The informal foundations of coalition management in Brazil

The informal foundations of coalition management in Brazil
Sérgio Praça 1 (UFABC/FGV-SP), Marcos Lopes (FGV-SP), Thiago Belmar
(USP/Fapesp)
Political institutions in Brazil have come a long way over the last twenty years.
Academic diagnosis of the political system in the late eighties and early nineties was
very pessimistic. As late as 1994, influential intellectuals were arguing that the military
could take advantage of a weak congress and even weaker president and take over the
country once again (Lamounier 1994). Four years later, even during the successful
Cardoso two-term presidency, some political scientists still argued that Brazilian
politics was rife with clientelism and corruption, in a way that rendered political parties
inefficient and unprogrammatic (Diniz 1997, p. 19).
Since the mid-nineties, however, political scientists have shown that Brazilian
presidents are able to govern – that is, pass legislation and constitutional amendments
through Congress (Figueiredo and Limongi 1999; Melo 2002) – with the help of
disciplined political parties (Figueiredo and Limongi 1995; Amorim Neto et al 2003)
and highly centralized congressional procedures (Figueiredo and Limongi 1999; Santos
2003). Although the “cost of governing” varies, it is not considered prohibitive, as
presidents obtain legislative support without sacrificing their policy goals (Figueiredo
and
Limongi
2008).
The “difficult
combination” of open-list
proportional
representation, multipartyism and presidentialism (Mainwaring 1991) has thus far had
“surprising success” (Pereira and Melo 2012).
There is, however, a puzzle that has largely been ignored by this literature.
While it is true that Brazilian coalitions work – in the sense that they organize the
content and legislative passage of substantive legislation in many policy areas –, the
mechanisms through which they do so are mostly acknowledged to be formal.
Specifically, the centralization of legislative procedures in the steering committee (Mesa
Diretora) and the party leaders (Colégio de Líderes), along with the constitutional
powers vested in the president, have been considered responsible for allowing
governments to govern.
Coalition management has also mostly been described as happening through
formal mechanisms, such as those that allow cabinet positions to be allocated to partisan
ministers (Amorim Neto 1994, 2006; Martínez-Gallardo 2010), budget amendments to
1
Corresponding author: [email protected].
1
be executed according to the support of specific parties on the floor (Alston and Mueller
2005; Raile, Pereira and Power 2011), and lower-level bureaucracy positions to be
doled out to certain coalition partners instead of others (Praça, Freitas and Hoepers
2011).
We argue that some informal underpinnings of coalition management in Brazil
have been overlooked by the literature. More specifically, budget amendments at the
federal level have, since 1996, been decided upon through informal mechanisms. A
quota for individual budget amendments is annually agreed upon in a budget report that
is not etched in parchment (Carey 2000). This delicate informal balance, sustained by
repeated interactions and externalities for members who cooperate (Bawn 1999), has
been very important for coalition management in Brazil, both at the federal and state
level.
Besides its relevance for the literature on coalition management in presidential
systems, our argument also contributes to an influential literature on informal
institutions in Latin America. At least since O’Donnell (1996), informality has been
considered a telling sign of weak democracy and malfunctioning of formal institutions.
A different literature, however, argues that informal institutions can help overcome
problems of collective action, standing in for formal institutions when the cost of
formalizing agreements is too high (Knight 1992).
Our article shows a new angle on the Brazilian budget process which challenges
much of the literature that associates informality in Latin America with weak
democratic institutions (O’Donnell 1996), with social and political relationships that
depend on creative (and often illegal) interpretation of rules (Rosenn 1971) and with
corrupt networks (Samuels 2006). We argue that informality is very important for
managing the president’s legislative coalition and facilitating intertemporal agreements
with different political parties. Our argument thus follows recent scholars (Mershon
1994; Siavelis 2006; Helmke and Levitsky 2006; Mejía Acosta 2009) in showing how
informal institutions can be “good” rather than “bad” for democracy.
The fact that informality is ubiquitous in legislative politics 2 strengthens our
position in arguing that informal institutions must be analyzed without necessarily
associating them with corruption and weak democracies.
2
In the United States, for example, informal agreements are responsible for committee decisions to be
followed on the floor (Krehbiel 1990, p. 79), for seniority rules to be respected (Polsby 1968, p. 106), for
budget subcommitees not to be packed with preference outliers (Rundquist and Ferejohn 1975, p. 90,
2
The article is divided in four sections. The first section shows the informal
underpinnings of coalition management in Brazil and how it has worked at the federal
and state level. We then argue that these informal agreements are unlikely to be made
formal in the future, due to the distributional consequences we spell out. In the third
section, we show the consequences of this arrangement for the “cost of governing” in
Brazil at the federal and state level. Section four concludes.
(I) Budget institutions and the informal amendment quota in Brazil
A few words about the overall budget process in Brazil are in order. It has three main
characteristics that we will describe referring to the federal level, but which hold true
with minor differences for the state and municipal level. The first is the exclusive
prerogative of the Executive to introduce budget legislation in this area. The 1988
Federal Constitution divides the public budget into three legislative pieces: the MultiYear Plan (MYP), the Budget Guidelines Statute (BGS) and the Annual Budgetary Law
(ABL). The MYP establishes goals to be achieved during a four-year period. The
Budget Guidelines Statute should establish goals and priorities for the forthcoming year,
and also be responsible for changes on the tax legislation, establish investment policy
guidelines for the official financial development agencies and guide the formulation of
the Annual Budgetary Law. The president is responsible for proposing these three bills
to Congress.
The second characteristic refers to the nature of the participation of the
Legislative branch in the process. The Joint Budget Committee formed by deputies and
senators is the main decision-making locus, along with the ministries, for budget
formulation in Brazil. It is a very centralized process. Working alongside the Finance
and Planning ministries, the budget committee chairman is responsible for a report on
the fiscal situation of the country for the oncoming year. It is a very important position
in Brazil’s fiscal governance.
The third characteristic is the non-mandatory nature of the budget law. Around 90%
of the Brazilian budget consists of earmarked revenues assigned to certain policy areas
citing Fenno 1966, p. 141-149), and for nominees for a state or district position to not be approved by the
Senate in case the home senator objects (Jacobi 2005, p. 193).
3
(Velloso 2006). However, the president can choose not to execute some of these
expenses and congressmen cannot formally react to this.
The 1988 Federal Constitution establishes some ground rules for amending the
budget. Congressional amendments must be presented to the budget committee and
approved by the floor. They cannot be incompatible with the Multi-Year Plan and the
Budget Guidelines Statute. They must operate within an off-setting mechanism, that is,
the amendment can only propose new expenditures if it cancels a piece of the budget
corresponding to the same value (Hallerberg et al 2007, p. 346). Transfers to states and
municipalities, civil service payments and debt payments cannot be canceled or
amended by Congress.
These rules leave a lot of leeway for creative interpretation and corruption. The
late eighties and early nineties were rife with budget corruption scandals of different
sorts, and there was even talk of a wholesale change of budget institutions which would
limit congressional interference. However, incremental measures were taken slowly.
Two of them are especially relevant here: measures limiting the number of individual
amendments and measures limiting the value of these amendments.
Until June of 1993, when congressional resolution 1/1993 passed, congressmen
could propose as many individual budget amendments as they wanted. When the first
signs of rampant corruption circulated through the media, Congress decided to limit
each deputy and senator to 50 individual amendments. During the parliamentary
committee that investigated wrongdoings of the sort in late 1993 and early 1994, there
was talk of prohibiting individual amendments altogether. This would, of course, hinder
congressional reelection strategies (Ames 1995; Samuels 2002b, Pereira and Rennó
2003, 2007), so it is not at all surprising that it did not pass. Congressional resolution
2/1995 established a limit of 20 individual amendments, jumping up to 25 in
congressional resolution 1/2006, which stands now.
Measures limiting the value that may be appropriated for individual amendments
have also been taken in a piecemeal manner. This idea was first considered in 1994,
right after the investigative parliamentary committee proposed prohibiting amendments.
Since this was clearly too radical a proposal, congressmen started thinking of ways to
tie their hands in order to make the amendment process less corrupt (this was much
4
needed after the CPI) and to facilitate the negotiation to execute amendments with the
Executive 3.
In December 1994, then, a congressional budget report opted to establish a quota
for individual amendments, due to the combination of “a dearth of resources” and
“multiple social demands to be addressed by the amendments” (Brazil Budget
Committee 1994). Interestingly, congressional resolution 2/1995, considered by many
scholars to have organized otherwise chaotic budget procedures, did not say anything
about a quota.
Two years later, an informal agreement between the budget rapporteur, the
Executive and members of the coalition and opposition was reached in order to limit the
value of individual amendments that could be proposed by each federal deputy and
senator.
This was repeated in the following years and, in 2001, the need of an informal
agreement to establish this limit was formally established in article 24 of congressional
resolution 1/2001. The value is defined in the budget rapporteur’s Preliminary Report.
From 1995 to 2001, the quota for each congressman was R$ 1million - R$ 1,5 million,
increasing in the following years until it reached R$ 8 million in 2007.
Why exactly do we call it an “informal agreement”? The quota is not established
through something similar to a “due legal process”. Let’s take a legislative bill as an
example. When a deputy has an idea for a bill in Brazil, she sends it to the Mesa
Diretora (an official channel for receiving bills), which in turn forwards it to
committees. Now let’s imagine that the deputy has an idea for an amendment, but
cannot formally send it to the Mesa Diretora nor the committees. She rather has to
convince a deputy in the role of rapporteur to support her idea. In other words, there are
no official legislative channels through which her idea might be considered.
The latter scenario is exactly how the quota is established. The budget
rapporteur negotiates with deputies, senators and the Executive a quota for individual
amendments to organize inclusion of amendments in the budget bill. The channel
through which he communicates the quota is the “parecer preliminar”, an important
report in which he explains how the budget was altered by the congressmen, but which
has no formal legal standing on its own. Nothing in the “parecer preliminar” can be
enforced through official channels.
3
Samuels (2002a, p. 324-325) makes a similar argument, though referring only to the limit on the number
of individual amendments.
5
From 2002 onwards, there is a formal rule inside congressional resolution
1/2001 appointing the rapporteur the person responsible for proposing the quota for
individual amendments. If this rule is broken (i.e., if the floor proposes the quota), any
deputy can move a “questão de ordem” against its non-observance. Therefore, the
enforcement of the quota – or at least, the enforcement of the rule indicating the actor
who has the power to define the quota – became formal.
However, the quota itself is still established informally. A quick counterfactual
to illustrate this is in order. Brazil lacks a modern Public Finance Statute to organize
many fiscal and budget provisions which are not contemplated in Bill 4.320/1964, the
closest thing there is to a statute of this sort. One could easily imagine a quota for
individual amendments to be included in a Public Finance Statute. It would be
organized under a section that establishes “rules for budget execution”. It could
determine a quota to be established according to various potential criteria such as
economic growth, or it could even be a fixed percentage of the whole budget –
something like 1% or 2%. But an informal agreement is cozier for all involved: the
Executive prefers to learn each year which resources are enough to manage the coalition
and to have the necessary flexibility to change if it needs to do so, and congressmen
prefer to pressure the Executive to increase the quota - as indeed has been the case in
the last few years.
To be even more precise, it is important to mobilize two different definitions of
informality and see how the quota fares. The first definition is Jack Knight’s (1992, p.
98), according to whom informal institutions are self-enforcing in the absence of an
external enforcement mechanism. The second is Helmke and Levitksy’s, for whom
informal institutions are “socially shared rules, usually unwritten, that are created,
communicated, and enforced outside officially sanctioned channels”. Table 1 organizes
this discussion and includes information for the São Paulo state.
6
Table 1: Quota for individual budget amendments as an informal institution
Is there
a quota?
Federal
19881996
Federal
19962001
Federal
2002-
No
Created
outside
officially
sanctioned
channels?
-
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
State
level,
São
Paulo,
19882005
State
level,
São
Paulo,
2006-
No
Yes
Communicated
outside officially
sanctioned
channels?
Distributional
consequences
HelmkeLevitsky’s
definition of
informality
Knight’s
definition of
informality
-
Enforced
outside
officially
sanctioned
channels?
-
-
-
-
Coalition
members are
privileged
Coalition
members are
privileged
-
Yes
Yes
No
Yes
Members of
the governor’s
party are
privileged
Yes
-
Yes
-
Yes
-
Yes
-
-
Yes
(II) Distributional consequences of informal budget institutions in Brazil
This section describes the distributional consequences of the informal budget quota in
Brazil over the last couple of decades. As Knight (1992, p. 40) argues, distributional
consequences – and not social efficiency – are what drive the maintenance and reform
of social institutions. Disadvantaged actors will attempt to dismantle existing
institutional arrangements, while the “winners” will do their best to maintain them.
Political reality is often murkier. The “winners” vs. “losers” framework is not
sufficient to describe what happens when actors pledge allegiance to multiple principals,
as in the case of legislators (Desposato 2004; Carey 2007; Saiegh 2010) and bureaucrats
(Alesina and Tabellini 2007). To illustrate our argument, let’s take a fictional PT federal
deputy, Marlos, as an example. Since 2003, when president Lula took office, Marlos has
been a member of the coalition. On the one hand, he wants to propose and have
executed budget amendments to please his constituents and campaign financers
(Samuels 2002b). On the other hand, he is part of a government responsible for keeping
the budget relatively balanced. Does Marlos want a small or a huge informal budget
quota? Hard to say.
7
The fact is that Marlos – and his coalition partners – will negotiate the informal
quota according to this tension between pleasing constituents and pleasing the president.
If Marlos can get enough amendments executed in order to please his constituents, and
if this money, when bundled up with money for other deputies amendments, does not
hinder the government’s quest for a good fiscal year, he will have reason to support the
informal quota as it stands. In other words, he will support the quota if its distributional
consequences please him and other coalition members.
There was a time in Brazil, from 1988 to 1994, when control over the
proposition and execution of individual budget amendments pleased no one except
those involved in a corrupt network nicknamed “budget dwarves” (anões do
orçamento). These corrupt legislators, which numbered 43 (Krieger et al 1994), hogged
the amendment process so much that other deputies were led to propose the cancelling
of sovereign debt payments in order to get money for individual amendments 4. As
described in the previous section, this ended after an investigation in 1994, and from
1996 onwards an informal budget quota started to be established.
Since its beginning, the quota has only shot up, from R$ 1 million in 1996 to
thirteen times that value in 2011. Table 2 shows the numbers.
Table 2: Informal quota for individual budget amendments (R$ million)
Brazil
São Paulo
1
1996
1
1997
1
1998
1,5
1999
1,5
2000
1,5
2001
2
2002
2
2003
2,5
2004
3,5
2005
5 (initial proposal: 3,5)
2006
6 (initial proposal: 5)
2
2007
8 (initial proposal: 6)
2
2008
10 (initial proposal: 8)
2,085 (196 million)
2009
12,5 (initial proposal: 10)
2 (Opposition suggested 15)
2010
13 (initial proposal: 12,5)
2
2011
Sources: Perezino 2008 (Brazil 1996-2008); Federal Budget Preliminary Reports 2006-2011; São Paulo
State Budget Preliminary Reports, 2007-2011.
From 2006 to 2011, it was possible to obtain the initial quota proposed by the budget
rapporteur. Without exception, the initial proposal was smaller than what was ultimately
4
Such was the case of PT federal deputy João Paulo, in 1989. See Brazil Budget Committee 1989.
8
agreed upon. Negotiations are tough. In 2009, for example, congressmen initially
expected to be unsuccessful in increasing the quota due to the international economic
crisis (FSP, Painel, October 2, 2008). Indeed, the execution of both individual and
collective amendments that year was smaller than usual (FSP, Painel, October 14,
2008). This did not stop the quota from going up to R$ 10 million per congressman
from 2008’s R$ 8 million, in an agreement that brought together members of the
opposition and the coalition (FSP, “Apesar de crise, congressistas aumentam valor de
emendas”, November 4, 2008).
As for the particular distributional consequences of the informal quota for
individual amendments, it is clear that members of the coalition obtain a larger share .
Between 1996 and 2001, the president executed around 70% of the individual
amendments proposed by coalition members and 40% of amendments by opposition
members (Figueiredo and Limongi 2008). From 2004 to 2007, the execution of
individual amendments by coalition parties was in average 45% superior to the ones of
opposition parties (Perezino 2008). Much anecdotal evidence from the media supports
this argument 5.
At the state level in São Paulo, members of the governor’s party (PSDB) are
comparatively favored when it comes to budget allocation. In order to explore this, we
propose a preliminary econometric model to obtain partial correlations between a
proposed budget amendment success ratio and dummy variables indicating affiliation
with either State government or opposition parties.
The main idea is to evaluate how much party affiliation and alignment with the
State government alter the bargaining power of State representatives in terms of
approving budget amendments. The estimates try to capture whether these
characteristics influence the success ratio of State representatives.
First, we define how the success ratio is calculated: each State representative is
entitled to have approved, yearly, up to R$ 2 million in budget amendments – this value
is the denominator of the success ratio. We fix the value of the denominator to capture
the rationality behind the budget process, considering it is not a rational behavior for
State representatives to propose lower values. The numerator is the sum of approved
budget amendments proposed by each State representative. To illustrate how the index
5
See, for example, FSP, Comissão do Orçamento aprova parecer que amplia para R$ 12,5 mi emendas
individuais”, November 12, 2009. See also OESP, “Comissão aprova LDO e flexibiliza regras para obras
da Petrobrás”, http://www.estadao.com.br/noticias/nacional,comissao-aprova-ldo-e-flexibiliza-regraspara-obras-da-petrobras,901496,0.htm
9
is calculated, suppose we have two State representatives, A and B. In 2010,
assemblyman A, a PSDB member, proposes and manages to approve six budget
amendments, with total value of R$ 1,500,000 (One million and five hundred thousand
reais); assemblyman B, an opposition member, proposes and manages to approve three
budget amendments, with total value of R$ 2,000,000 (Two million reais). The success
ratio for assemblyman A, in 2010, is equal to 0.75, and the success ratio for
assemblyman B is equal to 1.00.
After defining how we calculate the success ratio, we start out exploring basic
correlations using a log-linear model, estimated using ordinary least squares:
Success Ratioit = Constantit + ß1PSDB_Memberit + ß2Opposition_Memberit + ß3Xit
+ error_termit
Success Ratioit is the calculated success ratio for assemblyman i in year t,
PSDB_Memberit is a dummy variable indicating if assemblyman i is a member of
PSDB, State government party, in year t, Oposition_Memberit is a dummy variable
indicating if assemblyman i is an opposition member in year t, Xit is a vector containing
other relevant characteristics of assemblyman i in year t, and the error term is assumed
to have the usual desirable properties. The relevant characteristics of assemblymen
contained in vector X include age, educational attainment, a dummy variable indicating
previous experience in the State Assembly, total of votes received in the elections
(logarithm), a dummy variable indicating party leadership, a dummy variable indicating
whether the assemblyman has changed party affiliation since the electoral process, an
index calculated by and independent NGO – Movimento Voto Consciente - as a proxy
for assemblyman performance, a dummy variable indicating if the assemblyman
constituency is mainly located in the State capital and year fixed effects. The results are
shown in table 3.
10
Table 3: Estimation results for budget amendment execution in São Paulo, 2007-2010
Our preliminary estimates indicate that coalition members are not clearly favored when
it comes to the approval of budget amendments. The estimated coefficient for the
PSDB_Member variable, despite positive, is not statistically significant at the 10 percent
level – it is noteworthy, however, that the estimated coefficients on specifications 2 and
4 are significant at the 10.9% and 12.7% levels, respectively.
On the other hand, the estimates clearly indicate that opposition members are
strongly impaired: being a member of the opposition reduces the success ratio by
approximately 15 percentage points on specification 1, 2 and 3. The complete model,
used on specification 4, indicates a success ratio approximately 12.6% lower.
Considering the fact that the results are significant at the 5% level, for specifications 1
11
and 4, and at the 1% level for specifications 2 and 3, the issue clearly deserves deeper
analysis.
The effects of other assemblyman characteristics are not statistically significant,
except for educational achievement and age on specification 4. The effects captured by
year fixed effects (not shown here, available upon request) are consistent with the
electoral cycle theory, with success ratios increasing during the mandate, peaking in the
election year.
These results are preliminary, and not extremely robust. However, they are
suggestive, and confirm the need to perform deeper and more rigorous investigation on
following iterations of this article.
How are these distributional consequences maintained over time?
If the informal quota has privileged coalition members in a disproportional manner,
should there not be attempts to change the system? Talks of establishing mechanisms to
guarantee execution of all individual budget amendments, irrespective of the political
parties to which their proponents belong to, have come up time and again. The Budget
Guidelines Statute of 2002 included an article in this sense (art. 22 of Lei n. 10.524,
July 25, 2002); the rapporteur of a committee to reform budget institutions also
mentioned the idea in a 2005 meeting (Budget Committee 2005, April 7, 2005, p. 205);
and the government undertook somewhat serious talks to guarantee execution of budget
amendments as long as the informal quota were simultaneously decreased 6.
Formal attempts to guarantee execution of budget amendments were made by
nine different congressmen since 2000. These attempts are better understood as part of a
movement, in Brazil, to make the budget process “mandatory” instead of “authoritative”
– that is, to eliminate the presidential line-item veto on the budget 7.
How would different Mandatory Budget proposals –- affect the informal
political equilibrium in place today? There are four possible models of Mandatory
Budget institutions - all of which except one would somehow eliminate the presidential
line-item veto. Table 4 summarizes them 8.
6
See FSP, “Governo Dilma quer reduzir emendas de parlamentares”, April 7, 2011, and FSP, “Deputados
apoiam redução, mas exigem garantias de pagamento”, April 8, 2011.
7
The following paragraphs are based on Praça 2009.
8
The first three are described by Pontes Lima (2003, p. 8-9).
12
Table 4: Mandatory Budget institutions and political equilibria
Economic
accountability
Distributive
advantage
Strategic
advantage
Actor likely to
defend the
proposal
Presidential
Line-item Veto
Full Mandatory
Budget
Congress
Pork-only
Mandatory
Budget
Executive
Intermediate
Mandatory
Budget
Congress
Flexible
Mandatory
Budget
Executive
Executive
Considerable
(Coalition
Members)
Considerable
(Executive)
Executive,
Congressmen,
Coalition
Members
---
---
Small (Coalition
Members)
---
Congress
Partial
(Congress)
Congressmen,
mostly
opposition
members
Small (Executive)
Considerable
(Executive)
Executive
Congressmen,
mostly opposition
members
Congressmen,
both coalition and
opposition
members
With a Full Mandatory Budget (henceforth Full MB), the government would be forced
to execute the budget exactly as defined by Congress. If the president found himself
without the fiscal means to do so, he would necessarily incur in deficit in order to
execute the budget. With an Intermediate Mandatory Budget (henceforth Intermediate
MB), the government could choose not to execute certain expenses due to lack of
resources, provided it obtained explicit congressional approval. There is also the
Flexible Mandatory Budget (henceforth Flexible MB), in which the government would
not execute certain expenses for lack of revenue, communicating this to Congress, but
without needing its explicit authorization. Finally, there is a fourth budget model: the
Mandatory Budget for Congressmen’s Amendments (henceforth Pork-Only MB). In
this case, the Executive would be forced to execute all individual and collective budget
amendments. The Pork-Only MB, very much related to the debate on the informal
quota for amendments, would have certain consequences if passed. Congressmen of the
opposition would benefit, because the distributive advantage of the coalition strategically beneficiated by the Executive when the budget is executed - would vanish.
An Intermediate MB would require the Executive to seek parliamentary approval if it
wanted to suspend the execution of part of the budget.
Among the options mentioned, the Intermediate MB is a second-best alternative
for congressmen of the opposition. The Flexible MB would only add a formal
communication of the Executive to Congress regarding cancellations and transfers made
by the Executive. It would, in practice, mean the maintenance of the presidential lineitem veto.
Congressmen from the coalition and from the opposition have been equally
responsible for proposing budgeting institutional change in Brazil. This is hard to
13
understand in a strict “winners and losers” framework: only the latter should, according
to that argument, propose the end of the presidential line-item veto. But there are some
reasons why a member of the coalition could suggest a Mandatory Budget bill. He could
want to strengthen Congress as a collective actor (Schickler 2001) and win support of
his peers; maybe, considering the execution of his amendments, he would rather win X
than risk a gamble between X + 10 or X – 10; maybe he is trying to blackmail the
Executive into bargaining around other proposals or even get more of his amendments
executed (Tollini 2009, p. 25); maybe he wants to assert his independence from the
Executive at the end of his mandate to seek higher office.
The most persuasive argument might be the second. A member of the opposition
may have strong incentives to comply with the line-item veto if the Executive usually
executes the budget amendments proposed by him. In other words, this actor’s level of
compliance changes according to the execution of his amendments – or, at least,
according to the way he perceives his amendments are being treated by the Executive.
In the same manner, congressmen who belong to the coalition are “winners” under the
line-item veto provided the Executive actually executes their amendments or other
projects of interest. Under a Mandatory Budget, coalition congressmen would be
unconditional “winners”, but they would gain just as much as congressmen from the
opposition.
None of the Pork-Only MB proposals went very far. Only one of them – Bill
2/2000, by senator José Alencar (who would later be vice-president) – obtained a
positive report from the Comissão de Constituição e Justiça, the committee that
analyzes the legal and constitutional standing of every bill. The rest of the proposals
were not analyzed.
The lesson we can take away from this discussion is that the informal budget
quota does not operate in an entirely blissful setting, pleasing equally all political actors
involved. Rather it allows for the unequal distribution of power and resources, which
explains why, despite helping the government to effectively govern, it encounters
pointed (yet unsuccessful) opposition in some circles.
(III) The cost of governing Brazil at the state and federal level
Brazilian presidents and governors have very similar constitutional powers and
coalition-building tools, as well as relatively high legislative success rates. Most of
14
them navigate through a budget process that is highly centralized in a single budget
committee (Praça 2012). They are endowed with a very complete “toolbox” of coalition
goods that might attract the support of political parties and legislators towards the
presidential agenda. One might expect to see these goods fully mobilized by the
president especially because Brazilian presidents have very ambitious legislative
agendas. This is mainly due to the fact that the 1988 Constitution is highly detailed,
forcing presidents to propose and lobby for a number of constitutional amendments,
even when they might prefer otherwise (Couto and Arantes 2006). Although
constitutional amendments occupy a privileged spot in the political arena, there is much
to be done at the regular legislative level. And this is where extensive decree power and
the right to initiate certain pieces legislation exclusively are most important, allowing
the president to be a privileged agenda-setter (Limongi and Figueiredo 2003; Pereira,
Power and Rennó 2008), even if they might not always get what they want.
When a president takes office in Brazil, a number of coalition-building strategies
may be chosen. At a higher risk of lacking legislative support for her proposals, she
might keep a minority coalition. If she decides otherwise, she has four different basic
combinations for gaining and keeping support. She may distribute ministerial portfolios
in a proportional or non-proportional manner, and she may distribute a share of the
budget equally or unequally among legislators or among coalition partners. These two
mechanisms are, by far, the most relevant in the Brazilian presidential “toolbox” (Raile,
Pereira, and Power 2011) and have been largely analyzed by the recent political science
literature.
As has been made clear in the previous sections of the article, Brazilian
presidents and the São Paulo governor also have at their disposal an informal budget
quota agreement with the legislatures. With this informal agreement, they can distribute
the “coalition goods” that are budget amendments either equally among coalition
partners (as Brazilian presidents usually do) or disproportionality benefitting members
of their own party (as the São Paulo governor is prone to do).
At the federal level, the most common type of coalition since 1988, when the
Constitution was enacted, has been majority coalitions with a somewhat proportional
distribution of offices among coalition parties. This has occurred in 155 of 218 months
of Brazilian government – or 71,10% of the time 9. The exact opposite – minority
9
All of the José Sarney government, most of the Itamar Franco government, most of the Fernando
Henrique Cardoso governments and practically the whole of the first Lula government.
15
coalitions with non-proportional distribution of offices - occurred 15,6% of the time, in
the lame duck last year of the Cardoso presidency and in the beginning of the confusing
Collor presidency. Table 5 shows the numbers.
Table 5: Brazilian coalitions at the federal level, 1988-2007
President
Date of the coalition
Months
Proportional
share of
ministerial
offices?
High (.77)
Low (.19)
Low (.25)
Low (.19)
Medium (.42)
Medium (.48)
Medium (.44)
Low (.20)
Medium (.56)
Informal
budget quota
17
7
4
14
5
11
5
12
15
Majority
of seats in
the
House?
Yes
No
No
No
No
Yes
Yes
Yes
Yes
Sarney
Collor 1
Collor 2
Collor 3
Collor 4
Franco 1
Franco 2
Franco 3
Cardoso I1
Cardoso I2
Cardoso
II-1
Cardoso
II-2
Lula I-1
Lula I-2
Lula I-3
Lula I-4
Lula I-5
Oct 1988 – Mar 1990
Mar 1990 – Oct 1990
Oct 1990 – Jan 1991
Feb 1991 – April 1992
April 1992 – Sep 1992
Oct 1992 – Aug 1993
Aug 1993 – Jan 1994
Jan 1994 – Dec 1994
Jan 1995 – April 1996
April 1996 – Dec 1998
32
Yes
Medium (.59)
Yes
Jan 1999 – Mar 2002
38
Yes
High (.64)
Yes
Mar 2002 – Dec 2002
9
No
Low (.46)
Yes
Jan 2003 – Jan 2004
Jan 2004 - Jan 2005
Feb 2005 – May 2005
May 2005 – July 2005
July 2005 – Jan 2007
12
12
4
3
18
No
Yes
Yes
Yes
Yes
Medium (.59)
Medium (.53)
Medium (.54)
Medium (.53)
Medium (.53)
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
No
No
Sources: Figueiredo 2007.
At the state level, Brazilian governors have basically the same coalition goods to share
among coalition partners, with the difference that they have a less ambitious political
agenda than their federal counterparts. This is mainly due to the distribution of policy
turf by the 1988 Constitution, which endows cities and the federal government with the
control of a host of policy areas, while giving the state only “shared control” over
important policies such as urban transport (shared with cities) and health and education
(shared with cities and the federal government 10).
Despite the relative lack of policy relevance, governors have been usually
considered by the literature to be very strong players, taking advantage of the relative
weaknesses of checks-and-balances institutions such as the Public Prosecutor’s Office,
the Judiciary and the legislature (Abrucio 1988a, 1988b).
10
Paradoxically, in the early nineties governors were deemed to be very strong players in the Brazilian
political system due to their control of state-owned banks and the opportunity to use them to advance
favors without worrying about fiscal responsibility, at least until 2000 (Abrucio 1998).
16
It comes as no surprise that, as table 6 shows, both presidents and governors in
Brazil have had relatively high passage rates for their legislative proposals.
Table 6: Political institutions, agenda power and success rates in Brazil, 1988-2008
Nature of Executive
agenda
Executive powers
Federal level
Highly constitutional;
conflictual nature
State level
overall, 19992006
Much less conflictual
than at the federal level
* Decree power
* Right to initiate
certain pieces of
legislation
exclusively
* Budget proposal
and execution
* Right to initiate
certain pieces of
legislation
exclusively
* Budget proposal
and execution
* Right to initiate
certain pieces of
legislation
exclusively
* Budget proposal
and execution
Mostly minority
coalitions*; Nonproportional
offices; Distributive
budget advantage
for governor’s
party
Sources: Ricci and Tomio 2010; Tomio and Ricci 2012; Massonetto and Sandes 2012.
* With the exception of 2009.
State level São
Paulo, 20032008
Much less conflictual
than at the federal level
Nature of
coalitions and
distribution of
coalition goods
Mostly majority
coalitions;
Proportional
offices; Distributive
budget advantage
for coalition
members
?
Average
executive
success rate
75,1
87,6
90%
The puzzle that comes to mind with table 6 is: how does the São Paulo governor obtain
such high rates of legislative success at a much lower cost than his federal counterpart?
Let us concentrate on the fourth column. Although much research remains to be done in
order to assess which political parties belong to the coalition, at least in São Paulo
ministerial offices are distributed in a less proportional manner than at the federal level
(Massonetto and Sandes 2012). And rules for distribution of an equal share of the
budget among legislators are hardly ever in place – indeed, some states have very tight
rules restricting budget amendments (Praça 2012). There is some convincing evidence,
then, that governors use less coalition-building tools than their federal counterpart, and
that this does not come at the expense of low legislative success rates (Tomio and Ricci
2012) – quite the contrary is true, in fact, as column five shows.
Two interesting arguments can be taken from the discussion above. The first is
that the cost of governing at the state level is much lower than at the federal level in
Brazil. This is very much in line with Abrucio’s (1998b) “state ultrapresidentialism”
hypotheses, although we argue that the cost of governing is lower due to the less
17
conflictual nature of the legislative agenda that is carried out by the governor 11. The
second is the fact that the São Paulo state government could compensate, with the
execution of budget amendments, coalition partners unhappy with the lack of offices.
That this state government does not do so begs the question: are budget amendments
relevant “coalition goods” at the state level?
(IV) Conclusions
This article challenges one literature and complements another. The challenge is to the
literature on informal institutions and democracy. Many scholars, with notable
exceptions, associate informality with illegal acts, corruption, and the weakening of
democracy. We show that this is not always the case, even in a country rife with
informal rules such as Brazil’s. The informal budget quota agreement has eased
interbranch negotiation at the federal level at least since 1996. Thus, beyond the case of
Brazil, there is comparative analytic value in viewing “accommodating” (Helmke and
Levitsky 2006) informal institutions as mechanisms for coalition-building in
presidential systems.
We also complement the literature on coalition management in presidential
systems by showing that the “coalition-oriented” execution of amendments rests on
informal agreements, rather than on formal mechanisms specified by the Constitution
and other pieces of legislation. And this matters for the cost of governing Brazil and
other presidential systems.
11
Rather than, as Abrucio (1998b) argues, the overpowerwing influence of the governor on state
deputies’ reelection chances.
18
References
Abrucio, Fernando. Os Barões da Federação. São Paulo, Usp/Hucitec, 1998a.
______. “O Ultrapresidencialismo Estadual”, in Andrade, Régis (ed.). Processo de
Governo no Município e no Estado. São Paulo, Edusp, 1998b, p. 87-116.
Alesina, Alberto & Tabellini, Guido. “Bureaucrats or Politicians? Part I: a single policy
task”, American Economic Review, v. 97, n. 1, 2007, p. 169-179.
Alston, Lee J. and Mueller, Bernardo. “Pork for Policy: executive and legislative
exchange in Brazil”, Journal of Law, Economics and Organization, v. 22, n. 1, 2005, p.
87-114.
Ames, Barry. “Electoral Strategy under Open-List Proportional Representation”,
American Journal of Political Science, v. 39, n. 2, 1995, p. 406-433.
Amorim Neto, Octavio. “The presidential calculus: Executive policy making and
cabinet formation in the Americas”, Comparative Political Studies, v. 39, n. 4, 2006, p.
415-440.
______. “Formação de gabinetes presidenciais no Brasil: coalizão versus cooptação”,
Nova Economia, v. 4, n. 1, 1994, p. 9-34.
______. Cox, Gary W. & McCubbins, Mathew. “Agenda Power in Brazil’s Câmara dos
Deputados, 1989-98”, World Politics, v. 55, 2003, p. 550-578.
Bawn, Kathleen.“Constructing ‘us’: Ideology, coalition politics and false
consciousness”, American Journal of Political Science, v. 43, n. 2, 1999, p. 303-334.
Brazil Budget Commitee 1989. CMO. Projeto de Lei Orçamentária para 1990. Adendo
ao Relatório Preliminar de Eraldo Tinoco. Anexo I (Relatório do Assessor Raimundo de
Menezes Vieira sobre o último documento de Tinoco). 20 de novembro de 1989.
Brazil Budget Committee 1994. Congressional Budget Report for 1994. Congresso
Nacional. Comissão Mista de Orçamento. Parecer Final (Relatório Relator-Geral) da
Comissão Mista de Orçamento sobre o Projeto de Lei n. 3 de 1994 (CN) de iniciativa do
Poder Executivo que “estima a receita e fixa a despesa da União para o exercício
financeiro de 1995”. Autores: Gilberto Miranda e diversos relatores-adjuntos, que são:
senador Carlos Patrocínio; deputado Joao Faustino; deputado Roberto Balestra;
deputado Vadão Gomes; deputado João Almeida; deputado Oswaldo Coelho; senador
Ronan Tito. 2 de dezembro de 1994.
Brazil Budget Committee 2005. Notas taquigráficas da Comissão Mista destinada a
propor a reestruturação do processo orçamentário. Brasília, Congresso Nacional, 2005,
mimeo, 282 pgs.
Carey, John. “Parchment, equilibria, and institutions”, Comparative Political Studies, v.
33, n. 6/7, 2000, p. 735-761.
______. “Competing Principals, Political Institutions, and Party Unity in Legislative
Voting”, American Journal of Political Science, v. 51, n. 1, 2007, p. 92–107.
19
Couto, Cláudio Gonçalves e Arantes, Rogério Bastos. “Constituição, governo e
democracia no Brasil”, Revista Brasileira de Ciências Sociais, v. 21, n. 61, p. 41-62,
2006.
Desposato, Scott. “The Impact of Federalism on National Party Cohesion in Brazil.”
Legislative Studies Quarterly v. 29, n. 2, 2004, p. 259-85.
Diniz, Eli. Crise, reforma do Estado e governabilidade, Rio de Janeiro, Ed. FGV, 1997.
Fenno, Richard. The Power of the Purse: Appropriations Politics in Congress. New
York, Little Brown & Company, 1966.
Figueiredo, Argelina. “Government coalitions in brazilian democracy”, Brazilian
Political Science Review, v. 1, n. 2, 2007, p. 182-216.
Figueiredo, Argelina & Limongi, Fernando. "Partidos Políticos na Câmara dos
Deputados: 1989-1994”, Dados, v. 38, n. 3, 1995,pp. 497-524.
___________. Executivo e Legislativo na nova ordem constitucional. Rio de Janeiro,
Ed. FGV, 1999.
___________. Política orçamentária no presidencialismo de coalizão. Rio de Janeiro,
Ed. FGV, 2008.
Hallerberg, Mark; Strauch, Rolf & von Hagen, Jürgen. “The design of fiscal rules and
forms of governance in European Union countries”, European Journal of Political
Economy, v. 23, n. 2, 2007, p. 338-359.
Helmke, Gretchen & Levitsky, Steven. “Introduction” in Helmke, Gretchen & Levitsky,
Steven (eds.) Informal Institutions and Democracy: lessons from Latin America.
Baltimore, Johns Hopkins University Press, 2006, p. 1-30.
Jacobi, Tonja. “The senatorial courtesy game: explaining the norm of informal vetoes in
advice and consent nominations”, Legislative Studies Quarterly, v. 30, n. 2, 2005, p.
193-217.
Krehbiel, Keith. Information and Legislative Organization. Ann Arbor, University of
Michigan Press, 1990.
Lamounier, Bolívar. "A democracia brasileira de 1985 à década de 1990: a síndrome da
paralisia hiperativa" in João Paulo dos Reis Velloso (ed.) Governabilidade, sistema
político e violência urbana. Rio de Janeiro, José Olympio, 1994.
Limongi, Fernando & Figueiredo, Argelina. “Medidas Provisórias”, in Benevides,
Maria Vitoria; Vannuchi, Paulo & Kerche, Fábio (eds.) Reforma política e cidadania.
São Paulo, Perseu Abramo, 2003.
Mainwaring, Scott. “Politicians, Parties, and Electoral Systems: Brazil in Comparative
Perspective”, Comparative Politics, v. 24, n. 1, 1991, p. 21-43.
Martínez-Gallardo, Cecilia; Inside the cabinet: the influence of ministers in the
policymaking process, in: Carlos Scartascini, Ernesto Stein, and Mariano Tommasi
(eds.), How Democracy Works: political institutions, actors, and arenas in Latin
20
American policymaking, Cambridge: Inter-American Development Bank, 2010, p. 119145.
Massonetto, Ana Paula & Sandes, Vitor. “Formação de gabinetes e coalizões nos
governos subnacionais brasileiros (1995-2010)”. Paper presented at the VI Congreso
Latinoamericano de Ciencia Política, ALACIP, 2012.
Mejía Acosta, Andrés. Informal coalitions and policymaking in Latin America: Ecuador
in comparative perspective. London, Routledge, 2009.
Melo, Marcus André. Reformas constitucionais no Brasil: instituições políticas e
processo decisório. Rio de Janeiro, Ed. Revan, 2002.
Pereira, Carlos and Rennó, Lucio, “Successful Reelection Strategies in Brazil: The
Electoral Impact of Distinct Institutional Incentives”, Electoral Studies, v. 22, n. 3,
2003, p. 425-448
____. “O que é que o reeleito tem? O retorno: o esboço de uma teoria da reeleição no
Brasil”, Revista de Economia Política, v. 27, n. 4, 2007, p. 664-683.
Pereira, Carlos & Melo, Marcus. “The surprising success of multiparty presidentialism”,
Journal of Democracy, v. 23, n. 3, 2012, p. 156-170.
Pereira, Carlos; Power, Tim & Rennó, Lucio. “Agenda Power, Executive Decree
Authority, and the Mixed Results of Reform in the Brazilian Congress”, Legislative
Studies Quarterly, v. 33, n. 1, 2008, p. 5-33.
Perezino, Luiz Fernando de Mello. A preponderância do Poder Executivo no processo
orçamentário. Senado Federal, Unilegis, 2008
Polsby, Nelson. “The institutionalization of the U.S. House of Representatives”,
American Political Science Review, v. 62, n.1, 1968, p. 144-168.
Pontes Lima, Edilberto Carlos. “Algumas observações sobre o orçamento impositivo no
Brasil”, Planejamento e Políticas Públicas, n. 26, 2003, p. 7-16.
Praça, Sérgio. “The dynamics of inertia: stability and change in democratic Brazil’s
budgeting institutions”. Paper presented at the Annual Meeting of the American
Political Science Association, Toronto, September 2009.
_____. “A organização do processo orçamentário nas Assembleias Legislativas
brasileiras”, Cadernos Gestão Pública e Cidadania, v. 17, n. 60, Jan./Jun. 2012, p. 1-23.
_____; Freitas, Andréa & Hoepers, Bruno. “Political appointments and coalition
management in Brazil, 2007-2010”, Journal of Politics in Latin America, v. 3, n. 2,
2011, p. 141-172.
Raile, Eric, Carlos Pereira, & Power, Timothy. “The Executive toolbox: building
legislative support in a multiparty presidential regime”,Political Research Quarterly, v.
64, n. 2, 2011, p. 323-334.
Rundquist, Barry S. & Ferejohn, John. “Observations on a distributive theory of policymaking: two american expenditure programs compared”, in Liske, Craig; Loehr,
21
William & McCamant, John. (eds.) Comparative public policy: issues, theories, and
methods. New York, John Wiley & Sons, 1975, p. 87-108.
Saiegh, Sebastián M. Ruling by statute: how uncertainty and vote buying shape
lawmaking. New York, Cambridge University Press, 2011.
Samuels, David. “Progressive ambition, federalism, and pork-barreling in Brazil”, in
Morgenstern, Scott & Nacif, Benito. (eds.) Legislative Politics in Latin America. New
York, Cambridge University Press, 2002a, p. 315-340.
____. “Pork barreling is not credit claiming or advertising: Campaign advertising and
the sources of the personal vote in Brazil”, Journal of Politics, v. 64, n. 3. 2002b, p.
845-863.
Santos, Fabiano. O poder legislativo no presidencialismo de coalizão. Belo Horizonte,
Ed. UFMG, 2003.
Tollini, Helio. “Reforming the budget formulation process in the Brazilian Congress”,
OECD Journal on Budgeting, v. 1, 2009, p. 1-29.
Tomio, Fabricio & Ricci, Paolo. “Instituições e decisões: estudo comparativo do
processo legislativo nas Assembléias Estaduais”. Paper presented at the annual meeting
of the Associação Brasileira de Ciência Política, 2010.
______. “O governo estadual na experiência política brasileira: os desempenhos
legislativos das Assembléias Estaduais”, Revista de Sociologia e Política, v. 21, n. 41,
2012, p. 193-217.
22