75321 FEBRUARY 2013 • Number 106 Can Open Service Sector FDI Policy Enhance Manufacturing Productivity? Evidence from Indonesia Victor Duggan, Sjamsu Rahardja, and Gonzalo Varela Drawing on the findings of recent research, this note examines the extent to which changes to policy restrictions on foreign direct investment (FDI) in the Indonesian service sector affected the performance of downstream manufacturers during 1997–2009. The analysis uncovers two important findings: first, that relaxing restrictions toward FDI in service sectors was associated with improvements in the perceived performance of those sectors, and second, more importantly, that this relaxation accounted for 8 percent of the total observed increase in manufacturers’ total factor productivity (TFP) during this period. The results show that these TFP gains accrue disproportionately to those firms that are relatively more produc- tive and that gains are related to the relaxation of restrictions in the transport as well as the electricity, gas, and water sectors. TFP gains are associated, in particular, with the relaxation of foreign equity limits, screening and prior approval requirements, but less so with discriminatory regulations that prevent multinationals from hiring key personnel from abroad. Goods and Services: Mutually Reinforcing opening up the service and goods markets can be mutually Liberalization reinforcing, the full potential of each not being realized with- out adequate openness in the other. Increased openness in Just as Cinderella was the ignored sister with great potential, service sectors not only implies increased foreign presence, it so service sector FDI may be the neglected sibling of trade implies, more broadly, encouraging entry and inducing in- policy. As trade and FDI regimes have become more liberal- creased competition between foreign and domestic providers ized in recent decades, tariff reductions and FDI in manufac- alike. One would expect this competitive dynamic to deliver: turing have hogged the limelight, while FDI in services has (i) better and more reliable provision of existing services, (ii) often remained a wallflower. In fact, most restrictions on FDI new varieties of services, and (iii) competitive pricing in the flows today are in the service sectors (UNCTAD 2004), re- service sectors. Furthermore, one would expect increased flecting the fact that governments—particularly in developing productivity, trade and output in the service sectors, and im- countries—are not willing to allow unrestricted foreign entry proved economywide performance through links with the into sectors they consider sensitive or strategic.1 productive sectors. Openness in the service sectors is part and parcel of a Duggan, Rahardja, and Varela (2013) explore links be- comprehensive trade policy reform package. The benefits of tween changes to restrictions on service sector FDI in Indone- 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise sia and both the perceived performance of those service sec- tense in the immediate aftermath of the crisis (figure 1). tors and productivity in downstream manufacturing since Many of these policy changes were exogenously determined, the East Asian crisis. Their analysis uses firm-level data on driven by commitments to the International Monetary Fund, manufacturers’ productivity from 1997–2009, results from the World Trade Organization, and the Association of South- the World Bank Enterprise Surveys on performance percep- east Asian Nations. tions and the Organisation for Economic Co-operation and The reforms following the East Asian crisis had a tre- Development’s (OECD) FDI Regulatory Restrictiveness In- mendous impact not only on the development of Indonesia’s dex, combined with data from Indonesia’s input-output ta- service sectors, but also on overall economic growth. The re- bles regarding the intensity with which manufacturing sec- forms included abolishing state-owned enterprise (SOE) mo- tors use services inputs. nopolies, allowing more private operators, including FDI, and establishment of regulatory bodies for several service sectors. Indonesia: Reformed Service Sector Notably, Indonesia’s telecom and air transport sectors under- Driving Postcrisis Recovery went significant reform and have subsequently experienced Since the East Asian crisis, Indonesia has undergone political rapid transformation. As seen in figure 2, output from ser- and economic transformation, with stability becoming en- vices increased rapidly during the years of economic recovery trenched and the business environment improving signifi- (1999–2004), with a 6.5 percent average annual growth rate, cantly. It is one of the largest and fastest growing economies in compared to a GDP growth rate of 4.6 percent over the same Asia, and the world’s most dynamic growth pole. The service period. sectors combine to account for half of all those employed in The postcrisis relaxation of regulatory restrictions, along Indonesia, and more than half of gross domestic product with an increasingly attractive domestic market and relatively (GDP). With growth having averaged 7.2 percent over the lower labor costs, contributed to a strong recovery in FDI past decade, the service sectors are the most dynamic, driven flows across all sectors.  Despite Indonesia’s relatively restric- in large part by strong growth in the communications, trans- tive policy regime, FDI flows have been particularly strong in port, and trade sectors. Overall, the country has become more the service sectors (figure 3), accounting for 40.1 percent of open to trade and FDI, liberalization being particularly in- all FDI in 2011, driven in turn by investments in the trans- Figure 1. The Evolution of FDI Restrictiveness in Indonesia, Since 1997 overall FDI index all services real estate investment business services �nancial services communications media hotels & restaurants transport 2010 2006 distribution 2003 construction 1997 electricity 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Source: OECD FDI Regulatory Restrictiveness Index. Note: 0 = least restrictive, 1 = most restrictive. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 2. Services Played Important Role in Indonesia’s Economic Figure 3. Boom in Service Sector FDI, by Subsector Recovery 10 12 electricity, gas, and water supply construction 8 10 trade, hotel, and restaurant transport, storage, and communication annual real GDP growth (%) 6 8 real estate, industrial estate, US$ billions and business activity 4 6 other services 2 4 0 2 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 -2 0 agriculture & mining manufacturing 97 998 999 000 001 002 003 004 005 006 007 008 009 010 -4 services GDP 19 1 1 2 2 2 2 2 2 2 2 2 2 2 Source: Indonesia Investment Coordinating Board. Source: Staff estimates from “Statistics Indonesia� data. port, storage, communications, trade, hotel, and restaurant termediate manufacturing inputs, and 50 percent of interme- sectors. diate inputs into the service sectors themselves. While Indonesia is relatively open in terms of trade in These links underline the importance of services as in- goods, restrictions on foreign participation in the economy, puts for firms across different activities, and the potential for and the service sector in particular (either through trade or economywide productivity gains through improved service through FDI), remain pervasive despite broad-based liberaliza- sector performance. Changes in the regulatory environment tion efforts since 1997. Moreover, reforms have not been uni- for services therefore have the potential to have a strong econ- versally in the direction of increased liberalization. For exam- omywide impact. ple, having undergone significant opening to private sector Despite recent policy changes in limiting foreign owner- investment, both domestic and foreign, with the passage of the ship in the telecom and transport sectors, in general, Indone- landmark Telecommunications Act in 1999, the telecom sec- sian manufacturers have been increasingly presented with tor became relatively more closed through a tightening of for- better access to competitive services since 1997. The analysis eign equity limits after the passage of Presidential Regulation team calculated the weighted average restrictiveness of access 77/2007 and subsequently 36/2010 (that is, the investment to services faced by Indonesian manufacturers using plant negative list), which regulate investments in several sectors. level data and input-output tables, which use a methodology The OECD’s FDI Regulatory Restrictiveness Index re- presented in the next two sections. The calculations show flects these developments (figure 1): the communications that the average restrictiveness to services faced by Indonesian subindex fell significantly from 0.595 to 0.120 between manufacturing has declined steadily over the years (figure 4). 1997 and 2003 as a result of the 1999 reforms, but increased Access to more, better, or cheaper services can be expected to again to 0.410 between 2006 and 2010. The OECD index improve productivity at all Indonesian firms, enabling them also suggests that Indonesia’s service sector FDI regime is the to better compete in the global marketplace. second most restrictive, after China, of the 55 countries sur- Measuring the Impact: Service Sector veyed. Reform and Performance Indonesia: Links between Services and Given the important links between the service and manufac- Manufacturing turing sectors in Indonesia, key questions are: The individual service sectors combine to be not only the larg- (i) What are the productivity costs that manufacturers face est and most dynamic sector in Indonesia, but they are also due to policies restraining competition in the form of strongly linked with all of the productive sectors, accounting FDI restrictions in the service sectors? for 35 percent of overall intermediate inputs, including 27 (ii) What are the characteristics of the manufacturers most percent of intermediate primary inputs, 21.3 percent of in- affected by FDI restrictions in the service sectors? 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 4. Indonesian Manufacturers Have Better Access to subsector. The policy restrictiveness in services that is faced Competitive Services—Evolution in the Weighted Restrictiveness by each manufacturing sector user is calculated as a weighted Faced by Selected Manufacturing Sectors in Upstream Services average of each service sector’s OECD restrictiveness index, where the weights are given by the share in the total input bill of a given manufacturing sector. Figure 4 shows the evolution total manufacturing of this index over 1997–2010.2 The next step splits into two stages: first, links between this measure of policy restrictiveness and service sector per- formance as perceived by manufacturers are explored, and radio, television, and communication second, manufacturing firms’ TFP is related to the indicator equipment of policy restrictiveness in services.3 Reform and Perception-Based Performance in the Service Sector rubber & plastics For reforms in the services sector to have an effect on manu- facturing productivity, a preliminary condition is that these reforms somehow affect the performance of the service sec- tors themselves, either by affecting the quality, availability, or chemicals price of the services provided to manufacturers. This section examines these two relationships. The performance of the service sectors, as measured by the perceptions of its users in manufacturing industries, is paper negatively related to how restrictive it is toward FDI. Analysis showed that a 1 percent improvement in service sectors’ per- ceived performance was associated with a 1 percent reduction in the degree of restrictiveness toward FDI in those service sectors (figure 5). textiles Figure 5. Increased Service Sector Openness Associated with 2010 Improved Performance Perceptions 2006 food & beverages 2003 all manufacturing 1997 radio, TV, and others 0 0.05 0.1 0.15 0.2 Source: BPS, OECD FDI Regulatory Restrictiveness Index, and authors’ calculations. rubber & plastics (iii) In which service sectors are restrictions most costly? chemicals (iv) Which specific restrictions hurt manufacturers the most? paper Four data sets help provide answers to these questions: the OECD FDI Regulatory Restrictiveness Index for Indone- textiles sia for 1997–2009, input-output tables constructed by the Indonesian statistical office (BPS), Indonesian manufacturing food census data tracing all registered manufacturing firms in In- donesia with 20 or more employees (on average, containing - 0.02 0 0.02 0.04 0.06 information for about 20,000 firms per year), and perception increased openness of service sector data on service performance from the World Bank’s Enter- inputs prise Surveys for Indonesia (available for 2003 and 2009). performance improvement Input-output data provide a sense of the importance that each Source: Statistics Indonesia, OECD FDI Regulatory Restrictiveness Index, and service sector has for the input costs of each manufacturing authors’ calculations. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise The experience in freight and logistics services revealed Who gains the most from service sector reform? by data in Indonesia provides a concrete example of a typical Evidence suggests that in Indonesia, domestic and foreign- link between policy restrictiveness and performance. If ser- owned manufacturing plants benefit alike from reform in the vices related to freight and logistics remain sheltered from service sectors. This is important from a policy perspective, foreign investment and competition, costs may remain high because it suggests that local firms can benefit as much as for- and service quality poor, thus undermining the capacity of eign firms from further reform, and contrasts with evidence the private sector to benefit from business opportunities that found in different contexts. For example, in India, Arnold et involve long distance shipping. Logistics costs in Indonesia are al. (2010) find that gains accruing to foreign-owned manufac- estimated to be nearly double those in the Republic of Korea, turing plants were about 12 percent greater than those accru- and nearly three times those in Japan. Costs are even signifi- ing to domestic plants. cantly greater than those in neighboring countries. For exam- How productive you are matters for how much you ple, it costs US$750 to transport a container from Cikarang gain. The marginal return to service sector reform is higher to Tanjung Priok on the island of Java, but only US$450 be- among the best performers. The spillover effect from re- tween Pasir Gudang and Tanjung Pelepas in Malaysia, a rough- duced restrictiveness in the service sectors over the last 15 ly similar distance. years was not homogeneous across firms. While the least While the major importance of quantity and quality of productive 25 percent of firms did not benefit from service infrastructure should not be underestimated, differences in sector reform, the gains accruing to the 25 percent most pro- terms of openness and competition in the Indonesian service ductive were about double in size than those accruing to sector may also play a role. The road freight sector is relatively firms in the central part of the distribution. This result sug- more open in Malaysia, where they even allow full foreign gests that reforms in services disproportionately benefited ownership if services rendered include technologies, vehicles, those firms that had better capacity and technology to em- and expertise not available in Malaysia. Moreover, port han- ploy inputs. These more productive firms also tend to add dling charges are substantially higher in Indonesia than in more value per unit of output, generate more foreign ex- Malaysia, likely due, at least in part, to the fact that while a change in export markets, and create better paying jobs than monopoly operates the terminals of Tanjung Priok (Indone- less productive firms. sia), and cargo handling is closed to foreign ownership, several Not all reforms are created equal… firms (foreign and domestic) compete in Tanjung Pelepas From a policy perspective, it is important to know whether all (Malaysia), driving costs down and improving the quality of types of reforms conducive to reduced restrictiveness to FDI services. in services are equal from the point of view of their effects on Reform and Manufacturing Productivity manufacturing performance. Evidence suggests this is not the case. Post-Asian crisis reforms in the service sectors in Indonesia The sector in which restrictions are relaxed matters. have contributed to increased productivity of its manufac- Transport, electricity, gas, and water appear to be the service turing firms, adding roughly 0.4 percentage points to annual sectors in which reform most benefits manufacturers, while average productivity growth over 1997–2009. With Indo- the effects on manufacturers’ TFP derived from reform in the nesian manufacturing plants increasing their productivity communications, distribution, and construction sectors are by almost 43 percent over the period, and the total effect of indistinct, from a statistical point of view.5 relaxing restrictive policies toward FDI in the service sectors The type of restriction that is relaxed also matters. Anec- being close to 3.5 percent, the results here suggest that these dotal evidence suggests that foreign firms value having the reforms accounted for a sizable 8 percent of this total. This is majority on a firm’s board of directors. This allows them to be roughly in line with international evidence; for example, more effective in influencing firms’ operations and strategic Fernandes and Paunov (2012) find that reforms in the ser- vice sectors in Chile accounted for about 5 percent of the plans. Thus, restrictions in terms of maximum equity allowed total productivity growth of manufacturers over 1992– to be held by foreigners in particular sectors may be especially 2004. harmful in discouraging foreign entry and can discourage for- A numerical exercise is useful to give a sense of the rele- eign counterparts from deploying their best systems and tech- vance of this result. If Indonesia were to match the policies of nology. Results from this analysis confirm this to be the case. a service sector reform champion, such as Korea, the produc- While specific restrictions to hiring key personnel from tivity gains for manufacturing firms would be on the order of abroad by foreign firms in the services sector seem to be rela- 5 percent instead of the observed 3.5 percent.4 Matching Bra- tively innocuous for manufacturing performance, restric- zil’s level of restrictiveness would lead to TFP gains on the or- tions in terms of maximum foreign equity allowed are nega- der of 6 percent, on average (figure 6). tively associated with manufacturing firms’ TFP. The same 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 6. Estimated Effects of Service Sector Liberalization on There is nothing automatic about benefits from service sec- TFP, and Some Simulations tor FDI positively impacting economywide productivity. a. 14 Appropriate supply side supports are necessary to ensure the greatest possible welfare and productivity gains. Policies lower bound (90% CI) 12 on skills and education, investment in infrastructure, and mid-point estimate adequate protection for intellectual property all have a role 10 upper bound (90% CI) to play in maximizing the positive impact of opening the ser- vice sectors to foreign investment. Given the risks of market 8 failures, such as limited accessibility, poor quality and regu- 6 latory capture by big service providers, proper incentives and a robust regulatory regime have significant roles to play in 4 encouraging competition and ensuring universal access and quality services for users. Broad-based, productivity-enhanc- 2 ing service sector reform should then necessarily encompass 0 measures to increase competition, stimulate innovation and liberalization since matching Korea matching Brazil ensure judicious regulation, in addition to encouraging for- 1997 eign investment. Recent proposals to reform the Indian in- surance and pension sectors provide high profile examples of b. -0.1 efforts to package reforms aimed at simultaneously easing foreign equity limits in the service sector while strengthen- ing the regulatory framework. -0.08 Like Cinderella, Indonesia needs glass slippers for both feet. For Indonesia, reaping the maximum benefits from for- eign investment and global trade is likely to require the pro- -0.06 gressive opening of trade-related service sectors to foreign in- vestment, while maintaining an open posture on trade in commodities and manufactured goods, coupled with contin- -0.04 ued efforts to improve policy certainty, macro stability, and the business environment. -0.02 About the Authors Victor Duggan is a Consultant with East Asia Pacific PREM Indo- nesia. Sjamsu Rahardja is a Senior Trade Economist at East Asia 0 Pacific PREM Indonesia. Gonzalo Varela is a Consultant with all domestic the World Bank’s International Trade Department. lower bound (90% CI) mid-point estimate Acknowledgments upper bound (90% CI) The authors thank Beata Javorcik for very useful comments Source: Authors’ calculations. and suggestions on earlier drafts; Stephen Thomsen from Note: The red bars indicate the point estimates of the induced effects; the blue bars show the lower and upper bounds of 90 percent confidence intervals. OECD for providing information on the OECD FDI Restric- tiveness Index; and Ana Fernandes, Stephen Magiera, and holds for restrictions related to screening, prior approval re- Deborah Winkler for helpful comments. quirements, and other operational restrictions. Notes Conclusion 1. Many service sector activities are plagued with market fail- Substantial gains can still be achieved if countries move for- ures, for example, natural monopolies, incomplete markets, ward with their liberalization agendas in the service sectors. or asymmetric information. This may partially explain why Indonesia, in particular, has much to gain from tapping ser- governments often intervene heavily in these activities. vice sector FDI to drive the economywide productivity gains 2. This follows the approach of Arnold, Javorcik, and Mattoo needed for sustainable long-term economic development. (2011). 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise 3. This is calculated by regressing TFP on the manufactur- Arnold, J., B. Javorcik, and A. Mattoo. 2011. “Does Services Liberal- ing sector–specific index of policy restrictiveness in service ization Benefit Manufacturing Firms? Evidence from the Czech Republic.� Journal of International Economics 85 (1): 136–46. sectors and other controls that include firm-level fixed ef- Arnold, J., A. Mattoo, and G. Narciso. 2008. “Services Inputs and fects, year fixed effects, manufacturing input and output Firm Productivity in Sub-Saharan Africa: Evidence from Firm- tariffs, foreign presence in the manufacturing sector, and Level Data.� Journal of African Economies 17 (4): 578–99. policy restrictiveness toward FDI in the manufacturing sec- Blalock, G., and P. Gertler. 2008. “Welfare Gains from Foreign tors themselves. Direct Investment through Technology Transfer to Local Sup- 4. Korea speeded up its service sector reform agenda recently, pliers.� Journal of International Economics 74 (2): 402–21. at a later stage of development than Indonesia. Blalock, G., and D. Simon. 2009. “Do All Firms Benefit Equally 5. The estimated effects of restrictiveness on TFP are negative from Downstream FDI? The Moderating Effect of Local Suppli- in all cases, but for these three sectors, they are not statistically ers’ Capabilities on Productivity Gains.� Journal of International Business Studies 40: 1095–1112. significant. Evidence from India reported by Arnold et al. Duggan, V., S. Rahardja, and G. Varela. 2013. “Service Sector (2010) also points to reform in the transport sector as a top Reform and Manufacturing Productivity: Evidence from priority for manufacturers, as well as reform in communica- Indonesia.� Policy Research Working Paper 6349, World Bank, tions and finance. Washington, DC. Fernandes, A. M., and C. Paunov. 2012. “Foreign Direct Invest- References ment in Services and Manufacturing Productivity: Evidence for Amity, M., and J. Konings. 2007. “Trade Liberalization, Inter- Chile.� Journal of Development Economics 97 (2): 305–21. mediate Inputs, and Productivity: Evidence from Indonesia.� Hoekman, B. 2006. “Trade in Services at 25: Theory, Policy, and American Economic Review 97 (5): 1611–38. Evidence.� Mimeo, World Bank, Washington, DC. Arnold, J., B. Javorcik, M. Lipscombe, and A. Mattoo. 2010. “Ser- UNCTAD (United Nations Conference on Trade and Develop- vices Reform and Manufacturing Performance: Evidence from ment). 2004. World Investment Report: The Shift Towards India.� CEPR Discussion Papers, 8011, Washington, DC. Services. New York and Geneva. 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. 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise