GOV/PGC(2009)2 For Official Use English - Or. English For Official Use GOV/PGC(2009)2 Organisation de Coopération et de Développement Économiques Organisation for Economic Co-operation and Development 15-Apr-2009 English - Or. English PUBLIC GOVERNANCE AND TERRITORIAL DEVELOPMENT DIRECTORATE PUBLIC GOVERNANCE COMMITTEE GOVERNMENT AT A GLANCE For more information please contact: Mrs. Zsuzsanna Lonti +(33-1) 45 24 16 39/Zsuzsanna.Lonti@oecd.org or Mrs. Jordan Holt +(33-1) 45 24 78 91/Jordan.Holt@oecd.org JT03263039 Document complet disponible sur OLIS dans son format d'origine Complete document available on OLIS in its original format GOV/PGC(2009)2 TABLE OF CONTENTS FOREWORD ................................................................................................................................................. 4 I. INTRODUCTION ...................................................................................................................................... 5 Why examine government activities? ......................................................................................................... 5 What are the goals of the indicators in Government at a Glance? .............................................................. 5 What can you find in this publication?........................................................................................................ 6 Measurement challenges ............................................................................................................................. 8 Future plans.............................................................................................................................................. 10 How this publication is organized............................................................................................................. 10 II. POLITICAL AND ADMINISTRATIVE CONTEXT ............................................................................. 11 II.1 ELECTORAL SYSTEMS AND COALITION GOVERNMENTS .................................................. 12 II.2 INSTITUTIONAL CHECKS ON POLITICAL POWER .................................................................. 15 II.3 ADMINISTRATIVE STRUCTURES................................................................................................ 19 II.4 CORE PUBLIC SERVICE VALUES ................................................................................................ 22 III. GOVERNMENT REVENUES ............................................................................................................... 24 III.1. GENERAL GOVERNMENT REVENUES .................................................................................... 25 III.2. STRUCTURE OF GENERAL GOVERNMENT REVENUES ...................................................... 30 III.3. REVENUE STRUCTURE BY LEVELS OF GOVERNMENT ...................................................... 33 IV. GOVERNMENT EXPENDITURES ...................................................................................................... 37 IV.1. GENERAL GOVERNMENT EXPENDITURE .............................................................................. 38 IV.2. GENERAL GOVERNMENT EXPENDITURE BY SUB-SECTOR .............................................. 42 IV.3. GERNAL GOVERNMENT EXPENDITURE BY LEVELS OF GOVERNMENT ....................... 46 IV.4. GENERAL GOVERNMENT EXPENDITURES BY COLLECTIVE AND INDIVIDUAL GOODS AND SERVICES, AND CASH TRANSFERS AND GOODS AND SERVICES IN KIND.... 50 V. THE COST OF GOVERNMENT PRODUCED AND GOVERNMENT FUNDED GOODS AND SERVICES .................................................................................................................................................. 52 V.1 PRODUCTION COSTS IN GENERAL GOVERNMENT ............................................................... 53 VI. EMPLOYMENT IN THE PUBLIC SECTOR ....................................................................................... 56 VI.1. EMPLOYMENT IN GENERAL GOVERNMENT AND PUBLIC AND QUASI-PUBLIC CORPORATIONS ................................................................................................................................... 57 VI.2. DECENTRALIZATION OF EMPLOYMENT ............................................................................... 61 VII. WORKFORCE CHARACTERISTICS OF THE CENTRAL GOVERNMENT ................................. 64 VII.1. EMPLOYMENT OF WOMEN IN CENTRAL GOVERNMENT ................................................ 65 VII.2. AGEING WORKFORCE ............................................................................................................... 68 VIII. HUMAN RESOURCES MANAGEMENT PRACTICES IN CENTRAL GOVERNMENT ............. 71 2 GOV/PGC(2009)2 VIII.1 DELEGATION IN HUMAN RESOURCES MANAGEMENT AT THE CENTRAL GOVERNMENT ...................................................................................................................................... 72 VIII.2 OPENNESS IN RECRUITMENT IN CENTRAL GOVERNMENT ............................................ 75 VIII.3 STAFF PERFORMANCE MANAGEMENT IN CENTRAL GOVERNMENT .......................... 77 VIII.4 SENIOR CIVIL SERVICE ....................................................................................................... 80 IX BUDGET PRACTICES AND PROCEDURES ...................................................................................... 84 IX.1 Fiscal Sustainability .......................................................................................................................... 85 IX.2 BUDGET TRANSPARENCY .......................................................................................................... 88 IX.3 MEDIUM-TERM BUDGET PERSPECTIVE ................................................................................. 91 IX.4 PERFORMANCE-ORIENTED BUDGETING ................................................................................ 94 IX.5 EXECUTIVE FLEXIBILITY IN BUDGET EXECUTION ............................................................. 97 X. REGULATORY MANAGEMENT ....................................................................................................... 100 XI. INTEGRITY ........................................................................................................................................ 101 XI.1 CONFLICT OF INTEREST—DISCLOSURE OF FINANCIAL INSTERESTS BY DECISION MAKERS ............................................................................................................................................... 102 XI.2 PUBLIC INTEREST DISCLOSURE: WHISTLE-BLOWING ..................................................... 104 XI.3 PREVENTING CORRUPTION: PUBLIC PROCUREMENT ...................................................... 106 XII. E-GOVERNMENT ............................................................................................................................. 110 XII.1 E-GOVERNMENT READINESS................................................................................................. 111 XII.2 E-GOVERNMENT SERVICE MATURITY................................................................................ 114 XII.3 UPTAKE OF E-GOVERNMENT SERVICES ............................................................................. 117 XIII. OPEN GOVERNMENT .................................................................................................................... 121 XIII.1. OPEN GOVERNMENT LEGISLATION .................................................................................. 122 ANNEX A: CLASSIFICATION OF THE FUNCTIONS OF GOVERNMENT (COFOG) ...................... 125 ANNEX B: COMPOSITE INDICATORS FOR HRM, REGULATORY MANAGEMENT AND BUDGET PRACTICES .............................................................................................................................. 129 B.1. HRM Composites:Theoretical framework, construction and weighting..................................... 130 B.2. Budget Composites: Theoretical framework, construction and weighting .................................. 143 B.3. Regulatory Management Composites: Theoretical framework, construction and weighting ...... 182 ANNEX C CONFLICT OF INTEREST DISCLOSURE REQUIREMENTS FOR DECISION MAKERS ................................................................................................................................................................... 183 C.1 Communicating core values to central government public servants ................................................ 184 C.2 Conflict of interest disclosure requirements ..................................................................................... 185 C.3 Protection of Whistle-blowers .......................................................................................................... 187 ANNEX D: MEMBERS OF THE STEERING GROUP ........................................................................... 188 BIBLIOGRAPHY ...................................................................................................................................... 189 3 GOV/PGC(2009)2 FOREWORD 1. This is the first edition of the biennial Government at a Glance publication, a new addition to the OECD’s at a Glance series. Government at a Glance fills an important gap in internationally comparative data and general knowledge on how governments work, providing a suite of key indicators describing government institutions, structures, inputs and efficient and effective public management practices in OECD countries. Its publication is particularly timely as the current financial and economic crises have enhanced the role of government in the economy and society in many OECD member countries, while requiring that governments be more efficient, effective and citizen-centred. 2. People from all backgrounds interested in how their government works and how their government’s performance compares to other OECD countries - including politicians, policy makers, public sector practitioners, advocacy groups, academics, students and the general public - should find Government at a Glance informative. In addition, the indicators included in Government at a Glance could help non-OECD countries identify new ways to track and benchmark where they stand with the development and implementation of policies and reforms. 3. This work was led by Zsuzsanna Lonti under the direction of Martin Forst with major contributions from Jani Heikkinen, Jordan Holt and Laurent Nahmias; Dirk-Jan Kraan (Chapters 3 and 4 on Revenues and Expenditures); Elsa Pilichowski (Chapters 5 - 8 on Production Costs, Employment in the Public Sector, Workforce Characteristics of Central Government Employees and Human Resource Management Practices); Teresa Curristine, Ian Hawkesworth and James Sheppard (Chapter 9 on Budget Practices and Procedures); Christiane Arndt, Gregory Bounds and Stephane Jacobzone (Chapter 10 on Regulatory Management); Janos Bertok and Jean-Francois Leruste (Chapter 11 on Integrity); Jean- Francois Leruste, Barbara Ubaldi and Yih-Jeou Wang (Chapter 12 on E-government); and Joanne Caddy (Chapter 13 on Open Government). 4. Useful comments have been provided by members of the Steering Group (details in Annex D); Nick Manning (World Bank); Wouter van Dooren (University of Antwerp); Geert Bouckaert (Catholic University of Leuven); Joachim Wehner (London School of Economics); Knut Rexed; Nadim Ahmad and Enrico Giovanni (OECD Statistics Directorate); Jens Lundsgaard and Maurice Nettley (OECD Tax Directorate); Barry Anderson, Monica Brezzi, Maria Varinia Michalun, Elsa Pilichowski and Andrea Urhammer (OECD Governance Directorate). 4 GOV/PGC(2009)2 I. INTRODUCTION 5. Government at a Glance builds upon 20 years of accumulated expertise at the OECD in the area of public governance (defined as the exercise of political, economic and administrative authority) and public management, particularly in describing and analyzing government activities, developing benchmarks and internationally comparative data, identifying good practices and monitoring results. It also benefits from a practitioner-focus given the OECD’s unique access to senior level officials in member governments. 6. The work on Government at a Glance began in earnest in 2005. Since then, three working papers and four technical papers have been published, establishing a clear theoretical framework for the project and reviewing, and analyzing the existing literature on the measurement of government activity as well as the empirical evidence on the institutional drivers of efficiency in government. Indicators selected for this first edition of Government at a Glance are based on existing data identified in these papers and have been chosen in agreement with member countries. Why examine government activities? 7. Government is a major actor in modern society, contributing to economic growth, delivering goods and services, regulating the behaviour of businesses and individuals and redistributing income. Government activities affect people’s lives in countless ways from birth to death, by providing basic health care services and education, helping people when they lose their jobs, issuing driver’s licences and business permits, building roads and bridges and regulating the environment and workplace health and safety. A significant portion of the economy is devoted to public activities: in 2006, government expenditures ranged between 30 and 55% of Gross Domestic Product (GDP) in OECD countries. 8. Thus, good governance is critical to long-term economic, social and environmental development. What government does matters. How efficiently and effectively government performs these tasks is also critical, as citizens increasingly demand more from governments in terms of services that better meet their needs and the efficient and accountable use of resources. The ability of governments to operate efficiently and effectively is dependent on their management practices and policies, including budgeting practices that support fiscal sustainability; human resource management practices conducive to good performance; principles of public sector values and ethics; the wise use of e-government tools; and on the transparency and participatory nature of these practices. Thus, in order to evaluate government performance, its activities - including the resources it uses to produce goods and services and the underlying management processes and practices - need to be measured. As governments provide a large and evolving array of goods and services, quantifying and measuring government actions can help managers and leaders make better decisions, and can help to hold government accountable to its citizens. In addition, describing government structures and arrangements can illustrate important similarities and differences between countries, facilitating mutual learning. What are the goals of the indicators in Government at a Glance? 9. The objective of the indicators presented in Government at a Glance is to assist countries to: • Better understand their own practices. 5 GOV/PGC(2009)2 6 • Benchmark their achievements through international comparisons. • Learn from the experience of other countries facing similar challenges. 10. In the short term, the publication could help governments better assess, plan and measure their activities and their performance. Over time, it might allow governments to link their activities to overall performance and help evaluate the impacts of their reform agendas. What can you find in this publication? Focus on public administration 11. Government at a Glance focuses on the workings of public administration, or the “machinery of government,” in the 30 OECD member countries This includes the resources, practices and policies of government that support the goods and services (such as health and education) that it produces and delivers. The focus on public administration builds on the prior work of the OECD and benefits from the expertise of member country delegates to the OECD’s Public Governance Committee, the group guiding the OECD’s work on Government at a Glance. In particular, indicators on public management practices (including integrity, budget and regulatory management) are based on good practices and common standards that were developed and agreed upon by OECD countries. For example, indicators on integrity focus on government policies to require decision-makers to publicly disclose potential conflicts of interest, reflecting the principles included in the OECD’s 2003 Recommendations for Managing Conflict of Interest in the Public Service. While focusing on public administration builds on the capabilities of the OECD, it also fills a void. Internationally comparative data on the functioning of public administration are scarce and constitute a largely uncharted territory. Indicators on good government and public management practices 12. Government at a Glance is built on the following framework, which describes the public “production” process and identifies six major categories of indicators: context, revenues, inputs, processes, outputs and outcomes. Framework for understanding and measuring the activities of government Revenue Inputs Processes Outputs Outcomes How much money does government collect? How much and what kind of resources does goverment use? What does the government do, and how does it do it? What are the goods and services which the government produces? What is the resulting impact on citizens and businesses? Context Structure of government GOV/PGC(2009)2 Based on Hatry (1999), Pollitt and Bouckaert (2004) and W.K. Kellogg (2004) 13. The first edition of Government at a Glance includes four of the six types of indicators indentified in the framework: context, revenue, inputs and processes. (1) Contextual indicators 14. Contextual indicators include data on public service core values and major features of the political and administrative structures in member countries. By determining whose views are represented in the legislature and government and how decisions are made, these structures affect the types of goods and services that government offers, how these goods and services are paid for and even how they are produced. Situating policies and indicators within this contextual background can help us better understand differences between countries and identify countries with similar political and administrative structures that might serve as better comparators for benchmarking purposes. (2) Revenues 15. Data on the revenue structure provide insights into the incentives and constraints that governments face in determining what types of goods and services to provide. In addition, the amount and structure of revenues collected determine how the costs of past, present and future government activities are shared across society. Government revenues include social contributions and other taxes, and other sources such as fees, oil production and international grants. While revenues may not match expenditures in any given year due to governments’ ability to borrow, ultimately citizens must bear the cost of repaying debt. (3) Inputs 16. Input indicators include data on government expenditures, production costs, employment and workforce characteristics. These data allow countries to begin comparing the proportion of the economy devoted to producing different goods and services, as well as differences in the mix of inputs used in production For example, labour is a key input in the government’s production process, and the characteristics of the public workforce may affect government productivity and its capacity to provide goods and services. In addition, governments are increasingly outsourcing the production of goods and services, although the extent that private entities are involved in government activities varies greatly across countries. (4) Processes 17. Process indicators describe public management practices that influence government outputs and outcomes. Information on processes - such as human resource management (HRM), budget, regulatory management, integrity, e-government and open government – can allow countries to being to examine determinants of efficiency and effectiveness, and identify potential reforms and strategies to improve productivity. For example, delegating the ability to hire, fire and promote staff to line managers can enhance the flexibility of government to respond to changing circumstances and enable managers to obtain employees with the skills needed to improve the efficiency and effectiveness of the public services they provide. Likewise, the use of multi-year expenditure estimates in budgets can improve fiscal discipline and help to ensure that government resources are allocated productively and efficiently. Adopting strategies and processes to assess the impact of regulations on businesses and citizens can help to ensure that they meet the intended goals of government. Open and competitive procurement processes can reduce the incidence of corruption, increasing the resources available and devoted to producing goods and services and improving public trust in government. Similarly, the use of the Internet and other communications 7 GOV/PGC(2009)2 technologies to provide information and public services, such as tax payments or passport renewals, can both lower costs and save citizens and businesses time - increasing efficiency and productivity. Data presented for varying levels of government 18. In general, the data presented are based on the definition of “government” found in the System of National Accounts (SNA). In these terms, government encompasses ministries, departments and agencies, as well as non-profit institutions that receive a significant amount of financing from the government. Data on revenues and expenditures are presented for both the central and sub-central (state and local) levels of government, while data on workforce characteristics and public management processes and practices refer to the central level of government only. In addition, data on employment refer to the “public sector,” which expands the definition of government to include public and quasi-public corporations, such as publicly owned banks, harbours and airports. Other features of the data 19. Government at a Glance includes both financial (e.g. revenues and expenditures) and nonfinancial data (e.g. employment) as well as quantitative (e.g. workforce composition) and qualitative data (e.g. open government practices). The publication is based on primarily OECD data, which are provided by member country government officials. When OECD data were not available, the publication used international data from the European Commission, the World Economic Forum and the United Nations. 20. As part of the indicator set describing public management practices, Government at a Glance introduces several descriptive composite indexes in narrowly defined areas that were developed according to OECD guidelines. These composite indexes are a practical way of summarizing discrete, qualitative information on key aspects of public management practices, such as the openness of HRM systems to external applicants at all position levels. Details about the weights and variables used to construct these indicators are provided in Annex B, giving countries that want to take action on their indicator scores a clear indication of where to start. However, while the composite indicators were developed in cooperation with member countries and are based on best practices and/or theory, both the variables comprising the composites and their weights are offered for debate and, consequently, may evolve over time. How Government at a Glance differs from other data sets 21. Government at a Glance is unique in that: • The key aspects of public administration and the specific indicators presented were selected based on a consensus among member countries. • Data and qualitative information are provided by government officials of member countries. • It does not create a single aggregate indicator to evaluate government performance nor does it provide an overall ranking of countries based on their performance. Measurement challenges 22. The objectives and boundaries of government vary across countries, making identifying and collecting standardized data a key measurement challenge. For example, producing standardized, internationally comparable public sector employment data is complicated by country differences in, amongst others, the organizations of health systems and the use of private contractors. Across countries, and even within countries depending on the source of information, the definitions of “government 8 GOV/PGC(2009)2 organisations,” “the public sector” and “public services" vary significantly. Even sources that use a widely accepted definition like that of “General Government” in the System of National Accounts tend to provide data that are sometimes inconsistent with that definition. As a result, the OECD initiated its own data collection efforts in 2006 with the “Comparison of Employment in the Public Domain” (CEPD) survey. This survey collected data on employment in General Government (core ministries, departments and agencies, non market publicly owned hospitals, public schools, etc.) and employment in public and quasipublic corporations, which together define the public sector. The results of the survey are presented in Chapter 6. 23. In addition to measurement challenges associated with variations in the scope of government across countries, countries sometimes collect data in different ways, limiting comparability. For example, countries can either collect data on employment in full-time equivalencies or as total number of employees. To the extent possible, OECD survey instruments attempt to establish standardized definitions so that full comparisons can be made. In addition, the OECD cleans and verifies the data it collects by following up with member countries directly when there are potential inconsistencies or outliers, and, in some cases, by using other member countries as peer reviewers to verify responses. 24. Even in areas where standard definitions exist, such as expenditure data from the System of National Accounts, varying levels of detail are available for OECD member countries which can limit the type of analysis and comparisons that can be made. For example, detailed data on the purpose of government expenditures (called Classification of the Functions of Government, or COFOG level II) are currently only available for 13 OECD European member nations. These data could be used to match public spending components to economic and social objectives, such as comparing expenditures on research and development in different policy sectors or expenditures on environmental protection. Using the available second-level COFOG data, the OECD has developed a methodology to estimate government expenditures by individual and collective goods and cash and in-kind provision. This information helps to understand important differences in policy and service delivery choices of governments. This methodology is called COFOG- Special and is published in Government at a Glance. 25. In addition, it is challenging to develop valid indicators that truly measure the quality of public services (i.e. their ability to meet customer needs), the implementation of policies and practices and the outputs and outcomes of public administration. The first edition of Government at a Glance does not include output or outcome indicators, or indicators on service quality, such as customer satisfaction. Internationally there is still an extensive debate about how to measure outputs and outcomes in the public sector, and even leading countries in output and outcome measurement follow different paths. 26. The challenges associated with measuring outputs and outcomes go beyond those enumerated above, and are explored in depth in the recent OECD publication Measuring Government Activity. For example, output and outcome measurement may affect organizational behaviour. While the existence of output measures may lead staff to strive for improved performance, it may also lead to the neglect of nonmeasured dimensions or to “gaming,” in which either the output itself is adjusted or the measurements are distorted in order to achieve the appearance (rather than the reality) of “good” performance. In addition, while it is usually reasonable to hold government responsible for outputs, it is often not reasonable to hold it responsible for outcomes because many other factors beyond government’s control may influence the final impacts on society. 27. Due to these challenges in measurement and data collection, the indicators presented in Government at a Glance are just that—indicators of government activities and performance. Absolute levels or number should be interpreted with caution due to measurement errors. In addition, the raw data presented in Government at a Glance cannot in itself be used as a benchmark; rather comparisons among countries must be based on an analysis of the relevant indicators. The OECD’s in-depth public 9 GOV/PGC(2009)2 management reviews can also provide more nuanced details of individual country policies, practices and contexts that can be used to help better understand drivers behind performance differences. Future plans 28. Government at a Glance will be a biennial publication, and future editions will build upon the data presented in this first publication. Data contained in this edition are planned to be collected at regular intervals, utilising methods that will minimize the burden of data provision on member countries. A stable data set is needed to enable governments to compare their institutional arrangements and performance to other OECD countries not only at one point in time but over time; to shed light on the possible causes of performance differences among governments; and to facilitate more in-depth analysis of the impact of public sector reforms. A main goal of future data collection efforts is to improve the quality of the current data set. At the same time, the data set will not be static as new indicators can be added once their significance is established. 29. Future work may focus on expanding the availability of data on the characteristics of government inputs (e.g. the average educational attainment of government employees) and processes (e.g. how laws and regulations are implemented and/or enforced). In addition, future editions of Government at a Glance may further explore the relationship between and responsibilities of the different levels of government (central, state and local) across OECD countries. Despite the measurement challenges mentioned above, reliable and feasible internationally comparative output and outcome indicators for public administration and indicators of service quality (such as customer satisfaction) are being developed in cooperation with member countries. Intermediate outputs are being identified for different public administration processes, together with several “executive governance outcomes” such as public trust in government, equity and fiscal/economic stability. How this publication is organized 30. Government at a Glance comprises 13 Chapters that follow the framework of the production process, providing a holistic picture of government activity. Following the introduction, Chapter 2 provides indicators on the political and administrative context within which government decision-making and production occur. Chapter 3 looks at the amount and structure of revenues collected by government, and Chapter 4 examines what activities countries spend these resources on. Chapter 5 on production costs looks more closely at how goods and services are produced -i.e. whether the government produces weapons (goods) and/or provides health care (services) itself, or whether it contracts with non-profit or forprofit private entities to produce or provide them on its behalf to citizens and businesses. Chapters 6 and 7 look more closely at the size and characteristics of the public sector workforce, which is related to government decisions about how and who should be responsible for providing public goods and services. Chapters 8 and 9 examine the processes underlying government production, including human resource management practices that can influence the characteristics of the labour force and budgeting practices that can influence decisions on the amount of revenues to collect and the size and direction of government expenditures. Chapter 10 examines the quality of government processes to design and reform government regulation, looking at the extent to which countries have adopted best practice standards. Chapter 11 looks at government measures to promote integrity and prevent corruption, which can assume even more importance in countries that have outsourced the production of a larger proportion of goods and services to the private sector. Chapter 12 looks at the maturity of e-government services in OECD countries, and Chapter 13 describes the legislation and institutions that governments have enacted and created to help keep their actions accountable to the general public. Altogether, these Chapters provide insights on what government does, how it does it and why it does things certain ways. They allow countries to better understand their practices, benchmark their performance, identify areas for potential future reform and begin to assess the effectiveness and efficiency of their operations. 10 GOV/PGC(2009)2 II. POLITICAL AND ADMINISTRATIVE CONTEXT 31. Situating policies and indicators within a contextual background can help highlight and clarify differences between countries. Important contextual factors include electoral systems and coalition governments, political and administrative structures and core public service values. By determining what views are represented in government, how and by whom policy decisions are made and who administers programs and reforms, a country’s political and administrative context can clearly affect policy outcomes and ultimately government efficiency and effectiveness. Reflected in constitutions, laws and statutes, public service core values serve as the ethical infrastructure of public administration that guide the mission and daily operations of government organizations. Values both reflect and potentially reinforce recent public management reforms, such as those designed to increase transparency and improve service delivery. 32. While many contextual factors are products of a country’s historical development and cannot be easily changed by policy makers, they can be used to identify countries with similar political and administrative structures for comparison and benchmarking purposes. In addition, for countries considering different policies and reforms, they can illustrate structural differences that may affect their passage and implementation. 11 GOV/PGC(2009)2 II.1 ELECTORAL SYSTEMS AND COALITION GOVERNMENTS 33. The type of electoral system employed has a number of potential consequences on the nature and tenure of government, including the diversity of views represented and the ability of the legislature to create and amend laws. Major differences in legislative institutions that can affect the way a country’s bureaucratic system works include the existence of a bicameral system and an elected Upper House; the type of electoral system; the frequency of elections; and the frequency of coalition governments. 34. Twelve of the 17 OECD countries with bicameral legislatures elect the members of the Upper House. While bicameral legislatures often provide broader representation and a diffusion of power, policymaking may take more time due to the deliberation of both houses. The vast majority of OECD countries elect their Lower Houses through some sort of proportional representation system, which are usually characterized by multiple, smaller political parties. In comparison, single-member systems (i.e. where constituents vote for a single candidate rather than parties) are usually characterized by two strong parties. 35. Members of the Lower House are usually elected for shorter periods of time than Upper House members; on average, elections are required every 4 years for Lower House members and every 5 years for Upper House members. In countries with parliamentary systems, the legislature may be dissolved before its term expires. 36. The link between the electoral system and the frequency of coalition governments is clearly supported by the data, with coalition governments occurring relatively infrequently in single-member systems, such as Canada and the United Kingdom. The frequency of coalition governments in countries with multi-member systems depends largely on the strength and number of political parties. Although coalition governments may be more limited in their choices for action due to their need to work on a more consensual basis, policies resulting from coalition governments in some countries may be more resistant to change of government, and therefore more durable over time. Methodology and Definitions 37. Bicameral legislatures have both an Upper House and a Lower House whereas unicameral legislatures are comprised of only a Lower House. The types of electoral systems presented in figure II.1.2 are defined as follows: • Under First Past the Post, the winner is the candidate with the most votes but not necessarily an absolute majority of the votes. • Under Preferential and Two Round, the winner is the candidate who receives an absolute majority (i.e. over 50%) of votes. If no candidate receives over 50% of votes during the first round of voting, the Preferential system makes use of voters’ second preferences to produce a winner while the Two Round system uses a second round of voting. • Proportional Representation (PR) systems allocate parliamentary seats based on a party’s share of national votes. • Semi-proportional systems feature attributes of both single-member and PR systems, and allow two votes per person: one for a candidate running in the voter’s district and one for a party. As in PR, party seats are allocated proportional to the party’s share of national votes. 38. Data on the frequency of elections for Lower and Upper House members reflect statutory requirements. In actuality, elections may be held more frequently in parliamentary systems if governments 12 GOV/PGC(2009)2 collapse. Data on the frequency of coalition governments covers the period between January 1, 1988 and December 31, 2008, except for the Czech Republic (1992), Hungary (1990), Poland (1991) and the Slovak Republic (1993). A coalition government is defined as the joint rule of executive functions by two or more political parties. The number of governments is defined by the number of terms served by the head of the executive branch (where a term is either defined by a change in the executive or an election that renewed support for the current government). Source: Data are from member country government websites and are current as of December 31, 2008. Figure notes: (1) Thirteen countries have a unicameral legislature and do not have an Upper House: Denmark, Finland, Greece, Hungary, Iceland, Korea, Luxemburg, New Zealand, Norway, Portugal, Slovak Republic, Sweden and Turkey. (3) No value indicates a non-elected Upper House. In Austria and Ireland, the Upper House is appointed every 5 years. (4) Data are not presented for Korea, Mexico and the United States, which have a presidential system of executive power. In Switzerland, and agreement dictates the composition of the government, which is always a coalition of four parties. Data for France represent periods of cohabitation (coalition governments) and the number of prime ministers (number of governments). Figure II.1.1 Electoral system for Upper House (2008) Typology Total Country Elected Upper House 11 Australia, Czech Republic, France, Italy, Japan, Mexico, Netherlands, Poland, Spain, Switzerland, United States Partially elected Upper House 1 Belgium Non-elected Upper 5 Austria, Canada, Germany, Ireland, United Kingdom House Figure II.1.2 Electoral system for Lower House (2008) System Total Country Single Member First Past the Post 3 Canada, United Kingdom, United States Preferential 1 Australia Two Round 1 France Multi Member Proportional Representation Semi-Proportional Representation 17 8 Austria, Belgium, Czech Republic, Denmark, Finland, Greece, Iceland, Ireland, Luxembourg, Netherlands, Norway, Poland, Portugal, Slovak Republic, Spain, Sweden, Turkey Germany, Hungary, Italy, Japan, Korea, Mexico, New Zealand, Switzerland 13 GOV/PGC(2009)2 Figure II.1.3 Statutory requirements for the frequency of elections (2008) Figure II.1.4 Frequency of coalition governments at the central level between 1988 and 2008 14 GOV/PGC(2009)2 II.2 INSTITUTIONAL CHECKS ON POLITICAL POWER 39. Political institutions are fundamental to the proper functioning of democracy. The extent that power is shared between the legislative and executive branches, term limits for presidents or prime ministers, the ease of constitutional amendments and the ability of the judiciary to review the constitutionality of laws and actions set the constraints within which policies and reforms can be implemented and enacted. 40. Over two-thirds of OECD countries have a parliamentary system of executive power. Under this system, the executive is usually the head of the dominant party in the legislature and appoints members of his party or the coalition government to serve as ministers of the executive departments (ministries). Ten of the 23 OECD countries with parliamentary systems also have a president, whose powers are predominately ceremonial in nature. Under the presidential system, which is used in only three OECD countries, the executive and members of the legislature seek election independently of each other and ministers are appointed officials, not elected members of the legislature. The separation of the executive and the legislature provides checks and balances against the power of each structure. Four OECD countries operate under a dual executive system that combines a powerful president with an executive responsible to parliament, both of whom are responsible for the day-to-day activities of the state. It differs from the presidential system in that the cabinet, although named by the president, is responsible to the legislature, which may force the cabinet to resign through a motion of no confidence. 41. The existence of term limits influences the behaviour of executive governments, determining the amount of time available to design and implement policies. In most parliamentary systems, the prime minister and his/her cabinet remain in office as long as s/he enjoys the support of the legislature, but not beyond the duration of the legislative term. All of the OECD countries with presidential or dual executive systems have term limits for presidents. Most countries have two-term limits for presidents, although in Mexico, Korea and Turkey, the president is limited to one term in office. 42. The executive and legislature must enact and implement laws and policies within the framework of the constitution. In all but four OECD countries, the court has the power to annul acts of the executive and legislature where it finds them unconstitutional. While judicial review is a control device against possible abuses by the executive or legislature, courts need to be genuinely independent for it to work as a true check on power. 43. Laws and policies may be more rigid when constitutional amendment is more difficult. On the other hand, constitutions that are difficult to amend may safeguard the fundamental values of democratic society from the issues of the day. OECD member countries can amend constitutions with varying difficulty. In the United Kingdom, parliament can amend the constitution with a simple act of legislation. In comparison, amendment is difficult in the United States and requires a two-thirds approval of both houses and ratification by three-fourths of all states. In the four countries without systems of judicial review, constitutional amendment is also relatively simple, suggesting that the legislature is relatively powerful. Methodology and Definitions 44. Data are from member country constitutions and websites and are current as of December 31, 2008. Judicial review refers to the ability of the courts or a separate body to review the constitutionality of laws and actions and is usually enshrined in the constitution. In countries with limited judicial review, the courts only have the ability to review the constitutionality of specific types of laws or actions or under specific circumstances. 15 GOV/PGC(2009)2 Source: Member country websites and constitutions. Figure Notes: (2) Twelve OECD countries do not have Presidents: Australia, Belgium, Canada, Denmark, Japan, Luxembourg, Netherlands, New Zealand, Norway, Spain, Sweden and the United Kingdom. While the head of the Swiss government is called the “President,” s/he is elected annually by legislature. (3) The United Kingdom and New Zealand do not have a formal written constitution. Figure II.2.1 System of executive power (2008) System Total Country Parliamentary 23 Australia, Austria, Belgium, Canada, Czech Republic, Denmark, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Slovak Republic, Spain, Sweden, Switzerland, Turkey, United Kingdom Dual executive 4 Finland, France, Poland, Portugal Presidential 3 Korea, Mexico, United States Figure II.2.2 Existence of term limits for presidents (2008) Existence Total Country Term limit 15 Austria (12 yrs), Czech Republic (10 yrs), Finland (12 yrs), France (10 years), Germany (10years), Greece (10 yrs), Hungary (10 years), Ireland (14 yrs), Korea (5 yrs), Mexico (6 yrs), Poland (10 yrs), Portugal (10 yrs), Slovak Republic (10 yrs), Turkey (7 yrs), United States (8 yrs) No term limit 2 Iceland, Italy 16 GOV/PGC(2009)2 Figure II.2.3 Ease of Constitutional Amendment (2008) Legislative Approval Additional Approval(s) Country Majority (half circle) or Supermajority (full circle) Both houses Two consecutive legislatures Head of state Referendum Majority of states Australia .... Austria . Belgium ... Canada ... Czech Republic .. Denmark ... France ... Finland .. Germany .. Greece .. Hungary . Iceland ... Ireland ... Italy ... Japan ... Korea .. Luxemburg .. Mexico ... Netherlands ... New Zealand . Norway .. Poland .. Portugal . Slovak Republic . Spain .. Sweden .. Switzerland .. Turkey .. United Kingdom .. United States ... Required: . 17 GOV/PGC(2009)2 Figure II.2.4 Existence of system of judicial review of the constitutionality of laws and actions Typology Total Country Judicial review 19 Australia, Austria, Czech Republic, Denmark, Germany, Greece, Hungary, Iceland, Ireland, Italy, Korea, Luxembourg, Mexico, Norway, Poland, Portugal, Slovak Republic, United States Limited judicial review 7 Belgium, Canada, France, Spain, Sweden, Switzerland, Turkey No judicial review 4 Finland, Netherlands, New Zealand, United Kingdom 18 GOV/PGC(2009)2 II.3 ADMINISTRATIVE STRUCTURES 45. The way that governments are structured, including the division of responsibilities horizontally (between departments or ministries) and vertically (across levels of governments), forms the institutional basis for the functioning of public administration. 46. While a state’s structure is constitutionally defined, whether it is federal or unitary may have consequences for the degree of decentralization of authority and the effectiveness and efficiency of governments. Over two-thirds of OECD countries are unitary states, which have granted varying degrees of autonomy to local and regional governments. Eight OECD countries are federal states where regional governments enjoy constitutionally-defined autonomy and power, often including responsibilities for education, fiscal policy and policing. In recent decades, some unitary states have devolved more power to local governments while many federal states have become more centralized. While decentralization may allow governments to be more responsive to local needs, it may also lead to inefficiencies stemming from diverse laws and policies and duplicative ministries at the central and sub-central levels. Reforms implemented in more decentralized states may be narrower in scope and less uniform in practice, as different entities are able to go in different directions and different speeds (Pollitt et al., 2004). 47. OECD countries vary widely in the number of ministries, with New Zealand at one end with 35 ministries and Switzerland at the other with 7 ministries. The number of ministers does not necessarily correspond to the number of ministries and both may be influenced by political motivations; in some countries, ministers are responsible for more than one ministry, while in others, additional ministers do not have a ministerial portfolio. The political influence is less pronounced in presidential systems where ministers are not appointed by the legislature. For example, Mexico, the United States and Korea all have approximately the same number of ministries and ministers. Increasing the scale of operations by undertaking government activities through fewer, larger ministries can lead to efficiency gains due to savings in overhead and fixed costs—although creating an organization that is too large may make it difficult for managers to pay sufficient attention to all key issues. Likewise, recent research has found a positive association between the number of ministers and budget deficits and expenditures, which may be strengthened when there is partisan fragmentation in government. (Wehner, 2009) 48. Nineteen OECD member countries are also members of the European Union, where common policy goals regarding economic growth, agriculture, energy, infrastructure and research and development (among others) may affect the amount and structure of revenues and expenditures as well as a county’s public management processes and practices, such as its budgetary framework. Methodology and Definitions 49. Federal states have a constitutionally delineated division of power between one central and several regional or state governments. While unitary states often include multiple levels of government (such as local and provincial or regional), these administrative divisions are not constitutionally defined. 50. A ministry is an organization in the executive branch that is responsible for a sector of public administration. In some OECD countries, such as the United States, Sweden, Switzerland and Norway, ministries are called departments. Common examples in OECD countries include the Ministries (Departments) of Health, Education, Finance and Defence. While sub-central governments may also be organized into ministries, the figures and text only refer to ministries at the central level of government. Ministers advise the executive and are either in charge of one or more ministries or are responsible for a portfolio of government duties. In most parliamentary systems, ministers are drawn from the legislature and keep their seats. In most presidential systems, ministers are not elected officials and are appointed by 19 GOV/PGC(2009)2 the President. The data presented in Figure II.3.3 refer to the number of ministers that comprise the cabinet at the central level of government and exclude deputy ministers. Source: Data are from government websites and constitutions, and are current as of December 31, 2008. Further reading: Pollitt, Christopher et al. (2004), Agencies – How Governments Do Things Through Semi-Autonomous Organizations, Palgrave Macmillan, Bainstoke, UK. Wehner, J. (2009), “Cabinet Structure and Fiscal Policy Outcomes,” unpublished paper, Government Department, London School of Economics, http://personal.lse.ac.uk/wehner/cabinet.pdf Figure II.3.1 State structure (2008) State structure Total Country Federal 8 Australia, Austria, Belgium, Canada, Germany, Mexico, Switzerland, United States Unitary 22 Czech Republic, Denmark, Finland, France, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Netherlands, New Zealand, Norway, Poland, Portugal, Slovak Republic, Spain, Sweden, Turkey, United Kingdom Figure II.3.2 Number of departments or ministries at the central level of government (2008) 20 GOV/PGC(2009)2 Figure II.3.3 Number of ministers at the central level of government (2008) Figure II.3.4 OECD countries that are members of the European Union (2008) TypologyMembers of the European Union Total 19 Country Austria, Belgium, Czech Republic, Denmark, Finland, France, Greece, Germany, Hungary, Ireland, Italy, Luxembourg, Netherlands, Poland, Portugal, Slovak Republic, Spain, Sweden, United Kingdom Not members of the European Union 11 Australia, Canada, Iceland, Japan, Korea, Mexico, New Zealand, Norway, Switzerland, Turkey, United States 21 GOV/PGC(2009)2 II.4 CORE PUBLIC SERVICE VALUES 51. Values form the foundation of the public service. They are encoded in organisations’ culture and manifested through attitudes, employee conduct and decision making. Values guide judgment about what is good and improper in serving the public interest. In addition, values stated in public documents shape citizen’s expectations about the mission, vision and daily activities of public sector organisations. 52. Although values reflect national, social, political and administrative contexts, the data show a certain homogeneity of stated values across countries. Impartiality and legality have remained the top public service values over the past decade. They mark a distinctive feature from the private sector, which puts a particular emphasis on profitability and innovativeness. (van Der Wal and Huberts, 2008) 53. A significant shift can also be observed in stated core public service values as a result of new public management and good governance efforts. For example, the number of countries identifying transparency as a core public service value almost doubled in the past decade, and integrity is increasingly identified as a core value. Similarly, significantly more OECD countries stated efficiency as a core public service value, indicating an emerging convergence with the private sector. Only five OECD countries have developed a hierarchy of values to provide employees guidance in situations where values may conflict. These countries prioritize impartiality, legality and integrity. 54. Public service values are most commonly stated in a country’s legal framework, including the constitution, statutes and civil service or public service regulations. Almost three-quarters of OECD countries reviewed and revised the statement of core public service values in the past decade to meet the evolving expectations of society for good governance and for an increasingly results-based public service. Whereas public servants were involved in the update in the majority of OECD countries, the engagement of the public was not as common; only six countries consulted citizens in the revision process. 55. Providing stated values to employees when they join the public service has remained the most common dissemination method over the past decade. New technologies, such as the Internet, Intranet and CD-ROMS, are now widely used for communicating core values. Methodology and Definitions 56. The information on core values was collected in 2000 and 2009 using a survey instrument developed by OECD and was provided by public officials responsible for integrity issues in central governments. The data reflect the overall values of central government expressed in legal and policy documents rather than values of a particular ministry, department or agency. A total of 25 OECD member countries responded to both the 2000 and 2009 surveys. Thus, the data presented in figures II.4.1 and II.4.2 represents the percentage of the 25 countries for which data are available in 2000 and 2009. Annex C provides unique data for each country on how core values are communicated to central government employees. Source: OECD Survey on Integrity (2000 and 2009). Figure notes: Data are not available for Greece, Korea, the Slovak Republic, Spain and the United Kingdom. Further reading: OECD (2000), Trust in Government: Ethics Measures in OECD Countries, OECD, Paris. Wal, Z van Der and Huberts L.W.J.C (2008), “Value Solidity in Government and Business: Results for an Empirical Study on Public and Private Sector Organisational Values,” American Review of Public Administration, 38, pp. 264-285. 22 GOV/PGC(2009)2 Figure II.4.1 Frequently stated core public service values (2000 and 2009) Figure II.4.2 How countries communicate core values to central government public servants (2000 and 2009) 23 GOV/PGC(2009)2 III. GOVERNMENT REVENUES 57. The structure of revenues collected by governments (e.g. the proportion of revenues collected as social contributions and other taxes, grants, fees, etc.) determine who pays for public services and goods. Further, it affects the incentives and constraints that governments face when deciding what goods and services to produce and how to produce them. 58. This chapter compares the size and structure of government revenues across OECD countries, as well as how revenues are collected and shared across levels of government within each country. Differences in the amount and structure of revenues between countries reflect policy choices that have impacts on government efficiency, equity and effectiveness. Examining these differences can help policy makers determine the effects of reforms and identify alternative models of raising revenues. 59. It is important to note, however, that revenues in any given year do not necessarily match expenditures in that year. For example, revenues collected in 2009 may finance current, past (through interest payments on debt) and/or future (through budget surpluses) expenditures. In addition, OECD countries that are also members of the European Union have limits on the size of deficits and debts, as well as requirements for a value-added tax. These limitations and requirements may affect the amount and structure of revenues. 24 GOV/PGC(2009)2 III.1. GENERAL GOVERNMENT REVENUES 60. Governments must raise revenues to pay for their policies and programmes. Thus, the amount of revenues collected by countries is related to political decisions regarding the goods and services governments provide and the way that they are produced. For example, if governments provide support via tax breaks, revenue to GDP ratios will be lower. In addition, OECD countries that are members of the European Union have restrictions on the level of deficits and debt that may affect the size of revenues in any given year. 61. The size of government revenues varies greatly across OECD countries, comprising about a third of GDP in Korea to almost 60% of GDP in Norway. When grouped together, Nordic countries collect on average 10 percentage points more revenues per GDP than any other country group. As a group, Western European countries display the most variance in revenues as a share of GDP. Government revenues as a percentage of GDP in France and Belgium compare with Nordic countries, while revenues as a share of GDP in Luxembourg and Switzerland are more similar to those in Anglo-Saxon countries. 62. Between 1995 and 2006, government revenue as a share of GDP slightly increased for most OECD countries. When comparing countries according to government revenues as a share of GDP, country positions have remained fairly constant between 1995 and 2006. Government revenues as a share of GDP in Nordic countries were also high in 1995 when compared with other OECD countries. Likewise, government revenues comprise a relatively lower share of GDP in both 1995 and 2006 in the United States, Switzerland and Australia. While Korea still has the lowest government revenue to GDP ratio of all OECD countries, government revenues grew at a substantially faster pace (1.5 times) than GDP between 1995 and 2006, mostly due to changes implemented to the social security system. 63. The amount of revenues collected per person is an alternative way of comparing the size of government revenues across OECD countries, and provides a rough assessment of the need for public services, since every person is a potential patient, student or customer. Luxembourg and Norway are clear outliers, with government revenues greater than $30,000 per person in 2006. Nordic countries tend to collect above average government revenues per capita, while Central European countries collect comparatively less revenues per person. Only two countries (Czech Republic and Korea) experienced real annual growth in government revenues greater than five percent between 2000 and 2006. On average, per capita government revenues grew by about two percent a year in OECD countries during this period, while real government revenues remained the same in Germany. Methodology and Definitions 64. Government revenue data are derived from the OECD National Accounts Statistics and represent government revenues in 2006, the latest data available. OECD National Account Statistics are based on the System of National of Accounts (SNA), a set of internationally agreed concepts, definitions, classifications and rules for national accounting. Using SNA terminology, General Government revenue consists of central government, state government, local government and social security funds. Revenues encompass social contributions, taxes other than social contributions and grants and other revenues. Gross domestic product (GDP) is the standard measure of the value of goods and services produced by a country during a period. 65. Government revenues per capita were calculated by converting total revenues to 2006 US dollars using Purchasing Power Parities (PPP) for GDP and dividing by population. PPP is the number of units of country B’s currency needed to purchase the same amount of goods and services in country A. The annual real percentage change was calculated using a deflator for GDP and a base year of 2000 US dollars. 25 GOV/PGC(2009)2 Source: OECD National Account Statistics and OECD Population Statistics. Figure notes: (1) Data for New Zealand are from 2005. Data are not available for Mexico and Turkey. (2) Data for New Zealand are from 1995 and 2005. Data for Mexico are for 1995 and 2004. Data are not available for Greece, Hungary, Iceland, Japan and Turkey. There are small differences between figures III.1.1 and III.1.2 in revenues as a percentage of GDP in 2006 for the Slovak Republic due to small revisions in GDP. (3&4) Data for New Zealand are for 2005 and data for Mexico are for 2004. Data are not available for Turkey. Figure III.1.1 General government revenues as a share of GDP by country groups (2006) 26 GOV/PGC(2009)2 Figure III.1.2 General government revenues as a share of GDP (1995 and 2006) 27 GOV/PGC(2009)2 Figure III.1.3 Revenue per capita (2006) 28 GOV/PGC(2009)2 Figure III.1.4 Annual real percentage change in revenue per capita (from 2000 to 2006) 29 GOV/PGC(2009)2 III.2. STRUCTURE OF GENERAL GOVERNMENT REVENUES 66. A country’s revenue structure determines who pays for public services and goods, and by spreading revenues across different instruments, countries can distribute the burden across particular groups of citizens and/or sectors of the economy. 67. In all OECD countries, taxes other than social contributions represent the largest share of government revenues and this share has increased over the past decade. The proportion of general revenues financed by taxes other than social contributions increased in all but four countries between 1995 and 2006. In some cases, such as Italy, this is due to fiscal reform and the introduction of new taxes which increased total tax collections. In all OECD countries, grants and other revenues (such as fees and sales of natural resources) are the least common sources of revenue, usually representing between 10% and 15% of total revenues. Norway raises over 25% of revenues from other sources, mostly from the sale of oil resources. 68. While economic research suggests that the cumulative effect of taxes on economic growth is moderate, recent research has suggested that a relationship between the types of taxes imposed and economic growth exists. In general, property taxes seem to be the most growth-friendly followed by consumption taxes and then by personal income taxes. Methodology and Definitions 69. Revenue data are derived from the OECD National Accounts Statistics and represent government revenues in 2006, the latest data available. OECD National Account statistics are based on the System of National of Accounts (SNA), a set of internationally agreed concepts, definitions, classifications and rules for national accounting. Using SNA terminology, General Government revenue consists of central government, state government, local government and social security funds. Revenues encompass social contributions (e.g. contributions for pensions, health and social security); taxes other than social contributions (e.g. taxes on consumption, income, wealth, property and capital); and grants and other revenues. Grants can be from foreign governments, international organizations or other general government units. Other revenues include sales, fees, property income and subsidies. The aggregates presented (taxes other than social contributions, social contributions and grants and other revenues) are not directly available in the OECD National Accounts, and were constructed using sub-account line items. The data presented in figure III.2.4 are from OECD Revenue Statistics. 70. More information on the types of taxes levied by governments is available in OECD (2008) Revenue Statistics 1995-2007, OECD, Paris. 71. NOTE: Discussions are still in progress to identify the most appropriate data sources and methodology to present the structure of general government revenues. Source: OECD National Accounts Statistics and OECD Revenue Statistics. Figure notes: Australia does not collect revenues via social contributions because it does not operate government social insurance schemes. (1,2 &4) Data for New Zealand are from 2005. Data are not available for Canada, Mexico and Turkey. (3) Data for New Zealand are from 1995 and 2005. Data are missing for Canada, Greece, Hungary, Iceland, Japan, Korea, Mexico and Turkey. Further readings: Arnold, J. (2008), "Do Tax Structures Affect Aggregate Economic Growth?: Empirical Evidence from a Panel of OECD Countries", OECD Economics Department Working Papers, No. 643, OECD, Paris. 30 GOV/PGC(2009)2 Myles, G. (2008), “Economic Growth and the Role of Taxation”, report for the OECD, available at www.people.ex.ac.uk/gdmyles/papers/pdfs/OECDfin.pdf Figure III.2.1 Structure of General Government revenues as a share of GDP (2006) Figure III.2.2 Structure of General Government revenues (2006) 31 GOV/PGC(2009)2 Figure III.2.3 Structure of General Government revenues (1995 and 2006) Figure III.2.4 Tax structure of General Government (2006) 32 GOV/PGC(2009)2 III.3. REVENUE STRUCTURE BY LEVELS OF GOVERNMENT 72. Revenue structures and transfers between government levels illustrate the degree of fiscal autonomy exercised by sub-central governments and their ability to shape public service delivery and public policy. These abilities are also affected by a country’s institutional structure; federal states share sovereignty with sub-central governments who, consequently, may have more power to shape public policies. 73. In most OECD countries, the majority of revenues are collected by the central government. Taxes other than social contributions are the main source of financing for central government budgets and generally represents between 80 and 90% of revenues. In almost all countries, grants and other revenues form the second largest source of central government revenues, comprising around 10% of finances. Most social contributions finance social security funds and therefore form a small proportion of revenues collected by central and sub-central governments. 74. In contrast to the relative homogeneity of central government revenue structures, the discretion over fiscal resources available to state and local governments varies considerably. Limits on sub-central governments’ ability to set their own local tax bases, rates and reliefs reduce their power to generate their own revenue sources and potentially their ability to provide more tailored public services. Tax autonomy for sub-central governments can vary from full power over tax rates and bases to none at all. Of the OECD countries with federal systems, German states raise the most revenues via taxes other than social contributions, which represent over 70% of state finances. In contrast, Belgian states collect less than 20% of their revenues from taxes other than social contributions. Local governments’ ability to raise revenues through taxes also varies considerably; in Iceland, taxes comprise over 70% of local government revenues whereas Greek localities only receive about 10% of their revenues in the form of taxes. 75. Intergovernmental grants and other revenues are the predominant feature of local government finances in most countries, representing over 50% of revenues on average. Intergovernmental grants can be earmarked (tied to specific uses) or general, and can be used by central governments to compensate for a weak tax base in some localities, to avoid tax competition between local governments and/or to ensure territorial cohesion in policies and services. Methodology and Definitions 76. Revenue data are derived from the OECD National Accounts Statistics and represent government revenues in 2006, the latest data available. OECD National Account statistics are based on the System of National of Accounts (SNA), a set of internationally agreed concepts, definitions, classifications and rules for national accounting. Using SNA terminology, General Government revenue consists of central, state and local governments, and social security funds. State government is only applicable to the nine OECD countries that are federal states: Australia, Austria, Belgium, Canada, Germany, Mexico, Spain (considered a de facto federal state in the National Accounts data), Switzerland and the United States. To calculate the share in General Government revenues by level of government, transfers between the levels were excluded. Revenues encompass social contributions (e.g. contributions for pensions, health and social security); taxes other than social contributions (e.g. taxes on consumption, income, wealth, property and capital); and grants and other revenues. Grants can be from foreign governments, international organizations or other general government units. Other revenues include sales, fees, property income and subsidies. The aggregates presented (taxes other than social contributions, social contributions and grants and other revenues) are not directly available in the OECD National Accounts, and were constructed using subaccount line items. Source: OECD National Accounts Statistics. Data for Australia were provided by government officials. 33 GOV/PGC(2009)2 Figure notes: Data for New Zealand are from 2005. (1) Data are not available for Australia, Mexico and Turkey. Breakdown between state and local governments is not available for the United States. Social security funds are included in central government revenues in New Zealand, Norway, the United Kingdom and the United States. (2) Data are not available for Canada, Mexico and Turkey. (3) Data are not available for Canada and Mexico. (4) Data are not available for Canada, Mexico, Turkey and the United States. Figure III.3.1 Distribution of General Government revenues across levels of government (2006) 34 GOV/PGC(2009)2 Figure III.3.2 Structure of central government revenues (2006) Figure III.3.3 Structure of state government revenues (2006) 35 GOV/PGC(2009)2 Figure III.3.4 Structure of local government revenues (2006) 36 GOV/PGC(2009)2 IV. GOVERNMENT EXPENDITURES 77. The size and structure of public expenditures—both across sectors and levels of government— reflect the policy choices of governments. 78. Governments spend money on producing and purchasing goods and services (e.g. defence, education and health care) and on redistribution programs (e.g. pensions and unemployment insurance). After comparing the size of government expenditures across countries, this Chapter analyzes how much of public budgets are spent on various sectors. It also disaggregates expenditures by level of government, permitting comparisons of how responsibilities for providing different goods and services vary across countries. For example, in some countries, education is mainly the responsibility of central governments, whereas in others, local governments play a larger role in providing public schooling. 79. Finally, this Chapter presents detailed breakdowns in expenditures for 13 countries that allow for further analysis of economic and social policies. Distinguishing between government expenditures on collective goods (i.e. those that benefit the community at large, such as defence and regulation) and individual goods (i.e. those that benefit primarily individual households, such as pensions and health care) can help illustrate differences in government’s role in society. In addition, looking more closely at whether income is redistributed through cash or in kind benefits allows countries to compare their economic and social arrangements. 37 GOV/PGC(2009)2 IV.1. GENERAL GOVERNMENT EXPENDITURE 80. Government expenditures as a share of GDP and expenditures per capita indicate the size of the government and reflect political decisions about its role—both in providing services and in redistributing income. However, a large part of the variation in the government expenditure per GDP ratios across countries reflects the different approaches to delivering goods and services and giving social support rather than true differences in resources spent on public services. For instance, if support is given via tax breaks rather than direct expenditure, expenditure/GDP ratios will naturally be lower. In addition, it is important to note that the size of expenditures does not reflect government efficiency or productivity. Finally, OECD countries that are members of European Union have limitations on the amount that expenditures can exceed revenues in any given year. 81. In Nordic countries, government expenditures represent about 50% of the economy. In comparison, in Asian OECD countries, expenditures represent about 35% of GDP. On average, Western, Central and Southern European countries spend similar amounts in proportion to GDP. However, there is considerable variation in public spending relative to GDP within country groups. For example, Sweden, France and Hungary have relatively high levels of government spending (over 50% of GDP), whereas in Norway, Switzerland and the Slovak Republic, public expenditures range between 30 and 40% of GDP. 82. Since 1995, the size of government spending relative to GDP has converged in OECD countries. Whereas government expenditures ranged from about 20 to 65% of GDP in 1995, today spending comprises between 30 and 55% of GDP in OECD countries. Government expenditures as a share of GDP declined in all but two countries between 1995 to 2006. In all cases, this decline can be attributed to an increase in GDP rather than a contraction in general government spending in an absolute sense. Government expenditures as a share of GDP rose slightly in Portugal and by a larger margin in Korea during this period. Despite the major increase in government expenditures as share of GDP in Korea since 1995, expenditures relative to GDP in 2006 remain low in Korea when compared to other countries. 83. The range in government expenditures per capita is large; Luxembourg spent almost ten times as much per person than Mexico on public services and goods in 2006. Expenditures per capita have risen in all OECD countries since 2000, but have grown the fastest in Korea and the Central European countries of Hungary, the Czech Republic and Poland. Methodology and Definitions 84. Data are derived from the OECD National Accounts Statistics and represent government expenditures in 2006, the latest data available. OECD National Account statistics are based on the System of National Accounts (SNA), a set of internationally agreed concepts, definitions, classifications and rules for national accounting. In SNA terminology, General Government consists of central, state and local governments and social security funds. Gross domestic product (GDP) is the standard measure of the value of the goods and services produced by a country during a period. 85. Government expenditures per capita were calculated by converting total public expenditures to 2006 US dollars using Purchasing Power Parities (PPP) for GDP and dividing by population. PPP is the number of units of country B’s currency needed to purchase the same quantity of goods and services in country A. The annual real percentage change was calculated using a deflator for GDP and a base year of 2000 US dollars. Source: OECD National Account Statistics and OECD Population Statistics. Figure notes: Data for New Zealand are from 2005 and data for Mexico are from 2004. Data are not available for Turkey. 38 GOV/PGC(2009)2 (2) In addition to Turkey, time series data are not available for Greece, Hungary, Iceland and Japan. Figure IV.1.1 General Government expenditures as a percentage of GDP (2006) Figure IV.1.2 General government expenditures as a percentage of GDP (1995 and 2006) 39 GOV/PGC(2009)2 Figure IV.1.3 Government expenditures per capita (2006) 40 GOV/PGC(2009)2 Figure IV.1.4 Annual real percentage change of government expenditures per capita (from 2000 to 2006) 41 GOV/PGC(2009)2 IV.2. GENERAL GOVERNMENT EXPENDITURE BY SUB-SECTOR 86. The structure of government expenditures highlights the relative share of public resources spent on various goods and services, which reflect different political choices regarding the role of government in each country. In addition, for OECD countries that are members of the European Union, common policy goals regarding economic growth, agriculture, energy, infrastructure and research and development (among others) may affect the structure of expenditures. 87. The variance among countries in expenditures as a percentage of GDP is mainly explained by political differences regarding the role of government in providing social protections such as unemployment insurance, old age pensions and disability benefits. When government spending on social protection is excluded, expenditures range between 20 and 30% of GDP in all countries. In all but three countries, social protection is the largest category of expenses; Korea spends the most on economic affairs whereas the United States and Iceland spend more on health than any other government function. Aside from expenditures on social protection, OECD countries spend the largest percent of GDP on health, education and general public services (which includes public debt payments). Apart from Korea, social protection, health and education comprise 55 to 70% of all government expenditures; the average OECD country spends close to 60% of total expenditures on providing these services. Defence spending as a share of GDP is notably high in the United States, Korea, United Kingdom and Greece compared to other OECD countries. In general, OECD countries spend the least amount of government financial resources on environment protection and housing and community amenities. 88. Total expenditures as a share of GDP decreased in almost all countries over the past decade, and were largely driven by overall decreases in expenditures on general public services. The share of resources devoted to different activities has also shifted. OECD countries today spend a larger proportion of resources on social protection and health than in 1995. In most cases, the proportional increases in funds spent on social protection and health were balanced by proportional decreases in funds spent on general public services. In some countries, such as Italy, expenditures on general public services decreased due to reductions in interest payments on debt. The large decrease in expenditures on housing and community amenities in the Netherlands and on economic affairs in the Czech Republic and Germany between 1995 and 2006 are due to large, one-off capital expenditures that occurred in 1995. Methodology and Definitions 89. Data represent government expenditures in 2006, the latest data available. Data on expenditures come from the System of National Accounts (SNA) and are disaggregated according to the Classification of the Functions of Government (COFOG), which divides government spending into 10 purposes: general public services, defence, public order and safety, economic affairs, environment protection, housing and community amenities, health, recreation; culture and religion, education and social protection. Further information about the types of expenditures included in each category is available in Annex A. SNA is a set of internationally agreed concepts, definitions, classifications and rules for national accounting. In SNA terminology, general government consists of central, state and local governments and social security funds. State government is only applicable to the nine OECD countries that are federal states: Australia, Austria, Belgium, Canada, Germany, Mexico, Spain (considered a de facto federal state in the National Accounts data), Switzerland and the United States. Gross domestic product (GDP) is the standard measure of the value of the goods and services produced by a country during a period. Source: OECD National Account Statistics. Figure notes: Data for New Zealand, Norway and the United Kingdom are from 2005 and data for Mexico are from 2004. (1 and 3) Data are not available for Australia, Mexico, Switzerland and Turkey. 42 GOV/PGC(2009)2 (2 and 4) Time series data are not available for Australia, Hungary, Iceland, Japan, Mexico, New Zealand, Poland, Slovak Republic, Switzerland and Turkey. Figure IV.2.1 General Government expenditures by sub-sector as a share of GDP (2006) 43 GOV/PGC(2009)2 Figure IV.2.2 General Government expenditures by sub-sector as a share of GDP (1995 and 2006) Figure IV.2.3 Structure of General Government expenditures by sub-sector (2006) 44 GOV/PGC(2009)2 Figure IV.2.4 Structure of General Government expenditures by sub-sector (1995 and 2006) 45 GOV/PGC(2009)2 IV.3. GERNAL GOVERNMENT EXPENDITURE BY LEVELS OF GOVERNMENT 90. Analysis of expenditures by levels of government illustrates the relative responsibilities and policy priorities of central and sub-central administrations. 91. While central governments on average spend the largest proportion of total government resources, the level of fiscal decentralisation varies across OECD countries. In New Zealand, the central government accounts for 90% of all spending, suggesting a high level of centralization—although goods and services may be delivered locally, they are paid for by the central government. In contrast, the central government accounts for less than 15% of total expenditures in Switzerland, where state and local governments play a much larger role in financing goods and services directly. 92. In general, central governments spend relatively larger proportions of their budgets on general public services, social protection and defence than sub-central governments. Expenditures on social protection represent the largest proportion of central government budgets for about half of OECD countries. The central governments of Spain and Belgium are the most oriented towards general public services, where it amounts to over 50% of total expenditures. Although defence is predominately the responsibility of central governments in OECD countries, it averages less than six percent of total central government expenditures. 93. Education, recreation, environmental protection and housing and community amenities are generally the responsibility of sub-central governments, and comprise larger proportions of state and local spending relative to central government expenditures. For the six countries displayed, the fiscal responsibilities of state governments vary greatly. Education consumes the largest proportion of state expenditures in the United States and Canada, whereas spending on social protection is the predominate feature of state budgets in Germany and Austria. 94. Three-fourths of local governments allocate over 50% of all expenditures for health, education, social protection, housing and recreation. Local governments in Greece, Portugal and Spain spend notably more on general public services and local governments in Italy allocate much more resources to health (approximately 40% of total expenditures) than in other OECD countries. The Slovak Republic is the only OECD country that has significant defence expenditures by the sub-central government. Methodology and Definitions 95. Data represent government expenditures in 2006, the latest data available. Data on expenditures come from the System of National Accounts (SNA) and are aggregated according to the Classification of the Functions of Government (COFOG), which divides government spending into 10 purposes: general public services, defence, public order and safety, economic affairs, environment protection, housing and community amenities, health, recreation; culture and religion, education and social protection. Further information about the types of expenditures included in each category is available in Annex A. SNA is a set of internationally agreed concepts, definitions, classifications and rules for national accounting. In SNA terminology, general government consists of central, state and local governments and social security funds. State government is only applicable to the nine OECD countries that are federal states: Australia, Austria, Belgium, Canada, Germany, Mexico, Spain (considered a de facto federal state in the National Accounts data), Switzerland and the United States. 96. Data in Figure IV.3.1 exclude transfers between levels of government and thus provide a rough proxy of the overall fiscal burden for providing goods and services borne by each level of government. However, data on expenditures at the central and state levels (Figures IV.3.2 and IV.3.3) include transfers 46 GOV/PGC(2009)2 to local governments in addition to expenditures on goods and services, and therefore only illustrate how much is spent on each sector at each level of government. Source: OECD National Account Statistics. Figure notes: Data for New Zealand, Norway and the United Kingdom are from 2005. (1) Data are not available for Australia, Mexico and Turkey. For the United States, no breakdown between state and local governments is available. (2) Data are not available for Australia, Japan, Mexico, Switzerland and Turkey. (3) Data on state government expenditures are only relevant for OECD countries that have federal structures. Data are not available for Australia, Mexico and Switzerland. (4) Data are not available for Australia, Japan, Mexico, Switzerland, Turkey and the United States. Figure IV.3.1 Distribution of General Government expenditures by level of government (2006) 47 GOV/PGC(2009)2 Figure IV.3.2 Structure of central government expenditures by sub-sector (2006) Figure IV.3.3 Structure of state government expenditures by sub-sector (2006) 48 GOV/PGC(2009)2 Figure IV.3.4 Structure of local government expenditures by sub-sector (2006) 49 GOV/PGC(2009)2 IV.4. GENERAL GOVERNMENT EXPENDITURES BY COLLECTIVE AND INDIVIDUAL GOODS AND SERVICES, AND CASH TRANSFERS AND GOODS AND SERVICES IN KIND 97. Comparisons across countries of the proportion of benefits provided in cash by governments to eligible parties can illustrate differences in economic and social policies. Particularly within redistribution programs (such as unemployment and health assistance), governments can provide benefits in cash (e.g. pensions) or in kind (e.g. food stamps or housing vouchers). OECD countries spend more money on in- kind goods and services than cash transfers. While cash transfers generally have lower transaction costs, larger multiplier effects in the economy and provide individuals with more choice, governments may prefer in kind transfers because it may be hard to identify eligible individuals, they want to control the delivery process, and/or they want to ensure that individuals receive adequate food, medical services or housing. 98. In the 13 countries represented, cash transfers range between 30 and 50% of all expenditures, or between 10 and 20% of GDP. Germany, Greece, Austria and Italy provide the largest proportion of expenditures as cash transfers (over 40% of all expenditures are cash transfers in these countries). When scaled to GDP, cash transfers represent over 20% of GDP in Italy, Austria and Germany. Cash transfers are least popular in the Czech Republic, the United Kingdom and Spain. 99. Government expenditures can also be classified into spending on collective goods (a good that benefits society as a whole, such as defence and public order and safety) or spending on individual goods (a good that primarily benefits individual citizens, such as social services, health and education). Analysis of the proportional amounts that countries spend on collective vs. individual goods and services can help illustrate differences in the role of government in each country—countries that spend a relatively larger proportion of funds on individual goods tend to be considered “welfare states.” 100. In all 13 countries for which data are available, governments spend more resources on individual goods than on collective goods. On average, countries spend between 25 and 35% of GDP on individual goods, compared to 10-20% of GDP on collective goods. Only Hungary spends more than 20% of GDP (or approximately 40% of all expenditures) on collective goods. In comparison, Norway spends just over 10 % of GDP (or about 26% of all expenditures) on collective goods. Methodology and Definitions 101. Cash transfers refer to benefits provided to eligible individuals by governments that are not required to be spent on a specific good or service. Examples of cash transfers include pensions, unemployment benefits and development aid. In kind goods and services mean that the government provides (or contracts for the provision of) goods and services directly or reimburses households for their expenses. Examples of in kind goods and services include housing vouchers, police and most health and education services. In this context, in kind goods and services do not correspond one-to-one to the category of expenditures social transfers in kind contained in the National Accounts data, but are more expansive. Table 2 in Annex A provides more details about this relationship. Collective goods and services benefit the community at large and include expenditures on defence and public safety and order. Individual goods and services mainly benefit individuals and include education, health and social insurance programs. Gross domestic product (GDP) is the standard measure of the value of the goods and services produced by a country during a period. 102. Data represent government expenditures in 2006, the latest data available. Based on a methodology devised by OECD called Classification of the Functions of Government (COFOG)-Special, data were broken down into individual and collective goods, and cash and in kind transfers using secondlevel COFOG classifications which are currently only available for 13 OECD European member countries. See Annex A for more detail on the COFOG-Special methodology. 50 GOV/PGC(2009)2 Source: OECD National Account Statistics. Figure notes: Data are not available for Australia, Belgium, Canada, Denmark, France, Iceland, Ireland, Japan, Korea, Luxembourg, Mexico, Netherlands New Zealand, Slovak Republic, Switzerland, Turkey and the United States. IV.4.1 General government expenditures on cash transfers and goods and services in-kind as a percentage of GDP (2006) Figure IV.4.2 General Government expenditures on individual and collective goods as a percentage of GDP (2006) 51 GOV/PGC(2009)2 V. THE COST OF GOVERNMENT PRODUCED AND GOVERNMENT FUNDED GOODS AND SERVICES 103. Governments are concerned with their capacity to produce good quality goods and services at the lowest possible cost. Over recent decades, many governments have contracted with non-profit and forprofit private entities to produce goods and services that they formerly produced themselves. Outsourcing has affected most government sectors, including the development of weapons, care for the elderly, prisons and the delivery of foreign aid. 104. This chapter measures the total costs of producing goods and services for government, including those that are produced by government entities and those that are outsourced to non-governmental or private entities. The data allow countries to compare their relative use of outsourcing, and show trends over time. However, production cost data do not indicate the quantity or quality of goods and services produced and therefore are not an indication of efficiency or productivity. 105. Production costs are a subset of total government expenditures. Compared to total government expenditures, production costs of public services and goods exclude government investment (other than depreciation costs), interest paid on government debt and payments made to citizens and others not in exchange for the production of goods and services (such as subsidies or social benefits like unemployment insurance, family allowances and retirement pensions). 52 GOV/PGC(2009)2 V.1 PRODUCTION COSTS IN GENERAL GOVERNMENT 106. Decisions about the amount and type of services and goods to produce as well as how best to produce them are often political in nature and based on countries’ social and cultural context. While some governments choose to outsource a large portion of the production of goods and services to nongovernmental or private entities, others choose to produce the goods and services themselves. 107. The proportion of the economy devoted to producing public services and goods varies greatly across OECD members. For example, production costs as a percentage of GDP in Sweden are over double those in Mexico. The Netherlands, the United Kingdom, Germany and Japan rely comparatively more on private entities to produce goods and services than other OECD countries. Outsourcing is used to a much lesser extent in Mexico and Greece, where over 60% of the value of public services and goods are produced by the government. 108. Total production costs as a percentage of GDP decreased in most OECD countries between 1995 and 2007, although this could be attributed to increases in GDP rather than actual decreases in production costs. In countries where production costs as a percentage of GDP rose, the growth was primarily driven by increases in the costs of goods and services produced by private entities. In some countries, such as the Netherlands and Spain, these increases took place in parallel to a relative decrease in the proportion of goods and services produced by government employees, suggesting that some goods and services previously produced by government employees may have been outsourced to private sector entities. In other countries, such as the United Kingdom, Portugal and New Zealand, these increases occurred while the compensation costs of government employees remained relatively stable. Methodology and Definitions 109. The data use a methodology developed by the OECD Public Employment and Management Working Party that builds on the existing classification of organisations in the System of National Accounts (SNA). Specifically, government production costs include: • Compensation costs of General Government employees, including cash and in kind remuneration plus all mandatory employer contributions to social insurance and voluntary contributions paid on behalf of employees. Cross-country differences in how government employee pension schemes are funded can distort comparisons of compensation costs. • Costs of goods and services produced by private entities funded by government (these include goods and services provided to both the government and citizens). In SNA terms, this includes intermediate consumption (procurement of intermediate products required for government production such as accounting or IT services, including some relatively minor procurement between government-owned organisations) and social transfers in kind via market producers (a proxy for goods and services delivered by private actors directly to citizens, including those that are initially paid for by citizens who are ultimately refunded by government such as medical treatments refunded by public social security payments). • Consumption of fixed capital (indicating the level of depreciation of capital). This was originally excluded from the methodology published in OECD Working Paper on Public Government No. 8 and in The State of the Public Service. 110. The data include government employment and intermediate consumption for output produced by the government for its own use, such as roads and other capital investment projects built by government employees. 53 GOV/PGC(2009)2 Source: OECD National Accounts Statistics. Figure notes Data for Mexico are for 2004. Data for New Zealand are for 2005. Data for Switzerland are for 2006. (1 and 3) Data are not available for Australia and Turkey. (2) Data are not available for Australia, Austria, Greece, Hungary, Iceland, Japan, Korea and Turkey. Further readings: Pilichowski, E. and E. Turkisch (2008), “Employment in Government in the Perspective of the Production Costs of Goods and Services in the Public Domain,” OECD Working Papers on Public Governance, 8, OECD, Paris. OECD (2008), The State of the Public Service, OECD, Paris. Figure V.1.1 Production costs as a percentage of GDP (2007) 54 GOV/PGC(2009)2 Figure V.1.2 Production costs as a percentage of GDP (1995 and 2007) Figure V.1.3 Structure of production costs (2007) 55 GOV/PGC(2009)2 VI. EMPLOYMENT IN THE PUBLIC SECTOR 111. Labour is the most important input in the public sector production process and is a key factor in government productivity. Through its size and human resource management (HRM) practices and policies, government employment affects the functioning of the wider labour market, influencing wage levels in the economy; general working conditions; the supply and demand for various occupations, such as teachers and health care professionals; and the skill levels required in the various economic sectors. 112. This chapter compares the size of employment in the government and in public and quasi-public corporations across OECD countries (which together define the public sector), and looks at the proportion of total government staff employed at the central and sub-central levels. This information allows countries to compare the relative size and composition of their public sector staff, providing insight on the scope, level and delivery method of public services across countries and the relative decentralization of government employment. 56 GOV/PGC(2009)2 VI.1. EMPLOYMENT IN GENERAL GOVERNMENT AND PUBLIC AND QUASI-PUBLIC CORPORATIONS 113. Large differences in government employment reflect choices regarding the scope, level and delivery method of public services and, for some countries, differences in productivity. In terms of delivery methods, some countries prefer the work of government employees, while others choose to contract with the private sector. As a result, government employment numbers should be interpreted in the perspective of the costs of goods and services funded by government but produced by the private sector, a topic discussed in Chapter V. 114. The proportion of the labour force employed in government ranges from just over 5% in Japan and Korea to nearly 30% in Norway and Sweden. Since 1995, the proportion of the labour force working for the government has been relatively stable in most of the 15 countries for which trend data are available. Only Spain experienced a notable increase in the share of the labour force employed by the government. While the total number of government employees decreased in Austria, it increased in Canada and Sweden but was outpaced by increases in the labour force. 115. Public and quasi-public corporations can be a transitory stage towards a more privatized management mode of tasks. In some countries, the management of employees in these corporations is less flexible than in the private sector due to legal or historical reasons, or the political visibility of industrial relations. Except for a few countries, in particular the Netherlands and France, employment in (quasi) public corporations is a relatively minor part of the labour force. From 1995 to 2005, the share of the labour force employed in the public sector (government and (quasi) public corporations) declined in seven of the nine countries for which data are available, with the Netherlands and Spain being the two exceptions. Methodology and Definitions 116. Data refer to 2005 and were collected by the OECD 2006 Comparison of Employment in the Public Domain (CEPD) Survey. Respondents to the survey were predominately national statistical officers. Data for the United Kingdom were subsequently provided by government officials. 117. The data are based on System of National Accounts (SNA) definitions, and cover employment in General Government and public and quasi-public corporations. The General Government sector (referred to as “government” above) comprises all levels of government (central, state, local and social security) and includes core ministries, departments and agencies and non-profit institutions controlled by government. Public and quasi-public corporations are legal entities that are owned or controlled by the government (i.e. government has the right to control over 50% of the entity) and produce goods and services for sale in the market at economically significant prices. Countries that do not compile a specific sub-sector “public corporations/quasi-corporations” in the corporations sector of the National Accounts could provide data based on a pre-existing inventory of public corporations. 118. Data represent the number of employees, except for Austria, Netherlands, Sweden, Switzerland and the United Kingdom, where data represent full-time equivalents. As a result, employment numbers for these four countries are understated. The labour force comprises all persons who fulfil the requirements for inclusion among the employed or the unemployed. Source: OECD CEPD and Labour Force Survey (2006). Figure notes: Data for Austria do not include non-profit institutions financed by government, and data for employment in social security are missing for 1995. Data for France exclude some Public Establishments. Data for the Slovak Republic refer to ISIC categories L, M and N and assume that private institutions finance by General Government are of marginal importance. Data for 57 GOV/PGC(2009)2 Belgium, France and Poland are for 2004. Data for Austria and Finland are a mix of 2004 and 2005. Data for Mexico are for 2000. Data for Greece are for 2006 and include staff under private law. Data are not available for the Czech Republic, Denmark, Iceland, Ireland, Luxembourg and New Zealand. (2) Data are not available for Greece, Hungary, Japan, Mexico, Norway, Poland, the Slovak Republic and the United States. (3) Data are not available for Finland, Greece, Hungary, Italy, Luxembourg, Norway, Poland, Portugal, the Slovak Republic and Switzerland. (4) Data are not available for Austria, Belgium, Finland, Greece, Hungary, Italy, Japan, Mexico, Norway, Poland, Portugal, the Slovak Republic, Switzerland and the United States. Further reading: Pilichowski, E. and E. Turkisch (2008), “Employment in Government in the Perspective of the Production Costs of Goods and Services in the Public Domain,” OECD Working Papers on Public Governance, 8, OECD, Paris. OECD (2008), The State of the Public Service, OECD, Paris. Figure VI.1.1 Employment in General Government as a percentage of the labour force (2005) 58 GOV/PGC(2009)2 Figure VI.1.2 Employment in General Government as a percentage of the labour force (1995 and 2005) Figure VI.1.3 Employment in General Government and (quasi) public corporations as a percentage of the labour force (2005) 59 GOV/PGC(2009)2 Figure VI.1.4 Employment in General Government and (quasi) public corporations as a percentage of the labour force (1995 and 2005) 60 GOV/PGC(2009)2 VI.2. DECENTRALIZATION OF EMPLOYMENT 119. The proportion of government staff employed in sub-central levels of government is a potential indicator of the level of decentralization of public administration and the level of responsibility accorded to state and/or local governments. In general, the larger the staff of sub-central governments, the more responsibilities these governments would be expected to have for providing public services. While decentralization is generally thought to improve the abilities of governments to respond to local needs, conditions and priorities, it can lead to differences between central and sub-central government interests and human resource management practices, as well as variations in the efficiency of service delivery within countries. 120. Of the 18 countries for which data are available, the vast majority have more employees at the sub-central level than at the central level. In general, federal states have less than one-third of all government employees at the central level, confirming higher levels of decentralization. The range in the proportion of government employees at the central level of government is much wider for unitary states, from 15% in Sweden to almost 90% in Turkey. 121. For countries with data available, the proportion of government employees at the central and subcentral levels has been relatively stable between 1995 and 2005. Norway stands out as having experienced a relative re-centralisation of staff, whereas Spain experienced a relative decentralisation of staff. Methodology and Definitions 122. Data refer to 2005 and were collected by OECD’s 2006 Comparison of Employment in the Public Domain (CEPD) Survey. Respondents to the survey Respondents to the survey were predominately national statistical officers. Data for the United Kingdom were subsequently provided by government officials. 123. The data represent employment in central and sub-central government-owned or -controlled organisations. In System of National Accounts (SNA) terminology, they cover employment in General Government. The General Government sector includes core ministries, departments and agencies, and nonprofit institutions controlled by government. Sub-central government comprises state and local government including regions, provinces and municipalities. Most countries did not provide data on employment in social security. The following countries provided data on employment in social security, which were included in employment at the sub-central level: Austria, Belgium, Finland, France, Japan, Korea, Netherlands, Spain, Sweden and Turkey. Social security represents a small number of employees and is of minor importance as a percentage of the total government labour force. Source: OECD CEPD Survey (2006). Figure notes: Data for Austria do not include private non-profit institutions financed by government and data for public corporations are partial and only include universities that have been reclassified. Data for France exclude some Public Establishments and all teachers are central government employees.. Korean data include teachers and police officers at the central level (which account for 75% of the central government workforce). Data for Austria and Finland are a mix of 2004 and 2005. Data for Belgium and France are for 2004. (1) Data are not available for the Czech Republic, Denmark, Greece, Iceland, Ireland, Luxembourg, Mexico, New Zealand, Poland, the Slovak Republic and Switzerland. (2) Data are not available for Australia, the Czech Republic, Denmark, Germany, Greece, Iceland, Ireland, Luxembourg, Mexico, New Zealand, Poland, the Slovak Republic, Switzerland, the United Kingdom and the United States. Data for France, Hungary and Norway are from 2000. Further reading: 61 GOV/PGC(2009)2 Pilichowski, E. and E. Turkisch (2008), “Employment in Government in the Perspective of the Production Costs of Goods and Services in the Public Domain,” OECD Working Papers on Public Governance, 8, OECD, Paris. OECD (2008), The State of the Public Service, OECD, Paris. Figure VI.2.1 Distribution of employment between the central and sub-central levels of government (2005) 62 GOV/PGC(2009)2 Figure VI.2.2 Change in the percentage of government staff employed at the central levels (1995 and 2005) 63 GOV/PGC(2009)2 VII. WORKFORCE CHARACTERISTICS OF THE CENTRAL GOVERNMENT 124. Business transformations, fluctuating demand and social and technical evolution are constantly changing the government’s need for skills and competencies. Labour is the key input in public administration, and the characteristics of the workforce can determine labour productivity and the capacity of governments to deliver services. For example, a balanced age pyramid is necessary to promote both continuity and innovation, and to provide career opportunities to entry-level and mid-career staff. As the workforce becomes more diverse, government policies and practices often change to reflect the needs of new employees (such as child care and wheelchair accessible offices) and can affect employment practices in the broader labour market. In addition, staff that reflect the diversity of society (particularly in terms of gender, ethnicity and disability status) will better reflect the needs, aspirations and experiences of the citizens—enhancing the government’s ability to provide efficient and effective services and possibly citizens’ trust in government. 125. This chapter presents data on gender representation and age distribution in the central government public service. The data allow policy makers and citizens to determine the degree that the government workforce reflects the characteristics of the wider labour force and illustrate important difference in workforce characteristics across countries that could affect the capacity of government to deliver services. 64 GOV/PGC(2009)2 VII.1. EMPLOYMENT OF WOMEN IN CENTRAL GOVERNMENT 126. Many OECD countries recognise that increasing the participation of women in the public service is a priority for equity and efficiency purposes. As a result, many countries have established policies aimed at increasing female participation in the public workforce, especially at managerial levels. 127. The data show a persistent increase in women’s participation in public sector employment between 1995 and 2005. While women represent between 40 and 50 percent of the total labour force in most OECD countries, female participation in the government workforce varies much more across countries, from 70% in Poland to just over 10% in Turkey. In most OECD countries for which data are available, women are better represented in the central government workforce than in the general labour force. Women are relatively under-represented in the central government workforce in Switzerland, Germany, Japan and Turkey, where they represent less than one-third of all workers. 128. Women are less represented at more senior levels than they are in the general government workforce. Here again, the situation differs across countries. Whereas over a third of all senior employees in Greece, New Zealand, Mexico and Portugal are women, women represent less than five percent of senior managers in Korea and Japan. In comparison, women are usually more strongly represented at lower levels or in administrative posts. In 8 of the 13 countries for which comparable data are available, women are over-represented in administrative positions when compared to the general government workforce. Methodology and Definitions 129. Data refer to 2005 and were collected through the 2006 OECD Strategic Human Resource Management in Government Survey. Respondents to the survey are predominately senior officials in central government personnel departments. The Human Resource Management survey was completed by all OECD countries excluding Greece; Greece subsequently provided the OECD with data for 2005. Australia subsequently provided 2005 data which was initially missing from their survey responses. Countries missing from the figures include those that did not respond fully to the survey questions. 130. The data concern the core civil service at the central level of government. Definitions of the civil service, as well as sectors covered at the central level of government, differ across countries. The definitions of “senior positions” and “administrative tasks” were left to the interpretation of countries when filling the questionnaires, and are thus indicative of broad trends. The labour force comprises all persons who fulfil the requirements for inclusion among the employed or the unemployed. Source: OECD Strategic Human Resource Management in Government Survey (2006). OECD Labour Force Survey (2006). Figure notes: (1) Data are not available for Austria, Canada, the Czech Republic, Denmark, France, Hungary, Ireland, Luxembourg, Mexico, the Slovak Republic and Spain. Data for Greece are from 1996 and 2005. Data for the Netherlands are from 2000 and 2005. Data for Poland are from 1995 and 2004. (2) Data are not available for Austria, Canada, the Czech Republic, Denmark, Hungary, Ireland, Luxembourg, Mexico, the Slovak Republic and Spain. Data for Poland are from 2004. (3) Data are not available for Austria, Canada, the Czech Republic, Denmark, Hungary, Iceland, Luxembourg, Poland, the Slovak Republic, Spain, Sweden, Turkey and the United States. Data for Italy are from 2003. (4) Data are not available for Austria, Canada, the Czech Republic, Denmark, France, Hungary, Iceland, Luxembourg, Norway, Poland, the Slovak Republic, Spain, Switzerland, Turkey and the United States. Data for Italy are from 2003. Further Reading: OECD (2008), The State of the Public Service, OECD, Paris. 65 GOV/PGC(2009)2 Figure VII.1.1 Percentage of central government employees that are female (1995 and 2005) Figure VII.1.2 Percentage of employees that are female in the central government compared to the total labour force (2005) 66 GOV/PGC(2009)2 Figure VII.1.3 Percentage of senior positions in central government filled by women (2005) Figure VII.1.4 Percentage of central government administrative positions filled by women (2005) 67 GOV/PGC(2009)2 VII.2. AGEING WORKFORCE 131. While the government must respond to the changing demands made by an ageing society (such as declining demand for primary education and increasing demand for health and elderly care services), civil servants are themselves ageing. With a very large proportion of the public service retiring over a relatively short period of time, maintaining the capacity of the public sector to deliver the same level and quality of public services remains a complex issue, particularly if the wider labour market is not to be adversely affected by massive new hiring in the public sector. Significant staff departures are an opportunity to bring staff with new skills into government, downsize the workforce where needed, decrease staff costs (as entry-level salaries are lower) and to re-allocate human resources across sectors. However, this also poses a challenge, with the loss of capacity and the need to postpone the retirement of some key staff. In addition, given the large share of public employment in many OECD member countries, the high replacement needs in the public sector could risk pre-empting the private sector’s access to new labour market entrants, which in the past have been a source of flexibility. 132. With the exception of Korea, the civil service is getting older in OECD countries. The roots of this demographic profile lie in the rapid expansion of public services from the 1970s until the mid-80s and the massive hiring that took place at this time. This was followed by a period, in the 1980s and 1990s, when the numbers in the workforce were stabilised and hiring freezes occurred in many countries. 133. In many OECD member countries, government workforces are ageing more rapidly than the rest of society and the wider labour market. The difference is most pronounced in Belgium, where almost 45% of the central government workforce is over 50 compared with just over 20% of the total labour force. In 11 OECD countries, over 30% of the central government labour force will retire within the next 15 years. Methodology and Definitions 134. Data refer to 2005 and were collected through the 2006 OECD Strategic Human Resource Management in Government Survey. Respondents to the survey are predominately senior officials in central government personnel departments. 135. The Human Resource Management survey was completed by all OECD countries excluding Greece; Greece subsequently provided the OECD with data for 2005. Australia subsequently provided 2005 data which was initially missing from their survey responses. Countries missing from the figures include those that did not respond fully to the questions of the questionnaire or whose data were not really comparable with data of other countries. 136. The data concern the core civil service at the central level of government. Definitions of the civil service, as well as sectors covered at the central level of government, differ across countries. The labour force comprises all persons who fulfil the requirements for inclusion among the employed or the unemployed. Source: OECD Strategic Human Resource Management in Government Survey (2006). OECD Labour Force Survey (2006). Figure notes: (1) Data are not available for Belgium, Canada, the Czech Republic, Denmark, Greece, Hungary, Italy, Luxembourg, Mexico, New Zealand, Poland, the Slovak Republic, Spain and Switzerland. (2) Data are not available for Canada, the Czech Republic, Denmark, New Zealand, Poland, the Slovak Republic and Spain. Further Reading: OECD (2007), Ageing and the Public Service: Human Resource Challenges, OECD, Paris. 68 GOV/PGC(2009)2 Figure VII.2.1 Percentage of central government workers older than 50 years old (1995 and 2005) 69 GOV/PGC(2009)2 Figure VII.2.2 Percentage of workers older than 50 years old in central government and the wider labour force, 2005 70 GOV/PGC(2009)2 VIII. HUMAN RESOURCES MANAGEMENT PRACTICES IN CENTRAL GOVERNMENT 137. The ability of governments to recruit, train, promote and dismiss staff is a key determinant of their capacity to provide public services that meet client needs in an efficient and cost-effective manner. Over the past 20 years, many governments have reformed their human resource management (HRM) practices with the goal of improving the skill level and efficiency of their labour forces. The scope of reforms, however, has varied tremendously depending on the economic, social and cultural context. There is also no agreement or evidence that specific HRM features result in improved performance. 138. The composite indexes developed in this chapter focus on the most common features of HRM reforms, including: i) the delegation of HRM policies and practices from central bodies to line managers; ii) the opening of recruitment for positions of all levels to external candidates; iii) the focus on performance assessments and the use of performance-related pay; and iv) the use of different HRM policies and practices for senior public servants. 139. The composite indexes will allow countries to compare the main characteristics of their HRM systems and the extent that they have implemented reforms. Individual country scores should be interpreted with caution and only provide a general indication about where a country might stand relative to its peers. None of the indexes evaluate the performance of HRM policies, nor do they provide any information about the quality of the work performed by public servants. 71 GOV/PGC(2009)2 VIII.1 DELEGATION IN HUMAN RESOURCES MANAGEMENT AT THE CENTRAL GOVERNMENT 140. In many countries, centrally determined statutory employment conditions have been seen as an impediment to the continued development of an efficient and service-oriented administration, and to performance-oriented management. Public managers are expected to improve the performance of their organisations and the outcome of their activities, and have to work with their staff to encourage, enable and support them in a continuous quest for quality, efficiency, productivity and propriety. By delegating some authority for human resource management to line ministries, departments and agencies, governments aim to increase the ability of public sector managers to adapt working conditions to the business needs of their organisation and to the merits and performances of individual employees. As HRM responsibilities have been delegated to line ministries, the role of the central HRM body has changed to focus more on setting minimum standards and formulating policy rather than implementing them. However, delegation is not without risks, which can include an increased variability of conditions of employment across government organisations, decreased mobility of staff and difficulties in maintaining shared government values and a whole-of-government perspective. 141. Most OECD countries have moved towards a more delegated model of HRM. However, the extent of delegation varies across OECD countries, and sometimes even across government bodies within the same country. Thus, there is no single model or common standard. A few countries stand out as having a relatively high level of delegation, including New Zealand, Sweden and Australia. In these systems, ministries have more flexibility to identify their staffing needs, recruit staff and determine compensation levels and other conditions of employment. In comparison, Italy, Ireland, France and Turkey tend to have more centralised models of human resource management. The level of delegation does not indicate how well public service staff are managed. However, it does indicate the level of flexibility that departments and units have in adjusting their HR policies to the specificities of their organisation, the job or the individual. 142. Compensation levels are a key factor in managers’ ability to recruit, motivate and retain staff. Most OECD countries have delegated managers some authority to determine compensation levels. Across OECD countries, the basic level of pay is more likely to be determined by a centralised authority while public sector managers have more authority to determine the variable portions of pay, such as benefits or performance related increases. Methodology and Definitions 143. This index gathers data on the delegation of determining the number and types of posts needed in an organization, compensation levels, position classification, recruitment and dismissals, conditions of employment and the actual impact on conditions of employment. The index ranges between 0 (no delegation) and 1 (high level of delegation). For a detailed description of the methodology used to develop the composite index, please see Annex B.1. While it was developed in cooperation with member countries and is grounded in theory, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 144. Data refer to 2005 and were collected by OECD’s Strategic Human Resources Management in Government Survey. Respondents were predominately senior officials in central government personnel departments. The survey was completed by all OECD countries excluding Greece. Countries missing from the figures are those for which the OECD had incomplete or inconsistent data. 145. Data refer to HRM practices at the central level of government for the civil service. Definitions of the civil service, as well as sectors covered at the central level of government, differ across countries. 72 GOV/PGC(2009)2 Source: OECD Strategic Human Resource Management in Government Survey (2006). Figure notes: (1) Data are not available for the Czech Republic, Greece, Poland and the Slovak Republic. (2) Data are not available for Canada, the Czech Republic, Greece and New Zealand. Further Reading: OECD (2008), The State of the Public Service, OECD, Paris. Figure VIII.1.1 Extent of delegation of human resource management practices to line ministries in central government (2005) 73 GOV/PGC(2009)2 Figure VIII.1.2 Authority of central government managers to determine compensation levels (2005) Other types of Basic pay remuneration/social benefits Australia z . Austria { . Belgium . . Canada N/A N/A Czech Republic N/A N/A Denmark z . Finland . z France { z Germany { { Greece N/A N/A Hungary { . Iceland . . Ireland { { Italy . . Japan . . Korea { z Luxemburg { { Mexico { . New Zealand N/A N/A Netherlands . { Norway . { Poland z z Portugal { . Spain { . Slovak Republic . { Sweden z . Switzerland { { Turkey { . United Kingdom z z United States . . Managers have significant authority: .. Managers have some authority: .. Managers have no authority: .., N/A : Data not available 74 GOV/PGC(2009)2 VIII.2 OPENNESS IN RECRUITMENT IN CENTRAL GOVERNMENT 146. Public sector recruitment systems that are open to external candidates at any point in their careers provide managers with access to a more varied pool of applicants. While open recruitment systems offer managers flexibility to address evolving challenges to providing efficient, citizen-oriented services, they also create difficulties in maintaining cross-government values and require central bodies to pay more attention to recruitment processes to guarantee merit-based selection at all position levels. 147. This index focuses on the possibilities individuals have to become part of the civil service throughout their careers, providing an indication of the level of competition for government posts. A career-based system characterised by competitive selection early on in the public servants’ career with no posts open to external recruitment corresponds to a low degree of openness. In contrast, a position-based system where candidates apply directly to a specific post and with most posts open to both internal and external applicants ranks high on the index. The index does not take into account the recruitment of contractual or casual staff and, in many cases, does not take into account exceptions to recruitment processes that have been introduced in some OECD countries with career-based civil services. 148. The level of openness varies significantly across countries, reflecting a long-standing distinction between career-based and position-based systems. In Australia, Finland, the Netherlands, New Zealand, Sweden, Switzerland (except for diplomatic careers), the United Kingdom and the United States, all posts below senior management are open to external recruitment and applicants apply directly to a specific post. In other countries with career-based systems such as France, employees are recruited almost exclusively at lower levels and move up to higher positions throughout their time in their civil service. In some countries, including Belgium, Germany, Ireland, Japan, Korea, Luxembourg, Mexico, Portugal and Spain, recruitment is more open for top managers and special experts. 149. Countries with more open recruitment systems also appear to have moved to a more delegated model of HRM. Australia, New Zealand and Sweden score high on both the openness and delegation indexes, indicating that line managers in these countries have more flexibility in determining both who they hire and the conditions of employment. However, there are exceptions; while the Netherlands and Japan score similarly on the delegation index, the Netherlands’ uses a position-based system whereas Japan uses a career-based system to recruit employees. Methodology and Definitions 150. The composite index includes policies for becoming a public servant in general (e.g. competitive examination or direct application) and for recruiting senior civil servants, and systems for appointing entrylevel positions and for allocating posts across departments. The index ranges between 0 (no openness) and 1 (high level of openness). See Annex B.1 for a detailed description of the methodology used to develop the composite index. While it was developed in cooperation with member countries and is grounded in theory, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 151. Data refer to 2005 and were collected by OECD’s Strategic Human Resources Management in Government Survey. Respondents were predominately senior officials in central government personnel departments. Countries missing from the figures are those for which the OECD had incomplete or inconsistent data. 152. Data refer to HRM practices at the central level of government for the civil service. Definitions of the civil service, as well as sectors covered at the central level of government differ across countries. Source: OECD Strategic Human Resource Management in Government Survey (2006). 75 GOV/PGC(2009)2 Figure notes: (1) Data are not available for Canada, the Czech Republic, Greece and Spain. A very small change to the composite index has been made to Norway and Poland since the publication The State of the Public Service, slightly affecting their ranking. (2) Data are not available for Canada, Czech Republic, Greece, Poland, Slovak Republic and Spain. The coefficient of correlation is 0.672. Further Reading: OECD (2008), The State of the Public Service, OECD, Paris. Figure VIII.2.1 Openness of the recruitment process to external candidates in central government (2005) Figure VIII.2.2 Relationship between open recruitment processes and the level of delegation in HRM systems in central government (2005) 76 GOV/PGC(2009)2 VIII.3 STAFF PERFORMANCE MANAGEMENT IN CENTRAL GOVERNMENT 153. An increased focus on performance is at the core of modernising public administrations in most OECD countries. Assessing performance is the first step in recognising both individual and collective efforts in a more fair and objective manner. At the same time, performance assessments can be used to make the goals of the organisation clearer to staff, link their roles to organisational objectives and help implement changes. However, creating a performance management system does not in itself improve performance. Its success, in part, depends on goals and strategies being clearly defined and communicated to employees. 154. In recent years, efforts have been made to link incentives to the assessment of performance to increase the motivation and accountability of civil servants. A couple of decades ago, nearly all central government employees were given pay increases based on length of service, regardless of how well they did their jobs. As improving staff performance took on a new urgency in a context of economic difficulties and budget constraints, elements of performance related pay (PRP) were introduced in many countries along with an increased delegation of HRM responsibilities. PRP can vary along several dimensions, including: the range of staff positions it applies to; whether the targets and incentives apply to individuals or groups; the extent to which rankings are used; and the size of performance-related awards (one-off bonuses or merit increases in basic pay). PRP is not without disadvantages; its impact on trust, pay differentials (particularly for minority groups) and equity must be carefully monitored. 155. Almost all countries have formalised performance assessments for most core government employees. In only a few countries, performance assessments are seen as having only a limited influence on HRM decisions about individuals. Of the 26 countries for which there are comparable data, about onethird do not use PRP. There is no single model of PRP among the remaining two-thirds of countries. However, a trend is that PRP has spread from the management level to cover other categories of staff. Several countries in the middle, such as Canada, Ireland or Italy use PRP only for senior managers. In the Netherlands, only some government organisations use PRP. Neither of these indexes measures the capacity of the HRM system to motivate employees or the performance of employees. Further, many countries manage to nurture a performance culture without a formalised performance-focus. Methodology and Definitions 156. The performance assessment index indicates the extent to which it is used in career advancement, remuneration and contract renewal decisions, based on the views of survey respondents. The PRP index looks at the range of employees to whom PRP applies and the maximum proportion of base pay that PRP may represent. Both indexes range between 0 (no use) and 1 (high use). Annex B.1 provides a detailed description of the methodology used to develop these composites. While they were developed in cooperation with member countries and are grounded in theory, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 157. Data refer to 2005 and were collected by the Strategic Human Resources Management (HRM) in Government Survey. Respondents were predominately senior officials in central government personnel departments. Countries missing from the figures are those for which the OECD had incomplete or inconsistent data. 158. Data refer to HRM practices at the central level of government for the civil service. Definitions of the civil service, as well as sectors covered at the central level of government differ across countries. 77 GOV/PGC(2009)2 Source: OECD Strategic HRM in Government Survey (2006). Figure notes: (1) Data are not available for Canada, the Czech Republic, Greece, Japan, Luxembourg, New Zealand, Norway, the Slovak Republic, Switzerland, Turkey and the United States. A small change to the composite index has been made to Germany since the publication of The State of the Public Service, slightly affecting its score. (2) The OECD average includes eight countries that do not use PRP which are not depicted in the figure: Austria, Iceland, Luxembourg, Mexico, Poland, Portugal, the Slovak Republic and Sweden. Data are not available for Belgium, Greece, New Zealand and Turkey. Further Reading: OECD (2008), The State of the Public Service, OECD, Paris. Figure VIII.3.1 Extent of the use of performance assessments in HR decisions in central government (2005) 78 GOV/PGC(2009)2 Figure VIII.3.2 Extent of the use of performance-related pay in central government (2005) 79 GOV/PGC(2009)2 VIII.4 SENIOR CIVIL SERVICE 159. Countries need senior civil servants that are able to pursue performance-oriented governance and management, ensure sufficient cohesion across ministries, and at the same time preserve and protect the ethos of a politically neutral and professionally competent public administration. The senior civil service is the interface between politicians and the public administration. They are responsible for the proper and appropriate implementation of legal instruments and of political strategies and measures. They are also responsible for the coherence, efficiency and appropriateness of government activities. Thus, the capacity of the senior civil service has become a key public governance issue. 160. Due to the growing leadership expectations placed on senior civil servants, there is an increasing tendency in OECD countries to group and manage them separately under different HRM policies. The index shows the level of institutionalisation of specific HRM practices that apply to the group of senior managers. However, the existence of separate management rules and practices applying to senior civil servants does not indicate how well they are managed or how well they perform. The index only reflects the efforts made in recent years to adjust the rules and practices to the management needs created by the increased importance of senior civil servants. 161. In general, senior civil servants represent a very small percentage of all central government employees, and most are not politically appointed. There is a broad spectrum of senior management arrangements across OECD countries that reflect different administrative cultures and historical developments. Four OECD countries do not have a separate senior civil service: Austria, Denmark, Mexico and Switzerland. The Netherlands, the United Kingdom and the United States stand out as having institutionalised the management of their senior public servants the most, including the establishment of a separate formal senior executive service. In comparison, Italy, Ireland and Spain have little differentiation between their senior management group and other staff. Methodology and Definitions 162. The index is based on data about the existence of a separate group of senior managers, the identification of future leaders early in their careers, their recruitment process, the existence of a defined set of skills, the special emphasis put on the management of their performance and the term of their appointment. The index ranges between 0 (no separate management of the senior civil service) and 1 (high degree of separate management of the senior civil service). See Annex B.2 for a detailed description of the methodology used to develop the index. While it was developed in cooperation with member countries and is grounded in theory, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 163. Data refer to 2005 and were collected by the Strategic Human Resources Management (HRM) in Government Survey. Respondents were predominately senior officials in central government personnel departments. Countries missing from the figures are those for which the OECD had incomplete or inconsistent data. To calculate senior civil servants as a percentage of central government staff , 2005 data from the Comparison of Employment in the Public Domain (CEPD) Survey was used for the total number of central government employees, except for Australia who subsequently provided updated data. 164. Data refer to HRM practices at the central level of government for the civil service. Definitions of the senior civil service, as well as sectors covered at the central level of government differ across countries. 80 GOV/PGC(2009)2 Source: OECD Strategic HRM in Government Survey (2006) and the OECD CEPD Survey (2006). Figure notes: (2) The OECD average includes four countries that do not have a separate senior civil service or separate management of this group and are not depicted in the figure: Austria, Denmark, Mexico and Switzerland. Data are not available for Greece, Sweden and New Zealand. A small change in methodology was implemented since this composite index was first published in The State of the Public Service. Although the individual scores for several countries have changed, the overall grouping of the countries has remained the same (e.g. countries with higher index values are the same in both published versions). Further Reading: OECD (2008), The State of the Public Service, OECD, Paris. 81 GOV/PGC(2009)2 Figure VIII.4.1 Proportion of central government employees who are senior civil servants and number of senior civil servants who are appointed at the complete discretion of the ministers/head of state (2005) Number of senior civil servants Percentage of central government employees that are senior civil servants Scope or number of senior civil servants who are appointed at the complete discretion of the ministers/head of state Australia 2130 1.59% None Belgium 168 0.12% None Finland 130 0.11% There is new group of politically appointed state secretaries (10 appointees) that are not part of the career civil service and thus not considered senior civil servants. The Government or the President of the State appoints all the 130 senior civil servants who are nominated by ministers. Hungary 270 0.10% The senior civil servant staff is made up of civil servants who were appointed by the Prime Minister. The appointment as senior civil servant is for undefined time. Ireland 250 Not available Political appointees are not defined as part of a senior civil service and are appointed for a specified period of time generally equivalent to the period the minister is in office. Japan 861 0.16% Political appointment is conducted for Special Service such as Special Advisor to the Prime Minister, Senior Vice Minister and Parliamentary Secretary. For the other senior civil servants, although the power to appoint employees to designated service is vested with the minister of each ministry, it is conducted on a merit based system. Designated services are not subject to political appointment. Korea 1305 0.17% Staff belonging to the Office of the President and the Office of the Prime Minister(95 staffs), and Policy Advisor of Ministers (19 staffs) are appointed Luxemburg 200 Not available All senior civil servants are appointed by the Grand Duke on the basis of article 76 of the Constitution. Netherlands 800 0.28% None. No civil servant is appointed at the complete discretion of one minister. This is always done via a process with checks and balances to limit the possibility of political appointments. Norway 305 0.12% All are appointed by the Council of State/the King in Council. The highest posts are appointed by the Council of State. They are not political appointees, in the meaning that they shall leave their post when there is a change of government. Slovak Republic 250 Not available Pursuant to Article 25 clause 1 of the Civil Service Act they are professionals fulfilling tasks for a Government member, President, Chairman of the National Council or Vice-Chairman of the National Council. United Kingdom 3547 Not available 70 United States 6,700 0.25% 1,200 82 GOV/PGC(2009)2 Figure VIII.4.2 Degree to which senior civil servants are managed by separate HRM policies in central government (2005) 83 GOV/PGC(2009)2 IX BUDGET PRACTICES AND PROCEDURES 165. Ageing societies and the global economic downturn are putting pressure on government finances in many OECD countries. At the same time, governments are emphasising the importance of sustainable public finances to achieving long-term economic growth. Well-functioning budget institutions are those that can control fiscal aggregates, effectively allocate resources according to government priorities and deliver cost effective programmes. They are fundamental to supporting economic growth, improving fiscal health and achieving stable taxes and intergenerational fairness. 166. This chapter includes indicators on selected aspects of practices and procedures related to budget formulation and execution at the central level of government. Four composite indexes are presented on the extent to which: (i) the public is provided with information so that fiscal policies and spending priorities can be understood, monitored and evaluated; (ii) budgets incorporate a medium-term perspective to ensure that multi-year consequences of new spending measures are taken into account; (iii) performance information is used in the formulation of the executive’s budget proposal; and (iv) flexibility is provided to the executive and departmental managers to make changes during the fiscal year to the approved budget. 167. All results are based upon country responses to the OECD Survey of Budget Practices and Procedures. All responses represent the opinion of countries’ own assessments of the current practices and procedures. 168. NOTE: Discussions on the data are currently in progress and the composites may change. 84 GOV/PGC(2009)2 IX.1 Fiscal Sustainability 169. Fiscal sustainability encapsulates four interlinked dimensions: government solvency, continued stable economic growth, stable taxes and intergenerational fairness (Schick, 2005). Solvency is the ability of government to pay its financial obligations. Continued stable economic growth is the ability of fiscal policy - in concert with other policies (e.g. infrastructure and social policies) - to support and sustain economic expansion. Stable taxes are the ability to finance programmes without modifying the tax burden on citizens. Fairness is the capacity of governments to pay for current public services with today’s revenues, rather than shifting the cost to future generations or denying them the services available today. 170. Two practices to help countries achieve fiscal sustainability are fiscal projections to assess the likely long-term fiscal position and fiscal rules to constrain spending in the medium-term. Fiscal projections assess the likely consequences of continued current spending with the impact of demographic change and other factors. OECD Best Practices for Budget Transparency state that projections should cover 10 - 40 years, and be prepared/updated at least every 5 years or when major revenue/expenditure policy changes are made. Moreover, all key assumptions underlying the projections should be made explicit, together with a range of plausible scenarios. 171. A fiscal rule is a constraint on fiscal policy designed to curb excessive spending above the resource base. Rules may focus on different fiscal indicators: expenditure, the budget balance, debt and revenue. While evidence suggests that fiscal rules can assist governments to achieve fiscal objectives and discipline more generally, there is no one-size-fits-all rule for every country. 172. Two-thirds of all OECD countries produce long-term fiscal projections, the majority of which span 41- 50 years, though they extend up to 75 years in the United States. Half of the countries that prepare projections do so annually, although many do it more periodically (every three to five years). Balance and expenditure rules are the most common form of fiscal rules among OECD countries, while revenue rules are the least common. Six OECD countries do not use fiscal rules as a means to constrain spending. 173. In interpreting country’s reporting of these practices, however, it is important to consider how monetary union reporting obligations influence countries responses. The European Union convergence reporting obligations require EU member countries to report long-term fiscal projections and to adopt debt and balance rules. Some OECD countries report these as domestic practices; others do not. Methodology and Definitions 174. The indicators refer to the central government and draw upon country responses to questions in the OECD Budget Practices and Procedures database collected via a survey during the first half of 2007. Respondents were senior budget officials. Responses represent the countries’ own assessments of the current practices and procedures. Some European Union member country responses reflect practices for domestic procedures, while others capture countries’ practices for European Union Convergence Reporting. Source: OECD’s International Budget Practices and Procedures Database (www.oecd.org/gov/budget/database) (2007) Figure notes: (1) There are no long-term budget projections (i.e. spanning more than ten years) prepared in Austria, Canada, Denmark, Germany, Greece, Hungary, Italy, Japan, Luxembourg, Korea, Mexico, Slovak Republic, Spain and Turkey. Based upon Q. 12 “Are long-term fiscal projections prepared? How many years do they normally cover and how frequently are they prepared?” (2) Based upon Q. 14 “Are there any fiscal rules that place limits on fiscal policy? Which of the following defines the fiscal rule?” 85 GOV/PGC(2009)2 Further readings: Anderson, B. and J. J. Minarik (2006), “Design Choice for Fiscal Policy Rules,” OECD Journal on Budgeting, Vol. 5, No. 4, pp. 159 – 208. Anderson, B. and J. Sheppard (Forthcoming), “Fiscal Futures, Institutional Budget Reforms and Their Effects: What Can Be Learned?” OECD Journal on Budgeting. Schick (2005), “Sustainable Budget Policy: Concepts and Approaches,” OECD Journal on Budgeting, Vol. 5, No. 1, pp. 107-126. Figure IX.1.1: The coverage and frequency of long-term budget projections (2007) How many years do projections cover? 11 - 20 21-30 31 - 40 41 - 50 51 -60 71 -80 Total Every year Belgium, Czech Republic, Finland, France, Poland, Portugal, United Kingdom Norway United States 9 Frequency projections updated Every 3-5 years Switzerland Australia, New Zealand Ireland, Netherlands 5 Ad hoc basis Iceland 1 Total 2 0 2 9 1 1 86 GOV/PGC(2009)2 Figure IX.1.2 Use of rules that place limits on fiscal policy (2007) Country Expenditure rule Balance (deficit/ surplus) rule Debt rule Revenue rule Australia Austria Belgium . Canada . . . Czech Republic . Denmark . . . Finland . . France . . . Germany . Greece . . Hungary Iceland . Ireland . . Italy . . . Japan . . . Korea Luxembourg . . . Mexico . . . . Netherlands . . . New Zealand Norway . Poland . . Portugal . . Slovak Republic . . Spain Sweden . . . . Switzerland . Turkey United Kingdom . . United States Total 15 18 12 4 Used by countries: . 87 GOV/PGC(2009)2 IX.2 BUDGET TRANSPARENCY 175. Budget transparency represents openness about current policy intentions, budget formulation and execution. Publicly available, comprehensive budget documentation is important because it supports greater understanding of current fiscal policies and government priorities, and it enables governments to be held accountable for producing realistic budgets as well as for the social and economic impact of planned policy measures. Thus, budget transparency supports improved fiscal discipline, the effective allocation of resources and operational efficiency. However, while budget transparency is necessary for sound budget outcomes, it alone is insufficient. The availability of information within the budget document or other sources does not necessarily assure its accuracy. In addition, citizens, interest groups and legislators must use the information to hold the government accountable for achieving better budget outcomes. 176. As a group, Anglo-Saxon countries display high levels of budget transparency. In Australia, New Zealand and the United Kingdom, the government’s commitment to fiscal discipline is linked to fiscal responsibility legislation and a political commitment to transparency in budget and financial reporting. 177. A key aspect of transparency is the extent to which the executive’s budget discloses the underlying assumptions (macroeconomic and others) that set the fiscal framework within which government organisations formulate their spending proposals. Key assumptions include estimates of GDP growth, the current account, inflation and the interest rate. These estimates underlie forecasts for tax and non-tax revenues, government debt servicing obligations and debt servicing requirements. Arguably there is no single factor more responsible for “de-railing” budget outcomes and projections of deficits or surpluses than the use of weak macroeconomic assumptions. 178. Approximately half (18) of all OECD countries make their methodology underlying macroeconomic assumptions publicly available, either directly or upon request. Two countries, however, only report partial information. Countries not disclosing underlying assumptions may still score well on the transparency index due to the public availability of other information. Methodology and Definitions 179. Budget transparency is the full disclosure of all relevant fiscal information in a timely and systematic manner (OECD, 2002). The indicators draw upon the OECD Budget Practices and Procedures Database collected during the first half of 2007 and refers to budget practices in the central government. Survey respondents were senior budget officials. Responses represent the countries’ own assessments of the current practices and procedures. 180. The composite index draws upon the OECD Best Practices for Budget Transparency and reflects analysis conducted by OECD. It includes the public availability in the budget and supporting documentation of: the methodology used to establish economic assumptions; fiscal sensitivity analysis of the economic assumptions; differentiation between existing and new expenditure proposals in the budget documentation; the inclusion of multi-year expenditure estimates; the availability of government performance information; and the types of off-budget expenditures and contingent liabilities. In addition, it includes the frequency that in-year reports are published and the budgetary independence of the supreme audit institution. The index ranges between 0 (no budget transparency) and 1 (high level of transparency). For detailed information about the methodology used to develop the composite index, see Annex B.2. While the index is based on best practices and developed in cooperation with member countries, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 88 GOV/PGC(2009)2 Source: OECD’s International Budget Practices and Procedures Database (www.oecd.org/gov/budget/database) (2007) Figure notes: (2) In Belgium the methodology is available in a working paper on the Federal Planning Bureau website (www.plan.be) Further readings: OECD (2002), “OECD Best Practices for Budget Transparency,” OECD Journal on Budgeting, Vol.1, No. 3, pp. 7-14. OECD (forthcoming), Budgeting in OECD Countries, OECD, Paris. Figure IX.2.1 Extent of budget transparency (2007) 89 GOV/PGC(2009)2 Figure IX.2.2 Public availability of methodology used to establish the economic assumptions of the budget (2007) Number Countries Published 9 Austria, Canada, Denmark, Finland, Luxembourg, Netherlands, Norway, Switzerland, United Kingdom Available on request 7 Germany, Greece, Iceland, Ireland, New Zealand, Poland, Slovak Republic Other 2 Belgium, Italy Not available 12 Australia, Czech Republic, France, Hungary, Japan, Korea, Mexico, Portugal, Spain, Sweden, Turkey, United States 90 GOV/PGC(2009)2 IX.3 MEDIUM-TERM BUDGET PERSPECTIVE 181. In order to improve fiscal discipline and maintain sustainable public finances over the mediumterm, many countries have adopted medium-term budget and/or expenditure frameworks that typically span between three and five years, including the budgeted fiscal year. The medium-term perspective may include estimates of revenues and expenditures and/or targets or ceilings for spending. It signals the direction of policy and funding changes and gives actors time to adjust. It also helps identify fiscal space that can be allocated for new and existing programmes. Thus, it enables decision makers to consider the cost of competing programmes before they make funding decisions, while increasing the predictability of funding during programme implementation. OECD countries that are members of the European Union are required to report medium-term objectives; however, legislatures may or may not incorporate these when formulating the budget. 182. While there is a consensus about the importance of a medium-term perspective in the budget process to lend stability and credibility to a government’s fiscal objectives, different opinions exist on the importance of expenditure ceilings/targets. Whereas expenditure estimates capture information on the cost of existing policies and programmes and form the baselines for the following years’ budgets, expenditure ceilings provide a top-down constraint on spending in future years. 183. A medium-term budget perspective is neither necessary nor sufficient to achieve sound outcomes such as balanced budgets. Its impact is dependent on the credibility of the expenditure estimates (and ceilings/targets) as well as how this information is used by decision makers and by civil society observers. Failure to achieve medium-term budget objectives is often related to weak arrangements ensuring the effective preparation, legislation and implementation of budgetary targets. 184. Medium-term expenditure estimates are produced in all but one OECD country, most often at an aggregate level. In comparison, the use of multi-year ceilings is less common, although they are still used by two-thirds of OECD countries -most often at an aggregate level. The strength of the medium-term perspective adopted by countries varies greatly across OECD countries. Methodology and Definitions 185. The indicators draw upon country responses to questions in the OECD Budget Practices and Procedures database collected via a survey during the first half of 2007, and refer to the central level of government. Survey respondents were senior budget officials. Responses represent the countries’ own assessments of the current practices and procedures. Some European Union member country responses reflect practices for domestic procedures, while others capture countries’ practices for European Union Convergence Reporting. 186. The composite index measures the extent to which countries have developed a medium-term perspective in their budget process, and reflect analysis conducted by the OECD. It contains information on whether countries develop and present multi-year expenditure estimates in the annual budget, the number of years the estimate covers, how often estimates are updated; and whether there are multi-year targets or ceilings and how often these are revised. The index ranges between 0 (no medium-term perspective) and 1 (conceptually strong medium-term perspective). The composite index does not capture whether the framework includes performance information, or its impact on budget outcomes such as fiscal discipline and allocative efficiency. See Annex 2.B for a detailed description of the methodology used to construct the index. While the index is based on best practices and developed in cooperation with member countries, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 91 GOV/PGC(2009)2 Source: OECD’s International Budget Practices and Procedures Database (www.oecd.org/gov/budget/database) (2007). Figure notes: (2) Based on Q. 16 “Does the annual budget documentation submitted to the legislature contain multi-year expenditure estimates?” and Q. 20 “Are there multi-year expenditure targets or ceilings?” Further readings: OECD (2002), “OECD Best Practices for Budget Transparency,” OECD Journal on Budgeting, Vol. 1, No.3, pp. 7-14. OECD (forthcoming), Budgeting in OECD Countries, OECD, Paris. Figure IX.3.1. Strength of the medium-term budget perspective (2007) 92 GOV/PGC(2009)2 Figure IX.3.2: Existence of medium-term budget perspective (2007) None Total Level at which estimates are presented to legislators Aggregat e Austria, Czech Republic, France, Italy, Japan Germany Hungary Mexico, Poland, Switzerland 10 Ministry Ireland, United States 5 Line-item Luxembourg, Netherlands New Zealand, Sweden Canada, Denmark, 6 Other Australia, Norway, Portugal, Spain 5 None Greece 1 Total10 Level at which ceilings/targets are prepared Aggregate Ministry Line-item Other Slovak Republic, Turkey United Kingdom Finland, Iceland, Belgium Korea 12 2 1 5 93 GOV/PGC(2009)2 IX.4 PERFORMANCE-ORIENTED BUDGETING 187. OECD countries are under pressure to improve government efficiency and effectiveness while controlling expenditures. While all OECD countries are developing information to assess their government’s performance, there is a wide variation in the type of information produced and the extent of coverage. Formal performance information refers to both performance measures (outputs and/or outcomes) and evaluations. Advocates claim that the provision of “objective” performance information facilitates better decision making for the efficient use of resources, the management of programmes, centralised resource allocation and expenditure prioritisation decisions. Performance information can help to identify programmes that are working and those that are not, enabling managers and decision makers to take action to improve poor performing programmes. 188. The index examines the degree to which countries have established government-wide systems and frameworks for the development and use of performance information in the budget process. Countries receiving a high score have a comprehensive, government-wide framework for developing performance information (evaluations and performance measures); monitoring and reporting on results; and integrating performance information into the budget system and accountability process and use it in decision making. However, this is not a measure of how successfully any given system operates in practice. Success is better evaluated by whether the reforms are achieving their stated objectives and this cannot be captured in this index. 189. The approaches to developing and using performance information in the budget process vary across OECD countries and there is no one single model. For example, in most countries that use performance information in budgeting, there is a loose or indirect link between performance information and funding. These countries use performance measures and evaluations along with information on fiscal policy and policy priorities to inform, but not determine, budget allocations. In over 85% of countries for which data are available, failure to achieve a performance target does not result in the elimination of a program. Most OECD countries use performance information in budget discussions. Methodology and Definitions 190. The indicators draw upon country responses to questions in the OECD Budget Practices and Procedures database collected via a survey during the first half of 2007, and refer to the central level of government. Survey respondents were senior budget officials in OECD member countries. Survey respondents were senior budget officials. Responses represent the countries’ own assessments of the current practices and procedures. Countries missing from the figures are those for which OECD had incomplete or inconsistent data. 191. This composite index reflects analysis conducted by the OECD and contains information on what type of performance information is developed; the processes for setting goals and targets; the process for monitoring and reporting on results; and if and how performance information is used in budget negotiations and decision making processes by key actors including the central budget authority, the line ministries and politicians. The index ranges between 0 (no performance budgeting system) and 1 (highly developed and comprehensive performance budgeting system). Annex 2.B contains details of the methodology and questions used to construct this index. While the index was developed in cooperation with member countries, both the variables and their weights are offered for discussion and, consequently, may evolve over time. 94 GOV/PGC(2009)2 Source: OECD’s International Budget Practices and Procedures Database (www.oecd.org/gov/budget/database) (2007).Figure Notes: (1) Data are incomplete or inconsistent for France, Germany and Italy. (2) Based on Q. 83 “Is performance information used as part of the budget discussions/negotiations between the central budget authority and line/spending ministries?” Further Readings: OECD (2008), “Performance Budgeting: A Users' Guide”, OECD Policy Brief, March 2008. Curristine, T. (2005) “Performance Information in the Budget Process: Results of the OECD 2005 Questionnaire,” OECD Journal on Budgeting, Vol. 5, No. 2., pp. 87-132. OECD (forthcoming), Budgeting in OECD Countries, OECD, Paris. Figure IX.4.1 Level of development of performance budgeting system (2007) 95 GOV/PGC(2009)2 Figure IX.4.2 Use of performance information in budget discussions (2007) Types of performance information Country Evaluation reports Performance against targets Australia .. Austria .. Belgium .. Canada .. Czech Republic .. Denmark .. Finland .. France N/A N/A Germany .. Greece . N/A Hungary .. Iceland .. Ireland .. Italy .. Japan .. Korea .. Luxembourg .. Mexico .. Netherlands .. New Zealand .. Norway .. Poland .. Portugal .. Slovak Republic .. Spain .. Sweden .. Switzerland .. Turkey .. United Kingdom .. United States .. Total countries that use 20 18 Used by countries: .; Not used by countries: .; Data not available: N/A 96 GOV/PGC(2009)2 IX.5 EXECUTIVE FLEXIBILITY IN BUDGET EXECUTION 192. The ability of the executive, ministries and agencies to adapt spending to changing circumstances during the implementation of the budget is key to ensuring that funds are allocated efficiently and spent effectively. Even with a sound budget formulation process, economic assumptions can change, input prices can fluctuate and evolving political priorities can call for adjustments and reallocation of budgeted resources during budget execution. 193. Budget flexibility can occur at two levels of government: at the implementation level - within line ministries/agencies -and at the general executive level. The line ministry/agency flexibility index concerns their flexibility to use budgeted funds for different purposes or inputs (predominantly focused on discretionary spending). For example, the possibility to carry-over unspent appropriations from one fiscal year to another, borrow against future appropriations or reallocate in-year resources across ministerial lineitems facilitate the optimal use of public resources and seek to provide incentives to improve the efficiency of public expenditure. In addition, lump-sum appropriations or the right of agencies to use savings from efficiency gains for other purposes provide managers with flexibility to allocate funds to their best use. The general executive flexibility index is an extension of the ministry/agency flexibility index and concerns the government’s power to increase spending during the execution of the budget and the degree to which the legislature is involved in this process. 194. While the indicators capture the procedural flexibility of line ministries/agencies or the government more generally to reallocate budget resources, they do not measure the impact of flexibility on the provision of government goods and services, the internal management capacity to recognise when reallocations are necessary or the soundness of decision making processes that result in budget reallocations. It should be emphasised that if flexibility is applied unreserved and unchecked, it can result in sub-optimal budget outcomes. Potential risks include weakened operational and allocative efficiency, the accrual of arrears and opportunities for the abuse of power by government managers, as well as increased government deficit. Thus, any move to greater flexibility warrants clear and adequate controls and oversight in order to hold managers accountable. 195. OECD countries vary in the degree of flexibility they provide to line ministries/agencies and to the executive. Of interest is that some countries permit greater flexibility at the level of ministries/agencies relative to the executive as a whole (e.g. Korea, Luxembourg, Poland), whereas others permit greater flexibility for the executive as a whole relative to ministries/agencies (e.g. Denmark, Mexico, and Switzerland). Methodology and Definitions 196. The indicators draw upon country responses to questions in the OECD Budget Practices and Procedures database collected via a survey during the first half of 2007 and refer to the central level of government. Survey respondents were senior budget officials. Responses represent the countries’ own assessments of the current practices and procedures. 197. The ministry/agency flexibility index measures the extent to which lump sum appropriations are used; the number of line items in the budget; and agencies’ ability to carry-over unused budget allocations between years, borrow against future appropriations, reallocate between line items and keep efficiency gains. The executive flexibility index contains all variables in the ministry/agency flexibility index as well as: constraints on the executive’s ability to increase mandatory and discretionary expenditures during the budget year, and whether overspending can occur prior to Legislative approval. The indexes range between 0 (no flexibility) and 1 (high level of flexibility) and reflect analysis conducted by the OECD. See Annex 2.B for a detailed description of the methodology used to construct the indexes. While the index was 97 GOV/PGC(2009)2 developed in cooperation with member countries, both the variables and their weights are offered for discussion and, consequently, may evolve over time. Source: OECD’s International Budget Practices and Procedures Database (www.oecd.org/gov/budget/database)( 2007) Further reading: OECD (forthcoming), Budgeting in OECD Countries, OECD, Paris. Figure IX.5.1: Ministry/Agency flexibility index 98 GOV/PGC(2009)2 Figure IX.5.2: Executive flexibility index 99 GOV/PGC(2009)2 X. REGULATORY MANAGEMENT 198. This chapter will be posted separately on OLIS on April 20th 100 GOV/PGC(2009)2 XI. INTEGRITY 199. Fostering integrity and preventing corruption in the public sector is essential to maintain confidence in government and public decision making, as well as to support a level playing field for businesses. Integrity refers to the application of public sector values, principles and norms in daily operations of public sector organisations. Integrity requires public officials to use information, resources and authority according to intended official purposes in order to promote the public interest. 200. Achieving a culture of integrity requires coherent efforts to set and update integrity standards, provide guidance, monitor and enforce them in the daily practice. Maintaining public trust also requires a forward-looking approach that anticipates risks to integrity and applies tailored counter measures – such as specific guidelines or restrictions; increased transparency and tightened control; or regular rotation and background checks for certain positions. 201. This chapter examines member countries’ measures for promoting integrity and preventing corruption in central government. Data are presented on the scope and transparency of conflict of interest disclosures for decision-makers, and indicators on reporting procedures and whistle-blowing protection examine countries’ measures to enforce integrity standards by facilitating the detection of wrongdoing. The chapter also reviews the top government activities at risk for corruption, with a focus on public procurement markets. 101 GOV/PGC(2009)2 XI.1 CONFLICT OF INTEREST—DISCLOSURE OF FINANCIAL INSTERESTS BY DECISION MAKERS 202. Citizen’s trust in government is weakened when public officials allow personal bias to pervade their decision-making. At a time when the interchange between public and private sectors has significantly increased, standards and measures for preventing and managing conflict of interest are crucial to ensure that the integrity of decision making is not compromised by public officials’ private interests. 203. To maintain public confidence in the integrity of official decision-making, the vast majority of OECD countries have developed and implemented conflict of interest policies that require decision makers, in particular the president, prime minister and ministers as well as members of the Legislature, to disclose their financial and economic interests. Disclosure is a critical first step to determine whether private interests could improperly influence the performance of official duties. Disclosures may also support the detection of illicit enrichment. 204. Whereas Iceland, Luxembourg and Switzerland remain the only OECD countries that do not require decision makers in the executive and legislative branches to disclose private interests, a few countries have also begun requiring disclosures for officials in the judiciary, for example for judges in Finland and Hungary. Moreover, countries increasingly require officials in vulnerable positions, such as public procurement, customs and tax administration, to disclose their private interest. 205. Providing information on assets and liabilities has remained the primary focus of disclosure requirements. In the past decade, the number of countries requesting information on loans has increased the most significantly, almost tripling. Likewise, countries have placed an increased emphasis on the disclosure of ancillary employment arrangements and of previous and future employment. A more than double increase in countries requiring the latter indicates countries’ efforts to manage conflict of interest in the case of movements between the private and public sectors “revolving doors”, which recently received particular public attention. 206. Citizens’ rising demand for transparency in public life and closer scrutiny by the media and opposition parties have increased expectations for transparency of disclosures by high-level public officials. Although the vast majority of OECD countries require decision makers to disclose more and more of their private interest to avoid potential conflict of interest, it is much less common to make these disclosures available to the public. Twelve OECD countries provide access of disclosed private interest, in particular by legislators and ministers, to facilitate public scrutiny while in other countries the information provided is exclusive to internal official use and stays confidential. Methodology and Definitions: 207. The data focus on conflict of interest measures for decision makers in the central government that were in place in OECD member countries in 2000 and 2009. Data for 2000 were collected by OECD via a tailored survey questionnaire and were originally published in Trust in Government (2000). Respondents to the survey were OECD member country delegates in charge of integrity in central government. Data were updated in 2009 via a tailored survey questionnaire completed by members of the Expert Group on Conflict of Interest. A total of 25 OECD member countries responded to both the 2000 and 2009 surveys. Thus, the data presented in figure XI.1.1 represent the percentage of the 25 countries for which data are available in 2000 and 2009. Annex C provides unique data for each country on the types of private interests they require decision makers to disclose as well as the level of transparency provided to the disclosures. 102 GOV/PGC(2009)2 Source: OECD Survey on integrity (2000 and 2009). Further Reading: OECD (2003), Managing Conflict of Interest in the Public Service, OECD Guidelines and Country Experiences, OECD, Paris. OECD (2000), Trust in Government: Ethics Measures in OECD Countries, OECD, Paris. Figure Notes: Data are not available for Greece, Korea, the Slovak Republic, Spain and the United Kingdom. (2) In addition, data are not available for Denmark. Countries responding that disclosures are “partly” available to the public publish disclosures for some, but not all decision makers and/or some, but not all types of activities (such as assets and liabilities, but not loans). Figure XI.1.1 Percentage of countries that require decision makers in the central government to disclose potential conflict of interest (2000 and 2009) Figure XI.1.2 Public availability of conflict of interest disclosures of decision makers in the central government (2009 Level of transparency Fully available to the public Partially available to the public Not available to the public Total 12 8 2 Country Australia, Czech Republic, Hungary, Ireland, Italy, Japan, New Zealand, Poland, Portugal, Sweden, United States Austria, Belgium, Canada, Finland, France, Germany, Netherlands, Norway Mexico, Turkey Disclosures not required 3 Iceland, Luxembourg, Switzerland 103 GOV/PGC(2009)2 XI.2 PUBLIC INTEREST DISCLOSURE: WHISTLE-BLOWING 208. The risk of corruption is significantly increased in a secretive environment. Facilitating the reporting of misconduct could substantially help public institutions monitor compliance and detect malpractice or misconduct such as fraud and corruption. OECD countries have increasingly taken actions to provide accessible procedures for employees to raise concerns of violation of laws such as fraud and corruption in the public sector, or “blow the whistle.” Whistleblowing has increasingly been considers as a manifestation of an open organisational culture where employees are aware and trus reporting procedures to raise their concerns. It is also seen as an ultimate solution for safeguarding the public interest in order to maintain confidence in public organisations. Although whistleblowing remained a bonafied action to defend the public interest, a few countries, including Korea and Canada, have also introduced financial incentives to facitilate whistleblowing.. 209. Since 2000, many OECD countries have developed procedures for public servants to facilitate the exposure of misconduct. As of 2009, 24 OECD member countries either oblige their public servants to report misconduct and/or provide procedures to facilitate reporting, compared with 16 countries in 2000. In general, reporting procedures are defined in legal provisions and many countries supplement these legal provisions with internal rules. For example, in Australia, the Public Service Regulations incorporate the minimum requirements for whistleblowing programme while agency heads are required to establish procedures for dealing with disclosures alleging a breach of the code of conduct in their agency. In addition, in some OECD countries, public servants are obliged by law to report misconduct or any crime, including corruption. For example, Article 40 of the French Penal Procedure Code makes it compulsory for public officials to report suspected violation of the laws, including fraud and corruption, to the public prosecutor. 210. Providing adequate protection – including both legal safeguards and supporting institutional assistance – to whistleblowers has become an increased concern of OECD countries. As of 2009, over half of OECD countries provide some sort of protection to whistleblowers, most often legal. In addition, several countries provide anonymity and others protect whistleblowers against dismissal or other forms of retaliation. Methodology and Definitions: 211. Whistle-blowers are persons who expose wrongdoing in the public service. Data represent central government laws, policies or organisational rules on whistle-blowing that were in place in 2000 and 2009 and whether they provide procedures and protection for whistle-blowers. Data for 2000 were collected by OECD via a tailored survey questionnaire and were originally published in Trust in Government (2000). Respondents to the survey were OECD member country delegates in charge of integrity in central government. Data were updated in 2009 via a tailored survey questionnaire completed by members of the Expert Group on Conflict of Interest. A total of 25 OECD member countries responded to both the 2000 and 2009 surveys. Thus, the data presented in figures XI.2.1 and XI.2.2 represent the percentage of the 25 countries for which data are available in 2000 and 2009 that answered affirmatively to the survey question. Annex C provides unique data for each country on the procedures in place for public servants to report misconduct or suspected corruption and the types of protection offered to whistleblowers. Source: OECD Survey on integrity (2000 and 2009). Further Reading: OECD (2000), Trust in Government: Ethics Measures in OECD Countries, OECD, Paris. Figure Notes: Data are not available for Greece, Korea, the Slovak Republic, Spain and the United Kingdom. 104 GOV/PGC(2009)2 Figure XI.2.1 Procedures for public servants to report misconduct or suspected corruption (2000 and 2009) Figure XI.2.2 Percentage of countries that offer protection for whistleblowers (2000 and 2009) 105 GOV/PGC(2009)2 XI.3 PREVENTING CORRUPTION: PUBLIC PROCUREMENT 212. Knowledge about the main sources and forms of corruption helps decision makers to effectively focus anti-corruption efforts and efficiently disburse available resources. A survey of international business executives conducted by the World Economic Forum in 2006 identified public procurement as the government activity most vulnerable to corruption in OECD countries. 213. Public procurement, the purchase by governments and state-owned firms of goods and services, is a major economic activity of the government where corruption has a potentially high impact on tax payers’ money. The financial interests at stake, the volume of transactions on a global level and the closer interaction between the public and private sectors make it particularly vulnerable to corruption. 214. The size of public procurement markets vary across OECD member countries, ranging from under 10% of the GDP in Greece to over 25% in the Czech Republic, Netherlands and the Slovak Republic. Providing an adequate degree of transparency in the entire procurement cycle is critical to minimize the risk of corruption. Ensuring open and competitive tendering is crucial to maximise transparency in this key stage of the procurement process in order to provide a level playing field for business. The extent that competitive tendering is used by OECD member countries varies greatly; in Greece, approximately 60% of public procurements are competitive compared to less than 10% in the Netherlands and Austria. The frequent use of exceptions to competitive tendering may require other methods to ensure transparency and fairness in the procurement process, such as increased control and audit. Methodology and Definitions: 215. Data on the perception of bribery risk in selected government activities are from the World Economic Forum’s 2006-2007 Global Competitiveness Report and are based on a survey conducted between January and June 2006 of 11,297 top management business leaders in 125 countries. The respondents were asked how commonly firms would make undocumented extra payments or bribes connected with public utilities, tax payments, the award of public contracts and getting favourable judicial decisions (1 = is common, 7 = never occurs).Data on the size of the public procurement market relative to GDP and the degree that competitive tendering is used are from Eurostat and are only available for the 19 OECD member countries that are also members of the European Union. Eurostat defines competitively awarded as procurements that have been advertised in the Official Journal, which covers procurements whose value is above a defined threshold. Source: Eurostat (2006) and World Economic Forum, Executive Opinion Survey (2006) www.weforum.org/documents/gcr0809/index.html Further Reading: OECD (2009), OECD Principles for Integrity in Public Procurement, OECD, Paris OECD (2007), Integrity in Public Procurement: Good Practice from A to Z, OECD, Paris. Lopez-Claros, Augusto, Schwab, Klaus and Porter, Michael E. (ed.) (2006), Global competitiveness report 2006-2007: creating an improved business environment, Palgrave, Houndmills; Basingstoke; New York, N.Y. Figure Notes: (3 and 4) Data are provided for the 19 OECD member countries that are also members of the European Union. The following OECD countries are not included in the European Commission data: Australia, Canada, Iceland, Japan, Korea, Mexico, New Zealand, Norway, Switzerland, Turkey and the United States. 106 GOV/PGC(2009)2 Figure XI.3.1 Average perceived level of bribery risk in selected government activities in OECD countries (2006) 107 GOV/PGC(2009)2 Figure XI.3.2 Integrity in public procurement: Frequency with which firms make undocumented extra payments or bribes in connection with the award of public contracts for investment projects (2006) Figure XI.3.3 Total public procurement as a percentage of GDP (2006) 108 GOV/PGC(2009)2 Figure XI.3.4 Percentage of public procurement that is competitively awarded (2006) 109 GOV/PGC(2009)2 XII. E-GOVERNMENT 216. E-Government has become a key tool to improve performance and enhance innovation in public service development and delivery. E-Government is the use of information and communication technology (ICT) —particularly the Internet — to achieve better government. Beyond the simple exercise of putting information, data and services online, OECD countries are using e-government to transform the structures, processes and culture of government to make the public sector more efficient, effective, user-oriented and transparent. 217. The indicators presented in this chapter evaluate the readiness of a country for e-government development and implementation, and measure the supply, maturity and uptake of e-government services. As the OECD’s work on e-government concentrates on conducting detailed analysis of country practices through country reviews, and in response to member countries’ desires to minimise data collection burdens by using existing information collected internationally, several indicators presented in this chapter were developed by the UN and the European Union (the Commission and Eurostat). 218. The indicators allow readers to evaluate the extent that countries are using ICT to provide public services and benchmark the comparative use of e-government by OECD member nations. 110 GOV/PGC(2009)2 XII.1 E-GOVERNMENT READINESS 219. A high level of readiness to implement and develop e-government services is a prerequisite for a high-performing and innovative public sector that delivers integrated services, making life easier for citizens and businesses. E-Government readiness is therefore a significant indicator of whether a country is prepared to harvest efficiencies gained from ICT-enabled public administrations. 220. The UN’s e-government readiness index is a combined indicator of the supply of, the potential demand for and the maturity of e-government services provided by a country’s government. OECD countries exhibit a high capacity to implement and develop e-government services. This is generally characterized by an extensive broadband infrastructure; a repository of electronic information on government laws and policies, including links to archived information and downloadable forms; and a high level of comfort with ICT by citizens and businesses. Countries with the highest readiness index tend to also have a large amount of transactional and e-commerce features on their government websites. As noted by the UN in its 2008 e-government survey, the Scandinavian countries with the top three scores on the readiness index all generally share similar e-government environments (e.g. the accessibility and penetration of the electronic infrastructure) and web strategies (e.g. the online provision of services). Each country has two main government websites: one that is informative in nature and another that is a gateway for e-government services. In addition, citizens and businesses are able to access a large amount of services and complete many transactions online. (UN 2008) However, similar levels of e-government readiness can also result from different strategic approaches. 221. Internet access is a prerequisite for citizens and businesses to use e-government services, and thus a leading indicator of countries’ readiness to harness the potential efficiencies of ICT. Broadband penetration has increased dramatically in most OECD countries in the past five years as countries have made significant investments in their telecommunications infrastructure Methodology and Definitions 222. The UN E-Government Readiness Index is an internationally agreed-upon composite indicator that measures the capacity of governments to implement and develop e-government services. The index ranges from 0 (low level of readiness) to 1 (high level of readiness). Developed within the framework of the UN Global e-government survey, the indicator consists of three sub-indices: (1) the web measure index, (2) the telecommunication infrastructure index, and (3) the human capital index. The web measure index ranks countries based on the coverage, sophistication, and availability of e-services and e-products. The index categorizes countries as having either an emerging, enhanced, interactive, transactional, or networked e-government presence. The telecommunication infrastructure index is a weighted average of five measures of ICT infrastructure capacity: PCs/100 persons, Internet users/100 persons, Telephone lines/100 persons, Broad-banding/100 persons, and Mobile phones/100 persons. The human capital index is a weighted average of the adult literacy rate (two-thirds weight) and the combined primary, secondary, and tertiary gross enrolment ratio (one-thirds weight). For more information on the methodology used to construct this index see the source listed below. 223. Broadband penetration is the number of broadband subscribers (DSL, Cable, Fibre/LAN, Other) per 100 inhabitants and is based on data collected by the OECD from member country governments. Figure XII.1.2 shows both the cumulative and the incremental change in broadband subscriptions 20032008. Source: United Nations E-Government Readiness Knowledge Base, http://www.unpan.org/egovkb. Data extracted on 30 July 2008. OECD Broadband Portal (2008). Further reading: 111 GOV/PGC(2009)2 United Nations (2008) e-Government Survey, From e-Government to Connected Governance, United Nations Publication, New York. United Nations (2003) UN Global e-Government Survey 2003, United Nations Publication, New York. Figure note: (1) In the case of Austria, the 2008 UN survey did not include the main national e-government portal Help.gv.at. It complements the sites of central, regional and local authorities providing one-stop-shop access to information and transaction services for citizens and businesses thus playing a key role in Austrian e-government. Figure XII.1.1 E-Government Readiness (2008) 112 GOV/PGC(2009)2 Figure XII.1.2 Cumulative broadband subscriptions per 100 inhabitants (2003 - 2008) 113 GOV/PGC(2009)2 XII.2 E-GOVERNMENT SERVICE MATURITY 224. Since 2000, OECD countries have been working towards making all public services for citizens and businesses fully available online. In addition to the potential efficiencies gained by lowering administrative burdens on clients (such as one-stop shops that enable those not familiar with the structure of government to more quickly access online services), developing and implementing integrated, transaction-oriented e-government services often requires governments to standardise and align internal processes and data in order to integrate back-office functions across the public sector. However, some countries have legal or regulatory constraints that limit or prevent them from sharing data for service integration. Nevertheless, e-government service maturity can be an important proxy for the extent to which countries are generating internal efficiencies through the use of ICT and ICT-enabled governance structures. 225. The European Commission has developed an internationally recognized model and 3 indicators to analyse the maturity of 20 core e-government services: the sophistication, full online availability and national portal indicators. The sophistication indicator provides an overall portrait of a country’s progress towards making all services fully available online. The full-online availability indicator evaluates the number of public services that are fully mature and can be entirely handled online (i.e. citizens or businesses can submit completed forms or payments on-line, in addition to finding information about the service). The national portal indicator assesses the degree to which the main government website provides a “one-stop-shop” for users to access public services. 226. On average, OECD countries monitored by the European Commission exhibit a high degree of egovernment service maturity. Austria, Portugal and the United Kingdom are leaders with regard to sophistication and online availability. OECD European member nations show the largest disparity in the percentage of the 20 core public services that are fully available online, ranging from about 20% of services in Switzerland, to almost 100% of services in Austria. Methodology and Definitions 227. The European Commission’s indicators are based on the basket of the following 20 core public services: • 12 citizen-oriented services: income taxes, job search services, social security benefits, personal documents, car registration, building permits, police reports, public libraries, certificates, higher education enrolment, moving announcements and health-related services. • 8 business-oriented services: social contributions for employees, corporate tax, VAT, registration of a new company, submission of data to statistical offices, customs declarations, environment-related permits and public procurement. 228. The three indicators are based on a model defining five progressive stages of e-government service maturity: (1) information, (2) one-way interaction (downloadable forms), (3) two-way interaction (electronic forms), (4) transaction (full electronic case handling), and (5) personalisation (pro-active, automated). The sophistication indicator benchmarks countries according to the development phase of their e-government services, and the index indicates the average level of service maturity. The index ranges from 0 (no services available online) to 1 (a high degree of availability and maturity of e-government services). The full online availability indicator calculates the percentage of services that have reached the fourth stage of maturity and ranges from 0 (no services are fully mature) to 1 (all services are fully mature). The national portal indicator assesses the percentage of core public services accessible through the national portal; whether it provides personalisation options such as log in features; the number of ways users can 114 GOV/PGC(2009)2 access services (such as by department, type of service, or life event); and whether branding and graphics are consistent. The index ranges from 0 (low degree of accessibility) to 1 (high degree of accessibility). For more information on the methodology used to construct the indexes, see the source listed below. Source: EC DGISM (2007), The User Challenge: Benchmarking the Supply of Online Public Services, 7th Measurement, September, 2007. Figure notes: Data are provided for 23 European OECD countries monitored by the European Commission. The following OECD countries are not included in the European Commission data: Australia, Canada, Japan, Korea, Mexico, New Zealand, and the United States. (2) Data for the Czech Republic, Hungary, Poland, the Slovak Republic and Switzerland are from 2004. Figure XII.2.1 – Sophistication of e-government services (2007) 115 GOV/PGC(2009)2 Figure XII.2.2 –Full-online availability of e-government services (2007) Figure XII.2.3 – Assessment of the national portal (2007) 116 GOV/PGC(2009)2 XII.3 UPTAKE OF E-GOVERNMENT SERVICES 229. For e-government to be successful and for its efficiencies to be fully realised, citizens and businesses must be willing to use and have confidence in using e-government services on a regular basis. The maturity of those services and the internal efficiencies associated with providing them can only be realized if people use this delivery channel. 230. A significantly higher percentage of businesses use e-government services than citizens, in part because governments can more easily require the use of digital communications for businesses than they can for citizens. Both citizens’ and business’ use of e-government services has increased in recent years in most OECD countries monitored by the European Commission. However, when compared to the availability of e-government services, citizens’ take-up remains low even for leading countries. 231. Many OECD countries monitored by the European Commission are currently trying to determine why e-government services are not achieving higher rates of uptake and are working to identify the drivers behind higher uptake. Because access to the Internet is a pre-requisite for the use of online services, one driver for uptake is the penetration of broadband infrastructure in society. The data indicate a strong correlation between the penetration of broadband and the use of e-government services by citizens. Methodology and Definitions 232. The indicators on citizen and business uptake of e-government services are based on data collected by Eurostat. The data are part of Eurostat’s Information Society Statistics database which evaluates the share of citizens and businesses using the Internet to interact with public services. Data are collected for only European countries and are available for some countries beginning in 2002 for citizen uptake and in 2003 for business uptake; the figures present data for the earliest year available for each country. 233. The e-government take-up by citizens indicator measures the percentage of individuals (aged 1674) that used the Internet to interact with public authorities in the three months preceding the survey. Data are collected through Eurostat’s annual Community Survey on ICT Usage in Households and by Individuals. 234. The e-government take-up by businesses indicator measures the percentage of enterprises using the Internet to interact with public authorities. Only businesses with ten or more persons employed are included. Data are collected by national statistical offices based on Eurostat's annual model survey on ICT usage and e-commerce in businesses. 235. OECD calculated the correlation between broadband penetration and citizen uptake of egovernment services using an OECD-defined indicator of broadband penetration and Eurostat data on citizen uptake. The broadband penetration indicator is the number of broadband subscribers (DSL, Cable, Fibre/LAN, Other) per 100 inhabitants. The coefficient of correlation is 0.803. Source: (1&2): Eurostat (2008) http://ec.europa.eu/eurostat. Data accessed on 5th June 2008. (3): OECD 2007 Broadband statistics (www.oecd.org/sti/ict/broadband) and Eurostat (2008). Figure notes: Data are provided for 21 OECD countries monitored by the European Commission. The following OECD countries are not included in the European Commission data: Australia, Canada, Japan, Korea, Mexico, New Zealand, Switzerland, Turkey and the United States. 117 GOV/PGC(2009)2 Figure XII.3.1 Percentage of citizens using e-government services (2008 and earliest available year) . earliest available year: i:2002 ; ii:2003 ; iii:2004 ; iv:2005 ; v:2006 118 GOV/PGC(2009)2 Figure XII.3.2. Percentage of businesses using e-government services (2008 and earliest available year) . earliest available year: i:2002 ; ii:2003 ; iii:2004 ; iv:2005 ; v:2006 119 GOV/PGC(2009)2 Figure XII.3.3. Relationship between broadband penetration and citizen uptake of e-government services (2008) 120 GOV/PGC(2009)2 XIII. OPEN GOVERNMENT 236. Open government is now recognised by OECD member countries as an essential ingredient for democratic governance, social stability and economic development. Over the past 25 years, OECD countries have moved from a situation where governments chose what they reveal, to a principle of all government information being available, unless there is a defined public interest in it being withheld. Open government refers to the transparency of government actions, the accessibility of government services and information and the responsiveness of government to new ideas, demands, and needs. The trend towards more open government has come largely at the behest of the public. As the complexity of societies has increased, the reach of government regulation has penetrated deeper into public lives and the volume and scope of information about individuals collected by government has grown. Better-educated and informed citizens increasingly demand accountability for government decision making and the right to challenge administrative decisions and seek redress for failures of government. 237. This chapter tracks OECD member governments’ response to demands for greater openness by passing legislation on access to information, privacy and data protection and administrative procedures, and creating Ombudsman offices and Supreme Audit Institutions. Citizens can compare the legal framework for open government across countries. 121 GOV/PGC(2009)2 XIII.1. OPEN GOVERNMENT LEGISLATION 238. The scope, quantity and quality of government information provided to the public have increased significantly in the past 25 years as OECD countries have passed laws guaranteeing transparency and public access to information, and set up institutions to ensure that governments are accountable and responsive to the public. 239. Today, the legal framework for open government is largely in place in OECD countries, and consists of: • Laws on access to information which seek to give citizens the right to access information held by government. This is a precondition for public scrutiny and participation; • Laws on privacy and data protection set parameters directing when information cannot be provided to the public in the interest of protecting personal data; • Laws on administrative procedures provide some guarantees for citizens in their interactions with government and establish mechanisms for holding administrative bodies accountable; • Laws on Ombudsman institutions establish a point of contact for citizens’ complaints, appeals and claims for redress in their dealings with government entities; • Laws on Supreme Audit Institutions provide for an independent review of public accounts as well as of the execution of government programmes and projects. 240. This legal framework has evolved steadily over the past 50 years. Among the 24 countries who were OECD members in 1980, less than a third had legislation on access to information. In general, Nordic and Anglo-Saxon countries were the first to adopt these laws. The momentum for open government gained in the 1990s, and by 2001, the number of countries that had passed access to information legislation had almost doubled. By 2008, 29 of the 30 member countries had adopted access to information laws. Some OECD countries have even incorporated open government principles into their constitutions (e.g. Austria, Hungary and Poland). 241. The institutional framework needed to implement and enforce these laws has grown apace. In addition to the legal guarantees for the privacy of personal data adopted in most OECD countries, over two-thirds of OECD countries have established parliamentary commissioners for data protection and privacy. In 1960, only Sweden, Finland and Denmark had an Ombudsman office; today, 90% of OECD countries have such institutions or their equivalents. All OECD countries have a Supreme Audit Institution, in most cases an independent authority which reports directly to the legislature. Methodology and Definitions 242. The data are an excerpt of a wider set of data which were last updated in April 2008 by government representatives at a meeting of the OECD Public Governance Committee. 243. The two figures presenting cumulative trends in the adoption of laws on access to information and the establishment of an ombudsman institution in OECD countries are based on the date of first passage of legislation at the national level. The ombudsman is an independent investigator who intervenes on behalf of citizens alleging they have been treated wrongly by bureaucracy. 122 GOV/PGC(2009)2 Source: OECD (2009), Focus on Citizens: Public Engagement for Better Policy and Services, OECD Publishing, Paris. Figure notes: (1) While Italy does not have a national ombudsman, extensive coverage is provided by sub-national ombudsman institutions and a government-appointed commission oversees implementation of the law on access to public information. Turkey passed a law on a national ombudsman in 2006 which was subsequently suspended by a Constitutional Court ruling. Further readings: OECD (2009), Focus on Citizens: Public Engagement for Better Policy and Services, OECD, Paris. OECD (2005), Modernizing Government: The Way Forward, OECD, Paris. OECD (2001), Citizens as Partners: information, consultation and public participation in policy making, OECD, Paris. Figure XIII.1 Overview of current legislation and institutions for open government (2008) Australia Austria Belgium Canada Czech Republic Denmark Finland France Germany Greece Hungary Iceland Ireland Italy Japan Korea Luxembourg Mexico Netherlands New Zealand Norway Poland Portugal Slovak Republic Spain Sweden Switzerland Turkey United Kingdom United States European Union Freedom of information z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z Privacy/data protection z z z z z z z z z z z z z z z z z z z z z z z z z z z z z Administrative procedures z z z z z z z z z z z z z z z z z z z z z z z z z Ombudsman/ Commissioner z z z z z z z z z z z z z z z z z z z z z z z z z z z z Supreme audit institution z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z z Legislation in place: z 123 GOV/PGC(2009)2 Figure XIII.2 Number of OECD countries with laws on access to information, 1960-2008 Figure XIII.3 Number of OECD countries with laws on ombudsman institutions, 1960-2008 124 GOV/PGC(2009)2 ANNEX A: CLASSIFICATION OF THE FUNCTIONS OF GOVERNMENT (COFOG) First- and second-level COFOG Developed by the OECD, the Classification of the Functions of Government (COFOG) classifies government expenditure data from the System of National Accounts by the purpose for which the funds are used. As Table 1 illustrates, first-level COFOG splits expenditure data into 10 “functional” groups or subsectors of expenditures (such as defence, education and social protection), and second-level COFOG further splits each first-level group into up to nine sub-groups. While first-level COFOG data are available for 26 out of the 30 OECD member countries, second-level COFOG data are currently only available for 13 OECD European member countries1 . Table 1: First and Second Level COFOG First-level Second-level General public services • Executive and legislative organs, financial and fiscal affairs, external affairs • Foreign economic aid • General services • Basic research • R&D General public services • General public services n.e.c. • Public debt transactions • Transfers of a general character between different levels of government Defence • Military defence • Civil defence • Foreign military aid • R&D Defence • Defence n.e.c. Public order and safety • Police services • Fire-protection services • Law courts • Prisons • R&D Public order and safety • Public order and safety n.e.c. Economic affairs • General economic, commercial and labour affairs • Agriculture, forestry, fishing and hunting • Fuel and energy 1 First-level COFOG data are not available for Mexico, Switzerland and Turkey. Until recently, second-level COFOG data were available in some national statistical offices, but were not collected by international organizations. Moreover, the second-level COFOG data were not comparable between countries because the SNA guide and the IMF Manual on Government Finance Statistics do not provide much practical information on the application of COFOG concepts. However, in 2005, EUROSTAT established a task force to develop a manual on the application of COFOG to National Account expenditure data and to discuss the collection of second-level COFOG data for European countries. Second-level COFOG data are not available for Switzerland and all non-European member countries of the OECD: Australia, Canada, Japan, Korea, Mexico, New Zealand and the United States. Efforts are underway to reach agreement with these countries about the submission of these data to the OECD. 125 GOV/PGC(2009)2 • Mining, manufacturing and construction • Transport • Communication • Other industries • R&D Economic affairs • Economic affairs n.e.c. Environmental protection • Waste management • Waste water management • Pollution abatement • Protection of biodiversity and landscape • R&D Environmental protection • Environmental protection n.e.c. Housing and community amenities • Housing development • Community development • Water supply • Street lighting • R&D Housing and community amenities • Housing and community amenities n.e.c. Health • Medical products, appliances and equipment • Outpatient services • Hospital services • Public health services • R&D Health • Health n.e.c. Recreation, culture and religion • Recreational and sporting services • Cultural services • Broadcasting and publishing services • Religious and other community services • R&D Recreation, culture and religion • Recreation, culture and religion n.e.c. Education • Pre-primary and primary education • Secondary education • Post-secondary non-tertiary education • Tertiary education • Education not definable by level • Subsidiary services to education • R&D Education • Education n.e.c. Social protection • Sickness and disability • Old age • Survivors • Family and children • Unemployment • Housing • Social exclusion n.e.c. • R&D Social protection • Social protection n.e.c n.e.c means “not elsewhere classified.” 126 GOV/PGC(2009)2 COFOG-Special “COFOG-Special” is a method to classify government expenditures into individual and collective goods, and cash transfers and goods and services in kind2. It was developed by the OECD beginning in 2004 at the request of the OECD’s Working Party of Senior Budget Officials and is based on second-level COFOG expenditure data. Table 2 illustrates how second-level COFOG data are split into collective or individual goods and services, and as cash transfers or goods and services in kind. It should be noted that some first-level COFOG groups do not need to be split further because all of their expenditures are similar in terms of individual/collective and cash transfers/in kind goods and services. For example, all expenditures on Defence and Public Order and Safety can be classified as collective and in kind. Further details about the COFOG-Special methodology are available in OECD (2007), “COFOG- Special: Working paper on test procedure for five European countries”, OECD, Paris. Table 2: Methods used to construct COFOG-Special Data Collective/ Cash/in COFOG-Special Source (table 1101 and 1102) and National Accounts individual kind COFOG 1st and 2nd level Economic classification R+D Collective goods In kind General governance services Remainder General public services (701) All Basic research Basic research (7014)* All Defence Defence (702) All Public order and safety Public order and safety (703) all Infrastructure and network services Remainder Economic Affairs (704) All minus subsidies Environmental protection, development and community services Remainder Environmental protection (705) All minus subsidies Remainder Public housing and community amenities (706) All minus subsidies Service regulation General economic, commercial and labour affairs (7041)** Intermediate consumption, compensation of employees, gross capital formation Agriculture, forestry, fishing and hunting (7042** Intermediate consumption, compensation of employees, gross capital formation Mining, manufacturing, and construction(7044**) Intermediate consumption, compensation of employees, gross capital formation Other industries (7047)** Intermediate consumption, compensation of employees, gross capital formation R+D Economic affairs (7048)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Economic affairs (7049) ** Intermediate consumption, compensation of employees, gross capital formation R+D Environmental protection (7055)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Environmental protection (7056) ** Intermediate consumption, compensation of employees, gross capital formation R+D Housing and community amenities (7065)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Housing and community amenities (7066) ** Intermediate consumption, compensation of employees, gross capital formation 2 Collective goods and services benefit the community at large and include expenditures on defence, public safety and order, legislation and regulation. Individual goods and services mainly benefit individuals and include education, health and social insurance programs. Cash transfers refer to monies provided to beneficiaries by governments and are not required to be spent on a specific good or service. Examples of cash transfers include pensions and unemployment benefits. In kind transfers mean that the government pays only for specific goods or services, either by providing (or contracting for the provision of) goods and services directly or by reimbursing households for their expenses. Examples of in kind transfers include housing vouchers and most health and education services. 127 GOV/PGC(2009)2 R+D health (7075) Not elsewhere classified health (7076)** Intermediate consumption, compensation of employees, gross capital formation R+D Recreation, culture and religion (7085)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Recreation, culture and religion (7086)** Intermediate consumption, compensation of employees, gross capital formation R+D Education(7097)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Education (7098)** Intermediate consumption, compensation of employees, gross capital formation R+D Social protection(7108)** Intermediate consumption, compensation of employees, gross capital formation Not elsewhere classified Social protection (7109)** Intermediate consumption, compensation of employees, gross capital formation Cash Foreign aid transfers Foreign economic aid (7012) * Current and capital transfers General purpose and block grants Transfers of a general character between different levels of government (7018)*** Other current transfers between sub-sectors Interest National accounts, Interest payable Individual goods In kind Non-market recreation, culture and religion Remainder Recreation, culture and religion (708) All minus subsidies Social services Remainder Social protection (710) All Health Remainder Health (707) All Education Remainder Education (709) All Market subsidies Economic affairs (704) Subsidies Environmental protection (705) Subsidies Housing and community amenities. (706) Subsidies Recreation, culture and religion (708) Subsidies Cash Social cash transfers Cash part of Social protection (710) Social benefits * The general government figure is attributed to central government (source table 1102). ** The figures for the sub-sectors are computed on the basis of the proportion of service regulation in the COFOG-special groups in general government, separately for intermediate consumption, compensation of employees and gross capital formation (source table 1102). *** For the sub-sectors estimated by taking all Other current transfers between sub-sectors in 1st level COFOG Group General Public Services (706) except the imputed Other current transfers on Basic research and Foreign economic aid. 128 GOV/PGC(2009)2 ANNEX B: COMPOSITE INDICATORS FOR HRM, REGULATORY MANAGEMENT AND BUDGET PRACTICES The narrowly defined composite indicators presented in Government at a Glance represent the best way of summarizing discreet, qualitative information on key aspects of public management practices, such as the openness of HRM systems or flexibility in budgeting practices. “Composite indicators are much easier to interpret than trying to find a common trend in many separate indicators” (Nardo et. al., 2004). The composites presented in this publication adhere to the steps identified in the Handbook on Constructing Composite Indicators (Nardo et. al., 2008) that are necessary for the meaningful construction of composite or synthetic indicators: • Each composite index is based on a theoretical framework representing an agreed upon concept in the area it covers. The variables comprising the indexes were selected based on their relevance to the concept by a group of experts within the OECD and in consultation with country delegates to the relevant working parties. • Various statistical tools – such as factor analysis and the computation of Cronbach’s alpha— were employed to establish that the variables comprising each index are correlated and represent the same underlying concept. • Different methods for imputing missing values have been explored. • All sub-indicators and variables were normalized for comparability. • Several weighting options (e.g. equal weighting, factor weights and expert weights) were tested to examine each index’s sensitivity. • To build the composites, all sub-indicators were aggregated using a linear method according to the accepted methodology. • Sensitivity analysis was carried out in order to establish the robustness of the indicators to varying assumptions. While the composite indicators were developed in cooperation with member countries and are based on best practices and/or theory, both the variables comprising the composites and their weights are offered for debate and, consequently, may evolve over time. 129 GOV/PGC(2009)2 B.1. HRM Composites: Theoretical framework, construction and weighting All the data in the composite indexes for HRM practices derive from the OECD (GOV) 2006 Survey on Strategic Human Resource Management. The composites, including the variables comprising each index and their relative weights, are based on concepts that reflect contemporary public sector HRM developments and dilemmas on how best manage human resources in the public sector in the twenty first century (e.g. degree of openness of HRM systems, extent of decentralization, use of performance–based practices) and were reviewed by the OECD’s Working Party on Public Employment and Management. In addition, the variables’ statistical relevance to the underlying concept was verified using factor analysis3 and by computing a coefficient of reliability, called Cronbach’s alpha.4 Missing values were an issue for the Strategic Human Resource Management database. Different techniques to estimate missing values were applied based on the nature of the missing information, including mean replacement, expert judgment and/or eliminating the country from the calculation of the composite indicator. In order to eliminate scale effects, all the variables were normalized between “0” and “1” prior to the final computation of the index. After testing several weighting options (including equal weighting and factor weights), the final indexes were built on equal weights based on expert judgement. Aggregation is based on the linear method, as alternative methods – such as the geometric aggregation –were technically not relevant. Finally, robustness checks based on Monte Carlo simulations were conducted to test the sensitivity of the indexes to different weighting assumptions. 3 Factor analysis is a statistical check that looks at how a set of variables are associated and whether they are correlated with each other. Factor analysis is based on the idea that if there is a significant correlation among the variables for each composite, then no essential insight is lost by reducing this large set of variables into a smaller one (i.e. a composite). From a technical point of view, correlated original variables can be transformed through linear combinations into a new, smaller set of uncorrelated underlying variables that form a composite index. Variables with factor loadings less than 0.3 are statistically insignificant (i.e. not correlated with other variables) and thus not crucial theoretically. Such variables were omitted from the final computation of the index. 4 Cronbach’s Alpha is a coefficient of reliability based on the correlations between indicators. This statistic is generally used to investigate the degree of correlation among a set of variables and to check the internal reliability of items in a model or survey. A Cronbach’s alpha coefficient equal to zero means that the variables are independent (i.e. the selection is not correlated and therefore is statistically not relevant), while a coefficient equal to one means that the variables are perfectly correlated. In general, a coefficient of 0.6 or 0.7 is considered to be an acceptable indication that the variables are measuring the same underlying construct. 130 GOV/PGC(2009)2 B.1.1. Delegation in the public service Variables The following items have been used in the construction of this index, and were assigned equal weights: A. Existence of a central HRM body (Q.20). B. Delegation of establishment (Q.21). C. Delegation of decision on compensation levels (Q.24). D. Delegation of decisions regarding position classification, recruitment and dismissals (Q.27). E. Delegation of decisions related to other conditions of employment (Q.30). F. Impact of delegation for pay/terms and conditions of employment across government organizations (Q.36). Scoring A. Is there a central agency/department in charge of human resources at central/national/federal government level? (Question 20) • Yes: 1.000 • No: 0.000 • Not responsible, but a central agency/department aims to coordinate the HR policies across departments: 0.500 B. Delegation of establishment (Question 21) Primarily determined by: Central HRM body (which sets the rules and is closely involved in applying them)/Ministry of Finances Central HRM body but with some latitude for ministries/departments/ agencies in applying the general principles Ministries/ departments/ agencies, within established legal and budgetary limits Unit/team level Other/variable, depends largely on departments/ functions 131 GOV/PGC(2009)2 Numbers and types of posts within organisations 0.250 0.500 0.750 1.000 0.625 Allocation of budget envelope between payroll and other expenses 0.250 0.500 0.750 1.000 0.625 C. Delegation of decisions regarding compensation levels (Question 24) The overall score for question 24 was calculated by averaging the sub-scores. Primarily determined by: Central HRM body (which sets the rules and is closely involved in applying them)/ Ministry of Finances Central HRM body but with some latitude for ministries/departments/ agencies in applying the general principles Ministries/ departments/ agencies, within established legal and budgetary limits Unit/team level Other/variable, depends largely on departments/ functions General management of pay systems (fixed portion, progressions,…) 0.250 0.500 0.750 1.000 0.625 Management of the variable portion of pay - benefits - performance related pay 0.250 0.500 0.750 1.000 0.625 D. Delegation of decisions regarding position classification, recruitment and dismissals (Question 27). The overall score for question 27 was calculated by averaging the sub-scores. Primarily determined by: Central HRM body (which sets the rules and is closely involved in applying them) Central HRM body but with some latitude for ministries /departments/ /agencies in applying the general principles Ministries/ departments/ agencies, within established legal and budgetary limits Unit/team level Other/variable, depends largely on departments/ functions Post classification system – grades 0.250 0.500 0.750 1.000 0.625 Original individual recruitment into the civil service 0.250 0.500 0.750 1.000 0.625 Individual recruitment of casual staff 0.250 0.500 0.750 1.000 0.625 Individual duration of employment contract in the civil service 0.250 0.500 0.750 1.000 0.625 Individual duration of contract in specific posts 0.250 0.500 0.750 1.000 0.625 Individual career management 0.250 0.500 0.750 1.000 0.625 Individual dismissal 132 GOV/PGC(2009)2 - following lack of performance - following organisational restructuring - following misconduct 0.250 0.250 0.250 0.500 0.500 0.500 0.750 0.750 0.750 1.000 1.000 1.000 0.625 0.625 0.625 E. Delegation of decisions related to other conditions of employment (Question 30) The overall score for question 30 was calculated by averaging the sub-scores. Primarily determined by: Central HRM body (which sets the rules and is closely involved in applying them) Central HRM body but with some latitude for ministries/ departments/ agencies in applying the general principles Ministries/ departments/ agencies, within established legal and budgetary limits Unit/team level Other/variable, depends largely on departmental/ functions Flexibility of working conditions (numbers of hours, etc.) 0.250 0.500 0.750 1.000 0.625 Adjustments to working conditions (part time, etc.) 0.250 0.500 0.750 1.000 0.625 Performance appraisal systems 0.250 0.500 0.750 1.000 0.625 Code of conduct 0.250 0.500 0.750 1.000 0.625 Ethics, equal opportunity, equity issues 0.250 0.500 0.750 1.000 0.625 F. Do the levels of pay/terms and conditions of employment vary significantly across government organizations for the same level of posts? (Question 36) The overall score for question 36 was calculated by averaging the sub-scores. For each item below, the weightings are: • Yes significantly: 1.000 • Yes to some degree: 0.500 • Not at all: 0.000 Within central/national/federal government across ministries/departments/agencies: 1)Basic pay, 2) Other types of remuneration/social benefits, 3) Duration of individual contracts in specific posts (mandates), 4) General duration of employment in the public service, 5) Flexibility of working conditions (number of hours, etc.), 6) Adjustments to working conditions (part-time, etc.), 7) Performance appraisal systems, 8) Code of conduct, 9) Ethics, equal opportunity, equity issues. Within the same ministries/departments: 1)Basic pay, 2) Other types of remuneration/social benefits, 3) Duration of individual contracts in specific posts (mandates), 4) General duration of employment in the public service, 5) Flexibility of 133 GOV/PGC(2009)2 working conditions (number of hours, etc.), 6) Adjustments to working conditions (part-time, etc.), 7) Performance appraisal systems, 8) Code of conduct, 9) Ethics, equal opportunity, equity issues. 134 GOV/PGC(2009)2 B.1.2. HRM openness of the public service Variables The following items have been used in the construction of this index, and were assigned equal weights: A. Becoming a public servant in general (Q.37). B. Merit-based appointment at entry in the selection process (Q.38). C. Allocation of posts (Q.39). D. Recruitment of senior civil servants (Q.81). Scoring A. How does one become a public servant in general? (Question 37) • Through a competitive examination that provides for entry into a specific group of the public service: 0.000 • Through direct application to a specific post and interview (with minimum degree requirement) : 0.666 • It varies depending on the post: 0.333 B. If your system does not include competitive written examination how is merit-based appointment at entry guaranteed in the selection process? (Question 38) Equal weghting (0.200) was attributed to the following items that can be combined with each other: • All vacancies are published • Recruitments are made with panels • Recruitment firms are used 135 GOV/PGC(2009)2 • Shortlist of possible candidates prepared jointly by the HR department and the recruiting department • Other C. How does one individual get a specific position/post? (Question 39) • All posts are open to internal and external recruitment and applicants apply to the different posts: 1.000 • Most posts are open to internal and external recruitment and applicants apply to the different posts: 0.666 • Some posts are open to internal and external recruitment and applicants apply to the different posts: 0.333 • No posts open to external recruitment: 0.000 D. How are the rest of senior managers recruited? (Q.81) • Originally selected by competitive examination early on in their careers and managed as a group after: 0.000 • Through career progression within the public service only: 0.000 • All senior management positions are open to external recruitment: 1.000 • A good proportion of management positions are open to external recruitment: 0.500 136 GOV/PGC(2009)2 B.1.3. Intensity of performance-related pay (PRP) Variables The following items have been used in the construction of this index, and were assigned equal weights: A. The use of performance-related pay mechanism (Q.76). B. PRP and staff categories (Q.76.a). C. The use of one-off bonuses and/or merit increments (Q76.b). D. Maximum proportion of basic salary that PRP represent (Q.76.c.). Scoring A. Is performance related pay in use in your country? (Question 76) • Yes: 1.000 • No : 0.000 B. If PRP implemented, for which category of staff? (Question 76.a) • For most government employees: 1.000 • For senior staff only: 0.666 • Only in a few central/national/federal government organisations: 0.333 C. Do organisations mostly use: (Question 76.b) • One-off bonuses: 0.500 • Merit increments: 0.500 D. What is the maximum proportion of basic salary that PRP can represent?(Question 76.c.) 137 GOV/PGC(2009)2 • 1-5 % : 0.250 • 1-10%: 0.500 • 1-20%: 0.750 • 1-40%: 1.000 138 GOV/PGC(2009)2 B.1.4. Use of performance assessment in HRM decisions for government employees Variables The performance assessment index encompasses the following items, which were assigned equal weights: A. Career advancement (Q.74.1). B. Remuneration (Q.74.2). C. Contract renewal on the same job/ remaining in the same job (Q.74.3). D. Employment contract renewal in the public service (Q.74.4). Scoring A. Importance of performance assessment in relation to career advancement (Question 74.1) • High: 1.000 • Medium: 0.666 • Low: 0.333 B. Importance of performance assessment in relation to remuneration (Question 74.2) • High: 1.000 • Medium: 0.666 • Low: 0.333 C. Importance of performance assessment in relation to contract renewal on the same job (Question 74.3) • High: 1.000 139 GOV/PGC(2009)2 • Medium: 0.666 • Low: 0.333 D. Importance of performance assessment in relation to employment contract renewal (Question 74.4) • High: 1.000 • Medium: 0.666 • Low: 0.333 140 GOV/PGC(2009)2 B.1.5. Intensity of separated group management of senior civil servants Variables The following items have been used in the construction of this index, and were given equal weights: A. Group of staff in central/national government widely understood to be the “senior civil service” (Q.78). B. Identification of potential future leaders early in their careers (Q.82). C. Process of identification (Q.82.a). D. Existence of a centrally defined skills profile for senior staff (Q.82.b). E. The average age upon entering the senior management group (Q.83). F. The difference between the employment framework of senior management and regulalar staff (Q.84). Scoring A. Is there a defined group of staff in central/national/federal government who are widely understood to be the “senior civil service”? (Question 78) • Yes: 1.000 • No: 0.000 B. Does your government officially identify potential future leaders early on in their careers? (Question 82) • Yes: 1.000 • No: 0.000 C. Is this process used to identify potential future leaders early on in their careers formal or informal? (Question 82.a) 141 GOV/PGC(2009)2 • Informal: 0.500 • Formal: 1.000 D. Is there a centrally defined skills profile for senior staff? (Question 82.b) • Yes: 1.000 • No : 0.000 E. What is the average age upon entering the senior management group? (Question 83) • 20-30 years: 1.000 • 30-40 years: 0.750 • 40-50 years: 0.500 • More than 50 years: 0.250 F. The difference between the employment framework of senior management and regulalar staff (Question 84) Equal weighting (0.100) were attributed to the following items that can be combined at each other: • They are recruited with a more centralised process, • They are identified early on in their careers and more attention is paid to the management of their careers, • More emphasis is put into the management of their performance, • More emphasis is put into avoiding major conflicts of interest, • The part of their pay that is not basic salary but not performance-related is higher than for regular staff, • The part of their pay that is performance related is higher, • Their appointment contracts into a post has a specific term, • Their appointment term is shorter than for regular staff, • Their appointment into the senior management group is dependent on the renewal of their contract for a senior management post, • Other. 142 GOV/PGC(2009)2 B.2. Budget Composites: Theoretical framework, construction and weighting The composite indexes for budget practices and procedures, including the variables comprising each index and their relative weights, are based on political-economy literature on budget institutions (e.g. von Hagen, 1992, Alesina et.al, 1999, Hallerberg et. al, 2007). In addition, the variables’ statistical relevance to the underlying concept was verified using factor analysis5 and by computing a coefficient of reliability, called Cronbach’s alpha.6 There were relatively few missing values in the Budget Practices and Procedures Survey database and all missing data points were considered as “0” or negative responses. There were many indefinite “other” responses in the survey which were recoded to a definite answer based on the analysis of country comments available in database. In order to eliminate scale effects, all the variables were normalized between “0” and “1” prior to the final computation of the index. The weights for the underlying variables were selected based on the theoretical framework, and variables were aggregated using the linear method. Finally, robustness checks based on Monte Carlo simulations were conducted to test the sensitivity of the indexes to different weighting assumptions. 5 Factor analysis is a statistical check that looks at how a set of variables are associated and whether they are correlated with each other. Factor analysis is based on the idea that if there is a significant correlation among the variables for each composite, then no essential insight is lost by reducing this large set of variables into a smaller one (i.e. a composite). From a technical point of view, correlated original variables can be transformed through linear combinations into a new, smaller set of uncorrelated underlying variables that form a composite index. Variables with factor loadings less than 0.3 are statistically insignificant (i.e. not correlated with other variables) and thus not crucial theoretically. Such variables were omitted from the final computation of the index. 6 Cronbach’s Alpha is a coefficient of reliability based on the correlations between indicators. This statistic is generally used to investigate the degree of correlation among a set of variables and to check the internal reliability of items in a model or survey. A Cronbach’s alpha coefficient equal to zero means that the variables are independent (i.e. the selection is not correlated and therefore is statistically not relevant), while a coefficient equal to one means that the variables are perfectly correlated. In general, a coefficient of 0.6 or 0.7 is considered to be an acceptable indication that the variables are measuring the same underlying construct. 143 GOV/PGC(2009)2 B.2.1. Transparency Index Variables The transparency index is constructed from three sub-indexes of varying weights. Within each sub-index, questions are equally weighted. 1. Budget documentation (weight 50%) A. Public availability of economic forecast methodology (Question 7) B. Inclusion of a fiscal sensitivity analysis of the economic assumptions (Question 8) C. Distinction between current commitments and new policies (Question 11) D. Years covered by projections (Question 12.a). E. Level of detail of multi-year expenditures estimates (Question 16) F. Frequency that revenue and expenditure information are published(Question 56) G. Public availability of information on performance against targets (Question 82) 2. Specific disclosure (weight 25%) A. Types of off-budget expenditures included in budget documentation (Question 45) B. Notation of contingent liabilities (Question 48) 3. Accountability and control (weight 25%) A. Independent preparation of budget for Supreme Audit Institution (Question 30) B. Existence of a legislative budget research office (Question 34) C. Proportion of line ministries that have internal audit units (Question 66.b) D. Schedule for disseminating audits to the public (Question 70) 144 GOV/PGC(2009)2 Scoring Budget documentation (weight 50%) A. Is the methodology used for establishing the economic assumptions of the budget publicly available? (Question 7) • No: 0.000 • Yes, but only to certain parts of the Legislature: 0.333 • Yes, it is publicly available on request: 0.666 • Yes, it is published: 1.000 Countries reporting “other” were scored as follows: • Belgium: 0.500 • Italy: 0.000 B. Is fiscal sensitivity analysis of the economic assumptions included in the budget documentation? (Question 8) • No: 0.000 • Yes, a comprehensive analysis is included: 1.000 • Yes, but only for selected activities: 0.500 C. In the annual budget documentation presented to the Legislature are expenditures under current commitments in law and policy distinguished from new policies? (Question 11) • No: 0.000 • Yes, but not in all cases: 0.500 • Yes, comprehensively: 1.000 Countries reporting “other” were scored as follows: • Denmark: 0.500 D. How many years do the [long-term] projections normally cover? (Question 12.a) • Not applicable: 0.000 • Up to 10: 0.333 • 11-20: 0.333 • 21-30: 1.000 • 31-40 : 1.000 145 GOV/PGC(2009)2 • 41-50: 1.000 • 51-60: 1.000 • 61-70: 1.000 • 71-80: 1.000 • 81-90: 1.000 • 91- 100: 1.000 E. Does the annual budget documentation submitted to the Legislature contain multi-year expenditure estimates? (Question 16) • No: 0.000 • Yes, at the aggregate level: 0.333 • Yes, at the ministry level: 0.666 • Yes, at line item level: 1.000 Countries reporting “other” were scored as follows: • Australia: 1.000 • Belgium: 1.000 • Finland: 0.333 • Iceland: 0.666 • Norway: 0.666 • Portugal: 0.333 • Korea : 0.333 • Spain: 0.333 F. How frequently do you publish information on actual revenues and expenditures during the fiscal year? (Question 56) • Weekly: 1.000 • Monthly: 0.833 • Quarterly: 0.667 • Every six months: 0.500 • Annually: 0.333 • On an ad-hoc basis: 0.167 146 GOV/PGC(2009)2 • Not at all: 0.000 Countries reporting “other” were scored as follows: • Denmark: 0.600 • Finland: 0.600 • Korea: 0.600 • Switzerland: 0.600 G. Is information on performance against targets made available to the public? (Question 82) • No, it is not made available to the public: 0.000 • Yes, a Government-wide report on performance is published: 0.166 • Yes, individual ministries publish reports on their performance: 0.166 • Yes, as part of other Government-wide documents: 0.166 • Yes, as part of other ministry-specific documents: 0.166 • Yes, there is an internet site for this information: 0.166 • Other: 0.166 Specific disclosure (weight 25 %) A. Types of off-budget expenditures [exist, require legislative authorisation and are] included in the budget documentation (Question 45). This is a multiple choice question and respondents were to select all items that applied. Each country starts with a score of “1” and loses 0.067 points for each type of off-budget expenditure that is applicable but is not included in the budget documentation. Countries do not lose any points for items that are not applicable; thus, countries with high scores either have no or few off-budget expenditures and/or include most off-budget expenditures in the budget documentation. Countries with low scores have many off-budget expenditure items that are not included in the budget documentation. 45.a Tax expenditures (exemptions, deductions and credits) • Is it applicable? • Is it included in the budget documentation? 45.b Social security funds • Is it applicable? • Is it included in the budget documentation? 45.c Public Private Partnerships • Is it applicable? 147 GOV/PGC(2009)2 • Is it included in the budget documentation? 45.d Off-budget loans • Is it applicable? • Is it included in the budget documentation? 45.e Loan Guarantees • Is it applicable? • Is it included in the budget documentation? 45.f Public health care Funds • Is it applicable? • Is it included in the budget documentation? 45.g Public sector pensions • Is it applicable? • Is it included in the budget documentation? 45.i Funds for higher education establishments • Is it applicable? • Is it included in the budget documentation? 45.j Emergency/contingency funds • Is it applicable? • Is it included in the budget documentation? 45.h Donor funds • Is it applicable? • Is it included in the budget documentation? 45.k Stabilization funds • Is it applicable? • Is it included in the budget documentation? 45.l Expenditure financed by external loans • Is it applicable? • Is it included in the budget documentation? 148 GOV/PGC(2009)2 45.n Military expenditure • Is it applicable? • Is it included in the budget documentation? 45.m Special accounts managed by the Ministry of Finance/Treasury/Central Budget Authority • Is it applicable? • Is it included in the budget documentation? 45.o Other • Is it applicable? • Is it included in the budget documentation? B. Are Government’s contingent liabilities noted in the budget documentation (or in other documents) that is presented to the Legislature? (Question 48) • Yes: 1.000 • No: 0.000 Countries with missing answers were scored as follows: • Finland: 0.000 • Greece: 0.000 Accountability and control (weight 25%) A. In practice, which option most accurately describes the way in which the budget for the Supreme Audit Institution is prepared? (Question 30) • The Supreme Audit Institution prepares its budget and the Central Budget Authority includes it in the Government’s budget proposal without any changes: 0.333 • The Supreme Audit Institution is subject to the same procedures and policies as any other governmental organisation included in the Government’s budget proposal: 0.667 • The Supreme Audit Institution prepares its budget and submits it directly to the Legislature for approval: 1.000 B. Is there a specialised budget research office/unit attached to the Legislature to conduct analyses of the budget? (Question 34) • No: 0.000 • Yes, there is a specialised budget research office/unit : 1.000 Countries reporting “other” were scored as follows: • Canada: 1.000 149 GOV/PGC(2009)2 • Finland: 1.000 • New Zealand: 1.000 • Poland: 1.000 C. What proportion of line ministries have internal audit units? (Question 66.b) • Not applicable: 0.000 • 0-20%: 0.600 • 21-40%: 0.700 • 41-60%: 0.800 • 61-80%: 0.900 • 81-100%: 1.000 D. When are the accounts audited by the Supreme Audit Institution made publicly available? (Question 70) • 1 month after the end of the fiscal year: 1.000 • 2 months after the end of the fiscal year : 1.000 • 3 months after the end of the fiscal year : 1.000 • 4 months after the end of the fiscal year : 0.750 • 5 months after the end of the fiscal year : 0.750 • 6 months after the end of the fiscal year : 0.750 • 7 months after the end of the fiscal year : 0.500 • 8 months after the end of the fiscal year : 0.500 • 9 months after the end of the fiscal year : 0.500 • 10 months after the end of the fiscal year : 0.250 • 11 months after the end of the fiscal year : 0.250 • 12 months after the end of the fiscal year : 0.250 • More than 12 months after the end of the fiscal year : 0.000 • They are not publicly available : 0.000 150 GOV/PGC(2009)2 B.2.2. Medium-term budget perspective Variables The medium-term budget perspective index is constructed from two sub-indexes of varying weights. Within each sub-index, questions are equally weighted. 1. Estimates (weight 66.66%) A. Level of detail of multi-year expenditures estimates (Question 16) B. Period covered by the multi-year expenditure estimates (Question 17) C. Frequency with which the multi-year estimates are updated (Question 18) D. Basis for extrapolating multi-year expenditure estimates (Question 19) 2. Ceilings/Targets (weight 33.33%) A. Existence of multi-year expenditure targets or ceilings (Question 20) B. Period covered by the targets/ceilings (Question 21) C. Frequency with which the targets/ceilings are revised (Question 22) Scoring Estimates (weight 66.66 %) A. Does the annual budget documentation submitted to the Legislature contain multi-year expenditure estimates? (Question 16) • No: 0.000 151 GOV/PGC(2009)2 • Yes, at the aggregate level: 0.333 • Yes, at the ministry level: 0.666 • Yes, at line item level: 1.000 Countries reporting “other” were scored as follows: • Australia: 1.000 • Belgium: 1.000 • Finland: 0.333 • Iceland: 0.666 • Norway: 0.666 • Portugal: 0.333 • South Korea: 0.333 • Spain: 0.333 B. How many years does the multi-year expenditure estimates cover (including the upcoming budget)? (Question 17) • 2 years: 0.000 • 3 years: 0.200 • 4 years: 0.400 • 5 years: 0.600 • 6 years: 0.800 • 7 years: 1.000 • 8 years: 0.200 • 9 years: 0.200 • 10 years: 0.200 Note: no countries responded that their multi-year expenditure estimates exceeded 6 years; thus 0.800 was the highest score allocated for this question. Countries reporting “other” were scored as follows: • Mexico: 0.000 C. How often are the multi-year estimates updated? (Question 18) • Annually: 1.000 152 GOV/PGC(2009)2 • Every 2 years: 0.800 • Every 3 years: 0.600 • Every 4 years: 0.400 • Every 10 years: 0.200 Countries reporting “other” were scored as follows: • Australia: 1.000 • Canada: 1.000 • USA: 1.000 D. If applicable, from what basis do you extrapolate multi-year expenditure estimates? (Question 19) This is a multiple choice question and respondents were to select all items that applied. The final score is the sum of the scores for the items selected. • Expenditures in future years are based on current legislation: 0.167 • Expenditures in future years are based on anticipated legislative changes: 0.167 • Expenditures in future years are unchanged in nominal terms: 0.167 • Expenditures in future years are unchanged in real terms: 0.167 • Expenditures in future years are adjusted for the official macroeconomic forecast: 0.167 • Expenditures in future years are adjusted for demographic changes: 0.167 Scoring Ceilings/Targets (weight 33.33%) A. Are there multi-year expenditure targets or ceilings? (Question 20) • No, there are no such targets or ceilings: 0.000 • Yes, there are aggregate targets or ceilings : 0.333 • Yes, there are targets or ceilings for each ministry: 0.667 • Yes, there are targets or ceilings for line items: 1.000 Countries reporting “other” were scored as follows: • Belgium:1.000 • Canada: 0.500 • Denmark: 0.500 • Hungary: 1.000 153 GOV/PGC(2009)2 • Turkey: 1.000 B. How many years do the targets/ceilings cover (including the upcoming budget)? (Question 21) • 2 years: 1.000 • 3 years: 1.000 • 4 years: 1.000 • 5 years: 1.000 • 6 years: 0.000 • 7 years: 0.000 • 8 years: 0.000 • 9 years: 0.000 • 10 years: 0.000 C. How often are the targets/ceilings revised? (Question 22) • Every year: 0.200 • Every 2 years: 0.400 • Every 3 years: 0.600 • Every 4 years: 0.800 • Every 10 years: 1.000 • After each election: 0.800 • Every time a new Government is formed: 0.800 154 GOV/PGC(2009)2 B.2.3. Performance Budgeting Index Variables The performance budgeting index is comprised of the following 14 variables. Eight of the variables were considered to be relatively more important and were assigned double weights. As a result, six questions received weights of 4.56% (or 1/22) and eight questions received weights of 9.09% (or 2/22). A. Types of information produced to assess the Government's non-financial performance (Question 71) B. Types of evaluations commissioned and/or conducted (Question 72) C. Types of non-financial performance measures developed for central government (Question 73) D. Responsibility for setting performance targets Question 77) E. Responsibility for achieving performance targets (Question 78) F. Responsibility for achieving performance targets if they cut across organisational boundaries (Question 79) G. Bodies to which ministries/Government organisations are required to report (Question 80) H. Presentation of information on performance against targets (Question 81) I. Availability of information on performance against targets (Question 82) J. Use of performance information as part of the budget discussions/negotiations (Question 83) K. How performance is used by the central budget authority and line ministries (Question 84) L. Consequences if performance targets are not met (Question 85) M. Consequences if evaluations indicate poor performance (Question 86) N. Frequency with which performance information is used in budgetary decision-making (Question 87) 155 GOV/PGC(2009)2 Scoring A. What types of performance information are produced to assess the Government's nonfinancial performance? (Question 71). This is a multiple choice question and respondents were to select all items that applied. The final score is the sum of the scores for the items selected. • None: 0.000 • Performance targets: 0.200 • Performance measures: 0.200 • Evaluation reports (e.g. programme, sectoral, efficiency, or cost effectiveness reviews): 0.200 • Benchmarking: 0.200 • Other: 0.200 B. What types of evaluations are commissioned and/or conducted by the following institutions? (Question 72). These are multiple choice questions and respondents were to select all items that applied. The more items selected, the better the final score. The final score for question 72 is a simple average of the scores for each sub-question (72.a, 72.b, 72.c, 72.d & 72.e) 72.a Review of ongoing programmes • Central Budget Authority: 0.166 • Ministry of Finance: 0.166 • Line Ministries: 0.166 • Supreme Audit Institution: 0.166 • Legislature: 0.166 • Other: 0.166 72.b Ex post review of programmes • Central Budget Authority: 0.166 • Ministry of Finance: 0.166 • Line Ministries: 0.166 • Supreme Audit Institution: 0.166 • Legislature: 0.166 • Other: 0.166 72.c Review of new initiatives or programmes • Central Budget Authority: 0.166 156 GOV/PGC(2009)2 • Ministry of Finance: 0.166 • Line Ministries: 0.166 • Supreme Audit Institution: 0.166 • Legislature: 0.166 • Other: 0.166 72.d Sectoral reviews • Central Budget Authority: 0.166 • Ministry of Finance: 0.166 • Line Ministries: 0.166 • Supreme Audit Institution: 0.166 • Legislature: 0.166 • Other: 0.166 72.e Efficiency and/or cost effectiveness reviews • Central Budget Authority: 0.166 • Ministry of Finance: 0.166 • Line Ministries: 0.166 • Supreme Audit Institution: 0.166 • Legislature: 0.166 • Other: 0.166 C. What types of non-financial performance measures have been developed for Central Government? (Question 73). The final score for question 73 is an average of the scores for subquestion 73.a and 73.b. 73.a Output measures • Yes: 1.000 • No : 0.000 73.b Outcome measures • Yes: 1.000 • No : 0.000 Countries with missing responses were given scores of 0.000: 157 GOV/PGC(2009)2 • 73.a: Turkey • 73.b: Greece, Turkey D. Who has responsibility for setting performance targets? (Question 77) These are multiple choice questions and respondents were to select all items that applied. 77.a No one • Formally: 0.000 • In practice: 0.000 77.b The Prime Minister • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.c The President • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.d The Cabinet • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.e The relevant minister • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.f The Minister of Finance • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.g The administrative head of the relevant ministry 158 GOV/PGC(2009)2 • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.h The Minister of Finance together with the relevant minister • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.i The Legislature • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 77.j International organization • Formally: 0.500 • In practice: 0.500 • Not applicable: 0.000 Countries that answered positively to sub-question 77.a (i.e. that no one is responsible for setting performance targets) received a final score of 0.000. For all other countries, the final score for Question 77 is a weighted average of the scores for sub questions 77.b – 77.j. Weights were assigned as follows: Sub-question Weight 77.i & 77.j Legislature or International organization 7/28 (note: no country selected both of these options, i.e. a country either said that the Legislature or an International organization or neither had responsibility for setting performance targets) 77.b &77.c Prime Minister or President 6/28 (note: no country selected both of these options, i.e. a country either said that the Prime Minister or the President or neither had the responsibility for setting performance targets) 77.d Cabinet 5/28 77.f The Minister of Finance 4/28 77.h The Minister of Finance together with the relevant minister 3/28 77.e The relevant minister 2/28 77.g The administrative head of the relevant ministry 1/28 Countries reporting “other” received final scores as follows based on their comments: • Belgium : 0.250 • Canada: 0.500 • Denmark: 0.250 159 GOV/PGC(2009)2 • Finland: 0.500 • Sweden: 0.500 E. Who is responsible for achieving performance targets? (Question 78) This is a simple alternative choice question and respondents could only select one item from the choices presented. • No one: 0.000 • The President: 1.000 • The Prime Minister: 1.000 • The Central Budget Authority: 0.750 • The relevant minister: 0.500 • The highest civil servant in the relevant ministry: 0.250 Countries with missing responses were given scores of 0.000 based on their response to Question 75.a (i.e. these countries have not linked the numerical value of expenditures to performance targets or goals/objectives): • Spain: 0.000 • Switzerland: 0.000 F. If performance targets cut across organisational boundaries who is responsible for achieving them? (Question 79) • No one: 0.000 • One of the relevant ministers: 0.500 • All of the relevant ministers: 1.000 • The highest ranking civil servant in one of the relevant organizations: 0.500 • The highest ranking civil servant in each of the relevant organizations: 1.000 Countries reporting “other” received final scores as follows based on their responses: • Belgium: 1.000 (response: “The highest ranking civil servant in each of the relevant organizations”) • Canada: 0.500 (response: “Other”) • Finland: 1.000 (response: “The highest ranking civil servant in each of the relevant organizations ”) • Korea: 1.000 (response: “The highest ranking civil servant in each of the relevant organizations”) 160 GOV/PGC(2009)2 Countries with missing responses were given scores of 0.000 based on their response to Question 75.a (i.e. these countries have not linked the numerical value of expenditures to performance targets or goals/objectives): • Denmark: 0.000 • Ireland: 0.000 • Spain: 0.000 • Switzerland: 0.000 G. If ministries/Government organisations are required to report on performance against targets, to which bodies are they required to report? (Question 80). This is a multiple choice question and respondents were to select all items that applied. The final score represents the highest score of the items selected. • Not applicable, there are no targets: 0.000 • Not required to report: 0.000 • Internally within the relevant ministry/Government organization: 0.167 • Parent Ministry: 0.333 • Central Budget Authority: 0.500 • President’s Office: 0.833 • Prime Minister’s Office: 0.833 • Cabinet: 0.667 • Legislature: 1.000 • Supreme Audit Institution: 1.000 Countries reporting “other” received final scores as follows based on their responses: • Japan: 1.000 (response: “Ministry of International Affairs and Communications”) • Mexico: 1.000 (response: “Ministry of Civil Service”) • New Zealand: 0.666 (response: “Vote Minister”) • Portugal: 0.666 (response: “The Coordinator Minister”) • United-Kingdom: 1.000 (response: “Publicly”) • Countries with missing responses were given scores of 0.000: • Spain: 0.000 161 GOV/PGC(2009)2 H. Is performance against targets routinely presented to the Legislature? (Question 81) This is a simple alternative choice question and respondents could only select one item from the choices presented. • No it is not presented at all: 0.000 • No it is only presented on an ad hoc basis : 0.166 • Yes it is presented in a Government-wide report not accompanying the budget : 0.333 • Yes it is presented in a Government-wide report accompanying the budget : 0.600 • Yes it is integrated into the annual financial documents : 0.833 • Yes it is integrated into the main budget documents : 0.833 • Yes each ministry prepares performance reports accompanying the budget : 0.500 Countries reporting “other” received final scores as follows: • Finland: 0.500 • Ireland: 0.500 • Portugal: 0.166 • United States: 0.833 I. Is performance against targets made available to the public? (Question 82) This is a multiple choice question and respondents were to select all items that applied. The final score is a sum of all items selected. No, it is not made available to the public: 0.000 Yes, a Government-wide report on performance is published: 0.166 Yes, individual ministries publish reports on their performance: 0.166 Yes, as part of other Government-wide documents: 0.166 Yes, as part of other ministry-specific documents: 0.166 Yes, there is an internet site for this information: 0.166 Other: 0.166 J. Is performance information used as part of the budget discussions/negotiations between the Central Budget Authority and line/spending ministries? (Question 83) The final score for question 83 is a simple average of the four scores for sub-questions 83.a and 83.b. 83.a Evaluation reports • Yes: 1.000 • No: 0.000 162 GOV/PGC(2009)2 If yes could you indicate the proportion of ministries with whom evaluation reports are used as a basis for negotiations? • None: 0.000 • Up to 20%: 0.200 • 21-40%: 0.400 • 41-60%: 0.600 • 61-80%: 0.800 • 81-100%: 1.000 83.b Performance against targets • Yes: 1.000 • No: 0.000 If yes could you indicate the proportion of ministries with whom performance against targets are used as a basis for negotiations? • None: 0.000 • Up to 20%: 0.200 • 21-40%: 0.400 • 41-60%: 0.600 • 61-80%: 0.800 • 81-100%: 1.000 Countries with missing responses were given the following scores for both parts of question 83.b: • Greece: 0.000 K. How is performance information generally used by the following two actors? (Question 84) These are multiple choice questions and respondents were to select all items that applied. Countries that answered positively to sub-question 84.a (i.e. that there is no impact on decisionmaking) received a final score of 0.000. For all other countries, the final score for Question 84 is a simple average of the scores for sub questions 84.b – 84.i. 84.a No impact on decision-making • Central Budget Authority: 0.000 • Line Ministries: 0.000 84.b Allocating resources between ministries/agencies 163 GOV/PGC(2009)2 • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.c Allocating resources between programmes • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.d Allocating resources within programmes • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.e Justifying existing allocations to specific activities/programmes • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.f Cutting expenditures • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.g Compelling or pushing change in programmes • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.h Setting targets for next year • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 84.i Managing programmes/agencies 164 GOV/PGC(2009)2 • Central Budget Authority: 0.500 • Line Ministries: 0.500 • Not applicable: 0.000 Countries with missing responses were given final scores of 0.000: • Czech Republic: 0.000 • Greece: 0.000 • Switzerland: 0.000 L. If performance targets are not met, how likely is it that any of the following consequences are triggered? (Question 85) The final score for question 85 is a simple average of the scores for subquestions 85.a – 85.e. Q.85.a The programme is eliminated • Almost never 0-20%: 1.000 • Rarely 21-40%: 0.800 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.400 • Almost always 81-100%: 0.200 Q.85.b There is more intense monitoring of the programmes/activities in the future • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Q.85.c There are negative consequences for the size of the budget of the ministry responsible for delivering the target (i.e. the budget decreases) • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 165 GOV/PGC(2009)2 Q.85.d There are negative consequences for the pay of the head of the ministry/entity responsible for delivering the target • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Q.85.e There are negative consequences for future career opportunities of the heads of ministries/entities responsible for delivering the target • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Countries with missing responses were given final scores of 0.000: • Czech Republic: 0.000 • Greece: 0.000 • Switzerland: 0.000 M. If evaluations indicate poor performance how likely is it that any of the following consequences are triggered? (Question 86) The final score for question 86 is a simple average of the scores for sub-questions 86.a – 86.e. Q.86.a The programme is eliminated • Almost never 0-20%: 1.000 • Rarely 21-40%: 0.800 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.400 • Almost always 81-100%: 0.200 Q.86.b There is more intense monitoring of the programmes/activities in the future • Almost never 0-20%: 0.200 166 GOV/PGC(2009)2 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Q.86.c There are negative consequences for the size of the budget of the ministry responsible for delivering the target (i.e. the budget decreases) • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Q.86.d There are negative consequences for the pay of the head of the ministry/entity responsible for delivering the target • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Q.86.e There are negative consequences for future career opportunities of the heads of ministries/entities responsible for delivering the target • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Countries with missing responses were given final scores of 0.000: • Czech Republic: 0.000 • Greece: 0.000 • Switzerland: 0.000 167 GOV/PGC(2009)2 N. How often do the following actors use performance information in budgetary decisionmaking? (Question 87) The final score for question 87 is a simple average of the scores for subquestions 87.a – 87.e. 87.a The President • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.b The Prime Minister • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.c The Cabinet • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.d The Minister of Finance • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.e The Central Budget Authority • Almost never 0-20%: 0.200 168 GOV/PGC(2009)2 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.f The minister with responsibility for the ministry/entity which is supposed to deliver a performance target • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.g Members of the Budget Committee in the Legislature • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 87.h Members of sectoral committees in the Legislature • Almost never 0-20%: 0.200 • Rarely 21-40%: 0.400 • Sometimes 41-60%: 0.600 • Often 61-80%: 0.800 • Almost always 81-100%: 1.000 Countries with missing responses were given final scores of 0.000: • Czech Republic: 0.000 • Greece: 0.000 • Switzerland: 0.000 169 GOV/PGC(2009)2 B.2.4. Line Ministry / Agency Flexibility Index Variables The line ministry/agency flexibility index is comprised of the following equally weighted variables: A. Number of line item appropriations contained in the budget (Question 38) B. Receipt of lump sum appropriations (Question 49) C. Authority of Government to cut/cancel/rescind spending once the budget has been approved by the Legislature (Question 52) D. Ability of ministers to reallocate/vire funds between line items within their responsibility (Question 53) E. Ability of ministers to carry-over unused funds or appropriations from one year to another (Question 54) F. Ability of ministries/Government organisations to borrow against future appropriations (Question 55) G. Ability of Government organisation managers to keep any savings from efficiency gains that they have realized in order to finance other expenditures (Question 65) Scoring A. How many line item appropriations are contained in the budget? (Question 38) • 0-200 :1.000 • 201-500 : 0.800 • 501-1000 : 0.600 170 GOV/PGC(2009)2 • 1001-1500 : 0.400 • 1501-2000: 0.200 • >2000: 0.100 B. Do your agencies/executive organisations receive lump sum appropriations? (Question 49) • No, each agency/executive organisation receives an appropriation that specifies expenditures below the agency level: 0.000 • Yes, each agency/executive organisation receives a lump sum appropriation for operating expenditures only, without sub-limits: 0.500 • Yes, each agency/executive organisation receives a lump sum appropriation for operating expenditures only, with a sub-limit on wages: 0.250 • Yes, each agency/executive organisation receives a lump sum appropriation covering both operating and capital expenditures, without sub-limits: 1.000 • Yes, each agency/executive organisation receives a lump sum appropriation covering both operating and capital expenditures, with a sub-limit on wages: 0.750 Countries reporting “other” were scored as follows: • France: 0.500 • Luxembourg: 0.500 • Netherlands: 0.500 • Poland: 0.500 • Portugal: 0.500 • Spain: 0.500 • United-States: 0.500 C. Does the Government have the authority to cut/cancel/rescind spending once the budget has been approved by the Legislature? (Question 52) The final score for this question is a simple average of the scores for Q.52.a and Q.52.b. Q.52.a Is it possible? • No: 0.000 • Yes without restrictions: 1.000 • Yes, with restrictions : 0.500 Q.52.b Does it require any approval? Countries could select more than one of the items below as a response. If multiple responses were selected, the lowest score was used. • Approval is not required : 1.000 171 GOV/PGC(2009)2 • It requires Central Budget Authority approval prior to the fact: 0.600 • It requires legislative approval prior to the fact: 0.200 • It requires Central Budget Authority approval after the fact: 0.800 • It requires legislative approval after the fact: 0.400 Countries reporting “other” on question 52.b were scored as follows: • Australia: 0.500 • Canada: 0.700 • Luxembourg: 0.700 • Mexico: 0.400 • Spain: 0.600 • Sweden: 0.500 • USA: 0.300 D. Are ministers allowed to reallocate/vire funds between line items within their responsibility? (Question 53) The survey asked countries selecting the response “Yes, with restriction” to clarify their answers by selecting whether the carry-over of funds required approval of the Legislature or the Finance Minister. Countries responding that it required the approval of the Finance Minister received a score of 0.666. Countries responding that it required the approval of the Legislature received a score of 0.3337 . If a country replied “Yes, with restrictions” but did not provide further clarification, the question received a score of 0.500 (the average of the scores for the responses “With the approval of the Legislature” and “With the approval of the Finance Minister.”) • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 E. Can ministers carry-over unused funds or appropriations from one year to another? (Question 54). Each sub-question (54.a, 54.b & 54.c) received a weight of 1. The final score for question 54 is a simple average of the three sub-questions. The survey asked countries selecting the response “Yes, with restriction” to clarify their answers by selecting whether the carry-over of funds required approval of the Legislature and/or the Finance 7 According to the theoretical framework, the ex-ante approval of the Legislature reduces the margin of agency maneuvers and thus diminishes the flexibility 172 GOV/PGC(2009)2 Minister. Countries responding that it required the approval of the Finance Minister received a score of 0.666. Countries responding that it required the approval of the Legislature received a score of 0.3338. If a country replied “Yes, with restrictions” but without providing further clarification, the question received a score of 0.500 (the average of the scores for the responses “With the approval of the Legislature” and “With the approval of the Finance Minister.”) Q.54.a Operating expenditures • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 Q.54.b Investments • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 Q.54.c Transfers/subsidies • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 F. Is it possible for ministries/Government organisations to borrow against future appropriations? (Question 55) Each sub-question (55.a, 55.b & 55.c) received a weight of 1. The final score for question 55 is a simple average of the three sub-questions. Q. 55.a Operating expenditures • No: 0.000 • Yes, without approval: 1.000 8 According to the theoretical framework, the ex-ante approval of the Legislature reduces the margin of agency maneuvers and thus diminishes the flexibility 173 GOV/PGC(2009)2 • Yes, without approval if within a specified sub-limit: 0.667 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Q. 55.b Investment • No: 0.000 • Yes, without approval: 1.000 • Yes, without approval if within a specified sub-limit: 0.667 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Q.55.c Transfers, subsidies, etc • No: 0.000 • Yes, without approval: 1.000 • Yes, without approval if within a specified sub-limit: 0.667 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Countries reporting “other” to sub-question 55.c were scored as follows: • Belgium: 0.500 • Sweden: 0.500 G. In general, are Government organisation managers able to keep any savings from efficiency gains that they have realized in order to finance other expenditures? (Question 65) • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with some restrictions: 0.500 Countries reporting “other” were scored as follows: • Netherlands: 0.500 B.2.5. General Executive Flexibility Index Variables The general executive flexibility index is comprised of the following equally weighted variables: 174 GOV/PGC(2009)2 A. Number of line item appropriations contained in the budget (Question 38) B. Receipt of lump sum appropriations (Question 49) C. Authority of Government to increase spending once the budget has been approved by the Legislature (Question 51) D. Authority of Government to cut/cancel/rescind spending once the budget has been approved by the Legislature (Question 52) E. Ability of ministers to reallocate/vire funds between line items within their responsibility (Question 53) F. Ability of ministers to carry-over unused funds or appropriations from one year to another (Question 54) G. Ability of ministries/Government organisations to borrow against future appropriations (Question 55) H. Ability to overspend before a supplementary appropriation law/budget is approved by the Legislature (Question 57) I. Ability of Government organisation managers to keep any savings from efficiency gains that they have realized in order to finance other expenditures (Question 65) 175 GOV/PGC(2009)2 Scoring A. How many line item appropriations are contained in the budget? (Question 38) • 0-200 :1.000 • 201-500 : 0.800 • 501-1000 : 0.600 • 1001-1500 : 0.400 • 1501-2000: 0.200 • >2000: 0.100 B. Do your agencies/executive organisations receive lump sum appropriations? (Question 49) • No, each agency/executive organisation receives an appropriation that specifies expenditures below the agency level: 0.000 • Yes, each agency/executive organisation receives a lump sum appropriation for operating expenditures only, without sub-limits: 0.500 • Yes, each agency/executive organisation receives a lump sum appropriation for operating expenditures only, with a sub-limit on wages: 0.250 • Yes, each agency/executive organisation receives a lump sum appropriation covering both operating and capital expenditures, without sub-limits: 1.000 • Yes, each agency/executive organisation receives a lump sum appropriation covering both operating and capital expenditures, with a sub-limit on wages: 0.750 Countries reporting “other” were scored as follows: • France: 0.500 • Luxembourg: 0.500 • Netherlands: 0.500 • Poland: 0.500 • Portugal: 0.500 • Spain: 0.500 • United-States: 0.500 C. Does the Government have the authority to increase spending once the budget has been approved by the Legislature? For each type of change please indicate whether it applies, with what restrictions and whether any approval is required. (Question 51) The final score for this question is a simple average of the scores for Q.51.a and Q.51.b. 176 GOV/PGC(2009)2 Q.51.a. Increase mandatory spending. Final score is an average of the scores for sub-question 1 and sub-question 2. 1. Is it possible? • No: 0.000 • Yes without restrictions: 1.000 • Yes, with restrictions: 0.500 2. Does it require any approval? Countries could select more than one of the items below as a response. If multiple responses were selected, the lowest score was used. • Approval is not required : 1.000 • It requires Central Budget Authority approval prior to the fact: 0.600 • It requires legislative approval prior to the fact: 0.200 • It requires Central Budget Authority approval after the fact: 0.800 • It requires legislative approval after the fact: 0.400 Countries reporting “other” to sub-question 2 for Q.51.a were scored as follows: • Finland : 0.700 • Germany : 0.300 • Iceland : 0.200 • Spain : 0.500 Q.51.b. Increase discretionary spending Final score is an average of the scores for sub-question 1 and sub-question 2. 1. Is it possible? • No: 0.000 • Yes without restrictions: 1.000 • Yes, with restrictions: 0.500 2. Does it require any approval? Countries could select more than one of the items below as a response. If multiple responses were selected, the lowest score was used. • Approval is not required : 1.000 • It requires Central Budget Authority approval prior to the fact: 0.600 • It requires legislative approval prior to the fact: 0.200 • It requires Central Budget Authority approval after the fact: 0.800 177 GOV/PGC(2009)2 • It requires legislative approval after the fact: 0.400 Countries reporting “other” to sub-question 2 for Q.51.b were scored as follows: • Germany : 0.200 • Iceland : 0.400 • Portugal : 0.800 • Spain : 0.800 • USA: 0.200 D. Does the Government have the authority to cut/cancel/rescind spending once the budget has been approved by the Legislature? (Question 52) The final score for this question is a simple average of the scores for Q.52.a and Q.52.b. Q.52.a Is it possible? • No: 0.000 • Yes without restrictions: 1.000 • Yes, with restrictions : 0.500 Q.52.b Does it require any approval? Countries could select more than one of the items below as a response. If multiple responses were selected, the lowest score was used. • Approval is not required : 1.000 • It requires Central Budget Authority approval prior to the fact: 0.600 • It requires legislative approval prior to the fact: 0.200 • It requires Central Budget Authority approval after the fact: 0.800 • It requires legislative approval after the fact: 0.400 Countries reporting “other” on question 52.b were scored as follows: • Australia: 0.500 • Canada: 0.700 • Luxembourg: 0.700 • Mexico: 0.400 • Spain: 0.600 • Sweden: 0.500 • USA: 0.300 178 GOV/PGC(2009)2 E. Are ministers allowed to reallocate/vire funds between line items within their responsibility? (Question 53) The survey asked countries selecting the response “Yes, with restriction” to clarify their answers by selecting whether the carry-over of funds required approval of the Legislature or the Finance Minister. Countries responding that it required the approval of the Finance Minister received a score of 0.666. Countries responding that it required the approval of the Legislature received a score of 0.3339 . If a country replied “Yes, with restrictions” without providing further clarification, the question received a score of 0.500 (the average of the scores for the responses “With the approval of the Legislature” and “With the approval of the Finance Minister.”) • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 F. Can ministers carry-over unused funds or appropriations from one year to another? (Question 54). Each sub-question (54.a, 54.b & 54.c) received a weight of 1. The final score for question 54 is a simple average of the three sub-questions. The survey asked countries selecting the response “Yes, with restriction” to clarify their answers by selecting whether the carry-over of funds required approval of the Legislature and/or the Finance Minister. Countries responding that it required the approval of the Finance Minister received a score of 0.666. Countries responding that it required the approval of the Legislature received a score of 0.33310. If a country replied “Yes, with restrictions” but without providing further clarification, the question received a score of 0.500 (the average of the scores for the responses “With the approval of the Legislature” and “With the approval of the Finance Minister.”) Q.54.a Operating expenditures • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 Q.54.b Investments • No: 0.000 9 According to the theoretical framework, the ex-ante approval of the Legislature reduces the margin of agency maneuvers and thus diminishes the flexibility 10 According to the theoretical framework, the ex-ante approval of the Legislature reduces the margin of agency maneuvers and thus diminishes the flexibility 179 GOV/PGC(2009)2 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 Q.54.c Transfers/subsidies • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with restrictions: 0.500 • With the approval of the Legislature: 0.333 • With the approval of the Finance Minister: 0.666 G. Is it possible for ministries/Government organisations to borrow against future appropriations? (Question 55) Each sub-question (55.a, 55.b & 55.c) received a weight of 1. The final score for question 55 is a simple average of the three sub-questions. Q. 55.a Operating expenditures • No: 0.000 • Yes, without approval: 1.000 • Yes, without approval if within a specified sub-limit: 0.667 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Q. 55.b Investment • No: 0.000 • Yes, without approval: 1.000 • Yes, without approval if within a specified sub-limit: 0.667 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Q.55.c Transfers, subsidies, etc • No: 0.000 • Yes, without approval: 1.000 • Yes, without approval if within a specified sub-limit: 0.667 180 GOV/PGC(2009)2 • Yes, with approval of the Legislature: 0.333 • Yes, with approval of the Central Budget Authority: 0.500 Countries reporting “other” to sub-question 55.c were scored as follows: • Belgium: 0.500 • Sweden: 0.500 H. Can overspending occur before a supplementary appropriation law/budget is approved by the Legislature? (Question 57) • No: 0.000 • Yes, there are no limits on overspending without legislative approval: 1.000 • Yes, but only up to a certain limit: 0.500 • Yes but only for mandatory spending: 0.500 Countries reporting “other” were scored as follows: • Finland : 0.500 • Mexico : 0.500 • Netherland : 0.500 • Norway : 0.500 • Portugal : 0.500 • Sweden : 0.500 I. In general, are Government organisation managers able to keep any savings from efficiency gains that they have realized in order to finance other expenditures? (Question 65) • No: 0.000 • Yes, without restrictions: 1.000 • Yes, with some restrictions: 0.500 Countries reporting “other” were scored as follows: • Netherlands: 0.500 181 GOV/PGC(2009)2 B.3. Regulatory Management Composites: Theoretical framework, construction and weighting To be provided as a supplement on OLIS on April 20. 182 GOV/PGC(2009)2 ANNEX C CONFLICT OF INTEREST DISCLOSURE REQUIREMENTS FOR DECISION MAKERS This annex provides unique data for each country on how core values are communicated to central government employees; the types of private interests that they require central government decision makers to disclose as well as the level of transparency provided to the disclosures; procedures in place for public servants to report misconduct or suspected corruption; and the types of protection offered to whistleblowers. Data are from the OECD 2000 and 2009 Survey on Integrity. Respondents to the survey were OECD member country delegates in charge of integrity in central government. 183 GOV/PGC(2009)2 C.1 Communicating core values to central government public servants The detailed data presented below underlie the summary data presented in Figure II.4.2 in Chapter II.4 on core public service values, and indicate the different methods used by OECD countries to communicate core public service values to central government employees. Figure C.1.1 How countries communicate core values to central government public servants (2009) Countries communicating core values Values automatically provided Communicated by new technology Part of employment contract Distributed after revision Provided in new position Other measures used Australia X X X X X X X X X X Belgium Czech X Republic X X Finland X X X X Germany X Hungary X X X X Italy X X X X X Japan X X X X Ireland X Austria X X X X X X Canada X X X X X X Denmark X X X X France X X X Greece N/A N/A N/A N/A N/A N/A N/A X Iceland X X X X Luxembour g Netherland X X s X X Norway Portugal N/A N/A N/A N/A N/A N/A Spain X X X Turkey X X X X Switzerland Korea N/A N/A N/A N/A N/A N/A N/A Mexico X X X X X New Zealand X X X X X X X Poland X X X X X Slovak Republic N/A N/A N/A N/A N/A N/A N/A N/A Sweden X X X X X United N/A N/A N/A N/A N/A N/A Kingdom N/A United States X X X X X X: Method used N/A: Data not available Source: OECD 2009 Survey on Integrity. 184 GOV/PGC(2009)2 C.2 Conflict of interest disclosure requirements Figure C.2.1 provides unique data for each country on the types of private interests that they require central government decision makers to disclose as well as the level of transparency provided to the disclosures. The data presented in Figure C.2.1 underlie the aggregate, summary data presented in Figures XI.1.1 and XI.1.2 in Chapter XI.1 on conflict of interest disclosures. How to read the table: D: Disclosure is required, but it is not made publicly available DP: Disclosure is required, and is publicly available ND: Disclosure is not required NA: Not applicable. For example, decision-makers may be prohibited by law from accepting gifts or having outside employments. ID: Data are incomplete Data are not available for Denmark, Greece, Korea, the Slovak republic, Spain and the United Kingdom. Table C.1 Types of information decision makers are required to disclose, and level of transparency 185 GOV/PGC(2009)2 AUS AUT BEL CAN CZE FIN FRA DEU HUN ISL IRL ITA JPN LUX MEX NLD NZL NOR POL PRT SWE CHE TUR USA Assets and liabilities ND ND ND DP ND DP ND D D DP ND D DP Loans ND ND ND DP ND DP ND D D DP ND D DP President Income DP ND ND ND ND DP ND D ND DP ND D DP Outside positions NA ND ND ND NA DP ND D ND NA ND D DP Gifts NA ND ND ND ND DP ND D ND DP ND ND DP Previous Employment ND ND ND ND ND ND ND D ND DP ND ND DP Assets and liabilities DP D D DP DP DP D ND DP ND DP DP DP ND D DP DP DP DP DP D Loans DP D ND DP DP D D ND DP ND DP DP DP ND ND DP D DP DP ND D Executive Income DP DP DP D DP ND ND ND DP ND DP DP DP ND ND ND DP ND DP DP DPrime Minister Outside positions DP NA DP DP DP DP ND DP DP ND DP DP DP ND NA DP DP ND NA DP D Gifts DP NA ND DP DP ND ND DP DP ND DP D DP ND D DP DP DP DP DP ND Previous Employment DP ND ND D ND ND ND ND ND ND NA DP ND ND DP ND DP ND DP DP ND Assets and liabilities DP D D DP DP DP D ND DP ND DP DP DP ND D D DP DP DP DP DP ND D DP Loans DP D ND DP DP D D ND DP ND DP DP DP ND D ND DP D DP DP ND ND D DP Ministers Income DP DP DP D DP ND ND ND DP ND DP DP DP ND D ND ND DP ND DP DP ND D DP Outside positions DP NA DP DP DP DP ND DP DP ND DP DP DP ND D NA DP DP ND NA DP ND D DP Gifts DP NA ND DP DP ND ND P DP ND DP D DP ND D D DP DP DP DP DP ND ND DP Previous Employment DP ND ND D ND ND ND ND ND ND NA DP ND ND D DP DP DP ND DP DP ND ND DP Assets and liabilities DP ND D DP DP DP D DP DP ND DP DP DP ND D ND DP DP DP DP NA ND D DP Loans DP ND ND DP DP DP D ND DP ND DP DP DP ND D ND DP D DP DP ND ND D DP Income DP DP DP DP DP ND ND DP DP ND DP DP DP ND D D ND DP DP DP ND ND D DPLower House Outside positions DP DP DP DP DP DP ND DP DP ND DP DP DP ND D DP DP DP DP DP DP ND D DP Gifts DP ND ND DP DP ND ND DP DP ND DP ND DP ND D DP DP DP DP DP ND ND ND DP Legislative Previous Employment ID ND ND ND ND ND ND DP ND ND NA DP ND ND D DP DP DP ND DP DP ND ND DP Assets and liabilities DP ND D DP DP D ID ND DP DP DP ND D ND DP DP ND DP Upper House Loans DP ND ND DP DP D ID ND DP DP DP ND ID ND D DP ND DP Income DP DP DP DP DP ND ID ND DP DP DP ND D D DP DP ND DP Outside positions DP DP DP DP DP ND ID ND DP DP DP ND D DP DP DP ND DP Gifts DP ND ND DP DP ND ID ND DP ND DP ND D DP DP DP ND DP Previous Employment ID ND ND ND ND ND ID ND NA DP ND ND ID DP DP ND ND DP GOV/PGC(2009)2 C.3 Protection of Whistle-blowers The detailed data presented below underlie the aggregate summary data presented in Figures XI.2.1 and XI.2.2 in Chapter XI.2 on whistle-blowing, procedures in place for public servants to report misconduct or suspected corruption and the types of protection offered to whistleblowers. Figure C.3.1 Procedures for public servants to report misconduct or suspected corruption (2000 and 2009) Countries with reporting procedures Defined by legal provision Defined by internal rules Other 2000 2009 2000 2009 2000 2009 2000 2009 Australia X X X X X X X Austria X X X X Belgium X X X Canada X X X Czech Republic X X X Denmark X X X X X Finland X X X X X France X X X X Germany X X X X Greece N/A N/A N/A N/A Hungary X Iceland X X X Ireland X X Italy X X X X Japan X X X X Korea X N/A X N/A X N/A N/A Luxembourg X X Mexico X X X X X X Netherlands X X X X X X New Zealand X X X X X Norway X X X Poland X X X X Portugal X X X X Slovak Republic N/A N/A N/A N/A N/A N/A N/A N/A Spain X N/A X N/A N/A N/A Sweden X X X X Switzerland X X X X X X Turkey X X X X United Kingdom X N/A N/A X N/A N/A United States X X X X X X X X: Method used N/A: Data not available Source: OECD 2000 and 2009 Survey on Integrity. Note: Five countries did not complete the survey in 2009: Greece, Korea, Slovak Republic, Spain and the United Kingdom. 187 GOV/PGC(2009)2 ANNEX D: MEMBERS OF THE STEERING GROUP COUNTRY NAME TITLE/POSITION MINISTRY AUSTRALIA Ms Carmel McGREGOR Deputy Public Service Comissioner Australian Public Service Commission AUSTRIA Ms Elisabeth DEARING Director Federal Chancellery Administrative Development, Div III/7 CANADA Mr Bruce STACEY Executive Director for Results Based Management Treasury Board Secretariat FINLAND Ms Katju HOLKERI Head of Government Policy Unit Ministry of Finance FRANCE Mr Bernard BLANC Chef de la mission des relations internationals Ministére de l’Économie, des Finances et de l’Índustrie Direction générale de la modernisation de l'État (DGME) GERMANY Mr Andreas WEGEND Ms Christiane BOEDDING Head of Division for International Cooperation in Administrative Matters Federal Ministry of the Interior ITALY Dr Pia MARCONI Director General Department of Public Administration NETHERLANDS Mr Koos ROEST Advisor on Strategic Policy Ministry of the Interior & Kingdom Relations Directorate-General for Management of the Public Sector NORWAY Mr Bratveit KLENG Advisor Ministry of Government Administration and Reform SWEDEN Mr Claes ELMGREN Analyst Statskontoret 188 GOV/PGC(2009)2 BIBLIOGRAPHY Alesina, A., R. Hausmann, R. Hommes and E. Stein (1999), "Budget Institutions and Fiscal Performance in Latin America," Journal of Development Economics, 59(2), pp. 253-273. Anderson, B. and J.J. Minarik (2006), “Design Choice for Fiscal Policy Rules,” OECD Journal on Budgeting, Vol. 5, No. 4, pp. 159-208. Anderson, B. and J. Sheppard (forthcoming), “Fiscal Futures, Institutional Budget Reforms and Their Effects: What Can Be Learned?” OECD Journal on Budgeting. Arnold, J. (2008), “Do Tax Structures Affect Aggregate Economic Growth?: Empirical Evidence from a Panel of OECD Countries,” OECD Economics Department Working Papers, No. 643, OECD, Paris. Currestine, T. (2005), “Performance Information in the Budget Process: Results of the OECD 2005 Questionnaire,” OECD Journal on Budgeting, Vol. 5, No. 2, pp. 87-132. European Commission DGISM (2007), The User Challenge: Benchmarking the Supply of Online Public Services, 7th Measurement, September 2007. Hallerberg, M., R. Strauch and J. Von Hagen (2007), "The Design of Fiscal Rules and Forms of Governance in European Union Countries," European Journal of Political Economy, 23(2), pp. 338 359. Hatry, H.P. 1999. Performance Measurement: Getting Results. Washington, D.C.: Urban Institute Press. Lopez-Claros, Augusto, Schwab, Klaus and Porter, Michael E. (ed.) (2006), Global competitiveness report 2006-2007: creating an improved business environment, Palgrave, Houndmills; Basingstoke; New York, N.Y. Manning, Nick, D.J. Kraan and J. Malinska (2006), “How and Why Should Government Activity Be Measured in Government at a Glance?” OECD, Paris. Myles, G. (2008), “Economic Growth and the Role of Taxation,” report for OECD, available at www.people.ex.ac.uk/gdmyles/papers/pdfs/OECDfin.pdf Nardo, M., M. Saisana, A. Saltelli, S. Tarantola, A. Hoffman and E. Giovanni, (2005), OECD Handbook on Constructing Composite Indicators: Methodology and User Guide, OECD Statistics Working Paper 2005/3, OECD, Paris. Nardo, M., M. Saisana, A. Saltelli, S. Tarantola, A. Hoffman and E. Giovanni, (2008), Handbook on Constructing Composite Indicators, joint publication of the OECD and European Commission OECD (2000), Trust in Government: Ethics Measures in OECD Countries, OECD, Paris. 189 GOV/PGC(2009)2 OECD (2001), Citizens as Partners: information, consultation and public participation in policy making, OECD, Paris. OECD (2002), “OECD Best Practices for Budget Transparency,” OECD Journal on Budgeting, Vol. 1, No. 3, pp. 7-14. OECD (2003), Managing Conflict of Interest in the Public Service: OECD Guidelines and Country Experiences, OECD, Paris. OECD (2005), Modernizing Government: The Way Forward, OECD, Paris. OECD (2007), Ageing and the Public Service: Human Resource Challenges, OECD, Paris. OECD (2007), “COFOG-Special: Working paper on test procedure for five European countries,” OECD, Paris. OECD (2007), Integrity in Public Procurement: Good Practice from A to Z, OECD, Paris. OECD (2007), OECD Revenue Statistics 1995-2006, OECD, Paris. OECD (2008), “Performance Budgeting: A User’s Guide,” OECD Policy Brief, March 2008. OECD (2008), The State of the Public Service, OECD, Paris. OECD (2009), Focus on Citizens: Public Engagement for Better Policy and Services, OECD, Paris. OECD (2009), OECD Principles for Integrity in Public Procurement, OECD, Paris. OECD (forthcoming), Budgeting in OECD Countries, OECD, Paris. OECD (forthcoming), Measuring Government Activity, OECD, Paris. Pilichowski, E. and E. Turkisch (2008), “Employment in Government in the Perspective of the Production Costs of Goods and Services in the Public Domain,” OECD Working Papers on Public Governance, 8, OECD, Paris. Pollitt, Christopher et al. (2004), Agencies—How Governments Do Things Through Semi-Autonomous Organisations, Palgrave Macmillan, Bainstoke, UK. Pollitt, C. and G. Bouckaert. 2004. Public Management Reform: A Comparative Analysis. Oxford, United Kingdom: Oxford University Press. Schick (2005), “Sustainable Budget Policy: Concepts and Approaches,” OECD Journal on Budgeting, Vol. 5, No. 1, pp. 107-126. United Nations (2008), United Nations e-government survey 2008: from e-government to connected governance, United Nations, New York, N.Y. United Nations (2003), UN Global e-Government Survey 2003, United Nations, New York. 190 GOV/PGC(2009)2 Von Hagen, J. (1992). Budgeting Procedures and Fiscal Performance in the European Communities. Brussels, Directorate-General for Economic and Financial Affairs, Commission of the European Communities. W.K. Kellogg Foundation. 2004. Logic Model Development Guide. Michigan: W. K. Kellogg Foundation. Wal, Z van Der and Huberts L.W.J.C. (2008), “Value Solidity in Government and Business: Results for an Empirical Study on Public and Private Sector Organisational Values,” American Review of Public Administration, 38, pp. 264-285. Wehner, J. (2009), “Cabinet Structure and Fiscal Policy Outcomes,” unpublished paper, Government Department, London School of Economics, London. http://personal.lse.ac.uk/wehner/cabinet.pdf. 191