Long-Term Projections of Asian GDP and Trade

ISBN 978-92-9092-479-1 Publication Stock No. RPT114111

Long-term projections of Asian GDP and trade—November 2011 Mandaluyong City, Philippines: Asian Development Bank, 2011.

1. Macroeconomics. 2. International economics. 3. Labor and demographic economics. 4. Economic development, technological change, and growth. 5. Developing Asia. I. Asian Development Bank.

The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent.

ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use.

By making any designation of or reference to a particular territory or geographic area, or by using the term "country" in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.

ADB encourages printing or copying information exclusively for personal and noncommercial use with proper acknowledgment of ADB. Users are restricted from reselling, redistributing, or creating derivative works for commercial purposes without the express, written consent of ADB.

Photo source: ADB library.

Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines Tel: +63 2 632 4444 Fax: +63 2 636 2444 www.adb.org

Printed on recycled paper ii ║ Long-Term Projections of Asian GDP and Trade Contents

Acknowledgments ...... iv. List of Figures Executive Summary ...... v Figure 1: Sources of Gross Domestic Product Abbreviations ...... vi. Growth in Developing Asia, 1981–2007 ...... 7. Figure 2: Demographic Transition in East Asia I . Introduction ...... 1. versus South and Southeast Asia ...... 13. Figure 3: Growth of Gross Domestic Product per Capita II . Summaries and Key Findings ...... 3. in East Asia versus South and Southeast Asia . . . . 15. A . Summaries ...... 3. Figure 4: Sources of Gross Domestic Product Growth 1 . Human Capital Accumulation in Developing Asia, 2011–2030 (Baseline in Emerging Asia, 1970–2030 ...... 3. versus Reform Scenario) ...... 16. 2 . Past and Future of the Labor Force Figure 5: Average Gross Domestic Product Growth Rate in Emerging Asian Economies ...... 4. Projections: Baseline versus Reform Scenario . . . . 17. 3 . Economic Growth in Asia: Determinants Figure 6: Saving and Investment Gap of the Past 4 and Prospects ...... 4. Decades in Developing Asia (decade average) . . . 18. 4 . Projection of Long-Term Total Factor Productivity Figure 7: Domestic Saving Rate, Investment Rate, Growth for 12 Asian Economies ...... 4. and Current Account, Current and Future Trends . . . .19 5 . The Determinants and Long-Term Projections Figure 8: Cross-Country Income Dispersion Under of Saving Rates in Developing Asia ...... 5. the Baseline Scenario ...... 22. 6 . Physical Capital Accumulation in Asia-12: Figure 9: Cross-Country Income Dispersion Under Past Trends and Future Projections ...... 6. the Reform Scenario ...... 22. 7 . Asian Trade Flows: Trends, Patterns, and Prospects ...... 6. List of Tables B . Key Findings: Gross Domestic Product Table 1: Assumptions for Human Capital, Labor Force, and Trade Projections, 2011–2030 ...... 7. and Population, 2011–2030 ...... 8 1 . Gross Domestic Product ...... 7. . Table 2: Projections of Gross Domestic Product and 2 . International Trade Flows ...... 10. Gross Domestic Product per Worker in Developing Asia, Decade Averages (annual growth rates) . . . . 9. III . Selected Issues ...... 12. Table 3: Average Applied Tariff Rates in Developing A . Demographic Dividends and the Role of Asia, 1980–2006 (%) ...... 10. the Labor Force in Asia’s Growth ...... 12. Table 4: Projections for Intraregional Trade–Total B . Reforms for Efficiency-Driven Growth ...... 15. Trade Ratio in Developing Asia, 2011–2030 . . . . . 11. C . Developing Asia’s Saving and Investment Table 5: Projections for Trade Openness in Developing Gap ...... 17. . Asia, 2011–2030 ...... 11 D . Income Convergence in Developing Asia . . . . 20. 1 . β Convergence ...... 20. 2 . σ Convergence ...... 21.

References ...... 23.

Projections Overview ║ iii Acknowledgments

This report, a collection of research products, was prepared A special seminar was held at ADB headquarters on as part of a research and development technical assistance 30 September 2010, where valuable suggestions were project, Long-Term Projections of Asian Gross Domestic provided by Maria Socorro Bautista, Juzhong Zhuang, and Product (GDP) and Trade . Under the overall guidance of Joseph Ernest Zveglich, as well as by other participants . Jong-Wha Lee, former chief economist, Economics and The technical assistance project team also benefited from Research Department, Asian Development Bank (ADB), discussions at various conferences, including the Emerging the project was managed and coordinated by Akiko Terada- Asia 2030 Workshop, Tokyo, Japan, 24–25 January 2011; Hagiwara, economist, and Lea Sumulong, senior economics Asia–Pacific Economic Association Seventh Annual officer, both from Economics and Research Department, Conference, Busan, Republic of Korea, 24–25 June 2011; and ADB . International Economic Association 16th World Congress, Beijing, People’s Republic of China, 4–8 July 2011 . Key contributors to this project are Jong-Wha Lee and Ruth Francisco (Human Capital Accumulation in Emerging The final versions of the research products will be edited Asia); Jinyoung Kim (Past and Future of the Labor Force by Charles Yuji Horioka and Akiko Terada-Hagiwara to in Emerging Asian Economies); Jong-Wha Lee and be published in a special volume in Japan and the World Kiseok Hong (Economic Growth in Asia: Determinants Economy . Kiseok Hong prepared additional materials and Prospects); Jungsoo Park (Projections of Long-Term and significantly contributed to this report, Kimberly Total Factor Productivity Growth for 12 Asian Economies); Fullerton did the manuscript editing, and Ariel Paelmo Charles Yuji Horioka and Akiko Terada-Hagiwara (The was responsible for the layout and design . Technical Determinants and Long-Term Projections of Saving Rates in and research support was provided by Shiela Camingue Developing Asia); Etsuro Shioji and Tuan Khai Vu (Physical assisted by Elenita Pura and Maria Susan Torres . The Capital Accumulation in Asia 12: Past Trends and Future publication would not have been possible without the Projections); and Prema-chandra Athukorala (Asian Trade cooperation of the Department of External Relations and Flows: Trends, Patterns and Prospects) . the Office of Administrative Services, ADB .

Earlier versions of these research products were presented at an inception workshop held at the Bank of Korea, Seoul, Republic of Korea, on 1–2 March 2010, and at a finalization workshop held at the Chinese University of Hong Kong, Hong Kong, China, on 8–9 July 2010 . Jessie Pang provided administrative and technical support in organizing the finalization workshop . The following experts provided valuable comments during the two workshops: Shin-ichi Fukuda, Joonkyung Ha, Carsten Holz, Hyun Jeong Kim, Yong Jin Kim, Edward Leung, Pak-Wai Liu, Francis Lui, Weh-Sol Moon, Kwanho Shin, Yun-Wing Sung, Dennis Tao Yang, and Hongliang Zhang .

iv ║ Long-Term Projections of Asian GDP and Trade Executive Summary

In 2011, uncertainty has cast a dark shadow over global institutions—that can enhance economies’ productive economic prospects . Economic setbacks that are holding capacities and promote the efficiency-driven growth are back recoveries in many developed economies have been highlighted . Such reforms are found to be particularly expected to spill over to developing Asia . However, as effective in economies where these initial levels are the Asian Development Outlook 2011 Update confirms, relatively low, that is, where school enrollment is low, the impact to the region’s economies has been less research and development support is weak, and property severe . Developing Asia continued to exhibit steady rights protection is poor . growth, underpinned by ample fiscal space, low debt, and strong domestic demand . To sustain its growth Reforms are necessary not only for low-income economies, momentum, developing Asia must build on these gains and but also for those that have achieved substantial growth turn toward its unfinished growth agenda . already, such as in Asia’s newly industrialized economies . Reforms that facilitate a shift from input-driven growth Long-term growth in the region depends largely on supply- to efficiency-driven growth will be in increasing demand side factors, augmented by each economy’s productive as population aging progresses and the speed of factor capacity . Such growth reflects the combined effects of accumulation decelerates . production factors, such as capital and labor, as well as productivity improvements—not on the short-term business While these reforms already pose substantial challenges cycle . This distinction matters, as policies for sustaining to the region, the research products under the technical growth are different from policies for minimizing cyclical assistance project go even farther . From the seven fluctuations . The former involves structural supply-side papers produced, the report also discusses four crucial strategies that enable individuals, firms, industries, and the developmental issues: (i) impacts of demographic entire economy to become more productive on a sustained transitions and the labor force, (ii) policy reforms needed basis; the latter involves fiscal and monetary actions for to shift to efficiency-driven growth, (iii) the saving and influencing the level of aggregate demand only temporarily . investment gap in developing Asia, and (iv) convergence in income levels and income growth . In sum, the demographic The regional technical assistance project on Long-Term transition of the region—along with the efficiency Projections of Asian Gross Domestic Product (GDP) and enhancing reforms—is crucial in affecting growth dynamics Trade, provides careful analyses and projections of the of the region . supply-side factors of 12 economies in the region, as well as their GDP and trade growth, for the next 2 decades . This As demographic dividends that characterized developing report details the following key findings: (i) the vital driver Asia’s newly industrialized economies’ rapid growth fade, of the region’s robust growth during the past 3 decades the report concludes that it is urgent to increase the labor was primarily capital accumulation with the contribution participation of females and the elderly to moderate the of productivity improvements being relatively limited, decline in the region’s labor force . Yet, the anticipated (ii) GDP growth rates will gradually slow except for in a relatively strong income growth of the region can offset few economies during the next 2 decades as a convergence the negative aging effects on the saving rate to a certain phenomenon kicks in, and (iii) the surge in intraregional extent . The saving and investment gap of developing Asia trade of the past decades will continue but at a decelerated is projected to remain positive . Trade projects a similar rate . story, and a consistent picture can also be drawn from the projections for exports and imports . Central to these findings is the need for and urgency of growth friendly reforms in developing Asia . While the report confirms that poorer economies tend to grow faster, it also verifies that factors other than initial income are at work, generating income differences across economies . As a result, in the next 2 decades, per worker income is not expected to narrow in the region . Three possible areas—human capital, research and development, and

Projections Overview ║ v AbbreviationsAbbreviations

ADB – Asian Development Bank ASEAN – Association of Southeast Asian Nations GDP – gross domestic product PRC – People’s Republic of China TFP – total factor productivity

vi ║ Long-Term Projections of Asian GDP and Trade I. Introduction

he economies of Asia and the exposed to internal and external shocks Pacific have shown robust, that could affect their growth path and sustained growth during the 3 progress toward prosperity, equality, decades since 1980, with the and poverty reduction . The export-led region’sT real gross domestic product growth model followed by several Asian (GDP) in purchasing power parity terms countries is under reconsideration, as increasing more than sevenfold during the region’s economies are starting to 1980–2009, as compared to only threefold realize the limits of excessive dependency in the world economy as a whole . The on exports and the need to rebalance region’s real per capita GDP increased growth by increasing internal demand . more than fourfold, as compared to less More balanced growth will help solve than twofold in the world economy . This the problem of global imbalances and growth has raised living standards, lifted decrease the region’s dependency on final millions of persons out of poverty, and goods demand from the United States expanded the region’s global economic and Europe, although intraregional trade influence . is still too concentrated on intermediate products included in supply chains of While the economies of Asia and manufactured goods destined for non- the Pacific have shown considerable Asian markets . resilience to the global financial crisis of 2008–2009, they continue to be

Summaries and Key Findings ║ 1 Against this backdrop, the region’s project future trends in GDP and trade for The 12 economic interdependence is expanding the next 20 years (i .e ., until 2030), based both in the real and financial sectors, on the findings of (i) . developing and business cycles are becoming increasingly synchronized across the The 12 developing economies are the economies region .14 At the same time, following the People’s Republic of China; Hong Kong, are the recent adoption of its new long-term China; India; Indonesia; the Republic of strategic framework, Strategy 2020, Korea; Malaysia; Pakistan; the Philippines; People’s the Asian Development Bank (ADB) Singapore; Taipei,China; Thailand; and is moving toward expanding its role Viet Nam .36 These economies comprise Republic as a knowledge bank, with increasing 95% of the GDP of Asia and the Pacific . of China; advisory functions for member countries, Seven research products were prepared especially on critical economic issues and on GDP as a whole, saving rate, physical Hong Kong, policies . Thus, it has become increasingly capital accumulation, total factor important for ADB to have its own model productivity, labor force, human capital China; India; for long-term projections, especially of accumulation, and international trade Indonesia; the region’s GDP and trade . Such models flows . will help the region better understand the Republic the long-term trends of its economies and This report also discusses selected topics assist policy makers in coming up with based on the findings of the technical of Korea; strategic decisions that affect a wide assistance project . Four issues that have Malaysia; range of economic policies . important policy implications for the region’s long-term growth were identified Pakistan; the This report summarizes the findings of and will be discussed in the second half a research and development technical of this report . They are (i) impacts of Philippines; assistance project, Long-Term Projections demographic transitions and the labor Singapore; of Asian GDP and Trade, conducted from force, (ii) policy reforms needed to shift to 2009 to 2011 2. 5 This technical assistance efficiency-driven growth, (iii) the saving Taipei,China; has two objectives: (i) to analyze the and investment gap in developing Asia, determinants of each component of GDP and (iv) convergence in income levels and Thailand; and in 12 developing economies in the region income growth . Viet Nam. during the past 40 years; and (ii) to

1Evidence from an ADB study (2008) and other works show that Asia and the Pacific’s economic cooperation and integration has been rising rapidly during the last decade . 2ADB . 2009 . Technical Assistance for Long-Term Projections of Asian GDP and Trade. Manila (TA 7470-REG, for $225,000, approved 23 December 2009) . 3In this report, they are referred to as “developing Asia .”

2 ║ Long-Term Projections of Asian GDP and Trade II. Summaries and Key Findings

A. Summaries 1. Human Capital Accumulation in Emerging Asia, 1970–20304 Jong-Wha Lee and Ruth Francisco even research products were produced to analyze GDP, its Developing Asia has made rapid progress components (i .e ., saving rate, in raising the level of its educational physical capital accumulation, capital during the past 40 years, due in totalS factor productivity [TFP], labor force, large part to higher school enrollment and human capital accumulation), and rates, especially at the secondary level . international trade flow . In the following Lee and Francisco (2010) conducted a subsection, each paper is summarized . regression analysis of the determinants of school enrollment and found that better parental education and income, lower income inequality, declining fertility, and higher public educational expenditures contributed to the increase in school enrollment .

4Available at http://www .adb .org/documents/Working- Papers/2010/Economics-WP216 pdf.

Summaries and Key Findings ║ 3 The authors then projected future trends 3. economic Growth in Asia: Park (2010) in years of schooling based on the Determinants and Prospects6 results of the regression analysis and Jong-Wha Lee and Kiseok Hong support an found that average years of schooling in accumula- developing Asia are projected to increase Using a growth accounting framework, only from 7 .0 years to 7 .6 years by 2030, Lee and Hong (2010) found that tionist’s significantly slower than the 4 .1-year developing Asia grew rapidly over the increase during 1970–2010 . That would past 3 decades, mainly due to robust view during put the region’s educational capital in growth in capital accumulation . The the pre- 2030 at only the 1970 level of advanced contributions of education and TFP in countries such as the United States or the region’s past economic growth were 2000 period Japan, or still 3 .5 years behind the level relatively limited in the past decades . of advanced countries in 2010 . To sustain but show human development, the region needs to The authors then made long-term evidence of invest in improving educational quality growth projections for developing Asia and raising school enrollment rates at the by combining the growth accounting a growing secondary and tertiary levels . framework with a growth regression approach . Their baseline projections, contribution 2. Past and Future of the Labor based on the model of conditional of total Force in Emerging Asian convergence, showed that the growth Economies5 rates of GDP of the 12 developing factor Jinyoung Kim economies for the next 3 decades will be consistently lower than their historical productivity Kim (2010) investigated the determinants performance . However, policy reforms in during the of past changes in the labor force of the education, property rights, and research 12 developing Asian economies, and and development can substantially raise post-2000 made projections of the labor force in GDP growth in the region and partly those economies for 2011–2030 . Results offset the slowdown in growth caused by period. from the regression analysis indicate the convergence phenomenon . Yet even that the labor force has grown faster under the baseline scenario, the region’s than the population, has an inverted share in the world economy will increase U-shape relationship with per capita GDP, from 34% in 2009 to close to 50% in 2030 . and is smaller in more capital-intensive countries . 4. Projection of Long-Term Total Factor Productivity Growth for Using extrapolation, the author predicted 12 Asian Economies7 that the labor force will increase in all Jungsoo Park 12 economies during the first decade of 2011–2030 but will eventually decline Park (2010) examined past trends in the People’s Republic of China (PRC); and patterns in the growth of TFP in Hong Kong, China; Republic of Korea; developing Asia during the past 4 Singapore; and Taipei,China . The author decades . Growth accounting exercises, also projected the unemployment rate focused on the contribution of TFP and average working hours in those growth, were performed for the 12 economies . developing economies . The results support an accumulationist’s view during the pre-2000 period but show evidence of a growing contribution of TFP during the post-2000 period . The author

6Available at http://www .adb .org/documents/Working- Papers/2010/Economics-WP220 .pdf 5Available at http://www .adb org/documents/Working-. 7Available at http://www .adb .org/documents/Working- Papers/2010/Economics-WP218 .pdf Papers/2010/Economics-WP227 .pdf

4 ║ Long-Term Projections of Asian GDP and Trade The domestic saving rate in developing Asia as a whole will remain elevated in the next 2 decades, mainly estimated an empirical model of TFP showed that enormous variation exists driven growth augmented with human capital among the 12 developing economies in and research and development using a their domestic saving rates, with the by rising country-level panel dataset, and found nominal domestic saving rate ranging income that the major source of TFP growth of from 39 .8% in Singapore to 11 .2% in developing Asia in the early years was a Pakistan during 1966–2007 as a whole . levels, which catch-up effect driven by its low base . In fact, Viet Nam showed negative real domestic saving rates until it transitioned are only However, the author also found that to a market economy in the 1990s . partially the role of human capital in raising TFP growth has been gradually increasing The authors then conducted an offset for Hong Kong, China; the Republic of econometric analysis of the determinants Korea; Singapore; and Taipei,China, but of domestic saving rates in the 12 by aging has stagnated or weakened for other developing economies during 1966–2007, effects. developing Asia economies . Strong and found that the main determinants of growth in TFP is projected for the next the domestic saving rate in developing 2 decades, supporting the idea that Asia during this period appear to be the developing Asia will increasingly shift age structure of the population, especially to efficiency-driven growth from input- the aged dependency ratio, income levels, driven growth . and level of financial sector development . Their projections suggest that the 5. The Determinants and Long- domestic saving rate in developing Asia Term Projections of Saving as a whole will remain elevated in the Rates in Developing Asia8 next 2 decades, mainly driven by rising Charles Yuji Horioka and income levels, which are only partially Akiko Terada-Hagiwara offset by aging effects .

Against the common perception of developing Asia’s high saving rates, Horioka and Terada-Hagiwara (2010)

8Available at http://www .adb org/documents/Working-. Papers/2010/Economics-WP228 pdf.

Summaries and Key Findings ║ 5 and high investment goods prices, worked against it .

Projections for the next 2 decades suggest that the pace of capital accumulation will decline relative to its current level if no reforms are made to enhance potential growth rates . The authors asserted that if developing Asia wishes to maintain its current rapid pace of capital accumulation, the keys are to reduce distortions in the domestic market, as reflected in the relative price of investment goods, and to improve the quality of institutions, as reflected in the rule of law index .

7. Asian Trade Flows: Trends, Patterns, and Prospects10 Prema-Chandra Athukorala

Over the past 4 decades, the merchandise trade of developing Asia has grown at a much faster rate than growth in world trade, with a distinct intraregional bias . Global production sharing has become a unique feature of the region’s economic landscape, with the PRC playing a pivotal role as the premier assembly center within regional production networks . 6. Physical Capital Accumulation in Asia-12: Past Trends and Athukorala (2011b) estimated a standard Future Projections9 gravity model of trade flows for each Etsuro Shioji and Tuan Khai Vu of the 12 developing economies, which suggest that the two standard gravity Shioji and Khai Vu (2011) investigated the variables, GDP in pairs and geographic sources of rapid capital accumulation in distance, as well as the Logistics the region and projected future trends . Performance Index, an indicator of the The authors estimated a convergence quality of trade-related logistics, perform equation for physical capital per capita, remarkably well in explaining both which was derived from an open imports and exports . The projections economy growth model, using a pooled of trade flows based on the estimation cross-country, across-decade sample results find that the region’s total real for 1977–2007 . The authors found that nonoil trade will increase at an annual an economy with a low per capita rate of 8 .2% during the next 2 decades, capital stock tends to experience faster with a notable convergence of individual subsequent accumulation . Developing countries’ rates to the regional average . Asia certainly benefited from this convergence effect; however, other factors, such as a low rule of law score

9Available at http://www .adb org/documents/Working-. 10Available at http://www .adb .org/documents/Working- Papers/2011/Economics-WP240 .pdf Papers/2011/Economics-WP241 .pdf

6 ║ Long-Term Projections of Asian GDP and Trade Figure 1: Sources of Gross Domestic Product Growth in Developing Asia, 1981–2007 In general, Percentage points the 10 Capital Education baseline 8 Labor Total factor productivity forecasts of 6 annual GDP 4 growth for 2 2011-2030

0 are lower PRCHKG INDINO KORMAL PAKPHI SINTAP THAVIE than the HKG = Hong Kong, China; IND = India; INO = Indonesia; KOR = Republic of Korea; MAL = Malaysia; PAK = Pakistan; PHI = Philippines; PRC = People’s Republic of China; SIN = Singapore; TAP = Taipei,China; THA = Thailand; VIE = Viet Nam. Source: Lee and Hong (2010). historical averages in

B. Key Findings: Gross contributions of human capital and TFP 1981–2007. Domestic Product and growth to GDP growth during the same Trade Projections, period were relatively limited (Figure 1) . 2011–2030 A system of equations was set up to estimate the parameters of the growth model described above .13 The results 1. gross Domestic Product11 were then used to generate projections of GDP growth for 2011–2030 for each of the Using a growth accounting framework, 12 developing economies . In generating the sources of developing Asia’s GDP GDP projections for the technical growth in 1981–2007 were deconstructed assistance project, demographic profiles into growth in the different factors of were considered exogenously given . production—labor, human capital, and Population projections were sourced from physical capital—as well as in TFP 12. The the United Nations and are assumed to result shows that Asia’s rapid economic grow according to the medium-variant growth in nearly 3 decades has been fertility rate assumption of the United mainly due to robust growth in physical Nations . Labor force and human capital capital accumulation . In 10 out of the 12 are also considered exogenous to the GDP, economies studied, average growth in and their projections for the next 20 years physical capital stock was more than 5 .8% are generated in Kim (2010) and Lee and per year between 1981 and 2007, which Francisco (2010) (Table 1) . contributed more than 2 .3 percentage points to average GDP growth . The In general, the resulting baseline forecasts of annual GDP growth are lower than the historical averages in 11This section draws heavily from the material prepared for 1981–2007 (5 .5% versus 9 .3% for the PRC, Box 2 .1 .2 of Asian Development Outlook 2010 Update based and 4 .5% versus 5 .5% for India, for on Lee and Hong (2010) . example, Table 2) . Only the projections for 12It is assumed that the share of physical capital is 0 .4 and that the labor share is 0 .6 . Growth in physical capital must Pakistan and the Philippines are higher thus be multiplied by 0 .4 to get its contribution to GDP than the historical averages (by 1 .4 and growth . The contribution of labor and human capital is similarly calculated, multiplying their growth rates by the 0 .6 labor share . TFP is calculated by subtracting the contributions of capital, labor, and human capital from GDP growth . 13See Lee and Hong (2010) for details .

Summaries and Key Findings ║ 7 Table 1: Assumptions for Human Capital, Labor Force, and Population, 2011–2030

Projections Projections Economy 2011–2020 2021–2030 Economy 2011–2020 2021–2030 Schooling (mean years) China, People's Republic of 8.8 9.0 Pakistan 67.0 65.4 Hong Kong, China 11.0 11.7 Philippines 68.6 73.4 India 5.7 6.1 Singapore 66.7 55.7 Indonesia 6.6 7.0 Taipei,China 73.5 68.2 Korea, Republic of 12.1 12.5 Thailand 68.2 65.2 Malaysia 10.6 11.0 Viet Nam 71.6 70.3 Pakistan 6.4 7.1 Population growth (%)a Philippines 9.3 9.7 China, People's Republic of 0.4 0.1 Singapore 9.7 10.3 Hong Kong, China 1.1 1.0 Taipei,China 11.8 12.2 India 1.3 1.0 Thailand 8.2 8.9 Indonesia 0.9 0.7 Viet Nam 6.8 7.0 Korea, Republic of 0.5 0.3 Labor force participation rate (%) Malaysia 1.5 1.3 China, People's Republic of 82.9 83.4 Pakistan 1.7 1.3 Hong Kong, China 71.1 71.9 Philippines 1.6 1.4 India 62.4 64.1 Singapore 1.1 0.9 Indonesia 75.1 78.0 Taipei,China 0.1 0.0 Korea, Republic of 68.5 70.8 Thailand 0.5 0.3 Malaysia 71.9 78.0 Viet Nam 1.0 0.6 Pakistan 68.6 71.1 Population (million persons)a Philippines 66.5 77.4 China, People's Republic of 1,393.4 1,426.1 Singapore 83.1 94.3 Hong Kong, China 7.8 8.6 Taipei,China 64.6 65.6 India 1,326.2 1,480.2 Thailand 77.5 78.1 Indonesia 255.2 276.3 Viet Nam 77.0 77.0 Korea, Republic of 50.6 52.5 Working age population rate (%) Malaysia 31.2 35.8 China, People's Republic of 71.6 69.4 Pakistan 192.4 223.6 Hong Kong, China 70.2 60.2 Philippines 102.8 119.8 India 66.7 67.4 Singapore 5.5 6.0 Indonesia 67.4 67.3 Taipei,China 23.4 23.4 Korea, Republic of 65.8 66.1 Thailand 72.4 75.0 Malaysia 70.3 62.0 Viet Nam 94.1 101.2 aForecast figures are based on medium-variant projections. Sources: Lee and Francisco (2010); Kim (2010); United Nations (2010). World Population Prospects, the 2010 Revision. http://esa.un.org; and http://eng.dgbas.gov

2 .3 percentage points, respectively) . In the next 2 decades, labor force These results are largely driven by a growth is expected to decrease in most slowdown in labor force growth caused economies, mainly due to fertility decline by a demographic transition and the and population aging; Pakistan is the cross-country catch-up phenomenon only country where labor force growth where the lower-income economies of the is expected to increase in the next 20 region are expected to grow faster than years .14 The contribution of human capital the higher-income economies .

14See Section III on demographic dividends .

8 ║ Long-Term Projections of Asian GDP and Trade accumulation is also forecast to decline as Table 2: Projections of Gross Domestic Product and Gross Domestic Product per Worker in Developing Asia, Decade Averages economies move closer to best practice (annual growth rates) school enrollment and completion rates . Actual Projections Growth in capital stock is projected to Economy 1981–2007 2011–2020 2021–2030 contribute less to GDP growth in 2011– Gross domestic product 2030 than it did in 1981–2007 . This is China, People’s Republic of 9.4 6.1 5.0 because capital was the major source of Hong Kong, China 5.0 4.4 2.8 growth in the past, and as the marginal productivity of capital declines, the India 5.5 4.7 4.3 contribution of growth in capital stock to Indonesia 4.8 4.7 4.1 GDP growth tends to fall . Pakistan and the Korea, Republic of 6.3 4.4 3.4 Philippines are the exceptions, since they Malaysia 6.2 5.5 4.8 started off with lower growth in capital Pakistan 5.1 6.8 6.3 stock . These results suggest that, while Philippines 3.4 6.0 5.5 labor and physical capital have been a major source of growth in developing Asia Singapore 6.8 5.2 3.3 in the past, the same factors will turn Taipei,China 6.1 3.7 2.5 into a main cause of growth slowdown in Thailand 5.4 4.0 3.6 the future . Human capital growth is also Viet Nam 6.4 4.9 3.7 expected to decrease in every economy Gross domestic product per worker except for Viet Nam . However, due to its small magnitude, the growth rate China, People’s Republic of 7.7 5.8 5.1 of human capital makes only a limited Hong Kong, China 3.4 3.9 3.4 contribution to overall GDP growth in India 3.3 2.9 2.9 growth accounting . Indonesia 2.2 3.2 3.1 Korea, Republic of 4.5 4.0 3.7 In contrast to these factors, TFP is Malaysia 3.1 3.6 3.4 expected to maintain a relatively stable growth rate on average between the Pakistan 1.8 2.7 2.8 periods of 1981–2007 and 2011–2030 . Philippines 0.7 3.8 3.7 As a result, the relative contribution Singapore 3.9 3.7 3.3 of TFP growth in growth accounting Taipei,China 4.4 3.8 3.4 will become greater . On average, the Thailand 3.6 3.9 3.6 contribution of TFP growth remains Viet Nam 4.0 3.7 3.4 roughly the same magnitude in the next Source: Lee and Hong (2010). 2 decades; thus, its relative share in GDP growth should be increasing . In fact, it is only in the PRC and Singapore that These results generally support the the relative contribution of TFP growth conditional convergence theory . Faster does not increase between the periods rates of factor accumulation and of 1981–2007 and 2011–2030 . The PRC technology diffusion narrow the gap with may be an exception because many potential income, leaving little room to suspect that its historical TFP growth grow further . However, convergence does rate is overestimated .15 For Singapore, not imply that as Asia becomes more the relative contribution of TFP growth prosperous, it cannot keep growing fast remains almost constant in the two in the future . Institutional and policy periods at about 29% . variables can expand the potential income of an economy, widening the gap with the initial income level and allowing 15For example, Young (2003) argued that, after correcting for the achievement of a higher growth for measurement errors in deflators and labor inputs, TFP path . The potential impacts of policy growth in the nonagricultural economy of the PRC during 1978–1998 was only 1 .4% per year . Similarly, Woo (1998) reforms on GDP growth will be discussed reported TFP growth of 1 1%–1. .3% per year for 1979–1993 . in Section III .

Summaries and Key Findings ║ 9 2005 throughout the region . It is widely recognized that trade liberalization and the rapid trade expansion have played a key role in developing Asia’s economic growth . One of the most prominent patterns in developing Asia is the significance of intraregional trade, which has been growing much faster than total trade in the region . In fact, the rapid growth of total trade is to a large extent due to the growth of intraregional trade .

The fast growth of intraregional trade has been driven by global production sharing or vertical specialization and the increasingly deep integration of developing Asia, particularly the PRC, into the global production networks (Athukorala 2011a; Ng and Yeats 2003) . Vertical specialization, that is, the production of goods using imported Table 3: Average Applied Tariff Rates in Developing Asia, 1980–2006 inputs, has grown rapidly worldwide (%) since 1960s . Yet the significance of this 1980– 1985– 1990– 1995– 2005– new form of international specialization Economy 1984 1989 1994 1999 2006 has been particularly sizable in Asia . China, People’s Republic of 49.5 39.3 40.0 18.8 12.8 Global production sharing has changed India 74.3 93.5 57.0 33.7 16.8 the region’s economic landscape, with Indonesia … 13.7 13.4 6.4 8.5 the PRC playing a key role as the premier Korea, Republic of … 17.5 9.7 9.3 8.0 assembly center in the region and other Malaysia … 14.9 14.3 6.9 7.6 Asian economies specializing in the production of intermediate goods . A Pakistan … 66.7 58.5 41.6 13.1 notable outcome of this development Philippines 29.3 27.8 23.7 13.3 5.9 has been the rapid growth of cross- Singapore … 0.5 0.4 0.3 0.2 border trade within the region and the Taipei,China 26.5 16.8 12.5 8.4 5.5 advancement of the PRC as the region’s export hub . As the region’s export hub, Thailand 41.2 40.3 37.2 19.6 8.9 the country is importing more from other Viet Nam … … 13.4 13.7 14.4 Asian economies and exporting more to … = data not available. outside of the region . This runs counter Note: Simple averages of most-favored-nation rates. Source: Athukorala (2011b). to the popular perception that PRC’s global integration crowds out other countries’ opportunities for international specialization . 2. international Trade Flows Another important, related pattern is the Developing Asia has become increasingly sharp increase in the share of parts and open to foreign trade over the past 4 components in total trade . The increased decades . The average applied tariff rate trade in parts and components is driven among the 12 developing economies by the need for vertical specialization . decreased from about 30% in the late In PRC, the share of parts and 1980s to about 9% in 2005 (Table 3) . components in its total manufacturing Also, total trade as a percentage of GDP imports (44 .0%) is much larger than more than doubled between 1970 and the corresponding share in its exports

10 ║ Long-Term Projections of Asian GDP and Trade Table 4: Projections for Intraregional Table 5: Projections for Trade Openness in Trade–Total Trade Ratio in Developing Asia, Developing Asia, 2011–2030 2011–2030 Intra- 2011– 2021– 2011– 2021– Economy 2020 2030 regional Economy 2020 2030 China, People’s Republic of 49.6 70.5 Exports share in Hong Kong, China 162.8 184.7 China, People’s Republic of 54.3 58.5 India 8.6 11.1 total exports Hong Kong, China 68.4 70.7 Indonesia 44.7 63.8 and imports India 61.0 65.6 Korea, Republic of 50.3 65.3 Indonesia 75.7 87.7 to expand Malaysia 129.5 171.5 Korea, Republic of 62.7 67.2 Pakistan 18.7 23.0 further Malaysia 72.9 76.5 Philippines 152.1 231.9 albeit at Pakistan 45.8 49.1 Singapore 175.9 234.1 Philippines 71.8 76.8 a slower Taipei,China 56.4 75.0 Singapore 82.3 85.0 Thailand 73.8 95.7 speed. Taipei,China 60.1 63.4 Viet Nam 69.0 91.6 Thailand 73.9 77.1 Developing Asia 46.4 64.6 Viet Nam 67.1 71.3 Note: Trade openness = (exports + imports)/2 as percentage of gross domestic product. Imports Source: Athukorala (2011b). China, People’s Republic of 44.6 48.5

Hong Kong, China 90.7 92.6 Using standard gravity models estimated India 42.2 46.2 for exports and imports, Athukorala Indonesia 58.5 69.2 (2011b) extended the historical patterns Korea, Republic of 52.5 56.9 and trends in the series and produces forecasts . Projections are for total nonoil Malaysia 62.1 66.4 exports and imports from the region’s Pakistan 57.2 62.0 developing countries to increase at an Philippines 69.9 73.8 annual rate of 8 .5% and 7 .8%, respectively, Singapore 70.5 74.2 during 2011–2020 and 2021–2030, exhibiting a mild slowdown in the Taipei,China 63.0 68.2 rate of growth over time . The growth Thailand 68.7 72.2 of intraregional trade will be about 1 .2 Viet Nam 81.7 84.0 percentage points faster, resulting in Source: Athukorala (2011b). an increase in the intraregional share in total exports and imports (Table 4) . As a result, trade openness in developing (25 .6%) . This difference is consistent with Asia is project to average 46% and 65%, the production sharing in the region, respectively for 2011–2020 and 2011–2030 where Association of Southeast Asian Table 5) . Trajectories of the intraregional Nations (ASEAN) countries produce trade, however, will heavily depend on parts and components for final assembly how the regional production network activities in the PRC . Using the parts evolves in tandem with the increasing and components imported from ASEAN purchasing power of the region and countries, the PRC produces final goods exchange rate adjustments globally . that are exported to the rest of the world . In spite of its intrinsic comparative advantage, India still remains a minor player in this new form of international exchange (Krueger 2010) .

Summaries and Key Findings ║ 11 III. Selected Issues

A. Demographic Dividends in labor growth . Also, per capita GDP growth is expected to decrease by about 1 and the Role of the percentage point, even though per worker Labor Force in Asia’s GDP growth remains roughly constant, Growth because the ratio of the labor force to the total population decreases . This suggests that, while developing Asia so far has emographic changes have clear, been benefiting from the high growth of significant impacts on GDP its labor force, the demographic dividend growth . The growth accounting will gradually run out in the future . analysis undertaken by Lee and DHong (2010) shows that changes in the Across individual economies, however, labor force account for a large part of the the pace of the demographic change changes in GDP growth in developing will be diverse . In economies where Asia . For example, the average GDP the demographic dividend was started growth rate among developing economies earlier, such as Hong Kong, China; in the region is expected to decrease by the Republic of Korea; Singapore; and 1 .3 percentage points from 1981–2007 to Taipei,China, labor force growth is 2011–2030, and about 48% of the decline forecast not only to slow down but can be attributed to the slowdown even to turn negative during the next 2

12 ║ Long-Term Projections of Asian GDP and Trade decades . On the other hand, countries The pace of demographic transition in that entered the demographic transition East Asia has been one of the fastest in later, such as India, Indonesia, Malaysia, history . The transition from high fertility Pakistan, and the Philippines, will be and high mortality to low fertility and able to enjoy favorable demographic low mortality took more than a century in conditions throughout 2011–2030 . Europe but only several decades in East Asia . The rapid pace of transition implies The slowdown in labor force growth that the age structure of the population is caused mainly by mortality and and the growth rate of labor force should fertility decline as well as population manifest accordingly as large fluctuations aging . A fall in infant mortality, which in East Asia . typically precedes fertility decline by a few decades in each country, initially Changes in the age structure and the increases the labor force when more labor force may affect economic growth infants grow to working age . Also, a through many channels . First, an increase decline in fertility, which more than in labor force growth will directly offsets the effect of the preceding drop contribute to GDP growth . Second, an in mortality, leads to a decrease in young increase in the ratio of the labor force to dependents and thus an increase in the the total population can have a positive proportion of working-age people and the labor force in the total population . However, as the working-age cohort Figure 2: Demographic Transition in East Asia versus grows older, the proportion of the elderly South and Southeast Asia will increase, and the initial increase in the labor force will be reversed . Ratio A. Working age–total population ratio 80 The self-generated swing in the age structure of the population means that 60 a period of high labor force growth will necessarily be followed by a period of low labor force growth . The labor force 40 projection of Kim (2010) confirms these dynamics . According to his forecasts, 20 the labor force in the People’s Republic of China (PRC); Hong Kong, China; 0 the Republic of Korea; Singapore; and 1980 1990 2000 2010 2020 2030 Taipei,China will reach its peak around 2020 and then start to decrease . In Ratio B. Labor force–total population ratio contrast, other countries, including India, 0.6 Indonesia, Pakistan, and the Philippines, are projected to maintain a steady increase in their labor forces until 2030 . 0.5 The difference between the two groups of economies in the movement of the labor force is related with the difference 0.4 in the age structure of the population . As Figure 2 shows, the first group of economies (i e. ., East Asia) experienced 0.3 an early fall in their fertility rates and 1980 1990 2000 2010 2020 2030 thus had a higher initial value of the East Asia South and Southeast Asia working age to the total population in Note: Working age refers to ages 15–64 years. 1980 . Sources: Kim (2010) and author’s calculations.

Selected Issues ║ 13 When the labor force or working-age population grows faster than the total population, per capita GDP is positively effect on per capita GDP growth . This can analysis by Horioka and Terada-Hagiwara affected. be seen as follows: (2010) on the saving rate supports the prediction . They showed that the saving Labor force GDP per capita = GDP per labor * rate is negatively correlated with both Population the youth dependency ratio and the DIn (GDP DIn (GDP Labor force elderly dependency ratio . According to = + DIn ( ) per capita) per labor) Population their results, the demographic transition in Asia has contributed to capital The above equation shows that, given accumulation as well to raising the saving the growth rate of GDP per worker, an rate . An increase in the working-age increase in the labor force–population population may boost investment demand ratio will result in an increase in the at the same time by raising the marginal growth rate of GDP per capita . When the product of capital . Also, low fertility may labor force or working-age population lead to more educational spending per grows faster than the total population, child and thus greater human capital per per capita GDP is positively affected . worker . Bloom and Williamson (1998) argued that this demographic dividend has The demographic dividend, however, is characterized East Asia’s rapid growth only transitional . In time, the working-age since the 1970s . cohort will move into an elderly cohort, and the same forces that generated the Demographic factors can also indirectly demographic dividend will now work contribute to growth through changes to slow economic growth . As Figure 2 in saving and investment . An increase shows, the labor force as percentages of in saving and investment leads to an the total population is expected to start increase in the growth rate of GDP decreasing around 2020 among East per worker in the above equation, by Asian economies, while continuing to generating greater capital accumulation . increase among the other economies . The standard life-cycle hypothesis The average of the technical assistance predicts that the aggregate saving rate project’s 12 developing economies is will be systematically influenced by expected to peak around 2025 and then to the age structure of the population . The decrease slightly .

14 ║ Long-Term Projections of Asian GDP and Trade The difference between the two groups Figure 3: Growth of Gross Domestic Product per Capita in of economies with respect to the East Asia versus South and Southeast Asia movements of labor force and saving % rates translate differently in per capita 0.07 GDP growth . Driven by the stagnant growth in labor force participation rates, the per capita GDP growth rate will 0.05 exhibit a substantial fall in East Asian economies but a modest increase in South and Southeast Asian economies during 0.03 the next 2 decades (Figure 3) .

Economies in different stages of 0.01 1981–1990 2001–2010 2021–2030 demographic transition need different policy responses . In economies with a East Asia South and Southeast Asia rapidly aging population, it is urgent to Source: Lee and Hong (2010). increase the labor participation of females and the elderly to moderate the decline in the labor force . For relatively young economies, it may be more important to The growth-accounting analysis provided try to reap the demographic dividend in the technical assistance project while it lasts . The demographic dividend helps reveal from where the slowdown is not automatic, and the magnitude of originates, as discussed in Section II . The the demographic dividend will depend decomposition of the sources of growth on the economy’s ability to use the extra shows that the GDP growth rate will slow labor force . Latin American economies mainly due to a substantial fall in the experienced a similarly rapid increase in growth rate of the labor force and that of the working-age population around the physical capital stock . same period as East Asian economies, yet their economic performance was not as The observation that input-driven remarkable . Relatively young economies growth cannot be sustained signifies the such as those in South and Southeast importance of efficiency-driven growth Asia should be able to benefit from the via human capital accumulation and experiences of their aging predecessors total factor productivity (TFP) growth . in East Asia and Latin America . In principle, under the Cobb-Douglas representation of the production function B. Reforms for Efficiency- assumed in the technical assistance project, human capital is regarded as a Driven Growth factor input that contributes to production One of the most prominent features of the in the same way as physical capital baseline GDP forecasts by Lee and Hong and labor . Also, the growth-accounting (2010) is the slowdown in GDP growth . analyses in Lee and Hong (2010) show In almost all of the 12 economies covered, that the contribution by human capital is except for Pakistan and the Philippines, small and decreasing . However, this does the annual GDP growth rate forecast for not necessarily imply that human capital 2011–2030 is lower than the historical is irrelevant for efficiency-driven growth . average over 1981–2007 . For example, First, the measure of human capital forecasts for PRC GDP growth are 6 .1% used in the majority of the literature for 2011–2020 and 5 .0% for 2021–2030, is a function simply of the number of substantially lower than the historical schooling years and thus does not average of 9 4%. . Also, in all of the 12 incorporate quality of education . When developing economies, the GDP growth quality of education is taken into account, rate is projected to decrease between the the slowdown in human capital growth periods of 2011–2020 and 2021–2030 . may be reduced or even reversed .

Selected Issues ║ 15 Second, and more importantly, the Figure 4: Sources of Gross Domestic Product Growth in level of human capital may positively Developing Asia, 2011–2030 (Baseline versus Reform Scenarios) affect TFP growth by increasing the Percentage points A. Highest Impact research and development capacity 10 of the economy as in the endogenous growth theory (Romer 1990; Aghion 8 and Howitt 1998) or by facilitating the adoption and implementation of new 6 technologies (Benhabib and Spiegel 1994) . This suggests that human capital 4 may contribute to production in more comprehensive ways than assumed in the 2 standard growth-accounting analysis . 0 BR BR BR BR In both the neoclassical and the endogenous growth models, the ultimate Viet Nam PakistanIndonesia India source of sustained growth is TFP growth . TFP growth also may exhibit a slowdown Percentage points B. Medium Impact as developing economies catch up with 8 developed economies through technology imitation and learning . However, as 6 mentioned above, the average TFP growth rate in the region does not show a 4 substantial change between the periods of 1981–2007 and 2011–2030, suggesting that convergence may not be the 2 dominant force that drives the movement of TFP in the region . Thus, developing 0 Asia can partly offset the slowdown in BR BR BR BR growth by maintaining high TFP growth . People’s Philippines Republic Thailand Malaysia of China According to endogenous growth theories, TFP growth may be promoted Percentage points C. Lowest Impact through various policy measures and 6 institutional changes . For example, in the endogenous growth model of Aghion 4 and Howitt (2005), productive research and development, high education, and 2 protection of innovation rents against imitation all promote innovations and thus increase the productivity growth rate . 0 This suggests that policy reforms for a good educational system, property right (2) protection, and support for research and BR BR BR BR development investment should be given Hong Kong, Republic of high priority . China Korea Singapore Taipei,China Capital Labor Education Total factor productivity The reform scenario in Lee and Hong ( ) = negative (2010) represents this view . In particular, B = 2011–2030 baseline projection. R = 2011–2030 reform scenario projection. they considered simultaneous changes Source: Lee and Hong (2010). in the three measures: (i) the growth rate of research and development stock per worker in each economy will increase

16 ║ Long-Term Projections of Asian GDP and Trade from the historical average by 0 .02 * (log per worker research and development level of the United States – log per worker research and development level of each economy) as long as the level difference is positive; (ii) property rights of every economy reach the level of the United States by 2020 and remain constant afterwards; and (iii) school enrollment rates in each economy will grow logistically over time to converge to the best prevailing enrollment rates among the developing economies by 2030 . Under this scenario, the GDP growth rate in each economy was shown to increase by about 1 percentage point on average, with greater gains accruing to less developed economies with low initial values of policy variables (Figures 4 and 5) . This result suggests that, through the policy reforms, developing Asia can substantially increase its growth performance and partly offset Figure 5: Average Gross Domestic Product Growth Rate the slowdown caused by demographic Projections: Baseline versus Reform Scenario changes and the convergence phenomenon . % 8 To be sure, the list of policy variables relevant for sustained growth is not limited to the three measures mentioned 6 above . Openness, governance, entry costs to new firms, and other institutional 4 factors may be equally important for productivity growth . Depending on country-specific conditions, different 2 policy reforms may be needed . Also, a policy measure may interact with the effect of other variables on economic 0 growth . For example, Murphy, Shleifer, 1981–19901991–2000 2001–2007 2011–2020 2021–2030 and Vishny (1991) reported that countries Baseline scenario Reform scenario with more engineering students have higher growth than countries with Source: Lee and Hong (2010). more law students . Education may not contribute much to growth if countries lack the institutional environment that C. Developing Asia’s Saving promotes entrepreneurship against rent- seeking behavior . Due to the interaction and Investment Gap effects, simultaneous reforms in various measures may have accelerating impacts Global imbalances, measured as the sum on economic growth if undertaken in absolute terms of the current account properly . positions of all economies as a rate of the world GDP, fell by nearly one-half in the aftermath of the global financial crisis after reaching a high of over 5% of world

Selected Issues ║ 17 Figure 6: Saving and Investment Gap of the Past 4 Decades in Developing Asia (decade average)

% China, People's Rep. of Hong Kong, China India Indonesia 30 20 10 0 (10) (20) 1970 1980 19902000 1970 1980 1990 2000 1970 1980 1990 2000 1970 1980 1990 2000

% Korea, Rep. of Malaysia Pakistan Philippines 30 20 10 0 (10) (20) 1970 1980 19902000 1970 1980 1990 2000 1970 1980 1990 2000197021980 1990 000

% Singapore Taipei,China Thailand Viet Nam 30 20 10 0 (10) (20) 1970 1980 19902000 1970 1980 1990 2000 1970 1980 1990 2000197021980 1990 000

( ) = negative. Sources: Heston, Summers, and Aten (2009); and Horioka and Terada-Hagiwara (2010).

GDP in 2008 . The narrowing was largely and Singapore, while others such as India, due to the decline in oil prices that led to Pakistan, the Philippines, and Viet Nam a halving of the current account surplus always borrowed from abroad to finance of Middle Eastern economies and the their domestic investments (Figure 6) . rise in saving in the United States . Asia’s contributions to the narrowing trend have Population aging is projected to occur been limited, however, given its robust at a rapid rate in developing Asia as growth relative to the rest of the world . discussed in the previous subsection, which will presumably lead to a sharp Given the still significant contribution decline in saving rates . If so, the large, of the region’s current account surplus, current account surpluses that exist or surplus saving over investment, to in some economies will go away by the global imbalance, trajectories of themselves . However, if other factors, its saving and investment rates are of such as culture, market imperfections, great interest . Over the past decades, or a per capita income upsurge, are the saving and investment gap is found the dominant determinants of saving to be persistently positive in the PRC; rates, it is possible that saving rates will Hong Kong, China; Indonesia; Malaysia; remain high .

18 ║ Long-Term Projections of Asian GDP and Trade As discussed in the previous section, 20-year projections based on Horioka and Terada-Hagiwara (2010) suggest that the domestic saving rates in developing Asia will decrease in rapidly aging economies such as Hong Kong, China; the Republic of Korea; Singapore; and Taipei,China . On the other hand, in economies where aging of the population is not as imminent as those in East Asia (i .e ., the PRC), Indonesia, India, Malaysia, Pakistan, and the Philippines, the domestic saving rates are projected to increase moderately or to remain at a similar level of the recent past .

Meanwhile, Shioji and Khai Vu (2011) investigated and projected the capital accumulation and investment ratios for the next 20 years for the region . According to the study, capital accumulation is expected to accelerate through improvement in TFP while contributions from a catch-up effect and labor force increase will gradually fade . The study also confirmed the variation across the economies . For example, the PRC and India, where capital accumulation is still low, can expect a catch-up effect, say, relative to Singapore . Meanwhile, factors such as low TFP and higher relative price of investment goods would decelerate capital accumulation . Figure 7: Domestic Saving Rate, Investment Rate, and Current The study concluded that developing Account, Current and Future Trends Asia will continue with the per capita % Percentage points capital accumulation over the next 20 years but at a slower speed in some 45 6 economies such as the PRC, the Republic of Korea, and Viet Nam . It also projected a 30 4 decrease in the domestic investment rate for the PRC and the Republic of Korea, 15 2 which together comprise more than 60% of GDP in the region, and a moderate 0 0 downward trend throughout the region . Given that the saving rate is expected (15) (2) to remain at a similar level of the recent past, it suggested that a positive saving (30) (4) and investment gap or a current account 1986–1990 1991–1995 1996–2000 2001–2007 2011–2020 2021–2030 surplus of a similar or larger magnitude Saving–Investment gap (right scale) Saving late Investment rate will still remain in the next 2 decades (Figure 7) . Trade-side projections in ( ) = negative. Athukorala (2011b) also provided a Sources: Horioka and Terada-Hagiwara (2010); Shioji and Khai Vu (2011). consistent picture with persistent trade

Selected Issues ║ 19 surpluses projected over the next 2 75% on average between 2009 and 2030 . The decades . However, this remarkable growth will not be equally experienced by the 12 conventional The trajectory of the saving and economies in the region . An important 2% speed of investment gap in developing Asia as a question regarding the growth differential whole appears to be heavily influenced across economies is the issue of conditional by the trend in the PRC, which will convergence in income levels . Are poorer account for more than 50% of regional countries projected to grow faster than convergence GDP in the next 2 decades . Any new wealthier countries, with other things is embedded policy developments, such as increasing being equal? Will the dispersion of income expenditures on social services and levels across countries actually diminish in the GDP pensions and a reduced corporate saving over time? If so, what will be the speed of rate affecting the domestic saving rate convergence? The results of this project projection, in the PRC, are of great importance to suggest that, while the conventional yet the reduce the saving rate in the region as 2% speed of conditional convergence is a whole . Also, investment rates will embedded in the projection, the cross- cross- increase rather than decrease if the prices sectional dispersion of income levels may of capital goods fall, TFP improves, and/ remain . sectional or a structural formation leads to a larger dispersion industry sector . 1. β Convergence of income D. Income Convergence in The most commonly used definition of convergence, β convergence, measures levels may Developing Asia the relationship between the initial remain. income level and the subsequent income Long-term GDP projections portray a growth . A negative correlation between promising future for developing Asia . the two implies that, with other things According to the baseline projection, being equal, poorer countries will tend the region’s share in the world economy to grow faster . A conventional estimate will grow from 34% in 2009 to about 50% of β convergence is 0 .02; a doubling up in 2030 . Also, per capita GDP in each of the initial income level is associated country is expected to increase by about with a 2-percentage point decrease in

20 ║ Long-Term Projections of Asian GDP and Trade the growth rate (Barro and Sala-i-Martin results obtained from regression analysis 1995) . into the standard growth-accounting To narrow the framework . Thus, their projections can be Historical performance of the technical regarded as representing the conventional cross-country assistance project’s 12 developing view on income convergence . However, economies covered for this project seems the β convergence embedded in Lee and income gap to be consistent with the convergence Hong’s projection does not necessarily or foster hypothesis . First, even without imply that the cross-country income gap σ controlling for possible differences in will exhibit a decreasing trend over time convergence, the steady-state income levels across in their forecasts, as explained below . economies, there appears to be a developing negative relationship between the initial 2. σ Convergence Asia needs income level and subsequent income growth . For example, the PRC, which An alternative type of convergence, to exert its had the lowest initial level of GDP per σ convergence, measures the cross- worker, grew the fastest from 1981 to sectional standard deviation of GDP per efforts to 2007 . On the other hand, Singapore and capita (or per worker) among countries . catch up with Hong Kong, China, with significantly If poor countries actually catch up with higher GDP per worker in 1980, registered rich countries, the dispersion of income best practices relatively lower rates of growth . This levels across countries will decrease over unconditional convergence is more time . As is widely noted (Barro and Sala- in institutional apparent in capital stock per worker and i-Matin 1995; Young, Higgins, and Levy and human TFP, as shown in Figure 2 of Lee and 2008), β convergence and σ convergence Hong (2010) . are related but not equal—β convergence capital is a necessary condition for σ convergence Second, the multivariate regression but not vice versa . This can be illustrated development. results of Lee and Hong (2010) support by the following example of a stylized conditional convergence . While there is growth regression: no separate equation for GDP growth in

Lee and Hong’s model, Table 2 of their ΔInyi,t = –β Inyi,t–1 + εi,t

paper shows that for each growth of per Inyi,t = (1 – β) Inyi,t–1 + εi,t 2 worker capital stock and TFP growth, the var(Inyt) = (1 – β) var(Inyt–1) + var(εt) coefficient on the corresponding initial level is statistically significant and close where time subscript i,t denotes the

to –0 .02 . If GDP growth is given by the ith observation in period t, and var(xt) sum of capital stock growth and TFP denotes the cross-sectional variance of growth, the implied speed of convergence x in period t . The last equation shows for GDP growth is also close to 2% .16 4 that the cross-sectional variance of income level can unanimously decrease over time only if β convergence holds, Lee and Hong made GDP projections by that is, 0 < β < 1 . It is also clear that, explicitly incorporating the convergence even with 0 < β < 1, income variance may either decrease, increase, or stay

16Suppose that capital stock per labor and TFP each exhibits constant as long as var(εt) ≠ 0. In other β convergence of 2% and that GDP is given by a Cobb- words, even when there is β convergence, Douglas production function: σ convergence may well be absent . Cross-

ss ss sectional income dispersion is likely to ΔlnAt=0 .02[ln(A )–ln(At–1)], Δlnkt=0 .02[ln(k )–ln(kt–1)]

lnyt=lnA1+αlnki decrease over time when β is large and var(ε) is small .175 By combining the above equations, one can obtain the following: 17This can be seen in the following equation . If var(ε ) is lny = lnA + lnk =0 02[ln(. Ass) t Δ t Δ t αΔ t nonzero and constant, the following holds: ss –ln(At–1)]+0 .02α[ln(k )–ln(kt–1)] ss ss > < var(e ) =0 .02 [ln(A )+αln(k )– ln(A )–αln(k )] t t–1 t–1 Dvar(lny ) if and only if var(lny ) ss t+1 t 2 =0 .02[ln(y )–ln(yt–1)] < > 1 – (1 – b)

Selected Issues ║ 21 Figure 8: Cross-Country Income Dispersion Under the Baseline Scenario The movement of the historical standard A. log (GDP per worker) deviation of GDP per worker among the TA project’s 12 developing economies 0.89 can be understood within this context . The theoretical distinction between 0.87 β convergence and σ convergence implies that it is entirely possible for the cross- sectional income dispersion not to exhibit 0.86 monotonic dynamics .

0.84 In the next 2 decades, GDP projections 2010 2015 2020 2025 2030 by Lee and Hong provide a mixed picture B. log (GDP per capita) regarding income dispersion in the 0.94 region . The income dispersion (GDP per worker) exhibits an increasing trend over 0.92 time, suggesting that the cross-country income gap may actually increase in 0.90 the future despite the β convergence (Figure 8A) .618 In contrast, the standard 0.88 deviation of GDP per capita exhibits a 0.86 decreasing trend over time for most of 2010 2015 2020 2025 2030 the projection period (Figure 8B) . As Year described previously, the difference Source: Lee and Hong (2010). between GDP per worker and GDP per capita is the labor force–population ratio . Figure 9: Cross-Country Income Dispersion Under Since the labor force–population ratio the Reform Scenario A. log (GDP per worker) is projected to rapidly converge among economies, GDP per capita may exhibit 0.84 σ convergence while GDP per worker does not . 0.82 However, under the reform scenario 0.80 of Lee and Hong, the cross-country income gap actually decreases over 0.78 time for both GDP per worker and GDP per capita (Figure 9A and 9B) . This 0.76 result is not entirely surprising because 2010 2015 2020 2025 2030 the reform scenario considers cross- B. log (GDP per capita) country convergence in research and 0.95 development stock per worker, property rights, and school enrollment rates . To 0.90 narrow the cross-country income gap or σ convergence, developing Asia needs 0.85 to exert its efforts to catch up with best practices in institutional and human 0.80 capital development .

0.75 2010 2015 2020 2025 2030 18While the residual term is assumed to be zero, thus var(ε ) Year t = 0 in the projection, there are other explanatory variables Note: Income dispersion is computed as the standard deviation of log (GDP per worker) or log (GDP per than initial income that have nonzero cross-sectional capita). variances . Therefore, Figure 8 is not inconsistent with the Source: Lee and Hong (2010). β convergence embodied in the projection .

22 ║ Long-Term Projections of Asian GDP and Trade References

Aghion, P ., and P . Howitt . 1998 . Endogenous Growth Theory . Kim, J . 2010 . Past and Future of the Labor Force in Cambridge, MA: MIT Press . Emerging Asian Economies . ADB Economics Working ––––––––––– . 2005 . Growth with Quality-Improving Paper Series. No . 218 . Manila: ADB . Innovations: An Integrated Framework . In Aghion, P ., Krueger, A . 2010 . India’s Trade with the World: Retrospect and S . Durlauf, eds . Handbook of Economic Growth. and Prospect . In Acharya, S ., and R . Mohan, eds . Amsterdam: North-Holland Elsevier Publishers . pp . India’s Economy: Performance and Challenges. New 67–110 . Delhi: Oxford University Press . pp . 399–429 . Asian Development Bank (ADB) . 2008 . Emerging Asian Lee, J . W ., and R . Francisco . 2010 . Human Capital Regionalism: A Partnership for Shared Prosperity . Accumulation in Emerging Asia, 1970–2030 . ADB Manila . Economics Working Paper Series. No . 216 . Manila: ––––––––––– . 2009 . Technical Assistance for Long-Term ADB . Projections of Asian GDP and Trade . Manila . Lee, J . W ., and K . Hong . 2010 . Economic Growth in Asia: ––––––––––– . 2010 . Asian Development Outlook 2010: Determinants and Prospects . ADB Economics Macroeconomic Management beyond the Crisis. Working Paper Series. No . 220 . Manila: ADB . Manila . Murphy, K ., A . Shleifer, and R . Vishny . 1991 . The Allocation ––––––––––– . 2011 . Asian Development Outlook 2011 Update: of Talent: Implications for Growth . The Quarterly Preparing for Demographic Transition. Manila . Journal of Economics . 106(2) . pp . 503–530 . Athukorala, P . 2011a . Production Networks and Ng, F ., and A . Yeats . 2003 . Major Trade Trends in East Asia: Trade Patterns in East Asia: Regionalization or What are their Implications for Regional Cooperation Globalization? Asian Economic Papers. 10(1) . pp . and Growth? Policy Research Working Paper Series. 65–95 . No . 3084 . Washington, DC: World Bank . ––––––––––– . 2011b . Asian Trade Flows: Trends, Patterns Park, J . 2010 . Projection of Long-Term Total Factor and Projections . ADB Economics Working Paper Productivity Growth for 12 Asian Economies: The Series. No . 241. Manila: ADB . Past and the Future . ADB Economics Working Paper Barro, R ,. and X . Sala-i-Martin . 1995 . Economic Growth . Series. No . 227 . Manila: ADB . New York: McGraw-Hill . Shioji, E ., and T . Khai Vu . 2011 . Physical Capital Benhabib, J ., and M . Spiegel . 1994 . The Role of Human Accumulation in Asia-12: Past Trends and Future Capital in Economic Development: Evidence from Projections . ADB Economics Working Paper Series. Aggregate Cross-Country Data . Journal of Monetary No . 240 . Manila: ADB . Economics. 34(2) . pp . 143–173 . United Nations, Department of Economics and Social Bloom, D ., and J . Williamson . 1998 . Demographic Affairs, Population Division . World Population Transitions and Economic Miracles in Emerging Asia . Prospects: The 2010 Revision . http://esa .un .org/ World Bank Economic Review. 12 . pp . 419–455 . Woo, W . 1998 . Chinese Economic Growth: Sources and Heston, A ., R . Summers, and B . Aten . 2009 . Penn Prospects . In Fouquin, M ., and F . Lemoine, eds . The World Table Version 6 .3 . Center for International Chinese Economy. London: Economica . Comparisons of Production, Income and Prices at the World Bank . World Development Indicators . http://data . University of Pennsylvania . http://pwt .econ .upenn . worldbank .org/data-catalog/world-development- edu/php_site/pwt_index .php indicators Horioka, C ., and A . Terada-Hagiwara . 2010 . The World Trade Organization, International Trade Centre, Determinants and Long-Term Projections of Saving and United Nations Conference on Trade and Rates in Developing Asia . ADB Economics Working Development . 2007 . World Tariff Profiles 2006 . Paper Series. No . 228 . Manila: ADB . Geneva: World Trade Organization .

Summaries and Key Findings ║ 23 Young, A . 2003 . Gold into Base Metals: Productivity Growth in the People’s Republic of China during the Reform Period . Journal of Political Economy . 111 (Dec) . pp . 1,220–1,261 . Young, A ,. M . Higgins, and D . Levy . 2008 . Sigma Convergence versus Beta Convergence: Evidence from U .S . County-Level Data . Journal of Money, Credit and Banking. 40(5) . pp . 1,083–1,093 .

24 ║ Long-Term Projections of Asian GDP and Trade Summaries and Key Findings ║ 25 Long-Term Projections of Asian GDP and Trade

Long-Term growth in the region depends largely on supply-side factors, augmented by each economy’s productive capacity. Such growth reflects the combined effects of production factors, such as capital and labor, as well as productivity improvements—not on the short-term business cycle. This distinction matters, as policies for sustaining growth are different from policies for minimizing cyclical fluctuations. This report details the key findings of the Asian Development Bank's project on Long-Term Projections of Asian Gross Domestic Product (GDP) and Trade.

About the Asian Development Bank

ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains home to two-thirds of the world’s poor: 1.8 billion people who live on less than $2 a day, with 903 million struggling on less than $2 a day. ADB is committed to reducing poverty through inclusive economic growth, environmentally sustainable growth, and regional integration.

Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance.

Asian Development Bank ISBN 978-92-9092-479-1 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org ISBN 978-92-9092-479-1 Publication Stock No. RPT114111 9789290 924791 Printed on recycled paper. Printed in the Philippines

26 ║ Long-Term Projections of Asian GDP and Trade