-
Estimating InfrastructureEstimating Infrastructure’’s Potential
as a s Potential as a Catalyst for Asian Regional Integration,
Catalyst for Asian Regional Integration,
Growth, and Economic ConvergenceGrowth, and Economic
ConvergenceDavid RolandDavid Roland--Holst, UC BerkeleyHolst, UC
Berkeley
joint work withjoint work withIfzalIfzal Ali and Fan Ali and Fan
ZhaiZhai, ADB, ADB
presented at theADB/ERD Experts’ Meeting on Long Term
Scenarios
of Asia’s Growth and TradeNovember 10, 2005Manila,
Philippines
-
10 November 2005 Roland-Holst 2
Contents
1.Introduction and Motivation2.Overview of Initial
Conditions3.Estimation Strategies4.Discussion
-
10 November 2005 Roland-Holst 3
1. Introduction
• As Asia’s economic growth process matures, regional
integration offers important opportunities:– Geographic
diversification/new markets– Superior growth rates– Structural
differentiation – more rapid evolution
from established North-South patterns of trade and
specialization
• Propagating growth linkages across this diverse region will
also facilitate economic convergence.
• Infrastructure commitments will be an essential guarantor of
this process.
-
10 November 2005 Roland-Holst 4
Motivation
Infrastructure’s contribution can be seen from plays three
roles:
1. Keynesian – Aggregate demand and employment stimulus.
2. Ricardian – Reducing trade margins and intensifying
comparative advantage.
3. Neoclassical – Endogenous growth benefits.
-
10 November 2005 Roland-Holst 5
Keynesian Stimulus
• Infrastructure spending is a popular means of direct long term
or transitory employment stimulus– Examples: WPA (US), Work Relief
(PRC), Japan
(heavy counter-cyclical and recurrent fiscal commitments)
• Because of its generality, this kind of spending can be
targeted across a wide spectrum of regions and socio economic
groups
• For public good infrastructure multiplier effects are
generally quite substantial
-
10 November 2005 Roland-Holst 6
Ricardian Stimulus
By reducing trade margins, infrastructure:
1. Intensifies comparative advantage
2. Improves international terms of trade
3. Improves rural terms of trade (pro-poor)
and
4. Extends the horizon of profitable investment and marketing
(falling MC means economies of scale)
1⎯⎯⎯ →⎯++
∞→MF
H
MPMP
↓+
↑−
⇒↓DD P
MPWMandP
MPWEM
MPMP
PP
D
DR
U
RR
+−
==ρ 2)(2
MPP
M DD
+−=
∂∂ρ
-
10 November 2005 Roland-Holst 7
Neoclassical Stimulus
Modern economic theory recognizes many endogenous growth
factors, and these can be greatly facilitated by infrastructure:–
Productivity enhancement– Technology diffusion– Information
diffusion– Supply chain articulation and other
network externalities– Human capital development (migration)
-
10 November 2005 Roland-Holst 8
2. Overview of Initial Conditions
• Infrastructure conditions across Asia are highly variegated,
even between neighboring countries.
• Infrastructure expansion trends have been dramatic, but only
in a few countries.
• Public investment needs to extend its development leadership,
increasing returns to market participation and private investment
across the region.
• Public and private capital commitments must be complementary
if the former is to be effective.
-
10 November 2005 Roland-Holst 9
Paved Road Systems I
0
1
2
3
4
5
6
1960 1965 1970 1975 1980 1985 1990 1995 2000
Pave
d Ro
ad L
engt
h / L
and
Area
BangladeshChinaHong
KongIndiaIndonesiaJapanKoreaLaosMalaysiaMongoliaNepalPakistanPhilippinesSingaporeThailand
-
10 November 2005 Roland-Holst 10
Paved Road Systems II
0
1
2
3
4
5
6
7
8
1960 1965 1970 1975 1980 1985 1990 1995 2000
Pave
d R
oald
Len
gth
per C
apita
BangladeshChinaHong
KongIndiaIndonesiaJapanKoreaLaosMalaysiaMongoliaNepalPakistanPhilippinesSingaporeThailand
-
10 November 2005 Roland-Holst 11
Electrification
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
1960 1965 1970 1975 1980 1985 1990 1995 2000
Elec
tric
Capa
city
/ C
apita
BangladeshChinaHong
KongIndiaIndonesiaJapanKoreaLaosMalaysiaMongoliaNepalPakistanPhilippinesSingaporeThailand
-
10 November 2005 Roland-Holst 12
Mobile Telephony
0
100
200
300
400
500
600
700
800
900
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Mob
ile U
sers
per
Tho
usan
d Po
pula
tion
BangladeshChinaHong
KongIndiaIndonesiaJapanKoreaLaosMalaysiaMongoliaNepalPakistanPhilippinesSingaporeSri
LankaThailand
-
10 November 2005 Roland-Holst 13
Income and Infrastructure
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10000 20000 30000 40000 50000
Income per Capita
Nor
mal
ized
to M
axim
um
Pavement/AreaElec/CapMobile
Domestic private and public savings pose a serious
constraint.
-
10 November 2005 Roland-Holst 14
Regional Public and Private Investment(Asian inbound Aid and
FDI, USD Billions)
0
20
40
60
80
100
120
140
160
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
ODA-Asia FDI-Asia
Clearly, we have entered an Age of Complementarity.
-
10 November 2005 Roland-Holst 15
Infrastructure and Trade
• Trade has been a primary driver of Asian growth.•
Infrastructure is an essential complement to trade.• Past reliance
on demand outside the region was
facilitated by maritime expansion, with higher income countries
leading and limited spillovers.
• To support greater regional trade/integration, more diverse
infrastructure will be needed, financed collaboratively and
propagating growth externalities.
• This approach will facilitate economic diversification,
growth, and convergence.
-
10 November 2005 Roland-Holst 16
Asian Trade Flows, 2005(percent of total Asian trade)
Extra-regional demand remains a primary economic driver.
Asian regional trade is far from reaching its potential.
SoutheastAsia
South Asia
East andCentral Asia
14.2
4.6
0.5
29.3
5.0
3.2
0.4
0.4
0.2
0.1
22.1
6.9
2.0
2.4
8.8
-
10 November 2005 Roland-Holst 17
Baseline Per Capita GDP Growth(annualized percent change,
2005-2025)
-1 0 1 2 3 4 5 6 7 8
PRC
Sri Lanka
PhilippinesThailand
Viet NamHong Kong, China
Malaysia
IndiaKorea
Indonesia
BangladeshTaipei,China
Singapore
Australia, NZUnited States
Latin AmericaRest of World
Europe 17
Japan
Real GDP/Cap Pop
On a global basis, Asia continues to represent superior
growth.
Sources: DRI, Oxford Econometrics, IMF.
-
10 November 2005 Roland-Holst 18
Trade Distribution
Trade capacity and income are linked, particularly for regional
trade.
Infrastructure is a primary reason.BangladeshViet Nam
IndonesiaIndia Sri Lanka
PhilippinesChinaThailand
Malaysia
Korea
Singapore
Japan
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100
Cumulative Percent of Trade
Cum
ulat
ive
Perc
ent o
f Asi
an G
DP
All TradeEqualityAsian ImportsAsian Exports
-
10 November 2005 Roland-Holst 19
Notes on “Soft Infrastructure”
• Much emphasis is given to infrastructure goods, but there is
also a broad universe of infrastructure services.
• Concentrated in national and multilateral public institutions,
they constitute an essential element of trade facilitation.
• Trade negotiating institutions are only the most conspicuous
members of a large family of institutions promoting more coherent
market linkages and policy dialogue.
-
10 November 2005 Roland-Holst 20
Soft Infrastructure
• Our recent research indicates that structural barriers to
trade are now much more important regional growth impediments than
formal trade barriers.
• For this reason, more determined initiatives for both hard and
soft infrastructure are essential.
• In the latter category, an Asian OECD could make important
contributions.
-
10 November 2005 Roland-Holst 21
3. Estimation Strategies
• To assess the role of Asian infrastructure empirically, and
economywide and regionwide approach is needed.
• For this reason ADB/ERD has developed multi-country macro and
CGE models in parallel.
• Appropriate use of these tools depends on the kind of
infrastructure issues to be addressed.
-
10 November 2005 Roland-Holst 22
Keynesian Effects
• This category of effects focuses on fiscal commitments and
aggregation demand and employment linkages.
• At the national level, a standard macromodel can capture most
of this.
• For more targeted (sector and area) policies, project analysis
or partial equilibrium estimates are serviceable.
• Without a GE framework, however, many indirect effects may be
omitted.
-
10 November 2005 Roland-Holst 23
Ricardian Effects
• Any policies that significantly influence margins will induce
extensive market spillovers.
• Distribution effects are likely to be at least as prominent as
growth effects.
• In this context, only an economywide and regionwide GE model
will suffice.
-
10 November 2005 Roland-Holst 24
ERD Research on Structural Barriers to Trade
• An example of estimating Ricardian effects is our recent work
on structural barriers to regional trade.
• We found, among other things, that reducing trade margins
within the region achieved dramatically larger growth benefits than
removing nominal protection.
-
10 November 2005 Roland-Holst 25
0%
10%
20%
30%
40%
50%
60%
70%
80%
2005 2010 2015 2020 2025
GBL0 AFTA2 GBL2
East and Central Asia
0%
10%
20%
30%
40%
50%
60%
70%
80%
2005 2010 2015 2020 2025
GBL0 AFTA2 GBL2
Southeast Asia
0%
10%
20%
30%
40%
50%
60%
70%
80%
2005 2010 2015 2020 2025
GBL0 AFTA2 GBL2
South Asia
Income Growth(incremental percentage change from Baseline)
- WTO with Facilitation
- AFT with Facilitation
- WTO
-
10 November 2005 Roland-Holst 26
Neoclassical Effects
This is a vast area, that by definition is well suited to GE
analysis.
Some issues are more prominent than others:1. Economies of scale
– a primary driver of
modern regionalism2. Productivity growth3. Network
externalities, including supply
chains.4. Migration and Human Capital
-
10 November 2005 Roland-Holst 27
Economies of Scale
Ricardian comparative advantage emphasizes national
differentiation, but modern regionalism is increasingly driven by
shared characteristics, especially two catalysts for economies of
scale:– Multinational marketing: aggregating
regional consumer groups– Supply chains: aggregating
regional
component markets
-
10 November 2005 Roland-Holst 28
Economies of Scale
• We are currently working with an increasing returns,
differentiated product specification of the global GE model.
• Initial results confirm the importance of market
extensification to efficiency and growth.
• The main challenge is to improve empirical information for
calibrating scale effects.
-
10 November 2005 Roland-Holst 29
Productivity Growth
Infrastructure contributes to productivity growth in two ways:–
Direct increases in productivity for trade,
transport, and other wholesale and retail distribution
services.
– Indirect increase via technology diffusion.The former can be
modeled directly, and the results are
significant in terms of both extensive an intensive trade
facilitation.
The latter process is very diverse and best studied on a case by
case basis. In some cases, such as telecoms, benefits are pervasive
and obvious. In others, such as public health, economic benefits
are more complex.
-
10 November 2005 Roland-Holst 30
Network Externalities
• Supply chain decomposition has been a salient trend in recent
Asian growth
• The corresponding intermediate trade linkages are increasingly
responsible for the majority of value creation
• The foreign capital, technology, and demand allocation
accompanying this propagates growth and, because of wage
competition, can promote convergence
• Infrastructure is a prerequisite for effective participation
in this regional production sharing
-
10 November 2005 Roland-Holst 31
A Regional Example - Bamboo Capitalism
• Network externalities in local production and finance allow
complete markets to sprout from nodes in a global root system of
intermediate supply.
• This culminating aspect of global supply chain decomposition
has created a diverse and vibrant population of independent local
industries around the East Asian region.
• Many emergent enterprises are still bound to their roots by
ownership or contracts
• But increasingly they arise independently, promoting the
dynamics of global competitiveness and innovation.
-
10 November 2005 Roland-Holst 32
Estimating Network Externalities
• Evidently, this is an important issue for regional trade
generally and elucidating the role of infrastructure in
particular.
• Multiplier decomposition can shed some light, but has
conceptual limitations
• To model these effects more convincingly, requires attention
at all levels: data, specification, and experimental design.
-
10 November 2005 Roland-Holst 33
Discussion
-
10 November 2005 Roland-Holst 34
Overview of Conclusions
• Regional trade and integration offer Asia great potential for
more rapid and sustained growth. Indeed, much of Asia’s benefits
from global trade liberalization can be realized by regional
initiative alone.
• Structural barriers to trade are now more important than
tariffs. Policies and investments that facilitate trade can
accelerate regional growth dramatically.
• Regional integration can promote Asian economic convergence,
raising average growth rates and benefiting poorer countries.
Estimating Infrastructure’s Potential as a Catalyst for Asian
Regional Integration, Growth, and Economic Convergence Contents1.
IntroductionMotivationKeynesian StimulusRicardian
StimulusNeoclassical Stimulus2. Overview of Initial ConditionsPaved
Road Systems IPaved Road Systems IIElectrificationMobile
TelephonyIncome and InfrastructureRegional Public and Private
Investment�(Asian inbound Aid and FDI, USD Billions)Infrastructure
and TradeAsian Trade Flows, 2005�(percent of total Asian
trade)Baseline Per Capita GDP Growth�(annualized percent change,
2005-2025)Trade DistributionNotes on “Soft Infrastructure”Soft
Infrastructure3. Estimation StrategiesKeynesian EffectsRicardian
EffectsERD Research on Structural Barriers to TradeIncome
Growth�(incremental percentage change from Baseline)Neoclassical
EffectsEconomies of ScaleEconomies of ScaleProductivity
GrowthNetwork ExternalitiesA Regional Example - Bamboo
CapitalismEstimating Network ExternalitiesOverview of
Conclusions