POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise MAY 2010 · Number 14 54595 The Service Revolution Ejaz Ghani and Homi Kharas The growth experience of India and other South Asian countries suggests that a "Service Revolution"--rapid income growth, job creation, gender equality, and poverty reduction led by services--is now possible. What is a service revolution? Can services be as dynamic as manufacturing? Can latecomers to development take advantage of the globalization of services? Can services be a driver of sustained growth, job creation, and poverty reduction? What kind of policies and institutions do developing countries need to benefit from services-led growth? Both China and India have been recognized for rapid eco- ment that has held true for almost 200 years, since the start nomic growth. However, their growth patterns are dramati- of the Industrial Revolution (Chenery 1960; Kaldor 1966). cally different. China has a global reputation for exporting This law--which is now conventional wisdom--says that in- manufactured goods, and India has a reputation for exporting dustrialization is the only route to rapid economic develop- modern services. Indeed, India has sidestepped the manufac- turing sector and made the leap straight from agriculture into Figure 1. Comparing the Contributions of the Manufacturing and services. The differences in growth patterns between China Services Sectors to GDP in China, India, and OECD Countries, 2005 and India are striking and raise big questions in development economics. Can developing countries jump straight from 100 service value added (percent of GDP) HKG agriculture into services? Can services be as dynamic as man- MAC· · LUX ufacturing? Can services growth be a driver of job creation FRA · 80 USA · GBR · ITA · · and poverty reduction? ·PRT · · · ·· · ISL ·· · HUN · AUS · SGP · JPN · ·· South Asia LKA · IRL 60 · PHL · · CZE ·KOR · NOR Growth · BGD ·IND ·PAK · · NPL · THA IDN ·CHN · MYS Figure 1 compares the GDP shares of services value added 40 · VNM · East Asia in China, India, and countries of the Organisation for Eco- ·LAO 20 nomic Co-operation and Development. The share of the services sector in India and other South Asian countries is 0 much bigger than in China or other East Asian economies, 2.5 3.0 3.5 4.0 4.5 given the country's stage of development. Log real GDP per capita India's growth pattern in the 21st century is remarkable because it contradicts a seemingly iron-clad law of develop- Source: Ghani 2010. 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise ment for developing countries. It goes further to say that, as if the relationship is causal, then the effect of services growth a result of globalization, the pace of development can be ex- on aggregate economic growth seems to be stronger than is plosive. But the rule holds that the potential for explosive the effect of manufacturing growth. growth is specific to manufacturing only (UNIDO 2009). The trend over time to a higher services sector share in That is no longer the case. GDP shows that higher real growth in services has not been Figure 2a compares real GDP growth with services value- offset by price declines. There is no Dutch disease, whereby added growth in some 136 countries between 2000 and the price of a service falls with an increase in its supply (Bau- 2005. Each point on the graph represents one country. The mol 1967). India has a higher share of services and more positive relationship between the two variables implies that rapid growth in its services sector than does China, although countries with high growth in services also tend to have high the latter is richer and has grown faster. That fact suggests overall economic growth or, conversely, that countries with that services are responding not simply to domestic demand high overall economic growth have high services growth. (which would be higher in China), but also to export op- One cannot identify causality from this relationship. portunities (Ghani and Kharas 2010). Figure 2b depicts the relationship between manufactur- India's experience shows that growth has been led by ing growth and overall economic growth. Again, there is a services exports, that labor productivity levels in services are positive relationship. Comparing the two graphs, it is clear above those in industry, and that productivity growth in that the slope is steeper in figure 2a. That fact suggests that services sectors in India match labor productivity growth in manufacturing sectors in China. Furthermore, services-led growth has been effective in reducing poverty in India. Figure 2. GDP and Value-Added Growth in 136 Countries, 2000­05 a. GDP growth and services value-added growth Poverty 15 · We find that for a sample of 50 developing countries, growth · in the services sector is more closely correlated with poverty 10 · · reduction than is growth in agriculture. Visually, we can plot · · · GDP growth (%) · ·· the change in poverty between 1990 and 2005 against the · ·· · · · · · · · growth of services (figure 3). The negatively sloped line indi- 5 ·· · ··· · · · ···· · cates the strong association between services output growth · · ··· ·· · ·· ·· · · ·· ·· · ·· ·· · · ·· ·· ··· · · · · · · · and poverty reduction for our sample of developing countries. · ···· ·· ·· · ·· · · · · · · 0 · · · ·· · ·· · ·· ·· · · These results are further confirmed when we examine · · · · · · · the impact of sectoral growth patterns on poverty reduction · within India, using a panel of Indian state data from 1994 ­5 0 5 10 15 services output growth (%) Figure 3. Change in Poverty and Growth in Services Output, Selected Countries, 1990­05 b. GDP growth and manufacturing value-added growth 10 · GNB · ZMB · VEN · · · MDGBOL 10 · CIV · ZAF · 0 ·KGZ · · COL · MNG · KHM · change in poverty (%) · · · PER PRY · SLV LKA · · · · IND GDP growth (%) · ·· · · · UKR CAF · · ·ARG· · · ·BGD·· LAO · UGA 5 · ·· · · ·URY · THA ·IRN · · · ·· · ­10 · ·EGY· TUN · · · · ··· ·· · · ·· · · · · ·CRI · ETH · · · ····· ··· · · · · ··· ····· ··· · · · · ··· · · · KEN · · HND · MRT 0 · · ·· · · ·· ·· · ·· · · · · · · AZE ­20 · ········ · · ·· · · · SEN · ··· · ·· · · · IDN ­5 · ­30 CHN · ­5 0 5 10 15 ­40 manufacturing output growth (%) ­5 0 5 10 services output growth (%) Source: Ghani 2010. Note: Each point in the both charts corresponds to five-year growth for a specific country. GDP growth rates control for level of initial income per Source: World Bank, World Development Indicators. capita. All values are in constant 2000 U.S. dollars. Growth rates are Note: Illustration presents change in poverty between 1991 and 2205, after compounded annual averages. The sample consists of 136 countries. controlling for initial level of poverty. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 4. Contribution of Services Output Growth to Poverty transport, and even rural finance improved. For many sec- Reduction, Indian States, 1994­2005 tors, in fact, wages in rural areas may have increased faster 0.03 · Orissa than those in urban areas, possibly reflecting the rising rural- rate of change in poverty (%) Figure 5. Job Creation Rates in Agriculture, Industry, and Services Madhya Pradesh · Jharknand· · Bihar Sectors, Selected Countries, 1990­2006 · Maharashtra ·Uttar Pradesh · Chhattisgah 5.0 compounded growth in job creation (%) · Assam Rajasthan · · West Bengal 4.5 · Punjab Karnataka · · Gujarat 4.0 Tamil Nadu · 3.5 Kerala · · Haryana 3.0 ·Andhra Pradesh · Himachal Pradesh 2.5 ­0.5 ­.23 1.17 2.0 rate of change in service output (%) 1.5 1.0 Source: Ghani 2010. 0.5 0 to 2005. Our results show that the growth trend in the serv- Bangladesh India Pakistan Sri Lanka ices sector among Indian states is associated with a decrease country of almost 1.5 points in the trend of the head count poverty agriculture industry services rate during the sample period. In fact, the services sector is the only sector showing a statistically significant association Sources: Bangladesh: "Key Findings of Labour Force Surveys," http://www. bbs.gov.bd/dataindex/l_force_05.pdf; and various editions of the Statistical with a reduction in poverty. Similar results are found when Yearbook. India: Sivasubramonian (2004) for earlier years and Bosworth, differentiating into rural and urban poverty. Services sector Collins, and Virmani (2007) for the more recent period. Pakistan: Interna- growth is strongly associated with a reduction in both urban tional Labour Organization's LABORSTA, 2008; and State Bank of Pakistan (2005). Sri Lanka: up to 1989, International Labour Organization's Key In- and rural poverty rates. Some states--such as Andhra dicators of the Labour Market; from 1990 onward, various editions of Pradesh, Kerala, and Tamil Nadu--have experienced a sig- "Labor Survey Reports." nificant decrease in urban poverty that is associated with an increase in their services sector share of GDP (figure 4). Figure 6. Growth in Average Daily Earnings, India, 1999­ Services help lessen poverty through two channels. Di- 2000/2004­05 rectly, they provide the largest source of new job growth. In- 8 directly, they provide the income that, when spent, drives further demand for goods and services and for the jobs to 6 produce them. Figure 5 shows that the services sector in India has had the fastest growth in the number of jobs cre- 4 ated in recent years. In addition to direct job creation, some 2 percent estimates suggest that the indirect effect of a growing services sector can be larger than its direct effect. For instance, India's 0 information technology industry association (NASSCOM) estimates that for every job created in the information tech- ­2 nology sector, four additional jobs are created in the rest of ­4 the economy because of the high levels of consumption spending by professionals employed in this sector. ­6 agriculture mining manufacturing utilities construction trade transport finance Job Creation sector rural workers urban workers Services jobs are good jobs. Wage growth has been higher in Source: Kundu and Mohanan 2009. India's services sector than in manufacturing and agriculture Note: "Agriculture" includes agriculture, forestry, and fishing; "mining" in- over recent years (figure 6). Manufacturing wages fell in the cludes mining and quarrying; "utilities" includes electricity, gas, and water supply; "trade" includes trade, hotels, and restaurants; "transport" in- early 2000s in both rural and urban areas, despite rapid eco- cludes transport, storage, and communications; and "finance" includes fi- nomic growth; but services sector wages in utilities, trade, nancial intermediation, real estate, and business. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 7. Female Employment Goes Hand in Hand with Increased portable, nontradable, and nonscalable no longer hold for a Services Employment, 2005 host of modern impersonal services that are moved across · 80 borders over the Internet, digitized and stored electronically, · female population aged 15­64 (%) CHN · · SWE ·CHE and scaled into giant global businesses. Developing countries total female employment/total · ·· · USA · ETH · · FIN · ··GBR · · RUS ·PRT ··AUS can sustain services-led growth because there is enormous · AZE · · ·· · · JPN · · · · · · ROM 60 · ARG ·HKG space for catching up and convergence. BGD· · · ·PHL · ·ISR · SGP · ·GEO ·KOR · · · HUN ·ESP· ·LUX · India's development experience offers hope to develop- IDN · · ··BGR · · ITA · · · · MEX· · PAN ment latecomers. The marginalization of Africa during a pe- 40 · · IND LKA CHL · · · riod when China and other East Asian countries grew · · · · PAK · rapidly led some people to wonder if latecomers were MAR · ·TUR doomed to failure. Many economists considered the poorest 20 people--the "bottom billion"--to be trapped in poverty · SAU (Collier 2007). The process of globalization in the late-20th 10 20 30 40 50 60 century led to a strong divergence of incomes between those services employment/total population aged 15­64 (%) people in economies that industrialized and broke into global markets and the bottom billion people in some 60 Source: Ghani and Kharas 2010. countries where incomes stagnated for 20 years. It seemed as if the bottom billion would have to wait their turn for de- velopment until the giant industrializers, like China, became urban migration taking place in India. This internal rural- rich and noncompetitive in labor-intensive manufacturing. urban migration and links between rural and urban labor The promise of the Service Revolution is that countries markets allow the modern impersonal services sector in do not need to wait to get started with rapid development. India to contribute to overall poverty reduction, even There is a new boat that development latecomers can board. though modern services are concentrated in urban areas. The globalization of services provides alternative opportuni- Currently, South Asia suffers from one of the lowest fe- ties for developing countries to find niches beyond manufac- male labor force participation rates in the world. Only about turing where they can specialize, scale up, and achieve one third of all working-age women in India, Pakistan, and explosive growth--just like the industrializers. The core of Sri Lanka actually are working or looking for work. Interna- the argument is that as the services produced and traded tionally, countries with high employment in services tend to across the world expand with globalization, the possibilities have the highest participation of women in the labor market for all countries to develop on the basis of their comparative (figure 7). The development of services industries, therefore, advantages expand. Comparative advantages can just as easily brings new workers into the labor force, making the contri- be in services as in manufacturing or, indeed, in agriculture. bution to aggregate growth even larger.1 The employment of women has a special role in reducing Note poverty. Incomes of households where women have jobs are significantly higher. Higher household incomes and en- 1. In this regard, services play a role similar to labor-in- hanced economic status for women in turn reduce the num- tensive manufacturing exports in East Asia. There, garments ber of children per household, drive higher levels of and electronics assembly jobs have been filled largely by education and health care for those children, and increase women moving out of low-productivity self-employment on household savings and the ability to accumulate assets that family farms. generate additional income. About the Authors Is Services-Led Growth Sustainable? Ejaz Ghani is an economic adviser in the South Asia Poverty But can services-led growth be sustained? Yes, because the Reduction and Economic Management Network at the World current globalization of services is only the tip of the iceberg Bank; Homi Kharas is a senior fellow at the Brookings Institu- (Blinder 2006). Services is the largest sector in the world, tion and a former chief economist of the East Asia and Pacific accounting for more than 70 percent of global output. The Region at the World Bank, Washington, DC. "Service Revolution" has altered the characteristics of serv- ices. Now services can be produced and exported at low cost This note is a summary of The Service Revolution in South (Bhagwati 1984). The old ideas of services being nontrans- Asia, published by Oxford University Press. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise References Kaldor, Nicholas. 1966. Causes of the Slow Rate of Economic Growth of the United Kingdom. Cambridge, UK: Cambridge University Press. Baumol, William J. 1967. "Macroeconomics of Unbalanced Growth: The Kundu, Amitabh, and P. C. Mohanan. 2009. "Poverty and Inequality Out- Anatomy of Urban Crisis." American Economic Review 57 (3): 415­26. comes of Economic Growth in India: Focus on Employment Pattern Bhagwati, Jagdish N. 1984. "Splintering and Disembodiment of Services during the Period of Structural Adjustment." Presentation at the Or- and Developing Nations." World Economy 7 (2): 133­44. ganisation for Economic Co-operation and Development's seminar, Blinder, Alan S. 2006. "Offshoring: The Next Industrial Revolution?" Foreign "Employment Outcomes and Inequality: New Evidence, Links and Pol- Affairs 85 (2): 113­28. icy Responses in Brazil, China and India," Paris, April 8. http://www.oecd. Bosworth, Barry, Susan M. Collins, and Arvind Virmani. 2007. "Sources of org/dataoecd/60/42/42560797.pdf. Growth in the Indian Economy." Working Paper W12901. National Bu- Sivasubramonian, S. 2004. The Sources of Economic Growth in India: 1950­ reau of Economic Research, Cambridge, MA. 1 to 1999­2000. New York: Oxford University Press. Chenery, Hollis B. 1960. "Patterns of Industrial Growth." American Eco- State Bank of Pakistan. 2005. Handbook of Statistics on Pakistan Economy nomic Review 50 (4): 624­54. 2005. Karachi, Pakistan. Collier, Paul. 2007. The Bottom Billion: Why the Poorest Countries Are Failing UNIDO (United Nations Industrial Development Organization). 2009. In- and What Can Be Done About It. Oxford, UK: Oxford University Press. dustrial Development Report 2009: Breaking In and Moving Up: New In- Ghani, Ejaz, ed. 2010 The Service Revolution in South Asia. New York: Ox- dustrial Challenges for the Bottom Billion and the Middle-Income ford University Press. http://www.us.oup.com/us/catalog/general/sub Countries. Vienna, Austria. ject/Economics/Developmental/Regional/~~/dmlldz11c2EmY2k9OTc 4MDE5ODA2NTExMQ==. Ghani, Ejaz, and Homi Kharas. 2010. "The Service Revolution in South Asia: An Overview." In The Service Revolution in South Asia, ed. E. Ghani, 1­32. New York: Oxford University Press. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. It is produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at www.worldbank.org/economicpremise.