ADBI Working Paper Series
Navigating a Changing World
Economy: ASEAN, the People’s
Republic of China, and India
Peter A. Petri and Fan Zhai
No. 404
January 2013
Asian Development Bank Institute
Peter A. Petri is the Carl J. Shapiro Professor of International Finance at Brandeis
University, a senior fellow of the East-West Center in Honolulu, and a visiting fellow of
the Peterson Institute for International Economics in Washington. Fan Zhai is acting head
and managing director of Asset Allocation and Strategy Research Department at China
Investment Corporation (CIC).
This paper is part of the Asian Development Bank Institute study on The Great
Transformation: ASEAN, the PRC, and India.
The views expressed in this paper are the views of the authors and do not necessarily
reflect the views or policies of ADBI, the Asian Development Bank (ADB), its Board of
Directors, or the governments they represent. ADBI does not guarantee the accuracy of
the data included in this paper and accepts no responsibility for any consequences of
their use. Terminology used may not necessarily be consistent with ADB official terms.
The Working Paper series is a continuation of the formerly named Discussion Paper series; the
numbering of the papers continued without interruption or change. ADBI’s working papers
reflect initial ideas on a topic and are posted online for discussion. ADBI encourages readers to
post their comments on the main page for each working paper (given in the citation below).
Some working papers may develop into other forms of publication.
Suggested citation:
Petri, P.A. and F. Zhai. 2013. Navigating a Changing World Economy: ASEAN, the People’s
Republic of China, and India. ADBI Working Paper 404. Tokyo: Asian Development Bank
Institute. Available: http://www.adbi.org/workingpaper/2013/01/22/5457.navigating.changing.world.economy/
Please contact the authors for information about this paper.
Email: ppetri@brandeis.edu; zhaifan@china-inv.cn
Asian Development Bank Institute
Kasumigaseki Building 8F
3-2-5 Kasumigaseki, Chiyoda-ku
Tokyo 100-6008, Japan
Tel:
Fax:
URL:
E-mail:
+81-3-3593-5500
+81-3-3593-5571
www.adbi.org
info@adbi.org
© 2013 Asian Development Bank Institute
ADBI Working Paper 404
Petri and Zhai
Abstract
Most projections envision continued rapid growth in the members of the Association of
Southeast Asian Nations (ASEAN), the People’s Republic of China (PRC), and India
(collectively, ACI) over the next two decades. By 2030, they could quadruple their output,
virtually eliminate extreme poverty, and dramatically transform the lives of their more than 3
billion citizens. The impact will be felt across the world. This study—a background paper to an
Asian Development Bank report—used a Computable General Equilibrium model to examine
the likely effects of the region's growth on trade, resources and the environment, as well as the
implications of the many risks the region's growth path faces from its internal and external
environment.
JEL Classification: F02, F13, F33, F53
ADBI Working Paper 404
Petri and Zhai
Contents
1.
Introduction: Analyzing the Great Transformation ........................................................... 3
2.
Methodology ................................................................................................................... 4
2.1
2.2
2.3
Simulation strategy ............................................................................................... 4
Scenarios.............................................................................................................. 4
Modeling framework.............................................................................................. 5
3.
Projections of ACI growth ............................................................................................... 6
4.
Analysis of the Baseline: Dimensions of Transformation ................................................13
4.1
4.2
4.3
4.4
5.
A leap in the quality of life ....................................................................................13
New drivers of growth ..........................................................................................17
Resource and environmental challenges .............................................................22
Expanding global role ..........................................................................................24
Threats to the Transformation ........................................................................................25
5.1
5.2
5.3
The threat of the middle income trap....................................................................28
Structural shocks and policy alternatives .............................................................34
International linkages and cooperation .................................................................37
6.
Combinations of Risks ...................................................................................................40
7.
Conclusions ...................................................................................................................42
References ...............................................................................................................................45
Appendix 1: Regions and Production Sectors ...........................................................................47
Appendix 2: Technical Description of the CGE Model ...............................................................48
Appendix 3: Consumption Distribution Side Model ....................................................................50
ADBI Working Paper 404
Petri and Zhai
1. INTRODUCTION: ANALYZING THE GREAT
TRANSFORMATION
The Association of Southeast Asian Nations (ASEAN), the People’s Republic of China (PRC),
and India (collectively, ACI) have become the world’s principal growth engine. They avoided or
recovered robustly from the Asian Financial Crisis in the late 1990s and have been the world’s
most dynamic economies since the Global Financial Crisis. Over the past decade, their real
gross domestic products (GDPs) grew by 5.5, 9.6, and 7.1%, respectively (ABD 2009). By 2030,
the ACI countries are poised to quadruple their output. If they realize these extraordinary
projections, they will dramatically transform the lives of their more than 3 billion people and
indeed the structure of the world economy.
ADB’s Asia 2050 study argues that the prospects of the ACI economies remain very favorable
(ADB 2011). But the report also emphasizes that continued success cannot be taken for
granted; Asia in general and the ACI economies in particular face daunting risks. This study
analyzes those risks and potential policy responses in quantitative detail. This is an inherently
speculative undertaking—even with a shorter time limited to 2030—since the world economy is
likely to change dramatically over the next two decades, including in ways that we cannot
anticipate. The paper models probable structural changes in regional and industry detail and
illustrates the wider range of possible developments with scenario analysis.
Our central thesis is that ACI economies are now embarked on a major transformation of their
economies and role in the world economy—a development that the ACI study describes as the
“great transformation.” Their growth is likely to develop indigenous drivers, based on the
demand and production dynamics of the region’s middle income countries. Dynamism on the
demand side is likely to be driven by exceptional growth in the discretionary incomes of the
region’s vast middle income populations, and by related investment expenditures. Dynamism on
the production side should be led by high rates of investment, technological catch-up,
economies of scale and vigorous domestic, regional and international competition.
Any such vast transition involves risks. This study will consider potential shocks to the ACI
growth path associated with: (i) the viability of the new growth engines; (ii) structural issues
associated with the region’s food and energy requirements, environment impact and social
needs; and (iii) changes in the region’s international trade environment. Each of these areas
might be subject to shocks that adversely affect the region’s prospects. But growth also offers
tools for addressing these risks—it improves the possibilities for fund adjustments to challenges
in these and other areas of policy. The results suggest that despite the possibility of large
variations around our baseline projection, some key features of the “great transformation”—
including sizable improvements in the ACI region’s standard of living and in its global impact—
are reasonably robust.
What is certain is that navigating the transition will be difficult. The ACI economies will have to
find ways to grow despite the prospects of slow growth in advanced economies, which are
important destinations for their exports. The shift to internal drivers of growth, in turn, will raise
challenges in macroeconomics, microeconomics, finance, and politics. It will require rapid
technical progress in different industries and it will place new demands on asset markets and
financial systems. Meanwhile, the footprint of the ACI economies is becoming larger in many
dimensions of economic activity—in trade, finance, resources, and the environment—and thus
ACI policies face greater scrutiny and pressures from abroad.
3
ADBI Working Paper 404
Petri and Zhai
The ACI economies face a sea change in their domestic and international economic
environment. Do they risk falling into a “middle income trap”? Will their growth prospects be
dampened by high food prices, energy shocks, tighter environmental regulations, or rising social
spending? Will they fall prey to protectionism or can they improve their prospects through
regional integration? We develop quantitative evidence to assess these threats and potential
solutions to them.
2. METHODOLOGY
We conduct this analysis through the lens of computable general equilibrium analysis. The
approach allows us to examine predictable structural changes as well as shocks that could jolt
economic progress. The foundation for the study is a recently-developed computable general
equilibrium (CGE) model of the world economy which enables us to translate growth
assumptions into detailed sectoral and regional projections for the ACI region and other regions
of the world economy. The model incorporates established features of such modeling efforts, as
well as recent innovations in the economics of international trade based on the theory of
heterogeneous firms. Developed by Zhai (2008), the model has been used in several major
studies of Asian economic integration (Roland-Holst et al. 2005; Kawai and Zhai 2010; Petri,
Plummer and Zhai 2010; and Petri, Plummer, and Zhai 2012).
Simulation strategy
2.1
The model is not used to generate projections, in the sense of likely future outcomes, but rather
to analyze what developments could derail growth and what policies might keep it on track. We
hope to gain insight into sources of risk, their probabilities and importance, and the timing and
cost of the tools required to address them. We begin with projections based largely on the work
of others (including Asia 2050) and attempt to assess what actions, now and in the future, will
permit the great transformation to continue.
Our approach yields a cone of trajectories that describes how the regional and world economies
might evolve over the next two decades. Some paths assume favorable developments (such as
deeper integration of regional and global markets) while others examine tighter constraints
involving greenhouse gas emissions and resource supplies. Still others explore policy strategies
that could mitigate the impact of adverse developments.
Thus, we chart the future much as a mariner might prepare for a long, uncertain journey. The
wise mariner would not plan a single, rigid course, knowing that that much will depend on the
winds and other obstacles encountered. He would consider several possible routes, identifying
the shoals that have to be avoided along each of them. He would prepare his vessel for a range
of conditions and would be ready for course corrections that allow the voyage to proceed
despite inevitable surprises.
Scenarios
2.2
We pursue this strategy by constructing four types of scenarios:
•
Baseline. Building on ADB GDP projections we develop a consistent scenario of regional
and global growth. The resulting sectoral and regional details enable us to identify
structural changes implied by the baseline (they are substantial) and risks associated
with them. Risks arise, for example, from sustaining high productivity growth and
investment over unusually long periods of time, from adjustments required in the ACI
4
ADBI Working Paper 404
•
•
•
Petri and Zhai
economies and in trade partners, and from tightening global resource and environmental
constraints.
Productivity slowdown scenarios. In the wake of the global financial crisis, there is widespread concern about the possibility of a long-term, sustained deceleration of advanced
economies. There is also concern that the ACI economies will fall into the “middle
income trap.” These alternatives are examined with simulations that hypothesize the
deceleration in productivity growth and lower investment rates in various combinations of
economies.
Structural policy scenarios. Adverse developments might also selectively affect specific
sectors of vulnerability to ACI economies. For example, we explore the implications of
higher food prices, higher energy prices and the need to undertake substantially higher
environment and social expenditures.
International trade environment scenarios. Finally, the global environment of ACI
economies could experience worsening protectionism or, on the other hand, more
progressive policy responses, involving both the liberalization of trade and reductions in
current account imbalances.
Taken together, the simulations provide an overview of the trends, risks, and policy responses
that will shape the prospects of the ACI economies in the intermediate future. Of course, they
cannot cover all uncertainties. But scenarios can be framed in terms of general assumptions—
say, about negative productivity and resource shocks—that would be consistent with multiple
causes, including some that cannot be anticipated now.
One implication of the uncertain environment explored in this paper is the value of flexibility and
the ability to adjust to shocks. Important contributions can be made by competitive product,
labor, and capital markets, and by robust international linkages that enable economies to share
risks to demand and supply. Strong financial positions and adequate reserves also help. As the
world economy becomes more complex, resilience becomes a powerful asset.
2.3
Modeling framework
The CGE model tracks the evolution of demand, output, trade, technology, costs and prices in
26 sectors in 11 world regions over the next two decades (see Appendix 1). Sectors and regions
are linked through trade and capital flows and reflect optimizing decisions based on incomes
and prices. Labor endowments and productivity growth assumptions are set exogenously
(based on United National projections), but the model generates investment endogenously and
thus builds future capital stocks.
We solve the model under various assumptions to explore the effects of slowing or accelerating
productivity growth in regions and sectors, to study the impact of changing trade relations, and
to analyze policies affecting resources and pollution emissions. These analyses require, in
some cases, refinements in the structure and/or data of the basic model. For example, in the
simulations addressing energy issues, additional information needs to be introduced on energy
savings and on the technological profiles of low-carbon energy alternatives. Further iterations
may be needed to refine such aspects of the model based on the sectoral components of the
ACI study.
The model is solved year-by-year, allowing for gradual changes in parameters and policies, and
for capturing the cumulative impact of investment decisions. Foreign assets can be accumulated
and resource stocks depleted. As is usual in such general equilibrium studies, the model does
not provide information on business cycles or bouts of unemployment. Once the recovery from
5
ADBI Working Paper 404
Petri and Zhai
the global crisis is complete (as reflected in the IMF’s medium term projections), full
employment levels are maintained.
The structure of the model (Zhai 2008) follows a long tradition of multi-country, applied general
equilibrium models (Van der Mensbrugghe 2005; Shoven and Whalley 1992; Hertel 1997) but
also incorporates recent innovations in heterogeneous-firms trade theory, which account for the
exporting decisions of firms and the resulting intra-industry reallocation of resources. Thus, the
model can capture changes in both the intensive and extensive margin of trade. The approach
generates large impacts from trade integration than do conventional models, and are more
consistent with long-term historical experience.
The model is designed to address issues in economic integration and long-term development.
Its industrial structure tracks how changes in production patterns, production networks, and
market impediments lead to additional varieties of goods and to productivity gains associated
with growth. It also tracks how competitive pressures shift sales from relatively unproductive
firms to relatively productive ones in each regional economy. Thus, the model accounts for
several types of economic gains involved in the development process, including access to
broader varieties of products, economies scale, and changing intra-industry distributions of
productivity. The regions and sectors of the model are listed in Appendix 1 and its full
specification is discussed in Appendix 2.
The model was supplemented with an income distribution module to allocate total consumption
to four consumption classes: extremely poor (consumption below $1.25 per day in 2005 dollars),
low (between $1.25 and $10 per day), middle (between $10 and $100 per day) and high (above
$100 per day). This module takes as its input the country or region per capita consumption level
generated by the CGE model. It then determines the percentage of the population at various
consumption levels. The consumption levels—selected as absolute consumption levels for
defining extremely poor, middle income, and wealthy consumers, respectively—are those used
in ADB (2011), which appear to be based on Kharas (2010).
The distribution module provides plausible projections for the distribution of income along the
baseline of this study and permits the results of scenarios to be translated into effects on the
incidence of poverty and the emergence of the middle class. Although single-point poverty and
middle class estimates are available from other studies (e.g., ADB 2011), they do not always
provide plausible estimates or a way to estimate how distributions might be affected if the
projection trajectory changes. Future adaptations—not implemented at this time—could also
establish connections between the income distribution and growth, as recently proposed in
theoretical models.
3. PROJECTIONS OF ACI GROWTH
The GDP baseline projections were developed by ADB for this study. They show sustained and
relatively rapid growth for ACI economies. The methodology is based on the expected evolution
of the labor force, savings rates, and productivity growth based on historical statistical patterns.
Considerable work has been done by ADB on long-term growth projections (Lee and Hong
2010) and the details of the present methodology are explained in Box 1. The projections are
summarized in terms of the two measures commonly used in international comparisons:
6
ADBI Working Paper 404
Petri and Zhai
constant market prices in Table 1, and purchasing power parity (PPP) in Table 2 1. Scenarios
that challenge some of the assumptions of the baseline are presented later in the chapter.
Box 1: GDP Growth Projections
Economic growth projections are normally generated using production functions that
incorporate assumptions about the evolution of several factors of production and their rates
of productivity. The projections used in this study were developed by ADB for 171 countries,
based on a two-factor (capital and labor) Cobb-Douglas production function and an overall
total factor productivity index (Zhuang 2011). The methodology for projects productivity
gains based on the convergence approach, as discussed for example in Goldman Sachs
(2003). It was implemented with the World Bank’s World Development Indicators dataset.
The base year represents 2010 GDP levels converted into U.S. dollars at market exchange
rates. Some base year data points were adjusted after the completion of these simulations
to achieve consistency with ADBI’s ongoing ASEAN 2030 study. The adjustments involve
small economies and are unlikely to affect significantly the aggregate results reported in this
paper.
The three drivers of growth were modeled as follows:
Labor force projections were based on International Labor Organization population growth
and labor force participation rate forecasts to 2020, extrapolated by ADB to 2030.
The capital stock was modeled by adding investments over time and subtracting
depreciation. Investment was projected forward by multiplying income with an investment
rate and subtracting the projected current account surplus. The investment rate is the critical
element of this calculation and was estimated and forecast using a cross-country regression
model with explanatory variables consisting of income, growth and demographic factors.
Total factor productivity growth in advanced economies was assumed to be 1.3% per year.
To project future values of total factor productivity, emerging economies were divided into
eight convergence classes. The convergence rate for each class was determined from past
data, based on the annual reduction in the gap between the country’s total factor
productivity and that in the United States (taken to represent the frontier). The convergence
factors ranged from 0 to 1.8% per year and were assumed to remain fixed over the 20102030 period. Some baseline growth rates were also adjusted later to reflect judgments from
the ongoing ASEAN 2030 study but are unlikely to affect significantly the results reported in
this paper.
The projection methodology yields relatively rapid rates of growth for ACI economies,
reflecting the continuation of relatively high convergence speeds in recent decades. The
methodology essentially assumes “convergence as usual”—that is, that the process of
technological catch-up will suffer neither setbacks nor find opportunities for acceleration in
the future. By construction, growth rates decline as economies approach the technological
frontier.
Source: Authors’ description based on Zhuang (2011).
1
A third measure that is sometimes used, GDP at market exchange rates, is not reported here. That measure
accounts for the tendency of the real exchange rates of emerging economies to rise, and hence yields more rapid
estimates of their growth over time.
7
ADBI Working Paper 404
Petri and Zhai
Market price and purchasing power parity (PPP) measures provide somewhat different pictures
of the importance of ACI economies in the global economy (for an analytical description, see
Deaton and Heston 2010). The market price measure calculates real output with goods and
services priced at actual 2010 market prices. The PPP measure calculates real output in terms
of goods and services priced at 2010 “international dollar prices,” that is, with every country’s
output valued in terms of common prices. Since the market prices of non-traded goods and
services—the textbook example is haircuts—tend to be low in emerging economies compared
to those in advanced economies, market price estimates tend to yield lower GDP values for
emerging economies than PPP estimates. (By construction, the market price and PPP
measures are the same for the United States.) For the ACI region, for example, the market price
GDP in 2010 is only 55% as high as the PPP GDP. However, the real growth rates of the two
measures tend to be close.
8
ADBI Working Paper 404
Petri and Zhai
Table 1: Population and Output (Market Prices), 2010–2030
Population
ACI
ASEAN
Brunei Dar.
GDP
2010
World
Share
GDP/capita
3,167.0
8,781
14.80
593.4
1,566
0.4
11
Population
GDP
2030
World
Share
2010–30
GDP/capita GDP growth
2,773
3,631.8
33,287
27.72
9,165
6.9
2.64
2,639
706.0
4,634
0.02
27,277
0.5
23
3.86
6,564
5.6
0.02
43,587
3.8
Cambodia
14.1
12
0.02
826
17.4
56
0.05
3,244
8.2
Indonesia
239.9
571
0.96
2,380
279.7
1,700
1.42
6,080
5.6
Lao PDR
6.2
6
0.01
989
7.8
28
0.02
3,575
7.8
Malaysia
28.4
204
0.34
7,166
37.3
593
0.49
15,912
5.5
Myanmar
48.0
21
0.03
431
54.3
117
0.10
2,149
9.0
Philippines
93.3
170
0.29
1,828
126.3
660
0.55
5,224
7.0
Thailand
69.1
277
0.47
4,008
73.3
711
0.59
9,702
4.8
5.1
193
0.32
37,849
6.0
324
0.27
54,206
2.6
Singapore
87.8
102
0.17
1,160
101.5
420
0.35
4,136
7.3
PRC
Viet Nam
1,348.9
5,711
9.63
4,233
1,402.3
22,083
18.39
15,748
7.0
India
1,224.6
1,504
2.54
1,229
1,523.5
6,570
5.47
4,312
7.6
71.2
1,266
2.13
17,785
73.3
2,459
2.05
33,539
3.4
Japan
126.5
5,118
8.63
40,444
120.2
6,341
5.28
52,749
1.1
US
310.4
14,431
24.32
46,494
361.7
22,772
18.97
62,961
2.3
Europe
511.3
17,635
29.72
34,491
527.8
25,997
21.65
49,252
2.0
6,641.1
59,327
100.00
8,933
7,932.8
120,065
100.00
15,135
3.6
Korea; Taipei,China
World
Brunei Dar. = Brunei Darussalam; Korea = The Republic of Korea.
Population: millions; GDP: billions of 2009 U.S. dollars; GDP/capita: 2009 U.S. dollars.
Note: ADB projections as of 16 July 2011. The ASEAN aggregate includes Timor Leste, which is not an official member at this time. The projections were subsequently
revised slightly.
Source: ADB.
9
ADBI Working Paper 404
Petri and Zhai
Table 2: Population and Output (Purchasing Power Parity), 2010–2030
2010
Population
ACI
ASEAN
2030
World
Share
GDP
3,167.0
15,905
23.98
GDP/capita
Population
2010–30
World
Share
GDP
GDP/capita GDP growth
5,022
3,631.8
61,392
39.28
16,904
7.0
593.4
2,755
4.15
4,643
706.0
8,475
5.42
12,004
5.8
Brunei Dar.
0.4
18
0.03
46,362
0.5
39
0.02
74,082
3.8
Cambodia
14.1
28
0.04
1,968
17.4
134
0.09
7,729
8.2
Indonesia
239.9
926
1.40
3,862
279.7
2,760
1.77
9,868
5.6
Lao PDR
6.2
14
0.02
2,292
7.8
64
0.04
8,286
7.8
Malaysia
28.4
368
0.56
12,971
37.3
1,073
0.69
28,801
5.5
Myanmar
48.0
71
0.11
1,485
54.3
403
0.26
7,409
9.1
Philippines
93.3
313
0.47
3,354
126.3
1,211
0.77
9,587
7.0
Thailand
69.1
517
0.78
7,476
73.3
1,327
0.85
18,096
4.8
5.1
242
0.37
47,629
6.0
408
0.26
68,212
2.6
Singapore
87.8
256
0.39
2,913
101.5
1,054
0.67
10,388
7.3
PRC
Viet Nam
1,348.9
9,373
14.13
6,948
1,402.3
36,421
23.30
25,972
7.0
India
1,224.6
3,778
5.70
3,085
1,523.5
16,497
10.55
10,828
7.7
Korea; Taipei,China
71.2
2,000
3.02
28,093
73.3
3,893
2.49
53,097
3.4
Japan
126.5
3,824
5.77
30,220
120.2
4,738
3.03
39,414
1.1
US
310.4
13,104
19.76
42,220
361.7
20,678
13.23
57,172
2.3
Europe
511.3
14,452
21.79
28,266
527.8
21,682
13.87
41,077
2.1
6,641.1
66,329
100.00
9,988
7,932.8
156,308
100.00
19,704
4.4
World
Brunei Dar. = Brunei Darussalam; Korea = The Republic of Korea.
Population: millions; GDP: billions of 2009 international dollars; GDP/capita: 2009 international dollars.
Notes: ADB projections as of 16 July 2011. The ASEAN aggregate includes Timor Leste, which is not an official member at this time. The projections were subsequently
revised slightly.
Source: ADB.
10
ADBI Working Paper 404
Petri and Zhai
By 2030, the output of ACI countries will have grown roughly four-fold by either measure. In
market prices, ACI output will be larger than that of the United States or the European Union. In
PPP terms it will be larger than the two combined. These long term average growth rates
conceal gradual changes in the pace of growth over time (Figure 1). Interestingly, world
economic growth is projected to accelerate in coming decades, notwithstanding the aftereffects
of the global financial crisis, aging populations, and diminished prospects in advanced
economies. The reason is not faster growth in individual countries, but the great transformation.
More and more, world averages reflect rapidly growing emerging economies, including
especially the ACI countries, whose weight in global aggregates is already large and continues
to increase.
Figure 1: ACI Growth Rates will Slow but Stay High
Source: ADB projections.
The PRC’s growth rate is projected to be the highest initially among major countries and groups.
But the PRC is projected to decelerate—even after the effects of the current slowdown
dissipate—due to the slower growth and eventual decline of the PRC’s labor force, as well as
moderating productivity growth. Deceleration is inevitable given the PRC’s successful growth
trajectory and does not imply that the PRC will sink into a “middle income trap.” That term is
reserved for a more drastic, and hopefully avoidable, productivity slowdown that will be
examined as a special scenario.
India’s growth rate is projected to rise above past averages as the country emerges from the
current slowdown and embarks on full-speed catch-up. The acceleration that the projections
anticipate is based in part on solid labor force growth and progress in transferring workers from
agriculture to industry and services. In this scenario, India’s growth rate would overtake that of
the PRC sometime between 2015 and 2020. The NIEs are projected to decelerate as their
11
ADBI Working Paper 404
Petri and Zhai
income levels approach those of advanced economies, and the advanced economies
themselves are projected to grow at rates similar to long-term historical averages.
ASEAN’s growth is projected to remain relatively fast: low- and middle-income countries are
projected to grow in the 5–9% range. Growth rates are projected to be relatively high for the
least-developed countries (Cambodia, Laos, Myanmar and Vietnam), averaging above 7%.
Middle-income countries (Indonesia, Malaysia, the Philippines and Thailand) would expand with
growth rates in the 5–7% range. Only high-income Brunei Darussalam and Singapore are
projected to grow more slowly, at rates under 4%.
Investment and productivity improvements are likely to be the principal drivers of growth.
Figure2 shows that productivity growth will be the largest contributor to growth in India, while
capital will be more important in other ACI countries. The role of investment relative to
productivity growth will increase in all countries over time. The contribution of labor force growth
will be the least important, and will decline over time, actually tuning negative in the PRC toward
the end of the projection period.
Figure 2: Growth will be Driven by Investment and Productivity
Source: Authors’ calculations.
The prospects for sustaining growth consistent with historical experience are reasonably good,
but this path nevertheless represents an ambitious scenario. Such growth rates require
continued improvements in productivity and high rates of investment. Meanwhile, the
demographic dividend—the economic benefits of having a large working-age population relative
to young and aged populations—which played a role in some countries in the past, will turn
negative in many countries by 2030. But success also generates momentum; the ACI region is
seen, and is likely to remain, a propitious location for economic activities and investments that
benefit from scale and proximity to investment and consumption growth.
12
ADBI Working Paper 404
Petri and Zhai
4. ANALYSIS OF THE BASELINE: DIMENSIONS OF
TRANSFORMATION
Structural detail is useful for understanding the varied consequences of growth, ranging from
the demand for resources to international relationships through trade and finance. The model is
used to explore four major themes:
•
•
how economic growth will transform the lives of the ACI citizens,
•
how ACI growth will affect global resource and environmental constraints, and
•
how ACI economies will transition to new drivers of growth,
how the region’s role will change in the world economy.
The next sections probe these themes by considering alternative scenarios for ACI growth. the
global economic environment, and policies adopted in the region and beyond.
4.1
A leap in the quality of life
The baseline runs of the simulation model are calibrated to the ADB growth projections and help
to work out their detailed implications. These show, most importantly, that the region is poised
for an historic improvement in the quality of life. The ACI economies are converging toward the
world technology frontier and their incomes are rising accordingly (Figure 3). A detailed look at
country-level projections suggests that low- and middle-income ACI countries will grow faster
than the world average and outperform peers elsewhere at corresponding income levels.
13
ADBI Working Paper 404
Petri and Zhai
Figure 3: The Income Gap between ACI and Advanced Countries will Continue to Close
Note: GDP per capita at market exchange rates, 2010 US dollars.
Source: ADB projections.
Such sustained income gains should reach most segments of the ACI population. A sub-model
of expenditure distributions was used to translate average gains into estimates for various
points of subgroups of the income distribution. Most importantly, the model shows that the
incidence of poverty is poised to decline precipitously in the ACI region (Figure 4). By 2030 the
percentage of the ACI population below the $1.25 per day poverty line will have fallen from
more than 600 million across the region today to a little more than 30 million, accounting for only
one% of the region’s population. This also means that with modest-scaled targeted policies—
explored in more detail below—the region could fully eradicate extreme poverty by that time.
14
ADBI Working Paper 404
Petri and Zhai
Figure 4: Extreme Poverty can be Eradicated
Note: Percentage of population with consumption below $1.25/day (2005 US$).
Source: authors’ calculations.
Progress will be also dramatic higher up in the income distribution. By 2030, 64% of the
population of ASEAN, 79% of the population of the PRC, and 69% of the population of India
should move into the middle income brackets, defined as consumption expenditures between
$10/day to $100/day (Figure 5). As highlighted in Chapter 1, two billion ACI citizens will join the
ranks of middle income consumers over the next 20 years. These extraordinary developments
will double the size of the global middle class and by 2030, when about half of the world’s
middle class population will live in ACI. The ranks of the region’s truly affluent consumers—
those with consumption in excess of $100/day per person—will also swell; by 2030, more than
100 million people in ACI countries could enjoy such prosperity.
15
ADBI Working Paper 404
Petri and Zhai
Figure 5: Middle Income Groups will Become Dominant
Source: authors’ calculations.
However, large changes in private incomes and expenditures tell a partial story. They do not
accurately measure improvements in the quality of life, which also depend on goods and
services not provided by markets. That dimension is explored in Chapter 3. Critical public
services include education, health care, sanitation and transportation; measures to ensure
public safety in its many dimensions; and the protection of the environment. The simulations
cannot provide details on specific policies for managing these objectives, but they do show that
sufficient resources will be available to fund them. Governments will have ample incentives to
provide effective services: The quality of life is a compelling objective in its own right, and
making progress on it is also essential for sustaining support for leaders.
Urbanization is a systematic byproduct of rapid growth, and ACI cities will be among the
principal engines driving the transformation. According to United Nations (2012) estimates,
between 2010 and 2030 the ranks of urban dwellers in ACI countries will expand by 658 million
people and by 2030 a majority of the region’s population will be living in cities (Table 3). The
region has lagged world urbanization levels in the past, but the trends are now poised to push
ACI cities into the forefront of world development in the next two decades. Despite the region’s
relatively slow population growth, the United Nations estimates that its share of the world’s
urban population will rise from 37% to 39%.
As so many other dimensions of the transformation, urbanization has potentially positive and
negative implications. Cities are extremely productive: Incomes are typically high compared to
other places; scale advantages facilitate many kinds of production and innovation; cities are
good places to study and work and therefore attract talent and entrepreneurship; and dense
populations make it possible to offer private and public services in greater variety at lower cost.
But cities are also demanding and tough to manage: without good infrastructure and public
services, cities can become harsh places to live and even seedbeds of disease, dissatisfaction,
and violence.
16
ADBI Working Paper 404
Petri and Zhai
Table 3: Most ACI Citizens will Live in Cities
ACI
ASEAN
Brunei Dar.
Cambodia
Indonesia
Lao PDR
Malaysia
Myanmar
Philippines
Thailand
Singapore
Viet Nam
PRC
India
Korea; Taipei,China
Japan
US
Europe
World
Population
3,166
592
0
14
240
6
28
48
93
69
5
88
1,349
1,225
71
127
310
511
6,641
2010
Urban
1,308
262
0
3
120
2
20
15
45
23
5
27
668
379
57
115
255
391
3,559
%
41
44
76
20
50
33
72
32
49
34
100
30
50
31
80
91
82
76
54
Population
3,630
704
1
17
280
8
37
54
126
73
6
101
1,402
1,523
73
120
362
528
7,933
2030
Urban
1,966
393
0
5
176
4
30
24
71
32
6
44
967
606
62
116
311
429
4,984
%
54
56
82
26
63
52
81
44
56
44
100
43
69
40
84
97
86
81
63
Brunei Dar. = Brunei Darussalam;
Korea = The Republic of Korea.
Source: United Nations (2012).
In short, the ACI region is on the cusp of a giant advance in living standards—potentially freeing
hundreds of millions of people from poverty and building a huge middle class. The region’s
economic gains should generate solid increases in private consumption and living conditions.
They will also enable governments to fund major improvements in public services. In turn, the
region’s wide-ranging requirements and expenditures will create massive demand and
opportunities for investment and entrepreneurship. These trends should produce—given a
conducive economic environment—solid, indigenous foundations for growth.
4.2
New drivers of growth
For ACI growth rates to remain above world rates, the region’s output structure will have to shift
away from exports to advanced economies toward regional and domestic demand. An indirect
implication is that productivity gains can no longer depend on export-oriented production, which
is often associated with easy access to foreign technologies, partners and investors. Regional
innovation systems and markets will have to become the new drivers of ACI growth, providing
technology to expand output and markets to sell it. These shifts will be reinforced as the current
account surpluses of ACI economies continue to decline relative to GDP, due to the effects on
the financial crisis on expected returns abroad.
New drivers of growth will be needed on both the demand and supply sides of the ACI growth
equation. They are likely to emerge. The principal engines on the demand side will be rising
wages and policy shifts that direct a larger share of income toward private and public
17
ADBI Working Paper 404
Petri and Zhai
consumption expenditures. These effects will of course differ across the region. India’s current
account, for example, has been in balance or in deficit, and cannot be expected to decline. But
the current account surpluses have been high, and investment relatively low in several ASEAN
countries (relative to levels in the 1990s as well as international experience), so investment
represents a logical target for demand growth. Consumption increases are likely to be especially
important in the PRC, where the share of consumption expenditures is unusually low. The
components associated demand in ACI economies are illustrated in Figure 6. The net effect:
ACI producers will have ample opportunity to accelerate domestic and regional sales, including
especially of products that target the regional demand patterns.
Figure 6: Demand will Shift toward Consumption
Source: authors’ calculations.
Research by McKinsey & Co. (2011) has examined the opportunities in discretionary spending
in large ACI countries such as India and the PRC, and has singled out lucrative markets
associated with more prosperous urban lifestyles. Consider, for example, the implications of an
estimated 30 million people in ACI countries that will be added to urban populations each year.
This annual increase equals the population of the world’s largest city today and will generate
requirements for an additional 75 million square meters of housing and corresponding
opportunities in residential construction and in equipping and servicing these new residences
and their occupants. These sectors could alone add $100 billion to regional expenditures every
year.
Baseline simulations show similar results—more systematically generated by a multi-regional
and multi-sectoral general equilibrium model—suggest similar, large transformations. Services
will be an especially important component of rising demand; service inputs play a larger rule in
production in higher income economies, and consumer budgets have a large service
component relative to other demand components, such as exports and investment (Figure 7).
On a more micro-economic level, a range of new products will be required by middle-class
households with incomes still well below those in high-income countries. These “frugal
18
ADBI Working Paper 404
Petri and Zhai
innovations” could serve markets ranging from food products and clothing to motor scooters and
automobiles. ACI producers have a pole position in the race for these markets.
Figure 7: Production will Shift toward Services
Source: Authors’ calculations.
The simulations suggest sharp increases in the share of ACI economies in global expenditures
on consumption and investment across a wide range of sectors. These results (summarized in
Figure 8 for consumption and investment) will reinforce the within-economy demand shifts
described above. The markets that drive demand within ACI economies will become more
important also internationally. Located at the epicenter of world growth in several key sectors,
ACI producers have excellent opportunities to identify the goods and services that are best
suited for the region’s—and world’s—largest new markets.
19
ADBI Working Paper 404
Petri and Zhai
Figure 8: ACI’s Rising Shares of Global Expenditures
A. Shares of world investment
B. Shares of growth in world investment
D. Shares of growth in world consumption
C. Shares of world consumption
Source: Authors’ calculations.
Consider, for example, the interesting implications for investment and the sectors that provide
capital goods and services for it. ACI shares in world investment will rise from 24% in 2010 to
38% two decades later (Figure 8A). But what is still more remarkable is that the region’s share
of the growth of world investment—the new investment markets that will emerge in that period.
Figure 8B shows that ACI’s share in the growth of investment will rise from 40% in 2010 to 51%
in 2030. Further, if ACI results are combined with those of other emerging market economies,
which are also expected to grow at above average rates, the “South” will account for fully 85%
of the growth of world investment by that 2030. A vast majority of the new markets for
investment goods—for construction equipment, high-speed rail, energy plants, and so on—will
be created in emerging markets and especially in ACI. This will be true also for related services,
ranging from engineering and architecture to finance.
Especially important among investment sectors are those linked to infrastructure, such as roads,
ports, urban transport, communications, power generation, and a wide range of utilities and
amenities. The infrastructure needs in ACI economies range into trillions of US dollars. The role
of infrastructure in the transformation is important enough to warrant detailed analysis in
20
ADBI Working Paper 404
Petri and Zhai
Chapter 4. Infrastructure affects the prospects of the region not only through demand, but also
through contributions to productivity and innovation. In some countries—notably India,
Indonesia, and Philippines—infrastructure requirements are especially urgent since they already
appear to constrain growth.
As market opportunities and demand pull firms and resources into new sectors, so rising costs
and wages and international competition will help to push them. The trade effects of structural
change are shown in Table 4. The simulations suggest substantial migration of labor-intensive
processes from high- to low-wage economies, including from the PRC to India and Southeast
Asia. In turn, the PRC will move up the technology ladder into industries and segments now
served primarily by Japan, the Republic of Korea, and Western countries. Japan and the
Republic of Korea, in turn, will move into more advanced manufacturing and especially service
activities. But all of these activities are likely to be even more interconnected than now, with
production processes split into many steps and services. Asian production networks—as
examined in detail in Chapter 4—will facilitate these shifts and establish the linkages needed to
achieve minimum-cost production. These networks are powerful “clearing houses” for
organizing the deployment of technology and capital, designed to yield efficient systems and
secure niches in the global economy.
Table 4: Sectoral Shifts in ACI Output
Agriculture
Rice
Other grains
Crops
Livestock
Forestry, fisheries
Mining
Coal
Oil
Gas
Minerals
Manufacturing
Food, beverages
Textiles, apparel
Petroleum products
Chemicals
Metals
Machinery
Electrical equipment
Transport equipment
Other manufactures
Services
Electricity
Gas
Construction
Trade
Transport, communications
Financial services
Other services
Government services
GDP
ASEAN
8.9
1.9
0.3
3.2
1.1
2.4
4.3
0.4
1.9
0.9
1.1
33.2
3.9
3.0
0.4
4.9
3.0
4.8
7.2
2.4
3.7
53.6
1.2
0.4
6.1
13.1
6.8
4.3
12.5
9.3
100.0
2010
PRC
8.5
0.7
0.3
3.6
2.4
1.5
3.3
0.8
0.5
0.0
1.9
37.2
3.0
3.0
0.3
5.0
9.9
6.3
2.3
2.4
5.0
51.1
1.0
0.0
6.7
6.0
7.9
4.9
9.0
15.7
100.0
Source: Authors’ calculations.
21
India
20.8
3.2
2.2
11.4
2.1
2.0
2.8
0.3
0.6
0.3
1.6
20.4
5.6
2.3
0.1
2.2
3.6
2.3
0.3
1.4
2.5
56.0
1.8
0.0
9.2
12.7
8.2
5.0
8.6
10.4
100.0
ASEAN
5.5
1.0
0.1
2.1
0.7
1.7
2.9
0.4
0.9
0.7
0.9
36.8
2.6
3.0
0.3
5.4
2.8
7.5
9.0
2.2
3.9
54.8
1.2
0.3
9.6
9.4
6.3
3.1
12.9
12.0
100.0
2030
PRC
5.5
0.3
0.2
1.5
2.7
0.8
2.3
0.5
0.2
0.0
1.6
29.8
2.8
2.2
0.3
3.1
6.8
5.9
1.8
2.3
4.5
62.4
1.0
0.0
7.9
6.4
6.1
4.7
9.8
26.5
100.0
India
17.1
1.5
0.9
10.2
2.2
2.2
1.5
0.2
0.2
0.3
0.7
20.2
5.7
2.7
0.2
1.9
3.0
2.5
0.1
1.1
3.0
61.2
2.0
0.0
9.2
14.6
7.6
4.5
10.1
13.1
100.0
ADBI Working Paper 404
Petri and Zhai
The changes should help to sustain increases in productivity, derived increasingly from
productivity gains within sectors rather than shifts of resources among them. Competition and
wage increases provide incentives for such gains, and the national policy environment can
reinforce them by stimulating innovation and entrepreneurship. Key elements include a labor
force educated to levels consistent with intermediate-term skill requirements, the protection of
intellectual property, and financial markets that provide access to resources and reward risktaking. ACI’s urban centers will be a fundamental asset in the transformation; they attract talent,
encourage the acquisition of skills, facilitate competition and the exchange of ideas, and serve
as incubators for new products and business strategies.
4.3
Resource and environmental challenges
The ACI economies are so large that their growth will inevitably affect the resource sectors of
the world economy (Figure 9). The simulations suggest that world food and energy demand,
and thus prices, are likely to rise (Figure 10) in part due to ACI’s scale and growth. The ACI
economies are intense resource users because their economic structures are still weighted
toward basic materials and heavy industry, because their consumers still spend a relatively
large share of their income on food and fuel, and because they are (of course with notable
exceptions) densely populated and resource poor. Moreover, ACI consumers appear to be
following developed-country precedents in consumer demand, including motor vehicle
purchases. As ACI economies increasingly depend on international trade for food and energy,
their self-sufficiency rates will likely decline. International linkages will be important for ACI’s
continued development and their firms and governments can be expected to remain active in
global resource investments.
Figure 9: ACI’s Rising Share of Commodities Demand
Source: Authors’ calculations.
22
ADBI Working Paper 404
Petri and Zhai
Figure 10: Rising Commodity Prices
Source: Authors’ calculations
Rising primary goods prices would signal the end of an era—for many decades, energy and
food prices have been on a long declining trend. But in recent years they have begun to rise,
even spiking to much higher levels for shorter periods. The demand for primary materials is
bound to be robust as long as the world’s rising economies remain intensive users of resources.
Nevertheless, primary materials play a more limited role in economic growth now than in the
past, and potentially important supply and demand responses—including new technologies for
extracting natural gas and renewable energy—are also beginning to take shape. On the whole,
these trends are likely to limit the effects of tight primary goods markets on growth.
The effects of ACI growth on the environment are significant and potentially more damaging.
Because ACI energy supplies continue depend heavily on coal, the region has become an
unusually important source of greenhouse gases. ACI’s CO 2 emissions accounted for one-third
of global emissions in 2010, roughly equal to those of advanced economies. Over the next two
decades, its emissions are projected to increase 2.5 times and would account for an increasing
share of global emissions (Figure 11). These large changes would occur despite significant
advances in reducing the carbon intensity of ACI output.
23
ADBI Working Paper 404
Petri and Zhai
Figure 11: ACI’s Rising Share of World Carbon Emissions
Source: Authors’ calculations.
The implications of these trends are explored in scenarios below, but clearly greenhouse gas
emissions represent a fundamental risk to ACI growth. Business as usual—even with steady
improvements in the energy intensity of production—will be no longer viable if perceptions of the
global threat increase and/or the mitigation of emissions becomes a high priority for either the
region’s own citizens or other segments of the global community. Given ACI’s political and
economic interdependence with the global economy, conflict over the environment is an
important source of risk—arguably greater than the risks and costs involved in implementing an
early, efficient strategy to reduce the region’s carbon footprint.
4.4
Expanding global role
As the share of ACI economies in world GDP roughly doubles over the next twenty years, the
region’s global footprint will expand proportionately across a wide range of variables, including
consumption, investment, and resource requirements (as demonstrated in Figure 8). In turn,
these developments will reshape the region’s role in trade, finance, and the management of the
global economy. Due to the relatively slow growth of the region’s external markets, trade will
increase, albeit more modestly relative to ACI GDP than in the past. Since the composition of
the region’s economy will shift toward the PRC and India, which have relatively low trade/output
ratios due to their size, the overall ACI trade/GDP ratio will stay roughly constant at 2/3.
The structure of this trade, however, will be quite different. The bulk of the increase in world
trade over the next two decades will consist of South-South trade—trade among ACI partners
and other emerging economies. The share of these flows will expand from 20% to 36% of world
trade between 2010 and 2030 (Figure 12). Meanwhile, the share of North-North trade will
decline from 42% to 21% of world trade, while the share of North-South trade will rise slowly
24
ADBI Working Paper 404
Petri and Zhai
from 38% to 43% of the total. This also implies that South -North connections will remain critical
also for at least the next two decades.
Figure 12: World Trade will Turn South
Source: Authors’ calculations.
The rise of the ACI economies implies perhaps even stronger global connections through
financial flows. ACI have substantially higher savings rates than the rest of the world, so the
growing share of ACI economies in world GDP will mean an expansion in share of world income
devoted to savings and investment—increasing the world savings rate from 22 to 25%. These
high world savings are needed, of course, to drive ACI growth itself, but they will inevitably
affect global capital markets and transform the ACI’s economic role from “factory floor” to
“banker.” And as ACI countries become increasingly important issuers and owners of financial
assets, their regional institutions are also bound to play a larger role in intermediating financial
flows. Chapter 6 will assess the likely speed and implications of this shift, but among other
things, it should lead to the development of stronger Asian financial centers. These should thrive
not only because of the sheer size of their potential market, but also because institutions closer
to home often have an advantage in assessing regional information and serving the needs of
regional clients.
5. THREATS TO THE TRANSFORMATION
Steady, uninterrupted progress—as projected in the baseline—cannot be taken for granted and
is in fact unlikely. To achieve even the baseline results, the regional and international
environment will have to support the projected development trajectory in many ways, as it
generally has in the last several decades. The road to success is most likely narrow, but there
are many paths that could lead to less favorable outcomes. This section examines various
threats and their implications for ACI development and policy.
25
ADBI Working Paper 404
Petri and Zhai
Risks that threaten the transformation process could originate in ACI countries or in the global
environment; they could involve economic constraints or political conflict; and they could affect
variables ranging from market access to productivity change. It is impossible to address all such
risks, but selected “stress tests” involving key elements of the projections can provide insight
into the sensitivity of the results. The issues explored include:
•
•
A major slowdown in productivity growth in ACI and/or other countries
•
Rising energy costs
•
Rising distributional concerns
•
Rising food prices and food security concerns
•
Rising environmental concerns
Protectionism in ACI’s principal export markets
The following pages examine how these developments would affect ACI growth. The
assumptions of the scenarios are presented in Table 5. Aside from the baseline, they fall into
three groups. A first set (scenarios 2–4) explores alternative types of productivity deceleration. A
second set (scenarios 5–8) addresses structural issues affecting key sectors or types of
expenditure. The last set (scenarios 9–12) examines changes in the ACI region’s external trade
and capital flows. Given the uncertain outlook, more scenarios deal with adverse developments
(this is the case with scenarios 2–9) than with favorable ones (scenarios 10–12).
26
ADBI Working Paper 404
Petri and Zhai
Table 5: Alternative Scenarios
No.
Description
•
•
•
•
•
•
1
Baseline
2
ACI productivity shock
3
Advanced economies
productivity shock
•
•
4
5
Worldwide productivity
shock
Agriculture shock
•
•
•
6
Energy shock
•
7
Green growth policy
•
•
8
Social inclusion policy
•
•
9
10
Rising protection
Regional trade agreement
11
Global trade agreement
12
Fast rebalancing
•
•
•
•
•
•
Parameter changes
ILO population and labor force projections
ADB projections of investment rates, productivity growth
Sectoral productivity growth expectations based on history
Capital flows fixed at 2010 nominal level
TFP growth rate reduced by 25% in ACI countries
Investment rate reduced by 10% in ACI countries and 2% in
advanced countries
TFP growth rate reduced by 25% in advanced countries
Investment rate reduced by 10% in advanced countries and
2% in ACI countries
TFP growth rate reduced by 25% worldwide
Investment rate reduced by 10% worldwide
Agricultural productivity reduced worldwide to generate
annual agricultural price increases of 2% above baseline
Energy productivity reduced worldwide so as to generate
annual energy price increases of 2% above baseline
Carbon taxes set to $90/ton CO 2 in advanced countries and
$60/ton CO 2 in emerging economies
Government expenditures increased as estimated in the
International Energy Agency’s “450” scenario, financed by
reduced private consumption and investment
ACI government expenditures increased by 2% of GDP,
financed by reduced private consumption and investment
Expenditures result in 5%age point reduction in the Gini
coefficient
MFN tariffs and NTBs doubled in advanced countries
Intra-regional tariffs reduced 75% in ACI region
Intra-regional NTBs reduced 50% in ACI region
MFN tariffs reduced 75% worldwide
MFN NTBs reduced 50% worldwide
All net capital flows reduced linearly to zero by 2020
Notes: All parameter changes assumed to be phased in 5 equal steps 2010–2020 period.
Source: Authors’ assumptions.
Each of the adverse scenarios would reduce the income gains projected on the baseline. A
summary of the results is presented in Figure 13 for incomes in the ACI group under the
scenarios of Table 5. The most serious setback would result from a prolonged productivity
slowdown in the ACI economies, in the region alone or in combination with such slowdowns
worldwide. This “middle income trap” scenario would lower ACI incomes by nearly 25% in 2030.
Other shocks—to agriculture, energy, or environmental or social spending requirements—would
have smaller negative effects, in the order of 5% of incomes in 2030. In turn, improvements in
the regional and global trading system would raise incomes by 5–10%. Taken in combinations,
these results can be also used to develop a rough probability distribution of the possible range
of outcomes (presented later in Figures 14 and 15). The threats warrant attention individually
and in combination—much is at stake in preventing them or mitigating their effects.
27
ADBI Working Paper 404
Petri and Zhai
Figure 13: Effects of Shocks on ACI Incomes in 2030
Source: Authors’ calculations.
The analysis shows not only the effects of possible shocks, but also that the main elements of
the transformation of ACI economies would persist under varied shocks. The convergence
results of the baseline are reasonably robust. Even under unfavorable assumptions, per capita
incomes in the ACI economies would grow at an annual rate that is 2% faster than the rest of
the world economy, and would become more important in world output, trade, and finance.
Major structural changes within the ACI economies—in living standards and drivers of
production—would also likely persist. To be sure, the detailed results show that the harsher
outcomes would lead to much hardship and disappointment. The incentives to prevent or
mitigate such outcomes are great. But the relative rise of the ACI economies, with some
allowance for variations in its exact speed and scale, is likely to be a fixture of coming
decades—it is not a product of optimistic assumptions alone.
5.1
The threat of the middle income trap
The risk of a substantial slowdown in the ACI economies—the middle-income trap—is now
widely recognized, if poorly defined (see Box 2). The possibility that growth rates slow once
countries reach middle income levels has been widely discussed in the recent empirical
literature with the possibilities for deceleration attributed to a wide range of causes (see for
example Rodrik 1999, Ros 2005). The baseline projections of this study of course incorporate
deceleration in several of the region’s fastest-growing economies as their productivity gaps
relative to the global frontier narrow. But to be useful, the concept of the middle income trap has
to mean more than that. As argued in Chapter 1, a useful version of the concept refers to
impediment to changes in the drivers of growth, which imply that growth has to slow until those
impediments are rectified (which may take a long time).
28
ADBI Working Paper 404
Petri and Zhai
The detailed risks that lead to a middle income trap will vary across countries. In some countries
the trigger for deceleration might be the end of resource transfers from agriculture to industry,
coupled with the absence of capacity to generate within-sector productivity growth in industry. In
other countries the problem may be the lack of domestic or regional demand, once the economy
can no longer rely on low-wage exports to drive demand. In still others, it may mean that human
capital, or legal and economic institutions to permit industrial development, are insufficient to
enable an economy’s growth strategy to shift to more sophisticated industry. Later chapters
examine these risks in more detail; this discussion focuses on their overall implications.
Scenarios 2–4 examine the middle income trap effect by varying two key drivers of growth:
productivity and investment. These factors are often affected in parallel; a productivity shock, for
example, may both cause and be aggravated by lower returns on investment and diminished
entrepreneurial activity. The simulations assume that productivity growth declines by 25% and
investment by 10% in the economies assumed to enter the middle income trap. Three versions
of this deceleration are examined: deceleration in ACI economies (Scenario 2), deceleration in
advanced economies (Scenario 3), and deceleration worldwide (Scenario 4). Since such shocks
also spill over into the global economy—though the channels of global investment returns and
investor confidence—the scenarios further assume a 2% reduction in investment rates even
economies not directly affected by the assumed shock. Results are reported in Table 6.
Box 2: Understanding and Avoiding the Middle Income Trap
Policy makers in many rapidly developing countries are concerned that their economies might fall
into what Gill and Kharas (2007) and others have described as the “middle income trap.” Although
the term is now widely used, it is not well defined, and appears to be more an empirical possibility
rather than a theoretical necessity. The anecdotal examples of the middle income trap that are
often mentioned include Latin American economies such as Argentina and Brazil, and Middle
Eastern economies such as Egypt and Iran. Yet many East Asian economies, including notably
Japan, the Republic of Korea, and Singapore, have moved beyond middle income levels without
significant deceleration.
The middle income trap is sometimes described as the inability to from middle income to high
income levels. But this definition includes countries that reached middle income status slowly,
as well as those that are continuing to advance, albeit slowly. A more useful definition will
isolate cases that have more distinctive trajectories and more in common with the prospects
of Asia’s rapidly developing economies. One such definition involves substantial deceleration
from fast growth once middle income levels are reached. In the most extensive empirical
study so far, Eichengreen et al. (2011) found that the probability of growth deceleration (of at
least 2%age points from an sustained prior per capita income growth rate of at least 3.5%)
rises at a per capita income levels of around $10,000 and peaks at around $17,000 (see
Figure B1-1). Since several ACI economies will be in this range in the next two decades, it is
useful to ask what causes such deceleration and how they might prevent it.
29
ADBI Working Paper 404
Petri and Zhai
Figure B1.1: Probability of Growth Deceleration by Income Level
Source: Eichengreen et al. (2011).
Even if the empirical regularities of deceleration are well established, a theoretical structure is
needed to make the concept of the middle income trap useful for policy planning. The term
evokes earlier theoretical models of the “low income trap” identified in the development
literature (Nelson 1956, Leibenstein 1954). Those models were based on a clear theoretical
insight: at low income levels countries have to use their income to meet survival requirements
and cannot accumulate capital to raise productivity. Hence countries with very low incomes
are stuck at those levels until foreign aid, a harsh government investment or birth control
program, or a fortuitous technological breakthrough creates surplus to permit capital
accumulation to begin.
There appears to be no such common, model-based insight associated with the middle
income trap in the literature. Rather, discussions attribute slowdowns to multiple causes,
typically including a wide range of factors. As an economy’s development moves past the
middle-income level, the factors driving growth have to change. If a country has the newly
required factors in place, it continues to grow rapidly, but if it does not, growth decelerates.
But there is an important common thread that emerges from this perspective—one that could
lead to a more rigorous, operational definition of the middle income trap. The factors and
institutions required for growth have to be updated from time to time, and since it may take a
long time to develop the required factors, the process of updating them has to be launched
with a lead time, much as a manual automobile may need to shift gears before attempting to
climb a hill. Yet policy or market failures may prevent action in advance. By the time a country
experiences deceleration it is too late to fill the gaps without losing momentum and possibly
decades of progress. A smooth transition requires that countries anticipate shifts in growth
models and take early steps—that is, correct policy and market failures—to enable growth to
continue beyond middle income levels in the future. In this view, long institutional lead times
are the crucial problem that policy makers have to address to avoid future deceleration.
The transition from low-income to post-middle-income growth has been identified in various
ways. Some writers identify this point in terms of wages: once wages are high enough to
make labor-intensive production no longer competitive internationally, continued rapid growth
30
ADBI Working Paper 404
Petri and Zhai
requires the production of capital- or technology-intensive products. Others describe it as the
end of “easy” productivity gains achieved by transferring labor from agriculture to industry,
reached at the so-called “Lewis turning point,” when the supply of agricultural surplus labor is
exhausted. At that point, productivity gains have to be generated from productivity
improvements within sectors rather than inter-sectoral resource transfers.
The pre-requisites of post-middle-income growth have been also described in multiple ways.
High on the list are factors that support total factor productivity growth, including human
capital, and institutions such as efficient financial markets, and robust systems of
entrepreneurship and innovation. Also important may be income distribution and demand
policies that generate domestic demand once foreign demand for labor-intensive exports
dries up. Still other writers focus on deeper institutions (property rights, financial markets,
competition) that provide incentives and opportunities for transferring resources from low- to
high-productivity activities. Urbanization may be important too; it facilitates productivity growth
through information exchange, economies of scale, and the easy movement of resources
among activities.
As an example, Kharas (2009) has compiled a “short list” of policies that he believes will help
to avoid the middle income trap in the PRC (see Table B1-1). These include supporting the
continued transfer of resources from agriculture to industry by eliminating impediments to
migration, and stimulating productivity growth in industry by investing in human capital and
innovation.
Table B1-1: Policies for Avoiding the Middle Income Trap in the PRC
1. Gradually appreciate the exchange rate to raise real wages.
2. Pursue fiscal reform to shift profits to the household sector.
3. Increase agricultural productivity to permit greater flow of migrants to urban areas.
4. Eliminate the hukou system as a disincentive for migration.
5. Provide high quality junior high school education in rural areas to support migration.
6. Strengthen city governance to improve urban social and environmental conditions.
7. Expand access to finance for small and medium enterprises.
8. Remove political obstacles to internal market integration.
9. Reduce logistical barriers to internal market integration.
10. Accelerate innovation by firms and creation of high skilled jobs.
Source: Kharas (2009).
31
ADBI Working Paper 404
Petri and Zhai
Table 6: Productivity Simulations
Effects of productivity deceleration (%)
Scenario 2
Scenario 3
Scenario 4
In
In ACI
advanced
countries
countries
Worldwide
Baseline GDP
2010
Real GDP ($bill.)
ASEAN
PRC
India
ACI
World
Income/capita ($)
ASEAN
PRC
India
ACI
World
Extremely poor (mill.)
ASEAN
PRC
India
ACI
World
Middle income (mill.)
ASEAN
PRC
India
ACI
World
Affluent (mill.)
ASEAN
PRC
India
ACI
World
Total trade ($bill.)
ASEAN
PRC
India
ACI
World
Commodity prices
(2010=100)
Grains
Coal
Oil
Gas
Energy
CO2 emissions (2010=100)
ASEAN
PRC
India
ACI
World
2030
1,566
5,711
1,504
8,781
59,327
4,634
22,083
6,570
33,287
120,065
-17.8
-20.4
-20.9
-20.2
-6.0
-1.7
-2.8
-1.5
-2.3
-5.6
-18.4
-20.6
-21.3
-20.4
-14.0
2,639
4,233
1,229
2,773
8,933
6,564
15,748
4,312
9,165
15,135
-22.4
-26.2
-22.4
-24.8
-7.0
-2.6
-3.0
-1.4
-2.6
-6.4
-23.3
-25.9
-22.3
-24.7
-16.4
109
126
400
635
1,134
22
1
22
45
341
70.9
158.3
163.2
117.9
12.5
4.8
8.6
3.9
4.5
1.9
74.9
154.8
162.2
119.3
34.4
145
212
64
421
2,129
413
1,104
941
2,458
4,909
-13.4
-8.5
-19.8
-13.9
-7.5
-1.1
-0.5
-0.6
-0.7
0.1
-14.0
-8.4
-19.8
-13.9
-9.6
0.2
0.1
0.0
0.3
168
7
42
3
52
417
-9.5
-24.2
-72.2
-29.7
-7.1
-1.0
-3.2
-4.4
-3.0
-9.0
-9.9
-23.9
-72.1
-29.5
-19.3
2,377
2,833
724
5,934
35,628
10,141
15,377
3,713
29,231
105,320
-15.9
-15.0
-17.8
-15.7
-5.1
-1.8
-4.6
-2.3
-3.3
-5.7
-16.4
-17.7
-20.0
-17.6
-13.7
100
100
100
100
100
121
150
177
167
175
-3.7
-2.5
-6.3
-2.2
-5.7
-1.9
-1.5
-3.1
-3.5
-3.1
-7.0
-4.7
-13.0
-11.0
-12.4
100
100
100
100
100
214
192
234
201
161
-11.4
-10.1
-13.7
-10.9
-3.9
1.0
-0.1
1.0
0.2
-1.1
-7.4
-8.4
-11.3
-8.8
-6.3
Source: Authors’ calculations.
32
ADBI Working Paper 404
Petri and Zhai
These messages emerge from the simulations:
•
•
•
•
•
•
A deceleration of ACI economies would have the largest adverse effects; per capita
incomes would fall by 25% in 2030, representing a 1.5 percentage point reduction in the
growth between 2010 and 2030. The scale of this effect warrants serious attention to
factors that could lead to a middle-income trap, including insufficient institutional
progress, technological upgrading and innovation, and threats to regional and global
integration.
A deceleration affecting advanced economies would have milder effects on ACI growth. 2
Although the world incomes would be 6% smaller in 2030 under these assumptions, ACI
economies would experience per capita income losses of under 3%. In effect, the
simulation projects that ACI economies could replace lost markets in advanced countries
by depending more fully on their own and other emerging markets.
A worldwide deceleration would result in a 14% reduction in world incomes, but the
effects for ACI economies would be similar to those of its own middle-income trap. In this
case, the losses that might be expected due to the deceleration in external markets
would be offset by the one positive side-effect of a global slowdown: declining costs of
imported energy and other raw materials.
All three scenarios would have significant effects on the size of population consumption
groups. For example, productivity deceleration in the ACI economies or worldwide would
reduce the size of the ACI middle class by 14% and more than double the number of
people in extreme poverty (albeit from a small base in 2030).
While deceleration would be costly, it is unlikely to change general qualitative trends.
Even if the shocks originated in the ACI economies, by 2030 ACI per capita incomes
would rise considerably, increasing by a factor of 2.6, rather than 3.3 in the baseline.
While the incidence of poverty would rise relative to the baseline, it would remain far
below levels in 2010. The international trade of the ACI economies would be still 4.2
times as large as in 2010, compared to 4.9 times under the baseline.
Nor would deceleration significantly reduce the challenges of rising energy demands and
carbon emissions. These side-effects of growth would of course diminish with lower
output, but would still threaten sustained growth. Pressures on grain and energy prices
would ease, but still follow an upward trend. Deceleration would only postpone—perhaps
by no more than a few years—issues that will have to be addressed sooner or later with
targeted policy responses.
The middle-income-trap scenario (ACI productivity growth deceleration) poses the most severe
threat to ACI development among those examined in this study. All three types of deceleration
would impose new hardships on the poor and significantly impact all income classes in the
region. Moreover, growth would not slow enough to solve environmental sustainability issues,
nor ease significantly the scale of the transformation required in ACI economies.
2
In these simulations, the deceleration takes hold slowly, by reducing investment and annual productivity gains. Thus
countries have time to shift their trade to more robust markets. The results of such a slow transition are quite
different from those under a sudden, large unexpected shock, as experienced during the global financial crisis.
33
ADBI Working Paper 404
5.2
Petri and Zhai
Structural shocks and policy alternatives
The ACI economies also face specific structural challenges. These risks, explored in Scenarios
5–8, include adverse developments in global food markets, new energy price shocks, and
increasingly severe environmental stress of social inequality which require significant new policy
actions.
The baseline projections already incorporate economic adjustments and policy changes on
these fronts. For example, the projections envision rising food and energy prices that result in
conservation. They incorporate “new policies” (as defined in International Energy Agency [2011]
studies of climate change mitigation) to generate cleaner energy and reduce carbon emissions.
And they build in rising government expenditures on social programs, as reflected in recent
data. But these baseline changes may still understate the adjustments that would become
necessary given earlier or larger technological, production or social challenges than are now
expected. Such developments, or policy responses to them, could adversely affect the structure
and pace of the region’s economic growth.
The first major risk involves adverse developments in agricultural markets. Continuing pressures
on food production could lead to much less favorable agricultural prospects than represented in
the baseline scenario. The pressures include meat-rich diets that require additional agricultural
inputs, declining supplies of arable land that result from the growth of non-agricultural activities,
and perhaps adverse effects from climate change. Scenario 5 simulates these risks by
assuming deceleration in agricultural productivity growth large enough to increase agricultural
prices by an extra 2% per year over the baseline. Over the 20-year horizon of our analysis, this
would mean that agricultural prices would rise by 93% instead of 30% in the baseline solution.
Experts disagree on how agricultural prices are likely to evolve; some supporting an upward drift
of prices (McKinsey Global Institute 2011), but others are more optimistic that long-term trends
in productivity increases will resume and offset demand pressures (Anderson and Strutt 2011).
The effects of an upward agricultural price shock are shown in Table 7. Since ACI are net
importers of agricultural products and downstream products in the food manufacturing sectors,
the rise in agricultural prices would lead to a deterioration in their terms of trade and a 6%
decline in real incomes. The effects on income would be modest because, as noted above, by
2030 agriculture is projected to account for only 6% of output in ASEAN and the PRC and 17%
in India. As the most food-import-dependent economy among the three, the PRC would
experience the largest losses. The losses would shift income distributions downward, increasing
the number of people in low-income groups, including those in extreme poverty.
34
ADBI Working Paper 404
Petri and Zhai
Table 7: Structural Policy Simulations
Baseline GDP
2010
2030
Real GDP ($bill.)
ASEAN
PRC
India
ACI
World
Income/capita ($)
ASEAN
PRC
India
ACI
World
Extremely poor (mill.)
ASEAN
PRC
India
ACI
World
Middle income (mill.)
ASEAN
PRC
India
ACI
World
Affluent (mill.)
ASEAN
PRC
India
ACI
World
Commodity prices (2010=100)
Grains
Coal
Oil
Gas
Energy
Grain demand (2010=100)
ASEAN
PRC
India
ACI
World
Energy demand (2010=100)
ASEAN
PRC
India
ACI
World
CO2 emissions (2010=100)
ASEAN
PRC
India
ACI
World
Scenario 5
Agriculture shock
Effects of structural policies (%)
Scenario 6
Scenario 7
Energy shock
Green economy
Scenario 8
Social inclusion
1,566
5,711
1,504
8,781
59,327
4,634
22,083
6,570
33,287
120,065
-1.5
-2.8
-3.7
-2.8
-1.4
-2.1
-2.3
-3.1
-2.4
-2.1
-1.7
-2.1
-2.0
-2.0
-1.5
-3.2
-3.8
-3.0
-3.5
-1.0
2,639
4,233
1,229
2,773
8,933
6,564
15,748
4,312
9,165
15,135
-3.0
-7.0
-5.9
-6.1
-2.5
-4.1
-3.4
-4.5
-3.8
-2.8
-2.5
-2.8
-2.3
-2.7
-2.2
-5.3
-6.1
-4.4
-5.6
-1.5
109
126
400
635
1,134
22
1
22
45
341
7.5
27.5
25.8
16.9
5.1
10.6
13.4
19.3
14.8
4.9
6.1
11.3
10.5
8.4
5.8
-71.0
-92.5
-76.1
-74.2
-6.2
145
212
64
421
2,129
413
1,104
941
2,458
4,909
-1.7
-1.8
-4.0
-2.7
-1.8
-2.3
-0.9
-3.1
-2.0
-1.4
-1.3
-0.7
-1.7
-1.2
-1.2
10.6
8.5
8.6
8.9
4.6
0.2
0.1
0.0
0.3
168
7
42
3
52
417
-1.6
-7.8
-24.1
-9.4
-3.2
-2.1
-4.5
-19.0
-6.5
-4.1
-1.3
-3.9
-11.1
-4.7
-3.0
-23.6
-21.7
-81.9
-32.5
-7.0
100
100
100
100
100
121
150
177
167
175
21.1
-1.1
-1.7
-1.4
-1.6
1.4
46.2
48.1
46.3
47.8
-0.9
-11.5
-8.6
-21.7
-10.2
-0.7
-0.4
-1.1
-0.4
-1.0
100
100
100
100
100
149
209
261
216
176
-9.8
-13.0
-14.9
-13.4
-9.9
-1.0
-1.7
-2.3
-1.9
-1.5
-0.5
-1.4
-1.0
-1.1
-1.1
-1.6
-3.1
-2.0
-2.4
-1.3
100
100
100
100
100
173
178
221
185
135
-0.3
-2.4
-2.0
-1.8
-0.3
-35.1
-27.5
-33.8
-30.7
-29.6
-3.4
-22.0
-8.1
-14.5
-8.1
-1.4
-1.9
-2.1
-1.9
-0.2
100
100
100
100
100
214
192
234
201
161
-0.7
-2.6
-1.7
-2.2
-1.0
-30.3
-25.2
-33.7
-27.3
-25.4
-16.7
-40.0
-44.0
-38.0
-26.1
-2.1
-2.1
-2.4
-2.2
-0.8
35
ADBI Working Paper 404
Petri and Zhai
Source: Authors’ calculations.
A second important risk involves energy. The baseline suggests that energy prices will rise in
the future, reversing historical downward trends. These increases will lead to energy
conservation and greater energy efficiency, that is, declining energy use per dollar of output.
Even so, energy demand is projected to increase by 85% in ACI and by 35% worldwide
between 2010 and 2030. Ultimately, these large increases and resulting growth in import
dependence set the stage for serious risks to energy security. If conservation and supply turn
out less favorably than expected—or if access to supplies becomes difficult—sharp rises could
occur.
Scenario 6 represents a shock in the world energy sector by assuming a decline in productivity
in energy production that is large enough to raise prices by an extra 2% per year above the
baseline projection. The effects are again moderate, increasing energy prices by an additional
15% over the baseline and reducing world output by 0.3%. ACI economies would be clearly—
both not severely—affected, since they have relatively high energy/GDP ratios and are still in
the energy-intensive stage of their industrialization process. Income losses in 2030 would range
from 2% for ASEAN to 3% for India (Table 7).
A third source of risk is the environment. By 2030 ACI economies will account for 43% of global
carbon emissions under baseline assumptions, which include the International Energy Agency’s
(IEA 2011) “new policies” for mitigating emissions. These policies will substantially reduce the
carbon intensity of GDP. But they fall short of limiting absolute emissions to the extent required
by the climate-stabilizing “450 scenario.” Should internal and external political pressure require
ACI to limit emissions to the levels required by this scenario, ACI will need to adopt more
intensive policy changes and investments.
Scenario 7 simulates “green” policies based on IEA analysis. It sets carbon taxes to $90 per ton
of CO 2 in advanced economies and $60 per ton of CO 2 in emerging economies. It also
assumes substantial investments in energy conservation and alternative energy supplies. By
diverting investment from the non-energy sectors of the economy, the scenario forecasts
reductions in real incomes as conventionally measured. The results show that real incomes in
ACI economies would fall by 3%. The losses are lower than the cost of the mitigation strategies,
because the investments would be offset by energy savings and some decline in energy prices.
The scenario shows that global energy demand would decline by 8% and lower the cost of
various fuels (prior to carbon taxes) by 9 to 22%, helping to offset the cost of green policies.
Green policies would reduce CO 2 emissions relative to the baseline by 38% in the ACI
economies and by 26% for the world as a whole and, if the underlying IEA analysis is correct,
they would help to stabilize the climate. These reductions in CO 2 emissions would be
considerably greater than those observed in the energy shock scenario or the growth
deceleration scenarios—the mitigation of climate change would be achieved at a lower cost with
green growth. Put another way, any level of climate benefits can be achieved far more efficiently
by policies that broadly target, and selectively reduce, emissions rather than policies that
attempt to bring emissions under control indirectly, through piecemeal initiatives or slow growth.
Similar messages have also emerged from previous simulations of Asian climate change
policies (Van der Mensbrugghe 2010).
The fourth structural risk considered involves income inequalities. Since market-based growth
does not automatically produce equitable incomes, strategies to pursue inclusive growth—
examined in Chapter 3—also involve initiatives such as targeted health and education services,
and robust safety nets. The strategy also prioritizes investments in transport and
communications to help link disadvantaged areas to centers of economic growth, Given the
36
ADBI Working Paper 404
Petri and Zhai
spread of modern communications technologies, citizens increasingly expect—and often
effectively demand—governments to deliver these services.
As noted below, most ACI governments are steadily strengthening programs to help close
income gaps and planning is underway on a wide range of strategies that could generate further
progress. Since the baseline solution projects rapid growth in government expenditures, it
implicitly supports an expansion in the implementation of inclusive growth initiatives.
Nevertheless, given adverse market developments and/or mounting political pressures, or if the
strategies prove to be less effective than expected, governments may need to undertake more
intensive efforts. These initiatives may also include targeted transfers to the poor.
Scenario 8 quantifies an intensive “inclusive growth” strategy by increasing government
expenditures by 2% of GDP for related investments and services in each ACI economy. Since
government expenditures account for 12% of GDP for ACI countries as a group, this increment
represents a substantial addition to the resources that the public sector can direct toward social
expenditures. (These expenditures are assumed to be financed half by taxes and half by
borrowing.) Aside from simulated economic effects, we assume that these efforts could reduce
a country’s projected Gini coefficient by 5 percentage points, or approximately by one half of a
standard deviation of the distribution of Gini coefficients in World Bank statistics.
Since social policies are assumed to be financed in part by borrowing—which imply reductions
in investments in conventional production—their effects include a decline of 6% in ACI real
incomes in ACI by 2030. Despite these lower average consumption levels, successful policies to
combat inequality are likely to produce gains for many people. For example, the simulations
suggest that the number of those living in extreme poverty will decline by 74%—a result that
essentially means the eradication of extreme poverty. The numbers of those in the middle class
would increase by 9%. The negative effects would be felt only in the upper tail of the income
distribution: The number of affluent consumers would decrease by 33%. 3 These are large
changes in inequality at relatively low overall costs. Nevertheless, the output effects are
significant—for example, they are greater than those calculated for an energy shock. Yet they
are much lower than losses associated with the middle-income trap—so if social cohesion can
help countries avoid the middle-income trap, say by allowing governments to undertake greater
market liberalization, their cost would be amply repaid.
5.3
International linkages and cooperation
ACI’s future growth depends on deeper international trade and investment linkages. These
linkages are crucial to shifting demand from advanced economies to domestic and regional
sources; for sustaining the region’s central role in global manufacturing; and for meeting ACI’s
rising external energy, food, and resource requirements. These linkages, in turn, will require
vigorous intra-regional trade; strong ties with the emerging markets elsewhere, and sustained
trade, financial, and technology flows with advanced economies. From a policy perspective,
these intense trade relationships could lead either to greater cooperation or new tensions. The
risks would be exacerbated if growth in advanced economies remains sluggish, and especially if
global trade rules fail to be strengthened. These risks are assessed in Scenarios 9–12 (Table
8).
3
The largest negative effects are projected for India, where a significant part of the high income population is
projected to be close to the cutoff limit of the middle consumption group, and therefore moderate changes in
incomes have a large effect on shifting individuals from one group into the other.
37
ADBI Working Paper 404
Petri and Zhai
Table 8: Trade and Capital Flow Simulations
Baseline GDP
2010
Real GDP ($bill.)
ASEAN
PRC
India
ACI
World
Income/capita ($)
ASEAN
PRC
India
ACI
World
Extremely poor (mill.)
ASEAN
PRC
India
ACI
World
Middle income (mill.)
ASEAN
PRC
India
ACI
World
Affluent (mill.)
ASEAN
PRC
India
ACI
World
Total trade ($bill.)
ASEAN
PRC
India
ACI
World
Terms of trade (2010=100)
ASEAN
PRC
India
ACI
2030
Effects of international policies (%)
Scenario 9
Scenario 10 Scenario 11 Scenario 12
Rising
Regional
Global
Fast
protection
agreement
agreement rebalancing
1,566
5,711
1,504
8,781
59,327
4,634
22,083
6,570
33,287
120,065
-0.4
-0.7
-0.1
-0.5
-1.8
14.4
1.8
6.1
4.7
1.0
22.1
8.4
17.9
12.7
7.7
6.8
5.3
-3.4
3.5
0.1
2,639
4,233
1,229
2,773
8,933
6,564
15,748
4,312
9,165
15,135
-1.1
-1.0
0.0
-0.8
-1.5
15.3
1.3
2.2
3.6
0.8
22.2
7.6
9.3
10.2
6.5
8.7
7.2
-3.6
4.9
0.5
109
126
400
635
1,134
22
1
22
45
341
2.7
3.8
0.1
1.5
0.9
-30.0
-4.5
-8.8
-19.0
-1.6
-39.6
-23.1
-31.9
-35.3
-16.0
-19.9
-23.4
17.9
-2.0
4.0
145
212
64
421
2,129
413
1,104
941
2,458
4,909
-0.6
-0.2
0.0
-0.2
0.0
7.5
0.2
1.5
2.0
1.0
10.2
1.2
5.8
4.6
3.9
4.8
1.2
-2.9
0.2
-0.5
0.2
0.1
0.0
0.3
168
7
42
3
52
417
-0.6
-1.9
-0.1
-1.3
-2.4
9.0
2.8
11.2
5.5
1.0
13.5
16.7
53.6
-7.9
0.4
5.3
16.9
-17.8
8.6
1.0
2,377
2,833
724
5,934
35,628
10,141
15,377
3,713
29,231
105,320
-1.6
-5.4
-1.7
-3.6
-6.0
19.5
10.9
24.3
15.6
3.6
21.5
36.7
66.2
35.1
19.4
5.8
3.9
-2.7
3.7
0.9
100
100
100
100
122
95
88
103
-0.3
-0.8
-0.2
-0.5
-
5.2
0.6
-0.5
1.9
-
7.2
2.5
0.9
3.8
-
-0.4
-0.5
0.1
-0.3
-
100
100
100
100
100
100
100
121
150
177
167
175
-0.9
-1.3
-2.0
-1.8
-2.0
0.7
0.7
1.3
0.8
1.2
0.8
1.0
4.7
4.3
4.5
0.6
0.6
0.7
0.1
0.7
World
Commodity prices (2010=100)
Grains
Coal
Oil
Gas
Energy
Source: Authors’ calculations.
38
ADBI Working Paper 404
Petri and Zhai
Scenario 9 considers the possibility of increased protectionism in advanced economies. Such
policies would represents a costly, self-defeating response to challenges faced by the world
economy, but unfortunately cannot be ruled out as political reactions to increase economic
stress. They would reduce world trade by 6% and world real GDP by 2%, with the costs
disproportionately falling on those protectionist economies themselves. Such policies would also
generate losses for the ACI economies and especially the PRC, given its strong linkages with
advanced economies. The PRC’s trade would fall by 6%. Estimated global losses would be
around $2 trillion, with more than 90% falling on non-ACI economies. The stakes are high for
the world as whole; it would be hard to find significant groups of consumers or workers
anywhere who would benefit from such adverse developments in the global trading system.
At the other extreme, steps that reduce barriers and strengthen trade rules would benefit ACI
economies as well as other regions. Scenario 10 examines free trade among ACI economies,
and Scenario 11 hypothesizes substantial liberalization of world trade. In reviewing these
results, it is necessary to recognize that a global model produces only rough estimates of the
effects of trade liberalization. Its parameters are not sufficiently detailed on trade barriers and
the possibilities for reducing them. Yet such models provide useful estimates of overall effects,
and in this case the scenarios point to the possibility of large benefits.
A regional agreement that connects ACI economies would generate especially large income
gains for ASEAN and India, estimated at 15% and 2%, respectively. The trade effects would be
particularly large (suggesting a 24 increase) for India, which is now less closely connected with
the ACI region’s trading system. The results confirm a central theme of international economics,
that broad global liberalization is more beneficial than an effective regional agreement, even one
covering a large, dynamic economic zone. A global trade agreement would raise world incomes
by 7% and ACI incomes by11%. World trade would increase by 19%, with India increasing its
total trade by 66%. The results confirm a central theme of international economics, that broad
global liberalization is more beneficial than an effective regional agreement, even one covering
a large, dynamic economic zone.
Much attention has focused in recent years on the linkages of ACI economies to the global
economy through gross and net capital flows. The current account surpluses of ACI countries
and corresponding deficits in the United States, although they have diminished sharply since the
onset of the global financial crisis, have come under particular scrutiny. As many observers
have noted, it is difficult to justify a large “uphill” flow of capital from rapidly growing, capital-poor
economies to slowly-growing, capital-abundant ones. This report and other recent policy studies
have emphasized the benefits of “rebalancing” economies in ways that put capital to greater use
within emerging economies themselves, either to stimulate investment or raise living standards
directly. The baseline itself assumes that rebalancing efforts will be successful enough to keep
imbalances stable in nominal terms at 2010 levels, thus declining steadily and substantially
relative to GDP and other economic magnitudes.
To simulate the implications of rapid rebalancing, Scenario 12 imposes a faster adjustment path
on capital accounts that would eliminate all imbalances by 2020. This rebalancing program
would have a modest positive effect on ACI economies, yielding a 4% increase in GDP for the
region. The scenario generates increases for ASEAN and PRC, which have positive capital
flows along the baseline and so would have more resource available domestically under a faster
rebalancing scenario. It generates decreases for India, which has negative capital flows along
the baseline and would therefore have to reduce expenditures under stricter current account
balances. The results for economies that are initially in surplus are positive: Applying more
resources domestically—as opposed to low-interest foreign investments—would increase
welfare directly through consumption and indirectly through investment and growth. The
opposite is true for economies that are initially in deficit: Reduced domestic expenditures would
39
ADBI Working Paper 404
Petri and Zhai
lead to lower consumption and/or lower growth. At the micro-economic level, production in
surplus economies would shift from manufacturing to services, and vice versa in deficit
economies. Despite the considerable policy attention focused on rebalancing, the results show
modest effects on the size and structure of ACI economies in the long run.
The international simulations suggest more opportunities than threats. A turn toward
protectionism in advanced economies, to be sure, would reduce world output and negatively
affect ACI growth prospects, but the effects would be felt primarily in the protectionist
economies themselves. On the other hand, new liberalization measures, within the ACI region,
but especially those with global scope, could raise world incomes dramatically. And as in the
past rounds of global liberalization, the effects are likely to extend beyond the direct economic
gains to such indirect benefits as increased investor confidence, possibly more innovation, and
an improved climate for economic and political cooperation. More refined analyses are needed
to develop precise estimates, but the results confirm the continued importance of an effective
global trading system and the value of global negotiations. Renewed progress on global
economic integration could go a long way toward offsetting the effects of many possible
negative shocks in other dimensions of the world economy.
6. COMBINATIONS OF RISKS
The results of individual scenarios confirm the concerns of analysts and policymakers that the
region faces serious vulnerabilities. Yet they also show that most shocks could be managed at
moderate cost if addressed with timely policy. But the implications of shocks could be more
severe if they occurred suddenly and in combination—for example, if an unexpected
deceleration in productivity growth followed a sudden energy shock. Such interactions among
shocks are not unusual. An energy shock and a productivity shock combined to create a longlasting global economic slowdown in the late 1970s. In the current economic cycle, financial,
macroeconomic, food and energy shocks appear to be acting in combination. Thus, it is
important to assess not just the possibility of specific adverse developments, but also the
possibility that they might combine into “perfect storms” with severe impacts on growth.
Assessing the probability of combined shocks requires further, and necessarily speculative,
analysis. Are shocks independent of each other, with each following a separate probability
distribution? Or do shocks interact, so that a cluster of negative shocks, for example, is more
likely than the probability that they happen together by chance? The interaction hypothesis
represents the theory that “perfect storms” do not merely happen randomly—that adverse
developments raise the probability of other negative events (and in contrast favorable
developments tend to raise the probably of other positive ones). There are various mechanisms
that could lead to such results—for example, a system under pressure may put further stress on
weak components and thus fail in multiple ways. Positive results, too, can reinforce each other:
a generally strong economy will generate resources and good will that can be helpful in avoiding
problems.
The implications of multiple shocks are bracketed by two assumptions: full independence, which
implies that shocks occur according to separate probability distributions and interact only by
chance, and partial interdependence, which assumes that “like” shocks are more likely to occur
jointly. Even the assumption independence, of course, will give rise to rare “perfect storms” with
multiple negative or positive effects. The more worrisome assumption, however, is partial
interdependence, which implies that multiple shocks in the same direction—and hence larger
overall effects—have greater joint probability than mere coincidence would suggest. Although
the arguments for such correlations are familiar now, they were not widely analyzed in the run40
ADBI Working Paper 404
Petri and Zhai
up to the global financial crisis and may have contributed to pervasive underestimation of
financial risks by financial markets and governments.
Both possibilities were examined in simulations that combined the effects of the scenarios. This
exercise statistically simulated 10,000 future growth paths of the global economy with the
baseline subjected to various combinations of 10 of the shocks listed in Table 5 and Figure 13.
A probability of 25% was assigned to any of the shocks occurring independently. A benchmark
“independent” simulation was calculated by using a random draw to determine which particular
shocks occurred on each of the 10,000 paths. The results for each path were also calculated by
adding the effects of the active shocks to the baseline.
Under the benchmark simulations, the possibility that none of the 10 shocks would occur is
small—around 5% (0.7510). There is an even smaller chance, approximately one in a million,
that all 10 shocks would occur together (0.2510). Most results are in between, and there is a
large probability (around 40%) that the sum of the shocks turns out to be small, in the range
reported as “0” column in Figure 14A. These paths involve cases with either few shocks, or
more likely combinations of positive and negative ones. The probability rises to around 75% for
a somewhat wider range of outcomes, spanning minus 10% to 10% in income deviations from
the baseline. These middle outcomes have small shocks as well as combinations with canceling
positive and negative shocks.
Figure 14: Frequency of Combined Shocks
A. Independent shocks
B. Interdependent shocks
Source: Authors’ calculations.
Despite the “regression toward the mean” that is characteristic of statistical phenomena, around
25% of the distribution falls into the tails, representing significant deviations from the baseline.
Of this, only 1% is in the “right tail” of significant positive shocks, and 24% is in the “left tail,”
describing significant negative ones. Outcomes on the downside dominate those on the upside
because the scenarios considered cannot offer a positive counterpart to the middle income trap.
In the risky context of the current world economy, the plausible alternatives to baseline are on
the negative side. The net effect is an approximately one-in-four chance that ACI income levels
will fall well short of the baseline, with long-term growth rates at least one percentage point
lower.
The frequency of outcomes changes if the assumption of statistical independence is replaced by
partial interdependence—the expectation that similar shocks are likely to occur together. The
simulations described above were repeated under the assumption that four or more shocks in
the same direction were four times as likely to occur as their joint independent probabilities
would suggest. The new results in Figure 14B show a wider distribution of outcomes, with
41
ADBI Working Paper 404
Petri and Zhai
higher frequencies for both less and more favorable extremes. But while the more positive
extreme rises only slightly, significant negative shocks increase to 29%. This also pushes the
expected value of outcomes 2% below the baseline. Under the assumption of interdependence,
unfavorable shocks that dominate alternatives to the baseline now combine with greater
frequency.
These results are highly speculative—they attempt to attach probabilities to mostly unknowable
future events. Nevertheless, they offer some insight into the risks the surround the baseline
projections. They show, first, that in a dominant subset of cases (ranging perhaps from twothirds to three-quarters) the results would not differ significantly from the baseline, not because
all will go as projected, but because positive and negative deviations will often cancel each
other. Second, the results suggest that there is a relatively small chance of substantially better
outcomes, mainly because the outlook at this writing suggests more scenarios that lead to
worse than projected outcomes rather than better ones. Finally, the results show a modest, but
significant chance of significantly worse outcomes than the baseline indicates. Yet even these
adverse outcomes provide for substantial improvements in the quality of life in the ACI region. In
short, these stress tests of the projection suggest that while ACI economies face serious
threats, and could fall short of projections by significant margins, they are generally large and
diversified enough to withstand many of the shocks that are hypothesized, even if they occur in
adverse combinations.
This is not an argument for complacency: Downside risks are significant and their effect implies
substantial losses for all and great hardship for some. There are powerful reasons to prepare for
and to mitigate the shocks examined in this chapter, and especially to avoid the middle income
trap. Such policies are insurance against adverse outcomes, and given the frequency and cost
of adverse outcomes, they offer high expected return. But an important second message is that
even in unfavorable environments the ACI economies will make substantial progress. The great
transformation of the ACI economies is subject to risk, but is likely to withstand much turbulence
in its environment.
7. CONCLUSIONS
The baseline prospects of ACI economies are generally positive and are likely to lead to major
improvements in their citizens’ quality of life. Moreover, the region’s growth trajectory is likely to
fuel the engines that drive growth, generating new types of demand and greater competition,
innovation, and regional integration. Yet given the region’s scale, its projected growth will also
strain resources and the environment. These opportunities and risks were examined in this
chapter with several simulated alternatives.
The baseline growth path for the next two decades, based on a general equilibrium model and
ADB growth projections, suggests major changes in the region’s economy and in its role in the
world economy. ACI incomes are likely to increase around four fold. The region is already the
most populous in the world, and by many measures it will become the world’s largest economy
as well. In some markets—including especially investment goods—it will account for more than
half of global aggregates. The region’s standard of living will improve dramatically not just in
terms of per capita income, but also in terms of indicators such as the declining incidence of
poverty and the growth of the middle class.
As its growth continues, ACI’s interdependence with the world economy will deepen and evolve
in new directions. Trade will expand with the continued growth of production chains, increased
imports of primary products, and shifting patterns of comparative advantage within the region
and beyond. Intra-regional and South-South trade will be the most rapidly growing components
42
ADBI Working Paper 404
Petri and Zhai
of ACI trade, but North-South trade will remain important. Accordingly, trade liberalization
regionally and globally could generate major benefits to ACI and the world, raising incomes by
5–10% of GDP.
Taken together, these changes will also help to create new engines for ACI growth. The growth
in global investment and consumption goods markets will be led by the ACI region, and shaped
by the tastes and expectations of ACI investors and especially middle income consumers.
Global financial transactions will gravitate toward ACI’s large pools of savings and investment,
helping to give rise to deeper regional capital markets. And high rates of investment will enable
the region to create and adopt new technologies to build greener and more energy-efficient
infrastructure and industry.
This generally optimistic outlook is nevertheless subject to significant risks, involving factors
within the region as well as others in its international environment. There are few ways to meet
all of the requirements of rapid growth, but many ways to fall short, so numerous adverse
scenarios can be envisioned. Each could have a substantial impact on ACI growth, and
especially the region’s more vulnerable populations. The most negative among these could
depress ACI growth rates by 5 percentage points (below the 6.9% on the baseline scenario).
Simulations suggest that the chances for falling short of the baseline’s long-term growth rate by
at least 1 percentage point are in the neighborhood of 25–30%.
Significant decelerations in productivity growth—the middle income trap—pose the single most
important threat to the growth of ACI economies. (This threat is substantial in all ACI economies,
but takes different forms depending on their levels of development.) The costs of a prolonged
slowdown warrant wide-ranging preventive investments. Other risks—higher food prices, higher
energy costs, the need for substantial new expenditures to address climate change or
inequality—have significant sectoral implications, but have more modest overall effects. The
greatest threats involve combinations of shocks—“perfect storms” that increase strains on
economic systems and raise the probability of additional shocks or policy failures.
Despite the serious risks to the baseline, the simulations generally predict more moderate
deviations from the baseline. The growth of ACI incomes and their rising role in the world
economy remain salient features of all scenarios. Even scenarios with economic growth rates
substantially below that of the baseline yield large increases in ACI living standards and in ACI
shares of most important global aggregates. The scale and speed of the great transformation
are subject to risk, but their scope and directions appear robust.
As all economic results, these conclusions depend on the underlying model and the scenarios
examined. The general equilibrium model used in this study, as others of its type, is built on
assumptions of market interactions and represents the myriad of ways in which producers,
consumers, and traders adjust production and investments to price signals. When markets work
properly, these signals and market reactions enable economies to conquer adverse
developments at relatively low cost. Markets can of course fail, as financial markets have done
so spectacularly in recent years. But in the two-decade time horizon of this analysis, most
failures are likely to be corrected—or put another way, can be counted among the many risks
already captured in productivity deceleration scenarios. Serious downside risks that have not
been considered in this study include wars, epidemics, or the collapse of governance in major
countries, which lie beyond the scope of economic analysis.
The ACI economies face strong prospects and large challenges. They are embarking on a new
stage of development with unprecedented momentum, but in midst of great global uncertainty.
Reasonable projections indicate that their growth will continue and is, indeed, likely to be selffueling. This growth is likely to lead to major advances in the region’s standard of living and in its
global role. At the same time, wide-ranging structural changes will be required, in economies
43
ADBI Working Paper 404
Petri and Zhai
that are already large relative to the world economy. These changes imply significant, multiple
risks. This chapter has identified scenarios that could produce much less progress than
expected on the baseline, but has also found that the outline of the coming transformation of
ACI economies is robust, despite uncertainties about its timing and speed.
44
ADBI Working Paper 404
Petri and Zhai
REFERENCES
Anderson, K. and A. Strutt, 2011. Agriculture and Food Security in Asia by 2030. Background
paper prepared for this study.
Asian Development Bank. 2011. Asia 2050: Realizing the Asian Century. Manila: ADB.
Deaton, A. and A. Heston, 2010. Understanding PPPs and PPP-based national accounts.
American Economic Journal: Macroeconomics 2(4): 1–35. October.
Goldman Sachs, 2003. Dreaming with BRICs: the Path to 2050. Global Economics Paper No.
99. New York: Goldman Sachs.
International Energy Agency. 2011. World Energy Outlook 2011. Paris: International Energy
Agency.
Kawai, M. and F. Zhai, 2010. Asia’s Post-Global Financial Crisis Adjustment: A Model-Based
Dynamic Scenario Analysis. Asian Development Review 27(2): 122–151.
Kharas, H. 2009. China’s Transition to A High Income Economy: Escaping the Middle Income
Trap. Processed. Washington, DC: Brookings Institution.
———2010. The Emerging Middle Class in Developing Countries. Working Paper No. 285.
Paris: OECD Development Centre.
Lee, J. W., and K. Hong. 2010. Economic Growth in Asia: Determinants and Prospects. ADB
Economics Working Paper Series No. 220. Manila: Asian Development Bank.
Leibenstein, H., 1954. A Theory of Economic Demographic Development. Princeton, NJ:
Princeton University Press.
McKinsey & Co. 2011. New Frontiers of Growth: 2011 Annual Chinese Consumer Study.
Shanghai: McKinsey & Co.
McKinsey Global Institute, 2011. Resource Revolution: Meeting the world’s energy, materials,
Melitz, M. J. 2003. The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry
Productivity. Econometrica 71(6): 1695–1725.
Nelson, R. R., 1956. A Theory of the Low-Level Equilibrium Trap in Underdeveloped Economies.
The American Economic Review 46(5): 894–908. December.
Petri, P. A. and F. Zhai, 2012. Navigating change: Prospects and Challenges of ASEAN, the
PRC and India, 2010–2030. Background paper for this study.
Petri, P. A., M. G. Plummer and F. Zhai, 2012. The ASEAN Economic Community: A General
Equilibrium Analysis. Asian Economic Journal 26(2): 93–118.
Pinkovskiy, M. and X. Sala-i-Martin. 2009. Parametric Estimation of the World Distribution of
Income. NBER Working Paper No. 15433. October. Cambridge: NBER.
Rimmer, M. T. and A. A. Powell, 1996. An implicitly additive demand system. Applied
Economics 28(12): 1613–1622.
Rodrik, D., 1999. Where Did All the Growth Go? External Shocks, Social Conflict and Growth
Collapses. Journal of Economic Growth 4: 385–412.
45
ADBI Working Paper 404
Petri and Zhai
Ros, J. 2005. Divergence and Growth Collapses: Theory and Empirical Evidence. In: Beyond
Reforms: Structural Dynamics and Macroeconomic Vulnerability, edited by J.-A. Ocampo.
Stanford, CA: Stanford University Press.
Roland-Holst, D., J.-P. Verbiest, and F. Zhai, 2005. Growth and Trade Horizons for Asia: Longterm Forecasts for Regional Integration. ERD Working Paper No. 74. Manila: ADB.
United Nations, Department of Economic and Social Affairs. 2012. World Urbanization
Prospects: The 2011 Revision. New York: United Nations.
Van der Mensbrugghe, Dominique. 2010. Climate Change Policy Options for Asian Economies:
Findings from an Integrated Assessment Model. Asian Economic Policy Review 2010:
63–83.
Yu, W., T. W. Hertel, P. V. Preckel, and J. S. Eales. 2004. Projecting world food demand using
alternative demand systems. Economic Modelling 21(1)(January): 99–129.
Zhai, F. 2008. Armington Meets Melitz: Introducing Firm Heterogeneity in a Global CGE Model
of Trade. Journal of Economic Integration 23(3): September.
Zhuang, J., 2011. ASEAN, PRC and India Growth Projections 2010–2030. Powerpoint
presentation. Manila: ADB.
46
ADBI Working Paper 404
Petri and Zhai
APPENDIX 1: REGIONS AND PRODUCTION SECTORS
Regions
1. ASEAN
2. PRC
3. India
4. Latin America
5. Other emerging economies
6. Japan
7. Rep. of Korea and Taipei,China
8. Australia and New Zealand
9. Canada
10. US
11. Europe
Production sectors
1. Rice
2. Other grains
3. Other crops
4. Livestock
5. Forestry and fisheries
6. Coal
7. Oil
8. Natural gas
9. Other minerals
10. Food processing
11. Textiles
12. Apparel
13. Chemicals
14. Metals
15. Machinery
16. Electrical equipment
17. Transportation equipment
18. Other manufacturing
19. Utilities
20. Gas distribution and transport
21. Construction
22. Wholesale and retail trade
23. Transport and communications
24. Finance
25. Private services
26. Government services
Source: Authors’ specification.
47
ADBI Working Paper 404
Petri and Zhai
APPENDIX 2: TECHNICAL DESCRIPTION OF THE CGE
MODEL
The CGE model used in this study is a version of a global general equilibrium model developed
by Van der Mensbrugghe (2009) and Zhai (2008). The model incorporates recent
heterogeneous-firms trade theory into an empirical global CGE framework. It features intraindustry firm heterogeneity in productivity and fixed cost of exporting, which enables us to
investigate the intra-industry reallocation of resources and the exporting decision by firms, and
thereby capture both the intensive and extensive margin of trade. The model also incorporates a
greenhouse gas (GHG) emissions module and mitigation policies. The model is calibrated to the
GTAP (version 8 pre-release) global database with 2007 base year and implemented in GAMS
programming language. It includes 11 country/region and 26 sectors.
Dynamics of the model are driven by exogenous population and labor growth, capital
accumulation driven by savings and exogenous technological progress. Within each time period
a full equilibrium is achieved given the fixed regional endowments, technology and consumer
preferences. Production technology in each sector is modeled using nested constant elasticity
of substitution (CES) functions. At the top level, the output is produced as a combination of
aggregate intermediate demand and value added. At the second level, aggregate intermediate
demand is split into each commodity according to Leontief technology. Value added is produced
by a capital-land bundle and aggregate labor. Finally, at the bottom level, aggregate labor is
decomposed into unskilled and skill labor, and the capital-land bundle is decomposed into
capital and land (for the agriculture sector) or natural resources (for the mining sector). At each
level of production, there is a unit cost function that is dual to the CES aggregator function and
demand functions for corresponding inputs. The top-level unit cost function defines the marginal
cost of sectoral output.
Agriculture, mining and government services sectors are assumed to have perfect competition.
In each of these sectors, there is a representative firm operated under constant returns to scale
technology. Trade is modeled using Armington assumption for import demand. The
manufacturing and private services sectors are characterized by monopolistic competition, and
their structure of production and trade follows Melitz (2003). Each sector with monopolistic
competition consists of a continuum of firms which are differentiated by the varieties they
produce and their productivity. Firms face fixed production cost, resulting in increasing returns to
scale. There is also a fixed cost and a variable cost associated with the exporting activities.
On the demand side, the agents are assumed to have Dixit-Stiglitz preference over the
continuum of varieties. As each firm is a monopolist for the variety it produces, it sets the price
of its product at a constant markup over its marginal cost. A firm enters domestic or export
markets if and only if the net profit generated from its domestic sales or exports in a given
country is sufficient to cover the fixed cost. This zero cutoff profit condition defines the
productivity thresholds for firm’s entering domestic and exports markets, and in turn determines
the equilibrium distribution of non-exporting firms and exporting firms, as well as their average
productivities. Usually, the combination of a fixed export cost and a variable (iceberg) export
cost ensures that the exporting productivity threshold is higher than that for production for
domestic market, i.e., only a small fraction of firms with high productivity engages in exports
markets. These firms supply for both domestic and export markets.
Incomes generated from production accrue to a single representative household in each region.
A household maximizes utility using An Implicitly Direct Additive Demand System (AIDADS)
(Rimmer and Powell 1996). AIDADS is a demand system which allows the marginal budget
48
ADBI Working Paper 404
Petri and Zhai
shares to vary as a function of total expenditure. Empirical work by Yu et al. (2004) has
demonstrated the superiority of AIDADS over other demand systems in projecting food demand,
especially for long-term projections involving a wide range of countries. Investment demand and
government consumption are specified as a Leontief function. In each sector a composite good
defined by the Dixit-Stiglitz aggregator over domestic and imported varieties is used for final and
intermediate demand.
All commodity and factor markets are assumed to clear through price adjustment. There are five
primary factors of production. Capital, agricultural land and two types of labor (skilled and
unskilled) are fully mobile across sectors within a region. In natural resource sectors of forestry,
fishing and mining, a sector-specific factor is introduced into the production function to reflect
the resource constraints. For all primary factors, their stocks are fixed.
There are three macro closures in the model: the net government balance, the trade balance,
and the investment and savings balance. We assume that government consumption and saving
are exogenous in real terms. Any changes in the government budget are automatically
compensated by changes in income tax rates on households.
The second closure concerns the current account balance. In each region, the foreign savings
are set exogenously. With the price index of OECD manufacturing exports being chosen as the
numéraire of the model, the equilibrium of foreign account is achieved by changing the relative
price across regions, i.e., the real exchange rate.
Domestic investment is the endogenous sum of household savings, government savings and
foreign savings. As government and foreign savings are exogenous, changes in investment are
determined by changes in the levels of household saving. This closure rule corresponds to the
“neoclassical” macroeconomic closure in the CGE literature.
Emissions of CO2 have three drivers. Most are generated through consumption of goods—
either in intermediate of final demand. Some are driven by the level of factor input. And the
remainder is generated by aggregate output. The model has a flexible system of mitigation
policies. The simplest is a country or region specific carbon tax—that also allows for exemptions
for designated sectors or households. An alternative is to provide a cap on emissions at either a
country, regional or global level. The model will then produce the shadow price of carbon, i.e.,
the carbon tax, as a model outcome. If a global cap is imposed, a single uniform tax will be
calculated. This type of regime assumes no trading. A final option is to have a regional or global
cap with trading and assigned quotas. Similar to the previous regime, a uniform carbon tax will
be calculated (and would be nearly identical to the no-trade carbon tax), but emissions trading
would occur depending on the initial quotas and the shape of the individual marginal abatement
curves for each member of the trading regime.
49
ADBI Working Paper 404
Petri and Zhai
APPENDIX 3: CONSUMPTION DISTRIBUTION SIDE MODEL
The distribution side model permits the results of CGE simulations to be used to determine
impacts on the incidence of poverty and the emergence of the middle class. The side model
takes as its input the regional consumption levels generated by the CGE model. Its output
consists of the allocation of the population into four consumption classes: those that consume
less than $1.25 per day in 2005 dollars (extremely poor), those that consume between $1.25
and $10 per day (low); those that consume between $10 and $100 per day (middle), and those
with high consumption levels (more than $100 per day). These cutoffs are those used in ADB
(2011), which appears to be based on Kharas (2010). There is much variation in the literature
with respect to such groupings; some studies calculate divisions based on absolute measures
while others rely on relative measures, and there are differences in the cutoff values used to
define different groups. The $10–$100 is justified by Kharas (2010) and has the advantage of
representing absolute consumption levels that appear to be relevant across a wide range of
economies.
The distribution side model was built from the World Bank’s World Development Indicators data
on consumption shares, the source also used by most other studies. These data have been
translated into 20 percentile groups by Pinkovskiy and Sala-i-Martin (2009) for some 133
countries. These authors also report that the log-normal functional form appears to perform
better than alternatives. Therefore, population distributions by consumption level were modeled
with a log-normal function fitted to the World Bank data. The log-normal function has the
cumulative density function:
where c is per-capita consumption, μ is the mean of the log-normal distribution, σ is its standard
deviation, and Φ is the cumulative distribution function of the standard normal distribution.
Maximum likelihood estimates of the crucial standard deviation parameter σ can be easily
computed from distribution data:
where n is the number of income groups in the region (20 for a single economy). Estimates of σ
for regions that include more than one country include both intra- and inter-country variations,
and hence tend to be larger than for single countries, especially smaller ones. Due to the
apparent increase in inequality in many countries in recent observations, future values of σ were
increased by 0.5% per year (approximately 5% per decade).
An interesting feature of the log-normal distribution is that the Gini coefficient associated with a
distribution depends solely on the parameter σ:
Estimated σ, μ, and Gini coefficients are reported in Table A3.1.
While the log-normal function performs well in the middle of an income or consumption
distribution, it systematically underestimates the probabilities of being of extremely poor or
extremely prosperous. We therefore made corrections to both the left- and right-hand tails of the
distribution. We corrected the left-hand tail (the incidence of very low consumption) by
calibrating the model to actual poverty statistics from World Bank data. Specifically, we
50
ADBI Working Paper 404
Petri and Zhai
assumed that extremely poor individuals have access to k times the income predicted by the
log-normal distribution, perhaps due to transfers and non-reported income. The factor k was
found as the income level that would generate the observed poverty rate under the log-normal
function. Its estimated value hovers around 2 for emerging economies and is also reported in
Table A3.1. The factor (and hence correction) was assumed to diminish at higher consumption
levels to 1 at the low end of the middle consumption class. We corrected the right-hand tail
(incidence of very high consumption) by assuming that once the distribution reaches
consumption levels at the 95th percentile, the probability density function declines at half its
normal rate until the resulting probability density is equal to 5 times the density predicted by the
log-normal model (this occurs roughly at the 99.5th percentile).
The results of the model for the baseline are reported in absolute numbers and population
shares in Tables A3.2 and A3.3.
Table A3.1. Consumption Distribution Parameters
ASEAN
PRC
India
Latin America
Other emerging
Japan
Korea; Taipei,China
Australia, New
Zealand
Canada
United States
Europe
σ
μ
Gini
k
0.74
0.76
0.52
1.23
1.04
0.73
0.73
-0.28
-0.29
-0.13
-0.75
-0.54
-0.27
-0.27
0.40
0.41
0.29
0.61
0.54
0.39
0.39
2.51
1.49
2.74
1.27
2.55
1.00
1.00
0.73
0.68
0.82
0.73
-0.26
-0.23
-0.34
-0.27
0.39
0.37
0.44
0.39
1.00
1.00
1.00
1.00
Korea = The Republic of Korea.
Source: Authors’ calculations.
51
ADBI Working Paper 404
Petri and Zhai
Table A3.2: Population Numbers by Consumption Level
Consumption/Cap
ita
(PPP $2005)
ASEAN
PRC
India
Latin America
Other emerging
Japan
Korea; Taipei,China,
Australia, New
Zealand
Canada
United States
Europe
World (sum)
Memorandum: ACI
2010
2030
3,196
2,474
1,805
8,011
4,014
18,052
20,666
8,179
13,197
7,148
15,359
8,243
22,952
44,957
21,510
23,923
28,553
15,108
6,161
2,351
33,305
32,864
35,246
20,900
12,337
9,684
2010 population by consumption/day (mill.)
<$1.25
$1.25$10
593
1,349
1,225
534
1,860
127
71
93
124
400
34
465
0
0
328
974
759
227
772
4
2
172
251
65
251
585
111
61
0
0
0
22
38
12
9
27
34
310
511
6,641
3,167
0
0
0
0
1,116
617
1
0
6
29
3,102
2,061
22
28
230
453
2,229
488
4
6
74
30
194
1
Total
Korea = The Republic of Korea.
Source: Authors’ calculations.
52
$10$100
2030 population by consumption/day (mill.)
>$100
<$1.25
$1.25$10
706
1,402
1,523
636
2,509
120
73
16
1
16
25
319
0
0
213
220
435
214
882
4
0
454
1,110
1,053
336
1,207
97
41
22
71
20
60
101
20
32
33
40
362
528
7,933
3,632
0
0
0
0
378
34
0
0
7
20
1,996
868
23
28
242
435
5,026
2,618
10
12
113
72
533
113
Total
$10$100
>$100
ADBI Working Paper 404
Petri and Zhai
Table A3.3: Population Shares by Consumption Level
Consumption/Capita
(PPP $2005)
ASEAN
PRC
India
Latin America
Other emerging
Japan
Korea; Taipei,China
Australia, New
Zealand
Canada
United States
Europe
World (sum)
Memorandum: ACI
2010 population share by consumption/day
0.16
0.09
0.33
0.06
0.25
0.00
0.00
$1.25$10
0.55
0.72
0.62
0.43
0.42
0.03
0.02
$10$100
0.29
0.19
0.05
0.47
0.31
0.87
0.85
1.00
0.00
0.02
1.00
1.00
1.00
1.00
1.00
0.00
0.00
0.00
0.17
0.19
0.01
0.02
0.06
0.47
0.65
2010
2030
Total
3,196
2,474
1,805
8,011
4,014
18,052
20,666
8,179
13,197
7,148
15,359
8,243
22,952
44,957
1.00
1.00
1.00
1.00
1.00
1.00
1.00
21,510
33,305
23,923
28,553
15,108
6,161
2,351
32,864
35,246
20,900
12,337
9,684
<$1.25
Source: Authors’ calculations.
53
2030 population share by consumption/day
0.00
0.00
0.00
0.04
0.02
0.09
0.13
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.02
0.00
0.01
0.04
0.13
0.00
0.00
$1.25$10
0.30
0.16
0.29
0.34
0.35
0.03
0.00
0.84
0.14
1.00
0.00
0.01
0.69
0.30
0.82
0.74
0.89
0.34
0.15
0.17
0.24
0.06
0.03
0.00
1.00
1.00
1.00
1.00
1.00
0.00
0.00
0.00
0.05
0.01
0.00
0.02
0.04
0.25
0.24
0.69
0.67
0.82
0.63
0.72
0.30
0.31
0.14
0.07
0.03
>$100
Total
<$1.25
$10$100
0.64
0.79
0.69
0.53
0.48
0.81
0.55
>$100
0.03
0.05
0.01
0.09
0.04
0.16
0.44