POVERTY 73578 THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise JUN 2012 NOVEMBER 010 • Numbe 96 • Number 18 Service with a Smile Ejaz Ghani, Arti Grover Goswami, and Homi Kharas Can service be a growth escalator? The world is experiencing its third industrial revolution, and services are at the forefront of this revolution. Services have already surpassed industry as a source of economic growth and job creation, in both developed and developing economies. In the industrial sector, technologies have matured and employment is shrink- ing. However, services are getting more sophisticated and jobs are expanding. Services growth is also more inclusive and sustainable. It increases the participation of women in the labor force and places a lighter burden on natural resources. The promise of the services revolution is that countries do not need to wait to get started with rapid development. There is a new boat that development latecomers can take. Fundamentals and industry to gross domestic product (GDP) growth in the last 30 years for developed and developing countries. In both For more than 200 years, it was argued that economic devel- cases, the contribution of services to total growth is higher opment and growth were associated with growth of the man- than industry’s contribution. In developing countries, servic- ufacturing sector (Baumol 1967; Kaldor 1966). Services were es (and industry) contributed more to growth than in devel- considered as menial, low skilled, and not very innovative oped countries. The average growth of services exports from (McCredie and Bubner 2010). But today, services can be developing countries has exceeded that of developed coun- among the most dynamic sectors. The range of services that tries (Ghani 2010). Their services exports are growing faster can be digitized and traded globally is exploding—processing than goods exports. In brief, the globalization of services has insurance claims; call centers; desktop publishing; compiling enabled developing countries to tap into a new, dynamic audits; completing tax returns; transcribing medical records; source of growth. providing online education; and many more. Labor market What about jobs? Although the conventional wisdom is matching is increasingly conducted online, and platforms like that the manufacturing sector creates more jobs, recent data Odesk can connect employers and employees electronically suggest otherwise. Employment growth has been most rapid across national boundaries. The old idea of services being in the services sector in developed and developing economies nontransportable, nontradable, and nonscalable no longer (figure 1b). And in developed and developing countries alike, holds. Services can be unbundled and splintered in a value labor is being shed from both agriculture and manufacturing. chain just like goods, and exported at low cost (Bhagwati The services sector is also more gender inclusive. Internation- 1984). ally, countries with high employment in services tend to have But can developing countries benefit from the services the highest participation of women in the labor market (fig- revolution? Figure 1 compares the contribution of services ure 1c). 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1a. Sources of Growth in Developed and Developing Figure 1b. Sources of Job Creation in Developed and Developing Countries (1980–2009) Countries (1991–2006) 3.5 growth contribution (%) average employment growth (%) 3.0 2.5 4 2.0 3 1.5 2 1 1.0 0 0.5 -1 0 -2 developed low developing -3 services industry -4 Source: Authors’ calculations from World Development Indicators, World Bank. -5 developed developing service industry agriculture Source: Authors’ calculations from World Development Indicators, World Bank. Figure 1c. Service Growth Is More Gender Inclusive total female employment / total female 80 population 15–64 (%, 2005) China Switzerland Sweden Ethiopia Finland UK USA Portugal Australia Azerbaijan Russian Fed. Japan Macao 60 Romania Ireland Hong Kong, SAR Korea, Rep. Argentina Bangladesh Georgia Philippines Israel Singapore Luxembourg Indonesia Bulgaria Hungary Spain Panama Italy Mexico 40 Chile India Sri Lanka Pakistan Morocco Turkey 20 Saudi Arabia 10 20 30 40 50 60 service employment / total population 15–64 (%, 2005) Source: Ghani (2010). Are these good jobs? Figure 2 shows that the rise in ser- oped countries. Two things stand out: first, the tradability of vices’ contribution to growth is linked to a rise in productivity services has risen over time for both developing and devel- growth in that sector. Labor productivity growth in developed oped countries.1 Second, this ratio is higher for developing countries has been higher in services than in industry, and it countries. The figure also shows that a growing fraction of remains positive. That implies that the global technology total services trade is now accounted for by modern services frontier for services is still shifting out, while industry has (Ghani 2010). For developed countries, the share of modern stagnated. At the same time, productivity growth in develop- services in total service exports is now around 55 percent, ing countries in services is accelerating and appears to have while for poor countries the figure is 40 percent. Developing- outstripped productivity growth in industry. In 58 out of 94 country services exports are no longer just tourism, but are countries for which data are available, productivity growth in increasingly linked to modern services exported through the services exceeded that in industry. Internet. Figure 3 plots the difference between average labor pro- Export and Growth: PRODYs across ductivity growth in services and industry against the log of Sectors and Time real per capita incomes. There appears to be no tendency for this differential to be associated with per capita income levels. Economic growth can result from specialization and trade, Developing countries like Ethiopia, Moldova, and Lesotho are but the size of the growth dividend from specialization de- just as likely to excel in services, compared to industry, as de- pends on the technological sophistication of exports (Dalum, veloped economies like Hong Kong SAR, China. Laursen, and Verspagen 1999; Feenstra and Rose 2000). Em- Figure 4a plots services exports as a ratio of services val- pirically, an advanced export structure, higher productivity ue added on the left vertical axis for developing and devel- levels, and faster growth rates are linked. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 2. Comparing Labor Productivity across Services and Figure 4a. Tradability of Services Is Higher for Developing Manufacturing Sectors Countries (1990–2007) 10 services exports to VA (%) services tradability low developing productivity growth (%) services, low developing 12 5 10 services, developed developed 0 8 industry, developed industry, low developing 6 -5 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Source: Author’s calculation from World Development Indicators, World Bank. Source: Authors’ calculation based on World Development Indicators, World Bank. Note: Labor productivity is calculated as the sector value added per employee. Line shown is the best-fit quadratic function. Figure 4b. Modern Services Exported through Internet Are Increasing from Both Developing and Developed Countries Figure 3. Differences in Labor Productivity Growth across Services (1990–2007) and Manufacturing Sectors (1990-2009) aggregate services exports (%) modern services exports to modern services tradability average productivity growth difference: 55 services and industry, 1991–2009 10 50 developed ETH LSO MDA 45 BGR HKG 40 KGZ 5 35 UZB BLZ JAMBWA RUSCYP DZA COL MUS ECA GBR low developing MNG PHL SURMYS CHL EST GRC UKR URY SAUDEU PN AUS NOR NLD LUX 30 DOM SYR TJK BRA LVA PR VEN UMC N EMU ESPDNK SGP NC HND EGY PAN ZAF LAC ROM B L USA CHE 0 MAR BO PRY IBD TURMEX RL AUT MKD DN LKA PAK GEO ARG HRV LTU 1990 1995 2000 2005 2010 NIC SLVPER CRI ECU KAZ POLKOR SVN EAP CHN ALB THA HUN YEM SVKCZE FIN SWE TTO VNM JOR Source: Authors’ calculation from World Development Indicators, World Bank. 5 ARM AZE 6 7 8 9 10 11 sophisticated over time; (ii) modern services are the most so- log of real per capita income, 1990 phisticated component of trade; and (iii) developing coun- Source: Authors’ calculation from World Development Indicators, World Bank. tries now have higher RCA in modern services exports than do developed countries. Hausmann, Hwang, and Rodrik (2007) have introduced a method that gauges export sophistication in quantitative Measuring Sophistication: PRODY terms and have shown that their measure is positively corre- The starting point is the PRODY measure of sophistication lated with economic growth across countries. This metric, developed by Hausmann, Hwang, and Rodrik (2007). For- called PRODY, is defined as a weighted average of the per mally, the PRODY value for a product j is defined as capita incomes of all countries exporting a particular prod- N uct, with the weights given by the revealed comparative ad- PRODY j = ∑ s i , j . y i vantage (RCA) of the exporting country. PRODY is therefore i =1 an alternative to the engineering approach to technological where yi stands for the real per capita GDP of the i - th (i = 1, sophistication (Lall 2000) that tries to assign each sector to a 2, ..., N) country exporting in sector j, while the weight technological category based on its content. RCAi , j Hausmann and others only focused on trade in goods, a si, j = N significant omission considering that services are the fastest ∑ RCA i =1 i, j growing component of international trade and global growth, and services are the most important and possibly normalizes country i’s Balassa index of RCA with respect most sophisticated products being traded today. Recently, to those of all the countries exporting in the same sector. Mishra, Lundstrom, and Anand (2011) have extended the By construction, products with high values of PRODY analysis to cover services. Like earlier studies, their analysis are those where high-income countries play a major role in also suggests that services export sophistication is strongly world exports of that product. If one assumes that high-in- associated with growth in per capita income. This analysis come/high-wage countries primarily export products with goes deeper and shows that (i) services are becoming more significant practical knowledge or technological content, the 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise N N PRODY index is bound to be correlated with the unobserv- PRODYt 2 − PRODYt1 = ∑ s i , j ,t 2 .( y i ,t 2 − y i ,t1 ) + ∑ y i ,t1 (s i , j ,t 2 − s i , j ,t1 .) able degree of technological sophistication of the product. i =1 i =1 This formula is applied to the change in PRODY for The intuition is simple: products like airplanes are highly so- three sectors, modern services, traditional services and phisticated and so are only produced in rich countries. These goods, and then the change between 1987–89 and 1997– countries also have a high degree of RCA in the export of air- 99 is compared with the change from 1997–99 to 2007–9 planes, and so the PRODY for airplanes is high. Over time, the (figure 6). PRODY can grow because the income levels of the main ex- The decomposition exercise shows the following trends. porting countries grow, or because the RCA (the weights) of • First, all products have experienced a higher degree of so- richer countries grows. phistication over time as the per capita incomes of coun- The database for services exports comes from the bal- tries have grown. ance of payments (rather than customs data, as is the case for • Second, the RCA term is negative for five of the six calcu- goods) and is disaggregated by sector rather than product. lations, implying that developed countries are losing These sectors are divided into two aggregates: modern ser- their comparative advantage and developing countries vices and traditional services. Modern services are found in are becoming more important exporters. information and communication technology (ICT), busi- • Third, the exception to this trend is for modern servic- ness and finance, and other commercial services. Traditional es exports in this century. Despite the tremendous services typically require more face-to-face interaction: for growth in modern services exports from developing example, sectors such as government and community ser- countries, exemplified by business-process outsourc- vices, transport, trade, hotel, restaurant, and beauty shops ing, there has been even more rapid growth in modern and barbers. services exports in developed countries (relative to In 1990, modern services had a PRODY that was 10 per- their goods exports). cent higher than traditional services, but about 8 percent The PRODY decomposition assumes that the number of lower than that of goods. By 2007, the PRODY for modern countries reporting/participating in a sector’s exports re- services was 70 percent higher than for traditional services, mains the same throughout all periods. Since this does not and 40 percent higher than for goods (figure 5). All sectors hold in this case, the decomposition is inexact. To check ro- had increased the degree of technological sophistication over bustness, the analysis also applied the decomposition with a this period. balanced panel and for a different time period (change in PRODY from 1995–97 to 2001–3 and from 2001–3 to What Explains the Rising PRODY 2007–9) with minimal change in results. in Services? Conclusion A simple decomposition exercise explains the forces behind the rise in PRODY. Recall that the PRODY is calculated as the The pace of growth in developing countries has increased. But product of two variables, an exporting-country income level the potential for explosive growth was usually only seen in the and the weight of that country as given by its RCA. Accord- manufacturing sector: this is no longer the case. Countries ingly, the PRODY can change over time either because the weights shift for each exporting country (the RCA gets more Figure 6. PRODY Change Decomposition marked) or because the income levels of the exporting coun- Change in PRODY (in thousands) try rise. The decomposition of the change over time of the 7 PRODY is given by the following identity: 6 5 2199.551 4 Figure 5. Growing Sophistication of Modern Services, 1990–2007 3 2893 2 2828 2098 PRODY (in thousands) 20 1 1270 1398 modern services 0 15 -1 -4065.589 -377.3706 -810.9497 -725.2556 -744.3112 -1005.834 goods -2 10 modern traditional goods modern traditional goods traditional services services services services services 5 - 1987–89 to 1997–99 1997–99 to 2007–9 1990 1995 2000 2005 2010 sectors and time periods Source: Authors’ calculations from IMF balance of payments. per capita income term change RCA term change Note: Modern services are calculated by taking out travel and transport services exports from commercial services exports. Source: Authors’ calculations from IMF balance of payments. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise with high growth in services also tend to have high overall eco- References nomic growth. ADB (Asian Development Bank). 2012. Asian Development Outlook However, the causal connection remains uncertain: after 2012 Update: Services and Asia’s Future Growth. all, there is also a positive relationship—widely accepted in de- Alfaro, Laura, and Andrew Charlton. 2007. “Growth and the Qual- velopment economics—between manufacturing growth and ity of Foreign Direct Investment: Is All FDI Equal?� Centre for overall growth. But what has been overlooked is that the effect Economic Performance Discussion Paper 0830, LSE. of services growth on aggregate economic growth appears to Amin, Mohammad, and Aaditya Mattoo. 2006. “Do Institutions Matter More for Services?� Policy Research Working Paper be as strong, if not stronger, than the effect of manufacturing 4032, World Bank, Washington, DC. growth on overall growth. ———. 2008. “Human Capital and the Changing Structure of the Service-led growth is sustainable, because globalization Indian Economy.� Policy Research Working Paper 4576, World of services is still in its infancy. Moreover, the long-held view Bank, Washington, DC. that services are nontransportable, nontradable, and nonscal- Baumol, William J. 1967. “Macroeconomics of Unbalanced able no longer holds for a host of modern impersonal services, Growth: The Anatomy of Urban Crisis.� American Economic which can now be produced and exported at low cost. Devel- Review 57 (3): 415–26. Bhagwati, J. 2004. In Defense of Globalization. Oxford: Oxford oping countries can thus sustain service-led growth, given the University Press. huge room they have for catch-up and convergence. Bhagwati, Jagdish N. 1984. “Splintering and Disembodiment The process of globalization in the late 20th century led of Services and Developing Nations.� The World Economy 7: to a sharp divergence of incomes between those who indus- 133–44. trialized and broke into global markets and the “bottom bil- Blinder, Alan S. 2006. “Offshoring: The Next Industrial Revolu- lion� in some 60 countries where incomes stagnated for 20 tion?� Foreign Affairs 85 (2): 113–28. years. It seemed as if the “bottom billion� would have to wait Borchert, I., B. Gootiiz, and A. Mattoo. 2011. “Services in Doha: their turn for development, until giant industrializers like What’s on the Table?� In Unfinished Business?, ed. W. Martin and A. Mattoo, 115–144. Washington, DC: World Bank. China became rich and uncompetitive in labor-intensive ———. 2012. “Guide to the Services Trade Restrictions Database.� manufacturing. World Bank Policy Research Working Paper. No. 6108, Wash- The globalization of services, however, provides alterna- ington, DC. tive opportunities for developing countries to find niches, be- Bosworth, B., and A. Maertens. 2010. “Economic Growth and Job yond manufacturing, where they can specialize, scale up and Generation: The Role of the Service Sector.� In The Service Revo- achieve explosive growth, just like the industrializers. As the lution in South Asia, ed. E. Ghani. New York: Oxford Univ. Press. services produced and traded across the world expand with Dalum, B., K. Laursen, and B. Verspagen. 1999. “Does Specializa- tion Matter for Growth?� Industrial and Corporate Change 2: globalization, the possibilities for all countries to develop 267–88. based on their comparative advantage expand—and that com- Dasgupta, Sukti, and Ajit Singh. 2006. “Manufacturing, Services parative advantage can just as easily be in services as in manu- and Premature De-Industrialisation in Developing Countries: facturing or agriculture. The promise of the services revolu- A Kaldorian Empirical Analysis.� ESRC Centre for Business tion is that countries do not have to wait to get on the path to Research Working Paper 327. rapid development. There is a new way ahead. Eichengreen, B., and P. Gupta. 2009. “Two Waves of Services Growth.� National Bureau of Economic Research Working Acknowledgment Paper No. 14968, Cambridge, MA. Eschenbach, Felix, and Bernard Hoekman. 2005. “Services Policy This article draws upon The Service Revolution in South Asia Reform and Economic Growth in Transition Economies, (Ghani 2010), and columns in Project Syndicate and VOX EU. 1990–2004.� Policy Research Working Paper Series 3663, World Bank, Washington, DC. About the Authors ———. 2006. “Services Policies in Transition Economies: On the EU and WTO as Commitment Mechanisms.� World Trade Review Ejaz Ghani is Lead Economist in Economic Policy and Debt for 5 (3): 415–43. the Poverty Reduction and Economic Management (PREM) Net- Feenstra, R. C., and A. K. Rose. 2000. “Putting Things in Order: work. Arti Grover is a consultant at the World Bank. Homi Trade Dynamics and Product Cycles.� Review of Economics and Kharas is a Deputy Director for the Global Economy and Devel- Statistics 82 (3): 369–82. opment Program at the Brookings Institution. Francois, J., and B. Hoekman. 2010. “Services Trade and Policy.� Journal of Economic Literature 48 (3): 642–92. Note Francois, Joseph, and Bernard Hoekman. 2010. “Services Trade and Policy.� Journal of Economic Literature 48 (3): 642–92. 1. For an account of the differences between trade in goods Freund, Caroline, and Diana Weinhold. 2002. “The Internet and and services, see Copeland and Mattoo (2008), Hoekman International Trade in Services.� American Economic Review 92 and Mattoo (2008), and Francois and Hoekman (2010). (2): 236–40. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Ghani, Ejaz, ed. 2010. The Service Revolution in South Asia. Houseman, S., C. Kurz, P. Lengermann, and B. Mandel. 2011. New Delhi: Oxford University Press. http://sitere- “Offshoring Bias in U.S. Manufacturing.� Journal of Economic sources.worldbank.org/INTRANETTRADE/Resourc- Perspectives 25 (2): 111–32. es/239054-1239120299171/5998577-1254498644362/646 Jensen, R. 2007. “The Digital Provide: Information (Technology), 1208-1300395869284/SA_Service.pdf Market Performance, and Welfare in the South Indian Fisheries Ghani, Ejaz, Arti Grover Goswami, and Homi Kharas. 2011. “Ser- Sector.� The Quarterly Journal of Economics 122 (3): 879–924. vice with a Smile: A New Growth Engine for Poor Countries.� Kaldor, N. 1966. Causes of the Slow Rate of Economic Growth of the VOX EU, http://www.voxeu.org/article/service-smile-new- United Kingdom. Cambridge University Press. growth-engine-poor-countries. Kimura, F., and H. Lee. 2004. ‘The Gravity Equation in Interna- Gonzales, F., J. B. Jensen, Y. Kim, and H. K. Nordas. 2012. “Glo- tional Trade in Services.� European Trade Study Group Confer- balisation of Services and Jobs.� In Policy Priorities for Interna- ence, University of Nottingham, September 9–11. tional Trade and Jobs, ed. D. Lippoldt. Paris: Organisation for Klonner, S., and P. Nolen. 2008. Does ICT Benefit the Poor? Evidence Economic Co-operation and Development (www.oecd.org/ from South Africa. http://ifap-is-observatory.ittk.hu/node/583. trade/icite). Kolko, J. 2010. “Urbanization, Agglomeration, and Coagglomera- Goswami, Arti, Aaditya Mattoo, and Sebastian Saez, ed.2011. tion of Service Industries.� In Agglomeration Economics, ed. Exporting Services: A Developing Country Perspective. Washington, E. Glaeser, 151–80. Chicago: National Bureau of Economic DC: World Bank. Research and The Chicago University Press. Hausmann, Ricardo, Dani Rodrik, and Andres Velasco. 2005. Kox, H., and A. Lejour, A. 2005. “Regulatory Heterogeneity as “Growth Diagnostics.� Working Paper, Kennedy School of Obstacle for International Services Trade.� CPB Discussion Government, Harvard University, Cambridge. Paper, No. 49. Kuznets, S. 1959. Six Lectures on Economic Growth. New York: The Hausmann, Ricardo, Jason Hwang, and Dani Rodrik. 2007. “What Free Press of Glencoe. You Export Matters.� Journal of Economic Growth 12 (1): 1–25. Lall, S. 2000. “The Technological Structure and Performance of De- Head, Keith, Thierry Mayer, and John Ries. 2008. “How Remote Is veloping Country Manufactured Exports, 1985–98.� Oxford the Off-Shoring Threat?� CEPR Discussion Paper 6542 (forth- Development Studies 28 (3): 337–69. coming in the European Economic Review). McCredie, Andrew, and Darryl Bubner. 2010. Seven Myths about Hoekman, Bernard. 2006. “Trade in Services, Trade Agreements Services. Australian Services Round Table. and Economic Development: A Survey of the Literature.� Mirza, Daniel, and Giuseppe Nicoletti. 2004. “What Is So Special CEPR Discussion Paper 5760. about Trade in Services?� Leverhulme Centre for Research on Hoekman, Bernard, and C. Braga. 1997. “Protection and Trade in Globalisation and Economic Policy, Research Paper 2004/02. Services: A Survey.� World Bank Policy Research Working Paper Mishra, S., S. Lundstrom, and R. Anand. 2011. “Service Export So- 1747, Washington, DC. phistication and Economic Growth.� Policy Research Working Hoekman, Bernard, and Aaditya Mattoo. 2006. “Services, Econom- Paper 5606, World Bank, Washington, DC. ic Development and the Doha Round: Exploiting the Compara- Sridhar, K. S., and V. Sridhar. 2009. “Telecommunications Infra- tive Advantage of the WTO.� CEPR Discussion Paper 5628. structure and Economic Growth: Evidence from Developing ———. 2008. “Services Trade and Growth.� In Opening Markets for Countries.� Applied Economics and International Development 7 Trade in Services: Countries and Sectors in Bilateral and WTO (2): 37–56. Negotiations, ed. Juan A. Marchetti and Martin Roy, 21–58. World Bank. 2011. An Evaluation of World Bank Group Activities in Cambridge, UK: Cambridge University Press. Information and Communication Technologies: Capturing Technol- ———. 2011. “Services Trade Liberalization and Regulatory Reform: ogy for Development. Washington, DC. Re-Invigorating International Cooperation.� World Bank Policy ———. 2012. Information and Communications for Development: Maxi- Research Working Paper Series 5517, Washington, DC. mizing Mobile. Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are 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. 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise