ERVijt
  ß 15 CUijt ß16 Govijt ß17 FTAijt ß18
  Humcapjt ß19 Current ß20 Capital
• regional and time dummies ß21 EUij ß22
  MED ß23 CEECs ß24 YEARDummies
• where i and j denotes donor and host country
  respectively, t denotes time
• FDIij is the value of the FDI flow from country i
  (home country) to country j (host country)
• and the variables are defined as follows

31
Traditional gravity variables

• SUMGDPijt is the sum of nominal value of the
  gross domestic product in i and j
• POPi and POPj, is the population of i and j
• DiffGDPPCijt is the absolute difference in per
  capita income between i and j (a proxy for
  relative factor endowment)
• Distij is the Great Circle Distance between i and
  j in miles
• Areas is the sum of the areas of i and j in
  square kilometres (hence a proxy for distance
  within the country to the border),
• LLij is a dummy variable, which is 0 if no
  countries are landlocked, 1 if one partner is
  landlocked
• Borderij is a binary variable, which is 1 if i
  and j share a border and 0 otherwise
• Langij is a binary variable, which is 1 if i and
  j share an official language and 0 otherwise
• Colonialij is a binary variable, which is 1 if i
  colonized j

32
Further variables

• Regionalijt is a binary variable, which is 1 if i
  and j belong to a Regional Trading Agreement in
  year t
• IMP/GDPj a proxy for the openness of a country to
  foreign trade
• ERVijt is the volatility of the bilateral nominal
  exchange rate between i and j in period t
• CUijt is a binary variable, which is 1 if i and j
  use the same currency at time t
• Govjt is the sum of six governance indices of j
  at t
• FTA is a dummy variable defined as 1 if only one
  of the countries is in a regional trading
  agreement (and 0 otherwise) proxi measure of
  trade diversion
• Current and Capital are variables coded 1 if host
  country has current and capital account
  restrictions respectively
• MED is a dummy variable which is 1 when host
  countries are MED10 countries, CEEC is a dummy
  variable for belonging to CEEC10, EU is a dummy
  variable for EU membership of i.

33
Instability of FDI over time

• Data on FDI flows at which we are looking here
  may be considerably biased upward or downward in
  a particular year
• For instance, a large merger and acquisition deal
  has taken place or a substantial portion of the
  domestic corporate sector has been privatized.
• Great instability in the coefficients over time.
• Year dummies are introduced to solve this problem.

34
Empirical results

• Here we present the results of the regression
  analysis of bilateral FDI flows by major
  investing countries over 1994-2004.
• The gravity model introduced is used to define a
  "normal pattern" of bilateral FDI flows.
• Dummy variables are included for three groups of
  countries EU, CEECs, MED10 to get a very
  preliminary test for a possible divergence from
  this pattern.
• If the corresponding coefficients are
  significant and negative, we interpret this as
  evidence that the group has received less FDI
  than other countries after controlling for all
  the other factors.
• Therefore, the group concerned can expect to
  benefit from further large FDI inflows as foreign
  investors adjust their stocks to the new
  opportunities created by economic transformation.
  
• If the dummies are not significant, the future
  growth of the FDI flows can be expected to be in
  line with changes in the determinants of FDI.

35
Determinants I

• First focus on few variables at the core of
  gravity models sum of GDP (a measure of mass),
  Population, Distance (measure of transport and
  transaction costs).
• The gravity variables have all the expected
  sign
• Increased economic space (SUMGDP and
  Population) have a noticeable impact on FDI.
  Elasticity of bilateral FDI with respect to
  population larger than 1.
• Distance appears to harm FDI something which is
  more intuitive in the case of exports. However,
  non linear relationship squared distance
  positive. Also theory suggests that firms will
  tend to prefer FDI to exports as trade costs, as
  proxied by distance, rise. More distant markets
  will tend to be served by overseas affiliates
  rather than by exporting.
• Differences in relative factor endowments have a
  negative impact on FDI from the theoretical
  discussion above one can infer that, on average,
  EU investors are in general more prone to
  horizontal than to vertical FDI.

36
Regional Dummies

• MED negative and significant at 10
• CEECs positive but not significant
• EU positive and significant at 1 .

37
Determinants I

• -------------------------------------------------
  -----------------------------
• Log FDI Coef.
  Std. Err. z Pgtz
• ------------------------------------------------
  -----------------------------
• Log sum gdp .1151378
  .0075292 -15.29 0.000
• Log population host .8594187 .0551189
  15.59 0.000
• Log population donor 1.304252 .0645561
  20.20 0.000
• Difference of GDP p.c. -.9031703 .048244
  -18.72 0.000
• Distance -3.99e-07
  6.29e-08 -6.35 0.000
• Distance2 2.32e-14
  4.06e-15 5.72 0.000
• EU .648041
  .1534813 4.22 0.000
• MED10 -.5623256
  .314175 -1.79 0.073
• CEEC10 .0778871
  .2526697 0.31 0.758
• _cons -27.60273
  1.496819 -18.44 0.000
• -------------------------------------------
  -----------------------------

38
Determinants II enlarged gravity

• As expected Border, Language, Regional agreement
  all have positive and significant effects on FDI
  levels.
• CU is significant and positive.
• FTA not significant effect, indicating that for
  the full sample there has been no discernible
  trade diversion effect of FTAs.
• Colonial links not significant too.
• The import on GDP coefficient is also not
  significant.
• Volatility of exchange rate is positive and
  significant FDI more stable and less risky than
  portfolio and trade activities?
• Governance is highly significant with positive
  sign (very robust variable).
• Current and capital account restrictions are both
  negative and highly significant.

39
Regional Dummies

• MED no more significant all the negative
  difference with respect to other countries seems
  to be explained by the institutional and policy
  variables
• CEECs after introducing the institutional
  variables becomes negative and significant the
  group has received less FDI than other countries
  after controlling for all other factors.
• EU not significant the future growth of the FDI
  flows can be expected to be in line with changes
  in the determinants of FDI.

40
Log FDI Coef. Std. Err. z Pgtz
Log sum GDP .1102201 .0410843 2.68 0.007
Log population host 1.068.686 .0581647 18.37 0.000
Log population donor 1.181.239 .0662389 17.83 0.000
Difference in GDP pc -.3142086 .0527492 -5.96 0.000
Distance -3.24e-07 5.78e-08 -5.60 0.000
Distance squared 1.76e-14 3.69e-15 4.78 0.000
Log area -.2154785 .0396847 -5.43 0.000
Common border .8735025 .3147306 2.78 0.006
Common language .9506041 .3273679 2.90 0.004
Regional integration .293194 .0712112 4.12 0.000
Colonial links .6739344 .4622208 1.46 0.145
Import/GDP .002354 .0017827 1.32 0.187
Exchange rate volatility .223984 .0212081 10.56 0.000
Governance .0687912 .0037609 18.29 0.000
Current account restrictions -.1866332 .0440186 -4.24 0.000
Capital account restrictions -.0811416 .0455017 -1.78 0.075
Currency union .4337308 .0593678 7.31 0.000
FTA .053803 .0513012 1.05 0.294
EU -.1302178 .1475229 -0.88 0.377
MED10 -.306008 .2716849 -1.13 0.260
CEEC10 -.4018635 .218197 -1.84 0.066
_cons -3.325 1.462.512 -22.74 0.000
41
Simulation Analysis Actual vs. Expected FDI

• The simulation concerns the expected or normal
  FDI flows to each single MED and CEECs (that
  would be expected based on the empirical
  benchmark model described above) compared with
  actual FDI flows.
• Our expectation would be that, if the adjustment
  process was quite fast (slow) FDI flows are above
  (below) the average level of flows expected to
  countries with comparable attributes.
• For MED we would expect not to observe a strong
  catching-up effect in the past years in spite of
  the fact that these economies are underdeveloped
  as compared to average industrialised countries
  and in need to adjust and to catch up too.

42
Germany FDI flows to MED simulated in of
actual
43
Italy FDI flows to MED simulated in of actual
44
Simulation results by country

• The calculations for these four MED indicate that
  most of them attracted almost 100 of the total
  expected FDI inflows.
• A downward shift in inflows for Israel with
  slight increases for Egypt and Morocco.

45
Diversion of FDI from MED to CEECs?

• Here we use the gravity model to assess whether
  changes in FDI flows to CEECs appear to be
  associated with changes in FDI flows to MED.
• In particular, we check whether increasing CEEC
  integration over the 1990s, culminating into the
  accession for most of them, had any noticeable
  negative impact upon FDI flows from EU countries
  going to the MED10.
• We experiment by including the interaction of
  regional dummies with time dummies for particular
  sub-periods 1994-1998 (transition period),
  1998-2004 (pre-accession period) and for years.

46
Results by years

• Interaction of regional dummies with time
  dummies how they have been changing over two
  periods (1994-1998, 1999-2004)? and year by year?
• Negative and significant dummy for CEECs both in
  1994-1998 and in 1999-2004 (most years show a
  negative and significant dummy).
• For MED negative and significant for the first
  period but not significant and positive in the
  second period (in each year from 1999 to 2003).

47
(No Transcript)
48
Diversion of FDI from MED to CEECs?

• We find that MED countries in the 1994-1998 were
  receiving substantially less FDI than could be
  expected on the basis of their incomes and
  proximity to the EU and to other variables.
• However, the magnitude of this underpotential
  weakened in the late 1990s and in the first half
  of the 2000s which may suggest that the
  enlargement process did not adversely affected
  the magnitude of inward FDI from EU countries.
• Hence, our, albeit limited, analysis finds no
  evidence to suggest that the intensification of
  FDI in CEECs, following integration with the EU,
  has had a discernible dampening effect on FDI
  flows going to MED.

49
Conclusions

• The current situation of FDI in MED although does
  not correspond to an optimum allocation of
  resources corresponds to an equilibrium taking
  into account the distortions that economic agents
  are submitted to.
• Rejection of the hypothesis of FDI diversion away
  from the MED.
• FDI in these countries have come down but
  institutional and policy variables the main
  reason.
• It is from the issue of the FDIs that one can
  best perceive the necessity to modify the
  business environment and the behaviour of the
  enterprises but also the role of anchorage which
  EU can play.
• During the same period, Eastern Europe offered
  promising long term perspectives on this ground
  enhanced by the perspective of adhesion which
  also offered investors a guarantee of regulatory
  and institutional reforms
• Industries that faced difficulties (most of the
  time public ones), have been privatised and
  restructured under the impulse and with the help
  of Europe, a process that did not take place at a
  sufficient scale in the MED
• As a result CEECs moved towards a deeper
  integration not limited to a few tariff
  evolutions and are progressively adapting their
  legal framework and their practices to
  international standards