Some aspects of implementing Regulatory Impact Analysis in Brazil

In an increasingly complex society, regulatory polices emerge as an important tool in public management. Nevertheless, regulation per se is no longer enough, and the agenda for a regulatory reform is increasing. Following this context, Brazil has implemented Regulatory Impact Analysis (RIA) in its regulatory agencies. Thus, Brazilian specificities have to be considered and, in this regard, a systematic approach provides a significant contribution. This article aims to address some critical reflections about which policy-makers should ask themselves before joining the implementation of a RIA system in the Brazilian context. Through a long-term perspective, the implementation of RIA must be seen as part of a permanent change in the administrative culture, understanding that RIA should be used as a further resource in the decision-making process, rather than a final solution.


Introduction
In an increasingly complex society, regulation is an important tool in public management. Its growing use gave rise to the term "era of regulation" (Majone, 1994). In short, Majone (1994) claims that, due to the fiscal crisis of industrial societies in general, the State has found other ways to increase its influence by means of regulation. In other words, regulation emerges as an instrument to promote and guarantee State's influence, since other instruments, such as provision of public services and taxation, have reached their limits and there is no room for further growth (Müller, 2002).
The continued use of such an instrument creates the opportunity to investigate the nature of regulation, regulatory reform, regulatory management, and, more recently, an agenda to improve regulatory quality (Rodrigo, 2005). Accordingly, Kirkpatrick and Parker (2007a:2) advocates that regulation should be effective and efficient: Effective in the sense of achieving its planned goals and objectives, and efficient in terms of achieving these goals at least cost, in terms of government administrative costs and the costs imposed on the economy in terms of complying with regulation.
Damages brought by excessive regulation are not new. Starting from Ronald Regan and Margaret Thatcher, there was always a concern about the effectiveness of the regulatory system. Since then, literature notices the efforts to measure and assess the quality of this instrument (Kirkpatrick and Parker, 2007a).
Regulation itself is no longer enough. It is necessary to justify its use, demonstrating its benefits and impacts on our society. There is a need to use this instrument due to lack of options, its choice has to be reasonable and prove to be the most efficient way to achieve such Rev. Adm. Pública -Rio de Janeiro 48(2): 323-342, mar./abr. 2014 aim. In this context, regulatory reform emerges as a possibility to promote better regulation (Jacobs, 2006).
Nevertheless, the implementation and use of reform instruments need to be critically analyzed. Their adoption must be carefully adapted to each context, also considering advantages and limitations, as well as the environmental framework. Even in Europe, where Regulatory Impact Analysis (RIA) has been used for many years, the results brought by this instrument are controversial (Queiroz and Vieira, 2012). Thus, this article aims to provide some critical reflections about the implementation and assessment of RIA in Brazil, highlighting its particular features and challenges.
The next section analyzes the context of regulatory reform and the emergence of RIA, focusing on its definition and features. In the third section, RIA is contextualized within the Brazilian case. Therefore, descriptions of the regulatory situation in this country, as well as the development of regulatory performance measurement are addressed, including some prospects of a systematic approach. For answering to the question about the implementation of RIA in Brazil, the third section also presents an analysis of its specificities in the Brazilian context. Finally, the fourth section tackles the advantages and limitations of RIA, leading to the conclusion about the challenges of RIA in Brazil.

Regulatory reform and RIA
As argued by many authors (Kirkpatrick and Parker, 2007b;Jacobs, 2006), the aim of regulatory reform is not limiting State's role, but redefining State's capacity and role to meet evolving needs. Governments must find a more effective and efficient way to regulate, in the area where it is needed and shows better results. The focus of regulatory reform is not promoting a lesser government, but rather a better one (Kirkpatrick and Parker, 2007b).
The importance of regulation is undeniable. It is crucial for developing the economy of countries. When facing a weak or unsatisfactory regulatory system, investors and entrepreneurs are discouraged to invest, thus leading employment rates, tax revenues, and government budget to cope with consequences.
The Organisation for Economic Co-operation and Development (Oecd, 1997) reviews on regulatory reform have shown an empirical evidence of a strong connection between regulatory reform and better economic performance, reflecting gains in terms of productivity and economic growth. However, it is worth emphasizing that these improvements depend on a concrete and well established action program; "the more concrete and accountable the action program, the wider and more effective the reform" (Kirkpatrick and Parker, 2007a:10), with, certainly, better results.
In this regard, RIA stands out as a tool for measuring the performance of regulation on economy and society, in order to make it more efficient. The claim is: with an effective measurement system, clear identification of its options, reasons, costs, and benefits, govern-ment is better equipped to make a decision. Furthermore, RIA aims at bringing information, promoting debate, and providing the arena with clear rules to strengthen democracy and competitiveness.
In the last two decades, RIA has become a prominent tool through which "(…) governments learn how to deal effectively with increasingly complex public policy issues in an environment of competitive and open markets" (Jacobs, 2006:3). Since 1980, RIA has become a global phenomenon aimed at providing more effective and efficient governance. For instance, many nations (23 out of 30 OECD countries) and international bodies such as OECD, the World Trade Organisation (WTO), and the European Commission is using RIA or similar methods to process decision making (Jacobs, 2006).
RIA constitutes a tool used to assess the performance of regulation on economy and society, in order to make the best use of it. RIA is a methodology that examines action options based on their impacts and discussions. Then, the conclusion is communicated to the decisionmaker in charge.
Both the analysis and communication aspects are crucial. It is a flexible tool. Its objectives, design, and role in administrative processes differ among countries and even among regulatory policy areas. RIA is perhaps best understood as one "decision method" among several methods used to reach regulatory decisions (Oecd, 1997:14).
Nevertheless, it is worth emphasizing that drafting RIA could be a sophisticated and time-consuming exercise. It is, therefore, important to determine the circumstances where RIA is required. Accordingly, Oecd (2004:4) claims that the scope for RIA may be limited to the primary laws that are made by the legislative, or may also include the subordinate regulations, such as presidential decree, directives and guidelines that the executive makes in order to implement primary laws. RIA is usually required in the case of newly introduced regulations or strengthened regulations. In some countries, RIA is also required in the case of reviewing existing regulation.
Regarding RIA, the literature (European Commission, 2013;Kirkpatrick, 2007;Jacobs, 2006) addresses some key issues during its process: Functionally, RIA is seen as an important mechanism to allow further discussion and debate involving certain regulatory issues, which are followed by transparency and increased accountability. It provides a more systematic and consistent exam of the potential impacts arising from government action or non-action and makes the process more democratic, since it deals with many actors and stakeholders (Oecd, 2004).

Description
After conceptualizing and describing RIA, it is time to dive into an actual case. The analysis of RIA should not remain at an abstract and general level if we want to apply it to a concrete circumstance. After learning about the instrument, there is a need to understand the context and study the specificities of certain environments.
However, before analyzing the implementation of RIA in Brazil, we must have a brief overview of the current situation of the regulatory system in this country. Regarding its quality, Kaufmann and partners (2009), in a study for the World Bank, developed an index that relates regulation power to governance. This idea could be used as a starting point to examine the conditions and compare the regulatory system with that of other countries. In this research, six dimensions of governance are addressed, i.e. voice and accountability, policy stability, government effectiveness, rule of law, control of corruption and regulatory quality, and 212 countries are ranked according to the performance of their governance system. Data reflect the views on governance in the public and private sectors, as well as expert non-governmental organizations (NGOs) and thousands of citizens and survey respondents worldwide.
However, as far as this article is concerned, it is helpful to focus on one of these dimensions, namely, regulatory quality. According to Kaufmann and partners (2009:7), this dimension may be conceptualized as following: "(…) Regulatory Quality (RQ) -capturing perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development". Figure 1 shows the world rank of regulatory quality. Higher values indicate better governance ratings. Most of the developed countries and some developing countries (such as Chile) have a high regulatory quality, whereas Brazil occupies a middle position, at the same level of countries such as Mexico, Peru, Colombia, South Africa, Namibia, Saudi Arabia, and Turkey.
Analyzing regulatory quality, more specifically in the Latin American context (figure 2), Brazil occupies the 8 th position among 17 countries. Nonetheless, considering the economic and political leadership of Brazil in the region, this result seems to be below expectations. The Brazilian regulatory quality seems even more deficient when we consider that, in 2008, according to the International Monetary Fund (IMF), this country was among the biggest economic powers in the world (taking into account the gross domestic product (GPD), Brazil occupied the 11 th position). Comparing the regulatory situation to that of its neighbors, Brazil is at the level of countries that occupy the 96 th (Panama), 94 th (El Salvador), and 90 th (Costa Rica) positions in the IMF ranking. Considering per capita income among the countries with a better regulatory quality, Brazil is not above only Chile, Costa Rica, and Mexico. Moreover, considering other indexes, such as Human Poverty and Human Development (reported by United Nations (UN)), Brazil also stays above most of the countries. According to this rationale, the regulatory situation in Brazil does not correspond to its economic importance and power, thus, it seems an urgent need to pursue different approaches towards a new regulatory system. Figure 3 shows the evolution of the percentile rank on regulatory quality. There has been no clear pattern on the indicator over the previous years. Despite the decrease in the first three years (2003,2004,2005), the last 3 three years (2006,2007,2008) show an increase in regulatory quality, especially in 2007 and 2008. Nevertheless, the rate still did not reach the level of 2003, and there is much room for improvement.

Regulatory quality (2008) -evolution in Brazil
Source: Kaufmann and partners (2009). According to this description, there is a need to point out what has been done so far in order to develop the Brazilian regulatory system. As mentioned above, a regulatory reform seems to be an option; thus, this article focuses on the role played by RIA as a tool of this improvement process.
As Oecd (2008:319) notices, "Brazil does not use RIA systematically, but in the framework of the Programme for the Strengthening of the Institutional Capacity for Regulatory Management (PRO-REG) it intends to introduce a RIA system". Some authors also identify some experiences of the Brazilian government in this area. Peci (2009) identified some pioneering experiences in the National Civil Aviation Agency (Anac).
Besides, with the creation of the PRO-REG, in 2007, the government clearly shows its commitment to a better regulatory system. It may be seen as the main agent in the process to implement RIA, establishing general rules (Salgado, 2009).
Its purpose is to help improve the regulatory system and co-ordination among the institutions that participate in the regulatory process. The Programme aims at introducing new mechanisms for accountability, participation and monitoring by civil society and at strengthening the quality of market regulation (Oecd, 2008:39).

PRO-REG is structured into four pillars:
Thus, especially regarding the third and fourth topics, a major directive currently applied by the Brazilian government is the implementation of RIA. The Ministry of Chief of Staff is pushing regulatory agencies towards a broader adaptation and use of this tool. Regarding the implementation of RIA in Brazil, whose planning report is available at www.regulacao. gov.br, as PRO-REG shows, a major resource currently used by the Ministry of Planning focu-ses on capacity building. In this sense, it has often held courses and seminars addressing this topic, in order to introduce impact analysis to the regulatory agencies.
Most courses have been taught by well-known academicians and international professionals, not involved with the Brazilian framework. Seven courses were offered in 2010, five of them held by international professors or organizations. Specifically in RIA, two courses were offered and both by international experts. In 2013, the situation had a small change, most courses are still taught by international specialists, but there were one or two seminars on recent RIA experiences within Brazilian regulatory agencies.
Parallel to the training process, a group, consisting of representatives from regulatory agencies, Ministry of Planning, Ministry of Finance, and the Republic Presidency was formed to follow up the process of implementing RIA in Brazil. This group is discussing the draft proposal of legislative act for implementing RIA, which will subsequently be submitted to the agencies and ministries. Once RIA is implemented, it is expected that the group acts as a body for ensuring quality, according to the international best practices recommended by OECD (Brasil, 2013).
Furthermore, there are two agencies leading a pilot project in RIA. In each of these agencies, a working group was established to implement and analyze the features of this tool, how to involve the various regulation sectors, information gaps, public consultation forms and timing, etc. The Brazilian government signed a cooperation agreement with the United Kingdom to support the work on the implementation of RIA in Brazil, especially regarding the exchange of knowledge and ideas, as well as the measurement of costs and the benefits of regulation.
Recent studies (Salgado, 2009) identify some measurement activities that are not exactly taken by RIA, but include many of its phases. Many Brazilian agencies are aware of requirements by RIA and implement some of them. For instance, they keep a consolidated database with information about the stakeholders and the area of regulation; they also apply some transparency tools and are involved in a continuous debate with the target group through public consultation. They elaborate a regulatory agenda containing the aims and goals of regulation.
Some agencies are already involved in the process of RIA, implementing some of its procedures and principles. The National Health Surveillance Agency (Anvisa) has shown a great development in this area, including the implementation of a program that aims at incorporating RIA in its regulatory process, especially due to its high normative demand. Over its 13 years, Anvisa has published 2,400 normative acts and 1,200 public consultations (Silva, 2012). Recently, it published a guideline (Orientação de Serviço (OS) n. 2, enacted on July 27, 2009), which identifies RIA as an instrument to be used in the decision-making process. This internal guideline provides for a gradual implementation of procedures of RIA in policies. The guideline defines the unit in charge of coordination, promotion, guidance, and implementation of RIA. According to Ramalho (2009), RIA has the support of the head office of Anvisa and also a great support of PRO-REG. With this step, the agency shows an apparent recognition of the importance of this tool for the decision-making process.
Another important action has been taken by the Federal Administration: the adoption of consensual regulation mechanisms, as an alternative to traditional regulation. For instance, the Federal Administration is searching for "sectorial agreements" between the productive sector and government, in order to implement the National Policy on Solid Waste (Silva, 2012).
In spite of these experiences, Peci (2009) presents some suggestions to increase and strengthen the implementation of RIA in the Brazilian government. These results emerged from a research among the Brazilian government and agencies over 2008. The author suggests the Decree 4,176/2002 as a starting point for adopting the procedures of RIA.
According to Peci (2009), this decree provides the guidelines needed to draft normative acts, pointing out the responsible offices in the measurement process. However, it has no specification about the way how to measure the outcome or how to assess it, and RIA itself is not explicitly mentioned. Accordingly, Oecd recommends a new article in this decree, which explicitly points out the mandatory use of RIA, making sure that there will be enough time to prepare government before this new regulation enters into force.

Specificities and systematic approach
Although Peci (2009) identified a good acceptance of the RIA among the Brazilian regulatory agencies, there are many challenges to overcome. This section will address some of them, in order to deal in a better way with the implementation process, within the Brazilian context.
In Brazil, since the agencies work independently, there are various levels of development, expertise, resources, and revenues. There is no common language or strategy, and as far as RIA is concerned, each agency may adopt its procedures and requirements of its own. According to the Brazilian legislation, regulatory agencies are classified as special autarchies with independent or autonomous administration/decision-making. These differences are an additional challenge that needs to be acknowledged in order to implement RIA.
The establishment of a common language and procedure can facilitate the implementation and recognition of RIA among society and the various economic and social actors. RIA should be understood as a government program rather than the program of a specific agency, as it has done so far. It limits its scope and weakens its acceptance. A program that drives agencies to the same direction, with the same understanding, is very important, and this leading role could be played by the new institution PRO-REG, as Oecd has already pointed out.
OECD experience shows that RIA, to be effective, has to be in the hands of a body responsible for quality control and is able to challenge the use of RIA in other government institutions. In the Brazilian case, this task is closely linked to the creation of an oversight body for regulatory quality. This unit could be located in Casa Civil, since that institution has the political support and plays a key role for co-ordinating government policies. The oversight body should have as one of its main functions to revise and support the use of RIA across the whole administration. Minis-tries and agencies should be able to undertake RIA as early as possible in the decision-making process. This would imply training the responsible staff so they can successfully accomplish their task (Oecd, 2008:319).
Furthermore, it is worth knowing that there is no single generic RIA model used internationally; RIAs tend to include as a main structured plan: a clear identification of objectives, structured consultation of stakeholders, detailed examination of impacts; and consideration on the use of regulation alternatives. For implementing the RIA Programme successfully, Jacobs (2006) and Oecd (1997:22) suggest some conditions, which may reduce some problems: However, in the Brazilian case, some of these conditions stand out. First, regarding the institutional structure (point 4), we have to recognize the special situation of the institutional power and government capacity. Accordingly, as RIA in Mexico (Rodrigo, 2005) points out, the regulation process and its evaluation need to be aware of the probability of enforcing compliance and non-compliance. The enforcement power and institutional structures need to "be backed by sufficient budgetary and administrative resources to ensure the effective implementation of the regulatory process" (Rodrigo, 2005:14). Second, regarding the expertise and qualification of civil services (point 7), as Peci (2009) argues, there are some specialties in the Brazilian context that need to be taken into account. For instance, in the Brazilian context, the stronger and more specialized technocratic bureaucracy is within the regulatory agencies instead of the ministry, as in the case of the United Kingdom. Ramos (2005) also claims that, in fact, in the current praxis, the ministry is weaker, it only ratifies the policies made by the agencies. Due to this fact, the Brazilian agencies play a big role in shaping and interpreting laws and regulations. Taking this into account, ministries, and not only regulatory agencies, need to be considered, not to mention the civil society, which plays a very significant role and also needs to join the capacity building process. This item will be further addressed.
Considering this whole context, some system could also be used to interpret features and bring additional viewpoints for implementing RIA. If we portray regulatory polices as a social system, with autopoietic features, it becomes clear that new information or incentive for using a new tool, such as RIA, cannot be imposed from the outside into the system. Here, two main actors may be distinguished, (i) incentives for using RIA within sectoral agencies; and (ii) incentives for acknowledging RIA within civil society.
In the first subsystem, the government recognized the need to include agencies in its own internal changes, working together to establish new rules and processes. The organization of a working group, consisting of regulatory agencies and the central government, also includes other stakeholders, who will deal with the creation of new procedures of RIA. The actual approach recognizes the need to understand the system's rationale, when introducing a new tool, as it invites different stakeholders to the stage.
Regarding the second actor and the involvement of civil society in this process, there is a need for much more inclusion. The performance measurement process requires the participation and inputs of all stakeholders involved, thus, civil society must also be aware of the process and it has to be considered in the implementation of RIA. Although the Brazilian Institute for Consumer Defense (Idec) participates in the working group that represents civil society, its role and contributions are still unclear. Moreover, some other private actors could constitute a bigger group in this process, especially regulated companies, which are among the most affected groups. Their participation in RIA is crucial for gathering information and for the success of the tool. Without a proper inclusion of them, RIA compliance and acknowledgement by these groups gets weaker.
From an autopoietic view, the regulatory agency, civil society, and all stakeholders are seen as a learning system, having a set of skills and competences. This system must involve a continuous learning and information exchange process. Knowledge is at the core and only a constant interaction between government/regulatory agencies and government/civil society will allow a learning process and a better functioning of the system. By creating schemes of shared information, information flow, effective conversation and interaction, and team learning, the capabilities of knowledge management, system understanding, and the implementation of RIA could be maximized (Mele, Pels, and Polese, 2010). So far, few learning forums or knowledge management tools are presented in the RIA Programme in Brazil, and, officially, the information exchange process is also not being properly used, in order to allow continuous interaction and learning.
When focusing on communication, according to a systematic approach, as pointed out before, the key point is that the establishment of successful public policies also depends heavily on an effective communication and not only on people. In Brazil, the formal rule and procedures that establishes RIA as an official tool need much more emphasis than what has been done so far. It is crucial for the success of this tool to have a very good procedure that determines the ways how this tool should be used.
In Brazil, the role played by PRO-REG focuses mainly on capacity building courses. However, without the rules and the establishment of certain procedures to apply the new knowledge and adapt the organization, all these courses could be compromised. So far, no specific internal procedure was established, no central rule or recommendation was adopted. The agents have been qualified in performance measurement, but no specific guideline was defined. Many courses have been offered in the last years and some studies (Salgado, 2009) point that very little was done for the establishment of a central rule.
With only courses, members are not able to implement RIA, nor change and impose a role culture of cost-benefit analysis and performance measurement without changing the rules. Specific directives need to be done with the heads of each agency in order to establish procedures and norms that could be implemented by the agents.
Regarding value creation, the main insight from the theory of systems focuses on interconnection between some specific factors. Therefore, in the regulatory framework various actors, features, and organizations are contributing to the establishment of values and goals within their regulatory context. Here, there are two main points: (i) in a general sense, there is no established formula to implement performance measurement worldwide and each country, each framework, has its own goal and approach to adapt and use performance management of regulatory policies; (ii) second, despite the need to set a national body of standards and coordination, each regulatory agency is a subsystem embedded in a specific environment with specific stakeholders, goals, and, thus, specific values.
In Brazil, most of the expertise comes from outside, experts from United Kingdom have dominated most of the capacity building courses, and their contents need to be adapted to the Brazilian context and values. Accordingly, some efforts have been made in terms of procedure adaptation. The consultation process undertaken by the United Kingdom is combined to regular visits to this country by Brazilian experts, in order to understand their system and propose changes and improvements, according to the national value and reality.
Through the second approach, the creation of value by each sectoral regulatory agency, there is still a long way to go. Many agencies are not well integrated to the process and an understanding of few agencies has dominated the stage. The group created to prepare the legal framework mostly consists of the Ministry of Finance and the Ministry of Planning, and not all regulatory agencies are directly included in the discussions.
Furthermore, only a few agencies are working on the pilot project and committed to share their achievements and methodologies with the others. Since RIA is new in Brazil, there is not evidence enough to conclude how RIA will be adopted by the other agencies.
Lately, considering all these conditions, Radaelli and De Francesco (2008) point out important aspects to analyze RIA. The main claim of the authors is highlighting the various countries, which may have different goals with regard to the implementation of this tool and the use of measurement tools based on different pressure groups. An approach aims at achieving a better political control of bureaucracy. In another approach, neo-pluralist, RIA is adop-ted to provide pressure groups with equal opportunities, ensuring that a regulatory choice will collect information from different sources and balance different values. In this model, due to the importance of bargain, regulatory choices will be less aimed at measuring the costs and benefits of regulation, less aimed at market deals, and they will look more like deliberation on major trade-offs in multiple policy sectors. Finally, the authors turn to a governance model based on rationality and self-control of agencies. The idea, here, is making efficient decisions, for instance, by using cost-benefit analysis. If we think that technical decisions are the main focus of RIA, ministries and agencies are at the top position. If the focus lies on transparency and inclusion, civil society deserves close attention. Nevertheless, to have a more detailed overview, further research is required.

Why is it important (or not) to apply RIA in Brazil?
The first question to be asked is: do we really need regulation in this area/issue? Before analyzing the importance of applying RIA to the Brazilian context, there is a need to be aware of alternative tools, which could replace traditional regulation.
Information regarding what works and under what circumstances is also (…) a key tool for breaking the "regulate first" culture that may exist in some ministries and among some regulators. It is also important that information is made available to citizens in order to increase the acceptance of alternative instruments as legitimate and effective responses to policy issues. (Hepburn, 2006:17) According to Oecd, the main alternatives to traditional regulation are the following: market-based instruments, self-regulation and co-regulation approaches, and information and education schemes.
Market-based regulatory instruments, such as fiscal measures (taxes and subsidies), act to change or modify behaviors through economic incentives. It reflects individual decisions and tends to be more flexible. However, the use of fiscal instruments faces some problems related to fiscal crisis and its unpopularity, consequently bringing political struggles and disadvantages.
Self-regulation typically "involves a group of economic agents, (…) voluntarily developing rules or codes of conduct that regulate or guide the behavior, actions and standards of its members" (Hepburn, 2006:6). It is more often used where detailed technical knowledge is likely to be important. It has advantages such as flexibility, greater enforcement power, and ability to address specific controversy issues. However, there may also be negative consequences when the regime is captured by private actors, promoting their interests instead of those of society.
The third alternative aims at changing behavior over the provision and distribution of information, through education campaigns, labeling requirements, etc. However, with this tool the degree of direct government involvement in decision-making is more limited when compared to other instruments.
In order to help choosing the right instrument, some studies and measurements have to be done. The use of RIA in the decision-making process may provide some information and cost-benefit analysis to improve regulatory decisions. It seeks to make transparent the costs and benefits of regulatory and non-regulatory options and, thus, trade-offs between various policy options are more apparent to policy-makers (Hepburn, 2006).
Considering the context of regulatory agencies in Brazil, as already mentioned, they are mainly focused on their area, instead of being driven by a general guideline aimed at the whole country.
That fragmentation of regulatory reform has so far resulted in sub-optimal outcomes, and a lack of policy coherence. Most of the debate has consequently focused on the design of regulatory agencies, with less attention placed on the need to integrate a "whole-of-government" approach to regulatory quality that could involve setting up an oversight body responsible for regulatory reform (Oecd, 2008:46).
In this context, RIA could be one more linkage and bond between the agencies, in order to build a systematic and homogeneous policy of performance measurement. As stated above, it could be used as a further instrument to gather and bind agencies, so that all of them apply the same RIA structure This strategy may strengthen the regulatory system as a whole.
Another point stressed by RIA is accountability, meaning consultation and transparency. Some of its procedures highlight the participation of all stakeholders and a greater information disclosure. In spite of the significant value of its procedure in performance measurement itself, the worry about accountability helps disseminating a new culture, namely, responsibility and answerability. In a country like Brazil, with significant corruption rates, this kind of awareness plays a big role (75 th position in the Corruption Perception index 2009, developed by Transparency International).
It opens regulatory policies to different stakeholders and stimulates social learning. A consultation process with stakeholders, other ministries, and civil society has a wide range of information on costs and benefits (Oecd, 2004); "because it increases opportunities for debate, RIA contributes to the development of a degree of social consensus that allows difficult public policy decisions to be made" (Jacobs, 2006:14). Accordingly, RIA is totally in line with the idea of decentered regulation. The participation of different actors among the core points advocated for by Julia Black (2002), in order to increase compliance. The author advocates for the idea of a regulatory system where the government does not control everything, but instead depends on various stakeholders to deal with uncertainties and asymmetric information that frequently occur in this area.
However, we can notice that public consultation is more useful when it constitutes a means for the analysis. This means that regulators must be provided with accurate information over the decision-making process, it is not only a legitimacy process, but also a matter of gathering information. As it includes consultation of a wide range of stakeholders, it also provides those potentially affected by regulations with an opportunity to highlight any unforeseen consequences not previously taken into account (Ireland, 2004;Queiroz and Vieira, 2012). Here, it is worth underlining that consultation is not new in the Brazilian regulatory agencies, this is, however, a continuous practice, already used by Anac and the Brazilian Agency of Telecommunications (Anatel).
Nevertheless, RIA still faces many problems and it has many challenges to overcome. As Jacobs (2006) highlights, it is important to keep in mind what RIA can actually accomplish and what are its challenges. In summary: -The quality of analysis continues to disappoint. In country after country, RIA does not quantify enough impacts, and does not rigorously examine alternatives. -Quantification of benefits is an enormous problem affecting the majority of RIAs in every country. Part of the reason for this seems to be a lack of investment in skills and incentives, as discussed, and in part seems to be the inattention to key constraints on good quality analysis, particularly the availability of good data at affordable cost.
-Another problem is ineffective prioritization, or targeting, of RIA resources. (…) More investment in case studies, evaluation, and analytical criteria for assessment of alternatives are needed to help regulators do a better job in this area (Jacobs, 2006:8).
Other dilemmas are also pointed out by Pollitt (2000:4), considering the implementation of performance measurement. As the author recognizes, we should ask ourselves: Does what get measured, get managed? And what gets measured, gets attention, while the rest remains forgotten. Following author's argument, we may understand that RIA does not solve all decision-making problems. Accordingly, the author states that measurement instruments tend to stress measurable aspects and overlook other qualitative aspects. From another perspective, measurement will focus on aspects according to the bureaucratic interest. Bureaucracy could show only what it wants and selects the analytical indicator and criteria. In the Brazilian case, this danger means capturing the assessment processes by the agencies. Since regulatory agencies concentrate on expertise, the implementation of RIA could be less effective if various actors do not control the measurement processes, in order to avoid the direct influence of specialized bureaucrats. The selection of indicators and criteria should be also part of a collective discussion, so that many actors can contribute and many features of the regulatory process can be explored.
According to this argument, it is asked how far measurement may be understood in the various social groups, meaning that society shares divergent views about what should be measured and what is regarded as a good result. Performance measurement can hardly be a consensus, a stable, or an objective definition. The more actors are involved in the regulatory process, the more controversial the measurement will be. Particularly considering the Brazilian case, where we can notice various levels of expertise, knowledge, and information among stakeholders (central government, sectorial regulatory agency, firms, citizens…). Accordingly, the way to convey and share information, as well as public consultation, plays an important role to reduce these problems.
Overall, it needs to be understood that performance measurement is a useful tool for the government, but it is imperfect. All its failures and problems need to be acknowledged in order to promote an efficient implementation of RIA. In a country with many regulatory problems and discrepancies, this instrument needs to be used with more parsimonious communication, information, and qualification. More than a solution to regulation problems, RIA is just a further tool.

Conclusions
Is Regulatory Impact Analysis an effective tool for regulatory quality? How could we improve its implementation in Brazil? Is it effective in the Brazilian context? The aim of this article is providing some critical analysis on these questions. More than a final and definitive answer, the objective is looking at RIA in a critical way and understanding its advantages and limitations according to the Brazilian case.
The trend supported by Oecd regarding the implementation of this instrument overshadows important aspects of RIA. The use of new instruments, as regulatory reform proposes, aims at bringing many benefits. Starting from information collection to improve the decision-making process and going through transparency and accountability, until reaching social learning, the gains are significant and required. RIA can improve the actual impacts of actions taken by the government, by helping the decision maker to set regulatory priorities. Nevertheless, it needs to be regarded as an instrument instead of a final solution. There is still a long way to go in research and adaptation. More information needs to be conveyed as discussions evolve.
In the Brazilian case, as this article analyzes, many specialties need to be considered: the independent work of each regulatory agency, for instance, reveals the problems to coordinate and gather measurement policies, leading to a heterogeneous and negative perception of the instrument by society. Moreover, as described above, the concentration of expertise in agencies rather than in ministries increases the danger that the process is captured by the agencies.
According to this rationale, the implementation of RIA in Brazil asks for more joint effort among agencies, under a central and strong body, so that the implementation process can be more powerful and acknowledged by the stakeholders. A homogeneous RIA process in all regulation sectors improves its compliance and social learning. However, as this article pointed out, the various development stages in the agencies is a challenge that the policymaker has to overcome.
Furthermore, institutional power also plays an important role, and it needs to be considered in order to increase the success of RIA. In other words, support from political institutions, sufficient budget, and administrative resources are required.
The limitations of RIA also must be taken into account at the time of its implementation. Without having in mind the problems of this instrument, the chances of failure may sharply increase. Accordingly, the policy-maker should be aware of the possible obstacles, as discussed in the last section. For instance, the definition of what will be measured and by which means reveals the need for continuous discussion among stakeholders. Detailed information about the definition of indicators and values should be, therefore, shared among them.
In a more systematic approach, the implementation of RIA in Brazil also requires much advancement.
Autopoiesis: Bring all the actors within the process. Only by bringing them it will be possible to work within their rationality and achieve compliance. In the Brazilian case, social and private actors, as well as regulated bodies, are not fully included in the process. This will decrease the quality of RIA, since impositions by an outsider lack success if there is no internal acknowledgement.
Communication among actors was put aside, in order to focus rather on capacity building and people. There is a group working on the establishment of rules and procedures, but so far no concrete standard has been released. Besides, communication with civil society gets weaker without a core communication strategy headed by the central government. Only sectoral communications and relations have been set between agencies and society.
Value creation: the complexity and environment of each system need to be taken into account in order to implement RIA. The greater influence from international experts, who are not really aware of the Brazilian context, could bring insights that cannot be applied, based on various backgrounds and values. More adjustment needs to be done in the capacity building process. Regarding structure and rule making, further steps have been developed, the international consultancy group is working closely along with Brazilian officers, in order to better adapt the international guidelines to the national reality.
In short, the implementation of a RIA system can be successful only considering the design of every institutional framework. This article addressed some questions that policymakers should ask themselves before joining the implementation of a RIA system within the Brazilian context. From a long-term perspective, the implementation of RIA must be regarded as part of a permanent change in the administrative culture, understanding that RIA should be used as a further resource in the decision-making process rather than a final solution.