Report No. 54123-TR

Public Disclosure Authorized

TURKEY

Investment Climate Assessment

From Crisis to Private Sector Led Growth

May 2010

Public Disclosure Authorized

Public Disclosure Authorized

Europe and Central Asia Region

Public Disclosure Authorized

Document of the World Bank

CURRENCY EQUIVALENTS Currency Unit – New Turkish Lira (TL)

EXCHANGE RATE May 12, 2010 TL 1.53= USD 1.00

WEIGHTS AND MEASURES Metric System

FISCAL YEAR January 1 – December 31

Vice President: Philippe H. Le Houerou Country Director: Ulrich Zachau Sector Director: Fernando Montes-Negret Gerardo Corrochano Sector Manager Lalit Raina Task Manager: Donato De Rosa

ACRONYMS AND ABBREVIATIONS

Small and Medium Enterprises Development ABGS Secretariat General for EU Affairs KOSGEB Organization ABİGEM EU Turkish Business Centers KSS Small Industrial Estates ACTAL Dutch Advisory Board on Administrative Burden MAS Manufacturing Advisory Services Training and Small Industry Support Foundation of BILGE Computerized Customs Activity MEKSA Turkey BRSA Banking Regulation and Supervision Agency MEP Manufacturing Extension Partnership CBRT Central Bank of Turkey MLT Medium- and long-term CC Commercial Code MNC Multinational Corporation Second Competitiveness and Employment CEDPL2 MPM National Productivity Center Development Policy Loan CGF Credit Guarantee Fund MTP Medium Term Program CIF Cost, Insurance & Freight NIS Networks and Innovation Survey CMB Capital Markets Board NPAA National Plan for Adoption of the Acquis Japanese Council for the Promotion of Regulatory CPRR NPL Non-Performing Loans Reform DA Development Agency NSC National Steering Committee Organization for Economic and Cooperation D&Q Design and Quality OECD Development DB Doing Business OLS Ordinary Least Squares ECA Europe and Central Asia OSB Organized Industrial Zones EFCAS Enterprise Financial Crisis Assessment Survey PMR Product Market Regulation EIF European Investment Fund R&D Research and Development ES Enterprise Survey RIA Regulatory Impact Analysis EU European Union SBA Small Business Administration EURADA European Association of Development Agencies SCM Standard Cost Model FDI Foreign Direct Investment SIC Standard Industrial Classification FOB Free on Board SME Small and Medium Enterprise FX Foreign Exchange SPO State Planning Organization GDP Gross Domestic Product TASB Turkish Accounting Standards Board GERD Gross Expenditures on R&D TESK Merchants and Artisans Confederation of Turkey GoT Government of Turkey TFP Total Factor Productivity GOV Governance TFRS Turkish Financial Reporting Standards Union of Chambers and Commodity Exchanges of GVA Gross Value Added TOBB Turkey Small and Medium Industry Owners and Managers GVC Global Value Chains TOSYOV Foundation of Turkey IAC Investment Advisory Council of Turkey TPI Turkish Patent Institute ICA Investment Climate Assessment TSI Turkish Studies Institute Scientific and Technological Research Council of ICS Investment Climate Survey TUBITAK Turkey ICT Information and Communication Technologies TÜRKAK Turkish Accreditation Agency IFRS International Financial Reporting Standards TURKSTAT Turkish Statistical Institute IIPR Intellectual and Industrial Property Rights USPTO United States Patent and Trademark Office ISE Istanbul Stock Exchange VEDOP Tax Offices Automation Project ISO International Organization for Standardization WDI World Development Indicators

Investment Support and Promotion Agency of ISPAT WFE World Federation of Exchanges Turkey Japan Finance Corporation for Small and Medium JASME WIPO World Intellectual Property Organization Enterprises Coordination Council for the Improvement of the KADIM Struggle against Unregistered Employment Project YOİKK Investment Environment KKB Credit Bureau of Turkey

ACKNOWLEDGMENTS

This report was prepared in close cooperation with the YOİKK Secretariat. Preliminary and intermediate findings of the report were presented to YOİKK Steering Committee meetings on June 11 and November 13 2009. Intermediate findings were presented to various stakeholders in a series of roundtables held in November 2009. The World Bank team also performed field visits to Izmir and Adana, where it greatly benefited from discussions with local stakeholders.

The World Bank team was led by Donato De Rosa and included Dragana Pajovic, Paulo Correa, Murat Seker, Federica Saliola, Delia Rodrigo, Carlos Piñerúa, Jorge Peña, Alvaro Escribano, Manuel de Orte, Baris Dincer, Irem Guceri, Selma Karaman, Ayse Seda Aroymak, Zeynep Lalik, Erkan Erdil, Murat Ucer. The report was undertaken under the guidance of Ulrich Zachau, Fernando Montes-Negret, Gerardo Corrochano, Lalit Raina and Keiko Sato. Indispensable contribution and insight came from İbrahim H. Çanakcı, Undersecretariat of Treasury, Cavit Dağdaş, Deputy Undersecretariat of Treasury, Berrin Bingöl, Murat Alıcı, Mehmet Dündar, Özge Dumlupınar, Serenay Usta, Gamze Özdurgutlu, Gönül Bakır Kartal, Bahar Konak, Başak Ünal and Can Gürlek (Undersecretariat of Treasury team). In addition to the Treasury team, recognition is extended to all YOİKK member institutions. Willem van Eeghen and Stefka Slavova (World Bank) and Rauf Gönenç (OECD) were peer reviewers for the report. In addition, the team received helpful comments and suggestions from Mark Roland Thomas, Cihan Yalcin, Kamer Karakurum-Ozdemir, Mediha Agar, Muammer Komurcuoglu, Jesko Hentschel, Cristobal Ridao-Cano, Raif Can, Steen Byskov, Jean-Louis Racine, Cemile Hacibeyoglu, Andres Federico Martinez, Mahesh Uttamchandani and Anthony Ody.

TABLE OF CONTENTS

EXECUTIVE SUMMARY ...... i OVERVIEW AND POLICY OPTIONS ...... ii Chapter 1. CHALLENGES FACING THE TURKISH BUSINESS SECTOR ...... 1 1.1 Macroeconomic Setting ...... 4 1.2 Effects of the Global Crisis on the Corporate Sector: Preliminary Evidence ...... 9 References ...... 12 Chapter 2. EVOLUTION OF THE TURKISH INVESTMENT CLIMATE ...... 13 2.1 The Investment Climate and Business Sector Performance ...... 13 2.2 The Regulatory Environment ...... 21 2.3 Labor Market and Skills ...... 36 2.4 Innovation ...... 45 2.5 Access to Finance and Corporate Governance ...... 52 References ...... 62 Annex 2-A. Econometric Methods ...... 64 Annex 2-B. Investment Climate Variables Used for the Econometric Analysis ...... 76 Chapter 3. PROMOTING SME GROWTH ...... 79 3.1 Patterns of Firm Growth in Turkey ...... 80 3.2 Investment Climate Constraints to SME Growth ...... 84 3.3 Enhancing the Ability of SMEs to Access Bank Credit ...... 87 References ...... 94 Annex 3-A. Descriptive Statistics and Econometric Analysis ...... 95 Chapter 4. STIMULATING KNOWLEDGE FLOWS ...... 102 4.1 Production Networks and Knowledge Flows ...... 104 4.2 Increasing the Absorptive Capacity of SMEs...... 110 References ...... 117 Annex 4-A. Econometric Methodology and Results ...... 120 Annex 4-B. International and Domestic Trade Flows: A Regional Perspective ...... 121 Chapter 5. ENHANCING REGULATORY CAPACITY FOR BETTER REGULATION ...... 125 5.1 Regulatory Policies and Institutional Drivers in Turkey ...... 128 5.2 Administrative capacities to prepare new regulations ...... 134 5.3 Administrative capacities to review existing regulations ...... 136 5.4 Policy Options for Regulatory Reform...... 143 References ...... 147 Annex 5-1. OECD Principles of Regulatory Quality and Performance ...... 148 Annex 5-2. The Sources of Law in Turkey ...... 149

Figures

Figure 1-1: 5-year CDS spreads, Jan 2007-Dec 2009 ...... 1 Figure 1-2: Short-term external corporate debt in Turkey, million USD...... 1 Figure 1-3: GDP by sector in Turkey, 1998-2009, current prices, billion TL ...... 6 Figure 1-4: Value added as share of GDP by sector, 2008 ...... 6 Figure 1-5: Turkey‘s exports by industry 2000-2009 ...... 7 Figure 1-6: Turkey‘s exports by partner country, 2000-2009 ...... 7 Figure 1-7: Exports by technology level, share of GDP, 1996-2008 ...... 8 Figure 1-8: FDI net inflows/GDP, 2009 ...... 8 Figure 1-9: Effects of the Crisis ...... 10 Figure 1-10: Sales in the corporate sector: June 2008 - June 2009 ...... 10 Figure 1-11: Structure of corporate liabilities ...... 11 Figure 1-12: Firms‘ survival strategies, forms of adjustment ...... 11 Figure 2-1: Major obstacle levels for firms in Turkey ...... 15 Figure 2-2: The IC and TFP: Cross-country comparison ...... 16 Figure 2-3: TFP in Turkey ...... 17 Figure 2-4: IC percentage contributions to aggregate and average productivity ...... 19 Figure 2-5: Share of Investment Climate effects on the sample mean of employment and probability of exporting and receiving FDI ...... 20 Figure 2-6: Firms perceiving tax rates and tax administration as a major or very severe obstacle, by country...... 22 Figure 2-7: Tax rates as an obstacle, by firm size region and industry ...... 23 Figure 2-8: Corporate tax rates 2005-2009, Turkey and comparator countries, percent ...... 23 Figure 2-9: Tax wedge, 2008 ...... 24 Figure 2-10: Share of firms facing informal competition ...... 25 Figure 2-11: Development of product market regulation since 1998, Turkey and comparator countries ...... 29 Figure 2-12: Days to obtain operating license, country comparison ...... 30 Figure 2-13: Days to obtain operating license, by region, city, size, ownership and exporting status ...... 30 Figure 2-14: Days to obtain import license, by country ...... 31 Figure 2-15: Days to obtain import license, by size, ownership, exporting, region, city and industry ...... 31 Figure 2-16: Days to clear customs for imports...... 32 Figure 2-17: Access to land as an obstacle, by city ...... 32 Figure 2-18: Access to land as an obstacle, by firms size, ownership, exporting status and industry ...... 33 Figure 2-19: Days to obtain a construction-related permit, by country ...... 33 Figure 2-20: Number of inspections per year, selected countries ...... 34 Figure 2-21: Average time spent dealing with each inspection (days), by firm size ...... 34 Figure 2-22: Labor regulations as an obstacle, country comparison ...... 38 Figure 2-23: Temporary employment as a share of total employment, country comparison ...... 40 Figure 2-24: Labor regulations as an obstacle, by size, ownership and exporting status ...... 41 Figure 2-25: Population with at least upper secondary education, percent, 2007 ...... 42 Figure 2-26: Distribution of permanent workers, by type and firm size, 2005 ...... 42 Figure 2-27: Distribution of permanent workers, by type and firm size, 2008 ...... 42 Figure 2-28: Education of workforce as an obstacle, by country ...... 43 Figure 2-29: Education of workforce as an obstacle...... 43 Figure 2-30: Share of firms offering formal training to employees, by country ...... 44 Figure 2-31: Share of firms offering formal training to employees, by firm size, ownership form, exporting status, region and industry ...... 44 Figure 2-32: Share of production and non-production workers that received training, country comparison ...... 44 Figure 2-33: R&D expenditure as a share of GDP - total, private and public, 2000-2008 ...... 46 Figure 2-34: R&D investment as a share of total sales, 2005 and 2008 ...... 46 Figure 2-35: Patent applications, per million population, 2000-2008 ...... 47 Figure 2-36: Share of firms developing new products and R&D spending, by firm size, ownership form, exporting status and region ...... 48 Figure 2-37: Share of firms using technology licensed by foreign firms, country comparison ...... 49

Figure 2-38: Share of firms using foreign licensed technology, by size, foreign ownership, exporting status and region ...... 49 Figure 2-39: Share of firms with internationally recognized quality certifications, country comparison ...... 50 Figure 2-40: Share of firms with internationally recognized quality certifications, by region ...... 50 Figure 2-41: Use of websites in communication with clients and suppliers, country and small firm comparison ...... 51 Figure 2-42: Use of ICT in communication with clients and suppliers, by size, ownership form and exporting status ...... 51 Figure 2-43: Access to finance as an obstacle , country and small firm comparison ...... 53 Figure 2-44: Turkish banks by ownership, 2005 and 2009 ...... 54 Figure 2-45: Market capitalization of firms listed on the Istanbul SE/GDP, 2005, 2007, 2008 and comparator countries ...... 55 Figure 2-46: Firms listed on the ISE, by industry, 2009 ...... 55 Figure 2-47: Claims on the private sector, % of GDP, 2008 ...... 56 Figure 2-48: Private sector credit by recipient (biannually, billion TL) ...... 56 Figure 2-49: Share of firms with loans, country and small firm comparison ...... 56 Figure 2-50: Share of firms with loans, by region ...... 56 Figure 2-51: Sources for finance of fixed assets, country comparison ...... 57 Figure 2-52: Private credit coverage, percentage of adults ...... 58 Figure 2-53: Public credit coverage, percentage of adults ...... 58 Figure 2-54: Firms having financial accounts externally audited, country comparison ...... 59 Figure 2-55: Share of firms for which collateral was required for their latest loan, country comparison ...... 59 Figure 2-56: Type of collateral, 2005 and 2008 ...... 60 Figure 2-57: Protecting investors index, scale 1-10 ...... 61 Figure 3-1: Employment growth at different sizes and ages (2004-2007) ...... 81 Figure 3-2: Growth rates relative to micro firms (2004-2007) ...... 82 Figure 3-3: Percentage of firms that are above 16 years old ...... 82 Figure 3-4: Growth rates of firms: Cross-country comparison ...... 83 Figure 3-5: Single most severe investment climate obstacles, by firm size ...... 84 Figure 3-6: Access to Finance for SMEs ...... 85 Figure 3-7: Loans to SMEs ...... 86 Figure 3-8: Guarantees provided by the CGF as a percentage of the total SME loan volume ...... 91 Figure 3-9: Flow of funds in guarantee agreements under treasury support ...... 92 Figure 4-1: Industrial clusters in Turkey ...... 103 Figure 4-2: Value chain governance modes...... 107 Figure 4-3: Governance modes features, across regions ...... 108 Figure 4-4: Governance modes features, across industries ...... 109 Figure 4-5: Governance modes features, by firm size ...... 109 Figure 4-6: Trade and governance modes ...... 110 Figure 5-1: Phases of the preparation of laws in Turkey ...... 130 Figure 5-2: Institutional capacity for managing regulatory reform ...... 133 Figure 5-3: Explicit program for reducing administrative burdens ...... 137

Tables

Table 1-1: Turkey: Key Economic Indicators, 2002-2009 ...... 5 Table 2-1: Summary of econometric methods ...... 13 Table 2-2: IC variables‘ relative contributions to TFP in 2008 and 2005 (percent) ...... 18 Table 2-3: Investment climate variables‘ relative contributions to sample averages of employment, exporting and FDI in 2008 and 2005 (percent) ...... 20 Table 2-4: Summary of the effects of the regulatory environment in 2008 ...... 21 Table 2-5: Bankruptcy procedures (duration and recovery rate) comparison of selected countries ...... 36 Table 2-6: Summary of the effects of labor and skills ...... 37 Table 2-7: Key labor market indicators, Turkey and selected OECD countries, 2008 ...... 37 Table 2-8: Summary of the effects of innovation ...... 45 Table 2-9: Turkey's patent applications, 1995-2008 ...... 47 Table 2-10: Summary of the effects of finance and corporate governance ...... 53

Table 2-11: Structure of the financial system in Turkey, 2002-2009 ...... 54 Table 2-12: Protecting investors index, scale 1-10 ...... 61 Table 3-1: Growth rates by industry and region (2004-2007) ...... 81 Table 3-2: Summary of investment climate effects on firm growth, percent ...... 87 Table 3-3: Comparison of CGFs ...... 93 Table 4-1: Summary of investment climate effects on knowledge flows ...... 111 Table 5-1: Product Market Regulation - Barriers to Entrepreneurship 2008: Regulatory and administrative opacity ...... 126

Boxes

Box 1-1: Turkey‘s Income Gap ...... 2 Box 1-2: Recommendations in key reform areas from the 2007 Investment Climate Assessment ...... 3 Box 1-3: The Enterprise Financial Crisis Assessment Survey (EFCAS)...... 10 Box 2-1: The 2008-2009 Enterprise Survey ...... 14 Box 2-2: Tax reforms ...... 23 Box 2-3: The Government‘s struggle against informality ...... 26 Box 2-4: Development Agencies (DAs) ...... 27 Box 2-5: Indicators of product market regulation (PMR) ...... 29 Box 2-6: Employing workers in Turkey: rules of hiring, work schedules and redundancy ...... 39 Box 2-7: Labor market reforms in Turkey ...... 40 Box 2-8: Technology and innovation reforms ...... 52 Box 2-9: The Turkish banking sector in the wake of the crisis ...... 55 Box 3-1: Data for the analysis of firm growth ...... 80 Box 3-2: Main obstacles to SME lending: views from banks and SME representatives ...... 88 Box 3-3: The Credit Guarantee Fund (CGF) ...... 91 Box 3-4: Credit guarantee schemes: International comparison ...... 92 Box 4-1: Global value chains ...... 105 Box 4-2: The ―Networks and Innovation‖ Survey (NIS) ...... 106 Box 4-3: SME support programs: The international experience ...... 114 Box 5-1: Regulatory reform to improve the business environment: International experiences ...... 129 Box 5-2: International examples of oversight bodies for regulatory reform ...... 131 Box 5-3: Coordination between levels of government ...... 134 Box 5-4: Comprehensive administrative simplification efforts in OECD countries ...... 137 Box 5-5: Benefits of licensing reform ...... 138 Box 5-6: Dealing with construction permits ...... 139 Box 5-7: The use of SCM: International experiences ...... 141

EXECUTIVE SUMMARY This Investment Climate Assessment draws on the analysis of firm-level survey data collected during April 2008-January 2009, supplemented by other sources, to provide a comprehensive and up-to-date description of the investment climate facing Turkish firms of all size classes, including the impact of government regulations and recent reforms. An important feature of the analysis is extensive use of data from comparable countries to benchmark Turkey‘s performance. Beyond description, the report seeks to identify key priority areas where further policy reform and institutional development could help strengthen Turkish firms‘ performance in such areas as productivity, export competitiveness and employment creation. A special aspect of the report is its focus on Turkey‘s small and medium scale enterprise (SME) sector.

Since late-2007, global conditions have taken their toll on the Turkish business sector. Turkey‘s economy contracted by 4.7 percent, with unemployment reaching 14 percent in 2009. Sustainable growth in the post-crisis environment will require continuation of reforms aimed at promoting healthy business sector development.

Continued commitment to business environment reforms will help support a sustainable recovery. Building on recent reforms, actions to improve the regulatory framework need to be sustained to reduce incentives for firms to remain informal. Business registration has been simplified, but administrative procedures still impose a high ―time tax‖ on firms. Improved availability of skilled labor will be crucial for improving productivity. Sustained encouragement of firm-level innovation would also have positive effects on enterprise performance. The availability of credit to the corporate sector, especially SMEs, has been negatively affected by the crisis. With a competitive global environment expected in the post-crisis period, the report identifies three priority areas to help the economy achieve sustainable, broad-based growth that incorporates SMEs and is more evenly distributed across the country.

A first key priority is to alleviate the, mainly financial, obstacles that constrain SMEs‟ growth, thus preventing the largest portion of the enterprise sector from reaping scale economies. A healthy SME sector could improve employment opportunities and promote regional development. SMEs grow more slowly than micro or large firms in Turkey, unlike SMEs in comparator countries. Existing policies and regulations may impact SMEs more than either micro or large firms. Access to finance appears to be the single most important constraint to SME growth. Enhancing banks‘ ability to assess borrowers‘ creditworthiness, plus a more active Credit Guarantee Fund, should help ease lending to SMEs.

A second priority is to increase Turkish SMEs‟ competitiveness by enhancing their ability to adopt and use knowledge. Diffusion of the sources of growth beyond firms that are already sufficiently competitive to be direct exporters (and beyond already-successful manufacturing poles) will increase Turkey‘s resilience to future external shocks in global demand and ensure that the productive base of Turkish manufacturing is more evenly distributed geographically. The local business and institutional environment combines with country-wide features to determine firms‘ incentives to adopt innovative modes of production and organization. Analysis of Turkish production networks indicates that the absorptive capacity of local suppliers is key for successful participation in global markets. Existing government programs aimed at increasing the operational capabilities and absorptive capacity of SMEs could be improved at the local level in line with international best practice.

A third priority is to further reform and strengthen the regulatory capacity of the government. Turkey‘s recent important steps in establishing institutions and mechanisms for regulatory reform could be made more effective by refining their strategic vision, improving horizontal and vertical coordination, and enhancing consultation with the private sector. The establishment of Development Agencies (DAs) offers an opportunity to ease investment climate constraints through actions at the local level.

i

OVERVIEW AND POLICY OPTIONS 1. Since late-2007, adverse changes in global conditions have taken their toll on Turkey‟s previously booming economy and business sector. Prior to the global economic crisis, Turkey‘s economy had been thriving. Following the 2001 banking crisis associated with a sharp recession and a restructuring of the financial sector, Turkish GDP growth averaged nearly 7 percent per annum between 2002 and 2007. An important engine of growth was private investment, in part driven by large capital inflows, which contributed to a trebling of private sector Gross Fixed Capital Formation between 2002 and 2008. Since 2008, though, the external economic environment has deteriorated markedly, with falls in external demand and international capital flows – and associated declines in domestic demand and credit availability. Turkey‘s economy contracted by 4.7 percent, with unemployment reaching 14 percent in 2009.The corporate sector, in particular, has been hard hit by the slowdown in global demand. A survey carried out by the World Bank in the summer of 2009 shows that most enterprises experienced a sharp contraction in sales, with reported declines between 2008 and 2009 in the region of 40 percent. Almost half of the firms surveyed (46 percent) reported restructuring their liabilities, while one-third delayed payments to tax authorities and suppliers.

Tough global conditions heighten the need to attack constraints to firms’ productivity and growth.

2. Turkish firms cite a number of external constraints to their own performance. According to a survey conducted between April 2008 and January 2009, a majority of Turkish firms see themselves as held back by problems with access to finance (some 26 percent of firms cited this as their single most important constraint). Tax rates (18 percent) and political instability (18 percent) rank second and third, while other important factors are Figure 1: Top Five Investment Climate Obstacles competition from the informal sector and an inadequately educated 30 25.9 workforce (15 and 9 percent 25 respectively). Analysis of survey data 18.2 20 17.5 confirms a significant association 14.7 15 between the quality of the investment 9.1 10 climate and performance in areas like 5 productivity, job creation, export 0 competitiveness and attractiveness for Access to Tax Rates Political Practices Inadequately foreign investment. Productivity Finance instability Informal educated analysis shows that almost one-third Sector workforce of variation in the performance of the business sector in Turkey is explained Source: Turkey ES 2008 by investment climate factors.

3. The regulatory environment is the area of the investment climate with the largest relative contribution to productivity. Other relevant investment climate areas include infrastructure bottlenecks, access to finance and corporate governance, the availability of skilled labor and innovation. Aggregate productivity in Turkey appears to have increased since 2005, driven by improved allocation of resources towards firms with higher productivity. This has widened the gap between low and high productivity establishments, with larger firms appearing to benefit from the more positive aspects of the investment climate and smaller and less productive firms bearing the costs of its less positive features.

4. Analysis of survey data indicates that firm-level productivity in Turkey is negatively associated with a number of features of the regulatory environment. Some of these include formal bureaucratic requirements, such as the number of inspections to which businesses are subjected, the number of compulsory certificates required and the time necessary to obtain them, as well as time

ii consuming customs procedures for imports. Burdensome regulatory requirements constitute fertile breeding ground for the negative consequences of corruption, as it is exemplified by the negative association between productivity and informal payments to obtain power supply or a contract with the government. Inefficient regulations also provide incentives for firms to remain partially informal. Firms that are subject to competition from informal establishments are, in turn, associated with lower levels of productivity.

5. Problems for Turkish firms posed by tax rates and tax administration seem to have declined in importance since 2005, at least partly reflecting the impact of recent reforms. When examining firms‘ perceptions, the relative importance of tax rates in a ranking of obstacles has dropped, from being the largest obstacle in 2005, to where a relatively low 18 percent of firms in 2008 perceived tax rates as the single most relevant Figure 2: Share of firms facing informal competition impediment to business operations. The share of enterprises identifying 60% 54% 55% 52% 52% 49% tax rates as a major or very severe 50% 43% 44% constraint decreased from 81 percent 35% 40% 33% in 2005 to 50 percent in 2008. Tax 30% administration, viewed as a major constraint by 59 percent of 20% manufacturing firms in 2005, had 10% dropped to 19 percent by 2008. These 0% improvements between 2005 and Poland Romania Czech Turkey Hungary Turkey Turkey Bulgaria Brazil 2008 can, at least in part, be ascribed 2009 2009 Rep. 2005 2009 2008 2008 2009 2009 to tax reforms introduced since 2006. 2009 (manuf) (manuf) (all) Notable among the reforms are the Source: Turkey ES 2008 introduction of a new corporate tax code, the reduction of corporate income tax from 30 to 20 percent, and lower taxation on interest. On another front, despite the lower tax rates introduced with recent reforms and the fall in the headline measure of informality – from 53 percent to 44 percent between 2004 and 2008 – the share of surveyed firms complaining about informal competition increased from 44 percent in 2005 to 52 percent in 2008.

6. Turkey has made significant Figure 3: The „time tax‟ reforms in some areas of the regulatory climate, including easing business 30% 27% registration, but red tape still imposes 25% 19% 20% significant costs on businesses. The 13% 13% Government has made progress towards 15% 9% 9% 10% 11% 10% facilitating the process of business 5% registration. According to Doing Business 0% 2010, recent Turkish reforms have reduced the time it takes to register a business from 13 required steps in 2004, down to six steps in 2009. The Government has initiated e- Source: Turkey ES 2008 Judicial Registrations and Online Company Registration procedures, with a draft act currently being evaluated by the Prime Minister‘s Office and expected to be adopted in the National Assembly in early 2010. Manufacturing firms interviewed in the 2008 enterprise survey reported that the time it takes to acquire an operating license had decreased from 66 days in 2005 to 62 days in 2008. Even after the improvement, though, this is still higher than in comparator countries. More broadly, compliance with administrative procedures remains problematic for business operation. There is no clear framework for streamlining administrative procedures for businesses, and operating licenses are issued by various Ministries, each responsible for different business

iii areas. Additionally, in overall terms the perceived „time tax‟ – the share of management time spent dealing with government regulation – appears to have increased sharply since 2005, from 9 to 27 percent. The time tax is largest for medium and large enterprises (32 percent and 34 percent respectively), whereas managers in small firms report spending 23 percent of their time dealing with red tape.

7. A specific area of improvement has been in the cost and time invested in construction of business premises. According to the Doing Business‘ measure of the time involved in building a warehouse, Turkish firms were able to shorten their average site development time by 44 days between 2005 and 2009. This said, the time to obtain building permits varies significantly between Turkish cities, with firms in Istanbul seeming to struggle with construction permits more than firms located elsewhere. Additionally, a variation is notable by firm size, with SMEs spending nearly twice the time as large firms dealing with construction permits (60 versus 32 days).

8. Business inspections appear to be somewhat less burdensome in Turkey than in comparator countries. The average time all employees in a company spend with inspections per year is 6.6 days, which puts Turkey ahead of several comparator economies. The survey results are similar when firms are asked about the number of inspections taking place in their organization: the average annual frequency of inspectors‘ visits has diminished from 4 in 2005 to 2 in 2008. When examining the time that each firm spends on inspections, medium size firms seem to be more affected. Additionally, regional comparisons show significant variations in inspection times and duration. In general, the large regional variations observed for licenses, permits and inspections are a consequence of the fact that municipalities, often lacking adequate capacity, are frequently responsible for implementing regulations that are established at the central level. This institutional setup tends to create a burden for businesses and individuals, when having to comply with procedures established locally without prior agreement at the central level.

Better availability of skilled labor would help improve Turkish firms’ productivity.

9. Firms with a higher share of staff with university education tend to show higher productivity, according to econometric analysis. Survey results show that larger firms are in a better position to afford skilled staff with university education.

10. Education levels in Turkey lag behind other OECD countries. OECD data show that 26 percent of the Turkish adult population holds secondary education diplomas. This is well below the OECD average of 69 percent and the EU19 average of 70 percent. Graduates with tertiary education in Turkey are also scarce, and the entry rate to higher education programs (tertiary education) is low by international standards. Only 29 percent Figure 4: Education of workforce as an obstacle enroll in higher education in Turkey, compared to a 56 percent average in the 50% 42% 43% OECD countries. Nearly a quarter of 36% 40% 33% Turkish firms rate the education and 29% 30% 25% skills levels of the workforce as a 21% 21% 20% ―major‖ or ―very severe‖ constraint on operations and growth (See Figure). 6% 10% Although this is an improvement from 0% the 33 percent in 2005, the high rate still Hungary Bulgaria Turkey Turkey Czech Turkey Poland Chile Romania requires the attention of policymakers 2009 2009 2008 2008 Rep. 2005 2009 2006 2009 (manuf) (all) 2009 (manuf) and shows that measures need to be taken to better coordinate labor supply with the Source: Enterprise Surveys 2008 demands in the business sector.

iv

11. The share of Turkish firms offering formal training to their employees has increased slightly. Some 24 percent of manufacturing businesses reported offering training to their employees in 2005, and this number had risen to 29 percent three years later. The share of large firms offering training is three times higher than that among small businesses. Enterprises with exporting activities are also more active in offering training to workers.

12. Turkey‟s rate of labor force participation, at less than 50 percent, is low by international standards and has decreased somewhat since 2005. This places Turkey around 20 percentage points below OECD and EU-15 averages. The employment rate for women is particularly low, standing at 26.7 percent in 2007, far below the OECD and EU-15 averages of 61.3 percent and 65.3 percent respectively. The 2008 Enterprise Survey shows that on average only 16 percent of production employees and four percent of non-production employees are female.

13. The 2008 survey found firms in Turkey less likely to regard labor regulations as a serious obstacle than they had been in 2005, probably an effect of the timing of the survey. Turkey‘s score on this indicator appears to have improved substantially vis-à-vis the average for the Europe and Central Asia region. The survey results require cautious interpretation, however, as the inflexibility of labor regulations is still mentioned by enterprises in face to face interviews as the overarching constraint facing firm operation and growth. While firms‘ perceptions in the 2008 survey might have been influenced by labor reforms initiated in early-2008, it is also possible that the timing of the survey (from April 2008 to January 2009) found business more preoccupied with other, more immediate issues – e.g. loss of market share requiring downsizing or problems with access to finance – thus decreasing the perceived relative importance of labor regulations. These concerns may resurface as a major constraint for sustainable recovery.

Productivity and exports would benefit from policies encouraging innovation.

14. Turkish firms that invest in Research and Development (R&D) tend to show higher productivity levels, according to analysis of the 2008 survey of Turkish enterprises. Firms that had re- organized production processes to take advantage of outsourcing were also found to be more productive. Employment, exports and FDI are all positively associated with innovation at the firm level. Econometric analysis also points to a positive association between employment levels and the use of ICT in communication with customers and suppliers. Firms with quality certifications are also more likely to have a larger workforce. Finally, there is a significantly positive correlation between variables reflecting innovation (such as quality certification and the use of ICT) and the probability of exporting.

Figure 5: Share of firms with R&D spending 15. Turkey‟s total investment in R&D has nearly doubled over the past 30% 28% ten years, reaching 0.73 percent of GDP 25% 22% 23% in 2008. This is also reflected in the 19% 20% 16% relative high number of Turkish firms (23 15% 11% percent) that perform R&D expenditures 10% (See Figure).Nonetheless, Turkey still lags 5% behind other middle income countries and the OECD, which presented an average of 0% 2.29 percent in 2007, compared to 0.71 Hungary Poland Romania Czech Rep. Turkey Bulgaria percent for Turkey in the same year. Sources: Enterprise Surveys 16. Turkey‟s application of international quality standards (ISO 9001) has shown remarkable improvement over the past decade, with more than 13,200 certificates issued by the end of 2008. This performance compares

v

relatively well to that of other economies. Firm surveys in 2008 found 30 percent of Turkish firms reporting having an internationally recognized quality certification. This puts Turkey ahead of other middle-income countries, such as Brazil (26 percent), Bulgaria (20 percent) and Poland (17 percent). Certifications among small firms lag far behind medium and large companies: about 55 percent of large firms hold a quality certification, which is three times the share of small firms.

17. The Government has taken steps to encourage the use of ICT. With the implementation of the e-Transformation Turkey Project, the expansion of ICT in public services has received a boost. The Government has also stepped up initiatives to raise awareness of ICT among citizens and businesses. Further support is planned for enterprises in their use of ICT, as well as increased competition in the electronic communication sector.

Credit availability to firms, especially SMEs, has suffered during the crisis.

18. Turkish firms with good access to finance tend to show higher productivity. Analysis from 2008 indicates that higher productivity was related to several variables representing financial soundness (e.g., firms with a higher share of sales paid for before delivery, and firms with the ability to finance a higher proportion of fixed assets purchases with internal funds).

19. Turkey‟s financial sector is relatively small by comparative standards. According to a recent study by the Banks Association of Turkey, the ratio of financial assets to GDP in 2007 was 150 percent in Turkey, compared to 246 percent for emerging market economies and a global average of 421 percent. This said, according to the 2008 enterprise survey, 57 percent of Turkish firms had access to a loan, which compares fairly well to other middle income economies (see Figure). Nonetheless, firms of all size categories perceive access to finance as their single most severe obstacle, with medium-sized firms appearing to be particularly affected (34 percent), followed by micro (26 percent), small (24 percent) and large firms (19 percent). Turkish firms rely more on bank loans for investment financing (38 percent) than do firms in other countries. This is Figure 6. Share of firms with loans especially true for medium-sized firms, for which bank finance accounts for 47 80% 69% All firms Small firms 65% 65% percent of total investment funding. 57% 57% 60% 50% 50% Collateral requirements appear 46% 47% 48% 49% 46% 40% 43% 43% 35% 42% 40% particularly onerous for SMEs, 40% compared to both micro and large firms, amounting to 100 percent of 20% loan value for small firms and 91 0% percent for medium firms. The share of Bulgaria Hungary Turkey Czech Poland Turkey Turkey Brazil Chile loan applications that is rejected is also 2009 2009 2005 Rep. 2009 2008 (all) 2008 2009 2006 substantially higher for SMEs (17 (manuf) 2009 (manuf) percent) compared to large firms (12 Source: Enterprise Surveys percent).

20. The credit crunch following the global financial crisis has affected lending to the SME sector. Starting in late 2007, SMEs‘ share in total credit declined by about 5 percentage points to just over 20 percent, while SMEs‘ share in total corporate credit dropped from about 52 percent to some 44 percent. While growth in total banking sector credit remained relatively high until the escalation of the global crisis in late 2008, SME credit growth started to decelerate as early as the beginning of 2008. Between December 2006 and November 2009, cumulative growth in SME credit amounted to some 35 percent, which was only about half the rate of growth in other (non-SME) corporate credit. Non- performing loans SMEs rose from below 4 percent in the middle of 2008 to almost 8 percent.

vi

Actions in three priority areas will support broader-based growth that incorporates SMEs and is better distributed geographically.

A first key priority is to alleviate the, mainly financial, obstacles that constrain the growth of Turkish SMEs.

21. Difficult access to finance prevents the largest element in the Turkish enterprise sector from reaping scale economies. SMEs account for 79.4 percent of employment, 44.6 percent of total investments, 67.4 percent of total sales, 25-30 percent of total exports, 57.3 percent of total value added and 25 percent of bank credit (indeed, given data limitations and the size of the informal sector, the contribution of SMEs to the economy may well be somewhat underestimated). Given SMEs‘ scale, the development of a more productive and more outward-oriented SME sector is a crucial development challenge for Turkey. A healthy SME sector can not only provide increased employment opportunities for a rapidly increasing workforce and promote regional development, but is also crucial to increasing the resilience of the economy to future external shocks.

22. Turkish SMEs grow more slowly than other firms, the opposite of international experience. Analysis of firm dynamics Figure 7: Growth rates (percent) of Turkish SMEs relative to SMEs in indicates that small (11-50 other countries (2004-2007) employees ) and, especially, 0 medium firms (51-250 Russia Poland Ukraine Romania ECA EU-8 EU-10 -2 employees) grow more slowly than all other size categories, -4 with employment growth 16 -6 percent lower than micro firms and 5 percent lower than large -8 -7.6-7.7 firms. This is contrary to what -10 -8.5 -8.8 is observed in comparator -9.8 -9.8 -10.7 -10.8 -10.6 countries, where SMEs grow -12 -11.2 -11.3 -11.9 faster than large firms. -12.4 -14 -12.7 Comparison with other Small (11-50) Medium (51-250) countries also shows that Source: Turkey ES 2008 SMEs in Turkey are, on average, older. This is especially true for medium-sized firms, with 60 percent in Turkey being more than 16 years old, compared to 20 percent in the EU-10. This might indicate that SMEs in Turkey face barriers to their expansion that force them to remain at a smaller – and suboptimal – scale of operations. By contrast, the demographics of Turkey‘s large firms are in line with values in other countries. The slower growth of Turkish SMEs suggests that existing policies and regulations may have more distortionary effects for SMEs than for either micro or large firms. It seems likely that SMEs have neither the capacity of large firms nor the flexibility of micro firms to cope with the effects of these policies.

23. Problems with access to finance seem to be the most important constraint to the growth of SMEs. According to econometric analysis, one percent more usage of external finance for investment is related with 0.3 percent higher employment growth. The association of a loan or a line of credit with employment growth is even stronger and is estimated to have an effect on employment growth of 33 percent.

24. Improved ability on the part of banks to assess borrowers‟ creditworthiness, plus targeted interventions to ease collateral requirements, could help ease financial constraints to SME growth. Structural measures to enhance the ability of banks to assess the creditworthiness of SME borrowers appear necessary to help SMEs to tap into bank credit. Such measures would include (i) encouraging the vii expansion of coverage of existing credit bureaus, the Credit Registry of the Central Bank and the Credit Bureau of Turkey (KKB), and (ii) accelerating the adoption of the new Commercial Code, in order to enable SMEs to benefit from a simplified set of financial reporting standards.

25. Enhancing the role of the Credit Guarantee Fund (CGF) may also help improve SMEs‟ credit access. The CGF has played an important role in facilitating SMEs‘ access to credit by easing collateral requirements (especially since its recapitalization in 2007). The new CGF model, with Treasury involvement for a period of two years, is a positive initiative that expands the capacity of the CGF to serve the financing needs of SMEs following the credit crunch in the aftermath of the crisis. Considering ways to make the new CGF scheme more active would, hence, be a priority. Furthermore its scope could be enhanced to better target the needs of medium-sized firms. In fact, since 1994, the bulk of credit guarantees provided by the CGF has benefited micro and small enterprises, with only 11 percent of the guarantee fund being used by medium-sized firms.

A second priority is to enhance SMEs’ ability to adopt and use knowledge.

26. Access to the sources of higher efficiency needs to be extended to a wider share of firms. Turkey needs to open up the sources of growth beyond firms that are already sufficiently competitive to be direct exporters (and beyond already-successful manufacturing poles). Success in this effort will increase the resilience of the Turkish economy to future shocks in global demand. It will also ensure that the productive base of Turkish manufacturing is more evenly distributed across Turkish regions. Since the 1980s, the liberalization of the Turkish economy has offered new opportunities related to the general rise in trade in intermediate goods and in international capital mobility. Integration into global trade and investment flows has been accompanied by a significant spatial transformation of the Turkish economy, characterized by the emergence of a number of new industrial agglomerations far from the earlier manufacturing regions. These new centers are the so-called ―Anatolian Tigers.‖ Clusters of industries have formed in various parts of the country, with specialization in both traditional and more technologically advanced sectors, and have become the core of manufacturing and exporting activities. In response to these developments, the government has activated a number of instruments to foster the ability of SMEs to participate in global markets. The rationale of several such interventions has been to remove obstacles to the competitiveness of SMEs related to the business environment. For instance, SMEs in the manufacturing sector have been encouraged to locate in appropriately planned ―small industrial estates" (KSS) and ―organized industrial zones‖ (OSB) that can ease investment climate constraints by providing a number of advantages in terms of infrastructure services and regulation of business activity.

27. The local business and institutional environment combines with country-wide features to determine firms‟ incentives to adopt and use innovative modes of production and organization. The availability of a research base at the local level can, for example, encourage innovative behavior, if contacts exist between firms and local research organizations. The availability of a skilled workforce is also highly dependent on the quality of the local higher education and vocational training. The ease of access to bank finance is, on its part, conditional on the development of the local banking sector, as well as on personal contacts that may facilitate relational lending practices. Local conditions also influence the effect that the regulatory environment has on firm operations, since a large number of operating licenses are awarded at the local level. As a result, the effects in terms of knowledge transfer of linkages between globally connected firms and local suppliers may vary widely depending on local conditions. Analysis of Turkish production networks indicates that the absorptive capacity of local suppliers, especially SMEs – i.e. their ability to adopt and use knowledge – is key for successful participation in global markets. Specifically, together with a more efficient regulatory environment and easier access to finance for investment, the availability of technical skills and capacity ―to handle technology‖ is conducive to more knowledge-intensive value chain arrangements.

viii

28. In addition to wider investment climate reforms, scope exists to improve local-level programs aimed at increasing the operational capabilities and absorptive capacity of SMEs. Several governmental and non-governmental organizations provide support to firms, especially SMEs, with the objective of increasing their operational capabilities and absorptive capacity. The largest government program is offered by the Small and Medium Scale Enterprises Development Organization (KOSGEB), while the Union of Chambers and Commodity Exchanges of Turkey (TOBB) also provides such services to its associates. Following international best practices, the government could aim at reforming existing support programs by: (i) advancing the implementation of a flexible and decentralized management model to better serve the needs of SMEs on a local level; (ii) ensuring that the services on offer are not already available on market terms to SMEs, in order not to crowd out private providers; (iii) rationalizing the services on offer to create a single entry-point which could help SMEs better understand their business needs and opportunities; (iv) expanding the scope of support schemes beyond micro firms to better cater to the needs of larger SMEs.

A third priority is to further reform and strengthen the regulatory capacity of the government.

29. Turkey has taken important steps to improve the regulatory environment. The Government has paid particular attention to establishing institutions and mechanisms for regulatory reform; enacting legal reforms conducive to simplification of the legal framework; and introducing, through pilot projects, a number of regulatory tools to improve the quality of regulations. In this process, achieving EU harmonization has been a key driver of reform and Turkey has partially embraced the EU Better Regulation agenda in a number of areas. As a result, Turkey has established solid pillars of a regulatory system that have the potential to develop into a ―whole-of-government‖ approach to regulatory management and reform. The Coordination Council for the Improvement of the Investment Environment (YOİKK) has become a key structure where the private sector makes contributions to the process of improving the investment climate. The Council conducts its agenda with the help of 12 Technical Committees working on specific issues with participation of both public and private institutions. However, the different responsibilities allocated to institutions are not always linked towards a single regulatory reform strategy. This creates difficulties when it comes to establishing priorities and taking the lead for reform, and it often also results in overlapping responsibilities within and across levels of government that make implementation cumbersome, thus directly affecting business operation.

30. Building on recent progress, regulatory reform could aim at a clearer strategic vision, improved horizontal and vertical coordination among levels of government, and enhanced consultation with the private sector. In order for regulatory reform to produce substantial effects on the regulatory burden experienced by the business sector, a number of steps are necessary. First, there is a need for support at the highest political level that is translated into a clear, coherent and comprehensive strategy for regulatory reform (country-wide and including all components of a regulatory system). Second, coordination among different institutions and of different initiatives with similar objectives could be improved. Third, there is a need to link regulatory reform to clear and measurable economic targets and objectives in the medium and long term. Fourth, capacity building across the administration remains a crucial element for success. Efforts in different directions to train officials in the use of modern regulatory tools testify to the need to dedicate resources to this goal. Fifth, consultation with private sector stakeholders should be mandatory and institutionalized, with the existing YOİKK platform offering a good starting point.

31. Complementing other reforms, the establishment of Development Agencies (DA) could offer an opportunity to ease investment climate constraints by providing an interface for businesses at the local level. The Government, through the State Planning Organization, is currently in the process of making Development Agencies operational, with the goal of having 26 DAs that will cover the entire

ix country. DAs are potentially well placed to be an interface between business and government, provided that they can preserve a light organizational structure and remain at arm‘s length from the government, with the objective of minimizing the risk of capture by local interests. Additionally, an essential condition for DAs to be able to perform their tasks will be the existence of internal Government regulations recognizing their formal role vis-à-vis central and local government. Looking ahead, DAs could perform a number of useful functions. First, as already intended by the Government, they could act as “one- window” shops. Even short of radical reform of responsibilities for the issuance of licenses and permits, DAs could perform a useful facilitator role between issuing agencies and firms. Second, DAs could act as information points for businesses, in close cooperation with TOBB as well as with Government agencies, such as KOSGEB and TÜBITAK. The objective would be to rationalize financial and non-financial support initiatives – especially for SMEs who normally face high information costs – available at the local level. Third, the FDI promotion function via Investment Support Offices could be carried out in close coordination with ISPAT, the national FDI promotion agency.

x

Summary of Policy Objectives and Options

Objectives Options  Encourage the expansion of coverage of existing credit bureaus. Ease constraints on the ability of  Accelerate adoption of the new Commercial Code in order to SMEs to grow in size and generate enable SMEs to benefit from a simplified set of financial reporting employment standards.  Consider ways to make the new CGF scheme more active, expand its reach, and allow it to better reach the medium-sized firm segment.  Advance the implementation of a flexible and decentralized management model for SME support programs.  Ensure that the services on offer are not already available on market terms to SMEs in order not to crowd out private providers Increase the “absorptive capacity” of such services. of SMEs at the local level  Create a single entry-point for the various SME support programs, possibly in cooperation with DAs, in order to help SMEs better understand their business needs and opportunities.  Expand the reach of support programs beyond micro-enterprises, to better serve the needs of larger SMEs.  Map current regulatory reform initiatives in a single strategic document setting priorities and sequencing of reforms.  Strengthen the institutionalization of regulatory reform by creating a single oversight body for regulatory reform in the Prime Minister‘s office.  Strengthen YOİKK‟s role to improve the business environment and advocate for regulatory reform.  Design a comprehensive administrative simplification strategy Improve the regulatory with clear objectives, targets and review criteria for lower levels of environment for businesses regulation to improve the business environment.  Improve coordination mechanisms inside the administration when preparing laws and regulations.  Strengthen coordination and cooperation among levels of government.  Make consultation with stakeholders compulsory for the preparation of new and amended laws and regulations.  Continue implementation of Regulatory Impact Analysis (RIA).  Use existing e-government strategies to support regulatory reform and simplification efforts. Development Agencies (DAs) could serve as  ―one-window‖ shops, without radical reform of competences for the issuance of licenses and permits. Improve access to information on  information point for businesses, in close cooperation with regulatory requirements and business associations with a local presence, such as TOBB, as well government support programs as with Government agencies, such as KOSGEB and TÜBITAK, etc.  FDI promotion, with Investment Support Offices acting in close coordination with ISPAT.

xi

Chapter 1. CHALLENGES FACING THE TURKISH BUSINESS SECTOR

1.1 The global crisis is posing new challenges to the Turkish business sector. The external economic environment for developing countries has deteriorated markedly since 2008. World economic growth in 2009 was negative (-0.6 percent), with the Euro Area, Turkey‘s main export market, expected to experience a contraction in GDP of -4.1 percent.1 As a result, the demand for Turkish exports has fallen dramatically with serious consequences for industrial production. Exports fell by 23 percent in 2009 and industrial production by 9.6 percent compared with a year earlier. In addition to slowing demand for exports, capital flows to Turkey have fallen dramatically from USD 50.3 billion in 2007 to USD 14.7 billion in 20082 and risk premia have escalated significantly, as shown by the CDS spread for Turkey from 167 basis points end 2007 to 200 in late 2009 (Figure 1-1). During the period of high growth from 2002-07 Turkey relied heavily upon net capital inflows. The corporate sector now faces major challenges in continuing to attract external finance and in rolling-over its short term external debt, as shown by Figure 1-2.

Figure 1-2: Short-term external corporate debt in Figure 1-1: 5-year CDS spreads, Jan 2007-Dec 2009 Turkey, million USD

1000 60,000

800 50,000 40,000 600 30,000 400

Basispoints 20,000 200 10,000

0 0

07 08 09

07 08 09

07 08 09

07 08 09

- - -

- - -

- - -

- - -

2002 2003 2004 2005 2006 2007 2008 2009

Jul Jul Jul

Jan Jan Jan

Oct Oct Oct

Apr Apr Apr

Source: Bloomberg Source: CBRT

1.2 Effects on the growth prospects of the Turkish economy are likely to be durable. Turkey‘s economy contracted by 4.7 percent in 2009, and unemployment increased to 14 percent from 11 percent in the previous year. The Government‘s Medium-Term Program (MTP) sets out a realistic macroeconomic framework that, conservatively, foresees a rather slow recovery scenario.3 This mirrors projections for the world economy and for Turkey‘s major trading partners, and foresees a return to potential growth of the order of 5 percent only by 2012. This growth is associated with a decline in unemployment of only 1.5 percentage points from the peak of 14 percent in 2009. Economic activity is projected to recover weakly to 3.5 percent in 2010, 4.0 percent in 2011, and 5.0 percent in 2012. The growth process is expected to be led by the private sector, with an expected pick-up in private gross fixed capital formation to 8 percent in 2010.

1.3 Continued reform of the investment climate is key to mitigating the effects of the crisis and closing the income gap with more developed countries. The MTP outlines a post-crisis reform agenda for shared growth that follows the development axes established by the Ninth Development Plan for 2007-2013. Many of the planned actions aim at making the investment climate more conducive to private sector led growth, based on the notion that the investment climate – ―the set of location-specific factors

1 IMF World Economic Outlook Database, April 2010 2 Capital inflows including net errors and omissions, excluding change in official reserves and IMF credits (World Bank data). 3 On September 16, 2009, the government announced its new Medium-Term program (MTP). The MTP was followed on September 18 by the more detailed Medium-Term Fiscal Plan. The MTP gives aggregate fiscal targets for the period 2009-12.

1 shaping the opportunities and incentives for firms to invest productively, create jobs and expand,‖ as defined in World Bank (2005) – can significantly impact productivity, growth and economic activity. Important investment climate reforms contemplated in the MTP include further labor-market reform, tax administration reform, increased effectiveness of credit guarantees, expansion of education and vocational training and increased credit access for SMEs. Continued commitment to such reforms is crucial to narrow the income gap between Turkey and more developed OECD economies, which, despite sustained economic growth since 2002, remains wide (Box 1-1).

Box 1-1: Turkey‟s Income Gap Whereas Turkey has been successful since 2001 in securing a reasonable level of macroeconomic stability, convergence to the per capita income levels of OECD countries has been slow. In 2007, income per capita was 18 percent of the US level, up from 17 percent in 1960, while the gap relative to the EU15, as European countries were converging to US levels, has actually widened, with income per capita relative to the average of the EU15 economies decreasing from 26.5 percent in 1960 to 21.6 percent in 2007.

Figure A 100 1. Labor productivity as share of U.S. Value 80

60 % 40 26.7 20 13.3 EU15 Korea Turkey

0

1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 1950 Source: Penn World Table Version 6.3 80 2. Labor force participation rate 75 70 65 60 % 58.3 55 50.4 50 45 United States EU15 Korea Turkey

40

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1980 Source: WDI

The evolution of income per capita is driven by labor productivity and labor force participation (Figure A).4 Whereas low labor force participation appears to be the main driver of slow convergence in per capita incomes (Panel 2), labor productivity convergence has also been slow, especially when compared to countries, such as Korea, that had similar levels of labor productivity and income per capita until the 1970s (Panel 1). Capital intensity and total factor productivity (TFP) are the drivers of labor productivity and are affected by the incentives to invest and innovate associated with the policy and institutional framework (the investment climate) in which the business sector operates.

4 Labor productivity is measured as real GDP per worker in 2005 constant prices, USD, as share of U.S. value. Labor force participation rate is measured as the proportion of the population ages 15 and older that is economically active: all people who supply labor for the production of goods and services during a specified period.

2 1.4 The objective of this report is to explore empirically the role of the investment climate in determining the business sector‟s economic performance. The report uses the information contained in a 2008-2009 enterprise survey of Turkish enterprises to estimate the effects of the business environment on firm-level productivity and various measures of enterprise performance, such as the ability to generate employment or the probability to export and attract foreign capital. The main advantage of this type of surveys is that information is gathered directly from firms‘ managers on the quality of the physical and institutional infrastructure, as well as on basic economic performance measures at the firm level. Improved business sector performance, in turn, translates into aggregate economic improvements, with increased firm-level productivity and capital intensity translating into higher aggregate productivity, thus contributing to the reduction of income per capita differences.

Box 1-2: Recommendations in key reform areas from the 2007 Investment Climate Assessment Regulatory environment  Streamlining of business entry and exit  Reform of the firm registration, licensing and inspections regimes  Reform of taxation  Reform of access to land for business  Streamlining of public institutions involved in customs procedures

Labor market and skills  Reform of the fiscal and institutional framework governing labor  Increasing the flexibility of labor market regulations  Reform of the education and training system, in order to match workers‘ skills with the needs of enterprises

Innovation, technology adoption and ICT  Legal and institutional reform of the National Innovation System  Reform of the IPR legislation and alignment with the EU  Reform of the telecommunications sector

Adoption of quality standards and certification  Improving the legal and institutional framework governing firm access to and enforcement of standards

Access to finance and corporate governance  Strengthening financial reporting and credit information on firms  Increasing the use of collateral for lending transactions  Enhancing the legal and institutional framework for corporate governance

Access to infrastructure  Institutional reforms and capital investments in the electricity and transport sectors  Institutional reforms affecting the availability of trade-related services

1.5 Both in 2004-05 and 2008-09 total factor productivity in Turkey is highly correlated with the investment climate. Analysis of firm-level data collected in 2008 and 2009 and presented in this report finds a strong association between firm-level productivity and the investment climate, confirming the results of the 2007 Investment Climate Assessment based on 2004 and 2005 data. Both assessments pinpoint reforms aimed at improving the business environment by analyzing a number of key economic variables – productivity, employment, wages, exports and FDI – in relation to a number of investment climate variables. The 2007 study also identified labor productivity as the most critical challenge for the convergence of Turkey‘s income per capita, showing how it accounts for 80 percent of the per capita income gap between Turkey and the EU-15. Labor productivity improvements, in turn, are achieved by

3 increasing: (a) capital per laborer, that is, investments in physical assets such as machinery, infrastructure, and buildings, and (b) TFP, the residual contribution from factors such as technology adoption, labor skills, educational achievement, and competition among – and management of – firms.

1.6 Investment climate reforms implemented since 2007 acquire renewed significance in the wake of the recent crisis. Consistent with the Ninth Development Plan (2007-2013) and with the recommendations of the previous investment climate report (Box 1-2) the Government has undertaken a number of structural reforms in several areas of the business environment. Notable measures have included (i) tax simplification accompanied with a reduction of the corporate income tax rate from 30 percent to 20 percent; (ii) streamlining of procedures for firm start-up; (iii) adoption of a modern FDI promotion strategy coupled with the restoration of legal certainty for land ownership by foreigners; (iv) simplification of customs procedures and e-transformation of customs offices with the introduction of a Computerized Customs Activity System (BILGE); (iv) planned enactment of a new Commercial Code improving corporate governance, as well as the protection of investors‘ and minority shareholders‘ rights; (v) reform of R&D legislation aimed at increasing the private sector share of R&D; (vi) first phase of labor market reform geared to lower non-wage labor costs, accompanied by actions to strengthen the development of a competency-based skill-building system and continued reforms of curricula in secondary school. These reforms acquire renewed significance for the business sector in the current macroeconomic environment.

1.1 Macroeconomic Setting 1.7 In the aftermath of the 2001 crisis sound macroeconomic management and abundant global liquidity underpinned steady growth. After a banking crisis in 2001 that led to a sharp recession and a restructuring of the financial sector, GDP growth averaged nearly 7 percent per annum between 2002 and 2007 (Table 1-1). An important engine of growth was private investment, in part driven by large capital inflows, which contributed to trebling Gross Fixed Capital Formation by the private sector between 2002 and 2008 (from TL 43 billion, or less than USD 30 billion, to TL 152 billion, or USD 117 billion).

1.8 Sustained GDP growth failed to make a visible impact on the unemployment rate inherited from the 2001 crisis. Capital and exports intensive economic growth, combined with a rapidly growing and young labor force (an estimated 700,000 workers join the labor force each year), led to a decline in labor force participation, from 49.6 in 2002 to 46.9 percent in 2008, compared to an OECD average of 70.8 percent.5 Low labor force participation in Turkey reflects particularly low participation among the female population. Only 26.7 percent of women participated in the labor force in Turkey in 2008, compared to an OECD average of 61.3 percent and 65.3 percent of women in EU15.

1.9 Since 2008, the global turmoil has placed strains on Turkey‟s economy. Turkish manufacturing has been hard hit by the drop in global demand, with the effects felt in terms of unemployment, economic hardships at the household level and poverty. The Turkish economy had already begun to slow down from 2007. Annual growth in 2007 fell to 4.7 percent from 6.9 percent in 2006. This slowdown was reflected in, among other things, a build-up of inventories by Turkish firms in the order of 2 percent of GDP between the first quarter of 2007 and the third quarter of 2008. Unemployment has also risen sharply. After remaining stable at levels below 10 percent for several years, the unemployment rate in 2009 averaged 14 percent.

5 OECD Employment Outlook (2009) and World Bank calculations.

4 Table 1-1: Turkey: Key Economic Indicators, 2002-2009

2002 2003 2004 2005 2006 2007 2008 2009 Real GDP (bln USD, current prices) 230.5 304.9 390.4 481.5 526.4 648.8 742.1 617.6

Real GDP per capita (USD) 3,325.9 4,341.4 5,486.7 6.681.4 7,213.9 8,781.8 10,039.9 8,255.1

Private gross fixed capital formation (bln TL) 43.4 62.0 97.4 115.1 143.3 151.9 152.4 125.8

(Growth rate in percent)

Real GDP Growth 6.2 5.3 9.4 8.4 6.9 4.7 0.7 -4.7

CPI Inflation (%) 45.2 25.3 8.6 8.2 9.6 8.8 10.4 6.3

Long-term interest rate 63.5 44.1 24.9 16.2 18.0 18.3 19.2 11.9

Short-term interest rate 59.5 38.5 23.8 15.6 17.9 18.3 18.9 10.9

Exchange rate 1.5 1.5 1.4 1.3 1.4 1.3 1.3 1.6

(In percent of GDP unless otherwise indicated)

Labor force participation 49.6 48.3 46.3 46.4 46.3 46.2 46.9 47.6

Unemployment 10.3 10.5 10.8 10.6 10.2 10.3 11.0 14.0 Saving-investment balance

Domestic savings 18.6 15.5 16.0 15.9 16.5 15.5 16.9 14.2 Investment (Contributions to growth)

Public 0.3 -0.6 -0.2 0.7 0.1 0.2 0.4 -0.1

Private 2.0 3.1 5.6 3.1 3.1 0.6 -2.0 -4-4

Fiscal sector Primary balance (IMF defined) 3.3 4.8 5.5 5.0 4.6 3.1 1.7 -2.1 Public external debt 7.1 7.5 8.2 8.0 8.1 6.7 6.8 7.8

Monetary indicators

Broad money 39.9 35.2 34.6 40.5 42.4 43.9 48.7 50.01

Claims on private sector 14.5 14.5 17.3 22.2 25.9 29.5 32.6 33.01

External sector Current account balance (bln USD) -0.6 -7.5 -14.4 -22.1 -32.3 -38.3 -41.9 -14.0 Trade balance (bln USD) -6.4 -13.5 -22.7 33.1 41.1 -46.8 -53.0 -24.9

Exports (fob, bln USD) 40.7 52.4 68.5 78.4 93.6 115.4 140.8 109.7

Imports (cif, bln USD) 51.6 69.3 97.5 116.8 139.6 170.1 202.0 140.9

FDI 0.5 0.6 0.7 2.1 3.8 3.4 2.5 1.0 CBRT Reserves (bln USD) 28.3 35.3 37.6 50.2 60.7 74.7 72.9 74.8 Source: CBRT, IMF, OECD, SPO, Turkstat, World Bank. 1May 2009

5 1.10 The current account deficit and inflation persistence have now diminished in immediate importance while the fiscal challenge has increased. Turkey‘s current account deficits are not expected to return to pre-crisis levels in the medium term, although higher oil prices could still pose a challenge. Inflation persistence appears to have receded, with inflation in 2009 still below the inflation target. The monetary framework has not been altered and will continue to pursue unchanged inflation targets (6.5 percent in 2010 and 5.5 percent in 2011), in line with the framework in place since 2006. On the other hand, public debt and the budget deficit have increased as a result of the global crisis – the budget deficit stands at -2.1 percent of GDP in 2009 compared to 3.1 percent in 2007 and public sector debt at 7.8 percent. This has sharpened the challenge to fiscal management and underlined the central importance of the MTP and the associated budgets to reduce economic uncertainty.

1.11 Concerns over external financing were prominent before and as the global crisis hit. Since the 2001 crisis, the Turkish Treasury has pursued a conservative strategy of financing itself largely using domestic currency debt instruments, removing much of the foreign exchange risk from the public debt portfolio. After declining in 2001-05, the external debt-to-GDP ratio increased by more than 4 percentage points in 2006, driven by corporate sector external borrowing. External debt-to-GDP stood at 43.9 percent at end-2009. However, the subsequent sharp contraction in economic activity has reduced the overall need for foreign financing in 2009, while at the same time most large corporations have either retained market access or been able to draw on their own FX holdings. Turkey‘s foreign financing needs have fallen but will remain high in 2010-12. The current account deficit is contracted from USD 41.9 billion in 2008 to USD 14 billion in 2009. A sharp contraction in domestic investment has cut the demand for medium- and long-term (MLT) borrowing by the private sector, which has been reflected in lower rollover ratios. Repayments are also lower beyond 2010. The overall financing gap was USD 6 billion in 2009, and thus should easily be financed through reserves, which were USD 74.8 billion in 2009. Turkey should be in a position to attract further FDI given privatization efforts and the potential for mergers and acquisitions. Similarly, net portfolio flows are assumed to be positive, consistent with continued Eurobond issuance and the potential of the domestic capital market to raise financing. Projections on capital inflows are of course conservative if compared to the pre-crisis period of record global liquidity.

Figure 1-3: GDP by sector in Turkey, 1998-2009, Figure 1-4: Value added as share of GDP by sector, 2008 current prices, billion TL

150 100% 80% 100 60%

40% 50 20%

0 0%

Agriculture Industry Services Industry Services Agriculture Source: WDI

1.12 The services sector has not only been the fastest growing in the past decade, but is also hit by the economic crisis to a lesser extent than the manufacturing sector. Since 1998, the services sector has grown rapidly, with an average 30 percent annual growth and increased relative contribution to GDP from 51 percent in 1998 to 62 percent in 2008 (Figures 1-3 and 1-4). Meanwhile, the industry sector in Turkey remains small, both in relation to the domestic services sector as well as when comparing to

6 other emerging economies. Merely 28 percent of Turkey‘s GDP came from the manufacturing sector, with a real decrease from USD 58 billion in 2008 to USD 53 billion one year on.

1.13 Following high export growth since 2002, changed global demand conditions appear to have affected the destination of Turkish exports. Turkish exports averaged 20 percent annual growth in the period 2000-2008 (Table 1-1). Exports continue to be dominated by manufactured goods as well as machinery and transport equipment, each taking up 28 percent of total exports, with a total value of USD 4.8 billion in 2009. The machinery and transport equipment sector in particular has increased its relative share of exports, from 21 percent in 2000. For a considerable time, the majority of Turkish exports have been targeting the EU-market, with the relative share of total exports going to the EU27 region averaging 58 percent in the years 2000-2007. This flow has however experienced a significant drop to 46.3 percent in 2009. A similar negative development has been noticed in exports to the United States, with a decline from 11.3 percent in 2000 to 3.1 percent in 2009. Instead, the shares of exports to the Middle East and Africa have increased from 8 to 16.5 percent and from 4.9 to 9.9 percent respectively in the same period.

Figure 1-5: Turkey‟s exports by industry 2000-2009

100% Commodities and transactions, n.e.c 90% Miscellaneous manufactured articles 80% 70% Machinery and transport equipment

60% Manufactured goods 50% Chemicals and related products, n.e.s 40% 30% Mineral fuels, lubricants and related materials 20% Animal and vegetable oils, fats and waxes 10% Crude materials, inedible, except fuels 0% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Food and beverages

Figure 1-6: Turkey‟s exports by partner country, 2000-2009

100% Other 90% Ukraine 80% 70% United States 60% Russian Federation 50% Africa 40% 30% Asia 20% Middle East 10% 0% EU27 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Source: OECD

7 1.14 The technology content of Turkish exports remains dominated by low-technology manufactures with recent increases in 6 medium- and high-technology products. Figure 1-7: Exports by technology level, share of GDP, Turkey‘s medium- and high-technology 1996-2008 exports have increased significantly, with 8% real values reaching 26.5 billion in 2008, a 360 percent increase since the turn of the 6% century. Nonetheless, exports with higher technology content are a relatively low 4% contributor to Turkey‘s GDP, averaging 3.4 percent for the past five years. Medium- and 2% high-technology imports on the other hand contributed with 8 percent of GDP in 2008, 0% creating a substantial trade deficit for higher 1996 2000 2004 2008 technology goods and suggesting the need Low-tech Medium- & High-tech Resource-based for further actions in reforming export flows Source: UN COMTRADE, staff calculation from Turkey.

1.15 Inward FDI remains weak in an international comparison, with a recent shift away from telecommunication and finance. Net inflows of FDI into Turkey reached USD 6 billion in 2009, an inflow well below the USD 15.4 billion Figure 1-8: FDI net inflows/GDP, 2009 recorded in 2008 as well as the pre-crisis Slovakia* 0.5 level in 2007 (USD 19.1 billion). The sharp Turkey 1.0 decline is to a large extent a combination of Hungary* 1.1 the negative impact that the financial crisis Czech Rep. 1.4 has had on overall international investment Brazil* 1.6 flows and the stagnation of the large-scale Poland* 2.7 ECA average* privatizations that took place in Turkey in 2.7 7 Russia* 2.8 the years 2005 and 2006. Most FDI Turkey 2007 3.4 inflows during this period were in the Romania 4.2 financial services and telecommunications

0.0 1.0 2.0 3.0 4.0 5.0 sectors. Three years on, the relative % significance of FDI in these industries has * 2009 est. Source: IMF International Financial Statistics diminished and instead, electricity, gas and water supply as well as manufacturing hold the largest shares of investments from non-residents (33 percent and 28 percent respectively). Within manufacturing, the most significant increase in FDI in the first ten months of 2009 compared to the same period one year prior was in the chemicals and motor vehicles sectors (from USD 89 million to USD 306 million and USD 64 million to USD 208 million respectively). Meanwhile, the manufacturing industry has experienced a decline of FDI inflows into the sectors ‗other manufacturing‘ (USD 1,689 million to USD 382 million) and ‗food and beverages‘ (from

6 Technology level of exports is here defined according to UNIDO‘s methodology classifying manufactured goods into four sub- categories: resource-based, low-, medium- and high-tech exports, based on the Standards International Trade Classification (SITC) Revision 3. Examples of resource-based goods include: beverages, cut gems and glass, petroleum/rubber products, prepared meats/fruits, vegetable oils and wood products. Examples of low-tech manufactures include: clothing, footwear, furniture, headgear, leather manufactures, plastic products, pottery, simple metal parts/structures, textile fabrics, toys, and travel goods. Examples of medium-tech manufactures include: chemicals and paints, engines, fertilizers, industrial machinery, iron, motors, pipes/tubes, plastics, ships, switchgear, synthetic fibers, vehicles and watches. Examples of high-tech manufactures include: aerospace, cameras, office/data processing/telecom equipment, optical/measuring instruments, pharmaceuticals, power generating equipment, transistors, turbines and TVs. For detailed technological classifications of exports, see UNIDO (2009). 7 Privatization implementations in Turkey in 2005 and 2006 totaled USD 16.3 billion in domestic and international sales (Undersecretariat of Treasury)

8 USD 939 million to USD 120 million).8 The recent adoptions of a modern FDI promotion strategy by the Turkish government, along with the restoration of legal certainty for land ownership by foreigners, are steps in the right direction to improve investment levels in the country.

1.2 Effects of the Global Crisis on the Corporate Sector: Preliminary Evidence 1.16 Four factors appear to have combined to increase the impact of global events on Turkey.9 Although none of these factors is entirely specific to Turkey, their magnitudes and combination go a long way towards explaining the impact of global events on the real economy in Turkey.

1.17 First, the speed of the economic expansion in the preceding five year period had led to a marked inventory and capital build-up in Turkish manufacturing, with inventories accounting for fully one quarter of aggregate GDP growth in 2007 and the first three quarters of 2008, and capital formation growing at an average of 15.5 percent per year in 2002-2007.

1.18 Second, the high level of uncertainty experienced by all emerging markets during the crisis was compounded in Turkey by perceived vulnerability to the scarcity of external financing. Turkey‘s most notable macroeconomic challenge in the pre-crisis period, the current account deficit (averaging more than 5 percent of GDP over the five years from 2004-08), had been financed by increasing debt-creating flows to the private sector. By 2009, Turkish corporations needed to roll over total debt amortizations of the order of USD 100 billion. About a third of this total was not true foreign exposure, since it captured offshore lending by Turkish banks to their Turkish clients, and a further (hard to quantify) portion was secured by assets held overseas by Turkish nationals. Nonetheless, uncertainty over the magnitude of this exposure and its effect on the economy was high at the outset of the crisis.

1.19 Third, Turkey‟s banking system, having weathered several earlier crises, was quick to cut lending to all but the most creditworthy borrowers. There was no visible slowdown in credit intermediation to the private sector prior to September 2008. Domestic credit then retrenched significantly in late 2008. Turkey‘s domestic financial system, although it is well-capitalized and prudential regulations meet modern standards, is shallow for an economy of Turkey‘s size. After surviving crisis and restructuring in 2001-02, the Turkish banking sector was understandably conservative faced with the high uncertainty of late 2008 and early 2009.

1.20 Fourth, the composition of Turkey‟s exports exacerbated the demand shock. Turkey‘s exports – concentrated in hard-hit sectors such as automotive vehicles, consumer durables, and capital goods and machinery – made it vulnerable to a dip in export demand. Export volumes in the first half of 2009 were down 11 percent and this combined with price effects to create a loss in export earnings of more than a third (y/y). Exports had continued to perform strongly right up to the crisis: FOB export growth in 2008 (y/y) was 23 percent while imports (CIF) rose by 18 percent. Following the crisis, imports have contracted even faster than exports: as a result, net exports have contributed positively to GDP growth since the onset of the crisis.

1.21 A crisis impact survey fielded by the World Bank confirms that the global economic and financial crisis has had a significant impact on the Turkish enterprise sector. The Enterprise Financial Crisis Assessment Survey (EFCAS) was conducted in the summer of 2009 in six countries including Turkey (Box 1-3). Its objective was to identify the channels through which the global economic crisis is affecting the corporate sector and how firms have reacted to the shock.

8 See Undersecretariat of Treasury, International Direct Investment Information Bulletin (December 2009) 9 Restoring Equitable Growth and Employment Programmatic Development Policy Loan (REGE-DPL), draft December 14, 2009

9 Box 1-3: The Enterprise Financial Crisis Assessment Survey (EFCAS) The Enterprise Financial Crisis Assessment Survey (EFCAS) was carried out in June and July 2009 and covers 1,686 enterprises from Bulgaria, Hungary, Latvia, Lithuania, Romania, and Turkey. In all countries, the EFCAS covers a subsample of the 2008-2009 Business Environment Enterprise Survey (ES), carried out by the World Bank and the European Bank for Reconstruction and Development in 30 economies of Europe and Central Asia. Turkey‘s EFCAS sample is composed of 514 enterprises.

The ES and EFCAS samples are representative of the universe of non-agricultural private sector formal firms (with at least five employees) in the economy for groups D, F, G, H, I and subgroup 72 of the United Nations Statistics Division ISIC Rev. 3.1. Results are estimated through the application of sampling weights – that denote the inverse of the probability that the observation is included due to the sampling design – to the original data. Therefore, results are representative of the non-agricultural private economy in Turkey.

1.22 The main channel through which Turkish firms have been affected is a drop in demand. While a large majority of Figure 1-9: Effects of the Crisis firms across countries, including Turkey, declared Romania 78.47 increase the level of contraction in demand, the debt Bulgaria 78.12 most important effect of the increase input cost crisis on their business, the Latvia 75.43 share of enterprises citing a Turkey 71.3 reduce access to credit combination of supply-side factors in Turkey – notably Lithuania 70.77 drop in demand for its input costs, debt levels and Hungary 70.32 products or services access to credit (about 19 other percent) – was one of the 0% 20% 40% 60% 80% 100% highest among the surveyed Source: EFCAS countries.

1.23 Reported declines in volumes of Figure 1-10: Sales in the corporate sector: June 2008 - June sales between 2008 and 2009 have been 2009 substantial. Turkish firms reported a significant drop in sales amounting to an 0 average 38.7 percent in June 2009 relative -10 to June 2008, as shown in Figure 1-9. The decline presents wide sectoral variations; -20 with fabricated metal products (32.3 -30% -25.4 percent) and garments (31.3 percent) -40 -35.9 -35.5 being hardest hit compared to the food -38.7 and textiles sectors that have been the -50 least affected with a drop of 13.2 percent -48.4 -48.0 -60 and 16.6 percent respectively. This Latvia Lithuania Turkey Romania Bulgaria Hungary development is supported by the industrial Source: EFCAS turnover index for the Turkish manufacturing sector, when measured in the same time period. The sub-sectors of manufacturing showing most significant reduction in industrial turnover coincide with sectors in the survey reporting largest drop in sales (basic metals, fabricated metal products and non-metallic mineral products). The overall decline notwithstanding, a significant proportion of firms (about 15 percent) reported increases in sales (25.4 percent on average) in the same period, which may be a first indication that the crisis may be accompanied by structural adjustments, with a redistribution of market shares across firms.

10 1.24 Foreign currency Figure 1-11: Structure of corporate liabilities exposure and short term maturity 80.1 highlight potential risks for the 80 66.4 69.4 corporate sector. Confirming the 56.9 accumulation of foreign debt and 60 49.2 49.3 the prudent attitude of the banking % 39.4 35.3 40 31.2 system towards lending in the wake 21.4 21.8 20.7 of the crisis, survey findings in the 20 summer of 2009 indicate that the

0 share of foreign currency debt in the corporate sector is considerable (22 Latvia Bulgaria Romania Turkey Hungary Lithuania percent), while debt maturity is average % of total liabilities with a term of maturity less than one year concentrated in the short-term (66 average % of total liabilities denominated in foreign currency Source: EFCAS percent).

1.25 Debt restructuring has been the predominant form of adjustment on the part of firms. As a response to the current liquidity constraint, about 33.7 percent of Turkish firms delayed payments to tax authorities and suppliers. At the same time, 45.9 percent of surveyed companies attempted to restructure their debt while only 0.2% of firms have filed for insolvency or bankruptcy. A large proportion of firms (25.4 percent) benefited from some form of state-aid.

Figure 1-12: Firms‟ survival strategies, forms of adjustment

65.8 70 63.3 % of firms overdue on obligation in last year which have filed for 60 INSOLVENCY/BANKRUPTCY during the 50.7 50.5 47.9 last 12 months 50 45.9 % of firms overdue on obligation in last year which have RESTRUCTURED any of 40 33.3 33.7 their outstanding liabilities in the last % 27.4 28.3 12 months 30 23.4 25.4 % of firms overdue on obligation in last 21.0 year which have applied for STATE AID 20 in the last 12 months 8.7 10 4.1 % of firms which delayed payments to 1.4 3.6 0.2 tax authorities or suppliers for more 0 than one week Romania Bulgaria Turkey Lithuania Latvia Hungary

Source: EFCAS

11 References

Organization for Economic Cooperation and Development (OECD). 2009. OECD Employment Outlook 2009 – Tackling the Jobs Crisis. Paris

Republic of Turkey State Planning Organization. 2006. Ninth Development Plan 2007-2013. Decision No: 877

Republic of Turkey State Planning Organization. 2009. Medium Term Programme 2010-2012. Official Gazette No. 27351, 16.09.2009

Republic of Turkey, Undersecretariat of Treasury, International Direct Investment Information Bulletin (December 2009)

United Nations Industrial Development Organization (UNIDO). 2009. Industrial Development Report 2009 – Breaking In and Moving Up: New Industrial Challenges for the Bottom Billion and the Middle-Income Countries. Vienna

World Bank. 2005. World Development Report 2005: A Better Investment Climate for Everyone. Washington, DC

World Bank, Poverty Reduction and Economic Management, Turkey Country Management Unit, Restoring Equitable Growth and Employment Programmatic Development Policy Loan (REGE-DPL). Draft December 14, 2009

12

Chapter 2. EVOLUTION OF THE TURKISH INVESTMENT CLIMATE

2.1. Continued reforms of the investment climate are required to ensure that the Turkish enterprise sector remains competitive in a more challenging global environment. Since 2001 the Turkish Government has complemented sound macroeconomic management with more or less comprehensive reforms of various aspects of the investment climate – including taxation, business registration, customs, FDI promotion, R&D and labor legislation. Since the second half of 2007, changed global conditions characterized by declining domestic and global demand, reduced international capital flows and tighter credit conditions have already taken their toll on the Turkish business sector, as outlined in the preliminary evidence reported in the previous Chapter. Whereas the depth and duration of the aftereffects of the global turmoil are still uncertain, a number of investment climate areas emerge as crucial to make the Turkish business sector more resilient to future shocks and more competitive both domestically and internationally. As argued in this report, these are related to facilitating the growth of SMEs, intensifying knowledge flows to Turkish firms, and equipping policymakers with a system capable of ensuring consistent regulatory quality.

2.1 The Investment Climate and Business Sector Performance 2.2. Business sector performance is associated with the quality of the investment climate. Whereas Turkey has been successful since 2001 in securing a reasonable level of macroeconomic stability, convergence to the per capita income levels of OECD countries has been slow. In order to help close the gap in income per capita, it is essential to tackle the underlying policy and institutional framework that constitutes the investment climate conditions in which the business sector operates. Econometric analysis of firm level data collected in 2008 and 2009 (Box 2-1) confirms that productivity, job creation, the ability of exporters to be competitive on international markets and the attractiveness of the economy for foreign investment are significantly associated with the quality of the investment climate. Similar results were obtained by analyzing data of a previous survey, conducted by the World Bank in 2004 and 2005 (World Bank 2007). In addition, the analysis shows that firms with large market shares tend to cope better with the bottlenecks imposed by the investment climate and are even able to benefit from its more positive aspects. The analysis is based on the robust methodology summarized in Table 2-1.

Table 2-1: Summary of econometric methods i. Econometric model used to identify IC effects Data Economic base ii. Evaluation of IC contributions performance: Right hand side or (survey) Left hand side or explanatory variables dependent variables

- Productivity (TFP) 135 investment IC contributions to aggregate productivity through

climate variables + average productivity and allocative efficiency. [cross other controls and Employment more than 20 Probability of IC contributions to the sample means of industry/region/size exporting employment, probability of exporting and

datafor TFP] controls +

Probability of probability of receiving FDI. section 2008 for section IC for 2008 variables and recall simultaneous effects

Manufacturing Manufacturing receiving FDI I. identification and evaluation of IC effects on economic performance in 2008 II. Comparison of IC effects on economic performance from 2005 and 2008

13 Box 2-1: The 2008-2009 Enterprise Survey This report draws its data from a core enterprise survey that provides a standardized way of measuring and comparing investment climate conditions. Private contractors conduct the Enterprise Surveys on behalf of the World Bank in Turkey and in other countries globally, utilizing a standard instrument that enables comparisons of investment climate conditions across different regions within a given country, as well as of a country with its peers globally.

Data were collected in Turkey in between April 2008 and January 2009 as part of the fourth round of the Business Environment and Enterprise Performance Survey (ES), a joint initiative of the European Bank for Reconstruction and Development and the World Bank. The objective of the survey is to obtain feedback from enterprises on the state of the private sector as well as to help build a panel of enterprise data that will make it possible to track changes in the business environment over time, thus allowing, for example, impact assessments of reforms. Through interviews with Turkish firms in the manufacturing and services sectors, the survey assesses the constraints to private sector growth and allows estimation of statistically significant business environment indicators that are comparable across countries, across 5 Turkish regions, across industries and firms of different sizes.

The sample for the Turkey survey was selected using stratified random sampling. The whole population, or universe of the study, is the non-agricultural economy. Three levels of stratification were used in Turkey: type of industry, firm size, and geographic region. The universe was stratified into 11 manufacturing industries, 1 retail industry and 6 residual industries. Size stratification was defined following the standardized definition for the rollout: small (5 to 19 employees), medium (20 to 99 employees), and large (more than 99 employees). For stratification purposes, the number of employees was defined on the basis of reported permanent full-time workers. This seems to be an appropriate definition of the labor force since seasonal/casual/part-time employment is not a common practice, except in the sectors of construction and agriculture. Regional stratification was defined in 5 regions. These regions are Marmara, Aegean, South, Central Anatolia and Black Sea-Eastern. The Turkey sample contains panel data. The wave 1 panel consisted of 1325 establishments interviewed in 2005. Note that there are additional variables for location (city), industry, and size that reflect more accurately the reality of each establishment.

Table 2-2: The ES 2008-09 Sample Industry Region Central Black Sea - Marmara Aegean Anatolia South Eastern Total Food 40 22 41 34 21 158 Textiles 71 47 11 45 5 179 Garments 64 23 12 23 6 128 Chemicals 49 20 19 14 5 107 Plastic and rubber 21 6 2 9 5 43 Non-metallic mineral products 29 37 18 17 9 110 Basic metals 4 1 5 5 4 19 Fabricated metal products 11 3 11 8 5 38 Machinery and equipment 4 1 11 12 6 34 Electronics 3 4 1 3 2 13 Other manufacturing 20 13 21 6 7 67 Retail 51 7 17 18 14 107 Construction 3 3 1 4 2 13 Wholesale 37 10 6 31 6 90 Hotel and restaurants 0 0 0 1 0 1 Transport 3 0 2 3 0 8 IT 0 0 0 1 1 2 Other services 12 2 5 8 8 35 Total 422 199 183 242 106 1,152

The nature of the sample allows exploration of the effects of the investment climate over time and across regions. Econometric analysis uses the manufacturing subset of the sample (903 establishments). Of these, 425 panel establishments were present in the 2005 survey. This allows assessing the impact for firm

14 performance of the numerous investment climate reforms that have occurred between 2005 and 2008. Furthermore, the regional stratification of the sample will allow drawing statistically significant inferences of the effects of the investment climate across five broadly defined regions (Table 2-2). The regions are sufficiently diverse in terms of institutional settings, economic structure, income per capita and product specialization to warrant a meaningful assessment of the investment climate at the regional level.

The source of the sample frame was twofold. Universe estimates were taken from the TOBB database which contains a full list of establishments in manufacturing sectors. TOBB refers to the Union of Chambers and Commodity Exchanges of Turkey. Universe estimates for service sectors were taken from the Statistical Institute of Statistics (SIS) with additional information based on SIC code from the Turkish Studies Institute (TSI). Comparisons were made between estimates in TOBB and SIS to establish that the two sources are comparable and hence can be used side by side.

Three additional modules have been added to the ES survey with interviews of over 800 manufacturing firms in July-August 2009. These modules address (i) the effects of the current global financial crisis referred to in Chapter 1; (ii) firm-level innovation referred to in Chapter 2 and (iii) production networks used in Chapter 4.

2.3. Access to finance is perceived as the most serious obstacle by Turkish firms, a negative development since 2005 and a likely consequence of the credit squeeze in conjunction with the 2008- 2009 global crisis. Figure 2-1 depicts average firm responses when asked which elements of the business environment represent the biggest obstacle faced by the establishment. Despite the objective improvement in access to credit documented below, it is noticeable that a majority of firms are obstructed by access to finance in their business (26 percent), a likely effect of the more restrictive credit conditions in the aftermath of the 2008-2009 global economic and financial crisis. Tax rates (18 percent) and political instability (18 percent) rank second and third, while other important factors are informal competition and an inadequately educated workforce (15 and 9 percent respectively). Moreover, firms‘ opinion has shifted since 2005, where tax rates were identified as the main obstacle to operations and access to finance was depicted as the second most relevant.10 Scrutinizing the data by firm size, access to finance as an obstacle becomes even more evident. 29 percent of SMEs consider finance to be the chief obstacle to business. As will be discussed later in this Chapter, the econometric analysis has shown that many positive effects on productivity, as well as other key performance measures, derive from a sound financial system and wide access to finance for firms. It is therefore essential to recognize the negative development that the survey results present when addressing access to finance in relation to other key variables.

Figure 2-1: Major obstacle levels for firms in Turkey

30% 26%29% All firms SMEs 18% 20% 18% 18% 17% 15% 14% 9% 8% 10% 4% 3% 3% 2% 2% 2% 2% 2% 2% 1% 0%

Source: Turkey ES 2008

10 Due to methodology differences in 2008 and 2005, a comparison over time is only possible for the two largest obstacles, as perceived by Turkish firms.

15 Productivity

2.4. In 2008 the investment climate continues to be strongly associated with productivity. Total factor productivity (TFP) is an important driver of aggregate GDP growth in Turkey (World Bank, 2007). Several factors may influence the evolution of TFP, in other words the level of efficiency and technological content of production. Amongst these, the investment climate – related to the regulatory environment, the availability of skilled labor, the system of incentives to develop new products or experiment with novel production techniques, the ease of access to external finance – plays a critical role. This is confirmed by analysis of the 2008 enterprise survey of Turkish manufacturing enterprises. Based on the concept of demeaned TFP, defined as the share of aggregate log-TFP associated exclusively with the investment climate, and normalizing aggregate log-TFP to be 100, an estimated 31.4 percent of TFP is associated with the various aspects of the investment climate and the remaining 68.6 percent with other factors.11

2.5. The productivity of Turkish manufacturing is, overall, negatively influenced by the investment climate. Figure 2-2 compares the Olley and Pakes (1996) demeaned decomposition of Turkey‘s TFP with those of other countries, including Turkey in 2005, for which similar estimates are available. Both in 2008 and in 2005, the aggregate log-TFP of Turkish manufacturing industry is, overall, negatively influenced by the investment climate. This does not imply that Turkey is less productive than other countries but that investment climate conditions, on balance, are less conducive to the efficient use of resources. In other words, negative investment climate factors tend to dominate over positive ones, indicating that the investment climate available for doing business is preventing the economy from using resources as efficiently as it could.

Figure 2-2: The IC and TFP: Cross-country comparison

2 (Dem.) Aggregate log-TFP (Dem.) Average log-TFP (Dem.) Allocative efficiency - logs 1.50

1.6 1.52

1.00 0.77

TFP 1.2

0.70

-

0.63

0.60

0.50 0.50 0.50 0.50

0.8 0.37

0.27

0.12

0.10

0.07

0.03 0.03 0.02 0.4 0.01 0

-0.4

Demean log Demean

0.00 0.01

-0.8 -

0.37

0.40 -

-1.2 -

0.55

0.56

0.58

-

-

-

0.69

0.70

0.70

-

- -

Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details

2.6. The impact of the investment climate is asymmetrical with the productivity of larger firms positively associated with current investment climate conditions. The Olley and Pakes method allows decomposing aggregate TFP into an average component and an allocative efficiency component. The former reflects the productivity of the average firm, while the latter provides a measure of the efficiency with which resources are distributed among producers. Both in 2008 and 2005, the Turkish investment climate affects aggregate productivity mostly by negatively affecting the average firm. However, in 2008 the effect of the investment climate on the allocative efficiency component is positive, partially compensating the increased negative effect on the average firm. In other words, the positive effects tend to be concentrated in – conceivably larger – high market share firms and the negative effects in low market share ones. As discussed in Chapter 3, the need to increase average efficiency emphasizes the importance of policy and institutional mechanisms that stimulate the expansion of small- and medium- sized enterprises.

11 See Annex 2-A.

16 2.7. Improved allocation of resources appears as the main driver of increased aggregate productivity since 2005. Aggregate log-TFP of the manufacturing sector grew from 2.22 to 2.42 between 2005 and 2008 (Figure 2-3).12 The increase is driven proportionately more by improved allocative efficiency – from 0.29 to 0.44 – while average productivity has barely changed. This may be a sign of improved allocation of resources in Turkish manufacturing. The Black Sea-Eastern region is the most productive only partly because the average firm of this region is slightly more efficient – 9 percentage points more than the average firm of South and Central Anatolia regions, 13 percent more than in Marmara and 25 percent over the level of the average firm of Aegean. Most of the greater productivity derives from a substantially larger allocative efficiency effect, implying that resources tend to be concentrated in firms with high productivity. This efficiency effect is twice as large as in Central Anatolia, and more than three times the value in the other regions. Large variations also appear when examining sectors, with non-metallic mineral products, other manufacturing and garments presenting higher aggregate log-TFP. In all three cases, it is the allocative efficiency component that is driving higher aggregate productivity. The basic metals sector, on its part, stands out because its performance is driven by high average firm productivity.

Figure 2-3: TFP in Turkey

Aggregate log-productivity Average log-productivity Allocative efficiency - logs

3.6 3.42

3.2 2.83

2.73

2.67

2.65

2.50 2.47

2.8 2.45

2.42

2.39

2.26

2.26

2.22

2.22

2.13

2.11

2.09

2.07

2.07 2.04

2.4 2.04

2.03

1.98

1.98

1.97

1.93

1.93

1.89

1.87

1.81 1.79

2.0 1.78 TFP

- 1.6

1.19

log 0.88

1.2 0.85

0.71 0.60

0.8 0.44 0.43

0.41

0.38

0.30

0.29

0.29

0.24 0.24

0.4 0.13 0.0

-0.4

0.12 -

Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details

2.8. Investment climate reforms can positively affect both more efficient allocation of resources and increased average productivity. The importance of the allocative efficiency effect in Turkey indicates that the productivity improvements of the past few years have been driven by better allocation of resources towards the most productive firms (Figure 2-3). This emphasizes the persistence of a large gap between low and high productivity establishments within and across sectors and regions. Given the large negative association between current investment climate conditions and productivity (Figure 2-2), continued momentum in investment climate reforms will facilitate the reallocation of resources towards more productive firms (increasing allocative efficiency), while, at the same time, closing the gap between more and less productive firms by increasing the productivity of the average firm.

2.9. The regulatory environment is the largest relative contributor to productivity. Table 2-3 presents the relative contributions of various investment climate areas to demeaned aggregate productivity (the effect that can be associated with the investment climate). In order to avoid cancellations from positive and negative effects, the individual contributions are considered in absolute value. As in 2005,

12 Due to a number of methodological differences described in Annex 2-A, it is difficult to establish a direct equivalency between the demeaned aggregate log-TFP in 2005 and 2008. However, the computed results provide a reasonable approximation.

17 the regulatory environment remains the area of the investment climate with the largest relative contribution to productivity (Table 2-3).13 This motivates the closer examination, in Chapter 5, of the mechanisms underlying the formation and enforcement of regulations affecting business activity. Other relevant investment climate areas include infrastructure bottlenecks; access to finance and corporate governance; availability of labor and skills of the workforce; and quality and innovation capacity.

Table 2-2: IC variables‟ relative contributions to TFP in 2008 and 2005 (percent) Contributions to TFP Variable groups (percent) 2008 2005

Regulatory environment 46.82 72.97 Labor and skills 5.06 6.95 Quality and innovation 3.21 4.74 Finance and corporate governance 9.72 10.60 Infrastructure 11.28 4.08 Other control variables 23.91 0.68 Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details

2.10. Analysis of individual effects shows that most statistically significant variables have a negative effect on TFP. Figure 2-4 reports the percentage contribution to aggregate and average log-TFP of individual variables. Among regulatory variables, days to clear customs to import has a large negative contribution to average log-TFP (-73.0 percent). Since the contribution to aggregate log-TFP is almost identical, the negative effect on productivity is evenly distributed among firms. The number of compulsory certificates has a contribution to average log-productivity of -52.6 percent. The negative effect is, in this case, amplified by the allocation effect as the contribution of the variable to aggregate log-TFP is -60.4 percent, indicating that the negative effect is concentrated in high market share firms. Firms with larger market shares, on average, are required to obtain more compulsory certificates, and therefore the negative effect of this variable affects firms using a larger proportion of resources of the economy. Number of tax inspections accounts for -30.4 percent of aggregate log-TFP. The main contributor is the average effect with -28.1 percent, whereas the allocation effect contributes only -2.3 percent. Therefore, the problem is slightly concentrated in high market share firms, to some extent amplifying the negative average effect. Informal competition has a contribution to aggregate log-TFP of - 13.8 percent. In this case, the main contribution derives from average log-TFP (-13.1 percent), whereas the allocation effect accounts for 1.3 percent. That is, the negative effect is somewhat mitigated because the problem tends to affect low share of sales firms. The percentage of female workers in staff – an indication of labor intensive activities such as garments or textiles – is negatively associated with aggregate log-TFP (-14.2 percent). The positive allocation effect is less important as it accounts for only 0.5 percent of the total contribution to aggregate log-TFP. Exporting is positively related with TFP. Out of the total 13.6 percent contribution to aggregate log-TFP, 11.5 percent is due to the average effect and only 2.1 percent to allocation, indicating that the positive effect tends to be concentrated on high market share firms. Age of the firm, which is probably capturing effects such as the vintage of the firms or learning by doing externalities on TFP, has a clearly positive association with aggregate log-TFP, 78.4 percent. Furthermore, the contribution to average log-TFP and to the allocative efficiency are respectively 68.7 and 9.7 percent, implying that firm age affects more firms with high market shares.

13 Due to changes in the sampling methodology, analysis based on comparison of 2005 and 2008 results, can only provide an approximation of the effects of changes in investment climate conditions.

18 Figure 2-4: IC percentage contributions to aggregate and average productivity Infrastructure Regulatory environment Finance and corporate Quality and Labor Other control variables governance innovation and skills 100

% Contr. To (Dem.) Aggregate log-TFP 78.4

80 68.7 % Contr. To (Dem.) Average log-TFP 60

40

16.5

15.3

13.6 11.5

20 8.2

7.4

7.3

7.1

6.3 6.2

6.2

5.6

5.5

5.0

5.0

4.5

4.0 1.7

0

% Contribution %

0.4

0.4

0.6

0.7

1.0

2.1 2.1

2.1

2.3

-

-

-

-

-

4.2

4.7

- -

-

-

5.1 5.2

-20 5.4

6.7

-

-

-

7.9

-

-

-

-

13.8

14.2

14.7

15.1

-

- -

-40 -

28.1

30.4

-

-

38.0 40.7

-60 -

-

52.6 -

-80 60.4

-

72.1

73.0 - -100 - 1.1 1.2 1.3 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 3.1 3.2 3.3 3.4 3.5 4.1 4.2 5.1 5.2 6.1 6.2 6.3 6.4

1. Infrastructure 2.5 Compulsory certificates (days) 4. Quality and innovation 1.1 Duration of power outages 2.6 Compulsory certificates (number) 4.1 R&D (dummy) 1.2 Shipment losses, exports (interaction with 2.7 Informal competition (dummy) 4.2 Outsourcing (dummy) dummy for exporter) 2.8 Clearing customs for import (dummy) 1.3 Shipment losses, domestic 5. Labor and skills 3. Finance and corporate governance 5.1 Staff - female workers 2. Regulatory environment 3.1 Sales paid before delivery 5.2 Staff - university education 2.1 Payments for power supply (dummy) 3.2 New fixed assets finance - internal funds 2.2 Payments for government contract (dummy) 3.3 New fixed assets finance - state-owned 6. Other control variables 2.3 Security expenses banks 6.1 Share of exports 2.4 Tax inspections (number) 3.4 Largest shareholder 6.2 Decreased sales (dummy) 3.5 Subsidies (dummy) 6.3 Decreased prices (dummy) 6.4 Age

Notes: Contributions computed according to equation (9) in Annex 2-A. The IC contributions to the allocative efficiency term are not included in the table. Nonetheless, they can be obtained implicitly as the IC contribution to aggregate log-TFP minus the IC contribution to average log-TFP. The productivity measure used is the restricted Solow residual. 2% of upper and lower bounds of productivity are excluded of the calculations. Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details.

Other Performance Measures: Employment, Exports, FDI

2.11. The investment climate is also associated with employment, exporting and foreign direct investment. As shown in Figure 2-5, investment climate variables account for a sizeable share of the sample mean of employment, as well as of the probability of exporting and receiving FDI. The importance of the investment climate is greatest for employment (65.9 percent), it accounts for almost half of the probability of exporting, while is has a more limited contribution to the probability of receiving FDI. TFP, on its part, accounts for a large share of the probability of exporting (37.1 percent), while it plays a more limited role for employment and FDI.

19 Figure 2-5: Share of Investment Climate effects on the sample mean of employment and probability of exporting and receiving FDI A. Employment B. Exports C. FDI TFP: Wages: Other IC: 7.2% 16.6% variabl TFP: 10.1% es: 37.1% TFP: 15.6% 3.5%

IC: Other Other IC : 47.4% variable variable 65.9% s: 82.7% s: 14%

Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details

2.12. The relative importance of regulatory variables for employment, exports and FDI appears to have decreased since 2005, while the relevance of innovation and skills for exporting has increased. The total relative weights of the investment climate block of Figure 2-5 can be decomposed into the contributions of each group of investment climate variables. The absolute percentage contributions by blocks of variables (summing up to 100%) of Table 2-4 illustrate the variation between 2005 and 2008 of the estimated effects on average log-employment and on the probability of exporting and of receiving FDI inflows. In the case of employment, regulatory variables lose relative weight (from 46 to 4.2 percent), with, especially, labor and skills, real wages, and TFP gaining in relative importance for the sample mean of employment. A similar apparent decline in the relative importance of the regulatory environment can be observed in the case of exporting and FDI. Such a large fall in the bearing of regulatory variables may be in part due to the methodological differences between the 2005 and 2008 surveys. It is also worth noticing the large increase of the relative contribution of TFP to the probability of exporting and of receiving FDI, indicating that greater productive efficiency has become more crucial for the international integration of the Turkish manufacturing sector, as shown by the ability to compete in international markets and to attract foreign capital. The increase in the relative weight for exporting of variables related to innovation and quality (from 5 to 21 percent) is also worth mentioning. As argued in Chapter 4, in order to achieve broad-based business sector development, it is crucial that the investment climate is propitious to knowledge transfer from more productive and innovative internationally oriented firms towards less productive and innovative local suppliers. Table 2-3: Investment climate variables‟ relative contributions to sample averages of employment, exporting and FDI in 2008 and 2005 (percent) Probability of Probability of Variable groups Employment exporting receiving FDI 2008 2005 2008 2005 2008 2005

Regulatory environment 4.19 36.27 8.45 41.04 4.83 26.74 Labor and skills 19.15 10.44 0.00 6.10 7.95 38.85 Quality and innovation 7.75 7.68 20.83 5.17 2.94 13.74 Finance and corporate governance 21.74 21.57 2.78 2.12 4.52 0.00 Infrastructure 2.00 0.00 8.76 10.35 5.43 0.00 Real wages 19.82 13.30 0.00 0.00 0.00 0.00 TFP 4.14 2.17 43.91 13.38 43.27 17.83 Other control variables 21.20 8.57 15.27 21.85 31.05 2.83 Source: Staff calculations based on Escribano et al. 2009. See Annex 2-A for details

2.13. In order to trace the evolution of the business environment over the past few years, the remainder of this Chapter takes stock of various aspects of the investment climate and their effects on business sector operation and performance. The areas examined include the regulatory environment; the availability of labor and the level of skills; innovation capacity and technology; as well as access to finance and corporate governance.

20 2.2 The Regulatory Environment 2.14. By improving the efficiency and transparency of its regulatory environment the Turkish economy can diminish the overall risk of informality and boost productivity and growth. Evidence from econometric analysis of enterprise survey data supports the notion that business sector performance benefits from well-designed and enforced rules and regulations. While a detailed treatment of the mechanisms underlying the formation and enforcement of regulations is provided in Chapter 5, this section compares the development of the regulatory environment in Turkey between 2005 and 2008. Following examination of the econometric evidence of the effects of the regulatory environment on firm performance, the section offers a description of recent progress in the areas of taxation, informality, and administrative procedures.14

Table 2-4: Summary of the effects of the regulatory environment in 2008 Dependent variable Explanatory variable Probability Probability Productivity Employment of of receiving exporting FDI Tax inspections (number) - Tax inspections (days) - Business inspections (number) - Compulsory certificates (days) - - Compulsory certificates (number) - + Compulsory certificates (payments) - Customs clearance for imports (days) - Payments for power supply (dummy) - Payments for government contracts - - Security expenses - + Informal competition (dummy) - Source: Staff calculations based on Escribano et al. 2009. See Tables in Annex 2-A for detailed results.

2.15. The productivity of Turkish industry is negatively associated with formal and informal aspects of the regulatory environment. In addition to hampering the day to day operation of the business sector, cumbersome and inefficiently enforced rules and regulations de facto create an uneven playing field among producers, for example by implicitly or explicitly favoring larger or more well- connected firms with greater capacity to handle bureaucratic requirements. This dampens the benefits of market competition by reducing the incentives for both incumbents and other firms to adopt best practice production techniques, ultimately resulting in lower overall productivity.15 Such a mechanism may be at work in Turkey, where firm-level productivity is negatively associated with a number of features of the regulatory environment (Table 2-5). Some of these include formal bureaucratic requirements, such as the number of inspections to which businesses are subjected, the number of compulsory certificates required and the time necessary to obtain them, as well as time consuming customs procedures for imports.16 Burdensome regulatory requirements constitute fertile breeding ground for the negative consequences of corruption, as it is exemplified by the negative association between productivity and informal payments to

14 To enable comparison between 2005 and 2008, only the manufacturing sector is measured in the econometric analysis, whereas descriptive data for 2008 also embrace firms in the services industries. Due to differences in the variables used in 2005 and 2008, some homogeneity is lost when comparing 2005 and 2008 results. Nonetheless, the comparison presents an opportunity to assess key differences in the way investment climate variables affect firms‘ productivity, employment, propensity to export and to receive FDI in the two time periods. For a description of econometric methods and for detailed regression results, see Annex 2-A, as well as the background paper for this report by Escribano et al. (2009). 15 See Conway, De Rosa, Nicoletti and Steiner (2006). 16 The average number of days needed to clear customs to import is significant both in 2005 and 2008. Moreover, the coefficients are virtually identical, -.175 in 2008 and -.171 in 2005. However the contribution to average log-TFP became more negative, from -51.6 to -73, indicating that now the average firm suffers more this problem. See Table 2-A-2.

21 obtain power supply or a contract with the government. Inefficient regulations also provide incentives for firms to remain partially informal. Firms that are subject to competition from informal establishments are, in turn, associated with lower levels of productivity.

2.16. The regulatory environment seems to have a differentiated effect depending on firm size. For instance, larger firms must obtain a greater number of compulsory certificates but they spend less time doing so compared to firms with fewer employees.

2.17. The ability to export and the attractiveness for foreign investors of Turkish firms are negatively associated with formal and informal regulatory requirements. For instance, export propensity is negatively associated with the number of business inspections and with the informal payments required to obtain compulsory certificates or a government contract. At the same time, the probability of receiving FDI is negative associated with the duration of tax inspections.

Taxation 2.18. Despite the fact that tax inspections are still negatively correlated with the productivity of Turkish firms, tax rates and tax administration are perceived as less burdensome than in 2005. Econometric analysis of enterprise survey data shows that the high number of tax inspections experienced by firms has a negative impact on productivity (Table 2-5). However, when examining firms‘ perceptions, from being the largest obstacle in 2005, the relative importance of tax rates has dropped, with 18 percent of firms in 2008 perceiving tax rates as the biggest obstacle to business Figure 2-6: Firms perceiving tax rates and tax administration operations (Figure 2-1 above). At the as a major or very severe obstacle, by country same time, the share of manufacturing 100% companies identifying tax rates as a Tax rates 81% 84% 75% major or very severe constraint to their 80% Tax administration 59% 59% 59% business has decreased from 81 percent in 60% 50% 52% 2005 to 50 percent in 2008 (Figure 2-6).17 41% 42% 40% 27% 25% 22% 19% 22% Tax administration, viewed in 2005 by 59 16% 15% percent of manufacturing firms as a major 20% 8% constraint has now dropped to 19 percent. 0% Worth mentioning is also that only a Chile Bulgaria Czech Turkey Turkey Poland Hungary Turkey Brazil 2006 2009 Rep. 2008 2008 2009 2009 2005 2009 fraction of Turkish firms (0.3 percent) 2009 (manuf) (all) (manuf) indicate that tax administration is the biggest obstacle to their business. Source: Enterprise Surveys

2.19. As depicted in Figure 2-7, the share of firms perceiving tax rates to be a major or very severe obstacle to operations varies across firm size, with small and medium enterprises being more bothered by tax rates (55 and 51 percent respectively) than large companies (41 percent). Tax rates are also much less of an obstacle in Black Sea – Eastern and Central Anatolia (42 and 45 percent) than in for instance the Aegean region (61 percent). Variations across industries can also be noticed, with particularly firms in the construction sector being bothered tax rates (86 percent).

17 Note that in Figure 2-1 above, firms are asked to specify the one element in the business environment that represents the biggest obstacle faced by the establishment, whereas Figures 2-6 and 2-7 present the answers of firms when asked to rate how much of an obstacle each individual business element is to operations.

22 Figure 2-7: Tax rates as an obstacle, by firm size region and industry 90% 86% 80% 70% 61% 59% 55% 56% 59% 60% 51% 51% 53% 45% 48% 45% 46% 46% 48% 49% 50% 41% 42% 39% 42% 40% 34% 30% 20% 10% 0%

Source: Turkey ES

2.20. Since 2006 Turkey has embarked on a comprehensive overhaul of the taxation system. The improvement in the perceptions of enterprises between 2005 and 2008 can, at least in part, be ascribed to the tax reforms that have been introduced since Figure 2-8: Corporate tax rates 2005-2009, Turkey and comparator countries, percent 2006 (Box 2-2). Notable among these are the 60 Hungary introduction of a new 55 corporate tax code, Czech Rep. reduction of corporate 50 Romania income tax from 30 to 20 45 percent, as well as lower Turkey taxation on interest. 40 Poland These achievements 35 were recognized in the Bulgaria World Bank 2007 Doing 30 2005 2006 2007 2008 2009 Business Report, where Turkey resulted as one of Source: Doing Business 2010 the top reformers in 2006-2007 thanks to improvements in the ―Paying Taxes‖ indicator. The reduction in corporate tax rates has been significant and has improved Turkey‘s standing relative to a number of European neighbors (Figure 2-8).

Box 2-2: Tax reforms The Turkish government has recognized the importance of lower tax rates for a better business environment through the introduction of a new corporate tax code in 2006. The corporate tax rate was reduced by ten points to 20 percent, broadening the tax base and weakening the incentives for firms to work in the informal sector, and thus boosting overall growth and productivity. Moreover, the tax on interest was lowered from 18 to 15 percent.

As a part of the Tax Offices Automation Project (VEDOP), an online tax filing system was also introduced, reducing the time needed for complying with tax regulations by 31 hours as well as allowing companies to submit tax reforms and payments through the Internet Tax Office. This, along with other simplifications of the tax mechanism, initiated through Turkey‘s ―Ninth Development Plan 2007-2013‖ will reduce firms‘ compliance costs as well as provide a more efficient tax collection system, connecting a country-wide

23 network of 599 tax, inspector and regional finance offices. The nationwide communications network has been implemented by the Ministry of Finance in order to streamline administrative procedures and provide citizens with a secured and efficient online system to file taxes.

Additional tax reforms simplifying the activities of Turkish firms include amendments to the R&D law and allow for deduction of all R&D expenditures from the taxable income, with additional tax benefits after capitalization of profits. Moreover, firms with more than 500 R&D personnel and a 50 percent increase in R&D expenditure, compared to the year prior can deduct a further 50 percent of their revenues for tax purposes.

Amendments to the Personal and Corporate Income Tax Codes have also been made in order to cover anti- abuse legislation, restricting remittances to tax havens. Furthermore, business telecommunication became more cost effective in March 2009, as the Special Communication Tax was reduced to 5 percent from a previous 15 percent for land lines and 25 percent for mobile networks.

2.21. The tax wedge on labor – the gap between total labor costs and the employee‟s take home pay – has been reduced since 2005. An important disincentive to increase formal employment is the tax wedge, representing the non-wage cost of labor. This has also declined in the past three years from 42 percent in 2005 to 38 percent in 2008 for an average income earner, spouse and two children. When making an international comparison based on an average income earner with a non-working spouse and two children, Turkey has seen improvements, from previously having the highest taxation among the OECD countries (OECD 2005). Nevertheless, the tax wedge is still well above the OECD average of 27 percent. Comparing the tax wedge for a single worker with 67 percent of average earnings, the gap to the OECD average of 34 percent becomes considerably smaller.

Figure 2-9: Tax wedge, 2008 a) average income earner, spouse and two children, b) single with no children earning 67 percent of average 2008 wage, 2008

Mexico 15.1% Mexico 10.9% Korea 18.1% Korea 17.4% Czech Republic 20.6% Portugal 32.9% Slovak Republic 25.4% OECD 33.5% Portugal 27.2% Spain 33.8% OECD 27.3% Slovak Republic 36.1% Spain 31.8% Greece 37.6% Poland 33.7% Turkey 37.6% Turkey 38.5% Poland 38.7% Turkey (2005) 42.0% Czech Republic 40.0% Greece 42.7% Turkey (2005) 42.0% Hungary 43.9% Hungary 46.7%

0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% 50%

Source: OECD

Informality

2.22. Unfair competition from informal firms is highly obstructive of sound business development. Strongly connected to the topic of tax payments, and one of the critical issues examined in the 2007 Investment Climate Assessment, is the negative impact that the informal business sector has on productivity. Despite lower tax rates introduced with recent reforms, which could contribute to reducing the competitive advantage of informal firms, analysis of the Enterprise Survey shows that productivity is still negatively associated with competition from informal businesses (Table 2-5).

2.23. In Turkey the underreporting of revenues and wages and the non-registration of workers with the social security system are the most important forms of informality. Informality is usually

24 defined as legal economic activity taking place below the radar of government, covering categories that include unregistered firms, registered firms engaging in unrecorded economic activity, and the self- employed. Among the many categories, unrecorded sales (to evade taxes) and underreported/unrecorded employment (to avoid labor regulations and contributions) are the most common in Turkey. Although the headline measure of informality has been falling, this is almost entirely explained by migration out of agriculture into more formal sectors. The headline measure of informality fell from 53 percent to 44 percent between 2004 and 2008. 18 However, this aggregate decline hides important patterns. Most of the decline between 2001 and 2006 is explained by migration of the workforce out of agricultural employment (where nearly all workers are informal) to manufacturing and services, mainly in urban areas (where informality rates are below 20 percent for wage earners). Moreover, during this period of rapid expansion of the economy, urban and non-agricultural informality increased (from 29 percent to 34 percent for non-agricultural employment).

Figure 2-10: Share of firms facing informal competition By country By size

60% 54% 55% 56% 52% 52% 49% 55% 50% 55% 43% 44% 40% 35% 54% 33% 53% 30% 52% 52% 52% 20% 51% 10% 50%

0% 49% Poland Romania Czech Turkey Hungary Turkey Turkey Bulgaria Brazil Small Medium Large 2009 2009 Rep. 2005 2009 2008 2008 (all) 2009 2009 2009 (manuf) (manuf)

By region and industry 70% 62% 64% 64% 58% 59% 58% 58% 60% 54% 49% 51% 51% 45% 46% 50% 50% 44% 41% 43% 40% 30% 20% 10% 0%

Source: Turkey ES 2008

18 TURKSTAT measures informality as the proportion of workers unregistered for social security. Issues related to informality in Turkey are addressed in a forthcoming World Bank Economic Memorandum, Turkey Informality: Causes, Consequences, Polices.

25 2.24. The share of surveyed firms facing informal competition has increased substantially. 52 percent of all firms interviewed in the 2008 survey indicated that they compete with informal companies, a negative development since 2005, where the corresponding number was 10 points lower (Figure 2-10). Moreover, 32 percent of all firms claim practices of competitors in the informal sector to be a major or very severe obstacle. 15 percent of all firms interviewed agree that it is the most serious problem they are faced with in their business, positioning informal sector practice as the fourth largest obstacle. Figure 2- 10 also shows that larger firms experience more competition from the informal sector than SMEs (55 percent and 52 percent respectively), whereas companies in the South and in the Marmara region face considerably less informal competition than Black Sea – Eastern, Central Anatolia and Aegean Additionally, the survey data shows that informal competition is highest in the food and chemicals sectors.

Box 2-3: The Government‟s struggle against informality In 2006 the Government organized and initiated the two-year ―Struggle against Unregistered Employment (KADIM) Project,‖ with the main objective to reduce informal employment. The target was to prevent 100,000 illegal foreign workers and replace these jobs with a formal workforce, thus increasing domestic employment. Means to be applied include information and awareness activities, bureaucracy reduction, efficient inspection, audit and employee registration controls, as well as labor cost reductions. Between October 2006 and December 2007, more than 730,000 employees and nearly 200,000 workplaces had been audited. Moreover, the Turkish Social Security Institution increased the number of employees covered by 63 percent. The final results of the KADIM project have not yet been made public.

The latest initiative against informality was introduced in June 2008 through the Revenue Administration in cooperation with several key institutions. The development of the ―2008-2010 Action Plan of Strategy for Fight against the Informal Economy‖ covers 105 actions over a three-year period, with three key objectives: promoting formal activities; strengthening audit capacities; and procuring and strengthening organizational and societal consensus.

2.25. When considering that tax evasion is one of the main drivers of the informal sector, the Turkish government should continue its efforts in tax regulation reforms. Other recently implemented as well as planned actions are listed in Box 2-3.

Administrative procedures

2.26. Recent successful measures to streamline compliance with administrative procedures need to be expanded. There is today no clear framework for streamlining administrative procedures for businesses. Operating licenses are issued by various Ministries, each responsible for different business areas. For instance, the Ministry of Public Works is responsible for construction permits, in terms of their design and their applicability. Another example is the Ministry of Environment who is responsible for environmental licenses. However, there is at this point no single overall strategy for stepping up and facilitating licensing procedures. Both the Ministry of Agriculture and the Ministry of Health, for instance, are involved in the processing of licenses in the food industry, with the result of much overlapping and confusion.

2.27. Notwithstanding successful earlier actions towards regulatory reform, many concerns still remain to be managed. The last major governmental initiative to limit license and permit requirements was implemented in 2005, through the regulation on ―Opening a Business Place and Work Licenses‖ (CEDPL2), reducing the number of documents needed to obtain a license, as can be noticed in the survey 2005-2008 comparison. Nonetheless, the weaknesses in the Turkish regulatory system for licensing and permits are many and much remains to be done. As discussed in greater detail in Chapter 5, some major

26 difficulties include: insufficient coordination among agencies responsible for application processing; inadequately qualified municipalities in dealing with and applying secondary legislation; major gaps in sector-based strategies; and poorly constructed digitalized systems for application approvals from citizens and corporations. In addition, involved Ministries do not have access to any standardized procedure guidelines for implementing the regulatory framework.

2.28. The perceived „time tax,‟ management time spent dealing with government regulation, appears to have increased since 2005. Management personnel in Turkish businesses declare to spend on average 27 percent of their working time dealing with bureaucracy, a steep increase from the 9 percent average in 2005. The time tax is largest for medium and large enterprises (32 percent and 34 percent respectively), whereas managers in small firms declare to spend 23 percent of their time with regulations.

2.29. The establishment of Development Agencies (DA) could offer an opportunity to facilitate licensing procedures by providing an interface for businesses at the local level, provided that risks of capture are minimized. The Government, through the State Planning Organization, has established Development Agencies, with the objective of 26 Agencies covering the entire country (Box 2-4). Two Development Agencies are fully operational today – in the Çukurova region (Adana and Mersin) and Izmir, along with their accompanying Investment Support Offices. The objective is for agencies to be rolled out across the country building on the experience of Çukurova and Izmir, to serve as ―one-stop- shops‖ for companies in financial and technical matters and ―execute programs that will support development, competitiveness and local initiatives‖ (SPO 2009). In connection with the Development Agencies, and building on the experience of Izmir and Adana, Investment Support Offices are planned to gradually be established across the country with main focus on aiding investors and assisting them in licensing and registration procedures. DAs are well positioned to be an interface between business and government provided that they preserve a light organizational structure and remain at arm‘s length from the government, with the objective of minimizing the risk of capture on the part of local interests. Additionally, an essential condition for DAs to be able to perform their role is the existence of internal Government regulations recognizing their formal role vis-à-vis central and local government.

Box 2-4: Development Agencies (DAs) A Regional Development Agency – or Development Agency (DA) – can be defined as a regionally based, publicly financed institution outside the mainstream of central and local government administration designed to promote indigenous economic development through an integrated use of policy instruments. DAs have been operating in Western European countries since 1950s and 1960s in order to decrease the negative effects of the Second World War and keeping up with rapid technological developments in the world. According to European Association of Development Agencies (EURADA), ‗a DA is an operational structure that identifies sectoral or overall development problems, chooses a range of opportunities or methodologies for their solution and promotes projects which can maximize the solutions to the problems‘.

In Turkey, regional development policies have been developed in the quest to eliminate regional disparities, to accelerate local and regional economic development and to enable sustainable development. However, those policies developed towards particularly less developed regions have not been very successful, mainly due to the lack of institutional capacity at the local/regional level, i.e. the lack of effective institutional structures at the local level as well as that of sufficient financial resources. There had been a number of initiatives of regional development projects prior to Turkey‘s comprehensive adjustment process towards EU candidacy such as East Marmara Regional Planning Project (1963), Zonguldak Regional Planning Project (1963-64), Antalya Project (1960-65), Aegean Regional Development Project (1963-69), Çukurova Region Planning Project (1962), Keban Project (1964) all aiming at developing policies, plans and proposals for the problems of different regions.

Since Turkey gained European Union (EU) candidate status with the decision of Helsinki Summit in 1999, the momentum of legal and institutional reforms has increased unprecedentedly. Accession negotiations between the EU

27 and Turkey that started in 2005 and the issue of Turkey‘s compliance with the ―acquis communautaire‖ became even more significant for the legal and institutional reform processes. In its National Plan for the Adoption of the Acquis (NPAA), Turkey has committed itself to legal and institutional changes that will contribute to its adjustment to EU regional policy. Accordingly, in order to facilitate the development of structures of local/regional governance, ‗The Law on the Establishment, Coordination and Duties of Development Agencies‘, was ratified on January 25, 2006 (Law No. 5449, 2006) to facilitate and regulate the establishment of Development Agencies-DAs.

The Law establishes that the State Planning Organization (SPO) will be responsible for the coordination of DAs at national level. Agencies in all the 26 NUTS II regions were established. Their main purpose is to accelerate regional development, promote cooperation between the public and private sectors and contribute to the reduction of inter- regional disparities. The DAs will be funded in part from transfers from the national budget and in part by the special provincial administrations (local authorities) and municipalities. The DAs will also be expected to generate operating revenues, although this is not realistic in the poorer regions.

Two pilot DAs in Izmir and the Çukurova region started operating in 2008 and other agencies are now established in all provisional NUTS II-type regions. A total budget of nearly €125 million has been earmarked for the development agencies in the 2009 national budget. Relevant local and regional stakeholders are involved in establishing the budgets of individual DAs, but not in selecting the provinces to host the DAs.

At present DAs which are aiming to increase regions‘ competitive power and decrease regional discrepancies, focusing especially on high-technology, innovation and communication are still in their infancy and face criticisms and legal challenge on the grounds that they will weaken central government.

The organizational structure of the DAs includes a Developmental Board, a Management Board, a General Secretariat and Investment Support Offices for business support. The Developmental Board members are representatives of various public and private organizations, NGOs and universities within the region. This board functions as an advisory board. The Management Board is formed by governors, mayors of the metropolitan municipalities, the chairmen of the Chambers of Commerce and Industry and three representatives from NGOs or from the private sector. General Secretariat is the executive body in DAs. Moreover, Investment Support Offices- ISOs are located in each province of the NUTS II regions.

2.30. Looking ahead, DAs could more fully perform a number of useful functions. First, as already intended by the Government, they could act as “one-window” shops. Without radical reform of competences for the issuance of licenses and permits they could perform a useful facilitator role between issuing agencies and firms. Second, DAs could act as an information point for businesses, in close cooperation with business associations with a local presence, such as TOBB, as well as with Government agencies, such as KOSGEB and TÜBITAK. The objective would be to rationalize support initiatives – especially for SMEs who normally face high information costs – available at the local level. Third, the FDI promotion function via Investment Support Office should be carried out in close coordination with ISPAT, the national FDI promotion agency.

STARTING A BUSINESS

2.31. The Government has made progress towards facilitating the process of business registration. According to Doing Business 2010, recent Turkish reforms have reduced the time it takes to register a business from 13 required steps in 2004, down to six steps and six days required in 2009. This development is supported by the OECD Product Market Regulation (PMR) Indicators, which show a reduction in number of mandatory registration procedures, from 29 procedures in 2003 to 11 procedures in 2008, as well as a cut in days needed to complete registration, from 6 days to 2 days. Although Turkey has experienced an improvement since 2006 as well, the positive six-year development can mainly be attributed to the 2003 legal reforms, with a modern investment law and new legislation covering registration procedures (World Bank 2007).

28

2.32. The 2007 Investment Climate Assessment presented criticism of the lack of transparency in registration procedures as well as costs for Turkish firms, making among others recommendations for better use of information technologies. The Government has in fact initiated e-Judicial Registrations and Online Company Registration procedures, with a draft act currently being evaluated by the Prime Ministry and expected to be adopted in the National Assembly in early 2010 (Prime Ministry e- Legislation Seminar, September 2009).

Box 2-5: Indicators of product market regulation (PMR) The OECD PMR is a comprehensive and internationally-comparable set of indicators that measure the degree to which policies promote or inhibit competition in areas of the product market where competition is viable. The indicators measure the economy-wide regulatory and market environments in OECD countries in 1998, 2003 and 2008 and depict the strictness of the regulatory environment on a scale 0 to 6, with higher numbers indicating more restrictive policies towards competition. The PMR system has been revised and updated as of 2008 to better integrate sectoral information and cover additional key regulatory issues, such as governance and treatment of foreign entities. The measured variables in the areas of start-up and licensing systems are based on questions on how many days, procedures and the cost it takes to register a business as well as the presence or absence of ―one stop shops‖ for licensing information. As can be seen in Figure a. depicting Turkey‘s ranking in one of the measured variables, administrative burdens on start-ups, gradual improvements have been made, moving from 3.1 index points in 1998 to 2.6 index points in 2008. Turkey has however still ways to go to the OECD average of 1.53 points. The conditions for Turkish enterprises are thus more severe when analyzing indicators measuring license and permit systems, where the results have consistently been meager. For three consecutive measured periods, Turkey has performed poorly and remains to have a 6.0 index point ranking in the PMR indicators.

Figure 2-11: Development of product market regulation since 1998, Turkey and comparator countries Index scale of 0-6 from least to most restrictive a. Administrative burdens on start-ups b. License and permit systems 6 6 6 6 6 6

6 6

3.90 3.90

3.84 4 4 4 4

3.59

3.23

3.22 3.09

4 3.06 4

2.85

2.83

2.74

2.61

2.51

2.35

2.33

2.31 2.20

2.09 2 2 2

1.73 1.69 2 1.57 2 0 0 0 0 0 0 0 0 0 0

1998 2003 2008 1998 2003 2008 Source: OECD, Product Market Regulation Database Methodology derived from Wölfl et al. 2009

29 LICENSING

2.33. The time needed to obtain an Figure 2-12: Days to obtain operating license, country operating license19 remains high in Turkey comparison but tends to be shorter among firms in the 70 66.0 67.7 57.4 62.0 services sectors. Manufacturing firms 60 interviewed in the 2008 enterprise survey 50 35.6 36.0 report that the time it takes to acquire an 40 operating license has decreased from 66 days 30 20.8 23.7 in 2005 to 62 days in 2008. Albeit the 20 10 improvement, this is still relatively higher 0 than in comparison countries. However, as Bulgaria Romania Hungary Turkey Russia Turkey Turkey Chile seen in Figure 2-12, when considering the 2009 2009 2009 2008 (all) 2009 2008 2005 2006 (manuf) (manuf) total sample that includes the services sector, the number of days to obtain the license Source: Enterprise Surveys 2008 drops to 36 days, which internationally compared, is more favorable.

2.34. There are signs of variations among regions, as well as by size and ownership form of firms. Figure 2-13 shows that obtaining an operating license is nearly three times as time consuming in the South than in Central Anatolia. When comparing operating license procedures by city, the time in Istanbul has improved but is among the longest in the country. Furthermore, medium firms need more time than small and large ones to obtain necessary licenses. The same goes for domestic firms that have to wait 36 days for their operating license, versus 21 days for foreign-owned firms. By sector, enterprises in chemicals, textiles, and garments have to wait the longest to obtain their licenses, while the metals and machinery, as well as the service sector have a waiting time far below the average.

Figure 2-13: Days to obtain operating license, by region, city, size, ownership and exporting status

90 78.4 80 70 65.6 60 50.4 46.1 50 42.1 38.7 36.3 37.8 36.3 40 32.1 33.1 33.6 32.3 27.4 30 22.4 17.9 18.8 20.7 20 15.3 10 0

Source: Turkey ES 2008

19 An operating license is defined as a statement order by a government authority or other authorized legal entity that authorizes the business activity to be carried out in the country. An operating license is usually required as part of the business startup process and it verifies that the establishment possesses the operating business licensing requirement. The business activities which can be covered by an operating license may differ according to the country specific regulations.

30 2.35. Firms claim to need longer today to obtain import licenses20 than in 2005, but Turkey compares well internationally. The time has increased from 12 days to 18 days for the manufacturing sector and the number is even higher Figure 2-14: Days to obtain import license, by country when services industries are included (Figure 2-14). Medium-sized firms have 50 43.1 to wait on average 32 days, in comparison 40 with only 8 days for large enterprises. 32.7 29.6 Figure 2-15 also shows that there are large 30 variations among cities, with companies 21.2 17.7 20 16.6 in Denizli having to wait considerably 11.2 12.3 8.1 longer to obtain an import license than for 10 instance firms in Izmir (42 and 11 days 0 respectively. Furthermore, food and Bulgaria Poland Turkey Chile Turkey Turkey Russia Czech Brazil chemicals are some of the sectors that 2009 2009 2005 2006 2008 2008 2009 Rep. 2009 have to wait longer to obtain an import (manuf) (manuf) (all) 2009 license, while garments and non-metallic material experience shorter waiting Source: Enterprise Surveys 2008 periods.

Figure 2-15: Days to obtain import license, by size, ownership, exporting, region, city and industry

50 42.0 40 34.1 32.0 28.9 30 23.9 24.2 24.0 24.8 21.8 22.2 20.4 17.5 17.9 20 15.9 10.7 10.7 8.0 10 6.9

0

Source: Turkey ES 2008

2.36. Clearance of customs procedures has become less cumbersome than in 2005. According to survey data, firms claim that the time to clear customs for imports has increased from 7 days in 2005 to 10 days in 2008 for the Turkish manufacturing industry (Figure 2-16). However, when analyzing only panel firms (the firms interviewed in both 2005 and 2008) the survey data shows an improvement in customs clearance for imports, from 13 days to 8 days.21

20 An import license is defined as a required document issued by a government authority or other authorized legal entity that authorizes the importation of certain goods into its territory. 21 Considering the smaller sample of panel firms, some caution should be used when interpreting this data. Nonetheless, the comparison provides useful information when used in used together with other data sources, such as Doing Business.

31 Figure 2-16: Days to clear customs for imports

18 17 15 15 13 10 10 12 10 10 8 9 7 5 6 3 3 0 Hungary Bulgaria Turkey Turkey Poland Czech Rep. Turkey Turkey Turkey Brazil 2009 Chile 2006 2009 2009 2005 2008 (panel 2009 2009 2008 (all) 2008 2005 (panel (manuf) firms) (manuf) firms)

Source: Enterprise Surveys 2008

2.37. Doing Business data also present a more positive image, with the number of days to import decreasing from 25 in 2005 to 15 in 2009. In line with the findings of the EU Progress Report for 2010, Doing Business indicators also show that Turkey compares more favorably relative to other emerging economies (the average for the ECA region in 2009 is 18 days). It is worth mentioning that Turkey has initiated numerous recent actions to simplify firms‘ trading across borders. A Customs General Communiqué (SPO 2009) was published in April 2008 in the Official Gazette, covering the simplified practices when dealing with import and export documents. Since October 2008, the e-transformation of customs offices has been implemented through the Computerized Customs Activity (BILGE) System, connecting 120 customs directorates and allowing over 95 percent of all trade transactions to be carried out electronically today (European Commission 2009). Additionally, build-operate-transfer agreements for renovation of border gates have been initiated with the private sector as well as better coordination of agencies and tasks involved in the customs clearing process (World Bank 2008).

ACQUIRING LAND

2.38. Land access is in general not perceived as a major problem by Turkish firms. Only 5 percent of enterprises consider access to land to be a major or severe obstacle to operations. This is not only a significant improvement from the 20 percent response in 2005, but it also places Turkey in a positive international position. 2.39. Firms‟ perception of land access Figure 2-17: Access to land as an obstacle, by city varies by size and sector. The differences by region depict a mean of 3 percent in Black 9% 8% Sea - Eastern versus 7 percent in the South, 6% 6% 7% 6% 6% with a similar pattern depicted by city. As 5% 6% 5% Figure 2-17 depicts, in Kocaeli and Adana 4% the ability to acquire land is not seen as a 3% 3% major issue by many firms (3-4 percent), while 8 percent of corporations in Malatya do struggle with land access. A small variation 0% by firm size can also be noticed, where small and large firms (6 percent) deem land access to be a problem to a larger extent than

Source: Turkey ES 2008 medium enterprises (Figure 2-18). An explanation for the higher share for firms with large number of employees may be that these naturally acquire larger areas for production. Figure 10

32 shows that the differences among industries are more noticeable. Access to land is perceived as more of a hurdle in the machinery and equipment as well as basic metals sectors, whereas garments companies are not disconcerted by this matter.

Figure 2-18: Access to land as an obstacle, by firms size, ownership, exporting status and industry

25% 22% 20% 17% 15% 11% 11% 10% 7% 6% 6% 5% 6% 5% 5% 6% 3% 4% 3% 4% 4% 5% 2% 0%

Source: Turkey ES 2008

2.40. Turkey has also made improvements in the cost and time invested in constructions of business premises. According to the Doing Business‘ measure of the time involved in building a warehouse Turkish firms have been able to shorten their site development time by 44 days between 2005 and 2009. However, even though Turkey is below the Europe and Central Asia average, the country‘s results are mediocre in an OECD comparison. Also, it still takes Turkish firms six times longer to construct a building than for Finland, the EU best practice economy. The fees involved are also 23 times higher than in Hungary, for example. Overall, Turkey is ranked as 133 for dealing with construction permits among 183 measured economies.

Figure 2-19: Days to obtain a construction-related permit, by country

160 139.1 142.8 140 118.1 122.5 120 104.2 104.9 100 87.0 87.9 80 58.4 58.5 60 41.9 43.0 47.1 47.5 40 20 0 Turkey Hungary Turkey Bulgaria Turkey Czech Istanbul Istanbul Russia Istanbul Romania Poland Brazil Chile 2008 (all) 2009 2008 2009 2005 Rep. 2008 (all) 2005 2009 2008 2009 2009 2009 2006 (manuf) (manuf) 2009 (manuf) (manuf)

Source: Enterprise Surveys

2.41. The ES data also has a similar measure, where interviewed firms are asked how many days it took to obtain a permit that was requested over the last two years. The survey results present a more positive picture, with an improvement since 2005 from 58 days to 47 days in 2008, which is a leading ranking relative to other middle income economies, such as Brazil, the Czech Republic, Poland and Romania. Interesting to see is that the time to obtain building permits varies significantly between Turkish cities, with Kayseri and Bursa showing the least average time. Meanwhile, firms in Istanbul seem to struggle

33 with construction permits and have to wait on average 87 days to obtain a permit (Figure 2-19).22 Additionally, a variation is notable by firm size, with SMEs spending nearly twice the time as large firms dealing with construction permits (60 versus 32 days).

2.42. Some measures in improving land access conditions for businesses are underway. YOİKK‘s Location of Investment Technical Committee has set up goals for technical and legal actions in streamlining land title procedures as well as re-estimate land title fees, according to best international practice. Plans are also underway to evaluate the procedures involved in acquiring building permits and the costs for the construction process. The Government is preparing large plots of land for the development of industrial zones, aimed to attract high-value added or high-tech investments.23

BUSINESS INSPECTIONS

2.43. Business inspections appear to be less burdensome in Turkey than in comparator countries. The average time all employees in a company spend with inspections per year is 6.6 days, which puts Turkey ahead of such comparator economies as Bulgaria, the Czech Republic and Hungary. The survey results are similar when firms are asked about the number of inspections taking place in their organization. The average frequency of inspector visits has also diminished from 4 inspections in 2005 to 2 in 2008 (Figure 2-20). Additionally, regional comparisons show significant variations in inspection times and duration, with the average number of inspections per year being highest in Central Anatolia and lowest in the Aegean region. The differences can be explained by the fact that it is often the municipalities themselves enforcing regulations, without necessarily having higher approval.

Figure 2-20: Number of inspections per year, selected Figure 2-21: Average time spent dealing with countries each inspection (days), by firm size

7 6.1 4 3.7 3.7 6 2.8 5 3 3.7 4.0 4 3.5 2.2 2.8 3 2.4 2.4 2 2.0 2 1 1 0 0 Turkey Hungary Turkey Brazil Bulgaria Czech Turkey Chile Small Medium Large Total 2008 2009 2008, 2009 2009 Rep. 2005, 2006 manuf 2009 manuf

Sources: Enterprise Surveys and Turkey ES 2008

2.44. Time spent with inspections varies by industry and firm size. Inspections are least frequent in the services sectors whereas food and chemicals sectors are inspected more often. This is to be expected as the time corresponds to the health, environment, or safety risk levels. For the same reasons, large businesses are inspected more frequently than small and medium firms. However, when instead examining the time that each firm spends on each inspection, medium firms seem to be more affected. The average for medium and large businesses is 3.7 days, comparing with 2.2 for small firms (Figure 2-

22 The Enterprise Survey results for construction-related permits correspond better with the Doing Business indicator when taking into consideration the results for Istanbul, seeing as Doing Business methodology makes the assumption that the firm being analyzed operates in the country‘s most populous city, in this case, Istanbul (doingbusiness.org). This might suggest that Doing Business data is not necessarily representative of the regulations in other cities and regions. 23 2009 Technical Committee Action Plans of the Turkish Coordination Council for the Improvement of the Investment Environment (YOİKK)

34 21). The variations by ownership form and exporting status are on the other hand not significant for any of the variables.

2.45. Further harmonization among institutions and municipalities would further reduce the burden on businesses. Although the survey results from 2008 show signs of improvements in enforcement of and compliance with government regulation in Turkey, these remain quite weak and disorganized areas. As mentioned above, the municipalities are often the ones implementing regulation, while the inspection system per se is centralized. This institutional setup tends to create a burden for businesses and individuals, when having to comply with procedures established locally without prior agreement at the central level. At this stage, there are no clear governmental goals set for improvement of inspection procedures.

CLOSING A BUSINESS

2.46. Productivity and growth in an economy are to a large extent affected by the efficiency of firm entry and exit, through improved resource allocation and higher aggregate efficiency. A crucial feature of a friendly investment climate is the existence of efficient insolvency procedures. Just as markets with more net entry of firms over time become exposed to greater competitive pressure, so do economies allowing for efficient exit of less productive firms induce higher market competition, thus creating higher allocative efficiency and better long-term growth.24

2.47. By having well-designed insolvency and bankruptcy procedures, creditors are insured that debts from the insolvent firms can be collected efficiently and with the highest recovery rate possible. In this manner, an economy can improve the degree of competition in product markets and, as a result, productivity. Nevertheless, it is often the case, particularly in developing and emerging economies, that the insolvency procedures are too elaborate, with high administrative costs and long delays. The cause, as presented by Djankov, et al. (2008a), is the practice of developing countries copying industrialized ones in their often very formal bankruptcy procedures, without having the skills or the capacity to follow through with saving or swiftly closing the insolvent business.

2.48. The most recent reforms in Turkey in regards to insolvency measures took place in 2003 and 2004, with the main amendment in the Execution and Bankruptcy Law25 focusing on so called ―Suspension of Bankruptcy.‖ The procedure, applied from the Swiss bankruptcy code, presents the option of suspension of all insolvency proceedings when the possibility of restructuring the company can be presented to the court. The objective is to give the debtor a chance to restructure and avoid bankruptcy before creditors force the debtor into liquidation.

2.49. Turkey has to close the gap with OECD standards in insolvency procedures. When taking a first look at the results for Turkey in the Doing Business 2010 results, firms dealing with bankruptcy procedures seem to have experienced significant improvements since 2005. The time to close an insolvent business has in more than halved, from 5.9 years in 2005 to 3.3 years in 2009. This puts Turkey on par with the average of Europe and Central Asia. Furthermore, the recovery rate has risen from seven to 20 percent in the past four years. Although a strong improvement, this is still well below the average OECD level and international best practice (Japan). Out of 183 measured economies, this puts Turkey at rank 120 (Table 2-5).

24 See for instance Nicodeme and Sauner-Leroy, (2007) 25 Execution and Bankruptcy Act No. 2004

35 Table 2-5: Bankruptcy procedures (duration and recovery rate) comparison of selected countries Time (years) Recovery rate (cents on the dollar) Country 2005 2009 2005 2009 Turkey 5.9 3.3 7.0 20.2 International Best Practice 0.4¹ 0.4¹ 93² 92.5² Region (Europe & Central Asia) 3.9 2.9 29.8 31.6 OECD 1.5 1.7 74.0 68.6 Brazil 10.0 4.0 0.5 17.1 Bulgaria 3.3 3.3 33.0 32.1 Chile 5.6 4.5 23.0 21.3 China 2.4 1.7 31.0 35.3 Czech Republic 9.2 6.5 18.0 20.9 Hungary 2.0 2.0 36.0 38.4 Malaysia 2.2 2.3 39.0 38.6 Poland 3.0 3.0 26.5 29.8 Romania 4.6 3.3 17.0 28.5 Russia 3.8 3.9 27.0 28.2 ¹Ireland; ²Japan Source: Doing Business 2010

2.50. The time connected with bankruptcy procedure has been reduced, possibly because courts have applied the law in a strict way, demanding from the indebted company to present a serious change in being able to pay the debt in full before allowing extensions of the reorganization period. Nonetheless, practitioners seem to agree that the suspension amendment to the Bankruptcy Law has not been applied in an efficient manner and that many debtors have avoided serious efforts to reorganize, and instead stayed in business at the cost of their creditors. The possibility of long extensions of bankruptcy suspension is one of the main problems in the insolvency procedures, allowing for strong depreciation of the recovery rate over time. Other issues to be addressed include a lack of training among judges and often insufficient background in international standards in the case of insolvency procedures.

2.3 Labor Market and Skills 2.51. The skill level of the workforce is positively associated with productivity, while smaller firms are less likely to employ more educated workers and to provide training. Econometric analysis shows that firms with a higher share of staff with university education can expect a positive impact on productivity (Table 2-6).26 At the same time, a higher share of female workers is negatively associated with manufacturing productivity, reflecting the generally lower education levels of female workers, especially in labor intensive industries, such as textiles and garments. Survey results also show that larger firms have higher shares of non-production workers (e.g. managers, administration, sales), are in a better position to afford skilled staff with university education and also have the means to provide training to workers.

26 Results of the 2005 survey showed that more productive firms tended to have lower levels of part-time and unskilled employees. The length of the training provided to skilled workers – a variable not available in 2008 – also had a positive association with productivity levels. See Table 2-A-2.

36 Table 2-6: Summary of the effects of labor and skills

Dependent variable Explanatory variable Probability Probability of Productivity Employment of exporting receiving FDI Non-production workers + + Staff - female workers - Staff - university education + + Training (dummy) + Source: Staff calculations based on Escribano et al. 2009. See Tables in Annex 2-A for detailed results.

The labor market context

2.52. Strong economic growth, averaging 6.5 percent since the 2001 crisis, has not been accompanied by significant improvements in the labor market. A comparison with the OECD indicators four years ago as well as with OECD, EU-15 and selected economies shows that the overall participation rate of 46.9 percent is modest and has actually decreased since 2005 (Table 2-7). Turkey‘s labor participation is 20-25 percentage points below OECD and EU-15 averages. There has not been any improvement in unemployment in Turkey since the sharp increase during the 2001 crisis, but unemployment figures reached instead a record level of 11 percent in 2008.

2.53. The employment gap in an international context becomes even more noticeable when looking at the employment rate for women, which in 2007 was 26.7 percent, far below the OECD and EU-15 averages of 61.3 percent and 65.3 percent respectively. The 2008 Enterprise Survey also shows that on average only 16 percent of production employees and four percent of non-production employees are female. Earlier World Bank analyses have presented two main reasons for the low participation rates for women: the transition away from the agricultural sector, where women had higher participation rates, and selective participation in the urban labor market.

Table 2-7: Key labor market indicators, Turkey and selected OECD countries, 2008

Persons aged 15-64, percent Labor force participation rate Employment/population ratio Unemployment rate Total Males Females Total Males Females Total Males Females Turkey 46.9 74.8 26.7 41.7 66.6 23.5 11.0 11.0 11.9 Turkey (2005) 51.3 76.2 26.5 45.9 68.2 23.7 10.5 10.5 10.6 Hungary 61.5 68.3 55.0 56.7 63.0 50.6 7.9 7.7 8.1 Mexico 62.2 83.5 43.3 59.9 80.7 41.4 3.7 3.4 4.3 Poland 63.8 70.9 57.0 59.2 66.3 52.4 7.2 6.5 8.0 Portugal 74.2 79.5 68.9 68.2 74.0 62.5 8.1 6.9 9.4 Slovakia 68.9 76.4 61.4 62.3 70.0 54.6 9.6 8.4 11.4 Spain¹ 73.7 83.0 64.1 65.3 74.6 55.7 11.4 10.1 13.1 EU-15 72.5 79.7 65.3 67.4 74.4 60.4 7.1 6.6 7.6 OECD 70.8 80.5 61.3 63.9 75.7 57.5 6.0 6.0 6.2 Source: OECD, 2009 1. Indicators are based on values 16-64 years

37 2.54. Higher education levels are positively associated with employment and labor participation. In Turkey‘s case, this is particularly evident for female workers. Labor force participation for women with upper secondary education was merely 34 percent in 2008, whereas the corresponding rate for women with tertiary education was 71 percent. As Turkey‘s population expands by 800,000 every year, the young working population will increase rapidly, making focus on education on all levels and school- to-work programs even more crucial. Seeing as training opportunities after formal training is limited today (World Bank, 2008), Turkey needs to put more focus on relevant vocational in-class training as well as on work training programs.

2.55. Inclusion of workers reallocating from the agricultural sector and informality are important underlying causes of poor labor market performance. One of the main reasons for Turkey‘s low job creation in the past seven years can be widely ascribed to the significant transfer that the country has made away from the agricultural sector. Other factors include the high informal employment in Turkey, the share of unregistered workers averaging 50 percent in manufacturing and services (OECD 2007), as well as the high income tax rate and tax wedge. Although, as mentioned in the previous section, the tax wedge in Turkey has decreased since 2005, it still remains among the highest when compared to other OECD countries. OECD data also shows that restrictiveness of employment protection in Turkey is highest when measured against the same group of countries, with a 3.5 ranking on a scale of 0-6, with 6 being the most restrictive (Venn, 2009). Important mentioning is also that no change in employment protection has happened since 2003. The higher strictness does not allow for flexibility for firms in hiring new employees.

Employment trends and constraints

2.56. Firms appear less bothered by regulatory labor issues, although the result may be affected by the crisis environment and the issue is likely to become a concern for sustainable recovery. The 2008 survey shows that firms in Turkey to a much lesser extent than in 2005 regard Figure 2-22: Labor regulations as an obstacle, country labor regulations as a serious obstacle to comparison business operations. 8 percent perceived 70% 58% labor regulations to be a major or very 60% 46% severe obstacle, compared to 46 percent in 50% 2005 (Figure 2-22). Also by international 40% 27% 30% standards, this has been a great advance, 30% 20% 13% improving Turkey‘s position vis-à-vis the 8% 8% 9% 10% average of the Europe and Central Asia 10% region. Whereas a plausible explanation for 0% Turkey Turkey Hungary ECA Bulgaria Poland Chile Turkey Brazil the improvement in perceptions can be the 2008 2008 2009 Region 2009 2009 2006 2005 2009 labor reforms that were initiated in the (manuf) (all) (manuf) beginning of 2008 (Box 2-6), the Source: Enterprise Surveys inflexibility of labor regulations is still mentioned by enterprises in face to face interviews as the overarching constraint facing firm operation and growth. It is possible that the timing of the survey – from April 2008 to January 2009 – found business more preoccupied with other, more immediate constraints – e.g. loss of market share requiring downsizing or access to finance – thus decreasing the relative importance of labor regulations. Bearing the current financial crisis in mind, the 2005 survey results might, in fact, better reflect today‘s labor regulation environment. It is therefore a concern that should be given more attention as the economy improves.

38 2.57. The obstacle level does not vary by firm size, whereas ownership form appears to affect to what extent firms are burdened by labor regulations. While 8 percent of domestically owned firms find regulations to be an obstacle, close to none of the foreign companies cite them as a constraint. The 2005 Survey addressed the hypothetical issue whether firms would change the number of permanent full- time employees if labor restrictions did not exist (World Bank, 2007). The response was remarkable, showing that 62 percent of the companies would in fact increase their employment levels. Unfortunately this is not comparable with 2008 data, as the hypothetical question was not addressed. Although thought needs to be given to what the actual effect on employment growth would be if this were the case, the result still shows that firms were hindered in their operations by the strict labor regulations three years ago.

Box 2-6: Employing workers in Turkey: rules of hiring, work schedules and redundancy One of the indicators in the Doing Business assesses the regulation of employment, with the methodology specifically measuring the effects on hiring, the rigidity of working hours, difficulty of redundancy and redundancy cost.

In the case of Turkey, Doing Business depicts a limited liability company, operating in the manufacturing sector in Istanbul, fully domestically owned, with 60 employees. The business is also subject to collective bargaining agreements and is assumed to abide by every law and regulation, but does not grant workers more benefits than mandated by law. The worker is assumed to be a 42-year-old, nonexecutive, full-time, male employee, has worked at the same company for 20 years, and earns a salary plus benefits equal to the economy‘s average wage during the entire period of his employment. Additionally, the worker is not a member of a labor union, unless membership is mandatory.

Fixed-term contracts are not allowed for permanent tasks in Turkey, there is no limit for fixed-term contracts, and the mandated minimum wage relative to the average value added per worker is 41 percent. When measuring rigidity of hours, there are restrictions on night work but not on weekly holiday work, 6- day workweeks as well as 50 hour-workweeks (including overtime) for 2 months are allowed, and there is a requirement for paid vacation of 26 workdays per annum for an employee with 20 years of service.

Termination of workers due to redundancy is allowed in Turkey and the employer is not required to notify a third party nor does he need the approval of a third party before terminating one or a group of redundant workers. The law does, however, mandate retraining or alternative placement before termination and there are also priority rules applying to re-employment. When it comes to the cost of redundancy, these are more onerous in Turkey than in many comparator countries. An employer is required to give 2 months notice before a redundancy termination, and the severance pay for a worker with 20 years of service equals 20 months of wages. However, no penalty is levied for redundancy dismissal. The redundancy cost equals in total 23 months of salary, a rate quite high in particular when compared to the ECA and OECD averages of 6 to 7 months.

Turkey has an overall rating of 44 index points (on a 1-100 scale, with higher values representing more rigid regulation). Relative to other economies, Turkey ranks 145 out of 183 measured countries.

Source: Doing Business 201027

2.58. Another sign of the restrictive labor policies is the earlier discussed large informal sector. With 52 percent of the firms interviewed stating that they compete with unregistered and informal businesses, this shows that many establishments choose away business restrictions completely. At the same time, protection of workers through social security networks and unemployment insurance policies are very weak in Turkey.

27 Box 2-6 presents the methodology of World Bank Doing Business and can be found at www.doingbusiness.org, The specific labor regulations of the Turkish Labor Code are clarified in Article 11 and 17 of Labor Law No. 4857.

39

2.59. The survey results further indicate that the use of temporary employment among Turkish enterprises has decreased remarkably since 2005. In 2005, 31 percent of the workforce was temporary.28 This number has diminished to 4 Figure 2-23: Temporary employment as a share of total percent, when sampled in 2008. Moreover, employment, country comparison the average length of temporary employments has also decreased, from 7 months in 2005 to 35% 31% 30% 4 months three years later. This reduction in 25% use of temporary workers complements the 20% overall opinion in the survey that the 15% 10% 12% 6% 7% 10% 4% 4% perception of labor regulations as an obstacle 5% 2% has dropped from 46 to 8 percent. The change 0% in temporary employment is depicted in Bulgaria Turkey Turkey Brazil Romania Czech Poland Turkey Figure 2-23. Companies‘ relying less on 2009 2008 2008 2009 2009 Republic 2009 2005 (all) (manuf) 2009 (manuf) temporary workers today could de facto confirm that they are less burdened with the Source: Turkey ES 2008 obligations usually associated with hiring permanent employers, such as benefits and payroll taxes. Nevertheless, despite these positive implications and seeing as the Government is only in the first stages of a comprehensive reform package, more time needs to be dedicated in order to fully analyze the impact of the labor reforms in Turkey. The main state employment initiatives taking place are summarized in Box 2-7.

Box 2-7: Labor market reforms in Turkey  Earlier important legal amendments in Turkey include the introduction of the Labor Law No. 4857 from 2003, which harmonized legal and institutional regulation with the International Labor Organization (ILO) and the European Union, as well as the restructuring of the Turkish Employment Agency (ISKUR), which was given approval to the opening of private employment offices.

 The Ninth Development Plan 2007-2013 sets out a comprehensive labor market reform, focusing on three major areas of reform - Reducing the non-wage cost of hiring. This encourages companies to hire more unskilled employees and creates incentives for workers to participate in the labor force. - Increasing employment flexibility while improving worker protection regulations and supporting disadvantaged workforce groups. Severance payment obligations in Turkey in 2006 were among the highest in comparison with other OECD countries, creating further disincentives for formal employment. On the other hand, unemployment insurance policies only reached 4.4 percent of unemployed workers in 2007. - Improving coordination between employment and training programs, by adjusting existing vocational and secondary education programs to the needs of the labor market and providing assistance in finding jobs to minority groups, such as youth, women and disabled workers.

 The Government has initiated the first phase of the labor market reform through the introduction of the Employment Package, which covers amendments to existing laws, with the objective of increasing employment in the country. The main actions in the Employment Package include - Five percent reduction of employer‘s social security contributions - Further reductions in social security contribution for employment of the young, women and workers with disabilities - Legislative changes in employer obligations in regards to inter alia hiring quotas

28 Full-time temporary or seasonal employees are defined as all paid short-term (i.e. for less than a fiscal year) employees with no guarantee of renewal of employment contract and work 40 hours or more per week for the term of their contract (World Bank Enterprise Surveys, 2008)

40 - Transfer of funds from the Unemployment Insurance Fund (UIF) to ISKUR as a support for training programs and enable other unemployed, registered to ISKUR – alongside the ones who receive unemployment insurance – to benefit from fund financed education programs. - Broadening and sustainability of active labor market programs, with the aim of increasing the number of beneficiaries from 32, 200 in 2008 to 214,000 by the end of 2010. - Coooperation between the local Provincial Employment and Provincial Vocational Education and Training Councils has been initiated, with the aim of meeting the needs of the labor market more effectively

 Further flexibility measures have been put forward by the Government in the Medium Term Program 2010-2012, including severance pay reforms and additional modifications to the Labor Law no. 4857.

 YOİKK‘s Employment Technical Committee also specifies labor market reforms in its 2009 Action Plan. These include facilitation of use of the short-term work allowance, elimination of the burden for employers of filing recruitment and discharging document to multiple institutions and streamlining the process of work permit applications for foreign workers. The initiatives are planned to be completed in the fall of 2009.

2.60. The change in opinion regarding labor regulations and employment patterns with Turkish firms might, to some extent, reflect recent reforms. With clear plans on decreasing informality, applying flexible employment models and focusing on education-to-labor-demand compatibility, Turkey is taking earnest steps in improving the Figure 2-24: Labor regulations as an obstacle, by size, unemployment situation in the country. The ownership and exporting status shift in firms‘ opinion regarding labor regulations as a strain on operations could be 11% 12% an indication of the effects of the labor 9% 10% 8% 8% 8% 7% reform package that the Government has 8% initiated, as presented it the Ninth 6% Development Plan 2007-2013. The result 4% was particularly positive among foreign firms 2% 0% where almost two thirds of the companies 0% responded that regulations were no obstacle to business (Figure 2-24). This indicates that Turkey‘s efforts towards supporting investors, among others through incentives of insurance premium cuts may be paying off Source: Turkey ES 2008 and should be continued and expanded.29

2.61. Reductions in labor costs will have positive effects on employment levels. Although the cutback in non-wage labor costs, i.e. tax wedge reduction, is not at this stage reflected in the change in the employment numbers from 2005 to 2008, it can be expected that this action will result in higher employment levels. Turkey‘s Action Plan for the 60th Government (2007) in fact analyzes the impact of non-wage labor cost cuts, with the conclusion that employment in Turkey is positively receptive to changes in labor costs.

29 Cabinet Decision on State Aids for Investments, July 16 2009

41 Education, skills and training

Figure 2-25: Population with at least upper secondary 2.62. Education levels are lower than in education, percent, 2007 comparators. OECD data, illustrated in Figure 2-25, show that 26 percent of the 100 86 86 87 89 90 79 Turkish adult population holds secondary 80 69 education diplomas. This is not only lower 60 49 than in 2005, but well below the OECD 35 40 26 27 average of 69 percent and the EU19 average of 70 percent. Graduates with tertiary 20 education in Turkey are also scarce and the 0 entry rate to higher education programs (tertiary education) is well below international standards. Only 29 percent enroll in higher education in Turkey, compared to 56 percent Source: OECD average in the OECD countries.

2.63. The proportions of skilled and unskilled workers have tipped in favor of skilled employees. As seen in Figure 2-26, the distribution of permanent workers between non-production, skilled production and unskilled production workers was 16 percent, 42 percent, and 42 percent respectively in 2005. The corresponding shares in 2008 were 25 percent, 46 percent and 29 percent (Figure 2-27). When comparing worker types by firm size, the variations have remained marginal, with large firms having a higher share of unskilled production workers than small and medium firms.

Figure 2-26: Distribution of permanent workers, by Figure 2-27: Distribution of permanent workers, by type and firm size, 2005 type and firm size, 2008

60% 60% 48% 49% 50% 46% 50% 44% 46% 41%43% 42%42% 43% 39% 38% 40% 33% 40% 31% 27% 29% 30% 30% 24% 25% 25% 19% 20% 20% 16% 15% 16% 20%

10% 10%

0% 0% Small (5-19) Medium (20-99) Large (100+) Total Small (5-19) Medium (20-99) Large (100+) Total Non-production Skilled Production Unskilled production Non-production Skilled Production Unskilled production Source: Enterprise Surveys

42 2.64. Nearly a quarter of Turkish firms rate the education and skills levels of the workforce as a “major” or “very severe” constraint on operations and growth. Although this is an improvement from the 33 percent in 2005 (Figure 2-28), the Figure 2-28: Education of workforce as an obstacle, by country high rate confirms Turkey‘s serious problem with lagging education and skill 50% 42% 43% levels. This shows that much remains to 36% 40% 33% be done in coordinating labor supply 29% 30% 25% with the demands in the business sector. 21% 21% Rated by ownership form, foreign firms 20% perceive the education level to be a 10% 6% bigger constraint than domestic firms do 0% (19 percent and 15 percent respectively). Hungary Bulgaria Turkey Turkey Czech Turkey Poland Chile Romania Variation can also be seen among regions 2009 2009 2008 2008 Rep. 2005 2009 2006 2009 and industries in Turkey. Firms in the (manuf) (all) 2009 (manuf) Aegean and Black Sea – Eastern regions are especially concerned about Source: Enterprise Surveys 2008 inadequately educated workforce (30 and 31 percent respectively), while more than a third of enterprises in the chemicals, electronics, metals and construction industries perceive skills and education of workforce to be a major or very sever obstacle to operations (Figure 2-29).

Figure 2-29: Education of workforce as an obstacle

50% 42% 45% 37% 40% 33% 30% 31% 29% 26% 26% 26% 28% 30% 24% 27%23% 23% 23% 23% 24% 25% 19% 20% 21% 20% 20% 15% 13% 15% 10%

0%

Source: Turkey ES 2008

2.65. The share of the workforce with a university degree also remains low, with somewhat more positive results for the services industry. The survey results show that 10 percent of the total workforce in the manufacturing industry held a university degree in 2008, which is an actual deterioration from the 11 percent share in 2005. When including the services sector, the share of employees with a university rises to 18, putting Turkey in a relatively positive position compared to for instance Hungary (9 percent), Brazil (10 percent) and the Czech Republic (16 percent). The share of highly educated employees is particularly high among non-exporting firms (19 percent versus 14 percent for exporters). A higher share can also be found in the Central Anatolia (23 percent) compared to only 11 percent in the South. When scrutinizing education levels by industry, construction (39 percent), wholesale and retail (27 percent) and chemicals (22 percent) are the sectors that stick out, compared to relatively low shares in the textiles, garments and basic metals industries (7 percent, 8 percent and 9 percent respectively). There is no significant variation by firm size and ownership form. Taking into consideration the econometric results presented in Table 2-6, a highly educated workforce is a key issue for enterprises, as a high share leaves a significant impact on the firm‘s productivity and growth.

43 2.66. The number of Turkish firms offering formal training to their employees has increased slightly. Figure 2-30 demonstrates that 24 percent of manufacturing businesses claimed to offer training to their employees in 2005. This number rose to 29 percent three years later. Training is, as Figure 2-30: Share of firms offering formal training to can be expected, offered more often the employees, by country larger the firm (see also econometric analysis 70% 61% in Table 2-6), where the share of large firms 60% 53% offering training is three times higher than 50% with small businesses. Enterprises with 40% 29% 29% 31% 30% 24% 25% exporting activities are also more active in 20% 15% offering training to workers. Measured by 10% industry, training is provided to a lesser 0% extent in the machinery, garments, food and Hungary Turkey Romania Turkey Turkey Bulgaria Brazil Poland textiles sectors, while 83 percent of the firms 2009 2005 2009 2008 2008 2009 2009 2009 (manuf) (all) (manuf) in the electronics industry provide training for their workers (Figure 2-31). Source: Turkey ES 2008

Figure 2-31: Share of firms offering formal training to employees, by firm size, ownership form, exporting status, region and industry

90% 83%

60% 54% 53% 40% 39% 40% 41% 42% 34% 32% 29% 27% 30% 23% 22% 23% 22% 27% 30% 17% 18% 19% 20%

0%

Source: Turkey ES 2008

2.67. Survey data also shows that Figure 2-32: Share of production and non-production production workers are more likely to workers that received training, country comparison receive training than non-production 80% 71% 71% Production 67% workers. 67 percent of production workers 63% Non-production received formal training while only 50 60% 50% 51% 50% 50% 45% percent of non-production workers were 38% 38% 40% 32% 35% provided the same. Put in an international 27% perspective, both production and non- 24% production workers in Turkey are provided 20% 9% training to a larger extent than many other 0% comparative emerging economies, such as Romania Hungary Czech Chile Poland Turkey Bulgaria Brazil Romania and Hungary (Figure 2-32). 2009 2009 Rep. 2006 2009 2008 2009 2009 2009

Source: Turkey ES 2008

44 2.4 Innovation 2.68. Innovation is positively correlated with the productivity of Turkish firms. Innovation – encompassing product and process innovation, technology adoption, quality upgrading and the use of information communication technologies (ICT) 30 – has become crucial among world‘s economies and much focus is put by governments towards promoting and stimulating firm-based technological development. Analysis of the 2008 survey of Turkish enterprises confirms the importance of innovative behavior for firm performance. Table 2-8 shows a positive association between the occurrence of investments in R&D and productivity levels. At the same time, firms that re-organized production processes by resorting to outsourcing where also more productive.31

Table 2-8: Summary of the effects of innovation

Dependent variable Probability Explanatory IC variable Probability of Productivity Employment of exporting receiving FDI Website (dummy) + + R&D expenditure + + New product (dummy) + Foreign technology (dummy) + Outsourcing (dummy) + Quality certification (dummy) + + Source: Staff calculations based on Escribano et al. 2009. See Tables in Annex 2-A for detailed results.

2.69. Employment, exports and FDI are also positively associated with innovation at the firm level. Both the 2005 and the survey present a positive association of employment with the use of ICT (website n 2008, e-mail in 2005) in communication with customers and suppliers. Also, both periods measured showed that firms with quality certifications have more employees than firms without.32 Also for exports there is a significantly positive correlation between variables reflecting innovation and the probability of exporting. In particular, firms with quality certifications are more likely to engage in exporting activities, with an increasing importance since 2005. The use of ICT is also positively related with exports in both years. Additionally, firms that introduced new products in 2008 exported a higher share of their sales. The econometric analysis also shows that firms that use technology licensed by foreign firms have higher levels of foreign ownership, a correlation that is consistent with the 2005 Survey. FDI is also more present with companies that spend on R&D activities.

30 The World Bank ICT Sector Strategy define ICT as consisting of hardware, software, networks and media for collection, storage, processing and presentation of information, including computing, undersea cables, radio waves, spreadsheets and websites as well as post offices and newspapers. 31 Analysis of the 2005 survey also showed the relevance of ICT use (in the form of conducting business via e-mail) for productivity. The significance of ICT variables for productivity levels is lost in 2008. This may be explained by the fact that use of the internet has become more widespread since 2005, thus muting productivity differences due to ICT. The purchase of new technology also had a positive association with productivity in 2005 and turns out not to be significant in 2008. At the same time, R&D expenditures –-not significant in 2005—have a positive association with productivity in 2008. See Table 2-A-2 for details. 32 A difference between the two periods is that firms in 2005 that owned technology licensed by foreign companies had higher employment levels. This is no longer true in 2008, perhaps indicating that the use of foreign technology has become more common across firm sizes.

45 Performance in innovation

2.70. R&D spending has increased but is still lower than in comparator countries. Gross expenditures on R&D (GERD) in Turkey have seen a positive development (Figure 2-33). Overall investment in R&D has nearly doubled over the past ten years, reaching 0.73 percent of GDP in 2008. Although a clear improvement, Turkey still lags behind in comparison to other middle income countries and in particular to the OECD, where gross R&D expenditure amounted to 2.29 percent in 2007, compared to 0.71 percent for Turkey in the same year. Moreover, the share of R&D investments accredited to the private sector remains on a low level relative to international standards. ES data show an improvement in R&D investment as share of total sales in the manufacturing sector, from 2.8 percent in 2005 to 3.4 percent in 2008 (Figure 2-34). Figure 2-33 also illustrates that 0.34 percent of GDP in Turkey was contributed by industry R&D investments in 2008. The increasing private expenditure ahead of public R&D spending can to a large part be contributed to the 2007 R&D Law amendment on tax exemptions for R&D investment.

Figure 2-33: R&D expenditure as a share of GDP - Figure 2-34: R&D investment as a share of total total, private and public, 2000-2008 sales, 2005 and 2008 0.8 4% Total Private Public 3.41% 3.41% 2.84% 0.6 3%

0.4 2% % 0.2 1%

0.0 0% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2005, manuf 2008, all 2008, manuf

Source: Turkstat Source: Enterprise Surveys

2.71. Earlier World Bank research has showed that Turkey‟s dynamics in public-private and university-private collaboration covering innovation and technology needs to be boosted.33 This is confirmed by the WEF indicator, where Turkey received a mediocre rating in comparison to 132 other economies.34 Other areas of concern have been the lack of efficient intermediaries for transfer of publicly funded research to the private sector through for instance spin-off companies, joint research initiatives and technology transfer offices and relatively low number of patent applications by Turkish firms, both at home and internationally.35

33 Turkey CEM (2006) 34 The World Economic Forum Global Competitiveness Report 2009-2010 measures private cooperation with universities and research institutes. Turkey was given a score of 3.4 on a 1-7 scale (1=no collaboration; 7=extensive collaboration). 35 Issues related to innovation for Turkey are addressed in a forthcoming World Bank report, Turkey Innovation Policy for Competitiveness and Employment Generation.

46 2.72. Patent activity picked up after the 2001 economic crisis and has steadily been rising but it remains lower than in comparator economies. The patent activity in an economy is a valuable measure of a country‘s technological development and commercial applicability. While patent Figure 2-35: Patent applications, per million population, 2000-2008 applications by Turkish firms increased in the past six years, the country still shows lesser 140 activity than many other emerging 120 economies. The number of applications by 100 Turkish inventors per million inhabitants 80 amounted to 36 in 2008, an increase by over 60 50 percent in two years. It is also an 40 international improvement, where Turkey has 20 surpassed some comparator countries, for 0 example Brazil. This number is however still 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 low when compared to for instance 119 applications in the Czech Republic and 76 in Brazil Bulgaria Czech Rep. Poland Turkey Poland, indicating that Turkey still needs to Source: WIPO and WDI catch up internationally (Figure 2-35).

2.73. The overall number of patent applications by Turkish firms locally and abroad reached nearly 2,700 in 2008. Table 2-9 shows that 85 of these applications were filed to the U.S. Patent Office. It is however evident that Turkish businesses file considerably more patent applications with the European Patent Office, where 246 applications were made in 2007. Nonetheless, the application level to both offices is a remarkable improvement compared to a decade prior (10 patents in the U.S. and 20 in the European Patent Office). Additionally, the overall number of patents filed in Turkey‘s national patent office has picked up significantly since the 2001 crisis, from 1,152 in 2003 to 7,137 filed applications in 2008. Another good indicator of technological improvements is industrial design applications, which is another field where Turkey has advanced, from 2,326 applications in 2000 to 6,870 eight years later.

Table 2-9: Turkey's patent applications, 1995-2008

Patent applications with Turkey's patent office Patent applications, total US patents European patents Industrial designs Year (TPI) (WIPO) (USPTO) (Eurostat) (WIPO) 1995 1,690 185 6 5 1996 902 209 7 9 1997 1,531 226 10 20 1,995 1998 2,483 239 16 31 1,917 1999 3,020 332 29 22 1,779 2000 3,433 366 28 44 2,326 2001 3,214 421 28 45 2,708 2002 1,874 503 36 61 3,747 2003 1,152 603 28 85 4,208 2004 2,317 868 48 125 4,834 2005 3,518 1,120 46 163 5,380 2006 5,252 1,426 71 197 6,167 2007 6,247 2,184 66 246 6,552 2008 7,137 2,678 85 6,870

2.74. Turkish firms are to a larger extent introducing new products and are doing better in an international comparison. The level of innovation in the Enterprise Survey does not focus on patent

47 applications but instead measures the development of new products in firms as well as the level of R&D spending. When asked if the company in the last three years has introduced a new product, 45 percent of Turkish manufacturing firms active in the manufacturing sector responded positively, showing an improvement since the 2005 Survey as well as in relation to comparative economies.

2.75. There are some variations in the spread of R&D investments and introductions of new products by size, ownership and exporting status. The survey data, as depicted in Figure 2-36, shows that R&D spending is more widespread among medium and large firms, where on average 32 percent and 34 percent of the companies respectively invest in R&D, while only 16 percent of small firms do the same. The variations by size when considering the share of firms with new products or services were again more prominent the larger the firm. Businesses with exporting activities also tend to invest in R&D (52 percent) more than non-exporters (39 percent). Additionally, foreign firms tend to introduce new products to a much larger extent than domestic.

Figure 2-36: Share of firms developing new products and R&D spending, by firm size, ownership form, exporting status and region

60% 55% 58% 52% 50% 50% 45% 41% 43% 41% 43% 44% 38% 40% 32% 34% 28% 24% 24% 23% 23% 22% 21% 21% 16% 19% 20%

0%

New product R&D spending

Source: Turkey ES 2008

2.76. Investments in innovation activities are more widespread in industries with relatively advanced technologies and show some variation by region. Two thirds of the firms in the metals and machinery industries and as much as 90 percent of companies in the electronics sector have introduced new products in the period 2004-2007. Enterprises in the electronics industry have also shown a noteworthy spread of R&D spending. Regionally, the largest share of firms introducing new products could be found in the Aegean region and in Central Anatolia, while the South had the lowest share of innovation activities among firms.

Technology adoption

2.77. The use of foreign-licensed technologies by firms in Turkey is on par with comparable emerging economies. Using licensed technology can bring significant know-how benefits for a firm and is a prime source of technology adoption. In Figure 2-37 we can see that the share of Turkish companies using technology licensed by foreign-owned firms has increased, from 12 percent in 2005 to 16 percent three years later.

48 Figure 2-37: Share of firms using technology licensed by foreign firms, country comparison

25% 25% 20% 16% 16% 16% 15% 13% 13% 12% 12% 10% 6% 5%

0% Poland Czech Turkey Brazil Hungary Turkey Turkey Romania Russia 2009 Rep. 2005 2009 2009 2008 2008 2009 2009 2009 (manuf) (all) (manuf)

Source: Enterprise Surveys

2.78. There are significant variations by size, ownership form and region in the level of use of technology licensed from a foreign-owned company. More than 30 percent of large firms in Turkey apply foreign-licensed technology to their business, compared to only 8 percent of small enterprises. More foreign firms (26 percent) acquire Figure 2-38: Share of firms using foreign licensed technology, technology by foreign-licensing than by size, foreign ownership, exporting status and region domestic (15 percent) as well as exporting businesses (19 percent) 31% 30% 26% compared to 14 of non-exporters, 22% 19% 19% 20% suggesting that local firms lag behind 20% 15% 17% 14% 13% 14% foreign competitors in the use of 8% 10% technologically advanced equipment. Also here, the Central Anatolia and 0% Aegean regions stick out, with 20 and 22 percent of the establishments in these regions using foreign-licensed technology. The Marmara region lags behind, with a share of 13 percent Source: Turkey ES 2008 (Figure 2-38).

Quality standards

2.79. Firms in possession of quality standards tend to see a positive impact on productivity and competitiveness. Through the application of standards, a company can facilitate the adoption of technology and innovation as well as increase productivity and competitiveness by the embedded product and process information that usually can be found in standards. Additionally, the positive effects are even stronger when applying internationally recognized quality certification.36 It is therefore likely that Turkish firms with ISO certifications are more technologically advanced and thus more competitive globally.

36 Blind and Jungmittag (2005).

49 2.80. The application of ISO Figure 2-39: Share of firms with internationally recognized quality 9001 in Turkey has shown certifications, country comparison remarkable increases since the 50% 2001 crisis, with more than 13,200 43% All firms Small firms 39% 39% certificates issued by the end of 2008 40% and comparing relatively well to 30% 30% 26% 27% 27% other economies. The survey data 30% 25%24% 26% depicts additionally improving 20% 21% 20% 20% 17% figures, with 30 percent of all Turkish 12% 8% firms in 2008 claiming to have an 10% 8% internationally recognized quality certification (Figure 2-39). This puts 0% Turkey ahead of other middle-income Poland Bulgaria Turkey Brazil Romania Turkey Turkey Hungary Czech 2009 2009 2005 2009 2009 2008 2008 2009 Rep. countries, such as Brazil (26 percent), (manuf) (all) (manuf) 2009 Bulgaria (20 percent) and Poland (17 percent). Source: Enterprise Surveys

Figure 2-40: Share of firms with internationally 2.81. Certifications among small firms recognized quality certifications, by region lag far behind medium and large 47% companies. 55 percent of large firms hold a 50% 46% 45% 40% quality certification, which is three-fold of 35% 40% 33% 30% 31% the share of small firms. Nevertheless, Figure 30% 27% 22% 2-39 shows that small firms in Turkey seem 20% to be doing much better than small firms in for instance Brazil, Bulgaria and Poland. 10% Both foreign-owned companies (55 percent) 0% and firms that export (52 percent) tend to Aegean Marmara South Black Sea - Central have such certificates to a significantly larger Eastern Anatolia Total Manufacturing industry extent than domestic (29 percent) and non- exporting firms (24 percent). Regionally, a Source: Turkey ES 2008 larger share of quality certified enterprises can be found in Central Anatolia (40 percent). The share of firms with certificates in the Aegean region is on the other hand only 22 percent (Figure 2-40).

2.82. The industry variations in Turkey confirm that sectors with higher technologies tend to hold quality certifications more often than traditional lines of business. For instance, 90 percent of firms active in the electronics industry hold internationally recognized quality certifications, while the corresponding numbers in for example garments and textiles are 18 and 28 percent. When interpreting the survey data for Turkey it is however important to keep in mind that the firms‘ answers reflect not only formally recognized quality certifications, but can also include expired certificates as well as certification bodies not accredited by TÜRKAK, the national accreditation body. The market for certification services is in general relatively broad in Turkey but the quality of the certifications is not always strictly examined. Many smaller certification bodies are subsidiaries of European bodies accredited in their home market and seem to be approving certificates more quickly than the standard time for review and audit, thus creating an anticompetitive environment.

50 Information and communication technologies

2.83. Turkish firms continue to increase the use of ICT in their operations. The enterprise survey measures ICT among firms by the level of e-mail and website usage in communication with clients and suppliers. The results suggest that both ICT factors have increased since 2005, with 75 percent of Turkish firms having their own websites and as many as 89 percent communicating per e-mail.

Figure 2-41: Use of websites in communication with clients 2.84. The level of internet use is and suppliers, country and small firm comparison influenced by firm size as well as type. As expected, large and medium firms use 80% 75% 75% 75% 76% 76% All firms 68% 68% 66% ICT more extensively than small firms, 61% 65% Small firms 56% 56% although there has been a ―catch up‖ by 60% 48% 47% 41% SMEs to the level of larger enterprises. 35% 40% 29% This catch up is seen more clearly when comparing small enterprises in Turkey to 20% small firms in other emerging economies. As was the case with quality 0% certifications, small firms in Turkey seem Romania Bulgaria Turkey Poland Hungary Turkey Brazil Turkey Czech 2009 2009 2005 2009 2009 2008 2008 Rep. to use ICT to a larger extent than in for (manuf) (all) (manuf) 2009 example Brazil and Poland (Figure 2-41). By ownership form and exporting status, it is evident that foreign firms and Figure 2-42: Use of ICT in communication with clients and exporters are more advanced in their suppliers, by size, ownership form and exporting status application of internet. Especially in the

100% 96% 89% share of firms having their own website 85% 93% 93% 95% 92% 87% 91% 86% do non-exporters seem to be lagging 82% 75% 80% 68% 70% behind (Figure 2-42). 60% 2.85. There are regional and sectoral 40% differences in e-mail and website usage, 20% with firms the Aegean region using e-mail

0% in business communication the most. Interestingly enough, companies in the same region seem to have the least spread of website utilization. Finally, when E-mail Website Sources: Enterprise Surveys and Turkey ES 2008 looking at industry variations, it appears that the electronics and metals sectors in Turkey are very advanced in their use of ICT, while less technology-intensive industries, such as food, garments and textiles are well below the overall country average in both e-mail and website usage.

2.86. With the implementation of the e-Transformation Turkey Project, the expansion of ICT in public services has received a boost. The Government has also stepped up initiatives to raise awareness among citizens and businesses. Further support is planned for enterprises in their use of ICT as well as increased competition in the electronic communication sector. The overall reform initiatives by the Turkish Government are summarized in Box 2-7. One can see that the Ninth Development Plan 2007- 2013 among others includes ambitious objectives in the rise of internet usage in Turkey.

51 Box 2-8: Technology and innovation reforms The new R&D law in Turkey, implemented in March 2008, presents an extensive incentive and fiscal policy program, with the objective of reducing the R&D gap and encouraging economies of scale in the sector. Main features in the R&D law include (i) tax exemptions for R&D personnel; (ii) R&D and innovation expenditure discounts; (iii) social security payment deductions for R&D personnel; (iii) stamp tax exemption; and (iv) techno- entrepreneurship capital support for eligible recent university graduates.

The Government‘s comprehensive research, technology and innovation objectives are described in the Ninth Development Plan 2007-2013 (exact goal as share of GDP), the Medium Term Program 2010-2012, the National Science and Technology Strategy 2005-2010, TÜBITAK‘s National Innovation Strategy 2008-2010 as well as YOİKK‘s Action Plan for 2009.

Concrete policy goals in the area of R&D, stretching until 2013, are presented in the Ninth Development Plan and cover: (i) increase in the share of R&D expenditure from 0.7 to 2 percent of GDP; (ii) raise in the number of full time researchers from 24,000 to 80,000; (iii) upsurge in the share of R&D expenditures financed by the private sector from less that 29 to 60 percent; (iv) increase in the broadband subscriber penetration rate from 3.5 to 20 percent; as well as (v) improvement in the internet user penetration rate from 20 to 60 percent.

The YOİKK Technical Committees for Intellectual and Industrial Property Rights (IIPR) as well as for R&D have presented short-term goals for IIPR, innovation and technology in the 2009 Action Plan. Amendments to the Patent Law, Patent Authority Law and Trademark law will be drafted for harmonization with the EU Acquis and international regulations. Improvements to the university curricula on the subject of IIPRs and advancement of the qualifications and skills of IIPR courts and law enforcers are also proposed. The R&D Committee is also preparing a legal amendment, which will simplify the process of employing foreign research staff in Turkish firms and R&D institutions.

Strong emphasis has also been put on better university-industry cooperation. The Government is planning to establish Technology Development Regions, bringing together academic and private sector establishments as well as Technology Transfer Centers, conveying new technologies established at universities to the industry. Cooperation between the public and private sector in regards to IIPR and copyrights will also expand through Coordination Councils.

2.5 Access to Finance and Corporate Governance 2.87. The econometric analysis from the survey confirms the positive association of access to finance and corporate governance with productivity. There is much empirical evidence supporting the positive correlation between finance, productivity and growth, corroborating the idea that an efficient financial system helps firms to access credit and thus increase their investment and productivity.37 A sound financial system promotes economic growth by facilitating firms‘ access to external financing, allowing them to become more productive and expand.38 The analysis, as summarized in Table 2-10, indicates that higher productivity in 2008 was related to several variables representing financial soundness. Higher levels of productivity were observed in firms with higher share of sales paid for before delivery and with the ability to finance a higher proportion of fixed assets purchases with internal funds. It is also notable that firms are less productive when financing new investments through state-owned banks. Also, apparently, firms in receipt of state financial support were more productive. The level of ownership concentration also has a positive correlation with productivity levels.39

37 See Levine (2005) for a comprehensive survey of the theoretical and empirical literature. 38 See for instance Demirgüç-Kunt and Levine, 2008 39 In the analysis of 2005 data, the only significant variable in the finance and corporate governance block was the use of external auditory, with a positive effect on TFP. This variable was not significant in 2008 regressions.

52 Table 2-10: Summary of the effects of finance and corporate governance

Dependent variable Probability Explanatory IC variable Probability of Productivity Employment of exporting receiving FDI Sales paid before delivery + Purchases paid after delivery + New fixed assets finance - internal funds + New fixed assets finance - equity + New fixed assets finance - state banks - - Loan (dummy) + Land/buildings as collateral (dummy) + Subsidies (dummy) + Largest shareholder + External audit (dummy) + + Source: Staff calculations based on Escribano et al. 2009. See Tables in Annex 2-A for detailed results.

2.88. Financing structure presents a strong association with employment and exporting, while a more limited relationship emerges with foreign ownership. Larger firms are more likely to pay sales after delivery – with effects on the availability of working capital. Confirming the findings of the 2005 survey, employment levels are also positively associated with access to a bank loan and with externally audited accounts, which, in turn, favors the likelihood to access bank finance. By virtue of greater availability of internal and external financial resources, larger firms also less likely to finance their purchases of fixed assets with recourse to finance from a state bank. The probability of exporting is positively associated with the use of immobile assets, i.e. land and buildings, as collateral and with the ability to finance the purchase of fixed assets with new equity. Finally, there is a positive correlation between the level of FDI and firms‘ having their financial statements verified by an external auditor.

Overview of the financial sector

2.89. Despite the recent financial Figure 2-43: Access to finance as an obstacle , country and turmoil, Turkish firms‟ perception of small firm comparison access to finance improved 60% 56% significantly in the period 2005-2008. 50% All firms 47%49% The continued recovery of the economy Small firms 40% after the 2001 crisis, extensive 40% 37% privatization activities and more sound 22% 24% fiscal policies had a positive impact on 18% 22% 21% 16% 17% the enterprise sector. Turkey is also 20% 13% 16% 14% 12% 13% experiencing growing domestic credit to private enterprises and a transition from a 0% state-dominated financial system to more Hungary Turkey Turkey Bulgaria Poland Czech Romania Turkey Brazil private sector involvement. Improved 2009 2008 2008 2009 2009 Rep. 2009 2005 2009 access to finance for firms is today one of (manuf) (all) 2009 (manuf) the Government‘s main objectives set up Source: Enterprise Surveys in the Ninth Development Plan 2007- 2013, under the Increasing Competitiveness axis. The positive change in opinion can be noticed in the survey firm responses, where the share of firms considering access to finance to be a major obstacle to

53 operations had gone down from 47 to 13 percent in the measured periods (Figure 2-43). The financial system is also much less of an obstacle in Turkey than in comparative countries, such as the Czech Republic, Romania and Brazil. The data also indicate that smaller enterprises tend to be obstructed by financial factors to a larger extent than large and medium firms.

2.90. Nonetheless, the financial sector in Turkey is yet to reach international development standards. According to a recent study by the Banks Association of Turkey, the ratio of financial assets to GDP in 2007 was 150 percent in Turkey, while the responding share for emerging Figure 2-44: Turkish banks by ownership, 2005 and 2009 economies was 246 percent and the global average was 421 percent. Furthermore, the 60% Turkish financial sector remains dominated 36.2% 37.8% 40% 31.1% by banks, which amount to a total of 45, of 27.7% 29.8% which 32 are depository banks. The change 24.4% from 2005 has mainly consisted in a shift 20% 6.7% from domestic to a majority of foreign- 6.4% owned banks (Figure 2-44). Banks 0% contributed with nearly 78 percent of total State-owned Private, Foreign banks Development assets in the beginning of 2009, as seen in banks domestic banks and investment banks and SDIF Table 2-11. The Central Bank‘s assets 2005 2009 controlled banks amounted to 12.8 percent, while 3 percent and 1.8 percent came from securities funds Source: Banks Association of Turkey and financial leasing companies respectively.

Table 2-11: Structure of the financial system in Turkey, 2002-2009

2009 % Distribution (TL Billion) 2002 2003 2004 2005 2006 2007 2008 /03 Mar 09 CBRT 74.1 76.5 74.7 90.1 104.4 106.6 113.4 122.7 12.8 Banks 216.7 255.0 313.8 406.9 499.5 581.6 732.8 754.2 78.7 Financial Leasing Companies 3.8 5.0 6.7 6.1 10.0 13.7 17.2 16.9 1.8 Factoring Companies 2.1 2.9 4.1 5.3 6.3 7.4 7.8 7.2 0.8 Insurance Companies 5.4 7.6 9.8 14.4 17.4 22.1 26.5 14.8 1.5 Pension Companies 0.0 3.3 4.2 5.7 7.2 9.5 12.2 12.9 1.3 Securities Investment Funds 9.3 19.9 24.4 29.4 22.0 26.4 24.0 28.7 3.0 Other1 2.7 3.5 4.2 7.8 9.1 12.3 13.8 13.9 1.4 Total 314.1 370.4 437.7 560 668.6 770.1 935.5 958.4 100 Source: BRSA Financial Markets Report, Mar 2009 1Other include Consumer Finance Companies, Securities Intermediary Institutions, Securities Investment Partnerships and Real Estate Investment Partnerships

2.91. Market capitalization on the Istanbul Stock Exchange (SE) has been hit by the financial crisis limiting the possibilities of non-bank finance for the Turkish business sector. Market capitalization amounted to 15 percent of GDP at the end of 2008 ranking Turkey very low in comparison to other emerging economies (Figure 2-45). Nevertheless, up until one year prior, the Istanbul SE has experienced a steady growth, with a 10 percentage point cumulative increase in 2005-2007. Figure 2-46 illustrates that a majority of the 295 firms listed on the stock exchange today are manufacturing companies, followed by the financial sector with 33 percent, the two sectors best served by the exchange. The development since the 2001 economic crisis indicates an increasing interest in the Turkish stock exchange and motivates improved corporate governance, stronger regulations and market-related information. These types of financial intermediation are becoming increasingly important in today‘s volatile status of the global economy.

54 Figure 2-45: Market capitalization of firms listed on the Figure 2-46: Firms listed on the ISE, by Istanbul SE/GDP, 2005, 2007, 2008 and comparator countries industry, 2009

50% 44% Other 37% services, Technolog 40% 33% 5.4% y, 4.1% Wholesale 30% 22% , retail, 17% 18% hotel and Manufact 20% 15% 12% restaurant uring, s, 5.8% 51.9% 10%

0% Financial Hungary Turkey Poland Ireland Mexico Turkey Brazil Turkey institution 2008 2005 2007 s, 32.9%

Source: WFE and WDI Source: Istanbul Stock Exchanges

Box 2-9: The Turkish banking sector in the wake of the crisis The Turkish banking sector has proven resilient to the effects of the global credit crisis. The system remains highly-capitalized and profitable, despite a nontrivial deterioration in asset quality. Unlike in other emerging market economies, the Turkish banking system relies on a strong and stable domestic deposit base for funding, as reflected in relatively low loan/deposit ratio of 126 percent, compared to a regional ECA average of 171 percent. Wholesale funding, including syndications and securitizations, is relatively low at about 15 percent of non‐equity liabilities. The performance of the banking sector has been aided quite significantly by prompt and effective policy action on the part of the government and the monetary authorities to maintain stability in financial markets. Specifically, profitability has been supported in 2009 by a rapid widening of net interest margins as the policy rate has been cut, enabling banks to reduce their deposit costs and compensate for higher provisioning charges. Capital adequacy was relatively high at 19 percent as of end‐June 2009.

Nevertheless, as a result of the global credit crisis and the ensuing uncertainty, lending volumes have stalled and exposure to government securities increased. Since end‐2008, commercial bank loan volumes have been flat, whereas securities holdings increased by 14 percent (31.5 percent of total assets as of end-December 2009 compared to 26.5 percent end-2008). Lending to SMEs, a sector that had been growing significantly since 2005, has actually dropped in relative terms, now accounting for less than 22 percent of total lending, compared to almost 24 percent at end-2008. In this connection, while lending rates, after peaking in October 2008, have been on downward trend, there is now evidence of increased segmentation in the market, with rates charged to SMEs clients exceeding by far those offered to larger corporate clients. So far, the performance of corporate and commercial loans and mortgages has held up relatively well, whereas NPLs on SMEs, credit cards and unsecured consumer loans has grown considerably. The NPL ratio for the system as a whole was 4.9 percent at end-June 2009 (compared with rates around 3 percent before the crisis).

2.92. Bank credits to the private sector are increasing, but lag behind other emerging economies. After the record low of 12 percent of GDP in the aftermath of the 2001 crisis, claims on the private sector reached 32 percent in 2008 (Figure 2-47). Compared to other economies however, such as Russia, Brazil and Chile, Turkish credit to the private sector remains relatively low.

55 Figure 2-47: Claims on the private sector, % of GDP, 2008

100% 83.9% 80%

60% 50.4% 41.9% 40% 31.6% 25.0% 17.3% 20%

0% Mexico Turkey 2005 Turkey Russia Brazil Chile

Source: IMF International Financial Statistics

2.93. This growth in credit to the private sector is however increasingly targeted towards individual consumers rather than enterprises, especially SMEs. Figure 2-48 shows that consumer credits have surged by more than 700 percent Figure 2-48: Private sector credit by recipient (biannually, since 2004, compared to firm loans that grew billion TL) by 400 percent, the latter in fact showing a 100 decline since the beginning of 2009. This Consumer Loans could be a reflection of the improvement of 80 Firm loans lending technologies to consumers but at the 60 same time, the lack of adequate institutions for credit issuance to SMEs have slowed the 40 expansion of firm loans. Furthermore, 20 entrepreneurs with small startup that have trouble receiving firm loans can receive 0

credit for their business by instead using

05 06 07 08

04 05 06 07 08

05 06 07 08 09

- - - -

- - - - -

- - - -

- private immobile assets as collateral. This

Apr Apr Apr Apr Apr

Dec Dec Dec Dec Dec

Aug Aug Aug Aug would also as a result create a bias towards Source: Central Bank of Turkey consumer loans.

Figure 2-49: Share of firms with loans, country and small firm Figure 2-50: Share of firms with loans, comparison by region

80% 69% 70% 61% 62% All firms 65% 60% 70% 65% 55% Small firms 57% 57% 60% 60% 50% 50% 49% 46% 47% 48% 46% 50% 50% 40% 43% 40% 43% 38% 35% 42% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Bulgaria Hungary Turkey Czech Poland Turkey Turkey Brazil Chile Black Sea Central Marmara South Aegean 2009 2009 2005 Rep. 2009 2008 (all) 2008 2009 2006 - Eastern Anatolia (manuf) 2009 (manuf)

Source: Enterprise Surveys Source: Turkey ES 2008

2.94. The survey data show that the share of Turkish firms that do have loans is ahead of comparator countries, even when consider smaller firm sizes. Although the share of small enterprises

56 with loans (50 percent) is smaller than that of medium and large firms, amounting to 68 and 62 percent respectively, this is still better than in other comparative emerging economies, as for instance the Czech Republic, Poland and Brazil (Figure 2-49). There are significant differences in firms with loans measured by region, as depicted in Figure 2-50. 38 percent of enterprises in the Black Sea – Eastern region have loans while the corresponding number for firms in the Aegean region is 62 percent.

2.95. Turkish firms rely more on bank loans for investment financing than do firms in other countries but internal funds remain the main source of finance for investments. Internal funds and bank loans are by far the most important financial sources for purchase of investments, amounting together to 93 percent of financing. As shown in figure 2-51, more than half of the funds are internally generated, while issuance of new equity shares and advances from customers play a smaller role for firms funding.

Figure 2-51: Sources for finance of fixed assets, country comparison

100% Other 80% Advances from 60% customers

Bank loans 40%

New equity 20% shares issued

0% Internal funds Brazil Chile 2006 Turkey Poland Hungary Bulgaria Czech Rep. Russia 2009 2008 2009 2009 2009 2009 2009

Source: Enterprise Surveys

2.96. In spite of the severe effects of the crisis on Turkey, its private sector firms continued to have access to international borrowing. Corporate rollover of foreign debt amortizations fell below 100 percent in December 2008 for the first time since January 2007. However, firms continued to roll over the bulk of their debt service, with drawings tracking repayments quite closely month on month. Over the first nine months of 2009 the rollover ratio of the non-banking corporate sector‘s external debt was just over 70 percent.

2.97. Recent actions in facilitating credit access, in particular to SMEs, have been initiated, among others through increased funding to the public guarantee fund. As discussed in greater detail in Chapter 3, the Credit Guarantee Fund of Turkey (CGF) supports SMEs‘ access to credit, and through planned legislative action funding of CGF will be increased by an additional TL 1 billion in the next two years for leveraging additional resources for SMEs. The Government is mitigating risk-averse behavior in the banking system by limiting theses guarantees to 65 percent of loan amounts and sharing the burden with the Treasury and through participation of commercial banks for the guarantees. Funding to the Small and Medium Industry Development Organization (KOSGEB) has also been increased by 48 percent for support of existing credit subsidies and technical support programs. Moreover, the credit line with the CBRT of Turkey‘s Eximbank, which provides export credit guarantees for Turkish exporters was increased to USD 1 billion in 2008.

2.98. Planned cooperation between the Capital Markets Board and the Istanbul Stock Exchange could reap benefits for SMEs in future capital markets inclusion. Studies are underway to establish

57 the ISE Emerging Companies Market (ECM),40 as a projected collaboration between ISE and CMB. Companies that have been registered with the CMB but have not fully satisfied the listing requirements of ISE will be able to trade on the exchange. The initiative aims to create a transparent and organized floor where SMEs with growth potential can secure funding from the capital markets to lower registration and transaction costs and with more lenient auditory requirements. A market advisory mechanism will be created in order to provide firms with information on capital markets and exchange regulation, both prior to and during the application process, as well as regulatory support after admission. The technical and regulatory preparations of the ISE ECM are currently ongoing.

Credit Information

2.99. Turkey‟s two large credit bureaus are the Credit Bureau of Turkey (KKB) and the Credit Registry of the Central Bank. KKB is owned by nine banks and shares information among 38 financial institutions. The bureau has in a short time increased its coverage, from 27 percent in 2005 to 43 percent in 2009 (Figure 2-52). The Credit Registry of the Central Bank covers both individuals and firms, with information collected from both bank and non-bank financial institutions. However, the Credit Registry‘s coverage, as seen in Figure 2-53 is albeit improvements lower than KKB. Moreover, the Banks Association of Turkey is expected to outsource the technical operation to KKB in order to expand coverage of debtors (HEG-DPL PD 2009). The KKB has established the framework to expand beyond coverage of individual consumer credit information to also cover the corporate sector. Although this initiative has not yet been implemented, the current monitoring of individual consumers is also very useful, especially for information on small enterprises. Since entrepreneurs, as earlier mentioned, do not face easy approval of corporate loans, they often finance their investments through personal loans, and the credibility of the owner is thus often reflected on the credibility of their operations. As discussed in Chapter 3, continued progress in enhancing credit information is crucial to facilitate access to bank credit, especially for SMEs.

Figure 2-52: Private credit coverage, percentage of Figure 2-53: Public credit coverage, percentage of adults adults

73% 80% 68% 40% 33% 35% 59% 60% 30% 24% 43% 34% 16% 40% 27% 30% 20% 6% 20% 6% 10% 10% 5% 5% 0% 0% 0% 0%

Source: Doing Business 2010

Financial Reporting

2.100. Use of external auditors by Turkish firms has increased notably since 2005. 55 percent of the firms have their financial accounts externally audited, compared to 42 percent in 2005, showing a favorable international comparison (Figure 2-54). This has been reflected in an improvement in access to bank finance for firms and aided credit providers in their quest to assess firms‘ abilities in meeting credit

40 The legal background for the ECM initiative has been published on August 18 2009, as the ―Emerging Companies Market Regulation‖ in the Offical Gazette, no. 27323. More information can be found on the CMB website http://www.spk.gov.tr/duyurugoster.aspx?aid=200994&subid=0&ct=c&submenuheader=null (Turkish)

58 obligations. As is often the case in comparable economies, small and medium sized firms in Turkey use external auditors to a lesser extent than larger firms.

Figure 2-54: Firms having financial accounts externally audited, country comparison

80% 69% 64% 60% 55% 55% 47% 42% 37% 40% 28% 24% 26% 20%

0% Poland Brazil Bulgaria Czech Turkey Chile Turkey Turkey Argentina Hungary 2009 2009 2009 Rep. 2005 2006 2008 2008 (all) 2006 2009 2009 (manuf) (manuf)

Source: Enterprise Surveys

2.101. Audit reforms introduced in the past years in Turkey, have had a positive initial influence. Such reforms as required use of International Financial Reporting Standards and International Standards of Audit have had a strong impact on the increased number of externally audited financial accounts. Since October 2006, the Turkish Accounting Standards Board (TASB) has been publishing the Turkish Financial Reporting Standards (TFRS), based on the above mentioned IFRS. The TFRS is currently solely directed by such regulatory agencies as the Capital Markets Board (CMB) and the Banking Regulation and Supervision Agency (BRSA), but is expected to be applied in a larger scale with the ratification of the Commercial Code (CC), which has been pending in Parliament since 2005. The draft CC also stipulates external auditing of all firms, including those not covered by CMB and BRSA. As argued in Chapter 3, swift adoption of the Commercial Code with its provisions regarding the wide adoption of financial reporting standards will be beneficial to SME borrowing.

Use of Collateral

Figure 2-55: Share of firms for which collateral was required for their 2.102. About two thirds of latest loan, country comparison firms in Turkey provided collateral for their latest 83% 85% 90% loan in 2008, using mainly 73% 76% 80% 71% personal and immobile 70% 63% 65% assets. It is well known that 60% 47% 49% 50% firms that are able to commit 40% 32% collateral obtain larger loans 30% and on more beneficial terms 20% than firms without collateral. 10% Collateral not only provides 0% the creditor additional Brazil Chile Turkey Poland Turkey Turkey Czech Romania Bulgaria Hungary security in the case that the 2009 2006 2005 2009 2008 2008 Rep. 2009 2009 2009 loan cannot be repaid but (manuf) (all) (manuf) 2009 also improves the borrower‘s Source: Enterprise Surveys access to credit. Collateral

59 was used for 71 percent of loans in the manufacturing sector, an improvement from less than 50 percent in 2005 (Figure 2-55), placing Turkey on par with other middle income countries.

2.103. Meanwhile, the most common type of asset provided as collateral is still land and buildings, in fact gaining importance since 2005 (Figure 2-56). Figure 2-56: Type of collateral, 2005 and 2008 This could be a reflection in the numerous actions taken 2005 2008 61% to improve housing finance 55% and thus improving access to 43% 39% finance for firms pledging 31% 30% their real estate as collateral 19% 20% for smaller firms, where 15% entrepreneurs commit their 8% personal assets.41 These have become more common as use for collateral, while Machinery and Other Accounts Personal assets of Land and buildings equipment receivable and owner movables, which were the including movables inventories second largest group of Source: Turkey ES 2008 collateral in 2005 are today least significant. Considering that machinery and equipment were the most common asset type among firms in 2005, the shift away from this group may indicate a tightening of credit standards.

Corporate governance

2.104. Economies are increasingly recognizing the importance of corporate governance in achieving policy goals and improving growth and efficiency. The concept of corporate governance incorporates the relationship between a firm‘s management and its stakeholders, among others board, shareholders, employees, labor unions, customers and suppliers. It also determines the structure of a company‘s objectives and monitors overall business performance.42 Through its concept of accountability, corporate governance ensures that the interests of all stakeholders are taken into consideration by the management in its decision making process.

2.105. Strong corporate governance helps firms attract investment, improves private sector productivity and strengthens a country‟s financial stability. Through high levels of transparency and accountability in business procedures as well as improved monitoring of financial flows and firm performance, an economy can increase access to external financing by attracting investors; develop a more prosperous stock market, improve operational performance; reduce risk of financial crises; and build a better relationship of all stakeholders in a firm.

2.106. The development of Turkish corporate governance has been positive on the extent of disclosure, but remains poor in investor protection when considering director liability and the possibility of bringing shareholder suits. Doing Business measures the strength of minority shareholder protections against directors‘ misuse of corporate assets for personal gain. The indicators distinguish 3 dimensions of investor protection: transparency of transactions (extent of disclosure index), liability for

41 Turkey has in the past years taken several actions to promote housing finance. With the passing of a new mortgage law in 2007, the Government has taken an important step in the facilitation of the housing market. Banks are now allowed to charge floating rates on mortgages and to introduce prepayment penalties. A secondary regulation, allowing banks more flexibility in the design and securitization of mortgages as well as better risk management was issued in August 2007 (World Bank 2008). 42 See OECD Principles of Corporate Governance (2004) for a detailed overview in the subject as well as OECD Corporate Governance in Turkey (2006).

60 self-dealing (extent of director liability index) and shareholders‘ ability to sue officers and directors for misconduct (ease of shareholder suits index). The indices score each country on a scale of 1-10, where the highest represents the most favorable level.

Figure 2-57: Protecting investors index, scale 1-10 Table 2-12: Protecting investors index, scale 1-10

7 Extent 6.0 6.0 6.0 6.0 of Extent of Ease of 5.7 5.8 6 5.3 5.3 disclos director shareholder 5.0 5.0 Country ure liability suits 5 4.3 Hungary 2 4 7 4 Czech Rep. 2 5 8 3 Russia 6 2 7 2 Brazil 6 7 3 Turkey 2005 8.0 4 4 1 Turkey 9 4 4 0 OECD avg. 5.9 5.0 6.6 Bulgaria 10 1 7 Chile 7 6 5 Poland 7 2 9 Romania 9 5 4

Source: Doing Business 2010

2.107. The 2010 Doing Business Report, as illustrated in Figure 2-57, gives Turkey an average score 5.7. Table 2-123 depicts the Turkey has improved its level of disclosure, by requiring an external body to review business transactions before they are processed. The country also ranks well on immediately disclosing transactions to the public and/or shareholders through published periodic filings. However, Turkey does still not hold directors adequately liable, among others in the case of transactions that were unfair, oppressive or prejudicial to minority shareholders as well as when requiring beneficiaries to return excess profits realized through unfair transactions. Turkey also scores lower than most comparator economies in the possibilities for shareholders to pursue lawsuits. The low score comes from not allowing shareholders with less than 10 percent interest to request an inspector to investigate the transaction or inspect relevant documents before filing suit.

2.108. Several measures have been initiated recently to improve corporate governance among Turkish enterprises. The Capital Markets Board, responsible for monitoring and increasing transparency in Turkish capital markets, has introduced mandatory corporate governance compliance reports and requires the establishment of audit committees. Further improvements came through the accounting and auditing reforms discussed in the financial reporting section above. Moreover, rating principles for guarantee funds were brought up to international standards in 2008. The issue is widely covered by several recent Governmental policy publications, such as the Ninth Development Plan 2007-2013 and the YOİKK Action Plans 2009 and 2010. Upcoming reforms include legislative amendments on ―Director‘s Liability‖ and ―Ease of Shareholder Suits‖ as presented by Doing Business, the establishment of corporate reporting standards for firms not listed on the stock market, as well as preparation of a new corporate governance system for companies listed on the ISE.

61 References Banks Association of Turkey. 2009. The Banking System in Turkey. June 2009.

Blind, Knut and Andre Jungmittag. 2005. "Trade and the impact of innovations and standards: the case of Germany and the UK." Applied Economics, Taylor and Francis Journals. vol. 37(12)

Cameron, A.C. and P.K Trivedi, 2005. ―Microeconometrics: Theory and Applications‖. Cambridge University Press.

Conway, Paul, Donato De Rosa, Giuseppe Nicoletti, and Faye Steiner. 2006. ―Regulation, Competition and Productivity Convergence,‖ OECD Economics Department Working Paper, No. 509. Paris.

Demirgüç-Kunt, Asli and Levine, Ross. 2008. ―Finance, Financial Sector Policies, and Long-Run Growth.‖ The World Bank. Policy Research Working Paper No. 4469

Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes and Andrei Shleifer. 2002. ―The Regulation of Entry.‖ The Quarterly Journal of Economics, Vol. 116

Djankov, Simeon, Caroline Freund and Cong S. Pham. 2008. ―Trading on Time.‖ Review of Economics and Statistics. Nov 2008.

Escribano A. and J. Luis Guasch. 2005. ―Assessing the Impact of the Investment Climate on Productivity using Firm Level Data: Methodology and the Cases of Guatemala, Honduras and Nicaragua‖. Policy Research Working Paper # 3621, The World Bank, June.

Escribano A. and J. L. Guasch. 2008. ―Robust Methodology for Investment Climate Assessment on Productivity: Application to Investment Climate Surveys from Central America‖. Working Paper # 08-19 (11), Universidad Carlos III de Madrid.

Escribano, A., J. L. Guasch, M. de Orte and J. Pena. 2008a. ―Investment Climate and Firm‘s Performance: Econometric and Applications to Turkey‘s Investment Climate Survey.‖ Working Paper # 08-20 (12), Universidad Carlos III de Madrid.

Escribano, A., J. L. Guasch, M. de Orte and J. Pena. 2008b. ―Investment Climate Assessment Based on Demean Olley and Pakes Decompositions: Methodology and Applications to Turkey‘s Investment Climate Survey.‖ Working Paper # 08-20 (12), Universidad Carlos III de Madrid.

Escribano, A., J. L. Guasch and J. Pena. 2008. ―Assessing the Impact of Infrastructure Quality on Firm Productivity in Africa‖ Mimeo.

Escribano, Alvaro, Manuel de Orte, Jorge Pena. 2009. ―Econometric Analysis of Turkish Investment Climate – Analysis based on firm level data of the manufacturing sector in 2008 and 2005.‖ Background paper. World Bank: Washington, DC.

European Commission. 2009. eGovernment Turkey Factsheet.

European Patent Office. 2009.

Griliches Z. and J. Mairesse. 1997. ―Production Functions: The Search for Identification.‖ In S. Strom (ed.) Essays in Honor of Ragnar Frisch, Econometric Society Monograph Series, Cambridge University Press, Cambridge.

International Organization for Standardization. 2009. The ISO Survey of Certifications 2008. Genève

62 Levine, Ross. 2005. ―Finance and Growth: Theory and Evidence.‖ Handbook of Economic Growth, in: Philippe Aghion & Steven Durlauf (ed.), Handbook of Economic Growth, edition 1, volume 1, chapter 12, pages 865-934 Elsevier.

Nicodeme, Gaetan and Jacques-Bernard Sauner-Leroy. 2007. ―Product Market Reforms and Productivity: A Review of the Theoretical and Empirical Literature on the Transmission Channels.‖ European Communities. Brussels

Magnus J. R. and A. L. Vasnev . 2007. ―Local Sensitivity and Diagnostic Tests‖. Econometrics Journal, V. 10, 166- 192.

Marschak J. and W. H. Andrews. 1944. ―Random Simultaneous Equations and the Theory of Production‖. Econometrica, 12 (3,4), 143-205.

Olley, G. S. and A. Pakes. 1996. ―The Dynamics of Productivity in the Telecommunications Equipment Industry‖. Econometrica, Vol. 64, 6, 1263-1297.

Organization for Economic Cooperation and Development (OECD). 2004. OECD Principles of Corporate Governance. Paris

Organization for Economic Cooperation and Development (OECD). 2006. Corporate Governance in Turkey: A Pilot Study. Paris

Organization for Economic Cooperation and Development (OECD). 2007. ―Enhancing Turkey's Growth Prospects by Improving Formal Sector Business Conditions.‖ Economics Department Working Papers no. 542

Organization for Economic Cooperation and Development (OECD). 2009. Population and Employment Statistics

Organization for Economic Cooperation and Development (OECD). 2009. Science and Technology Statistics

Organization for Economic Cooperation and Development (OECD). 2009. ―Ten Years of Product Market Reform in OECD Countries - Insights from a Revised PMR Indicator.‖ Economics department working Papers no. 695

Republic of Turkey State Planning Organization. 2006. Ninth Development Plan 2007-2013. Decision No: 877.

Prescott, Edward C. (1998) "Needed: A Theory of Total Factor Productivity". International Economic Review, 39, 525-552.

Republic of Turkey State Planning Organization. 2008. 2009 Annual Programme. Decree Number: 2008/14200.

Republic of Turkey State Planning Organization. 2009. Medium Term Programme 2010-2012. Official Gazette No. 27351, 16.09.2009

Republic of Turkey. 2009. Technical Committee Action Plans of the Turkish Coordination Council for the Improvement of the Investment Environment (YOİKK).

Republic of Turkey. 2010. Technical Committee Action Plans of the Turkish Coordination Council for the Improvement of the Investment Environment (YOİKK).

Schafer, J, 1997. ―Analysis of Incomplete Multivariate Data‖. London: Chapman and Hall.

United States Patent and Trademark Office. 2009.

Venn, Danielle. 2009. ―Legislation, collective bargaining and enforcement: Updating the OECD employment protection indicators.‖ OECD Social, Employment and Migration Working Papers no. 89

63 World Bank. 2006. Report on the Observance of Standards and Codes (ROSC) - Accounting and Auditing. Washington, DC

World Bank. 2007. Turkey Investment Climate Assessment. Washington, DC

World Bank. 2008. ―Investing in Turkey‘s Next Generation: The School-to-Work Transition and Turkey‘s Development.‖ Report no. 44048 – TÜ. Washington, DC

World Bank. 2008. Program Document for a Second Competitiveness and Employment Policy Loan to the Republic of Turkey. Report No. 46436-TR Washington, DC

World Bank. 2008. Turkey Country Economic Memorandum – Sustaining High Growth: Selected Issues. Washington, DC

World Bank. 2009. Turkey National Innovation and Technology System Report. Washington, DC

World Bank. 2009. Programmatic Human Capital, Employment and Growth Development Policy Loan for Turkey. October 1, 2009. Concept Draft

World Bank, 2009, Turkey Innovation Policy for Competitiveness and Employment Generation. June 2009. Decision Meeting Draft

World Bank Doing Business. 2010

World Economic Forum. 2009. The Global Competitiveness Report 2009-2010. Geneva

World Intellectual Property Organization. 2009

Annex 2-A. Econometric Methods In the identification of the statistically significant investment climate effects on productivity and other performance measures, the main concern is to use the information contained in the enterprise survey to perform the estimation and, at the same time, control for all the contingent problems that may be encountered: measuring productivity, simultaneity, endogenous regressors, selection of the relevant model, as well as data quality issues such as missing values, outliers or measurement errors. This Annex briefly describes the methodology used and proposes solutions for a number of methodological problems. The methodology, together with complete results, is detailed in the background paper for this report by Escribano, de Orte and Pena (2009).

Summary of the econometric methods used to assess IC effects The econometric methodology consists of two steps:

a. Identification of statistically significant IC effects on productivity and other performance variables. For the identification of the statistically significant IC effects on economic performance, analysis uses a simultaneous equations system that relates the interactions between the investment climate (IC) with productivity, demand for labor, exports and FDI inflows. Estimation always controls for firms‘ size, region and sector and yields elasticities and semi-elasticities of investment climate variables with respect to productivity, employment, wages, export propensity and FDI propensity. The IC elasticities and semi- elasticities provide a measure of the sensitiveness of outcome variables when the IC changes marginally.

b. Evaluation of relative IC contribution to aggregate (weighted average) productivity. Analysis also evaluates the IC in terms of the Olley and Pakes (O&P) decomposition of aggregate productivity (or weighted average using as weights the share of sales) into average productivity and allocative efficiency.

64 Estimation of IC effects on productivity and other performance measures In the system of equations investment climate information (IC) is used as proxy of idiosyncratic firm level differences. We assume that cross-sectional TFP is determined by a wide set of firm level attributes such as location, sector of activity and firm size (all represented by dummy variables included in the matrix D),  TFP by other firm level attributes (ai) normally unobserved by the econometrician, and by an error term ( i ). In this case we use the information of the survey to approximate ai and therefore can represent firm level productivity processes as follows:

logTFP aDTFP    TFP iPi D i i (1a)

aTFP IC TFP v TFP i IC i i (1b)

Similarly, the demand for labor is assumed to be determined by firm level productivity (logTFPi), by real wages in logs (logWi), by firm level differences approximated by the IC, sector/location/size information  L (D) and by an error term ( i ): logL=  aLL   log TFP   logW   D   i L i P i w i D i i (2a) aLLL  IC v i L i i (2b)

The probability of firms entering the export market is described by the next pair of equations yExp a Exp  log TFP D  Exp i Exp i P i D i i (3a) aExp IC Exp v Exp i IC i i (3b)

Finally, the probability of receiving foreign direct investment (FDI) is given by yFDI a FDI  log TFP D  FDI i FDI i P i D i i (4a) aFDI IC FDI v FDI i IC i i (4b)

y Exp yFDI Note also that i and i are used as covariates in all the equations as both are included in the matrix r IC, within the group of other control variables. On the other hand, since the variables y it, with r = Exp or P(/)(/) yrr1 x E y x FDI, are binary random variables taking 0 and 1 values, it it , the conditional probability is equal to the conditional expectation which is usually assumed to follow a Probit or a Logit model. In general, linear probability models (LPM) approximate well Probit and Logit nonlinear models when the variables are evaluated close to their sample means. Since we are interested in the mean IC contribution relative to the mean values of the dependent variables of (1a) to (4a), we will concentrate only on linear probability specifications, like (3) and (4).

The estimation of the productivity (TFP) equation is at the core of this econometric process of estimating the ,  ,  ,  relationship of the investment climate with the firm´s economic performance ( IC L IC IC ). We follow a sequential procedure in which we first estimate the IC effect on TFP and then we proceed by estimating the remaining equations in the system.

65 The Productivity equation derives from a structural system of equations, such as:

logY log L   log M   log K log TFP i L i M i K i i (5a)

logTFP aDTFP    TFP iPi D i i (5b)

aTFP IC TFP v TFP i IC i i (5c)

Where, (5a) is the production function, (5b) relates productivity with firm/sector/region/size fixed effects and (5c) is a formulation for firm-level fixed effects in terms of the investment climate. Y is firm output (sales), L is employment, M denotes intermediate materials, K is the capital stock, IC and C are time-fixed effect vectors of other investment climate and control variables and D are the vectors of state, industry and size dummies. Since the current literature does not provide a univocal theory of TFP (Prescott, 2004), TFP obtained as a residual from the production function (5a) is a ―black box‖ that may contain any factor that affects the way firms transform inputs into outputs. Against this backdrop, the approach implicit in system (5) is to use the information contained in the enterprise survey to empirically test TFP. That is, the initially undefined TFP measure is filled with enterprise level data reflecting the way firms transform inputs into output, as well as the factors underlying differences observed between more and less efficient firms. The robust IC elasticities and semi-elasticities with respect to productivity are obtained from the following regression equation, obtained by combining (5b) and (5c):

logTFP IC  D    uTFP i IC i D i P i (6) and are reported in Tables Annex 2-A-2 and 2-A-6. Parameter estimates are interpreted as the average individual IC effects on productivity (TFP), after controlling for the other IC and control variables. Whereas a causal interpretation is not possible, the robustness of empirical results allows interpretation of the estimated coefficients.

Summary of econometric issues in productivity estimation Econometric issues in the productivity equation:  No single salient productivity measure Since there is no single salient measure of productivity, any empirical evaluation of the productivity impact of the IC may critically depend on the particular productivity measure used. o Solution : Escribano and Guasch (2005, 2008) – following the literature on sensitivity analysis of Magnus and Vasnew (2006) – suggest looking for empirical results (elasticities and semi-elasticities) that are robust to several productivity measures. The different productivity measures derive from: i) different functional forms of the production function: Cobb-Douglas and Trasnlog; ii) different sets of assumptions (technology and market conditions) yielding two different approaches to estimate the system: the two step estimation by applying the Solow residual to estimate firm level productivity and the single step estimation by parametric techniques of; iii) different levels of aggregation in measuring input-output elasticities of equation (6) (at the industry level or at the aggregate country level). The productivity measures used in this report are summarized below:

Table 2-A-1 Functional forms of Estimation Aggregation level of Result production function procedure coefficients of PF Two-step 1.1 Unrestricted coefficients 2 (P) measures; 2 (IC) 1. Solow´s Residual estimation 2.2 Unrestricted coefficient elasticities Single-step 2.1 Restricted coefficient 2 (P) measures; 2 (IC) 2. Cobb-Douglas estimation 2.2 Unrestricted coefficient elasticities Single-step 3.1 Restricted coefficient 2 (P) measures; 2 (IC) 3. Translog estimation 3.2 Unrestricted coefficient elasticities 6 (P) measures; 6 (IC) Total elasticities

66  Endogeneity of the inputs. There is an identification issue separating TFP from the production function (PF); when a PF input is influenced by unobserved common causes affecting productivity – such as a firm‘s fixed effects – there is a simultaneous equation problem in equation (6) and hence in the single step estimation procedure. o Solution : To address this well-known problem (Marschak and Andrews, 1944, and Griliches and Mairesse, 1995) analysis follows the approach proposed by Escribano and Guasch (2005, 2008). That is, the usually unobserved firm-specific fixed effects (which are the main cause of inputs‘ endogeneity) are proxied by a long list of observed firm-specific fixed effects deriving from the enterprise survey. Controlling for the largest possible set of IC variables and plant characteristics (C), under standard regularity conditions, estimation yields consistent and unbiased least squares estimators of the parameters of the PF and the IC elasticities.  Endogeneity of IC variables: For consistency in estimation, the error term must be uncorrelated with any variable contained in the IC vector. It can be argued that the error term may contain alien, unmeasured effects correlated with IC, thus rendering the OLS estimator of IC effects inconsistent. o Solution 1: Correction for observable fixed effects; by using the full set of information contained in the IC variables we are able to control for more than 130 variables in the estimation, eliminating a large degree of endogeneity and spurious correlations as expectation of the outcome variables is conditioned on as much information as possible. o Solution 2: In spite of the observable fixed effect correction, the exogeneity condition does not hold for all the variables in the model. In this case industry-region (or industry-region-size) counterparts of firm level IC variables are used, which is equivalent to applying a general IV estimator.  Selection of the relevant model. The population model is unknown and needs to be approximated based on a broad set of more than 130 variables, including IC and other controls. o Solution: The econometric methodology applied for the selection of the variables goes from the general to the specific. Otherwise, an omitted variables problem would generate biased and inconsistent parameter estimates. Estimation proceeds by removing from the regressions the less significant variables one by one, until a final set of variables is obtained, all significant in at least one of the regressions and with parameters varying within a reasonable range of values.  Heteroskedasticity in the error term. o Solution: the heteroskedasticity of the error is addressed by using robust (White) standard errors. In response to the fact that data was collected using a random sampling by clusters, cluster standard errors are also computed, allowing for correlation within industry and region.

Summary of econometric issues regarding data quality Although enterprise surveys are valuable instruments improving our understanding of the investment climate factors affecting economic growth, particularly in emerging and transition economies, at the same time they present a number of issues related with the quality of the information provided; measurement errors, outlier observations and missing data are frequently found in this datasets. The enterprise survey of Turkey is not an exception.

Econometric analysis uses the subset of the ES sample (1152 firms in total) covering manufacturing firms. In order to ensure a sufficient representation of large establishments in the sample, a sampling approach which oversampled large firms was applied. The result is a sample with 903 manufacturing establishments. Due to the stratified sampling structure, use of proper weighting to correct for oversampling of large firms when doing descriptive analysis is advisable. However, the regression analysis uses un-weighted estimation based on the fact that stratification variables are controlled for in the estimation and that stratification is not based on the dependent variable of the regression.43

The first Turkey survey was performed in the summer of 2005 surveying 1,323 manufacturing establishments. The sampling structure was based in the same stratification variables used in 2008. Likewise, the sampling process was based on the Industry Database of the Union of Chambers and Commodity Exchanges of Turkey (TOBB). However, there are some differences in the industries included in 2008 and 2005, and in the percentage of establishments of each category, while the same regions are used in both surveys, and percentages of each region are almost unchanged in 2008. In 2008 135 IC variables are used, out of which 44 are common to 2008 and 2005, and the

43 See Cameron and Trivedi (2005) for more details.

67 remaining are available in 2008 only. Due to these differences some homogeneity in the methodology applied is lost. However, comparison of econometric results still allows identification of key differences in the way the investment climate affected firms in the two years. Econometric issues:  Missing data. The number of missing values reduces the sample available from 903 to 443 observations in the complete case (49 percent of the sample). Operating with the complete case is only acceptable if missing data comprise a small percentage, say 5% or less, of the size of the sample (Schafer, 1996), and preserves the representativeness of the original sampling frame. In models with a large number of regressors, missing data force the exclusion from regression analysis of explanatory variables with a high proportion of missing values. As Cameron and Trivedi (2005) point out, this practice may lead to an omitted variables problem. o Solution for production function variables: We impute those missing values with the objective of preserving the sample representativeness, gaining efficiency in the estimation and retrieving for the analysis a large number of costly interviews. Basically, the imputation mechanism replaces the missing values by the expectation of the variable conditional on the information we have on sector, size and region (see Escribano and Pena, 2008 for more details). To check robustness we also estimate the productivity equation under different imputation procedures and different assumptions on the missing data mechanism (MDM), see Escribano, de Orte, Pena and Guasch 2008. After the imputation we are able to use 768 observations, 85 percent of the sampling frame. o Solution for IC variables: Industry/region/size averages are a good solution also for the missing IC values. Nonetheless, using too many IC variables in average form introduces a high degree of multicollinearity to the regression. In addition, not all industry-region-size averages are good instruments of plant-level IC variables. As an alternative solution for the missing values in IC variables, we replace only the missing values by the expected value conditional on the information on industry, region and size.  Endogenous missing data mechanism (MDM). As Escribano and Pena (2008) signals, the missing data problem may also have important implications for the consistency of the IC parameters estimates when the pattern of missing values is determined by the IC variables. o Solution: We need to control for any IC variable correlated with the MDM to achieve consistency in the estimation of the system.  Outliers. o Solution: We exclude from the analysis outlier observations, i.e. observations with ratios of materials and/or labor cost to sales larger than one.

International comparisons: demean log-productivity Interpretation of TFP is assumed conditional on the understanding of firms‘ operating conditions. Any productivity measure is subject to measurement errors, unmeasured effects, differences in the deflators used, etc. To make cross- country comparisons based on IC impacts on productivity it is desirable create an index (demean productivity). After subtracting the mean (that is, the constant term, time effects, industry effects and country-specific effects) from firm level log-productivity we can concentrate on the part of log-productivity explained by the IC variables. Thus, demean log-productivity at the firm level is simply:

D TFP logTFPiˆ IC IC i (6)

Expression (7) is comparable across countries because the set of IC variables is very similar in all countries (with some slight modifications depending on the specific characteristics of each economy) and the same methodology is applied to select the set of significant IC variables. In addition, the O&P decompositions can be easily computed based on the demean portion of productivity allowing international comparisons of IC impacts on aggregate productivity.

Assessment of IC effects on the Olley and Pakes (1996) decomposition The Olley and Pakes (O&P) decomposition of aggregate productivity in logs is

Y logTFP log TFPN cov(ˆ sii ,log TFP ) (7)

68 Where logTFP is aggregate log-productivity (or weighted average productivity, where the weights are given by the shares of sales), logTFP is un-weighted average log-productivity and the last term is the covariance between share of sales and firm level productivity, or allocative efficiency term, describing the ability of the markets to reallocate resources from less to more productive establishments.

The useful additive property of equation (7) in logarithms, allows obtaining an exact closed form solution of the decomposition of aggregate log productivity. Following Escribano et al. (2008b), we can express aggregate log productivity as a weighted sum of the average values of two composite terms; a) the IC and dummy D variables, the intercept and the productivity residuals; and b) the sum of the covariances between the share of sales and IC, D and the productivity residuals:

TFP TFP Y TFP Y Y TFP (8) logTFPˆICi IC   ˆ Dipit D ˆ uuˆˆ NNN  ˆ IC cov(ˆ s ii , IC )   ˆ Ds cov(ˆ s ij , D )  cov(ˆ s ii , )

The contributions of IC variables to aggregate log-productivity of equation (8) can be computed for the whole sample, by industry/sector, by region, by firm size, etc. In particular, for international comparisons we compute the IC contributions relative to demeaned aggregate TFP as follows:

100 ˆTFP ˆ ˆ Y TFP (9) 100D [ICICN IC cov( s it , IC i )] logPq

Note that from (9): i. We can compare net contributions by isolating the impact of IC variables from the impact of industry dummies, the intercept, and the residuals; ii. We can differentiate the portion of aggregate productivity explained by IC, and C variables (demean logP), from the part attributable to the constant term, industry dummies, etc.; iii. We can carry out international comparisons of the effects of IC on aggregate productivity; iv. We can neutralize the different directions (positive or negative) of the various IC effects by taking the percentage contributions in absolute value; v. Finally, we can compute the absolute percentage contributions to the average log-productivity by blocks of IC variables or by regions, sectors or sizes.

IC impact on average employment and on the probability of exporting and of receiving FDI In order to evaluate the impact of the average IC variable on the sample average values of the dependent variables of the system, we substitute unknown parameters from the system (1)-(4) by their corresponding 2SLS estimated values. The labor demand equation evaluated at the sample means and expressed in relative terms is

100 100ˆ   ˆ logTFP   ˆ log W   ˆ ICˆ D L P W IC D log L . (10)

y Exp yFDI Since it and it are binary variables, evaluating the impact at the sample mean implies evaluating the probability (frequency) of exporting and receiving FDI, as follows: 100 100ˆ   ˆ logTFP   ˆ ICˆ D ˆ Exp P IC Ds P( Exp 0) (11) 100 100ˆ   ˆ logTFP   ˆIC  ˆ D ˆ FDI P IC D P( FDI 0) . (12)

69 Results of the econometric analysis

Table 2-A-2: Comparison of IC effects on TFP in 2008 and 2005 2008 2005 IC coefficient (IC % cont. on [IC % cont. on IC coefficient (IC % cont. on [IC % cont. on on TFP aggr. log-TFP) avg. log-TFP] on TFP aggr. log-TFP) avg. log-TFP] Regulatory Payments for power supply (dummy) -0.413 (-0.4) [-0.7] N.A (.) [.] environment Payments for government contracts -0.023 (-5.4) [-7.9] N.S (.) [.] Losses due to criminal activity N.S (.) [.] -0.097 (-150) [-148.9] Payments for protection N.A (.) [.] -0.254 (-32.4) [-33.7] Security expenses -0.002 (-0.4) [-0.6] N.S (.) [.] Tax inspections (dummy) -0.17 (-30.4) [-28.1] N.S (.) [.] Inspections (number) N.S (.) [.] -0.032 (-11.3) [-12.4] Compulsory certificates (days) -0.031 (-5.1) [-4.2] N.A (.) [.] Compulsory certificates obtained (number) -0.363 (-60.4) [-52.6] N.A (.) [.] Time tax N.S (.) [.] -0.021 (119.9) [123.8] Informal competition (dummy) -0.158 (-13.8) [-15.2] -0.1 (-6.5) [-5.7] Sales declared to taxes N.A (.) [.] 0.013 (-13.4) [-15.1] Absenteeism N.A (.) [.] -0.271 (-16.3) [-10.8] Lawsuit (dummy) N.A (.) [.] -0.147 (59.5) [54] Customs clearance for imports (days) -0.175 (-72.1) [-73] -0.171 (-51.7) [-51.6] Total % contributions regulatory environment (-188) [-182.3] (-102.2) [-100.4] Labor and Staff - female workers -0.003 (-14.2) [-14.7] N.S (.) [.] skills Staff - unskilled workers N.S (.) [.] -0.182 (-3.4) [-1.9] Staff - part-time workers N.A (.) [.] -0.005 (21.5) [23] Staff - university education 0.002 (6.2) [5] N.S (.) [.] Training of skilled workers (weeks) N.S (.) [.] 0.041 (-0.6) [-0.6] Total % contributions labor and skills (-8) [-9.7] (17.5) [20.5] Quality and New technology purchased (dummy) N.A (.) [.] 0.187 (-14.1) [-13.3] innovation R&D (dummy) 0.078 (5.6) [4] N.S (.) [.] Outsourcing (dummy) 0.137 (7.3) [6.2] N.A (.) [.] E-mail (dummy) N.S (.) [.] 0.074 (11.9) [10.6] Total % contributions quality and innovation (12.9) [10.2] (-2.2) [-2.7] Finance and Sales paid before delivery 0.003 (7.4) [6.3] N.A (.) [.] corporate New fixed assets finance - internal founds 0.001 (7.1) [5.5] N.A (.) [.] governance New fixed assets finance - state-owned banks -0.005 (-1) [-2.1] N.A (.) [.] Largest shareholder 0.002 (15.3) [16.5] N.A (.) [.] Subsidies (dummy) 0.292 (8.2) [5] N.A (.) [.] External audit (dummy) N.S (.) [.] 0.769 (14.7) [14.6] Total % contributions finance and corporate governance (37) [31.2] (14.7) [14.6] Other control Share of exports 0.003 (13.6) [11.5] N.S (.) [.] variables Decreased sales (dummy) -0.194 (-2.3) [-6.7] N.A (.) [.] Decreased prices (dummy) 0.274 (1.7) [4.5] N.A (.) [.] Age 0.143 (78.4) [68.7] 0 (3.3) [0.7] Privatized firm (dummy) N.S (.) [.] 0.344 (-8.2) [-8.7] Duration of power outages -0.244 (-38) [-40.7] -0.332 (-12.5) [-14.2] Shipment losses, exports -0.054 (-5.2) [-4.7] N.A (.) [.] Shipment losses, domestic -0.012 (-2.1) [-2.1] N.A (.) [.] Phone connection (days) N.S (.) [.] -0.005 (-10.4) [-8.6] Total % contributions other control variables (46.1) [30.5] (-27.8) [-30.8] Grand Total % contribution (-100) [-120.1] (-100) [-98.8] "." contribution statistically non different from zero; N.S not significant variable; N.A not available variable

70 Table 2-A-3: Comparison of IC effects on employment in 2008 and 2005 2008 2005 IC coefficient (IC % cont. IC coefficient (IC % cont. on on av. log- on on av. log- Employment employment) Employment employment) Productivity -0.152 -9.3 -0.072 -92.1 Real wages -0.171 -44.5 -0.101 -563.6 Regulatory Court action (dummy) 0.321 3.5 N.S. . environment Security expenses (dummy) 0.193 2.6 N.S. . Compulsory certificates (dummy) (b) 0.139 1.5 N.A. . Compulsory certificates (days) (b) -0.077 -1.8 N.A. . Time tax (a) N.S. . 0.018 106.4 Informal competition (dummy) N.S. . -0.077 -24.6 Sales declared to taxes (a) N.A. . -0.015 -510.0 Labor costs declared (a) N.A. . 0.019 876.7 Payments for government contracts (dummy) N.S. . -0.112 -19.5 Transaction fees to obtain a land or a building (a) N.A. . -0.075 -407.9 Total % contributions red tape, informality and 5.8 21.0 others Labor and Staff - nonproduction workers (b) 0.016 36.1 N.A. . skills Staff - university education (b) 0.008 2.9 -0.013 -93.1 Training (dummy) (b) 0.374 4.0 N.A. . Management education (dummy) N.A. . 0.451 202.3 Internal training (dummy) N.A. . 0.2 67.9 External training (dummy) N.A. . 0.325 79.0 Total % contributions labor and skills 43.1 256.2 Quality and Quality certification (dummy) 0.41 5.4 0.448 118.2 innovation FDI (dummy) 0.41 0.4 N.S. . Website (dummy) 0.463 11.0 N.S. . New technology purchased (dummy) N.A. . 0.226 66.4 E-mail (dummy) N.S. . 0.267 140.8 Discontinued (dummy) -0.144 -1.0 N.A. . Total % contributions quality and innovation 15.8 325.4 Finance and Purchases paid after delivery (a) 0.019 34.3 N.S. . corporate Purchase fixed assets (dummy) 0.327 5.1 N.S. . governance Loan (dummy) 0.193 3.4 0.156 219.2 Loan - state-owned banks (dummy) -0.32 -1.1 N.A. . External audit (dummy) 0.275 4.9 0.239 64.1 Credit line (dummy) N.S. . 0.214 71.0 Outstanding loan (dummy) N.A. . 0.333 43.1 Loan in TL (dummy) N.A. . -0.278 -55.4 Collateral (dummy) N.S. . -0.269 -53.3 Rent land (dummy) N.A. . -0.21 -235.6 Total % contributions finance and corporate 46.6 53.1 governance Other Incorporated company (dummy) 0.743 0.3 0.253 6.7 control Limited company (dummy) 0.232 5.2 N.A. . variables Exporting experience (years) (b) 0.191 4.3 N.A. .

Capacity utilization (b) 0.006 12.5 N.S. .

Shipment losses, domestic (b) -0.016 -0.5 N.S. .

Power outages (dummy) -0.222 -4.0 N.S. .

Age 0.298 24.9 N.S. .

Privatized firm (dummy) N.S. . 0.722 5.1

Exporter (dummy) N.S. . 0.353 109.0

Previous public ownership (dummy) N.A. . 0.755 6.7

Dummy for young firms N.S. . -0.311 -27.4

Total % contributions other control variables 42.6 100

Grand Total % contribution 100.0 100.0

71 Table 2-A-4: Comparison of IC effects on probability exporting in 2008 and 2005 2008 2005 IC (IC % cont. IC (IC % cont. coefficient on av. log- coefficient on on av. log- on exporting) exporting exporting) exporting Productivity 0.251 83.7 0.178 50.1 Regulatory Inspections (dummy) -0.047 -6.5 0.046 6.6 environment Payments for government contracts (dummy) -0.221 -5.2 N.S. . Payment for compulsory certificates (dummy) -0.027 -4.4 N.A. . Sales declared to taxes N.A. . -0.003 -21.7 Security expenses N.S. . 0.084 103.8 Customs clearance for imports (dummy) N.S. . -0.075 -18.3 Informal competition (dummy) N.S. . -0.05 -3.2 Total % contributions regulatory environment -16.2 67.2 Labor and Management education N.A. . 0.128 12.6 skills Staff - skilled workers N.S. . 0.089 5.6 Training of unskilled workers (weeks) N.A. . 0.015 4.6 Total % contributions labor and skills 0.0 22.8 Quality and Quality certification (dummy) 0.203 14.4 0.064 3.7 innovation New product (dummy) 0.113 8.2 N.S. . Website (dummy) 0.134 17.1 N.S. . E-mail (dummy) N.S. . 0.136 15.7 Total % contributions quality and innovation 39.7 19.3 Finance and New fixed assets finance - equity 0.002 1.1 N.S. . corporate Land/buildings as collateral (dummy) 0.106 4.2 N.A. . governance Loan (dummy) N.S. . 0.046 3.6 External audit (dummy) N.S. . 0.073 4.3 Total % contributions finance and corporate governance 5.2 7.9 Other control Increased prices (dummy) -0.082 -3.4 N.A. . variables Age -0.057 -25.7 N.S. . Inventory (days) 0.032 16.7 N.A. . Duration of power outages N.S. . -0.123 -38.7 Incorporated company (dummy) N.S. . 0.178 1.0 Competitors (number) N.A. . -0.141 -44.1 Capacity utilization N.S. . 0.003 25.5 Unionized workers N.A. . -0.014 -11.1 Total % contributions other control variables -12.5 -67.4 Grand Total % contribution 100.0 100.0

Table 2-A-5: Comparison of IC effects on the probability of receiving FDI in 2008 and 2005 2008 2005 IC (IC % cont. IC (IC % cont. coefficient on av. log- coefficient on on av. log- on FDI FDI) FDI FDI Productivity 0.067 93.5 0.037 38.3 Regulatory Tax inspections (number) -0.028 -10.4 N.A. . environment Customs clearance for imports (days) N.S. . -0.04 -36.5 Payments for government contract (dummy) N.S. . -0.166 -20.9 Total % contributions regulatory environment -10.4 -57.5 Labor and Staff - nonproduction workers 0.001 17.2 N.A. . skills Management education N.A. . 0.105 37.9 Internal training (dummy) N.A. . 0.148 39.0 Staff - university education N.S. . 0.001 6.6 Total % contributions labor and skills 17.2 83.5 Quality and Foreign technology (dummy) 0.031 4.1 0.028 2.2 innovation R&D (dummy) 0.005 2.3 N.A. . New product (dummy) N.S. . 0.143 27.3 Total % contributions quality and innovation 6.4 29.5 Finance and Loan - non-financial institutions (dummy) -0.078 -0.7 N.A. . corporate External audit (dummy) 0.023 9.0 N.S. . governance Total % contributions finance and corporate governance 8.3 0.0 Other Importer firm (dummy) 0.03 11.1 N.S. . control More than 5 competitors (dummy) 0.038 20.9 N.A. . variables Old firm (dummy) -0.055 -33.4 N.S. . Decreased prices (dummy) -0.023 -1.8 N.A. . Duration of power outages -0.015 -11.7 N.S. . Incorporated company (dummy) N.S. . 0.065 1.4 Exporter (dummy) N.S. . 0.019 4.7 Total % contributions other control variables -14.9 6.1 Grand Total % contribution 100.0 100.0

72 Table 2-A-6: Robust IC elasticities and semi-elasticities with respect to productivity – OLS Estimation

Restricted estimation Unrestricted by industry estimation Two step estimation One step estimation Two step estimation One step estimation Solow's Residual Cobb-Douglas Translog Solow's Residual Cobb-Douglas Translog Coef. S.E Coef. S.E Coef. S.E Coef. S.E Coef. S.E Coef. S.E Regulatory Payments for power supply (dummy) (b) -0.413 [0.149]*** -0.179 [0.218] -0.228 [0.223] -0.441 [0.155]*** -0.107 [0.263] -0.161 [0.294] environment Payments for government contracts (b) -0.023 [0.011]** -0.019 [0.011]* -0.021 [0.010]** -0.022 [0.011]** -0.022 [0.011]** -0.025 [0.012]** Security expenses (b) -0.002 [0.012] -0.016 [0.012] -0.015 [0.011] -0.002 [0.013] -0.018 [0.012] -0.016 [0.011] Tax inspections (number) (a) -0.17 [0.093]* -0.154 [0.091]* -0.122 [0.091] -0.181 [0.093]* -0.168 [0.087]* -0.134 [0.086] Compulsory certificates (days) (b) -0.031 [0.031] -0.032 [0.024] -0.051 [0.022]** -0.031 [0.031] -0.036 [0.023] -0.046 [0.024]* Compulsory certificates (number) (a) -0.363 [0.107]*** -0.383 [0.110]*** -0.294 [0.109]*** -0.373 [0.105]*** -0.417 [0.109]*** -0.345 [0.115]*** Customs clearance for imports (days) (a) -0.175 [0.175] -0.312 [0.169]* -0.252 [0.161] -0.188 [0.177] -0.312 [0.189] -0.17 [0.187] Informal competition (dummy) (b) -0.158 [0.068]** -0.169 [0.066]** -0.142 [0.063]** -0.156 [0.069]** -0.133 [0.070]* -0.106 [0.073] Labor and Staff - female workers (b) -0.003 [0.002]* -0.005 [0.002]*** -0.005 [0.002]** -0.003 [0.002] -0.006 [0.002]*** -0.005 [0.002]*** skills Staff - university education (b) 0.002 [0.003] 0.003 [0.002] 0.003 [0.002] 0.002 [0.003] 0.004 [0.003] 0.003 [0.002] Quality and R&D (dummy) (b) 0.078 [0.087] 0.114 [0.075] 0.11 [0.073] 0.069 [0.088] 0.106 [0.077] 0.143 [0.081]* innovation Outsourcing (dummy) (b) 0.137 [0.095] 0.161 [0.084]* 0.159 [0.078]** 0.143 [0.096] 0.15 [0.085]* 0.101 [0.073] Finance and Sales paid before delivery (b) 0.003 [0.002] 0.003 [0.002] 0.003 [0.002] 0.004 [0.002]* 0.002 [0.002] 0.003 [0.002] corporate New fixed assets finance - internal founds (b) 0.001 [0.001] 0.001 [0.001] 0.001 [0.001] 0.001 [0.001] 0.001 [0.001] 0.001 [0.001]** governance. New fixed assets finance - state-owned banks (b) -0.005 [0.004] -0.005 [0.003] -0.005 [0.003] -0.005 [0.004] -0.006 [0.003]* -0.005 [0.003]* Largest shareholder (b) 0.002 [0.002] 0.002 [0.001] 0.002 [0.001] 0.002 [0.002] 0.002 [0.001] 0.002 [0.001] Subsidies (dummy) 0.292 [0.120]** 0.205 [0.112]* 0.185 [0.111]* 0.302 [0.122]** 0.198 [0.116]* 0.175 [0.120] Other control Share of exports (b) 0.003 [0.001]*** 0.003 [0.001]*** 0.003 [0.001]*** 0.003 [0.001]*** 0.003 [0.001]*** 0.003 [0.001]*** variables Duration of power outages (a) -0.244 [0.158] -0.282 [0.162]* -0.264 [0.154]* -0.25 [0.159] -0.278 [0.165]* -0.213 [0.175] Shipment losses, exports (a) - interaction with dummy for -0.054 [0.023]** -0.055 [0.019]*** -0.06 [0.019]*** -0.054 [0.023]** -0.042 [0.028] -0.054 [0.024]** exporter Shipment losses, domestic (b) -0.012 [0.003]*** -0.012 [0.004]*** -0.013 [0.004]*** -0.012 [0.003]*** -0.013 [0.005]*** -0.011 [0.004]*** Decreased sales (dummy) (b) -0.194 [0.095]** -0.26 [0.095]*** -0.228 [0.097]** -0.192 [0.094]** -0.267 [0.094]*** -0.176 [0.097]* Decreased prices (dummy) (b) 0.274 [0.139]* 0.12 [0.142] 0.114 [0.139] 0.27 [0.136]** 0.108 [0.145] 0.113 [0.144] Age 0.143 [0.059]** 0.161 [0.056]*** 0.156 [0.057]*** 0.142 [0.059]** 0.161 [0.056]*** 0.13 [0.061]** Observations 780 780 780 780 780 780 R-squared 0.13 0.76 0.78 0.13 0.77 0.80 NOTES: Restricted: equal input output for all the establishments in the country; Unrestricted: equal input-output elasticities for all the establishments in the same sector. Two steps estimation: in the first step estimation of equation (3.5a) by non-parametric techniques to compute productivity (Solow residual), in the second step estimate (3.5b) and (3.5c) by OLS using as dependent variable the Solow residual from the first step, either restricted or unrestricted. Single step estimation: estimate (3.5a), (3.5b) and (3.5c) in a single step by OLS, where (3.5a) can be a Cobb-Douglas Production function or a Translogarithmic. *significant at 10%; ** significant at 5%; *** significant at 1% given by robust standard errors corrected for correlation between cluster (industry and region) in brackets. Each regression includes a set of industry, size and region dummies and a constant term. (a) Variables instrumented with the industry-region-size average; (b) Variables approximated with a proxy (only missing values replaced by the industry-region-size average). Source: Authors‘ estimations with Turkey ES 2008 data.

73

Table 2-A-7: IC percentage contributions to aggregate demeaned log-productivity in 2008 Aggregate log-TFP Average log-TFP Allocative efficiency

Regulatory Payments for power supply (dummy) (b) -0.10 -0.20 0.38 environment Payments for government contracts (b) -1.31 -2.31 3.61 Security expenses (b) -0.11 -0.18 0.23 Tax inspections (dummy) (a) -7.43 -8.28 -3.29 Compulsory certificates (days) (b) -1.25 -1.25 -1.26 Compulsory certificates (number) (a) -14.77 -15.49 -11.21 Customs clearance for imports (days) (a) -17.63 -21.50 1.33 Informal competition (dummy) (b) -3.37 -4.46 1.94 Labor and Staff - female workers (b) -3.46 -4.34 0.84 skills Staff - university education (b) 1.51 1.46 1.72 Quality and R&D (dummy) (b) 1.37 1.17 2.32 innovation Outsourcing (dummy) (b) 1.78 1.84 1.52 Finance and Sales paid before delivery (b) 1.81 1.84 1.65 corporate New fixed assets finance - internal funds (b) 1.73 1.61 2.33 governance. New fixed assets finance - state-owned banks (b) -0.25 -0.61 1.49 Largest shareholder (b) 3.73 4.85 -1.77 Subsidies (dummy) 2.01 1.48 4.63 Other control Share of exports (b) 3.34 3.40 3.03 variables Duration of power outages (a) -9.29 -11.99 3.92 Shipment losses, exports (a) - interaction with dummy for exporter -1.28 -1.39 -0.75 Shipment losses, domestic (b) -0.51 -0.63 0.09 Decreased sales (dummy) (b) -0.56 -1.98 6.42 Decreased prices (dummy) (b) 0.43 1.31 -3.93 Age 19.16 20.22 13.95 Total contribution of IC (demean log-productivity) -24.45 -35.40 29.19 Other stuff Industry/region/size controls -31.85 -33.60 -23.30 Constant term 141.86 170.80 0.00 Residual 14.45 -1.81 94.11 Total contribution of other stuff 124.45 135.40 70.81 Total 100.00 100.00 100.00

Table 2-A-8: IC elasticities and semi-elasticities with respect to employment – IV Estimation Dependent variable: log-employment (demand for labor) Restricted Solow residual Unrestricted Solow residual Blocks Explanatory ICA variables Coefficient % Contrib Coefficient % Contrib Productivity -0.152* -8.8 -0.153* -8.8 Real wages -0.171*** -41.8 -0.171*** -41.9 Regulatory Court action (dummy) 0.321*** 3.3 0.320*** 3.3 environment Security (dummy) 0.193** 2.4 0.193** 2.4 Compulsory certificates (dummy) (b) 0.139*** 1.4 0.138*** 1.4 Compulsory certificates (days) (b) -0.077** -1.7 -0.076** -1.7 Labor and Staff - nonproduction workers (b) 0.016*** 33.9 0.016*** 33.9 skills Staff - university education (b) 0.008*** 2.8 0.008*** 2.8 Training (dummy) (b) 0.374*** 3.8 0.375*** 3.8 Quality and Quality certification (dummy) 0.410*** 5.1 0.409*** 5.1 innovation Discontinued (dummy) -0.144** -0.9 -0.143* -0.9 Website (dummy) 0.463*** 10.3 0.462*** 10.3 Finance and Purchases paid after delivery (a) 0.019* 32.2 0.019* 32.3 corporate Purchase fixed assets (dummy) 0.327*** 4.8 0.326*** 4.7 governance Loan (dummy) 0.193*** 3.2 0.194*** 3.2 Loan - state-owned banks (dummy) -0.320*** -1.0 -0.321*** -1.0 External audit (dummy) 0.275*** 4.6 0.275*** 4.6 Other control Incorporated company (dummy) 0.743*** 0.3 0.742*** 0.3 variables Limited company (dummy) 0.232*** 4.9 0.233*** 4.9 FDI (dummy) 0.410*** 0.4 0.408*** 0.4 Exporting experience (b) 0.191*** 4.0 0.191*** 4.0 Power outages (dummy) -0.222*** -3.8 -0.221*** -3.8 Shipment losses, domestic (b) -0.016*** -0.5 -0.016*** -0.5 Capacity utilization (b) 0.006*** 11.8 0.006*** 11.7 Age 0.298*** 23.3 0.298*** 23.4 Instruments First stage R-squared: productivity2 0.324 0.318 evaluation Partial R-squared: productivity3 0.27 0.264 Partial R-squared F test (p-value): productivity4 0 0 Hansen test (p-value)5 0.591 0.639 Observations 779 779

74

Table 2-A-9: IC linear probability coefficients with respect to the probability of exporting – IV Estimation Dependent variable: probability of exporting Restricted Solow residual Unrestricted Solow residual Blocks Explanatory ICA variables Coefficient % Contrib Coefficient % Contrib Productivity 0.251** 124.6 0.243** 119.9 Regulatory Inspections (number) (b) -0.047** -9.7 -0.046** -9.6 environment Payments for government contracts (b) -0.221*** -7.8 -0.222*** -7.9 Compulsory certificates (days) (b) -0.027** -6.6 -0.026** -6.6 Quality and Quality certification (dummy) (b) 0.203*** 21.4 0.204*** 21.5 innovation Website (dummy) (b) 0.134** 25.5 0.135** 25.8 New product (b) 0.113*** 12.2 0.113*** 12.3 Finance and New fixed assets finance - equity (b) 0.002 1.6 0.002* 1.6 corporate Land/buildings as collateral (dummy) (b) 0.106*** 6.2 0.102*** 5.9 governance Other Increased prices (dummy) -0.082* -5.1 -0.083* -5.2 control Age -0.057** -38.2 -0.055* -37.4 variables Inventory (days) (b) 0.032** 24.8 0.032** 24.7 Large firm (dummy) 0.211*** 10.7 0.214*** 10.9 Instruments First stage R-squared: productivity2 0.214 0.223 evaluation Partial R-squared: productivity3 0.0428 0.0436 Partial R-squared F test (p-value): productivity4 0.0046 0.003 Hansen test (p-value)5 0.431 0.412 Observations 636 636

Table 2-A-10: IC linear probability coefficients with respect to the probability of receiving FDI – IV Estimation Dependent variable: probability of receiving FDI Restricted Solow residual Unrestricted Solow residual Blocks Explanatory ICA variables Coefficient % Contrib Coefficient % Contrib Productivity 0.067* 528.9 0.066* 518.1 Regulatory Tax inspections (days) (b) -0.028*** -59.0 -0.029*** -59.3 environment Labor and Staff - nonproduction workers (b) 0.001* 97.2 0.001* 98.2 skills Quality and Foreign technology (dummy) (b) 0.031* 23.2 0.031 23.3 innovation R&D expenditures (b) 0.005*** 12.8 0.005*** 12.7 Finance and Loan - non-financial institutions (b) -0.078*** -4.2 -0.078*** -4.2 corporate External audit (dummy) (b) 0.023* 51.0 0.023 51.3 governance Other control Importer firm (dummy) (b) 0.030* 62.7 0.030* 63.2 variables More than 5 competitors (dummy) (b) 0.038** 118.1 0.038** 116.8 Old firm (dummy) -0.055* -188.8 -0.055* -190.6 Decreased prices (dummy) -0.023* -10.0 -0.023* -9.8 Duration of power outages (b) -0.015** -66.4 -0.015** -65.9 Large firm 0.056** 23.5 0.057** 23.7 Instruments First stage R-squared: productivity2 evaluation Partial R-squared: productivity3 0.0233 0.231 Partial R-squared F test (p-value): productivity4 0.0121 0.01 Hansen test (p-value)5 0.987 0.987 Observations 778 778

75

Annex 2-B. Investment Climate Variables Used for the Econometric Analysis

Table 2-B-1: Definition of IC Variables – Regulatory environment Survey Name of the variable Description of the variable availability 2005 ICS & Time tax Percentage of managers' time spent in dealing with bureaucratic issues. 2008 ES Informal competition (dummy) Dummy variable that takes value 1 if the firm competes with informal (no registered) firms. Inspections (number) In the last year, total Inspections (number) (log). Payments for government contracts (dummy) Dummy variable that takes value 1 if in plant's sector it is common to pay an extra amount of money in order to obtain a contract with the government. Payments for government contracts Illegal payment in order to obtain a contract with the government. Related as percentage of contract value. Construction related permit (days) Actual delay to obtain a construction related in days (log). Operating license (days) Actual delay to obtain a main operating license in days (log). Customs clearance for exports (days) Average number of days to clear customs to export (log). Customs clearance for imports (days) Average number of days to clear customs to imports (log). Only in Import license (days) Current delay to obtain an import license related in days (log). 2008 ES Permit (days) Number of days it takes to obtain a permit Payment for import license (dummy) Gifts expected or requested to obtain an import license, conditional on submit an import license. Share of exports Share of exports over total annual sales. Time tax Percent of managers' time spent in dealing with bureaucratic issues per week (log) Inspections (days) Number of working days spent with inspections (log) Tax inspections (dummy) Dummy variable that takes value 1 if the firm has been visited by tax officials during last year. Tax inspections (number) Total Inspections (number) of tax officials received by the plant in 2007 (log) Tax inspections (days) Number of working days spent with tax inspections (log) Security expenses (dummy) Dummy taking value 1 if the plant has security expenses. Payment for construction permit (dummy) Gifts expected or requested to obtain a construction permit, conditional on submit a construction permit. Compulsory certificate (dummy) Dummy variable that takes value 1 if the firm has to have any compulsory certificate to produce or sell any product Compulsory certificates (number) Number of compulsory certificates obtained (log) Compulsory certificates (days) Number of working days spent when obtaining compulsory certificates (log) Payment for compulsory certificates (dummy) Gifts expected or requested to obtain a compulsory certificate, conditional on submit a compulsory certificate. Only in Payments for protection Cost due to protection payments e. g. to organized crime to prevent violence (bribery) (log). 2005 ICS Sales declared to taxes Percentage of total sales declared to taxes. Labor costs declared Percentage of workforce declared to taxes. Absenteeism Days of production lost due to absenteeism (log). Lawsuit (dummy) Dummy variable that takes value 1 if the firm has been involved in a lawsuit in the last three years.

Table 2-B-2: Definition of IC Variables – Labor and skills Survey Name of the variable Description of the variable availability 2005 ICS & Staff - skilled workers Percentage of skilled workers in firm's staff. 2008 ES Staff - unskilled workers Percentage of unskilled workers in firm's staff. Staff - female workers Percentage of female workers in firm's staff. Staff - university Percentage of staff with at least one year of university. Labor regulation Share of firms perceiving labor regulation as a major or very severe obstacle Only in Training (dummy) Dummy taking value one if the firm provides formal (beyond on the job) training to its employees. 2008 ES Manager experience (years) Number of years of experience of Top Manager in the sector Staff - production workers Percentage of production workers in staff. Staff - nonproduction workers Percentage of nonproduction workers in staff. Training - production workers Percentage of production workers receiving formal (beyond on the job) training Training - non-production workers Percentage of non-production workers receiving formal (beyond on the job) training Only in Staff - part time workers Percentage of part time workers in firm's staff. 2005 ICS Internal training (dummy) Dummy variable that takes value 1 if the plant provides internal training to its employees. External training (dummy) Dummy variable that takes value 1 if the plant provides external training to its employees. Weeks of training of skilled workers Number of weeks of training received by the skilled workers during last year.

76

Table 2-B-3: Definition of IC Variables – Quality and innovation Survey Name of the variable Description of the variable availability 2005 ICS & Quality certification (dummy) Dummy variable that takes value 1 if the plant has a quality certification. 2008 ES New product (dummy) Dummy variable that takes value 1 if the plant has developed a new product or product line. E-mail (dummy) Dummy variable that takes value 1 if the plant uses email. Website (dummy) Dummy variable that takes value 1 if the plant has a website. Product upgraded (dummy) Dummy variable that takes value 1 if the plant upgraded an existing product last year. Foreign technology (dummy) Dummy variable that takes value 1 if the firm used a licensed technology of a foreign company in the last year. Only in Sales of new products Percentage of total sales corresponding with new products 2008 ES R&D (dummy) Dummy that takes value 1 if the firm performed R&D activities during last year. R&D expenditures Total R&D expenditures as percentage of annual sales Computer Percentage of staff using computer at job. Outsourcing (dummy) Dummy taking value 1 if the plant subcontracts any part of the activity. Discontinued (dummy) Dummy taking value 1 if the plant has discontinued at least one product line Only in New technology purchased (dummy) Dummy variable that takes value 1 if the firm purchased any new technology during last year. 2005 ICS

Table 2-B-4: Definition of IC Variables – Finance and corporate governance Survey Name of the variable Description of the variable availability 2005 ICS & Loan (dummy) Dummy variable that takes value 1 if the plant reports that it has a bank loan. 2008 ES Loan – bank (dummy) Dummy variable that takes value 1 if the firm has a loan from a domestic private commercial banks. Loan – state-owned bank (dummy) Dummy variable that takes value 1 if the firm has a loan from a state owned banks. Collateral (dummy) Dummy variable that takes value 1 if the loan requires collateral. Collateral Value of collateral as share of loan value (conditional on having loan with collateral) External audit (dummy) Dummy variable that takes value 1 if the plant uses an external auditory. Only in Purchases paid after delivery Percentage of annual purchases paid for after the delivery. 2008 ES Sales paid before delivery Percentage of annual sales paid for before the delivery. Purchase fixed assets (dummy) Dummy variable that takes value 1 if the firm has purchaser fixed assets during last year. New fixed assets finance - internal funds Percentage of firm's working capital financed with funds from informal sources. New fixed assets finance - equity Percentage of firm's working capital financed with funds from equity. New fixed assets finance - state-owned banks Percentage of investments in new fixed assets financed with funds state owned banks. New fixed assets finance - state-owned banks Percentage of investments in new fixed assets financed with external funds Loan - non-financial institution (dummy) Dummy variable that takes value 1 if the firm has a loan from a non-financial institutions. Land/ buildings as collateral (dummy) Dummy that takes value 1 if the firm uses land or buildings as collateral (conditional on having a loan with collateral). Largest shareholder Percentage of firm's capital owned by the largest shareholder. Overdraft facility (dummy) Dummy that takes value 1 if the firm has access to an overdraft facility Subsidies (dummy) Dummy variable that takes value 1 if the firm receives any subsides from the national, regional and local government or EU. Only in Outstanding loan (dummy) Dummy variable that takes value 1 if the firm has a loan outstanding from a financial institution. 2005 ICS Rent land (dummy) Dummy variable that takes value 1 if the firm has a loan from a leasing arrangement. Loan in TL (dummy) Dummy variable that takes value 1 if the loan is denominated in Turkish Lira.

Table 2-B-5: Definition of IC Variables – Other Survey Name of the variable Description of the variable availability 2005 ICS & Duration of power outages Average duration of power outages suffered by the plant in hours (log). 2008 ES Loss from power outages Value of the losses due to power outages as share of percentage of sales. Incorporated company (dummy) Dummy variable that takes value 1 if the plant is an incorporated company. Public ownership (dummy) Dummy variable that takes value 1 if the firm belongs to the government.

77

Age Difference between the year that the plant started operations and current year. Capacity utilization Average percentage of capacity used during last year. Inputs imported Share of inputs imported directly Importer (dummy) Dummy variable that takes value 1 if imports are greater than 10%. Exporter (dummy) Dummy variable that takes value 1 if exports are greater than 10%. Phone connection (days) Current delay to obtain a phone connection in days (log). Only in Power outages (dummy) Dummy variable that takes value 1 if the plant has suffered any power outages during last year. 2008 ES Duration of power outages Total duration of power outages suffered by the plant by month, in hours, conditional on the plant reports having power outages. Shipment losses, exports Fraction of the value of the plant‘s average cargo consignment that was lost in transit due to breakage, theft or spoilage in the international market. Shipment losses, domestic Fraction of the value of the plant‘s average cargo consignment that was lost in transit due to breakage, theft or spoilage in the domestic market. Limited company (dummy) Dummy variable that takes value 1 if the plant is a limited company. Exporting experience Number of years of exporting experience.(log) More than 5 competitors (dummy) Dummy taking value one if the plant has less than 5 competitors in the local market. Domestic sales (dummy) Firms selling more than 90% of their output in the domestic market Decreased sales (dummy) Dummy taking value 1 if the plant has decreased its sales Increased prices (dummy) Dummy taking value 1 if the plant has increased its prices Decreased prices (dummy) Dummy taking value 1 if the plant has decreased its prices Only in Competitors (number) Number of competitors in the main market (log). 2005 ICS Unionized workers Percentage of workers that belongs to a syndicate. Privatized firm (dummy) Dummy variable that takes value 1 if the firm previously belonged to the government.

78

Chapter 3. PROMOTING SME GROWTH

3.1. Sustainable growth in the post-crisis environment will crucially depend on easing the constraints to the expansion of SMEs. SMEs in Turkey account for 79.4 percent of employment, 44.6 percent of total investments, 67.4 percent of total sales, 25-30 percent of total exports, 57.3 percent of total value added and 25 percent of bank credit. Given the rate of tax evasion and the size of the informal sector, the contribution of SMEs to the economy may be somewhat underestimated. This implies that the development of a more productive and more outward-oriented SME sector is a crucial development challenge for Turkey. A healthy SME sector can not only significantly contribute to capital accumulation, provide increased employment opportunities for a rapidly increasing workforce, and promote regional development, but is also crucial to increase the resilience of the economy to external shocks, like the one represented by the recent global crisis. This Chapter, using data from the most recent Enterprise Survey for Turkey and other countries,44 contributes to the understanding of the role of the investment climate in determining the expansion of the Turkish SME sector.

3.2. The 2007 Investment Climate Assessment highlighted the importance of investment climate factors in constraining the expansion of firms. In addition to the 2005 enterprise survey, the report based its analysis on 1996-2001 TURKSTAT firm level data,45 finding that the main constraints to productivity and output growth in Turkey may come not so much from barriers to start up and exit, but rather from barriers to expansion. The report also pointed out that Turkish firms are heterogeneous, showing large disparities in size, growth and productivity performance. New entrants in Turkey were found to grow at a slower pace than firms in comparator countries. Furthermore, in Turkey a large share of productivity benefits derives from firm churning in low-tech industries, rather than, as it is the case in most comparator countries, from entry and exit in high-tech sectors, where new technologies are often better harnessed by new firms. The analysis also indicated that firms face significant barriers to expansion, suggesting reforms aimed at improving a wide range of investment climate conditions, including access to credit, adoption of quality standards, technologies and innovation, quality of the labor force and reduction of administrative procedures.

3.3. Analysis of 2008 data confirms that the current investment climate negatively affects the productivity performance of small and medium-sized firms. As argued in Chapter 2, analysis of 2008 data shows that, while larger firms benefit from the more positive aspects of the investment climate, smaller firms bear the burden of its less positive features. The need to increase average efficiency of the Turkish business sector emphasizes the importance of policy and institutional mechanisms targeted towards easing the efficient operation and expansion of SMEs.46

44 The Enterprise Survey was conducted in 29 other countries in the East Europe and Central Asia (ECA) region in the same time period it was conducted in Turkey. This allows comparison of firm evolution in Turkey with other countries in the region. The analysis compares Turkey with Russia, Ukraine, Poland, Romania, EU-10 (Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia), EU-8 (EU-10, excluding Bulgaria and Romania), as well as with the average for the entire ECA region. See www.enterprisesurveys.org for the complete dataset and for a detailed description of the data and methodology used in sampling. 45 The TURKSTAT database draws from annual surveys of all Turkish manufacturing firms with more than 10 employees conducted between 1996 and 2001. It is representative of the whole population of manufacturing firms in Turkey and includes information on micro units. 46 Until a few years ago, the definition of what constitutes an ―SME‖ varied from one agency to another (see OECD (2004)). Since 2005, however, the Turkish Ministry of Industry and Trade has consolidated the concept by defining SME as a firm with a less than TL25 million turnover and 250 employees. (In the European Union, the turnover figure is about 4 times the Turkish figure [€50 million] but the employee size is the same.) This SME definition is further broken down in the Turkish context to 3

79

This Chapter adds to the productivity analysis of Chapter 2 by examining the dynamics of evolution of Turkish SMEs in terms of employment growth. It finds that (i) medium-sized firms are the slowest growing group; that (ii) SMEs grow at a slower rate in Turkey than in comparator countries; and that (iii) improved access to finance significantly increases employment growth rates. It also suggests a number of measures to ease financial constraints to the expansion of SMEs. It should be noted that the enterprise survey tracks formal full-time employment, hence investment climate constraints to firm growth could be more deep-rooted and far-reaching than our analysis suggests, since reduced incentives to employ formally imply that firms have access to lower skill levels, which, in turn, negatively affects their productivity potential.

3.1 Patterns of Firm Growth in Turkey 3.4. Smaller and younger firms can be expected to grow faster, but in Turkey there seems to be a “missing middle” with small and medium firms falling short of prediction. Using the size and age groups described in Box 3-1, the growth rates of firms are calculated as the annual increase in full-time employment between 2004 and 2007. Figure 3-1 shows that the growth rates of Turkish firms do not monotonically decrease as size increases. Micro (1-10 employees) and large firms (more than 250 employees) appear as the fastest growing groups in 2004-2007, while small (11-50 employees) and medium firms (51-250 employees) display far slower growth rates across different age groups. Young (1- 5 years) medium-sized firms, in particular, display no growth at all in the period under examination. These findings are in opposition with regularities observed in other countries and predicted in the theoretical literature on firm growth. For instance, firm growth rates should decrease as size and age increase.47 This is based on the observation that diminishing returns to scale or bounded efficiency may limit the rate of expansion for larger firms. Similarly, diminishing returns to learning can explain the inverse relation between growth and age, with older firms less prepared to implement new modalities of operation.48 In addition to size and age, numerous other factors can affect firm growth, such as the levels of technology and human capital, the development level of the country, or the regulatory environment. Such broad investment climate factors are even more critical for smaller firms, who may lack the capacity to cope with distortions in access to finance, availability of technology and skills or the regulatory environment.49

Box 3-1: Data for the analysis of firm growth The analysis of firm growth is based on the 2008 Enterprise Survey data used in the rest of this report (Box 2-1). In analyzing how firm evolution changes with size, we divide firms into four size groups measured in fulltime employment levels: Micro: ≤10, Small: 11-50, Medium: 51-250, and Large: ≥251. These groups are constructed according to the employment level of the firms in 2004. The relevant questions in the survey refer to the last complete fiscal year which is 2007. The survey asks the number of full-time employees in the last fiscal year and sub-segments, as micro (turnover less than TL1 million), small (turnover between TL1-TL5 million) and medium (turnover between TL5-TL25 million) enterprises. 47 Dunne, Roberts, Samuelson (1989), Evans (1987a, 1987b) and numerous other studies have provided evidence on these regularities. Studies like Jovanovic (1982), Cooley and Quadrini (2001), Klette and Kortum (2004), Klepper and Thompson (2007) construct structural models that can explain these regularities. 48 Jovanovic (1982) argues that firms learn their efficiency levels over time. Least efficient firms exit and more efficient firms adjust their scale of operations through learning. Hence small and young firms grow faster because they are in the initial process of uncovering their efficiency levels. 49 Beck, Demirguc-Kunt, Maksimovic (2005) show that financial, legal, and corruption related problems distort firm growth. They also show that whether these factors affect growth depends on firm size and among all firms, small firms are the most constrained group. In a more recent study Aterido, Hallward-Driemier, and Pages (2009), analyze how business environment affects firm growth using a wide set of objective measures for obstacles. They show that poor business environment hurts firm growth and the amount of distortion varies by firm size.

80

three fiscal years ago, which is 2004. Annual growth rates are calculated using these questions. Firms are also divided into three age groups as: 1-5, 6-15, and ≥16 years of age. In addition to the cross-sectional data from 2008 survey, 419 firms were also surveyed in 2005, thus giving a panel dimension to the data. In all tabulation and regressions use is made of probability weights, to ensure that results are representative of the population of firms. Annex 3-A provides a full description of the data used.

3.5. The pattern of firm growth presents sectoral and regional variations, with several sectors and regions where SMEs grow more slowly than large firms. Table 3-1 shows the growth rates of firms in different size groups in 2-digit Figure 3-1: Employment growth at different sizes and ages (2004-2007) industries and across five macro-regions. In 35 the food sector both small and medium size 29 30 firms grow at a slower rate in 2004-2007 26 than large firms – 1 percent and 5 percent, 25 respectively, compared to a growth of 8 20 18 14 14 14 14 percent for large firms. In the garments % 15 sector, medium firms have not only grown 10 6 more slowly than large firms, but their 3 3 5 2 1 2 employment has contracted by 11 percent in 0 the period. In the plastics and rubber sector, 0 -5 Micro Small -2 Medium Large small firms have shed employment at a faster -4 pace (-5 percent) than large firms (-1 1-5 6-15 >=16 Total percent), while in machinery and equipment Source: Turkey ES 2008 employment in small firms has declined by 9 percent, compared to an increase in all other size categories. Turning to the regional breakdown, South and Marmara have the fastest, while Aegean has the slowest, rate of employment growth in micro firms. In Marmara and Central Anatolia, small and medium firms have slower growth rates than large firms.

Table 3-1: Growth rates by industry and region (2004-2007)

Micro Small Medium Large Food 0.12 0.01 0.05 0.08 Textiles 0.17 0.09 0.01 0 Garments 0.04 0.07 -0.11 0 Chemicals 0.13 0.07 -0.03 -0.12 Plastics & rubber 0.26 -0.05 0.05 -0.01 Non metallic mineral 0.02 0.05 0.1 0 2-Digit Basic metals 0.09 0.06 0.19 0.17 Industry Fabricated metal products 0.18 0.06 0.05 0.03 Machinery and equipment 0.2 -0.09 0.1 0.16 Electronics 0.27 0.01 0.02 -0.34 Services of motor vehicles 0.13 0.1 0.71 0.1 Wholesale 0.12 -0.03 0.13 0.59 Retail 0.21 0.03 0.11 0.28 Marmara 0.21 0.01 -0.01 0.05 Aegean 0.06 0.03 -0.02 -0.04 Region Central Anatolia 0.14 -0.01 0.1 0.15 South 0.22 0.14 0.06 0.03 Black Sea - Eastern 0.15 0.03 0.08 0.06

81

3.6. Medium firms (51-250 employees) are the slowest growing group, with employment growth 16 percent slower than micro firms and 5 percent slower than large firms. Based on the size categories described above, regression analysis allows analyzing how growth rates of firms differ by size controlling for other factor that may influence employment growth.50 Growth can be affected by firm characteristics, such as age, trade orientation, and ownership status. For instance, as mentioned earlier, it is well established that younger firms experience faster growth rates, or that firms that are globally engaged through exporting or foreign ownership are presented with more Figure 3-2: Growth rates relative to micro firms (2004-2007) opportunities to expand than non-trading Small (11-50) Medium (51-250) Large (≥251) and fully domestically owned firms. To control for these factors, regression analysis 0 includes dummy variables controlling for

-5 firms that generate more than 10 percent of their sales from exports and for foreign -10 ownership defined as 10 percent or more share in firm ownership. Firms with -15 government ownership might also grow at different rates because their objectives -20 might differ from a fully privately owned All Data Panel Firms firm. To control for this factor, a dummy -25 variable is included representing 10 percent Source: Turkey ES 2008 or more government ownership51. In addition, analysis controls for external conditions that may have an impact on employment growth, such as the industry and the region in which the firm operates. Regression results indicate that micro firms are the fastest growing group, allowing for benchmarking of the growth rates of other groups to that of micro firms (10 employees or less). Contrary to prediction and as already noted above, the growth rate does not decrease monotonically as size increases. In particular, medium-sized firms are the least growing group, growing 16 percent more slowly than micro when the full sample is considered and 23 percent more slowly when looking only at panel firms (Figure 3-2).

Figure 3-3: Percentage of firms that are above 16 years old 3.7. Comparison with other countries indicates 70 that SMEs in Turkey are, 60 62 60 on average, older, i.e. they 49 48 remain small for a longer 50 46 time. The cross-country 40 31 32 nature of the ES data allows 30 26 25 25 comparison of the dynamics 19 19 20 20 of firm growth in Turkey with 20 15 13 9 11 9 other countries in Europe and 10 5 1 1 2 1 Central Asia. In all 0 comparator countries larger Turkey Russia Ukraine Romania EU10 All ECA firms are more likely to have a higher proportion of firms Micro Small Medium Large that are older than 16 years.

Source: Turkey ES 2008 Turkey stands out for the

50 See Annex for detailed econometric results. 51 The firms that are 100 percent government owned are excluded from the survey.

82

longer average life of firms in all size categories, which can, to a large extent, be explained by the fact that, until the early 1990s, private enterprise was not allowed in most of the countries in the region, whose economies were governed under central planning arrangements. However, the far larger proportion of old (above 16 years of age) micro, small and, especially medium-sized firms in Turkey is still remarkable (Figure 3-3). In Turkey, 26 percent of micro firms and 31 percent of small firms are more than 16 years old, compared to 11 and 9 percent, respectively, in the EU-10 countries. Even more striking results emerge for medium-sized firms, with 60 percent in Turkey being more than 16 years old, compared to 20 percent in the EU-10.52 This observation might indicate that micro firms and SMEs in Turkey grow at a much slower pace than firms in other countries. In other words, they face barriers to their expansion that force them to remain at a smaller –and suboptimal – scale of operations. The demographics of large firms, on the other hand, are in line with average values of comparable firms in other countries.

Figure 3-4: Growth rates of firms: Cross-country comparison 3.8. Small and medium firms have much lower growth a: Growth rates relative to micro firms, percent (2004-2007) rates, while large firms

0 grow at similar rates as counterparts in other -2 Small Medium Large Russia countries. In order to -4 Poland verify whether SMEs in -6 Turkey face an Ukraine -8 environment that is more % Romania hostile to expansion than -10 All_ECA -11 for comparable firms in -12 EU10 other countries, -14 -14 Turkey regression analysis -16 -16 allows benchmarking the growth rate of Turkish -18 firms of various size b: Growth rates (percent) of Turkish SMEs (2004-2007) categories to that of a 0 number of comparators. Russia Poland Ukraine Romania ECA EU-8 EU-10 Whereas, in all countries -2 in the region, the rate of -4 expansion decreases as size increases, as -6 mentioned above, -8 Turkey stands out for the -7.6-7.7 slow growth of its SMEs -8.5 -10 -8.8 (Figure 3-4a). Contrary -9.8 -9.8 -10.7 -10.8 -10.6 to what is observed in -12 -11.2 -11.3 -11.9 Turkey, in all -12.4 -14 -12.7 comparator countries in Small (11-50) Medium (51-250) the region small- and

Source: Turkey ES 2008 medium-sized firms grow faster than large firms. Turkish SMEs in 2004-2007 grew 14-16 percent more slowly than micro firms, while growth of

52 Dunne, Roberts, and Samuelson (1989) perform a similar exercise for the US firms with a slightly different size and age classification. They find that the fraction of firms with 5-19 employees who are between 11-15 years old is 7 percent. In the same age group, the fraction for firms with 20-49, 50-99, and 100-249 employees vary around 9 percent. These values are consistent with the average values ECA countries reported here.

83

SMEs in comparator countries was in the range of 5-6 percent slower than that of micro firms. Figure 3- 4b shows that, overall, Turkish SMEs grow substantially more slowly than SMEs in comparator countries. The difference for both small (11-50 employees) and medium firms (51-250 employees) is of the order of 10-12 percent, with medium-sized firms confirmed as the slowest growing group.53 These results can explain why there are more SMEs at old ages in Turkey relative to the comparator countries, while both the age composition and the growth rates of large Turkish firms are internationally comparable.

3.2 Investment Climate Constraints to SME Growth 3.9. Reasons for the slow growth of Turkish SMEs can be found in features of the business environment.54 Comparison of the evolution of Turkish firms with several comparator countries has shown that growth rates of SMEs relative to micro firms are significantly lower in Turkey. One possible explanation is that existing policies and regulations can have more distortionary effects for SMEs than for both micro and large firms, since SMEs have neither the capacity of large firms nor the flexibility of micro firms to cope with the effects of these policies.55 For instance, incentives to hire new workers and expand may directly be influenced by labor regulations, which impinge on the cost of hiring and firing, by broader regulatory issues affecting the cost of doing business or by access to finance, with the availability of funding determining the pace at which firms can expand production and, in consequence, employment.

Figure 3-5: Single most severe investment climate obstacles, by firm size

40 34 Micro Small Medium Large

30 26 24 25

19 2019 19 20 18 18 1515 15 14 13 13 12 11 10 8 5 5 4 3 4 4 3 2 3 3 3 3 2 2 2 2 1 1 1 1 0 1 0 0 1 0 Access to Tax rates Political Informal Unskilled Licenses Corruption Customs Electricity Transport Labor finance instability competition workforce and permits regulations regulations

Source: Turkey ES 2008

3.10. Access to finance is perceived as the single most severe obstacle by firms of all sizes, and especially by medium-sized firms. According to the 2008 enterprise survey, firms of all size categories perceive access to finance as the single most severe obstacle (Figure 3-5), with medium-sized firms appearing to be particularly affected (34 percent), followed by micro (26 percent), small (24 percent) and large firms (19 percent). Tax rates also rank quite high as an obstacle, especially for micro and small firms. Adequate skill levels are perceived as a problem mainly by large firms. The stringency of labor regulations, a possible direct cause of slow employment growth, rank relatively low among firms‘ concerns, being cited as the most severe obstacle by only 4 percent of medium-sized firms, 3 percent of

53 Table A-3-7 in the Appendix shows that the difference in growth rates between Turkish SMEs and SMEs in comparator countries is statistically significant. 54 Aterido, Hallward-Driemier, and Pages (2009), using data on more than 56,000 enterprises in 90 countries, confirm the importance of the business environment for employment growth, with effects varying across firms of different sizes. 55 This ―missing middle‖ hypothesis is discussed in several studies. See, for example, Gauthier and Gersovitz (1997), Sleuwaegen and Goedhuys (2002), and Van Biesebroeck (2005).

84

micro firms and by a negligible proportion of firms in other size categories. As previously mentioned, this result should be interpreted with care, since the perceptions of respondents may have be influenced by the timing of the survey.

3.11. Other indicators in the survey confirm that Turkish SMEs are dependent on bank finance but their applications for bank credit are faced with onerous collateral requirements and high rejection rates. The perceived severity of access to finance as an obstacle broadly decreases with firm size, with 16 percent of micro firms citing it as a major or very severe obstacle, compared to only 10 percent of large firms, and with medium-sized firms appearing marginally more concerned (14 percent) than small firms (13 percent) (Figure 3-6a). At the same time, Turkish firms – as discussed in Chapter 2 – are more dependent than peers in other countries on bank finance to fund their investments in fixed assets. This is especially true for medium-sized firms, for whom bank finance accounts for 47 percent of total funding (Figure 3-6b). Collateral requirements also appear particularly onerous for SMEs, compared to both micro and large firms, amounting to 100 percent of loan value for small firms and 91 percent for medium firms (Figure 3-6c). Notwithstanding the higher collateral requirements, the amount of rejected loan applications is also substantially higher for SMEs (17 percent) compared to more creditworthy large firms (12 percent) (Figure 3-6d).

Figure 3-6: Access to Finance for SMEs

(a) % of firms perceiving access to finance as a "major" (b) Sources of funding for the purchase of fixed assets or "very severe" obstacle 100% 2 18 16.4 5 2 Other means 90% 5 16 14.1 80% 29 34 14 13.3 34 Credit from 70% 42 suppliers/customers 12 10.5 60% State-owned banks 10 50% 8 40% Private banks 6 30% 4 20% 2 10% New equity shares issued 0 0% Internal funds Micro Small Medium Large Micro Small Medium Large

(c) Value of collateral as % of loan value (d) % of firms with loan applications rejected

120 25 100 100 91 19 20 17 17 76 80 70 15 12 60 10 40

20 5

0 0 Micro Small Medium Large Micro Small Medium Large

Source: Turkey ES 2008

85

3.12. Access to bank credit for SMEs has further deteriorated in the aftermath of the global financial crisis, but SME lending still constitutes a strong growth area for banks. In terms of loans, the SME sector has started to lose ground since late 2007, with the slowing of domestic growth and escalation of the global credit crisis. During this period, SME share in total credit has declined by about 5 percentage points to little over 20 percent, and its share in comparison to total corporate credit dropped by even more from about 52 percent to some 44 percent. In terms of growth rates, growth in total banking sector credit remained relatively high and unchanged until the escalation of the global crisis in late 2008, whereas SME credit growth started to decelerate as early as the beginning of 2008. The differences become more dramatic when SME and non-SME corporate credits are compared (Figure 3-7). For the whole period for which data are available (from December 2006 through November 2009) cumulative growth in SME credit amounted to some 35 percent, but that was about half the rate of growth in other (non-SME) corporate credit. In terms of problem loans, SMEs also looked to be faring much worse than the rest of the sector, with the gross NPL ratio of the sector rising to almost 8 percent, from below 4 percent in the middle of 2008. However, despite the fact that SME credit appears more affected than other types of credit as a consequence of the crisis, SME lending still constitutes a strong growth area for banks, and banks are probably eager to tap that potential.

Figure 3-7: Loans to SMEs

SME Loans SMEs Loans (as % of Total Loans) (as % of Total Corporate Loans) 35 56 Total TRY 54 FX 30 52

25 50

48

20 46

44 15

42 Loans=Cash Loans+Non-Performing Loans 10 40 Dec-06 Apr-07 Aug-07 Dec-07 Apr-08 Aug-08 Dec-08 Apr-09 Aug-09 Dec-06 Apr-07 Aug-07 Dec-07 Apr-08 Aug-08 Dec-08 Apr-09 Aug-09

Source: World Bank (2010b)

3.13. Econometric results confirm that access to finance is the area of the investment climate most consistently associated with the ability of firms to generate employment and grow in size.56 In order to test the hypothesis that the investment climate – and finance in particular – have an effect on firm growth, regression analysis is conducted incorporating variables from the 2008 survey.57 Results indicate that one percent more usage of external finance for investment – including all sources other than internal funds or retained earnings, such as issuance of new equity, public and private banks, purchases on credit from suppliers and advances from customers, money lenders, and non-banking financial institutions – is related with 0.3 percent higher employment growth. The association of a loan or a line of credit with employment growth is even stronger and is estimated to have a positive effect on employment growth of

56 Inclusion of the IC variables as explanatory variables is not straightforward. As discussed in the previous Chapter and in this Annex, the inclusion of the investment climate variables can cause endogeneity problems even when the chosen variables are not subjective measures that depend on firms‘ perceptions of the business environment. For instance, faster growing firms might be more likely to have access to external finance or they might be interacting more with government officials. To address this issue, in this Chapter investment climate variables are used as industry-region averages. 57 See Annex 2-B for a complete list of investment climate variables used for the analysis.

86

33 percent (Table 3-2).58 Furthermore, when controlling for access to finance, medium-sized firms are confirmed as the slowest growing group.59 Estimation based on principal component analysis adds robustness to the results indicating that a one percent increase in the block of variables connected with access to finance is associated with 6 percent faster employment growth. 60 At the same time, faster growing firms are more likely to perceive labor regulations as an obstacle and to spend time clearing customs to export. Among other investment climate variables the perceived stringency of labor regulations is strongly and positively associated with firm growth. Unlike other variables in Table 3-2, this captures the subjective sentiment of firm managers and may indicate that faster growing firms are more bothered with labor regulations, given their needs to expand employment. The number of days required to clear customs for export is also positively associated with employment growth. This result may indicate that exporters, who naturally spend time at the customs to ship their products, are also the faster growing firms.

Table 3-2: Summary of investment climate effects on firm growth, percent Effect of IC Small Firms: variable on Medium Firms: Large Firms: growth rate growth rate growth rate growth rate IC Variable relative to relative to relative to micro relative to micro micro firms micro firms firms (%) firms (%) (%) (%) Labor regulation +41.4 -13.1 -14.8 -10.7 New fixed asset finance - external 0.3 -13.0 -15.2 -11.3 Collateral (*) NS NS -15.6 NS Purchase paid after delivery NS -14.0 -15.7 -11.4 Overdraft facility (dummy) NS -13.7 -15.4 -11.0 Loan (dummy) +33.4 -14.5 -16.3 -12.5 Loss from power outages (*) NS -21.8 -26.6 -22.0 Informal payments (*) NS -15.7 -16.7 -9.6 Tax inspections (number) (*) NS -19.5 -19.9 -15.5 Time tax NS -11.6 -12.3 -8.6 Customs clearance for export (days) (*) +1.0 NS -24.7 NS Customs clearance for import (days) (*) NS -37.9 -48.9 -50.1 Inspections (number) NS -13.0 -15.9 -12.1 Notes: See Table 3-A-8 in the Annex. (*) Less than 600 observations. NS “Not Significant” Source: Turkey ES 2008

3.3 Enhancing the Ability of SMEs to Access Bank Credit 3.14. Improving the connection between the banking sector and SMEs is a crucial development challenge. It was mentioned that large Turkish enterprises are able to benefit from the more positive aspects of the investment climate, including the unprecedented expansion of credit to the private sector from about 15 percent of GDP in the late 1990s, to over 30 percent today. At the same time, micro firms are often sufficiently flexible to overcome the more negative aspects of the investment climate and are able to meet their financial needs with internal resources. Within this environment, small and medium

58 See Tables 3-A-8 and 3-A-9 in the Annex for detailed results. Micro firms are confirmed as the fastest growing group, followed by small firms, while, in most specifications, medium firms are the slowest growing group. 59 Results in Table 3-A-9 indicate that access to finance does not have a strong differentiating effect on the growth of firms in different size groups, with some evidence that small and large firms find the absence of a line of credit more constraining than firms in other size groups. Given the relatively small number of observations available we prefer to use as main reference the data in Table 3-A-8 where the absence of interaction terms allows preserving more of the information contained in the data. 60 The Principal Component method transforms a number of possibly correlated variables – as the variables in the block of access to finance – into a smaller number of uncorrelated variables named principal components, which are ranked depending on the amount of variability in the data each of them contains. See Table 3-A12 in the Annex for details.

87

firms appear to bear a large portion of the negative effects of an unfriendly investment climate. This affects the ability of SMEs to grow and the factor most clearly associated with firm growth appears to be inadequate access to finance. These conditions have certainly been exacerbated by the repercussions of the crisis, which has caused a drop in domestic and external demand and a tightening of credit markets. As a consequence, lending to SMEs, a sector that had been growing significantly since 2005, has actually dropped in relative terms,61 now accounting for less than 22 percent of total lending, compared to almost 24 percent at end-2008. At the same time, while lending rates, after peaking in October 2008, have been on a downward trend, there is anecdotal evidence that there is segmentation in the market, with rates charged to SME clients exceeding by far those offered to larger corporate clients. Moreover, lending maturities have been shortening and collateral requirements increasing.

3.15. The preceding analysis revealed that access to finance is the investment climate obstacle most consistently associated with the ability of SMEs to generate employment and grow in size. In the context of Turkey, with a reasonably developed and sophisticated banking system, increasing the ability of SMEs to have access to affordable bank credit at convenient maturities appears as a crucial development challenge. Banks, however, are reluctant to cater to SMEs due to the difficulties they face in assessing their creditworthiness and their ability to repay debt (Box 3-2).

Box 3-2: Main obstacles to SME lending: views from banks and SME representatives Banks are aware that the credit boom of the past several years is over, and the credit environment will be tighter going forward. But they seem to have nevertheless positioned themselves for stepped up lending to the SMEs. The lessons of Turkey‘s devastating 2001 banking/financial crisis appear learnt. Banks are conscious of risks, while at the same time they seem aware of the need to enhance capacity in the SME sector, and are therefore willing to act as the financial advisory to potential borrowers. Having been in the field for several years, they feel they know the clientele reasonably well despite the many shortcomings related to transparency, lack of collateral and so on. Bank‘s preparations in the run-up to Basel II transition, which was expected to result in a tightening of credit to the SMEs, also helped to enhance the general level of knowledge on the sector.

From the SME side, banks pointed out three major obstacles to lending. Interestingly, the first obstacle has been put forward – broadly defined – as the “absorptive capacity” of the sector. Banks are aware that credit is just one ingredient to a very complex problem of supporting viable and productive firms. The SMEs in Turkey are typically seen strong on the technical side (especially family businesses that have been going on for generations with expertise transferred from father-to-son) but weak on the managerial side. Through what they call ―academies‖, several banks help to develop this expertise and raise awareness in their clientele through intensive training programs, offered on a regional basis.

Second and third constraints in SME lending are very well-known: lack of transparency of financial statements and lack of collateral. As both are closely related to the informality issue in Turkey (tax and wealth evasion), a quick resolution to these problems is not expected too soon. Banks are therefore taking a pragmatic stance: as much as possible, they stick to the usual procedure of ―credit scoring‖ in their lending practices to SMEs, but relationship- building and establishing a track record for the client seem to be on their agenda as well. The issue of collateral also appears to be pervasive. Real property (homes, shops, factories) is the most common form of collateral, but personal letters, guarantees through third persons are also used. Support from government agencies also help to alleviate problems like KOSGEB support, which primarily provides interest subsidy, and collateral provided through the Credit Guarantee Fund (CGF) are helpful but they do not yet make up a significant portion of the SME credit market. According to KOSGEB data for 2004 for instance, only 0.5 percent of the guarantees are provided by the CGF. The new leg of the CGF that involves a TL1 billion contribution from the Treasury has recently become operational but is yet to be fully utilized.

61 Akbank, the third largest private bank, reported a cut of almost 50 percent in its lending to SMEs in the nine months ending in September 2009.

88

SME representatives mention a host of obstacles that complicate the operating environment of a typical SME, such as very high labor taxes, energy costs, the difficulty of adhering to ILO standards (as part of EU accession) and the statutory minimum wage that drives many entities to lesser developed countries like Egypt (in textiles), or Bulgaria in agriculture. But finance appears to be a central problem, too, and complaints there seem two-fold: insufficiency of funds, and as a corollary to that the steep interest rates the banks charge on SME clients, and inadequacy (or inaptness) of banks‘ risk metrics in evaluating SMEs‘ prospects.

Why did SME credit suffer more than the rest during the past two years? The understanding from interviews with banks is that both supply and demand factors were at play. One large bank has explicitly emphasized this by stating that as a business strategy (i.e. to reduce exposure to the more vulnerable clientele) they have curbed lending to the SME sector more than to the rest of the corporate sector. At the same time, they noted, demand for certain types of credit collapsed from SMEs too, because of production cuts (a result of a wait and see attitude) and the collapse in global trade (affecting demand from those engaged in foreign trade). In this connection, banks mentioned that micro firms (those with turnovers less than TL1 million) were the most affected segment, and that having crossed a certain threshold in size appears to have increased the survival rate of firms during the crisis. But bank behavior appears to have played a role in this outcome as well. For instance, some SME representatives expressed the view that banks were willing to extend credit, but that their methods of risk analysis clashed with the dynamics of SME behavior, especially the very small ones, where borrowers were judged on the basis of their turnovers. They also stated that as the size of the firm declines, banks asked more personal (owners‘) collateral, regardless of the business prospects.

Source: World Bank (2010b)

3.16. Proposal: Enhance the ability of banks to assess the creditworthiness of SME borrowers by expanding the coverage of credit bureaus and ensuring the wide adoption of financial reporting standards. Transparent and accessible credit information on firms and improved accounting and auditing standards are crucial for financial institutions extending loans, particularly to SMEs. Empirical research has found that the presence of transparent and accessible credit registries is associated with higher share of private credit to GDP.62 Lenders are able to access borrowers‘ credit history through credit bureaus and are thus able to estimate their ability and willingness to pay back loans. This information is especially important to SMEs, which have smaller loan volumes and often shorter credit backgrounds, making easy to access information essential to lenders. At the same time, ensuring the wide adoption of financial reporting standards would increase the ability of banks to assess the creditworthiness of SME borrowers. In particular, the following developments would be appropriate:

 Encourage the expansion of coverage of existing credit bureaus, the Credit Registry of the Central Bank and the Credit Bureau of Turkey (KKB). Whereas the Credit Registry of the Central Bank covers both firms and individuals, the private Credit Bureau of Turkey (KKB) has not yet expanded to monitoring of the corporate sector. Although a planned expansion to also cover firms has not yet been implemented, the current monitoring of individual consumers is also very useful, especially for information on small enterprises. Since entrepreneurs, as earlier mentioned, do not necessarily get their business loans approved, they often finance their investments through personal loans, and the credibility of the owner is thus often reflected on the credibility of their operations. Nonetheless, a completion of the planned expansion will significantly improve Turkish financial institutions‘ possibilities to efficiently direct loans to the most creditworthy customers, both individuals and corporations.

 Accelerate adoption of the new Commercial Code in order to enable SMEs to benefit from a simplified set of financial reporting standards. The Turkish Accounting Standards Board (TASB) established in 2002, has been publishing Turkish Financial Reporting Standards (TFRS) based on International Financial Reporting Standards (IFRS) since October 2006. TASB regularly updates

62 See for instance, Djankov et al. 2008.

89

and publishes the standards in line with the revisions in the IFRS. The draft Commercial Code (CC), which has been pending enactment in Parliament since November 2005, accepts TASB as a sole and exclusive authority to set Turkish Accounting Standards and requires that the financial statements of all financial entities regardless of public interest or not be prepared in accordance with TFRS. The code enables simplified set of standards to be adopted by SMEs if endorsed by TASB. This authority that will be appointed to TASB with the enactment of the CC requires TASB to have adequate technical and financial resources. However, currently TASB lacks such resources that will enable it to provide technical support and guidance to the market for the adoption of TFRS.

The role of Government: expanding the reach of existing schemes in support of SME lending

3.17. A number of government schemes are aimed at addressing financial bottlenecks to the expansion of SMEs. Several organizations currently provide financial and non-financial assistance for the development of SMEs. They include the Small and Medium Enterprises Development Organization (KOSGEB)63, Halkbank, Turkey's SME and Tradesman Bank, and the Credit Guarantee Fund (CGF).64

3.18. The Credit Guarantee Fund supports SMEs by providing a guarantee for their financing thus increasing their ability to benefit from bank credit. The CGF became operational in the early 1990s, but it has grown substantially since 2007. Access to CGF support makes it possible for SMEs with inadequate collateral to apply for bank credits. The CGF provides guarantees through banks to SMEs with fewer than 250 employees and turnover below TL 25 million. Application for CGF support is carried out with partner banks as well as through a network of banks with which CGF has an agreement. The upper limit for guarantees provided by CGF amounts to TL 1 million. In its mission, CGF gives high priority to supporting young and female entrepreneurs, promoting innovative investments, supporting export, increasing the rate of employment, and contributing regional development. There is no restriction imposed on the type of credit for which enterprises seek a CGF guarantee. In order to ease the impact of the crisis on SMEs, the Medium Term Program for 2010-2012 calls for a stronger and more influential CGF backed by the Undersecretariat of Treasury. (Box 3-3).

63 The mission of KOSGEB and its focus on ―SME Development and Support‖ is specified in the Council of Ministries‘ Decision No. 2008/13524 as ―promoting investment, export and production effectiveness of SMEs.‖ Since 2004, KOSGEB provides financial support by covering interest payments on credits, while leaving the credit risk to the bank. The upper limit for this type of credit is TL 300,000. Applications are processed through a network of public, private and branches of foreign owned banks and the benefiting SME remains responsible for the principal amount according to a predetermined payback schedule with the limit of 48 months. In 2009 the Turkish government increased its funding to KOSGEB by 48 percent. Moreover, through the use of a web-based system, application processes and monitoring of loans has become highly efficient. 64 Halkbank provides financial support to SMEs for export credit and capital investment along with credit for investment and operations to improve productivity and increase employment. Credit is given for 5 years for investment purposes, without requesting any repayment in the first year, and for 4 years for operations purposes. Credit on similar terms – 5 years of maturity without repayment in the first year – is offered to manufacturing firms for purchases of machinery and equipment. Firms in manufacturing sector, agribusiness, tourism, education, energy, IT sectors can benefit from medium- to long-term credit with 7 year maturity and no repayment required for the first 2 years. A detailed treatment of KOSGEB‘s activities is provided in Ch. 4.

90

Box 3-3: The Credit Guarantee Fund (CGF) The idea of a credit guarantee scheme in Turkey was Figure 3-8: Guarantees provided by the CGF as a initiated in 1991 with the motivation to ease the credit percentage of the total SME loan volume65 constraints faced by SMEs. A partnership was later formed 250,000 0.30% between TOBB, KOSGEB, occupational groups ―TOSYOV, 0.249% 66 TESK, MEKSA‖ and Halkbank to realize this initiative. 200,000 0.25% CGF became operational in early 1990s, but it has been 0.20% growing substantially since 2007. By 2009, the CGF had 150,000 0.15% accumulated a total capital of USD 165 million, generating a 100,000 217,561 0.10% 0.068% 0.060% guarantee volume of USD 426.5 million, and with a 0.051% potential67 of USD 586 million. Figure 3-8 demonstrates the 50,000 0.05% ThousandTL 37,280 42,735 44,590 increased capacity of the CGF after a capital injection from 0 0.00% its shareholders in 2007. 2005 2006 2007 2008 Real guarantee volume (2005 prices) (LHS) Under the scheme, only loans with a maximum volume of Ratio of CGF collaterals / SME loans about USD 835,00068 were collateralized, with the amount Source: Credit Guarantee Fund, BRSA of support provided by the CGF ranging between 50-80 percent of the loan, in return for a commission of 1-2 percent of the outstanding loan amount. In this setup, CGF worked with 38 financial institutions through individual protocols, and was the responsible body for undertaking the follow-up procedures of non-performers of its collateral share of the loans.

With the aim of providing liquidity to the real sector, in June 2009, the Parliament approved the law69 amending the Law on Public Finance and Debt Management70 and called for the provision of guarantees totaling USD 650 million via a reformed Credit Guarantee Fund71. 20 banks became shareholders to the CGF, and its structure changed radically. In this new setup, the CGF is envisaged to undertake two separate functions:

(i) Traditional role: The CGF will carry on extending guarantees under the traditional scheme. Its capacity will allow for holding a maximum risk of about USD 825 million.72 For the utilization of this fund, banks may send a demand to CGF for it to support the debtor under its own risk, abiding by the previous protocols. Decisions over loan guarantees are taken by credit committees, which are composed of 5 members, namely, the General Manager of the CGF and one representative from each of TOBB, KOSGEB, banks and other shareholders.

(ii) Fund under Treasury support: After a long period of negotiations at the ministerial level between the Treasury, CGF (including each of its shareholders) and banks (under the umbrella of the Banks‘ Association), all parties agreed on the establishment of a new credit guarantee system, for which the Treasury support under Law no.4749 will remain in effect for 2 years following its establishment. A major reform in the new setup is, in addition to the endorsement of the Treasury for extending guarantees, the participation of the banks to the shareholding structure of the CGF. Under this arrangement, 20 banks have now become shareholders in the CGF with equal shares, quadrupling its paid-in capital and lending

65 Total loan volumes are calculated including cash and non-cash loans. End-year values were taken into account. 2005 SME loan volume uses an estimate based on the SME loan volume/Total loan volume in the following years. 66 TOSYOV: Small and Medium Industry Owners and Managers Foundation of Turkey TESK: Merchants and Artisans Confederation of Turkey MEKSA: Occupational Training and Small Industry Support Foundation of Turkey 67 Banks accepted guarantees up to 5 times the total capital of CGF (Leverage multiplier is 5). 68 Maximum loan volume was TL 750,000. In this Chapter, amounts legally declared in TL terms have been converted into US dollars using the interbank rates on dates which apply to associated legal arrangements. 69 Law no. 5909 published on Official Gazette no. 27268 dated 06/24/2009. 70 Law no. 4749. 71 The principles of the reformed CGF is set out in the Council of Ministers Decision no. 2009/15197 published on Official Gazette no. 27289 dated 07/15/2009. 72 As the multiplier effect generates TL 1.2 billion of guarantees out of TL 240 million, roughly.

91

capacity. This gave these banks the right of representation in the CGF Board of Directors with 2 members out of 9. As concerns the remaining members, 2 are appointed by the Treasury, 2 by KOSGEB and 3 by TOBB, with TOBB also appointing the chairman of the Board.

Figure 3-9: Flow of funds in guarantee agreements In this new fund structure, all loan applications are under treasury support planned to be channeled through the 20 banks which participate in the system. Remaining banks that have a protocol with the CGF may obtain guarantees for their creditors through the traditional scheme. Decisions on loan guarantees will be made by ―credit approval committees‖, composed of 5 members, namely, the General Manager of the CGF and one representative from each of (i) TOBB, (ii) KOSGEB, (iii) banks and (iv) Treasury. The representative of the Treasury is empowered with a veto power over decisions related to the loans under its guarantee.

For each ―Treasury-supported‖ loan, 65 percent of the risk will be borne by the Treasury and the remaining 35 percent will be collateralized, leaving no risk to the CGF.

Loans with maturities less than 4 years and a maximum volume of about USD 685,000 will be supported. Unlike the traditional role of the CGF, follow-up procedures will be fully undertaken by the associated banks, with a payback mechanism channeled from the Treasury to the banks. The banks will later transfer the collected collateral amount to the CGF, for it to pass on to the Treasury. The support will be provided to beneficiaries in exchange for a ―guarantee commission fee‖ of 0.5-1.5 percent on the loan amount.

Source: World Bank, REGE-DPL

3.19. Placing Turkey in an international perspective, as presented in Box 3-4, the Turkish CGF seems to have lower share of GDP in guarantees, with the outstanding stock amounting to 0.09 percent of GDP. However, combined capacity of the traditional and Treasury-funded loans is 0.19 percent, putting Turkey closer to the average of higher-income countries. As to funding and management of the Turkish CGF, responsibilities are shared, with the Government – through KOSGEB – together with TOBB representing almost exclusive involvement in all four areas (funding, management, risk assessment and recovery). Additionally, the European Investment Fund (EIF) has committed to counter-guaranteeing up to 50 percent of CGF‘s investment credit guarantees.

Box 3-4: Credit guarantee schemes: International comparison The form and scope of credit guarantee funds (CGFs) has been proven to vary significantly across countries, both in coverage and actual implications on firm performance. A World Bank survey on CGFs from 2008 indicated that outstanding guarantees in relation to GDP averaged 0.35 percent for transition economies, whereas the corresponding share in high income countries was 0.21 percent. The survey also shows that the median age of interviewed CGFs is 15 years, with an average of 27 years in high-income and 13 years in low-income countries. As presented in Table 3-3, Asian economies have particularly active CGF presence, with the guarantee stock reaching 3.5 and 5 percent of GDP in Japan and Korea.

There is a large variation in ownership structure, pricing control and risk management but the World Bank survey does show an overall important role of government in funding of CGFs. Nevertheless, relative to the private sector, government seems to have less influence in management, risk appraisal and recovery, indicating that the lending banks are usually left with the responsibility of credit risk assessment and recovery of defaulting loans.

92

Table 3-3: Comparison of CGFs Maximum Total loan; max as outstanding share of GDP Cost of loan guarantee Guarantee per capita guarantee Target firms Industries value/GDP Risk management Turkey CGF Up to 80% USD 835,000; 1-2% per Less than 250 Agriculture, 0.09% (2009) Counter guarantee 53.5% annum employees and manufacturing, by Treasury and less than TL services and remaining part by 25m annual mining sectors, the guaranteed turnover limited to firms. capital formation U.S. Small 75%; USD 2.2m; 1.3% one- Less than 100 Emphasis on N.A. Lender oversight Business 46.2% time subsidy employees women, office supervises Administration of the veterans & lender activity & Preferred Loan guaranteed undeserved more than half of Program loan value markets guarantees extended to preferred lenders. UK Enterprise 75% USD 1.6m; 2% per Less than GBP All but 0.07% Risk assessment of Financing 35.3% annum 25m annual agriculture, (Guarantee borrower delegated Guarantee turnover transport, stock 2005- to lender. Time Scheme (increased finance and 2008) restricted program from GBP insurance until March 2011. 5.6m in 2009) sectors Japan JASME 100% USD 1-6m, Set by 50-300 Manufacturing, 5% (2007) Mainly providing Agency Loans depending on JASME employees, wholesale, guarantees during end use; 26- depending on retail & periods of tight 157% sector services credit. No direct risk monitoring system. Korea CGF 50-100% USD 1.9m, 0.5-2% per Unspecified, Emphasis on 3.5% (2008) Lower guarantees with annum based but 99% of innovative over time to induce exceptions to on risk guarantees to enterprises and lenders' monitoring USD 7.5m; SMEs new firms in all and credit 88.2% sectors, except evaluation of luxury goods guaranteed firms. Sources: Honohan (2008), Heron & Co. (2007), Beck et al. (2008), Turkey CGF, Korea CGF, JASME

Firms regarded as eligible for guarantees also differ greatly across countries, with dissimilarities in firm size, annual turnover and industry. Maximum loan guarantees reach as high as USD 2.5 million in the United States (firm size is however limited to 100 employees). In regards to risk assessment and credit appraisals of final borrowers, this is in many best-practice cases not done by the CGFs, but instead performed by the intermediary financial institutions giving out the actual loan (Honohan, 2008), usually creating higher efficiency and lower operating costs for the CGFs. An incentive for lenders to carry out credit appraisals is shared risk with the CGF (partial credit guarantee funds). On average, the CGFs guarantee 70-80 percent of loans, with shares going as high as 100 percent in Japan and Korea. In the case of very high guarantees, the presence of broad and high-functioning credit information systems is desirable in order to ensure eligibility of borrowers.

3.20. Proposal: Consider ways to make the new CGF scheme more active, expand its reach, and allow it to better reach the medium-sized firm segment. Collateral requirements, traditionally onerous for smaller enterprises, have become even more of a constraint to bank lending in the aftermath of the 2008-2009 economic and financial crisis. For many years, and especially since its recapitalization in 2007, the CGF has played an important role in facilitating access to credit for SMEs. The new CGF model, with Treasury involvement for a period of two years, is a positive initiative that expands the capacity of the CGF to serve the financing needs of SMEs following the credit crunch in the aftermath of the crisis. Furthermore, a collateral rate of 35 percent demanded of beneficiary firms is a wise measure that will minimize the risk of misuse of funds. The Government could build upon these positive

93

developments by accelerating the implementation of the new CGF scheme. Furthermore, the CGF could more fully perform its role if it was better targeted to serve the needs of medium-sized firms. In fact, since 1994, the bulk of credit guarantees provided by the CGF has benefited micro (40 percent) and small (49 percent) enterprises. Since 1994, 23 percent of the guarantee fund has been used by medium-sized firms (50-249 employees).

References Aterido, Reyes, Mary Hallward-Driemeier, Carmen Pages. 2009. ―Big Constraints to Small Firms‘ Growth? Business Environment and Employment Growth across Firms,‖ World Bank Policy Research Working Paper 5032.

Beck, Thorsten, Asli Demirgüç-Kunt and Vojislav Maksimovic. 2005. ―Financial and Legal Constraints to Growth: Does Firm Size Matter?‖ Journal of Finance 60(1): 131-177

Beck, Thorsten, Leora F. Klapper and Juan Carlos Mendoza. 2008. ―The Typology of Partial Credit Guarantee Funds around the World,‖ World Bank Policy Research Working Paper 4771.

Cooley, Thomas F. and Vincenzo Quadrini. 2001. "Financial Markets and Firm Dynamics," American Economic Review, 91(5):1287-1310.

Dunne, Timothy, Mark J. Roberts, Larry Samuelson. 1989. ‖The Growth and Failure of U.S. Manufacturing Plants,‖ The Quarterly Journal of Economics, Vol. 104, No.4, pp.671-698.

Djankov, Simeon, Oliver Hart, Caralee McLiesh, Andrei Shleifer. 2008. ―Debt Enforcement around the World.‖ Journal of Political Economy, 2008, vol. 116, no. 6.

Evans, David S. 1987a. ―The Relationship between Firm Growth, Size, and Age: Estimates for 100 Manufacturing Industries,‖ Journal of Industrial Economics, Vol. 35, pp.567-82.

Evans, David S. 1987b. ―Tests of Alternative Theories of Firm Growth,‖ Journal of Political Economy, Vol. 95, pp.657-74

Gauthier, Bernard and Mark Gersovitz. 1997. ―Revenue Erosion through Exemption and Evasion in Cameroon 1993,‖ Journal of Public Economics, Vol.64, pp.407-24.

Heron & Company, for Canadian Small Business Financing Program. 2007. ―Review of SME Loan Guarantee Programs: Analysis and Comparison of Administrative Features.‖ Ottawa

Honohan, Patrick. 2008. ―Partial Credit Guarantees: Principles and Practice,‖ Prepared for the World Bank Conference on Partial Credit Guarantees, March 13-14, 2008

Jovanovic, Boyan. 1982. "Selection and the Evolution of Industry," Econometrica, 50(7): 649-67

Klepper, Steven and Peter Thompson. 2007. "Submarkets and the Evolution of Market Structure," Rand Journal of Economics, 34(4):862-888.

Klette, Tor J, and Samuel S. Kortum. 2004. "Innovating Firms and Aggregate Innovation," Journal of Political Economy, 112(5):986-1018.

Sleuwaegen, Leo and Micheline Goedhuys. 2002. ‖Growth of Firms in Developing Countries, Evidence from Cote d‘Ivoire,‖ Journal of Development Economics, Vol. 68, pp.117-35.

Van Biesebroeck, Johannes. 2005. ―Firm size matters: growth and productivity growth in African manufacturing,‖ Economic Development and Cultural Change, Vol. 53(3), pp.545– 83.

94

World Bank. 2007. Turkey Investment Climate Assessment. Washington, DC

World Bank, Poverty Reduction and Economic Management, Turkey Country Management Unit, Restoring Equitable Growth and Employment Programmatic Development Policy Loan (REGE-DPL). Draft Dec. 14, 2009

World Bank, 2010a. Innovation Policy for Competitiveness and Employment Generation. Forthcoming

World Bank. 2010b. ―The Credit Environment of Small and Medium Enterprises in Turkey: Some Observations,‖ Mimeo, Prepared by Murat Ucer.

Annex 3-A. Descriptive Statistics and Econometric Analysis Table 3-A-1 shows the number of firms in the sample in each size and age group for both all data and for panel data only. As expected, larger firms are more likely to be older.

Table 3-A-1 Size and age distribution ALL FIRMS PANEL FIRMS ONLY Employ \Age 1-5 6-15 ≥16 Total 1-5 6-15 ≥16 Total Micro 0.30 0.42 0.28 206 0.18 0.42 0.40 60 Small 0.18 0.48 0.33 374 0.15 0.48 0.37 158 Medium 0.09 0.38 0.53 223 0.09 0.36 0.55 139 Large 0.05 0.38 0.57 111 0.03 0.29 0.68 62 Total 155 395 364 914 50 169 200 419

The survey includes data from manufacturing, retail, and other service sectors. The two digit classification of the industries is made according to ISIC revision 3.1. The list of industries covered in the data and the number of firms at different size groups are given in Table 3-A-2. The largest industry in terms of number of observations is Textiles which is followed by Food industry. For most industries, size distribution has a long right tail. Small firms make the largest group of firms in almost all industries. The textile industry has the highest share of large firms. The survey data is also inclusive of all regions in Turkey.

Table 3-A-2 Proportion of firms in industries included in the survey 2 Digit Industries Micro Small Medium Large Total Food 0.22 0.42 0.26 0.10 130 Textiles 0.07 0.30 0.35 0.28 149 Garments 0.13 0.38 0.35 0.15 96 Chemicals 0.26 0.44 0.22 0.07 95 Plastics & rubber 0.19 0.37 0.33 0.11 27 Non metallic mineral 0.28 0.49 0.22 0.01 83 Basic metals 0.23 0.38 0.31 0.08 13 Fabricated metal products 0.23 0.57 0.10 0.10 30 Machinery and equipment 0.34 0.28 0.28 0.09 32 Electronics (31 & 32) 0.33 0.11 0.44 0.11 9 Services of motor vehicles 0.39 0.45 0.03 0.13 31 Wholesale 0.32 0.57 0.09 0.02 53 Retail 0.38 0.33 0.13 0.16 93 Others* 0.13 0.53 0.25 0.09 68 Total 208 375 223 114 920 * Other industries include Other manufacturing, IT and Transportation sectors

There are five macro- regions covered which are listed in Table 3. Around 40 percent of firms in all regions are small firms. 40 percent of firms in Black Sea-Eastern region are micro firms. This region has the lowest share of large firms.

95

Table 3-A-3 Proportion of firms in regions included in the survey Regions Micro Small Medium Large Total Marmara 0.14 0.39 0.28 0.19 334 Aegean 0.18 0.39 0.29 0.14 156 Central Anatolia 0.28 0.39 0.26 0.08 145 South 0.29 0.45 0.19 0.07 196 Black Sea - Eastern 0.40 0.44 0.12 0.03 89 Total 208 375 223 114 920

Table 3-A-4 Growth rates of firms at different size and age groups ALL FIRMS PANEL FIRMS ONLY Growth\Age 1-5 6-15 ≥16 Total 1-5 6-15 ≥16 Total Micro 0.29 0.14 0.14 0.18 0.22 0.13 0.24 0.19 Small 0.14 0.02 -0.02 0.03 -0.02 0.02 0.03 0.02 Medium 0.00 0.03 0.01 0.02 -0.31 -0.08 -0.05 -0.08 Large 0.26 0.14 -0.04 0.06 -0.09 -0.15 0.00 -0.03

Table 3-A-5 Fraction of firms in each size-age cell in comparator countries Turkey Russia Employ\Age 1-5 6-15 >=16 Total Employ\Age 1-5 6-15 >=16 Total Micro 0.29 0.45 0.26 16,572 Micro 0.37 0.58 0.05 19,554 Small 0.15 0.54 0.31 17,710 Small 0.28 0.63 0.09 28,318 Medium 0.09 0.31 0.60 5,065 Medium 0.08 0.73 0.19 21,734 Large 0.06 0.45 0.49 1,932 Large 0.02 0.50 0.48 7,387 Total 7,945 19,511 13,822 41,279 Total 17,258 48,659 11,077 76,994

Ukraine Poland Employ\Age 1-5 6-15 >=16 Total Employ\Age 1-5 6-15 >=16 Total Micro 0.36 0.63 0.01 20,933 <=10 0.09 0.63 0.28 34,911 Small 0.24 0.58 0.19 16,489 11-50 0.05 0.62 0.32 15,928 Medium 0.15 0.60 0.25 4,887 51-249 0.04 0.46 0.50 5,920 Large 0.10 0.28 0.62 1,920 >=250 0.13 0.56 0.31 1,176 Total 12,297 26,218 5,713 44,229 Total 4,336 35,210 18,389 57,935

Romania All ECA Employ\Age 1-5 6-15 >=16 Total Employ\Age 1-5 6-15 >=16 Total Micro 0.36 0.63 0.01 26,556 Micro 0.25 0.64 0.10 269,735 Small 0.17 0.81 0.02 20,608 Small 0.19 0.68 0.13 225,957 Medium 0.11 0.73 0.15 5,195 Medium 0.11 0.63 0.25 76,951 Large 0.31 0.49 0.20 1,010 Large 0.09 0.45 0.46 20,678 Total 13,944 37,720 1,706 53,370 Total 121,983 384,250 87,087 593,320

EU8 EU10 Employ\Age 1-5 6-15 >=16 Total Employ\Age 1-5 6-15 >=16 Total Micro 0.18 0.69 0.13 124,967 Micro 0.21 0.68 0.11 166,805 Small 0.18 0.71 0.11 88,097 Small 0.18 0.73 0.09 121,066 Medium 0.11 0.67 0.21 25,969 Medium 0.12 0.68 0.20 32,964 Large 0.17 0.48 0.35 4,501 Large 0.18 0.50 0.32 5,938 Total 42,038 167,904 33,592 243,534 Total 61,424 227,221 38,128 326,773

Results of econometric analysis The specification used in the regression analysis is given in equation (1)

(1)

Annual growth rate of full time employment of firm i at time t (between 2004 and 2007) is represented by git. Smallit-3, Mediumit-3, and Largeit-3 stand for the indicators of firm size in 2004, obtained from the retrospective question on employment asked in the 2008 survey. The omitted group is micro firms. Ageit-3 is the age of the firm in 2004. Exportit, Foreignit, and Govtit are the other firm level control variables which are only available in 2007. In the

96

regressions for the panel data use is made of the values of the control variables from the 2005 survey. The size groups in the regressions for panel data are constructed according to the values in the 2005 survey. Finally Industryi and Regioni control for the 2-digit industry and the region of the firm. In all regression analyses performed in this study, probability weights are used in order to obtain results that are representative for all firms in Turkey. All standard errors are clustered to allow for possible correlations in growth rates across firms within the same industry and region. The regression results are given in Table 3-A-6.

Table 3-A-6 Regression of employment growth rates All Data Panel Firms Small (10-50) -0.14 -0.149 (0.039)*** (0.083)* Medium (51-250) -0.155 -0.233 (0.033)*** (0.068)*** Large (≥251) -0.114 -0.191 (0.060)* (0.079)** Age -0.003 0.002 (0.001)** (0.002) Foreign -0.032 0.02 (0.072) (0.096) Govt 0.037 0.23 (0.045) (0.041)*** Exporter 0.018 -0.047 (0.033) (0.052) Const 0.321 0.184 (0.089)*** (0.125) Obs 668 416 R2 0.201 0.144 Robust standard errors which are given in parentheses are clustered by region and 2-digit industry for all data and by 2-digit industry for the panel data. Additionally control for 2-digit industry and region fixed effects. *** p<0.01, ** p<0.05, * p<0.1.

In order to analyze whether the growth rates of firms in Turkey are significantly different from comparator countries, a dummy variable indicating Turkish firms is included. Interaction of this dummy variable with the size groups reveals whether the average growth rate in each size group in Turkey differs from the growth of firms in the same size group in the comparator countries. Extending equation (1) to include TRt, as the dummy variable for Turkey, we obtain

(2)

Regression results are given in Table 3-A-7. Each column shows the result of the regression that includes only the data from the comparator country (or group of countries) and Turkey. All_ECA column includes all countries in the region.

Table 3-A-7 How firms in turkey differ from the firms in comparator countries in their growth rates Russia Poland Ukraine Romania All_ECA EU8 EU10 Small -0.073 -0.059 -0.077 -0.077 -0.055 -0.056 -0.057 (0.025)*** (0.022)*** (0.023)*** (0.021)*** (0.011)*** (0.019)*** (0.016)*** Medium -0.058 -0.038 -0.066 -0.064 -0.056 -0.051 -0.050 (0.022)*** (0.022)* (0.032)** (0.042) (0.012)*** (0.019)*** (0.017)*** Large -0.063 -0.088 -0.068 -0.182 -0.069 -0.058 -0.071 (0.024)*** (0.040)** (0.037)* (0.049)*** (0.017)*** (0.036) (0.031)** Small*Turkey -0.085 -0.098 -0.088 -0.076 -0.108 -0.106 -0.098 (0.039)** (0.036)*** (0.036)** (0.035)** (0.031)*** (0.035)*** (0.034)*** Medium*Turkey -0.112 -0.127 -0.107 -0.077 -0.119 -0.124 -0.113 (0.041)*** (0.035)*** (0.042)** (0.049) (0.031)*** (0.034)*** (0.033)*** Large*Turkey -0.081 -0.048 -0.074 0.069 -0.073 -0.083 -0.060 (0.048)* (0.057) (0.050) (0.063) (0.045) (0.055) (0.052) Turkey 0.037 0.148 0.109 0.073 0.068 0.120 0.109 (0.046) (0.032)*** (0.032)*** (0.034)** (0.037)* (0.033)*** (0.033)*** Constant 0.103 0.111 0.061 0.190 0.105 0.064 0.075 (0.037)*** (0.025)*** (0.045) (0.032)*** (0.033)*** (0.052) (0.030)**

97

Obs 1594 1088 1462 1128 7392 2292 2758 R2 0.188 0.191 0.187 0.188 0.117 0.130 0.121 Robust standard errors clustered by region and 2-digit industry are in parentheses. Additionally control for foreign and government ownership, age, export status, 2-digit industry, and region fixed effects. Regressions for EU8, EU10 and All_ECA include controls for countries. *** p<0.01, ** p<0.05, * p<0.1.

In order to account for investment climate effects, investment climate variables are included one at the time, with ICvarjct representing the value of the average investment climate variable at industry j, region c, and time t:

. (3)

Since the main data source is cross-sectional, inclusion of the IC variables can cause endogeneity problem even when the chosen variables are not subjective measures that depend on firms‘ perceptions of the business environment. For instance, faster growing firms might be more likely to have access to external finance. To alleviate this problem, we use industry-region averages for the IC variables, capturing the total effect of the business environment in which firm operates. Cells with less than 5 observations are excluded from the analysis. Table 3-A-8 Firm growth and investment climate variables

IC Variable Used Smallt-3 Mediumt-3 Larget-3 Aget-3 IC Vart Const Obs R2 Labor regulations -0.131 -0.148 -0.107 -0.003 0.414 0.170 668 0.214 (0.040)*** (0.034)*** (0.060)* (0.001)** (0.195)** (0.085)* New fixed asset -0.13 -0.152 -0.113 -0.003 0.003 0.316 668 0.21 finance - external (0.040)*** (0.032)*** (0.060)* (0.001)** (0.001)** (0.090)*** Collateral -0.10 -0.156 -0.153 -0.005 -0.001 0.475 219 0.35 0.102 (0.092)* 0.127 0.004 0.001 (0.129)*** Purchase paid after delivery -0.14 -0.157 -0.114 -0.003 0.000 0.333 660 0.20 (0.039)*** (0.033)*** (0.060)* (0.001)** 0.002 (0.150)** Overdraft facility -0.137 -0.154 -0.11 -0.003 -0.144 0.348 656 0.21 (dummy) (0.040)*** (0.033)*** (0.062)* (0.001)** 0.122 (0.076)*** Loan (dummy) -0.145 -0.163 -0.125 -0.003 0.334 0.216 665 0.22 (0.039)*** (0.034)*** (0.064)* (0.001)** (0.079)*** (0.087)** Loss from power -0.218 -0.266 -0.22 0.000 0.003 0.272 223 0.33 outages (0.094)** (0.073)*** (0.081)*** 0.002 0.006 0.204 Informal payments -0.157 -0.167 -0.096 -0.003 0.029 0.281 551 0.23 (0.044)*** (0.041)*** (0.053)* (0.001)** 0.024 (0.084)*** Tax inspections -0.195 -0.199 -0.155 -0.004 0.000 0.525 324 0.30 (number) (0.084)** (0.057)*** (0.084)* (0.001)*** -0.001 (0.133)*** Time tax -0.116 -0.123 -0.086 -0.003 0.002 0.207 617 0.18 (0.037)*** (0.032)*** 0.063 (0.001)** -0.002 (0.104)* Customs clearance -0.171 -0.247 -0.229 -0.004 0.010 0.316 336 0.38 for exports (days) 0.13 (0.114)** 0.138 (0.002)** (0.004)** (0.112)*** Customs clearance -0.379 -0.489 -0.501 -0.005 0.002 0.414 208 0.52 for imports (days) (0.142)** (0.125)*** (0.122)*** (0.002)** 0.003 (0.139)*** Inspections -0.13 -0.159 -0.121 -0.003 -0.005 0.253 636 0.19 (number) (0.043)*** (0.037)*** (0.063)* (0.001)** 0.003 (0.065)*** Robust standard errors clustered by region and 2-digit industry are in parentheses. Additionally control for foreign and government ownership, export status, 2-digit industry, and region fixed effects. *** p<0.01, ** p<0.05, * p<0.1. Each row shows the IC variable included in the regression equation.

In order to analyze whether the IC variables have any differential affect on the growth rates of firms of different sizes, interaction terms of the IC variables with the size dummies are included in the equation, yielding:

. (4)

98

Table 3-A-9 Firm growth and investment climate variables with interactions (all data) IC Variable Small Medium Large IC IC*Small IC*Medium IC*Large Const Obs R2 Labor -0.167 -0.122 -0.098 0.230 0.457 -0.238 -0.076 0.181 668 0.220 regulations (0.070)** (0.052)** (0.089) (0.378) (0.466) (0.431) (0.555) (0.093)* External -0.031 -0.223 -0.149 0.004 -0.005 0.003 0.002 0.158 668 0.23 finance 0.078 (0.089)** (0.082)* (0.002)* 0.004 0.004 0.004 (0.074)** Collateral -0.086 -0.244 -0.044 -0.001 0.000 0.001 -0.001 0.086 219 0.35 0.196 0.168 0.208 0.001 0.001 0.001 0.001 0.219 Purchase -0.251 0.015 0.32 0.000 0.002 -0.002 -0.006 0.319 660 0.21 paid after delivery 0.196 0.232 0.234 0.003 0.003 0.003 (0.004)* (0.175)* Overdraft -0.112 -0.426 -0.32 -0.161 -0.038 0.382 0.286 0.352 656 0.21 facility 0.132 0.28 0.265 0.168 0.196 0.387 0.412 (0.103)*** (dummy) Line of 0.261 0.019 0.378 0.783 -0.629 -0.294 -0.79 -0.105 665 0.24 credit 0.185 0.165 0.236 (0.252)*** (0.320)* 0.262 (0.413)* 0.163 Loss from -0.263 -0.362 -0.255 -0.001 0.01 0.023 0.007 0.165 223 0.33 power out. (0.119)** (0.109)*** (0.117)** 0.009 0.015 0.019 0.019 0.11 Informal -0.054 -0.133 0.006 0.14 -0.171 -0.08 -0.196 0.226 551 0.28 payments 0.036 (0.054)** 0.051 (0.051)*** (0.059)*** 0.048 (0.052)*** (0.085)*** Tax -0.193 -0.2 -0.148 0.000 0.000 0.000 -0.002 0.524 324 0.31 inspections (0.088)** (0.058)*** (0.087)* 0.001 0.001 0.001 (0.001)** (0.135)*** Time tax -0.094 -0.109 0.338 0.003 -0.001 -0.001 -0.015 0.174 617 0.19 0.095 0.136 0.216 0.003 0.003 0.005 (0.007)** 0.132 Customs cl. -0.095 -0.237 -0.162 0.016 -0.013 -0.001 -0.012 0.296 336 0.39 - exports 0.171 0.151 0.181 0.012 0.012 0.011 0.013 (0.121)** Customs cl. -0.606 -0.796 -0.903 -0.025 0.025 0.032 0.045 0.885 208 0.55 - import (0.218)*** (0.198)*** (0.212)*** (0.013)* (0.014)* (0.015)** (0.017)** (0.194)*** Inspections -0.111 -0.162 -0.09 0.003 -0.008 0.001 -0.013 0.236 636 0.19 (number) (0.054)** (0.056)*** 0.095 0.016 0.015 0.017 0.024 (0.070)*** Robust standard errors clustered by region and 2-digit industry are in parentheses. Control for foreign and government ownership, export status, 2- digit industry, and region fixed effects. *** p<0.01, ** p<0.05 p<0.1 Each row shows the IC variable included in the regression equation.

Robustness Although averaging the firm level values of IC variables over industry-region alleviates the endogeneity problem, it might not remove it completely. We perform the same regression analysis specified in equation 3 restricting the sample to panel firms. In the panel regressions, we use the values of firm characteristics and IC variables from the 2005 survey. Here we cluster the standard errors at 2-digit industry level. Despite of the small sample size for the panel data and the inclusion of the controls for IC variables, for most of the regressions, medium size firms have the lowest growth rate. About the IC variables, time to clear customs for both exporting and importing is negatively related to firm growth.

Table 3-A-10 Effects of IC Variables on firm growth – panel firms Small Medium Large Age IC Const Obs RSqr New fixed asset financing -0.149 -0.233 -0.19 0.002 0.000 0.196 416 0.14 - external finance (0.084)* (0.067)*** (0.079)** (0.002) (0.001) (0.100)* Collateral -0.177 -0.235 -0.098 0.002 0.000 0.212 116 0.51 (0.067)** (0.043)*** (0.040)** (0.002) (0.000)** (0.101)* Purchase paid after delivery -0.148 -0.233 -0.19 0.002 -0.001 0.232 414 0.15 (0.084)* (0.067)*** (0.080)** (0.002) (0.000) (0.142) Overdraft facility (dummy) -0.143 -0.241 -0.207 0.002 0.048 0.127 416 0.15 (0.081)* (0.068)*** (0.082)** (0.002) (0.039) (0.110) Loan (dummy) -0.143 -0.241 -0.207 0.002 0.048 0.127 416 0.15 (0.081)* (0.068)*** (0.082)** (0.002) (0.039) (0.110) Loss from power outages -0.202 -0.343 -0.276 0.003 0.004 0.171 308 0.22 (0.080)** (0.071)*** (0.080)*** (0.003) (0.002) (0.143) Informal payments -0.15 -0.23 -0.208 0.002 -0.001 0.205 368 0.15 (0.089) (0.073)*** (0.089)** (0.002) (0.003) (0.145) Tax inspections -0.145 -0.23 -0.2 0.002 0.001 0.175 406 0.14 (0.087) (0.070)*** (0.083)** (0.002) (0.002) (0.079)** Time tax -0.144 -0.241 -0.202 0.003 0.001 0.198 403 0.14 (0.088) (0.077)*** (0.082)** (0.002) (0.001) (0.121) Customs clearance for export -0.111 -0.295 -0.188 0.003 -0.008 0.25 201 0.27 (0.049)** (0.139)* (0.111) (0.001)** (0.002)*** (0.085)** Customs clearance for import -0.065 -0.091 -0.027 0.001 -0.004 1.405 139 0.37

99

(0.095) (0.133) (0.097) (0.002) (0.002)** (0.128)*** Robust standard errors clustered by region and 2-digit industry are in parentheses. Control for foreign and government ownership, export status, 2-digit industry, and region fixed effects. *** p<0.01, ** p<0.05 p<0.1 Each row shows the IC variable included in the regression equation.

As an alternative to four size groups, we constructed three size groups measured as total fulltime employment. In this grouping, size classes are Small: ≤19, Medium: 20-99, Large: ≥100. This is the classification used in ES surveys for stratifying the population of firms in size. In Table 3-A-11, for brevity, we only present the results that include external finance and line of credit as IC variables. These IC variables were the ones that had significant coefficients in the main regression results and they are the only variables that persisted to be significant in this alternative size grouping. In these regressions, the group of firms with less than or equal to 19 employees is omitted. The non- monotonic decrease in growth rates as size increases is observable in this size grouping as well.

Table 3-A-11 Employment Growth with Three Size Groups (1) (2) Medium (20-99) -0.097 -0.103 (0.032)*** (0.031)*** Large (≥100) -0.085 -0.088 (0.043)* (0.044)** Age -0.003 -0.003 (0.001)** (0.001)** External Fin 0.004 (0.001)*** Loan (dummy) 0.311 (0.073)*** Constant 0.197 0.09 (0.084)** -0.081 Observations 668 665 R-squared 0.188 0.185 Robust standard errors clustered by region and 2-digit industry are in parentheses. Additionally control for foreign and government ownership, export status, 2-digit industry, and region fixed effects. *** p<0.01, ** p<0.05, * p<0.1.

To provide further support to the significance of access to finance for firm growth, we used principal component method. First we determined the principals of the IC variables that measure access to finance. The first regression uses six of the seven IC variables that are related to access to finance. The excluded variable is the use of collateral. Collateral variable is included in the second estimation. Since very few firms answered the question for collateral usage, the sample size in the second regression is much smaller. The principal component analysis allows explaining the overall effects of variables representing access to finance. In the regressions, only the first component is included, since higher ranked components were not statistically significant. In both regressions, access to finance contributes positively to growth.

Table 3-A-12: Access to finance and growth: principal component method (1) (2) Small -0.163 -0.177 (0.075)** (0.145) Medium -0.135 -0.249 (0.055)** (0.103)** Large -0.159 -0.253 (0.097) (0.159) Age -0.004 -0.004 (0.002)** (0.005) Principal component 0.06 without collateral (0.020)*** Principal component 0.067 with collateral (0.029)** Constant 0.304 0.002 (0.115)** (0.191) Obs 419 170 R2 0.248 0.435

100

Robust standard errors clustered by region and 2-digit industry are in parentheses. Additionally control for foreign and government ownership, export status, 2-digit industry, and region fixed effects. *** p<0.01, ** p<0.05, * p<0.1.

We also used several alternative definitions of growth rates. One is log of employment growth and another is an extensively used measure of job creation and destruction in labor economics. It is found as dividing the difference in employment levels over the past three years by the simple average of employment levels over the same period. This measure is bounded by values of -2 and 2 and provides more robust estimates to outliers. In either case the relation between size and employment growth was not sensitive to these alternative measures of growth.

101

Chapter 4. STIMULATING KNOWLEDGE FLOWS

4.1. Since the 1980s, the liberalization of the Turkish economy stimulated the creation of regional production networks centered on internationally competitive manufacturers. The 1980s, characterized by a transition from protectionist attitudes towards increased reliance on market forces, are considered a turning point for the international integration of the Turkish economy.73 The completion of trade liberalization in 1984 – requiring the dismantling of foreign exchange controls, the abolition of quotas on imports, and a downward revision of tariffs – was accompanied by the enactment of pro-active export promotion policies and by the depreciation of exchange rates. This presented the Turkish economy with new opportunities related to a generalized rise in international capital mobility and trade in intermediate goods. As a result, Turkey‘s share of world manufacturing exports increased six-fold from 0.15 percent in 1980 to 0.90 percent in 2008.74 Integration into global trade and investment flows has been accompanied by a significant spatial transformation of the Turkish economy characterized by the emergence of a number of more recent industrial agglomerations far from the earlier manufacturing regions, the so called ―Anatolian Tigers.‖ Clusters of industries, with specialization in both traditional and more technologically advanced sectors, have formed in various parts of the country and have become the core of manufacturing and exporting activities (Figure 4-1).

4.2. In response to these developments, the government has activated a number of instruments to foster the ability of SMEs to participate in global markets. Beginning in the mid-1990s the attention of policy has turned to the triple challenge of more growth, greater competitiveness, and more jobs through the development of SMEs.75 In this spirit, a number of instruments have been activated with the intent of stimulating the participation of SMEs in regional production networks.76 The rationale of several such interventions has been to remove obstacles to the competitiveness of SMEs related to the business environment. For instance, SMEs in the manufacturing sector have been encouraged to locate in appropriately planned ―small industrial estates" (KSS) and ―organized industrial zones‖ (OSB) that ease investment climate constraints by providing a number of advantages in terms of infrastructure services and regulation of business activity. In a similar vein, the government has encouraged the development of regional innovation strategies aimed at sustaining the competitive advantage of regions. In the future, primary responsibility for the preparation of regional innovation strategies is expected to lie with the newly established development agencies discussed in Chapter 2.

4.3. Global conditions in the aftermath of the crisis highlight the need to further improve the international competitiveness of smaller local firms in order to achieve sustainable and broad-based growth. Chapter 2 emphasized how larger Turkish firms are able to benefit from the more positive aspects of the investment climate. Foreign-owned firms and successful exporters appear at the forefront of this pattern, with TFP and investment climate variables related to innovation and quality appearing as crucial determinants of the probability of exporting and, to a more limited extent, to receive FDI (Figure 2-5). This is in line with the empirical evidence across developed and developing countries indicating that larger and more productive firms more easily manage to penetrate export markets, and that foreign owned firms display higher productivity and a higher knowledge content of production relative to domestic counterparts. 77 The 2008-2009 economic crisis – with its widespread decline in international trade and

73 Eraydın and Armatlı (2005). 74 Export data are compiled from the World Bank WDI database. 75 Kuruüzüm, 1998. 76 See Doloreux and Parto, 2005; Bathelt et al. (2003); Malecki and Oinas (2000); and Tallman et al. (2004). 77 See Wagner (2007) for an overview of the literature on the link between productivity and exporting and Markusen (2005) for a discussion of the peculiarities of foreign-owned vis-à-vis local firms.

102

capital flows likely to continue in the near future – reinforces the relevance of productive efficiency and of the ability to innovate as crucial conditions for the international competitiveness of the Turkish manufacturing sector. Whereas these challenges are present for successful Turkish exporters it is even more urgent, as already emphasized in the 2007 ICA, to ensure that smaller firms that are not directly integrated into global markets are able to increase the knowledge content of their production in order to participate in global value chains. Diffusion of the sources of growth beyond firms that are currently sufficiently competitive to be direct exporters and outside more successful manufacturing poles will increase the resilience of the Turkish economy to future shocks in global demand and guarantee that the productive base of Turkish manufacturing is more evenly distributed across Turkish regions.

Figure 4-1: Industrial clusters in Turkey

Structural characteristics of selected clusters Factors/ Conditions Denizli Bursa Ankara The type of the Industrial district Innovative manufacturing High-tech industrial cluster manufacturing cluster cluster Area of Textiles, especially towels Textiles for home furnishing Machinery, electronics, specialization and bathrobes defense industry and software The main character Traditional Traditional/Modern Modern/High-tech of the cluster Small artisanal, and highly Small Artisanal, and highly High-tech firms of different specialized family owned specialized firms as well as size firms located in close large multinational companies proximity co-operating with these small enterprises Main observed Co-operation in production Collective competition in Weak collaborative benefit and specialized fields environment Market relations marketing for international with state institutions markets Technical Complementarities Specialization increasing Adaptation of new dynamic Social shares of export in engineering technologies for national capital Collaborative action, trust industries, Adaptation and market and reciprocity Strong product development for social networks international markets Access to qualified labor Source: Öz (2004) and Eraydın and Armatlı (2005)

103

4.4. Analysis of the case of Turkish production networks indicates that the absorptive capacity – i.e. the ability to adopt and use knowledge – of local suppliers, especially SMEs, is key for successful participation in global markets. Global buyers (domestic or foreign-owned manufacturers, retailers or brand-name companies) have come to play an increasingly important role in the organization of global production and distribution systems. In this light, this Chapter examines knowledge flows in Turkish manufacturing by focusing on the ability of local manufacturers to participate in global markets via their participation in production networks. A Global Value Chains (GVC) perspective is used to analyze the mechanisms of knowledge transfer at firm level, and the intensity with which buyers – usually larger firms with direct links to domestic or international markets –are engaged in transferring knowledge and technology to smaller suppliers. Analysis shows that (i) suppliers endowed with some technical skills and capacity “to handle the technology” are more likely to enter knowledge intensive value chain relationships. Furthermore, (ii) more capable suppliers are more likely to engage in value chain relationships characterized by transfer of design, quality standards and technology from the buyer to the supplier. A more efficient regulatory environment, specifically operating licenses and permits, and higher shares of investment financed from bank loans are also positively associated with knowledge intensive value chain arrangements. The latter result confirms the central role of access to external finance already highlighted in Chapter 3.

4.1 Production Networks and Knowledge Flows 4.5. Production networks can play a crucial role in diffusing technological knowledge and enhancing learning and innovation, with the intensity of knowledge transfer depending on the complexity of the task, the extent to which information can be codified, and the capabilities of potential suppliers. Recent developments in the GVC literature have shown the importance of value chain governance modes, especially those between suppliers and buyers, in enhancing technology diffusion, learning and innovation (Box 4-1). 78 Value chain governance modes refer to linkages among buyers, sellers and service providers that influence the range of activities required to bring a product or service from inception to its end use. The key parameters shaping the nature of the underlying arrangements among firms are what, how, when, and how much is to be produced. Alternative types of governance modes can thus be defined based on the ―intensity‖ of knowledge transfer between buyers and suppliers participating in a value chain and are determined by: (i) the complexity of information and knowledge transfer required to sustain a particular transaction, especially with respect to product and process specifications; (ii) the extent to which this information and knowledge can be codified and, therefore, transmitted efficiently and without transaction-specific investment between the parties involved in the transaction79; and (iii) the capabilities of actual and potential suppliers in relation to the requirements of the transaction.80 Governance modes are thus considered a key vehicle for knowledge transfer from global buyers towards local firms, as well as key determinants of entrepreneurial growth for local suppliers.81

78 Governance of the value chain can be defined as ―authority and power relationships that determine how financial, material, and human resources are allocated and flow within a chain‖ (Gereffi, 1994). Also see Altenburg, 2006; Gereffi, 1999; Giuliani et al., 2005; Kaplinsky, 2000; Humphrey & Schmitz, 2002a,b; Pietrobelli and Rabellotti, 2007. 79 Typically, the knowledge base of traditional industries is highly dependent upon local and tacit forms of knowledge, whereas the knowledge base of firms in high-technology sectors is more codified allowing firms to establish networks to access distant knowledge sources (Vale and Calderia 2006). 80 Sturgeon, 2001; Gereffi and Kaplinsky, 2001; Saliola and Zanfei, 2009. 81 Kappel and Brach 2009.

104

Box 4-1: Global value chains Recent developments in the studies on Global Value Chains (GVC) have shown the importance of value chain governance modes, especially those between suppliers and buyers, in enhancing technological knowledge diffusion, learning and innovation. Governance modes are considered to be a key vehicle for knowledge transfer from global buyers towards local firms and to enhance entrepreneurial growth of local suppliers. From this perspective, the GVC literature has used the tool of the governance to examine how firms active in global markets organize the transfer and coordination of complex and strategic information along value chains and their implications for knowledge transfer and development.

There is a long tradition of studies investigating the channels through which knowledge is transferred to local companies in global markets. In particular, a plethora of contributions have focused on Multinational Companies (MNCs) generating knowledge (and pecuniary) spillovers to the host economy through linkages with local firms (Hirschman (1958), Lall (1978), Rodriguez-Clare (1996), Markusen and Venables (1999), Görg and Strobl (2001) and Jarvocik-Smarzynska (2004).). The theoretical debate is dominated by arguments predicting positive effects, including the transfer of new technologies and management skills and the increase in the level of competition. However, there is little empirical evidence suggesting that domestic firms benefit from multinational presence, and the controversy is far from resolved.

The GVC approach can provide complementary insights to these studies. The GVC perspective is useful for various reasons. First, because the focus moves from manufacturing only to the other activities involved in the value chains of goods and services, including distribution and marketing. Second, GVC emphasizes the nature of the relationships among the various actors involved in the chain. Moving beyond firm-specific analysis and concentrating on inter-firm linkage allows to better capture dynamic flows of economic and organizational activities between producers within different sectors even on a global scale. Finally, GVC studies identify distinct types of relationships characterized by a different level of involvement of suppliers in knowledge intensive activities. This helps capture the essential features of knowledge dissemination and absorption, which are likely to enhance the development of local suppliers.

4.6. Turkey offers an interesting case study to analyze knowledge flows within production networks. First, the increasing presence of firms, both domestic and multinational, linked to different global value chains has provided learning opportunities for local Turkish enterprises. Such firms act as focal or lead firms for the local network, and perform the key functions of generating new knowledge and technologies, spinning out innovative companies, attracting researchers, performing investments in R&D and related research facilities, enhancing R&D activities in networked firms, and stimulating demand for new knowledge.82 Second, networks of firms, with the knowledge flows that characterize them, have proved to be a valid alternative to vertical integration in a number of sectors. A case in point is that of Turkish machinery producers, which, despite weak vertical linkages in globally integrated supplier networks, which are considered the primary providers of innovation, have been able to integrate alternative global value chain arrangements that, short of vertical integration, offer SMEs an opportunity to expand their ability to innovate. 83 Finally, international retailers play an important role in the development of Turkish industrial districts, by transferring tacit expertise in related fields, such as marketing, to firms participating in a local production network.84

4.7. Turkish subcontracting relationships can be portrayed to have three main characteristics.85 First, some of the networked relationships are long-term and duration is connected to product-life cycles. Each time a new product is designed and manufactured, the large firm makes a call for the best

82 Erdil 2009; Autio et al., 2004; Bergman and Feser, 1999 and Ceglie and Dini, 1999. 83 See Kozan et.al. 2006, Ulusoy 2003, Erdil and Çetin 2008. 84 See Öz 2004, Nadvi 1995. 85 See Ulusoy 2003.

105

offer from suppliers. At that stage, suppliers are put into competition. However, the firm generally continues subcontracting relationships with the same suppliers from a product to another, in order to avoid costly and time consuming renegotiations. Such duration of relationships allows deriving some of the benefits of vertical integration. Second, some of the Turkish networked relationships are institutionalized and hierarchical. Such hierarchy of subcontractors is defined according to the type of product bought by the large firm. In this case, the subcontractors are autonomously chosen on the basis of quality. Product design is often carried out jointly by the supplier and the main firm. In the latter case, the supplier only executes orders from the firms according to its production definitions, and is highly dependent on the large firm. Third, the Turkish networked relationships are contractual and characterized by specific procedures. Usually a supplier is contracted when the new product is in the development phase (with no specification of quantities to be delivered, nor of prices, etc.) providing flexibility and adaptation capability to possible changes in the specification of products.

4.8. Analysis of survey data confirms that buyer-supplier relationships in Turkey are quite long- term with an average duration of 12 years. The NIS survey (Box 4-2) allows mapping the nature of buyer supplier relationships in Turkey. By looking at the sample distribution by firms‘ age, on average, young companies (six years or less in business) have worked with their largest buyer for 5 years (73 percent of their business life) and firms with more than 40 years of activity for 22 years (55 percent of their business life time). This result is in line with the argument that Turkish buyers, once they have chosen their suppliers, tend to continue working with them.86 Relationships tend, therefore, to be based on trust between the parties, which avoids costly and time-consuming renegotiations and reduces the time needed to learn the specifications of the product.

Box 4-2: The “Networks and Innovation” Survey (NIS) This Chapter is based on the ―Networks and Innovation‖ survey (NIS) implemented by the World Bank in the summer of 2009 which represents a complementary study of the Enterprise Survey (ES) implemented in 2008 (See Box 2-1). The NIS survey was administered to 514 establishments in the manufacturing sector and it represents a sub-sample of firms drawn from the set of firms previously interviewed in the ES Turkey 2008. The ES Turkey 2008 data has been used for comparisons in the analysis. The database contains also retrospective information for up to three periods before the year of reference for 231 establishments.

Sampled firms were stratified by size, location, and sector to ensure that most major types of firms are covered. 15 percent of firms in the sample are classified as micro (less than 10 employees), 41 percent small (11-50 employees), 30 percent medium (51-249) and 13 percent large (250 employees or more). Firms surveyed are active in the following industries: Food processing (18 percent), Textiles (21 percent), Garments (10 percent), Chemicals (12 percent), Non-metallic mineral products (11 percent), and a residual category labeled ―other manufacturing‖ and comprising rubber and plastics, basic metals, fabricated metal products, machinery and equipment, electrical machinery and apparatus, motor vehicles, trailers and semi-trailers, furniture (28 percent). The regions covered are five: Marmara, Aegean, Central Anatolia, South and Black-Sea Eastern. Marmara includes Bursa, Istanbul and Koaceli; Aegean includes Denizli, Izmir and Manisa; Central Anatolia includes Ankara, Eskisehir, Kayseri and Konya; South region includes Adana, Gaziantep, and Kahramanmaras city; and Black Sea-Eastern includes Erzurum, Malatya, Samsun, and Trabzon city. Only firms with five employees or more were included in the sample.

4.9. The degree of buyer involvement is shaped by the sector in which firms operate, with low- tech sectors characterized by a more disintegrated supply chain showing greater buyer involvement and a higher degree of knowledge transfer. On average, 44 percent of firms‘ sales are made according to largest buyer‘s unique specification. This varies substantially across firms in different sectors. For example, firms in textile and garments industries make about 55 and 51 percent of their sales,

86 See Kozan 2006.

106

respectively, according to largest buyer unique specification compared to 25 percent of firms in food and chemicals industries. In 61 percent of the firms in the sample the largest buyer gets involved in defining design & quality. Labor-intensive industries – such as food, textiles, and garments – show greater buyer involvement in design and quality standards. Over time, based on the development of relationships based on mutual trust, suppliers and buyers are expected to share sensitive strategic data. This sharing process is associated with a number of benefits, notably encouraging suppliers to invest in buyer‘s future needs, generating entrepreneurial growth, and creating and expanding competitive advantages and synergy effects.87 33 percent of firms are characterized by buyer involvement in R&D activities, with firms in food and textiles more likely to receive inputs on process or product R&D activities from their largest buyer. On average 17 percent of enterprises in the sample received technology inputs from buyers. The Marmara region stands out in as being the region with the smallest share of firms receiving buyer support for technology dissemination. Furthermore, buyer involvement in dissemination and diffusion of new technologies affects sectors unequally, with firms in garments the most likely to receive technology and firms in chemicals the least likely. This result is obviously linked to the nature of the technology and of the value chain arrangements typical of each sector, with a sector like garments having a highly disintegrated value chain and a definition of technology that may include the acquisition of low-tech equipment from a lead buyer.

4.10. Buyer influence on sales tends to become more prominent as firms become larger, whereas inputs on R&D are more common for SMEs. Across firm size, large firms exhibit the largest share of sales made to meet buyer‘s unique specification (51 percent, versus 54 percent for SMEs. On the other hand, medium firms are more likely to receive inputs from buyers on design and quality standards (77 percent) compared to small and large firms (61 and 70 percent respectively). Inputs on R&D from the largest buyer are more common with small and medium businesses (38 and 37 percent respectively), compared to 20 percent for micro firms and 30 percent for large enterprises. Finally, the percentage of large firms with buyer involvement in technology dissemination is significantly higher than for medium and small firms (28, 14 and 20 percent respectively). By and large, the survey findings show that micro businesses are significantly less involved in a close relationship with the main buyer, relative to the other size groups.

4.11. While all value Figure 4-2: Value chain governance modes chain relationships imply some transmission of Dissemination new information between the Process and Product technologies into firm’ GOV0 Design and Quality R&D Activities parties, the extent to facilities which knowledge is

actually created, Dissemination new transferred and adopted Process and Product Design and Quality technologies into firm’ along value chains varies GOV1 R&D Activities facilities substantially. From this perspective, one can characterize at least four Dissemination new Process and Product Design and Quality technologies into firm’ different types of value GOV2 R&D Activities facilities chain agreements, based on the intensity of knowledge transfer (Figure Dissemination new Process and Product R&D GOV3 Design and Quality technologies into firm’ 4-2). The first, which we Activities facilities shall name GOV0, is characterized by no

87 Kappel and Brach 2009.

107

knowledge transferred from buyer to supplier. The second, GOV1, is characterized by the transfer of a product specification from the purchasing firm to the supplier, and implies that the latter will carry out a number of autonomous tasks which will eventually lead to the provision of the required good (or service), following the directions given by the buyer. Under this circumstance, knowledge transferred is kept to a minimum, will be mostly codified, and it will flow in one direction only (from the buyer to the supplier). A third mode of governance, called GOV2, takes place when the buyer provides product design and quality standards to be followed in the production process. Knowledge transfer is uni-directional here too, but it may well be partly tacit and it is more detailed than in the former case. Under this circumstance, the supplier is given strict and mandatory instructions, and is thus granted very little degrees of freedom when executing his/her tasks. A fourth typology of value chain relationship, GOV3, can be observed when the buyer disseminates specialized competencies, and involves the supplier in R&D and knowledge development. One should notice it is not only a matter of ―quantity‖ of knowledge flowing between firms: a substantial part of the transfer takes place through the mobility of personnel. The knowledge involved is thus largely tacit in nature, and flows in both directions, although not necessarily in a balanced way. These typologies of value chain relationships thus correspond to different modes of organizing knowledge transfer and diffusion.

4.12. The intensity of knowledge transfer varies substantially across regions. Looking at regional features depicted in Figure 4-3, the data reveal that Marmara, Figure 4-3: Governance modes features, across regions Aegean and Central Anatolia 40% 38% 35% are mainly associated with less 33% 35% 30% knowledge intensive buyer- 29%29% 29% 30% 27% suppliers relationships (GOV0 25% 25% 24% 23% 25% 23% and GOV1). This is an 21% 20% 21% 20% 20% indication of the fact that firms 20% 14% in these more developed 15% 13% regions are, on average, more 10% capable of harnessing 5% knowledge without the need 0% for buyer involvement and also Marmara Aegean Central Anatolia South Black-Sea of the prevalence of certain Eastern sectors. Buyer-supplier Gov0 Gov1 Gov2 Gov3 relationships featured by Source: Turkey NIS 2009 transfer of design and quality standards, and a direct involvement in R&D activities but not in technology dissemination are more common among firms located in the Black-Sea Eastern and South regions.

108

4.13. Governance modes are also different depending on the organization of production typical of each industrial sector. Metallic products and other manufacturing industries are highly associated with little or no knowledge transfer (GOV0). This is an indication that these sectors are, by nature, characterized by more standardized products which require little or no knowledge transfer from buyers and that the firms in these sectors that are competitive internationally rely merely on their productivity – as testified by the vast literature on the self-selection of more efficient firms into export markets – the quality of their products achieved via successful R&D expenditures, and pricing advantages. Firms in the textile industry tend to be associated mainly with transfer of design and quality standards and a direct involvement in R&D activities by the buyer (GOV1 and GOV2). This is also linked to the specific characteristics of these products, where Turkish manufacturers are often sub-contractors for main buyers, domestic or foreign, that possess and maintain within their boundaries the functions associated with higher value added, such as design, quality control and marketing. However, these main buyers have the need to transfer some of their R&D and quality control competences backward to their suppliers in order to ensure a certain standard for the product being subcontracted. In food processing 30 percent of firms in the sample receive transfer of design and Figure 4-4: Governance modes features, across industries quality standards, and 45% 42% 41% 45 percent of firms 38% 39% 40% receive also buyer‘s 35% 35% 34% 35% 31% 32% involvement in R&D 30%30% 29% 30% 25% and technology 25% 24% 25% dissemination activities. 19% 19% 20% 17% 17% Firms in the garments 15% 15% industry show the 11% 10% 7% highest share of firms 3% 2% characterized by buyer 5% transfer of precise 0% requirements in terms Manufacture of Textile Garments Chemicals Non metallic Other Food mineral Manufacturing of design and quality products standards, and Gov0 Gov1 Gov2 Gov3 involvement in Source: Turkey NIS 2009 technology dissemination and R&D (GOV3). In both of these industries, the buyer is a large retailer or distributor closer to the final consumer, that needs to ensure maximum consistency of its products with expected standards and therefore has an interest in maximum hands-on involvement with production in supplier firms. 4.14. Medium and large firms stand out Figure 4-5: Governance modes features, by firm size for the relatively higher share of suppliers 60 involved in knowledge intensive 50 relationships. This is featured by transfer of 50 D&Q inputs for both medium and large 40 37 36 enterprises (GOV1), whereas input on R&D 31 investments (GOV2) is more prominent for 27 27 29 28 30 24 medium firms. Additionally, technology 22 20 21 20 dissemination from the main buyer into 14 14 13 9 firms‘ facilities (GOV3) seems to be more 10 frequent in large enterprises. This is probably 0 an indication of firms active in labor- micro small medium large intensive sectors, such as textile, garments or food, which receive a large amount of Gov0 Gov1 Gov2 Gov3 Source: Turkey NIS 2009 knowledge from their buyers. Again, the results show that knowledge transfer is

109

relatively less frequent in micro enterprises, whereas a significant number of small firms (27 percent) receive buyer inputs in D&Q and R&D investments.

4.15. No major differences are observed across firms in terms of share of direct exports based on governance features. Figure 4-6 shows that the more knowledge intensive the buyer-supplier relationship is, the higher the share of indirect exports (sold domestically to third Figure 4-6: Trade and governance modes party). Firms characterized by GOV3 (transfer of design, quality standards, R&D 80 75 67 66 64 and technology dissemination) sell 17 percent to distributors compared to 8 percent 60 of firms characterized by no transfer of knowledge (GOV0). By restricting the 40 % analysis to those firms that were interviewed 17 19 18 17 19 20 15 15 in 2005 and in 2009 (panel firms), we notice 8 that firms characterized by GOV1 and GOV3

0 in 2009 have more than doubled the share of indirect exports. Also, we observe that the GOV0 GOV1 GOV2 GOV3 share of direct exports has increased for all Domestic sales Indirect exports Direct exports the categories since 2005 except for firms Source: Turkey NIS 2009 featured by GOV1.

4.16. Interaction with MNCs is associated with more intensive knowledge transfer. When we examine the correlation between buyer‘s characteristics and governance features for domestic sales flows only, we observe that firms characterized by GOV3 are more likely to deal with MNCs operating in Turkey, and parent companies or subsidiaries: on average 24 percent of their sales is made for these types of buyers. Firms characterized by GOV2 are less likely to sell their products to government and state owned firms. Looking at how domestic sales destination for firms interviewed in 2005 and in 2009 have changed in the past three years, we observe an overall increase of sales made for MNC operating in Turkey. Firms characterized by transfer of design and quality standards and R&D inputs (GOV2) sell significantly less to Government and State owned firms company and more to MNCs operating in Turkey compared to 2005, while firms characterized by GOV3) have increased their sales to MNCs operating in Turkey and to parent companies or affiliated subsidiaries.

4.2 Increasing the Absorptive Capacity of SMEs 4.17. The local environment can influence the incentives and ability of firms to adopt and employ new knowledge. The extent to which exposure to advanced technology translates into broad diffusion across the economy and technological upgrading depends on the absorptive capacity of an economy‘s firms. For example, the business climate may be too poor to stimulate investment, or the technological literacy of the local labor force may be too low to successfully adapt the machinery to local conditions. As a result, local firms may not realize the potential productivity improvements. The absorptive capacity of a firm, in turn, is determined by two main groups of factors, the first being internal to the firm, in particular (i) the technological literacy of the firm‘s labor force and management; and the other being external, notably (ii) the investment climate or broader environment in which firms operate that affects their incentives to invest and innovate. Furthermore, the local business and institutional environment overlaps with country-wide features to determine the incentives that firms have to adopt and use innovative modes of production and organization. The availability of a research base at the local level can, for example, encourage innovative behavior on the part of enterprises, if contacts exist between firms and local research organizations. Access to a skilled workforce is also highly dependent on the quality of the local higher education and vocational training. The ease of bank finance, in turn, is conditional on the

110

development of the local banking sector, as well as on personal contacts that may facilitate relational lending practices. Finally, the regulatory environment is also largely local since a large number of operating licenses are awarded by municipalities. As a result, the effects in terms of knowledge transfer of linkages between globally connected firms and local suppliers may vary widely depending on local conditions.

4.18. Econometric results indicate that the ability of firms to adopt knowledge transferred in production networks is strongly associated with features of the investment climate that affect the firms‟ absorptive capacity. Combination of the 2009 NIS survey with the main 2008 enterprise survey allows focusing on the effect that investment climate conditions may have on the intensity of knowledge transfer within production networks. To gauge the relative importance of various investment climate features, the analysis uses only objective, as opposed to perception based, measures. Various investment climate features concur to affect a firm‘s absorptive capacity. Some are directly related to the internal capability of firms, in terms of managerial experience, skill levels of the workforce, provision of training, prevalence of ICT, possession of an internationally recognized quality certification, use of foreign licensed technology and R&D expenditures. Other factors that affect a firm‘s ability to adopt and use knowledge are related to the external environment, namely to the regulatory framework governing the issuance of licenses and permits or inspections; the availability of bank finance or receipt of direct government subsidies. Regression analysis also controls for firm size, region and industry and industry conditions, implying that the results hold regardless of these characteristics.88 The summary results in Table 4-1 indicate whether each explanatory variable has a positive, negative or insignificant association with the likelihood of an outcome implying some degree of knowledge transfer (GOV1, GOV2 or GOV3), relative to the baseline outcome of no knowledge transfer (GOV0).

Table 4-1: Summary of investment climate effects on knowledge flows GOV0 GOV1 GOV2 GOV3

R&D (dummy) - - + + Foreign technology (dummy) - + NS + New product (dummy) NS NS - + INNOVATION AND SKILLS Computer - + NS + Quality certification (dummy) NS NS NS + Manager experience (years) NS NS NS + Staff - skilled workers - NS + + Training (dummy) NS - NS + Time tax - + + - Operating license (days) NS NS NS - REGULATORY ENVIRONMENT Permit (days) + - NS - Business inspection (number) + NS NS NS Loan (dummy) NS NS + NS FINANCE New fixed asset finance - bank NS + - + Collateral NS - + + Domestic sales - NS + NS TRADE Inputs imported + - NS - OTHER Subsidies (dummies) + NS NS NS Notes: See Table 4-A-1 in the Annex. NS “Not Significant”. Regressions include region, industry, size, exporter and foreign ownership dummies.

4.19. Suppliers endowed with some technical skills and capacity “to handle the technology” and able to diversify their product offer are more likely to enter knowledge intensive value chain relationships, because the payoff they can expect to obtain from access to external sources of

88 See Annex for details on the econometric methodology and results.

111

knowledge will be higher.89 It is apparent from Table 4-1 that more capable suppliers are more likely to engage in value chain relationships characterized by transfer of design, quality standards and technology dissemination (GOV2 and GOV3). Namely, there is a positive correlation between R&D investments and GOV2 and GOV3, and a negative correlation between R&D investments and less knowledge intensive relationship. In the same vein, the usage of foreign technology has a positive correlation with GOV3 and a negative correlation with GOV0. The share of workforce using computers is positively correlated to GOV1 and GOV3 and negatively correlated to GOV0. The share of skilled workers turns out to be significant too, with a positive correlation to GOV3. These results, along with the ones obtained for innovation variables, seem to confirm the hypothesis, following Cohen and Levinthal (1989), that domestic firms will need some competencies not only to handle the knowledge they are already endowed with, but also as a means to gain access to external sources of knowledge. From this perspective, the greater the technical competencies of local suppliers the more value chain agreements are likely to be an effective vehicle of knowledge transfer and adoption. Finally, firms that have more diversified products are also more likely to enter knowledge intensive value chain agreements (positive correlation between ―new products‖ and GOV3).

4.20. A cumbersome regulatory environment negatively affects the likelihood of knowledge intensive buyer-supplier relations. The time spent by senior management dealing with regulations is negatively associated with GOV3. An inefficient licensing environment, specifically operating licenses and permits, does not encourage value chain relationships characterized by high transfer of knowledge. The higher the number of days to obtain operating licenses, the less the likelihood of observing types of governance featured by buyers‘ involvement in technology dissemination and R&D activities (negative correlation with GOV3). In the same vein, the higher the number of days to obtain permits, the less the likelihood of observing buyer-supplier relationship featured by transfer of knowledge. Finally, the number of business inspections is positively correlated to GOV0.

4.21. Easier access to bank finance is connected with more intensive knowledge transfer, while government subsidies appear directed to firms engaged in less knowledge intensive relationship. Providing further support to the importance of bank finance (Chapter 3), higher shares of investment financed from bank are positively associated with knowledge intensive value chain arrangements (GOV3). Evidently, access to bank finance goes hand in hand with high collateral requirements, since the value of the collateral required is also positively associated with more intensive modes of knowledge transfer (GOV2 and GOV3). Subsidies from the national, regional or local governments instead show a positive correlation with GOV0.

4.22. Suppliers directly related to international markets tend to be characterized by limited transfer of knowledge. We find in fact a positive correlation between GOV0 and the shares of inputs imported directly, and a negative correlation between GOV0 and the share of domestic sales. Turkish exporters and inputs importers are likely to characterize their cooperation models by arm‘s-length market coordination or by participation in vertical relationships in which they are allowed a high level of autonomy. Domestic sales are instead positively correlated with GOV2 (buyer‘s involvement in D&Q and R&D). From this perspective, as highlighted above, it should be considered the presence of firms linked to different global value chains (e.g. Turkish companies with extended networks or MNCs located in Turkey) which provide learning opportunities to local enterprises that are in economic relation with them. These results suggest that the local relationships among focal firms and other networked local firms (as in the supplier-buyer network) facilitate inter-firm learning and knowledge exchanges.

4.23. Governments globally have traditionally played a role in facilitating firms‟ access to knowledge by focusing public interventions in areas where private investment is either sub-optimal

89 Cohen and Levinthal 1989.

112

or non-existent. Firms can acquire technological capability through a number of learning mechanisms. The most important are private learning mechanisms. These include: (i) internal efforts by firms, such as in-house training of workers, R&D activities, and hiring of staff with knowledge of more advanced technologies; and (ii) external learning mechanisms, including contacts with foreign buyers and suppliers of equipment, interaction with other firms in the industry, training courses and hiring of consultants. When firms cannot manage the technology transfer process themselves, they can turn to collective mechanisms, if available, such as technical support services offered by government, business associations, and NGOs. These collective learning mechanisms can be broad based (e.g., private sector training courses and consulting services) or they can be highly-focused (e.g., direct technical support to individual firms).

4.24. In Turkey several governmental and non-governmental organizations provide support to firms, especially SMEs, with the objective of increasing their operational capabilities and absorptive capacity. A list of these institutions and of their programs cannot be exhaustive, since several government agencies and other organizations have established such initiatives. Here, it is worth mentioning the initiatives undertaken by the Union of Chambers and Commodity Exchanges of Turkey (TOBB), and by two publicly funded programs, under the Small and Medium Enterprises Development Organization (KOSGEB), and the National Productivity Center (MPM).

4.25. TOBB plays a role in supporting SMEs, by making information on available financial and non-financial resources more accessible and through the network of EU business centers (ABİGEM). TOBB has coordinated the establishment by the Prime Ministry of an information site for SMEs90 in order to provide a foundation to more efficient promotion of publicly funded support programs for SMEs and to facilitate access to information. Another initiative led by TOBB is the establishment of European Union Business centers (ABİGEM) in 15 provinces around Turkey with funds provided by the EU. These institutions provide professional training, consultancy and information services in the regions since 2002. Their aim is to help SMEs improve their competitiveness within the national and international markets and increase SMEs‘ contribution to both local and national economies. The total EU financial investment in the network is exceeding 50 million Euros, making it one of the major EU projects in the world.

4.26. KOSGEB is a government institution aimed at enhancing the efficiency and competitive ability of SMEs, including by supporting their technological capacity and innovation potential. KOSGEB was established in 1990 in affiliation with the Ministry of Industry and Trade. It provides services and programs to SMEs in areas linked to management and business development, as well as technology upgrading. In April 2009, it included SMEs in the service sector in their scope of firms to provide support. Support is mainly offered in the form of loans and grants to purchase goods and services from external providers, but KOSGEB also organizes general training programs. With an operational budget in the order of USD 60 million, it plays an important role in SME support. Its operational budget is larger than most comparable non-lending support, including MPM in Turkey, MEP in the US and MAS in the UK. However, KOSGEB operates a relatively small program when compared to the Small Business Administration (SBA) in the US, which mostly provides loans to SMEs. While the SBA provided financial assistance equivalent to more than 3 percent of US GDP in 2008, KOSGEB provided loans of only 0.02 percent of Turkish GDP. Approximately 70,000 companies have registered in KOSGEB‘s SME database, out of a total of about 250,000 in Turkey. Support is provided following the completion of Strategic Roadmaps by candidate enterprises, and reached over 10,000 SMEs in 2008. KOSGEB has established 22 different targeted programs delivered through 35 Enterprise Development Centers across

90www.kobi.org.tr. The information is provided under 11 headings: ―Enterprise Establishment, Transfer and Dissolution,‖ ―Construction and Opening Permits,‖ ―Regulations about Taxes, Health, Labor and Social Security,‖ ―Information About Production and Quality,‖ ―Foreign Trade,‖ ―Cooperation Between Enterprises,‖ ―Databases,‖ ―Subsidies for SMEs,‖ ―Information About Financing,‖ ―Information About EU for SMEs‖ and ―R&D, Innovation and Technology Transfer.‖

113

Turkey. In cooperation with universities, KOSGEB has also established 20 Technology Development Centers.

4.27. MPM is a smaller organization with a mission to conduct research, disseminate knowledge and provide technical assistance on productivity to both public and private sector organizations. MPM was established by law in 1965, and operates as an autonomous organization under the responsibility of the Ministry of Industry and Trade. It performs a broad scope of functions, involving theoretical, empirical and practical research, as well as monitoring, consultancy and training work. As a result, the roughly 100 experts employed at MPM – out of a total staff of around 170 – represent a wide breadth of fields, including various engineering disciplines, economics, statistics, agriculture, management, sociology, psychology, education, international relations and communications. Consultancy and training services are offered through a central Ankara office and four regional offices. Productivity consultancy services are billed at cost to the client, while group training services are offered free. Training projects are conducted at the provincial level over specific time periods, and each project is tailored to specific needs of the region. From 1998 to 2009, training projects have covered 50 different provinces.

Box 4-3: SME support programs: The international experience SMEs are often unaware of the benefits of available support choices and adopting new modes of production may appear risky. They are constrained to the information they receive from restricted sources, such as personal contacts and a limited number of suppliers. Successful programs use two types of approaches to help SMEs understand their needs. The first more traditional approach consists of outreach activities that convey general knowledge about various topics to firms. The second comprises diagnostic interventions that provide firms with more specific knowledge about their own potential areas of improvement. Outreach activities based on diffusion of general knowledge can take multiple forms. These include free seminars and workshops with visual presentations to explain the use of different technologies or organizational methods to potential users. Extension programs can also promote awareness of available technologies through practical demonstrations. The diagnostic approach to outreach is extensively used by the US MEP and UK MAS programs. The UK MAS program provides a free one-day on-site diagnostic visit by a MAS manufacturing specialist to review a company's entire manufacturing operation, with 3,700 such visits having taken place annually during 2002-05. In 2007, 12 percent of the US MEP‘s assistance was provided through initial technology assessments, provided at no cost to firms. The diagnostic approach can also take the form of performance-benchmarking to compare a firm's use of a specific technology to that of a best-practice firm.

Tailor Interventions to Individual Firm Needs A key characteristic of the most successful technology extension programs is that they offer services targeted at problems that are firm-specific. This is in addition to general group awareness and training activities. In the US, the MEP offered consulting projects to just over 7,500 clients in 2008. During 2002-05, the UK MAS delivered just over 1,000 consulting projects for clients on average, and 3,051 overall. The average company was involved in two consulting projects, with each project on average lasting 10.7 days.

Take a Comprehensive Approach International experience shows that programs need to target a broad range of technological and business areas to be effective. In the US MEP program, most assistance in 2007 was provided in the areas of manufacturing systems (41 percent) and business services (23 percent), but help was also provided on quality systems (11.5 percent) engineering services (12 percent), human resources and organizational development (9.4 percent) and IT (3.1 percent). Programs from the US and UK systems have shown that the extension centers are more effective when they develop a close relationship with their clients. Offering multiple services at the same center, rather than discrete services in different locations, has the advantage of creating longer periods of interaction between the SME and the center, strengthening the ties between two, and building the mutual trust necessary for an effective intervention. Another reason is that many of the barriers to technology absorption are complex and may require addressing different technological and non-technological aspects within a firm. Moreover, the overlap between management processes and technology is blurry.

114

Use Different Forms of Interventions International experience has shown that many problems faced by SMEs are of a very general nature and can be solved through a single phone call or visit from the manager. Technology extension centers typically find that a gradual approach to assistance provision avoids wasting valuable resources. The traditional and most common form of public technology extension services is the provision of technical assistance. Extension services are increasingly being combined with a variety of other types of assistance, such as demonstration programs, workforce training, organizational management and networking services that facilitate a more comprehensive approach to technology diffusion.

Support a Broad Range of SMEs Many international programs are either tailored to SMEs or offer their services at a discount to SMEs. Their clients are predominantly in traditional, medium-technology manufacturing sectors, though this can vary widely in certain countries. Given that these programs are designed to target SMEs, it is interesting to note the range of size of firms served. The US MEP targets primarily SMEs but also works with a number of large firms. In 2007, 21 percent of firms served had between 1-19 employees, 19 percent between 20-49 employees, another 19 percent had between 50-99, 23 percent firms had 100-249 employees and the remaining 18 percent had over 250 employees. The UK MAS program targets small firms more specifically, with 25 percent of its clients during 2002-05 having 1-9 employees, 45 percent of beneficiaries between 10-49 employees, and another 25 percent 49-250, with only 5 percent of clients having more than 250 employees.

Provide Services with Unique Value Added for SMEs Two factors heavily influence whether a technology extension programs has a unique value added for SMEs. The first concerns the quality of the knowledge offered through the program. A reasonable proxy for this is the knowledge and experience of the personnel providing services to the firms. Evaluations from a range of programs globally show that the most critical factor for the success of the program is the quality of program experts, both in terms of their technological competence, as well as the quality of their inter-personal relationship with the client. Personnel should be rigorously screened based on their understanding of local industry needs, technical knowledge and their willingness to follow general extension programs guidelines during their intervention and to engage the firm in building self-knowledge. US MEP staff are generally engineers with more than 15 years of industry experience, sometimes close to retirement age. UK MAS advisors are industry experts including engineers, process managers, and specialists in specific areas of demand by firms, including exporting/marketing, sales and part procurement. The second factor influencing the unique value-added of extension programs is whether similar services are already available to SMEs in the market.

Leverage External Sources of Knowledge International experience shows that partnerships with external sources of support in the region play an important role in the delivery of support to enterprises. Technology extension programs make extensive use of networks of external consultants and experts. In many cases, technology extension programs are affiliated with or managed by regional technology institutes or universities and can draw heavily on their expertise. This is the case with the Japanese Kohsetsushi and German Steinbeis programs.

Minimal Response-time Both the US MEP and UK MAS programs include direct helplines to regional centers where manufacturers can ask quick questions to an expert. In Japan‘s Kohsetsushi Centers, the standard assistance procedure is similar to the MEPs in the United States. Many of the consulting services are performed over the phone and last no longer than 20 minutes. If the problem is more complex, the client sends staff members to the center for consultation sessions. If the problem is still not resolved, a center agent, a registered private manufacturing consultant or a university professor will visit the firm for an average of 6 days.

Decentralized, Flexible, Autonomous Organization Decentralization of public extension programs can result in many benefits, such as enhancing efficiency, service quality and outreach. Comparisons between the more centralized Japanese Kohsetsushi and the US MEP models find that the US programs exhibit far more in the way of experimentation and innovation than seen in Japanese efforts, for example in developing assessment tools, benchmarking measures, and telecommunication techniques. A decentralized operation which maintains network ties maximizes adaptability to local needs and flexibility. Furthermore, local service providers can focus on industrial sectors relevant to their regions while offering

115

specialized services to all sectors by using other centers' resources. A common feature of extension programs is that they emphasize the geographical proximity of the local delivery system to the area it is serving. This ensures an understanding of local industrial needs, and a close and easy interaction with firms that fosters the creation of tight relationships. Geographical proximity and autonomy also create more of a buy-in from the local community, which feels that the services are tailored to its needs. In the US, a central MEP agency coordinates and funds a network of independent ―MEP centers‖ (MECs) mainly state universities and nonprofit organizations. The central MEP provides program guidelines, training, marketing and research on best practices. Overall, the MEP program is decentralized and flexible, with individual centers able to develop strategies and services which are appropriate to state and local conditions. MECs were established in state universities and nonprofit organizations, based on local needs and the host organization‘s past performance, and they receive funding based on their initial ability to match federal funding at the time of the application process. The MECs receive one-third of their budget from federal sources (NIST), one-third from state and local organizations, and one-third from fees collected from client firms. In order to favor a cooperative relationship with the private sector, MECs generally have governing or advisory boards which include local public- and private-sector representatives. The decentralization, flexibility and adaptability of the MEP has enabled centers to respond to local needs and has given the program strong public and private local support. In Japan national guidance and coordination is provided by the Ministry of Economy, Trade and Industry (METI) through its Small and Medium Enterprise Agency (SMEA) and by the Agency of Industrial Science and Technology, but the centers are administered by the prefectural and municipal governments, which provide most of the funding.

Integrate Evaluation and Assessment in the Program International experience shows that mechanisms that monitor the impact of interventions on client firms and periodic independent program assessments provide opportunities to experiment with new program offerings and continuously improve performance. Both the US MEP and UK MAS programs monitor their effectiveness through follow-up surveys of participating companies. Such surveys consist of client valuation such as the impact of the center's intervention on sales, labor costs, material costs, inventory costs, capital investments and jobs, and client progress such as changes in total sales, productivity and income. The MEP monitors center performance standards, such as the number of technical projects initiated. Centers also undertake studies in which the performance of participating manufacturers is compared to others, and case studies linking services to program outcome. In addition, there have been many third party evaluation of the MEP. The UK Government mandates an annual evaluation of MAS against yearly targets that relate to numbers of interventions and generation of additional gross value added (GVA).

Source: World Bank (2010a)

4.28. Existing SME support programs could be improved following international best practices. The study of knowledge flows within Turkish production networks in this Chapter has shown that the absorptive capacity of firms is essential for the adoption of more knowledge intensive modes of organization and production. In addition to addressing broader investment climate constraints related to regulation, finance or the availability of skilled labor, the absorptive capacity of Turkish SMEs would be further enabled by upgrading and expanding the scope of existing programs. In this regard, a number of actions inspired by international best practices (Box 4-3) would contribute to enhance the effectiveness of the assistance received by firms, especially SMEs.

4.29. Proposal: Further decentralize the management model of existing SME support programs in order to better serve local needs and increase the reach beyond micro firms to cater to the needs of larger SMEs. Both KOSGEB and MPM have quite centralized organizational model. KOSGEB, with its 25 regional offices is more able to reach out to different regions, however the programs that are implemented are generally centrally designed leaving few margins of adaptation to local needs. On the other hand most KOSGEB support is demand-driven, since it requires a financial commitment from the beneficiary by being provided in the form of a loan or a matching grant. Moreover, grants for consultancy support can be used according to pre-defined rules and at support ratios determined according to regions, which in turn are defined in KOSGEB‘s Supports Regulation and Application Guidelines. One important limitation is that the firm is limited to using consultants in KOSGEB‘s network for assignments of above

116

US $10,000, and so there is the risk that these consultants may not always have the specific expertise requested by the firm. In this case, firms may hire outside-network consultants, however in line with additional KOSGEB requirements. A number of interventions would enhance the effectiveness of both KOSGEB and MPM programs:  Advance the implementation of a flexible and decentralized management model. A program to address the specific needs of thousands of clients is best implemented in a flexible and decentralized way. With this goal in mind, KOSGEB could better leverage existing resources by decentralizing further the network of technology extension services to universities, industry associations and private providers, as is the case in a number of OECD countries. It could also devise a system of monitoring, evaluation and feed-back mechanism to ensure that organizations delivering the services at the regional level have the ability to tailor programs according to local needs, as in the case of similar programs in the US, the UK and Japan.

 Ensure that the services on offer are not already available on market terms to SMEs in order not to crowd out private providers of such services. The risk of crowding out private service providers is less strong in less developed areas, especially for smaller enterprises, where the availability of managerial and technical skills may be less abundant on the market and SME clients may not be able to afford market rates. Granting greater autonomy to local offices would help tailor public services to local demand and supply conditions.

 Create a single entry-point which could help SMEs better understand their business needs and opportunities. This implies improving the capacity to provide a complete diagnosis of clients‘ business needs and opportunities. In the case of KOSGEB, this would require upgrading the existing ―Business Existing Situation Determination‖ and ―Strategic Road Map‖ questionnaires into a deeper diagnosis and benchmarking activity. It would also require making corresponding investments in human resources. The benefit would be the supply of a service that is more tailored to the specific needs of each company, which has been the main component of programs worldwide that have been internationally recognized as successful. The newly established Development Agencies could be a key local partner to coordinate various SME support programs and better target end-users at the local level.

 Expand the reach of support programs beyond micro-enterprises, to better serve the needs of small (10-49 employees) and medium (50-249 employees) firms. While SME support programs have been able to reach out to many Turkish firms, a large number of companies still remain excluded. In Turkey, there are more than 20,000 small enterprises with between 10 and 49 employees, and more than 10,000 medium-sized firms with 50-249 employees, and they are potentially under-served, since the largest government program, KOSGEB, is mostly used by microenterprises (those with 1 to 9 employees).

References Altenburg, T. 2006. ―Governance patterns in value chains and their development impact,‖ The European Journal of Development Research (18), pp. 498–521

Asanuma, B. 1989. ―Manufacturer-supplier relationships in Japan and the concept of relation-specific skill,‖ Journal of the Japanese and International Economies, 3, pp. 1-30.

Autio E., Sapienza H.J. and Parhankangas A. 2004, "Knowledge relatedness and post-spin-off growth", Journal of Business Venturing, Vol. 19 pp.809 - 829.

117

Bathelt, H. 2003. ―Geographies of production: growth regimes in spatial perspectives on innovation, institutions and social systems.‖ Progress in Human Geography 27 (6), pp.789–804.

Bergman, E.B and E.J. Feser. 1999. ―Industrial and Regional Clusters: Concepts and Comparative Applications. The Web Book of Regional Science.‖ Regional Research Institute, West Virginia University

Brach J., Kappel R. 2009. "Global Value Chains, Technology Transfer and Local Firm Upgrading in Non-OECD Countries," GIGA Working Paper Series 110, GIGA German Institute of Global and Area Studies

Ceiglie G. and Dini M. 1999. ―SME Cluster And Network Development In Developing Countries: The Experience Of UNIDO‖, PSD Technical Working Papers Series, UNIDO

Doloreux D. and Parto S. 2005. ―Regional innovation systems: Current discourse and unresolved issues‖, Technology in Society 27 (2005) pp.133–153

Eraydın A. and Armatlı-Köroğlu B. 2005. ―Innovation, networking and the new industrial clusters: the characteristics of networks and local innovation capabilities in the Turkish industrial clusters‖, Entrepreneurship and Regional Development, 17 (2005), pp.237–266

Erdil, E. And Çetin D. 2008. Innovation and relationships in an organized industrial district: Ankara Sincan Industrial District, TEKPOL Working Paper Series, 2008/02

Erdil, E. .2009. ―Review of international production and global value chain studies: the case of Turkish regional networks,‖ prepared for the World Bank Turkey Investment Climate Assessment

Gereffi, G. 1994. ―The organization of buyer-driven global commodity chains: how US retailers shape overseas production networks,‖ in: G. Gereffi and M. Korzeniewicz (Eds) Commodity Chains and Global Capitalism (London: Praeger)

Gereffi, G. 1999. 'International trade and industrial upgrading in the apparel commodity chain', Journal of International Economics 48, pp. 37–70.

Gereffi, G. and Kaplinsky, R. 2001. ―The value of value chains‖, Special Issue, IDS Bulletin, 32

Giuliani E., Pietrobelli C., Rabellotti R. 2005. Upgrading In Global Value Chains: Lessons From Latin American Clusters, World Development, Vol.33(4), 2005, pp. 549-73.

Göksidan, H.T. 2006. Inter-Organizational relations in an organized industrial district: Ostim case, Unpublished Master Thesis, METU Source: http://etd.lib.metu.edu.tr/upload/2/12607044/index.pdf

Görg H. & Strobl E. 2001. "Multinational Companies, Technology Spillovers and Plant Survival: Evidence for Irish Manufacturing," European Economy Group Working Papers 8, European Economy Group.

Hausman J., McFadden D. 1984. Econometrica, Vol. 52, No. 5, pp. 1219-1240

Hirschman A.O. 1958. The Strategy of Economic Development, Yale University, New Haven.

Humphrey, J. and Schmitz, H. 2002a ―How does insertion in global value chains affect upgrading industrial clusters?‖ Regional Studies, 36, pp. 1017–1027

Humphrey, J. & Schmitz, H. 2002b. ―Developing Country Firms in the World Economy: Governance and Upgrading in Global Value Chains,‖ INEF Report No. 61, University of Duisburg, Duisburg

Kaplinsky, R. 2000. ―Globalisation and unequalisation: What can be learned from value chain analysis?‖ Journal of Development Studies, 37, pp. 117–146

118

Kozan M.K., Wasti S.N, and Kuman A. 2006. Management of buyer–supplier conflict: The case of the Turkish automotive industry, Journal of Business Research, Volume 59, Issue 6, June 2006, pp. 662-670

Kuruüzüm O. 1998. ―Bilimsel Bilginin Ticarileştirilmesinde Yeni Ufuklar: Teknoloji Geliştirme Bölgeleri,‖ Akdeniz Üniversitesi Yayınları

Lall S. 1978. ―Transnational, Domestic Enterprises and Industrial Structure in Host LDCs. A Survey.‖ Oxford Economic Papers , vol. 30, n.2, pp.217-248

Malecki, E., Oinas, P. (Eds.) 2000. ―Making Connections: Technological Learning and Regional Economic Change‖ Ashgate, Aldershot

Markusen, J. 1995. The boundaries of multinational enterprises and the theory of international trade. Journal of Economic Perspectives, 9, 169–189.

Markusen J., Venables A.J. 1999. ―Foreign Direct Investment as a Catalyst for Industrial Development.‖ European Economic Review 43 (1999), pp.335-356.

Nadvi, K. (1995), Industrial Clusters and Networks: Case Studies of SME Growth and Innovation, UNIDO, Vienna Oz O. (2004), Clusters and Competitive Advantage: The Turkish Experience, New York: Palgrave Macmillan, 2004.

Özkanlı Ö., Erdil E., and Akdeve E. 2008. Innovation and Relationships in Industrial Districts: The Case of Turkey, Innovative Marketing, 4, pp.114-119

Pietrobelli, C. and Rabellotti, R. 2007. ―Upgrading to Compete. Global Value Chains, Clusters and SMEs in Latin America,‖ Cambridge, MA: Harvard University Press.

Rodriguez-Clare, Andre. 1996. ―Multinationals, Linkages, and Economic Development,‖ American Economic Review 86, pp.852-873.

Saliola,F. and Zanfei, A. 2009. ―Multinational firms, global value chains and the organization of knowledge transfer.‖ Research Policy 38 (2009) pp. 369–38

Smarzynska-Javorcik B. 2004. "Does Foreign Direct Investment Increase the Productivity of Domestic Firms? in Search of Spillovers Through Backward Linkages." American Economic Review, American Economic Association, vol. 94(3), pages 605-627, June

Sturgeon, T. 2001. How Do We Define Value Chains and Production Networks, Industrial Performance Center, Massachusetts Institute of Technology.

Tallman, S., Jenkins, M., Henry, N. and Pinch, S. 2004. Knowledge, clusters and competitive advantage. Academy of Management Review, 29, (2), 258-271.

Ulusoy, G. 2003. ―An assessment of supply chain and innovation management practices in the manufacturing industries in Turkey.‖ Int J Prod Econ, 86, pp. 251–70.

Vale M., and Caldeira, J. 2008. ―Fashion And The Governance Of Knowledge In A Traditional Industry: The Case Of The Footwear Sectoral Innovation System In The Northern Region Of Portugal,‖ 2008, Volume 17, Issue 1-2, pp. 61-78.

Wagner, J. 2007. ―Exports and Productivity: A Survey of the Evidence from Firm-level Data,‖ The World Economy, Blackwell Publishing, vol. 30(1), pages 60-82, 01

World Bank, 2010a. Innovation Policy for Competitiveness and Employment Generation. Forthcoming

119

Annex 4-A. Econometric Methodology and Results Based on the insights from the GVC literature, we define a measure of value chain governance based on three variables:91 (i) Buyer‘s provision of information on Design/Quality (i.e. product characteristics) and product quality standards imposition; (ii) Buyer‘s engagement in the supplier process or product R&D activities; (iii) Buyer‘s provision of employees to (or organized personnel exchanges with) suppliers as a means to disseminate and diffuse new technologies into the local firms‘ production facilities.

GOV is regressed on the different measures of IC. The equation to be estimated will thus be of the following type:

ICvarjc is the average value of the average investment climate variable at industry j and region c. We also control for size, region and industry fixed effect, – which means that results hold regardless of these characteristics. Additional controls are included for the share of foreign and domestic ownership, and for the share of output produced for export. To account for endogeneity, we construct a measure of the IC conditions faced by firm i on dimension k, by averaging the responses of firms in the same location- sector. To ensure adequate numbers of firms in each location- sector cell average, we drop those cells with less than 5 observations. Standard errors are also clustered at location- industry level. We use a multinomial logit model approach to investigate the correlation between value chain governance modes and the characteristics of firms and IC in Turkey. Since GOV0, GOV1, GOV2 and GOV3 represent qualitatively different modes of organizing knowledge transfer, which may each be influenced by different combinations of explanatory variables, a multinomial logit model is the most suitable for this analysis. We estimated the multinomial logit model by maximum likelihood method. As we are interested in the change in absolute probability of the outcome governance induced by the regressors, we shall calculate the marginal effects. We define a multi-category variable (y) as dependent variable. Let x be the vector of explanatory variables, the multinomial logit model response probability takes the following form:

K Pr (yn  k)  Pnk  exp(xk ) /[1  exp(xk )] k0 for k=0, 1, …K where Pnk is the probability that the dependent variable (Yn) takes value k at nth observation, with k ranging from 0 to K. In our case k will take values 0, 1 or 2 to identify four different governance modes: GOV0 (k=0), GOV1 (k = 1), GOV2 (k = 2) and GOV3 (k = 3). It should be mentioned that the point estimates of a multinomial logit tell us, for each choice k, the change in probability of the outcome k, relative to the baseline outcome (k = 0), induced by a unit change in the explanatory variables. In a multinomial framework, this does not assure that the absolute probability of outcome k will increase or decrease, but that k will be more or less likely relative to baseline outcome. As we are interested in the change in absolute probability of the outcome k induced by the regressors, we shall calculate the marginal effect:

P nk  k xn Estimation of the multinomial logit model is based on the assumption that probabilities of the alternative choices are independent of each other. This property is called the independence from irrelevant alternatives (IIA). The validity of this assumption is checked using the test introduced by Hausman and McFadden (1984). The available data allow

91 The survey included questions on the establishment‘s sales made to the largest buyer, the number of years of the relationship with the largest buyer, and the percentage of sales made according to largest buyer‘s unique specification. In additions, firms were asked whether the largest buyer was engaged in the provision of information on Design/Quality (i.e. product characteristics) and imposed product quality standards, whether the largest buyer provided inputs on firm‘s process or product R&D activities, and whether the buyer provides employees to (or organized personnel exchanges with) the firms as a means to disseminate and diffuse new technologies into the local firms‘ production facilities. Using this information, we investigate the characteristics of buyer-supplier inter-firms linkages and the intensity with which buyers are engaged in transferring knowledge and technology to Turkish suppliers. The analysis is restricted to those firms which claimed to have a largest buyer and to sell at least 5 percent of their sales to the largest buyer (90 percent of firms in the sample).

120

us to carry out only a cross-section analysis which highlights simple correlations and no causal links between variables. Along with multinomial logit regressions, we run probit regressions for each of the three governance components separately. The purpose of the additional empirical exercise is to complement the analysis by investigating possible additional features in the relationships between IC variables and governance components, if any.

Table 4-A-1: Summary of multinomial logit results GOV0 GOV1 GOV2 GOV3

Variable Coef. P>z Coef. P>z Coef. P>z Coef. P>z

Business inspections (number) 0.025 0.048 -0.008 0.459 -0.009 0.354 -0.007 0.489 REGULATORY Operating licenses (days) 0.029 0.454 0.009 0.108 0.002 0.475 -0.010 0 ENVIRONMENT Permits (days) 0.009 0.002 -0.004 0.037 0.000 0.708 -0.005 0.08 Time tax -0.006 0 0.005 0 0.003 0.081 -0.002 0.055 Skilled workers -0.012 0.005 0.002 0.475 0.008 0 0.002 0.227 LABOR & SKILLS Training (dummy) 0.010 0.8 -0.098 0.002 -0.025 0.403 0.113 0 Manager experience -0.024 0.474 -0.010 0.724 0.006 0.716 0.028 0.088 Quality certifications (dummy) -0.139 0.791 -0.156 0.647 -0.524 0.321 0.819 0.004 R&D (dummy) -0.174 0.000 -0.037 0.126 0.137 0.000 0.074 0.000 INNOVATION Foreign technology (dummy) -1.938 0.000 0.881 0.000 0.178 0.516 0.878 0.000 New product (dummy) -0.109 0.598 -0.125 0.466 -0.419 0.038 0.653 0.000 Computer usage (share of workforce) -0.025 0.001 0.011 0.024 -0.004 0.518 0.019 0 Bank financing 0.312 0.675 1.343 0 -3.990 0.001 2.334 0 ACCESS TO FINANCE Loans (dummy) 0.444 0.478 -0.335 0.576 0.472 0.062 -0.581 0.46 Collateral (value as share of loan) -0.001 0.252 -0.002 0.015 0.002 0 0.001 0.12 Inputs imported 0.818 0 -0.256 0.047 -0.133 0.272 -0.429 0.005 Crime (dummy) 0.144 0.863 0.503 0.166 -1.142 0.315 0.496 0.039 Losses from outages 0.033 0 -0.017 0.056 0.001 0.889 -0.017 0.105 OTHER Subsidies (dummy) 2.310 0.014 -0.703 0.396 -1.316 0.191 -0.291 0.705 Domestic loss -0.088 0 0.037 0 0.038 0 0.012 0 Domestic sales, more that 90% of -0.499 0 0.055 0.519 0.496 0 -0.051 0.483 total (dummy) Note: Regressions include region, industry, size, exporter and foreign ownership dummies

Annex 4-B. International and Domestic Trade Flows: A Regional Perspective The Aegean and Marmara regions are the most export oriented, but both experienced a decrease in exports between 2007 and 2008, with Central Anatolia being the only region showing an increase. In each of the five regions included in the survey sample, firms were asked to provide the share of sales made for direct exports, exports through a distributor (indirect exports) and domestic sales in fiscal year 2008. Overall, firms in the sample sell 71 percent of their sales into the domestic market, 12 percent of sales are made through distributors and 17 percent of sales are exported. In addition to having the highest share of exporters (50 percent of the total number of firms), the Aegean and the Marmara regions are the most export oriented by value, showing the highest share of direct exports and exports made through distributors (Figure 4-B-1a). It is interesting to note that Central Anatolia exhibits the lowest share of exports through distributors and the highest portion of national sales (85 percent). The Black-Sea Eastern region places right after Central Anatolia with a share of 79 percent of domestic sales and 12 percent indirect exports.

121

Figure 4-B-1: International and domestic trade flows (sales) a. Fiscal year 2008 (%) b. Fiscal year 2008 vs. 2007 (% change) 100 85 79 25 74 19.6 80 71 66 57 20 60 15 40 26 10 6.8 7.8 17 1520 17 5.3 20 12 12 1313 129 3 5 0.30.1 0 0 -0.5 -0.4 -5 -2.4 -1.8 -3.4 -4.7 -10 -6.3 -5.4 -15 -15.0 National Indirect exports Direct exports -20

National Indirect exports Direct exports

Source: Turkey NIS 2009

Figure 4-B-1 provides a regional synthesis of the redirection of exports occurred as a consequence of the decline in global demand discussed in Chapter 1. Between 2007 and 2008 firms in Marmara experienced a decrease in their direct exports and an increase in domestic sales. Firms in Central Anatolia are the only ones which experienced an increase in their direct exports. Interestingly, Aegean, firms experienced an increase in their exports through distributors, while firms in the South region did not experience any significant change in the composition of their sales flows.

The South is the region that is most Figure 4-B-2: International and domestic trade flows (input dependent on foreign inputs. In terms of purchases) number of firms importing inputs, firms in 100 90 91 84 84 the South region are the most exposed to 82 78 80 trade (55 percent of firms are importers), followed by Aegean and Marmara. Firms in 60 Central Anatolia and Black Sea Eastern confirm to be the least exposed to trade with 40 external markets. Paralleling the analysis of 22 exports, firms were also asked to provide the 16 18 16 20 10 9 value share of inputs imported directly and purchased in the domestic market. Businesses 0 in the sample import 16 percent of their Total Marmara Aegean Central South Black Sea- inputs directly and purchase 84 percent in the sample Anatolia Eastern domestic market (Figure 4-B-2). Central Purchased in the domestic market Imported inputs Anatolia and the Black-Sea region exhibit the highest share of national inputs purchases. Source: Turkey NIS 2009 The South region seems to be the most dependent on foreign inputs with 22 percent of inputs imported directly, followed by Marmara (18 percent).

Central Anatolia, Black-Sea Eastern and Marmara are the most closed in terms of inter-regional sales flows, selling about three-quarters of their output within regional borders. An interesting finding is that Marmara represents the second most important domestic sales destination market for each of the other regions in the sample. In addition, Aegean and South regions are the most connected areas to the Marmara market, selling on average 26 and 27 percent of their total domestic sales to the Marmara region, respectively. It is also worth mentioning that regions with the least diversification in inter-regional sales flows – Central Anatolia and Black Sea Eastern – are also the least exposed to foreign trade. On the other hand, the Aegean and South regions show significant shares of inter-regional trade flows (about 50 percent). The Aegean region also represents the most open area with respect to export flows.

122

Figure 4-B-3 Intra and inter-regional trade flows: sales and input purchases

Source: Turkey NIS 2009

Marmara represents the main site for sales and input purchases for a majority of industries. Marmara stands out for the high portion of domestic sales in the Textile sector directed to the region (57 percent). The South region represents a very important site along with Marmara for the non-metallic products industry receiving about as much domestic sales as the Marmara region (31 percent Marmara and 29 percent South). In the Chemicals industry, the Central Anatolia region represents the most important destination of sales (50 percent Central Anatolia and 24 percent Marmara). Central Anatolia represents also the second most important region, after Marmara, for Food (28 percent), Garments (27 percent) and other manufacturing (20 percent) sales. Regarding input purchases, especially in the Garments industry, the percentage of inputs purchased in Marmara is substantial (81 percent on average). Central Anatolia represent the second most important site, after Marmara, in the Food, Chemicals and Other

123

manufacturing industries, while the South region represents the second most important site, after Marmara, in the Textile and Non-metallic industries.

Marmara is the core location for domestic input purchases while Black–Sea Eastern displays the highest share of regional diversification in terms of inputs purchases. In terms of material input purchases, in the Aegean, Central Anatolia and South regions at least 50 percent of inputs are purchased within regional borders. At the same time Black–Sea Eastern appears as the region that is most connected to other regions, as its enterprises purchase only 37 percent of inputs internally, acquiring a significant share from Marmara (35 percent) and from Central Anatolia and South (23 percent in total). It is also interesting to note that Black–Sea Eastern is the region with the highest share of non imported inputs and with highest regional diversification in terms of inputs purchases. The region of Marmara, on the other hand, appears to be the most self-sufficient in terms of inter-regional inputs purchases supporting 87 percent of its 82 percent of non-imported inputs. The Marmara region is the also core location for domestic input purchases for all the regions included in the sample. On average, in fact, each region purchases about 30 percent or more of its inputs from Marmara

124

Chapter 5. ENHANCING REGULATORY CAPACITY FOR BETTER REGULATION 5.1. Productivity, employment and the ability to export as well as to receive FDI in the Turkish business sector are all negatively affected by long and complex regulatory procedures with large regional differences. The econometric analysis of the Turkey Enterprise Survey (ES) confirms that the measured regulatory environment variables have a significantly negative impact on productivity, as well as other performance measures, such as employment and export propensity. The business sector overall proves to function on a higher productivity level with well-functioning regulatory mechanisms in place. As presented in Chapter 2, examples of variables with a negative impact on productivity include more frequent inspections, a higher number of compulsory certificates, longer time dealing with these, as well as long customs clearance procedures. Additionally, the econometric results indicate that smaller firms struggle more obtaining necessary certificates, whereas businesses‘ export propensity is negatively associated with business inspections and payments for compulsory certificates. There are also variations across the five Turkish regions (Aegean, Black Sea-Eastern, Central Anatolia, Marmara and the South) when looking at such concerns as obtaining operating and import licenses, construction permits, clearing goods through customs, getting access to land and dealing with inspections. This could be seen as a sign of inconsistency and lack of coordination across Turkish regions, a subject that will be discussed in detail later in this chapter.

5.2. Turkey still lags behind developed OECD economies in offering efficient license and permit procedures. A measurement of regulatory systems in OECD countries is available through the Product Market Regulation (PMR) indicators (see Box 2-5 for description of PMR). When examining regulatory and administrative opacity, there seem to have been significant improvements in Turkey in regards to communication and simplification of rules and procedures in the period 1998-2008. Although this has contributed to an overall improvement in lowering barriers to entrepreneurship for Turkish enterprises, dealing with licenses and permits remains cumbersome, with a continuous 6.0 index point ranking. Furthermore, Turkey is still lagging behind higher income OECD economies, such as France, Germany and the UK, as can be seen in Table 5-1.

5.3. This quantitative background prepares for a discussion on the actual policies in place in Turkey and the country‟s capacities to organize and implement high quality regulation. In recent years, Turkey has made progress in improving its business environment, tackling several areas identified as constraints for doing business, such as taxation procedures and customs regulations.92 Through Turkey‘s e-Government program, an online tax filing system has been introduced, significantly cutting the time to comply with tax regulations. Furthermore, an automated customs clearance system has been implemented, allowing for most customs procedures to be carried out electronically.

5.4. One area that has been highlighted as particularly cumbersome is the amount of time that businesses spend in complying with paperwork and business regulations. According to responses in the ES conducted for Turkey in 2008, the share of senior management‘s time spent dealing with requirements imposed by government regulations averaged 27 percent, a steep increase from the 2005 Survey, where the average was 9 percent. Moreover, Turkey ranks much higher than any other comparative emerging economy interviewed for the ES, where the average for the ECA region is 15 percent. Accepting a degree of subjectivity in survey responses, this concern has led to investigating the causes for this perception in terms of the institutional framework underlying the formation and implementation of regulation affecting businesses.

92 The issues of tax and customs procedures are discussed among others in the World Bank Country Economic Memorandum – Sustaining High Growth: Selected Issues, 2008

125

Table 5-1: Product Market Regulation - Barriers to Entrepreneurship 2008: Regulatory and administrative opacity

France Germany Hungary Italy Portugal Spain Turkey UK

Licenses and permits system Is the ―silence is consent‖ rule (i.e. that licenses are issued no no yes yes yes yes no yes automatically if the licensing office has not acted by the end of the statutory response period) used at all?

Are there single contact points (―one-stop shops‖) for getting yes yes yes yes yes yes no yes information on notifications and licenses? Are there single contact points (―one-stop shops‖) for issuing or yes no yes yes yes yes no no accepting notifications and licenses?

Country scores (0-6) 2.00 4.00 0.00 0.00 0.00 0.00 6.00 2.00

Communication and simplification of rules and procedures

Communication Are there systematic procedures for making regulations known yes yes yes yes yes yes yes yes and accessible to affected parties? Is there a general policy requiring "plain language drafting of yes yes yes yes yes yes yes yes regulation? Do affected parties have the right to appeal against adverse all all cases some all all cases all all all enforcement decisions in individual cases? cases cases cases cases cases cases Regulations published or otherwise communicated to the public yes yes yes n.a. yes yes yes yes in a manner accessible at the international level: There are inquiry points for information on the operation and enforcement of regulations Does government policy impose specific requirements in relation yes yes yes yes yes yes yes yes to transparency and freedom of information government wide?

Simplification of rules and procedures Does the national government (all ministries and agencies) keep a yes no no yes yes yes no no complete count of the number of permits and licenses required?

Is there an explicit program to reduce the administrative burdens yes yes yes yes yes yes yes yes imposed by government on enterprises and/or citizens?

Is there a program underway to review and reduce the number of yes yes yes yes yes yes yes no licenses and permits required by the national government?

Country scores (0-6) 0.00 0.09 1.11 0.00 0.00 0.00 0.50 0.23 Source: OECD. Note: The indicators depict strictness of regulatory environment on a scale 0 to 6 (0=Yes; 3=In some cases; 6=No), with higher numbers indicating more restrictive policies towards competition. For a detailed description of methodology, see Wölfl et al. (2009)

5.5. Regulatory quality is generally recognized not only to be a key component of good governance, but also clearly correlated to measures of growth and productivity.93 For example,

93 Several authors and institutions have explored these links. According to Kaufman et al (2002), the quality of regulation and governance is correlated to better economic outcomes. OECD has found that ―reforms that reduce competition-restraining regulations, cut tariff barriers and ease restrictions on foreign direct investment to ―best practice‖ levels in the OECD area could lead to gains to GDP per capita of up to 2 to 3 per cent in the European Union, where productivity is already higher than in most developing countries‖ OECD (2005b). The 2005 Doing Business report found that indicators of cost to business in the areas covered by the report seem correlated with higher rates of job creation. However, rather than showing that cost-cutting in these areas can produce macroeconomic effects, the correlation might be due to the fact that countries regulate consistently, using a national style that is reflected in the Doing Business indicators. This would support the OECD‘s 1999 conclusions about the pro- competitive style of regulation in the United States. A group of researchers finds a strong correlation between regulation of entry

126

administrative, compliance and economic costs generated by government regulations are large. In developed countries, calculations of administrative costs to business of complying with regulatory obligations suggest that a 25 percent reduction in the Netherlands will reduce administrative burdens from 3.7 percent to 2.8 percent of GDP. This is still significantly above the UK (1.5 percent), Denmark (1.9 percent) as well as several other European countries.94 In Australia, regulatory compliance costs on business are estimated to be 4 per cent of GDP.

5.6. Turkey has made important steps in establishing a system and key institutions for regulatory reform. Since the 2001 OECD Report on Regulatory Reform,95 the Government has paid particular attention to establishing institutions and mechanisms for regulatory reform; enacting legal reforms conducive to simplification of the legal framework; and introducing, through pilot projects, a number of regulatory tools to improve the quality of regulations. In this process, achieving EU harmonization has been a key driver of reform and Turkey has partially embraced the EU Better Regulation agenda in a number of areas. As a result, Turkey has established solid pillars of a regulatory system that have the potential to develop into a ―whole-of-government‖ approach to regulatory management and reform.

5.7. After an economic crisis in 2000, Turkey embarked in a deep and comprehensive economic reform program comprising institutional changes in the way regulations were prepared and revision of the existing regulatory framework.96 Among the different challenges Turkey still faces to increase productivity and employment to EU levels97, regulatory reform and in particular reducing red tape, appears to be fundamental. A number of formal mechanisms have been established to promote a better investment climate in Turkey. In order to rationalize bureaucratic procedures and reduce red tape, a comprehensive reform program was launched in 2001 by a Council of Ministers Leading Decision. The reform program has been implemented by two platforms which provide a channel to reflect views and priorities of the international and national private sector representatives:

5.8. The Coordination Council for the Improvement of the Investment Environment (YOİKK)98 has become a key structure where the private sector makes contributions in the process of improving the investment climate. The Council conducts its agenda with the help of 12 Technical Committees working on specific issues with participation of both public and private institutions.99

into markets and the regulation of labor. As they explain this: ―countries have regulatory styles that are pervasive across activities and shaped by the origin of their laws.‖ Juan Botero et al (2004). 94 Kox (2005) 95 OECD (2001) 96 In many countries, crises open up the possibility to be innovative and introduce changes in the system. Korea, for instance, designed a comprehensive strategy for regulatory management and reform after the Asian crisis in 1997. At the core of the reforms introduced in Korea, regulatory reform was fundamental to overcome the financial crisis. The 1998 regulatory reform program included two key initiatives. The first was a massive deregulation initiative in which the president ordered each government ministry to eliminate 50 percent of its regulations. The second was an enduring institutional reform that established institutions and mechanisms at the center of government to promote reform and monitor and guarantee the quality of regulations and the regulatory process. World Bank (2008), p. 1 97 World Bank Group (2007) 98 The following institutions are part of YOİKK: Undersecretary of the Ministry of Finance, Undersecretary of the Ministry of Industry and Trade, Undersecretary of the State Planning Organization, Undersecretary of the Treasury, Undersecretary of Foreign Trade, Chairmen of the Technical Committees, Chairman of The Union of Chambers and Commodity Exchanges of Turkey (TOBB), Chairman of The Turkish Industrialists‘ and Businessmen‘s Association (TÜSİAD), Chairman of the International Investors Association (YASED), and Chairman of the Turkish Exporters‘ Association (TİM). The secretariat works of YOİKK Platform are conducted by the Undersecretariat of Treasury. 99 The technical committees are: Technical Committee on Company Establishment, Technical Committee on Employment, Technical Committee on Sectoral Licenses, Technical Committee on Investment Location, Technical Committee on Taxes and Incentives, Technical Committee on Foreign Trade and Customs, Technical Committee on Intellectual Property Rights, Technical Committee on Investment Promotion, Technical Committee on Foreign Direct Investment Legislation, Technical

127

YOİKK aims to rationalize the regulations on investments in Turkey, develop policies by determining the necessary arrangements that will enhance the competitiveness of the investment environment, generate solutions to the administrative barriers encountered by the domestic and international investors in all phases of the investment process including the operating period. The Investment Advisory Council of Turkey (IAC) is an international platform established to receive the recommendations of executives of high ranking multinational companies and international institutions regarding the Turkish business environment. The recommendations stated in annual meetings by the Council members regarding business environment in Turkey are taken to the agenda of YOİKK Technical Committees and to the following Investment Advisory Council meeting. YOİKK Technical Committees have conducted studies to identify solutions to the different constraints that businesses face in Turkey. The YOİKK Technical Committee Action Plans, initially prepared in 2007, were reviewed on the 14th YOİKK Meeting on March 2010 to tackle the bottlenecks faced by the investors. As a result, the YOİKK Technical Committee Action Plans for 2010 have been prepared, comprising the activities planned to be realized in 2010 by YOİKK Technical Committees to streamline the investment environment.

5.9. Another recent development in the interface between the government and the business sector is the set up of development agencies at provincial level. As discussed in Chapter 2, these agencies could act as an information point and to rationalize initiatives, especially for SMEs, at the local level, in close cooperation with TOBB and other government agencies. These agencies could act as ―one- window‖ shop, facilitating the acquiring of licenses and permits between issuing agencies and firms.

5.10. A fundamental driver for regulatory reform in Turkey is commitment to EU harmonization. The National Program for the Adoption of the EU Acquis (2007-2013) aims at achieving harmonization with the acquis communautaire with the perspective of full membership of the EU. The Program includes all chapters of the acquis envisaged to be adopted following a screening process. Legislative measures, secondary legislation, and the main strategy and policy papers required in the relevant chapters and financial requirements and resources needed are included in this Program. Furthermore, the Program has identified the responsible institution and the timeframe for the adoption of legislation. Eleven chapters are open for EU accession. Legal reforms continue in accordance with the National Program. The quality of new or amended regulations has also benefited from the EU accession process: any law or regulation amended for harmonization purposes is sent to the Secretariat General for EU Affairs (ABGS) in accordance with the Decree Law No. 2008/14481 and Regulation No 2005/9986 (Article 6 (f)), before being sent to Prime Ministry for review of compliance with the acquis communautaire. ABGS has also permanent contact points at different political levels in all Ministries and institutions to ensure that EU legislation is considered when preparing regulations. ABGS also participates in commissions when draft laws are discussed in the National Assembly. Despite these efforts, the EU Better Regulation agenda has not been fully adopted in Turkey.

5.1 Regulatory Policies and Institutional Drivers in Turkey 5.11. Pursuing private sector development and reforming regulation are two interconnected policy areas that should have clear objectives. Some principles and objectives of reform, as well as the responsibilities of the groups involved in reform should be identified by governments. The most effective way to do this is to establish an explicit regulatory reform policy, based on internationally accepted principles of good regulation (Annex 5-1). Despite the achievements mentioned above, a number of institutional challenges remain to improve the quality of regulation and regulatory processes in Turkey. In particular, regulatory reform is not yet an integral policy area linked to the improvement of the business environment, as it is the case of many other OECD countries (Box 5-1). Even if efforts have

Committee on Small and Medium Size Enterprises, Technical Committee on Corporate Governance and Technical Committee on Research and Development.

128

been undertaken to link both issues, and YOİKK seems to be an appropriate platform for this purpose, the lack of a clear strategy for regulatory reform to promote a friendlier business environment in the country is evident.

Box 5-1: Regulatory reform to improve the business environment: International experiences Many OECD countries have established wide and comprehensive regulatory reform programs with a strong focus on establishing a more friendly environment for businesses:

In Australia, governments have undertaken major policy reforms over the past two decades, which have contributed to the country‘s strong economic performance. These reforms have included regulatory changes to expose the economy to greater competitive pressures and to provide firms with greater flexibility to respond. In the same period, however, Australia has experienced a dramatic rise in the volume and reach of regulation, in response to a variety of social, environmental and economic issues. Since 1990 the Australian Parliament has passed more pages of legislation than in the nine preceding decades. In this context, Australian governments have acknowledged that the country could not function well without regulation. But there is too much regulation and, in many cases, it imposes excessive and unnecessary costs on businesses. The Regulation Taskforce established in 2006 identified a forward agenda comprising some 100 reforms to existing regulation that could provide relief to businesses. It also considered how the processes and institutions responsible for regulation could be improved. As a result of this research, the Australian Government, the Council of Australian Governments (COAG) and most Australian states and territories have integrated approaches recommended by the Taskforce into their best practice regulation processes.

The Netherlands has focused its regulatory reform program in reducing the regulatory burden for businesses and society. Between 1994 and 2003, a reform approach was incrementally developed, consisting of the following characteristics: a quantitative method, quantified and individualized targets, external pressure from a watchdog (Actal), internal organization coordination (the IPAL-unit) and monitoring using the budget cycle. The program was reviewed in 2007 and the current Dutch Cabinet has committed to simplify the regulatory burden by another 25 percent in 2011, for example by reviewing contradictory rules, cancelling permits and licenses, reducing administrative and supervisory burdens and improving services. Some of the innovations included in the current program are: the program was taken beyond administrative burdens, involving the measurement, quantitative targeting and reduction of substantive compliance costs and increasing the quality of services delivered to businesses with regards to regulation affecting them. At institutional level, the coordination was facilitated by merging the project groups in a new unit dealing with all aspects of regulatory reform for businesses. The tasks and responsibilities of Actal, the watchdog, were also broadened, including substantive compliance costs and an audit role of regulatory impact analysis.

Source: Government of Australia (2006) and Nijland (2008)

5.12. In Turkey, there is an important disconnection between the efforts undertaken to improve the legal aspects of laws and regulations and the purpose of regulatory reform, which is to create a legal framework that is conducive to sound economic activity and social welfare. The gains made in improving the legal aspects of regulation have not been embedded in the way regulations impact the economy. The Turkish legal system is composed of a number of regulatory instruments stemming from various sources of law (Annex 5-2). In the preparation of laws and regulations, the following phases can be identified (Figure 5-1).

129

Figure 5-1: Phases of the preparation of laws in Turkey

Submission for Preparation by Publication in Official ratification to the Ministries Gazette/Notification President

Request of opinion from other Submission to institutions and National Assembly stakeholders

Ratification of line Submission for minister signature to Cabinet

5.13. Following the OECD Principles for Regulatory Quality and Performance, a number of important challenges remain in order for regulatory policy to find its place as a key priority for government action. One initial constraint is to find high political support for regulatory reform as a key priority area of government action. One of the fundamental elements for success in any regulatory reform is the strong commitment and political support from the highest political level. This has been particular important in regulatory reform programs established in emerging countries such as Mexico or Korea, where the involvement of the President made possible the introduction of a regulatory policy. In Turkey, regulatory reform has permeated certain senior management levels in different ministries, a number of technical experts and various stakeholders that are aware of the importance to embark in broad regulatory reform. However, one of the remaining challenges is to ensure ownership of regulatory reform at high political level in the country.

5.14. Regulatory policy should serve clearly identified policy goals, and be effective in achieving those goals. In many OECD countries, such as the UK, regulatory policy is clearly defined as a way ―to eliminate obsolete and inefficient regulation, create user-friendly new guidelines and tackle inconsistencies in the regulatory system.‖100 In Turkey regulatory policy remains a fragmented area in which different institutions establish their own policy goals and agenda.

5.15. Regulatory policy should also include the establishment of frameworks for implementation. This means that governments should ensure that institutional frameworks and resources are adequate, and that systems are in place to manage regulatory resources effectively and to discharge enforcement responsibilities. In Canada, for instance, The Regulatory Affairs Sector, transferred in July 2006 from Privy Council Office, supports the Treasury Board Committee in its role as the "Queen's privy council for Canada" by providing advice to the Governor General and by providing management and oversight of the government's regulatory function. In addition, it provides policy leadership on the federal regulatory policy contained in the Cabinet Directive on Streamlining Regulation. In Turkey the institutional framework for regulatory policy remains weak and strong leadership is needed. Turkey has the opportunity to build over the existing pillars to link the private sector and investment climate initiatives with a more comprehensive approach to regulatory policy.

100 The Better Regulation Executive in charge of the regulatory reform program in the UK has identified the following actions in order to implement regulatory policy: i) using targeted measures to simplify and improve existing regulation; ii) communicating more clearly with businesses, to help them understand what they must do to comply; iii) carefully assessing the impact of any new regulations and iv) working with the EU to improve European guidelines. http://www.berr.gov.uk/whatwedo/bre/policy/page44059.html

130

5.16. Regulatory reform needs appropriate institutions at the centre of government to lead this process. Many emerging countries are setting up regulatory reform units that establish mechanisms with procedures that spell out explicit responsibilities for managing and tracking regulatory reform inside the administration. This is done with the objective to keep reform on track and on schedule, and to ensure that regulatory quality standards improve. Many developed countries have strong oversight bodies for regulatory management and reform that are responsible for a ―whole-of-government‖ approach to this policy area.

5.17. Different regulatory institutions exist in Turkey, having different roles and responsibilities, with mandates covering different aspects of regulatory reform. There is, however, no single oversight body for regulatory reform that concentrates all responsibilities to oversee the quality of the regulation, as it is the case in other OECD countries (Box 5-2).

Box 5-2: International examples of oversight bodies for regulatory reform Many OECD countries and emerging economies have established oversight bodies for regulatory reform with the aim of institutionalizing regulatory reform, by allocating clear responsibilities to a single unit in charge of regulatory policy and its implementation:

In the UK, the Better Regulation Executive (BRE) is the institution responsible for regulatory reform issues across government. Located in the Department for Business, Enterprise and Regulatory Reform (BERR), it works with government departments and regulators to scrutinize new policy proposals, to achieve effective new regulations, to make it easier to change or remove regulation where beneficial, to reduce existing regulatory burdens affecting business, the third sector and frontline staff in the public sector, to improve transparency and accountability for regulation, to effectively communicate regulatory changes and to drive forward the better regulation agenda in Europe.

In the United States, the Office of Information and Regulatory Affairs (OIRA) is the oversight body for regulatory reform. As a federal office that Congress established in the 1980 Paperwork Reduction Act, it is part of the Office of Management and Budget, which is an agency within the Executive Office of the President. In addition to reviewing collection of information, OIRA reviews draft regulations and develops and oversees the implementation of government-wide policies in the areas of information technology, information policy, privacy, and statistical policy. OIRA reviews agency draft regulations before publication to ensure agency compliance. The review includes consideration of alternatives to the rulemaking and analysis of the rule‘s effects on society, both its benefits and costs. OIRA has more than 50 full-time professionals who work with agency professionals on specific issues and decisions. The majority of OIRA employees are career public servants.

In Italy, the Simplification Unit (Unità per la Semplificazione) has been moved from the Legal Department of the Prime Minister‘s Office to the newly created (May 2008) Ministry of Simplification. The Unit is in charge of cutting red tape, simplification measures, very basic quantitative assessments linked to simplification and the implementation of the taglia-legge (guillotine law) at federal level. The Better Regulation Unit is composed of 30 experts nominated for three renewable years, with proved expertise on legal, economic and organizational issues. In Korea a Regulatory Reform Committee (RRC) was set up by law in 1997 with a ―general mandate to develop and co-ordinate regulatory policy and to review and approve regulations.‖ Its main functions are to give some strategic perspective in the regulatory reforms, to undertake research, to monitor improvement efforts of each agency and to make sure there is coherence between their actions. The RRC is composed of 25 members, 18 of whom are from the private sector and 7 are government officials from various departments. The RRC is jointly chaired by the Prime Minister and one member from the private sector appointed by the President. It is one of the cases where more power has been given to this kind of institution, multiplying the „engine of reform‟ effect. The secretariat function supporting the RRC is undertaken by the Regulatory Reform Office which is located in the Prime Minister‘s Office. This unit includes around 40 civil servants and 3 professional experts, under the direction of the Deputy Minister for Regulatory Reform.

131

5.18. Four main institutions are active within the Prime Ministry‟s office in charge of improving the quality of new and existing laws and regulations. These institutions have made progress in introducing principles of high quality regulation and establishing clearer criteria to prepare and review laws and regulations. These are:

 General Directorate for Laws and Decrees. This Directorate is mainly responsible for reviewing the quality of new laws and decrees in Turkey. Their review concentrates on the legality of the drafts and on the consistency of regulation with the Government‘s main development plans. The Directorate also ensures that consultation among government agencies has been conducted during preparation of new laws and decrees.

 General Directorate for Legislative Development and Publication. This Directorate is in charge of reviewing quality of by-laws proposed by ministries and public agencies; developing and managing the Legal Information System (electronic inventory of stock of all primary and secondary regulations) and keeping it up-to-date by collecting, consolidating and identifying existing regulations; introducing simplification measures to existing regulation; and publishing the Official Gazette.

 Regulatory Reform Group. The Group is mainly concerned with the introduction of Regulatory Impact Assessment (RIA) in Turkey. It is currently in charge of the implementation of a training program of government officials, financed by the EU. It is a small group of experts with primary responsibilities in other departments. The duties and responsibilities of the Group have been defined by a memo which was signed by the Undersecretary, but in the current structure the Group is not attached to other formal institutions.

 Quality Legislation and RIA Group. This most recent regulatory oversight body was established under the Prime Ministry in March 2010. The Group is responsible for preparing draft legislations in line with recognized national and international standards as well as evaluating and guiding ministries in RIA processes. The institution will be chaired by Director General of General Directorate of Laws and Decrees, with two vice presidents from the Directorate of Laws and Decrees as well as the Diretory of Legslative Development and Publication.

5.19. Institutional fragmentation remains an important challenge to establish a single, comprehensive and coordinated regulatory policy. Compared to other OECD countries, the current structure puts Turkey in a weak position of institutional capacities for managing regulatory reform. The OECD Principles of Regulatory Quality and Performance suggest that countries should ―create effective and credible coordination mechanisms, foster coherence across major policy objectives, clarify responsibilities for assuring regulatory quality, and ensure capacity to respond to a changing, fast-paced environment‖. It is good practice for governments to support reform at all levels. This difficult task is increasingly important as regulatory responsibilities are shared among many levels of government, including supranational, international, national, and sub-national levels. High quality regulation at one level can be undermined by poor regulatory policies and practices at other levels, while, conversely, coordination can vastly expand the benefits of reform. The policies and mechanisms for coordination across, as well as between, levels of administration are thus becoming increasingly important for the development and maintenance of an effective regulatory framework.

5.20. Coordination mechanisms across the administration are fundamental to have a comprehensive regulatory reform strategy that applies to the whole administration and intends to achieve common objectives. In many OECD countries, coordination among ministries is ensured through clear mandates and allocation of responsibilities at high political level. In countries with more consensual traditions, coordination is ensured through soft mechanisms in the system. Several institutions

132

deal with regulatory reform in Turkey and they do not always act in a coordinated manner. Some of the programs aimed at improving the regulatory system overlap, such as some areas of e-government and revision of procedures for economic licenses. The different responsibilities allocated to institutions are not always linked towards a single regulatory reform strategy. This creates a challenge when it comes to establishing priorities and taking the lead for reform, and it often also leads to overlapping responsibilities that make implementation cumbersome, thus directly affecting business operation.

Figure 5-2: Institutional capacity for managing regulatory reform

18 16 14 12 10 8 6 4 2

0

2008 2008 2008 2008 2005 2005 2005 2005 2008 2008 2008 2008 2008 2005 2005 2005 2005 2008 2008 2005 2008 2005 2008 2005 2005 2005 2005 2005 2008 2005 2008 2005 2008 2008 2008 2008 2008 2005 2008 2005 2008 2005 2005 2005 2005 2005 2005 2008 2005 2008 2005 2008 2008 2008 2008 2008 2005 2008 2005 2008 2005 2005

LUX NOR TUR FIN NZL GRE HUN POL AUT CZE DNK FRA PRT DEU ESP SVK CAN CHE JPN AUS ICE IRL ITA SWE USA BEL KOR EU MEX NLD UK

5th-95th percentile Index 2005 Index 2008 Weights:

Is there a dedicated body (or bodies) responsible for promoting the regulatory policy and monitoring and reporting on if yes, weight=3 regulatory reform and regulatory quality in the national administration from a whole of government perspective? Is this body consulted as part of the process of developing new regulation? if yes, weight=2 Does this body report on progress made on reform by individual ministries? if yes, weight=2 Is this body entrusted with the authority of reviewing and monitoring regulatory impacts conducted in individual ministries? if yes, weight=2 Can this body conduct its own analysis of regulatory impacts? if yes, weight=1 Is this body entrusted with an advocacy function to promote regulatory quality and reform? if yes, weight=1 Is there an advisory body that receives references from Government to review broad areas of regulation, collecting the if yes, weight=1 views of private stakeholders? (e.g. past bodies have included the Better Regulation Task Force in the UK, the External Advisory Council on Smart Regulation in Canada and the Regulatory Reform Council in Korea) If the answer is “yes”: b(i) Does this body have a degree of independence from government (e.g. through a board or if yes, weight=1 commission structure)? If the answer is “yes”: b(ii) Does this body report its findings publicly? if yes, weight=1 Is a specific minister accountable for promoting government-wide progress on regulatory reform? if yes, weight=2 If the answer is “yes”: c(ii) Is the Minister required to report to Parliament on progress? if yes, weight=2

See Question 14 / 2008 OECD Regulatory Indicators Questionnaire

Note: The sample includes 31 jurisdictions for 2005 and 2008. Data for 1998 are not available for the European Union, Luxembourg, Poland and the Slovak Republic. Results have been adjusted for 26 countries following the results of a peer review process. Results for Spain are pending confirmation. Results for Belgium, Germany, Ireland, Luxembourg and Switzerland represent the initial answers to the surveys and will be adjusted within the next weeks. Source: OECD Regulatory Management Systems‟ Indicators Survey 1998, 2005 and 2008. www.oecd.org/regreform/indicators, in OECD (2009)

5.21. Turkey has gone through a decentralization process in which municipalities play a more active role in implementing regulation. The Ministry of Interior plays a role in dealing with these issues, but there is no official forum where different levels of government can discuss implementation issues. As a result, an important number of conflicts arise, in particular dealing with construction permits, settlement and land issues.

5.22. Conflict resolution mechanisms are costly and time consuming. Businesses can appeal a municipal decision to Court, but also file a formal complaint to prosecutors, governors or the Ministry of Interior. This Ministry is the main body responsible for mediating and solving conflicts that arise from these relationships. This substantially adds to the regulatory requirements – in terms of both time and cost – faced by businesses.

133

Box 5-3: Coordination between levels of government Co-ordination between different levels of government is a political priority for many OECD countries. The principle of subsidiarity reflects a real concern for clarity and calls for finding more appropriate co-ordination mechanisms that can help to avoid overlapping and duplication. The devolution process also speeds the need for co-ordination.

Most OECD countries dealing with a multi-level dimension have set up co-operation and co-ordination mechanisms and permanent institutional bodies to streamline the relationship between levels of government. Those mechanisms are either formal or informal, depending on the political and legal tradition and tend to have a more permanent structure, rather than an ad hoc basis. For instance, in Norway, several mechanisms are in place to ensure co- ordination of regulatory proposals affecting local governments. First, regular formal meetings are held between representatives from central and local government. At the political level a process of four consultative meetings per year (since 2000) brings together key ministries of the central government with high level representatives from the Norwegian Association of Local and Regional Authorities (Kommunenes Sentralforbund, KS). Similar meetings are held addressing issues pertaining specifically to county and municipality issues. Second – as part of the public consultation on draft laws and regulations – local government and local government organizations (KS) receive for comments those government draft regulations considered of special relevance for local governments. Third, continuous informal dialogue takes place between central and local government representatives at different levels, in many different forms, and on political as well as technical and professional issues.

In most countries, regulatory co-ordination has been promoted by associations and local authorities. This has provided a good basis for advice and better understanding of the needs and problems at different levels of government. But co-ordination has been improved mainly by special bodies and institutional mechanisms that serve lower levels of government to submit comments, to put forward specific measures and to communicate with the central level. Co-operation agreements have also improved co-ordination by establishing specific plans with clear frameworks for implementation and financing. In Denmark, for instance, as a result of a sustained process of decentralization, particularly since the fusion of local authorities in 1970, much government service delivery is carried out at lower levels of government. Regulatory policy remains concentrated at the national level, although there is significant consultation with local government as a result of its major role in implementation. From the perspective of local government, the key regulatory issue is that of increasing the freedom to act to be able to achieve efficiency gains needed to allow services to be delivered within tight fiscal restraints. To achieve this goal, the Government initiated a local government reform and a five-year work reform took place in 2007. Structural setting and relations between local and central government were redefined. According to the new system, there are new mechanisms and areas in which national and central governments co-operate and co-ordinate their service delivery.

Source: Rodrigo, D., L. Allio and P. Andres-Amo (2009)

5.2 Administrative capacities to prepare new regulations 5.23. Current processes for making legislation and subordinate regulations should support applications of core principles of good regulation. Good practice shows that systematic capacities to generate high quality regulation are essential and they should ensure that both processes and decisions are transparent to the public.

5.24. Clear guidelines are essential to establish systematic procedures to prepare and amend regulations. A skilled and well-trained civil service recruited on the basis of merit is also a prerequisite for developing and maintaining high-quality regulations and regulatory policies. Law-making in Turkey follows many high quality standards. The by-law on the ―Preparation Method and Principles of Legislation‖ (2005/9986) has established procedures and principles regarding drafting of laws, decree laws, regulations, by-laws, annexed decisions of Cabinet decrees and other regulatory proceedings to be prepared by the Prime Ministry, ministries and their affiliated, related and associated public institutions

134

and agencies. The by-law was enacted by the Decision of the Council of Ministers on December 2005. The main principles that apply in the preparation of drafts are the following:

- Drafts shall not be contrary to high norms of law; - Drafts shall be prepared in line with purposes of regulation; - Judicial decisions shall be taken into account in the preparation of drafts; - All of the legislation on regulated areas shall be reviewed during the preparation of drafts; - The scope of drafts shall be clearly regulated without leaving any room for misgivings; and - Articles of draft shall be concise and clear, there shall not be explanatory provisions in parenthesis.

5.25. Turkey has a good pool of legal drafters that work in line ministries, preparing drafts of laws and regulations. Good expertise can also be found in institutions dealing with the quality control of such drafts, such as the Ministry of Justice and the different institutions at the Prime Ministry‘s Office. This is, however, not the rule and in many cases there is a lack of professionals with other backgrounds, in particular economics, that could incorporate more evidence-base information to the preparation of laws and regulations with high potential economic impact for society and businesses. It is therefore not uncommon that some drafts relay on the work of consultants and are sometimes inspired by laws from other countries, which reveals the need to strengthen capacities in the country to prepare laws on complex issues.

5.26. Turkey has devoted particular attention to creating capacities inside the administration to introduce new tools for regulatory management. This is the case of the use of RIA and Standard Cost Model, which are essential to improve the quality of regulations affecting business operations by reducing the time and cost of doing business. The efforts have been limited so far, but this approach has showed the importance of building capacities and creating awareness at different levels: technical experts, decision-makers, politicians and the private sector.

5.27. Transparent and consistent processes for making and implementing laws and regulations are fundamental to ensure public confidence in the rulemaking process. Consultation is a systematic attempt to discover the opinions of groups affected by regulation and to obtain data useful in regulatory development and analysis. It may be general (e.g. advertisement for comment) or specifically targeted (e.g. focus groups, working parties), depending on the type and impacts of a regulation issue or proposal.

5.28. The Turkish Government has created some public-private mechanisms to deal with the business environment and reduce costs for doing business. These instances participate in consultation mechanisms about the way regulations affect them. The YOİKK and the IAC have proven to be very helpful in terms of identifying constraints for doing business in Turkey and increase private sector participation in the design of solutions to improve the investment climate in the country. They have set a basis for trust between the government and the business sector. But businesses have reported that not all of the Committees work at the same pace.

5.29. Public consultation is essential to improve the quality of new regulations. In terms of primary and secondary laws and regulations, Turkey has made it compulsory to consult draft proposals among government agencies. Any draft that comes to the Prime Ministry has to be submitted with an explanation about the consultation among government bodies. But consultation with external stakeholders is not yet compulsory and the degrees to which ministries consult vary across the administration.

5.30. Regulatory Impact Analysis (RIA) is a fundamental tool to assess the likely impacts of future laws and regulations. RIA provides decision-makers with valuable empirical data and a comprehensive framework in which they can assess their options and the consequences their decisions

135

may have. A poor understanding of the problems at hand or of the indirect effects of government action can undermine regulatory efforts and result in regulatory failures. RIA is used to define problems and to ensure that government action is justified and appropriate.

5.31. A by-law passed in February 2006 introduced the use of RIA in Turkey for all draft laws and statutory decrees that could have an impact above TL 10 million. The article dealing with the regulatory impact assessment of this by-law came into effect as of February 2007.101 Since then, RIA is required by all ministries before a draft law proposal is sent to the Prime Ministry. Following a Prime Ministry‘s circular in April 2007, RIA guidelines were prepared, including procedures and processes on how RIA should be conducted. The Better Regulation Group has been responsible for the RIA work in Turkey. Its main responsibilities are increasing capacity in the administration, particularly in line ministries; ensuring quality control of submitted RIAs; raising awareness of RIA issues across the administration; creating a RIA network in the administration; and maintaining a RIA Turkey website.

5.32. The Better Regulation Group obtained a grant from the EU to conduct a project called “Introducing Regulatory Impact Analysis into the Turkish Legal Framework,” mainly based on training and capacity building activities. The project has financed different RIA events (workshops, training, etc.) and has allowed the training of 370 government officials in the use of RIA. Within the framework of this project, two pilots have been conducted, on taxation for electricity with the Ministry of Finance and on rural incentives with the Ministry of Agriculture. The project has also set up a web site for RIA issues in Turkey102 and a RIA Network to facilitate communication and effective implementation among ministerial RIA units. The finalization of this project has allowed reviewing the current set up for RIA. Different recommendations have been made to ensure that RIA is a tool to support policy making. These recommendations are aligned with international good practice and take into consideration Turkey‘s current capacities of implementation. They cover a wide range of issues, from design and focus, improvement of guidelines, communication strategy, institutionalization of the process, data collection strategies, capacity building and improved coordination.

5.3 Administrative capacities to review existing regulations 5.33. Ensuring regulatory quality refers not only to new, but also to existing regulation. Keeping regulations up-to-date is important to ensure that the legal framework is relevant and effective for policy purposes. A number of different techniques and mechanisms exist to conduct periodic evaluations of whether existing regulations still constitute the best available solutions to the problems they seek to address.

5.34. Governments worldwide are adopting programs to reduce the administrative burdens associated with regulatory requirements. Regulatory paperwork and government formalities can be unnecessarily burdensome on regulated groups if coordination between regulators is lacking, new technologies are not used to assist in information gathering, and if unnecessary information is sought by regulators.

5.35. Turkey has made progress in simplifying certain procedures, regulations and laws. This has however mostly been on an ad-hoc basis and without having a comprehensive approach to administrative simplification, as it has been the case of many OECD countries (Box 5-4 and Figure 5-3). As part of the efforts undertaken by the Turkish government, there are a number of issues that can be highlighted:

101 Under Article 24 of the ―Regulations on the Procedures and Principles for Preparation of Legislation‖ published on 17/2/2006, it is provided that RIA shall be carried out for acts and statutory decrees, and other regulatory acts as deemed appropriate by the Prime Ministry, to be prepared as of 17/2/2006, with the exception of those on national security matters, as well as budget and final accounts acts and statutory decrees. 102 www.riaturkey.org

136

- Certain laws and regulations have been amended after some simplification processes; - Some new laws have been enacted to solve legal ―grey areas‖; - Several pilot projects have been undertaken in the use of tools for administrative simplification, such as the Standard Cost Model (SCM); - A new inventory of public services has been prepared for the entire administration; and - Establishment of one-stop shops to support businesses to deal with procedures.

Figure 5-3: Explicit program for reducing administrative burdens

6

5

4

3

2

1 1998 2005 2008 0 FIN HUN IRL SVK POL CZE LUX CHE NOR TUR ICE JPN PRT DNK NZL AUS ESP USA EU DEU SWE BEL GRE AUT ITA MEX CAN KOR FRA NLD UK

Weights: Is there an explicit government programme to reduce the administrative burdens imposed by if yes, weight=2 government on enterprises and/or citizens? If the answer is “yes”: Does this programme include quantitative targets? if yes, weight=1 Which of the following strategies are used? - Information and communication technologies for regulatory administration (e.g. electronic databases, if yes, weight=1 online formats) - Other streamlining of government administrative procedures if yes, weight=1 - Reallocating powers and responsibilities between government departments and/or between levels of if yes, weight=1 government

See Question 12 / 2008 OECD Regualtory Indicaotrs Questionnaire

Note: The sample includes 31 jurisdictions for 2005 and 2008. Data for 1998 are not available for the European Union, Luxembourg, Poland and the Slovak Republic. Results have been adjusted for 26 countries following the results of a peer review process. Results for Spain are pending confirmation. Results for Belgium, Germany, Ireland, Luxembourg and Switzerland represent the initial answers to the surveys and will be adjusted within the next weeks. Source: OECD Regulatory Management Systems‟ Indicators Survey 1998, 2005 and 2008. www.oecd.org/regreform/indicators in OECD (2009)

Box 5-4: Comprehensive administrative simplification efforts in OECD countries In many OECD countries, administrative simplification is becoming increasingly embedded within the overall regulatory quality systems of respective countries. In the past, administrative simplification was often undertaken on an ad hoc or sectoral basis. In most countries there is now more of a ―whole-of-government‖ approach to reducing burdens. Simplification is being increasingly embedded in the policy-making process. Simplification strategies focus on two dimensions: ex ante control of the burden introduced by new regulations (a flow concept) and the reform ex post of existing burdensome regulation (a stock concept).

Measurement has also become an important part of the burden reduction programs of many countries. The focus of the measurement exercise (and subsequent burden reduction programs) tends to be on business, often with special consideration for small and medium sized businesses, but there has also been a trend towards measuring and reducing the burdens imposed on others, including private citizens and the not-for-profit sector. The sophistication of the measurement techniques varies between countries, but the trend is clearly towards more sophisticated and accurate techniques that allow a very detailed examination of the source of administrative burdens.

In the UK, for instance, research and surveys have shown that businesses spend at least £1.4 billion each year on advice to help them comply with regulation. Businesses will pay for advice if they feel that this is cheaper or easier

137

than following regulations on their own. The government has decided to make real reductions in how much businesses need to spend on regulatory advice by tackling, volume and complexity, low awareness of government guidance, regulatory change, poor quality government guidance as well as uncertainty, risk and lack of confidence. As a result, the current strategy applied in the UK relies on the following actions:

- Improving the regulatory process: Plan guidance at an early stage of the policy process and issue guidance earlier - Improving communication on regulation: Increase the market penetration of businesslink.gov.uk; communicate directly with businesses using high-quality, simple guidance and communicate with businesses through intermediaries - Improving the quality of government advice on regulation: Improve feedback mechanisms on guidance and consider joint-badging or outsourcing the design of guidance - Improving the environment for business advice on regulation: Help businesses become informed consumers of advice services by increasing understanding of regulatory requirements; take advantage of online forums for businesses to share information on regulations and provide dedicated guidance for advisors where appropriate.

Source: www.berr.gov.uk/whatwedo/bre/reviewing-regulation/reducing-cost-business/page44090.html

5.36. In many emerging economies, streamlining some regulatory instruments has been an initial point for broader regulatory reform efforts. By targeting particular regulations, governments ensure that the political support is concentrated in few, but sometimes very pervasive regulations that might cause a great impact in economic terms by imposing excessive costs to businesses. This is the case of licenses and construction permits. In terms of regulatory instruments, licenses and construction permits have been identified as two important areas for further improvement by YOİKK. Both regulatory instruments represent the complexities of the regulatory problems in Turkey and have the potential to increase a positive direct impact on economic activity if they are properly streamlined.

5.37. Licenses are one of the main areas of complaint by Turkish businesses. Some initial improvements have been made by the Government, such as having a single Business Opening License, but there has not been a comprehensive approach to streamline, simplify or eliminate licenses. There is currently no single institution responsible for licenses, which are granted by Ministries, regulatory agencies and municipalities. The State Planning Organization, which is responsible for the YOİKK Technical Committee on Sectoral Licenses, conducts some work on the licenses delivered by Ministries. In addition, the Ministry of Public Works and Settlement is also involved in the work related to construction permits, as the Chair of the YOİKK Technical Committee of Investment Locations. However, no single comprehensive strategy for licensing reform has been undertaken in Turkey and it is not clear how much licenses cost to businesses and government.

Box 5-5: Benefits of licensing reform In many emerging and developing countries, such as Korea, Mexico, Croatia, Ukraine, Moldova, Hungary, etc., licensing reform has been a trigger for further reforms because a well designed licensing program can create momentum and political appetite for other regulatory reforms. Licensing reform is also pertinent as a starting point for a broader program of regulatory reform, since licenses cut across policy areas and mobilize most government agencies dealing with economic activities.

Business licensing is a commonly used form of regulation which affects specified businesses and occupations by regulating entry into markets and conduct within markets. Licenses typically impose on businesses a range of conditions, obligations and rights — often in the form of a specific license, permit or concession. Licensing can be distinguished from other types of regulatory requirements by usually obliging the regulated parties to obtain a certification of compliance with regulatory obligations prior to the commencement of a given business activity. From an economic and business point of view, licensing is a potentially much more costly and potentially damaging

138

regulatory intervention when compared to other types of regulation (such as broadly applied competition law and accounting rules, or other ‗lighter‘ forms of regulation etc). This is because licensing requirements not only impose regulatory compliance burdens (as do most types of regulation), but also can restrict healthy competition by establishing significant and unnecessary entry barriers to particular economic activity and markets. These restrictions can include: - grant exclusive rights for a supplier to provide a good or service; - affect the ability of some types of firms to participate in public procurement; - significantly alter costs of entry or exit to a market; and/or - create a geographic barrier to the ability of businesses to supply goods or services, invest or supply labor.

Licensing can restrict or reduce the ability of businesses to compete and innovate through: - control or substantially influence the price at which a good or service is sold; - altering the ability of suppliers to advertise or market their products; - setting prescriptive standards for product/service quality that are significantly different from current practice; and/or - significantly altering costs of some suppliers relative to others.

Licensing reform advocates a comprehensive, top-down, and institutionalized approach. Through this approach, reforms are driven forward by an explicit political mandate, if possible also by a quantitative target for the reform, and by strong incentives for regulating agencies to participate constructively in the review process. Licensing reform can only be sustained over time if other tools for regulatory management are gradually integrated in the system and if certain institutions and capacities are created inside the administration.

Source: World Bank Group (2009)

5.38. According to the Doing Business Indicators, Turkey ranks 133 among 183 countries in obtaining construction permits. As mentioned in Chapter 2, DB shows that the site development time in 2009 amounts to 188 days, a figure that is not equally negative when analyzing the corresponding data in the Enterprise Survey. Average time spent dealing with construction-related permits was in fact lower than in most comparator countries. Instead, there seem to be significant variations among regions and cities, indicating that the system of permit processing has not been sufficiently streamlined across Turkey. Several areas can be identified as problematic in this field:

- There are major gaps on sector based strategies and on urban development plans from several Ministries. Coordination mechanisms to prepare strategies and plans are weak. - Traditional methods to deal with permits are slow. There are not well established digitalized systems that can ensure faster processing of applications from businesses and citizens. - Similar authorities conflict on areas of action. Sometimes two agencies deal with the same issue and there is no planning authority that can help solving the differences. - The approval of investors‟ plans falls into the responsibility of different agencies, which might have different solutions for different plans. - Inspections related to construction permits are in the hands of the Ministry of Interior, which does not have a specialized group of experts on planning and risk related issues. The Ministry of Interior inspects only from the administrative side. - Municipalities are responsible for delivering some permits and licenses but they do not have adequate capacity to understand and apply secondary legislation. - There is no standardization of procedures and steps in Ministries involved in construction permits. There are no clear guidelines and handbooks that can help to implement the regulatory framework.

Box 5-6: Dealing with construction permits A building permit grants legal permission to start construction of a building project. It is usually granted by a

139

municipality or a specialized local authority. The main objective of building permits is to ensure the health and safety of the community. This has important implications for policy-makers who need to strike the right balance between the cost imposed on industry (including the checks imposed through the building permitting process), and the real benefits in safety and health standards.

The building permit process also plays a critical ―gate-keeper‖ role in protecting a range of other public goods such as preventing construction close to airports, and protecting the environment or preventing potentially harmful industries to locate in residential areas. These public goods are more fragmented, and not related to structural properties or the risks directly associated to the structural characteristics of the building. When this ―gate-keeper‖ function in not carefully managed and coordinated with the relevant authorities, an insurmountable bureaucracy may emerge, which is likely to discourage investment, and increase the level of informality. Carefully managing the ―gate-keeper‖ role is an important factor in the success of building permit reforms.

Reformist countries are gradually adding new policy objectives in their reform efforts, such as energy conservation and environmental sustainability. These are important public goods, but should not translate into a more complex process for investors. Best practice reform experience shows that new policy objectives, including those going beyond the improvement of public safety, can be combined with effective red tape reduction programs and more efficient and streamlined processes. In fact, building permit reforms, as observed in good practice countries, have all generated positive impacts on processes, although streamlining procedures might not have been the original or main focus.

New Zealand and Canada have both reformed their construction building systems. The objectives of the reforms in both countries were primarily driven by the need to increase safety, improve the standards of building practitioners, and turn local authorities into efficient enforcement bodies. In pursuing this strategy, both countries have considerably streamlined their processes, and made it easier for investors and developers to go through the permitting process. Both countries have introduced a common set of reforms, i.e., pro-actively engaging private building practitioners in the permitting process and inspections, introducing risk management, supporting innovative one-stop-shops, consolidating pre-approval requirements, and improving appeal mechanisms for developers and investors, etc. To address the dysfunctional relationship between central authorities and local permitting authorities while still managing high standards of enforcement, New Zealand engaged in one of the most original and daring reform in establishing a compulsory accreditation process for local permitting authorities. Both reform efforts are starting to pay off with lower rates of accidents, and a faster turnaround of building permit processes. For example, there has been a 40% reduction of accidents within the regulated industries in building sites in Ontario, and fires have declined by 15% since the introduction and enforcement of the new law.

Source: Moullier (2009)

5.39. Both licenses and construction permits are potential regulatory instruments for comprehensive review. They could also serve as platform for developing a focused, targeted strategy on simplification with clear baselines and reduction targets. For this purpose, the use of Standard Cost Model (SCM) seems an interesting regulatory tool that could be employed in Turkey. The SCM is a methodology used in many developed and, increasingly, emerging countries to measure the administrative costs – both monetary and in terms of time – required for compliance with regulations. It is a key tool for broad administrative simplification efforts since it leads to the establishment of clear targets for the reduction of administrative burdens. Results in the use of the SCM have shown that reduction in administrative costs has an important impact on GDP. In the European Union, for instance, achieving the objective of reducing administrative burdens by 25 percent could lead to an increase in the level of EU GDP of approximately 1.4 percent or € 150 billion in the medium term.103

5.40. Turkey has started the use of SCM with its application in two pilot projects on registration of vehicles and on establishing businesses. The use of SCM has been introduced primarily by the Department of Administrative Development of the Prime Ministry through the MATRA project. Some

103 European Commission (2007)

140

training and awareness activities have been conducted to inform the political and technical level of different ministries about the use of SCM. In addition, in order to measure cost of sectoral licenses which is one of the YOİKK agenda, a pilot project called ―Measurement of Administrative Burdens in Export Procedures‖ has been conducted in cooperation with TOBB. Business associations have also participated in dissemination activities. Some constraints faced in Turkey to introduce the use of SCM are the financial resources and technical skills needed for its implementation.

Box 5-7: The use of SCM: International experiences Complying with regulations usually involves costs for businesses, which can be divided into various different categories: - Financial costs are the result of a concrete and direct obligation to transfer a sum of money to the government or the competent authority. - Compliance costs are all costs to businesses of complying with regulation, with the exception of the financial costs. Compliance costs can be divided into ‗substantive compliance costs‘ and ‗administrative burdens‘. - Substantive compliance costs are the costs that businesses make in order to comply with the content obligation that legislation and regulations require of a production process or a product. - Administrative costs are the costs imposed on businesses, when complying with information obligations stemming from regulation.

The Standard Cost Model (SCM) is a method for measuring the administrative burdens imposed on businesses through legislation, regulations and other requirements. The SCM has been developed to provide a simplified, consistent method for estimating the administrative costs imposed on business by central government. It takes a pragmatic approach to measurement and provides estimates that are consistent across policy areas. The starting point of SCM analysis is the identification of ―information obligations‖ that businesses are required to provide to the government and other bodies. The SCM can measure information obligations arising from different sources such as all existing laws and regulations; a specific field of laws and regulations (like fiscal rules, the transport sector, starting a business, employment procedures); or requirements imposed by a selected government body.

Since the 1990s, SCM has been developed and modified for use in OECD countries. The Czech Republic was measuring the baseline of overall administrative burden. The measurement was undertaken between March 2005 and September 2005. The baseline measurement included a measurement of all business related generally binding regulations in 12 ministries and 10 central administrative authorities. The results of the measurement were sent to the Department of Regulatory Reform and Central State Administration Reform by the end of September 2005. Subsequently, the Department carried out the analytical report of overall administrative burdens on businesses entitled ―Analysis of the administrative burdens on businesses‖, elaborated in February 2006 on the basis of the information collected by ministries and other central state administration authorities. As part of the recommendations of the report, a number of measures were adopted by the Government Resolution No. 759 of 11th July 2007: i) Introduction of the obligation to assess administrative burdens ex ante in case of new legislation drafts; and ii) Preparation of overviews enumerating the concrete legislation proposed for the purpose of the reduction. In its latest Resolution, No. 446 of 21 April 2008, the Czech government confirmed the reduction target set in 2005; e.g. all the legislative amendments should be accomplished by 2010 with the aim to reduce the burden on businesses by 20 percent.

Source: www.administrative-burdens.com

5.41. ICT solutions make a valuable contribution to improve the regulatory environment of a country. Many international experiences show that the use of ICT in relation to transactions within and between government bodies and between government bodies and business and citizens, is a key enabler of administrative simplification, a fundamental element of regulatory reform.

5.42. Important steps have been taken to integrate the use of ICT to improve services delivery both to citizens and businesses and to create a basis for an information society. The State Planning

141

Organization is responsible for setting the policies and coordination of implementation of e-government in particular and information society in general under the umbrella project named e-Transformation Turkey. Lately, an ad-hoc e-Government Group at the Prime Ministry was tasked to tackle with the legislative and administrative barriers for selected eleven e-government projects. e-Government activities had gained momentum after Turkey became a party to the eEurope+ Initiative, which was designed for EU candidate countries in 2001. The State Planning Organization has been leading the e-government work since 2003. The eTransformation Turkey Project was launched in 2003 aiming to carry out the process of transformation into an information society. Since the inception of this Project, two action plans have been launched and implemented successfully. An Information Society Strategy covering 2006-2010 was prepared under the coordination of State Planning Organization to enable Turkey to benefit from ICT effectively and to identify the middle and long term strategies and targets for the realization of the ambitious transformation. The current Information Society Strategy establishes targets and implementation measures to transform Turkey into an information society.

5.43. The e-Government Group at the Prime Ministry is an ad-hoc group established to deal with eleven on-going e-government projects. In addition to this task, the e-Group has prepared a draft of the ―E-Government and Information Society Act,‖ covering two main aspects: organizational (creation of an Information Society Act and institutional architecture for e-government and information society in Turkey) and framework articles (project management, performance evaluation, data sharing and ownership, common e-government services, authentication and authorization, e-government portal, e- archive, use and protection of personal data, liability and secrecy, etc).

5.44. Another dimension of transparency is the need for the government to effectively communicate the existence and content of all regulations to the public. This means that the regulations are available to the public at reasonable cost, in a language that can be easily understood. Communication is also essential to achieving effective compliance. The by-law on the ―Preparation Method and Principles of Legislation‖ (2005/9986) encourages legal drafters to use plain language in the preparation of draft proposals. In terms of communication, Turkey has a Regulatory Information System104 which contains all published regulations, e.g. laws, decrees and other secondary legislation. All citizens and businesses have right to access to it freely. Amendments are consolidated daily and republished the principal act as a single text. There is no legal basis for courts and public agencies to accept it as ―prima facie evidence of the law‖. All regulatory instruments are published in the Official Gazette which is available both on-line105 and in hard copy.

5.45. The quality of existing regulation is as important as the assessment of new regulations to ensure a high quality regulatory structure. In many cases and as technology, the economy, and society change, existing regulations often become less relevant and effective. It is therefore essential to maintain a periodic reevaluation of whether existing regulations still constitute the best available solutions to the problems they seek to address. A systematic approach is required to ensure that all regulations are regularly subjected to this reassessment and several techniques exist for this purpose. In Turkey, the first review of existing regulation took place in 1986 and included the codification of all laws, decrees having force of law, regulations (Tüzük), and by-laws issued by the Council of Ministers. A second effort was undertaken in 2005, when the Government identified that 13,967 by-laws were issued between 1970 and 2005 and all were reviewed. This led to a reduction to 4,510 by-laws. The most recent effort is currently under way and comprises the review of 4,795 by-laws, in which regulatory burdens imposed to society and businesses have been identified. The review has led the following results:

104 http://www.mevzuat.gov.tr 105 http://rega.basbakanlik.gov.tr

142

- Preparation of drafts to amend 155 by-laws; - Removal of around 400 document and information obligations; - Replacement of 38,000,000 obligations for submitting the copy of the ID card by declaration of the ID number; and - Delegation of authority to lower levels in 25 procedures.

5.46. Enforcement and compliance with regulations are two areas that remain relatively weak. Municipalities play a role in implementing regulations and most of the inspections system remains an issue at the central level. The impact of this institutional set up is significant in terms of the burden that is imposed on citizens and businesses, which sometimes have to comply with procedures established by the local levels without prior agreement at the central one. This is also a consequence of limited coordination when preparing laws and regulations and of the duplication of responsibilities among ministries in specific areas of action.

5.47. The role of the judiciary is essential for regulatory quality control and better economic performance. The effectiveness of the process arises from the ability of the judiciary to consider regulations‟ consistency with principles of constitutionality, including notably proportionality and the right to be heard. It also arises from courts‘ scrutiny of whether delegated legislation is fully consistent with primary legislation. Administrative justice and judicial review are important elements for compliance and enforcement of regulation. In Turkey, both elements remain a challenging area for further improvement that imposes costs and delays to economic activity.

5.48. Public redress can be made directly to the administrative authority, which has officially 60 days to respond to the interested party. Given the little responsiveness in the system, most people appeal decisions to Court. This procedure is not expensive, which makes justice accessible. However, this creates a workload that is difficult to manage for authorities involved. In addition, the lack of specialized courts to deal with particular ―cases‖ brought by investors reduces the effectiveness of the system. In case the Ministry of Interior has to intervene with local authorities, delays are to be expected since the Ministry is not equipped with the adequate resources.

5.49. The Ministry of Justice is actively participating in a number of proposals to improve judicial review. There are current efforts to introduce new conciliation mechanisms, expanding on the successful experience with tax system reform, and to streamline administrative and judicial review. There is a current draft of a new General Administrative Procedure Law which will contribute to improve the situation at the intermediary level. There is also an initiative to create an Ombudsman mechanism under the public authority to introduce a new conflict resolution system. There are also proposals to increase the number of courts, judges and prosecutors.

5.4 Policy Options for Regulatory Reform 5.50. Regulatory reform has the potential to be a high political priority, in particular, linking regulatory management and reform to the current efforts of improving the investment climate. This task, however, requires strong leadership and political back up, since it requires consolidating the current efforts and establishing good co-ordination mechanisms between different institutions. A comprehensive strategy for regulatory reform would make the Turkish economy more competitive and dynamic by reducing costs on businesses, eliminating incertitude and unnecessary obligations, and providing better services for citizens.

5.51. Turkey has adopted many good international practices in the way regulations are prepared and implemented but still much remains to be accomplished. In particular more systemic changes that can establish more transparent procedures, consistent consultation mechanisms, more evidence-based

143

decision making, reduced discretionary decisions and comprehensive administrative simplification efforts could be considered. This could make Turkey a more attractive destination for investment and could encourage better levels of compliance and enforcement, which remain low compared to international standards. A number of actions could help Turkey to improve its regulatory management system and strategy.

5.52. Proposal: Revise current efforts for regulatory reform to set clear priorities, objectives and targets within a comprehensive strategy. Current efforts on regulatory reform could be combined in order to establish clear priorities that can translate in improvements of the investment climate and business environment. The existing work done by YOİKK Technical Committees could be a platform to identify priority areas and start implementation efforts. In the medium term, it could be helpful to establish a single dialogue mechanism between the Prime Ministry and YOİKK, such as regular high level meetings, to set priorities. A decision on the sequencing of regulatory reform would also help to better allocate resources, build capacities inside the administration and raise awareness about regulatory reform in the public. It would be advisable to devise a strategy mapping the current efforts and priorities to define how they could be integrated in broader regulatory reform objectives. Such a strategic document could include not only principles of quality regulation to prepare new and amend existing regulation, but also a clear indication to link this area to reform of the business environment and to improvement of service delivery for citizens. A national regulatory policy could then follow a ―whole of government‖ approach that is applied by all government agencies and all levels of government. Such a document would help (i) allocate clear responsibilities to different institutions in order to facilitate coordination and cooperation between government agencies; (ii) establish clear objectives and targets that can be measured over time to show progress; (iii) increase transparency and accountability in the way regulations are prepared and implemented.

5.53. Proposal: Strengthen the institutionalization of regulatory reform by creating a single oversight body for regulatory reform. Turkey has a number of institutions that deal with regulatory reform from different perspectives. Most of the traditional roles that are conducted by oversight bodies for regulatory reform in other OECD countries remain in hands of various institutions. At least three institutions in the Prime Ministry deal with the quality of laws and regulations and some others are in charge of regulatory reform issues, apart from other bodies with regulatory responsibilities across the administration. Turkey could consider the possibility to bring together the existing expertise and entrust a single unit with clear responsibilities for regulatory reform. The Prime Ministry could establish an oversight body responsible for moving forward the better regulation agenda in the country, which could incorporate the different roles and responsibilities spread across the administration, in order to implement a single strategy for regulatory management and reform. The benefits of having a single oversight body for regulatory reform have been documented by international experiences. Strong oversight bodies, such as in the case of UK, USA, Canada, Mexico and Korea, have been key actors in the process of regulatory reform, working as ―engines of reform‖, maintaining a whole-of-government approach and coordinating inside the administration. The main role of oversight bodies is to ensure regulatory quality. This is done through supervision, control and coordination of the regulatory program and system. Regulators are then forced to demonstrate and justify the relevance of their regulatory actions (newly proposed and existing), using accountability and assessment mechanisms.

5.54. Proposal: Strengthen YOİKK‟s role to improve the business environment and advocate for regulatory reform in Turkey. YOİKK role in identifying constraints to investment climate issues in Turkey has been very helpful. This platform has established a clear mechanism for dialogue between the public and the private sector which could be strengthened over time. In particular, YOİKK could play a significant role in moving forward the regulatory reform agenda, by linking the investment climate constraints to the systemic challenges of improving the quality of regulation. YOİKK has the potential to participate in the definition of priority areas for regulatory reform, given the existing work undertaken and

144

the different action plans already in place for policy action. The existing structure that YOİKK has built in Turkey constitutes a solid basis for further coordination and consultation among the private and the public sectors. It also provides a forum for identification of common challenges for different ministries. This could be used to encourage the needed coordination that Turkey requires to implement regulatory reform. YOİKK could also be a strong advocate for regulatory reform. In many OECD countries, advocacy and/or advisory bodies have played a key role in shaping the regulatory reform agenda. In Japan, for instance, the Council for the Promotion of Regulatory Reform (CPRR), composed of members from the business sector and civil society, was established in 1994 to provide input to a 3-year action plan for regulatory reform. In the UK, the Better Regulation Commission was a fundamental actor for regulatory reform during its ten year existence (1997-2007). The Commission provided independent advice and challenges to the UK government on its management and delivery of better regulation, as well as independent scrutiny of departments' plans for regulatory simplification. With its current structure, YOİKK could play some of these roles by setting up priorities for government action and making recommendations about regulatory reform with a clear involvement of the private sector. In the medium and long term, YOİKK could be a platform with an advocacy role for regulatory reform in Turkey.

5.55. Proposal: Design a comprehensive administrative simplification strategy with clear objectives, targets and review criteria for lower level of regulation to improve the business environment. A particular area with great potential that has not been fully explored in Turkey is the design and implementation of a comprehensive administrative simplification strategy for lower level of regulation. Such a strategy could be relevant for Turkey for several reasons. First, it could establish a clear link between the importance of improving the quality of regulation and the improvement of the business environment, which is in fact one of the central purposes of regulatory reform. By looking at lower level of regulation, such as licenses or administrative decrees, the Government could produce important economic gains that would translate into a better economic performance. Second, such a strategy could also be the impulse that Turkey needs to fully embark in the implementation of a better regulation agenda. Third, such a strategy would be an excellent opportunity to strengthen public and private dialogue by establishing clear reduction targets of administrative burdens. This could strengthen existing consultation mechanisms, encouraging the private sector to play a leading role in the implementation of the strategy. This would also facilitate setting up monitoring and evaluation mechanisms to measure progress. Fourth, this strategy could make it possible to sequence regulatory reform in a more strategic way. For instance, it would facilitate the implementation of a number of regulatory tools that have been tested in pilot projects in Turkey, such as the use of Standard Cost Model and Regulatory Impact Analysis, which require time and capacity building to be embedded across the administration. Building on the work carried out by YOİKK and in particular by the Technical Committee on Licenses, an action plan could consider the following steps:

- Identify a political and technical platform (Committee, unit and/or oversight body for regulatory reform) to lead the administrative simplification and reform program, supported at the highest political level - Prepare and/or review an inventory of selected lower regulation with input from all authorities involved - Based on the inventory, select a number of lower level regulation subject to measurement with the Standard Cost Model - Use working groups to accompany the process and validate data - Based on measurement results and impact on the economy, select lower level regulations to be reviewed in a first review process (―quick wins‖) - Establish clear criteria for review and discuss provisions of lower level regulation in working groups with public and private participants - Working groups make recommendations to simplify, streamline or eliminate procedures to Committee at political level

145

- Committee revises and makes recommendations to be taken at high political level - Working groups work on the rest of licenses - After recommendations taken, list of ―clean‖ lower level regulation is put on an e-registry for the particular lower level regulation - Once the stock has been ―cleaned‖, Regulatory Impact Analysis is used to review the new regulations

5.56. Proposal: Improve coordination mechanisms inside the administration when preparing laws and regulations. Since a number of institutions are responsible for regulatory issues and there is no high level authority imposing a single strategy, it is difficult to ensure coordination among various institutions. Several countries have improved coordination at technical level by nominating ―regulatory reform experts or champions‖ in each ministry or regulatory agency. The creation of networks of experts dealing with regulatory reform facilitates dialogue among institutions and ensures that regular meetings are the platform for discussions and sharing experiences. Turkey could design a network of experts on regulatory reform to build capacities, discuss priorities and communicate in a strategic way what could be implemented. In the medium and long term, the establishment of an oversight body for regulatory reform, with a clear mandate about its responsibilities, could also help improve coordination among institutions. Turkey has already relevant institutions with skills and expertise in improving the quality of regulation, there would be no need to create new institutions, but to reengineer the existing mechanisms.

5.57. Proposal: Strengthen coordination and cooperation among levels of government. Being provinces and municipalities the main levels responsible for implementing regulations, it is essential to establish formal mechanisms to discuss the way their implementation and enforcement can be improved. In particular, provinces and municipalities could be systematically consulted in the preparation of laws that can have a direct impact in their roles. Turkey could reinforce the current consultation mechanisms with provinces and municipalities and strengthen the participation of lower levels of government in the preparation and design of laws and regulations. Consultation could be undertaken in a formal and systematic way, using approaches already proved in other countries, such as permanent roundtables or discussion meetings. The Government could also encourage the development of capacities at local level to improve implementation. Lower levels of government could be encouraged to respect the obligations imposed on businesses and citizens when delivering permits or licenses. The central government could also establish mechanisms to supervise accurate implementation of national directives. For instance, a revision of the current inspections system could be introduced, in order to improve enforcement and compliance. Inspection services, which are currently cumbersome and lack sufficient skilled staff, could be implemented based on assessed risk and the likelihood of violations, increasing transparency in the process and accountability of the concerned agencies.

5.58. Proposal: Make consultation with stakeholders compulsory for the preparation of new and amended laws and regulations. Consultation in the preparation of laws and regulations has improved over time in Turkey since consultation among institutions was made compulsory. Nonetheless, many gains could be obtained from making it compulsory for government to consult with external stakeholders. In particular, consultation would help obtain a better understanding of the way regulations affect citizens and businesses. This would require the Government to establish clear guidelines for consultation, clarifying the steps to follow, such as deadlines for comments, and ensuring that regulators are accountable for the comments received. Training regulators on consultation techniques could also be encouraged, as well as making use of the data gathered during the process.

5.59. Proposal: Continue implementation of Regulatory Impact Analysis (RIA). Turkey has made important steps in the introduction of Regulatory Impact Analysis (RIA). The work done so far by the Regulatory Reform Group has raised awareness of the importance to make wiser decisions based on evidence. The training program on RIA has proved to be useful in developing skills inside the

146

administration, but it is certainly only an initial step to a deeper use of this tool. The Government could ensure that the work done by the Regulatory Reform Group is continued and supported at high political level. The expertise gained in the introduction of RIA could be transferred to a future oversight body in charge of regulatory reform, which could be entrusted with the responsibility to challenge those RIAs that do not comply with the quality criteria established by the government. This function is essential to improve the quality of the Turkish regulation in the future. Capacity building in RIA would also be necessary with training programs linked to the use of other tools, such as consultation techniques and the use of Standard Cost Model. It would be important to revise the current guidelines to update them with the most recent lessons learned in the use of RIA in Turkey.

5.60. Proposal: Use existing e-government strategies to support regulatory reform and simplification efforts. Turkey has made improvements in the application of e-government strategies to deliver better services for its citizens. One of the main goals is to reach an automatization of procedures that could facilitate entrepreneurship. The Government is committed to ensure that every document is available to citizens in soft copy and transactions can be done electronically. This is in line with good international practice, but it requires adequate resources, skills and ensuring that all society takes advantage of the use of ICT mechanisms. It is essential for Turkey to link these efforts to a broader regulatory strategy. There are already good examples in Turkey in the way ICT can be used to improve the quality of the regulation. The Regulatory Information System is an initial step that could be further developed. In the same way, a broad administrative simplification effort could rely in the use of ICT tools to be implemented.

References Botero, Juan, Simeon Djankov, Rafael La Porta, Florencio Lopez-de-Salinas, and Andrei Shleifer (2004), ―The Regulation of Labor,‖ The Quarterly Journal of Economics

Department of Administrative Development. 2009. Standard Cost Model and Administrative Simplification, Ankara, September

European Commission. 2007. Action Programme for Reducing Administrative Burdens in the European Union, COM(2007) 23 final, Brussels

Government of Australia. 2006. Rethinking Regulation, Taskforce on Reducing Regulatory Burdens on Business, Canberra

Kaufmann, D., Kraay, A. and Zoido-Lobatón, P. 2002. ‗Governance Matters II: Updated indicators for 2000/01‘ Policy Research Working Paper, World Bank, Washington

Kox, Henk. 2005. Intra-EU differences in regulation caused administrative burden for Companies. CPB Memorandum Number, No. 136. Rev.1

Nijland, Jeroen. 2008. The Dutch Approach, EIPASCOPE, No. 2

Moullier, Thomas. 2009. Reforming Building Permits: Why It Is Important and What Can IFC Really Do?, IFC Advisory Services, Cairo

OECD. 2001. Regulatory Reform in Turkey, Paris

OECD. 2005a. OECD Principles for Regulatory Quality and Performance, Paris

147

OECD. 2005b. ―The benefits of liberalizing product markets and reducing barriers to international trade and investment: the case of the United States and the European Union‖, Economics Department Working Paper, No. 432, OECD, Paris

OECD. 2009. Indicators of Regulatory Management Systems. Update and Preliminary Results, Working Party for Regulatory Management and Reform, GOV/PGC/REG(2009)2, Paris

Prime Ministry. 2007. Circular Letter 2007/6. Regulatory Impact Analysis Guide, Directorate General of Personnel and Principles, Ankara

Rodrigo, D., L. Allio and P. Andres-Amo. 2009. "Multi-Level Regulatory Governance: Policies, Institutions and Tools for Regulatory Quality and Policy Coherence", OECD Working Papers on Public Governance, No. 13, Paris

State Planning Organization. 2006. Information Society Strategy, Ankara

World Bank. 2007. Turkey: Investment Climate Assessment, Washington

World Bank. 2008. Regulatory Transformation in the Republic of Korea, Case Studies on Reform Implementation, FIAS, Washington

World Bank Group. 2009. Policy Framework Paper on Business Licensing Reform and Simplification, Investment Climate Advisory Services, Washington

YOİKK. 2009. Activities of the Coordination Council for the Improvement of the Investment Environment and Technical Committees Action Plans, Ankara

Annex 5-1. OECD Principles of Regulatory Quality and Performance The 2005 OECD Guiding Principles for Regulatory Quality and Performance capture the dynamic and on-going whole-of-government approach to implementation of regulatory quality. These principles state that governments should: - Adopt at the political level broad programs of regulatory reform that establish clear objectives and frameworks for implementation. - Assess impacts and review regulations systematically to ensure that they meet their intended objectives efficiently and effectively in a changing and complex economic and social environment. - Ensure that regulations, regulatory institutions charged with implementation, and regulatory processes are transparent and non-discriminatory - Review and strengthen where necessary the scope, effectiveness and enforcement of competition policy. - Design economic regulations in all sectors to stimulate competition and efficiency, and eliminate them except where clear evidence demonstrates that they are the best way to serve broad public interests. - Eliminate unnecessary regulatory barriers to trade and investment through continued liberalization and enhance the consideration and better integration of market openness throughout the regulatory process, thus strengthening economic efficiency and competitiveness. - Identify important linkages with other policy objectives and develop policies to achieve those objectives in ways that support reform.

Source: OECD (2005a)

148

Annex 5-2. The Sources of Law in Turkey The Turkish legal system is composed of the following sources of law:

Constitution. The Constitution is the highest legal rule which binds all state organs and individuals. The provisions of the Constitution are fundamental legal rules binding upon legislative, executive and judicial organs, and administrative authorities and other institutions and individuals

Statutes. Statutes lay down principles, but leave to the judge the problem of interpreting and applying these principles to concrete facts. Statutes written down in a systematic fashion to regulate specific areas of law are given the title of "Codes". Statutes enacted by the Parliament must comply with the Constitution.

International Agreements. International law is also a direct source of Turkish legal system. In line with art. 90 of the Turkish constitution, approval of international agreements signed with foreign countries and international organizations are valid hence they are approved with a law enacted by the National Assembly.

Decrees having force of statutes. According to art. 91 of the Turkish constitution, the National Assembly can authorize the cabinet for enacting a decree law. This authorization determines the aim, scope, principles, using time of this authority and if more than one decree law can be enacted. Decree laws are in force when they are published in official gazette, but another date can be shown in decree law as an enforcement date. Decree laws are submitted to the National Assembly on the same day they are published in the official gazette.

Regulation (Tüzük). In accordance with art. 115 of the Turkish constitution, the cabinet can make regulations to specify the matters in which it is obliged by laws. Regulations should not be contrary to laws and they should be examined by the Council of State

By-Law. In accordance with art. 124 of the Turkish constitution, ministries and public administrations prepare by- laws to implement laws and regulations related to their sphere of duties. The Cabinet can make by-laws on the basis of general regulation authority.

Administrative regulation. The Cabinet and other administrative institutions can issue administrative regulation in accordance with the upper regulations such as laws, regulations (tüzüks), etc. These instruments can be called decree, circular, principles, command, etc. Cabinet decree is a common form used by authorities, in particular when the laws gives the authority to regulate certain area. Even if there is not a specification about its authority to make legal arrangements in laws, cabinet can make regulations to implement the laws based on its general regulation authority. The subject of cabinet decisions can be general regulations as a matter of public concern like tax rates or changing the tax amounts, or can be personal proceedings like appointment decisions.

Source: General Directorate for Laws and Decrees

149