Tsebelis G. Veto Players: How Political Institutions Work

Подождите немного. Документ загружается.

321

FIGURE 9.2

“Executive

dominance“

Agenda

setting

Veto

Players

Ch. 7

Ch. 10

Lijphart

Causal relationshipsbetween VPs,governmentagendasetting,

government duration, and “executive dominance”

FIGURE 9.1

Government

duration

322

CHAPTER 10: JUDICIARY AND BUREUCRACIES

In the introduction I connected the legislative game and the capacity of political actors to

change the status quo with the independence and significance of the judiciary and bureaucracy in

a country. The reasoning was simple: both the judiciary (when making statutory interpretations)

and the bureaucracies can be legislatively overruled if they make choices the (legislative) veto

players disagree with, so they are likely to avoid such choices. In fact, both the judiciary and the

bureaucracy will try to interpret the law according to their point of view (interests?) while

eliminating the possibility that they will be overruled. So, high policy stability will give more

discretion to both bureaucrats and judges.

In game theoretic terms I describe a sequential game were the bureaucrats or judges make

the first move (interpret the existing laws) and the veto players make the second (decide to

overrule or not and how). This description can be found in the literature quite frequently.

128

I am

discussing only the mechanism of legislative overrule, and I am not addressing other factors (like

length of tenure) that presumably also affect independence.

In this chapter my goal is to discuss this literature and present empirical evidence

corroborating the expectations. It should be noted that we are in the beginning of the empirical

search, and we have advanced more on judges in a comparative perspective than on bureaucrats.

So, the empirical analysis will depend either on indicators developed as proxies, or assessments

by experts which sometimes turn out to be conflicting. My presentation will be in three parts: the

first presenting the decision-making problem of the first mover when he chooses with the

possibility of being overruled, the second applying the model to the judiciary, and the third to

bureaucracies.

128

Gely and Spiller (1990), Mikva and Bleich (1991), Ferejohn and Weingast (1992a, 1992b), Eskridge (1991)

Cooter and Drexl (1994) and Bednar, Ferejohn and Garrett (1996) about judges; McNollgast (1989), Hammond

323

1. How to Avoid Legislative Overrule.

Let us assume that there are three legislative veto players. The triangle 123 that they

define is their core, that is, the set of points that they cannot agree to change. Consequently, if the

first mover selects one of the points of the core, there will be no legislative overrule. Figure 10.1

presents three different possibilities. In the first two cases the first movers’ ideal points J and K

are outside the legislative core and they select the closest core point to them (J’ and K’

respectively). Despite the fact that these two choices are significantly different from each other,

the veto players are incapable of changing either of them. In the third case, the first mover is

located inside the legislative core but changes her mind and moves from point L1 to point L2.

Since the first mover is inside the core she can select her own ideal point.

INSERT FIGURE 10.1 HERE

These idealized stories are close to political realities. Think about the following cases:

In the United States (a country with three veto players) the Supreme Court decided on several

extremely important issues that in most other countries would have been the prerogative of the

legislative branch. Desegregation and choice come immediately to mind. But in the (near?)

future tobacco and guns may join the list of political decisions delegated to courts because the

political system is unable to legislate on the issue.

As an example of the change in mind of the first mover consider the issue of sexual

harassment where the burden of proof requirements changed. In the past she needed to show that

as a result of the behavior of a superior or co-worker she was very disturbed, she lost days of

work, she visited doctors etc. After the Harris v. Forklift Systems (1993)

129

court decision any

behavior that would have disturbed an average person was defined as sexual harassment.

(1996), Hammond and Knott (1999) about bureaucrats.

129

I thank Eugene Volokh for the reference.

324

However, the model I present here is very simple, and the theoretical argument needs to

be buttressed (on top of the supporting stories). The first question is what happens if the veto

players are not individual but collective, the dimensionality of the policy space high, and, as a

consequence, there is no core as in Figure 10.1. Then, no point is invulnerable from a legislative

overrule. Does this mean that the first mover (judiciary or bureaucracy) has no agenda setting

power? Not exactly.

In the case of absence of a legislative core the winset of the status quo is not empty but,

as demonstrated in Chapter 2, it can be quite small for certain positions of the status quo. A

bureaucratic or judicial decision that has such a small winset may not be worth the effort of

legislative overrule. Indeed, there are serious transaction costs for every legislative decision: take

the initiative to present a bill, put together a coalition to support it, eliminate opponents who may

have a different opinion by buying them out, or by solidifying your allies etc. etc. If the

difference between the judicial or bureaucratic decision and the outcome a particular veto player

is likely to obtain is not big enough, such an enterprise may not be worthwhile.

The discussion of existence of the core brings us to another interesting point: the

legislative overrule may require different majorities in which case the legislative core may have

significant size. For example in the US, think of decisions of the Supreme Court that are

constitutional instead of statutory interpretations. After President Clinton signed the Freedom of

Religion Act into law (an act that he had practically initiated and commanded almost unanimous

support in both Houses of Congress), the Supreme Court decided that the law violated the

Constitution because it was legislating in an area that was the jurisdiction of States. All the

proponents of the law changed their course of action and decided to introduce fifty such laws,

one in each state, rather than try to modify the law. The reason is that very few people think that

325

a modification of a Supreme Court Constitutional decision can come in any other way but from a

change in the Court’s collective mind (think of Roe vs. Wade the only way to change it is either

to wait for the Court to change its mind, or with a constitutional amendment).

130

There are two more objections concerning the above simple game theoretic account

raised in the literature. The first is that given that the first movers in the game presented above

will be able to select a policy close or identical to their own ideal point, what will the legislative

branch do to prevent this event from materializing? There is extensive literature arguing that

legislation will be more restrictive when there are many veto players (McCubbins, Noll and

Weingast (1987), (1989), Moe (1990), Moe and Caldwell (1994), Epstein and O’Halloran (1999)

etc). I will demonstrate that my analysis only appears to contradict this argument. The second is

that there may be significant differences between parliamentary and presidential systems with

respect to delegation of powers: presidential veto players some of the literature argues (Moe

(1990), Moe and Caldwell (1994), Strom (2000), Ackerman (2000)) have explicit assignments to

oversee the bureaucracy, while parliamentary veto players practice oversight collectively. As a

result, this view contends that political systems differ from each other not because of the number

of veto players but because of regime type.

It is interesting to note that these objections have been raised only about bureaucrats and

not about judges. To my knowledge while many American researchers have made the argument

that more detailed legislation is designed to restrict the role of bureaucrats, none has made the

same argument about the role of judges.

131

Similarly, the difference between presidentialism and

130

I will make the point that Constitutional interpretation may turn the Court into an additional veto player in the

following section.

131

For a European exception see Fritz Scharpf (1970) who has made precisely this argument about the German

legislature. His point is that German law is very detailed for a series of resons among which the restraint of judges

(who decide on both procedural and substantive grounds) as well as the restraint of state bureaucracies (who are

independent from the federal government). In Scharpf’s analysis American courts do not have substantive review of

bureucratic decisions, and there are federal bureaucracies. These differences may account for the absence of the

326

parliamentarism has appeared in the literature on bureaucracies and not in the literature about the

judiciary. Given that my presentation on bureaucracies and the judiciary was symmetric, I do not

know the reason for this differential treatment in the literature. However, I will respect it, and

address these points in the third section on bureaucracies.

II. Veto players and the Judiciary

a. Traditional Theories of the Judiciary. The usual distinction in comparative

law is in common law and civil law tradition countries. In common law countries (UK, and all its

ex-colonies like US, Australia, New Zealand, Ireland, Malta etc) "laws" are seen less as the acts

of parliament and more as the accumulation of decisions and interpretations of the judges. The

central rule in common law countries is stare decisis (let the decision stand), the Latin words for

the importance of precedent. Decisions made by previous judges in similar cases are binding for

a judge. As a result, judges create the law as well as apply and interpret it.

In countries that follow the civil law tradition, the foundation of law is a comprehensive

and authoritative legal code. It is upon this code that legislatures build a superstructure of

statutes. The most frequently used of such codes is the Napoleonic Code (used in France,

Belgium, Greece, Luxembourg, Netherlands, Italy, Spain, and Portugal). A second such code is

the German Civil Code (used in Germany, Norway, Sweden, Finland, Denmark, Iceland). In

civil law countries the judges interpret the law, they do not make it.

According to this classic distinction, the role of the judiciary should be more important in

common law countries. However, more recent analyses indicate convergence of the two systems.

Gallagher, Laver and Mair (1995: 62) cite a series of authors who explicitly discuss convergence

argument that detailed legislation restricts judges in the US.

327

(Waltman and Holland (1988: 85)), or describe the behavior of the judiciary in civil law systems

in terms of “precedent”, and discuss “statutes” in common law systems.

Convergence theories agree with the analysis in the first section of this chapter.

According to the veto players theory what matters for the independence and significance of the

judiciary is not the legal system of a country, but whether courts are constitutional or not and the

difficulty of the political system to overrule a statutory or constitutional interpretation.

We already discussed the issue of statutory interpretation as a matter of policy stability of

the corresponding political system; let us now focus on the question of constitutional

interpretation. This is a major issue, because if courts can interpret the constitution and base their

decisions on it, they cannot be overruled by the political system. The only exception would be by

a modification of the constitution, which would mean that the judiciary of a country would be a

veto player, since a decision by the judiciary could invalidate a law.

b. Are judges veto players? While it is clear that the judiciary of a country is not

a veto player when performing statutory interpretations because it can be overruled by

legislation, the opposite is true with respect to constitutional interpretation. Indeed, a rejection by

a constitutional court is sufficient to abrogate legislation approved by the legislature. Some

countries like France have a priori and abstract review of legislation, invalidating laws on

constitutional grounds before they are applied. In this case, the Conseil Constitutionel of France

acts as an additional chamber of Parliament and can abort whole pieces of legislation or parts of

them just before this legislation is signed into law by the President of the Republic (Stone 1992).

Stone Sweet (2000) argues that the introduction of scrutiny by constitutional courts has

profoundly altered the role of both courts and legislatures, and has introduced a constant

interaction between the two institutions. According to this interaction the legislatures are always

328

aware that their actions can be overruled by constitutional courts, and sometimes even ask the

courts for instructions in order to immunize their decisions from judicial abortion. According to

Stone, as courts become increasingly more elaborate in different areas, the discretion of

legislatures is reduced. As a result, we are in the process of the formation of a government of

judges. Volcansek (2001) makes similar arguments in the Italian case, and explains how and why

the Italian constitutional court has made important interventions in the questions of divorce and

executive decrees.

The essence of these arguments is correct: constitutional courts can abort legislation, and

consequently they are veto players. However, the conclusions and predictions about governments

of judges seem exaggerated. Why? My answer is based on the discussion of the absorption rule

in Chapter 1. While constitutional judges are veto players, most of the time they are absorbed.

As we discussed in Chapter 1, for a veto player to make a policy difference it has to be

located outside the unanimity core of the other existing veto players (see Proposition 1.3). I will

argue that Constitutional Courts very often are located inside the unanimity core of the other veto

players. The main reason is the appointment process to the highest positions. The only major

country without any restrictions to a purely politicized appointment process is France, where the

nine members of the Conseil Constitutionel are appointed by the President of the Republic

(three), the President of the National Assembly (three) and the President of the Senate

(three)without any specified qualifications or approvals.

In the US the President’s nominees have to be approved by the Senate. In Italy one third

of the members of the Constitutional Court (five) are appointed by the President of the Republic,

one-third by the Judiciary, and one-third by the Parliament by a 2/3 majority in a joint session of

the Chamber of Deputies and the Senate; all appointments have to be judges with 20 years

329

experience or tenured law professors. In Germany eight members are elected by the Bundestag,

and the other eight by the Bundesrat with 2/3 majority; all members must be qualified to be

federal judges (and 6/16 must actually be federal judges). In Spain two of the 12 judges are

appointed by the Government, two by the judiciary, and four by each chamber of Parliament by

3/5 majority; their judicial competence must be well known (Stone Sweet 2000: 49).

The restrictions imposed upon the selection of members of the highest institution of the

judiciary eliminate extreme positions, and practically guarantee that the median of the court will

be centrally located in the policy space. However, the means used by American political

scientists and judicial scholars to study the US Supreme Court are unavailable for other courts

because deliberations are secret and most of the time dissents (if any) remain unknown and

certainly are not signed and published.

The above account generates another question: how is it possible for constitutional courts

to ever veto legislation under these conditions? That is, under what conditions is the median of

the supreme court not included in the unanimity core of the existing veto players? I will provide

two plausible answers to this question, and remind the reader that plausibility is the most one can

expect unless the black box of judicial deliberations opens.

First, judges are selected for competence and for their (known) policy positions. Some of

their positions may be unknown because they have not deliberated on every issue, and some

positions may be considered secondary so that they are not subject to litmus tests. One of these

issues may become important someday and controversial, like the issue of gay marriages or

euthanasia, and on this issue the Supreme Court may be in disagreement with the policymakers,

but this was not the criterion of selection in the past. In fact, this is likely to be a new dimension

that cuts across party lines.

330

Second, a veto by the Supreme Court should not necessarily be considered as opposition

to government action. It may be the expression of procedural preferences (Rose-Ackerman

(1990) and Ferejohn and Weingast (1992a)), like the introduction of technical restrictions. The

court may be indicating to the government that this particular way of reaching its goal violates

the constitution, so a different course of action is necessary. Stone (1992) provides several

examples when the French Parliament asked the Constitutional Council to provide specific

wording so that legislation would survive the Court’s Constitutional scrutiny. Finally, courts

through their own interpretation of the law might also provoke new more desirable legislation.

This is the way Van Hees and Steunenberg (2000) explain the famous decision by the Dutch

Supreme Court permitting euthanasia under the specific conditions that occurred in the case

under review, which then provoked additional legislation on the issue.

In conclusion, judges are not veto players when they take statutory decisions. They are

veto players when they make constitutional interpretations, but most of the time they are

absorbed by the existing political veto players. The only exceptions would be if the existing veto

players are located in extreme policy positions (the case of France under the first Mitterrand

government (1981) comes to mind where the government wanted to implement a series of

significant policy changes while the constitutional court had been appointed by right wing

governments), or if new issues come under consideration. In these cases constitutional courts

should be counted as additional veto players. However, given the black box that contains

decision-making by the judiciary it would be impossible to attribute to this veto player policy

positions. As a result, I have not included the judiciary (not even in the form of constitutional

courts) as a veto player into my accounts in this book.

c. Empirical evidence. If the judiciary is not an additional veto player, then we can