Revista Internacional de Contratos Públicos
RICP

Governing By Collaborative Agreements - a Case Study On Partnerships Established Between The Federal Government And Non - profit Organizations In Brazi

Natasha Salinas 2013 nº 2

Nº 02 – Dezembro de 2013

Prefere a versão histórica em PDF? Baixe aqui (paginação original)

GOVERNING BY COLLABORATIVE AGREEMENTS: A CASE STUDY ON PARTNERSHIPS ESTABLISHED BETWEEN THE FEDERAL GOVERNMENT AND NON-PROFIT ORGANIZATIONS IN BRAZIL1

Natasha Salinas INTRODUCTION The Brazilian federal government has funded nonprofit organizations (NPOs) activities throughout its history.2 At an early stage, the funding of NPOs primarily served the purpose of sponsoring independent private endeavor. In this institutional arrangement, organizations were seen as mere recipients of governmental financial aid. As the federal bureaucracy expanded, the government began purchasing goods and services provided by both nonprofit and for-profit private institutions. For this type of government-funding, nonprofit and for-profit organizations competed against each other on equal terms. Since the early-1990s, however, government-funding of nonprofit institutions has assumed a different shape: besides awarding grants and contracting goods and services, government has delegated exclusive administrative authority to nonprofit institutions by granting them a prominent role in public policy formulation and implementation.3 In this new institutional scheme, NPOs are funded by the federal government to provide a vast array of social services and administer core aspects of federal programs. They do so through the establishment of collaborative agreements specifically designed to control government-NPO partnerships.

The legal regime applied to these collaborative agreements is characterized by a wide range of controls which aim to shape the behavior of administrators and nonprofit organizations to ensure that the public interest, not particularistic interests, prevails.

In this regard, I intend to explore to what extent the current mechanisms of control over government-NPO partnerships constitute appropriate instruments for the implementation of federal programs in Brazil. In other words, the purpose of this paper is to analyze whether the legal regime applied to government-NPO collaborative agreements contributes to - or undermines - the achievement of public policy goals and, more importantly, how and why this happens.

In this paper, I argue that government-NPO collaborative agreements are inadequate instruments for the implementation of federal programs by the government in conjunction with nonprofit organizations. This happens because collaborative agreements rely on a legal regime which provides poor guidance for administrators to analyze and approve proposals filed by nonprofit institutions with the aim of implementing federal programs. In this regard, legislation regarding collaborative agreements confers too much discretion on administrations for monitoring NPOs activities on substantive grounds. Conversely, legislation is excessively intrusive in imposing procedural obligations for the spending of public funds transferred by the government to NPOs. In other words, rather than controlling NPOs by the activities they realize, federal agencies monitor their nonprofit partners by the way they manage public funds, an approach which has proved not to foster effective policy implementation.

With these objectives in mind, in part I of this paper I will identify the purposes implied or declared by the enactors of collaborative agreement laws. In part II, I will describe the legal regime of collaborative agreements in its core aspects. In part III, I will present a critical perspective of the issue, pointing out some of the loopholes and drawbacks of this legal regime which I identified through the conduction of qualitative research.

I. PURPOSES OF GOVERNMENT-NPOS COLLABORATIVE AGREEMENTS For a better understanding of the main requirements that both the government and nonprofit institutions must meet to comply with the existing laws, I will first comment briefly on the problems or social needs that the current legislation that governs

grants, subsidies and tax exemptions, whereas government outsourcing of goods and services´ delivery takes the form of contracts. Finally, cooperation and coordination between government and NPOs in planning and implementation of social policy programs take the form of certain types of collaborative agreements. Since I am mainly concerned with the functionality of the government-NPO partnership legal forms, the second framework is more suitable for the objectives of this paper.

collaborative agreements aimed to solve when it was passed. Such a description is important in order to assess the internal consistency between the instruments and goals of public procurement laws involving nonprofit institutions.

This paper will deal with two legal regimes that apply to collaborative agreements between federal agencies and NPOs. A collaborative agreement may either take the form of a convênio or the shape of a termo de parceria. Convênio is to date, the most important type of collaborative agreement that governs collaborative initiatives between the Brazilian federal government and nonprofit institutions. Termo de parceria is a new form of agreement targeted at one specific group of NPOs which under Brazilian law are entitled to apply for the qualification of “Civil Society Organization for the Public Interest.”

Starting with the convênio, it should be noted that this type of agreement was initially designed to promote decentralization of administrative actions from the federal government to the states and municipalities. Decentralization was the major platform used in a broader reform movement that took place in 1967 in Brazil with the purpose of reorganizing the entire bureaucracy. The framers of this administrative reform movement, which culminated in the enactment of a statute entitled “Administrative Reform Act of 1967”4 advocated for decentralization of administrative authority from the federal to the local governments in order to maximize efficiency.

It should be noted that this type of delegation was originally designed only to affect government entities: collaborative agreements were to be concluded between state actors, not between the federal government and nonprofit institutions. The focus of this type of delegation was to use state or local government capacities to implement federal programs or to provide services to others on the federal government’s behalf. This information is important because the legal regime applied to public delegations of government authority was later on extended – with no major adaptations – to private delegations of administrative actions. As it will be discussed later, this lack of adaptation produces nonefficient results, since it forces nonprofit institutions to deviate their time and energy from the accomplishment of their mission to the compliance of excessively burdensome procedural rules.5 Another preliminary clarification that should be made is that the types of collaborative agreements – the convênio and the termo de parceria - that will be discussed in this paper should not be confused with a much more widespread type of contract applied for the procurement of ordinary goods and services by the federal government from the private sector (I will refer to these as “government contracts”).6 The federal government purchases from the private sector a myriad of goods and services that it is either not deemed reasonable or feasible for the government to produce and/or provide directly. The rationale underlying this contractual practice relies on the presumption that the government should only spend its time and resources in the execution of its core functions - namely planning, coordination, oversight, controlling and regulation activities.7 It is undisputed that nonprofit institutions may also engage in this second type of government contract. Unlike collaborative agreements, however, in which the government is only allowed to establish partnerships with nonprofit institutions (besides state actors)8, nonprofit institutions will have to compete on equal terms against for-profit institutions if they want to engage in ordinary government contracting.

Implications of this ordinary type of government contract will not be discussed in this paper. However, a brief explanation should be made regarding some core differences between ordinary government contracts and the collaborative agreements that constitute the object of this paper. One of the remarkable differences between collaborative agreements and regular government contracts is the method by which the government may choose partners in the former and contractors in the latter. While the federal government has more flexibility to choose its partners in collaborative agreements, it is much more constrained to select its contractors in government contracts. As a mandatory requirement to formalize the latter, the government is constrained to select contractors through bid processes whose stages and procedures are prescribed in minute details in legislation.

Another striking difference is the way private contractors are entitled to manage funds received by the government. In ordinary government contracts, the government pays contractors in return to the acquisition of goods and services. In this contractual relationship, the government retains control over the manner and method in which the work is performed, but does not trace the use of the money paid as compensation to private contractors. Therefore, any money received by private contractors in return to the services and goods they provide is considered their personal property and cannot be controlled by the government.9 On the other hand, in collaborative agreements any money transferred by the federal government to nonprofit institutions or state actors remains public in its nature and can only be managed according to stringent statutory provisions.10 Consequently, parties that engage in collaborative agreements with the federal government not only must follow directives on how their

nonprofit institutions in the U.S. by local governments, see Maureen Berner, Meredith Archer Hatch, Eileen Youens, Two Models for Nonprofit Funding Allocation: Lessons for Nonprofit Managers. 14 Journal for nonprofit management 1 (2010).

services should be rendered, but also on how they should spend the money they receive to execute those services.

Since private entities end up wielding government power in collaborative agreements, it is understandable that the federal government exercises close oversight of their private partners in order to to prevent them from advancing their own financial or partisan interests at the expense of program participants and the public. In Brazil, this control is mainly characterized by the enforcement of stringent public spending procedures on nonprofit institutions that engage in collaborative agreements with the federal government. Since convênio became regulated in the 1960s, legislation has gradually become more focused on public spending procedures to ensure that funds transferred to decentralized actors remain recoverable.11 Legislation has become more stringent on public spending control as scandals and misuse of public funds became more common. This concern with non-compliance that led to more stringent concurrent and ex post controls should be understood in a context where delivery of social services through collaborative agreements has commonly been exercised in a particularistic way, where the president and politicians exchange policy-implementation authority with local governments and nonprofit organizations in return for political support.12 Frequently, the distribution of individual benefits to state and nonprofit actors takes place irrespective to their capacity to undertake policy programs, which raises inevitable control concerns.

Instruments of control were, therefore, gradually included in legislation governing convênio to prevent fraudulent activities by those who manage public funds. As it will be shown, these instruments of control are not, however, governed by standards of goal fulfillment; they are predominantly governed by standards of compliance with procedural norms. Rather than assessing the outcomes of projects carried out by nonprofit institutions, administrators usually restrict themselves to analyzing whether these organizations have spent the resources they received from the government in a correct fashion. In this respect, burdensome financial rules are applied to nonprofit organizations, which must meet rigid accounting and record-keeping requirements – the same ones imposed on any administrator who deals with public resources within the administration.

During the 1990s, a segment of the nonprofit sector13 began to advocate for changes in the legal regime applied to collaborative agreements. It campaigned for better opportunities for NPOs whose missions were related to the promotion of the public interest. Traditionally, the nonprofit sector in Brazil is characterized as encompassing various different types of organizations with completely different structures and missions, including religious groups, political parties, hospitals, clubs, professional associations, institutions of higher learning, research institutes etc. All of these groups, with the exception of clubs, are in principle eligible for establishing convênio with the federal government as long as they pursue a non-profit mission. This wide eligibility was then attacked by a group of actors who argued that the government should stop decentralizing administrative authority to organizations which do not pursue the public interest, named by these actors as the “negative applicants.”14 According to these actors, mutual benefit organizations, which include clubs and professional associations, do not pursue the public interest. Neither do religious groups or political parties, whose interests are presumably biased. Even organizations which provide services to the entire population, such as hospitals and clubs, would not, in the view of these actors, fit into the public interest category if the services they provide to the population were priced.15 This proposal, if carried out to its ultimate consequences, would exclude the great majority of Brazilian nonprofit organizations from establishing collaborative agreements with the federal government. In addition to that, such a move would be virtually impossible, since, as mentioned, NPO government-funding has been used by politicians in exchange for political support.16 The actors which lobbied for legislative reforms then relied on a different strategy: rather than advocating for an exclusion of “negative applicants’’17 organizations from the establishment of partnerships with the government, which would implicate a reform in the legal regime applied to convênio, they proposed the creation of a new type of collaborative agreement which would only be applied to organizations qualified as Civil Society Organizations for the Public Interest (CSOPI). Since privileges could not be eliminated, these actors decided to create new ones for themselves.

These actors succeeded in influencing the enactment of Law n. 9.790/99, which created a new type of collaborative agreement – the termo de parceria – designed to be only applied to CSOPI. The termo de parceria differentiates from convênio in two major aspects. First, it prescribes more relaxed controls on NPOs regarding public spending. Second, it introduces new outcome-oriented controls on the implementation activities of NPOs. Public spending controls were relaxed because they were deemed too burdensome by the supporters of this law.18 According to them, nonprofit institutions were struggling to comply with these rules and those who succeeded were becoming too bureaucratized. On the other hand, outcome-oriented controls were introduced in order to enhance NPO’s accountability, as well as policy effectiveness derived from NPO-government partnerships.19 Before undertaking an empirical analysis on how these two types of collaborative agreements have been enforced in actual and specific contexts, I will briefly describe the instruments set by political actors20 to control the behavior of administrators, as well as nonprofit activities, who engage in collaborative agreements.

II. INSTRUMENTS OF CONTROL OF GOVERNMENT-NPO PARTNERSHIPS Nonprofit organizations are subject to a wide range of controls while dealing with the federal government. Controlling of nonprofit institutions is exercised in three different stages: 1) before a collaborative agreement is signed (“ex ante control”); 2) while implementation is in progress (“concurrent control”); 3) when implementation is concluded or subject to a periodical review (“ex post control”).

These instruments of control may vary between convênio and the instrumento de parceria. As mentioned, convênio is the most common type of agreement established between the government and nonprofit institutions because it entitles virtually all nonprofit organizations to establish contracts with federal agencies. On the other hand, instrumentos de parceria can only be established between the government and organizations possessing the status of Civil Society Organization for the Public Interest (CSOPI).

The description that follows will take into account these differences.

2.1 Instruments of ex ante control Ex ante control is regarded as an essential form of control to ensure the effectiveness of collaborative agreements. As it will be discussed in part III, The Tribunal de Contas da União (TCU), an arm of Congress with responsibilities equivalent to the Government Accountability Office in the U.S., has consistently advised executive federal agencies to improve their instruments of ex ante control. According to the TCU, ex ante controls are much less imperfect mechanisms for detecting noncompliance than ex post controls.21 This is so because concurrent and ex post controls, as it will be discussed, focus exclusively on spending rules, whose noncompliance is extremely hard to detect. Moreover, the TCU reached the conclusion, after auditing various nonprofit organizations which are funded by the federal government, that the effectiveness of concurrent and ex post forms of control is entirely dependent on how agencies perform ex ante control. In other words, if an agency exercises ex ante control properly, the likeliness that it will perform concurrent and ex post controls effectively will increase substantially.22 That said, in the following subsections, two instruments of ex ante control will be analyzed – criteria for selecting proposals and proposal drafting requirements.

2.1.1 Criteria for selecting proposals I mentioned in the introduction that collaborative agreements established between the federal government and nonprofit organizations are considered means of achieving policy goals. In this way, a collaborative agreement must govern projects and actions undertaken by nonprofit institutions which are deemed appropriate to achieve policy objectives set by the government.

In this respect, the allocation of public resources aimed at funding the activities of nonprofit institutions has to meet programmatic criteria, which means that any allocation of funds must be aimed at serving the purposes of federal government programs.23 It should be noted however, that federal programs usually are set in budgetary laws in very broad terms. This program framework does not provide the specifics for the implementation of projects and therefore do not provide meaningful standards for the selection of nonprofit partners. The agencies have therefore absolute discretion to set the substantive standards which will govern bid procedures for the selection of nonprofit partners.

Besides substantive discretion, federal agencies were, up until January of 2012, also free to establish the methods and procedures for the selection of nonprofit partners. Although legislation governing convênio and termo de parceria determined the procedures for the selection of nonprofit partners, these norms were entirely optional until the passage of a circular norm which came into force in January of 2012.24 This non-mandatory character of bid processes raised criticisms on various grounds. Scholars argued that there was no reason to maintain that agencies should have wide discretion to establish the methods for selecting their collaborative agreement partners, whereas they had no equivalent autonomy when establishing regular government contracts.25 Other legal scholars, on the other hand, feared that a mandatory bid process would unduly bureaucratize the system.26 But it was the precise fact, as shown in Part III, that federal agencies had systematically abused of their discretionary power in selecting nonprofit institutions which were not acting on behalf of the public interest that led bid procedures to become bounding to the entire federal administration.

Legislation stipulates that federal agencies running bid processes must provide a description of the program whose implementation will be delegated to nonprofit institutions and must set objective criteria for the selection of projects. Additionally, nonprofit institutions will only be entitled to compete if they prove to have at least three years of experience in the area to be promoted by the government funding.27 In addition to these broad mandates concerning the mentioned bid process, legislation imposes an additional substantive constraint for its establishment between federal agencies and NPOs. Federal agencies cannot contract NPOs whose directors (or their immediate relatives) are also members of the legislative power, the Ministério Público (equivalent to the Attorney General Office in the U.S.) or heads of any agency of all levels of government.28 This measure intends to prevent nepotism in federal contracting, a fairly common practice in Brazilian public administration.

These ex ante forms of control apply not only to convênios, but also to termos de parceria. The main difference, however, between these two forms of collaborative agreement is that only for the latter civil society can participate in the selective process of nonprofit partners. The bid committee for the celebration of any termo de parceria must be composed by one representative of the federal administration, one person specialized on the subject of the bid process and one representative of a public policy management council29, when existent.30 This reflects a main difference on the rationale applied to these two legal forms of nonprofit-government partnership: while the former reflects a model of fund allocation fundamentally controlled by government staff, the second is based on a model of shared control between officials and outside interest groups.31 This main difference will also prevail between the mechanisms of concurrent and ex post control of convênio and termo de parceria, since the same committee in charge selecting NPOs will also be in charge of conducting the other monitoring activities for the latter.

2.1.2 Proposal drafting requirements a) Convênio Before concluding the bid process or accepting a proposal spontaneously submitted by an organization, the agency needs to examine the adequacy and feasibility of the project to be carried out by the nonprofit institution. Presidential decrees and ministerial circulars prescribe, in essentially procedural terms, how a nonprofit must draft a work plan. Since, as mentioned, budgetary laws are broad and vague in stating the federal program´s principals and goals, nonprofit institutions do not face many constraints on how to design their programs substantively. Legislation is relatively more detailed, however, in establishing constraints on one specific component of the work plan – the budget.

Budget formulation rules certainly represent the most striking constraints on nonprofit institutions. Legislation, presidential orders and federal grant manuals anticipate the types of expenditures that must, may and must not be included in the work plan budget. Nonprofit institutions are required, for example, to fund by their own means an average equivalent to 20% of the budget expenditures32. In order to meet this requirement, they may either pay for these expenditures or render services which they would outsource otherwise.33 On the other hand, nonprofit organizations may assign up to 15% of the federal funds they receive to cost administrative expenses necessary for the implementation of the work plan. However, they cannot include in the work plan budget line items such as administration fees34, banking fees35, payments to mutual benefit organizations such as clubs36, payments to individuals who work in the federal public office37 and expenditures with marketing.38 b) Termo de Parceria For the establishment of a termo de parceria, similar proposal drafting requirements apply.39 It is true, however, that budget formulation rules applied to the termo de parceria are relatively more flexible than the ones applied to the convênio.40 Unlike convênio, termo de parceria preceded by bid processes may be established for the procurement for services provided by nonprofit institutions such as consultancy.

Another remarkable difference between drafting proposal requirements in convênio and in termo de parceria is the fact that work plans ruled by the latter must expressly indicate objective criteria for assessing nonprofit performance.41 Outcome and performance indicators must then be provided by the nonprofit organization which submits a work plan that will be ruled by a termo de parceria. It should be noted, however, that even in this case legislation gives no major guidance on how these indicators should be formulated.

Legislation governing termo de parceria is therefore more flexible in some aspects and more stringent in some others when compared to convênio rules. This can be explained by the fact that the design of legislation governing termo de parceria was influenced by the lobbying of civil society actors who advocated for more relaxation of public spending rules in collaborative agreements in return to more stringency in accountability and nonprofit performance assessment.

Regarding social control, it should be noted that the establishment of a termo de parceria is preceded by an analysis from an Evaluation Committee composed by two agency officials, one representative of the nonprofit organization and one representative of the public policy management council of the area covered by the termo de parceria, when existent. This same Committee also oversees nonprofit institutions during the implementation process and must also assess projects after their execution is complete.42 Therefore, only CSOPI which establish termo de parceria with the government may have the privilege to provide non-traditional services such as consultancy for the government. On the other hand, CSOPI are supposed – at least in theory - to be evaluated by their performance in much more sophisticated terms than other types of organizations that establish convênio with the government.

2.2 Instruments of concurrent control Once a collaborative agreement is established, the contractual parties must comply with a wide range of rules established in legislation, presidential orders and circulars with the aim of controlling the behavior of nonprofit organizations during the implementation process. This control can be essentially procedural – when focused, for example, on how nonprofit organizations should manage the project’s financial resources. It can, however, be also more substantive – when emphasizing outcomes originated from the implementation of collaborative agreements. Both of these types of control, with an emphasis to the former, are present in the legal regime governing collaborative agreements.

2.1.1 Public spending requirements a) Convênio When a collaborative agreement is established, separate bank account is created for the management of the funds that support a project, service or activity carried out by the selected NPO. Government funds are transferred to this specific bank account43 whose funds can only be used to pay for goods and services related to the object of the collaborative agreement. Nonprofit organizations must also deposit their own funds (usually limited to 20% of the whole budget of the project) on this bank account before making payments for the goods and services they need.44 All payments made by nonprofit organizations under this regulatory scheme must be easily traceable. For this reason, payments made by nonprofit organizations must in principle be credited to bank accounts of third persons.45 The nonprofit organization must also log in on a specific portal46 to register all the payments made. These payment statements must also be accompanied by supporting documentation which must be uploaded by the nonprofit organization.

Nonprofit organizations are strictly bounded by their spending budget work plan, which means that they cannot use project’s funds to pay for goods and services not previously authorized by the federal grantor. Rules applied to the convênio are very rigid in this respect. When establishing convênio with the government, nonprofit organizations can only pay for goods and services that explicitly constitute line items of the budget authorized by a federal agency.47 If an organization intends to make payments that were not included in the initial work plan authorized by the federal government, it will then have to submit an amendment proposal to the budget whose implementation will depend on previous agency approval. Even in emergency cases nonprofit organizations must not take any actions without agency approval,48 irrespective to the fact that usually government agencies take considerable time to review organizations proposals of any sort. Moreover, any proposed amendment to the budget must not modify the nature of the object of the collaborative agreement, although proposals for expansion of the object or reductions or exclusions of implementation stages may be approved.49 Non-profit organizations must also conduct simplified bid processes for payments to subcontractors whose amount is superior to eight thousand Reais.50 These bid processes must be run through the portal in which all relevant acts involving collaborative agreements are registered.51 Nonprofit organizations are in principle constrained to contract with interested suppliers which submit proposals through this portal.52 In this case, they must contract with the supplier which submits the least costly proposal. In some exceptional circumstances, nonprofit organizations are exempt from running bid process. This happens in cases where competition is neither feasible nor desirable. Regardless of the type of bid process adopted, organizations must follow the principles of “impartiality, morality and economicity” to select their goods and service suppliers.53 Nonprofit organizations are only allowed to make payments during the term of the agreement.54 Organizations may ask for an extension of this term no later than 30 days before its expiration.55 Between the end of the term of the agreement and the act by which the government formally authorizes its extension, organizations are forbidden to spend any remaining resources of the project’s bank account.

Nonprofit organizations must comply with these and various other rules to manage the funds of their projects.56 These rules have become more stringent along the years in order to prevent frauds and misuse of government funds.

b) Termo de Parceria However, the actors involved in the drafting of Law n. 9.790/99 were successful in relaxing some of the spending rules just described. Therefore, for a CSOPI which establishes a termo de parceria with federal agencies, some of the mentioned public spending requirements do not apply. Organizations which establish a termo de parceria with the government may, for example, use their federal grants to pay the salaries of its directors, employees and consultants57, something that institutions which engage in convênio cannot do. Moreover, legislation governing termo de parceria does not require that payments be necessarily credited to bank accounts of third persons or that resources deposited in the project’s bank account are invested in pre-determined ways. Organizations which establish termo de parceria have also more autonomy to run their bid processes as long as they follow the “principles of legality, impartiality, morality, publicity, economicity and efficiency”.58 2.1.2 Reporting requirements of activities in progress a) Convênio Besides meeting a wide range of public spending requirements, nonprofit organizations are constrained in some circumstances by a requirement to report their work in progress before an agreement reaches its end. This happens especially when parties engage in agreements whose objects are designed to be implemented in multiple stages, such as the construction of public works. At the end of each stage of the construction, agencies will assess the work of nonprofit organizations before the latter can proceed with the implementation of the next stage. In practice, government attempts to guarantee concurrent control of nonprofit organizations actions by paying the latter in installments.59 In agreements established for the construction of public works, agencies usually delegate monitoring authority to a Brazilian Public Banking Company – Caixa Econômica Federal. In this way, nonprofit organizations will only receive the funds necessary to proceed with the implementation of the next stage of their projects after they report to Caixa Econômica Federal that they have met expectations in the preceding stage.

b) Termo de Parceria CSOPI which establish termo de parceria with the government are supposed to be subject to more stringent ongoing oversight. After a termo de parceria is established, the Evaluation Committee that had been consulted before the agreement was signed is supposed to monitor NPOs during the implementation process.

The idea that control should focus on projects outcomes rather than on how money should be spent by NPOs was key for the actors involved in the enactment of Law n. 9.790/99. This instrument of concurrent control was thus established to compensate the relaxation of public spending requirements which are deemed burdensome and intrusive by the nonprofit sector. Legislation is vague, however, on how the oversight by the Evaluation Committee should be exercised. “In truth, it is not totally clear in which moment – if in stages previously stipulated or if only at the closing of the project – the Evaluation Committee and external auditing organizations should act.”60 2.3 Instruments of ex post control When the term of a collaborative agreement reaches its end, nonprofit organizations have to meet reporting obligations in order to demonstrate they complied with procedural and substantive rules imposed on them due to the establishment of a collaborative agreement. Moreover, in addition to reporting obligations, authorities also employ other monitoring tools, such as government audits, especially to monitor government-funding of nonprofit organizations. Both types of ex post control serve two main objectives: detecting non-compliance that should be corrected and promoting knowledge about institutional deficiencies that should be ameliorated.

2.3.1 Reporting requirements a) Convênio Nonprofit organizations which establish convênio with administrative agencies are required to file a report within days of the expiration of the term of the collaborative agreement.61 These declarations must be accompanied by copies of supporting documents whose original versions must be retained by the nonprofit organization for ten years after being reviewed by the government.62 Reporting manuals issued individually by government agencies and departments are however much more detailed in prescribing how nonprofit organizations should report the way they managed projects financially. In this regard, government agencies usually require nonprofit organizations to provide a list with all the payments they made for the implementation of a project, accompanied by a myriad of supporting documents, which range from banking company statements to tax receipts. They also require detailed reports concerning the bidding processes ran by nonprofit organizations for expenditures superior to eight thousand Reais.

When a nonprofit institution fails to file a report or when this report is not approved by public officers due to alleged non-compliance with substantive and/or procedural rules, agencies must inform the Tribunal de Contas da União (TCU) immediately so that it can initiate and run an investigative process63 to identify whether there was a misuse of public funds.

An organization that fails to file a report or that is subject to investigation by the TCU is prevented from establishing other collaborative agreements with the government and from receiving federal grants of any sort.64 However, a report filed by a nonprofit organization but pending of evaluation by the administration does not prevent the former from establishing other collaborative agreements.

b) Termo de Parceria Nonprofit organizations that establish termo de parceria with administrative agencies are also required to file a report which, when compared to the convênio´s requirements, is more outcomeoriented.65 Organizations qualified as CSOPI, irrespective of whether they establish collaborative agreements with federal agencies or not,66 must also produce annual reports of their activities and finances to the Department of Justice. In other words, this reporting requirement is tied to having the status of a public interest organization, which demands a higher level of accountability from organizations qualified as CSOPI. It should be noted, however, that Law n. 9.790/99 does not provide any methodology for the preparation of financial statements by the nonprofit organizations, which tend to rely on accounting regulation designed for the for-profit sector.67 If a CSOPI fails to file a report or if this report is not approved by the agency, TCU will run an investigation in exactly the same way as described for convênio.

2.3.2 Audit works of collaborative agreements When a report filed by a nonprofit organization is reviewed and approved by an agency, this information will be included in the annual report that the latter submits to the TCU.68 In this respect, it should be noted that the TCU is the body responsible for overseeing the application of resources transferred by the federal government through collaborative agreements.69 It does so by first conducting an investigation when nonprofit organizations do not comply with their duties, as previously mentioned; secondly by reviewing annual reports filed by each agency/department of the federal government – which necessarily includes evaluation statements of all the collaborative agreements established with nonprofit organizations; and thirdly by undertaking monitoring, inspection and audit works of administrative activities involving collaborative agreements.

The TCU has undertaken various regularity and operational audits involving collaborative agreements with administrative agencies and NPOs over the last decade. These audits were intensified between 2005 and 2007, when scandals involving government funding of nonprofit organizations were revealed to the public.

The TCU usually formulates recommendations for government agencies on how they should improve their regularity and finalistic controls over nonprofit organizations, and some of these suggestions end up being incorporated in legislation governing collaborative agreements. For example, recent modifications in legislation governing convênio, such as the promulgation of Decree nº 6.170/07 and Portaria nº 507/11, were actually responses to many of the problems identified in the conformity and operational audit works conducted by the TCU in that period.

2.4 Summary The table below summarizes the instruments of control described above:

CONVÊNIOS TERMOS DE PARCERIA Optional Optional Bid Process for selection of NPOs Limitations on budget expenditures Ex ante Control Proposal drafting Relaxation on budget expenditures requirements Outcome and performance indicators must be included in the proposal Stringent requirements Relaxed requirements Concurrent Control Public Spending - Payments to directors and - Payments to directors and Requirements employees are prohibited employees are allowed - Expenditures must be pre-approved - Administrators must not preby the administrator approve any individual - Payments must be traceable (usually expenditure made by check) - Payments by check are not - Bid processes must be conducted mandatory through a governmental portal - Relaxed bid processes Reporting requirements Reports should focus on the management Reports should focus on the management of financial resources of activities in progress of financial resources and the outcomes accomplished by Evaluated by administrative agencies the NPO or, through delegation, by Caixa Evaluated by administrative agencies Econômica Federal in conjunction with the Evaluation Committee composed by state and non-state actors Ex post Control Reporting requirements Reports should focus on the management Reports should focus on the of financial resources management of financial resources and the outcomes accomplished by the NPO

| ________________ Audits______________ Run by TCU randomly _____________ Run by TCU randomly |

III. COLLABORATIVE AGREEMENTS IN CONTEXT: A CRITICAL APPROACH After describing how the legal regime governing collaborative agreement is structured, I turn now to explore how it works in practice.

The legal regime just described was designed to shape the behavior of the two parties which engage in a collaborative agreement – government agencies and nonprofit organizations. It is impossible to determine, however, by simply analyzing the statutes and regulations described above, whether administrators and nonprofit organizations behave in accordance with these norms. In order to do that, I rely on primary and secondary sources which can tell something about the actors’ actual behavior. With his account I do not intend to fully assess the effectiveness of the ex ante, concurrent and ex post instruments of control mentioned above. Rather than evaluating to what extent these instruments are effective in controlling agency and nonprofit organization, I intend to discuss how and why these instruments, under specific circumstances, shape the behavior of these actors the way they do.

That said, I begin by addressing the fact that the regulatory framework of collaborative agreement gives leeway for public administrators to fail in exercising ex ante control. This conclusion was reached by the TCU in 2006, when it produced a joint report of ten audits of government-funding of nonprofit organizations which took place from 1999 to 2005.70 This joint audit-work involved 28 collaborative agreements (convênios), whose values totaled 150,698.123.9371 These agreements were covering various areas, including health care of indigenous people, granting of scholarships for masters’ and doctoral students, support of minors, construction of a hospital, maintenance of a basic health care unit, conservation of natural resources, training of representatives of municipal councils, as well as training of young people to enter into the workforce.

TCU audit officers identified deficiencies in the legal and technical review made by administrators of the work plans filed by nonprofit organizations. The most striking shortcomings identified in the 28 analyzed collaborative agreements were: (i) insufficient and superficial analysis of work plans submitted by nonprofit organizations; (ii) nonexistent analysis of the technical expertise and the operational capacity of nonprofit organizations.

Regarding the former, the TCU officers found that administrative agencies approved work plans with imprecise objects and lacking essential information required by law. Consequently, work plans did not objectively describe the goals of the project. The coordinator of this joint audit work commented:

The lack of clarity on how the goals and actions should be effectively implemented generates unrealistic expenditure schedules, with no correlation between the stages of physical implementation and the grants solicited (…) implicating, as a consequence, excessive or insufficiency release of resources to the detriment of a rational administration of the services intended to be provided to the population72 The TCU found that the work plans submitted by nonprofit organizations were poorly drafted, and lacked a clear explanation of the means and objectives of the project they were proposing. The description of implementation stages, in almost all the cases analyzed, was very superficial, with no straightforward explanation about the means to be used to achieve the proposed objectives. Moreover, none of the analyzed proposals assessed the necessities of the public beneficiary of the projects proposed by the nonprofit organizations.

TCU audit officers found that ex ante technical and legal analysis was in many cases merely pro forma, since administrators reproduced in their opinions the terms of the work plans proposed by nonprofit organizations without providing any relevant comments on the merits of the projects. Administrators generally did not proceed to an evaluation of the necessity, opportunity, convenience and/or technical feasibility of the objects of work plans proposed by nonprofit organizations.

In other audit works, TCU audit officers also found that some collaborative agreements were established with no preceding technical and legal analysis whatsoever.73 This TCU’s audit report concluded that 15 (fifteen) out of the 28 (twenty eight) analyzed agreements were established with organizations with no operational conditions or with no object-related missions for executing the proposed objects.

Similar results revealed in the audit reports of agreements established between agencies and nonprofit organizations are present in audits involving agreements established only between government actors, such as federal departments and municipalities.74 These results signal that administrators behave in similar ways when they exercise ex ante control of agreements established between public entities and agreements established between government bodies and nonprofit organizations.75 Besides the technical analysis conducted by agencies officials, NPOs which establish termo de parceria with federal agencies must use an additional instrument of control. As already mentioned, an Evaluation Committee composed by government and civil society actors is supposed to intervene in all stages of the implementation process by monitoring CSOPI’s activities on substantive grounds. Legislation, however, was designed in a way that it actually undermines the role played by this Evaluation Committee. First, recommendations issued by the Evaluation Committee are not binding. Thus, an agency may establish a termo de parceria with a CSOPI even if the Evaluation Committee recommends it not to do so.76 In practice, the Evaluation Committee has not taken its work seriously, as one can infer by the fact that Committee members usually sign recommendations for the establishment of termo de parceria inside their offices, without even meeting with each other.77 Consequently, one can easily conclude that these statutory provisions make the ex ante control of the Evaluation Committee essentially toothless.78 Regarding the analysis of the technical expertise and the operational capacity of nonprofit organizations, a TCU’s audit report found that [i]n none of the analyzed convênios the contracting bodies were concerned in evaluating whether the contractor institutions had the conditions (administrative, technical, operational, experience, qualified personnel etc.) to execute the proposed objects. This qualification, completely omitted in the technical analysis of the proposals, is almost treated as a given, regardless of the object agreed.79 In this audit force, 15 out of the 28 analyzed collaborative agreements were established with nonprofit organizations whose operational capacity to implement the objects they proposed was not demonstrated at all or “demonstrated” by the mission statements in their founding documents which were unrelated to the project covered by the agreement.80 In this respect, the disconnect between the founding documents of the nonprofit organization and the collaborative agreement may be read as a signal of lack of operational capacity. TCU read in this way when it found that a culturerelated organization was hired by the Fundação Nacional da Saúde (Agency for the National Health) to render health care services.81 However, the fact that the nonprofit organization´s mission is declared in its founding documents is not sufficient in the TCU´s view to assess organization´s operational capacity. In one of its audit task forces, TCU found that three collaborative agreements, whose values totaled 8 million Reais, were established with nonprofit organizations with only three months of legal existence and which had not received any other revenue but federal funds.82 In an empirical study involving 266 termos de parceria, a significant portion of the agreements analyzed were established with CSOPI which had less than one year of existence. Considering that several other CSOPI reported unsuccessful attempts to establish termo the parceria with the government, the evidence seems to indicate that some entities obtain their CSOPI qualifications with the sole purpose of establishing an already promised agreement,83 which contradicts the goal of legislature to encourage competition within the nonprofit sector.

In this respect, it should be noted that this audit report was publicized almost two years before the promulgation of Portaria n. 127/08, which introduced a three year minimum of experience for nonprofit organizations to become eligible to establish convênio with the federal government.84 Not only this provision, but also the circular norm as a whole, can be seen as a response to some of the institutional deficiencies detected by TCU officers in the audits of collaborative agreements between federal agencies and nonprofit organizations.

Insufficient and superficial analysis of work plans submitted by nonprofit organizations and nonexistent analysis of the technical expertise and the operational capacity of nonprofit organizations are not, however, the only shortcomings of ex ante control detected by the TCU in its audit works. Another problematic factor pointed out by TCU is the lack of transparent criteria in the selection processes of NPOs by federal agencies.

As previously mentioned, legislation governing convênio and termo de parceria confers great levels of discretion on federal agencies regarding this matter. The greatest risk associated with the lack of adequate criteria for the selection of NPOs by the government is that individual interests may take over the implementation process of social policy programs.

In practice, the great majority of federal agencies which rely on delegation of implementation authority to NPOs have neither promulgated rules nor created manuals establishing routines and procedures to be observed for the selection of proposals. They simply rely on the existing legislation which, as already observed, is notoriously broad and vague on this matter.

The broad mandates of this legislation allow agencies to establish long-term relationships with certain NPOs, restricting the opportunity to other NPOs of having a chance to collaborate with the government.85 Although long enduring relationships might be desirable and the best solution for the procurement of certain goods and services by the government,86 the fact of the matter is that the collaborative agreements’ regime is not an adequate instrument to support this type of relationship. Until very recently, collaborative agreements could only be established on an annual basis, in which an agreement could only remain in effect for one fiscal year. Therefore, if a federal agency was willing to continue working with a specific NPO, it would have to establish a new collaborative agreement, with new clauses, for every subsequent fiscal year. This annual requirement has affected the nature of collaborative agreements, which must rely on short-term implementation processes. This limitation in the duration of collaborative agreements is in many ways problematic, not only because the welfare enhancing impact of short-term actions is very limited, but also because there are actions that cannot be interrupted. This is the case when the National Health Foundation,87 an organization subordinated by the Health Care Department, contracts NPOs for the rendering of health care services to indigenous people.88 Since these services cannot be interrupted, the National Health Foundation has year after year concluded new convênios with the same NPOs. One problem identified by the TCU with this contractual renewal practice is that the National Health Foundation was proceeding with inadequate legal and technical analysis before signing a new agreement with the same NPO.89 Therefore, ex ante control, in this case, has been merely pro forma, since agreements are signed without a substantive analysis of NPO’s proposals. Moreover, TCU officers detected that NPOs were unresponsively assuming financial obligations for a prospective agreement prior to its enter into force.90 In response to these problems, the Planning, Budget and Management Department authorized the government to establish multiannual agreements with NPOs. However the multiannual agreement may be revoked since it is funded by federal resources whose spending must be annually approved by Congress and the President as a line item of the federal budget. Fearing the threat of this political veto, federal agencies have been establishing multiannual agreements only in rare circumstances, which means that the tensions between the necessity of establishing long-term relationships and the inadequate legal institutions to support it remain.

The deficiencies highlighted in the audit reports just described refer to situations in which agencies have discretion to establish collaborative agreements with nonprofit organizations. However, as already mentioned, there are situations in which budget resources are earmarked as mandatory transfers to specific nonprofit organizations. When that is the case, agencies are constrained to establish partnerships with organizations whose names are explicitly referred in line items of the federal budget.

Specific ex ante control problems may arise in this context. These problems arise from the way budget is formulated and implemented in Brazil. In Brazil, the executive branch proposes the budget. Legislators than propose amendments to the budget and then pass it back to the executive branch. After the president approves the final proposal, with the president still retains a line-item veto power over the budget, since it is up to the executive power to determine the release of budget funds. Therefore, during the budgetary year, congressmen spend a significant portion of their time seeking to convince the president to release funds which will ultimately be transferred to its constituencies, which include nonprofit organizations. In this process, the executive power, under the justification that it must meet fiscal surplus projections, authorizes the releases of resources at a very slow pace.91 Considering that releases of budget resources occur only after long negotiation processes between the executive and the legislative branches, a considerable portion of budget’s earmarked funds will only be available for transfer to nonprofit organizations at the end of the fiscal year. Since government agencies must await the release of resources to establish collaborative agreements with nonprofit organization, they have to fast track this process so they can transfer public resources to nonprofit organizations before the end of the fiscal year.

In an audit task force undertaken with the purpose of investigating failures in urban policy agreements, TCU audit officers found that 168 out of the 393 collaborative agreements established by the Municipalities Department from 2000 to 2006 were concluded within the last ten days of the fiscal year.92 Approximately 90% of the resources transferred by the Municipalities Department in the analyzed period represented revenues associated with budget earmarks. The audits´ reporter maintained that the great majority of the investigated collaborative agreements were established without adequate prior technical analysis of work plans and operational capacity of nonprofit organizations. These analyses were sped up due to the late release of resources by the president.

The mentioned TCU audit´s report also reveals that failures in the ex ante control of this type of agreement is not only caused by lack of time. It also originates from the fact that agencies tend to ignore their obligation to exercise ex ante control of projects, which are expressly authorized in the federal budget, as it is implied from the following TCU´s Audit Report statement.93 The results of these audit reports show us that administrators deviate quite dramatically from the process prescribed by legislation for the exercise of ex ante control. Administrators tend to use their discretion for selecting proposals as a free pass to behave in an unconstrained matter. This abuse of discretion is also stimulated by the tendency of the public watchdog bodies – the TCU and the CGU – not to investigate administrators when they do not exercise ex ante control of NPOs properly.94 In this respect, the TCU and the CGU have interpreted these administrative failures as mere “formal irregularities” with no major consequences for those who commit it. A TCU audit officer conceded in an audit report that “ ‘formal irregularities’ actually constitute blatant omissions or even deliberately actions undertaken to undermine the effectiveness of control in subsequent stages (…).”95 The officer then concluded that public officials should be held accountable for their shortcomings in the exercise of ex ante control.

TCU has favored a stricter enforcement of accountability rules on administrators because it perceives a straightforward cause and effect relationship between shortcomings in ex ante control and failures in concurrent and ex post control. This causal relationship was evidenced by the fact that in virtually all collaborative agreements which were audit, when administrators failed to examine the adequacy of proposals´ work and budget plan, as well as the fiscal and legal regularities of proponents, misuse of public funds committed by NPOs were also identified. Shortcomings at the ex ante control stage increase government´s risk of accepting NPO´s proposals that are either poorly-formulated or created with the sole intention of embezzle public funds. When proposals are poorly-formulated, administrators are incapable of assessing the outputs of the projects carried out by NPOs, since they are left with no adequate basis to compare what was proposed with what was achieved by the latter.96 Moreover, inadequacies and fraud committed in the implementation process are rarely detected because concurrent and ex post forms of control are reactive measures which are far more costly and far more complex than preventive measures such as ex ante monitoring.97 Therefore, negligent ex ante monitoring of administrators, coupled with insignificant or mitigated punishment, affect the process of implementation as a whole.

Even when ex ante technical and legal analysis is exercised properly by administrators, there is still a great deal of uncertainty and lack of regulatory guidance in the other stages of control. In this regard, I interviewed several NPOs which have established convênios with at least three different federal agencies and they unanimously answered that there is no uniformity across the federal administration in the procedures of control exercised by administrators. In their view, it is very difficult to adapt to very diverse institutional environments. The most inconsistent practices undertaken by agencies officials involve the analysis of work plan’s budget. I previously mentioned that NPOs are required to self-fund the equivalent to 20% of the budget expenditures. In order to meet this requirement, they may either pay for these expenditures or render services which they would outsource otherwise. Federal agencies, however, have discretion to establish the NPO’s assets that are going to be accepted as a contribution to the budget.98 While some agencies accept either payment or services, others only accept payment or require a mixture of both forms of private contribution.99 Agencies also have different interpretations on which expenditures should be authorized or prohibited to be included in the work plan’s budget. While some agencies authorize the funding of salaries of NPO’s employees, others ban it outright. Moreover, interpretation of spending rules varies not only across agencies, but also within one single government body. This happened, for example, in the implementation process of the program entitled “Points of Culture”, in which different divisions of the Culture Department in charge of exercising ex ante and ex post controls had totally different views on which activities should be funded with public resources.100 The NPOs I interviewed also faced particular difficulties to comply with the burdensome spending rules described in the previous section. They have faced particular difficulties to run bid processes to purchase goods and to hire subcontractors to execute policy programs. In this process, NPOs are left with no flexibility to purchase goods and services, which in their view has the potential to affect negatively the implementation process, since they are preventing to hire the suppliers they trust. These rules also end up ossifying the NPO’s operational and structural mechanisms. In this regard, all the NPOs I interviewed complained that for the compliance of these rules they had to spend significant time and resources which otherwise would have been spent in the accomplishment of their missions and in the implementation of policy programs. Moreover, these rules are in their perception imperfect monitoring mechanisms, since they do not prevent NPOs from embezzling funds when they want.

One could imagine that problems of this sort will no longer affect CSOPI. Indeed, one of the main purposes of Law n. 9.790/99 was to relax some of the mentioned spending rules so that government and NPOs could better focus on the substantive outcomes of collaborative agreements. The rationale underlying this new legal regime is that constraints on spending of public funds are worthless if they do not lead to the achievement of policy goals. Therefore, Law n. 9.790/99 was created to avoid interpretation disputes such as the one just described by relaxing some of the spending requirements. This law, however, did not solve this problem for two simple reasons. First, Law 9.790/99 applies only to CSOPI, leaving out various other types of organizations – including religious organizations, public benefit organizations, labor unions, political parties, non-public hospitals and educational institutions – which are also part of the nonprofit sector. Second, and more importantly, the implementation actors of termo de parceria are the same ones who deal with convênio. In practice, this overlapping function of public officials has undermined the implementation of termo de parceria. Public officials are procedurally-minded and do not have the capacity to deal with termo de parceria, which relies on substantive and outcomeoriented monitoring. In addition to that, government has not allocated resources to promote the training of these officials in order to change their traditional behavior in monitoring and controlling government-NPO collaborative agreements. But perhaps the leading factor that undermines implementation of termo de parceria is the fact that legislation is too broad and vague to provide any guidance to administrators on how to apply outcome-oriented controls. In this respect, public officials who work at the Especial Secretary of Human Rights complained of the non-existence of a “norm which details all the steps to establish a termo de parceria,”101 as, in their view, does the legislation governing convênio. As a result of that, public officials tend to avoid the establishment of termo de parceria and resort to convênio instead. Or, when they opt for termo de parceria, they tend to follow the same procedures and apply the same rules meant to be applied only to convênio.102 In this respect, it should be noted that public officials extend the application of convênio rules for the controlling of termo de parceria not only because they prefer to work with familiar procedures but also because other public bodies told them to do it. For example, the Decree n. 3.100/99, which implement and regulates some of the provisions of Law n. 9.790/99, allows that termo de parceria be established for contracting services like technical consultancy of any sort. The contracting of this service is prohibited under convênio regime, because TCU maintains that this activity should be acquired by federal agencies through regular government contracting, which is preceded by a bid process and is mostly targeted at for-profit organizations. Consequently, despite the fact that Decree n. 3.100/99 explicitly allows the procurement of consulting services through termo de parceria, this provision turned out to be ineffective because TCU expressly condemns it.103 Part of this resistance of watchdogs’ organizations to accept the application of the termo de parceria is due to the fact that its legal regime does not provide an accurate description on how outcome-oriented instruments of control should be implemented. For example, Law n. 9.790/99 established that an Evaluation Committee composed by government and civil society actors would monitor CSOPI’s actions on substantive grounds during and after the implementation of government programs. However, legislation is not clear on the binding character of the Evaluation Committee’s opinions, since agency administrators may not take them into consideration when making a final assessment of NPO´s implementation actions. Legislation is also vague on which moment the Evaluation Committee should actually act – before the establishment of the termo de parceria, in the beginning, in the middle or in the end of the implementation process.104 Making the scenario even worse, legislation also did not mobilize the resources so that the Committee members could, for example monitor CSOPI’s activities. In this respect, government lacks personnel with expertise to monitor and assess the implementation of federal programs. As a result, vagueness in statutes has ultimately turned the termo de parceria ineffective.

CONCLUSION The patterns of behavior identified in this paper cannot be used to measure compliance with collaborative agreement procedures. Indeed, the sources analyzed tell us nothing about what percentage of the total number of collaborative agreements established by the government present the problems described above. However, important conclusions emerge from this case study. First, the legal regime applied to collaborative agreements between the federal government and nonprofit institutions is excessively formalistic, since it requires that nonprofit organizations follow burdensome expenditure responsibility steps for the use of grant funds. On one hand, nonprofit organizations face particular difficulties to comply with these expenditure rules, which ultimately lead them to deviate from their missions. On another hand, administrators fail in many aspects to control NPOs on these formalistic grounds.

Second, attempts of shifting the controlling of collaborative agreements from formalistic to substantive grounds have collapsed. Inconsistencies and limitations of legislative design seem to obstruct this project of reform. Administrators not only lack capacity, but also lack incentives to behave on more substantive grounds. As a result, administrators have not taken significant action to formulate the substantive standards that the legal regime of the collaborative agreements seems incapable to provide.

Taken to the ultimate consequences, these conclusions may challenge the adequacy of governmentNPO partnerships structured around the collaborative agreement legal regime analyzed in this paper.

Este texto está publicado na Revista Internacional de Contratos Públicos, disponível no endereço : HTTP://www.direitodoestado.com.br/ijpc. Se deseja colaborar com a Revista, envie seu trabalho para o Comitê de Avaliação com resumo em inglês : ijpc@direitodoestado.com.br.

Cet article est publié par la Revue Internationale de Contrats Publics, disponible sur le site: HTTP://www.direitodoestado.com.br/ijpc. Si vous voudriez collaborer avec la Revue, veuillez envoyer votre travail au Comité de Rédaction avec un résumé en français: ijpc@direitodoestado.com.br This article is published by the International Journal of Public Contracts, available at the website: HTTP://www.direitodoestado.com.br/ijpc. If you would like to collaborate with the Journal, please send your work to the Editorial Board with an abstract in English: ijpc@direitodoestado.com.br

Notas

  1. Earlier drafts of this paper were discussed at the Administrative Law Seminar conducted by Professor Jerry Mashaw at Yale Law School in 2011 and at the Yale Doctoral Scholarship Conference in 2012. I thank Jerry Mashaw, Jocelyn Stanley, Danieli G. Evans, Ruth A. French-Hodson, Alex C. Gesch, Alexander B. Schwab and Paul J. Slattery, Diego McCawley and Chen Zhao for their valuable comments and suggestions at that opportunity. ↩
  2. Brazil is a federal state with three different levels of government – the Union, the states and local governments. These national, state and local government units engage autonomously in relationships with nonprofit institutions for social service delivery. In this paper, however, I will only analyze the federal legal system. Legislation addressing the federal funding of nonprofits has become a model which has been replicated by state and local governments, so the account herein provided can be up to a certain extent generalized to all government-nonprofits relations in that country. ↩
  3. There are different ways of framing the relationships between the state and nonprofit organizations. A well- known framework is the one based on state and NPO´s institutional interests and preferences for policy ends and means. According to this framework, the nature of the relationship between government and NPOs may be complementary, supplementary or adversarial, or, to use a similar terminology, cooperative, confrontational, complementary or co-optative. See Dennis R.Yong, Complementary, Supplementary, or Adversarial? A Theoretical and Historical Examination of Nonprofit-Government Relations in the United States, in NONPROFITS AND GOVERNMENT: COLLABORATION & CONFLITC, 2006); Adil Najam, The Four-C´s of Third Sector-Government Relations: Cooperation, Confrontation, Complementarity, and Co-optation. 10(4) NONPROFIT MANAGEMENT AND LEADERSHIP 375 (2000). Another way of framing these relationships is based on the different types of government-funding of NPO´s activities and their underlying legal forms. For example, government contributions to the nonprofit sector take the forms, in the Brazilian legal system, of ↩
  4. Administrative Reform Act of 1967 is the popular name of Decree-Law n. 200/67. ↩
  5. One of the main criticisms made by the nonprofit sector is that the current legal regime imposes stringent duties and obligations on them. As it will be discussed later, the regulation of collaborative agreements is stringent when it obligates the nonprofit private partner to follow burdensome procedures to manage the public resources transferred by the government. For a study on the perception of NPOs about the legal regime governing collaborative agreements, see Natasha Salinas, Avaliação Legislativa no Brasil: um estudo de caso sobre sobre as Normas de Controle das Transferências Voluntárias de Recursos Públicos para Entidades do Terceiro Setor (Apr. 16, 2011, 4:30 PM), http://www.teses.usp.br/teses/disponiveis/2/2139/tde-31032009-153959/pt-br.php. ↩
  6. These government contracts would be equivalent, for example, to purchase-of-service and purchase-of-goods contracts by which U.S local governments pay another entity (for-profit, nonprofit or public) to perform a specific service or provide a specific good. For an instruction explanation on the different types of funding of ↩
  7. Hélio Beltrão, Considerações preliminares. Documento I apresentado à Comestra em 20 de outubro de 1964, in REFORMA ADMINISTRATIVA. (Departamento de Imprensa Nacional ed., 1968) at 36-37. ↩
  8. For-profit institutions cannot be involved in collaborative agreements. This does not mean that private entities with forprofit missions do not undertake responsibilities in the implementation of specific government programs. Governmentfunding of these activities, however, are formalized by other legal institutions and will not be discussed in this paper. ↩
  9. MARIA SYLVIA ZANELLA DI PIETRO. PARCERIAS NA ADMINISTRAÇÃO PÚBLICA: CONCESSÃO, PERMISSÃO, FRANQUIA, TERCEIRIZAÇÃO E OUTRAS FORMAS 251 (5th ed. 2006). ↩
  10. Id. at 251. ↩
  11. JORGE MIRANDA RIBEIRO & MARIA MOTA PIRES, CONVÊNIOS DA UNIÃO: TEMAS POLÊMICOS, DOUTRINA, JURISPRUDÊNCIAS DO TCU E PODER JUDICIÁRIO, RECOMENDAÇÕES 141 (Brasília Jurídica eds., 2005). This seems also to be a trend in the U.S., in which the government also imposes a strict regime of reporting and auditing requirements on nonprofit institutions. See. Peter Manzo, Privatizing Better: Representing Nonprofit Service Providers Contracting with the Government. 35 CLEARINGHOUSE REV. 666 (2001-02) at. 670. ↩
  12. The exchange of programs that benefit particularized groups of citizens for political support is a fairly common strategy adopted by Brazilian presidents. Barbara Gueddes demonstrates that political uses of the bureaucracy made by Brazilian presidents has undermined dramatically effective governance. Barbara Gueddes, supra note 2, at 131-180. ↩
  13. This segment was composed by (i) the Brazilian government, through the Solidarity Community Council; (ii) international funding agencies (especially the Inter-American Development Bank); (iii) a segment of the Brazilian business community; (iv) public universities; (v) workers in the nonprofit sector. Mário Alquino Alves, Uma boa festa, mas ninguém apareceu: resistência à nova qualificação legal do terceiro setor. (Apr. 22, 2007), http://integracao.fgvsp.br/ano6/01/administrando.htm. ↩
  14. Paulo Modesto, Reforma do Marco Legal do Terceiro Setor no Brasil (Apr. 20, 2007), http://www.bresserpereira.org.br/Terceiros/Autores/Modesto,Paulo/terceirosetorreforma.PDF. ↩
  15. Law n. 9.790/99, article 2nd, explicitly excluded the following types of organization from the CSOPI qualification and the possibility of engaging in termo de parceria: (i) business corporations; (ii) unions and professional associations; (iii) religious organizations; (iv) political parties and political parties-related organizations; (v) mutual benefit organizations; (vi) health insurance companies; (vii) hospitals which provide non-free services; (viii) private educational institutions; (ix) social organizations; (x) cooperative organizations; (xi) public foundations and organizations created by public foundations; (xii) credit organizations. ↩
  16. Law n. 9.790/99 could only be passed after a compromise in Congress in which the privileges of traditional nonprofit organizations were preserved. Elizabete Ferrarezi, O Novo Marco Legal do Terceiro Setor no Brazil (Mar. 16, 2011, 5:52 PM), http://www.lasociedadcivil.org/uploads/ciberteca/ferrarezi.pdf. ↩
  17. Paulo Modesto, supra note 14. ↩
  18. Purpose Declaration n. 20, 23rd July 1998, Diário da Câmara dos Deputados, Oct. 14. 1998, at 23,000. ↩
  19. Id, at 23,000; Elisabete Ferrarezi, supra note 16. ↩
  20. The most important political actor in instruments-setting of control is the president. Instruments of control are mainly regulated by presidential decrees and Portarias, which are regulations issued by the Treasury and Planning departments. This public procurement regime is constantly characterized by regulatory inflation, instability, contradiction between regulation and constitutional rules and lack of legitimacy. ↩
  21. TCU’s Decision n. 641/2007. En banc. Decision’s reporter: Benjamin Zymler. DOU, Apr. 23, 2007. ↩
  22. BRASIL. TRIBUNAL DE CONTAS DA UNIÃO. RELATOR: MINISTRO UBIRATAN AGUIAR. RELATÓRIO E PARECERES PRÉVIOS ↩
  23. The regulatory framework of a given government program is solely characterized by nomination of the government agencies involved in its implementation, a list of projects, very broadly described, which shall be undertaken to achieve the program goals, the amount of funds which can be allocated to each one of these projects and a list of the organizations entitled to execute these projects in partnership with the federal government - which may or may not encompass nonprofit institutions. Information about all the Brazilian federal programs is available at: <http://sidornet.planejamento.gov.br/docs/cadacao/>. ↩
  24. See Portaria Interministerial nº 507/11. ↩
  25. See Roberto Domingos Todero, Dos Convênios da Administração Pública 49-50 (2006) (unpublished Master’s thesis defended at Faculdade de Direito, Pontifícia Universidade Católica do Rio Grande do Sul); José Anacleto Abduch Santos, Licitação e Terceiro Setor, in TERCEIRO SETOR, EMPRESAS E ESTADO: NOVAS FRONTEIRAS ENTRE O PÚBLICO E O PRIVADO 281-308 (Fórum ed., 2007); Maria Nazaré Lins Barbosa, A Experiência dos Termos de Parceria entre o Poder Público e as Organizações da Sociedade Civil de Interesse Público (OSCIPS), in PARCERIAS PÚBLICO-PRIVADAS 487-523 (Malheiros ed., 2005). ↩
  26. Maria Nazaré Lins Barbosa, supra note 25 at 507. ↩
  27. This minimum period of experience was another major change in legislation, which came into force as of July 1st, 2008, as a response to scandals involving partnerships between the federal government and unqualified nonprofit institutions, as it will be discussed in part III. ↩
  28. See Article 10. II, Portaria n. 507/11. ↩
  29. Public policy management councils are forums compounded by civil society representatives and government staff which have played since the promulgation of the Constitution in 1988 an auxiliary role in the formulation and monitoring of social policy programs in Brazil. Over than 28.000 management councils, organized at all levels of government, have been established in important areas such as health policy, education, the environment, and other matters. See. Vera S. P. Coelho, Barbara Pozzoni, Mariana Cifuentes, Participation and Public Policies in Brazil, in THE DELIBERATIVE DEMOCRACY HANDBOOK at 175. ↩
  30. Article 30, Statute nº 9.790/99. ↩
  31. Maureen Berner et al., supra note 6, named these two models, respectively, as Internal Control Model and as External/Shared Control Model. In the first model, all the. ↩
  32. The percentage amount to which nonprofit organizations must contribute varies from 3% to 40%, depending on the HDI of the municipality where the project will be implemented. ↩
  33. Article 24, Portaria nº 507/11. ↩
  34. Article 52, I, Portaria nº 507/11. ↩
  35. Article 52, VII, Portaria nº 507/11. ↩
  36. Article 52, VIII, Portaria nº 507/11. ↩
  37. Article 52, II, Portaria nº 507/11. ↩
  38. Article 52, IX, Portaria nº 507/11. ↩
  39. Requirements such as that 20% of the budget be funded by means provided by nonprofit organizations, or prohibition of use of federal funds to pay individuals who work in the federal public office also apply to projects rules by a termo de parceria. ↩
  40. See article 23, Decree nº 3.100/99. Expenditure rules are also more flexible concerning line items of the work plan budget. In this way, expenditures such as administration and banking fees, which cannot be funded when a convênio is established, can be funded in the case of a termo de parceria. ↩
  41. Law nº 9.790/99, article 10, III. ↩
  42. Law nº 9.790/99, article 11, § 1, and Decree nº 3.100/99. ↩
  43. The bank in which the funds will be deposited must be owned or controlled by the federal government, pursuant to article 10, Decree nº 6.170/07. Usually resources used to fund the object of collaborative agreements are deposited at the Banco do Brasil (Bank of Brazil). ↩
  44. Decree nº 6.170/07, Article 7. ↩
  45. Decree nº 6.170/07, Article 10, § 1. Exceptionally, nonprofit organizations may make payments to third persons who do not own a bank account but in this case they must do so in a way that allows the banking company to recognize the identity of the recipients or they must provide this information directly to the agency. This is fairly common, for example, when a project is being implemented in underdeveloped regions of the country, or in projects dealing with indigenous people. See Decree nº 6.170/07, Article 10, § 2 and 3, II; Portaria nº 127/08, Article 50, § 4 and Article 50, § 2, II. ↩
  46. Available at https://www.convenios.gov.br/portal/. ↩
  47. Portaria nº 507/11, Article 64. ↩
  48. Portaria nº 507/1, Article 52, IV. ↩
  49. Portaria nº 507/11, Article 52, III. ↩
  50. Portaria nº 507/11, Article 58, δ. 1, I. This amount is equivalent to U$ 4,705.00. ↩
  51. Portaria nº 507/11, Article 58. ↩
  52. In order to be able to submit proposals through the portal of collaborative agreements, suppliers must be first registered with the government, which requires them to attest they comply with a certain number of regularity requirements. In cases where no suppliers submit their proposals through this method, nonprofit organizations are presumably authorized to quote directly the services and goods they need with at least three suppliers in the market. See Portaria nº 507/11, Article 57, Parágrafo único. ↩
  53. Decree nº 6.170/07, Article 11 and Portaria nº 507/11, Article 57. ↩
  54. Portaria nº 507/08, Article 52, V and VI. ↩
  55. Portaria nº 507/08, Article 50. ↩
  56. Funds not planned to be spent within a period equivalent or superior to a month must necessarily be kept in a saving account, while funds planned to be spent within less than a month must be transferred to a short-term investment fund or to an open market operation on the Brazilian government bond market. (Decree nº 6.170/07, Article 10, §4 and Portaria nº 502/11, Article 54, § 1, I and II). Gains originated from these investments must be entirely used in the implementation of the project undertaken by the nonprofit institution (Decreto nº 6170/08, Article 10 § 5 and Portaria n. 507/11, and Article 54, § 2). ↩
  57. Lei nº 9.790/99, Article 10, § 2º , V. ↩
  58. Lei nº 9.790/99, Article 4, I. ↩
  59. Portaria nº 507/11, Article 54. ↩
  60. Natália Massaco Koga, As Organizações da Sociedade Civil de Interesse Público (OSCIPS) e os Termos de Parceria: uma reflexão sobre a relação entre Estado e Sociedade Civil 134 (2004) (unpublished Master’s thesis defended at Escola de Administração de Empresas de São Paulo, Fundação Getúlio Vargas, São Paulo). ↩
  61. Portaria nº 507/11, Article 72, II. ↩
  62. Portaria nº 507/11, Article 3, § 3. ↩
  63. Normative Instruction of the TCU nº 35, promulgated on August 23rd, 2000, regulates this investigative process. ↩
  64. Portaria nº 507/11, Art. 82, § 3, I and II, ↩
  65. This report must contain: (i) a comparison between program’s objectives and program’s outcomes; (ii) a statement including all revenue and expenses over the period in which the collaborative agreement was in effect; (iii) an independent auditor’s report and opinion, applicable to the cases in which government funds are equal or superior to six hundred thousand Reais (iv) a summary report of physical and financial implementation of the program, which must be published at the Federal Official Gazette of Brazil. ↩
  66. In many situations, as it will be discussed in the next section, government has preferred to establish convênio rather than termo de parceria with CSOPI. ↩
  67. Cássia Vanessa Alves Olak, Organizações da sociedade civil de interesse público e termos de parceria: um estudo sobre o processo de avaliação de resultados dos projetos sociais desenvolvidos com recursos governamentais no Brasil 136 (unpublished master’s thesis defended at Faculdade de Economia e Administração, Universidade de São Paulo, São Paulo). ↩
  68. Portaria nº 507/11, Art. 84, I, d. ↩
  69. Federal Constitution, Article 71. ↩
  70. TCU’s Audit Report nº 2066/2006. Decision’s reporter: Marcos Bemquerer Costa. DOU, Nov. 13, 2006. The TCU chose to audit ten nonprofit organizations (spread out all over the country, having headquarters in the states of Amazonas, Bahia, Ceará, Goiás, Pernambuco, Rio de Janeiro, Rondônia and Roraima) which were selected based on the following criteria: (i) volume of government-funding; (ii) agreements under investigation by Congress; (iii) exposure of nonprofit organization’s image in news media. ↩
  71. One hundred and fifty million, six hundred and ninety eight thousand, one hundred and twenty three Reais and ninety three centavos. ↩
  72. TCU’s Audit Report n. 2066/2006, supra note 70. ↩
  73. See TCU’s Audit Report n. 2.261/2005. En banc. Decision’s reporter: Guilherme Palmeira. DOU, Jan 3rd. 2006, concerning the audit of collaborative agreements involving agricultural reform, and TCU’s Audit Report n. 1.933/2007. En banc. Decision’s reporter: Walmir Campelo, DOU, Sept. 21st, 2007, regarding audits of agreements involving digital inclusion. ↩
  74. A joint audit work analyzed 46 collaborative agreements established from 2003 to 2005, between 17 federal agencies and 21 local governments and 2 nonprofit organizations, with headquarters in 15 states. These agreements involved the transfer of R$ 15, 3 million (fifteen million and three thousand hundred reais) to municipalities and nonprofit organizations, according to TCU’s Report n. 641/2007. En banc. Decision’s reporter: Benjamin Zymler. DOU, April 23rd, 2007. ↩
  75. That happens, for example, when government agencies establish collaborative agreements with both public and private actors for the implementation of a program. See, e.g., TCU Audit´s Report n. 1.933/2007 supra note 94. This conformity audit taskforce investigated collaborative agreements established between the Science and Technology Department, The Research and Project´s Financing Agency, the Department of Planning, Budget and Management, the Department of Communications, and public and private partners for the implementation of projects related to the promotion of digital inclusion, between 2004 and 2006. This audit task force analyzed 30 agreements, involving 11 states, 27 contractors, and R$ 82,7 million in total. ↩
  76. In this respect, TCU has already recommended legislators to amend the statute establishing that “occasional rejection of a Council’s opinion should be justified by the federal agency.” TCU’s Decision n. 931/1999. En banc. Decision’s reporter: Marcos Vinícios Rodrigues Vilaça. BTCU, n. 78, 1999. ↩
  77. Valéria Maria Trezza, O termo de parceria como instrumento de relação público/privado sem fins lucrativos: o difícil equilíbrio entre flexibilidade e controle 91-107 (2007) (unpublished master’s thesis defended at Escola de Administração de Empresas de São Paulo, Fundação Getúlio Vargas, São Paulo) at 121. ↩
  78. TCU’s Decision 931/1999, supra note 76. ↩
  79. TCU’s Audit Report n. 2066/2006, supra note 70. ↩
  80. Id. ↩
  81. Id. ↩
  82. A nonprofit organization admitted in a correspondence sent in response to auditors´ audit questions that it had been created with the sole purpose to establish a partnership with the Agency for the National Health, thus revealing that it did not have autonomy and independent institutional mission. Id. ↩
  83. Valéria Trezza, supra note 77 at 199. ↩
  84. Portaria nº 127 de 29 de maio de 2008, Art. 18, VII. This provision was superceded by Portaria n. 507/11, Art. 8, δ 6. ↩
  85. Valéria Trezza’s study involving 266 termos de parceria identified recurrent and enduring relatioships between federal agencies and CSOPI. Supra note 77, at. 199. ↩
  86. This conclusion is reached by Steven Kelman in his work PROCUREMENT AND PUBLIC MANAGEMENT: THE FEAR OF DISCRETION AND THE QUALITY OF GOVERNMENT PERFORMANCE (University Press of America ed., 1990). ↩
  87. Fundação Nacional de Saúde – FUNASA as it is called in Portuguese. ↩
  88. TCU’s Decision no 2075/2007. En banc. Decision’s reporter: Raimundo Carreiro. DOU, Oct. 5, 2007. ↩
  89. Id. ↩
  90. Id. ↩
  91. For an analysis on how the president and Congress use budget formulation procedures in exchange of political support, see, e.g., David Samuels, AMBITION, FEDERALISM AND LEGISLATIVE IN BRAZIL (Cambridge University Press, 2003). ↩
  92. TCU´s Audit Report n. 347/2007. En banc. Decison’s Reporter: Ubiratan Aguiar. DOU, March 16, 2007. ↩
  93. Officers of the Municipalities Department should realize that they are not legally exempted from proceeding with technical analysis, even when the beneficiary and the object are already set (…) it is their responsibility to evaluate the opportunity and technical appropriateness of the works which will be constructed. Id. ↩
  94. Scholars who address the relationship of legislators and bureaucrats as a principal-agent problem emphasize the availability of rewards and punishments and the likelihood that improper behavior will be punished affect the incentives facing administrators. See, e.g., Matthew D. McCubbins, Roger G. Noll and Barry R. Weingast, Administrative Procedures asInstruments of Political Control. 3 J. LAW, ECON. & ORGANIZATION, 243 (1987). ↩
  95. TCU’s Decision n. 2066/2006, supra note 70. ↩
  96. Id. ↩
  97. Matthew D. McCubbins, Roger G. Noll and Barry R. Weingast, supra note 126, at 179, call the attention for the high costs of forms of control such as concurrent and ex post monitoring, which rely on a system of rewards and punishments. ↩
  98. The great majority of federal agencies seem to accept both options as there are no manual instructions stating the contrary. ↩
  99. The Special Secretary for the Human Rights and the National Fund for Education Development require both services and payment, as pointed out by one interviewed NPO. ↩
  100. This case was mentioned by one NPO whose proposal, including work plan and budget, had been approved by the competent authority. Its budget reporting was, however, later on rejected on the basis that the NPO could not have paid for certain expenses (which bizarrely had been previously approved by the Culture Department). Natasha Schmitt Caccia Salinas, supra note 5, at 217-218. ↩
  101. Valéria Trezza, supra note 77 at 100. ↩
  102. The empirical research conducted by Cássia Vanessa Alves Olak concluded that emphasis on evaluation remains focused on compliance with public spending rules rather than on outcomes derived from the implementation of the project. Cássia Vanessa Alves Olak, supra note 67 at 143-147. ↩
  103. Decision n. 1.777/2005. En banc. Decision’s reporter: Marcus Vinicios Vilaça. DOU, Nov 22, 2005. ↩
  104. Natália Massaco Koga, supra note 68, at 134. ↩