55636 Better Regulation for Growth Making It Work: ‘RIA Light’ for Developing Countries Investment Climate Advisory Services of the World Bank Group With funding from FIAS, the multi-donor investment climate advisory service in partnership with BETTER REGULATION FOR GROWTH GOVERNANCE FRAMEWORKS AND TOOLS FOR EFFECTIVE REGULATORY REFORM MAKING IT WORK: ‘RIA LIGHT’ FOR DEVELOPING COUNTRIES INVESTMENT CLIMATE ADVISORY SERVICES WORLD BANK GROUP ©2010 The World Bank Group 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission to repro- duce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. About the Investment Climate Advisory Services of the World Bank Group The Investment Climate Advisory Services (IC) of the World Bank Group helps governments implement reforms to improve their business environment, and encourage and retain investment, thus fostering competitive markets, growth and job creation. Funding is provided by the World Bank Group (IFC, MIGA, and the World Bank) and over 15 donor partners working through the multi- donor FIAS platform. The findings, interpretations and conclusions included in this note are those of the author and do not necessarily reflect the view of the Executive Directors of the World Bank Group or the governments they represent. Better Regulation for Growth Program The Better Regulation for Growth (BRG) Program was launched in 2007 by the Dutch Ministry of Foreign Affairs, the UK Department for International Development (DFID) and IC, the investment climate advisory services of the World Bank Group. The objective of the BRG Program is to review and synthesize experiences with regulatory governance initiatives in developing countries, and to develop and disseminate practical tools and guidance that will help developing countries design and implement effective regulatory reform programs. Reports and other documentation developed under the BRG Program are available at: www .ifc.org/brg Table of Contents Acknowledgments .............................................................................. vi Acronyms .......................................................................................... vii Executive Summary ...........................................................................viii Introduction ........................................................................................ 1 What is RIA and What are its Benefits? ................................................ 3 What is RIA? .....................................................................................................3 Extent of Use of RIA ............................................................................................4 Who conducts RIA? .............................................................................................5 Potential benefits of RIA ......................................................................................6 Regulatory scrutiny vs. Budget scrutiny ...................................................................7 Does RIA work in practice? ..................................................................................8 Major Elements of a Best Practice RIA System ..................................... 11 Formally established RIA policy, with high-level political endorsement .....................11 Integration of RIA into the policy process ..............................................................12 Consistent application of RIA .............................................................................12 Appropriate methodological requirements ...........................................................12 Targeting of RIA efforts ......................................................................................13 Consultation mechanisms ..................................................................................14 Data collection strategies ...................................................................................15 Central oversight of RIA ....................................................................................15 Development of adequate capacities ..................................................................16 Oversight of legal quality ..................................................................................16 Applying RIA thinking to existing, and proposed laws ...........................................16 Adapting RIA best practice to developing countries: Challenges in implementation .................................................................................................18 Making it Work: Identifying Basic Requirements for a Functional RIA System ........................................................................ 19 RIA in developing countries: Evidence from current experiments ..............................19 Promoting a new approach to impact analysis: RIA Light........................................21 Political commitment to establish and operate an effective, self-sustaining RIA process ......................................................................................................22 A unit or group of regulatory reformers and quality control by an expert supervisory body .....................................................................................23 Clear criteria: What to subject to RIA and what questions to address .......................24 Transparency in the process and consultation .......................................................27 iii Building RIA capacities within government............................................................29 RIA Light for developing countries: Summing up the new approach .........................32 Implementing and Institutionalizing RIA ............................................. 33 Initial implementation issues................................................................................33 Options for institutionalizing RIA .........................................................................34 Legal requirement to undertake RIA .....................................................................34 Options for RIA oversight ..................................................................................36 RIA oversight within the administration .................................................................36 Oversight by independent or external bodies........................................................36 Integrating RIA with policy decision-making process ..............................................37 Challenges and obstacles in implementing RIA Light in emerging and developing countries ...................................................................................38 Lack of compliance............................................................................................38 Institutional resistance ........................................................................................39 Lack of credibility of RIA ....................................................................................40 Lack of influence of RIA outcomes on regulatory decisions ......................................40 Conclusions ........................................................................................ 42 Bibliography ...................................................................................... 44 Annex 1: Defining Basic Terms in Regulatory Reform .......................... 45 Annex 2: Better Regulation for Growth Program ................................ 47 iv List of Boxes and Figures Boxes Box 1: Features of High-Quality Regulation ............................................................4 Box 2: The Regulatory Reform Toolkit: Focusing on Different Elements of the Reform Process ....................................................................................9 Box 3: Capacity Building in a RIA System ............................................................16 Box 4: Elements to Consider When Looking for Political Support to Integrate RIA Light in the Policy Process ...................................................................22 Box 5: Important Elements, Responsibilities of Regulatory Reform Units ....................23 Box 6: Clear and Consistently Applied Criteria and Rules Employed to Screen Regulatory Proposals...............................................................................28 Box 7: Elements of Transparency and Consultation in a RIA Light System .................29 Box 8: Capacity-Building and Monitoring and Evaluation Programs ........................31 Box 9: Legal Underpinning of RIA in Emerging and Developing Countries................34 Box 10: RIA Adopted Through Government Policy Statements.................................35 Figures Figure 1: The RIA Process - Seven Essential Phases ..................................................4 Figure 2: Components of a RIA Document ..............................................................4 Figure 3: Regulatory Impact Analysis: Requirement for RIA, 1998, 2005 ...................5 Figure 4: Elements of RIA Best Practice and Challenges for Developing Countries .....17 Figure 5: RIA Light Elements ...............................................................................32 v ACKNOWLEDGMENTS The report has been prepared by Stephen Rim- Buitenkamp, Fabrizio de Francesco, Greg Bounds, mer, Delia Rodrigo and Peter Ladegaard of the Jeroen Nijland, Josephine Kanyi, Ksenija Vidulic, Investment Climate Department of the World Lars Grava, Margo Thomas, Mustafizur Rahman, Bank Group. Nick Godfrey, Richard Sandall, Stephen Rimmer, Scott Jacobs and Ulrich Ernst. An early draft of this document was discussed at the Technical Meeting of the Advisory Panel of the Early and advanced drafts of the report benefited Better Regulation for Growth Program held in The from additional comments from Andreja Maru- Hague on November 13–14, 2008. Comments to sic, Collin Kirkpatrick, Edward Donelan, Margo this draft were provided by participants to this Thomas, Rex Deighton-Smith, Richard Sandall, meeting: Andre Nijsen, Andreja Marusic, Andrea Sanjar Ibragimov, Scott Jacobs and Vsevolod Renda, Claudio Radaelli, Colin Kirkpatrick, Delia Payevskiy. Zai Fanai supported the publication of Rodrigo, Edward Donelan, Esra Yilmaz, Eva the report. vi ACRONYMS BEST Business Environment Strengthening for Tanzania BRG Better Regulation for Growth Program COFEMER Federal Regulatory Improvement Commission (Mexico) DFID British Department for International Development EU European Union IC Investment Climate Advisory Services OECD Organization for Economic Cooperation and Development RBP Regulatory Best Practice Program (Uganda) RIA Regulatory Impact Analysis UMI Uganda Management Institute vii EXECUTIVE SUMMARY Regulatory Impact Analysis (RIA) is recognized by from external donors, consultants, and other advi- the Organization for Economic Cooperation and sors. The latter group does not always seem to Development (OECD) and most developed coun- fully appreciate the long-term complexities of tries as a key tool to improve the efficiency, trans- establishing functioning RIA systems. This has parency and accountability of regulatory decision led to a broader question of whether and how RIA making. RIAs assess the likely impacts of new reg- can or should be transferred and adapted to devel- ulations in both quantitative and qualitative terms, oping country contexts. helping decision makers to make good choices. By providing a systematic, evidence based and con- Extensive literature has been produced over the sultative framework for regulatory policymaking, last two decades about RIA in developed coun- well-functioning RIA systems typically encourage tries. However there is very little evidence about “good governance” features and contribute to a its relevance to developing countries and how it better business-enabling environment. might be best used in such countries. The exist- ing literature on RIA provides good insights into An increasing number of developing and transi- the introduction and implementation of RIA sys- tion countries are developing and implementing tems, including challenges and lessons learned. new RIA systems, by adapting and integrating Still, this literature is often based on the implicit RIA within their existing policy-making processes assumption of the end-goal being a “gold-plated and institutions. Several of these RIA initiatives best practice model.” When viewed from the per- look promising. However, some have faced set- spective of implementing RIA in developing backs and a slower- than-expected implementa- countries, this approach is problematic for several tion. Evidence suggests this has been due to a reasons. First and foremost, it may lead to an combination of factors, including overly ambi- overly rigid and inefficient RIA system that can- tious implementation targets, and a lack of under- not be appropriately sustained in the low-capacity lying capacities and governance and/or support context of many developing countries. Second, viii the so-called “best practice” end-target may not political commitment and a cadre of reformers be achievable or relevant to strive for in part and experts focused on implementing the RIA because there may be other higher priority Light system. Clear criteria for what should be reforms. Indeed, there are a range of different subject to RIA should then be established – based regulatory review and reform tools available to on local priorities and conditions. It is also governments and RIA is just one of these. important that there are transparency mecha- nisms, including consultation processes which This paper aims at identifying a set of minimum make sure that stakeholders’ views are taken into requirements for a well-functioning RIA Light- account when developing and implementing system that is tailored to the requirements of regulations and capacity building processes, such developing countries. Embracing the overall as training for officials. objectives and relevance of RIA, the paper explores the fundamental set of building blocks The paper provides specific guidance and options and activities required to establish and maintain for each of the five key requirements. Taken a RIA Light system, taking into account what it together, these institutional and conceptual is considered as good practice. building blocks provide the basis for establishing the framework for a “RIA Light,” which is con- The paper argues that the following five basic cri- ducive to the objectives, needs and capacities of teria have to be in place for a functioning RIA developing countries. system, which is referred to as “RIA Light”: Indeed, this paper argues that, on balance, there 1) Political commitment to establish and operate are significant benefits of applying empirically an effective and self sustaining RIA process. based scrutiny of the impacts of new regulatory proposals in both developed and developing 2) A unit or group of regulatory reformers – countries. However, the approach taken to design- preferably based in a central area of govern- ing and implementing the RIA system must be ment – which oversees, comments and reports cognizant of the specific circumstances of the on the quality of regulatory proposals before country in question, of different reform priorities decisions are made about regulation. and of the resource constraints that apply. 3) Clear and consistently applied criteria and rules employed to screen regulatory proposals. RIA Light also implies that a number of compo- nents considered as key and integral parts of a 4) A transparent regulatory policy development well-functioning RIA system might not be process, which includes consultation with included given institutional, technical and mate- stakeholders. rial constraints in developing countries. For example, the paper argues that a functional RIA 5) A capacity building program, involving Light system would not require a full integra- preparation of guidelines; training of officials tion of RIA in the policy development processes preparing RIA and facilitating the required (clearly a desirable objective, but not fully cultural changes, and establishing monitor- achieved even in developed countries), the use ing, evaluation and reporting systems. of cost/benefit analysis, the heavy quantification of impacts, or a requirement to systematically The paper also suggests that these five criteria consider alternatives. Having a RIA Light sys- ideally should be established in sequential order. tem can still add value to regulatory decision In other words, the set-up and sustainability of making through enhanced transparency, better any RIA Light system is contingent on strong quality information for decision makers and ix stakeholders, and improved consultation regard- regulatory reform, and in setting up RIA systems ing regulatory issues, among other things. in particular, in which IC has been involved. This paper is based not only on field research The BRG is a joint initiative of the Dutch Minis- conducted by the IC of the World Bank Group try of Foreign Affairs, the British Department for in the framework of the Better Regulation for International Development (DFID) and IC, the Growth (BRG) Program, but also on experiences multi-donor investment climate advisory service accumulated over years of practical projects on of the World Bank Group. See Annex 2 for a detailed description of the BRG Program. x INTRODUCTION The economic, social and environmental costs to which examines the implications of potential society of poor quality regulations are substantial. regulatory policy options. A key feature of RIA is Governments need to work systematically to its consideration of the potential economic ensure that the regulations they develop and impacts of regulatory proposals. implement are of high quality, achieving intended objectives in a transparent manner while also The discussion provided in this paper is based on minimizing costs. In particular, poor quality reg- lessons learned from existing literature, practical ulations lead to increased compliance costs for implementation and challenges encountered by business and other groups. These costs are passed governments as well as development organiza- through to consumers in the form of higher prices tions supporting these governments. for goods and services, such as food and energy. Poor quality regulations also substantially reduce The main objective of this paper is to explore and levels of economic activity and impede employ- identify “minimum requirements” for a func- ment and wealth creation. tional RIA system. The paper is based on the dual premise that RIA systems provide net benefits RIA is a fundamental tool to help governments and quality to regulatory decision making, but assess in a systematic manner the impact of regu- that a more simplified and streamlined model of lations and the different options they might have RIA than that usually seen in developed coun- to achieve policy objectives. RIA is used to exam- tries may be more appropriate for developing ine and measure the likely benefits, costs and countries. This approach is called “RIA Light.” effects of proposed new regulations. The imple- mentation of RIA supports existing processes of Part One of this paper identifies the key features policymaking by contributing valuable evidence- of RIA and discusses the objectives and benefits based data to policy decisions and through the of this tool. It compares the RIA process, which construction of a rational decision framework scrutinizes regulatory policy proposals, with 1 similar processes used in all countries to scruti- Part Three discusses the key building blocks and nize spending proposals, as part of the budget elements that can be pragmatically adapted to process. It then focuses on how RIA works in establish and operate a successful and sustainable practice, including other regulatory review and RIA process, called RIA Light, which matches reform tools which can also be employed by the capacities and constraints often experienced governments. in developing countries. This discussion includes options for institutionalizing RIA Light, oversee- Part Two presents the key features, elements and ing the RIA Light process and integrating RIA dimensions of best practice RIA systems. This with existing policy decision-making processes. discussion is based on extensive research and analysis undertaken by a range of governments Part Four concentrates on practical challenges and international organizations, particularly the and impediments to establishing, operating, and OECD. institutionalizing a successful RIA Light process. 2 WHAT IS RIA AND WHAT ARE ITS BENEFITS? What is RIA? Indeed, decisions about regulation are often based on limited information and in some cases RIA is a tool that helps policymakers ask system- guesses regarding who is affected and how. atic questions about the different policy options Therefore, a systematic approach is needed to and consequences of government interventions. identifying and weighing regulatory effects. The output is an assessment report that provides Only in this way can policymakers be confi- high quality evidence comparing different policy dent that the benefits of a policy action are options. The final objective of RIA is to improve likely to be greater than the costs. Only if total the quality of regulation.1 benefits exceed costs will society as a whole be better off as a result of regulation. Regulation only maximizes community welfare when it is effective, efficient and transparent (see While there is no one single definition of RIA, Box 1). However, the impact of regulations, there are several common elements and features both positive and negative, is not always appar- evident in all countries with functioning ent. For example, the behavior of firms and RIA systems. The first is that each country with individuals will often change in response to reg- RIA has used this process to strengthen existing ulations, but these changes are often subtle and decision-making processes, not to replace difficult to predict or measure. The problem of them. Second, existing RIA systems have two understanding regulatory impacts is especially components: acute when the longer-term is considered, because the impact of regulations can often 1) RIA process, which is the process of systemati- change substantially as the economic and social cally identifying policy options and assessing environment in which it operates changes. the expected effects of regulatory proposals, 1 See Annex 1 for a basic definition of regulatory reform using a consistent analytical method, linked terms, in particular regulatory quality. to policy decision making (Fig. 1); 3 2. RIA document, which is the final product of evidence based information about the different the process and is presented to policymakers policy options for government intervention. summarizing potential alternatives, their impacts and implementation aspects of pro- posed measures (Fig. 2). Extent of Use of RIA The RIA process, materialized in the RIA docu- Over the last few decades there has been a signifi- ment, feeds the decision making process with cant increase in the use of RIA in developed coun- Box 1: Features of High-Quality Regulation Effectiveness and efficiency. Effective regulation is that which achieves its objective(s). Regulation is effi- cient if it achieves the objectives of regulation at the lowest possible cost and at a cost that is also smaller than the benefit of achieving the objective. Transparency and accountability. Transparency in regulation-making means that interested parties have the opportunity to provide their views to government via an open consultation process while the regulation is being developed. Transparency in regulatory implementation also means that people who must comply with regulation have access to the regulations and can readily understand their requirements. Accountability means that regulatory policy development, administration and enforcement are subject to public scrutiny, so that regula- tors are accountable for their actions. Proportionality. The principle of proportionality means regulation should be proportionate and targeted at addressing the specific risks and problems that have been identified as requiring government action. Consistency. Consistency requires that regulators take account of related areas of regulation and ensure similar treatment. It also focuses on the avoidance of regulatory duplication and overlap. Figure 1: The RIA Process - Seven Essential Phases Regulatory Impact Assessment Presenting a Enforcing and Defining the Identifying Thinking Conducting Consulting conclusion reviewing the policy the policy about policy an impact with and a way proposed problem objective options analysis stakeholders forward option Figure 2: Components of a RIA Document RIA document contains: Conclusion and Enforcement and Policy problem Policy objective Policy options Impact analysis Consultation way forward review 4 Figure 3: Regulatory Impact Analysis: Requirement for RIA, 1998, 2005 Requirement to demonstrate that benefits 12 of regulation justify the costs 15 Requirement to identify the benefits 17 of new regulation 25 Requirement to identify the costs 20 of new regulation 25 Requirement for draft subordinate 18 regulations 25 18 Requirement for draft primary laws 25 12 Formal requirement by law 20 Number of countries 1998 2005 Source: Adapted from OECD Regulatory Indicators Questionnaire, GOV/PGC/REG(2005)ANN1. tries. For example, in the early 1980s only a hand- As the table shows, RIAs are implemented differ- ful of OECD countries employed wide-ranging ently in each country. Countries focus on particu- RIA processes. This increased to 50 percent by lar components of RIA. But the table also suggests 2000 and at the present time over 90 percent of that a RIA that had none of these features would OECD countries claim to have RIA systems. A barely qualify itself as a RIA. further 20 developing and transition countries, such as Poland and the Czech Republic, have recently established RIA processes. In addition, a large number of developing and emerging coun- Who conducts RIA? tries are implementing or actively exploring the RIA is prepared by regulatory departments, agen- potential to establish RIA type systems, includ- cies or ministries – sometimes called regulators – ing Macedonia, Turkey, Bangladesh, Kenya and which sponsor new or amended regulation. The Sri Lanka. regulators responsible for areas of regulation are generally best placed to understand regulatory Figure 3 illustrates the features of RIA processes problems, issues and possible solutions in their used in OECD countries in 1998 and 2005.2 area of responsibility. Regulators also typically RIA programs have expanded over time in scope have links with affected stakeholders, a relatively in the last few years. In 2005, 25 countries use it good understanding of the impact of regulations for primary laws and subordinate regulations. on them, and are well placed to lead consultation Thirteen countries had RIA units providing cen- processes on regulatory issues. tral oversight of the RIA process, which were based outside regulatory agencies sponsoring reg- Such regulatory agencies also can have an ulatory proposals. entrenched culture that is risk averse or conserva- tive and, therefore, is not open to new ideas or 2 Jacobzone, S., C. Choi and C. Miguet (2007), “Indicators of Regulatory Management Systems”, OECD Working Papers approaches to regulation. Regulators can be “cap- on Public Governance, 2007/4, OECD Publishing. tured” by the businesses they regulate and seek to 5 benefit these businesses, even if this is at the Only if governments choose a policy response – expense of consumers and broader society. Indeed, or mix of responses – that is based on high qual- regulators may also benefit from particular regu- ity and reliable information can governments be latory outcomes, for example, because their bud- confident that they are making the right choices. gets or staff can gain from particular regulatory Indeed, if government decisions are based on solutions or approaches. poor quality information then such decisions are essentially based on guesses and, therefore, are For these reasons, the RIA process is usually over- likely to result in regulatory failure. seen by an independent unit or group based at the center of government. This unit should not be In addition to limited information, governments involved in regulating business and should not also have limited capacities to make, administer have vested interest in particular regulatory out- and enforce regulation. This is particular acute in comes. Therefore, it can provide regulators and developing countries where enforcing and com- ultimately decision-makers with high quality, pliance with regulation is a challenge that requires trusted and impartial advice about regulatory issues major changes in the regulatory system. This, in and the quality of analysis contained in RIAs. itself, means that governments must take care before committing to new regulatory require- ments. Indeed, there are limits to the amount of Potential benefits of RIA regulation that any government can impose; however, regulatory inflation remains a trend in As noted above, the main objective of RIA is to many countries. improve the quality of regulation. A common feature of poor quality regulation is that the Some critics of the use of RIA have questioned underlying problem and objective of an existing whether it is appropriate to all legal and constitu- (or proposed) regulation are not clearly identi- tional contexts or whether its development simply fied. The RIA process facilitates the identification reflects the specific circumstances of the common- of both the underlying policy problem and also law countries in which it has historically been an objective that is focused specifically on address- largely developed. However, it is important to ing that problem. Only if the policy problem and bear in mind that the expected result of applying objective are properly identified can a menu of the RIA model is the provision of better informa- relevant and feasible solutions be identified. tion to political decision-makers. This outcome is desirable regardless of political, legal and institu- The RIA process is also inherently an evidence- tional contexts. based approach to scrutinizing and comparing several policy options. It involves not only com- Another potential benefit of making use of RIA is paring the impacts, positives and negatives (e.g. that better quality regulation is less restrictive of benefits and costs) likely to be associated with a business, effectively protects consumers, better particular policy action, but also compares these protects citizens’ rights, especially those of vul- expected outcomes with those that would result nerable groups, and reduces opportunities for from other possible policy actions that could be corruption. RIA offers an opportunity to identify taken in pursuit of the same objective. RIA the possible costs and benefits of regulation on helps identify information gaps, thus highlight- affected groups, relating to consultation mecha- ing where information collection programs nisms that increase transparency in the regulatory should be focused. For example, the RIA pro- decision making process. cess typically encourages the use of consultation with stakeholders in considering how identified A further benefit of RIA is that it does not neces- problems might be fixed. sarily involve making significant institutional and 6 process changes. Rather, in many countries that budgetary processes and that applied to regula- use RIA, the process supports, reinforces and tory actions. Budget decisions are generally sub- strengthens existing legal, political, economic ject to detailed scrutiny by Parliament and within and institutional processes. the government administration, according to clearly established criteria and processes. The Furthermore, RIA might not require significant existing budget scrutiny processes have developed additional resources or fiscal outlays. The RIA progressively over the last century and are essen- process itself is typically managed and overseen tial to the operation of government. There is also by a small unit of several regulatory experts (or a a well-established “industry” of academia, con- small advisory group) which is usually based in a stantly developing and improving tools of national central area of government.3 accounting and budgeting. At the political and administrative level, budget policy has its own Finally, RIA is intended to result not only in more ministry and usually a very strong position among and better information being made available, but peer ministries. also that this information is presented in a sys- tematic, logical manner. Indeed, according to the This is far from the case for regulatory systems and OECD: regulatory scrutiny, despite the importance of reg- ulatory policy to the economic potentials of any “…RIA’s most important contribution to the given country. Regulatory decisions are often quality of decisions is not the precision of the made on an ad hoc basis with little information calculations used, but the action of analyz- about likely impacts, little scrutiny or government- ing–questioning, understanding real-world wide co-ordination. Thus, the sophistication and impacts and exploring assumptions.” institutionalization of regulatory processes can be said to be at a much earlier stage of development than the equivalent tools used in budget and Regulatory scrutiny vs. Budget national accounting processes. This is to a large scrutiny extent due to the relatively recent expansion in governments’ use of regulation as the preferred Regulation imposes significant costs on govern- tool of intervention.4 ments, businesses and other groups who must comply with it. In this respect, the act of making Nonetheless a similar approach, based on the long regulation is conceptually very similar to that of term development of effective and appropriate raising taxes and adopting government spending scrutiny processes, is needed to support regulatory programs. That is, in both cases private resources decision making. This becomes especially impor- are being diverted to achieving public goals tant given that governments are typically using through government action. regulation more often and more extensively than in the past.5 This conceptual similarity strongly suggests that similar scrutiny and accountability processes should be applied to both types of government 4 Cf. Majone and the “Rise of the Regulatory State.” 5 The greater use of regulation is a result of several factors. For action. However, in most countries, there is a very example, community expectations tend to rise over time and substantial gap between the level of sophistication also the growing complexity of technology and markets also generates additional demands for new regulation. Further- and scrutiny applied through the government’s more, existing disciplines on fiscal spending and the absence of effective quality control over the creation of new regula- tions can create a bias to regulate first and ask questions 3 See Part Three of this paper for further information on the about the impacts of those regulations at a later stage. The core elements and building blocks required to have a func- rapid and exponential growth in regulation in many coun- tioning RIA system. tries is sometimes called “regulatory inflation.” 7 In this regard the adoption of a RIA system should In the second case, RIA is not particularly a con- be seen as an important step in a long-term pro- sequence of other regulatory reforms, but linked cess of developing better controls over the use of to existing regulatory processes or ongoing the regulatory tool and thereby improving the reforms, such as in the case of Brazil. In this quality of regulatory decision making and regula- country, RIA is foreseen as a mechanism that can tory outcomes. In other words, RIA can help empower regulatory agencies to make better deci- ensure that regulatory quality control processes sions and increase transparency in the regulatory are brought progressively into line with those process, including stakeholders’ views. adopted in other areas of government decision making, such as the budget and related processes. The following paragraphs describe why RIA is conceptually an attractive regulatory tool to be used, even in developing country contexts, despite the challenges it could impose. Does RIA work in practice? The preceding sections have outlined the con- RIA systems have been promoted for many years ceptual rationale for RIA and its potential bene- as key contributors to economic growth and good fits. It is also noted that scrutiny of the use of governance. RIA has been seen by many observ- regulation is significantly less developed than ers as an obvious and logical component of sound scrutiny of the budgetary process. However, the regulatory management systems. In fact, despite experience of RIA in both developed and devel- the intuitive logic and intellectual attraction of oping countries has led many to question the RIA, there is relatively limited evidence of the practical merits of this approach to improving success of RIA systems, as implemented to date, the quality of regulation. in improving regulatory quality. IC work on regulatory governance includes sup- Despite many years of implementation, trials and porting developing countries in their efforts to errors, the evidence of their positive contribution establish RIA systems. There has been a trend in to better regulation and lower costs and risks for recent years of establishing RIA systems, some- businesses is relatively scarce. This conclusion is times as a complement to ongoing regulatory true even for many developed countries that have reforms, but also as a single component of exist- extensive experience and relatively well-developed ing mechanisms. RIA systems. Critics point out that proposed regulation is halted or amended following RIA In the first case, RIA is sought as a mechanism to scrutiny in only a minority of cases. Even where maintain the gains achieved through reforms that regulation has been improved following RIA have improved the stock of regulations, such as scrutiny, the causal role of RIA in achieving the licensing reform. In these cases, RIA will be used change is often debatable. to ensure that new regulation meets certain qual- ity criteria to keep the future regulatory stock up RIA proponents tend to argue that the presence to date. This is the case of Kenya, Rwanda and of the RIA discipline means that ill-justified pro- Bangladesh, where RIA systems have not yet been posals are less likely to be brought forward in the implemented, but some form of impact assess- first place, suggesting that many RIA benefits ment seems appropriate to screen new regulatory may be hidden. However, this dynamic is inher- proposals and ensure the wins obtained do not get ently almost impossible to measure. Indeed, a key lost. It has been the case of Croatia, for instance, issue in this debate is that information about the where RIA was introduced after a guillotine pro- impacts of RIA are very difficult to monitor and cess that eliminated business regulations. measure, because RIA is only one part of the 8 complex machinery of government involved in In many ways, these different regulatory quality reviewing, making and applying regulations. tools are complementary. For example, RIA as a Separating the impacts of RIA from other possi- methodology is equally applicable to the task of ble impacts is in practice very difficult to do. revising and streamlining existing regulation as it is to ensuring the quality of proposals for new Therefore, in considering the usefulness and regulation. However, where limited resources are applicability of RIA in developing country con- available to support regulatory quality programs, texts, donors and governments interested in the merits of these different programs will often implementing or strengthening RIA systems need to be weighed comparatively. should consider the potential benefits of RIA against such criticisms. Should RIA systems be Some have argued that RIA imposes significant dismissed as a utopian, over-rational approach fiscal and resource costs of governments. However, to improving decision making? Furthermore, as implied above, there is little evidence to support even if RIA does generate net gains for a coun- this claim. The main cost of a RIA process is estab- try using RIA, it is also important to weigh the lishing and operating a central unit to oversee and merits of RIA (as one means of improving regu- report on the RIA process. Regulators may incur latory quality) against those of other available costs in preparing RIA, but these costs are typi- approaches and tools. Indeed, a range of regula- cally not significant. For example, in Australia in tory tools are available that can be either com- 2004-5, the average labour cost to departments plementary to RIA or in some cases alternatives and agencies preparing RIA for decision makers to RIA. Some of these options are provided in averaged around US$3500 per RIA. Box 2 and are discussed in depth in the sister paper “Tools and Approaches to Review Exist- This paper argues that, on balance, there are sig- ing Regulation.” nificant benefits of applying empirically based Box 2: The Regulatory Reform Toolkit: Focusing on Different Elements of the Reform Process Focus on strengthening Focus on the flow of new Focus on the stock of Focus on reform Institutions regulation existing regulation process as a whole ■ Central Unit ■ Regulatory Impact ■ Scrap and build ■ Mandatory public ■ Regulatory reform Assessment (RIA) ■ Processre-engineering consultation committee ■ Forward planning ■ Staged repeal ■ Transparent ■ Business advisory ■ Business cost calcula- ■ The guillotine decision making panel tor ■ Regulatory perfor- ■ Reviews of ■ E0regustry ■ Administrative mance indicators regulatory system ■ One stop shop procedure law ■ Administrative ■ Improved data ■ Improved ■ Silent is consent simplification collection communication strategies ■ Better enforcement ■ Better monitoring ■ Evaluation and reporting 9 scrutiny of the impacts of new regulatory propos- must be cognizant of the specific circumstances als in both developed and developing countries. of the country in question, different reform pri- Despite the lack of strong evidence, the value and orities, and the resource constraints that apply. cost-effectiveness of RIA appear extremely plau- In developing country contexts, these issues have sible because impact assessments introduce: particular relevance, as a sustained RIA system with a lower degree of institutional capacities ■ awareness of unintended effects of regula- and limited resources is more challenging to tions that otherwise could result in regula- introduce. tory failures; Little evidence, however, is available to confirm ■ transparency and efficiency in the way regu- that RIA makes a difference in the policymaking lations are prepared and communicated; and process of developing countries. In many coun- tries, such as Tanzania, Uganda, Serbia, Croatia, ■ accountability in the policy process. Vietnam, Uzbekistan, etc., this tool has been integrated only recently in the regulatory process For RIA to work, however, the approach taken and results have not been properly monitored to designing and implementing the RIA system and evaluated. 10 MAJOR ELEMENTS OF A BEST PRACTICE RIA SYSTEM A number of OECD publications6 have sought some of the main linkages between RIA system to identify the elements of a best practice RIA elements and features. system, based on extensive analysis of the expe- riences of member countries in implementing RIA since the 1980s. These publications and reports argue strongly that a best practice RIA Formally established RIA policy, system contains a number of interconnected with high-level political endorsement elements. This means that the most successful A formally announced policy, which is endorsed RIA systems, in terms of their credibility and by the head of government or senior minister, is capacity to improve regulatory quality, are those essential – but not sufficient – to ensure that in which the various elements are designed and regulators actually comply with the RIA require- implemented to be mutually supportive of each ment. A formal policy also helps ensure that other. In addition, such best practice RIA sys- consistent approaches to RIA are taken, so that tems support and strengthen existing regula- the same disciplines are applied to different pol- tion making processes. icy areas. Endorsement at the highest political level ensures that the policy has adequate author- The following discussion identifies each of the ity within government to support its effective main elements of a ‘best practice’ RIA system, implementation. highlighting the importance of each as well as The RIA policy must be contained in an authori- 6 See, in particular, Regulatory Impact Analysis: Best Practices in tative document. In some cases, RIA requirements OECD Countries, 1997, and Regulatory Policies in OECD Countries: From Interventionism to Regulatory Governance, are established in law, providing them with the 2002. See also Building an Institutional Framework for Regu- highest possible level of formal authority. More latory Impact Analysis. Guidance for Policy-Makers, 2008 and Regulatory Impact Analysis: A Tool for Policy Coherence, 2008, commonly, the policy is contained in a Presiden- forthcoming. tial Decree, or Prime Ministerial Instruction, or 11 in the Cabinet Handbook setting out the proce- Another important contributor to integrating dural arrangements for government decision mak- RIA with the policy process is to ensure that prac- ing. The specific mechanism used to establish the tical benefits for regulatory quality achieved due political authority of the policy must be appropri- to the use of RIA are documented and dissemi- ate to the broader political and institutional cul- nated, enhancing understanding and underpin- ture of the country. ning support for its use. For example, partial measures of the impact on RIA on regulatory Strong support from senior government minis- outcomes include the proportion of RIA where ters is needed, in particular in the early stages of the recommended options change during the RIA implementation, to establish the authority policy development process, or where decision- of the policy and encourage officials to incorpo- makers reject a regulatory proposal that is accom- rate it effectively into existing policy development panied by RIA which fails to meet minimum processes. requirements established by the government. Integration of RIA into the policy Consistent application of RIA process If possible, RIA should be applied to major regula- RIA is most effective when it forms an integral tions in all policy areas rather than only to selected part of the existing regulatory policy develop- portfolios or minor regulations. Only in this way ment, rather than a procedural hurdle towards can RIA maximize its potential to improve the the end of the policy development process, when quality of regulations. In practice, RIA require- ministers and/or regulators are committed to par- ments are sometimes not applied in particular pol- ticular regulatory positions. RIA has to be under- icy areas (e.g., environmental regulation) because taken early in the policy development process and of concerns about the feasibility of assessing bene- before a decision to regulate has been made. fits, or even costs. However, while such practical problems may be significant, the process of sub- Integrating RIA into existing policy development jecting policy proposals to the basic RIA logic (i.e., processes is usually a long-term process and is the clearly specifying policy problems, objectives, iden- result of several factors. A necessary part of this tifying alternative options, weighing all benefits process is to maximize understanding of those and costs and seeking the option with the most net involved in regulatory policy of the underlying benefits) will still improve decision making and logic and elements of the RIA process. That is, policy outcomes. senior government officials as well as those directly responsible for preparing RIA documents should understand the objectives and core features of Appropriate methodological RIA, as part of a good decision-making process. requirements RIA can only become integrated into the existing Consistent decision making on regulatory issues, policy process if the primary responsibility for its across the range of policy areas, can only be implementation lies with line ministries and regu- applied if broadly consistent methodological latory agencies themselves. While a central agency approaches are also taken in RIA. This means with specific RIA responsibilities can act as a that the basic elements of the required methodol- source of technical advice and perform an impor- ogy and approach should be specified in the RIA tant oversight and quality control function, regu- policy and/or supporting guidelines. This includes latory agencies should prepare and complete RIA requirements to specify the policy problem and in relation to their own regulatory proposals. objective, identify all feasible policy tools for 12 achieving the objective – both regulatory and processes to focus on the most significant regula- non-regulatory – and assess the impact of each in tions. Alternatively some governments establish- a consistent, evidence-based way. ing RIA processes for the first time focus RIA on particular areas of regulatory activity. For exam- Any important impacts on business and other ple, one approach to targeting is to apply RIA stakeholders should be documented in the RIA. only to primary legislation, since it is usually at This analysis should also include impacts on dis- this level that the most far-reaching impacts are advantaged or vulnerable groups. The weights found. Another approach is to require the gov- given to impacts, benefits and costs could also be ernment body responsible for overseeing RIA to identified in a transparent manner in the RIA. conduct a preliminary analysis and decide which proposals will need to be subjected to a full RIA. Where major impacts cannot be quantified, a sys- Several criteria can be used to determine whether tematic approach (such as Multi-Criteria Analysis)7 a RIA is needed. The most important are likely to should be used to maximize the objectivity and be the size of the expected monetary impacts, transparency of decision making. whether there are major competition concerns, and whether there are important citizen’s rights Whatever methodology is employed with RIA, issues or sensitive policy issues affecting equity, the core criteria and principle underpinning the environment or other issues. RIA is that regulations should only take place where their benefits are greater than the costs. Such targeting is essential to ensure that RIA Furthermore, the preferred approach should efforts are focused on the most important areas of also be the best way forward (e.g., maximize regulatory policy, where improvements in the net gains) compared to other options consid- quality of regulation will have the greatest impacts ered in the RIA. in terms of enhancing economic activity, trans- parency and community welfare. Therefore, in considering how to target RIA, it is important to Targeting of RIA efforts focus on the size of the expected impacts of the policy, rather than the formal status, or legal Depending on the approach taken, RIA can gener- form, of the policy instrument. For example, in ate resource demands both at the implementation the European Union, RIA can apply to most stage and on an ongoing basis. For example, RIA types of regulation, but a screening process is can create significant demands for analytical skills employed so that RIA processes focus only on the that may be in short supply within government. most important regulatory issues. In Australia, RIA is applied to the most important 5 per cent This is one reason why all governments that use and 10 per cent of all regulations made. RIA focus their efforts on particular types and areas of regulation, often by employing screening Embedding this important distinction in a for- mal policy statement on the application of RIA 7 Multi-Criteria Analysis (MCA) is a methodology that allows systematic and transparent decisions to be made even where can prove difficult. Therefore, one response is to quantification of major regulatory impacts is not possible. give the RIA oversight body (or some other MCA involves identifying the underlying policy objectives appropriate organization) responsibility for con- and then determining all of the factors (the criteria) that would indicate achievement of these objectives. These crite- ducting preliminary assessments to determine ria are then ranked in terms of their relative importance. whether RIA is required in particular cases. Such Once this has been done, each of the available policy options can be “scored” on each individual criterion. The weighted a role should be conducted in accordance with scores can then be added together to determine which option published guidelines setting out the broad criteria best meets the policy objectives. OECD, Introductory Hand- to be used in determining when RIA is required. book for Undertaking Regulatory Impact Analysis (RIA), Paris (forthcoming). The criteria employed should be transparent and 13 objective, in part, to minimize the risk that the Enhancing business and community confidence process could be politicized. is a very important ingredient in increasing investment. For example, in many cases stake- Also important in terms of targeting is the ques- holders will be primarily concerned with how a tion of proportionality. Generally, this means proposed regulation is to be implemented, rather that the amount of resources devoted to conduct- than whether a particular regulation should be ing RIA on a policy proposal should be propor- applied. In such cases, consultation provides tionate to the size of the expected impacts of the insightful information about the views of stake- proposal. Therefore, a far-reaching proposal holders and can facilitate a dialogue focused on should be subject to a more detailed RIA than a how the agreed objectives of the regulation can proposal whose impacts on stakeholders are less be achieved, while minimizing administrative broad and/or intense. and compliance costs on business and other affected parties. Consultation mechanisms There are many different types of consultation, including informal consultation, notification of Consultation with affected parties and other stake- pending regulatory reviews (or decisions), public holders is an essential element of a best-practice notices, circulation of draft regulations and RIA RIA system. Consultation with stakeholders and for comment, public hearings, focus groups, the public supports the decision-making process business test panels and business and/or commu- and RIA in several important ways. First, con- nity advisory bodies, etc. In best practice RIA sultation provides a cost-effective way to gather systems, consultation should commence early in the information and data needed to complete a the decision-making process and before a deci- high quality RIA. Those affected by proposed sion to regulate is made. For example, where regulations will often be best placed to provide appropriate, consultation can be assisted by pub- information that will allow likely impacts to be lic release of a draft RIA document as part of the documented and understood. However, when consultation process. This can be followed by conducting RIA, officials must be aware that the release of a more detailed final RIA informing information sought may be biased to favor the stakeholders of the government’s final decision, interests of the parties supplying it. Therefore, it reasons for the decision and how the views of is important in a consultation process to not only stakeholders were considered during the policy collect information, but also assess, consider and development process. test its validity. A second benefit of consultation is that it is a very There are potential risks that should be taken effective way of identifying faulty reasoning, as into account when planning and implementing well as those proposals that are likely to fail consultation. Consultation processes should because they lack acceptability by major stake- focus on the most important issues, to minimize holders who must comply. resources costs and avoid “consultation fatigue.” It is important that the consultation process not A third benefit is that consultation increases simply reinforce vested interests, is accessible to a transparency and helps build trust and under- wide range of community and business interests, standing. This in turn generates several impor- and communicated effectively. Furthermore, tant benefits, including improving business and failing to take account of comments collected community confidence in governments, their through a consultation process can significantly regulatory systems and particular regulations. increase the risk of regulatory failure. 14 Data collection strategies in RIA. Political backup for such a role is funda- mental. Central units that are powerful are better While consultation with stakeholders can pro- positioned to impose quality criteria and create vide an essential source of information and data incentives for better regulations prepared by reg- for RIA, other strategic approaches are also ulatory agencies. needed to ensure that the RIA is based on high quality, accurate and timely information. In a best practice system, the oversight body should be located in a central agency (i.e., in the The RIA oversight body should provide general President or Prime Minister’s Ministry, or in the guidance on sources of data that regulators should Finance Ministry). In most cases, such central seek out in preparing RIA, while the formal RIA agencies are neither regulators, nor advocates for policy should set out expectations regarding the particular sectional groups. Therefore, they are quality and type of information that should be better able to exercise the RIA scrutiny role cred- included in RIA. For example, in many countries ibly, without major concerns over conflict of regulators preparing RIA are expected to estimate interest. Another important requirement is that administrative burdens on business of the various the oversight body brings a “whole of govern- options considered in the RIA, often by using the ment” perspective to its role, rather than favoring Standard Cost Model.8 There are several other ways any sectional interest. to collect data for RIA. For example, a good start- ing point can be government statistics and data Central oversight also provides an opportunity for bases, which often identify the number of busi- technical assistance to be provided to improve the nesses, size of particular sectors of the economy, RIA. Particularly in the early stages of RIA imple- features of the operation of markets, etc. Expert mentation, regulators sometimes need assistance academic, consultants and research organizations in developing high-quality analysis. The central can also be consulted and, where appropriate, tools oversight body is in the best position to develop such as surveys, synthetic analysis (e.g., estimating specific expertise in RIA-related topics and dis- impacts on a hypothetical business) and related seminate this knowledge to regulatory agencies tools can also be used. Some countries, such as preparing RIA. Indeed, a related role is that the Denmark, have also adopted formalized mecha- central RIA authority can and should focus on nisms for data gathering, including standing busi- capacity building within regulatory agencies. This ness consultative groups. will improve RIA standards over time and improve compliance by encouraging greater understand- ing and acceptance of RIA within government. Central oversight of RIA The oversight unit should be creative in improv- ing RIAs based on technicalities and encouraging Regulating ministries often have a range of incen- regulators to perform better in the future by pro- tives to proceed with regulations, even where they viding guidance and comments (see Box 3 for are not aware of their likely impacts, or where some key capacities required from regulators). regulations impose significant and unnecessary costs on stakeholders and the community. This The central RIA authority can also act as an advo- means that scrutiny of RIA by a central unit, cate for RIA, encouraging compliance and high- body or agency is essential to ensure that high lighting cases of non-compliance. More broadly, it quality analysis is undertaken in a timely manner can also assess the performance of the RIA system overall and recommend improvements that can 8 A detailed explanation of the Standard Cost Model (SCM) can be found in the paper Tools and Approaches to Review enhance the performance of RIA over time and Existing Regulation prepared for the BRG Program. ensuring it remains a best practice system. 15 officials responsible for RIA. Such training should Box 3: Capacity Building provide a broad understanding of the role and in a RIA System importance of RIA, the specific procedural requirements in place and the technical skills The success of a RIA system depends substantially required to prepare RIA. on regulatory officials having effective incentives to prepare high quality RIA in a timely manner. Offi- Training efforts must also be maintained over the cials also require capabilities to play their roles effectively. The main capacities in question are: long term. Movement of officials between minis- tries and changing roles means that there will be 1. An understanding of the RIA system that has a constant supply of new official demand for RIA been adopted by government and their respon- sibilities within it. training. Moreover, RIA standards should be pro- gressively increased over time as expertise devel- 2. Understanding of the logic of RIA as a tool to improve policy. This involves understanding the ops. This also means that additional training will limits to government regulatory action and the be needed. need to make rational choices between alterna- tives based on comparative analysis. An important corollary to training provision is 3. Understanding of the specific methodological the publication of relevant guidance materials on tools used in RIA, such as benefit/cost analysis. RIA processes and methodological issues. While 4. More broadly, developing the RIA system must a wide range of guidance materials is available, include making sure there are adequate links RIA authorities should focus in particular on between government and stakeholders to allow ensuring that the guidance they provide is suited for the necessary consultation and information- to the specific RIA context involved and the par- gathering to underpin the system. ticular needs of their RIA officials. The transparency and credibility of the RIA pro- Oversight of legal quality cess can be enhanced further if there is an inde- pendent advocacy body, often comprising a mix Regulation must be of high legal quality if it is to of senior officials and representatives from busi- be fair, enforceable and consistent. Proposed reg- ness and other stakeholders, who are able to ulation should also be scrutinized for legal qual- monitor and report to senior political leaders on ity during the development process. This is a role the effectiveness of the RIA process (including for a specialist agency, with specific training in the role of the RIA unit, quality of RIA, etc). high-quality legislative drafting. This issue is, in itself, outside the scope of the formal RIA system, in part, because legislation is usually drafted only after a RIA has been considered and a decision to Development of adequate regulate has been made. However, checking to capacities ensure that this element of quality control has The skills required to complete high-quality RIA also been considered is usually seen as forming are specialized and may not be widespread within one part of the RIA system. the administration. Consequently, a major goal of a best practice RIA system must be to develop the relevant capacities of regulators progressively Applying RIA thinking to existing, and over time. In addition to providing technical proposed laws expertise on an ad hoc basis, as suggested above, a central agency unit with responsibility for the Reforming existing regulation is a high priority RIA program should focus strongly on training in most countries and may be particularly so for 16 Figure 4: Elements of RIA Best Practice and Challenges for Developing Countries 1. Formally established RIA policy, with endorsement at a high political level ■ Key in both developed and developing countries, but in the latter even more relevant as these countries are characterized, in most cases, by top-down policy and decision-making process. RIA champions inside the administration in developing countries are fun- damental to introduce the use of this tool and to support it over time. 2. Integration of RIA into the policy process ■ One of the major challenges for a succesful RIA system, both in developed and developing countries. Hardly achievable in the short term, but fundamental to make RIA relevant in the medium and long term. In developing countries it is difficult to expect that line ministries and regulators can take full responsibility of RIA and make use of it in the decision process. This is something that gradually has to be integrated and improved. 3. Consistent application of RIA ■ Key in establishing clear criteria that have to be followed by all regulators. In developing countries, this is fundamental to introduce changes in the administrative culture and provide guidance on screening mechanisms. 4. Appropriate methodological requirements ■ In developing countries, a full cost/benefit analysis is difficult to conduct. Some methodological requirements are needed, but they should be expanded gradually over time. 5. Targeting RIA efforts ■ In developing countries, it is fundamental to apply RIA initially to the most significant regulations to show the usefulness of the tool. A good selection of draft proposals to be subject to RIA is key in developing countries to measure the magnitude of the efforts needed. 6. Consultation mechanisms ■ Key in any RIA system, and fundamental in developing countries in order to open up the decision-making process, to reduce discre- tionality in decisions and to ensure proper inclusion of stakeholder’s views. 7. Data collection strategies ■ In developing countries, data collection might be challenging as information is not available and data are not systematically gathered. It is however essential to introduce some basic estimations in the initial RIAs to provide decision-makers with evidence based information. 8. Central oversight of RIA ■ Key in any regulatory system to ensure the quality of RIAs. In developing countries is particularly relevant to promote coordination, to establish clear criteria to be applied across the administration and to build capacities inside the administration. 17 9. Development of adequate capacities ■ Key in developing countries, whose public administrations lack in most cases of experts and capacities to undertake tasks associated to RIA. 10. Oversight of legal quality ■ This element is present in most countries, and some developing countries have well-staffed bodies looking at the legal quality of law proposals. A challenge in developing countries is to link the existing capacities to oversight legal quality with the RIA process. 11. Apply RIA thinking to existing, as well as proposed laws ■ In developing countries, a choice on RIA application has to be made, starting with new proposals, and gradually be expanded to other existing laws. In the short term, it is difficult to envisage a full application of RIA to all regulations. developing countries. The principles of RIA are most of the best practice elements are not in place equally applicable to the review of existing laws as and RIA systems are not fully operational. OECD it is to the assessment of new regulatory propos- countries show a great degree in diversification of als. When applying RIA to existing laws the same RIA systems. Best practice elements have been an issues arise in relation to its use with new regula- important driver to improve them, but limita- tory proposals. Thus, the best practice elements tions have been documented.9 For developing highlighted in the above sections remain gener- countries, the benchmark is certainly relevant, ally relevant to both the stock of existing regula- but difficult to be measured against it. tion and the flow of new regulation. Figure 4 presents some of the challenges of However, processes for reviewing existing regula- adapting RIA best practice elements to devel- tions will often be fundamentally different from oping country contexts. Acknowledging the using RIA to review the flow of new regulations. relevance of each one of them, the table is an A “stock review” for example, will often be attempt to identify which ones are essential to anchored to a specific decision to review a defined start designing a RIA system and which others number or types of regulations, with a dedicated can be integrated gradually, once implementa- task force charged with the assignment. Further- tion has taken place: more, reviews of existing regulation generally – and appropriately – are undertaken at a much The following section focuses on identifying more aggregated level than is RIA for specific new basic requirements for RIA in developing coun- regulatory proposals. tries, given the above-mentioned challenges to establish far-reaching RIA systems that are fully operational in very few developed countries. Adapting RIA best practice to developing countries: Challenges in implementation 9 OECD country reviews on regulatory reform, and particular the background chapters on “Government Capacities to OECD experience shows that RIA best practice Assure High Quality Regulation” assess practices against is difficult to achieve. In many OECD countries RIA Best Practices. 18 MAKING IT WORK: IDENTIFYING BASIC REQUIREMENTS FOR A FUNCTIONAL RIA SYSTEM The previous section identified the major elements them needs time and perseverance, as well as of a best practice RIA system. However, where strong commitment to introduce changes in the RIA processes are being established or significant administrative and political culture. capacity constraints exist, less ambitious approaches will be needed. Indeed, the development of a RIA system is a long-term undertaking and ana- RIA in developing countries: Evidence lytical and process standards should be progres- from current experiments sively improved over time as experience and capacities allow. RIA can yield important benefits In emerging and developing country contexts, few for decision-makers even where a best practice examples of successful RIA implementation have system is not yet feasible. been documented. A number of emerging and transition economies, such as Mexico, Korea, Developing countries can take advantage of RIA Poland and the Czech Republic have set up RIA systems if they are not only adapted to their par- systems that mirror those of highly advanced ticularities, but also integrated gradually in the OECD countries. Results in countries that recently policy-making process and ensure that existing joined the EU, such as Latvia, Slovenia, and Hun- mechanisms can be used and improved over time. gary, have often indicated that RIA processes exist Best practice RIA does not seem to be the best way more “in form, rather than in substance.”10 In to introduce and design RIA systems in develop- most cases, existing formal mechanisms are some- ing countries, as they do not have the same amount times ignored due to the lack of appropriate skills, of resources, capacities, technical skills and degree support structures and real commitment to their of transparency in decision-making processes. RIA operation. This makes RIA implementation chal- is also a tool that requires democratic principles lenging, as many countries try to reinvent the and mechanisms to be effective. These constraints are acute in developing countries and overcoming 10 SIGMA (2007), p. 51 19 wheel, instead of making use and improve existing to integrate RIA into their policy-making process, mechanisms already in place. are embarked in a number of policy reforms (civil sector reform, decentralization, anti-corruption A small number of developing countries are now programs, etc) that are difficult to manage and also trying to set up RIA processes, but little evi- coordinate at the same time. The amount of dence has been compiled on the results and impacts resources that can be particularly dedicated to of such systems. In some cases, such as Tanzania establish RIA systems gets reduced by an over- and Uganda, RIA experiments have replicated whelming number of reforms that also require OECD best practice with very few positive results attention by government officials. RIA seems to to document. Countries driven by a possible EU compete with other reforms and it is not always a partnership are closer to the OECD benchmark, top priority in the reforms to be implemented. but still far away from showing results. It is there- fore relatively early to extract conclusions, since in In many developing countries, and in particular most cases these process are at early stages of design in the African ones, the actions of leading deci- and implementation. However, even at this early sion-makers within the government have to stage actual and likely challenges are becoming match their public commitment to regulatory identified that suggest that best practice RIA as reform. A commitment to evidence-based deci- described in Section 2 is unlikely to be achievable sion making and consultation with all stake- in the short to medium term. holders introduced into individual regulatory areas has to be supported by clear and broad Concerning some of the key elements that any adoption of such an approach by the champi- RIA system should contain, such as political sup- ons themselves. However, commitment to this port and building capacities, experiences in principle has not always been observed in the emerging and developing countries seems to be developing countries reviewed by IC. In the challenging. Even if political support for regula- case of Uganda, for example, incentives for tory reform has been observed in some develop- lower-level staff in the ministries to promote ing countries, such as Tanzania, Kenya and Ban- the regulatory reform agenda have been signifi- gladesh, there is no clear evidence of the creation cantly weakened by the fact that in several high- of capacities within governments to integrate the profile cases the evidence-based, decision-mak- use of this tool in the long term. These kind of ing process has been replaced at crucial times by concerns are also seen in middle income coun- political considerations. tries such as Mexico, Korea and Croatia, where developing skills among regulators in charge of Another region attempting to establish RIA is RIA has been a core task of the institution in southeast Europe. While significant progress has charge of RIA, but results have been uneven. been made, each country reviewed12 has found the application of RIA to more than a limited number Another evidence observed in many developing of laws extremely challenging. Serbia is the most countries refers to the role of regulatory reform as advanced in the region, with the introduction of a stand-alone policy area in the reform process of mandatory RIA for legislation proposed by the these countries. In many cases, regulatory reform government. Although the implementation of this initiatives are only a minimal part of broader pub- requirement has been hampered by limited lic and private sector reforms. Countries such as resources, RIAs are now being selectively conducted Tanzania, Uganda and Kenya11, which are trying by the Council for Regulatory Reform. In the cases Zambia, as part of the evidence base of regulatory reforms 11 In the framework of the Better Regulation for Growth in developing countries. (BRG) Program, IC has conducted country regulatory 12 Albania, Bulgaria, Bosnia and Herzegovina, Croatia, capacity reviews of Tanzania, Kenya, Uganda, Rwanda and Moldova, Macedonia, Romania, Serbia and Montenegro. 20 of Bulgaria and Moldova, although there are regulations by making decisions based on more requirements for mandatory RIA for primary and transparent mechanisms and evidence-based secondary legislation, this tool is not being applied information, as well as in consultation with rel- systematically. Croatia has completed pilot RIA evant stakeholders. This should acknowledge exercises and is currently applying the RIA to assess the particularities of developing countries, espe- the potential impact of selected draft laws.13 cially the limited institutional and human capac- ities, the discretionary nature of policymaking In southeast Europe, the view that regulatory qual- and the scarcity of reliable data. ity is a determinant for private investment and growth has increased over time. These countries The following discussion identifies the five mini- have made use of opportunistic approaches, build- mum requirements for a functional RIA Light ing on administrative simplification initiatives and system: taking advantage of opportunities to conform to international practice with the assistance of inter- 1) Political commitment to establish and oper- national donors. Another important driver has ate an effective and self sustaining RIA pro- been the “pull” of EU accession, and the need to cess. comply with international membership require- ments. Clear impacts on the use of RIA, however, 2) A unit or group of regulatory reformers– are not recorded so far. preferably based in a central area of govern- ment – which oversees, comments and Looking at the different efforts across countries, reports on the quality of regulatory propos- RIA in emerging and developing countries als, before decisions about regulation are requires adaptation from the OECD “best prac- made. tice” discussed in Section 2. Having the interna- 3) Consistent criteria and rules employed to tional benchmark is important, but for RIA to be screen regulatory proposals. feasible and effective in other contexts, it has to be adapted to the limited capacities that most 4) A transparent regulatory policy development developing countries face. process that includes consultation with stake- holders. Promoting a new approach to impact 5) A capacity building program, involving analysis: RIA Light preparation of guidelines, training of officials preparing RIA, and establishing monitoring, There is considerable merit in focusing on estab- evaluation and reporting systems. lishing “RIA Light” systems that perform the basic functions of a RIA system and are self sus- Taken together, these institutional and concep- taining. The RIA process and documents under tual building blocks provide the basis for estab- this system are nevertheless limited in scope and lishing and operating the framework for a RIA application because of capacity constraints and/ Light system that is conducive to the objectives, or political priorities. needs and capacities of developing countries. However, it is important to note that no country The rationale behind establishing RIA Light today with a RIA system employs all of the con- systems is to improve gradually the quality of cepts and issues identified here. Therefore, the building blocks contain a menu of RIA elements that can be used both as a diagnostic tool and also 13 Thomas, Margo (2007), “South East Europe: opportuni- ties and challenges for improving regulatory quality” in to help identify elements that could be consid- Kirkpatrick, (2007), pp. 253. ered and adopted. 21 Furthermore, it is important to note that these be set up. RIA Light systems in developing coun- issues and elements where they are to be used tries will have some of those elements, not all of need to be considered and employed in a planned, them. The boxes essentially provide a menu for sequential and phased manner. Ideally, these five developing countries to select RIA components minimum requirements for establishing and that will work best for them when establishing or operating a RIA Light system should be applied strengthening their RIA processes. sequentially. It is important to note, however, that many countries have some of the compo- Political commitment to establish and nents of RIA – such as consultation policy and operate an effective, self-sustaining process – but not a functioning RIA Light sys- RIA process tem. In this case, the existing institutions and processes can provide a basis for the establish- Formally establishing RIA policy with high-level ment of RIA Light. and ongoing political support is a fundamental first step in implementing RIA (see Box 4 for In the following subsections the building blocks some issues to consider when looking for politi- are developed. The information presented in each cal support). The importance of high-level politi- one of the boxes of the different building blocks cal support is at its greatest at the time that RIA is intended to provide more operationally rele- is being introduced as a new concept to govern- vant information on how RIA Light systems can ment administrations and policymakers. Box 4: Elements to Consider When Looking for Political Support to Integrate RIA Light in the Policy Process ■ There must be high level political commitment from the head of government or senior minister that can be sustained over time. ■ There should be high-level endorsement, usually through a formal government policy, with a senior minister being responsible for the RIA process. ■ RIA can be established under law, or through a policy statement of the executive arm of government (eg., Presidential Decree, Prime Ministerial Instruction, Cabinet Handbook etc). ■ RIA law or policy statements should establish the core elements of RIA, and also provide flexibility for the RIA process to be amended over time, as priorities change. ■ Policy statements could include the following details: ■ who is responsible for the policy at the political level. ■ area of government that is (or will be) responsible for administering the policy, governance and reporting arrangements – including who should provide central oversight of the RIA process and RIA documents. ■ when RIA will be prepared and by whom, rules about the targeting of RIA efforts. ■ how RIA will be integrated within policy-making consultation and decision- making processes (eg., pre- sentation of RIA to decision makers for consideration before regulations are developed). ■ how RIA documents are to be used in public forums (eg., should draft RIA be used as part of community consultation processes; should final RIA be published after a decision has been made, etc.) ■ clear criteria for deciding and indicating in government policy when RIA should be prepared. It is not feasible to prepare RIA for all new regulatory issues and proposals, so RIA efforts need to focus on par- ticular types or areas of regulation (see Box 5 below for further information). (Continued ) 22 Box 4: Elements to Consider When Looking for Political Support to Inte- grate RIA Light in the Policy Process (Continued) ■ broad methodology to be used as the basis of RIA (eg., cost, benefit and risk analysis, cost effectiveness analysis, etc.) ■ core issues and types of impacts that should be considered in RIA, such as economic, impacts on business, administrative and/or compliance costs, social, environmental, regional, equity etc ■ Transitional arrangements should be made. Establishing a functioning RIA process takes time and needs to be implemented sequentially over time. Often a set of milestones and target dates are included in the RIA policy. Box 5: Important Elements, Responsibilities of Regulatory Reform Units Important elements of regulatory reform units: ■ The unit needs clarity on which area (or areas) of government will oversee, provide quality control and report on the RIA process and/or of RIA documents. ■ The unit should have a high level of independence from regulatory agencies and be able to provide impar- tial, objective and expert advice about RIA processes and documents. This is one reason why such units are usually based in a central agency not actively involved in making and administering regulations. ■ The RIA unit, as in some countries, could focus on assessing and commenting on RIA documents, while a separate high-level advisory body – often comprising senior business and community leaders and indepen- dent experts – could monitor and report on the operation of the RIA process. In other countries the RIA unit performs both functions. Possible responsibilities of regulatory reform units: ■ advise departments, agencies, ministries on when RIA should be prepared and type of information that should be included in the RIA. ■ examine draft RIA and advise whether they meet minimum requirements. ■ advise decision makers about the quality of analysis contained in the final RIA. This role does not usually extend to verifying the accuracy of data included in the RIA, as this is the responsibility of the sponsoring agency. ■ provide training to officials preparing the RIA. ■ provide technical assistance and advice to officials preparing the RIA. ■ prepare reports on the functioning of the RIA process. For the same reason, having ministers take a for- A unit or group of regulatory reformers mal role during the early stages of implementa- and quality control by an expert tion may be a particularly useful strategy. In supervisory body Brazil, the Civil House of the Presidency of the Republic is the body in charge of RIA design The role of a central unit or authority with respon- and implementation. In Moldova, the Ministry sibilities for overseeing and reporting the RIA of Economy has taken the lead in assessing draft program is of even greater importance when a laws for their potential impact on business. new RIA system is being developed and capacities 23 are relatively low. Thus, particular attention institutions being reformed and with political should be paid to the responsibilities given to this power inside the government. Mexico assembled body and to the resources provided to it to carry a small, high-level group of professionals outside out its functions. the bureaucracy, called the Economic Deregulation Unit, at the early stages of RIA implementa- In the best practice model discussed in Part tion. Over time, this institution became the Two of this paper, it is argued that regulators Federal Regulatory Improvement Commission should conduct RIA themselves, subject to (COFEMER). quality control at the center of government. This is seen as an important means of integrat- In other developing countries, institution build- ing RIA disciplines with the broader policy ing has to increase gradually. In Macedonia, for process. The potential benefit is in ensuring instance, the government created in 2006 a that the quality of analysis is high enough to Committee for Implementation of Regulatory have real impacts on the quality of information Reform. This committee supported the guillo- provided to decision makers. A key point is to tine process, coordinated by a Sector for Regula- ensure that the RIA unit works cooperatively tory Reform. This institution, which consists of with relevant regulatory agencies in developing 34 working groups (14 within ministries and 20 the RIA. Regulating agencies must have a clear within other government bodies), will also be understanding of the policy proposal and pro- responsible for introducing RIA, developing a vide information inputs into the RIA. Further- RIA manual and RIA forms. Kenya, for instance, more, RIA units must also involve regulators in started a licensing reform process in 2005. The the analytical process as a practical way of Ministry of Finance established a small unit, the enhancing RIA skills. Business Regulatory Reform Unit (BRRU) in 2007, which will be in charge of introducing RIA units also need to have the option of request- RIA. In these cases, the role of the central units ing that regulators commence the RIA, with the has been essential to move forward the reform analysis subsequently being refined by the RIA agenda.14 units. Practices in this area should evolve having regard to the need to balance several factors: to ensure timely information is provided to deci- Clear criteria: What to subject to RIA sion-makers, that capacities are not over-stretched and what questions to address and that learning occurs over time. Establishing clear criteria to conduct RIA is an important step to create a new administrative cul- Finally, RIA units should be able to report to the ture. Most countries using RIA have established decision-makers, and identify areas where RIA those criteria in a legal document (primary law, does not meet minimum standards, without fear presidential decree, Cabinet order, etc.) to make or favor. In many cases, such units also produce them compulsory for all regulatory bodies, defin- publicly available reports that describe the opera- ing in this way how the RIA process has to take tion and impacts of the RIA process over time, place. Criteria refer not only to what is going to be the role of the RIA unit, and compliance with the subject to RIA (regulatory proposals concerning RIA requirements (see Box 5 for a list of elements primary laws, secondary laws, sectoral regulations, and responsibilities of RIA units). licenses, etc), but also those questions that need to In developing countries, this criteria has been met in leading reformers. Croatia, for example, estab- 14 For detailed information on the role of institutions for lished the Office of RIA, a single reform unit, at regulatory reform, please refer to the BRG paper on Build- ing and Strengthening Institutions for Regulatory Gover- the center of the government, independent of the nance at the Center of Government. 24 be addressed during the RIA process and have to a magnitude in change that is unattainable. RIA be part of the RIA document. Light should be designed to prevent this. Governments typically provide written guidance Once this criteria has been formally established, describing how the RIA process works, with screening becomes a necessary tool to make the checklists and criteria establishing minimum best use of existing capacities to undertake the standards for the RIA. As noted above, the impor- analysis per se. When regulators approach an tance of targeting RIA efforts is greatest where issue for the first time, they should consider all capacities are low. It is preferable that RIA Light options to undertake RIA. However, it is not fea- is applied to major regulatory proposals for sev- sible in a RIA to consider 10 or 15 options and eral reasons, including: undertake a full impact analysis on each. The screening process is designed to weed out options ■ a more consistent application of RIA to major that are simply not sensible to consider further. issues will help to establish it as an integral Unrealistic options that usually involve a dispro- part of the policy process; portionate or high risk response should be dis- carded. The RIA Light then should consider in a ■ even relatively limited scrutiny can be effec- more careful manner the remaining options and tive in screening out some of the least well- the use of alternatives. A consideration in detail considered policy proposals; and between two to five options seems reasonable in countries that are starting the process. ■ the policy will be more credible if applied to major issues in all portfolio areas. The screening process can help determine the “significance of impacts.” Even if there is degree Given the above, targeting means that RIA should of subjectivity in this process, in the context of be applied only to the most important policy RIA Light the significance is generally considered proposals. Therefore, some kind of screening to be the scope of impacts (e.g. wide impacts or mechanism is needed to ensure that RIA efforts narrow impacts) compared to the intensity of are properly targeted. This should be based on impacts (e.g. high, medium, low intensity). A the application of known rules and should occur matrix can be used to identify in the horizontal at the center of government (Prime Minister’s axis the intensity of impacts and the scope of Department or Ministry of Finance) so that the impacts in the vertical axis. The most significant decisions made are seen as authoritative. regulations in terms of impact will therefore qualify for a RIA Light. Screening mechanisms: Where to start? Consistent application of RIA and use of ‘pilots’ RIA Light can be applied to a legal instrument (law or regulation) or to a policy proposal. It is It is a good practice to apply RIA in consistent important to assess the amount of possible ways across all policy areas. This includes the levels impacts, so efforts to apply RIA Light can con- of analysis in RIA being proportional to the likely centrate on the most significant legal instrument size of the regulatory impacts in question. Regula- or policy proposal. tory proposals of similar importance should there- fore be treated consistently under RIA. A first step in conducting RIA Light is to identify the policy problem and policy objectives. The key Some countries have chosen to adopt pilot pro- point is to identify the size of the gap between the grams, which may apply in only one or two policy problem and the objective of government actions. areas, or to a defined set of policy proposals, A common error is to set an objective that requires before fully implementing their RIA processes. 25 The advantages of RIA pilots are, among others, tries. However, in a newly established RIA system the following: or in circumstances of limited capacities, like the case of transition and developing countries, a for- ■ They may be used to provide practical evi- mal benefit/cost analysis may not be feasible due dence of the benefits of RIA in the particular to the lack of proper data, skilled expertise, etc. country context before the government com- mits itself to full introduction of the policy. Efforts to introduce RIA methodologies in devel- oping countries have not proceeded smoothly to ■ They may reveal weaknesses in the proposed date. Less progress has been made in the area of RIA system design, or highlight areas in which analysis of options and cost benefit calculations changes may better adapt the policy to the rel- in countries such as Tanzania and Uganda. The evant institutional and cultural contexts. entry point of regulatory reform has been licens- ing systems (in the case of Tanzania, the entry ■ They help develop RIA expertise and capaci- point has been wider, involving land and labor ties that can then be applied to a wider range regulations), and RIA as a technique has not been of regulatory issues. broadly applied. The effort to measure costs and benefits has been at best limited to application of The question of whether a pilot program is desir- the standard cost model, where calculations have able must be considered in the specific context of been possible to date only through use of heroic the individual country. A pilot program may assumptions. Efforts to introduce RIA in Uganda result in useful lessons as to how best to tailor the and Tanzania, for example, have been character- RIA system to the country’s specific circumstances ized by a large gap between best RIA practice and and so increase the prospects of success of the what is feasible in a data-poor economy. larger RIA program. Moreover, where support for RIA is limited or uncertain, a carefully designed This evidence shows that an exhaustive impact pilot program may help to make the case for its analysis, such as the kind recommended by OECD wider adoption by helping to demonstrate the best practice, is rarely feasible in developing coun- practical benefits of RIA in the specific country tries. In the case in Uganda, for instance, the RIA context. A pilot may also highlight specific imple- Handbook includes the impact of regulation on mentation issues and help RIA authorities to “fine groups that are “particularly susceptible to dispro- tune” RIA requirements to better suit the circum- portionate regulatory impact,” namely SMEs, stances of the individual country. vulnerable groups (the poor, women, children, the elderly, the disabled, people living with HIV/ That said, expectations must be carefully man- AIDS, etc), different districts, tribes and religions, aged, given the difficulties of demonstrating and civil society and NGOs. It is not clear whether concrete benefits from RIA in the very short any RIAs that incorporated such a distributional term. Furthermore, such approaches, if employed, analysis were carried out in fact so far. Even if pov- should have a finite lifespan and ultimately be erty reduction is a goal in many developing coun- replaced with a broadly applicable and consis- tries, there is no evidence about the particular use tently applied program. of RIA to tackle this concern.15 15 It has been observed that one of the possible reasons why Appropriate methodological requirements RIA has not been widely used in developing countries is that “the methodology proposed in the OECD guidelines The choice of a method to conduct the RIA is of does not readily transfer to these countries with their very key importance. A variety of methodological different economies and their greater focus on sustainability and poverty goals.” The suggestion is therefore to have RIAs approaches exist, the benefit/cost principle being that are “pro-poor, placing an explicit heavy weighting on the most widespread and used in OECD coun- poverty reduction and skewing the assessment in favor of 26 Therefore, a basic RIA Light methodology should identify how it will be monitored and reviewed be used that is tailored to the newly implemented in the future. RIA system and/or available capacities and resources, by: The analysis itself has to be of qualitative and quantitative nature. As RIA Light is intended to ■ identifying clearly the problem and then the establish a process of thinking about policy and objective of the policy proposal (i.e., clarify- regulatory options, a discussion of major issues is ing the end to be achieved, and avoiding fundamental. This qualitative assessment has to confusing the end with possible means of be complemented with quantitative information achieving it); that over time, as resources and skills increase, can support more evidence-based decision mak- ■ identifying and considering feasible alterna- ing (Box 6 describes some issues involved in set- tives (including no action) and their expected ting clear and consistent criteria for screening). impacts. In the relatively rare cases where it is concluded there is no alternative, RIA should If the RIA policy has high-level political author- explain how this conclusion was reached; ity (as identified above as a fundamental system element) the likely political costs of ignoring RIA ■ using quantitative data and estimates (espe- requirements will help ensure that regulators cially monetary costs) where available to iden- comply with these requirements. Moreover, if the tify and discuss the impacts of each option. It central unit or body has sufficient authority and is likely that only quite limited quantification acquits its function properly, there will be a pow- will be achievable in new RIA systems, particu- erful mechanism to encourage preparation of a larly in developing countries. However, oppor- high quality RIA. tunities to present quantitative data should be taken, even where significant gaps will inevita- Even in developing country contexts then, clear bly exist. Furthermore, qualitative data can and consistently applied criteria should be also be useful in identifying impacts;16 employed to screen proposals. Where this has not happened, RIAs have been rather poor in data ■ Documenting the extent of consultation, and systematic analysis has not been conducted. along with a description of the views of major In Mexico, for instance, even with criteria estab- stakeholders; and lished by law, the quality of the RIA analysis has been poorly rated by the COFEMER, which ■ Including in the RIA a conclusion identifying accepts that much remains to be done to improve the preferred option and explaining why it the level of quantification and data analysis. has been recommended. This discussion There are challenges to securing this, but these should also focus on who, how and when the can be overcome by extensive training of govern- preferred options could be implemented and ment officials and a rather modest methodologi- cal component for the impact analysis with quan- tifications that can be expanded over time. regulatory changes that assist the poor” (Kirkpatrick, 2003, p. 11). However, in Uganda there was explicit incorpora- tion of poverty issues in the criteria to apply RIA. The results in that respect have been rather disappointing. Transparency in the process and 16 Some have argued that the identification of impacts on vul- nerable groups is an essential part of a credible RIA method- consultation ology in a developing country context, since distributional factors are likely to be particularly important in government The role of consultation and transparency is par- decision making. However, it must be recognized that this is a demanding task and one that may be beyond the capabili- ticularly important in newly implemented RIA ties of a newly adopted RIA system. systems in low-capacity contexts. As noted above, 27 Box 6: Clear and Consistently Applied Criteria and Rules Employed to Screen Regulatory Proposals Issues involved in setting clear and consistent criteria and rules to screen regulatory proposals: ■ When should RIA be prepared? It is not possible to prepare RIA for all new regulations, so RIA efforts need to be focused on important high-priority regulatory issues and proposals. It is very important that clear criteria for deciding when RIA should be prepared are set. Such criteria can include: ■ particular types of regulation, such as primary legislation considered by parliaments; ■ particular forms of regulation, such as licenses, restrictions on competition; ■ significant regulations that are of particular concern to business or the community; ■ regulations that impact businesses; and ■ regulations administered by a particular area of government, such as a particular agency or level of government. ■ Who should prepare RIA and when? Draft RIA should be prepared early in the policy development process when regulatory issues are being considered and before a decision is formally or informally taken to regu- late. ■ How should effectiveness of RIA processes be ensured? Ensuring effectiveness requires setting incentives as well as sanctions for non compliance with the RIA process. These include: ■ Decision makers being advised by the RIA unit that a RIA document does not meet minimum requirements established by government policy; ■ public identification of regulatory proposals where the RIA was inadequate, in the Parliament or through public reporting processes (eg., annual reports etc.); ■ mandatory review of regulations, maybe within two years, where a regulation is made but the RIA was inadequate; and ■ performance contracts and appraisal processes for senior officials that explicitly include reference to complying with RIA policy and processes. consultation generates information about stake- effects on the level of analysis and preparation put holders and impacts on them. This information into draft regulatory proposals and RIAs. should be included in RIA and will help address information gaps. More generally, capacity building efforts may need to be undertaken with major stakeholders Formal requirements to publish easy-to-read RIAs (e.g., business associations, consumer groups) to in parallel to consultation can assist the consulta- help them play an effective role in consultation tion process and help focus stakeholders on key conducted as part of RIA – and so contribute to issues and impacts. Presenting final RIA to the RIA (hence regulatory) quality. For example, pro- Cabinet or Parliament may also create strong viding information explaining the nature of the incentives for regulators to observe due process and regulatory proposal, the reasoning underlying it not submit proposals that would be potentially and the alternative proposals that have been con- embarrassing and insufficiently thought through. sidered will provide a basis for stakeholders to Experiences from Mexico have shown that such respond. Asking specific questions can also help publication obligations had strong disciplinary guide their responses and ensure their relevance. 28 This synergy between RIA and consultation means including groups that do not have ready access to that implementing RIA may require simultaneous the government and in educating all parties in action on both policy fronts. Where consultation how to conduct consultation, but the principle of processes are not well-established, consideration consultation that is an essential element of RIA is should be given to developing such processes con- accepted and is being implemented. This should currently with the RIA policy, while the needs of provide a good basis for RIA going forwards. RIA should be reflected in the design of the consul- tation process. Developing a “consultation culture,” with stakeholders being confident and willing to Building RIA capacities within engage with government on policy issues may also government be a medium-term task. That said, practical con- straints on the development of a consultation pol- The capacity-building role of the RIA unit is icy should not be seen as an absolute impediment especially important when a RIA system is being to implementing RIA Light. Box 7 describes some established and, in particular, where this is occur- issues to consider regarding transparency and con- ring in a context of low policy, institutional, and sultation in RIA Light. resource capacities. This implies that the RIA unit should devote a significant proportion of its In the case of Rwanda, Tanzania, Kenya, Uganda, resources to training and support functions. The and Zambia, there has been clear focus and prog- various strategies cited in the previous section ress in the area of consultation. Consultative pro- are all relevant to this context. However, the cedures have been introduced for reforms well issue also arises as to whether RIA training beyond the strictly regulatory arena – all five should be integrated with broader training in countries have, for example, produced long-term public policy issues. That is, if policy develop- “vision” statements, and they have all involved ment and advice capacities in general are low, intensive and broadly inclusive consultation teaching RIA related skills might be most effec- within government and with business and civil tively done if integrated into a broader program society groups. Much work remains to be done in of training officials on policy skills. Box 7: Elements of Transparency and Consultation in a RIA Light System Transparency and consultation are essential for a functional RIA Light system. Some of the issues to consider in this area: ■ Transparency and reporting processes generate the right incentives for participants in the policy development process and should apply to all participants. ■ Processes can include public reporting and oversight by independent advisory bodies, reporting require- ments that apply to the regulation review unit and to regulatory agencies, departments and ministries, via publication of draft and final RIAs, annual reports, etc. ■ RIA processes should be integrated with a broader government policy on community consultation This policy should include information about who should be consulted, when consultation will occur, how consultation will occur, and how to manage any risks that arise. ■ Transparent and effective consultation provides valuable information about the views of stakeholders on regulatory issues, including the impacts and risks of regulatory options under consideration. ■ Consultation generates greater trust between governments and the community, creating a greater sense of shared ownership and responsibility for regulatory issues and how such issues should be best managed. This also helps generate higher levels of compliance with existing and proposed regulatory requirements. 29 In the medium term, as practical experience with ability in these fields as a criteria for promotions RIA develops, a number of supporting strategies to high-level jobs in the civil service. Between can be adopted to help improve the quality of 2000 and 2003, a major emphasis of the pro- RIA outcomes. The role of the central RIA over- gram was to build understanding and awareness sight body in proposing data collection strategies of regulatory reform among top-level officials. is also particularly important when RIA is first Training courses and materials were developed, being introduced, since regulatory officials will including modules on social and environmental have little experience in obtaining and interpret- dimensions of RIA, and were delivered to gov- ing RIA related data. A related function that the ernment, private sector, and civic society. More RIA oversight body should help regulatory offi- than 350 persons received training, including cials gather and analyze data in the course of the 65 policy analysts, 60 or so media and public development of specific RIA. This kind of practi- relations representatives of regulatory agencies, cal assistance can potentially function as a very senior business reporters/editors, and parlia- effective means of developing regulators’ skills in mentary technical and analytical staff. this area. The RBP project developed the capacity of a Some examples of the importance of training gov- local training organization, the Uganda Man- ernment officials to conduct RIA are found in two agement Institute (UMI), to deliver training in emerging countries: Mexico and Korea. In Mexico, RIA to Ugandan officials. A business plan for for instance, COFEMER trained more than 740 UMI‘s RBP/RIA training was developed, and public employees from October 2001 to February the initial training curriculum was modified so 2005. The objectives of the RIA courses covered that it could meet the diverse requirements of different aspects: public servants were thought not the market. One core program was a one-week only how to use the online system for RIA devel- training course in basic RIA, which was offered opment and how to elaborate a RIA, but also to to government’s policy analysts before the RIA improve communication and the relationship with requirement was formally introduced. However, the oversight body itself and the public servants in following the end of the RBP Program, funding charge of the proposals. They also developed skills from government for training courses dried up, in quantifying effects of regulation and how to and it appears that UMI is no longer active in integrate regulatory and non-regulatory alterna- providing training to civil servants. The RBP tives in the options assessed. The Korean regula- Program provided evidence that, when an insti- tory reform, for example, had as shortcomings the tution is at the very early stages of its develop- lack of investment in expertise at initial stages of ment, it is necessary to couple advocacy capacity the introduction of RIA. Since 2005, the Korean building with institutional and other forms of Government has introduced a range of training capacity building. programs for civil servants on RIA. The Prime Minister’s Office has been instrumental in estab- These examples illustrate the fact that building lishing many of them. capacities within the administration to ensure that officials are able to carry out impact analysis In Uganda, where a RIA system has been intro- following established criteria and are trained on duced, capacity building activities are at the core regulatory reform issues on a regular basis is fun- of the regulatory reform program. From the damental in the medium and long term. A RIA beginning of the Regulatory Best Practice (RBP) Light system will depend highly on the efficiency Program, attention was given to training and and sustainability of those capacities created creation of skills related to RIA. In addition, inside the administration. Box 8 lists some issues steps were taken to emphasize the importance of to consider when strengthening capacities in the skills in RBP and RIA by including demonstrated administration. 30 Box 8: Capacity-Building and Monitoring and Evaluation Programs Communicating rules and processes underpinning RIA systems is fundamental for strenghthening capacities within the administration. Some issues to consider in this respect are: ■ Government-endorsed manuals, provided in most countries, which contain detailed information about the rules and RIA processes, such as: ■ objective of the RIA process, why there is a RIA process etc; ■ how the RIA process operates; ■ role clarity – who does what and when; ■ the rules and requirements regarding RIA, when a RIA should be prepared; ■ if a RIA needs to be prepared, the steps involved in preparing RIA; ■ consultation processes that should be used within government and with the community; ■ the information that needs to be included in RIA – structure of RIA, methodology, data etc.; ■ how the central unit will assess each RIA, what criteria are used to determine whether a RIA meets minimum standards; and ■ examples of real RIA provided to assist officials. ■ RIA training program and RIA pilots, which can be provided by central RIA units, or by other bodies (such as consultants). Such training programs build internal capacity within governments by providing helpful infor- mation to officials working on regulatory policy issues. This can include information about how to correctly apply the RIA requirements, where to go for help when deciding if a RIA should be prepared and how to prepare a RIA document. ■ Monitoring, evaluation and reporting processes, including evaluation of the RIA process and RIA docu- ments (as discussed above). ■ Internal monitoring and reporting systems that can include: ■ the operation of the broader regulatory system, such as the number and type of existing regulation and new regulations made, regulatory “hot spots” and issues of concern to stakeholders and broad regulatory trends (such as a country performance according to international measures of the quality of regulations and regulatory systems, including Doing Business, Transparency International, etc.); ■ the number of RIAs prepared and considered by decision-makers; ■ number of these RIA that met minimum requirements; ■ how RIA documents are used, types of issues undergoing RIA analysis; ■ impacts of RIA process on decision-making processes and outcomes, such as cases where the recom- mended option changes during the policy development process, citations of RIA documents by ministers, senior officials, in the Parliament and media, etc; ■ views of key stakeholders about the RIA process; and ■ RIA unit performance, including responses from officials about the quality of RIA training, quality and timeliness of advice about RIA to agencies, departments and ministries etc. 31 RIA Light for developing countries: decision making. The RIA Light approach Summing up the new approach acknowledges that rationalizing the decision- making process is not an easy task and requires RIA Light is a new approach to integrate evidence much more than simple politial will. It offers the base to decision making without imposing heavy possibility for countries to make use of existing burdens in terms of resources and capacities to and new processes that can contribute to the countries where institution building, transpar- improvement of their regulatory decisions. ency and skills remain limited. This approach is not free of challenges, but has Figure 5 presents a summary of the five bulding the advantage of advocating for a gradual integra- blocks of any RIA Light system, assessing their tion and improvement of capacities and expertise, viability in developing country contexts and as well as an adaptable framework to improve making links between each element. Figure 5: RIA Light Elements RIA Light Elements 1. Political 2. A unit or group of commitment to 3. Clear criteria: regulatory reformers 4. Transparency in 5. Building RIA establish and what to subject to and quality control the process and capacities within operate an effective RIA and what by an expert consultation government and self sustaining questions to address supervisory body RIA process As some country exam- For RIA Light to be opera- In developing countries, a In developing countries, RIA Light does not mean ples have shown, politi- tional, a unit in charge of it careful management of transparency in the regu- an extensive assessment cal commitment in devel- has to be in place, with the what is subject to RIA is latory process and con- of costs and benefits of oping countries can be goal to look after its qual- fundamental to make reg- sultation with stakeholders regulatory proposals. But found in a number of key ity. In developing countries, ulators acquainted with cannot be taken for even qualitative and institutions and in almost these units have to be the new tool. Clear crite- granted. RIA Light can some quantitative analy- all cases where other entrusted with basic func- ria have to be established only be effective if the sis requires the develop- public and private sector tions, such as to establish by the regulatory reform central unit develops a ment of skills within the reforms have been previ- clear criteria for RIA, ensur- unit to ensure that the strategy to improve both administration. Capaci- ously introduced. Sustain- ing consultation with stake- administration follows the aspects and ensures that ties are therefore also ability of the political holders and building same instructions. Those more open mechanisms linked to the clear criteria commitment and support capacities within the gov- criteria have to be to discuss regulatory pro- that regulators have to over time remains how- ernment. Units need contin- adapted to the capacities posals are put in place. comply with and to the ever challenging. Expand- ued political support to in place. They also need RIA Light requires as pre- continued support that the ing support over the initial operate. to consider the way they condition some existing central unit can provide. stages has to be linked to can be improved over fora that can be rein- the creation of capacities time, once regulators start forced over time. and the role of the regula- using the tool. tory unit in charge of RIA quality control. 32 IMPLEMENTING AND INSTITUTIONALIZING RIA The previous sections identified five key dimen- requiring a screening of regulatory proposals, sions of a functional, self-sustaining RIA system. which employs the logic of RIA but does not This section explores practical issues and options demand sophisticated analysis. governments should consider in the process of designing and establishing the RIA Light system. As noted above, another means of ensuring lim- ited RIA resources are best employed is to adopt a well-considered targeting of RIA requirements. This could involve an initial piloting of RIA in Initial implementation issues one important policy area. This may be an appro- As stated previously, a fundamental requirement priate approach where there is substantial doubt when first implementing RIA is to ensure that the or concern as to the capacity of the government policy has clear political authority, which is sup- process to adopt RIA successfully, or where there ported at the highest levels of government. Only is a lack of clear support for the introduction of this level of authority will ensure high levels of RIA. A successful pilot program can provide tan- compliance, particularly in the early years when gible evidence of improvements in regulatory many regulators will not perceive the intrinsic quality due to RIA that will increase support for benefits of the RIA process. It is also essential that its wider use. the analytical requirements and standards set are appropriate, given the capacities that exist within Securing compliance with the assessment require- government and the other environmental con- ments is likely to be a major challenge for the RIA straints likely to exist (e.g., limited support for, unit, particularly when RIA is first implemented. and understanding of, RIA). Thus, a careful review Implementation efforts must be designed in ways of the country’s policy and administrative pro- that encourage and require high levels of compli- cesses is needed as a first step in designing a RIA ance. An essential first step is to educate regulators system. RIA Light implementation might involve about the logic of impact assessment. Regulators 33 should understand the purpose of the policy, while there should be a focus on its potential to assist Box 9: Legal Underpinning of RIA them in completing their policy roles more effec- in Emerging and Developing tively and efficiently, particularly by using RIA in Countries ways that are integrated with their own processes for developing proposed laws. This implies pro- Emerging and developing countries have estab- viding general information on RIA to public sec- lished the use of RIA mostly in laws, which might be tor managers, as well as providing training and understood as a way to show strong commitment to supporting materials to officials with specific RIA the use of the tool and as making its amendment or elimination more difficult. These variables are rele- responsibilities within regulatory agencies. vant in developing country contexts, as the political volatility in these countries can interfere in the sus- As noted in previous sections, developing a best tainability of the RIA system. For example: practice RIA program is a long-term undertak- ■ In Vietnam, a draft Law of Laws (2008) requires ing, while initial RIA implementation often must RIA to be conducted for legal instruments issued be based on a limited RIA model, particularly by the National Assembly, the Standing Com- where capacity constraints are significant. This mittee of the National Assembly and the govern- implies that there will be a need for the initial ment. RIA program to be reviewed and revised relatively ■ In Kenya, a draft Business Regulation Bill envis- soon after implementation. This review process ages the introduction of RIA to support the licens- should focus on questions such as: ing reform. ■ In Moldova, a Guillotine Law approved in ■ whether the requirements are set at the right 2006 introduced basic principles for regulatory level and, if not, what changes are required; quality, such as the use of RIA. ■ what specific problems can be identified with the institutional model of RIA Light adopted, RIA policy and, consequently, the degree of com- and how can they be addressed; and pliance achieved. It can also allow Parliaments to reject regulations that have not met RIA require- ■ how can required standards (and hence the ments, potentially providing a further means of benefits of RIA Light) be increased progres- quality control that is outside, or independent, of sively as capacities and experience with the the government. It can allow affected (i.e., regu- system develop. lated) groups to challenge the validity of such laws in the courts, with the possibility of the laws being struck down as a result. This option pro- Options for institutionalizing RIA vides the highest level of authority for the RIA policy and has significant advantages in terms of ensuring that a high level of compliance can be Legal requirement to undertake RIA achieved within a short period after the RIA pol- The option of establishing RIA requirements in icy is adopted. laws is not widely used internationally in the OECD context, though it appears to be more On the other hand, if regulations are frequently common in developing countries (see Box 9).17 and successfully challenged in the courts on the This approach can enhance the authority of the grounds of failure to meet legislated RIA require- ments, there may be problems of confusion and 17 Very few OECD countries have established RIA by law, uncertainty as to compliance obligations. This such as Mexico (Federal Administrative Procedures Law), Italy (Law No. 50 of 1999), Korea (Basic Act on Adminis- approach can also be inflexible, for example not trative Regulation), Hungary (Act on Legislation), etc. being able to cater for genuine emergencies, etc. 34 Another potential problem is that it is often dif- ficult to amend laws frequently. Box 10: RIA Adopted Through Government Policy Statements However, the option of setting out only the broad outlines of the RIA system in legislation, while In Canada, RIA has been adopted in the framework supplementing these provisions with more detailed of the Cabinet Directive on Streamlining Regulation material contained in guidelines, directives and which came into effect on April 1, 2007. This Direc- the like can enable the benefits of enshrining RIA tive introduced several key improvements, including a more comprehensive management approach with in law to be attained, while minimizing or avoid- specific requirements for the development, imple- ing any problems of lack of flexibility. mentation, evaluation and review of regulations. Denmark adopted a RIA requirement in a revised A second and more commonly used option is to Prime Ministerial circular on drafting laws in 1998. adopt the RIA requirement through an explicit Thus, the RIA requirement is established among a policy statement by the head of government, or broader range of procedural requirements for law- by the Cabinet. RIA based on a Cabinet decision making. Ministries are advised to use RIA where or directive, or equivalent, is used in a wide range proposed laws are expected to have substantial economic impacts. The Ministry of Finance guides of countries including Canada, Denmark, Fin- to benefit/cost analysis are expected to be used as land, Germany, Ireland, Japan, New Zealand, the basis for conducting RIA. However, other assess- Norway, Poland, Portugal, Sweden and the United ments of administrative impacts on government and Kingdom (see Box 10 for some examples). citizens, the environment, business, and industry are also required. The circular requires RIA to be RIA is based on a Prime Ministerial decree or included among explanatory notes circulated with guideline in countries including Australia18, Italy draft laws. and the Netherlands, and by Executive Order in Poland adopted in 2006 its Regulatory Reform Pro- the United States.19 These instruments are more gram, which was the first comprehensive program defining an integrated approach to regulatory man- readily amended than Acts. Indeed, the U.S. RIA agement policy. Among the priorities, a revised RIA system has seen several amendments to the main methodology was adopted by the Council of Minis- Executive Order setting out its essential require- ters, and integrated in the Guidelines for Regulatory ments. Nonetheless, this option ensures that a Impact Assessment. Institutional arrangements for high level of authority is still accorded to the RIA RIA were implemented in 2006 to strengthen the policy, including the possibility of challenging RIA system, including placing responsibility for the review of RIA in the Chancellery of the Prime Minis- regulations on the basis of non-compliance with ter instead of the government Legislation Center RIA requirements. where it was until July 2006. The main potential disadvantage of this approach is that RIA requirements can be changed easily and significant changes can be made, say after a been evolutionary and have involved progres- new government is elected. This could reduce the sively strengthening the RIA process. continuity and effectiveness of the RIA process. That said, in most developed countries RIA is Adopting RIA Light through non-legislative supported by all major political groups and, as a arrangements effectively requires monitoring and result, changes to RIA over time have typically enforcement activity to be conducted within gov- ernment, with the RIA unit performing this role. However, in some cases other areas of govern- 18 A Prime Ministerial order is used by the national gov- ment can also do this. For example, a Cabinet ernment in Australia, whereas most States establish RIA in laws. office or similar body can be given responsibility 19 See: http://www.whitehouse.gov/omb/inforeg/eo12866.pdf for ensuring regulatory proposals meet the RIA 35 requirements before they are considered and traditional primacy of legal perspectives in consid- approved by the Cabinet. In this context, detailed ering regulatory quality issues. However, law draft- requirements may be set out in more general pro- ing experts do not usually have strong RIA-related cedural guidance documents, such as a “Cabinet skills, while the regulatory quality concepts under- Handbook” or equivalent document summarizing pinning RIA are much broader than those relating all process requirements for making laws. This can to law drafting. This can mean that RIA does not ensure that the RIA requirement is widely known receive adequate prominence and that the RIA and helps to enhance its perceived authority. body does not function as a powerful advocate of the importance of RIA in regulatory quality. In all options described above the importance of Options for RIA oversight locating the unit at the center of government lies primarily in ensuring its credibility vis-à-vis other RIA oversight within the administration institutions and stakeholders and to use a system of checks and balances to prevent any other min- The OECD best practice is that there should be a istry or regulator from dominating the choice of body within the government administration that regulatory options. The unit can serve to balance has specific responsibility for exercising oversight opposing interests by providing credible and and quality control over RIA. The RIA oversight impartial opinion on the impact assessment and unit, body or authority should be located at the requesting regulators to improve the analysis. center of government, such as in the President/ Prime Minister’s Department or the Ministry of As discussed above, RIA oversight bodies in most Finance. This ensures that it wields appropriate countries have both a role in assessing the quality of authority and helps prevent conflicts of interest, RIA (e.g.,. approving RIA documents as meeting since these agencies do not have direct regulatory quality standards) and also advising and training functions in most cases. It also helps underline regulators in the preparation of their RIA. There the view that RIA (and regulatory reform) is not can be a perception of a conflict between the for- a program intended to confer benefits on any one mer policing role and the latter role of providing sector of society but, rather, is intended to enhance positive assistance. Moreover, this latter role poten- regulatory quality form the viewpoint of society tially requires the RIA unit to have a wide range of as a whole. expertise in particular regulatory issues and areas. One option adopted in the Netherlands to address The main alternative adopted in some countries these issues is to establish a separate ‘RIA Helpdesk’ has been to locate the RIA body and other regula- to focus on providing positive assistance. The Help- tory reform functions within the industry depart- desk has been structured to ensure that a wide ment. This has generally reflected a primary focus range of expertise is available by including staff of regulatory reform policies on improving busi- from the Ministries of Economic Affairs, Justice ness conditions. However, a major problem with and the Environment. However, this option may this approach is that it reduces the authority of be too resource intensive for many developing the RIA body within government. Also, the RIA countries. It is therefore preferable to reinforce unit and process can be seen to be an advocate of resources in the oversight body. sectional (i.e., business) interests, rather than tak- ing a broader view and promoting regulatory Oversight by independent or policy that is in the interest of society as a whole. external bodies In some other cases, the RIA oversight body has The main reasons for adopting the concept of been located in the law department, reflecting the a central RIA oversight body are that it can 36 bring specialist expertise to bear in assessing RIA Integrating RIA with policy quality and provide an independent perspec- decision-making process tive, rather than approaching RIA as regulators with a specific area of policy responsibility and The OECD best practice states that RIA must be interest. This logic can also be applied to bod- integrated with the existing policy decision- ies other than RIA units within the govern- making process if it is to be fully effective in ment administration. Some OECD countries improving regulatory quality. This means that have established different forms of RIA over- those responsible for developing regulation sight, either as an alternative to a body within should understand the RIA and prepare RIA, the government administration or as an addi- where appropriate, early in the policy develop- tional level of scrutiny. ment process and before a decision is made. In practice, this can be difficult to achieve, especially One example was the Better Regulation Com- in the short term. Indeed, OECD documents mission in the United Kingdom.20 The Commis- speak of the need to achieve a process of “cultural sion was composed of external stakeholders from change” through a medium to long-term effort. various parts of society including business, chari- ties and other non-profit organizations. Similarly, Practical initiatives that can help facilitate cul- the Netherlands has used an external watchdog tural change: (Actal – the Dutch Advisory Board on Adminis- trative Burdens) to provide advice to government ■ ensuring that RIA training is provided at on reducing the administrative burdens imposed middle management levels, so that those by regulation. In some cases, these bodies report responsible for developing regulation under- annually to Parliament, providing another form stand RIA and how it can help them to work of independent scrutiny. Similarly, in some coun- better; tries (e.g., Australia) there are specific parliamen- tary committees with responsibility for RIA over- ■ monitoring the impact of RIA and publiciz- sight. In the Australian model, this oversight is in ing cases in which it has improved regulatory addition to that of a body within the government outcomes, to reinforce understanding of its administration. merits; and In a RIA Light system, this independent over- ■ ensuring that RIA process requires prepara- sight can bring positive results, but it is not tion of RIA at a relatively early stage of the compulsory at early stages of the reform pro- policy process. cess. A body that is entirely independent of government has the advantage of being freer to RIA conducted at an early stage is likely to be provide objective advice on RIA issues. How- influential in that the results are considered before ever, a problem with many such bodies is that there is any firm commitment to a particular pol- their scrutiny is exercised late in the regulatory icy approach. On the other hand, RIA conducted process – often after a regulation has been pub- at this point is likely to be assessing general policy lished. This tends to reduce their effectiveness options, rather than well-developed regulatory in practice. proposals, and is therefore limited in the firmness of conclusions that can be drawn. By contrast, RIA conducted later in the policy process can 20 The Better Regulation Commission has been replaced by usually provide a more precise analysis of expected the Risk and Regulation Advisory Council (RRAC), which impacts, but may be less effective in altering pol- is also an independent advisory body responsible for devel- oping some work on public risk and how best to respond icy outcomes if an explicit or an implicit decision to it. has already been taken to regulate. 37 This trade-off must be borne in mind when and decision-makers understand the rationale for designing RIA system requirements. In some conducting RIA. OECD countries, this issue is addressed by requiring that RIA be commenced at an early stage in policy development and then continually Challenges and obstacles in refined as more detailed proposals are developed. Thus, an internal RIA might be developed, based implementing RIA Light in emerging on limited consultation, to guide broad policy and developing countries decisions. A more detailed RIA might subse- Despite the clear conceptual merits of RIA, expe- quently be released for public consultation once a rience in a wide range of developed and also firm proposal has been developed. developing countries shows that there are consid- erable practical challenges in implementing an For example, this approach is required in the effective functioning RIA process. In some cases, United Kingdom, where the RIA guidelines21 spe- the achievements of RIA have fallen short of ini- cifically state that RIA activity should commence tial and often high expectations. Furthermore, early in the policy process and should continually developing the RIA system and increasing its abil- be updated over time. In some Australian RIA ity to affect policy outcomes also often takes lon- contexts a two part RIA process is required. In this ger than expected. case, there is a requirement that a draft RIA be released for public consultation. Comments The following discusses some of the most impor- received are used to improve the analysis and, in tant issues and challenges to establishing func- many cases, drive revisions to the regulatory pro- tioning and self-sustaining RIA processes, based posal before the final RIA is presented to ministers on experiences in developed countries and devel- or other decision-making body for decision.22 oping countries that have already set up a RIA system . It is important to note that many of these This approach has been criticized for increasing challenges are operational and can be managed or the resource costs of completing RIA. However, resolved through administrative decisions. By as noted above the fiscal cost to government of contrast, some require decisions to be made at the preparing RIA is modest. Furthermore, the cost political level. This is particular relevant when a will be negligible where regulators are already RIA Light process is designed and implemented applying good practice in developing regulatory as expectations have to be properly managed and options and proposals, since in such cases the its sustainability will depend highly on the way RIA simply documents the existing policy devel- obstacles are overcome. opment process. Developing countries, introducing a RIA Light system, are not excluded from this trade-off. Lack of compliance They can, however, mitigate the problem by The most basic problem is the need to ensure selecting RIA pilots in which particular attention that regulators comply with the RIA require- is paid to the way RIA results can be integrated ments. Because of institutional resistance (see into the decision-making process, so regulators below), voluntary compliance with RIA is gener- ally low. This means that attention must be paid, 21 Impact Assessment Guidance, Better Regulation Executive, available at: http://www.berr.gov.uk/files/file44544.pdf. firstly to incentives for officials to prepare RIA 22 This process is used whenever State/Territory Ministers when they are required to do so. These incentives agree to adopt nationally uniform regulation. See: http:// www.coag.gov.au/docs/COAG_best_practice_guide_ can include monitoring and encouraging compli- 2007.pdf ance with the RIA requirements and process. Sec- 38 ondly, there should be incentives which ensure In Serbia, for instance, the regulator must recieve that there are effective sanctions available for the opinion of the Council for Regulatory Reform non-compliance. for the legislative proposal prior to submitting such proposal to the government. When the High level political support for the RIA program opinion of the Council is negative, the govern- is particularly important in ensuring that RIA ment refers the legislation back to the regulator compliance is a pre-condition for regulatory pro- or ministry with a request to revise the RIA. If posals being considered by decision-makers. the analysis is inadequate and the RIA unit finds that it is not difficult to improve the analysis, e.g. This is particularly problematic in developing to collect additional data and to calculate countries where RIA might be seen as a new additional costs, the RIA unit remains authorita- bureaucratic layer. Changes in the administrative tive and the RIA does not go through. culture are difficult to introduce in the short term. In Tanzania, for instance, training and RIA Where RIA documents are reluctantly prepared, awareness were stimulated at the beginning of the it is often a case of “formal compliance,” or “going BEST Program by the RIA champion, the Cabi- through the motions,” rather than RIA actually net Secretariat. Even if the Principles of Good influencing policy thinking and outcomes. Action Regulation called for the use of RIA whenever to minimize this obstacle includes ensuring that “the proposed new law has significant compli- senior officials, and not simply officer-level staff ance costs for businesses or significant economic responsible for conducting RIA, are given training or social consequences”, results indicate that RIAs in the nature and importance of RIA. Demon- have not been conducted neither in general nor strating how RIA has improved policy outcomes for the components of the BEST Program. and even how it has benefited regulatory agency officials can be an important means of helping to overcome institutional resistance. Furthermore, it is Institutional resistance important that senior officials have effective incen- RIA can often be seen by regulators as undermin- tives to properly implement RIA. These can ing or limiting their ability to serve their ministers include RIA compliance being included as part of and other constituencies. RIA is also sometimes the criteria used to assess the performance of seen as time consuming and expensive, while its senior staff, and reporting to decision makers (and benefits may not be understood or accepted. This ultimately the public) when a particular RIA fails means there are strong incentives in Ministries to meet minimum standards. and regulatory agencies to avoid the RIA require- ments. While looking at RIAs prepared by regula- That said, the most important incentive for offi- tors, the unit has to explain why assessments have cials and regulators to properly implement RIA is not been properly done, based on the established having high level and ongoing political support criteria that apply to all regulators. This is a way for the RIA process. Such support is essential to to build authority over time. the establishment and operation of a successful, functioning and self-sustaining RIA process. To counteract those negative incentives, the qual- Such support is attainable in developing coun- ity control unit has to be proactive and creative. tries, but lack of commitment over time has been Initially, RIA can be used as a “damage control” observed in many cases as well. For instance, in tool, more than a systemic tool. It is always easier Uganda, the Ministry of Finance, Planning and to pinpoint the damage that has been prevented Economic Development (MFPED) played a piv- than the hardly measured benefits that will require otal role in supporting the introduction of RIA a lot of persuasion. into public policymaking. It had a strong team of 39 senior managers and staff who benefitted from or unrealistically high initial expectations and the project and gave active support to RIA imple- disappointments due to limited early successes. mentation. As the project entered its second phase, however, it became clear that MFPED was Minimizing these obstacles involves managing not demonstrating commitment and ownership initial expectations, including emphasizing the needed to guarantee sustainability. MFPED had medium-long term nature of the commitment to many calls on its time and resources, and its RIA that is needed for success. Indeed, RIA is not responsibility for economic management was a quick fix for existing regulatory problems. diverted increasingly to fiscal and macro issues; Rather, it is about strengthening capacities and changes in key staff also resulted in less engage- institution over time, with the ultimate objectives ment with private sector issues. of creating a more effective and efficient regula- tory system. Lack of credibility of RIA Furthermore, it is important that early gains due Capacity and resource constraints can mean that to RIA are clearly documented and communi- the quality of RIA documents in the short term cated to ministers, senior officials and other stake- can be too low to achieve credibility and have a holders. This can include documenting individual meaningful input into decision making. RIA cases where RIA added significant value to the officials and donors can address this challenge by quality of a submission to decision-makers. More ensuring that a high level of technical assistance is complex and comprehensive monitoring and provided to regulators conducting RIA. This evaluation systems focusing on RIA can also be might be achieved either through making their developed and applied over time. In countries own staff available, providing training and/or such as Uganda and Tanzania, monitoring and providing resources to enable external consultants evaluation have been rather limited on the quality to be used. of RIAs conducted so far. The lack of evaluation of the experiences has also limited the communi- cation concerning RIA to a broader audience. In Tanzania, for instance, a RIA was conducted on two labor laws in 2004, and it was discussed at Indeed, broadening the constituency in favor of one or more workshops with stakeholders RIA, by encouraging business and other stakehold- involved in the reform process. The RIAs have ers to see its benefits and argue in favor of it can not been widely publicized; they appear to have also be very important to ensuring the viability and been qualitative and general, oriented to the con- success of RIA in the medium and longer terms. sultation process, and they did not contain detailed information on costs or savings of the proposed reforms, did not specify the institu- Lack of influence of RIA outcomes on tional consequences of the reform and did not regulatory decisions indicate implications for the operation of labor markets. These features have contributed to Regulatory decision making often fails to reflect undermine the credibility of the tool. the conclusions of RIA. To some extent, this is because politicians retain the power and respon- sibility to make regulatory choices and RIA is Inconsistent implementation only one source of information to decision-makers. For example, in some cases RIA results can be High-level support for RIA can often weaken, or ignored because of strong lobbying from interest even be withdrawn, early in the implementation groups that can allow them to prevail over broader process. This is often the result of a combination community interests. 40 Meeting this challenge involves designing the publicly releasing the opinions of the RIA unit RIA process so that it and its results are as trans- regarding each RIA, publishing final RIA, requir- parent as possible, and regulatory decision mak- ing ministers to personally sign-off and endorse ing is made as accountable as possible. These the final RIA and requiring that regulations transparency mechanisms can include providing agreed without support from a RIA be publicly a high level of independence to the RIA unit, reviewed after their implementation. 41 CONCLUSIONS Integrating RIA into the decision-making process efficient RIA processes. International benchmark is one of the tools available to countries to improve has paved the way for setting up high standards their regulatory decisions. RIA offers decision of what is considered RIA best practice. In the makers a range of options and a systematic analy- OECD context, many countries find challeng- sis of the benefits of each one, while preventing ing moving towards those standards, as RIA regulations that are not well designed from being requires a high amount of political support, approved. Many emerging and developing coun- changes in the administrative culture and ensur- tries are confronted with the need to make wiser ing that the policy-making process takes into decisions in order to avoid regulatory failures, account RIA proposals. attract more investment, and stimulate entrepre- neurship by reducing costs to businesses. These goals are even more difficult to achieve in countries with low capacities, as well as limited Regulatory reforms are currently under way in resources and skills. However, emerging and devel- many emerging and developing countries. In those oping countries still can make use of such a tool if that have already made an effort to tackle the stock it is conceived in a less rigid way that very sophis- of existing regulations, for instance through licens- ticated RIA systems of OECD countries and it is ing reform, RIA is an appropriate tool to deal with built over existing mechanisms that can be the flow of new regulations. It might also useful to improved over time, such as consultation with reinforce capacities to strengthen regulatory man- stakeholders or existing legal quality controls. This agement systems. RIA Light approach is pertinent for a number of countries that want to move from discretionary to Establishing a RIA system is, however, a chal- more evidence based decision making in a gradu- lenging task. International experience shows that ally way and without compromising sustainability many countries still struggle for effective and over time. 42 As this paper has shown, a RIA Light system is capacities to deal with the application of this tool. comprised of five building blocks that are mutu- RIA Light will highly depend on choosing a right ally supportive. They correspond to those indis- institution to take over these responsibilities. pensable elements of good practice that any RIA system requires to be operational, effective and The selection of clear criteria for RIA has to be efficient. RIA Light differs from OECD best measured against country’s expectations. A mod- practice in reducing the scope and limiting the est approach can be more suitable than using RIA methodological approach of the impact analysis, for any single piece of regulation. Selecting a pilot in acknowledging more limited capacities for based on a real case brings the opportunity to test data collection, and in advocating pilot projects the way such a system could work. as a basis for further improvement and maturity of the RIA system. Application of RIA Light systems will start in the coming months in several emerging and For RIA Light to be implemented, emerging and developing countries with the support of IC. developing countries have to show continued Given the experience accumulated until now, it political support for integrating this tool. A care- is worth noting that moving gradually is essen- ful selection of the group of reformers in charge of tial to keeping momentum for reform and introducing RIA is also important, in particular expanding the constituencies needed to make entrusting existing bodies dealing with regulatory changes in the regulatory decision-making pro- reform with expanded responsibilities and more cess of these countries. 43 BIBLIOGRAPHY Better Regulation Executive, Impact Assessment OECD (1997), Regulatory Impact Analysis: Best Guidance, London. Practices in OECD Countries, Paris. Kirkpatric, Colin and Parker, David (2003), OECD (2002), Regulatory Policies in OECD “Regulatory Impact Analysis: Developing Its Countries: From Interventionism to Regulatory Potential for Use in Developing Countries”, Governance, Paris. Working Paper Series, Center on Regulation and Competition, University of Manchester, Work- OECD (2004), Taking Stock of Regulatory Reform: ing Paper No. 56, Manchester, July. A Multidisciplinary Synthesis, Paris. Kirkpatrick, Colin and Parker, David (ed.) OECD (2007), Regulatory Management Capaci- (2007), Regulatory Impact Assessment. Towards ties of Member States of the European Union that Better Regulation?, Edward Elgar, Cheltenham. Joined the Union on 1 May 2004, SIGMA Work- ing Paper No. 42, SIGMA, Paris. Jacobs, Scott (2006), Current Trends in Regulatory Impact Analysis: The Challenges of Mainstreaming OECD (2008), Building an Institutional Frame- into Policy-Making, Jacobs & Associates, Wash- work for Regulatory Impact Analysis. Guidance for ington. Policy-Makers, Paris. Jacobzone, S., C. Choi and C. Miguet (2007), OECD (2009), Regulatory Impact Analysis: “Indicators of Regulatory Management Systems”, A Tool for Policy Coherence, Paris. OECD Working Papers on Public Governance, 2007/4, OECD Publishing. 44 ANNEX 1: DEFINING BASIC TERMS IN REGULATORY REFORM Regulation: The diverse set of instruments by which governments set requirements on businesses and citizens. Regulations include laws; formal and informal orders and subordinate rules issued by all levels of government; and rules issued by non-governmental or self-regulatory bodies to whom governments have delegated regulatory powers. Deregulation: Elimination of regulatory requirements for which social welfare costs are judged to be higher than social welfare benefits. Regulatory relief: Cutting regulatory costs to businesses with the intent of stimulating business growth. “Regula- tory relief” initiatives do not assess the benefits of regulation, it merely focuses on cost reductions. Regulatory quality: A regulatory framework in which government agencies seek to develop and implement regulations and regulatory regimes that are efficient in both a static and dynamic sense in terms of using eco- nomic, social, and environmental resources to their greatest value; effective in terms of achieving a clear public policy purpose; transparent; and accountable for results.23 To these quality standards, this report adds flexibility, since regulatory rigidities in the face of changing context and needs are common and among the main con- tributors to regulatory failures. Regulatory reform: This refers to a wide range of measures of deregulation, regulatory relief, regulatory quality initiatives, re-regulation, and institution-building. The term is a generic reference to any change in regulatory policies, functions, procedures, instruments, or capacities. Regulatory management: Refers to the construction and exercise of a management capacity in the machinery of government to control the quality of regulatory activities. A key feature of good regulatory management is the capacity to design and manage policy mixes. Good regulatory management is not about choosing one par- ticular instrument (i.e. self-regulation), but often about managing complex mixes, where one instrument works alongside others. 1 Adapted from OECD (2004), Taking Stock of Regulatory Reform: A Multidisciplinary Synthesis, Paris. 45 Regulatory policy: This term has two distinct meanings. 1) It refers to the substantive policy content of regulation. Some reforms seek to distinguish between regulatory policy and regulatory design. For example, the Doing Busi- ness indicators and the Standard Cost Model are based on the assumption that regulatory costs can be reduced while leaving regulatory policy unchanged; 2) “Regulatory policy” is also used by the OECD as a meta-narrative for the multifaceted program of a government to improve its use of regulation. The national regulatory policy agenda aims to improve four major elements: regulatory policies, regulatory tools, regulatory development (policy) processes, and regulatory institutions. Regulatory governance: Describes the systematic implementation of government-wide policies on how govern- ments use their regulatory powers to produce quality regulation within the procedural values of the governing system (such as democratic processes). Good regulatory governance is grounded in the view that ensuring the quality of regulation is a permanent and essential role of government, not a one-off set of improvements, and that institutional capacities should be designed around a clear view of the appropriate use of regulation in society. 46 ANNEX 2: BETTER REGULATION FOR GROWTH PROGRAM The Better Regulation for Growth (BRG) Pro- The BRG Program has resulted in preparation of gram was launched in 2007 by the Dutch Minis- eight policy papers on regulatory governance try of Foreign Affairs, the UK Department for issues, covering a broad spectrum: from regula- International Development (DFID) and IC, the tory governance, links to competition policy, investment climate advisory service of the World regulatory institutions and tools to indicators for Bank Group. regulatory quality. It has also involved prepara- tion of five country case studies on regulatory The objective of the BRG is to improve the regu- capacities in selected African countries. latory and investment climate in developing countries, thereby stimulating private sector The web portal www.ifc.org/brg is part of the investment, economic growth and poverty reduc- BRG Program and contains key documents, tion. The BRG program aims to achieve this by including references extracted from a compre- developing and disseminating for the first time hensive compendium of resources on regulatory widely practical and operational guidance that management and reform and a newly developed will help developing countries design and imple- Regulatory Impact Analysis (RIA) database. ment effective regulatory reform programs. 47