IACC-Logo, back to IACC-Home

Programme Papers from the 9th IACC
past IACCs



Privacy Policy


The 9th International Anti-Corruption Conference

The Papers

The Federal Court of Audit in Brazil TITLE
Institutional Arrangements and its Role in Preventing Fraud and Abuse of Public Resources

Bruno Wilhelm Speck


In Brazil, like in other part of the world, recent scandals of corruption and misuse of public funds put in doubt the efficiency of the financial control system. Like any other political system Brazil has a complex set of external and internal control institutions supposed to prevent misuse of public funds and identify responsibilities if bad use occurred. Public opinion criticised these control mechanisms for not being able to prevent major scandals involving corruption and fraud with public money. The so called Collorgate led to the impeachment of President Collor in 1992. A second scandal involving the Budget Commission of the Congress made several Congress members loose their seats in 1994. These are only two major examples in a long array of corruption cases, misuse and waste of public money uncovered in Brazil in the 1990's. Why did the system of internal control fail to detect scandals which badly damaged the image of judiciary, legislative and executive power? Why couldn't institutions like the Federal Court of Account (Tribunal de Contas da União, further on as TCU) prevent or even alarm the political system from these cases of grand corruption? This paper intends to give some preliminary answers related to the TCU. It is based on an ongoing research on its institutional mechanisms and the analysis of its role in the Brazilian politics.

Even if social science paid little attention to the issue of financial control, institutions in this field have a long tradition. In fact all modern political systems made different efforts to control the use of public resources. They developed a more or less sophisticated system of resource control. Many of the superior institutions have a long tradition which began in the XIXth century, not to mention earlier steps in the systematisation of the administration of public funds since ancient times. One of the oldest institutions is probably the General-Rechen- Kammer, introduced in Prussian administration in 1714 by King Wilhelm I. 1 On the other hand institutions like the General Courts representing the Estates in medieval society were concerned about how monarchs used the taxes. As soon as public funds were raised on a regular base the Courts began to link revenues to special expenditures to limit executive power. At the same time they began to set up special commissions to control the Kings compliance with these rules.2 We do not intend to solve the question whether external or internal control systems came first. But financial control institutions might belong to the traditional organisations in modern political systems. During the XIXth century, most of the Republics as well as the Constitutional Monarchies introduced mechanisms like Comptrollers, Inspectors or Courts of Account to increase the rational use of public funds. The foundation of supreme financial control institutions spread rapidly in Europe, beginning with France under Napoleon (1807), Netherlands (1820) and Belgium after independence (1846), Italy during the process of nation building (1862) and England (1866).

In Brazil the debate on control on public expenditure began during the Monarchy in the XIXth century. Several Brazilian Secretaries of Finance repeatedly made an alerted about the misuse of public resources. The First Republican Constitution of Brazil (1891) finally introduced an auxiliary institution to the Congress, called Court of Accounts (Tribunal de Contas).3 This institution has similar competence until the present days. Together with the internal control system founded in 1967, the TCU is one of the most important institutions to control the use of public resources.


Social science didn't make a great effort to understand the dynamics of control systems despite the long tradition of institutions in this field. Comparing to other areas of research like the decision making process or parties and voting behavior, the logic of control institutions has not had the same attention by political scientists. Most of the studies on this subject came from law studies and administrative science. Research on the Brazilian control institutions is even narrower. Almost all the studies published about internal and external control in Brazil have been written by members of the control institutions.4

Institutions of financial control like the TCU of Brazil, the Bundesrechnungshof in Germany or the General Accounting Office of the United States do not fit easily into the model of separation of powers. In some aspects these superior control institutions became important auxiliary bodies for the exercise of control by the Legislative. They give assistance to Legislative investigations and Commissions when expertise or investigation is required. In other cases control institutions continue strongly linked to the executive and its internal control of public resources. A close contact with administrators and an attempt to improve the efficiency of the system is more important than pointing faults or defining responsibilities. A practical question in this case is the relation between the internal and external control. A third trend is to turn the external control institutions into an independent authority responsible for the good use of public resources. In this case, control institutions frequently gain power to give a final verdict on accounts. The relationship to the judiciary system turns to be an important question.

In fact all financial control institutions seem to have a hybrid character and to a different degree share elements of all three models. All of them seem to be divided between different activities of power sharing, investigation, assessment, prevention, expertise etc.


Another approach to clarify the role of control institutions is to link their control efforts to two main concerns: first the liberal concern with the limitation of power and prevention of abuse and second the managerial concern about the best use of short resources. These two categories appear in a number of attempts to analyse control systems. While the older literature emphasises primarily the first aspect, recent studies give a special attention to the second one.5

The liberal concern about the limitation of power is behind a number of classical arrangements like separation of powers and limitation by legal rules. The mechanisms of power sharing or mutual vetoes is known as a system of checks and balances. Liberal concern is about the abuse of power and the non-compliance with all these rules. Allocation of financial resources plays a major role for the exercise of political power. Therefore control of compliance with norms is essential not only to guarantee an efficient use of public funds but also to keep power holders within a near bandwidth of norms and make them responsible for their acts. This liberal concern about limitation of power is one source of legitimation for the institutions of public financial control.

The managerial concern about the best use of short resources tries to improve the system as a whole. It does not aim at the identification of power holders individually. For Karl Deutsch this kind of permanent improvement was a central element of the ongoing process of goal achievement which characterises political systems. In his cybernetic model of politics Deutsch did not identify clearly the kind of deviance or misleading behaviour he had in mind, but we see no problem to include fraud, waste and corruption as possible examples for the problems political systems have to deal with. A systematic information feedback is a core element of self regulating political systems. 6 It allows detection and correction of damaging dynamics as soon and adequate as possible. This concern about the quality of political outcomes reappeared in the recent discussion about the new public management and the evaluation of politics. It is the second source of legitimation for financial control institutions (for a systematic comparison see table 1).

The application of these two categories might clarify the discussion about the role of these major control institutions in the political system. Internal control certainly follows more closely the idea of managerial concerns about improvement of the use of resources. External control institutions seem to lack clarity about what their role actually is, should be or could be. In the following chapter I will take a closer look on the TCU in Brazil. In the final section we will return to the question of liberal and managerial concerns.


Like many other control institutions the TCU has different assignments linking it to all three constitutional powers: the legislative, the executive an the judiciary. Since its foundation one major task of the TCU was the elaboration of a previous analysis of the accounts rendered by the government to Congress. Congress makes the final judgement on the discharge of the government, but the analysis of the TCU is conclusive and has his own weight. The reproval of accounts by the TCU causes a major impact in the political arena. In June 1992 the TCU for the first time since more than five decades gave a negative previous judgement on the accounts rendered by President Collor to the Congress for the year of 1991.7

Besides this main annual task TCU also gives support to Congress and its Commissions, if any kind of investigation is required. In the 1990's Congress gained visibility because Parliamentary Commissions of Investigations played a major role in the disclosure of corruption scandals. Congress also asked for a report on unfinished public works which has been done by the TCU. While nearly all these investigations based on the assistance of the TCU, credits have been attributed to the Congress.

Besides this role of TCU close to legislative power, the institution acts as an independent authority when it makes a final verdict on the accounts given by the administrators of public resources. The judgements have major consequence for the administrators, either discharging them definitively or turning them responsible for bad use of resources. In the later case they might suffer severe punishment by the proper TCU, including financial penalty and restriction of access to public functions. When TCU gives his final verdict on the correctness of administrators, it is undeniably close to judicial power. The guarantees given to assure the independence of the members of TCU and the organisational autonomy of the institution are grounded on this role.

Since its foundation (1891) another assignment turns the TCU close to the executive power. All government expenditures depended on a previous approval by the TCU. The model of previous control had been introduced earlier in Belgium and Italy.8 Alike the Presidents right to veto laws approved by the Congress, the TCU checked the legal origin of any expenditure and had the right to veto this acts if he concluded there was any kind of improper behaviour involved. This veto was absolute when public funds were empty or the expenditure was not based in the budget law. In all other cases, the veto could be overruled by a decision of the President and later on the issue would be appreciated by Congress. In fact this arrangement followed more closely the logic of power sharing with administration than the idea of control for liberal or managerial concerns. TCU was a kind of independent branch of the executive, although it was not integrated in the hierarchical model of the administration. In practice it stood above all state secretaries but below the chief of government. In 1967 the TCU lost this power of previous control almost completely. Only some in some minor areas it continues to share administrative power, like in the case of admission and retirement of public officials which only is effective after TCU's approval. TCU also fixes the quotient for the repartition of federal funds for States and Towns.

In the following sections I will analyse a number of central institutional arrangements in order to define TCUs place in the control system of public resources in Brazil. The categories are (a) the range of control authority, (b) the question of selection of court members, (c) the autonomy of programming the control tasks, (d) techniques and criteria of control and (e) the power to turn control results relevant.


Concerning internal control systems the range and intensity of control is mainly defined by government's interest in efficiency and prevention of scandals. On the contrary, the definition of the limits of external control institutions has always been a controversial issue. Initially the TCUs authority was limited to control direct administration, but the shift from the liberal state to a state promoting industrialisation and individual welfare opened a series of new fronts of resource allocation. TCU was excluded from the control of these new fields and extended its authority step by step. One of the areas excluded from external control were powerful state enterprises which do not belong to direct federal administration. Other autonomous bodies like Pensions Funds and Institutions of Public Welfare responsible for large amounts of public resources were also excluded.

One of the main concern of TCU up to the 1980's was this limit of control authority to the very centre of state administration. The new Brazilian Constitution of 1988 settled this question based on a large definition of responsibilities. All persons or institution participating in collection, administration and expenditure of public resources are obliged to render accounts on its activities to the TCU.9 This obligation does not make any reference to the legal status of the person or institution, so public administrators are included as well as private institutions financed by public sources. A public servant is as responsible for the funds of the institution he manages as is an volunteer heading an entity receiving public funds.10 Following this criteria more than 3.200 persons and institutions are obliged to render accounts to the TCU.

Some minor questions of access to relevant information still remain unresolved. In recent years TCU did not succeed in implementing an evaluation of tax collection, because the Federal Office of Revenue refused to open his files due to his obligation to preserve confidential personal data. Another special investigation concerning the Central Bank and its policy to support financial institutions in crisis could not be completed for the same reason. But in a general way the question of control authority has been settled in a satisfactory way. The question of control areas nowadays is not a legal but a practical one. The TCU is entitled to investigate all areas where public resources come from and go to.


The form of selection of the members of superior control institutions is one of the classic issues when the independence of these institutions is at question. The existing models of selection cover a wide variety from professional recruitment, promotion by merit and experience to nomination by government or by legislative authorities. Another issue linked to the question of independence are the guarantees given to the members of control institutions. Here models go from nomination for an extended period of time or for a lifetime.

Since its foundation the members of TCU in Brazil have been selected by the President with ratification by the Senate (see table 2). Nomination ever since has been for lifetime and members of the TCU have a legal status equal to the judges of the highest courts. The strong influence of government on the selection of the members of TCU was supposed to be counterbalanced by a long time commitment of the members to their control institution. While governments come and go, the TCU with its members indicated by different Presidents was supposed to build up an own institutional identity. The political loyalty to the government who appointed a member would be substituted in the long run by a professional identity. Further on, judgements of the TCU are based on group decisions rather than on individual opinions. This, too, should prevent members form external pressures.

In fact, for several reasons these concepts did not work out in the way they have been conceived. First, real shifts in governmental power are rarer in Brazil than the regular changes of Presidents in power suggest. In fact, the same political group remained in power for long periods. During the First Republic (1891-1930) the political elite of Minas Gerais and São Paulo dominated national politics for several decades. In the Populist Period (1945-1964) Getúlio Vargas and his political inheritors governed Brazil for a decade and a half. And the following two decades were dominated by the influence of military elite on politics (1964-1984). The result was a homogeneous composition of the TCU by members appointed by the same political group. As a consequence of extended periods with the same political group staying in power this limitation of plurality and political independence of Courts of Account frequently happens in several states of the Brazilian federation.

A second fact seems to have weakened the independence of the TCU. Despite the appointment for a lifetime, many members of the institution stay in office only for a surprisingly short period. The average permanence in office of the 68 members of TCU appointed between 1893 and 1980 was 9 years. Obviously this number varies considerably if we compare different periods (see table 3). While the average permanence of members of the TCU was climbing slowly from nine up to twelve years, during the recent period of military rule it fell abruptly to half of this time. A different aggregation of these data shows that only a small group of 12 out of all 68 members of the TCU remained in the institution for 15 years of more, whereas nearly half of them (30) quitted their jobs after five years or less.11 In fact, the Brazilian Courts of Account both on national and state level served for the governments either to store politicians who in a later moment might come back into the political arena, or to reward political allies with an office that allowed an early retirement with a generous pension. This tendency can be illustrated by a law approved by the State Assembly of São Paulo in 1970, which limited explicitly the retirement with full pension to members who had stayed in office at least for a period of five years. The legislators wanted to slow down the high rate of replacement in the Court of Accounts which rendered work of the institution difficult.12

This misuse of the TCU appointments by government to reward political allies had a deeper impact on the institution than political manipulation could have. Political loyalty as a criteria for the selection of candidates nominated by government might lose its effect after a while. But the expectation of early retirement seems to inhibit a major commitment with the institution. It turned a dynamic process of control in search of more efficiency and accountability more difficult than political loyalty could have done. The new Brazilian Constitution of 1988 changed the selection criteria of the members of TCU. The executive lost partially its influence in nominating them to the Congress. The new rules allow the President to select only one third out of a total of nine members of the TCU while the other two thirds will be appointed by the two Houses of the Brazilian Congress. But presidential nomination power has further limits. Two of three candidates he indicates must be selected out of the technical staff of TCU itself. This should lead to a more professional recruitment of the members. If one of the main threads to the TCU is retirement and not political conformity, we might ask critically why only the president is suspect to make bad use of his powers and not Congress too.

On the state level most of the federal rules concerning the external control institutions have been simply copied, but the shift in formal power from executive to legislative has very little effect. The influence of the Governor on State Legislatives is very strong and the division of power is more an ideal than reality. The Governors continue to have a great influence on the nomination of the members of State Courts of Accounts. Major changes seem to occur when change of political power takes place regularly. In recent years, several cases of deep conflict between external control institutions and governments in Brazilian states and towns show that formal powers of the Courts of Account are strong enough if there is someone interested in making use of them.13 This kind of accountability rendered not to an electorate but to an independent public institution like the Courts of Account is a new experience in Brazilian democracy. Politicians and Auditors have to get acquainted with it step by step.


Like financial control institutions in other political systems, not only the independence of TCU members is guaranteed. The institution has a wide autonomy to structure its organisation, to apply resources and to program its control activities. This begins with the selection of its president who is elected by his peers.

A decisive element of the autonomy is the power of TCU to set priorities for its control activities. Contrary to other institutions of the judicial system TCU is not obliged to legitimate its selection. Police investigators or prosecutors in part react on external impulses like denunciation. But their own initiatives of investigation have to be bases on some kind of crime suspect. On the contrary, financial control institutions do not need any suspect to legitimate an investigation. This wide definition of their autonomy in self programming the execution of control tasks is a decisive characteristic of the TCU. We will come back to it. But there are several limiting factors.

The autonomy of self programming has practical limits because the TCU has to carry through a series of constitutional tasks like the previous register of all cases of admission and retirement of public servants. These are several thousand processes every year and as mentioned above TCU is obliged to appreciate them because they turn to be effective only with his previous approval. A further task which takes time and resources to the TCU is the examination and judgement of the annual accounts of administrators. More than 3.000 accounts are rendered annually to the TCU. Finally the main task of the examination of the accounts of government for approval or rejection by the Congress takes resources of the TCU.

This obligatory tasks take the TCU's human and organisational resources nearly fully. Even if the calculus of work on a base of number of processes may not translate the real amount of work, it can serve as an illustration on the distribution of different tasks. In 1996 TCU analysed 15.000 processes, half of them related to the field of admission and retirement of federal servants and another third related to the examination of accounts submitted to TCU's examination and judgement.14 Out of several control techniques, power sharing by obligatory registration of administrative measures and the verification of annual accounts are passive forms of control. The controller has a limited influence on the presentation and selection of data.

On the contrary, investigations and audits programmed autonomously by TCU represent a minor part in the activities of the institution. The practical problem of TCU's capacity to make use of its wide reaching autonomy seems to be this overload with superficial control activities. In 1996, only 3,5% of the processes of the TCU refer to self programmed audit activities which have been implemented by its own staff in loco. The Zeitgeist of times when the TCU had to approve all financial expenditures of federal administrators on a base of second hand information still reigns in the institution until recent days. This is the case even after abolition of previous approval in 1967 when TCU has been given the opportunity to perform investigations and audits on his own initiative.

Between the constitutional control tasks with low autonomy and the audits programmed and implemented by the TCU lie a number of obligations of institutional assistance to the Congress and its Commissions. During the 1980's specially Special Commissions of Investigation of the Brazilian Congress played a mayor role in the uncovering of responsibilities for corruption scandals. As mentioned above, in this case the TCU has given great contributions by investigating financial corrupt practices. The merits of the investigation have gone to the Congress itself. Ineffectiveness attributed to TCU seems to be in part a problem of visibility of the whole range of its activities.

The TCU's lack of initiative is related to the question of overload with bureaucratic procedures. In recent years TCU makes a great effort to cut down the burden of routine control tasks by standardisation and computerisation. Freeing capacities for self-controlled audits is a prerequisite for an efficient use of the authority given to the TCU. The next problem on the way to an efficient control is selection. On the contrary to previous registration and submission of accounts, audits are activities based on selective rather than a full range approach. As fraud, misuse and waste of public funds are faults discovered by investigation rather than by denounces or complaints, a policy of investigation is a decisive element for its detection. Basic tools of such a policy of detection are (a) the identification of high risk areas, (b) the development of fraud indicators, (c) the allocation of special resources for the investigation and (d) the development of human expertise and institutional tools to deal with special issues.15 Recent experiences with special working groups on detection of fraud show how all these four issues are interrelated and can transform bureaucratic procedures into a creative way of detection and analysis without waste of public funds. The German Bundesrechnungshof in 1996 published the results of an in depth analysis on corruption in the area of contracts for construction of federal highways. The results include the expertise of several institutions which have experience in this field and led to an identification of a series of indicators for the detection of fraud in this area.16 The Landesrechnungshof of Sachsen-Anhalt decided to create a special unit for corruption control and developed fraud indicators for the area of public procurement.

In the Brazilian case, the experiences on the field of identification of problems have not grown too far. In fact, some efforts have been made to identify high risk areas. The TCU's investigation of the process of privatisation and the audit of Brazilian embassies, consulates and other institutions abroad seems to follow this philosophy. But fraud indicators to identify relevant cases of waste and abuse of public funds have not been developed systematically. Also allocation of resources and building up teams for special tasks have not grown very far. The state of Minas Gerais is the only one which set up a special department responsible for public procurement.17 An intensification of the efforts of the Brazilian TCU could improve the capacity of the TCU to detect grand corruption and lead it beyond its role of investigation of scandals detected by other institutions. But this does not mean TCU does not try to improve its performance. The next chapter shows how it participates in the debate on modern control standards.


For a long period, financial control was closely associated with formal correctness of accounts and accurate application of legal standards. Public expenditures of administrators were discharged if they could prove a legal origin in the budget law and followed a model of universal bureaucratic procedures. The techniques of the control institutions were limited to an analysis of the accounts rendered by the authorities to justify financial expenditure. The analysis of formal documentation prevailed over examination of material implementation.

Until the late 1960's this control pattern was nearly exclusive in the Brazilian TCU. The new Constitution of 1967 under authoritarian rule in Brazil, from one side shut down the model of previous approval for all public expenditure by the TCU. From the other side it opened the way for modern techniques of control. The introduction of techniques of investigation and audit for the external control permitted auditors to go beyond passive verification of documents. By the same time the government introduced a system of parallel internal control. Besides this separation of two different branches of control and the new techniques for its implementation another idea gained force. Since 1964 the government has introduced a new notion of program budget and has opened the discussion on an evaluation of public politics beyond the compliance of formal and legal standards.18 The proper efficiency of administration and the effectiveness of politics came to discussion and should be evaluated by internal control systems. Even if the TCU had no constitutional mandate to extend his control criteria into this area, an internal discussion on this new issue took place within the institution.

But it was only in 1982 when the TCU made its first practical experiences in the field of audit by criteria of efficiency and effectiveness of policies. This experience occurred even without an explicit authorisation by the Constitution of 1967 which still limited the criteria to standards of accountancy and legality. In 1985, after the shift to civilian rule, but still under the same Constitution, 28 federal institutions were examined by the TCU with the new criteria of efficiency and effectiveness. And in 1989 the TCU started an intense debate on the implementation of efficiency and effectiveness as new criteria of external control.19 The background was the explicit amplification of control criteria by the new Constitution of 1988. Constitution legitimates TCU to exercise control in the light of the three main criteria of legal, legitimate and economic use of public resources.20 From that moment on the question of implementation of controls dedicated to efficiency and effectiveness of programs are a permanent theme of discussion and one of the areas of main practical experimentation of the TCU.

In the mid 1990's this discussion on the issue of evaluation criteria received an further impulse by the reform of public administration introduced by the government of President Cardoso. Inspired by the discussion about new public management the reform searches to introduce a new philosophy of managerial ethics and diminish bureaucratic norms of behavior. The idea got strength because the internal control system of federal government has been centralised and completely restructured and plays a major role to make the new philosophy of control by criteria of efficiency and effectiveness work. 21 In addition, new kinds of contracts of the government including criteria like goal achievement oblige the external control institutions to familiarise urgently with this new philosophy in order to permit the control of these institutions.

In several countries the discussion on new public management received main impulses from external control institutions.22 In Brazil, the TCU so far played no major role in this process. But it managed to modernise its procedures and showed its capacity to innovation and creative adaption to new situations.


As mentioned above one of the original tasks of the TCU was to support Brazilian Congress in its role of controlling the federal administration. The presentation of a critical appraisal of the accounts of government was the basis for the final discharge by the Congress. Also the TCU shared executive powers as an institution not submitted to hierarchy when previous registration of financial expenditure became obligatory. We showed how this task obliged the TCU to a logic of action close to the bureaucratic structure of the executive and how the amount of processes still overload the institution.

Since its foundation, TCU has been additionally authorised to discharge administrators after the examination of their accounts. In case of rejection of accounts, the TCU could oblige the administrators to restitute the funds and even fix penalties. When the system of previous registration fell in 1967, this final verdict on the accounts of administrators remained as a major source of autonomous powers of the TCU. The Constitution of 1988 strengthened this position considerably. The judicial powers of the TCU are composed by several powers. First, the administrator can be condemned to restitute the full amount of losses he caused. Second, he might be punished to pay a penalty of up to 100% of losses or up to some US$ 6.000 if there are no material losses. Third, in cases of severe faults he might be temporarily suspended from office and his property be confiscated, if the investigation or restitution are in acute danger. Moreover, he may be prohibited to exercise any kind of public role on the base of appointment and can be declared ineligible. The powers of enforcement to prevent any form of non-cooperation and obstruction are also considerable.

An overview on the reproval of accounts and the resulting financial obligations of administrators to restitute losses shows a TCU which made use of its new attributions (see table 4). The tendency of financial charges against administrators seems to grow and reaches a level of 20 million US$ per year. But the same statistic reveals an extremely low rate of effective restitution of losses and punishment of about 1% of the total amount. The reason is a series of problems in the judicial procedures following the TCU verdict. It includes suspension of the accusation for lack of means for restitution, contestation of the decision before the courts of common justice and others.23 As a result, where TCU is efficient and uses its powers, other institutions of the judicial system slow down or even abort the condemnation.


If we go back to our initial question about the lack of awareness of financial control institutions to the major cases of misuse of public funds in Brazil we might state the following. Since the constitution of 1988 no major legal constraints limit the control activities of the Brazilian courts of accounts. The range of control areas is widely defined and the control criteria include traditional legal standards as well as modern criteria of efficiency and effectiveness. Despite TCU's historical burden of bureaucratic procedures it is on his way to cut down low intensive control routines and create new space for more intense forms of investigation and audits. The effort of the TCU to modernise control criteria towards efficiency and effectiveness is certainly a hopeful sign for TCU's capacity for innovation. On the other hand a policy for the detection of fraud, corruption and misuse of public resource could improve results in a field where Brazilian political system seems to depend on occasional denounces by the press and civil society. Foreign experience prove that identification of high risk areas and fraud indicators can improve the results in this area.

A further argument for more investment in this area of classic legal control might be the strength of TCU in this area once the final verdict on accounts lies in the hands of the TCU. An acceleration of the processes of judicial contest of TCU verdicts before Courts lies outside the jurisdiction of the TCU but significant modifications are on the way in the Brazilian judicial system. On the other hand the chances to turn findings in the area of effectiveness and efficiency relevant for practical politics still remain obscure. Finally, I would like to raise the question about the repetitive or complementary character of internal and external control systems. If the new internal control system has been designed to detect lack of efficiency in goal achievement, should the TCU really insist to repeat this task or wouldn't it be more appropriate to do a complementary work oriented by the traditional criteria of norm complaint?

The improvements in selection of court members have to show results. Once the new criteria had a minor effect on the composition of the TCU the effect specially of the shift of power to Congress has to be examined. As we saw, the major problem of TCU seemed to be the high fluctuation of its members due to its role in rewarding political allies with high pensions. Part of the new selection criteria since 1988 have been a first step to a more professionalised recruitment policy. But there are still spaces for recruitment of more auditors out of the proper institution or other control organs. Especially Courts of Accounts in the states of the Brazilian federation need to improve selection criteria, once the long standing influence of the same political group in power diminishes the role the Courts of Account could play as an independent authority of control.


 Managerial ModelLiberal Model
Why Control?diminish losses (material and image) limitation of power
When Control?parallel controlposterior control
Area of Controlall aspects of financial administration in practice conflict about areas of authority
Criteria of Controlformerly: account and legal standards; today efficiency and effectivenessformerly: account and legal standards; today efficiency and effectiveness
Powers of Control
Information of executive, parliament, citizensPreparation of decision on responsibility of public agent
Control aimChange execution of politicsJustice
LegitimacyDerived from executive or legislative Autonomous power, based in independence and autonomy

Federal Constitution (Chapter)Authority of Nomination Previous SelectionGuarantees
FC 1891, Art. 89 (General)President with approval of Senatenihil nomination for a lifetime
FC 1934, Art. 100 (Coordination)President with approval of Senatenihilnomination for a lifetime; guarantees equal to Ministers of Supreme Court
FC 1937, Art. 114 (Judiciary)President with approval of Federal Counselnihilnomination for a lifetime; guarantees equal to Ministers of Supreme Court
FC 1946, Art. 76 (Legislative)President with approval of SenateBrazilian; minimum age 35 years; political rightsnomination for a lifetime; guarantees equal to Ministers of Supreme Court
FC 1967, Art. 72 (Legislative)President with approval of SenateBrazilian; minimum age 35 years; moral integrity; expertise nomination for a lifetime; guarantees equal to Ministers of Supreme Court
FC 1969, Art. 70 (Legislative)President with approval of Senate Brazilian; minimum age 35 years; moral integrity; expertise nomination for a lifetime; guarantees equal to Ministers of Supreme Court
FC 1988, Art. 73 (Legislative)1/3 President with approval of Senate, 2/3 CongressBrazilian; age 35-65 years; moral integrity and reputation; expertise based on minimum of 10 years experience nomination for a lifetime; guarantees equal to Ministers of Supreme Court

periodnumber of new nominations permanence in power (means)

yearpersons condemned financial penalties (in US$, approx.)rate of execution (% or penalties)

Bruno Wilhelm Speck, Political Scientist, PhD
Work: State University of Campinas (UNICAMP), Brazil
e-mail: bws@unicamp.br
Address: Rua Judith Passald Esteves, 296, ap. 101
05625-030 São Paulo - SP, Brazil

Notes and References

  1. Heinz Günther Zavelberg: 275 Jahre staatliche Rechnungsprüfung in Deutschland, in: idem (Hrsg.): Die Kontrolle der Staatsfinanzen. Geschichte und Gegenwart. 1714-1989.Festschrift zur 275. Widerkehr der Errichtung der Preuwssischen General-Rechen-Kammer, Dincker & Humblot, Berlin, 1989.

  2. Alfredo Cecílio Lopes: Ensaio sobre o Tribunal de Contas, São Paulo, 1947.

  3. Antônio Roque Citadini: O controle externo da administração pública, Max Limonad, São Paulo, 1995.

  4. The overview of Antonio Roque Citadini: O Controle Externo da Administração Pública, São Paulo, Max Limonad, 1995, stands for many others. Citadini is a member of the Court of Accounts of São Paulo. The trend to limit the discussion to professionals in audit institutions is confirmed by the division of scientific journals dealing with control questions. Discussion is limited to a number of journals edited by the proper Courts of Account (Revista do Tribunal de Contas da União, Brasília). Some articles have been published in the Journal of the Legislative (Revista de Informação Legislativa, Brasília) and in the Journal of Administrative Science (Revista de Administração Pública, Rio de Janeiro). Brazilian Social Science journals until now do not include any contributions to the theme until now.

  5. See for example Richard Mulgan: The Process of Public Accountability, in: Australian Journal of Public Administration, vol. 56, n° 1, 1997, p. 25-36; Hans Blasius: Finanzkontrolle und Gesetzgebung. Können die Rechnunshöfe einen Beitrag zur Gesetzgebung leisten, in: DÖV, H. 7, 1989, S. 289- 306.

  6. Karl W. Deutsch: The Nerves of Government. Models of Political Communication and Control, New York, The Free Press of Glencoe, 1963

  7. At this time the impact of this verdict was minor if compared to the process of parliamentary investigation which resulted in the impeachment of the president.

  8. Alfredo Cecílio Lopes: Ensaio sobre o Tribunal de Contas, São Paulo, 1947, p. 211ff.

  9. Constitution of the Federal Republic of Brazil, October 1988, Art. 70.

  10. Since 1988 the one constitutional amendment made to the articles about the role of external control is related to the definition of the range of persons and institutions obliged to render accounts to the TCU."Prestará contas qualquer pessoa física ou entidade pública..." was changed to "Prestará contas qualquer pessoa física ou jurídica, pública ou privada..." See João Henrique Pederiva: Accountability, Constituição e Contabilidade, in: Revista de Informação Legislativa, Brasília, ano 35, n° 140, 1998, p. 19.

  11. Source for proper calculus: Ministros do Tribunal de Contas Da União. Dados Biográficos. 1893-1990, 4ª edição, Brasília, 1990

  12. O Estado de São Paulo, 29/12/1970.

  13. Conflicts between the local Courts of Account and Government occurred in the town of São Paulo, when the leftist majors Luisa Erundinas accounts were rejected by the Court of Accounts, whose members had been nominated bey her political enemies, for three consecutive years (1990, 1991, 1992). Other similar cases were the vetoes of the Courts of Account of the state of São Paulo (1996/1997) and Rio de Janeiro (1996) bringing privatization of railroad systems and public financial institutions to a halt for formal failures. Jornal do Brasil 01/09/1996, 22/10/1996, O Estado de São Paulo 22/10/1996, 28/02/1997).

  14. Relatório das Atividades do TCU. Exercício 1996, Brasília, 1997, p. 15.

  15. GAO United States: Methods and Techiques used in Preventing and Detecting Fraud and Corruption, paper prepared for the XVI INCOSAI, Montevideo 1998).

  16. Bundesrechnungshof. Sondertagung Korruption und Vergabemanipulation im Bundesfernstrassenbau. Dokumentation, Frankfurt, 1997.

  17. State Constitution of Minas Gerais, 1989, Art. 21, § 6.

  18. Luis Carlos Braga de Figueiredo: Auditoria operacional e avaliação de desempenho, in: Revista do Tribunal de Contas da Unição, ano 22, n° 49, 1991, p. 39-55

  19. Several seminaries have been organized by TCU on this topic. See for example Auditoria Operacional e Avaliação de Impacto e Desempenho, Brasília, setembro 1991.

  20. Constitution of the Federal Republic of Brazil, 1988.

  21. Secretaria Federal de Controle: http://www.sfc.fazenda.gov.br/

  22. Andreas W. Freiherr von Gall: Der lange Weg zum Ziel einer schlanken Verwaltung. Verwaltungsreform aus der Sicht einen Verwaltungscontrollers, in: VOP 9, 1997, S. 13-16.

  23. Relatório das Atividades do TCU. Ecxercício 1996, Brasília, 1997, p. 18.

return to table of contents