There are many definitions of judicial independence. This chapter explains why I use only one of them – an independent judiciary delivers decisions that do not consistently reflect the preferences of a particular group of actors. The emphasis is on consistently because each decision will reflect somebody’s preferences – the winner’s. The key is that when courts are independent, no one can expect to win all the cases that they would like to win. The chapter also proposes that we should be aware that independent judicial output can result either when actors are unwilling or incapable of imposing their preferences in court.

Institutional Judicial Independence

In institutional terms, judicial independence refers to the insulation or autonomy of the judiciary from the other branches of government or from the public (Fiss, Reference Fiss1993; Ferejohn, Reference Ferejohn1999; Russell & O’Brien, Reference O’Brien, Ohkoshi, Russell and O’Brien2001; Chavez, Reference Chavez2004; Finkel, Reference Finkel2004). Institutional judicial independence is higher when there are more structural safeguards against interference by nonjudicial actors in judicial decision making. Institutional shields include life tenure guarantees for judges, judicial control over appointment, promotion and dismissal, as well as the drafting and administration of the judiciary’s budget. Institutional insulation definitions of judicial independence are particularly prevalent in policy analysis of levels of the rule of law around the world. For example, in its annual reports on the East European states’ progress toward meeting the accession criteria, the EU has consistently criticized certain candidate states for providing institutional opportunities for executive interference in judicial affairs, such as vesting appointment powers with a political actor, rather than a corporate judicial body, such as a Judicial Council (Open Society Institute, 2001; ABA/CEELI, 2002a).

Focusing on institutional guarantees against undue influence on judicial decision making may seem practical because identifying these structural shields is a fairly straightforward measurement task. However, as Ferejohn (Reference Ferejohn1999) points out, institutional judicial independence is not intrinsically valuable (p. 353). The desirability of independent courts hinges squarely on the assumption that the institutional shields prevent external actors from influencing judicial output, which in turn ensures equal responsibility and protection under the law (i.e., rule of law) and provides stable protection of property rights (i.e., promotes economic growth).

However, there is mounting empirical evidence from around the world that makes this assumption problematic. Clark (Reference Clark1975) devises a “judicial effectiveness score,” which measures the institutional independence of Latin American Supreme Courts. His measure includes indicators such as tenure guarantees, method of appointment and removal, and salary guarantees for each court. However, he is surprised to find that courts generally perceived as less independent (the Argentinean and the Brazilian ones) score much higher than seemingly more independent ones (such as the Costa Rican one). Hilbink (2007) and Helmke and Rosenbluth (2009) also note that during the tenure of the Latin American military dictatorships (Chile, Brazil, and Argentina are the main cases), institutionally independent judiciaries failed to constrain the military juntas and protect human rights that were officially guaranteed by the countries’ constitutions. In the Asian context, Ramseyer and Rosenbluth (1993), Ramseyer and Rasmussen (Reference Ramseyer and Rasmusen1997), and O’Brien and Ohkoshi (2001) have shown that despite the standard institutional safeguards against judicial dependence on the executive, Japanese judges are rather deferential toward incumbent politicians. Finally, in the European context, Guarnieri (Reference Guarnieri, Russell and O’Brien2001) argues that the institutional autonomy of the Italian judiciary from the executive has led to judicial bias in favor of the political opposition and undue influence of the media and the mafia on individual judges. Smithey and Ishiyama (Reference Smithey and Ishiyama2000) build a “judicial power” indicator for postcommunist high courts, which also suggests that institutional independence does not co-vary with behavioral independence. According to their index, which includes variables such as judges’ terms relative to those of other political actors, conditions for judicial removal, and number of actors involved in the selection of judges, Poland, Estonia, the Czech Republic, Hungary, and Slovenia have much weaker courts than Armenia, Romania, Bulgaria, and Moldova. Judicial assessments conducted by the EU, the World Bank, and the Open Society Institute, however, suggest that the courts in the first group of countries exhibit much more independent behavior than their counterparts in the second group. Herron and Randazzo (Reference Herron and Randazzo2003) test the hypothesis that postcommunist high courts that are institutionally more insulated from the executive will be more likely to exhibit independent behavior and find that formal independence is a significant but negative predictor of independent behavior.

One way to interpret this evidence is to conclude following Kornhauser (Reference Kornhauser, Burbank and Friedman2002) that judicial independence is a useless concept, which should be abandoned, because we cannot show that it is either necessary or sufficient for the rule of law, economic prosperity, or good governance – the inherently desirable ends, which justify the need to promote and study judicial independence in the first place. Alternatively, this evidence could simply convince us that we should stop using the institutional setup of the judiciary as a measure for its independence. Rather the institutional configuration of the judiciary should be treated as an independent variable in the analysis of judicial independence. In other words, institutional set up could be used to try to explain why some courts are independent and others are not.

Behavioral Judicial Independence

An alternative approach to institutional definitions is a judicial behavior conceptualization of judicial independence. If judicial behavior is the focus of interest, then the relevant unit of analysis becomes the individual judge (e.g., Becker, Reference Becker1970; Salzberger, Reference Salzberger1993; Ramseyer, Reference Ramseyer1994, Reference Ramseyer and Rasmusen1997; Ramseyer & Rasmusen, Reference Ramseyer2001a, Reference Ramseyer2001b). We could compare judiciaries according to the percentage of judges who generally adjudicate cases independently. Becker (Reference Becker1970), in an often-quoted definition, argues that independent judges are those who believe that they can and do decide cases consistently with their interpretation of the law rather than that of any other actor. In particular, independent judges follow this strategy even when a decision averse to the desires of other power holders could possibly result in retribution either on the judges personally or on the power of their court. According to this definition, a country has high judicial independence if a large percentage of its judges both feel independent and act independently¹.

The first drawback of the behavioral definition stems from the tension between feelings of independence and independent actions. It seems that the two things do not necessarily go together. In particular, how would we classify cases, in which judges do not feel independent, that is they know that ruling against the regime could and probably would lead to retribution, either against them or against the court, but they act independently nonetheless? What is the level of judicial independence in these cases?

Consider the historic Orange Revolution case, in which the Ukrainian Supreme Court canceled the results of the presidential election runoff and ordered a rerun. Retribution from the incumbent Kuchma regime, whose selected candidate, Viktor Yanukovych would have become president, had the Supreme Court not canceled the Central Election Commission decision, was certainly possible. In fact, a judge from the Supreme Court explained that his colleagues were acutely aware that the Kuchma regime would try to coerce or coax them into delivering a pro-Yanukovych decision. As a result, the twenty judges hearing the case decided to isolate themselves completely from the outside world during deliberations in order to thwart a possible divide-and-rule strategy by the regime, in which individual judges may have been targeted for pressure (A. V., 2004). In other words, the judges clearly did not feel independent from the regime, but they did rule against it. Does the ruling indicate behavioral judicial independence or not?

In addition, reliable data on how judges feel seems elusive. Presumably, we could use surveys to get judges to tell us whether they feel independent or not, but even sincere judges themselves might not be able to distinguish between their true interpretation of the law and the myriad of influences from a variety of sources that might affect their judgment. Finally, we would have to come up with an inherently arbitrary threshold to distinguish between an independent judge and a dependent judge. Should judges who feel or rule independently 51 percent of the time qualify as independent judges?

Even if we could reliably measure behavioral judicial independence, we run into a normative problem. Behavioral judicial independence might not be desirable at all from a rule of law perspective. Judges, like the rest of us, have complex personalities and diverse ideologies. One possibility is that given complete decisional independence, they may completely disregard the law and impose their ideological preferences on the rest of society. As the attitudinal model of decision making at the U.S. Supreme Court has shown, judges’ personal ideology plays a significant role in their decision making process (Segal & Spaeth, Reference Segal and Spaeth1993). It seems that in a system characterized by highly independent judges, equal responsibility and protection under the law can no longer be universally guaranteed because the outcome of individual cases would depend greatly on the personality and biases of the presiding judge. Most scholars in fact agree that granting sweeping decisional independence to individual judges is undesirable, because as fallible human beings they should not be fully unaccountable for their decisions (Burbank & Friedman, Reference Burbank and Friedman2002).

Decisional Judicial Independence

Decisional judicial independence exists if no actor can consistently secure judgments that are in line with his or her preferences. Quite a few studies conceptualize judicial independence as a feature of judicial output (Landes & Posner, Reference Landes and Posner1975; Shapiro, Reference Shapiro1981; Rosenberg, Reference Rosenberg1992; Hanssen, Reference Hanssen2004). For example, Shapiro’s often-cited definition of judicial independence as the “inability of government to influence the outcome of any individual case” points to individual cases as the appropriate unit of analysis (Shapiro, Reference Shapiro1981, p. 8). Rosenberg also suggests looking at individual decisions to see whether they reflect the preferences of elected officials (Rosenberg, Reference Rosenberg1992, p. 373).

“Consistently” is the key word in the definition. It means that we need to look at the outcomes of multiple cases involving a given actor in order to determine reliably whether the courts are independent from that actor or not. It seems inherently impossible to determine on an individual case basis whether a ruling reflects the judge’s own interpretation of the law or someone else’s. Each case has a winner and a loser, and we have no way of knowing whether the judge sides with the winner due to a good faith belief that the winner’s case was stronger or out of a fear of retribution. Power differences between the litigants might help to clarify things somewhat. For example, when ordinary citizens win cases against powerful state agencies, it is unlikely that the judge was dependent on the citizen. However, powerful state actors surely have meritorious cases too, so a victory by such actors does not automatically mean that the court lacks independence. Also, how do we interpret the outcomes of individual cases, where both litigants are potential principals on whom the court could be dependent?

To go back to the Orange Revolution case – the Ukrainian Supreme Court judges had to choose whether to side with the million-strong crowd gathered under their windows or with a regime that had shown readiness to go after judges who rule against it². Both actors, at the moment that the judges were deliberating, had significant, albeit different, sources of power vis-à-vis the court. Or, what should we make of the decision in a defamation case, which pits a powerful incumbent politician against a newspaper, owned by a politically active oligarch. Both Russian and Ukrainian courts heard numerous cases like that where we would not be able to glean the level of judicial independence from just looking at the outcome of an individual case.

This caveat can be overcome by systematic analysis of many, similar cases. I propose that we should define decisional judicial independence in terms of the number and the identity of the actors who do not exercise power over the adjudication process, as well as in terms of the magnitude of the power of the actors who do manage to impose their preferences. According to this definition, judicial independence would be low if pressure is exerted routinely in a vast majority of cases and multiple actors can impose their preferences over judicial outcomes. Conversely, judicial independence would be high if fewer actors have power over the outcome and only in a few cases.

Decisional judicial independence seems preferable to behavioral and institutional definitions because it is certainly a necessary component of the rule of law. The emphasis of the rule of law doctrine on equal protection and responsibility for all under publicly promulgated laws essentially calls for establishing a legal system that would guarantee that the preferences embodied in publicly promulgated laws would be consistently imposed on individual case outcomes. In addition, if no one actor can consistently impose his or her preferences on the court, this would be an indication that judicial output reflects the “true” preferences of the law.

Independent from Whom? Judicial Independence from Different Principals

However it is defined, judicial independence is a relational concept so we should always specify the potential source of dependence (Russell & O’Brien, Reference O’Brien, Ohkoshi, Russell and O’Brien2001). More normative studies should explicitly address the question of which type of judicial independence is most desirable – independence from politicians, from the public, from interest groups, from the presiding judge’s personal ideology, from the judges sitting at the higher courts, from litigants, and so on. The emphasis in many judicial independence definitions on minimizing meddling by politicians in judicial decision making suggests that independence from politicians is at the top of the desirability list. In addition, scholars have pointed to the drawbacks of judicial dependence on public opinion or on individual litigants (Shapiro, Reference Shapiro1981; Burbank & Friedman, Reference Burbank and Friedman2002).

Once we recognize this additional dimension of variation, the next question that arises is whether and how independence from different external actors (i.e., different principals) might vary within the same judicial system at any given point in time. We should investigate whether different types of independence always go together, in which case we can talk about a general level of judicial independence. Or there might be a trade-off – maximizing independence from one actor may automatically open the door for dependence on another actor. For example, we could imagine a judiciary that is simultaneously highly independent from politicians (i.e., politicians cannot consistently impose their preferences on judicial output) and highly dependent vis-à-vis organized crime (i.e., rich mobsters can get favorable rulings in most or all cases that interest them).

In addition, both the degree and the source of judicial dependence may vary across different legal issue areas within the same judicial system. By a legal issue area, I mean the universe of cases litigated with reference to a specific law or group of laws. The area could be as narrow as defamation law involving media outlets as defendants or litigation arising from privatization deals, or as wide as criminal law or civil law. The crucial point is that different issue areas have a different set of actors with preferences about judicial outcomes. For example, the executive might not have strong enough preferences to attempt to secure favorable judicial decisions in disputes that are not politically salient.

Capacity Vs. Willingness Conceptualization of Judicial Independence

The last example raises the question: Should we always define judicial output that does not systematically reflect the preferences of a given principal as judicial independence? Does it matter whether the principal cannot or would not impose his or her preferences on the outcomes of cases that interest him or her? If the principal seems unwilling to interfere in judicial decision making, does it matter whether the restraint is due to strong belief in the rule of law doctrine or to a self-interested rational calculation that interference would be too costly? I argue that the definition of judicial independence should incorporate all three scenarios, but scholars should be cognizant of the distinctions between them.

Most studies of judicial independence do not touch upon this issue at all. Some scholars talk about the government’s inability, rather than unwillingness to influence judicial decision making (Shapiro, Reference Shapiro1981; Russell & O’Brien, Reference O’Brien, Ohkoshi, Russell and O’Brien2001; Chavez, Reference Chavez2004). Others discuss circumstances under which politicians provide or offer independent courts, which implies that politicians in such cases refrain from pressuring the courts to deliver favorable rulings (Salzberger, Reference Salzberger1993; Ramseyer, Reference Ramseyer and Rasmusen1997; Hirschl, Reference Hirschl2000; Stephenson, Reference Stephenson2003). Often times, though, the implications seem to be the unintended result of particular word choice, rather than explicit assumptions or intentional definitional choices.

The distinction may be consequential, however, so it deserves some attention. For the purposes of clarity, I will discuss politicians, who are probably the principal whose influence on the courts is most consequential for the rule of law. In the capacity conceptualization, we can categorize judicial output as independent only when politicians simply cannot influence judicial output even if they tried. The willingness conceptualization views judicial independence as the result of a conscious choice by politicians to not interfere in judicial decision making; they have the capacity to influence judicial output but refrain from exercising this power. The willingness conceptualization itself can be divided in two scenarios. First, politicians may be unwilling to lean on the courts as a result of a strategic calculation that it is not in their best interest to try exerting pressure – either because the potential costs of doing it are too high or because the potential benefits of doing it are insignificant. Alternatively, politicians may be unwilling to pressure the courts for ideological reasons (i.e., they may espouse a strong belief in the rule of law and judicial independence and as a result would not even consider pressuring the courts).

The three conceptualizations can be hard to distinguish observationally. Even when politicians tout their staunch commitment to independent courts and, indeed, refrain from attacking the courts to get their way, we cannot be sure of their sincerity. It could be that they forego applying pressure because it is too costly or not beneficial enough. Surveys about politicians’ attitudes and commitment to the rule of law are unlikely to be productive. The billions of dollars spent around the world on rule-of-law promotion have probably ensured that virtually any politician around the world knows the “right” answer to questions about judicial independence and the rule of law.

The distinction between capacity and willingness is also tricky to observe. When politicians make obvious attempts to pressure the courts, but judicial output does not reflect their preferences, it is clear that the principals simply cannot impose their will on the courts. But such failed attempts at interference are not likely to be sustained practice and we rarely see them. Politicians at some point either figure out a more effective way to pressure the courts or give up trying to interfere. When politicians do not engage in obvious influence peddling, determining whether the restraint is due to a realization that pressure would be futile or to a calculation that the costs of interference are too high might be quite hard. Rather than look at the actual judicial output, we would then need to look closely at the behavior of politicians and at the existence of informal or formal institutional channels, through which they could affect judicial decision making. For example, throughout its EU accession process, the Czech Republic was repeatedly criticized for formal institutions, which facilitate potential executive pressure on the courts. The EU monitoring reports, however, never provided evidence or even claimed that the executive actually took advantage of these potential institutional channels. Thus, this seems like a scenario, which allows us to distinguish observationally between judicial independence due to lack of willingness, rather than lack of capacity by politicians to subordinate the courts.

At first glance, the distinction between willingness and capacity seems irrelevant to the relationship between judicial independence and the rule of law and the desirable phenomena that follow. Whether politicians do not interfere because they cannot or because it is too costly to them, the politicians would not be receiving preferential treatment. As a result, all actors involved in the adjudication process will expect litigant equality. Therefore, we would conclude that in both cases the courts are de facto independent and the rule-of-law doctrine is in place.

Within the willingness conceptualization, however, the distinction between “unwilling because of high costs” and “unwilling because of low benefits” seems normatively important. The second scenario seems normatively “inferior” and may create different expectations for all participants in the adjudication process. It is debatable whether we could call the courts independent if the only reason for noninterference by the principal is the lack of motivation on the principal’s part. Consider the following analogy – in this situation the politicians have the courts on an invisible leash; unless they pull it, no one knows how long it is or whether it is even there. In other words, unless politicians choose to exercise their influence on the courts, the courts (and any other actor, for that matter, including the politicians) have no idea how strong that influence is and whether it would get them consistently favorable rulings.

Epistemological conundrums aside, the more important issue is whether this scenario is as likely as the others to promote the rule of law. First, judicial independence that stems from politicians’ apathy appears to be less sustainable in the long run because it is dependent entirely on the incumbents’ decision to impose their preferences or not. In addition, this type of judicial independence might make a smaller contribution to economic growth because presumably societal actors would realize that the judiciary’s ability to withstand pressure has not been tested and this realization might reduce long-term investment. Finally, the fact that neither the courts nor other societal actors are sure of the boundaries of how much judicial independence the politicians would tolerate probably decreases the level of predictability of state action.

Others may argue that even this diminished form of judicial independence is conducive to the rule-of-law project. Independent judicial output which does not systematically reflect the preferences of any principal boosts both the real and the perceived equality of litigants. And equality under the law is what it is – a central component of the rule of law – regardless of how it comes about. I leave the normative issue to be debated by normative theorists. The details of the Russian case will illustrate in the following chapters that we should be aware of the conceptual difference between judicial independence provided by the principal and judicial independence wrested from the principal. Clearly, all three types of judicial independence deserve scholarly attention.

Judicial Independence at Different Levels of Abstraction

A final definitional conundrum is whether the concept of judicial independence should refer to unencumbered decision making at the level of the case or at the level of the rule.³ Scheppele (Reference Scheppele, Burbank and Friedman2002) makes a convincing case that it is hard to distinguish conceptually between judges taking orders in specific cases and strictly following a rule that is so straightforward and so detailed that it predetermines the outcomes of specific cases. In other words, politicians can impose their preferences on judicial output either through “telephone law” when they just call up the judge and demand a specific ruling, or by passing laws that if followed would maximize the politicians’ utility. I argue, however, that the rule of law requires decisional judicial independence at the level of the case, but not necessarily at the level of the rule.

The legal saga involving Italian premier Silvio Berlusconi illustrates the conceptual blurring that Scheppele describes. Before becoming the head of government, Italy’s most powerful tycoon was facing criminal prosecution for false accounting, money laundering, bribery, and tax fraud. During his term in office, his parliamentary faction managed to pass amendments to the criminal code, which decreased the statute of limitations on false accounting and granted the prime minister immunity from prosecution while in office. These measures resulted in the dismissal of some charges and the postponement of others. Other measures, however, such as a proposed amendment to the rules regarding the admissibility of documents obtained from other countries did not pass despite the efforts of Berlusconi’s faction to push them through parliament (Economist Intelligence Unit, 2002). Scheppele basically questions whether there is a fundamental difference between these events and the Soviet experience with “telephone law” in which some judges received direct instructions from the executive on how to adjudicate specific cases.

I argue that the two scenarios are different from the standpoint of the rule-of-law doctrine. “Telephone law” would benefit only Berlusconi, whereas the selective amendments to the criminal code would now benefit all fraud perpetrators. It may seem that the second scenario is even less desirable than the first because it would let more criminals off the hook and contribute to a long-term increase in levels of accounting fraud. However, only the first scenario violates the rule of law’s emphasis on equal responsibility and protection under the law because it gives an advantage in court to one litigant. The second scenario is in line with the tenets of the rule-of-law doctrine because it increases litigant equality under the law.

Another difference is that promulgating laws in a politically competitive environment is costly in terms of time and efforts. Ad hoc interference in specific cases, on the other hand, especially if it is an informally established prerogative of the incumbents, requires virtually no political resources and has no time lag, so it is a much more effective tool for imposing the politicians preferences on individual trial outcomes. We should then expect that dependence at the level of the case would affect a much larger set of cases than dependence at the level of the rule. Since I argue that we should measure decisional judicial independence through the percentage of cases adjudicated independently from a given actor, then dependence on politicians at the level of the case would result in a much lower level of judicial independence than dependence on politicians at the level of the rule.

Conclusion

This chapter argues that the definition of judicial independence should stem from the link between judicial independence and the rule of law. It also argues that decisional judicial independence is necessary for the rule of law, and institutional and behavioral judicial independence are not. Therefore, the book seeks to explain variation in decisional judicial independence, which is defined as judicial output, which does not systematically reflect the preferences of non judicial actors. Under this conceptualization, courts are independent from a given actor if this actor does not consistently impose his or her preferences on trial outcomes. The definition also suggests that analyses of judicial independence should identify the principal whose real or potential influence on the courts is the focus of study. Finally, studies should also specify whether independent judicial output results because politicians (1) lack the capacity to interfere in court decision making; (2) rationally conclude that it is not in their interest to interfere; or (3) espouse a strong ideological commitment to independent courts.