Report No. 49371-KH Cambodia A Better Investment Climate to Sustain Growth Second Investment Climate Assessment April 2009 Poverty Reduction and Economic Management Sector Unit East Asia and Pacific Region Document of the World Bank and the International Finance Corporation Table of Contents Preface ............................................................................................... iii Acknowledgments ................................................................................... v Executive Summary ................................................................................ vi 1. The Private Sector in Cambodia ............................................................ 1 A. A Largely Informal Private Sector with a Small Modernizing Formal Sector ............................. 1 B. Performance of the Private Sector ..................................................................................................... 5 C. Sector Highlights .................................................................................................................................11 D. Summing Up ........................................................................................................................................18 2. Constraints to Productivity and Growth ................................................. 19 A. Overall Constraints and Impact on Productivity ...........................................................................19 B. Potential Constraints to Investment and Productivity ..................................................................22 C. Summing Up: Binding Constraints ...................................................................................................45 3. The Reform Agenda ......................................................................... 48 A. Ongoing Efforts ..................................................................................................................................48 B. Key Directions .....................................................................................................................................50 C. Concrete Priorities ...............................................................................................................................52 Annexes ............................................................................................. 57 A. Summary Recommendations .............................................................................................................58 B. Summary Statistics ...............................................................................................................................62 C. The 2007 Investment Climate Survey ..............................................................................................68 D. The Doing Business Indicators .........................................................................................................71 E. Productivity Analysis ..........................................................................................................................74 F. References.............................................................................................................................................85 LIST OF BOXES Box 1: Cambodia's Private Sector and its Business Environment in Numbers ....................................................... vii Box 2: Growth and Governance in Cambodia .............................................................................................................. xi Box 1-1: Analytical Tools Supporting this Report ......................................................................................................... 4 Box 2-1: Key Findings from the 2003 Investment Climate Survey ........................................................................... 21 Box 2-2: Transparency of Export Procedures .............................................................................................................. 29 Box 2-3: Setting up a Business in Cambodia ................................................................................................................. 36 Box 2-4: Breach of Trust ................................................................................................................................................. 38 Box 2-5: Alternative Dispute Resolution ....................................................................................................................... 39 Box 2-6: The Government-Private Sector Forum........................................................................................................ 44 Box 3-1: The Trade SWAp .............................................................................................................................................. 50 LIST OF FIGURES Figure 1: Constraints on the Business Environment in 2004 and 2008 ..................................................................... ix Figure 1.1: Foreign Direct Investment ............................................................................................................................. 2 i Figure 1.2: Connection to Global Markets ...................................................................................................................... 3 Figure 1.3: Labor Productivity and Wages Across Countries (2004 US$) .................................................................. 7 Figure 1.4: Labor Productivity vs. Capital Intensity in the Garment Sector ............................................................... 8 Figure 1.5: Productivity Gap between Mean and Top Performer across Sectors ...................................................... 9 Figure 1.6: Distribution of TFP in the Garment Sector .............................................................................................. 10 Figure 1.7: Standard Cut, Make, Trim, Trade and Transport Costs........................................................................... 12 Figure 1.8: Value-chain for Aluminum 330ml Can (Body) Production in Cambodia ............................................. 15 Figure 1.9: Inland Freight Costs (US$/km/ton) .......................................................................................................... 16 Figure 1.10: Cost of Electricity (US$/kWh) .................................................................................................................. 16 Figure 2.1: Constraints on the Business Environment in 2004 and 2008 ................................................................. 19 Figure 2.2: Firms using Banks to Finance Investment (%) ......................................................................................... 22 Figure 2.3: Has your Firm a Loan? ................................................................................................................................. 23 Figure 2.4: Firms Offering Formal Training (%, manufacturing sector)................................................................... 24 Figure 2.5: Days of Inventory of Most Important Input ............................................................................................ 27 Figure 2.6: Average Time to Clear a Shipment (number of days) .............................................................................. 28 Figure 2.7: Cost and Timeliness of Shipments ­ Garments and Other Sectors ....................................................... 30 Figure 2.8: Electricity ........................................................................................................................................................ 32 Figure 2.9: Regulations ..................................................................................................................................................... 35 Figure 2.10: Company Registrations............................................................................................................................... 35 Figure 2.11: Mitigating Risks of Commercial Disputes ............................................................................................... 37 Figure 2.12: Corruption.................................................................................................................................................... 40 Figure 2.13: Consumer Price Index ................................................................................................................................ 42 LIST OF TABLES Table 1.1: Firm Distribution (among those with 5 employees or more) ..................................................................... 2 Table 1.2: Drivers of Informality ...................................................................................................................................... 5 Table 1.3: 2005/06 Labor Productivity (2004 US$) ....................................................................................................... 6 Table 1.4: 2005/06 Value Added per Worker, Wages per Worker, and Mark-Ups (2004 US$) .............................. 6 Table 1.5: Costs for Denim Jeans and Polo Shirts (US$ / piece)............................................................................... 11 Table 1.6: Cost and Source of Packaging Material Used by Major Supermarkets ................................................... 15 Table 1.7: Different Entrepreneurs ­ Different Potentials ......................................................................................... 18 Table 2.1: Impact of Investment Climate Constraints on TFP .................................................................................. 22 Table 2.2: Logistics Performance Index ........................................................................................................................ 28 Table 2.3: Telecommunications and Internet Access in East Asia ............................................................................ 33 Table 2.4: The Many Facets of Corruption and their Impact ..................................................................................... 41 Table 2.5: Cost of Doing Business ................................................................................................................................. 46 Table 2.6: Characteristics of the private sector and constraints ................................................................................. 47 ii Preface On behalf of the Royal Government of Cambodia, I would like to thank the World Bank and IFC for producing this report about Cambodia's investment climate. This publication offers a wide range of information about factors which help and hinder the development of Cambodia's private sector and the actions which key stakeholders ­ government, the private sector and development partners ­ must undertake collaboratively to continue improving the investment climate and the environment for doing business. As this report shows, Cambodia has made great progress in private enterprise development in the few short years since decades of civil war and isolation ended in the 1990s, but we still face important challenges. Private companies, and especially the small and medium enterprises, need better access to finance, to information on regulations and procedures, to export opportunities, to reliable dispute resolution and to efficient, transparent and honest government services. Improving the investment climate and the environment for doing business has always been a high priority for the Royal Government of Cambodia as we see private enterprise development as a key factor in sustaining economic growth and reducing high rates of poverty. Private enterprise development is one of four key pillars in the Royal Government's Rectangular Strategy and we support private enterprise through offering investors among the most attractive incentives in Asia. Cambodia was also the first low income country in world to join the World Trade Organization and we are keen participants in ASEAN and other trade-related bodies. The report also documents progress in a number of areas under the leadership of the Royal Government. As this report shows, our efforts to support private enterprise have paid off in nearly 10 percent annual GDP growth from 1997 to 2007, and poverty rates that declined 1 percent per year over the same period. Since 2008, when the global financial crisis began to seriously impact our key sectors of growth and threaten to reverse the important gains we have made in recent years, the Royal Government has been working even harder to address investment climate problems. Although we still have differences with the Bank and IFC team on some of the methodologies, data and the assessment we appreciate the work done and take it proactively in addressing the challenges we face now. This is important not only to mitigate the impact of the current global downturn but also to help Cambodia achieve its very considerable potential and continue improving the livelihoods and lives of all Cambodians. We have many opportunities for regulatory reforms that would make the business environment more conducive to a healthy and diversified growth of the business sector. iii As numerous studies have shown, Cambodia has great potential in a number of sectors. Due to our worldwide reputation for good labor standards in the garment industry, we still have strong potential in garment and footwear manufacturing that should be realized once economic recovery starts and consumers start buying again. We also have a wealth of tourist attractions in addition to Angkor Monuments, and excellent potential in agriculture and agro- processing. Recent exploration shows potential too in developing oil, gas and mineral resources. Cambodia has important geographic advantages as well. We are well located in the center of Southeast Asia, we have river as well as sea ports and our population is comparatively small and young, with a lifetime of productive years ahead of it. Very importantly, we also have a stable society. The global financial crisis has not diminished Cambodia's considerable potential and we can use this time of slower growth to our advantage if we all work together to make even greater efforts to reduce investment climate problems. This investment climate assessment is an important management tool for all stakeholders to focus on opportunities for regulatory reforms, monitor the implementation of our Rectangular Strategy, and position Cambodia for sustained growth and prosperity. Phnom Penh, 29 April 2009 Keat Chhon Deputy Prime Minister Minister of Ministry of Economy and Finance iv Acknowledgments 1. The Investment Climate Assessment that led to this report was initiated in August 2007, under the guidance of His Excellency Hang Chuon Naron (Permanent Vice-Chairman of the Supreme National Economic Council). Early findings were discussed with researchers from the Supreme National Economic Council in August 2008. The final report was reviewed by the Private Sector Development Steering Committee, chaired by His Excellency, Deputy Prime Minister and Minister of Finance, Keat Chhon, on March 31, 2009. Comments from this review have been incorporated in the final draft, presented to His Excellency Keat Chhon on April 30, 2009. 2. The World Bank team was led by Stéphane Guimbert. The team was composed of: Huot Chea, Sophorn Kith, Yasuo Konishi, Andrew McNaughton, Remco Oostendorp, Peng Seng Tan, Tilahun Temesgen, and, from the IFC, Soneath Hor, Khemara Ros, Charles Schneider, Lili Sisombat, and Khy Touk. Ann Bishop provided technical editing and proofreading for the report. The team worked under the guidance of Ian Porter (Country Director), Nisha Agrawal and Qimiao Fan (Country Manager), and Kazi Matin and Mathew Verghis (Lead Economist). Peer reviewers ­ Magdi Amin (IFC) and Syed Mahmood (IFC) ­ provided comments. Albert Zeufack provided guidance on the design of the survey. James Brew provided invaluable insights on the private sector and suggestions on the final report. The analysis benefited from very close collaboration with the team at the Garment Productivity Improvement Center (GPIC), supported by USAID, with significant insights and contributions from Mona Tep, Jane O'Dell, and Peter Minor. Finally, comments from ILO's Tuomo Pountainen and John Ritchotte are gratefully acknowledged. 3. The team would like to thank as well all the private sector contributors to the 2007/08 Investment Climate Survey. The survey was undertaken by Indochina Research Limited, under the outstanding leadership of Charis Wuerffel. Senior managers from some 500 firms provided their valuable time and insights in completing the questionnaire. Follow-up meetings were held in September 2008. Feedback from Van Sou Ieng (GMAC/CAMFEBA) is gratefully acknowledged. 4. The concept note for the work program was reviewed in September 2007. This report was reviewed in October 2008 within the World Bank Group. The analysis fed into the broader framework of the "Sustaining Rapid Growth" analysis prepared by the World Bank and its development partners in 2008, in preparation for the 2008 Cambodia Development Cooperation Forum and the 2009 Cambodia Economic Forum. v Executive Summary 1. The objectives of this Investment Climate Assessment (ICA) are to: (i) provide an up-to-date, fact-based analysis of the business environment for policy-makers in the Royal Government of Cambodia (RGC), the private sector, civil society, and development partners; and (ii) outline priorities for improving the business environment and suggest policy options. This updates the ICA published in 2004. 2. A key finding of this ICA is that Cambodia has been growing at double-digit rates since 2004 with progress along a number of dimensions of the investment climate. Sound macroeconomic management and deeper integration into the world economy continue to be hallmarks of the period of sustained growth. Progress in trade facilitation has been noteworthy, with faster border clearance in 2008 than 2004. With the development of the country, constraints such as security have become much less of a concern as well, while access to finance and information technologies has improved significantly. 3. Nevertheless firms' concerns about business environment constraints are very similar in both the 2004 and 2008 ICA surveys. One could draw from this, the simple conclusion that the economy will continue to grow in spite of concerns about the business environment. However this would be misleading for three reasons: (i) characteristics of the private sector suggest that business environment constraints are actually limiting some forms of growth which are likely to be important now that the global economy is shrinking; (ii) the persistence of some constraints indicates they are deep rooted, hence challenging to alleviate; and (iii) despite the similarity of constraints identified in 2004 and 2008, there have been significant improvements in the business environment that suggest how more improvements could be made. The 2008-09 global economic crisis, to which Cambodia's private sector is exposed through international trade, tourism, and foreign investments, makes further improvements to the business environment even more urgent, to weather the global crisis and encourage new local and foreign investors to further develop Cambodia's considerable assets in tourism, agriculture, minerals, and related manufacturing. 4. The investment climate covers a range of indicators that determine whether it will be viable and profitable to invest in a country. These include contextual issues (such as macroeconomic management or concerns about security); regulatory issues (such as levels of tax, unofficial fees, incidence of land disputes); and need for infrastructure, skills, access to finance, etc. In this report, the investment climate has been assessed primarily through a survey of 500 firms, conducted between October 2007 and February 2008. This focused on establishments that had been operating for at least one full year, had five employees or more, and were located in one of the five main urban centers in Cambodia. The results of 2008 ICS are complemented with results from the same survey conducted in Cambodia in 2004; similar surveys conducted in other countries; the Doing Business Indicators (which mainly use a case-study approach); case studies on garment manufacturing and food packaging; and direct feedback from the private sector. 5. The body of this report is comprised of three chapters. Chapter 1 gives a picture of the private sector in Cambodia in early 2008, including its structure, main features, and overall performance across key sectors and nationwide. The sectors featured are: garments, food packaging, telecoms, and finance. Chapter 2 reviews the business environment, starting with existing firms' vi perceptions of the business environment, and then goes into more depth on each dimension. The chapter concludes by discussing which of these challenges are truly binding constraints on growth. Chapter 3, after taking stock of on-going reform efforts, concludes with some key directions for reform and a few concrete policy options for the short and medium term. Five Features of Cambodia's Private Sector 6. Cambodia's achievements on the economic front are remarkable, as evidenced by ten years of rapid growth (more than 9½ percent per annum between 1998 and 2007). This has been entirely driven by the private sector, with the public sector providing support. Until 2008, macroeconomic indicators showed continued growth in exports as well as in foreign direct investment (FDI), both of which are positive signs for the investment climate. A number of microeconomic indicators (see also Box 1) provide clear evidence of a vibrant and modernizing private sector. This evidence includes an increasing number of firms registering every year (from 720 in 2004 to 2,890 in 2008); an increasing proportion of firms using e-mail and websites; and rapid growth in technology-based sectors such as telecoms and banking. In addition, the firms surveyed for the 2008 Investment Climate Survey were very upbeat about their plans for expansion. Nevertheless, five features of Cambodia's private sector suggest avenues to further develop the country's economic potential. Box 1: Cambodia's Private Sector and its Business Environment in Numbers The industrial and non-public service sectors grew at 16 percent and 11 percent, respectively, between 1998-2007 (against 4 percent for agriculture). As of early 2008, an impressive 83 percent of firms said they intended to increase their investment in the next three years (37 percent intended to "increase drastically"). Among firms with five employees or more, 71 percent are small (fewer than 10 employees) and 7 percent are large (more than 100 employees); and 9 percent are in manufacturing (mainly garments), against 29 percent in trade and 25 percent in tourism. A total of 19 percent of firms have foreign owners and 43 percent of inputs in the manufacturing sector are imported. This connection to global markets is strongest in the garment industry. In total, 56 percent of firms use e-mail to contact suppliers and clients, and 39 percent of firms have a website. Only 14 percent of firms have audited financial statements. Only 10 percent of firms use the banking sector to finance investments. Women's participation in the workforce is 37 percent, with 30 percent in management positions. Value added per worker was around US$5,000 in 2008, the annual wage per worker was US$1,600, and capital investment per worker was around US$9,600. However, there were significant differences across sectors for these statistics (for instance, in the garment sector, these three statistics were respectively, US$2,800, US$1,200, and US$1,400). A total of 36 percent of firms use a generator, and firms report that 3.8 percent of their sales are lost due to power outages. Evidence of business owners concerns about uncertainty includes: firms keeping 41 days of inventory; 21 percent of sales being pre-paid; only 22 percent of sales being made on credit; and only 0.3 percent of business disputes taken to court. Senior managers spend 9 percent of their time dealing with Government regulations and 84 percent of the firms make informal payments to "get things done". 7. First, and most importantly, entrepreneurs have concerns that hold them back. This was noted in the 2004 ICA, and four years of double-digit growth (and significantly lower concerns about security and political stability) have not changed attitudes and behavior with regard to taking business risks. Business owners' concerns about uncertainty are visible in the high level of inventory that firms keep (highlighting their lack of trust in logistics services providers); the high level of pre-paid sales; the low level of sales made on credit; the reluctance of firms to use the courts to settle business disputes (pointing to lack of trust in the conflict resolution system); and the low level of capitalization (pointing to similar problems, as well as vii lack of skilled human resources and reliable electricity). This translates into two risk avoidance behaviors of particular importance: high levels of informality and lack of diversification. 8. Second, the private sector remains largely informal and firms are more generally small. Even among firms with five employees or more, 71 percent have fewer than 10 employees. They tend to significantly under-report income for tax purposes (78 percent of firms state that a typical firm reports less than 100 percent of its income). Analysis shows that the decision to remain informal and/or small is largely based on costs (cost of registration; cost of becoming more visible to authorities) and few benefits from becoming formal or larger (including limited access to credit; no benefit from courts). While a significant part of the informal sector and many small businesses belong to the subsistence economy, and are simply a reflection of the economy's level of development, it is likely that a number of these enterprises would have strong growth potential if owners and managers were convinced of the net benefits of becoming formally registered. 9. Third, lack of diversification also shows business owners' concerns. Even though diversification has long been important, garments remain Cambodia's dominant export. And even within the garment sector, product and market diversification have been limited, with most firms undertaking only basic `cut-make-trim' activities. The economy's limited capital intensity ­ itself impacted by risk avoidance, the high cost of electricity, and the lack of skilled workers ­ continue to make diversification a challenge. 10. A fourth feature of the Cambodian private sector is its strong connection to the global economy. In addition to macroeconomic data, survey data also confirm this finding. This results from a number of successful government policies including: liberal trade and investment regimes; expanded quotas for the garment sector; accession to the World Trade Organization; and, more recently, progress in trade facilitation (see para 18). Foreign direct investment has played a critical role in bringing financing, know-how, and connections to global supply chains. However this also reflects that local entrepreneurs find it difficult to access sufficient finance and connect to international markets ­ factors that together discourage investment. 11. The fifth and final feature of the private sector is its low and variable productivity. For understandable reasons, given Cambodia's decades of war and isolation which only ended in the 1990s, productivity in Cambodia remains low (Box 1), despite significant improvements in recent years. For instance, across the economy, productivity is below that of six comparators, with the exception of Bangladesh. Importantly, this gap reflects, in large part, thin capitalization. However efficiency is, on average, close to competitors. Indeed, some Cambodian firms are at the cutting-edge in their sector. This is particularly true for the garment sector, for which comparative cross-country data is available. This unevenness of performance across sectors suggests that: (i) Cambodia has some of the important assets needed to be competitive, but (ii) the incentives to be more competitive and productive are, in some cases, lacking. 12. Adjustments to these five features could boost Cambodia's growth potential. Risk- taking is essential for economic activity and growth. If business owners have concerns about the business environment, it reduces their willingness to take productive risks and, at the same time, constrains growth. There is no firm consensus on the overall need for `formalization', and it should be recognized that a large informal sector is typical of a low-income country. Nevertheless, too large an informal economy is likely to discourage the economies of scale that large firms achieve internally, or a network of well-established firms achieves by trading with each other. Comparison of data from other countries also suggests that poor diversification limits growth. Too narrow an economy is also risky as little else will generate earnings if key sectors decline, as is apparent in 2009. Lack of innovation and investment by local entrepreneurs is another concern. Finally, stronger incentives for workers to increase productivity would also viii drive growth. If all firms in each industry were as productive as the top Cambodian performers in their industry, labor productivity would more than double. In the garment sector alone, bringing up total factor productivity to the level of Thailand would increase production by 15 percent. This could be done by improving efficiency, and not necessarily by increasing capital intensity. Characteristics of the Private Sector and Investment Climate 13. Corruption remains the Number 1 concern of firms (Figure 1), despite the good performance of the economy (Box 2). The Government fully recognizes that governance and corruption are key issues for the economy and places better governance as a top policy priority in the Rectangular Strategy, Phase II and the National Strategic Development Plan. Related issues, such as regulatory uncertainty and informal practices also rank high among firms' concerns. It is important to unpack the many facets of corruption. There is evidence of corruption at the service-delivery level (e.g. making informal payments to speed up connection to the electricity grid); at the regulatory level (e.g. informal payments for tax transactions or to influence the courts); at the public procurement level; and likely too at the level where regulations are interpreted. Each of these translates into financial costs and unpredictability, with firms wasting time and money on mitigating strategies. While some sectors have been less exposed to, or been more capable of dealing with corruption, corruption's many facets certainly have had a cost in terms of low capitalization, lack of diversification, a large informal sector, and other strategies for avoiding risk. 14. Macroeconomic uncertainty has also emerged as an important concern. This concern might seem at odds with the Government's continuing good macroeconomic management. However, at the time of the survey, inflation was rising (mainly driven by external factors): this concern has entirely receded since then. This concern, which was particular high for garment Figure 1: Constraints on the Business Environment in 2004 and 2008 60 Less severe constraint in 2008 51.8 More severe constraint in 2008 S i 1 50 38.2 40 31.8 30 26.7 20.9 18.1 20 16.0 15.6 15.5 12.8 12.0 11.5 11.5 10.8 10.6 9.1 10 7.2 3.3 - Elec tric ity Labor regulations Macro Ec onomic Legal system/conflict Corruption Skills & Education Tax Administration Acc ess to Financing Transportation Cus toms and Trade Anti-competitive or Tax Rates Licensing and Permits Economic & Regulatory Cos t of Financing Crime, theft and disorder Telecommunications Access to land informal practices Policy Unc ertainty Uncertainty regulations resolution Note: Share of firms that respond "severe or very severe" to the question "how problematic are the following issues to the operation and growth of your business". Source: 2003 and 2007 Cambodia Investment Climate Surveys. ix firms, also likely reflect uncertainty in the global economy. 15. Electricity is also a growing concern for firms. Electricity from the grid has been expensive and unreliable. As a result, many Cambodian firms have coped by owning a generator, but this is a major investment, and expensive to run (e.g. the cost of diesel) and maintain (e.g. cost of imported parts). Lack of skilled workers, as well as the high cost of electricity, have discouraged investment in capital intensive (and typically more electricity-intensive) industries. With declining oil prices in 2009, and additional electricity becoming available mid-year from Vietnam, electrical supplies and price should improve at least for Phnom Penh. 16. Labor market regulations are not viewed as a significant constraint (and could be an asset, as is the case with good labor standards in the garment industry). However, lack of skills and poor industrial relations are becoming more urgent concerns. Between 2004 and 2008, the number of days lost to labor disputes increased from 2.6 to 6.0 days per establishment, per annum (and as high as 8 days per factory in the garment sector). 17. Access to finance was not viewed as a major constraint by most firms. Firms noted progress in the financial sector, both in terms of access to and cost of finance. However, very high collateral requirements remain a barrier, with the result that only 1 in 5 firms has a loan. In late 2008, as the macroeconomic environment began to deteriorate rapidly, access to credit likely became a more acute problem. 18. Significant progress has been made in the area of trade facilitation. The average time required to clear a shipment declined to 4.3 days for exports and 5.1 days for imports. This has been of considerable benefit to the garment sector. Progress still needs to be made across all sectors though, to stimulate diversification. Greater availability of logistics services providers such as trucking firms and customs brokers is also needed in Cambodia. Five Goals to Improve Investment, Productivity, and Diversification 19. The constraints faced by the Cambodian private sector explain many of its features such as high informality, lack of diversification and lack of large firms. But it is unlikely that all constraints are binding at the same time, and given limited Government resources to undertake reforms, these need to be prioritized. A `Growth diagnostics', a methodology developed by economists Ricardo Hausmann and Dani Rodrik, reflects on key features of the private sector and the constraints which best explain these. These features include risk avoidance, informality, lack of diversification, limited domestic investment, and low and uneven levels of productivity. 20. Is finance a constraint? Do entrepreneurs fail to invest (at all, let alone in diversification or productivity improvement) because they lack good projects, or are they unable to finance viable projects? Analysis shows that, in most cases (and at the time of the ICA 2008 survey), finance was not a binding constraint. Most firms interviewed for the ICA indicated that they did not feel constrained by lack of finance. Many are foreign-owned, with parent company financing provided. Also, real interest rates were negative in 2008. That said, finance could well be a constraint for more risky projects, such as venturing into new production; for investors lacking foreign investment or not backed by a Cambodian conglomerate with retained earnings. Finance is also possibly a problem for small investors that would like to expand but lack the necessary collateral. Lack of export financing continues to be a constraint too, especially since the external environment changed so dramatically in late 2008. Lack of access to finance is likely a disincentive for firms to formalize. In summary, while access to finance is not a major barrier, the absence of some specific financial products such as long-term finance and export finance could be an important constraint. 21. Is it possible to generate high returns in Cambodia? If lack of finance is not the key issue, then do entrepreneurs find it difficult to identify high-return projects? This could be either x because it is simply impossible to generate high returns in Cambodia, or because while high returns are possible, certain factors limit profitability. Lack of skilled workers and lack of infrastructure (especially electricity) are certainly constraints for some types of projects (such as more capital intensive manufacturing). However, although currently a binding constraint, as stated previously, access to affordable electricity is expected to improve starting in 2009. Labor intensive industries should continue to be a viable option in Cambodia, but (i) industrial relations problems need to be addressed, and (ii) productivity needs to be improved through pay practices that reward greater effort. In addition, further progress in trade facilitation and the provision of logistics services would benefit most sectors since Cambodia relies much on international trade. 22. Then, could it be that entrepreneurs cannot get adequate returns from their projects? If, aside from issues of labor costs/productivity and trade facilitation/logistics, entrepreneurs can conceive of high-return projects but still do not undertake them, the problem must be that they cannot retain adequate returns. Inflation was a temporary concern at the time of the survey. Cumbersome procedures, lack of Government services, and corruption create uncertainty for most firms and add costs that reduce returns. These costs are high by international standards and contribute to keeping vibrant informal businesses out of the formal sector. Finally, there are a number of coordination and market failures that make it challenging for entrepreneurs to realize good returns on their investments (e.g. lack of information on markets; lack of logistics services). 23. To build on achievements to date, the foregoing analysis suggests five main goals for action: Goal #1: Encourage entry and diversification. Taking into account all the various types of entrepreneurs (foreign investors; established Cambodian entrepreneurs; small new entrants), there is a need to facilitate entry through simpler and more transparent processes. Easier entry will need, in addition, support for the development of public goods such as standards, information about markets and technologies, etc. More privately-provided logistics services such as truckers, packers and customs brokers should also be encouraged (see logistics below). Making electricity accessible and affordable will also be very important if Cambodia is to diversify through more capital intensive industries. Particular attention Box 2: Growth and Governance in Cambodia There is much empirical evidence worldwide that good governance is an important contributor to development. Cambodia has been one of the best achievers in terms of growth over the last decade (1998-2007). However, many indicators suggest that governance is weak and the 2008 investment climate survey, like the 2004 survey, found that corruption is the main concern of businesses in Cambodia (and presumably an even higher concern for potential investors). Unpacking the many forms of corruption, their incidence, and firms' response helps to explain why growth has been strong, despite firms' concerns about corruption. First, the incidence of corruption varies across types of firms. Smaller firms, for instance, are less exposed to some bureaucratic corruption, limiting the cost to them. Second, firms have developed risk prevention strategies to limit their exposure to corruption, for instance avoiding risks (i.e. avoiding going to court and instead using other mechanisms of dispute resolution), remaining in the informal sector (or under-reporting revenues, or negotiating specific arrangements (e.g. through business associations as in the garment industry). Third, firms have developed risk coping strategies, such as using private sector intermediaries (in fact, in many cases, reported concerns about corruption might refer to these intermediaries rather than to bureaucratic corruption). These various risk prevention and coping strategies suggest why the private sector continued to grow impressively for ten years despite its concerns about corruption. But these strategies also show the constraints that result from corruption. First, risk aversion strategies reduce efficiency and investment, in particular, investment in diversification and in more capital intensive enterprises. Foreign investors appear to be especially deterred by the absence of a good court system and the perception that corruption is common. Second, corruption is one of the factors explaining the large size of the informal economy. Informality is a key risk avoidance behavior, and also a key barrier to higher productivity, diversification, investment, and Government revenue collection. Third, corruption adds to the cost of doing business in Cambodia. In part, these costs of corruption are offset by generous tax holidays, but this reduces taxation and, in turn, the Government's financial capacity to improve skills and infrastructure. xi should continue to be paid to foreign investors who have, until now, been the main source of diversification and growth in Cambodia. Goal #2: Reduce business owners' concerns. For established firms, in particular for foreign investors, several costs should be reduced to lower uncertainty for business owners. Maintaining macroeconomic stability continues to be critical. Improving commercial dispute resolution, as the Government is doing with the launch of the National Arbitration Center in 2009, also appears to be an obvious priority. Some financial sector reforms initiated by Government (such as the Credit Information System and the Secured Transactions Law) will also let entrepreneurs focus on normal business risks, rather than undue efforts to secure financing. Regulatory reform ­ through improving transparency, and greater simplification of procedures ­ would reduce costs too (and, in particular, remove barriers that discourage some informal firms from registering). Goal #3: Improve logistics. A major way to reduce costs and uncertainty is to accelerate the Government's successful efforts in trade facilitation, in particular through rolling out risk management and creating the National Single Window (based on the ASYCUDA platform). More information about processes and fees would also pave the way for large as well as small investments. Encouraging the development of logistics service providers would complement the Government's ongoing efforts on trade facilitation. This will benefit not only exporters, but also domestic trade and services such as hotels and restaurants. Goal #4: Strengthen labor productivity. For better performing firms, improving labor productivity is becoming a priority. Two immediate strategies consist of improving pay practices to stimulate productivity and improving industrial relations to avoid losses due to slowdowns and strikes. The scaling-up of vocational and on-the-job training should also be a priority, although the full impact of this will take time, given the Government's broader efforts to improve the education system as a whole. In particular, developing vocational training and apprenticeships could have a significant pay-off. Goal #5: Increase participation of the private sector in identifying and solving problems. The private sector can be mobilized to resolve market failures by working through the eight private sector working groups of the Government-Private Sector Forum, provincial-level public/private dialogues and Cambodia's increasingly numerous business associations. 24. Specific short- to medium-term priorities are identified in Chapter 3 for each of these goals. Five Lessons from Change and Stability 25. Finally, it is important to take note of lessons learned in reforming the business environment in Cambodia (and other countries). Although the top constraint (corruption) was the same in 2004 and 2008, the ranking of the next five constraints has changed. The top constraint and related issues of economic and regulatory policy uncertainty (down from 3rd to 4th in rank) and anti-competitive and informal behaviors (up from 4th to 3rd in rank) remain important concerns. But rankings have improved with regard to crime and theft (down from 2nd to 6th in rank), conflict resolution (down from 5th to 11th in rank), and customs and trade regulations (down from 6th to 17th in rank). In parallel, macroeconomic uncertainty (up from 8th to 2nd in rank) and electricity (up from 10th to 5th in rank) ­ and to a lesser extent, the cost of finance and lack of skills ­ are more severe concerns now than they were in 2004. 26. These changes (or lack of change) provide five valuable lessons for the future development of the private sector in Cambodia. Some changes were driven by factors that were external to the business environment ­ for example, the reduction in concerns about crime and insecurity, or the rise in concerns about macroeconomic uncertainty. But many other xii changes ­ and even more those things that have not changed ­ suggest deeper constraints that demand a fresh look at how to achieve reform. 27. First, change is possible when driven by Government leadership, external commitments, and private sector demand. The agenda derived from the accession to WTO is a good example of a number of reforms achieved by the Government in recent years (such as the Law on Secured Transactions, which enabled Cambodia to raise its Doing Business ranking from 150th to 135th in the 2009 report). Contributions from the private sector, in particular the garment industry, and analytical work from development partners on trade facilitation, also seem to be paying off. ASEAN members are also exerting pressure in a number of areas. With Government encouragement, G-PSF private sector working groups have been another effective promoter of reform, even in contentious areas such as labor regulations. 28. Second, a sound approach to reform is piloting, and when successful, scaling up. China is increasingly seen as a model for reform because it tests reforms on a small scale and then replicates broadly when pilots prove successful. Cambodia, through the creativity of many of its leaders and extensive external assistance, has piloted a number of schemes, which can now be evaluated and, when successful, scaled up. 29. Third, having a coherent reform agenda, with a time-bound action plan and simple coordination and monitoring mechanisms, has been critical in achieving reforms. The Government's 12-point action plan on trade facilitation provides a good example of well- coordinated reform. But, mainly the customs automation component, which had its own coordination and monitoring structure, has moved steadily. This demonstrates the need for clarity about implementation arrangements due to the overlapping mandates of different Government agencies. To move some important successes beyond sector-specific achievements, a more comprehensive set of reforms is called for. Experience shows that reforms which level the playing field drive diversification and eventually, growth. Priorities for reform should therefore begin with those reforms that have been on the agenda for some time (for example, in the areas of setting standards and testing). 30. Fourth, among core reforms are many governance issues, in particular the need for public sector reform. Achievements in the area of public sector reform have resulted from the Government's commitment to simplifying procedures, increasing transparency, and increasing accountability to the private sector. However, the impact has in some areas been weakened by Government's limited technical and institutional capacity. The Public Financial Management Reform Program provides an example of successful civil service, human resources management and pay reform (through the Merit-Based Performance Initiative, MBPI). Additional promising instruments and pilots are being developed in other sectors, and these need to be packaged in an ambitious reform program that targets private sector problems. 31. Finally, since a one-time study cannot go into depth, it is important to encourage a good policy-making/monitoring system. A report like this is unable to address all possible problems, and also problems change quickly. For example, concerns about electricity (and most recently, skills) are long standing in Cambodia, but have lately become more pressing as other concerns are alleviated. The G-PSF is an obvious place to identify problems and possible solutions and has had well-documented success in achieving reform. To that end, it requires greater support for conducting research and effectively advocating reform. Seeking feedback on problems from business associations and SEZs, or from the Cambodian Investment Board and the Trade Promotion Department could also prove valuable. Within Government, broader public sector reform (see above) that clarifies roles and responsibilities would contribute to good policy-making and improved capacity to implement reforms. The proposed Sector Wide xiii Approach (SWAp) for trade will help coordinating development partners' efforts and support some of the strategies proposed here to accelerate reform. 32. The need to accelerate progress toward a better investment climate has rarely been so urgent. Experience from the past decade of rapid growth has proven that a good investment climate ­ e.g. in the garment sector ­ does lead to faster growth. But rapid deterioration in the external economic environment is making profound improvements in the investment climate a critical necessity to sustain growth. With so much risk in the global economy, problems in the business environment must not create further uncertainties. Pursuing the five goals outlined above, bringing all stakeholders on board, and monitoring accomplishments and impact, will go a long way towards improving the investment climate. Given the country's good potential in under developed sectors such as tourism, agriculture, and minerals, with fewer barriers, lower costs, and more support services, Cambodia should be able to attract greater investment, diversify into more sectors, and most importantly, create more jobs for Cambodia's fast growing labor force. xiv 1. The Private Sector in Cambodia Key Messages Although quickly modernizing, the private sector remains largely informal and composed of small firms. Labor productivity remains below that of comparators, except Bangladesh. This reflects, in large part, low capitalization, while efficiency is, on average often close to competitors (at least in the garment sector). Wage levels are comparable to those of competitors. There are significant variations in performance within each sector, with the average firm only 50 percent as effective as the top-performer in its sector. Some well-established sectors (garments) or emerging sectors (food packaging, telecoms and finance) provide valuable insights into the important role Government can play in promoting growth in Cambodia. 1.1 Cambodia has established a remarkable track record over the past decade, growing at an average of 9 ½ percent per annum over 1997-2007. With a small public administration and almost no state-owned enterprises, this growth has been entirely driven by the private sector. Most growth has been in the industry and services sectors, accounting for 3 ½ and 4 ½ points of growth per annum, respectively (against 1 ½ for agriculture).1 This chapter reviews the structure of the private sector in industry and services, with a particular focus on the formal sector. Section A describes the key features of the private sector, in terms of size, sectors, integration with global markets, etc. Section B takes a look at various measures of performance to benchmark Cambodia's firms across sectors and with key country comparators. Section C draws lessons from a few sector case studies. A. A Largely Informal Private Sector with a Small Modernizing Formal Sector 1.2 Cambodia's private sector is primarily small and informal (Table 1.1).2 Of the 63,500 establishments identified by IFC in 2008,3 67 percent employ only the owner, 22 have between one and four employees, and 1 ½ percent have more than 20 employees. Even 1 The contribution of services includes the growth of tax on products, net of subsidies. Cf. national accounts. 2 Statistics on firms are extremely limited in Cambodia. The SME Secretariat has made an attempt to collect statistics on small and medium enterprises but faced difficulties due to the fragmentation of licensing requirements. The registry of the Ministry of Commerce has improved, but still provides little data on the population of firms, as evidenced by the ICS sampling frame (Annex A). The National Institute for Statistics has included a census of firms in its Statistical Master Plan. 3 For the second round of the Provincial Business Environment Scorecard survey (PBES), IFC and The Asia Foundation conducted a census of establishments from May to July 2008. Although this was not a countrywide (town-by-town) survey, it provides a good indication of business activities in Cambodia as it covers all provincial capitals and ten additional major cities in the ten most economically-active provinces. The PBES covers `establishments', as opposed to `firms' in the ICS (one firm can have several establishments). 1 within firms of five employees or more ­ those covered by the 2007 Investment Climate Survey (ICS, cf. Box 1-1) ­ firms with fewer than 20 employees, comprise 71 percent of firms and those with more than 100 employees account for only 7 percent of firms. In line with Cambodia's level of development, the private sector has many features of an informal sector: (i) limited tax compliance (78 percent of firms consider that a typical firm reports less than 100 percent of its income, against a median of 43 percent in developing countries); (ii) poor accounting practices (14 percent of firms with audited financial statements, against 47 percent in developing countries); and (iii) limited use of the banking sector (10 percent for investments and 13 percent for general expenses, against a median of 23 and 23 percent, respectively, in developing countries). 1.3 The tourism and Table 1.1: Firm Distribution (among those with 5 trade sectors are among employees or more) the largest sectors, in % in terms of number of firms % GDP (*) terms of numbers of Small Medium Large Total firms. Although the Manufacturing 4.3 1.7 2.7 8.8 28.2 garment sector represents a Trade 23.3 4.5 1.0 28.8 14.5 larger part of GDP, the Tourism 18.4 5.8 0.4 24.5 7.0 Other 25.2 9.7 2.9 37.8 50.3 manufacturing sector Total 71.2 21.8 7.0 100.0 100.0 accounts for only 9 percent (*) as a proportion of GDP for private industry and services (i.e. excluding of firms. Most of the large agriculture and public administration). "Tourism" measured only with firms are in the garment "Hotels and Restaurants". Source: Based on sampling for ICS 2007, see sector. Within the whole Annex C. NIS for 2007 National Accounts. population of firms covered by the Provincial Business Environment Scorecard (PBES), commerce and services represent 48 and 43 percent of these establishments, respectively. Hence, there are very few manufacturing firms. 1.4 In the formal sector, a few sectors have emerged over the last decade (see Section C). The garment sector is the best known, with more than 300 firms employing more than 300,000 employees in 2008. Garment manufacturing accounts for 16 percent of GDP and some 80 percent of merchandise exports. In the food packaging sector, some more capital intensive industries have recently emerged. The policy of setting up Special Economic Zones has led to foreign investment in footwear and assembly factories (bicycles and soon motorbikes). In the services sector, the telecoms and financial industries have grown 18 percent and 16 percent, respectively, over the last decade (against 10 percent for overall GDP between 1997 and 2006). At the frontier between the formal and the informal sector, tourism remains a key sector in Cambodia. 1.5 The formal sector ­ which is Figure 1.1: Foreign Direct Investment the focus of the 2007 ICS ­ exhibits a 10.1 % GDP number of modernizing features. 1,000 900 10.0 9.0 8.4 1.6 First, Cambodian firms are 800 8.0 increasingly connected to global 700 6.0 7.0 markets (Figure 1.1 and Figure 1.2). 600 500 6.0 5.0 4.4 More firms have foreign ownership 400 4.0 (from 16 to 19 percent between 2003 300 2.8 3.0 and 2007, which is above most 200 <= US$ 1.6 2.0 comparator countries).4 Foreign Direct 100 million 1.0 0 - Investment (FDI) net inflows, an 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 important indicator of the business Source: NBC and NIS. environment, peaked at 10.1 percent of 4 Throughout the analysis, firms are classified as having foreign ownership if the share of foreign capital in their capital is strictly positive. 2 GDP in 2007 (above US$800 million). While still concentrated in hotels and garment factories, and with investment mainly from countries in the region, there are some modest signs of diversification. Since 2003, this acceleration has been supported by political and macroeconomic stability, and by the 2003 amendment to the Investment Law, which provides tax incentives and ensures equality of treatment for domestic and foreign investors. The proportion of imported inputs has also increased (from 33 to 43 percent in the manufacturing sector, and to almost 90 percent in the garment sector). The garment sector, however, continues to dominate exports. In addition, the growth of imported inputs highlights the limited capacity of the economy to capture larger parts of value chains (e.g. in garments, fabrics continue to be almost entirely imported). Figure 1.2: Connection to Global Markets a/Firms with foreign ownership (%) b/ Imported inputs in manufacturing (%) Cambodia Cambodia 2007, 19.1 2007, 42.7 Cambodia 2003 Cambodia 2003 Vi Th La La Vi - 10.0 20.0 30.0 40.0 50.0 - 20.0 40.0 60.0 80.0 100.0 Note: the box represents the 25th and 75th percentile of the distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. Cambodia's performance in 2003 and 2007, as well as those of Lao PDR, Thailand, and Vietnam are also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. 1.7 Second, modern ITC technologies have gained ground significantly in recent years. Some 56 percent of firms use email to contact suppliers and clients (against 41 percent in 2003 and a median of 54 percent in developing countries) and 39 percent of firms have a website (against 23 percent in 2003 and a median of 32 percent in developing countries). This has been supported by rapid development in the telecoms sector (see Section C). Beyond the potential positive impact on productivity, this trend also enables a number of very valuable e-government initiatives (Chapter 2 highlights a few of those, including the computerization of customs). 1.8 Third, the quality of labor is also gradually increasing. Although improving, education levels remain low (one-third of employees have grade 12 or more, against 16 percent in 2003, and 16 percent have less than grade 6, against 21 percent in 2003). Formal training is now offered in 48 percent of the firms (twice as high as in 2003 and above comparator countries, Figure 2.4). 1.9 Finally, Cambodian firms also have a high proportion of women in their labor force. The proportion of women is 37 percent on average (and above 80 percent in the garment sector). The proportion of women in management positions is reported to be 30 percent, significantly above other developing countries. Cambodian women entrepreneurs also make significant contributions to the development of Cambodia through their ownership of private enterprises. Based on indicative data from a census of Cambodian businesses in the major cities of Cambodia's 24 provinces, 62 percent of private enterprises are owned by women.5 Considering that job creation in rural areas is critical for poverty reduction, women-owned businesses could become important job creators if they faced fewer challenges to growth. 1.10 However, for a variety of reasons, women-owned businesses face more challenges than is the case with businesses owned by men. In addition to the regulatory and business operation challenges faced by all business owners, women say they face 5 Cf. IFC, Voices of Cambodian Women Entrepreneurs, (September 2008) and Census of 2008 Provincial Business Environment Scorecard, see footnote 3). 3 additional challenges in starting, operating, and expanding their businesses. For example, as a result of women's usually lower levels of education and training (especially for women over 40), women say they find it even more challenging to access and interpret laws and regulations. Lower levels of education also make it more difficult for women to keep the type accounts banks require of would-be borrowers. In addition, women say they find it harder to negotiate with male authorities and to gain respect from male employees. Women say as well that their unequal share of family responsibilities is a barrier to success as it prevents them from networking and other business development activities that take place outside working hours. Conservative notions about appropriate female behavior are one more challenge as they make it difficult for business women to join the `old boy's network' and use this to get ahead. With challenges such as these, it is not surprising that women-owned businesses usually remain small and largely in the informal sector. Thus, it is important for Government to improve the enabling environment for women entrepreneurs. Among recommended policy actions are making regulations easier to understand, easily available and simplifying business registration, tax payment and exporting procedures. 1.11 As expected at Cambodia's level of development, the economy remains largely informal ­ primarily because the benefits of `being formal' are unclear. Based on 2006 PBES data, 42 percent of establishments are not registered; 26 percent do not pay taxes; and 34 percent have three employees or less (with these three possible definitions of informality being significantly correlated). Overwhelmingly, non-registered firms view registration as Box 1-1: Analytical Tools Supporting this Report The 2007 Investment Climate Survey The main instrument for this report was the 2007 investment climate survey. It was initiated in August 2007 and completed in early 2008, with further work until May 2008 to clean the data and calculate weights. The sampling methodology was based on a double stratification: (i) by sector (manufacturing, trade, tourism, and others); and (ii) by size (small: 5 to 19 employees; medium: 20 to 99 employees; and large: 100 employees or more). The approach also ensured the sampling of a large number of exporters (more than 120). The survey focused on five key urban areas: Phnom Penh (including greater Phnom Penh), Battambang, Siem Reap, Kampong Cham, and Sihanoukville. It only included establishments that were in operation in April 2006 and had five or more full-time employees at that time. In contrast with the 2003 survey, weights were calculated to make averages nationally representative (and to correct for the oversampling of large firms and the garment sector). Locating firms turned out to be very difficult because of limited data for the sampling frame (mainly the registry of the Ministry of Commerce). Only about 235 (or 6 percent) of the 4,072 randomly selected firms were successfully interviewed. In addition, only 86 firms interviewed for the 2003 survey were re-interviewed. Finally, additional garment manufacturers, tourism firms, and medium and large businesses active in business forums were interviewed to ensure adequate coverage of these segments. The questionnaire was essentially based on the 2003 questionnaire to enable easy comparison. Despite attempts at simplification, it remained rather long and future surveys should aim to simplify the questionnaire. Although the quality of the data is rather good, those collecting data faced a number of difficulties related to the business environment: (i) absence of an accurate, updated business registry to locate firms; (ii) lack of management systems in firms to record financial and operational data; and (iii) firms' concerns about the confidentiality of data, especially financial data, and of their judgments of Government performance. Doing Business Indicators The Doing Business methodology consists of an assessment of 10 dimensions of the business environment. Analysis is done annually for more than 180 countries. It is largely based on case studies (e.g. what is the cost of tax for a given type of business in each of the countries around the world), as opposed to surveying firms. Other Sources In addition, the analysis builds on investment climate surveys available in many countries (see www.enterprisesurveys.org, including the survey done for Cambodia in 2003). Econometric analysis is based on data from Bangladesh (data for 2007), China (2003), Indonesia (2003), Lao PDR (2005), the Philippines (2003), Sri Lanka (2004), Thailand (2004), and Vietnam (2005). The analysis also uses the 2006 survey conducted for the Provincial Business Environment Scorecard (PBES, see IFC, 2007), as well as the listing of firms undertaken for the 2009 PBES. Finally, the report builds on case studies discussed in Section 1.C and on direct private sector feedback from various consultations. See Annex C to E for more details. 4 "not necessary" (84 percent, against 7 percent for "too expensive", 5 percent for "too complicated", and 4 percent for "would bring attention to business from local authorities"). Of those who registered, 32 percent sought help to complete registration, two-thirds of which asked an individual working for the provincial or district governor's office. 1.12 Looking at the drivers of informality (Table 1.2): Size indeed matters (the smaller the establishment, the higher the likelihood of being informal) and informal firms are often young firms. There are significant differences across provinces, highlighting the potential for provincial-level administrations to make a difference. Firms with plans to expand are also less likely to be informal; in turn, firms in provinces with a weaker business environment are less likely to have plans to expand (cf. PBES 2006). Manufacturing Table 1.2: Drivers of Informality firms are less likely Independent Variable to be informal, Not No tax Micro firm (1- registered payment 3 emp) (+) everything else being equal. This Size 0.328 ** 0.416 *** 1.192 *** suggests the need Age -0.014 -0.028 ** -0.044 *** for capital Manufacturing -0.385 * -0.141 -0.248 Severe obstacle? investment and for Tax admin. about consistency Perception 0.295 -1.306 *** -0.001 -0.180 -0.140 0.022 help with contract Perception about connections 0.070 -0.055 -0.117 enforcement for Bank loan -0.446 -0.375 -0.532 ** these firms. Use of email -0.589 ** -0.441 * -0.651 ** Plan to expand -0.282 ** 0.101 -0.005 Firms with access Control for region Yes Yes Yes to loans are also less Constant -0.062 -0.770 ** -0.662 ** likely to be Observations 500 500 500 Results of a probit regression with 3 different independent variables. The informal. variable on size is equal to 1 for firms with fewer than 3 employees and 0 Generally, otherwise. (+) the third definition refers to the size of the firm when it significant at the perceptions about started operations. *, **, *** forbased on PBES data. 10%, 5%, and 1% levels respectively. Source: ICA analysis the advantages of political connections and about consistent application of policies do not appear correlated to informality. In a separate analysis, using the 2007 ICS, informal firms indicated they are less likely to find electricity and corruption to be severe constraints. B. Performance of the Private Sector 1.13 Section B reviews the performance of the Cambodian private sector. The initial performance measure is labor productivity. The distribution of labor productivity, between wages and a mark-up for the investors, is also analyzed. Drivers of labor productivity ­ i.e. capital intensity and total factor productivity ­ are then reviewed. The Section concludes with a discussion on private sector performance with regard to diversification. Labor Productivity and Wages 1.14 At the macroeconomic level, growth has been rapid and sustained over the past decade (9 ½ percent over 1997-2007). GDP per capita has increased, although it remains significantly below comparator countries. Labor productivity in the industrial and services sectors is now closer to comparator countries (with labor productivity in agriculture further behind). This growth in income per capita has, however, not been driven by significant increases in investment, which remained around 20 percent of GDP over the 10-year period. Exports have increased significantly (18 ½ percent per annum since 1998), albeit from a low 5 base, but diversification remains limited (the top five products still account for 68 percent of exported goods). 1.15 At the microeconomic Table 1.3: 2005/06 Labor Productivity (2004 US$) level,6 mean labor productivity Mean Median N ­ measured as value added 2005 2006 2005 2006 1,565 1,727 936 1,127 72 divided by the average Manufacturing Trade 1,991 1,727 1,498 1,317 62 number of employees ­ Tourism 1,414 1,427 1,115 1,108 94 amounted to around US$1,900 Other services 2,369 2,348 2,011 2,011 62 Total 1,941 1,873 1,460 1,408 290 in 2005 and 2006 (cf. Annex E Export status for a discussion on the definition Exporter 2,689 2,678 2,349 2,321 69 of productivity, the specific Non-exporter 1,875 1,801 1,426 1,372 220 FDI status methodology used, and some FDI 3,437 3,112 2,298 2,130 93 detailed results). The median Non-FDI 1,585 1,578 1,228 1,233 197 labor productivity is significantly Firm size Small (1-19 employees) 1,599 1,588 1,284 1,342 124 lower, reflecting that: (i) smaller Medium (20-99 employees) 2,963 2,749 1,746 1,859 85 firms are less productive than Large (100 or more employees) 2,081 1,903 1,801 1,580 81 medium and large firms and (ii) Note: see Annex E for details. the distribution of firm size is skewed toward small firms. A number of groups have above-average productivity: (i) non- tourism/non-trade services in general, travel services, food manufacturing, and wholesale trade; (ii) exporters; (iii) foreign-owned firms; and (iv) medium, and to a lesser extent, large firms. 1.16 These differences are largely Table 1.4: 2005/06 Value Added per Worker, reflected in wage differentials ­ in Wages per Worker, and Mark-Ups (2004 US$) other words, firms with high VA / Wage / Mark-up productivity typically also pay higher worker ($) worker ($) (%) wages (either because marginal Manufacturing productivity is also high and wages Food 2,650 1,103 58.4 Garments 1,993 996 50.0 reflect marginal productivity, with well- Other manufacturing 1,315 1,126 14.3 functioning labor markets, or because Trade workers share part of the rent of highly Wholesale 3,501 1,829 47.8 productive firms). The annual average Retail 1,415 1,082 23.5 wage was US$1,343 ­ slightly above Tourism US$100 per month (Table 1.4). Hotels and restaurants 1,191 1,159 2.7 1.17 Wages were somewhat higher Travel services 2,140 1,566 26.8 in the wholesale trade and transport Other services Transport 3,498 2,880 17.7 sector, and lower in the garment Other 2,032 1,844 9.2 sector. In part, these differences reflect Total 1,907 1,343 29.6 disparities in the skills' structure. In Note: see Annex E for details. garments, the proportion of the labor force with education below grade 6 is above 55 percent, with only 3 percent above grade 12. These proportions are 11 and 26 percent in trade, 22 and 24 percent in tourism, and 11 and 51 percent in other services. The other explanation is that there are different mark-ups across sectors. 1.18 Indeed, mark-ups ­ defined as labor productivity minus wage per worker, divided by labor productivity ­ vary widely around a mean of 30 percent. Markups are particularly high in the food industry, garment industry and in wholesale trade, and 6 For microeconomic analysis, the comparator countries include Bangladesh (data for 2007), China (2003), Indonesia (2003), Lao PDR (2005), the Philippines (2003), Sri Lanka (2004), Thailand (2004), and Vietnam (2005). All data are converted in 2004 US$, using a manufacturing price deflator and the 2004 US$ deflator. In some parts of the analysis, countries are excluded due to lack of comparable data. 6 particularly low in hotels, restaurants and other services. These differences could reflect different degrees of capitalization and competition across these sectors, with higher capital investment, higher rents, and weaker competition in sectors with high mark-ups. The extremely low mark-up in hotels and restaurants is surprising (especially given the high capital per worker ratio in the sector) and could be driven by two factors: (i) significant investments have been made, but occupancy rates remain low, hence generating low mark- ups; and (ii) the opportunity cost of finance was low for the investors, who are not expecting significant mark-ups. 1.19 Compared to other countries, Cambodia is trailing in terms of labor productivity, but is closer to average in terms of wages (Figure 1.3). Labor productivity across the economy is above that of Bangladesh and below all other comparator countries (though close to Lao PDR, Sri Lanka, and Vietnam). However, the gap in terms of wages is smaller ­ i.e. wages in Cambodia are relatively less low than is the case with labor productivity. This could signal concerns about the competitiveness of Cambodia. (see also Chapter 2). Figure 1.3: Labor Productivity and Wages Across Countries (2004 US$) a/ Across sectors b/ In garments 7,000 Mean labor productivity ($) 6,000 Mean labor productivity ($) China Thailand 6,000 Thailand 5,000 5,000 4,000 4,000 Indonesia Philippines 3,000 China Average Philippines 3,000 Average Vietnam Cambodia Sri Lao {DR Lanka 2,000 Indonesia 2,000 Vietnam Lao PDR Bangladesh Cambodia 1,000 1,000 Bangladesh Sri Lanka - - - 500 1,000 1,500 2,000 2,500 3,000 3,500 - 500 1,000 1,500 2,000 2,500 3,000 3,500 Mean wage per worker ($) Mean wage per worker ($) Note: see Annex E for details. 1.20 Some of the differences could result from differences in industrial structure. Looking at the garment sector only, Cambodia has more competitive labor productivity7 ­ comparable to Vietnam and Indonesia, above Bangladesh and Sri Lanka, although below China and Thailand. Wage differentials are consistent with labor productivity ­ with the significant exception of China, where labor productivity is very high but wages are at a level found in countries with much lower labor productivity (such as Cambodia and Indonesia): hence, in the garment sector, China has lower wages than Cambodia, but significantly higher labor productivity. This difference is visible in the mark-ups that are significantly higher in China (and to a certain extent in Vietnam) than in Cambodia. 1.21 The finding that labor productivity is similar in Cambodia and Vietnam, is somewhat at odds with the general perception in the sector. This finding might result, in small part, from a measurement error. As explained in Annex E, the measurement of value-added for Cambodia is based on output, while it is based on sales for other countries, including Vietnam. Although an increase in inventory is unlikely to be significant in a zero- inventory industry like garments, it could explain part of the difference. This finding might also be explained by the existence of a few high performers in Cambodia. While many Cambodian garment establishments would be struggling against Vietnamese competitors, on average, the performance of the industry might be better because of these high performers. Finally, the difference between labor productivity and competitiveness means that despite similar labor productivity, Cambodian firms might still be disadvantaged with regard to 7 It should be noted, however, that this Section focuses only on productivity, not on competitiveness. Labor productivity is an important ingredient of competitiveness, but far from the only one. Other elements include the cost of sourcing fabric and exporting; the capacity to deliver orders on time; high quality standards; etc. 7 competitiveness because Vietnamese firms perform much better on other elements of competitiveness (such as the cost of raw materials and logistics, as well as quality and timeliness). Capital Investment 1.22 Macroeconomic data and anecdotal evidence suggest that the Cambodian private sector remains weakly capitalized. This is evident from the weak investment to GDP ratio over the last decade, as well as from anecdotal evidence highlighting the limited number of capital-intensive firms in Cambodia. Figure 1.4: Labor Productivity vs. Capital Intensity in the Garment Sector 10 9 8 9 7 8 log YL log YL 7 6 6 5 2 4 6 8 10 2 4 6 8 10 log KL log KL Cambodia Fitted values Cambodia Fitted values Thailand Fitted values Indonesia Fitted values 9 9 6 7 8 7 8 log YL log YL 6 5 4 5 4 6 8 10 2 4 6 8 10 log KL log KL Cambodia Fitted values Cambodia Fitted values Sri Lanka Fitted values Philippines Fitted values These four charts plot value added per worker (log YL) against capital per worker (log KL) for four countries and Cambodia. Each chart includes the distribution of observations and the fitted value of a linear regression. See Annex E. 1.23 Measuring a capital stock is however very difficult, making cross-country comparisons difficult. Across sectors, garment firms are the least capitalized, with a stock of only US$1,400 per worker (against US$9,600 across sectors). The ratio is particularly high in tourism (US$10,170). Capital intensity is also more than four times as high in high performing firms than in low performing firms. Surprisingly, across the firms surveyed in Cambodia and Thailand, the capital intensity (or capital per worker) is comparable in these two countries. This could, however, reflect different samples and industries covered by the surveys. By looking at the garment sector only, to address this bias, capital intensity is indeed lower in Cambodia. At US$1,036 in 2004 terms, Cambodia is below Sri Lanka (US$1,656), the Philippines (US$3,543), and Thailand (US$8,844) ­ only above Indonesia (US$627). 1.24 In the garment sector, by comparing labor productivity (value added per worker) to capital intensity (capital per worker), Cambodian firms also appear poorly capitalized, but quite efficient (Figure 1.4) consistent with the finding on TFP (Figure 1.6). Compared to Thailand, garment firms are much less capitalized (at a ratio of 1 to 8) and are less efficient (the fitted line is significantly below that of Cambodia: even an increase in 8 capital intensity would not bring Cambodian firms close to the labor productivity in Thailand). Cambodian firms are, through this view,8 more productive than Indonesian firms by being more capital intensive, but with little difference in efficiency. This view suggests that Cambodian firms are more efficient than Sri Lankan firms. The gap with the Philippines is to the advantage of Cambodian firms at low levels of capital intensity, but disappears at higher levels. 1.25 Taking a forward looking approach, the private sector had in 2007, a very optimistic view of the future. Some 83 percent of firms intended to increase their investment in the next three years (37 percent intended to "increase drastically"), with only 5 percent planning to close or reduce their investment. Even firms in the garment sector, which were already subject to intense competition and suffering a downturn in their main export market, the United States, were optimistic, with 77 percent planning to increase investment. Interestingly, non-registered firms (surveyed by the 2006 PBES) were less likely to plan on increasing their investment, highlighting that informality is a barrier to growth. The relatively low level of capacity utilization could also be used to generate growth (the level is 66.4 percent in the manufacturing sector, against 69.1 in 2003, and a median of 72.8 in developing countries). However, the worldwide economic crisis in 2008 has most likely tempered this optimism. Technology and Performance 1.26 Total Factor Productivity Figure 1.5: Productivity Gap between Mean and (TFP) is the most widely used Top Performer across Sectors measure among the multifactor 100% Top Performer in Each Sector productivity measures (as opposed 80% 66% 66% to a single factor productivity 60% 50% 47% 54% 49% 49% 43% 40% 45% measure, such as labor productivity). 40% TFP captures an efficiency effect in 20% the use of production factors that is 0% not caused by inputs or economies Retail Total Food Wholesale Garments manufacturing Other Transport Hotels and Travel services restaurants Other of scale. Since it is estimated by econometric methods as a residual Mean firm-level TFP across sectors, as % of TFP in best of an estimated production function, performing firm in the sector. See Annex E. it is less intuitive to interpret than simpler measures, and subject to a degree of uncertainty. TFP remains, however, a critical measure of technology and efficiency. A number of other measurement issues are discussed in Annex E: the results in this Section are based on a production function specified as a translog function, and TFP is estimated as the residual of sector-specific regressions, with a number of investment climate variables and firm characteristics to address simultaneity biases. 1.27 Across sectors, the key finding is that of significant disparities within each sector: TFP is, on average, only 49 percent of the best performer (Figure 1.5). Mean firm-level TFP is even smaller compared to the best performer in transport (40 percent) and garments (43 percent), highlighting inefficiency in the sector and potential productivity gains through more intensive competition (see next Section on garments). On the other hand, disparities are much less in travel services and other manufacturing (mean firm-level TFP is 66 percent of the best performer), possibly indicating more intensive competition in these sectors. 8 This method is different from the TFP analysis since it only compares labor productivity to capital intensity, i.e. it uses a very simplistic approach to the production function. Although less satisfactory, this approach is useful through the simple graphic representations it enables. 9 1.28 Across countries, focusing on the garment sector,9 10 Cambodian firms appear 15 percent less productive on average than firms in Thailand (Figure 1.6). Productivity is comparable to other countries ­ slightly better than the Philippines and slightly below Indonesia and Sri Lanka. The distribution of TFP also shows that the majority of Cambodian garment firms have productivity similar to that of Sri Lanka and Indonesia, but a minority of Cambodian firms have productivity closer to top performers in other countries (see, however, footnote 7 on the difference between productivity and competitiveness). Hence better incentives for instance through more competition, could generate productivity gains by pushing a higher proportion of garment firms toward the higher end of the distribution. Figure 1.6: Distribution of TFP in the Garment Sector a/ Average productivity gap (Thailand = 100%) b/ Density functions 100% 8% 9% 15% 19% 75% 50% 25% 0% Cambodia Indonesia Sri Lanka Philippines Thailand Mean firm-level TFP in garments across countries, as % of TFP in Thailand. The right panel shows the density function of the TFP in garments across 5 countries. Note: see Annex E for details. 1.29 In summary: Labor productivity in Cambodia is lower than comparator countries. This, in part, reflects differences in industrial structure. Within a particular sector (garments), Cambodia is closer to average labor productivity. This outcome reflects lower capital intensity in Cambodia than in more developed countries such as Thailand or the Philippines, but a level of efficiency that is closer to the average. One option for improving productivity is therefore to enable deeper capital intensity. Nevertheless, the dispersion of performance is significant and the top performers in Cambodia's garment sector are much closer to international standards. A second option for improving productivity is to strengthen the incentives for all firms to reach the level of Cambodia's best performers. Finally, across sectors, low capital intensity also materializes through a higher share of labor productivity being allocated to wages ­ with a lower-than-average mark-up. This might not be attractive for foreign investors and could raise competitiveness issues. Diversification 1.30 As noted above, an important feature of Cambodian private sector performance is lack of diversification. This is evident for exports, where the garment industry is the main sector. In addition, very little diversification has happened within the sector: despite some recent diversification out of the US market into the EU market, the types of garments made in Cambodia remain very narrow. Poor diversification also seems to be the case for the 9 Using economy-wide comparisons could be misleading by reflecting differences in industry structure, as opposed to differences in TFP. 10 Capital stock data from China and Vietnam are not directly comparable with those in other surveys, and these two countries are not included in these comparisons. 10 domestic sector. Although difficult to measure, there are evident symptoms (a large and growing share of imported inputs, Figure 1.2). 1.31 Lack of diversification is one of a number of risk avoidance behaviors. As will be discussed in Chapter 2, a number of features of the business environment create uncertainties that make investors wary of taking risks. Lack of diversification-oriented finance might also be a constraint (financing production of a new product or exports to a new market is a risky activity, which might not be easily financed, given the limited depth of the financial sector, see Chapter 2). 1.32 Poor diversification also relates to the large size of the informal economy. Both can be explained by the many drivers of risk avoidance (see Section A and Chapter 2). Another factor is the weak market linkage between the formal and the informal economy. 1.33 Another constraint to diversification is the need for predictability. Reaching new markets requires a reliable logistics chain (including financial and transport services and trade facilitation), which is not strong in Cambodia (Section B). These elements can be overcome, either through foreign ownership (bringing financing capacity and connection to global supply chains) or through a well-organized supply chain (such as the garment industry, that has reached the critical mass necessary to improve parts of its logistics chain). However, for most new local investors, the absence of financing and good logistics exerts a serious pressure on the process of diversification. C. Sector Highlights Garments11 1.34 The Cambodian garment industry has grown very rapidly over the last decade12 ­ but growth is decelerating. In 2007, its 300 firms contributed US$2.7 billion in exports (above 80 percent of total merchandise exports), US$1.0 billion in value-added (12 percent of GDP), and almost 350,000 jobs (5 percent of total employment).13 The sector grew at 40 percent per year between 1995 and 2006, but decelerated in 2007 ­ most likely a reflection of: (i) Vietnam's WTO accession (Vietnam has attracted significant foreign investment in recent years), and (ii) slower growth in the US market, Cambodia's main destination for exports. Another potential factor is the lifting in 2009 of safeguard measures imposed on China's garment exports by the US and the EU. This could further erode Cambodia's competitive position given China's advantages that result from major economies of scale, large production capacity, high productivity from recent investments in modern facilities, and abundant local textile production. However, some analysts suggest that the gradual increase in wages in China will erode its competitive edge. Another important development is the decline in prices observed in the global garment market, indicating increased competition and even greater pressures to lower production costs. Table 1.5: Costs for Denim Jeans and Polo Shirts (US$ / piece) 11 This Section is largely based on USAID (2007a), which uses a rigorous cost accounting method, based on six Cambodian garment firms, to analyze the cost structure and efficiency of the industry. 12 The industry's take off began after negotiations with the US that granted significant growth in Cambodian quotas in the US market if Cambodia adopted sound labor policies and demonstrated adherence to international standards (achieved through ILO's Better Factories Cambodia, the institution in charge of monitoring and reporting on industry labor standards). Although quotas ended at the end of 2004, in 2005, safeguard measures were imposed on Vietnam until January 2007 and on China until December 2008. 13 This is based on Cambodian data. Export partners report exports some 25 percent higher. USAID (2007a) calculated value-added in garments at around US$1.6 billion. 11 1.35 The industry has specialized Denim primarily in Cut-Make-Trim (CMT), Polo Shirt Jeans i.e. firms performing only cutting, Fabric, trims, thread and packaging 4.14 1.39 sewing, finishing and packaging on CMT (35% efficiency) 1.15 0.43 behalf of an overseas intermediary that Trade and inland transport 0.22 0.06 deals with sourcing fabric and FOB (Sihanoukville) 5.51 1.88 Ocean shipping 0.15 0.06 negotiates with buyers. Around a US import duty 0.91 0.32 quarter of establishments cover full Landed Duty Paid 6.56 2.26 operations, which include purchasing Note: CMT refers to `Cut-Make-Trim'; 35% efficiency refers the fabric, packaging, and shipping the to the labor productivity, see para. 1.37. Source: USAID orders to wholesalers or retailers (and (2007a). possibly sample making and direct negotiations with buyers). These activities capture a larger part of the benefits in the value- chain, but also entail significantly higher risks (e.g. financial risk if an order is not completed as planned) and require higher working capital. Some 15 percent of factories are sub- contractors, hence covering CMT operations, but not the importing of fabric and the exporting of products. This distribution, across CMT, full operation, and sub-contractors, is in part a reflection of Cambodia's factories being mainly subsidiaries of, or connected to, regional groups (half the firms are owned by Hong-Kong, Taiwan, or Chinese interests, with less than 9 percent of firms owned by Cambodians). This also reflects the cost structure of the industry. 1.36 Looking at the cost of cut, make, trim, trade and transport ­ i.e. excluding the cost of fabrics and trims (Table 1.5) ­ there are only four cost items that the firm can manage in this highly competitive environment (Figure 1.7): Figure 1.7: Standard Cut, Make, Trim, Trade and Transport Costs a/ Denim jeans b/ Polo shirts Inbound Export fees Inbound Export fees Outbound Outbound transport 8% transport 8% transport transport 1% 2% Import fees 1% 2% Import fees 3% 5% Embroidery (external) Direct 12% Labor Direct Labor Embroidery Finishing 38% 39% (external) 7% Rent 31% 5% Indirect Indirect Labor Labor Finishing Rent Energy General and 4% Energy 3% 0% 3% General and 7% 5% Admin Admin 10% 6% Source: USAID (2007a). The main cost (39 percent) is that of direct labor (for cutting, sewing, finishing, and quality control). The cost of unskilled labor was found lower in Cambodia than in China and Indonesia, slightly lower than in Vietnam, and higher than in Bangladesh. Factoring in other regulations (such as overtime, vacations and other benefits), Cambodia's wages appear more significantly below those of Vietnam (34 cents per hour versus 41 cents). However, this was more than offset by lower productivity (see 1.37). In addition, as noted in Table 1.4 above, firm surveys indicate higher average wages in Cambodia than in a number of other countries, highlighting the impact of indirect labor (management and supervision). Indirect labor accounts for 4 percent of the unit costs. These 20-30,000 managerial and supervisory jobs are in large part performed by foreigners as firms find it difficult to identify Cambodians with the skills and attitudes for such jobs. 12 Energy costs account for 5 to 7 percent of unit costs, a high proportion for an industry with limited demand for electricity. This reflects the high cost of power from the grid and a high proportion of self-generation, given the unreliability of the grid. Although the price of electricity from the grid has decreased since 2003 (50 to 18 cents per kwh, with small and medium commercial consumers paying 37 cents per kwh as they source electricity from subcontractors), grid reliability seems to have decreased and the price of diesel has doubled, hence increasing the cost of self-generation (42 to 78 cents per liter, JDI, 2007b). Trade and transport costs account for 13-16 percent of unit costs. Benchmarking against a 2003 value-chain analysis indicates some progress in reducing the costs of inspections and fees, but continued high uncertainty (GDS, 2003, and JDI, 2007b, see Figure 2.7). 1.37 Productivity is weak overall, but varies greatly. Efficiency ­ measured by comparing the time required for constructing a pair of basic five-pocket denim jeans as a percentage of the international standard for time ­ was in 2005, only 35 percent on average (92 minutes),14 however, with some factories operating in the 60-80 percent range, or 30-45 minutes (the normal high achievable production rate in a well-run factory). 1.38 There are several hypothesis behind the generally low efficiency observed in garment firms: Low productivity could result from a poor mix of labor and capital inputs. As discussed above, capital intensity is low in Cambodia. Although levels of employment do not seem the issue (in fact, according to the ICS, without constraints, garment firms, like other Cambodian firms, say they would raise employment by 11 percent), the skill level of workers is low, with 55 percent of the workforce having less than a Grade Six education. Poor individual productivity is somewhat offset by a typical worker's work year being significantly higher in Cambodia than in Vietnam (2,836 hours against 2,174 hours), but this comes at a cost in terms of workers' quality of life and opportunities for upgrading skills. Low productivity could result from the abnormally low utilization rate (57 percent in the garment sector, against 66 percent across the economy). This is due to significant disruptions (power cuts; strikes; etc.). According to GMAC, there are over 1,000 unions in the 300 operating garment firms, and a significant number of working hours are lost to strikes, and ­ perhaps even more importantly ­ strikes introduce an element of uncertainty that forces factories to decline orders. In fact, garment factories indicate that they refuse orders when uncertainty related to strikes is so high that they cannot guarantee delivery. Finally, low productivity might reflect poor management processes, such as weak quality control. Failure to control fabric quality before cutting can generate significant waste as defects are detected at the end of the production line. Poor cost accounting, and lack of attention to time lost throughout the production line, could be another problem. 1.39 Some progress in productivity has been made over the years, but could be jeopardized by recent developments. Since 2003, productivity gains of 15-25 percent have been achieved (around 5 percent per year, JDI 2007b). This is the result of shop floor technical training paid for by the factory owner. However, the recent sharp increase in consumer prices has led to: (i) an increase in the minimum wage; and, (ii) difficulties in retaining workers, given lower real wages. Rapid development of the garment sector in Vietnam plus the possible lifting of US and EU safeguards on Chinese imports at the end of 2008 could result in a serious decline in Cambodia's garment industry in 2009. 14That is only five pairs per worker per working day (8 hours), against 25 pairs per worker in China, 21 in India, 18 in Kenya, 14 in Uganda, and 10 in Ethiopia (data compiled by Global Development Solutions). 13 1.40 Better management of factory floor processes and capacity utilization rates, combined with more efficient pay incentives, would generate a double dividend. Paying workers by the piece and engineering an increase in efficiency from 35 to 55 percent would: (i) bring the average CMT, trade and transport of a pair of denim jeans down from US$1.4 to US$1.2 (a 12 percent gain) and (ii) double operator wages (part of these gains could be given up to free up time for skill upgrading). Such significant gains would come from spreading fixed costs (e.g. indirect labor or electricity) over more variable costs. In Vietnam, a substantial part of the salary is based on piece-rate incentive payments, while in Cambodia, only a quarter of firms pay such incentives and do so for a smaller part of the paycheck. Achieving this increase in efficiency would require better factory floor management, greater quality control (to reduce waste), and a higher capacity utilization rate. 1.41 Beyond such improvements, a more radical option for growth is upgrading the industry, but this is not happening much. Upgrading could come from finding new export markets beyond the US and EU (which only began significantly in 2007/08) and through initiating a number of value-adding activities in Cambodia. These include special finishing (e.g. jeans) and printing or embroidery (e.g. on shirts). More radical upgrading would involve structural changes in the value chain (e.g. by bringing the designing or financing of products to Cambodia or initiating backward integration by growing cotton and/or setting up textile mills). 1.42 A few factors explain why such upgrading has been limited: (i) radical upgrading is time and capital intensive (including greater demands for electricity), and perhaps this is not in the interest of establishments that use their foreign headquarters for design, financing, connection to buyers, sourcing, etc.; (ii) Cambodia lacks the economies of scale of China, Vietnam and other countries that would justify developing value-adding services; (iii) even smaller forms of upgrading would involve taking greater financial risks (and hence having access to the necessary low-cost capital to finance upgrading) and bringing down waste rates; and (iv) foreign managers are rarely trained in, or have few incentives from their headquarters, for process controls or industrial engineering. Instead they focus on maintaining production levels and can be risk adverse. Food Packaging15 1.43 The food packaging sector in Cambodia is only nascent. Limited investment in food packaging is viewed as a constraint for other sectors such as agribusiness and food processing, including for export. Trends and constraints in food packaging also have some economy-wide lessons. 1.44 The market for food packaging is already significant and growing fast, with demand in large part met by imports or low-quality domestic production. Supermarkets are growing fast in cities where a growing affluent population (especially those who are younger) prefers shopping in a brightly lit, clean and air-conditioned environment, rather than in less pleasant outdoor wet markets. Supermarkets have increasing need for packaging as they buy bulk food from local wet markets and package this for sale. They also import large packages and repackage in smaller units to fit local preferences. Thus, supermarket demand is increasing for shrink wrap, vacuum packaging, plastic shopping bags, foam trays, and labels. The market for bags, for example, is some 10 million bags per month. The market for aluminum cans is also growing fast, with demand estimated at around 470 million cans per year (350 for beer and 120 for soft drinks). The market for polyethylene (PET) bottles is very significant too, at some 200 million bottles (500 ml) per year. Aseptic packaging (Tetra Pak), which was introduced in 2003, is used by a soy milk company. However, the use of glass jars and hard transparent plastic containers remains limited, in part reflecting lack of knowledge about the potential for such packaging, as well as a range of technical issues (e.g. the capacity of local printing companies to produce quality labels). Demand for food packaging in Cambodia is 15 This Section is based on JDI (2007a). 14 expected to continue growing fast, driven by population growth, urbanization, and various changes in dietary habits. These changes in diet include a decline in per capita rice consumption; increasing consumption of wheat (typically in a processed form); and higher demand for temperate zone products. 1.45 Three types of food Table 1.6: Cost and Source of Packaging Material packaging are currently made Used by Major Supermarkets in Cambodia: (i) aluminum cans Country of production Cambodian Thailand Direct Cambodia Singapore Cambodian Local Producer Exporter for beer and other beverages Importer Import Informal Market (with one firm established in Shrink wrap (6x600mx45cm) 16.0 14.0 2007); (ii) PET bottles, primarily Vacuum pack ($/100 bags) 200g 4.0 5.0 for bottled water; and (iii) plastic 2 kg 8.0 10.0 shopping bags. The latter is an Plastic shopping bag ($ /2 color / 100 interesting example of spillover Foambags) box/tray ($/100 trays) 2.6 2.2 from the garment sector, which Black 7.5 6.0 has a high demand for plastic Label w/print ($/1,000 labels) White 2.1 6.0 20.0 25.0 packaging. Recently a local packaging firm moved into the Note: JDI (2007a). plastic bag market. These products highlight the challenge of an informal market, which can provide goods at a lower price (Table 1.6); the potential for domestic production to replace some simple imports; and the challenge of developing a regulatory framework for quality control. Figure 1.8: Value-chain for Aluminum 330ml Can (Body) Production in Cambodia Source: JDI (2007a). 1.46 The key costs drivers in the food packaging industry are imported raw materials, electricity, and labor (Figure 1.8 for aluminum cans). For the whole sector, all raw materials are imported (aluminum sheets, synthetic resins, ink, etc.).16 Given the absence of brokers/importers with an established overseas network, each company in the sector typically acquires an import license and registers as an importer for each of its necessary inputs, including parts for equipment. Firms also have to maintain a large inventory of these 16 For aluminum cans, the import of aluminum alloy, sourced from Korea or Australia, accounts for almost 76 percent of the total cost. Other imports (mainly chemicals for coating/drying and printing) account for 7 percent. A major cost driver was the decision to allow a zero-rated tariff on imports of aluminum alloy ­ although this is somewhat offset by high tariff rates on a number of other inputs such as chemicals. For PET bottles, imports of pre-formed bottles account for 79 percent of the unit cost. For plastic bags, imported plastic pellets account for 82 percent of the unit cost. 15 materials, with associated costs for warehousing and working capital. Enabling the establishment of support institutions and services to facilitate such imports should therefore be a priority. The cost of inland freight (from port to factory) is also high due to the high costs of container clearance and rising fuel costs for shipping (Figure 1.9). Factories highlight as well the lack of transparency and predictability of logistics services ­ a significant weakness compared to other countries in the region. 1.47 The cost, and quality, of Figure 1.9: Inland Freight Costs (US$/km/ton) infrastructure, and of electricity in 0.20 particular, is the next key cost 0.200 0.17 0.15 driver. The example of aluminum can 0.150 production is enlightening: 0.100 Capital cost: an upfront capital 0.04 0.06 0.06 0.050 investment of some US$2 million 0.02 0.02 0.02 was necessary. As none was - available, the factory financed a Cambodia Mozambique India Ethiopia Tanzania Kenya Pakistan Indonesia South Africa transmission line from the grid to the production site (and similarly financed water supply). Due to Source: JDI (2007a). unreliable electricity and water supply, the factory also bought a 3 MW generator, dug its own well, and installed several water tanks. Operating cost: the cost of electricity Figure 1.10: Cost of Electricity (US$/kWh) from the grid is high by regional standards (Figure 1.10). In addition, 0.2 0.18 the cost of electricity from a 0.15 business' own generation is also 0.09 high (operating the 3 MW generator 0.1 0.06 0.07 0.07 0.05 requires 300 l of diesel fuel/hour). 0.05 0.04 Operating costs for electricity and water account for 5 percent of total 0 Cambodia Taiwan PDR Lao China Vietnam Africa South unit cost (11 percent for PET Pakistan bottles). Maintenance cost: the factory has to Source: JDI (2007a). import parts itself for the generator (cf. above, on imports) and always maintains a set of spare parts. Recruiting and training qualified staff for maintenance also adds to the cost. Waste cost: blackouts occur 2 or 3 times a week, lasting between 1 and 8 hours. Since it takes a few seconds for the back-up generator to start up, the factory suffers substantial spoilage and wastage rates (some 10 percent of production). Despite repeated requests, the factory has yet to have EDC call 15-30 minutes before an expected power cut. 1.48 Labor, including overheads, accounts for 3 percent of unit costs for aluminum cans, and 10 percent for PET bottles. Poor institutional support for technical and vocational training, especially for mechanical, technical and administrative staff (bookkeeping, accounting) has increased costs. The aluminum can factory had to develop its own in-house training and send many of its staff for training in parent company factories overseas. The factory has also been forced to develop its own testing system, as it could not rely on the certificates provided by Government. Retaining workers, particularly trained workers, has been another cost. 1.49 In the food packaging industry, two examples highlight the potential for Cambodia to leapfrog more traditional technologies. Tetra Pak, which keeps drinks fresh without refrigeration, was introduced in 2003 by a soya milk producer because 16 Cambodia lacks reliable refrigeration, as well as bulk distribution (the shape of the packaging allows efficient stacking). Modified atmosphere packaging (MAP), which was developed in the early 1970s, also enables distribution of food products when a cold/chill chain is lacking. This reduces waste, increases marketability and increases revenues for farmers.17 MAP is being tested by one or two firms growing organic vegetables and herbs in northern Cambodia. 1.50 The packaging industry also highlights some of the regulatory constraints of Cambodia's business environment. A first example is the difficulty of getting Certificates of Origin. A streamlined process would be useful as agri-business firms could then include the value of packaging in their rules-of-origin calculations. Another example is the under- regulation of the PET bottle market. In the absence of a quality control framework, it is difficult to ensure good standards (hence, all major water bottlers import PET pre-forms) and quality issues raise a number of health concerns. This example also suggests the potential benefits of an association to develop some of these standards and advocate for appropriate regulations. Fast Developing Sectors: Telecoms and Finance 1.51 Telecoms is an interesting example of a rapidly growing sector (20 percent per annum growth since 2000).18 In the 1990s, fixed line phones grew rapidly, but slowed considerably in recent years as mobile phones took over. Cambodia has only 0.2 fixed lines per 100 habitants (against 34 in Vietnam, cf. Table 2.3). The fixed phone sector is dominated by a public company, Cambodia Telecoms, which also operates the connection to international calls. There are now 18 mobile phone subscribers per 100 habitants (against 28 in Vietnam), with the sector growing at 49 percent per year. CamGSM (Mobitel), created in 1996 through a joint-venture between Sweden's Millicom International and Cambodia's Royal Group of Companies, remains the leader in mobile phones in Cambodia. But competition has increased with new entrants (most with foreign investment) and competition in terms of services (3G, Voice over Internet, and mobile banking). Competition is also visible in the rapidly growing market of internet providers. In addition to growing demand, expansion in the telecoms sector was enabled by the allocation of several licenses for mobile phones and the development of a regulatory framework. 1.52 The financial sector has also been expanding recently (16 percent per annum since 2001). This is another example of rapid growth enabled by a sound and open regulatory framework. Since some rationalization of the sector in the early 2000s (through higher capital requirements), the sector has been growing rapidly and some innovation has been introduced. The strategies of two large banks highlight interesting directions. One bank, ANZ Royal, a foreign investor with a local partner, has introduced ATM machines and a more aggressive approach to reaching out to clients. Growth has been mainly in Phnom Penh and a few large urban centers. ANZ Royal's innovations have, in turn, spread to other banks and stimulated further competition and growth. Another major bank, ACLEDA, which specializes in micro-finance and SME lending, has become a commercial bank with a large network of branches across the country. Thanks to its origins in microfinance and its strengths, ACLEDA has been able to cater to underserved segments of the market, in particular, small businesses and agriculture. Through providing a range of services to clients, ACLEDA has been able to get to know them and their business needs well and tailor its lending accordingly. The result has been a high-quality and expanding portfolio, while at the same time providing access to finance for tens of thousands of poorer and rural individuals who previously had no access to finance and financial services. 17 MAP entails injection of nitrogen into a sealed plastic bag which displaces the oxygen, thus helping to preserve both the color and flavor of perishable produce. The "ballooning" of the package also helps prevent damage to a perishable when it is boxed for shipping. 18 See Naron (2007). 17 D. Summing Up 1.53 The foregoing analysis describes a number of key characteristics of the private sector and suggests the existence of five categories of entrepreneurs, with different potentials ­ hence requiring different approaches (Table 1.7). 1.54 Three of the groups of entrepreneurs should be the target for a growth strategy. First, as Sections B and C highlight, there are strong performers already in the economy. They have potential for further growth through productivity improvement, capacity expansion, and innovation. Second, part of the informal economy is vibrant and should be encouraged to join the dynamic formal sector through simplification of registration procedures and higher benefits to be realized from formalization (such as access to finance and useful Government services). Third, there are entrepreneurs that remain outside the market, for which facilitating entry is key. 1.55 Two other groups of entrepreneurs are less promising. In the formal sector, a group of firms is not performing strongly. Second, part of the informal sector is in fact a subsistence economy. Being critical to the livelihoods of many Cambodians, it should be encouraged as a safety net, but not be a major part of the growth strategy. Table 1.7: Different Entrepreneurs ­ Different Potentials Type of firms Characteristics Potential Approach 1. Formal sector: Usually exporters; often have Significant potential for Need to favor productivity good performers, access to finance and growth and innovation improvements (reduce costs productive and technology (through own and uncertainty; improve growing resources, banks, or foreign logistics; upgrade skills) and investors) capacity to expand 2. Poor Usually not exporters; coping Limited potential because Favor productivity performers from with a difficult business weak incentives to improve improvement through the formal sector environment and poor productivity competition, or exit management capacity 3. Small, vibrant Thriving on informal sources of Perceive costs and Increase registration (simplify informal sector financing, usually not procedures as the main processes; increase benefits exporters; growing rapidly, but barriers to becoming formal from formalization), to enable without diversification growth and productivity 4. Small, Basic shops and other activities Limited capital and know- Encourage mainly as social informal, with 1-3 employees max; how; unwelcoming business safety net, through lowering subsistence sector limited growth environment regulatory pressure 5. Outside market Entrepreneurs with good a Potentially high Facilitate entry (reduce idea, but not in the market barriers to entry; facilitate access to finance; facilitate support services) Source: based on Chapter 1 analysis. 1.56 Chapter 2 will analyze the constraints facing entrepreneurs in these five groups in order to prioritize strategic objectives for developing Cambodia's private sector. 18 2. Constraints to Productivity and Growth Key Messages: As in 2003/04, corruption remains Cambodian firms' main concern. Once unpacked, corruption translates into additional financial costs and unpredictability for most firms. While some sectors have been less exposed or have been more capable of dealing with it, corruption has had a cost in terms of low capitalization and lack of diversification. Significant progress has been made by Government in the area of trade facilitation for garment firms. This progress now needs to be generalized to other sectors to promote diversification. It also needs to be deepened to improve overall logistics in Cambodia. Issues of coordination and market failures are pervasive in Cambodia and should be addressed through a variety of policy responses. Labor market regulations are not viewed as a significant constraint (and could be an asset, as is the case with good labor standards), but lack of skills and poor industrial relations are becoming more pressing problems. A number of other problems are being addressed or are constraints only for some categories of firms (e.g. finance, infrastructure). They are, however, important barriers to diversification. 2.1 Cambodian firms are at differing levels of productivity, with significant disparities within the country and gaps with competitors. Chapter 1 also highlighted a number of characteristics of the private sector: a high degree of informality; thin capitalization; and, lack of diversification. To distill priorities for policy action, this Chapter reviews the constraints in the business environment and their impact on productivity. Section A summarizes the perceptions of firms about key constraints and links these to productivity. Section B takes a closer look at all potential constraints. Section C summarizes key lessons learned from this Chapter by assessing the differentiated impact of the business environment on these constraints and prioritizing the key constraints that should be addressed through five priority goals. A. Overall Constraints and Impact on Productivity 2.2 The five main constraints perceived by firms are: corruption; macroeconomic uncertainty; anti-competitive or informal practices; economic and regulatory policy uncertainty; and electricity (Figure 2.1). At the outset, however, it is important to note two main caveats with data based on perceptions. First, the intensity of each constraint might not be entirely comparable as each question varies in span: for instance, "tax rates" is quite specific, while "corruption" can cover a number of areas. Second, perceptions are influenced by the `public good' aspect of constraints and the characteristics of existing firms: for instance, firms might not raise electricity as a constraint if the firms are not electricity intensive. In some cases the perception might even be misleading. As for example, if only poor-performing firms raise issues of access to finance (which should indeed be the case, if the financial sector works well). Figure 2.1: Constraints on the Business Environment in 2004 and 2008 19 60 Less severe constraint in 2008 51.8 More severe constraint in 2008 S i 1 50 38.2 40 31.8 30 26.7 20.9 18.1 20 16.0 15.6 15.5 12.8 12.0 11.5 11.5 10.8 10.6 9.1 10 7.2 3.3 - Macro Economic Elec tricity Labor regulations Legal system/conflict Corruption Tax Adminis tration Acc ess to Financing Transportation Cus toms and Trade Sk ills & Education Anti-competitive or Economic & Regulatory Tax Rates Licensing and Permits Cos t of Financing Crime, theft and disorder Telecommunic ations Access to land informal practices Policy Unc ertainty Uncertainty regulations resolution Note: Share of firms that respond "severe or very severe" to the question "how problematic are the following issues to the operation and growth of your business". Source: 2003 and 2007 ICS. 2.3 Although corruption remains the main concern, important changes are noticeable since 2003 (Box 2-1). Crime and security concerns have significantly declined, and perceptions of customs and trade matters, as well as legal system/conflict resolution, have improved. On the other hand, concerns about macroeconomic uncertainty have increased and electricity has emerged as a major concern. These changes are reviewed in more detail in Section B. 2.4 Important nuances are also apparent across firms: Examination of the top one-third of the most productive firms (measured by TFP) finds that lack of skills is an obstacle for 24 percent (against 15 percent for others). Similarly, labor regulations and anti-competitive/informal behaviors are more of a concern for large firms. Also, top performers are more concerned than others about the weakness of the legal system. Electricity supply is an even more important issue for the garment and tourism sectors. Regulatory uncertainty is a major issue for large and garment firms, while macroeconomic uncertainty is an issue mainly for these same firms and low-productivity firms. Finance is an issue both for poor performers and top performers, presumably because poor performance of the former makes them unattractive to finance institutions, and for the latter, finance is a problem because top performers have more projects than the financial sector can finance. Enterprises that export also view access to finance as a more severe constraint than other businesses, possibly pointing to the need for export financing. The 2006 PBES data show that informal or micro-enterprises have more concerns about macroeconomic uncertainty and the cost of finance, and less about corruption and infrastructure (electricity, road quality and telecommunications). New firms (created in 2005 or after) are much more concerned about macroeconomic and regulatory/policy uncertainty, and somewhat less about corruption (with the three comprising around 48-49 percent of negative perceptions). 20 Box 2-1: Key Findings from the 2003 Investment Climate Survey Seizing the Global Opportunity: Investment Climate Assessment and Reform Strategy used an investment climate survey of 800 urban, rural, and informal firms to identify how the policy and institutional environment impacts individual firms, and benchmarks the responses of Cambodian firms with those in China, Bangladesh, India, and Pakistan. In spite of Cambodia's rapid growth and a remarkable recovery in the years preceding the ICA, the data suggested that both total factor productivity and labor productivity were low in comparison with countries that Cambodia must compete with in international markets, including Bangladesh, India, China, and Pakistan. The report highlighted key factors contributing to low productivity, including corruption, weak rule of law, informal practices, and complex and costly regulation. Among the more surprising results, the share of sales revenue paid by Cambodian firms in the form of bribes was over twice that of Bangladesh and by far the highest among comparators. Cambodia also had the most annual inspections, the highest cost per capita to officially register businesses, and the second highest amount of time for managers dealing with officials (after China). Trade facilitation practices were particularly constraining, with Cambodia having the second longest clearance times after Pakistan and some of the highest observed incidence of unofficial costs. In contrast with Cambodia's reputation as a liberal environment, in several key respects, Cambodia appeared to be a restrictive market due to the cumulative impact of multiple agencies operating in an uncoordinated fashion. The governance findings were so acute that they overwhelmed visible deficiencies such as finance, infrastructure, and human capital/skills. This was not to suggest that these factors were not relevant, but rather that the private sector was not in a position to be constrained by them because of more urgent concerns. The structure of Cambodia's private sector was already a reflection of the investment climate. There was a high degree of informality, and little long-term investment in productive assets outside of sectors which enjoy either policy-based market advantages or unique cultural assets. The strategy of remaining informal appeared to make short-term sense: informal firms as well as rural firms were less constrained. For the long term, this strategy was limited as the informal sector had a reduced capacity to trade with the formal sector, to obtain credit and to grow. Overall, the result appeared to be a negative cycle in which firms remain small and informal, denying Government the revenue base needed to improve public sector performance, which in turn contributes to weaknesses in the investment climate. The report made a number of recommendations, with a particular focus on improving trade facilitation practices, integrating markets through institutions such as private value chains, and injecting competition and transparency in private participation in infrastructure. The response by the Government to this report indicated its political will to confront the main challenges and "seize the global opportunity." Source: World Bank (2004). Impact of the Investment Climate on Productivity 2.5 Differences in the investment climate have an impact on Cambodian firms' productivity. To show this, productivity, as measured by TFP, can be regressed on key investment climate indicators (Table 2.1, see Escribano and Guasch, 2005). These indicators are based on actual firms' experience, as opposed to perceptions, and cover (i) infrastructure; (ii) red tape, corruption and crime; and (iii) access to finance and corporate governance. The results are of an order of magnitude similar to that found in several other countries and suggest: The impact of power outages on productivity is very significant. Based on semi- elasticity, a reduction of 50 percent, compared to the current mean, would increase productivity by 3.25 percent. In other words, with no power outages, productivity would be 3.67 percent higher. The impact would be even greater for firms that make above-median use of electricity (5.77 percent). Customs procedures also reduce productivity: cutting 50 percent of delays would increase productivity by 1.15 percent. Again, for firms that significantly use customs processes, they lose 3.78 percentage points in productivity due to problems with customs. Firms that pay 50 percent less for "getting things done" have, on average, a productivity 3.75 percent higher. 21 The impact of having an overdraft is limited overall, because few firms have one, but this is very significant for firms with an overdraft (which have productivity 11.9 percent higher than firms without an overdraft). Table 2.1: Impact of Investment Climate Constraints on TFP Regression Impact (standard (semi-) At mean For high Coef. error) elasticity value use Infrastructure Total hours of power outages -0.09 0.04 ** -0.065 -3.67 -5.77 Total hours of unavailable mainline phone services -0.01 0.02 -0.004 -0.2 -0.52 Av. number of days to clear customs for imports -0.13 0.07 * -0.023 -0.8 -3.78 Red tape, corruption, crime Number of days spent in inspections/official meetings 0.00 0.03 0.000 0.01 0.02 % time spent by senior management dealing with bureaucracy/regulations 0.00 0.02 0.001 0.06 0.14 % sales paid in gifts or informal payments to "get things done" 0.17 0.11 0.075 2.71 5.65 % sales lost to theft, robbery, vandalism, or arson -0.03 0.11 -0.002 -0.04 -0.14 Finance and corporate governance Existence of overdraft facility or line of credit (1=yes) 0.67 0.82 0.945 0.83 11.9 Existence of loan from bank or financial institution (1=yes) -0.05 0.14 -0.052 -0.2 -0.91 Access to loan is constrained (1=yes) -0.03 0.06 -0.025 -0.05 0.46 Hiring of outside accounting/audit firm (1=yes) 0.02 0.08 0.021 0.07 0.36 Observations 272 2 R 0.97 Note: TFP is measured from a sector-specific translog production function, see Annex E. (**) ­ resp. (*) ­ signals variables significant at the 5% level. A few variables are instrumented with sector-specific dummies to address their endogeneity. The (semi-)elasticity is estimated at sample means. The impact is measured as the gain/loss of productivity for the mean firm in Cambodia (e.g. the mean power outage, vs. having zero outage). The "High Use" column measures the productivity impact for firms that have above-mean use of a particular investment climate variable (e.g. have above-mean power outages). See Annex E for more details. 2.6 As noted in the previous chapter (para. 1.12), some of these variables also have an impact on the decision to stay in the informal sector. B. Potential Constraints to Investment and Productivity 2.7 To deepen the analysis of investment climate constraints, this Section reviews in a systematic way, the key drivers of investors' decision to create or expand capacity, or to make investments in productivity. The structure is based on the framework of `growth diagnostics' developed by Haussman and Rodrik (2005). Finance 2.8 The first key issue is whether Figure 2.2: Firms using Banks to Finance entrepreneurs are able to finance Investment (%) projects. Subsequent sub-sections will Cambodia focus on whether it is possible in 2007, 10.8 Cambodia 2003 Cambodia for entrepreneurs to develop projects that will generate sufficient returns. Insufficient access to finance T Vi h La could result from lack of savings or liquidities, or poor intermediation - 20.0 40.0 60.0 80.0 100.0 between the supply of, and demand for, Note: the box represents the 25th and 75th percentile of the finance. distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. 2.9 Access to finance in Cambodia Cambodia's performance in 2003 and 2007, as well as those is very limited and uneven. Overall of Lao PDR, Thailand, and Vietnam are also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. only 10 percent of firms use banks to 22 finance investment (Figure 2.2). This puts Cambodia behind its neighbors and most other countries. Thirty-six percent of large firms (above 100 employees) and 24 percent of trade firms have a loan (against 21 percent across the economy). The top one-third of TFP performers also have better access, at 26 percent. Smaller firms were more likely to point to collateral requirements and complexity of processes as constraints, while larger firms tend to be concerned about the excessive cost of financing. 2.10 Some improvements were nevertheless made through rapid development of the financial sector (cf. Section 1.C). Following some restructuring of the sector in the early 2000s, progress in bank supervision and strong demand, deposits and loans have increased from 15.1 and 8.5 percent of GDP, respectively, in 2004, to 26.8 and 18.3 percent of GDP in 2007. Firms also noted improvements in the quality of service. For instance, delays to clear a check or a wire transfer were significantly reduced between 2003 and 2007. Macroeconomic data in 2007 also suggest ample available liquidities in the financial sector. Some decrease in nominal interest rates and the recent spike in inflation have also caused real interest rates to become negative. Hence, the supply side of finance seems to have improved rapidly in the last couple of years. 2.11 Nevertheless, firms that have access to a loan have drastic conditions. Although the terms improved between 2003 and 2007 (the average nominal interest rate went down from 13.4 to 10.7 percent and the average duration for loans went up from 7 to 20 months), collateral requirements became much tighter. The proportion of loans requiring collateral went up from 51.5 to 86.6 percent and the value of the collateral went up from 47 to 353 percent of the loan value. The recently approved Secured Transactions Law, which was hailed in the 2009 Doing Business report as a world-class framework, should help in expanding the type of collateral that can be used, enabling moveable and intangible assets (including equipment, inventories, accounts receivables, and agriculture commodities) to be used as security and registered with minimal formalities. The internet-based registry office (www.setfo.gov.kh) now needs to become fully operational for financial institutions. The newly established Credit Information System (www.cis.org.kh) will also help borrowers develop another form of collateral (their credit history), although the current setup of the CIS needs to be greatly expanded and to cover positive as well as negative information (i.e not only data on defaults). Full implementation of these two reforms would make access to finance easier for many firms. 2.12 Overall, it is unclear whether Figure 2.3: Has your Firm a Loan? firms actually feel constrained by Do you have a loan? such limited access. Only 1 in 10 firms sees access to finance as a severe No: 79.3% Yes: 20.7% constraint (Figure 2.1), which is below most other countries. Most firms in fact Never applied: 96.8% Rejected: 3.2% did not even try to access finance (Figure Too complex: 4.8% 2.3): only 3 percent had their application No need: 86.3% Collateral requirts: 4.9% rejected. And most of those that did not High interest rates: 4.0% apply, did not feel the need to apply, Source: Cambodia 2007 survey. while traditional constraints (complexity; collateral requirements; high interest rates) explained only a small proportion of the lack of applications.19 2.13 Part of the explanation might be the structure of the private sector. On the one hand, the high proportion of firms with foreign ownership (Figure 1.2) means that many firms have access to internal financing through their parent company. Indeed, less than 15 percent of firms with foreign ownership, despite their relatively large size, have a loan. The 19It is recognized that interpreting these questions is complex, as firms could claim no need for loans only because of the features (e.g. high interest rates) of the existing credit supply. 23 other 85 percent of firms overwhelmingly report that they do not need a loan. However, the fact that a large number of firms are still informal leads many to invest from their retained earnings or cash available to their owners (firms report that two-thirds of their working capital, and of financing for investment, comes from retained earnings, family or friends). 2.14 However, it is likely that finance is a more severe constraint for (i) new entrants and (ii) technology upgrading. In both cases, the risks are higher. Also, foreign parent companies might not have the incentive for upgrading (cf. Section 1.C on garments), but new foreign investors might still be looking to leverage their equity with in-country debt, which is difficult to raise. The fact that firms in the trade sector have more access to finance, but still raise this as a severe concern, suggests that export financing might also be a constraint. More than a real financial constraint, this might reflect firms' lack of capacity to prepare business plans and banks' lack of capacity to assess those business plans. Well- targeted advisory services could be helpful; matching grant schemes might also be an attractive option.20 2.15 In summary: The rapid development of the financial sector, both in terms of available financing and quality of services, and the low demand for financial sector services because of the structure of the economy (informality; foreign investment), mean that access to finance is unlikely to be a binding constraint in the short term across firms. However, thoroughly implementing some recent innovations (such as the Secured Transactions Law and the Credit Information System) could bring down the cost of financing. Finally, some specific needs are not particularly well served by the current system. New entrants and innovators, especially for trade, face more constraints, which is a possible explanation for the economy's lack of diversification. Favoring new entrants and innovators will require progress with the Credit Information System, full implementation of the Secured Transactions Law, and building private sector access to trade finance. Labor 2.16 The next three sub-sections turn to the ingredients necessary to generate projects with sufficient returns: labor; logistics; and infrastructure. 2.17 While labor regulations remain Figure 2.4: Firms Offering Formal Training unproblematic for most firms, skills (%, manufacturing sector) are emerging as a significant concern Cambodia (Figure 2.1). There has been a general Cambodia 2003 2007, 53.6 increase in the level of education, and the demand for skills continues to increase. Firms have responded through La Vi Th quickly increasing the provision of training to their employees (Figure 2.4). - 20.0 40.0 60.0 80.0 100.0 Nevertheless, the effectiveness of the Note: the box represents the 25th and 75th percentile of the current approach, which relies distribution and the middle line, the median. The extreme min and max of significantly on internal training, is of the bar represents thein 2003 and 2007,the well as those Cambodia's performance as distribution. probably low, reflecting a market failure of Lao PDR, Thailand, and Vietnam are also displayed. to provide relevant training for the Source: Cambodia surveys, and www.enterprisesurvey.org. private sector. In addition, firms report significant problems in retaining trained staff, weakening the incentive to provide training. 20 Such a scheme has been piloted by the Ministry of Commerce through an Export Market Access Fund (financed by IDA). AusAID also has a regional facility, the Enterprise Challenge Fund, to help commercially viable projects. 24 The time to fill vacancies has increased significantly, from 4.2 to 11.5 weeks for skilled workers and from 4.9 to 7.1 weeks for unskilled workers. 2.18 Addressing this persistent issue will take time and requires a multi-pronged approach. On the demand side, CAMFEBA, the main employers' association, has initiated a program to better understand the needs of the firms. In the garment sector, a preliminary needs' assessment has been conducted (IFC, 2007). On the supply side, also in the garment sector, four institutions provide training (Better Factories Cambodia, the Garment Industry Productivity Center, GMAC, and IFC), although these partly donor-funded schemes need to find a sustainable model. In factories covered by IFC's skills training program, for instance, absenteeism went down 8 percent and both in-line and shipment rejection rates went down by 40 percent. Nevertheless, there are still gaps on the supply side with these programs, having limited coverage and missing some components (such as attitudes to work). In summary, the private sector has a good sense of its needs and trainings have been piloted, but the supply needs to be scaled up through an appropriate business model. There are also coordination issues among multiple Government stakeholders (ministries concerned with labor, education, and commerce). The accreditation system appears ineffective as most firms cannot rely on resumes or degrees presented by job applicants and have to finance their own testing. Finally, there seems to be a mismatch between the specific skills produced by the education system (a large proportion of business degrees, for instance) and demand from employers (for mechanics, accountants, etc.). This will require innovative solutions to link supply and demand (such as industry-specific centers, as in the garment sector, or linking universities and business associations, or providing training through Special Economic Zones). Developing vocational training and creating real opportunities for apprenticeship could be targeted at sectors where lack of skills appears more constraining than in others (or where skill needs have been better identified, as in the garments and tourism industries). 2.19 Wages remain competitive, but are increasing rapidly. The minimum wage, which effectively applies only to the garment and shoe sectors, is at a level similar with competitors such as Vietnam, but productivity in Cambodia is lower (USAID, 2007a and Section 1.B). But recent pressures of inflation, including those generated by some shortages of labor in a context of rapid growth, are generating demands for higher wages. 2.20 Improved pay practices could generate a double dividend. As discussed in Section 1.B, paying employees by the piece and generating efficiency gains could lead firms to become more competitive and employees to increase their take-home pay. The practice is already used in several factories, but this could be scaled up in a way that does not further increase conflicts on this issue (in particular by making sure that pay-by-the-piece practices are used transparently ­ which is necessary if they are to provide incentives for productivity, are fair to workers, and do not circumvent regulations such as the minimum wage). 2.21 Labor regulations are not viewed as a constraint. Asked whether they would adjust employment if there was no restriction, the median firm indicates that it would make no adjustment (the overall average adjustment would be a 10 percent increase, i.e. if anything, firms slightly refrain from creating jobs because of labor regulations). This is consistent with the generally positive assessment of the 1997 Labor Law. Earlier constraints on night shifts were lifted in July 2007. However, the process of registering a firm at the Ministry of Labor is consistently ranking as a long and expensive process (including informal fees) (see also Doing Business indicators). 2.22 Higher labor standards in the garment industry are considered positive. The vast majority of firms tend to agree with the notion that "improvements in working conditions improve productivity and exports". The garment industry is subject to close monitoring from Better Factories Cambodia (Section 1.C) and to significant pressure from unions (see next paragraph). As a result, in the garment sector, workers benefit on average from one rest day per week (while 50 percent of firms still have no rest day) and maternity benefits (on average 3 months with pay at 70 percent of the salary, i.e. slightly above the legal 25 requirement of 3 months at 50 percent of the salary). Despite significant debates about the future of Better Factories Cambodia (in particular whether monitoring should be mandatory), the survey does not raise significant objections and supports the Government's decision to sustain compulsory monitoring after the end of the quotas. 2.23 The incidence of labor disputes, strikes, and civil unrest has increased considerably. The number of days of production lost to these causes increased from 2.6 to 6.0 days per establishment, per annum, between 2003 and 2007. This is particularly true in the garment sector, with 8 days of production lost per factory and per year, on average (ILO reported 175 strikes in 2006 and 2007, costing 600,000 lost workdays). This is in particular reflected in the very low capacity utilization rate in the garment sector (less than 57 percent). This is in part related to the combination of a rapid increase in union membership (35 percent of the workforce in garments, against less than 1 percent in other sectors, according to the ICS survey) and the lack of effective mechanisms at the enterprise level to prevent and resolve disputes, including the use of collective bargaining agreements (CBAs).21 2.24 Independent arbitration, union representativity, and collective bargaining agreements are several options to address this situation. First, since 2003, an independent Arbitration Council (AC) has been established (also with support from the ILO).22 It has heard almost 500 cases of disputes between unions and employers, and has successfully resolved more than two-thirds of them. One of its challenges is that the Labor Law lets the parties choose whether the award is binding or non-binding (and in 90 percent of the cases, they choose non-binding). While this option is important for the impartiality of the AC (non-binding awards remove any incentive to bribe arbitrators), it obviously weakens effectiveness. A second challenge to improved industrial relations in the garment industry is multiple unions in the workplace. According to GMAC, some factories have half a dozen or more different unions and these compete with each other. Currently, a union can be formed with as few as three employees. Determining which union is the most representative (has 51 percent or more of the workers as members) is an important first step. Management and the most representative union should then negotiate collective bargaining agreements. The Government, supported by the ILO, is strengthening the system for determining the most representative union in each factory and promoting the use of CBAs with standard clauses such as dispute resolution through binding arbitration; no strikes for the life of the agreement; and no further claims for the life of the agreement. Over 35 of the most- representative-union certifications in the garment industry where issued by MOLVT in 2008, compared to 11 in 2007. Also, since 2004, the use of CBAs has succeeded in eliminating strikes in hotels that have such agreements. 2.25 In summary: Labor remains in excess supply, wages relatively low, and regulations supportive. Hence the Cambodian labor market still has many features of a low income country, where labor is a key asset and not a major constraint. Nevertheless, three issues are becoming more pressing. First, the level of skills is poor, and poorly adapted to new demands from the private sector, and the market for skills has many market failures. Recent private sector 21 It also is partly an unintended effect of external monitoring of labor standards and should therefore be addressed in that context as well. 22 The creation of an AC was mandated by the 1997 Labor Law. Several ingredients have contributed to its outstanding reputation with employers and employees: (i) institutional and financial autonomy from the Government; (ii) external oversight from the ILO; (iii) well-qualified arbitrators, with a rigorous selection of the three-member panel (for each case, the employer chooses one arbitrator out of a list agreed on by employers; the employees choose one arbitrator out of a list agreed on by employees; and these two choose one arbitrator out of a Government list). Once chosen, arbitrators act independently and make a decision by a two-third majority; and (iv) speedy resolution of cases, with clear, transparent and widely disseminated documentation of the decision. The fact that decisions are non-binding also reduces any incentive for bribery. 26 initiatives should be supported, and scaled up, to better link the demands of the private sector with the offerings of the education system. A strong accreditation system is also key to fixing some of the market failures. Second, pay practices are not sufficiently supportive of productivity improvements. A simple approach with a double dividend would be that of expanding pay-by-the-piece, hence, raising incentives for productivity and increasing take-home pay. Finally, specifically in the garment sector, but potentially a generic constraint for larger- scale manufacturers, poor industrial relations have become a major detriment to productivity and morale. Logistics and Trade 2.26 Access to global markets and effective internal logistics should be key assets for Cambodia. As a small economy, Cambodia has to take advantage of the rapid growth of the global economy. Its location in Southeast Asia is a key asset from that point of view. The rapid growth of exports since the mid-1990s, mainly driven by the garment sector, is also a reflection of capacity to access global markets. Further growth is unlikely to come from preferential tariff access, as was the case for garments until 2004.23 Connection to global markets is also visible in the share of inputs sourced from abroad (Figure 1.2). Cambodia joined the World Trade Organization (WTO) in October 2004 (and ASEAN in the late 1990s). This has generated a large agenda of legal and other reforms (CLRDC, 2004). Cambodia also agreed to follow the Trade-Related Aspects of Intellectual Property Rights (TRIPs); WTO's Customs Valuation Agreement; WTO's SPS agreement; and the Technical Barriers to Trade Agreement. 2.27 However, firms in Cambodia Figure 2.5: Days of Inventory of Most raise a number of concerns related to Important Input logistics. Among the ICS respondents, Cambodia 7 percent had domestic deliveries that Cambodia 2003 2007, 40.7 were late and 15 percent experienced delays for exports; some 3.5 percent of the value of consignments suffered cargo Th ViLa loss or damage on domestic transport and 1.7 percent on international - 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 transport. Another 1.8 percent of sales Note: the box represents the 25 and 75 percentile of the th th were lost because the goods did not distribution and the middle line, the median. The extreme min and max of the reach the buyer on time. Since issues are of the bar represents the in 2003 and 2007, as distribution. Cambodia's performance well as Lao, similar with inputs. Firms have to Thailand, and Vietnam are also displayed. maintain some 41 days of inventory, Source: Cambodia surveys, and www.enterprisesurvey.org. which is very high by international standards (Figure 2.5). 2.28 While Cambodia is doing well compared to low-income countries, progress in improving customs, infrastructure, and the quality of the logistics industry is necessary (Table 2.2). Based on feedback from global freight forwarders and express carriers, the Logistics Performance Index compares performance across various dimensions and countries. Cambodia ranks 10th among low income countries and 81st overall. Despite this positive performance, competitors like Vietnam, China, and Thailand are doing much better. Areas for particular improvement are: customs, infrastructure, international shipments, and logistics competence. 23 This is the conclusion of the Government's Diagnostic for Trade Integration Strategy, DTIS (RGC, 2007). Notwithstanding this general conclusion, there will be a few specific opportunities for targeted sectors in the coming years. 27 Table 2.2: Logistics Performance Index Logistics Performance International Logistics Domestic logistics Customs Infrastructure Tracking & tracing Timeliness Index shipments competence costs Hong Kong, China 4.00 3.84 4.06 3.78 3.99 4.06 2.66 4 Thailand 3.31 3.03 3.16 3.24 3.31 3.25 3.21 3 China 3.32 2.99 3.20 3.31 3.40 3.37 2.97 3 Vietnam 2.89 2.89 2.50 3.00 2.80 2.90 3.30 3 Cambodia 2.50 2.19 2.30 2.47 2.47 2.53 3.21 3 Lao PDR 2.25 2.08 2.00 2.40 2.29 1.89 2.13 2 East Asia & Pacific 2.58 2.41 2.37 2.64 2.54 2.53 3.04 3 Low income 2.29 2.12 2.06 2.32 2.29 2.25 2.99 2 Note: The Logistics Performance Index is based on a survey of operators on the ground worldwide (global freight forwarders and express carriers), providing feedback on the logistics "friendliness" of the countries in which they operate and those with which they trade. They combine in-depth knowledge of the countries in which they operate, with informed perceptions of other countries with which they trade, and experience of the global logistics environment. Feedback from operators is supplemented with objective data on the performance of key components of the logistics chain in the home country and data collected for 100 countries. The indicators are ranked on a 1 to 5 scale, with higher scores indicating better performance. Source: World Bank (2008). 2.29 Government progress in trade facilitation has been significant in recent years. The time to clear a shipment (between the port and the factory) has decreased for exports (to 4.3 days), bringing Cambodia to slightly below the median country (Figure 2.6). Performance for imports is 5.1 days, on average (3.7 days for manufacturers), which is now closer to the top 25 percent.24 These results translate into customs and trade being less of a constraint on the business environment ­ even though 19 percent of exporters still feel it is a severe constraint (against 9 percent across all firms).25 Figure 2.6: Average Time to Clear a Shipment (number of days) a/direct exports b/ imports (manufacturing sector) Cambodia Cambodia 2007, 4.3 2007, 3.7 Cambodia 2003 Cambodia 2003 ThLa Vi La Th Vi - 5.0 10.0 15.0 20.0 - 5.0 10.0 15.0 20.0 25.0 30.0 th th Note: the box represents the 25 and 75 percentile of the distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. Cambodia's performance in 2003 and 2007, as well as those of Lao PDR, Thailand, and Vietnam are also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. 2.30 This reflects a range of ongoing Government initiatives, in large part the initial implementation of the 12-point action plan on trade facilitation, adopted in 2004. Following the 2004 ICA, the Government adopted a trade facilitation strategy, with a 12- point action plan. One of these points, for which implementation is well underway, is a risk management framework. A clear strategy was formalized in sub-decree 21, dated March 2006. An audit and risk management unit in the General Department of Customs and Excise (GDCE) has been created and staffed, and the list of prohibited and restricted goods has been streamlined (although the list would benefit from further streamlining). Risk and selectivity criteria have been set up; and a profiling of traders has been prepared. Since this progress has largely been driven by the GDCE, it now needs to be matched by progress in other trade-related agencies. The second output was the agreement among trade agencies to develop a Single Administrative Document (SAD) to streamline documentation requirements 24 Given the limited volumes of trade, there will be limited congestion problems. Hence a high relative performance would be expected. 25 The Doing Business Indicators highlight a similar trend (Annex D). The "Trading across Borders" indicator however shows a longer time for imports and exports because it includes a larger part of the process (e.g. preparing documentation), not just the time at the border. 28 at the border.26 The third, related output has been the computerization of customs with the ASYCUDA World system. The first site, at the Port of Sihanoukville, went live on May 1st, 2008 and, after some initial adjustments, is delivering faster processing, stronger governance, and increased revenue collection. This system also enables a more rigorous implementation of the risk management strategy. In future, ASYCUDA will be rolled out to other sites and other agencies are expected to step up their efforts to implement the risk management strategy. In addition to the ASYCUDA roll-out, the Government intends to develop a National Single Window (NSW) based on the ASYCUDA platform, to provide a single entry point for trade-related transactions with the Government. The Government targets 2012 to have the NSW in place, in conjunction with the ASEAN Single Window. Finally, the enactment of the Customs Law in 2007 is a major milestone, paving the way for the adoption of some 34 regulations; tariff adjustment (in particular to fulfill commitments to ASEAN to move to the Common Effective Preferential Tariff Scheme by 2010); adherence to the 1999 Revised Kyoto Convention of the World Customs Organization; and creation of a clear status to professionalize customs brokers. The GDCE adopted a revised 5-year strategy in 2008. Box 2-2: Transparency of Export Procedures In an effort to provide a central access point for information on export procedures and regulations, the Government developed, with support from the World Bank Group, a Handbook on Export Procedures: Practical Guide for Small and Medium Enterprises in Cambodia which explains the export procedures of respective Government agencies required for a wide range of products. The handbook also includes procedures at five main export points in Cambodia: Bavet, Poipet, Sihanoukville port, Phnom Penh port, and Pochentong International Airport. The booklet was reviewed by a joint meeting of the Government and the private sector, as part of the Sub-Steering Committee on Export, and widely disseminated. The Ministry of Commerce's website is also being updated and is expected to be then linked to a comprehensive Trade Information Website. The private sector values this first handbook on export procedures and this has generated demand for other such booklets, for instance on the standards used in Cambodia (for trade-related issues).() The impact of such transparency will depend on whether processes are implemented as described in these handbooks. The Government-Private Sector Forum should serve as a forum for channeling such feedback from the private sector. () A handbook on business registration was also published by MIME, with support from ADB. 2.31 Progress in trade facilitation needs to go further. First, positive outcomes are different across sectors. The garment sector has benefited from significant progress, reflecting the active role of its industry association and recognition by Government of its key role in the country's economy. However, other sectors have not achieved as impressive results (Figure 2.7). In fact, the top one-third of performers (in terms of TFP) have documentation problems with their containers in only 11 percent of cases, against around 40 percent for other firms. Second, the cost of shipping remains high and, again, differentiated across sectors, with progress for the export of garments but deterioration for the export of rice, for example. The cost of getting an import license is also high due to a significant incidence of informal payments (although delays in getting a license are lower than in other countries). The cost of exports is also worsened by delays and informal payments to get VAT rebates. Third, uncertainty remains high: the longest time is on average six days for export and nine days for import. Anecdotal evidence also points to significant variability in processing time for import and export. Finally, lack of transparency remains a concern (Box 2-2): according to the ICS. While 50 percent of firms have good access to information on various regulations (taxes, environmental and labor regulations), only 10-20 percent have good access to information on trade fees and processes. Obtaining certificates of origin and 26The SAD is another example of the need for dissemination. The 2007 survey asked about the SAD, but 54 percent of the firms (including 35 percent of exporters) have never heard about it. Of those who knew about it, 1 out of 2 non-exporters thought it would facilitate trade, and 2 out of 3 exporters thought so as well. 29 complying with rules of origin remains complex, as evidenced by the analysis of the food packaging industry (Section 1.C). Figure 2.7: Cost and Timeliness of Shipments ­ Garments and Other Sectors a/ Cost for a 40' container of b/ Cost to export a c/ Time to clear at border (days) garments ($/container) 20' container of rice ($/ton) 1200 1,126 60 57 10 8.4 1000 858 879 50 41 8 800 655 40 5.5 6 600 30 4 2.8 400 20 2.1 2 200 10 0 0 0 Av. # days to clear imports Av. # days to clear exports 2003 2007 Import Export Garment Other 2003 2007 Source: JDI, 2007b. 2.32 Awareness of sanitary and phyto-sanitary standards, as well as other technical standards, is limited. Most non-exporters are not at all aware of such trade-related requirements and the overwhelming majority of exporters do not see them as particularly constraining. Ironically, despite this, firms are concerned that their competitors in the domestic market undermine them by not respecting their copyright, patent or trademark. While not constraining the current main export, garment products, lack of awareness about standards and intellectual property rights will be a major constraint for many potential exports. 2.33 This points to several gaps in implementing the 12-point action plan and the need to address a few other priorities: The strategic review of CamControl was undertaken and, among the possible options, the Council of Ministers decided to strengthen CamControl. The ICA survey shows little change between 2004 and 2008. Although the decisions to have GDCE as the lead agency at the border, and to conduct joint inspections, had a positive impact, firms continue to report many problems. GDCE has made progress toward the introduction of an ASYCUDA fee. However, no progress has been made with the introduction of a flat fee. Progress in streamlining registration at the Ministry of Labor was limited (see Doing Business Indicators, and paragraph 0 about the registration process). Progress on the transparency agenda needs to be accelerated. In parallel, the capacity of the private sector to be aware and comply with regulations in the trade area should be supported. The proposed `System of Transparent Performance Measurement, including Private Sector Monitoring' led to a one-time `Reform Scorecard' in 2005, but it was not institutionalized. 2.34 Another area where progress should accelerate is the efficiency of ports. Although the presence of a deep-sea port is a major asset for Cambodia, despite recent progress and ongoing reforms (such as computerization), the efficiency of the Port of Sihanoukville is still low. UNCTAD's line shipping connectivity index is 3 for Cambodia, against 18 for Vietnam and 35 for Thailand. The airfreight cost of a shipment from the US is 21 percent of the import value, against 15 percent in Vietnam, 12 percent in Lao PDR, and 3 percent in Thailand (World Trade Indicators, 2008). It costs US$600 to ship a dry, 40-foot 30 container from Sihanoukville to Singapore, against US$220 from Ho Chi Minh City to Singapore.27 2.35 Despite rapid growth of trade, the supply of trade-related services is weak. As noted in the case of aluminum cans (Section 1.C), a manufacturer often has to import inputs itself (raw materials; parts for equipment; etc.). In addition to incurring the costs of getting each import license, and undertaking customs procedures themselves, companies incur the cost of maintaining high inventories to avoid the consequences of import delays. 2.36 In particular, logistics services remain underdeveloped in Cambodia. This, in part, reflects the relatively recent development of the economy. Although only 13 percent of firms rate transportation as a severe constraint (and the network of main roads has indeed considerably improved over the last few years), logistics firms raise a number of issues that impede their development: Informal fees on roads remain widespread and impose a double penalty on the industry: a direct financial cost, as well as the indirect cost created by significant uncertainty over potential delays. Payments en route from Phnom Penh to the border amount to US$250 (World Bank, 2008a). This is in addition to the costs at the border, as discussed above. The cost of petroleum products is high in Cambodia by regional standards.28 While this, in part, reflects the fact that a few other countries in the region subsidize petroleum products ­ an unsustainable and environmentally unfriendly approach ­ it also reflects the small size of the market and potentially high margins. Competition in the transport sector is also said to be somewhat limited. In particular, the allocation of bilaterally- negotiated quotas for trucks allowed to operate on both sides of a border (e.g. in Thailand or in Vietnam) is not transparent. With some of the key road infrastructure in place, trade is now hindered by a range of non-physical barriers to the cross border movement of goods. Although distances in the Greater Mekong Sub-region (GMS) are not particularly long compared to more integrated regions such as Europe (500 to 2,000 km, against 2,000-3000 km), intra-GMS trade is very limited ­ almost non-existent, in the case of Cambodia (World Bank, 2008a). A well-established industry like garments continues to source fabric from outside the region because of these constraints. The cost of moving a semi-trailer with a 40 foot container from Ho Chi Minh City to Phnom Penh (a 2-day trip) is US$5.65 per km, against US$0.76 per km for Bangkok to Singapore (itself not very competitive compared to other more integrated regions). Building on existing agreements, such as the Cross-Border Transport Agreement and the Strategic Framework for Action on Trade Facilitation and Investment (both sponsored by the ADB), a few actions could be taken: (i) create a regional transit system (in the spirit of the Transport International Routier or TIR) with a clear protocol for transit; (ii) improve processing at the border (possibly through joint facilities); (iii) develop agreements for trucks to operate in several countries, and develop financial products that will support this); and (iv) improve the capacity and coordination of customs agencies, ministries of transport, and the private sector. 2.37 In summary: While Cambodian firms do not rate transportation or customs as major concerns, the persistent high levels of inventories, formal and informal fees, plus anecdotal evidence on variability of delivery times are important signals for an economy that needs to expand trade in order to develop. 27 Significant investments financed by JICA to upgrade the Port of Sihanoukville should help improve the situation in the coming years. 28 Although the mode of taxation (fixed amount, as opposed to ad valorem) has led the gap to shrink while global prices were increasing. 31 Due to concerted Government efforts, much progress has been made in the area of trade facilitation, although the achievements should be expanded to benefit all traders. Further progress is therefore required to extend the roll out of ASYCUDA, create a NSW, and extend the risk management approach. A quick win would be achieved by improving the level of transparency on trade-related processes and fees. The export guide should be disseminated more widely and complemented by more trade-related booklets, websites, and dissemination events. The next priority will be to expand services supporting trade. A few of these services are public goods (such as creating standards, testing, etc.). Although delivery by the private sector might be the most effective approach, these services will need support from the Government, in particular to facilitate intra-GMS trade. Infrastructure 2.38 Electricity ­ limited access, high cost, and unreliability ­ is becoming a major issue for Cambodian firms. For the average Cambodian firm, almost 4 percent of sales were lost to power outages in 2007, against 2.2 percent in 2003 (Figure 2.8). The losses are as high as 6.7 percent of sales in the tourism sector. More than one-third of the electricity used by manufacturing firms comes from generators (26 percent, economy-wide). The proportion is as high as 54 percent for the garment industry, even though, among manufacturing industries, garment factories are not particularly intensive users of electricity. A total of 82 percent of garment factories have their own generators (against 39 percent across the economy). As noted for aluminum cans, the costs of generators include capital, operating, maintenance, and waste costs. The percentage of firms with a generator (which tends to be large in the garment and tourism sectors) also tends to have a higher-than-average cost for electricity (28 cents per kwh in garments, against 21 cents across the economy). This is another sign of the opportunity cost resulting from poor access and quality, i.e. a willingness to pay significantly more for self-generated electricity than for electricity from the grid (at around 15 cents per kwh, Figure 1.10). Figure 2.8: Electricity a/Value Lost to Power Outages (% sales) b/ Electricity from Generator (%, manufacturing sector) Cambodia Cambodia 2007, 3.8 2007, 36.2 Cambodia 2003 Cambodia 2003 Th La Th La - 5.0 10.0 15.0 20.0 25.0 - 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 th th Note: the box represents the 25 and 75 percentile of the distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. Cambodia's performance in 2003 and 2007, as well as those of Lao PDR and Thailand, is also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. 2.39 The Government is very aware of concerns about electricity and a number of investments are underway.29 Starting 2009, a new transmission line enabled the import of electricity from Vietnam. Other transmission lines and new capacity (hydro and coal) have also been contracted and should come on line between 2010 and 2015. This should cover a large part of current unmet demand. 29 This ICA did not review in detail the issue of private participation in infrastructure (PPI), which was discussed in detail in the 2004 ICA. Some progress has been made, in particular with the enactment of the Law on Concessions in 2007. However, much more is needed to build a modern regulatory framework, build capacity, and make sure that Cambodia gets value for money out of public/private participation in infrastructure. 32 2.40 Investments in electricity generation need to be complemented by an adequate distribution system and improvements in the regulatory environment in order to keep the price low and improve the quality of services. Two shortfalls in the current approach make it difficult for firms to bet on better access in two or three years (and hence invest immediately). First, because development plans for the grid are unclear and connections remain very expensive,30 firms are reluctant to invest on the assumption that they will be connected to the grid by 2010 or 2011. Second, because the investments made in the power sector ­ and their terms ­ are not disclosed and the EDC's plans are unclear, investors are still not sure whether they will benefit from lower prices in the future. For instance, recent improvements in the supply of electricity in Siem Reap has reportedly not yet translated into lower costs or more reliability. 2.41 Firms also have concerns about water supply. Respondents report on average 11 days of interruption in water supply per annum and almost 1 percent of sales lost due to insufficient water supply. 2.42 Telecoms are Table 2.3: Telecommunications and Internet Access in East emerging as a Asia constraint mainly International Price basket for Mobile phone Telephone Internet Internet users because of their cost. Country bandwidth Internet (US$ per (bits per month) (per 100 people) subscribers (per 100 people) mainlines (per 100 people) person) In particular, more than Cambodia 1.29* 33.02 0.48 17.88 0.26 20 percent of exporters Korea, Rep. Indonesia 9.22 1027.84 18.27 34.56 5.76 71.75 36.27 89.63 7.90 47.99 raise the sector as a Lao PDR 32.08 27.60 1.71 25.23 1.62 severe constraint (against Malaysia Mongolia 124.47 116.07 7.61 10.57 56.50 12.25 87.94 29.99 16.38 6.15 11 percent across the Papua New Guinea Myanmar 1.94 0.99* 46.28 25.10 0.08 1.77 0.44 1.24 1.05* 1.05* economy). The sector Philippines 38.01* 1.95 6.03 49.70 4.21 has developed rapidly, ThailandVietnam 389.99 147.76 7.37 10.36 21.02 20.99 80.49 27.87 11.00 33.51 but costs remain high East Asia and Pacific 158.81 14.42 14.44 41.42 24.03 Source: International Telecommunication Union. Data are from 2007. Data in (Table 2.3), bandwidth italics are from 2006. Data with an * are from 2005. limited, and weak connections across networks remains a concern. 2.43 Across infrastructure, firms give positive feedback on the rapidity of getting connections, but negative feedback on the incidence of informal payments.31 Delays are short, from four days, on average, to get a phone connection to 15 days for electricity. But the cost is high (as high as US$1,300 for an electricity connection). And with the exception of water connections, more than 50 percent of firms report having to make informal payments. 2.44 In summary: Although not a new issue, infrastructure (mainly electricity) is more of a concern than a few years ago, most likely reflecting the further increase in global oil prices and the shifting structure of the economy (i.e. higher demand for electricity). The Government's plans appear to largely respond to the urgency of the situation in terms of creating generation capacity. This still leaves the need to deal with the delivery side through an appropriate regulatory framework and better governed delivery. This is also true for telecommunications. 30 It still costs more than US$1,300 to get a connection to the grid (up from US$182 in the 2003 survey), with informal payments in 55 percent of the cases. 31 Also the proportion of firms that find the Government "effective" in delivering services has increased from 13 percent in 2004 to 23 percent in 2007 (with an additional 47 percent characterizing it as "somewhat effective", against 41 percent in 2004). 33 Taxes, Property Rights, and Regulations 2.45 Finally, the next four sub-sections discuss problems that could prevent entrepreneurs from reaping returns from their projects. Even with profitable ideas which they can finance, entrepreneurs might decide not to invest because their profits will be captured by the State ­ through legal means (e.g. taxes) or illegal means (e.g. bribes) ­ or by others (e.g. if intellectual property rights are not protected). 2.46 Tax rates continue to be at acceptable levels for most firms. Only 16 percent of firms view taxes as a severe constraint, with a higher proportion of concerns expressed by large firms and exporters (other than garments). The better situation in the garment sector can be explained by the continued tax incentives granted by the Government to the industry. Using the Doing Business methodology, Cambodia taxes profits at 22.6 percent, against 35- 40 percent in neighboring countries.32 2.47 Firms do not rate tax administration as a major concern. The required number of meetings with tax officials has declined from 7 per year in 2003 to 3 in 2007, bringing Cambodia close to the median of comparator countries. This reflects some of the recent reforms implemented by the General Department of Taxation, including: (i) the restructuring of the department, enabling the broadening of self-assessment with more offices in Phnom Penh, and an increase in the human and technical capacity of the department, supported by better IT capacity; and (ii) a growing program of tax audits and collection of arrears. Progress in legislation is also being made, and a process of dialog with the private sector (as part of the Government-Private Sector Forum) enables private firms to get clarification on tax matters. 2.48 However, a number of concerns remain, especially for larger firms and exporters, other than garments, that report more issues. For instance, in 60 percent of the tax inspection visits (up from 42 percent in 2003), firms had to pay informal fees and these informal fees have increased. However, the nature of these payments ­ which could include payments to private intermediaries whose capacity to advise on tax issues is unproven ­ is not identified by the survey. Firms also indicate that the cost of registering for tax and VAT has increased. In addition, despite a slight improvement, most firms do not report all their revenues (some of them because they indeed benefit from tax holidays) and few firms have audited accounts, reflecting the limited capacity of the private sector in accounting. This also reflects the fact that many firms, and especially smaller firms, use the estimated tax regime. This differs from the `real regime' (or self-estimated regime) in that the firm's profits are `estimated' and agreed on with the tax authorities. This creates obvious opportunities for under-reporting and informal fees and may explain why firms do not rate tax administration as a severe constraint. 2.49 Only 10 percent of firms view land as a severe constraint. This is consistent with findings from household surveys on the incidence of land disputes. But, despite the sharp increase (up from 3 percent in 2003), such a low incidence is at odds with the perception of major issues with land administration and dispute resolution. Part of the explanation might be the sharp decrease between 2003 and 2007 of firms owning land (down from 54 to 36 percent of firms) and owning buildings (down from 53 to 39 percent of firms). Through renting, fewer firms may be directly exposed to land disputes, while now being exposed to other risks (such as sudden rent increases). Also, as in the case of finance, the issue might not be a severe constraint for established firms, but is still a major barrier for new firms. 2.50 Regulations continue to be a significant burden for firms, with senior management spending on average 9.3 percent of its time on "requirements imposed by Government regulation" (Figure 2.9). This places Cambodia close to the bottom 25 percent 32 Despite the tax regime being business-friendly, Cambodia offers generous tax incentives to `qualified investment projects' (QIPs). QIP status, which is granted by the CDC, grants investors tax holidays, exemption from import duties for production equipment and raw materials, and exemption from most export duties. 34 in global performance. In addition, 27 percent of the firms identify regulatory uncertainty as a severe constraint. As noted above, although regulations on trade are even less transparent, 50 percent of Cambodian firms cannot access regulations as they want. Although only 11 percent of firms have delayed investments because of regulations ­ half because of excessive compliance costs, and half because of excessive uncertainty about regulations ­ the real impact is probably more significant given: (i) the risk-avoidance that this generates; and (ii) implications for the formal/informal economy. That said, across all regulations, firms complained less in 2007 than in 2003 and there are many positive examples of regulations that were supportive of private sector development (e.g. telecoms and finance, Section 1.C). Figure 2.9: Regulations a/ Senior management time spent in dealing b/ Procedures and regulations viewed as severe with Government regulation requirements (%) constraints (%) Cambodia 2007 - 5 10 15 20 25 30 2007, 9.3 2003 Cambodia 2003 Tax administration 14.8 Customs regulations 12.9 Registering a new enterprise 12.0 Procedures for access to land and premises 11.7 Th Vi La Business/sectoral licensing 7.7 Price regulations 7.0 - 5.0 10.0 15.0 20.0 25.0 30.0 35.0 6.9 Business inspections (of all types) 6.6 Note: the box represents the 25th and 75th percentile of Fire/Safety and sanitary regulations 4.8 the distribution and the middle line, the median. The Labor regulations extreme of the bar represents the min and max of the Standards and certification 4.4 distribution. Cambodia's performance in 2003 and 2007, as Environmental regulations 3.1 well as those of Lao PDR, Thailand, and Vietnam are also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. 2.51 Starting a business has Figure 2.10: Company Registrations improved in Cambodia.33 A 3,000 2,886 Commercial Law was enacted in 2005 to 2,784 streamline the startup process (see 2,500 USAID, 2007b, p. 10-16). This has 2,000 reduced the official fees for registration 1,500 (from US$600 to US$105 as of August 2007). Also all fees are now posted in 1,000 723 large print at the Ministry of Commerce 500 and a handbook on commercial registration (plus a website) has been - 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 developed by the MoC, with support Note: includes all new registrations of companies, from the ADB. These steps have led to investment, representative offices, and brand offices. an increase in company registrations Source: Ministry of Commerce. since 2003 (Figure 2.10). Nevertheless, in 2007, firms reported that the creation of a business has a median cost of US$240 at the Ministry of Commerce, US$300 each for tax and VAT registration, and US$630 for labor registration, with informal payments (`facilitation fees') in 46 percent of the cases at the MoC, 80 percent of the cases for tax purposes, and extremely high payments for labor registration. This is also evident from the first of the Doing Business indicators, on which Cambodia ranks 169th out of 181 countries (Annex D). There has also been progress for foreign investors, as the Council for the Development Cambodia acts as a `one-stop shop' with a statutory deadline of 28 days, as per the 2003 amendment to the Investment Law. However, survey data and anecdotal evidence indicates that more progress could be achieved as many 33 This paragraph and the next two focus on the regulatory constraints. Finding land and mobilizing finances are also potential constraints in starting a business and are discussed in paragraph 2.49 and paragraphs. 2.9-2.14. 35 firms still have to get permits from various line ministries, and informal payments are frequent and high there as well. 2.52 There are significant disparities in start-up costs across provinces. The Provincial Business Environment Scorecard (PBES) ­ a key benchmark tool for provincial economic governance and business friendliness developed by IFC (IFC 2007a) ­ highlights that the time and cost to set up a firm varies greatly between good performers. 2.53 There are a number of options to ease business entry. Based on the experiences of other countries, options for addressing this include the creation of a `one-stop shop' (for company approvals or for other regulatory approvals); the elimination of some steps such as obtaining a company seal and the publication of an abstract of company documents, which are steps less and less used around the world; and the elimination of minimum paid-in capital (which does not exist in Thailand, Lao PDR and Vietnam, as well as many other countries).34 The recent initiative ­ piloted in Battambang with support from the ADB ­ to decentralize the registry (and allow local authorities to immediately issue a temporary certificate) is also very positive. Full implementation of the Investment Law by CDC would also ease entry. 2.54 The number of licenses that firms need for operating is high (3.7 per firm). The Box 2-3: Setting up a Business in Cambodia Setting up a business in Cambodia can be difficult, expensive, and time consuming, creating a strong incentive to stay small and informal. For example in Kampot and Banteay Meanchey, the waiting time for Business Registration Certification was up to 30 days (PBES 2006). In Siem Reap, formal business start-up may involve registration and application for licenses from as many as five line departments and authorities. For example in 2007, to start a new restaurant in Siem Reap, the entrepreneur needed to get at least five licenses and permits (the Department of Land Management, Construction and Urbanization, the Provincial Hall, the Department of Commerce, the Department of Tourism, and the Tax Department). In some departments, the process took up to 30 days and the official fee could be up to US$300. Though central-level rules and regulations that govern business start-up processes across the country are the same, practices at the provincial level vary from one province to another. Administrative steps for the process of obtaining licenses and permits are mostly decided by the provincial departments of various line ministries. Provincial and local authorities sometimes also add additional administrative steps to the ones already required by line departments. The Royal Government of Cambodia recognizes these constraints and is committed to reform. At the national level, the Ministry of Commerce has introduced measures to reduce cost, simplify procedures, and make the registration process more transparent. Clear guidelines on the costs of registering have been established, significantly reducing official fees, but informal fees still exist. The estimated total cost of incorporation has dropped from almost US$1,000, to approximately US$300 and the minimum capital requirement has been reduced from CR20 million (approximately US$5,000) to CR4 million (approximately US$1,000). Registration application requirements were also streamlined by eliminating the need for a statement of conformity and local approvals, and the requirement that companies publish a newspaper announcement when they register has been abolished as well. In addition, the Ministry of Industry, Mines, and Energy has introduced a mechanism for regulatory reform. A Regulatory Review Team, an inter-ministerial team established under the SME sub- committee, will implement a licensing review and recourse mechanism to reduce the burden of licensing and inspection procedures for the private sector. This measure is expected to cut overlapping functions of different Government institutions at different levels and eliminate unnecessary licensing and inspection requirements, thus reducing the time and cost for business compliance. However, implementation has not started yet. At the provincial level, an effort to re-engineer the business start-up process has been piloted in Siem Reap. The Siem Reap provincial administration, with support from IFC, has streamlined four administrative procedures related to new business registration and licensing: (i) the Sole Proprietorship Business Registration; (ii) the Operating License for a Hotel, Guest House and Restaurant; (iii) the Operating License for an Agricultural Inputs Supplier; and (iv) the Letter of Exploitation of Place (location permit). Recent data suggest that this has significantly reduced the time and cost to start up a business in Siem Reap Province. For example, the number of steps required in registering a sole proprietorship at the Department of Commerce are down by 25 percent (from 12 to 9 steps) and the cost down by 70 percent (from US$150 to US$45). Source: IFC, The Provincial Business Environment Scorecard in Cambodia (2006), and MIME, Cambodia Small and Medium Enterprise Development Program (2006). 34 The minimum capital requirement has been reduced from US$5,000 to US$1,000. 36 Figure 2.11: Mitigating Risks of Commercial Disputes a/ Share of sales that are pre-paid (%) b/ Share of sales sold on credit (%) Cambodia Cambodia 2007, 20.7 2007, 22.4 Cambodia 2003 Cambodia 2003 Th Vi La LaVi Th - 10.0 20.0 30.0 40.0 50.0 - 20.0 40.0 60.0 80.0 100.0 Note: the box represents the 25th and 75th percentile of the distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. Cambodia's performance in 2003 and 2007, as well as those of Lao PDR, Thailand, and Vietnam are also displayed. Source: Cambodia surveys, and www.enterprisesurvey.org. setting up of a restaurant is a good case study (Box 2-3). Only 11 percent of firms identify permits and licenses (e.g. a construction permit, an import license, etc.) as a severe constraint ­ most likely a reflection of the fact that most firms acknowledge that the delays to get these licenses are not too high. Nevertheless, they find that in most of the cases, they are charged informal fees. The Government has created, with ADB support, and under the leadership of MIME, a process of licensing reviews and this needs to be implemented. 2.55 Inspections also remain widespread and costly. The average number of tax inspections went down from 7.2 to 2.9 per annum (and the average fine declined from US$4,700 to US$230); but the incidence of informal payments rose from 42 to 60 percent of cases, and the average value rose from US$100 to US$390. Larger firms (and in particular garment firms) are more subject to inspections and informal payments. 2.56 Trust in the judicial system remains limited. Only 25 percent of firms are "confident that the judicial system will enforce their contractual and property rights in business" (an insignificant improvement over 2004). Courts are widely perceived as corrupt. According to the Doing Business indicators, it takes 401 days to resolve disputes through the formal court system and costs are approximately 102.7 percent of the claim (on this indicator, Cambodia ranks 136th out of 181 countries). Nevertheless, the proportion of firms that see the legal system as a severe constraint to their operations went down significantly, from 31 percent in 2003 to 11 percent in 2007. Three explanations can be proposed for this decline: Concerns related to crime and safety have improved significantly. This is evident from the proportion of firms rating this as a severe constraint to business. Also, losses due to theft and vandalism amount to 0.6 percent of sales, against 1.6 percent in 2003, and firms pay 1.2 percent of their sales in security costs, against 13.9 percent in 2003. Firms report very few disputes with the Government (less than 2 percent of firms had a legal dispute in the last two years). Of this very small amount, only a handful of disputes go to the court. Almost half are solved through arbitration or mediation, and half are not resolved. Firms also have few non-commercial disputes, given their limited knowledge of the legal system (in part the result of the recent proliferation of new laws and the absence of a central repository for laws, decrees, and other regulations) and the low level of legal expertise (in 2006, there were only 20 law firms in Cambodia, none of them with more than five lawyers, only one notary, and 150 Cambodian judges, cf. USAID, 2007b). Disputes mainly concern contractual issues: 16 percent of firms report payment delays of more than a month on average ­ this is up from 11 percent in 2003. But only 0.3 percent of the cases were reported to the courts. Also, until the recent enactment of the Civil Code, commercial contracts were governed by an obsolete 1988 decree and there was a weak tradition of written contracts. The excessive interpretation of breach of trust further reduces the incentives to use the court system (Box 2-4). Hence, it is likely that firms do not identify legal issues as a severe constraint because they are simply not using 37 the legal system, even though commercial disputes are numerous. To mitigate this risk, firms use two approaches: (i) they have increased the proportion of pre-paid sales and reduced the proportion of sales on credit (Figure 2.11), hence preventing disputes35, and (ii) they are using alternative dispute resolution mechanisms, hence coping better with disputes (Box 2-5). Box 2-4: Breach of Trust Breach of trust is a criminal offense against property that occurs when the accused person, who lawfully possesses property under an agreement with the owner, intentionally and permanently deprives the owner of the property. At issue is the way Cambodian courts apply the breach of trust provision when there is a contractual dispute. Failures to fulfill contractual obligations (i.e. in a sales contract) do not constitute a breach of trust offense. However, there are many incidents where the courts treat simple contractual disputes as ones involving criminal breach of trust. The reason is simple. Disputes involving contractual arrangements are tried as civil cases, whereas disputes involving breach of trust are tried as criminal cases. When breach of trust cases are heard in criminal proceedings, enormous pressures are put on the accused to pay whatever is required to avoid going to jail, especially when the accused party is put in pre-trial detention, as often occurs. Based on a small-scale study by IFC, individuals who had been accused of breach of trust spent between US$5,000 and US$50,000 to get themselves out of a breach of trust case. Because business owners fear that they may be subjected to a breach of trust law suit, many of those surveyed by IFC said that they passed up business opportunities with people they do not know well. This is a cost to the Cambodian economy in terms of lost investments (estimated by IFC as between US$ 3 million and US$12 million). More important than the precise figure, is the negative signal that abuse of the breach of trust law sends to the business community. If investors are subject to improper breach of trust law suits, this could drive them away from investing in Cambodia. This further increases risk avoidance behaviors, including refusing contracts in order to limit risks. Source: IFC 2008. Breach of Trust: Its Improper Use and Impact on the Cambodian Economy. 2.57 In summary: Weak regulation and red tape continue to be a concern for firms. Although existing firms have found ways to deal with this (and, as a result, do not rate this as a severe constraint), this has a clear negative impact, including (i) risk-avoidance behaviors by existing firms, and (ii) reduced incentives for firm creation or formalization. This is true with respect to tax, firm creation processes, and courts. Similar issues were raised in earlier Sections on trade and credit. Three key themes emerge for improving this situation: (i) simplify (e.g. reduce the number of steps required to set up a firm or establish a `one-stop shop' mechanism for registration, and review and streamline licensing requirements); (ii) disclose (e.g. publish documentation on fees and licenses; clarify rules for the estimated tax regime); and (iii) solve disputes (the priority is to formalize alternative dispute systems, while a longer- term solution will consist of setting up a commercial court). Corruption 2.58 Corruption remains the most severe constraint identified by existing firms. More than 50 percent of firms rate corruption as a severe issue, with an even higher proportion for large firms (60 percent), exporters (66 percent), and foreign investors (64 percent), and in the garment (64 percent) and trade (59 percent) sectors. As noted above, however, this category can cover a wide range of issues and needs to be unpacked. 35 Among other things, this reflects the poor protection of creditors in the current system in Cambodia. Although a well-drafted Bankruptcy Law (with a clear hierarchy of creditors' rights, clear power and duties of administrators, and a rehabilitation process for saving viable businesses, USAID, 2007b) was enacted late in 2007, there is still no practice of insolvency. Full implementation of the law will require extensive dissemination, training of judges and administrators, and reform of the courts. It will also require addressing to create trust in administrators. 38 Box 2-5: Alternative Dispute Resolution In any country, when faced with disputes that arise from failure to fulfill contractual obligations (including failure to pay money owed), lack of clarity of the terms of the contracts, or any dispute related to business operations, parties generally negotiate to find acceptable settlements for all concerned. If negotiations fail, parties generally turn to their respective trusted third party or authority figure (usually Government officials and police officers) to help resolve disputes in their favor. This normally comes at a price, both financially and in terms of a favor owed. This practice leads to the expectation that winning a dispute is a function of knowing the person with the most influence or power and not on the merits of the case. For a party that does not have high social and political standing or cannot afford to buy influence, `cutting one's losses' by walking away from a dispute is the only alternative. None of these options are ideal for business people operating in Cambodia. To manage risks, businesses are reluctant to enter into contracts or commercial arrangements with those whom they do not know or who have not been referred by trusted parties. This limits the potential of their business. To foster the growth of business, a fair, transparent, less costly and less time consuming conflict resolution system is imperative. One option is to promote the use of alternative dispute resolution (ADR), a system of resolving disputes outside the purview of the court (see also para. 2.23 about ADR for labor disputes). The Government clearly recognizes this need and in 2006 adopted a Law on Commercial Arbitration to promote the settlement of disputes through arbitration. With support from development partners -- including IFC and the ADB -- to implement this law, the Government is currently developing a sub-decree on the establishment of a National Arbitration Center (NAC.) The sub-decree is expected to be adopted by the end of 2008. Soon after, an arbitration center will be in place to provide arbitration services to the business community in Cambodia. While establishing this center is a critical first step, real improvement in commercial dispute resolution can happen only if the center is independent of undue interference from the Government and its services are of high quality, comparable to those of reputable foreign arbitration centers. Independence requires that the initial set up of the center (selection of the first batch of arbitrators and organizational arrangements) is undertaken properly, in a transparent manner. To ensure high quality, arbitrators must receive the best possible training available and continue to receive such training on a regular basis. Moreover, the center should be well-promoted from the start to ensure that the private sector uses the arbitration services on offer. Source: IFC 2007. Business Issues Bulletin: Commercial Disputes and the New National Arbitration Center 2.59 The incidence of corruption is perhaps best known in the area of red tape (Table 2.4). More than 8 out of 10 firms claim being expected to pay to "get things done". As in the 2003 ICA survey, firms pay on average 5 percent of their sales in informal payments to officials. This practice affects a range of Government services, including connecting to an electric line and getting a telephone line. Corruption also affects many regulatory processes as discussed above, including taxation, business registration, trade processes, construction permits, etc. Finally, corruption is also reportedly high in the courts, although the survey itself does not generate data on this issue. The incidence of corrupt practices is particularly high in the garment sector, for larger firms, and for high performers. 2.60 A second important form of corruption occurs in contracting with the Government (Figure 2.12). Although the incidence of corruption has been reduced somewhat (below the levels observed in 2003 and below the median across developing countries), the average cost of this form of corruption has increased considerably to a median value of 8 percent of the contract. Foreign investors and exporters appear more subject to this practice than other firms. 2.61 The third area relates to the influence on policy decisions and interpretations. One-third of firms feel that "people with personal connections to political persons have major influence". 2.62 Unpacking the many forms of corruption, their incidence, and firms' response, helps reconcile strong growth with firms' complaints about corruption. First, the incidence of corruption varies across types of firms. Smaller firms, for instance, have been less exposed to some `red tape' corruption, limiting the cost to them. Second, firms have 39 developed risk prevention strategies to limit their exposure to potential corruption.36 Avoiding risks and remaining informal are two obvious risk prevention strategies. Avoiding going to court by using alternative dispute resolution instead, or even in some cases, forgoing potentially profitable ventures, is another example. Well-organized groups (e.g. business associations) have also reduced risks through negotiating specific arrangements. For example, GMAC has negotiated lower export management fees. Third, firms have developed risk coping strategies, such as limiting the reporting of revenues to tax authorities. Figure 2.12: Corruption a/Firms identifying corruption as a severe b/ % of Firms expected to make informal constraint (%) payments to public officials (to "get things done") Cambodia Cambodia 2007, 51.7 2007, 84.1 Cambodia 2003 Cambodia 2003 La ViTh La Vi - 20.0 40.0 60.0 80.0 100.0 - 20.0 40.0 60.0 80.0 100.0 c/ % of Firms expected to give gifts to secure a d/ Value of gift expected to secure Government contract (%) Government contract (% of contract) Cambodia Cambodia 2007, 17.4 2007, 8.0 Cambodia 2003 Cambodia 2003 La Vi La Vi - 20.0 40.0 60.0 80.0 100.0 - 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Note: the box represents the 25th and 75th percentile of the distribution and the middle line, the median. The extreme of the bar represents the min and max of the distribution. Cambodia's performance in 2003 and 2007, as well as those of Lao PDR, Thailand, and Vietnam are also displayed. For chart d/, the Cambodian value for 2007 is the median value (with the mean being even higher, at 17% of the contract). Source: Cambodia surveys, and www.enterprisesurvey.org. 2.63 Unpacking corruption also helps explain how it undermines growth. The various risk prevention and coping strategies adopted by businesses suggest how the private sector can develop, despite corruption. But these strategies show, more specifically, the constraints that result from corruption. First, corruption is one of the factors explaining the high levels of risk described in this report and how entrepreneurs respond to risk. As noted elsewhere, risk avoidance strategies reduce efficiency and investment, in particular greater capital investment and investment in diversification. This is especially the case for foreign investors for whom the absence of a reliable court system and the perception of rampant corruption are strong deterrents. Second, corruption is one of the factors explaining the large size of the informal economy, a key risk avoidance behavior. Risk aversion is also a key barrier to higher productivity, diversification, investment, and revenue collection by the Government. It links as well to anti-competitive practices. These anti-competitive practices, which are ranked the third most severe constraint (Figure 2.1), weaken competition, and hence incentives for growth and productivity. Third, corruption adds to the cost of doing business in Cambodia (Table 2.5). In part, these costs are offset by generous tax holidays, but 36 In their seminal work, Holzmann and Jørgensen (2000) make a distinction between risk reduction (reducing the probability of a downside risk), mitigation (reducing the impact of a future downside risk), and coping (reducing the consequences after a risk has materialized). This highlights that risk of corruption has a direct impact (for example, the financial cost of informal payments), an indirect impact due to coping strategies (e.g. reporting only part of income to tax authorities), and an indirect impact due to ex-ante risk reduction or mitigation measures (e.g. using pre-paid sales to avoid disputes). 40 this has an obvious cost to the national budget and its capacity to address constraints such as lack of skills and lack of infrastructure. Table 2.4: The Many Facets of Corruption and their Impact Incidence and Direct Cost Indirect Impact Facets Overall Variations Risk prevention Risk coping 1. Red tape and 84% of firms expected to pay to Higher for hassle get things done; 5% of annual garment sales for informal payments to industry, high- officials performers, medium and large firms a. For services 30-90% of firms expected to make Ensure services informal payments depending on are delivered process despite informal payments (cf. time for grid connection) b. In regulations Informal payments and delays to More inspections Stay informal Only 59% of setup a business in garments and revenues reported Excessive inspections high-performers to tax authority Increased lobbying for tax holidays c. In courts Limited direct cost given very low High performers, Avoid courts (0.3% Limited risk coping use of courts exporters, of commercial given low use of foreign investors disputes reported courts more concerned to the courts) and about dispute use alternative resolution dispute resolution; system 52% choose not to go to courts because of bribes; use pre-paid sales 2. Government 17% of firms expected to pay for Higher for - Increases cost of procurement contracts; median value of exporters and goods and services payment is 8% foreign investors for Government 3. Undue 30% of firms feel that people with Engage in lobbying - influence on personal connections to high activities (17% of policy decisions ranking officials have major firms) and /interpretation influence on policies cultivating 32% of firms view anti- network competitive behaviors as severe Stay away from constraint activities linked to 27% of firms view policy/ regulations/stay regulatory uncertainty as severe informal constraint Based on 2007 ICS. 2.64 Understanding corruption suggests solutions. In the area of red tape, reforms such as improving transparency, simplifying regulations, and increasing accountability are the way forward. The recent progress in trade facilitation is encouraging. Reforming public procurement is critical, as is planned as part of the Public Financial Management Reform Program now under implementation. Finally, with regard to unfairly influencing Government decision-making, greater accountability is crucial, either through simpler processes, with less discretionary power, or through autonomous bodies such as the Arbitration Court. This obviously reinforces the need for broader governance reform in the areas of public sector management and accountability. 2.65 In summary: Corruption remains a significant concern for Cambodian firms, and this covers a wide range of problems, including red tape (in accessing utilities, complying with regulations, paying taxes, using the court system, etc.) and unfair influence over public procurement and decision-making. 41 Although corruption is an obvious financial burden, firms have mitigated and coped with it through a variety of strategies. These explain how the economy has developed rapidly in spite of the constraints caused by corruption. However, these strategies have their obvious downside and are contributing to: (i) the high cost of doing business in Cambodia; (ii) high levels of uncertainty and numerous risk avoidance behaviors; and (iii) the large size of the informal economy and low degree of diversification. Macroeconomic Uncertainty 2.66 Macroeconomic uncertainty Figure 2.13: Consumer Price Index emerged as a major issue in 2008 45 (rated as a severe constraint by 38 40 percent of firms, against 19 percent in 35 2003). An explanation for this concern 30 25 could be the level inflation at the time of 20 the survey, i.e. between October 2007 15 10 and February 2008 (Figure 2.13). A 5 second possible explanation relates to - (5) the global environment and uncertainty (10) about the value of the US dollar 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Headline Food Housing / Utilities Core (Cambodia has an almost fully dollarized Source: National Institute of Statistics. economy); and, fears of a decline in orders from Cambodia's main export market, the United States. Across firms, concerns are highest in the garment sector, among large firms, and among firms in the lower one-third tier of productivity.37 2.67 Continued sound macroeconomic management will therefore be a key driver of future growth. There were pressures for higher wages in 2008 (in particular a 10 percent increase of the minimum wage, as agreed in the first quarter). Yet, with a high proportion of value added allocated to labor, there is little room for wage increases to be absorbed through lower profitability. In other words, wage increases will have to be accompanied by productivity improvements. Continuing good management of the macro-economy is crucial to avoid a significant adverse impact on competitiveness, and hence on job creation and growth. Coordination and Other Market Failures 2.68 Finally, numerous market failures could prevent entrepreneurs from investing and innovating. These include the shallowness of the input market (which makes starting a new activity unprofitable because the business must also invest in all the inputs or import them at high cost); lack of protection for investors (once a project has demonstrated viability, others will copy it, hence reducing returns), and anti-competitive practices (see para. 2.63). 2.69 Lack of diversification is a symptom of such market failures. Inadequate inputs (including raw materials and business services) require entrepreneurs to invest in more areas than they would want, hence making business prospects unviable. The example of the aluminum can company (Section 1.C) highlights how this firm had to become an importer for many parts and invest as well in electricity and water supplies. The fact that most innovations are undertaken by foreign investors or through the few Cambodian conglomerates, also supports this conclusion. Weak protection of intellectual property rights, which Cambodian firms complain about,38 is another symptom of such market failures. 37 An explanation for this latter group is that these firms are competing more on price, as their poor performance does not allow them to compete on quality; hence their higher sensitivity to inflation. 38 The main problem, which 39 percent of firms see with regard to their competitors, is that competitors "violate [their] copyrights, patents or trademarks". 42 2.70 There is limited connection between formal firms and the informal sector. Hotels and restaurants should be able to source food locally, and after 15 years, the garment sector should be able to source fabrics domestically,39 etc. This is a typical market failure where no individual establishment has the incentive to finance developing local suppliers, organizing the supply chain, and taking the associated risks. Yet, once such start-up costs are paid, many firms would benefit from such local suppliers. The example of the garment industry's packaging supplier spinning off into production of plastic bags (Section 1.C) is an exception. 2.71 The small business services sector is another challenge (IFC, 2005). This reflects, in part, the relative newness of the private sector in Cambodia, and hence the limited demand for business services. The 2004 IFC survey on business services providers also identified problems such as delayed payment and lack of knowledge about the value of business services. Yet, again, once developed, business services providers would benefit many firms. Options for addressing this problem include the establishment of a network of business consultants, providing training for them, and promoting them to business owners and managers. Business services providers would also benefit from actions that reduce risk avoidance behavior. 2.72 Limited knowledge of foreign markets and new value chains is another challenge. Building up knowledge about new markets and new products is expensive, and the investor that does so faces the risk of loosing returns once an innovation is proven successful. It is striking that the main innovation over the last 20 years (the garment industry) was developed by foreign investors and sparked by a policy decision (the quota agreement with the United States). It is likely that foreign investors will continue to bring in new ideas, and connections to new markets. Given the millions of dollars40 invested in support for the development of 20 different value-chains, the Government's proposed Value Chain Unit41 would be an effective way to make information on value chains public knowledge. The function of promoting exports ­ which is located in MoC's Department of Trade Promotion ­ can address some of these problems that are due to lack of knowledge. Nevertheless, global experience suggests: (i) providing a holistic approach to export promotion activities and bundling interventions (as opposed to fragmenting interventions along the value chain); (ii) combining this export promotion function with functions such as investment promotion and export financing; and (iii) targeting activities, including business services, (see previous paragraph) as opposed to targeting specific sectors. Also, as discussed above, there might be a place for a scheme financing research on export markets such as the MoC's Export Market Access Fund (EMAF). Finally, there will also certainly be a role for technological leapfrogging as discussed in the case of food packaging (Section 1.C). 2.73 Corruption and other governance problems are a challenge. `It takes two to tango' is a well-known saying with regard to corruption. A potential way forward is exemplified by the new Clean Business Initiative (launched in September 2008), which proposes to create a `good governance' umbrella for firms to act collectively against corruption. The same issue of coordination exists in many corporate social responsibility (CSR) areas. Recently, a number of Cambodian businessmen have developed an interest in CSR, which should be promoted through collective action. 2.74 Another challenge is Cambodia's image as an investment destination. Despite continued rapid growth and some success in terms of policy reforms and achievements (e.g. 39 In the case of garments, this is compounded by logistics constraints, which also make sourcing from the region expensive. 40 There was US$415 million (from 13 development partners, working with 6 different ministries) programmed in the area of Private Sector Development as of September 2007 (USAID, 2007c). 41 This is one of the activities proposed under the Trade SWAp (Box 3-1). This unit would be housed in the Trade Promotion Department of the Ministry of Commerce and would be a repository and promoter of value chain knowledge. 43 rapid development of the garment, telecoms, and financial sectors; increasing foreign investment), Cambodia is still regarded as a country where doing business is difficult. This is linked to international rankings, where Cambodia does not do well (135th out of 181 countries in the World Bank's Doing Business rankings, see Annex D; 110th out of 131 countries on the World Competitiveness scale). As an illustration, a survey of Japanese investors in 2005 showed that they were focused on: (i) security and post-conflict issues; and (ii) Angkor Wat. Taking symbolic action in the area of corruption; improving Cambodia's international rankings (as was done in 2008 for the Doing Business indicators); and accelerating the investment and promotion activities of the Cambodia Investment Board (CIB) are three options to improve the country's image for investors. 2.75 On the positive side, the investor community is generally satisfied with their capacity to engage Government on policy issues and comment on laws and policies. Although some of this access is informal, the Government-Private Sector Forum (G-PSF) is considered to be excellent institutionalized practice and has already had significant impact (Box 2-6). The G-PSF has the potential to work beyond its current trouble-shooting mandate. One option would be to pilot a more systematic approach by having one the working groups identify key strategic objectives (e.g. reducing costs and delays for imports and exports; making business entry easy for small and medium enterprises), and then pursuing these objectives in a systematic manner. Other possible options for private sector feedback include: having managers of SEZs give feedback on policy issues through the SEZ Board; having the CIB play a policy advocacy role, based on its investment servicing and Box 2-6: The Government-Private Sector Forum Cambodia's Government-Private Sector Forum (G-PSF), launched by the Royal Government of Cambodia in 1999, is a mechanism for public/private sector consultation on investment climate issues ranging from long range policy to day-to-day operations. The G-PSF gives the private sector a reliable means of raising and resolving problems, and Government a channel for getting private sector feedback on draft policies, laws and regulations. Twice a year, the G-PSF holds a formal, nationally-televised meeting which brings together the Prime Minister, key cabinet ministers, and some 600 business leaders, Government officials, journalists and development partners. Since these have the status of cabinet meetings, decisions made by the Prime Minister are binding. Throughout the year, eight private sector Working Groups (WGs) meet regularly to identify and prioritize common problems, and negotiate solutions with Government counterparts. These WGs concern: Banking and Finance; Tourism; Manufacturing and SMEs; Agriculture and Agri-business; Export Processing and Trade Facilitation; Energy Infrastructure and Transport; Law, Tax and Governance; and Industrial Relations. IFC has been supporting a G-PSF Coordinating Bureau since 2002. Through this bureau, IFC facilitates dialogue and informal exchange within and among the Working Groups, and broadly between the Government and the business community. IFC also conducts selected research to aid working groups in advocating for reform. On an ongoing basis, the WGs and Government counterparts are negotiating solutions for more than 300 different problems. In an independent evaluation in 2007, the Forum received high marks for organizational effectiveness and impact on the reform process. The evaluation also cited criteria for judging the Forum's economic impact. These include: An estimate of US$350,000 per year for pro bono consultation that business leaders provide to the Forum through their participation in the Working Groups. This compares very favorably with annual donor funding for the Forum of US$160,000. An assessment of a sample of only nine reforms out of the nearly one thousand raised with Government since 2002, found that the nine reforms resulted in nearly US$70 million in private sector savings. A return of US$105 for every dollar invested in the Forum by IFC and other donors. Since IFC's involvement began in 2002, the dialogue between all participants has matured and the private sector now shows sufficient capacity to directly engage with their Government counterparts without donor support. Therefore, over the next three years, IFC will work with key Cambodian business associations to develop their capacity to play a greater role in administering the G-PSF, commissioning research and undertaking advocacy. IFC is also working with other development partners such as USAID, GTZ, UNDP, and The Asia Foundation to expand dialogue at the provincial level and provide more support for research related to reform proposals. 44 promotion activities; and, institutionalizing user feedback surveys and surveys such as the ICS. 2.76 In summary: Issues of coordination and market failures appear pervasive, as should be expected in a small, nascent economy. This includes the issue of inadequate market information, lack of business services, etc. A range of collective actions, to be organized by the Government or sponsored by development partners, could help address some of these issues. This includes export promotion (or a value-chain unit) and investment promotion; as well as collective actions such as the new Clean Business Initiative. Since identifying all possible challenges is impossible, and these are also likely to change over time, having several ongoing feedback mechanisms (such as the very successful G- PSF), will be critical in identifying and designing solutions to the challenges identified in this report. C. Summing Up: Binding Constraints 2.77 This concluding Section highlights an approach to prioritize constraints. In short, the methodology reflects on the features of the private sector (in particular its low productivity; limited investment; large informal sector; and poor diversification), and identifies which constraints best explain these challenges. 2.78 Is finance the constraint? The first question is whether entrepreneurs do not invest (at all, let alone in diversification or productivity improvement) either because they have no good projects, or because they are unable to finance viable projects. In most cases, firms say they do not feel constrained by lack of finance. Many firms are foreign-owned and have adequate financing from the parent company or investors. Real interest rates were negative in 2008. That said, lack of finance could well be a constraint for more risky projects, such as venturing into new production, or if not backed by foreign investment or a Cambodian conglomerate with retained earnings. Lack of finance is also possibly a problem for small investors that would like to expand but do not have adequate capital. Export financing might may be a constraint too. In addition, lack of access to finance is likely one of the missing incentives for firms to formalize. In other words, while the financial sector would not be a major barrier, the absence of some specific products could be an important constraint. 2.79 Is it possible to generate high returns in Cambodia? If finance is not the key issue, then the problem must be that entrepreneurs do not identify high-return projects. This could result from one of two problems: (i) it is simply impossible to generate high returns in Cambodia; or (ii) while high returns are possible, entrepreneurs cannot realize these returns. Lack of skills and infrastructure are certainly constraining certain types of enterprises (such as more capital intensive manufacturing). Although currently a binding constraint, access to sufficient and affordable electricity is expected to improve in the near future. In the meantime, labor-intensive industries should continue to be a viable option in Cambodia, but: (i) industrial relations problems need to be addressed; and (ii) pay practices should be changed to improve productivity. In addition, trade facilitation and logistics are areas that seem to constrain most sectors. While progress has been made in trade facilitation at the border, this progress has not benefited all firms and more remains to be done. 2.80 Then, is the problem that entrepreneurs cannot get the returns they expect from projects? If, aside from concerns about labor costs/productivity and trade facilitation/logistics, entrepreneurs can conceive of high-return projects but still do not undertake them, the problem must be inability to retain their returns. Here a key, though recent, problem is high inflation. While good macroeconomic management has been a major Government achievement over the last decade, the conjunction of external inflation 45 (depreciation of the dollar and high food and oil prices) and domestic inflation (extremely rapid growth of credit) are shocks that have made inflation a major concern for most firms. Although improving, lack of access to regulatory information, cumbersome procedures, ineffective services, and corruption create uncertainty for most types of firms and add costs that reduce returns. Importantly, they also keep a vibrant informal sector from formalizing. These costs remain high by international standards (Table 2.5). Finally there are numerous coordination and market failures that make it difficult for entrepreneurs to realize good returns on their investments. Table 2.5: Cost of Doing Business Cambodia Cambodia China Kenya Tanzani India (2006) (2007) (2003) (2002) (2007) (2006) Production lost to crime 0.6 1.6 0.3 0.2 3.0 1 Payment for security 1.2 13.9 0.8 1.3 2.9 2 Bribes 4.7 5.2 1.9 2.1 3.6 3 Production lost due due power outages 3.8 2.2 2.0 7.8 7.1 10 Total 10.2 23.0 5.0 11.4 16.6 17 Source: Cambodia surveys, and www.enterprisesurvey.org. 2.81 These questions lead to a shorter list of binding constraints. Together with the features of the Cambodian private sector described in Chapter 1, five goals need to be achieved to reduce the problems currently constraining different types of firms and overall growth of the economy (Table 2.6): Goal #1: Encourage entry and diversification. Taking into consideration various types of entrepreneurs (foreign investors; established Cambodian entrepreneurs; small new entrants), there appears to be a need to facilitate formalization through much simpler and more transparent processes. Some entrants will need additional support for some public goods, such as standards, information about markets and technologies, etc. Diversification could be improved by access to affordable electricity. Goal #2: Reduce uncertainty. For established firms, in particular for foreign investors who may be less at ease in the current environment, several costs could be reduced to make `risk less risky'. Continued sound macroeconomic management is a sine qua non condition. Commercial dispute resolution is another obvious priority. Some improvements in the financial sector (such as improving the Credit Information System and implementing the Secured Transactions Law) would allow entrepreneurs to focus on business risks, as opposed to financial risks. Simplification of regulations and increasing transparency would also reduce the cost of doing business in Cambodia. Goal #3: Improve logistics. Another major way to reduce costs and uncertainty is to implement the trade facilitation agenda, in particular through rolling out the risk management approach and creating the National Single Window (based on the ASYCUDA system). More transparent procedures and fees would help pave the way for these major investments. The trade facilitation agenda will also need to be complemented by a broader effort to improve logistics. This will benefit not only exporters, but also domestic trade and services such as hotels and restaurants. Goal #4: Strengthen labor productivity. For the better performing firms, improving labor productivity is becoming a priority. Two immediate strategies are improving pay practices to stimulate productivity and improving industrial relations to avoid strikes. Improving vocational training should be a priority too, although achieving the full impact of this will take time, given the need to improve education over all. Goal #5: Deepen participation of the private sector in identifying and solving problems. Both the promise of the G-PSF and incomplete implementation of the 12- point trade agenda highlight the potential role of the private sector in identifying and solving problems. 46 Table 2.6: Characteristics of the private sector and constraints Characteristics Risk Largely Low International Low and of private avoidance informal diversification orientation heterogeneous sector behaviors productivity Constraints Barriers to Limited Limited FDI main entry and benefits and incentives and source of diversification high costs of few public diversification formalization goods supporting diversification; makes complex value-chains difficult High Lead firms stay informal or Diversification Strongly Weaken unproductive avoid taking productive risks would be an deters foreign incentives for risks additional risk, investors productivity hence is discouraged by high risks Poor logistics Leads to Makes diversification in Reduce Direct loss of high complex value chains difficult potential for competitiveness inventories (hence weakens incentives to further formalize) export- oriented FDI Barriers to Weaken Undermine Direct impact higher labor incentives for Cambodia's on productivity productivity formalization comparative advantage in low-cost labor Limited Uneven pace Difficult to Uneven pace Reduce improvements of reform get informal of cross- potential for in business weakens firms to cutting reform further FDI environment expected participate in makes returns and policy dialog diversification leads to risk more risky avoidance behaviors Note: each row explains how each constraint leads to observed characteristics of the private sector. 47 3. The Reform Agenda Key Messages: The Royal Government of Cambodia has embarked on a number of initiatives to improve the business environment. Many of these have been targeting some of the constraints discussed in Chapter 2. There are important lessons from such experience (as well as from the experience of other countries). This includes the critical role of public sector reform; the benefits of a time-bound action plan; the role of piloting, evaluating, and scaling up successful initiatives; and the decisive role of the private sector in identifying issues, designing solutions, and monitoring reform efforts. Develop specific short- and medium-term policy options and an initial set of monitoring indicators for each of the five goals of reform proposed in Chapter 2 by: (i) encouraging entry and diversification; (ii) lessening uncertainties to reduce costs; (iii) improving logistics; (iv) strengthening labor productivity; and (v) deepening participation of the private sector in identifying and solving problems. 3.1 Many of the issues identified in Chapter 1 and 2 are familiar to policy makers and the private sector in Cambodia, and a number of initiatives are ongoing to address these. This Chapter first reviews these initiatives (Section A). In light of the lessons learned from these ongoing initiatives, and from the priorities identified in Chapter 2, Section B outlines some possible key directions for reform. Section C concludes with some concrete immediate policy options to reach the objectives of a more vibrant private sector, with more investment, more products, and more jobs. A. Ongoing Efforts 3.2 The reform agenda in private sector development (PSD) is a key priority of the RGC, although progress is hindered by significant fragmentation. The Government's Rectangular Strategy and the National Strategic Development Plan (NSDP) have a vision of private sector-led growth, with a strong role for exports. Better governance, including in private sector development, is a key objective of these strategies. Nevertheless, despite important achievements, progress in the sector is constrained by institutional fragmentation on the Government side, with a large number of agencies involved and an elaborate structure to address this. This is compounded by significant fragmentation of efforts on the development partners' side (despite recent progress through bi-monthly meetings of locally- based development partners). 3.3 The PSD reform agenda is led by a Steering Committee on PSD, which contributes to partly overcoming institutional fragmentation. This committee is led by the Ministry of Economy and Finance (MEF), and includes all other relevant agencies (Ministry of Commerce; Ministry of Industry, Mines, and Energy; Ministry of Agriculture, Forestry and Fisheries; Ministry of Transport; Ministry of Tourism; Ministry of Public Works and Transport; Ministry of Social Affairs; Ministry of Labor and Vocational Training; the National Bank of Cambodia; and the Council for the Development of Cambodia). The committee meets on an ad hoc basis. Although the Steering Committee is expected to play the 48 role of a Technical Working Group,42 its meetings are usually internal for the Government (as opposed to joint with development partners and stakeholders). The PSD Steering Committee has three sub-committees.43 Development partners and stakeholders are usually invited, and participate actively in the sub-steering committee meetings. The PSD Steering Committee also works with the Government-Private Sector Forum to include the private sector's contributions to the reform agenda. 3.4 The Sub-Steering Committee on the Investment Climate and Private Participation in Infrastructure is chaired by MEF. Its meetings are not regular. It deals with investment climate issues, in particular legal and tax reforms, and with the regulation of public/private partnerships. Achievements in this area include: (i) enactment of the Law on Customs and many related regulations (and progress towards AFTA commitments in terms of trade duties); (ii) progress on the WTO legal reform agenda (including the Law on Secured Transactions; the Law on Bankruptcy; and the Law on Commercial Arbitration); (iii) initial implementation of ASYCUDA; (iv) establishment of several SEZs, each with a One-Stop Service; and (v) enactment of the Law on Concessions (which guides public/private partnerships in infrastructure). Current priorities include: (i) full implementation of the Law on Customs (including valuation issues; meeting AFTA commitments); (ii) design and implementation of the National Single Window (to meet Cambodia's commitment to join the ASEAN Single Window by 2012); (iii) further progress on the WTO legal reform agenda; (iv) adoption of the Sub-Decree on SEZs and development of existing SEZs (several have land allocated, but have not opened yet); and (v) adoption and implementation of the Sub- Decree on the Law on Concessions (consistent with international standards, including competitive bidding and disclosure). 3.5 The Sub-Steering Committee on Trade Development and Trade-Related Investment is chaired by MoC. It meets on a regular basis, including with development partners and with the G-PSF Working Group on Export Processing and Trade Facilitation, representing the private sector. Achievements in this area include: (i) progress in customs (see previous para.) and implementation of the Single Administrative Document (SAD); (ii) approval and initial implementation of a risk management framework (para. 2.30); and (iii) restructuring of the MoC. Current priorities include: (i) full roll-out of the risk management framework for exports; (ii) introduction of a flat fee for import and export; (iii) institutional streamlining and strengthening for the Sanitary and Phyto-Sanitary and Technical Barriers to Trade agenda; (iv) facilitation and streamlining of trade and company registration procedures and resolution of many trade-related bottlenecks; and (v) further institutional and capacity development reforms. This Sub-Steering Committee has also launched a new approach, the Trade SWAp, to coordinate efforts on the RGC side and align development partners' support (Box 3-1). Although still at an early stage of preparation, and short of certain features of a sector-wide approach, this is a promising initiative. 3.6 The Sub-Steering Committee on Small and Medium Enterprises is chaired by MIME. It sometimes meets with the private sector, such as the G-PSF Working Group on Manufacturing and SMEs. Achievements include: (i) some streamlining of business registration and adoption of a handbook on business registration (para. 0); and (ii) preparation of a licensing review approach. Current priorities include: (i) implementation of the licensing review; and (ii) establishment of the Institute of Standards. 42 Aid coordination in Cambodia is structured around 19 Technical Working Groups, chaired by a Government agency and facilitated by a lead development partner. The apex of the system is the Cambodia Development Cooperation Forum (held every year or year and a half), together with the Government Donor Coordination Committees (meeting three times a year). 43 These portfolios have some overlaps. For instance, progress in trade facilitation and investment promotion is discussed at the first two sub-steering committees. It is also unclear where the issue of business entry regulations belongs. The Sub-Steering Committee on SMEs has areas of focus overlapping with others since almost the entire private sector is comprised of SMEs. 49 Box 3-1: The Trade SWAp Building on the 2002 Integrated Framework (IF), the MoC, with support from development partners, led by the UNDP, finalized in 2007, a Diagnostic for the Trade Integration Strategy (DTIS). The DTIS reviews developments in key trade-related policies, including tariff and other restrictions; trade facilitation; legal reform; Technical Barriers to Trade (TBT) and Sanitary and Phyto-Sanitary (SPS) measures; Intellectual Property Rights (IPR); and investment promotion. It also reviews the global environment for Cambodia and concludes that tariff advantages are unlikely to drive future export growth. It derives from this, the need to diversify exports and export markets, and identifies 19 products and services for potential diversification. In view of the challenges of coordinating a significant number of Government agencies and development partners, the RGC has started to develop a Trade SWAp led by the Ministry of Commerce (MoC). As a vehicle to implement the DTIS, the Trade SWAp seeks to bring together activities in the trade area around a common monitoring framework and with a consolidated three-year rolling reform plan. It is structured to address issues in three broad areas called strategic `pillars': (i) Reforms and cross-cutting issues for trade development (legal reforms, trade facilitation, technical barriers to trade, improving sanitary and phyto-sanitary regulation and practice); (ii) Product and service sector export development (sector-specific reforms, with a focus on the products identified in the DTIS); and (iii) Capacity building for trade development and for the management of trade development (building competencies, institutional functionality, and information and accountability systems). The design of each pillar is led by a senior Government official, and facilitated by a development partner. Pillar teams meet on a regular basis to prepare logframes for each sub-program. These logframes, once vetted at the technical level, get approved by the Sub-Steering Committee on Trade. Most development partners in the sector have been trying to align their support to this emerging sector-wide approach. A small group of development partners (Danida, EC, and UNIDO) will pilot an approach that pools their funds for a flexible financing mechanism, administered by the World Bank (the "Trade Development Trust Fund"). 3.7 There are a number of other initiatives that relate to the PSD reform agenda. Most importantly, in the financial sector, a blueprint has been adopted (and revised in February 2007), with ADB support. It is a very ambitious plan that lays out key steps for the development of the financial sector. A number of other sectoral initiatives also have obvious connections with the PSD agenda, including those in agriculture and tourism. 3.8 One of the most important reform tools is the G-PSF (Box 2-6). Chaired by the Prime Minister, the G-PSF is a public-private dialogue mechanism which enables the private sector to directly raise concerns with Government and provide inputs on enterprise-related policies, laws and regulations. As noted above, some working groups of the G-PSF are connected to steering committees and have development partner participation. Represented by business associations such as the Cambodia Federation of Employers and Business Associations (CAMFEBA), the Garment Manufacturers Association in Cambodia (GMAC), the Cambodia Hotel Association (CHA), and the Freight Forwarders Association, the private sector can advocate with Government through the G-PSF. B. Key Directions 3.9 Based on an analysis of the main constraints to the business environment, Chapter 2 identified five priority goals: (i) encouraging entry and diversification; (ii) reducing unproductive risks to reduce costs; (iii) improving logistics; (iv) strengthening labor productivity; and (v) deepening participation of the private sector in identifying and solving problems. 3.10 Many of these problems have been persistent constraints for the private sector. It is therefore important to recognize the efforts made so far (Section A) and why the constraints continue to persist. Such analysis in fact meshes with international experience in managing private sector development reform (see World Bank, 2005, for instance). 50 3.11 A first lesson is the need for public sector reform. In the case of Cambodia, the capacity of the public sector is hampered by two major issues. First, is an issue of coordination and fragmentation. The example of standards is particularly relevant. The Ministry of Commerce (and several departments within the Ministry, including CamControl); the Ministry of Economy and Finance; the Ministry of Industry, Mines and Energy; the Ministry of Agriculture, Forestry and Fisheries; the Ministry of Health; the Ministry of Culture; as well as a few others, are all involved in the standards issue. Attempts to eliminate this barrier to effectiveness have been made through the creation of various committees (Codex Alimentarius; the PSD Steering Committee and its three Sub-Steering Committees; Trade SWAp management; etc.). But progress will not be significant without clear roles and responsibilities that avoid duplication. This calls for a functional review of all Government actors with overlapping responsibilities and strong leadership from the top to eliminate duplication. Rationalization of the various coordination committees might also generate efficiency gains. Second, civil service reform has been a lagging area, with poor human resource management practices, poor accountability, and insufficient pay. 3.12 Some building blocks are being created to address this issue. Targeted efforts can have a high pay-off. For instance, the GDCE recently piloted an interesting approach (USAID, 2007b, p. 81-82). It recruited 124 highly capable officials through a competitive and transparent process, from more than 5,000 applicants. It adopted an official code of conduct in 2007, that is very clear and strict, and which officers must carry. It has a two-year rotation policy. More generally, the RGC has adopted the Merit-Based Performance Initiative (MBPI). Through a better definition of roles and responsibilities, stronger human resource management practices, and higher pay levels, this scheme helps address some bottlenecks. The RGC has also adopted the principle of Special Operating Agencies (SOAs), whereby entities within ministries are given enhanced autonomy, against clear performance criteria. Although untested, this is a promising approach. An ambitious reform program in the PSD area should make more strategic use of these instruments. 3.13 The second lesson from experience in Cambodia and elsewhere is the role of a time-bound action plan, with strong oversight, monitoring and disclosure of results. The slow progress (and failure in a few areas) of the 12-point action plan, is a good reminder that implementation must be subject to thorough monitoring, with pressure to stay on track from both political leaders and the private sector. Other areas have made less progress than trade facilitation due to lack of a way to monitor progress. With the steering committee structure, the RGC has developed a sound system of oversight. 3.14 The third important lesson is to build on successful experience. China is increasingly viewed as a model for piloting policy reform first, and then replicating on a broader scale if the pilot proves successful. Cambodia, through the creativity of many of its leaders, has tried a number of reform initiatives, but these need to be properly evaluated and, if appropriate, scaled up. 3.15 The fourth lesson is the role of a combination of Government leadership, external pressures, and private sector demand. The agenda derived from accession to WTO is a good example of a number of reforms achieved by the Government in recent years (such as the Law on Secured Transactions, which enabled Cambodia to improve its Doing Business ranking from 150th to 135th in the 2009 assessment). Pressures from the private sector, in particular the garment industry, and analytical work from development partners on trade facilitation seems to be paying off. Drive from the private sector through the G-PSF has also proved that change is possible, including in contentious areas such as labor regulations. 3.16 The fifth important lesson is that, more broadly, the private sector has a key role in identifying issues, designing solutions, and monitoring their implementation. Cambodia has already made significant progress since the early 2000s through the G-PSF. However, as discussed above, this progress can be deepened through (i) nurturing a similar 51 dialog at the provincial level, and (ii) making the public/private sector dialog more systematic at the national level. C. Concrete Priorities 3.17 Operationalizing the five goals through these key directions and lessons from experience suggests the following immediate priorities. This concluding Section highlights some policy options, mainly for the next 2-3 years, with some indication of a longer-term agenda. It also flags how progress can be measured. This is summarized in a table in Annex A. Goal #1: Encourage Entry and Diversification 3.18 To encourage entry and diversification, a first priority is to accelerate the process of simplifying business registration. Two options are available: (i) streamline steps (as part of the Doing Business Reform Memo, specific steps have been proposed such as the elimination of minimum capital requirements), or (ii) create a one-stop shop for this (possibly building on the experience of a one-stop window in Battambang, at the Ministry of Commerce). The creation of a one-stop shop for business registration has been one of the most frequent reforms documented by the Doing Business Indicators (World Bank, 2008b). In the same area, making processes transparent is critical, and handbooks and website dissemination should be further encouraged. An important area to improve is the quality of this service at the provincial level (see also goal #5). 3.19 A second priority is to make establishment of special economic zones (SEZ) more effective. At the moment, despite having 19 SEZs approved by the RGC, only a few are actually operating. SEZ `lessons from experience' are being captured by CDC, with support from IFC, and a law on SEZs is being prepared. In particular, the regulatory framework, which provides for CDC to revoke licenses if SEZ managers are not investing as planned, should be fully enforced. An evaluation of the services offered by the SEZs (such as the existing One-Stop Shop) should also lead to improving these services (e.g. streamlining processes at these One-Stop Shops and empowering them, when possible, to take action without the interventions of outside agencies). 3.20 A third priority is to greatly expand the information available to entrepreneurs. Creating and making the MoC value chain unit's work available is important. Mechanisms such as the Export Market Access Fund can also contribute. The investment promotion function of the CDC should also be greatly expanded. The Cambodia Investment Board strategy needs a thorough review to ensure it becomes an effective investment promotion agency. In particular, the CIB needs to refocus its efforts to not only serve investors as a one-stop shop for licenses, but also to become: (i) an effective promoter of Cambodia as an FDI destination, and (ii) an effective policy advocate within RGC to represent foreign investors' views on reform priorities (for instance through aftercare services to investors). 3.21 A fourth priority is to develop the effective public services related to setting and enforcing standards. The absence of standards (in particular sanitary and phyto-sanitary standards) is a major constraint for diversification into new sectors and products. Progress in the area has been hampered by the lack of a clear strategy and significant fragmentation of responsibilities. The Trade SWAp can play a major role in this regard to better coordinate Government actions and development partners' support. 3.22 A fifth priority is to improve access to and the quality of electricity, and reduce its costs. Although not reviewed in detail in this report, this issue requires careful planning to ensure that the ongoing investments in electricity generation and transmission translate into high-quality, affordable electricity. 3.23 A range of other options also need further exploration. 52 First, the generous tax holidays available in Cambodia could be reoriented to favor diversification (and productivity improvement). This could include moving from tax holidays to more active tax incentives such as accelerated depreciation or carrying tax deductions forward. Second, also in the tax area, several concerns have been raised that the import tariff structure (combined with informal practices that have led to difficulties in claiming VAT refunds) is a disincentive to diversification, with tariffs on raw materials higher than the tariffs on the related processed goods. Third, in the financial sector, further analysis should be undertaken to see how to increase risk-taking for innovative projects (most likely through a combination of building entrepreneurs' capacity as well as banks' confidence and capacity). Finally, the current focus on priority sectors (in the Trade SWAp) should be complemented by a focus on particular activities that facilitate entry (such as business services or logistics ­ see below on the latter). In particular, the lack of support services in training is raised by many firms. The lack of professional importers is also a concern, forcing businesses such as the aluminum can company to become importers in order to get the raw materials and parts required for production. 3.24 Progress in these areas can be measured by (i) the number of firms registered every year, and (ii) indicators of diversification into new sectors, including firm registration, exports, and FDI. This, in itself, requires improvements in the business registry maintained by MoC ­ or any other alternative (such as the census of firms planned in 2009 by the National Institute of Statistics). The planned value chain unit can also play an important role. Goal #2: Reduce uncertainty 3.25 To reduce risk avoidance, the first priority is obviously to continue macroeconomic stability. 3.26 The second priority is to accelerate the creation of a credible alternative dispute resolution system. The law is in place and the sub-decree needs to be adopted promptly. Key actions must be taken to build the credibility of the National Arbitration Center, including: (i) statutory autonomy from the MoC; (ii) sound processes for recruiting arbitrators (iii) extensive training for arbitrators; (iv) disclosure of judgments; and (v) outreach and capacity building for the private sector to encourage use of the NAC. 3.27 In the same area of legal issues, scaling up dissemination of new laws and regulations should be considered a priority. In recent years, some major legal changes have been enacted, in particular, most recently, the new Civil Code. These changes are not well understood by the private sector and information should be properly disseminated. A stock-taking exercise of WTO commitments should be undertaken to ensure that the remainder of this ambitious agenda is properly prioritized and implemented. 3.28 In the area of regulations and service delivery, mechanisms to increase accountability include simplification, transparency, and third-party monitoring. Some countries (e.g. in Eastern Europe) have used the so-called `guillotine approach' to drastically reduce the burden of regulations. This involves requiring all public agencies to justify every single regulation by a tight deadline and removing all un-justified regulations. At the very least, MIME's licensing review mechanism should be implemented. Transparency in fees should also be implemented, both for regulatory requirements and utility services (such as grid connection). Options include the use of websites to submit applications (cf. also analysis in Chapter 1, showing that the use of emails is negatively correlated with informality); or mandating that payments are made directly to a bank (with no cash handled by employees performing transactions); finalizing and disseminating the compendium of public services prepared by the Council for Administrative Reform; etc. The One-Stop Window being piloted by the Ministry of Interior provides a good example of transparency and 53 simplification, and also facilitates third-party monitoring since all applications can be lodged in the same office (where civil society can review). Finally, the practice of `rulings' (in particular on tax administration) could be developed to reduce uncertainties and increase transparency. 3.29 A longer-term agenda, which needs to get started now, is reform of the Contract Law, establishment of the commercial court, and more generally, reform of the court system. This area is not reviewed in detail in this report, but will be critical in the longer term. 3.30 Progress in the area can be measured by subsequent investment climate surveys, measuring (i) informal payments and other similar costs, and (ii) risk avoidance symptoms (such as levels of inventory, of pre-paid sales, etc.). More intermediate outcomes can be measured by assessing the usage and effectiveness of the National Arbitration Center, and through private sector feedback. Goal #3: Improve Logistics 3.31 The unfinished agenda of trade facilitation is the main priority to improve logistics. First, the National Single Window should be designed and implemented to meet the ASEAN deadline of 2012. There are three immediate priorities: (i) roll out ASYCUDA, from the Port of Sihanoukville to the four other main customs offices; (ii) design the National Single Window, based on the platform of ASYCUDA, to meet private sector requirements and be in line with the ASEAN approach; and (iii) automate permit-issuing agencies other than GDCE (probably starting with MoC). 3.32 Second, transparency needs to be greatly improved. The export booklet already prepared needs to be more widely disseminated, and the private sector needs to be given opportunities to send feedback on its implementation (obviously the value of such a booklet depends on its actual implementation). Other guides should be published on imports, standards, etc. More fundamentally (and to meet Cambodia's obligations to WTO), a Trade Information Website should be created, connecting existing websites and adding new material to provide a convenient repository of trade-related information. 3.33 Third, the risk management strategy needs to be fully implemented. GDCE's approach to risk management, now piloted at the Port of Sihanoukville with ASYCUDA, needs to be monitored so that GDCE can fine tune the parameters in the ASYCUDA system to optimize both the facilitation of trade and the protection of Cambodia. GDCE's pioneer role needs to be extended to other agencies, including CamControl, MoC, MoH, MIME, and MAFF. 3.34 There is a range of options to further improve logistics, including making the SEZs work; improving the efficiency of the Port of Sihanoukville; and seizing quick-wins by simplifying rules for GMS intra-trade. A range of other options and quick-wins are also very possible. Given the importance of logistics in firms' cost structure, the Government might want to consider a dedicated in-depth review of logistics in Cambodia. This would generate a prioritized short- to medium-term agenda beyond trade facilitation. 3.35 Progress in the area can be measured by (i) costs (financial and time) for imports and exports; (ii) private sector feedback on transparency; and (iii) administrative data on the proportion of containers inspected at the border. Some of this will require repeated investment climate surveys. Higher frequency data could be obtained through user feedback surveys. Administrative data will come from the GDCE and the ASYCUDA system and should be made available transparently. Goal #4: Strengthen Labor Productivity 3.36 The short-term priority to strengthen labor productivity seems to be reform of industrial relations, particularly in the garment sector. The first priority is to make the 54 system of `the most representative union' work; the second is to ensure the development of collective bargaining agreements. 3.37 A second priority is generalizing productivity-enhancing pay practices. A few garment firms have introduced pay-by-the-piece practices. This needs to be generalized, possibly through proactive encouragement by GMAC (and possibly BFC to show the upside in terms of labor standards). An in-depth review might be necessary to understand possible bottlenecks in such generalization. 3.38 Third, enhancing skills needs a two-pronged approach to address immediate needs of the private sector.44 On the one hand, a key constraint that firms face relates to basic training, including overall attitudes towards work. This means very basic training on a massive scale. On the other hand, firms also face difficulties acquiring some very specialized skills, such as those of mechanics or accountants. Individual sectors, such as the garment sector also have specific skill needs. The challenge here is to find the most effective delivery mechanisms. These could include: creating training centers in SEZs; partnering business associations with training centers at universities; creating a new training institute managed by a coalition of private-sector firms (e.g. a public/private partnership similar to the Penang Skills Development Center in Malaysia); and broader development of technical and vocational education. 3.39 Obviously, the longer-term agenda is to prepare Cambodia for the medium-term through a gradual increase in skills. This calls for a much improved education strategy. In particular, the technical and vocational education strategy needs further work and added support in the budget to ensure implementation. 3.40 Progress in this area can be measured by: (i) labor productivity; (ii) the incidence of strikes; and (iii) private sector feedback on skills. Other indicators include the time it takes to fill vacancies, and feedback from young people about skills and jobs. Goal #5: Deepen Participation of the Private Sector in Identifying and Solving Problems 3.41 Further improving the Government-Private Sector Forum would deepen private sector participation in identifying and solving problems. The G-PSF could adopt a more systematic approach to reform, in addition to its current approach of: (i) solving problems faced by certain firms; and (ii) providing feedback on draft laws. This will require developing a sustainable financing mechanism that would provide the G-PSF with an adequate secretariat and research capacity. As a pilot, one working group could identify two or three areas of concern to its members (e.g. transparency for the export group and the law and tax group; or standards for the agriculture group); and, with inputs from expert studies and lessons from cross-country experience, seek to develop a comprehensive reform program and then monitor it. The G-PSF could also monitor broad issues such as competitiveness or the impact of development partners' work on value chains. 3.42 Under the framework of the Trade SWAp, product-specific groups could lead efforts to identify constraints, develop plans, and monitor their implementation. Development partner participation in these groups would enable rapid mobilization of resources. 3.43 This should be complemented by expanding the role of the SEZ Board, the CIB, and MoC's Trade Promotion Department in policy-making. These bodies are in direct contact with investors and exporters and should become advocates for reform within the Government. 3.44 Similar forums to the G-PSF should be developed at the provincial level. The impact and sustainability of ongoing experiments sponsored by various development 44Obviously, these points, which are based on the foregoing analysis, will need to be incorporated into a broader education strategy. 55 partners should be assessed before being scaled up. Adequate linkages between these provincial initiatives and the national G-PSF should be created and institutionalized. 3.45 There is also a role for single-topic private sector initiatives, such as the Clean Business Initiative. More generally, such `demand-side' initiatives ­ such as demand for better governance ­ are important to further development, and could include a variety of actions on corporate social responsibility. 3.46 Progress in the area can be measured by (i) further evaluation of the impact of the G-PSF (and other demand-side initiatives); and (ii) feedback from the private sector on its capacity to influence policies. 3.47 Pursuing these five goals, bringing all stakeholders on board, and monitoring accomplishments and impact will go a long way towards improving the investment climate. With fewer barriers, lower costs, and more support services, Cambodia should be able to get more investment, move into more products, and create more jobs. 56 Annexes 57 A. Summary Recommendations Table A.1: Summary of Recommendations Goals Recommendations Measuring progress and impact 1. Encourage entry and diversification, in particular through easing the cost of starting a business 1a. Ease business entry Short-term: - Adopt an action plan to simplify business entry, including through one-stop window; reduction in unnecessary steps; reduction in costs; abolition of minimum capital requirement - Implement action plan 1b. Accelerate development of Short-term: SEZs - Draft and approve law on SEZs, based on 2008 review of experience Medium-term: - Scale up SEZ implementation and make SEZs more effective 1c. Provide information and Short-term: Number of firms registered every year financing for diversification - Create Value Chain Unit in MoC Indicators of dispersion across sectors (for - Adjust Cambodia Investment Board strategy to give additional emphasis to investment firm registration, exports, and FDI). promotion and implement strategy (a priority could be the creation of an aftercare unit) Medium-term: This in itself requires progress in the business - Consider synergies between trade promotion and investment promotion activities and further registry maintained by the MoC ­ or any other improve promotion function alternative (such as the census of firms planned - Assess performance of Export Market Access Fund and scale up if successful by the National Institute of Statistics). The 1d. Provide public goods for Short-term: planned value chain unit can also play an diversification - Develop a strategy to create standards (technical and sanitary and phyto-sanitary) important role. - Review tax incentives and import tariff structure - Review constraints in financing innovation Medium-term: - Implement strategy on standards - Develop and implement a strategy to promote the provision of business services 1e. Improve accessibility, quality, and affordability of electricity 58 Table A.1: Summary of Recommendations Goals Recommendations Measuring progress and impact 2. Reduce uncertainty to reduce costs of doing business in Cambodia 2a. Maintain macroeconomic - stability 2b. Develop dispute resolution Short-term: system - Pass sub-decree to establish a National Arbitration Center (NAC) that is independent and effective - Set up and disseminate processes; train arbitrators to establish NAC's credibility; disclose NAC's judgments Medium-term: - Enact and implement Commercial Court Law - Draft, enact, and implement a new Contract Law - Reform overall court system 2c. Simplify and clarify Short-term: Measure through subsequent investment regulatory framework - Expand use of booklets, websites, and dissemination events to increase transparency (e.g. climate surveys: export processes; registration; licensing requirements) (i) informal payments and other similar - Accelerate implementation of MIME's licensing review process and agree on target to reduce costs and; licensing requirements (ii) risk avoidance symptoms (such as levels of inventory, of pre-paid sales, etc.). - Take stock of legal agenda and prioritize before introducing further changes Medium-term: More intermediate outcomes can be - Complete WTO legal agenda measured by assessing the usage and - Continue process of monitoring complexity of regulations effectiveness of the NAC, and through 2d. Increase accountability in Short-term: private sector feedback regulations and service - Consider drastic methods (e.g. `guillotine' approach) delivery - Develop transparency for regulatory requirements and fees, ministry by ministry: Options include the use of websites to submit applications; or mandating that payments are made directly to a bank (with no cash to the employee performing the transactions); finalizing and disseminating the compendium of public services prepared by the Council for Administrative Reform) - Continue expansion of MoI One-Stop Window and evaluate - Develop monitoring system for accountability in regulations and service delivery - Introduce the practice of rulings in tax administration Medium-term: - Replicate successful initiatives 59 Table A.1: Summary of Recommendations Goals Recommendations Measuring progress and impact 3. Improve logistics 3a. Further facilitate trade Short-term: - Roll out ASYCUDA to other GDCE posts - Prepare and adopt National Single Window action plan - Design and introduce flat fee for services - Initiate automation of permit-issuing agencies other than GDCE (possibly starting with the MoC) Costs (financial and time) for imports and Medium-term: exports - Implement National Single Window action plan to meet the 2012 ASEAN deadline Private sector feedback on transparency 3b. Strengthen transparency in Short-term: Administrative data on proportion of trade - Further disseminate export handbook and give opportunities to private sector to send containers inspected at the border feedback about its implementation - Develop other booklets on imports, standards, etc. Some of this will require repeated - Continue to increase transparency in regulations, processes, and fees through creating a Trade investment climate surveys Information Website 3c. Implement risk Short-term: Higher frequency data could be obtained management strategy - Initiate role-out risk management strategy in agencies beyond GDCE through purposive user feedback - Fine tune usage of risk management at the Port of Sihanoukville surveys Medium-term: - Continue implementation of, and adjustment to, risk management strategy Administrative data will come from the 3d. Develop logistics and other Short-term: GDCE and the ASYCUDA system and services - Undertake a comprehensive assessment of logistics services in Cambodia should be made available transparently - Continue reforms at the Port of Sihanoukville - Accelerate development of SEZs (see 1b. above) - Seize quick-wins by simplifying rules for intra-GMS trade Medium-term: - Implement strategy to develop logistics (lower costs, higher reliability, higher volumes) 60 Table A.1: Summary of Recommendations Goals Recommendations Measuring progress and impact 4. Strengthen labor productivity 4a. Improve industrial relations Short-term: - Support rapid identification of most representative union in each factory - Build capacity of firms and unions to develop collective bargaining agreements. Medium-term: Labor productivity - Continue strengthening labor dispute resolution mechanisms Incidence of strikes 4b. Increase incentives for Short-term: Private sector feedback on skills productivity - Support scaling up of pay-by-the piece practices in the garment sector with appropriate safeguards for workers Other indicators include the time it takes 4c. Enhance skills Short-term: to fill vacancies or youth's feedback - Develop strategy to address basic training needs ­ e.g. attitude to work ­ and specialized about skills and jobs needs (e.g. through supporting business associations; creating training centers in SEZs; or creating a public/private partnership approach) Medium-term: - Implement strategy 5. Deepen participation of the private sector in identifying and solving problems 5a. Develop monitoring tools Short-term: and other demand-side - Introduce reform scorecard with regular monitoring through the PSD Steering Committee instruments and the G-PSF - Introduce user feedback surveys on key processes (e.g. registration; trade) - Support private sector initiatives such as the Clean Business Initiative Medium-term: Further evaluations of the impact of the G-PSF (and other demand-side - Ensure sustained use of surveys, scorecards, and other demand-side interventions initiatives) 5b. Deepen existing Short-term: Feedback from private sector on its consultation mechanisms - Accelerate creation of provincial level forum for public-private dialog capacity to influence policies - Pilot a more systematic monitoring of progress in one of the G-PSF working groups - Increase participation of CDC (SEZ Board and Investment Promotion) in policy-making to benefit from their contacts with investors Medium-term: - Consolidate progress in national and provincial G-PSF 61 B. Summary Statistics In Tables B.1-B.8, the columns include: - overall average, weighted or not (W vs. NW); all columns other than the second are weighted; - a breakdown by size: small (5-19); medium (20-99); and large (100 or more employees); - a breakdown by sector (garments, trade, tourism and other); - a breakdown by exporter (yes/no) and foreign ownership (yes/no); - a breakdown by three groups of TFP performance, from low to high. Table B.1: Descriptive Statistics (Correlations) Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Sector Garments 2.8 19.6 0.00 0.08 0.91 100.00 0.00 0.00 0.00 0.98 0.90 0.48 0.29 0.24 Trade 28.9 20.8 0.81 0.16 0.04 0.00 100.00 0.00 0.00 0.02 0.14 0.40 0.31 0.29 Tourism 24.4 28.7 0.70 0.28 0.02 0.00 0.00 100.00 0.00 0.05 0.11 0.34 0.31 0.35 Other 37.8 23.7 0.65 0.28 0.07 0.00 0.00 0.00 100.00 0.14 0.32 0.28 0.37 0.35 Non-garment manuf. 6.1 7.2 0.66 0.30 0.04 0.00 0.00 0.00 0.00 0.19 0.11 All manufacturing 0.06 0.09 0.40 0.44 0.36 Size Small 68.8 40.7 100.00 0.00 0.00 0.00 0.34 0.25 0.36 0.09 0.11 Medium 24.3 30.3 0.00 100.00 0.00 0.01 0.19 0.28 0.44 0.07 0.36 Large 6.9 29.1 0.00 0.00 100.00 0.37 0.16 0.06 0.38 0.40 0.73 Exporter (1=Yes) 0.1 0.3 0.58 0.17 0.26 0.25 0.04 0.11 0.49 100.00 0.56 Foreign (1=Yes) 0.2 0.4 0.36 0.41 0.23 0.12 0.18 0.12 0.55 0.28 100.00 Performance Low 23.6 32.1 0.43 0.44 0.14 0.28 0.28 0.08 100.00 0.00 0.00 0.04 0.38 0.27 Medium 34.3 33.6 0.66 0.30 0.04 0.26 0.37 0.10 0.00 100.00 0.00 0.03 0.29 0.22 High 42.1 34.3 0.90 0.08 0.02 0.29 0.36 0.07 0.00 0.00 100.00 0.02 0.28 0.08 62 Table B.2: Investment Climate Constraints (% of firms rating each constraint as "severe" or "very severe" Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Skills 15.4 15.7 12.8 21.8 18.8 15.1 15.0 11.8 17.5 12.8 15.7 22.0 13.5 23.6 14.0 15.0 Labor 3.2 5.0 1.5 6.7 7.9 8.6 3.2 1.8 3.1 4.6 3.1 5.9 2.5 6.6 5.9 0.3 Anti-comp / informal 32.4 30.5 33.9 29.8 26.3 30.9 33.7 30.4 31.6 39.4 31.5 38.8 30.5 37.8 38.2 29.0 Telecoms 11.4 9.5 12.8 9.2 5.6 7.6 7.2 14.2 13.6 22.4 10.1 9.4 12.0 8.1 8.0 12.3 Electricity 21.1 22.4 21.3 19.9 24.4 25.8 17.9 26.0 18.1 27.2 20.4 23.7 20.4 22.1 28.2 14.0 Transportation 12.7 12.2 12.6 13.0 13.1 11.4 12.0 8.7 15.7 17.0 12.2 18.2 11.1 13.0 12.0 12.7 Crime 18.4 17.2 19.1 18.2 12.2 14.5 18.3 21.2 17.9 24.9 17.6 14.9 19.4 25.3 9.1 15.7 Legal 12.6 11.4 11.3 17.5 7.7 13.6 14.3 8.6 15.9 16.0 12.2 21.7 10.1 24.4 13.5 6.4 Land 10.1 10.0 12.6 3.2 12.3 11.9 14.2 11.3 5.1 8.4 10.3 7.4 10.9 10.0 4.3 7.0 Tax rates 16.1 15.3 16.4 12.7 25.2 14.0 11.4 11.0 24.4 30.6 14.3 20.8 14.7 18.9 14.5 12.3 Tax admin 13.6 12.7 12.7 14.6 18.0 9.6 10.4 8.8 18.8 29.6 11.7 20.8 11.5 20.1 11.8 11.8 Licenses/permits 12.4 13.1 11.0 13.3 22.5 15.0 8.9 9.5 17.8 12.0 12.4 11.3 12.7 14.7 16.2 10.3 Uncertainty 26.7 29.7 27.9 24.0 46.5 40.4 27.9 20.9 34.5 30.8 27.9 27.9 28.3 26.0 21.6 36.3 Macro 39.8 41.7 39.8 35.9 53.0 48.3 38.9 39.1 41.3 36.5 40.2 41.7 39.2 35.2 37.9 49.3 Corruption 53.7 55.0 52.8 54.1 60.3 64.0 59.2 50.3 52.3 65.5 52.2 63.6 50.8 53.4 56.9 55.7 Access to finance 15.5 14.3 16.4 14.7 10.0 14.6 18.0 13.5 15.2 14.9 15.5 12.1 16.5 16.9 12.6 19.7 Cost of finance 12.8 9.6 15.4 9.4 2.4 6.8 18.9 8.0 12.6 8.4 13.3 6.8 14.7 15.3 12.9 16.7 Customs/trade 9.4 10.0 9.5 8.8 10.6 17.3 10.4 4.3 12.2 19.1 7.9 12.4 8.3 14.0 10.2 18.6 63 Table B.3: Productivity and Performance Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Will increase 0.8 0.8 0.8 0.9 0.9 0.8 0.8 0.8 0.9 0.9 0.8 0.9 0.8 0.9 0.8 0.8 investment somewhat or drastically (%) Capacity utilization (%) 66.4 64.2 76.8 56.9 66.5 62.3 61.9 71.4 65.3 67.5 52.7 71.2 57.0 Labor prod (US$) 5019 5038 2284 6908 6221 2827 4007 3039 11474 5679 4811 9282 2640 3713 2865 1007 Wage per worker (US$) 1572 1552 1274 1848 1675 1172 1423 1370 2346 1484 1602 2057 1306 1947 1439 875 Capital per worker 9595 10453 8124 12177 8933 1367 8180 10173 16013 9014 9781 10983 8900 8436 3340 1867 (US$) TFP (see Annex E) 0.49 0.47 0.44 0.55 0.51 0.45 0.52 0.53 0.43 0.47 0.5 0.49 0.5 0.7 0.47 0.33 Table B.4: Finance Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low With loans (% of all) 20.8% 21.8% 19.8% 19.4% 36.2% 12.3% 24.3% 15.9% 21.2% 17.3% 21.3% 14.5% 22.6% 26.3% 18.1% 18.1% Turned down (% of those with loans) 3.2% 2.6% 3.7% 2.6% 0.6% 1.1% 5.4% 3.9% 1.5% 3.3% 3.2% 0.1% 4.2% 0.0% 3.6% 2.9% Reasons for no loan (% of those that never applied) No need 86.5% 86.9% 87.4% 82.9% 91.5% 86.3% 83.7% 90.8% 84.9% 90.0% 86.1% 91.1% 85.1% 92.1% 86.1% 78.3% Complex 4.5% 4.2% 4.4% 5.3% 1.7% 2.2% 6.7% 4.0% 3.9% 0.6% 5.0% 2.1% 5.2% 0.0% 7.1% 8.5% Collateral req 4.9% 3.7% 4.9% 5.8% 1.2% 2.2% 8.3% 0.5% 6.5% 0.6% 5.5% 0.3% 6.4% 5.1% 4.8% 8.2% Cost 4.1% 5.3% 3.2% 6.2% 5.6% 9.3% 1.4% 4.7% 4.7% 8.8% 3.5% 6.5% 3.3% 2.8% 2.0% 4.9% Corruption 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 64 Table B.5: Infrastructure Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Electricity costs c/kwH 0.208 0.086 0.206 0.206 0.275 0.273 0.219 0.214 0.196 0.201 0.209 0.235 0.204 0.218 0.194 0.203 % electricity from gen 26.2 38.3 20.5 31.7 33.3 54.5 19.3 29.1 23.2 35.6 24.7 34.6 22.9 37.1 20.4 23.6 % firms with gen 38.9% 49.8% 28.5% 56.3% 80.5% 81.5% 27.2% 50.7% 34.3% 50.4% 37.5% 51.7% 35.3% 44.7% 35.7% 31.4% Loss due to power 3.8 8.7 4.2 2.9 3.0 1.7 2.2 6.7 3.0 7.2 3.4 2.3 4.1 2.8 4.2 3.9 outages Proportion of imported 33.4 46.1 32.1 32.4 50.3 89.0 56.7 8.6 29.5 42.0 32.4 42.0 31.1 36.0 32.6 35.3 inputs (%) Security exp (% sales) 1.3 1.0 1.1 0.7 4.6 0.4 0.2 1.0 2.5 0.3 1.4 2.0 1.0 0.6 0.8 0.6 Losses due to crime (% 0.6 0.4 0.8 0.1 0.0 0.1 0.3 1.2 0.5 2.8 0.3 0.1 0.7 0.1 0.3 0.1 sales) Table B.6: Customs Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Average time to export 4.3 2.4 . 1.6 2.2 2.0 . . . 4.3 . 1.9 10.3 Longest time to export 5.7 3.0 . 3.3 2.7 2.7 . . . 5.7 . 2.6 14.3 % containers with 31.4 37.5 . 1.5 39.3 39.6 . . . 31.4 . 37.4 0.2 19.2 24.1 34.2 documentation problem for export Average time to import 5.1 4.2 5.4 4.6 5.2 2.8 5.8 . 4.4 3.4 5.4 4.2 5.6 3.6 6.1 8.4 Longest time to import 9.2 7.8 9.1 9.7 8.7 4.9 9.9 . 5.6 9.9 6.8 10.7 7.8 13.2 12.8 9.3 % containers with 36.4 36.3 37.6 32.5 39.2 35.8 37.4 33.7 40.6 35.6 32.0 39.4 10.9 46.1 34.9 documentation problem for import 65 Table B.7: Regulatory Issues and Corruption Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Consistent 0.2 0.4 0.2 0.1 0.2 0.2 0.1 0.1 0.2 0.1 0.2 0.1 0.2 0.1 0.2 0.1 interpretations of regs (% agree fully or in most cases) Number of licenses 3.7 5.6 3.1 4.8 6.2 7.8 3.0 4.5 3.5 5.5 3.5 4.7 3.4 4.6 3.6 3.5 Senior management % 8.9 20.7 7.2 12.7 12.1 11.3 7.2 8.9 10.1 8.2 9.0 7.2 9.4 14.9 5.4 10.1 time with regulations High cost of compliance 0.5 0.5 0.4 0.8 0.4 0.3 0.6 0.3 0.5 0.6 0.5 0.6 0.5 0.7 0.2 0.6 Unreliable regulations 0.5 0.5 0.6 0.2 0.6 0.7 0.4 0.7 0.5 0.4 0.5 0.4 0.5 0.3 0.8 0.4 Incidence of informal 0.6 0.4 0.6 0.7 0.9 0.9 0.7 0.6 0.6 0.7 0.6 0.7 0.6 0.8 0.7 0.6 payments (frequently, mostly, or always) % annual sales paid for 4.7 8.7 4.5 5.5 3.8 3.1 2.9 4.2 6.8 11.3 3.9 6.0 4.3 4.3 4.8 3.9 informal payments Share of revenues 58.6 30.8 52.0 70.7 70.5 72.2 54.9 59.3 60.7 62.4 57.9 65.2 56.8 64.4 61.9 60.8 reported for tax (%) Incidence of inspections 0.5 0.5 0.5 0.6 0.7 0.6 0.4 0.5 0.6 0.6 0.5 0.6 0.5 0.6 0.5 0.5 (all inspections) Total number of 4.8 9.3 3.7 6.1 7.4 12.1 4.0 6.2 4.0 5.3 4.7 5.1 4.7 7.1 4.8 5.4 inspections (all types) Cost of fines (all types, 200 2626 62 327 697 689 203 21 254 188 201 715 44 192 106 26 US$) Unofficial payments 354 2518 248 358 1075 1357 298 211 398 844 268 683 234 502 366 268 (US$) Contract value under 17.4 23.2 19.0 17.6 7.3 7.3 16.8 21.5 16.4 49.8 13.9 23.9 15.3 10.9 33.7 8.8 the table (%) Incidence of lobbying 0.2 0.4 0.1 0.2 0.3 0.2 0.1 0.2 0.2 0.2 0.2 0.3 0.1 0.2 0.2 0.1 66 Table B.8: Labor Overall Size Sector Exporter Foreign Performance W NW Small Med Large Garments Trade Tourism Other Y N Y N High med Low Training offered (% 67.9 65.1 67.4 67.3 74.3 53.2 68.5 62.3 72.9 72.0 67.3 78.7 63.1 60.0 79.6 62.4 firms) Maternity benefits (# 3.2 3.2 2.9 3.2 3.6 3.2 3.1 2.8 3.3 2.9 3.2 3.0 3.3 3.4 3.7 3.0 months) Maternity benefits (% 69.4 64.7 68.2 72.6 65.4 49.9 75.4 71.5 69.0 59.5 71.8 68.9 69.8 67.6 64.0 66.9 salary) Education (% below 16.3 24.2 17.0 11.8 25.6 55.4 10.8 22.4 10.6 20.2 15.9 12.4 17.4 11.6 11.4 19.9 grade 6) And ( % above grade 12) 33.2 25.5 33.8 33.1 28.2 3.5 26.1 24.2 50.9 48.7 31.3 47.4 29.3 40.0 34.5 22.8 % women 37.2 46.7 36.6 34.8 51.7 83.2 22.5 50.6 38.5 47.8 35.9 44.4 35.2 31.6 32.8 38.4 % skilled i.e. 0.6 0.6 0.6 0.5 0.6 0.8 0.6 0.4 0.7 0.7 0.6 0.6 0.6 0.6 0.7 0.6 management, professional, and skilled workers (employment; wages) 0.8 0.9 0.7 0.7 2.4 0.9 0.7 0.5 1.1 0.7 0.8 1.2 0.7 0.7 0.7 0.6 % workers in union 1.4 8.1 0.0 1.2 16.5 34.9 0.3 0.5 0.3 9.7 0.4 5.0 0.4 2.5 0.9 1.1 Incidence of strikes 1.5 4.4 0.7 2.4 2.9 8.2 0.5 1.9 0.1 6.4 0.9 2.8 0.6 0.5 2.1 2.7 (days lost) Employment w/o 110.6 110.6 110.4 110.4 113.1 110.5 111.5 106.7 112.7 109.6 110.7 111.6 110.3 constraints (100=today) 67 C. The 2007 Investment Climate Survey 1. The 2007 Investment Climate Survey was conducted by Indochina Research Limited (IRL) Methodology 2. The objectives of the sampling approach were to: (i) cover large parts of key sectors (e.g. garments); (ii) include a large sample of exporters (more than 120); (iii) enable panel data with the 2003 survey (this turned out to be impossible since no proper record of the firms interviewed in 2003 had been kept); and (iv) include adequate coverage of firms by size (small: 5 to 19 employees; medium: 20 to 99 employees; and large: 100 employees or more). The survey focused on five key urban areas including Phnom Penh, Battambang, Siem Reap, Kampong Cham, and Sihanoukville. It only included establishments that were in operation in April 2006 and had five or more full-time employees at that time. 3. Firms were drawn from a sampling frame composed of over 14,000 firms that was created from the following lists: (i) the sample of the 2003 survey; (ii) the registry of firms maintained by the Ministry of Commerce; (iii) the listing of firms undertaken for the 2006 PBES. 4. The sampling frame had two important shortcomings. First, because registered firms with 50 or fewer employees are included in both the PBES and the Department of Commerce lists of formally registered firms, the sampling frame consisted of many duplicates. Also, the firms included in the ICA 2003 were likely to appear multiple times in the sampling frames. Second, because the Department of Commerce list does not include good information on the sector and number of workers, the sampling frame did not provide the population shares across sectors and firm sizes. 5. Therefore, sampling was done in two steps. In the first step a simple random sample was taken to estimate the population shares across sectors and firm sizes, correcting for duplicates. 6. However, attempts to locate firms from the sampling frame proved extremely difficult. Many of the firms randomly drawn from the sampling frame could not be found or located (i.e. the contact phone did not work and the firm was not found when a messenger was sent to the firm's registered address).Of the 4,072 firms that were randomly selected from the sampling frame, only 877 (21.5%) could be located. Of these, only 835 firms (95%) could be contacted; the remaining 42 firms (5%) were closed. A total of 37% of the contacted firms refused to be interviewed. Only firms that met the screening criteria (i.e., were in operation in April 2006, had at least five permanent employees at that time, were a private company and were located in one of the five survey locations) were asked to participate in the survey. Of the firms contacted and in operation, which were willing to provide initial screening data over the phone or for which we already had accurate screening data from the PBES database, only 76.7% met the screening criteria. Based on the sector and size information for all firms meeting the selection criteria, the population shares were estimated. 7. In the second step, a sample was created from panel firms from the ICA 2003 and a random sample of newly selected firms from the sampling frame. In order to achieve the sampling objectives stated under point 2, firms from several segments were oversampled, particularly larger firms and firms in the garment and tourism sectors. Because it was difficult to locate enough of these firms from the sampling frame, additional firms were identified for interviews from the following lists: (i) the registry of firms maintained by the Garment Manufacturers Association in Cambodia (GMAC), (ii) a list of tourism firms provided by IFC, (iii) a list of medium and large businesses provided by IFC. 8. Of the 500 firms interviewed for the 2003 ICS, only 86 firms were interviewed again for the 2007 ICS. Part of the difficulty in securing more interviews with panel firms is that we were only able to locate and contact 165 of the 445 firms for which we had contact information (260 firms could not be located, and 20 firms were closed). The refusal rate of those firms that could be found and were in operation was 48%. It should also be noted that for the 86 firms that were interviewed in both 2003 and 2007, there was no unique firm identification number available to match the information from both surveys for these firms. 68 9. Using the estimated population shares from the first sample, post-stratification weights were calculated with respect to size and sector to make the second (and final) sample representative of the population of firms in Cambodia. 10. The questionnaire was essentially based on the 2003 questionnaire to ensure maximum comparability. A few questions were dropped after noting the low response rate in 2003; a few other questions were adjusted to allow comparison with other countries. A few questions were added at the request of the Government (e.g. on recent policy initiatives). Although the simplification of the productivity module turned out to work well, the questionnaire was eventually found too long (especially for larger firms) and should be further streamlined in future surveys. Fieldwork and Data Processing 11. Fieldwork involved many steps, from training enumerators and piloting the survey, to carrying out the survey with selected firms. To be selected for the interview, a firm had to meet the following screening criteria: (i) it was in operation in April 2006; (ii) it had at least five permanent employees in April 2006; (iii) it was a private company; and (iv) it was located in one of the five survey locations. Firms were asked a series of questions during an initial phone call, to confirm that they met the screening criteria. If they did, they were asked if they would be willing to participate in the survey. Information from the screening questions was maintained in a separate screening database. 12. Key steps and lessons from the fieldwork include: Need for longer than expected training of enumerators, due to the complexity of the survey. Delays in reaching the target of 500 firms also generated the need to train a new batch of enumerators, further adding to costs and delays. Initial pilot tests, carried out as part of training, were essential, but took much longer than expected as the initial interviews lasted over four hours, and interviews with some of the large pilot firms took two days. Although the questionnaire was subsequently somewhat streamlined, further shortening should be considered for future surveys. The questionnaire, tested in Khmer, was also translated into Chinese. Locating firms and confirming interviews was very time-consuming because of the poor quality of the sampling frame (the MoC registry) and the difficulty of getting proper contact details. There were also concerns on the part of firms about the objectives of the survey, the confidentiality of the data, and the qualifications of the enumerators. Financial data and questions on taxes and informal fees were particularly sensitive. To reduce survey time (on average two hours), multiple senior managers were interviewed in large firms (e.g. admin/HR; finance/accounting). A few firms refused to be interviewed, because they did not believe: (a) that the results would be useful; and (b) that their responses would remain confidential. Detailed quality control procedures were put in place: (i) observation forms were used to determine the quality of the interview by each enumerator; (ii) each interviewer was observed for at least one full interview by a fieldwork supervisor; (iii) fieldwork supervisors checked completed questionnaires within one day of their completion; (iv) survey materials were checked on a daily basis by the fieldwork manager/supervisors, before enumerators left for interviews; and (v) the fieldwork manager provided weekly updates to the operations manager and social research director on scheduling and fieldwork progress. 13. For data entry and cleaning of the data, a number of steps were taken, including: Using a computerized data entry system with built-in controls and entering at least 10 percent of all completed surveys twice, to ensure quality. Running frequency tables and crosstabs to check data range and consistency and ensure that skip patterns were followed accurately. Having batches of 100 questionnaires reviewed by World Bank staff as a further consistency check. 69 Weighting 14. Weights were created for the data to ensure that two objectives were met: (i) the weighted sample corrected for the under- or oversampling of the sampling strata, and (ii) the sum of the weights added up to the total number of firms in the population. It should be noted that the 2003 survey did not have weights. 15. The sample of surveyed firms was as follows: Sample - surveyed firms Size Small Medium Large Total Count Count Count Count % of % of % of tot % of tot tot tot Sector Manufacturing 10 19 105 134 0.02 0.04 0.21 0.27 Trade 68 31 6 105 0.14 0.06 0.01 0.21 Tourism 69 57 18 144 0.14 0.11 0.04 0.29 Other 57 39 23 119 0.11 0.08 0.05 0.24 Total 204 146 152 502 0.41 0.29 0.30 1.00 16. The weights are as follows (additional details available upon request; see also the distribution of firms in Table 1.1): Weighting - based on sample and population estimate Size Total Small Medium Large Count Sector Manufacturing 37.78 8.09 2.28 771 Trade 30.16 12.87 15.52 2,543 Tourism 23.48 8.92 1.90 2,163 Other 39.00 21.99 11.00 3,334 Total 6,272 1,919 620 8,811 70 D. The Doing Business Indicators C....... . _ . _ .... ,. .... . H- - . . . . . . . . . . . H........, / I I I I I I c .......· .·_ . _ _ _ ---" -'- - _.- --,- - · - - · -- -- ... -- · · · · - -.- - 71 ,_,, _ c-. _ -'- --. · -- ~~~~- ,," · -.-<-.. . ~ ~- ·· --. -'~-- -- ~~.-~- -" ·· ._.- -'-- · -~-- · -'-- · --_..- -~-- -_. · -- -- · · · .. -- --- ~~ '-'- - --- --.-... - · · -- - - -""- -- ~- -· · 72 -- --- --- · · --- - '- -- -- " " --. - .. - -- '" -, .. --....- ._-- .. " ,,--_. ---- " ---- , ~.--~- · ---- " ---- - " '- - --, ~.--- - -- .. " .. .-.- -- ~~ .- ~~ . - ------ --.. -- 73 E. Productivity Analysis 1. Productivity is commonly defined as a ratio of a volume measure of output to a volume measure of input (OECD 2001). Consequently, there are many measures of productivity, depending on the exact choice of output and input volume. These measures can, in general, be divided into two broad groups, namely single factor productivity measures and multifactor productivity measures. Single productivity measures relate the measure of output to a single measure of input, while multifactor productivity measures relate to multiple inputs. Labor Productivity 2. Labor productivity is the most well-known single factor productivity measure, and is given by the ratio of gross output or value added and labor input. It is a partial measure, in the sense that high levels of productivity not only reflect the productivity of labor, but also that of other inputs, such as capital and intermediate inputs. Nevertheless, labor productivity is a central component of economic welfare, given that changes in output per capita may arise from either changes in the labor force participation rate or changes in labor productivity.45 Given that labor force participation rates are bounded from above, large improvements in economic welfare will depend on labor productivity improvements. 3. Labor productivity is measured as value added divided by the average number of (permanent and temporary) employees (in fulltime equivalents). Value added is measured as total output minus costs for raw material and purchases of goods for resale, electricity, fuel, water, communication, transportation maintenance, and repairs. Total output is typically measured by total sales rather than total production because firms find total sales easier to report, inventory data are often poor quality and firms are often involved not only in production but also in trading activities. Moreover, given that purchases of goods for resale are typically included in the costs of raw materials, sales figures rather than production figures are appropriate to calculate total value added. 4. However, we find that the mean sales figures exceed mean production figures by 68 percent, which is exceptionally high according to international standards (0.46) At the same time, the share of intermediates in total sales is only 0.45 and only marginally higher at 0.47 for manufacturing firms. This suggests that the sales figures are too high or the intermediates figures too low. One plausible explanation for the high sales figures compared to the production figures is that firms are not only involved in production but also in trading to a significant extent. And if firms do under-report or exclude the costs of purchasing goods for resale from the reported costs of raw materials, then this also explains why the share of intermediates in total sales is rather low in the survey. And indeed, if we calculate our measures of value added and productivity, they are much too high if we use the sales figures, and they are plausible if we use the production figures. Therefore, we believe that total production is the preferred measure of output for the analysis. 5. We apply the following cleaning rules for the analysis (see World Bank 2006, pp.141-42). First, all firms with negative value added are omitted. Second, all firms with a share of 45 Where labor force participation rates are in full-time equivalents, controlling for changes in unemployment rates and hours of work. 46 The gap between mean sales and mean production figures varies somewhat across sectors (manufacturing 71%, trade 47%, tourism 44%, other 55%), but especially across firm size (<20 employees 19%, between 20 and 100 employees 38%, 100 or more employees 71%). 74 intermediates in output less than 0.1 or more than 1 are omitted where intermediates are the sum of raw materials and goods bought for resale, electricity, fuel, water, communication, transportation and maintenance, and repairs. Third, if the intermediates share in output deviates more than three standard deviations from the industry-mean (after applying the first two cleaning rules), then the observation is omitted as an outlier. Fourth, the bottom or top five percent of the (weighted) labor productivity distribution for each industry and year is eliminated. Finally, if we report figures for both 2005 and 2006, we omit any firms for which labor productivity increased or decreased by more than a factor five over the one-year period.47 6. The value added numbers were converted into (end of) 2004 prices using a manufacturing price deflator from the World Development Indicators. This conversion into 2004 prices was done to allow international comparisons at current price levels (see further). We included only those firms for which we have observations for both 2005 and 2006 to eliminate changes due to changes in the sample. 7. The analysis of wages is Table E.1: Labor Productivity (2004 US$) based on the total wage bill Mean Median N (including all benefits) 2005 2006 2005 2006 divided by the average Manufacturing 1,565 1,727 936 1,127 72 Food 2,323 2,977 1,109 1,231 5 number of (permanent and Garments 1,994 1,991 1,378 1,415 52 temporary) workers (in Other manufacturing 1,228 1,401 780 803 15 fulltime equivalents). Wages Trade 1,991 1,727 1,498 1,317 62 have been converted to Wholesale 4,027 2,975 2,092 1,859 19 Retail 1,440 1,389 1,284 1,213 43 2004 prices using the same Tourism 1,414 1,427 1,115 1,108 94 manufacturing price Hotels and restaurants 1,205 1,177 942 899 78 deflator, and the bottom and Travel services 2,068 2,211 1,609 2,065 16 top 5 percent of the wage Other services 2,369 2,348 2,011 2,011 62 distribution in each sector Transport 3,478 3,518 2,263 2,247 13 Other 2,051 2,012 1,801 1,908 49 have been trimmed. Also the Total 1,941 1,873 1,460 1,408 290 firm-level observations are Export status weighted by the number of Exporter 2,689 2,678 2,349 2,321 69 employees to estimate the Non-exporter 1,875 1,801 1,426 1,372 220 average wage per worker in FDI status FDI 3,437 3,112 2,298 2,130 93 each sector. Non-FDI 1,585 1,578 1,228 1,233 197 Firm size Total Factor Small (1-19 employees) 1,599 1,588 1,284 1,342 124 Productivity Medium (20-99 employees) 2,963 2,749 1,746 1,859 85 Large (100 or more employees) 2,081 1,903 1,801 1,580 81 8. Differences in labor Note: see Annex E for details. productivity reflect differences in the choice of factor intensities (because of differences in factor prices), differences in technology (production functions) and differences in efficiency (total factor productivity). 9. While labor productivity is the most well-known single factor productivity measure, total factor productivity (TFP) is the most widely used measure among the multifactor productivity measures. TFP addresses any effects in total output that are not caused by inputs or economies of scale, and is typically estimated as the residual in a production function estimate: (1) 47 There are valid observations for 316 and 362 firms in 2005 and 2006, respectively, and for 290 firms with valid observations in both years. 75 where Y is output, F a production function, M intermediates, L labor inputs, K capital inputs and the estimated parameters of the production function. However, straightforward OLS estimates of the production function parameters will be inconsistent because of simultaneity bias in the inputs ­ managers will choose inputs based on their private knowledge of productivity levels. Different approaches have been used in the literature to address this simultaneity bias. One possibility is to include firm fixed effects, but this requires that the (unobserved) productivity level is time-invariant. Another approach is to use instrumental variables, but valid instruments are often difficult to find. Olley and Pakes (1996) and Levinsohn and Petrin (2003) have developed an innovative nonparametric technique to control for the omitted unobservable productivity level. However, this approach relies on a structural model of optimal firm investment behavior whose assumptions remain untested. Also, Escribano and Guasch (2005) argue that the approach assumes a single unobserved source of simultaneity and a detailed sequential timing of the inputs decision which is more appropriate if we have high frequency data (daily, weekly, monthly). However, standard investment climate surveys collect data on an annual basis and productivity levels are assumed to depend on multiple (usually unobserved) investment climate variables. 10. We therefore follow the approach of Escribano and Guasch (2005) assuming that total factor productivity is a function of firm characteristics (C), investment climate investment variables (IC) and an idiosyncratic error term (u):48 (2) where and are the number of firm characteristics and investment climate variables included. We assume that the error term u is uncorrelated with the inputs, firm characteristics and investment climate variables, and therefore that any correlation between the inputs and TFP is captured by the firm characteristics and investment climate variables. As long as we include a sufficient number of investment variables and firm characteristics, this may not be an overly restrictive assumption, although panel data would be required to test the validity of this assumption. 11. We estimate a production function relating the total output of a firm to three inputs, namely intermediates, labor and capital, and including a number of firm characteristics and investment climate variables. Among the firm characteristics, we include the age of the firm (in log years), firm size dummies for medium (between 20 and 100 employees) and large firms (100 or more employees), a dummy for whether the firm exports (directly or indirectly), and a dummy for whether the firm is (partly) foreign owned. Among the investment climate variables, we select eleven indicators for infrastructure, red tape, corruption and crime, as well as for finance and corporate governance (discussed in detail in Section 3 below). We also include location dummies. We estimate two specifications of the production function, namely a Cobb-Douglas specification as well as the more general Translog specification. Also, for each production function, we estimate two variants ­ one without and one with sector-specific input coefficients.49 Statistical tests and residual analysis will be applied to choose the best specification among these alternatives. 48 Escribano and Guasch (2005) actually call the residual from the production function `multifactor productivity' rather than total factor productivity, given that the productivity term is not always fully separable from the production function (ftn.4). However, we make the common assumption of separability and interpret the residual term as total factor productivity. 49 Without a long panel there are insufficient degrees of freedom to estimate a firm-specific production function. 76 12. Total output is measured by Table E.2: OLS Production function estimates for total production. Total intermediates Cambodia, 2006 (standard errors in parentheses) are measured by the total cost of raw Panel A. Cobb-Douglas materials, electricity, fuel, water, 1 2 3 4 5 aggregate manufacturing trade tourism Other communication, transportation and M 0.69 0.72 0.76 0.6 0.7 maintenance and repairs. Labor is [0.02]*** [0.02]*** [0.03]*** [0.03]*** [0.03]*** measured by the total number of L 0.25 0.17 0 0.14 0.22 [0.04]*** [0.07]** [0.12] [0.06]** [0.06]*** permanent and temporary K -0.03 0.03 -0.05 0.01 -0.1 employees (in fulltime equivalents). [0.03] [0.02]* [0.01]*** [0.01] [0.03]*** Capital is measured as the estimated Observations 272 272 R-squared 0.97 0.96 sales value of machinery, vehicles F (p-value) 0 S and equipment. It is also assumed Panel B. Translog that all firms within each sector face 1 2 3 4 5 the same prices for output, M aggregate manufacturing -0.26 0.39 trade -0.29 tourism -0.34 Other 0.08 intermediates and capital, so that all [0.19] [0.48] [0.10]** [0.50] [0.47] prices will be captured by the sector- L 0.64 0.57 0.11 0.82 0.86 [0.18]*** [0.69] [0.77] [0.67] [0.17]*** specific intercepts.50 In the different K 0.16 0.03 0.68 0.15 0.03 country surveys, the value of [0.09] [0.14] [0.48] [0.06]** [0.22] machinery and equipment has been M2 0.06 0 0.09 0.06 0.05 measured as acquisition value, (net) 2 [0.01]*** [0.06] [0.02]*** [0.03]* [0.03] book value, sales (or market) value, L 0.04 [0.02]** 0.04 [0.03] 0.15 [0.14] 0.08 [0.05] 0.05 [0.03] and/or replacement value. K2 -0.02 -0.05 -0.04 -0.01 -0.01 Unfortunately, the exact [0.01]** [0.04] [0.02] [0.01] [0.01] measurement varies across the M*L -0.11 [0.01]*** -0.09 [0.05] -0.19 [0.14] -0.12 [0.05]** -0.13 [0.02]*** countries, and therefore there is no M*K 0 0.07 -0.03 0 -0.01 internationally consistent measure [0.01] [0.06] [0.02]* [0.02] [0.03] for the capital stock. For Cambodia, L*K 0.04 [0.01]*** 0.04 [0.02]** 0.13 [0.04]** 0.03 [0.02] 0.05 [0.02]** we have the sales value as well as the Observations 272 272 net book value, but the latter is R-squared 0.97 0.97 available for only a small subset of FCD (p-value) 0 0 0 0.23 0 FS (p-value) 0 firms and a poor indicator of the the economic value of the capital stock. Notes: the dependent variable is The log of total output in 2006. The inputs are measured in logs. aggregate regressions include Given that we are primarily sector and region dummies, the other regressions include region interested in the productivity dummies. Standard errors correct for clustering within location- sector groupings. * significant at 10%; ** significant at 5%; *** position of Cambodia vis-à-vis the significant at 1%. Null hypothesis of the F-test FS: equal coefficients other countries, we look at country for the input variables across sectors. Null hypothesis of the F-test differences in factor intensities for FCD: the coefficients for the quadratic and interaction terms of the input variables are jointly zero (per sector if sector-specific input all countries reporting the sales (or coefficients) market) value of the capital stock, namely Cambodia, Indonesia, the Philippines, Sri Lanka and Thailand. Because the sales value of capital is unavailable for 2005, the production function is estimated for 2006. For this analysis, we apply one additional cleaning rule, namely if the capital-labor ratio, intermediates-labor ratio, and/or capital-output ratio deviate more than three standard deviations from the industry-mean, then the observation is omitted as an outlier (see World Bank 2006, pp.141-42). 13. Table E.2 reports the OLS estimates for the parameters for the production inputs. Firm characteristics and (instrumented) investment climate variables were also included, as will be 50 Firm-specific prices are not available. 77 discussed in detail in Section 3.51 Panel A reports the Cobb-Douglas estimates, and the goodness of fit is high with a R2 in excess of 0.95. The coefficients for the intermediate and labor input variables are positive in the aggregate production function, as expected. This is also generally the case for the regression with sector-specific input coefficients. The coefficient for the capital input variable is, however, insignificant in the aggregate production function, and has conflicting signs across the sector-specific production functions, suggesting serious measurement error in the capital stock, biasing the coefficient downwards. Importantly, however, the F-test for the null hypothesis that each input coefficient is the same across the sectors, is strongly rejected ­ the four sectors are clearly heterogeneous in terms of production technology. 14. Because the Cobb-Douglas function is relatively restrictive, assuming a constant and unitary elasticity of substitution, we also estimate a Translog production function. The Translog production function can be viewed as a second-order approximation to any production function, and is therefore more flexible and also includes the Cobb-Douglas function as a special case. The regression coefficients for the input coefficients are reported in Panel B. The results show that many of the additional interaction and square terms for the inputs are significantly different from zero. Moreover, the F-test for the null hypothesis that these additional interaction and square terms are zero, is rejected for each sector, except for the tourism sector. This implies that the Cobb-Douglas function is too restrictive compared to the Translog specification. The F-test on the equality of each input coefficient (including the interaction and square terms) across sectors is strongly rejected as well. This implies also that the aggregate Translog specification is too restrictive compared to the sector-specific Translog specification.52 15. In Chapter 2, Figure 1.5 reports the resulting measures for TFP, based on the different specifications of the production function, as reported in Table 6. We do not report TFP directly but as an index of TFP given by , where is the highest TFP within a sector. This index indicates the efficiency of a firm compared to a firm with the highest TFP, keeping technology and input levels constant.53 International Comparisons 16. The data for international comparisons are from the PICS surveys and cover the period 2002-2006. For each survey we use the most recent year for which data are available (reducing bias due to recall error). In order to make the data comparable across years, we convert the monetary values to constant 2004 prices, using a manufacturing price deflator obtained from the World Development Indicators (WDI) for each country. However, for China, no manufacturing price index is available and we use the GDP price deflator from the WDI. These values are subsequently converted into U.S. dollars using the average exchange rate between the local currency and the U.S. dollar in 2004. Also, because in some countries the survey includes only firms with at least 10 employees, we limit the comparisons to firms with at least 10 employees. Table E.3: National Accounts and Labor Productivity 51 In cases where the instrumented investment climate variables were included, then the regressions were estimated with 2SLS, with OLS estimates (with corrected standard errors) from the second stage reported in table 9. 52 Also, only if we use the sector-specific Translog production function, the distribution of the error term (u) is virtually symmetric and normally distributed. For the other specifications the error distribution is asymmetric and left-skewed. This suggests that the sector-specific Translog production function captures most of the systematic factors affecting productivity. 53 78 17. We compare GDP per capita Overall GDP per worker Agri Industry Services Labor productivity (ICS) Mean Median aggregate labor Bangladesh 397 872 1,388 663 877 1,490 1,137 productivity across China Cambodia 511 2,016 1,045 3,388 511 n/a 2,140 n/a 1,724 n/a 1,714 6,487 1,314 3,189 countries. However, given Indonesia 1,636 3,342 2,655 3,851 3,164 3,946 1,887 that labor productivity Lao PDR Philippines 597 1,363 1,459 3,064 n/a 3,004 n/a 2,045 n/a 4,393 2,127 3,613 1,397 2,163 varies across industries Sri Lanka 1,356 3,221 2,515 2,346 5,327 2,312 1,389 and the industrial Thailand Vietnam 3,252 725 5,661 1,361 2,844 1,706 6,622 2,216 6,196 893 5,781 2,258 4,299 1,582 composition varies across Source: WDI for national accounts (2006 data) and investment climate surveys for countries, it is also useful labor productivity to do comparisons for given sectors. The best comparison can be done for the garment sector as this sector is well- represented for each country in the survey. Other important sectors covered in the Cambodia survey, particularly (retail) trade and hotels/restaurants, cannot be compared with most comparison countries because of limited sectoral overlap. Table E.3 also compares these results with data from national accounts (from the World Development Indicators). 18. As before, labor productivity is measured as value added divided by the number of (permanent and temporary) employees.54 Value added is measured as total output minus costs for raw materials, electricity, fuel, water, communication, transportation and maintenance and repairs.55 Output is measured by total sales, except for Cambodia, where it is measured by total production for the reasons stated in the previous Section.56 The same cleaning rules were applied as in the previous analysis for Cambodia only. Table E.4: Labor Productivity, Wages, and Mark Ups (2004 US$) Labor Productivity Wages Mark Up All Sectors Garments All sectors Garments All sectors Garments Mean Median Aggregate N Mean Median N Mean Median N Mean Median N Mean Mean Bangladesh 1,490 1,137 1,402 962 1,223 1,129 233 630 591 1,090 598 591 233 75.6 51.1 Cambodia 1,714 1,314 1,826 228 1,957 1,355 60 1,417 1,009 324 1,025 970 60 22.6 47.6 China 6,487 3,189 7,565 1,131 2,810 1,828 218 1,484 1,333 1,688 985 805 218 156.9 64.9 Indonesia 3,946 1,887 4,485 385 1,909 1,302 72 1,481 1,216 582 1,242 978 72 130.6 34.9 Lao PDR 2,127 1,397 2,589 121 1,881 1,555 12 824 748 167 1,072 973 12 93.3 43.0 Philippines 3,613 2,163 6,343 431 2,845 1,862 40 2,731 2,582 570 2,295 2,369 40 40.8 19.3 Sri Lanka 2,312 1,389 2,108 274 1,364 910 60 862 805 368 970 809 60 104.4 28.9 Thailand 5,781 4,299 7,353 1,061 5,601 4,306 147 3,309 2,882 1,246 3,001 2,488 147 57.5 46.4 Vietnam 2,258 1,582 2,440 900 1,761 1,380 59 884 817 1,025 779 749 59 86.9 55.8 Average across countries 3,303 2,040 4,012 5,493 2,372 1,736 901 1,514 1,331 7,060 1,330 1,192 901 85.4 43.6 19. International comparisons of TFP raise a number of issues. First, inputs and outputs should be measured in terms of volume, but we only know their monetary values for lack of suitable input and output price indices. Therefore, we need to assume that firms face the same international prices for their inputs and outputs (Dollar et al. 2005). Second, wages do diverge between countries, and therefore labor inputs should be measured in physical units. Third, differences in human capital may be corrected for by including the percentage of the workforce with primary, lower secondary, upper secondary, and higher education as input variables.57 Fourth, the estimated value of the capital stock should be accurate and comparable across 54 For Indonesia, there is information on the total number of peak season workers, but for an unspecified period, and no information on the average number of temporary workers in the year. Therefore, for Indonesia, we use the number of permanent workers only ­ biasing the estimated labor productivity in Indonesia upwards. 55 In case a firm did not report any raw materials costs, the observation was eliminated. For the other cost items, it was assumed that the amount was effectively zero if none was reported. 56 Sales figures are the preferred measure, and therefore used for all countries except for Cambodia, for the reasons stated previously. 57 For The Philippines and Thailand we include the average years of education of the workforce as proportions are unknown. 79 countries, and for the reasons stated before, this limits the comparison to Cambodia, Indonesia, the Philippines, Sri Lanka and Thailand. Fifth, because technology varies across sectors, one should control for sectoral differences in the production function or compare TFP within comparable sectors. The above analysis showed that this is an important issue for Cambodia, and therefore we will estimate a Translog production function for the garment sector. Sixth, all monetary values should be comparable, and thus, we convert them into constant 2004 prices and in US$. Seventh, one should control for firm characteristics and investment climate variables in order to avoid endogeneity bias in the estimation of the input coefficients. We include among the firm characteristics: the age of the firm (in years); firm size dummies for medium-sized firms (between 20 and 100 employees) and large firms (100 or more employees; a dummy for whether the firm exports (directly or indirectly); and, a dummy for whether the firm is (partly) foreign owned. We control for the investment climate by including location dummies, where locations are given either by cities (Cambodia) or provinces (Indonesia, the Philippines, Sri Lanka, and Thailand).58 Eighth, the raw data should be treated as the same across countries and therefore the same cleaning rules used for Cambodia also apply to the other countries. 20. Table E.5 reports the country- Table E.5: OLS Translog production function specific and pooled Translog estimates for garments sector in Cambodia, production function estimates for Philippines, Sri Lanka and Thailand (standard the garment sector in Cambodia, errors in parentheses) Variable Cambodia Indonesia Philippines Sri Lanka Indonesia, the Philippines, Sri Lanka M -0.58 -0.7 -1.7 1.33 Thailand 0.25 All 0.02 and Thailand. The F-test Fpool shows [1.23] [0.67] [1.22] [0.58]** [0.30] [0.14] that the coefficients for the input L 2.82 [2.36] 1.22 [0.87] 4.54 [1.01]** 0.58 [1.06] 0.76 [0.41]* 1.2 [0.19]*** variable are not significantly K -0.29 0.48 0.55 -0.67 0.04 0.04 [0.77] [0.42] [0.45] [0.41] [0.11] [0.08] different from each other and that M 2 0.08 0.1 0.42 -0.02 0.03 0.05 we can therefore pool the different 2 [0.03]** [0.07] [0.12]** [0.05] [0.02]* [0.01]*** countries. The pooled regression L -0.15 0.1 [0.18] 0.74 [0.16] 0 [0.27]* [0.13] 0.02 [0.03] 0.05 [0.02]*** shows that a number of terms K 2 0 -0.01 0.04 0.04 0 0 [0.06] [0.03] [0.03] [0.02] [0.00] [0.00] involving labor and intermediate M*L -0.14 -0.18 -1.15 -0.01 -0.05 -0.12 inputs are significantly different [0.16] [0.15] [0.34]** [0.13] [0.04] [0.02]*** M*K -0.02 -0.03 -0.25 0 -0.01 -0.01 from zero, but none of the terms [0.06] [0.08] [0.10]* [0.04] [0.01] [0.01] involving capital inputs are found to L*K 0.11 0.03 [0.14] 0.33 [0.10] -0.03 [0.12]* [0.08] 0 [0.02] 0.01 [0.01] be significant. This again, probably reflects problems with measurement Observations R-squared 49 0.93 56 0.99 25 1 56 0.99 132 0.99 318 0.99 error in the capital stock. The F-test F (p-value) pool 0.46 FCD indicates that the Cobb-Douglas F (p-value) CD 0 Notes: the dependent variable is the log of total output in constant specification is rejected in favor of 2004 prices converted in US$. The regressions also include controls the Translog specification. for firm characteristics, the investment climate and location inputs in significant at 21. Based on the above regression dummies. The 5%; *** are measured 1%.logs. *null hypothesis 10%; ** significant at significant at The of the coefficients, we can estimate the Fpool-test is that the coefficients of all inputs are the same across Null hypothesis mean TFP in each country, where countries. and interaction of the F-test FCD: the coefficients for the quadratic terms of the input variables are jointly TFP is given by equation (2). Figure zero coefficients) 1.6a uses a TFP index to measure the mean firm-level efficiency in a country relative to the mean firm-level efficiency in the most efficient country. 58 Alternatively, we can include firm-specific measures of the investment climate controlling for endogeneity using location dummies as instruments (such as in the analysis for Cambodia). Because the focus is now on cross-country differences in productivity, possible firm-level differences in the investment climate are ignored. 80 Impact on Productivity 22. Almost all decisions of a firm critically depend on the surrounding investment climate. The investment climate provides opportunities and incentives for firms to invest productively, create jobs and expand, but the high costs of doing business in a less-than-ideal business environment may also constrain firms (World Bank 2004). Firms that face much red tape and bureaucratic harassment when becoming larger and more `visible' might choose to invest sub-optimally (from a social welfare point of view), while infrastructure problems may reduce the efficiency of production and exchange. In this Section we therefore take a closer look at how the investment climate affects the productivity of firms in Cambodia. 23. The 2007 ICS includes a large number of indicators that can be used to measure the investment climate in Cambodia. Following the taxonomy used by Escribano and Guasch (2005), we include a number of indicators for three areas in particular, namely: (1) infrastructure, (2) red tape, corruption and crime, and (3) finance and corporate governance, and we analyze how productivity is affected by these three problems.59 The following table reports the mean level of a number of indicators in these three areas by tercile of TFP, where TFP tercile has been defined within each sector. Hence, the first tercile includes all firms that are among the one-third of most productive firms in their sector, while the third tercile includes the one-third of the least productive firms in their sector. Also, TFP is derived from the sector-specific Translog production function specification, as reported in columns (2)-(5) of Panel B in Table E.x. The analysis uses: Three indicators for infrastructure, namely: (1) the total hours of power outages in 2006, (2) the total hours of unavailable mainline telephone services in 2006, and (3) the average number of days in 2006 that it took a firm to clear customs from the time the firm's goods arrived in their point of entry (e.g. airport) until the time the firm could claim them from customs;60 Four indicators for red tape, corruption and crime, namely: (1) the number of days spent by the firm in 2006 on inspections or required meetings with Government officials from national, provincial or municipal agencies, (2) in a typical week, the percentage of time spent by senior management in dealing with bureaucracy/regulation, (3) the percentage of sales paid by a typical comparable firm in gifts or informal payments to public officials to "get things done" with regards to customs, taxes, licenses, regulations, services, etc., and (4) losses from theft, robbery, vandalism or arson (as a percentage of sales); and Four indicators for finance and corporate governance: (1) whether the firm has an overdraft facility or line of credit, (2) whether the firm has a loan from a bank or other financial institution, (3) whether the firm has constrained access to loans, and (4) whether the firm has hired an outside accounting/audit firm or individual auditor in 2006. A firm is regarded as having constrained access to loans if the firm does not have a term loan (more than 6 months) from a bank or financial institution, either because the loan application was turned down or because the firm never applied for a bank loan because it considered the application procedures too cumbersome, the collateral requirements too stringent, interest rates too high and/or because of corruption in the allocation of financial institution credit. 59 See table A.2 in Escribano and Guasch (2005). Some of their indicators are not available in the Cambodia ICA 2007. Also, some of the indicators are strongly related (e.g. there are five indicators related to power outages), but we include only one of them in the analysis. We do not include the variable `fraction of sales undeclared to the IRS' because it refers to the typical firm and is based on perception, rather than observation. 60 For firms that did not (directly) import goods in 2006, the last indicator has missing values. There are no missing values for the first two indicators because all firms had power and a mainline telephone connection. 81 24. Table E.6 reports the mean for each of these indicators by TFP tercile. In order to eliminate the impact of outliers, we have eliminated all observations in the top and bottom 5 percent of the distribution of all variables that are not dummies. Although the patterns are not always monotonic across the terciles, the table suggests a number of findings. First, the most productive firms face fewer infrastructure problems, particularly fewer power outages and fewer days to clear customs. Second, the most productive firms face more red tape and corruption, as indicated by more days spent on inspections, senior managers spending a higher percentage of time in a typical week dealing with bureaucracy/regulation, and larger payments needed to get `faster action' from bureaucracy. Third, the most productive firms score better in terms of finance and corporate governance ­ they are more likely to have an overdraft or a formal loan, are less likely to have constrained access to loans, and more often use an external auditor of financial statements. Table E.6: Impact of Selected Investment Climate Variables on Productivity (TFP) Average value of investment climate indicator, from most to least productive % Average (log) productivity gains and firms losses due to investment climate I II III Most Least Investment climate indicator Mean Low use High use productive productive Infrastructure Total duration of power outages (hours) 16.89 28.82 25.67 -3.67 -1.27 -5.77 Total duration of unavailable mainline telephone services (hours) 6.14 7.83 5.25 -0.2 0 -0.52 Days to clear customs for imports (if direct imports) 2.89 3.26 3.24 -0.8 0 -3.78 Red Tape, Corruption and Crime Number of days spent by the plant in inspections 1.2 0.46 0.52 0.01 0 0.02 Percentage of time in a typical week spent by senior management dealing 8.73 3.05 5.69 0.06 0 0.14 with bureaucracy/regulation Payments to deal with bureaucracy 'faster' (% of sales) 2.56 2.29 2.43 2.71 0.75 5.65 Loss from theft, robbery, vandalism or arson (% of sales) 0.04 0.04 0.03 -0.04 0 -0.14 Finance and Corporate Governance Firm has overdraft (dummy) 0.14 0.07 0.02 0.83 0 11.9 Firm has loan from bank or financial institution (dummy) 0.32 0.16 0.18 -0.2 0 -0.91 Firm has constrained access to loan 0.07 0.14 0.16 -0.05 0 -0.46 External audit of financial statements (dummy) 0.3 0.15 0.18 0.07 0 0.36 25. Although the above findings are suggestive, we cannot conclude that many of these investment climate variables indeed have a major impact on firm-level productivity in Cambodia. First, the table shows only bivariate correlations, while the investment climate variables are clearly correlated. Second, the observed patterns may also reflect other factors, particularly firm characteristics. Third, the observed correlations between productivity and investment climate variables may reflect reverse causality as well, with more productive firms `demanding' more investment climate `services'. 26. Therefore, it is useful to analyze the relationship between productivity and the investment climate within a regression framework. Note that the estimated efficiency levels in the previous analysis have been calculated under the assumption that they are a function of firm characteristics and investment climate variables (equation (2)). This assumption was made to control for possible bias due to endogeneity in the input variables. However, the estimated relationship between TFP and the investment climate variables directly indicates how firm-level efficiency depends on the investment climate. Therefore, the next table reports the coefficients of the investment climate variables for the sector-specific Translog production function specification (the coefficients of the input variables are reported in column (2)-(5) of Table 6). All continuous investment climate variables are taken in logs after adding one to reduce the impact of outliers. For the firms which did not import, and for those lacking observations on the average number of days to clear customs, we assigned a zero. 27. An important concern is the possible presence of endogeneity in the investment climate indicators. A number of previous studies have used location-industry averages for the firm-level 82 investment climate variables instead of the firm-level variables themselves to reduce the endogeneity problem (Dollar et al. 2005, Escribano and Guasch 2005). If we follow this approach, only a few investment climate variables can be included in the analysis because of the small number of location-industry cells. We therefore apply a slightly different approach. First, we test for the presence of endogeneity in each of the investment climate variables with a Hausman endogeneity test using as instruments, sector-specific location dummies (Cameron and Trivedi 2005, p.276). Second, we instrument only those variables which indicate a significant endogeneity bias (at 10%) with sector-specific location dummies. This approach is tantamount to using location-industry averages of the firm-level investment climate variables as a number of previous studies have done, but only for those variables for which a significant endogeneity bias is observed. Moreover, this approach has the added advantage that the standard errors will take into account the sampling error in the estimated location-industry averages. The Hausman endogeneity test was significant for three investment climate variables, namely; `Total duration of power outages', `Payments to deal with bureaucracy 'faster'' and `External audit of financial statements', and therefore all of the discussed investment climate variables could be included in the analysis. 28. Dollar et al. (2005) correctly note that a possible concern is that efficient firms may tend to concentrate in locations where the investment climate is better and that we cannot control for all variables governing self-selection to better climate conditions. They therefore estimate firm performance for two samples of garment firms in Bangladesh, China, India and Pakistan, namely the full sample and a sample of (relatively) `immobile' firms (c.q. domestically-owned small firms employing no more than 150 workers). The results for both samples are quite similar suggesting that mobility bias does not play an important role in their estimations. We have performed the same robustness test for our sample of Cambodian firms and find also that mobility bias does not appear to play a role. 29. Because the investment climate variables are in logs after adding one or dummy variables, Table E.6 (and Table 2.1) also includes the (semi-)elasticities evaluated at the sample means. The table shows the following features. First, productivity in firms which experience more power outages is lower, and if power outages are reduced by 50 percent, productivity will increase by 3.25 percent. Second, customs procedures reduce the productivity of firms, and if the number of days it takes to clear customs for imports is reduced by 50 percent, then productivity will increase by 1.15 percent. Third, firms which pay a lower percentage of sales to deal with bureaucracy `faster', appear to become less productive than firms that pay more, although the coefficient is insignificant (p-value 0.15). This result is not due to reverse causality as we have instrumented this variable to control for endogeneity bias. In terms of semi-elasticities, the productivity in firms that pay 50 percent less greasing money to bureaucracy is reduced by 3.75 percent. The other investment climate variables have very low t-values and therefore do not indicate a statistically significant link with productivity. This may also reflect the relatively small sample size of 272 observations. It should also be noted that the estimated elasticities may appear low but they are of similar order of magnitude as those found for Guatamala, Honduras and Nicaragua (Escribano and Guasch 2005, table D). 30. We can also evaluate the average productivity contribution of each investment climate variable (Escribano and Guasch 2005). Note that because we estimate the regressions by (two- stage) least squares, the mean of the error term is zero and therefore we can evaluate equation (2) at the mean values without an error term: (3) 83 31. Dividing the whole expression by the dependent variable and multiplying by 100, we get (4) 32. The last term represents the sum of the productivity gains and losses from all investment climate variables, relative to the average log total factor productivity. The next table reports the contribution of each investment climate variable to this total sum. For each variable, we report the mean impact for all firms as well as the mean impact for the firms with below and above median values for the investment climate variables (`low use' versus `high use'). 33. Power outages reduce average (log) total factor productivity by 3.7 percent for the average firm (compared to no power outages), and the loss in (log) TFP is 5.8 percent for the high use firms. Custom procedures for imports reduce average (log) TFP by 3.8 percent for firms that do import directly, compared to firms that do not import directly (and 0.8 percent for firms in general). Also, payments to deal with bureaucracy faster have a relatively large impact on TFP (compared to no payments at all) ­ increasing average (log) total factor productivity by 2.7 percent for the average firm and 5.65 percent for the firms paying above median amounts. The largest impact is found for firms with an overdraft ­ their (log) total factor productivity is 11.9 percent higher, compared to the counterfactual of no overdraft. 84 F. References Cameron, A. and P. Trivedi (2005), Microeconometrics. Methods and Applications. Cambridge University Press. CLRDC (Cambodian Legal Resources Development Center), in collaboration with MoC (2004): "Study on the Domestic Economic Impact and Social Costs of Adjustment to the WTO Membership of Cambodia", March. Dollar, D., M. Hallward-Driemeier, T. Mengistae (2005), "Investment Climate and Firm performance in developing Countries", Economic Development and Cultural Change, vol. 54. Escribano, A. and J. Guasch (2005), "Assessing the Impact of the Investment Climate on Productivity using Firm-Level Data: Methodology and the Cases of Guatemala, Honduras and Nicaragua", World Bank Policy Research Paper 3621. Global Development Solutions (2003), "Towards Private Sector Led Growth for Cambodia, Volume I Value Chain Analysis", draft for the World Bank. Holzmann, R. and S. Jørgensen (2000), "Social Risk Management: A New Conceptual Framework for Social Protection and Beyond", World Bank Social Protection Discussion Paper Series, 6, February. IFC (2005), "Consulting Services in Cambodia: an Overview of the Industry", Private Sector Discussions, 19, March. IFC (2007a), "The Provincial Business Environment Scorecard in Cambodia ­ a Measure of Economic Governance and Regulatory Policy", technical report, March. IFC (2007b), "Cambodia Garment Industry Training Strategy", draft, July. JDI (2007a), "Value Chain Analysis of Cambodia's Food Packaging Sector: Current Status and Future Prospects", prepared for the World Bank, December 2007. JDI (2007b), "Benchmarking Selected Variables from the 2003 Value Chain Report: Has the Enabling Environment Improved in Cambodia to Strengthen the Competitiveness of Strategic Sectors?", prepared for the World Bank, December 2007. Hausmann, R., Rodrik, D. (2005), "Growth Diagnostics", draft. Levinsohn, J. and A. Petrin (2003): "Estimating Production Functions Using Inputs to Control for Unobservables", Review of Economic Studies, Vol. 70, Issue 2, pp.317-41. Naron, Hang Chuon (2007), "Le Secteur de Télécommunications au Cambodge", draft. OECD (2001): Measuring Productivity. Olley, G. and A. Pakes (1996): "The Dynamics of Productivity in the Telecommunications Equipment Industry", Econometrica, Vol. 64, No. 6, pp. 1263-1297. RGC (2007), Diagnostic for Trade Integration Strategy, Royal Government of Cambodia. USAID (2005), "Competitiveness of Cambodia's Garment Industry", prepared by Nathan Inc. USAID (2007a), "Factory-Level Value Chain Analysis of Cambodia's Apparel Industry", prepared by Nathan Associates Inc. and Werner International, September. USAID (2007b): "Southeast Asia Commercial Law and Institutional Reform and Trade Diagnostics ­ Cambodia", final report, December. USAID (2007c): "Cambodia Development Partner Value Chain Activity & Coordination Study", final draft, September. World Bank (2004a), Cambodia: Seizing the Global Opportunity: Investment Climate Assessment and Reform Strategy for Cambodia, Report No.27925-KH. World Bank (2004b), World Development Report 2005. A Better Investment Climate for Everyone. Co-publication of The World Bank and Oxford University Press. World Bank (2006), Thailand Investment Climate, Firm Competitiveness and Growth, Report No.36267-TH. World Bank (2008a), "GMS Regional Trade Facilitation Policy Note", draft, January. World Bank (2008b), Doing Business 2009, September. 85 103° E 104° E 105° E 106° E 107° E 0 20 40 60 80 Kilometers THAILAND L A O P. D . R . To CAM BODIA 0 20 40 60 Miles To Pak Charang Khu Khan To SELECTED CITIES AND TOWNS Champasak Kon To g ODDAR Phiafai Samraong Cheom Ksan PROVINCE CAPITALS MEANCHEY Siem Pang NATIONAL CAPITAL ng Kompong Sralao Sre ng 14° N 14° N Ko RATAN AKI RI RIVERS PREAH VIHEAR To To BANTEAY Play Cu MAIN ROADS Buri n Phum Kompadou Phnum Tbeng Melouprey Sa MEANCHEY Meanchey STUNG TRENG Bun RAILROADS Kralanh Long Sisophon PROVINCE BOUNDARIES SIEM REAP Stung Treng Srepok Lomphat Siem Reap INTERNATIONAL BOUNDARIES Rovieng Battambang Tonle Se n g 13° N 13° N ng B AT TA M B A N G Sap Kohnieh Poro Pailin Chas Mekon KAMPONG THOM MONDOL PA I L I N Kampong Thom K R AT I E K I R I Chbar Pursat Kratie Senmonorom Peam Ton le P U R S AT KAMPONG To KAMPONG CHAM Duc Lap Sap Kampong 12° N Ca Phnum CHHNANG Mekong Cham 12° N rd Aoral am (1,810 m) To om Ba Ra PHNOM PENH M Krong Koh Kong t KAMPONG s. PHNOM PENH SPEU To Kampong Speu Ta Khmau Prey Veng KOH KONG SVAY Tay Ninh G ul f of KANDAL PREY RIENG V I E T N AM T ha i la n d Chambak VENG To ssa Ba k Svay Ho Chi Minh City Rieng Takeo 11° N 11° N Chhak Kampong CAMBODIA Saom TAKEO This map was produced by To the Map Design Unit of The KAMPOT Cao Lanh World Bank. The boundaries, Sihanoukville colors, denominations and Kampot any other information shown SIHANOUK- on this map do not imply, on Kep To the part of The World Bank VILLE Long Xuyen Group, any judgment on the legal status of any territory, To or any endorsement or Rach Gia acceptance of such KEP boundaries. 103° E 104° E 105° E 106° E 107° E JUN E 2006 IBRD 33381R