Content uploaded by Sebastian M. Saiegh
Author content
All content in this area was uploaded by Sebastian M. Saiegh
Content may be subject to copyright.
The Role of Legislatures in the Policymaking
Process∗
Sebastian M. Saiegh
Department of Political Science
University of Pittsburgh
sms47@pitt.edu
February 21, 2005
Abstract
Legislatures are critical institutions in making a democratic system function.
In most Latin American countries, however, the role of legislative institutions
in the policymaking process is marginal at best. This paper attempts to
evaluate the main factors that affect the role of Latin American legislatures
in the policy-making process. I identify the contribution of legislatures to the
overall policy-making process and the manner in which they exercise those
roles. Next step I discuss some of the factors that contribute to differences
among legislatures with respect to their ability to play an active role in the
policymaking process. Finally, I summarize the analysis by classifying Latin
American legislatures into different “types”.
∗Prepared for delivery at the Workshop on State Reform, Public Policies and Policymaking
Processes, Inter-American Development Bank, Washington D.C., Feb. 28-Mar. 2, 2005.
1
1 Introduction
Legislatures are critical institutions in making a democratic system function. In
most Latin American countries, however, the role of legislative institutions in the
policymaking process is marginal at best; legislators are inexperienced and unac-
customed to their roles and responsibilities; and the power and authority of the
legislature vis-`a-vis the other branches of government, particularly the executive,
are often weak or not clearly defined.
This paper attempts to evaluate the main factors that affect the role of Latin
American legislatures in the policy-making process. The first step is to identify
what is the contribution of the legislature to the overall policy-making process.
The goal is to establish a set of criteria that will enable us to assess the extent to
which a particular legislature fulfills the prescribed roles or not. The purpose is
to move away from characterizations such as the legislature is a “strong” or “im-
portant” player and to propose a clear “metric” to evaluate whether a legislature
is an active player or just a marginal one in the policymaking process. In other
words, once we characterize what a particular legislative body may be able to do,
and once we establish empirically what it is actually doing, we can then make
a judgment call on how active such legislature is in a given country/period with
respect to its specified role.
Once we have surveyed the extent to which different legislatures play more or
less actively some of their roles, the task is to identify the manner in which they
exercise those roles. For example, some legislatures may be able to produce their
own legislative proposals and thus challenge the executive branch, while some other
legislatures may not. Alternatively, policy initiatives may be drafted by political
parties and then presented to the party’s legislative contingent for their approval.
In this case, we should examine the interaction between parties and legislatures.
Notice also that a legislature can be very active either by blocking everything the
executive branch proposes, or by reformulating and/or amending those initiatives.
In this case, the key is to look at the interaction between the two branches of
government.
2
The next step is to discuss some of the cross-legislature factors that contribute
to differences among legislatures with respect to their ability to play an active
role in the policymaking process. Among the several factors that drive a leg-
islature’s role in the policymaking process, we should take into account (i) the
extent of its formal powers; (ii) the adequacy of the capacity provided by its pro-
cedures/structures/support; (iii) the amount of political space/discretion afforded
by other power holders (executives, parties); (iv) the goals of the members and
leaders of the legislative bodies themselves.
Finally, the last step is to summarize our analysis by identifying a few dimen-
sions so that we can establish some sort of typology of the cases. Namely, the idea
is to classify the different legislatures into a set of categories. For example, we may
be able to identify a “very active, obstructionist and unrepresentative” legislature
and distinguish it form a “very active, transformative and representative body”,
or from a “marginal, non-obstructive, representative” legislature.
In order to reach the enumerated goals, several key questions should be an-
swered. These questions are both positive and normative. An important question
arises first: how good is our existing knowledge on how legislative institutions op-
erate? Next, three broad sets of questions will be useful in orienting our research
on legislatures’ roles, and three additional sets of questions will useful for deter-
mining the factors behind the differences in legislatures’ abilities to play an active
role in the policymaking process:
1. How much power does the existing institutional framework grant to the leg-
islature?
2. What scope of action does the political system reserves for the legislature?
3. How much do legislators want to do to advance the institution as a significant
actor in the policymaking process?
4. How well does the legislature interact with society?
5. How well does the legislature perform its lawmaking/oversight functions?
6. How well do systems of legislative management and infrastructure help the
legislature perform its functions?
Following are some types of questions this paper seeks to answer under each
area. Answers to these questions will allow us to evaluate the different types of
3
legislatures with respect to their role in the policymaking process.
The first question has to do with how much authority do the constitution and
other organic laws grant the legislature: are members entitled to introduce legisla-
tion? If so, what is the scope of that power? Conversely, how much authority does
the executive have to introduce and influence the course of legislation? What are
the executive’s veto powers? Can the legislature override an executive veto – and
how? Can the executive rule by decree? What oversight powers do the legislature
posses? Does the legislature have the right to compel the executive to produce in-
formation about government operations? If so what is the scope of those powers?
What powers does the legislature have to call to account and remove executive
officials?
More power to legislators may mean less for someone else. Who are the other
significant actors in the political system? The list of potential rivals is long and
includes other governing institutions and political parties. What is the role of
regional interests? Do they think that a more active legislature would better rep-
resent them? The third question deals with the extent to which there is support
for a larger role for the legislature among the legislators themselves: does a sup-
port base for legislative strengthening within the legislature exist? Can one be
developed?
The next question refers to the extent to which citizens are able to observe their
legislature in operation: may citizens visit the legislature and attend committee
or plenary sessions? Do legislators meet with citizens regularly? What is the level
of access to information on the legislature? Does the legislature have a bill status
system which allows legislators and the public to know where legislation is in the
legislative process, and who is responsible for particular pieces of legislation? Does
the legislature keep accurate records of votes and meetings, and is this information
available to the public?
In terms of lawmaking and oversight performance, the following questions are
pertinent: are legislators equipped to draft legislation? Do they? Does the legis-
lature have professional staff – either partisan or nonpartisan – to assist them in
developing legislation? Is the staff adequately trained? Does the legislature have
adequate information upon which to base decisions? Does the legislature have
adequate financial resources for carrying out its responsibilities effectively? Are
legislators willing to exercise the oversight authority they possess? Are staff and
4
information resources sufficient for the legislature to exercise effective oversight?
Do political concerns preclude legislators utilizing the powers they possess?
Finally, we want to know how well do systems of legislative management and
infrastructure help the legislature perform its representation, lawmaking and over-
sight functions. Do legislative rules and procedures tend to help or hinder the
legislature from performing effectively? Does the legislature keep accurate records,
and are they available to the public? Are there regular policies in place for the
management of the legislature?
The answers to these questions will depend on each legislature’s formal powers
and the political context in which it operates. For example, the structure and
tenor of legislative-executive relations characterize the environment in which leg-
islators operate. However, the shape of interbranch relationships has to be seen
as the consequence of the branches’ capacities, and not as the reason why some
legislatures are not so active in the policymaking process. Namely, the degree to
which legislatures perform such functions as lawmaking, oversight and representa-
tion depend in great degree on their own organizational structure. On the other
hand, the “activism” of a particular legislature will in turn depend on the structure
of the legislature, as well as a country’s broader political institutions, its party and
electoral system, and the preferences of those who run it.
This initial study does not seek to cover everything in detail, though. The pri-
mary focus is to get a broad sense of the present circumstances of Latin American
legislatures, and their potential to become an important actor in the policymak-
ing process as defined by the laws, by politics, and by the incentives faced by its
members. At this stage, thus the project’s value lies in two distinct areas. First, it
yields extensive information on the internal organization of Latin American legis-
latures. This information is not easy to acquire and is absolutely necessary for any
intelligent thinking about legislative policy-making processes. Thus, the data that
we are in the process of collecting for this project is by itself a significant contribu-
tion. Most importantly, our findings will help us make informed recommendations
about how Latin American legislatures can become more representative, more ef-
fective at lawmaking, and more capable of oversight.
We should keep in mind, though, that we will not be able to empower legis-
latures if we do not establish the right incentives for individual legislators first.
Many of the resources that make a legislature a proactive player in the policymak-
ing process are “endogenous” to how much the legislators are motivated to develop
5
them. The institutionalization of any legislature starts with the recognition on the
part of its members that they need to spend part of their time and effort on build-
ing a stronger collective body. Hence the focus should be on the incentives that
legislators have to have in order to “invest” in the legislature.
The remainder of this paper is organized as follows. In section 2 I discuss
our existing knowledge on how legislative institutions operate in Latin American
countries. Section 3 describes the role of the legislature in the overall policy-making
process and the way in which they exercise those roles. Section 4 focuses the factors
that affect how active a role in the policymaking process different legislatures have.
In Section 5 I discuss briefly what type of generalizations can be made based on
the available evidence. Conclusions follow.
2 The Stock of Knowledge on Latin American
Legislatures
The enormous literature on the United States Congress provides a detailed and
relatively comprehensive understanding of this institution. In fact, we likely know
more about it than about any other political institution in the world. Venturing
outside of the United States however, one most commonly encounters legislatures
that do not resemble the US Congress.
This is particularly the case in the party-centered systems that dominate Latin
American democracies. However, our knowledge on how legislative institutions op-
erate in Latin America is still very rudimentary. Indeed, for many years the study
of legislatures in Latin American countries was thought to be irrelevant and even
frivolous so long as they had little (or no) significant in dictator-ridden polities.
Fortunately, the Third Wave of democratization brought, among other things, a
renewed interest in the role of legislatures. When democratic regimes started to
show symptoms of consolidation, most students of Latin America turned from the
problems of regime transition to the analysis of “democratic routines”, paying spe-
cial attention to the capacity of institutions to shape public policy-making through
their effects on political actors’ behavior.
There are now several studies dealing with constitutional issues, and the re-
lationship between the different branches of government. For example, most of
6
the recent studies on legislative performance in Latin America (both scholarly and
policy-oriented) have focused almost exclusively on the disproportionate power of
the executive branch vis-´a-vis the legislature. According to the conventional view,
Latin America countries are characterized by “proactive” executives and “reactive
assemblies” (Cox & Morgestern, 2002; Amorin Neto, Cox and McCubbins, 2003).
However, a closer examination of legislative-executive relations in some of these
countries, shows that legislatures do not solely exist to “rubber stamp” executive
decisions. By looking at the proportion of legislative initiatives of the executive
that are approved by the lower house of the national legislature we can observe that
Latin American presidents experience numerous legislative defeats. Take the case
of Ecuador. During the 1979-1986 period, the legislature only approved an average
42% of the bills introduced by the president, with a maximum of 65 percent in
1992 and a minimum of 10.7 percent in 1986. The Venezuelan congress, in turn,
approved on average 67% of Executive-initiated bills between 1959 and 1988, but
it only passed an average 50% of those bills in the 1979-1983 legislative period. In
fact, the success rate of Executive-initiated bills varies considerably across Latin
American countries and through time within these countries.
This evidence casts doubts on prior perceptions that legislatures in Latin Amer-
ica “rubber stamp” presidential initiatives. Of course, this does not mean that in
these countries legislative bodies play a central role in the policymaking process.
We also know that, in general, Latin American legislatures lack the organization,
financial resources, experienced members and staff to serve as a mature and au-
tonomous point of deliberation in the policy process. Thus, these legislatures –
very much like the U.S. Congress at the end of the nineteenth century –, operate
more as a blunt veto player (exercising gatekeeping and some other types of delay
strategies) than as effective policy making bodies.
We we know so far, thus is that the executive branch does not entirely domi-
nate the legislature, but that legislatures in these countries need to enhance their
capacity in matters such as policy-making, oversight, and representation. This
characterization is good, but it is not good enough. Further research is needed.
In particular, we need more studies that gives us a better understanding of the
micro-dynamics of the legislative policy-making process. Researchers have recently
became aware of this need. Not surprisingly, some of them have turned to theoret-
ical models developed towards the analysis of the U.S. Congress. However, while
studies of the US Congress are very helpful in generating general theory, they are
7
also quite limited in terms of extending general legislative theory due to their sta-
tus as case studies of an atypical case. One of the goals of this paper is to fill this
gap in our knowledge of how Latin American legislatures operate and their role in
the overall policymaking process.
3 The Role of a Legislature in the Policymaking
Process
Modern democracies are characterized by shared decision-making by the legislative
and executive branches: most policy initiatives are enacted into law, and thus must
be approved by the legislature. It becomes very important then to identify what
exactly is the contribution of the legislature to the overall policy-making process
should be.
USAID’s Handbook on Legislative Strengthening identifies three primary func-
tions of a democratic legislature. These functions are briefly described as follows:
1. Representation. Legislatures listen to, communicate with, and represent
the needs and wishes of citizens in policymaking; and intercede with govern-
ment on behalf of the citizens.
2. Lawmaking. Legislatures identify problems, study issues, receive expert
and public input, formulate or approve policies, and implement those polices
through laws designed to address or remedy the problem or issue.
3. Oversight. Legislatures oversee the implementation of laws, policies, and
programs by monitoring, reviewing, and investigating government activities
to ensure that government actions are transparent, accountable, consistent
with, and uphold existing laws and regulations.
In addition, we can identify some other functions performed by at least some of
the legislatures under study: (i) playing roles in executive removal (impeachment,
votes of no confidence, censure), (ii) serving as a recruiting pools for other govern-
ment positions.
In general, a process of democratic governance is most legitimate when it char-
acterized by transparency, pluralism, citizen involvement in decision-making, rep-
resentation, and accountability. In the case of legislatures, their operations are
8
transparent when legislative sessions and members are accessible to the public,
public information about legislative actions is readily available, and media cover-
age of the legislature is widespread.
A legislature that is representative and participatory is characterized by citizens
having access to, and input in, the policy making process through the legislature;
there is high degree of interaction between members and citizens; issues are ad-
dressed or legislation amended because of public input or pressure; and civil society
organizations, advocacy organizations, and/or interest groups are active partici-
pants in the legislative process.
In terms of enacting policy, a proactive legislature is one in which the policy for-
mulation and law making processes are the product of informed decision-making,
and legislation is well-thought out and drafted. In terms of policy implementation,
an accountable legislature ensures that laws and government programs are being
implemented fairly and effectively; the national budget is scrutinized and agreed
upon; public revenues and expenditures are monitored; and issues of public cor-
ruption and mismanagement are addressed.
As mentioned in the introduction, the way in which these functions are carried
out depends on the structure of the legislature, as well as a country’s broader po-
litical landscape. For example, when the PRI lost control of the Mexican Congress
in 1997, different parties controlled the executive and legislative branches for the
first time in the country’s modern history. Since then the Mexican Congress has
increased its independence, assertiveness and willingness to criticize the execu-
tive. On the other hand, while the environment in which legislators operate is
important, these actors do not necessarily need to be passive. They can change
the environment in which they are operating. Take the case of Bolivia. In 1997
changes to the electoral system for its House of Deputies from pure party list to a
mixed system were introduced, at least in part, to respond to the perception that
the Congress was unresponsive to the needs of the electorate.
The “activism” of a particular legislature thus cannot be isolated from its coun-
try’s broader political institutions. But the degree to which a legislature performs
such functions as lawmaking, oversight and representation depend in great degree
on its own organizational structure.
9
4 Explaining the Variation in What Legislatures
Do
I turn now to some of the cross-legislature factors that contribute to differences
among legislatures with respect to their ability to play an active role in the policy-
making process. Several factors drive a legislature’s role in the PMP: the ex-
tent of its formal powers; the adequacy of the capacity provided by its proce-
dures/structures/support; the amount of political space/discretion afforded by
other power holders (executives, parties); and the goals of the members and leaders
of the legislative bodies themselves.
Let me start by looking at the constituent units of a legislature: its members.
The degree to which legislators influence and active in policymaking varies consid-
erably. Their behavior is shaped by a variety of factors. These include personal
motivations, how they view their jobs, and the variety of ways that they can re-
spond to constituents. Member-constituent relations can affect each of the role of
the legislature in the policymaking process by shaping member motivations and
incentives, by providing local content and human context to decisions, and by pro-
viding a way for constituents to measure performance of legislators and to assess
government actions. In other words, legislative behavior and the organization of
legislative institutions are affected by electoral rules.
4.1 The Electoral Disconnection
The idea that the organization of the US Congress reflects the electoral incentives
faced by its members has been a cornerstone of the American politics literature.
The key question for comparative work is whether the assumptions behind the US
case are appropriate in understanding how electoral rules shape legislative incen-
tives elsewhere. In principle, non-US politicians are as strategic in their actions
as their US counterparts. However, the political institutions that shape legis-
lators’ incentives do vary across countries; career structure, electoral laws, and
party rules can be very different. The question narrows, then, to the incentives
these politicians face in different contexts. For example, if party nomination is
inconsequential for electoral success, as is the case for incumbents in the Brazilian
Chamber of Deputies, party renomination will play no substantial role in shaping
legislators’ behavior. In contrast, there are situations, as in closed-list PR systems,
where nomination at the top of a major party list can virtually guarantee electoral
success. In this latter case, legislators’ behavior will be constrained by the renom-
10
ination rules but essentially unconstrained by the electoral process.
Take the case of Argentina, a federal country with a closed-list PR electoral
system. As Jones et. al. note, the process by which the provincial (district) party
lists are formed tends to affect which candidates run on each party list, what order
they occupy, and, consequently, their chances of winning a seat in the legislature.
Moreover, these lists are often assembled at the district level by the respective local
party boss. Therefore, legislators will have little incentive to work hard to improve
their visibility in the eyes of the voters, and no incentive to develop legislative
policy expertise. The aim of this example is to make the point that a legislature’s
organization reflects the electoral incentives faced by its members.
Figure 1 shows reelection rates for a cross-section of Latin American legislators.
The change in legislative personnel is important because high turnover seriously
curtails the chances of providing good representation. As figure 1 shows, reelection
rates are very low in Latin America. On average, less than 40% of the legislators
return to their seats. Uruguay shows the highest reelection rates with less than
60%. While in Argentina the typical legislator does not spend more than a 4-year
term in office.
<Figure 1 Here >
The high levels of turnover may be seen as a sign of a “healthy” political sys-
tem: the voters are voting the rascals out of office. However, this is seldom the
case. In general, the biggest hurdle in the career path of Latin American legislators
is posed by their own parties. Most party or executive dominated systems often
have features that limit the development of constituency ties. These include pro-
hibitions on re-election (Mexico), and devices that reduce the control constituents
have over candidates (party list systems, large, at-large constituencies and tightly
controlled candidate selection processes). But even when legislators develop close
ties with constituents, these relationships may still compete with other loyalties,
including those to their parties and leaders. Legislators are most likely to repre-
sent constituent interests when (i) they know precisely who their constituents are,
(ii) they interact with them frequently, and (ii) their political futures depend on
gaining and keeping constituent support.
In terms the legislature’s own accountability, accessible or inadequate meeting
facilities and/or insufficient time for legislators to meet with constituents also act as
11
constraints on relations between legislators and their constituents. In many coun-
tries citizens may not even enter the national legislature without an appointment.
It is necessary that the legislature and legislators, as representatives of the people,
communicates their deliberations and decisions to the public. Unfortunately, here
the record of most Latin American legislatures is also wanting.
<Table 1 Here >
Table 1 shows the availability of public information regarding the legislature’s
proceedings in 16 Latin American countries. Except for a few cases (Brazil, Chile,
Peru, and Mexico), in most countries it is very difficult to obtain a record of the
roll call votes in the legislature.
4.2 Agenda Control
Representing the public, though, means more than articulating citizen preferences.
It also involves having a say in translating preferences into policy through enacting
legislation. The representative and lawmaking functions co-exist in an uneasy but
necessary relationship. Lawmaking requires reconciling of differences once articu-
lated, as well as pressing the legislature’s claim to power against the executive and
other power-holders, such as political parties. This requires legislative processes
capable of reconciling conflicts and bringing to bear enough expertise to be taken
seriously by the executive branch and other actors in the system.
In general, for a bill to become law, it first must be approved by the legislature.
Next, the executive branch has to enact or promulgate it, and then a bill becomes a
law. In presidential regimes, the executive has the opportunity to reject legislation
through the veto process. In turn, legislators may override a president’s veto and
enact legislation without his/her signature. Besides this general characterization,
though, legislative processes can and do vary. There is variation in the range
of options on how to shape legislation (amendment, rejection, referral back to
executive branch), in the finality of legislative passage (executive veto powers and
override possibilities and requirements), and the impact of legislative rejection or
inaction (including the possibility of executives ruling by decrees which have the
force of law). For example, in some legislatures, bills submitted by the executive
must be considered and acted on by the legislature, and the executive may even set
a deadline for the legislature to act. This is the case in Chile, where the executive
12
determines the “urgency” of legislation, setting a timetable for legislative action.
The following table gives an account of the rules and practices of agenda setting
that are currently in force in 18 Latin American countries.
<Table 2 Here >
The table illustrates the effect of different veto procedures on the balance of
power between the executive and legislative branches of government. In particu-
lar, it shows how amendatory veto power gives the executive branch substantial
leverage in the legislative decision-making process. As it can be seen, in 10 of the
countries (El Salvador, Nicaragua, Peru, Venezuela, Ecuador, Uruguay, Bolivia,
Costa Rica, Mexico, and Chile) the prediction is that the executive branch can
use amendatory observations to obtain its preferred policy. In five countries, the
predicted outcome is a compromise between the policy positions of the legislature
and the executive (Colombia, Paraguay, Panama, Brazil, and Argentina). In only
three of the 18 countries the prediction is that the legislature will be able to impose
its ideal policy.
Agenda-setting rules that give to much power to the executive may end up
undermining legislators’ interest in developing the capacities of the legislatures.
In the case of Chile, the 1980 constitution sought to broaden and consolidate the
powers of the military government. When the country transitioned to democrat-
ically elected government in the late 1980’s, many of the provisions of the 1980
constitution remained intact. Pinochet remained the head of the Armed Forces for
eight years and the position of Senator for Life. Nine Senate seats were set-aside
for appointed military leaders. This gave the right wing forces (who were mili-
tary supporters) a majority of seats in the upper house. The constitution however
severely limited the role of Congress in legislative matters relative to earlier legis-
latures in Chilean history. In several areas, the president was given sole authority
to introduce bills. The president could grant certain initiatives priority status,
requiring that Congress act within a short period. This gave the president the
exclusive power to set the legislative agenda and, therefore, the political agenda.
The legacy of Pinochet’s government contributed to the strong unity and discipline
of the Concertacion parties. Both presidents and legislative leaders sought con-
sensus and implemented moderate policies in the post-authoritarian era. However,
some observers believe that this political accommodation will show cracks in the
long-term. In particular, members of Congress may start to resent their lack of
significant input into the policy process.
13
4.3 Legislative Success Rates
The authority that these institutional rules grant the legislature are particularly
important in the case of Latin America because most of them play a dominant role
in the law-making process. They introduce a significant proportion of bills, and in
some countries, they even have the monopoly of legislative initiative on important
issues. For example, the role of the legislature with regard to budgetary legislation
is highly limited everywhere except for Bolivia, Costa Rica, Guatemala, Honduras,
and Paraguay. In all the other countries, the legislature can propose only those
amendments that do not increase the deficit or the spending, and in several cases
it can do so only with the approval of the president.
Figure 2 shows the distribution of Latin American presidents’ legislative success
rates. Legislative success is measured by the number of executive proposals ap-
proved in the lower house of the national legislature, divided by the total number
of proposals introduced by the executive in a given period. The sample includes
11 countries for the period 1946-2000, including 207 country-year observations.
Sometimes the sources reported the data in legislative terms, so to create annual
observations, longer periods were apportioned to years taking as the criterion the
state of affairs as of December 31 of each year.
<Figure 2 Here >
On average, presidents get 64% of their bills approved by the lower house of
their respective national legislatures. The lowest success rate, 8% corresponds to
Colombia in 1993, followed by Brazil in 1954/55 (9.8%), Ecuador in 1986 (10.7%)
and Costa Rica in 1989 (11.6%). Cardenas et. al. (2004) argue that Colombia’s
1991 constitutional reform had very deep and important effects on the country’s
policy-making process. In the case of Brazil, Cheibub Figueiredo suggests that
the meager success rate of the Caf’e Filho and Nereu Ramos administrations is
explained by the limited powers awarded to the Chief Executive under the 1946
constitution. In fact, the average legislative success rate for different adminis-
trations during the whole 1947-64 period was less than 30% (Cheibub Figueiredo
2000: 10). With regard to Ecuador, Mej´
ia Acosta (2000) argues that presidents
have been very ineffective in getting their legislation passed due to the timing of
elections. He claims that they have little time to learn how to make correct predic-
tions about the expected support for their policies in congress. As for Costa Rica,
according to Taylor-Robinson (2001), term limits explain low levels of legislative
14
support. She argues that deputies’ best opportunity for continuing in politics is
to transfer their loyalty to new presidential candidates rather than to cooperate
with incumbent presidents. Thus, even legislators from presidential parties have
an incentive to oppose the president and wait to the next election.
At the other end of the distribution, there are 19 observations were presidents
got more than 95% of their bills approved: Mexico (1982-1996), Brazil (1973)
and Paraguay (1990-1992). These are cases where the president could exert a lot
of pressure over its legislative contingent. In fact, many observers would even
argue wether we should consider these countries to be democratic in these years.
In general, though, Latin American presidents do not seem to be less successful
legislatively than presidents in other parts of the world. Table 3 displays descriptive
statistics for the legislative success measure conditioned on regime type for a sample
of non-Latin American countries. As the table shows, Chief Executives do not get
their legislation approved all the time. Moreover, Latin American presidents are
not “exceptional” with respect to their legislative success rates.
<Table 3 Here >
Instances where government bills make it all the way to the floor and are voted
down are exceptional, but do occur. In general, tough, executive-initiated bills
“die” in the legislature in other ways. Their fate depends on legislative procedures
in each country, and in general, can be any of the following: a negative committee
report which kills the bill; no committee report, so the bill never leaves the commit-
tee; a positive committee report, which makes it possible for the Floor to debate
the bill, but does not guarantee that the bill will come up for debate; and outright
defeat of the bill by the Floor. In most countries, bills that are not considered
during a legislative year expire and have to be introduced again to the legislature
in the following period. But in some other countries, legislative proposals can carry
over into the following year. Thus, if a bill does not reach the floor, its “death”
can be automatic after some period of time or can “languish” in the legislature for
a longer period. The lifetime of bills before lapsing if they are not adopted varies
considerably between countries. In some countries a bill must be passed within
the legislative session in which is introduced. Otherwise, the bill has to be reintro-
duced in the next session. In some other countries, the lifetime of a bill is equal to
that of a legislative term. Therefore, bills do not lapse unless the legislative term
expires. Finally, in some other countries bills never lapse, except when they are
explicitly rejected by a vote. This is the case, for example, in Argentina.
15
Other bills do not die, but mat be amended by the legislature. Unfortunately
I do not have data on legislative amendments to executive-initiated bills for these
countries. Nonetheless we can identify some potential sources of variation by look-
ing at the institutional rules that regulate the amendment process in different
legislatures. For example, some legislatures require amendments to be supported
by a particular number of legislators (in Brazil the number amounts to 26). My
conjecture is that this type of restrictions will tend to affect smaller parties in the
legislature and restrict their ability to influence the policy making process.
Now, legislators can also propose their own legislation. Here the picture is even
bleaker. Casar reports that in Mexico between 1982 and 1988, about six out of ten
of all legislative proposals originated from legislative initiative, but of those only
one out of twenty became law (Casar 2002: 126). This seems to be the pattern in
most Latin American countries: while individual legislators have everywhere the
right to introduce bills, the legislation most likely to be enacted is initiated by the
executive branch. Table 4 presents data from Argentina, Honduras, and Paraguay.
<Table 4 Here >
We can see that in Honduras, there is a clear dominance of the executive over
the legislature. The president was responsible for almost 75% of the laws that were
passed by the Congress between 1990 and 1995. In contrast, both in Argentina
and Paraguay, the majority of the laws enacted by the legislature were of legisla-
tive origin. In Argentina the president only is responsible for 48% of the total
number of approved bills by Congress during this period. Likewise, in the case
of Paraguay, legislators were responsible for roughly 60% of the approved bills.
By looking at these data one may conclude that the executive is “dominant” in
Honduras, and less so in Argentina or Paraguay. However, a different way to in-
terpret these data is to look at the ratio of bills passed to bills introduced. The
data show that in Honduras the legislature passed 80% of the executive-initiated
bills, while individual legislators converted 43% of their bills into laws. In the case
of Paraguay, the executive had 75% of its bills approved by the legislature, and
legislators got 47% of their bills approved during the 1992-2003 period. While in
Argentina, the ratio of bills passed to bills introduced amounted to almost 60%
in the case of the executive and it was less than 5% for individual legislators.
Now, a different picture emerges: in all countries the success of private member
bills is lower that that of the executive. This is most notable in the Argentine case.
16
Finally, one may look at total legislative output. The Argentine Congress passed
127 laws per year on average: 59 of them were initiated by the executive, and 68
were initiated by legislators. In the case of Paraguay, the average number of laws
enacted by year amounted to 208, of which 84 were executive-initiated and the
remaining 124 were of legislative origin. And in Honduras, the yearly average
legislative output was 207 laws, of which 156 were initiated by the executive and
the remaining 51 by members of the legislature. If one looks at total legislative
output, yet another conclusion can be drawn from the data. The Paraguayan
legislature seems to be the most “productive,” followed closely by Honduras, and
then by Argentina. However, it is not a very good idea to compare legislative
output across nations (D¨oring 1995; Trantas 1995). The main concern is that legal
instruments are often defined in distinct way across nations. For example, in some
countries, very trivial and non-controversial matters are subject of legislation driv-
ing legislative output up. In contrast, in some other countries, only relevant and
controversial matters reach legislative stage, while all other matters are handled
by the administration.
4.4 Committees
In most legislative bodies, committees are the workhorses of the legislative process.
Referral to committee is usually a formal step in the process of adopting a bill.
The possible roles of the committees, though, vary from country to country, de-
pending upon the governing system, strength and organization of political parties,
available resources and other political factors. In some legislatures, committees
have the power to prevent unwanted legislation from being further considered, in
others; the committee stage may be merely a formality. Committees usually reach
decisions by a vote of the majority. The “majority” may not mean a majority
of committee members, but a majority of a certain minimal number of members
present (a quorum). Following the committee review proposed bills are then re-
ported to the floor for further debate and voting. Bills may also be discharged
from the committees and be submitted to the floor. However this usually entails
a special procedure that requires the approval of all the legislative parties’ leaders
in conjunction with the Speaker.
In Latin American most legislatures have permanent committees with specific
policy jurisdictions and legislation is routinely referred to these committees. The
jurisdiction of the committee is usually defined by subject matter. A well-designed
17
system is one in which this definition tends to parallel the structure of the admin-
istrative or cabinet agencies. However this is the exception (i.e. Colombia) rather
than the rule in Latin America.
In general in most Latin American countries committee jurisdictions should be
streamlined. Some committees have far too vast, and others far too narrow, a focus.
As a result when many bills are referred to multiple committees, generating either
turf fights or legislative paralysis ensues. For example, in the case of Argentina,
according to the Chamber rules whenever a bill is referred to more than one com-
mittee, a committee on the whole (composed of all the members of the committees
involved) has to agree before a bill is sent to the floor (article 102/Chamber of
Deputies). This procedure creates multiple veto points. When the Appropriations
committee is involved, the committee on the whole has to make a decision within
a month. Otherwise, the lead committee can report the bill to the floor. Most
bills dealing with economic issues are referred to the Appropriations committee.
Therefore, a bill’s likelihood of approval depends on its content, but ceteris paribus
tends to decrease as more committees are involved in the process.
The number and size of committees also vary greatly and does not necessarily
correspond to the size of the legislature. One would expect legislatures with many
members to have relatively more committees than legislatures with few members.
However, some small legislatures have relatively too many committees. The fol-
lowing figure shows the relative size of committees in 16 Latin American lower
houses of their respective national legislatures. The size of these bodies range
from 513 (Brazil) to 57 (Costa Rica) members and the number of committees from
7 (Colombia) to 48 (Honduras). It also includes, as a benchmark, the average ratio
of the number of committees to the size of the legislature for a cross-section of non-
Latin American countries. Compared to this benchmark, except for Colombia and
Brazil, all the countries in this sample have too many committees relative to their
size. The worst case is Paraguay with 25 standing committees in a 80-legislator
body. Notice that too many committees vie for legislators’ effort. So in terms of
the legislature role in the policymaking process, a system with too many commit-
tees overextend legislators and/or create duplication in their work.
<Figure 3 Here >
Committee composition and membership is thus very important. Legislative
rules often require each legislator to sit on at least one committee, but they differ
18
in how members are selected and on the number of committees on which they
can serve. These rules are important because the size of a committee may affect
its ability to develop expertise and to function effectively. The following table
contains information about specific committee rules based on 2004 data for 16
Latin American legislatures.
<Table 5 Here >
The first thing to notice is that these rules affect the number of committee slots.
The size of committees can vary from countries like Argentina, which has more 675
slots, to Colombia, which has less than 250. In Argentina, for example, commit-
tees must normally have a minimum of 15 and a maximum of 25 members. The
rules do not restrict multiple assignments. Moreover, during the 1983-2004 period
the number of standing committees from 27 to 45, while the number of deputies
increased only slightly from 254 to 257. And not only there has been a monotonic
growth in the supply of committee slots, but also in the demand for them. The
large number of Chamber committee assignments in the 2004 legislative period,
for example, required every member to fill an average of 2.63 slots. Note that this
is the minimum number of slots required on average to fill the 650 existing slots.
However, the average Argentine deputy decided to serve on 4.5 committees!
Comparative research shows the centrality of legislative committees as deter-
minants of the approaches to policies and of the interests that prevail in processes
of legislative policy-making. Strengthening the committee system is often seen
as the appropriate vehicle to intensify a legislature’s capacity. However – as a
well-established literature in political science shows – the stronger the committee
system, the greater the risk of allowing special interests to dominate the legislative
policy-making process. Thus, in order to make a correct evaluation of a legisla-
ture’s strength (and how to enhance it), we need to know more about how the
particular committee systems works, and specially whose views and interests pre-
vail in the legislative policy-making process. Committees with a large number
of members may be unwieldy and ineffective. As Jones et. al indicate, in the
Argentine chamber of deputies committees are too large and to unwieldy for the
development of genuine expertise. The information in table 5 also gives us some
understanding regarding the role of committees in the legislative policy-making
process. For example, if committee members tend to serve particular constituen-
cies (through specialization); or whether political parties and not committees have
an upper hand in the policy-making process. I look at the legislature in each of
these and focused on the number of committees each legislator serves on. Then, I
19
computed for each legislator the number of committees he/she serves on. Finally
I compared the average number of committees the typical legislator serves on with
the minimum and the maximum number of slots required on average to fill the
existing slots. With this information I constructed an index of “specialization.”
The data are presented in figure 4.
<Figure 4 Here >
We can see from the figure that Colombia, Brazil and Peru show the most
specialized committees, while Paraguay, Guatemala and Honduras have to many
committees with far too many members each. A further “test” of this lack of
specialization is the relationship between committee membership and the electoral
incentives of legislators. As posited above, organization of legislative institutions
are affected by electoral rules. In those countries where legislators depend on
particular constituencies to survive politically we should expect to see higher degree
of specialization.
<Table 6 Here >
Table 6 presents the index on specialization alongside data on electoral insti-
tutions. In particular, whether ballot access is controlled by parties or is in the
hands of individual candidates (Ballot=1 and 0, respectively). And an aggregate
measure, PARTY, indicating the relative incentives of a given system for a “par-
tisan” vote (it is constructed as the opposite of Hallerberg and Marier’s measure
of “personal vote.” (also shown). It is worth noting that the relationship between
electoral rules and specialization is statistically strong and significant: the index
and the variable PARTY are correlated at .78.
So what/who determines committee assignments? Unlike in the United States,
there is no such thing as a seniority system. Committee and leadership assign-
ments are made on a partisan basis. This partisan composition of committees
usually reflects the relative strengths of the different parties represented in the leg-
islature. Most chambers’ rules state that the composition of each committee shall
proportionally reflect the partisan composition of the bodies as a whole. However,
in some countries there are additional rules to ensure the representation of every
party in the committees. For example, in El Salvador, small parties are assigned
a seat on every committee. Thus, parties have a central role in this process.
20
In the case of committee chairs, in many countries, these are chosen only from
the majority party. In a few cases, such as in Argentina, members from opposi-
tion or minority parties may chair committees. In terms of their tenure, in some
countries, for example Brazil, committee chairmanships rotate on an annual basis,
therefore there are no longstanding powerful committee chairs, as in the US.
With respect to the committees’ technical capacities, these also vary across the
different legislatures. The extent to which a legislature is able to influence the
policy making process depends on its capacity to gather and process information
independent of the executive. It is often agreed that an effective legislature requires
a highly complex structure that is capable of articulating conflict and reconciling
differences, as well as information capacities that will enable it to initiate its own
legislation and to analyze executive’s initiatives. Most studies suggest, thus, that
a legislature’s strength, thus, depends on the possession on a number of critical
resources that will allow its members to perform such tasks. One would expect
that a well-developed committee system would also bring an additional resource,
a competent staff, for the legislature. Unfortunately may be also be true as the
quantity and quality of committee staff tend to reflect the relative importance of
the committee system to the legislative institution. Arrangements for committee
staff also vary greatly from one legislature to another. Some legislatures have no
professional committee staff, other have large, qualified staff who are accorded sub-
stantial responsibility. In general, most committees have a secretariat that assists
it with its functions. These secretariat functions include administrative, research
and document preparation. The scope and range of this administrative assistance
vary widely from one country to another. For example, in Argentina each commit-
tee has access to a secretary, an administrative secretary and two clerical assistants.
However these personnel only perform administrative functions. In contrast, in El
Salvador each committee has only 1 technical assistant and 1 secretary, but they
perform all three secretarial functions.
A study of the legislatures of Argentina, Bolivia, Brazil, Chile and Honduras
conducted by Rundquist and Wellborn in 1994 concluded that the committees of
these countries lacked skilled staff. The authors found that most committees had
a single nonpartisan staff professional employed by the secretariat, supplemented
by party-employed staff controlled by the chamber or committee party leadership
(Rundquist and Wellborn 1993). Their assessment of the Argentine congress is
gruesome: “rarely do committee (partisan and non-partisan) staffs number more
than a half dozen.” (Rundquist and Wellborn 1993: 393). Wynia (1995) paints a
21
very similar picture: argentine legislators keep large congressional staffs, however
these staff members are “under-prepared and inadequately financed” to conduct
research on the executive’s proposals (Wynia 1995: 79). In the case of Mexico,
according to Camp (1995), legislators are bereft of professional assistance and are
staffed only by secretaries. Therefore, they usually rely on the executive branch
for information, interpretations, and policy recommendations, rather than on their
own or independent sources (Camp 1995: 24).
The reliance on executive agencies’ expertise negatively affects a legislature’s
role in the policymaking process. Unless they develop comparable levels of tech-
nical capacity, legislatures are at a disadvantage in evaluating executive-initiated
bills. They are also at a disadvantage when it comes to holding the executive
branch accountable. Usually besides their role in the lawmaking process, com-
mittees also play a critical function in the legislative oversight of the executive.
Most oversight activities take place within permanent committees or in some cases
committees specially assembled to that effect.
Technical limitations aside, in spite of the importance of providing effective
oversight, legislators may also lack the incentives to carry out this responsibil-
ity. Moreover, note that the existence of these kind of resources is “endogenous”
to how much the legislators are motivated to develop them. The institutional-
ization of any legislature starts with the recognition on the part of its members
that they need to spend part of their time and efforts toward building a stronger
collective body. In other words, it is erroneous to think of a legislature’s lack of
capacities (inadequately prepared legislators, lack of staff, inadequately trained
staff) without taking into account political party dynamics and electoral incen-
tives. Human and material resource constraints limit effectiveness of legislators in
performing their mandated tasks and make difficult the practical implementation
of weak and vaguely formulated laws. But improvements in human resources and
physical equipment should be means to specific ends, not ends in themselves. New
computers, for example, should fulfill a specific purpose in helping the legislature
better fulfill its representation, lawmaking, and oversight functions. Demands for
a more active legislature may be initiated from within the legislature. Legislative
leaders may recognize the need for strengthening their institution, and establish
a legislative modernization group responsible for institutional strengthening. On
the other hand, legislative majorities may try to prevent legislative resources from
getting to the minorities and this may conflict with goals like building non-partisan
institutional capacity.
22
The point is that the political context in which the legislature operates plays a
very important role in this story. For example, majority control by the government,
coupled with party loyalty may limit a legislator’s motivation to criticize executive
policies. This last point leads to the role of political parties in the legislative
policymaking process.
4.5 Political Parties
(incomplete)
Partisan politics significantly influence legislative policy-making. Along with
committees, political parties are one of the major means of organizing the work
of the legislature. Parties usually set policy agendas, nominate candidates and
monitor the work of elected representatives. These activities usually take place in
regular meetings of party caucuses. They usually take place once a week and it is
here where party policy is established, party and legislative leaders (including in
some cases committee chairs) are elected, internal party differences are resolved,
strategies for passing legislation or publicizing important matters are developed.
In other words, in most of these countries the overwhelming majority of decisions
are made by party leaders or in party meetings. Moreover, as shown above, most
committee chairpersons and members rotate on a regular basis. Thus, this gives
them little opportunity to develop substantive expertise in the committees juris-
diction relative to the party cadres.
The party system also tends to shape the degree of independence and assertive-
ness a given legislature exhibits. When a single party controls both the executive
and legislature, then the chances for independent legislative decision-making di-
minish. For example, a governing party (or parties) may circumscribe its role in
the legislature to merely transforming government policy into law. This task, of
course, will be subject to a number of constraints: proportion of seats, intra-party
cohesion, the committee system, and the strength of the opposition. Take the case
of Mexico, even though the constitution grants the Congress a substantive role in
the PMP, when the PRI controlled Mexican politics it was hardly an active player.
But when control over the government was divided between the PRI and their
opponents in the House of Deputies, the House became more assertive.
Leaders of ruling and opposition parties can each also be constrained by back-
23
benchers. In addition, coalition governments face the challenge of managing a
relationship among different parties in the coalition. However, from the individual
legislator’s point of view, the costs for challenging one’s party are likely to be very
high - especially in party list systems where members depend on the good graces of
their party for a good position on the list in the next election. When the legislative
career of a legislator depends on his/her placement on the party list, the last thing
he/she may want to do is to engage in activities that challenge the policies and
actions of his/her own party’s government.
5 A Typology of Latin American Legislatures
TO BE WRITTEN.
6 Conclusions
TO BE WRITTEN.
24
Tables and Figures
Table 1. Electronic voting systems in Latin American legislatures
Country Chamber Installed In use? Availability of electr. vote records
Argentina House Pre-1983; 2001 Yes Very Few
Senate Pre-1983 No
Bolivia House 2001 No
Senate 2001 No
Brazil House Pre-1987 Yes Camara dos Deputados Plenario
Senate 1990s Yes
Chile House 1990 Yes In text of Boletin de Sesiones
Senate None No
Colombia House 1996(ish) Occas. Not published.
Senate none No
Costa Rica Unicameral mid-1970s No
Ecuador Unicameral None No
El Salvador Unicameral None No
Guatemala Unicameral None No
Honduras Unicameral None No
Mexico House 1998 Yes Gaceta Parlamentaria
Senate 2002 Yes Senado de la Republica website
Nicaragua Unicameral 2000 Yes Not published
Panama Unicameral
Peru Unicameral 1998 Yes Congreso de la Republica
UruguayeHouse None No
Senate None No
Venezuela Unicameral 1997 No
Source: Carey (2002).
25
Table 2. Conditional Agenda Setting
Country Cond. A-S Override Default Pred. Outcome
Dom. Rep. No 2/3 of Votes Status Quo Bill as originally passed by Congress
Guatemala No 2/3 of Votes Status Quo Bill as originally passed by Congress
Honduras No 2/3 of Votes Status Quo Bill as originally passed by Congress
Colombia No Majority of Members Status Quo Partially Vetoed Bill
Paraguay No Majority of Members Status Quo Partially Vetoed Bill
Panama No 2/3 of Votes Status Quo Partially Vetoed Bill
El Salvador Yes Majority of Members Status Quo Amended Bill
Nicaragua Yes Majority of Members Status Quo Amended Bill
Peru Yes Majority of Members Status Quo Amended Bill
Venezuela Yes Majority of Members Status Quo Amended Bill
Brazil No Majority of Members Partially Vetoed Bill Partially Vetoed Bill
Argentina No 2/3 of Votes Partially Vetoed Bill Partially Vetoed Bill
Ecuador Yes 2/3 of Votes Amended Bill Amended Bill
Uruguay Yes 3/5 of Members Amended Bill Amended Bill
Bolivia Yes 2/3 of Votes Status Quo Amended Bill
Costa Rica Yes 2/3 of Votes Status Quo Amended Bill
Mexico Yes 2/3 of Votes Status Quo Amended Bill
Chile Yes 2/3 of Votes Partially Vetoed Bill Amended Bill
Source: Aleman (2004).
26
Table 3. Distribution of Chief Executive’ Legislative Success, by
Regime Type (in country-years)
Mean Std. Dev. Minimum Maximum Obs.
Parliamentarism 81.76 12.46 34 100 314
Mixed 76.08 18.01 22.2 98 57
Presidentialism
(excluding Lat. Am.) 70.22 16.49 36.2 93.1 61
27
Table 4. Legislative Production
Argentina, Honduras, and Paraguay
Government Government Private Member Private Member Total
Bills Laws Bills Laws Laws
Argentina (1983-1997) 1485 887 21196 1015 1902
Honduras (1990-1995) 1150 935 708 305 1240
Paraguay (1992-2003) 1348 1009 3177 1486 2495
28
Table 5. Committee Membership Rules
Size Leg. No. Comm. Size Committees Multiple Member Rest.
Min. Max
Argentina 257 45 15 31,35,40,45 No
Brazil 513 20 0.03 0.12 No
Chile 120 19 13 13 No
Costa Rica 57 16 5/9 5/9 No
Ecuador 100 18 7 7 Yes(2)
Guatemala 158 35 7 11/15 No
Mexico 500 42 30 Yes(3)
Panama 78 20 7 7 No
Paraguay 80 25 6 No
Peru 121 24 Yes(5)
Uruguay 99 16 No
Venezuela 165 15 7 25 No
Honduras 128 48 3 7
El Salvador 84 18
Colombia 166 7 Fixed Fixed
Nicaragua 92 21 Yes(2)
29
Table 6. Committee Specialization and Electoral Rules
RANK BALLOT POOL VOTE PERSONAL PARTY
Paraguay 1.28 1 1 1 0.08 0.92
Guatemala 0.805 0 0 1 0.24 0.76
Argentina 0.41 1 1 1 0.11 0.89
Panama 0.07 0 0 0 0.39 0.61
Uruguay 0 0 0 0 0.38 0.62
Costa Rica -0.08 1 1 1 0.11 0.89
Chile -0.11 1 1 1 0.43 0.57
Ecuador -0.37 1 1 1 0.4 0.6
Venezuela -0.485 1 1 1 0.23 0.77
Peru -0.56 1 0 1 0.5 0.5
Mexico -0.57 0.608611 0 0 0.17 0.83
Brazil -0.635 0 1 0 0.62 0.38
30
Figure 1: Reelection Rates
Reelection Rate
0 10 20 30 40 50 60
Mexico(1934-1997)
Argentina (1983-2003)
Ecuador (1996-2002)
Venezuela (1958-2000)
Paraguay (1998-2003)
Average
Brazil (1985-2002)
Colombia (1994-2002)
Panama (1999)
Uruguay (1989-1999)
Reelection Rate
31
Figure 2: Distribution of Latin American Presidents’ Legislative Success (in
country-years)
0 .05 .1
Fraction
0 20 40 60 80 100
LEGEFF
32
Figure 3: Ratio Committee/Size
Ratio COMM/SIZE
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Honduras
Paraguay
Panama
Guatemala
Nicaragua
El Salvador
Peru
Ecuador
Argentina
Average (LatAm)
Uruguay
Chile
Costa Rica
Bolivia
Venezuela
Mexico
Average (Non-LatAm.)
Colombia
Brazil
Ratio COMM/SIZE
33
Figure 4: Committee Specialization
SPECIAL.
-1.5 -1 -0.5 0 0.5 1
Colombia
Brazil
Peru
Venezuela
Ecuador
Chile
Costa Rica
Nicaragua
Uruguay
Panama
Argentina
El Salvador
Honduras
Guatemala
Paraguay
34
