“Conceptualizing and Operationalizing Judicial Independence Globally”
C. Neal Tate
Professor and Chair
Department of Political Science
Vanderbilt University
VU Station B # 351817
Nashville TN 37235-1817
Phone: 615-322-6241 Fax: 615-343-6003 E-mail: n.tate@vanderbilt.edu
and
Linda Camp Keith
Assistant Professor of Political Science
University of Texas at Dallas
Green Hall 3.232
Box 830688
Richardson, TX 75083
Phone: 972-883-6481 Fax: 972-883-6297 E-mail: lck016000@utdallas.edu
INTRODUCTION
Our initial interest in judicial independence grew out of the increasing attention given to
the concept by international lawyers, academics and aid organizations interested in the
rule of law, democratization, and/or states’ human rights behavior. The association
between an independent judiciary and rights protection is not a new one. Alexis de
Tocqueville in his Democracy in America noted that an independent judiciary
empowered with judicial review was "one of the most powerful barriers erected against
the tyranny of political assemblies" (1966, 261). The current focus on judicial
independence and rights protection has largely stemmed from the renewed emphasis on
constitutionalism in the democratizing world of the post Cold War era. As Russell
(2001) notes, within the general trend toward liberal democracy in the world today,
judicial independence is viewed as an “essential feature of liberal democracy” (2). This
view is accepted by many scholars, such as Ackerman (1991) who argues that an
independent judiciary is “the ultimate guarantor of constitutionalism” and by
organizations such as the American Bar Association whose Code of Judicial Conduct
states that an “independent and honorable judiciary is indispensable to justice in our
society” (as cited in Ramseyer 1994, 721). Larkins (1996) goes even farther, arguing that
judicial independence is “not meaningful if the courts cannot exercise it to check the
arbitrary or unjust exercise of power by political and social actors” (611). Additionally
the United States Agency for International Development (USAID 2002) links U.S.
democracy and governance assistance to the promotion of rule of law principles, which
include respect for human rights protected in part through judicial independence. These
expectations are based on the assumption that a truly independent judiciary should be
1
able to withstand incursions upon fundamental rights because 1) the courts’ power and
fiscal well-being are protected, 2) the courts have some ability to review the actions of
other agencies of government, and 3) the judges’ jobs are constitutionally protected.
These expectations are not entirely universal; there are skeptics (especially in
regard to Latin American and African courts) who doubt that judges on the bench will be
inclined to or be able to challenge a dictatorial regime1 or who fear that an overly
independent judiciary can be as abusive in its exercise of power as the legislative or
executive branches (e.g. Moderne 1990; Frühling 1993; Prillaman 2000; Mutua 2001).
Other skeptics include rational choice proponents such as Ramseyer (1994; see also
Ramseyer & Rasmussen 2003) who argues that “independent judiciaries are not common
to freedom-loving countries everywhere” and in fact, he argues they “are not as likely as
we hope” because of the willingness of the legislature to keep the judiciary independent
depends upon the expected electoral fortunes of politicians in power (722).
While our initial interest in judicial independence related to the empirical question
of its influence on human rights behavior, in this paper we report on our efforts to
conceptualize and operationalize judicial independence so as to understand the concept
both as an independent variable and as a dependent variable. Specifically, we explore
and use factor analysis to examine the structure of nine formal measures of judicial
independence and a behavioral indicator of judicial independence, along with three
additional formal provisions that are strongly linked to judicial independence even
though they are not themselves principal elements of the concept. We find that the
formal provisions cluster strongly on two distinct dimensions, which we delineate as
1
One might think, for example, of the recent decision by the High Court of Zimbabwe not to intervene in
the handling by the national election commission of the election dispute between President Robert Mugabe
and his opposition.
2
judicial decision making and organizational factors, and we find that the behavioral
measure, while correlated with the formal measures, loads strongly on its own factor. We
then create measures of overall, decisional, and structural judicial independence using the
factor scores from our analysis of the formal constitutional provisions measures and
examine their relationship across time with our behavioral measure.
CONCEPTUALIZING JUDICIAL INDEPENDENCE
Some scholars, even those who deem judicial independence to be an essential
feature of liberal democracy, argue that there is little agreement on just what this
condition is or on what kind or how much of it is required for a liberal democratic regime
(Russell 2001, 1). Others such as Larkins (1996) claim that judicial independence “may
be one of the least understood concepts in the fields of political science and law.”
Larkins also points out that “on some level, as Theodore Becker (1970) noted more than
thirty-five years ago, ‘we all know what it means,” yet its full significance, intricacies,
and implications still seem beyond our reach” (607). Kornhauser (2002) goes so far as to
argue that “the confusion over the meaning of judicial independence cannot be
eliminated” and thus “judicial independence is not a useful, analytic concept” (45). We
do not accept this extreme assertion but rather believe that despite the variety of proposed
definitions or conceptualization of the elements of judicial independence, there is
significant agreement on its basic principles. Indeed, globally, there has been substantial
movement, as we will show below, towards formally adopting these principles.
If we examine the political science and law literature there is actually rather
substantial conceptual overlap in defining judicial independence, with most scholars
building upon Becker’s 1970 definition:
3
a) the degree to which judges believe they can decide and do decide consistent
with their own personal attitudes, values and conceptions of the judicial role (in
their interpretation of the law), b) in opposition to what others, who have or are
believed to have political or judicial power, think about or desire in like matters,
and c) particularly when a decision adverse to the beliefs or desires of those with
political or judicial power may bring some retribution on the judges personally or
on the power of the court (144).
Rosenn (1987) suggests that Becker’s definition needs further refinement and
recommends improving the definition by broadening its scope and simplifying the
language. He then offers his revised definition of judicial independence: “the degree to
which judges actually decide cases in accordance with their own determination of the
evidence, law and justice, free from coercion, blandishments, interference, or threats of
government authorities or private citizens” (7).
The definition put forth by Ferejohn, Rosenbluth, and Shipan (2004) is similar to
Becker’s and Rosenn’s; they say, “We take judicial independence to mean court
autonomy from other actors,” and they further delineate it saying that “to the extent that a
court is able to make decisions free of influence from other political actors, and to pursue
its goals without having to worry about the consequences from other institutions, it is
independent” (3).
Some scholars define judicial independence by specifically parsing the definition
into categories or its core dimensions. Rosenn’s examination of Latin America
judiciaries finds two broad but somewhat overlapping categories: 1) protection of the
judicial decision making process from outside pressures and 2) protection of the personal
4
independence of the judge. He identifies the common structural measures that seek to
insure independence in each of the categories. Dimensions of judicial independence
related to protecting the integrity of judicial decisions include 1) guaranty of noninterferences with judicial proceedings, 2) jurisdictional monopoly (no special tribunals),
3) requirement of a reasoned opinion, and 4) requirement of public trials. Measures
protecting personal independence include 1) irreducibility of salaries, 2) guaranteeing the
judiciary a fixed percentage of the budget, 3) protected tenure in office, 4) specified and
transparent selection and reappointment processes, 5) protected transferability of judges,
6) avoidance of conflicts of interest, and 7) judicial immunity.
Larkins (1996) defines judicial independence as “the existence of judges who are
not manipulated for political gain, who are impartial toward the parties of a dispute, and
who form a judicial branch which has the power as an institution to regulate the legality
of government behavior, enact ‘neutral” justice, and determine significant constitutional
and legal values” (611). He posits three dimensions similar to Rosenn’s: impartiality, 2
insularity and scope of authority (609). The first two he identifies from within the
literature and notes that they are obviously linked to goal of the judiciary being a neutral
third party: 1) impartiality requires that judges will base their decisions based on law and
facts not predilection toward one of the litigants (Fiss 1983; Shapiro 1981) and 2)
political insularity which requires that judges should not be used as tools to further
political aims nor punished for preventing their realization (Clark 1975; Rosenn 1987;
and Fiss 1993) . He notes that the first quality, impartiality, is difficult to identify and the
2
Becker argues and we agree that impartiality is different from and should not be confused with
independence. It may not be possible for a non-independent judge to be impartial, but it is certainly
possible for an independent judge (in the sense used in these definitions) to be partial (say because he/she is
racist or hates business people, etc., etc.). Nevertheless, writers on independence frequently do lump
impartiality with independence. The practice seems to do no harm in the context of our research.
5
second component is typically achieved through widely accepted formal and structural
safeguards like life tenure, protected salary, and so forth. He adds a third component—
the scope of the judiciary’s authority as an institution which he describes as “the
relationship of the courts to other parts of the political system and society and the extent
to which they are collectively seen as a legitimate body3 for the determination of right
and wrong, legal and illegal” (610, emphasis in original).
Russell (2001) also synthesizes the literature, noting that in political science
judicial independence basically encompasses two concepts: 1) autonomy of judges, both
collectively and individually from other institutions and individuals and 2) judicial
behavior, by which he ultimately means the individual judge’s capacity for independent
thought and judgment (6). He notes that the two concepts are closely related in that one is
the means to the other; we want judges to enjoy a high measure of autonomy so that they
can think and act independently rather than being controlled or influenced by other actors
(6). Russell ultimately argues that judicial independence does not refer to a single lack of
dependence, but, rather, is best understood as two-dimensional, having both an external
dimension (all those forces outside the judiciary itself that can encroach on the autonomy
of the judiciary collectively or individually) and an internal dimension (sources of
influence and control within the judiciary itself). He also argues that in terms of the
sources of dependency, external controls and influences must be distinguished from those
that are internal, and that in terms of the targets of influence or control, the individual
judge must be distinguished from the judiciary as a collective whole or institution. He
3
As with impartiality, we think it best not to conflate legitimacy and independence, but the practice seems
to do no harm in our research.
6
confirms that the external dimension is the most widely recognized dimension as it
embraces the principle of separation of powers (11).
While the dimensions of judicial independence conceptualized by each of these
scholars and others do not fit together perfectly, we do see a common core across them
that allows us to identify two somewhat overlapping sets of distinctions. The first is the
distinction between (1) institutional (or collective) independence from the other branches
or private and public actors and (2) the independence of individual judges from the same
influences. The independence of the individual judges then is further distinguished by
the source of influences or dependence, which either can arise from external sources
(other branches or outside actors) and internal sources within the judiciary itself. Based
on these conceptualizations one might predict that the exploratory factor analysis of the
formal indicators of judicial independence that report below will identify two dimensions.
Our next task is to identify the principal elements or standards that would be
closely linked to the concept of judicial independence and that yield suggestions as to
how the concept can be effectively operationalized. Russell (2001) argues that the most
pressing question is apt to be how to secure the minimal requirements of judicial
independence. International organizations such as the International Bar Association and
the United Nations have set forth minimum standards or basic principles of judicial
independence, which represent substantial consensus in the international legal community
on the priorities for shaping structural safeguards for the judiciary.
7
ELEMENTS OF JUDICIAL INDEPENDENCE
The International Bar Association adopted a set of minimum standards in 1983.
The Delhi standards, as they are called, set forth minimum expectations rather than
ideals, as Justice King (1984) notes. They identified primarily essential standards for
personal and substantive independence of the individual justice and, to a lesser degree,
standards for collective and institutional independence. They included such standards as
guaranteed terms and protected salaries and fiscal autonomy for the judicial branch.
More usefully, the United Nations General Assembly adopted in 1985 a set of
basic principles which we believe now best represent international consensus on what the
basic elements of judicial independence are or should be; therefore, we base our measures
on these conceptualizations.
The United Nations' recommendations are outlined in two documents: the UN's
Basic Principles on the Independence of the Judiciary and the 1995 report by Dato'
Param Cumaraswamy, the U.N. Special Rapporteur on the independence of judges and
lawyers.4 Both documents explicitly call for the formal enshrinement of the principles of
judicial independence in written constitutions or law.
4
The U.N. Basic Principles are as follows:
1. The independence of the judiciary shall be guaranteed by the State and enshrined in the
Constitution or the law of the country. It is the duty of all governmental and other institutions to
respect and observe the independence of the judiciary.
2. The judiciary shall decide matters before them impartially, on the basis of facts and in accordance
with the law, without any restrictions, improper influences, inducements, pressures, threats or
interferences, direct or indirect, from any quarter or for any reason.
3. The judiciary shall have jurisdiction over all issues of a judicial nature and shall have exclusive
authority to decide whether an issue submitted for its decision is within its competence as defined
by law.
4. There shall not be any inappropriate or unwarranted interference with the judicial process, nor
shall judicial decisions by the courts be subject to revision. This principle is without prejudice to
judicial review or to mitigation or commutation by competent authorities of sentences imposed by
the judiciary, in accordance with the law.
8
5.
Everyone shall have the right to be tried by ordinary courts or tribunals using established legal
procedures. Tribunals that do not use the duly established procedures of the legal process shall not
be created to displace the jurisdiction belonging to the ordinary courts or judicial tribunals.
6. The principle of the independence of the judiciary entitles and requires the judiciary to ensure that
judicial proceedings are conducted fairly and that the rights of the parties are respected.
7. It is the duty of each Member State to provide adequate resources to enable the judiciary to
properly perform its functions.
United Nations High Commissioner for Human Rights. Basic Principles on the Independence of the
Judiciary. Adopted by the Seventh United Nations Congress on the Prevention of Crime and the Treatment
of Offenders held at Milan from 26 August to 6 September 1985 and endorsed by General Assembly
resolutions 40/32 of 29 November 1985 and 40/146 of 13 December 1985;
The Special Rapporteur’s principles are as follows:
1) All necessary measures should be taken to respect, protect, and promote the independence of
judges.
2) In particular, the following measures should be taken:
a) The independence of judges should be guaranteed pursuant to the provisions of the
Convention and the constitutional principles, for example by inserting specific provisions in
the constitutions or other legislation or incorporating the provisions of this recommendation in
internal law. Subject to the legal tradition of each State, such rules may provide, for instance,
the following:
(1) decisions of judges should not be subject to any revision outside any appeals procedures
as provided by law;
(2) the terms of office of judges and their remuneration should be guaranteed by law;
(3) no organ other than the courts themselves should decide on its own competence, as
defined by law;
(4) with the exception of decisions on amnesty, pardon or similar, the Government or the
administration should not be able to take any decision which invalidates judicial
decisions retroactively;
b) The executive and legislative powers should ensure that judges are independent and that steps
are not taken which could endanger the independence of judges.
c) All decisions concerning the professional careers of judges should be based on objective
criteria and the selection and career of judges should be based on merit, having regard to
qualifications, integrity, ability and efficiency. The authority taking the decision on the
selection and career of judges should be independent of government and administration. In
order to safeguard its independence, rules should ensure that, for instance, its members are
selected by the judiciary and that the authority decides itself on its procedural rules.
However, where the constitutional or legal provisions and traditions allow judges to be appointed
by government, there should be guarantees to ensure that the procedures to appoint judges are
transparent and independent in practice and that the decisions will not be influenced by any
reasons other than those related to the objective criteria mentioned above. These guarantees could
be, for example, one or more of the following:
(i) special independent and competent body to give the Government advice which it follows
in practice; or
(ii) the right for an individual to appeal against a decision to an independent authority; or
(iii) the authority which makes the decision safeguards against undue or improper influences.
d) In the decision-making process, judges should be independent and be able to act without any
restrictions, improper influence, inducements, pressures, threats or interferences, direct or
indirect, from any quarter or for any reason. The law should provide for sanctions against
persons seeking to influence judges in any such manner. Judges should have unfettered
freedom to decide cases impartially, in accordance with their conscience and their
interpretation of the facts, and in pursuance of the prevailing rules of the law. Judges should
not be obliged to report on the merits of their cases to anyone outside the judiciary.
e) The distribution of cases should not be influenced by the wishes of any party to a case or any
person concerned with the results of the case. Such distribution may, for instance, be made by
9
There is considerable overlap in the principles discussed in the two documents,
but the U.N.'s Basic Principles appear to be broader than the Special Rapporteur's
principles. They go beyond listing the formal attributes of an independent judiciary to
prescribing certain actions that the judiciary should take - specifically that the judiciary
should ensure fair trials and should protect individual rights. For now we limit our
analysis to the key formal provisions that the U.N. suggests will produce an independent
judiciary
A Merged Set of Principles
Merging the two sets of UN principles produces the following criteria for an independent
judiciary:
1. a) Terms of office and b) remuneration are constitutionally guaranteed, regardless
of whether appointed or elected.
2. The decisions of judges are not to be subject to any revision outside any appeals
procedures as provided by law.
3. The courts have exclusive authority to decide on their own competence, as
defined by law -- their decisions are made without any restrictions, improper
influences, inducements, pressures, threats or interference, direct or indirect,
from any quarter or for any reason.
drawing lots or a system for automatic distribution according to alphabetical order or some
similar system.
f) A case should not be withdrawn from a particular judge without valid reasons, such as cases
of serious illness or conflict of interest. Any such reasons and the procedures for such
withdrawal should be provided for by law and may not be influenced by any interest of the
Government or administration. A decision to withdraw a case from a judge should be taken by
an authority which enjoys the same judicial independence as judges.
3) Judges, whether appointed or elected, shall have guaranteed tenure until a mandatory retirement
age or the expiry of their term of office, where such exists.
The Report of the Special Rapporteur on the independence of judges and lawyers, Dato' Param
Cumaraswamy, submitted pursuant to Commission on Human Rights resolution 1995/36.
10
4. The courts have jurisdiction over all issues of a judicial nature. This criterion
seems to relate to the criterion that everyone shall have the right to be tried by
ordinary courts or tribunals using established legal procedures. This criterion
could be condensed to the requirement that civilians not be tried in military
courts or exceptional courts.
5. The courts must have adequate resources.
6. The executive and legislative powers should ensure that judges are independent
and that steps are not taken which could endanger the independence of judges.
This criterion could be condensed to the formal separation of powers of the
branches of government.
7. The selection and career of judges should be based on merit: qualifications,
integrity, ability and efficiency. The authority taking the decision on the
selection and career of judges should be independent of government and
administration. If the constitutional or legal provisions and traditions allow
judges to be appointed by government, there should be guarantees to ensure that
the procedures to appoint judges are transparent and independent in practice and
that the decisions should be based on the above criteria.
The first element seeks to protect individual judges from possible professional
and personal retribution from the other branches of government. It also attempts to limit
improper influences generally that might interfere with judicial impartiality. The next
two elements, finality of decisions and exclusive authority, are aimed at safeguarding
judges’ ability to carry out their functions fully without incursions from other state actors,
and the elements are intended to enhance the probability that judicial decisions are not
11
hollow attempts to protect human rights. The fifth element, fiscal autonomy, is directed
both at protecting the courts from the financial retribution of an abusive regime and at
ensuring that the court has adequate resources to fully carry out its functions. The sixth
element, formal separation of powers, further removes judicial power from arbitrary
abuses of the other branches since, under such a structure, the judiciary’s power is a
constitutional grant of power rather than grant of power from the state that could be
removed at will. The final element, enumerated qualifications, is directed at establishing
institutional professionalism and impartiality. This provision should lead to judges that
are more competent and that have been socialized to the norms of judicial independence,
which should make them more willing and capable to withstand the incursions from other
branches upon human rights. We conclude that these seven elements represent from a
theoretical perspective the best guide to operationalizing international and academic
consensus about what constitutes judicial independence.
Two Problematic Elements
In addition to the seven elements that emerge from the United Nations’ principles
we consider two more questionable elements: hierarchical system and judicial review.
Neither of these would have the level of consensus for their inclusion that the above
elements have.
In part our inclusion of these elements as potential indicators of formal judicial
independence rests on the work of Blasi and Cingranelli (1996), who developed an index
based on their conceptualization of judicial independence. This index was composed of
eight factors that overlap substantially with our list of principles based on UN
recommendations. Their eight elements included 1) the judiciary is housed in a separate
12
branch; 2) the judiciary is fiscally autonomous; 3) the judiciary has the power of judicial
review of the actions of the legislative and executive branches; 4) judges have life tenure
(at least for the highest level judges); 5) judges are protected by restricted removal
procedures; 6) the judiciary is organized in a hierarchical system; 7) public hearings are
required; 5 and 8) the judiciary is composed of professional judges.
The most notable difference between the U.N. principles and Blasi and
Cingranelli's index is the political scientists’ inclusion of judicial review, which is not
explicitly specified in the United Nations' lists. In fact, authorities disagree as to whether
judicial review is a power that contributes to an independent judiciary or whether an
independent judiciary is a necessary component for the real exercise of judicial review
(contrast Blasi and Cingranelli 1996 with Rosenthal 1990). However, in the international
community and to some extent in the U.S. as well, judicial review is linked to rights
protection by an independent judiciary. For example, Keck (2002) suggests that in the
context of the United States “the very mission of an independent Supreme Court [has]
come to be identified—in the minds of ordinary citizens and the justices themselves—
with the enforcement of rights-based limits on political action” and “for the justices to
abandon this role would be to call into question the very justification for their office”
(135). International proponents of constitutional or judicial review have expected this
independent rights-protective role for the judiciary, especially in periods of crisis or
instability (e.g. Becker 1970; International Commission of Jurists 1983; Caine 1988;
Chowdury 1989; Maduna 1989; Chowdury 1989; Ackerman 1989; Stotzky 1993; and
5
We have elsewhere conceptualized the right to a public trial or hearing as an individual right and have
(Keith 1999) treated it a component of bills of rights rather than judicial independence. We will discuss
this provision in the next section where we discuss the three related factors that we eventually factor
analyzed as potential indicators of judicial independence.
13
Garro 1993). In this paper we treat the inclusion of this potential element of judicial
independence as an important empirical question to be explored.
The second notable difference between Blasi and Cingranelli’s and the UN
principles is their inclusion of the requirement for a hierarchical system in which the
judiciary is structured in multiple layers with the highest level court exercising final
control/review of lower court decisions. In theory this arrangement should ensure that
the decisions of lower level judges, who are arguably more susceptible to corruption or
other outside influence, would be subject to review by the higher courts. Compared to
higher court judges, lower court judges are presumably more numerous, less qualified at
least in terms of experience, less socialized toward judicial norms, and less well paid and
more susceptible to bribery. Thus an effective judicial hierarchy would potentially
strengthen overall independence vis-à-vis external actors. Additionally, this dimension
would seem to strengthen judicial impartiality and the stability of the courts’
jurisprudence.
At the same time, a hierarchical system potentially weakens the independence of
these lower court judges to the extent that career advancement, remuneration, or the
security or amenability of the work environment is dependent upon higher level judges.
Russell (2001) notes that judicial independence can be threatened, not only from outside,
but from within by senior judges using administrative controls (see also Ramseyer and
Rasmussen 2003; Guarnieri and Pederzoli 2002). Certainly Prillaman’s work and that of
other Latin Americanists would suggest that a hierarchical structure can be abused,
especially in systems that are strongly lacking transparency in their selection processes
and personnel decision making. Rosenn (1987) on the other hand, argues that judicial
14
independence is not necessarily inconsistent with judiciaries where senior judges make
decisions regarding lower court judges’ promotion, transfers, and salaries, presumably
based on account of judicial performance (5).
Three Additional Judicial Provisions to Consider
In other research we examine formal provisions for fundamental individual rights
typically associated with bills of rights and provisions for states of emergencies (Keith
2002, Keith and Poe 2004; and Keith, Tate and Poe 2007). Three of these provisions are
linked to judges’ behaviors and the concept of judicial independence: the right to a fair
trial, the right to a public trial or hearing, and the right to the writ of habeas corpus.6
Typically, these provisions are conceptualized as individual rights and are usually
housed in the bill of rights section rather than the judiciary section of constitutions, with
the sometime exception of public trials provisions, which do occasionally fall within the
courts section of constitutions. In fact, these provisions may be necessary if judicial
independence is to be in any sense effective. We have noted the connection of these
provisions to the concept of judicial independence in the other work cited above,
particularly that of the trials provisions. Our subsequent discussion explains that we
found empirical evidence of this connection in our exploratory factor analysis, so we
have added these three “rights” provisions as candidate indicators of formal judicial
independence here.
Like Russell (2001) and the UN, we believe that one of the most fundamental
ways to protect judicial independence is through formalized constitutional guarantees and
we would agree that newly emergent democracies would be well advised to write strong
6
Blasi and Cingranelli actually included public hearings as a part of their judicial independence index, as
we note above.
15
principles of judicial independence into their constitutions along with a catalog of
fundamental rights (Russell 2001:22-23). We, too, recognize that constitutional language
may be valuable from a symbolic and educational perspective and agree with Siegan
(1994) that constitutions may be "the most important legal document for a nation that
subscribes to the rule of law" because they define the relationship between the nation's
people and their government (72) and because they impose constraints upon government
and protect the individual's freedoms from arbitrary or abusive state action (see Andrews
1964; Finer 1974; Stotzky 1993; Siegan 1994; Elster 1993).7
7
This may be the appropriate place to acknowledge the obvious fact that a formal written constitution is
not the only mechanism through which nations structure courts, judicial independence, and rights. Clearly
statutes, legal codes and doctrines, and even customs also structure a nation’s “constitution” and the
interpretation of its written “Constitution.” Our reliance on the latter is primarily a matter of necessity. We
must tolerate the irony of a using a data source that leads us to code one of the world’s oldest
“constitutional” regimes, the United Kingdom, as having “no constitutional provisions” because it has no
written constitution and the birthplace of judicial review, the United States, as having no constitutional
judicial review because the practice was established though Supreme Court doctrine. It is simply
impossible to implement and code a broader definition of national constitutional practices. The importance
still placed on written constitutions and their formal provisions is evident in the large scale constitution
drafting that has gone on during the Third Wave of democratization and in such ongoing research efforts as
the ambitious Comparative Constitutions Project in the Cline Center for Democracy at the University of
Illinois. See https://netfiles.uiuc.edu/zelkins/constitutions/index.htm accessed April 29, 2008. (For some
preliminary reports from this unfinished project, see Elkins, Ginsburg and Melton 2008; Elkins and Sides
2007; Ginsburg, Elkins and Chernykh 2008.)
16
FORMAL (DE JURE) AND BEHAVIORAL (DE FACTO) JUDICIAL
INDEPENDENCE
We certainly recognize that formal or de jure judicial independence does not
necessarily equate to de facto judicial independence. A variety of circumstances and
factors may intervene that negate the actualization of constitutional promises. And thus
we also have created a measure of judicial independence that attempts to capture the
actual behavior of the judiciary. We return to this measure below. First, however, we
describe the distribution of the measures of formal constitutional provisions for judicial
independence that we have created to operationalize the principles discussed earlier..
Constitutional Provisions of Formal Judicial Independence (De Jure)
We have constructed nine ordinal measures for constitutional provisions that reflect the
international principles of judicial independence we described earlier.8 Since it is
possible that a constitution may meet a criterion to some degree but not fully, we have
coded the variables as follows:
No provision (0): no constitution or constitution does not provide this element;
Qualified provision (1): constitution provides for this element to a limited
degree or provides for this element vaguely but not fully; and
Explicit and full provision (2): constitution provides for this element fully and
explicitly.
The data sources we used to code these variables were Blaustein’s Constitutions of the World (1993 and
continuing) and Flanz's (1997 and continuing) Constitutions of Countries of the World. This series contains
English translations of the vast majority of the current constitutions of the world and, in most cases,
contains information on past constitutions as well. These are also the fundamental sources for the
Comparative Constitutions Project (see note 4).
8
17
In addition, for two measures we defined a fourth code (-1) for constitutional
provisions that explicitly deny an international principle of judicial independence. The
nine measures are:
Guaranteed Terms: The constitution guarantees terms of office, regardless of
whether appointed or elected, and restricted removal of judges. This measure reflects the
first part of the first element above.
Finality of Decisions: The decisions of judges are not be subject to any revision
outside any appeals procedures as provided by law. This measure reflects the second
element above.
Exclusive Authority: The courts have exclusive authority to decide on their own
competence, as defined by law -- their decisions are made without any restrictions,
improper influences, inducements, pressures, threats or interference, direct or indirect,
from any quarter or for any reason. This measure reflects the third element above.
Ban against Exceptional or Military Courts: The courts have jurisdiction over
all issues of a judicial nature (civilians are tried by ordinary courts or tribunals, not
military or exceptional courts). This measure reflects the fourth element above. This
measure includes an additional score beyond 0, 1, and 2. Constitutions are coded (-1) if
they explicitly condone trying civilians in military courts or explicitly allow the
formation of exceptional courts. We have chosen to give these constitutions a negative
score, because not only do the constitutions fail to specifically ban these types of courts
(which would have earned a score of zero), they go so far as to explicitly legitimize the
use of such courts).
18
Fiscal Autonomy: The courts are fiscally autonomous. Their salaries and/or their
budgets are protected from reduction by the other branches. This measure reflects the
fifth element and the second part of the first element.
Separation of Powers: The courts are housed in a separate branch from the
executive and legislative powers. This measure reflects the sixth element above.
Enumerated Qualifications: The selection and career of judges are based on
merit: qualifications, integrity, ability and efficiency. This measure reflects the seventh
element above.
Judicial Review: Courts exercise judicial or constitutional review of legislative
and executive branches.9 We recorded a -1 for constitutions in which courts were
explicitly forbidden to exercise judicial or constitutional review.
Hierarchical System: Courts are structured in multiple layers with the highest
level court exercising final control/review of lower court decisions.
Additionally, we constructed three measures that are traditionally housed in a bill
of rights, but that are also theoretically associated with the behavior of independent
judges. As discussed above, we find them to be relevant empirically to our analyses of
the structure of judicial independence. They include the provision for:
Writ of Habeas Corpus the requirement that government promptly present
evidence before a judge or magistrate sufficient to justify holding a prisoner and to make
known the crimes which with a prisoner is being charged.
9
In many constitutions it was easy to identify the constitutional provision for judicial review because the
document explicitly stated that courts exercise "constitutional review" or "judicial review." In other cases,
we used Ackermann's (1989) definition as the standard for determining whether constitutional or judicial
review existed: right and duty conferred on a court or particular courts of a country to interpret
authoritatively the constitution of that country, to decide authoritatively the constitutionality of laws,
executive and administrative acts, and in appropriate cases, to declare such laws and acts invalid and
unenforceable when they conflict with the country’s constitution (p.60).
19
Right to Public Trials: the right to public trials or hearings.
Right to a Fair or Trail: the right to a fair or impartial trial or hearing.10
The frequency distributions for these twelve constitutional provisions are given in
Table 1. The results in Table 1 come from our cross-national dataset that covers a twentyfive year period; therefore, our observations represent “country years” (for example,
Ethiopia in 1981, Ethiopia in 1982…France in 1981, France in 1982, and so forth). To
help readers evaluate the results it presents, the modal categories for each variable have
been highlighted in Table 1.
10
Again, we have coded the level of constitutional provision for these measures as (2) explicit guarantee or
mention in the constitution; (1) explicit guarantee or mention in the constitution but notes exceptions or
qualifications, such as a public interest clause; (0) no mention in the constitution.
20
Table 1: Frequency Distributions for Formal Judicial Independence Measures
Constitutional
Provision
Denies
Provision
(-1)
Guaranteed Terms
Finality of Decision
Exclusive Authority
Ban against
Exceptional or
Military Courts
Fiscal Autonomy
7.0%
(345)*
-
Separation of Powers
Enumerated
Qualifications
Judicial Review
4.6%
(227)**
Hierarchical System
Right of Habeas
Corpus
Right to Public Trial
-
Right to Fair Trial
-
No
Provision
(0)
40.9%
(2011)
66.7%
(3275)
48.2%
(2366)
Qualified/
Limited
Provision
(1)
17.0%
(835)
6.1%
(299)
15.1%
(743)
Explicit &
Full Provision
(2)
70.8%
(3477)
78.2%
(3840)
40.1%
(1967)
61.2%
(3007)
51.2%
(2514)
37.0%
(1817)
50.1%
(2468)
52.2%
(2570)
72.3%
(3558)
4.3%
(210)
4.3%
(212)
9.0
(42)
9.4%
(464)
15.3%
(753)
5.6%
(276)
26.0%
(1281)
39.6%
(1951)
17.1%
(841)
17.9%
(879)
17.5%
(859)
50.9%
(2502)
29.3%
(1440)
28.9%
(1417)
57.4%
(2818)
23.8%
(1174)
8.2%
(402)
10.7
(524)
42.0%
(2065)
27.2%
(1337)
36.7%
(1802)
N = 4911 country years
*Explicitly condones trying civilians in military courts or explicitly allows the formation
of exceptional courts.
**Courts explicitly forbidden to exercise judicial or constitutional review.
“No Provision” is the dominant category for nine of the twelve provisions in
Table 1 (finality of decision, exclusive authority, the ban against exceptional courts, fiscal
21
autonomy, enumerated qualifications, judicial review, right to habeas corpus, right to
public trial, and right to fair trial). “Explicit and Full Provision” is the modal category for
three provisions (guaranteed terms, separation of powers and hierarchical system). Only
for the latter three provisions do the combined categories of “full provision” and
“qualified provision” represent fifty percent or more of the county years. The two least
prevalent constitutional provisions are a ban against exceptional courts and provision for
fiscal autonomy, both of which see full or partial provision in barely 20 percent of the
country years under study here.
Behavioral (De Facto) Judicial Independence
As Howard and Carey (2004) note, prior to their work, there had not been a
systematic replicable measure of judicial independence available for the global set of
countries that went beyond formal provisions. A few studies have created subjective
ratings for a significant albeit non-random cross-section of countries; however, these
evaluations are generally not replicable (e.g. Johnson 1976; Cross 1999 and La Porta,
López-de-Silanes, Pop-Eleches, and Shleifer 2004). We agree with Howard and Carey
that it should be possible to create a measure that is replicable and valid and that the U.S.
State Department’s annual Country Reports on Human Rights Practices11 offer the best
resource to code judicial independence for the global set of countries. We share their
concern that the reports are potentially thinner in their reporting of judicial independence
backwards in time, and we have kept this in mind in our coding and analyses as well.
Another caveat we have tried to keep in mind in creating our judicial
independence measures is that they are measures of judicial independence as reported by
11
Available at http://www.state.gov/g/drl/rls/hrrpt/ accessed April 29, 2007.
22
the State Department. As such, they represent the considered judgment of State
Department field officers who have been charged by Congress and their superiors with
delivering annual assessments of human rights practices (and judicial independence) for
the countries where they are stationed. Just how “expert” their assessments are might be
debated. Nevertheless, we would strongly defend these assessments as providing the best
data source currently available for evaluating levels of judicial independence for the
global set of nations across a significantly long time period: whether the judicial
independence measure derived from them is useful for research and evaluation purposes
is largely an empirical question that we begin to address in this article.12
In constructing our measure of behavioral judicial independence we used Howard
and Carey as a reference point. They described their measure, which they were able to
create for ten years, as follows:
Our full, or high, independence measure corresponds to judiciaries that
function in practice independent of the executive and legislature, are relatively
free from corruption and bribery, and afford basic criminal due process
protections to criminal defendants. Our middle measure of independence
corresponds to judiciaries that have one, but not both, of these two characteristics,
or that have partial measures of both (287-88).
Carey and Howard code the two measures as independent dichotomous variables;
presumably the default or reference category is those without judicial independence.
12
We are pleased that Cingranelli and Richards are also currently creating measures of judicial
independence as part of their CIRI project (see http://ciri.binghamton.edu/ accessed April 29, 2008), albeit
with a somewhat different coding scheme. Thus scholars interested in assessing judicial independence
from a global perspective will have multiple indicators available to them. While their data are not available
yet, they have generously shared with us their initial data and codebook. We do not give the details of their
measure in this paper since they are still in the process of data collection. However preliminary analyses,
encouragingly, confirm that their measure and our own appear to be valid, if somewhat different, indicators
of the same phenomenon.
23
While we applaud Howard and Carey’s measure of judicial independence, we
find it less than ideal for our purposes because one of its three components considers
whether the state “afford[s] basic criminal due process to criminal defendants.” This
component seems at best indirectly related to consensus definitions of judicial
independence and seems to conflate independence and impartiality, which we prefer not
to do. In addition, our strong interest in subsequently exploring the potential influence of
judicial independence on human rights leads us to be very sensitive to possible overlaps
between operationalizations of independent variables used in human rights models and
dependent variables such as the Political Terror Scale (Gibney and Dalton 1996) and
Cingranelli and Richards’ measures of physical integrity rights
(http://www.humanrightsdata.org). These dependent variables assess the degree of
abusive or politically motivated state behavior including imprisonment, torture,
disappearance, and extrajudicial killings, which seem by their very nature to deny the
basic due process rights included in Howard and Carey’s operationalization of judicial
independence. Therefore, we were reluctant to follow Howard and Carey’s
operationalization because ultimately we wanted a measure that would be fully
independent of human rights protection variables.
In creating our “Behavioral” or de facto judicial independence indicator, we first
compiled a document containing all portions of the State Department country reports that
dealt with the judiciary for the entire set of countries for two sample years, 2003 and
1993. We then read this compilation to ascertain what coding specificity was possible
given the level of detail reported across countries for the two decade-separated sample
years. While it might be desirable to create a richer measure than the Howard and Carey
24
indicator, the level of information in the reports simply did not promote confidence that
we could do so for the entire set of countries, a conclusion that we assume Howard and
Carey reached before us.
From our reading of these reports we created a measure that captured our
conceptualization of judicial independence, one that we believe is appropriate for the data
available in the State Department reports, closely reflects the language of the State
Department assessments themselves.13 We created a trichotomous measure that assigns
scores as follows:
2) Fully Independent Judiciary: The judiciary is reported as “generally
independent” or is independent in practice with no mention of corruption or outside
influence.
1) Somewhat Independent Judiciary: The judiciary is reported to be somewhat
independent in practice with reports of (some) pressure from the executive “at times” or
with occasional reports of corruption.
0) Non-Independent Judiciary The judiciary is reported as not being
independent in practice; is reported to have significant or high levels of executive
influence or interference; or is reported to high levels of corruption.15
Using these definitions, we had two research assistants code all countries for the
years 2003 and 1993. Their independently coded scores produced a reliability coefficient
of .91. We then coded the State Department reports backwards in time, so that we could
assess initially whether the reports were becoming too thin as we moved back into the
The State Department reports are summative and typically use phrases such “has a fully independent
judiciary” or “has a judiciary that is somewhat or partially independent” or that through the year “there was
pressure from the executive” or that “there was high levels of corruption within the judiciary.” For
examples of the actual language in the reports, see the Appendix to this paper.
15
Examples of country coding are also included in the Appendix.
13
25
1980s. Coders were instructed not to force the coding if the reports became insufficiently
detailed to determine the appropriate score, and we reemphasized this message as the
coding moved to earlier years.
While we cannot be sure that coders always explicitly noted that the information
present in a report was insufficiently detailed to allow a confident coding, for the years
1981-1984, the four years for which we expected information to be least adequate, they
did do so, and reported insufficient information in less than 1% of the country-years. At
the time of this writing, the coders had compiled judicial independence scores for all
countries included in the State Department Reports for the years 1981-2006, inclusive.
The frequency distribution for our behavioral judicial independence measure is
presented by year in Table 2. The overall mode for all countries in all years is “Fully
Independent” (39.1%). The percentage of countries within a single year achieving the
rating of full judicial independence ranges from a high of 46.8% in 1990 to a low of
34.9% in 2002. The percentage of countries within a single year failing to achieve any
observable degree of judicial independence ranges from a high of 44.3% in 2004 to a low
of 25.2% in 1981.
26
Table 2: Frequency Distributions for Behavioral Judicial Independence
Not
Independent
Somewhat
Independent
Fully
Independent
Total
1981
25.2% (40)
30.2% (48)
44.7% (71)
100.0% (159)
1982
28.4% (46)
31.5% (51)
40.1% (65)
100.0% (162)
1983
29.8% (48)
31.7% (52)
38.5% (62)
100.0% (161)
1984
31.1% (51)
31.1% (51)
37.8% (62)
100.0% (164)
1985
31.1% (52)
29.3% (49)
39.5% (66)
100.0% (167)
1986
29.9% (50)
27.5% (46)
42.5% (71)
100.0% (167)
1987
29.4% (50)
31.8% (54)
38.8% (66)
100.0% (170)
1988
30.2% (51)
29.6% (50)
40.2% (68)
100.0% (169)
1989
28.1% (48)
31.0% (53)
40.9% (70)
100.0% (171)
1990
29.8% (51)
23.4% (40)
46.8% (80)
100.0% (171)
1991
29.9% (52)
28.2% (49)
42.0% (73)
100.0% (174)
1992
26.3% (50)
31.1% (59)
42.6% (81)
100.0% (190)
1993
31.1% (60)
25.9% (50)
43.0% (83)
100.0% (193)
1994
32.8% (63)
26.0% (50)
41.1% (79)
100.0% (192)
1995
35.8% (69)
24.9% (48)
39.4% (76)
100.0% (193)
1996
34.0% (66)
25.8% (50)
40.2% (78)
100.0% (194)
1997
37.1% (72)
25.8% (50)
37.1% (72)
100.0% (194)
1998
39.7% (77)
22.7% (44)
37.6% (73)
100.0% (194)
1999
36.1% (70)
26.8% (52)
37.1% (72)
100.0% (194)
2000
38.0% (73)
26.6% (51)
35.4% (68)
100.0% (192)
2001
39.1% (75)
25.5% (49)
35.4% (68)
100.0% (192)
2002
42.7% (82)
22.4% (43)
34.9% (67)
100.0% (192)
2003
40.8% (78)
23.6% (45)
35.6% (68)
100.0% (191)
2004
44.3% (85)
19.8% (38)
35.9% (69)
100.0% (192)
2005
42.2% (81)
20.8% (40)
37.0% (71)
100.0% (192)
2006
39.4% (76)
24.4% (47)
36.3% (70)
100.0% (193)
Total
34.2%
(1616)
26.6%
(1258)
39.1%
(1849)
100.0%
(4723)
27
We plot the frequency distributions across time in Figure 1.16 This plot reveals
significant trends, which we will discuss in conjunction with a subsequent plot of the
annual and overall means (Figure 2).
Figure 1: Categories of Behavioral Judicial Independence, 1981-2006
Figure 217 plots annual mean and the overall mean (all countries, all years) of
judicial independence over time. The overall mean is 1.05. Its standard deviation is .082
16
Because the number of countries for which data exist increased over time from 159 to as many as 194,
we decided to plot the raw rather than the relative frequencies of countries receiving each code. We have
also plotted the relative (percentage) frequencies and found a similar pattern, albeit one in which the
percentages of fully independent and somewhat independent nations fairly clearly declines over time, rather
than most fluctuating around a steady level, as in Figure 1.
28
Figure 2: Mean Behavioral Judicial Independence by Year, 1981-2006
Plotting the annual means of the behavioral judicial independence across time
shows that the level of judicial independence fluctuated somewhat but was consistently
above the overall mean from 1981 through 1994. . With the dissolution of the Soviet
Union and the concomitant increase in newly independent states in the world, however,
we see (starting in 1993) an immediate decrease in the mean level of judicial
independence globally that continues for over a decade. Similarly, in Figure 1 we can see
a substantial increase (from a relatively stable 50 in 1992 to a peak of 85 in 2004) in the
Readers should note that the Y axis scale in Figure 2 has been truncated to a “convenience” range of .9 to
1.2. The full range of the judicial independence measure is of course 0 to 2. We have presented the
truncated scale because it illustrates the pattern of change in judicial independence more clearly.
17
29
number of non-independent judiciaries following the dissolution of the Soviet Union.
We do see a simultaneous decrease in the numbers of countries with fully independent
and somewhat independent judiciaries, but the changes in these categories are more
moderate.18 We know from our previous work that the proportion of states adopting
constitutional provisions for an independent judiciary actually increased dramatically
during the period of constitution writing and democratization that occurred with the
break-up of the Soviet Union, so it is somewhat startling to note the decrease in actual
achieved judicial independence that occurred during the same time period. We will return
to this gap in a later section after addressing the structure of formal and behavioral
judicial independence.
THE STRUCTURE OF JUDICIAL INDEPENDENCE
22 Constitutional Provisions
In other work (Keith, 2002; Keith, Tate, and Poe 2007), we have created and
analyzed a set of 22 formal indicators of the existence of constitutional provisions
relevant to the promotion or protection of fundamental human rights for a worldwide set
of countries. These indicators currently exist for the period 1976-2005. They are divided
into indicators of personal rights and liberties (9 measures), limits on the declaration and
exercise of emergency rule (4 measures), and elements of judicial independence (9
measures). Naturally, this paper focuses on the latter. However, before restricting our
attention to measures of judicial independence, we began by examining the extent to
which all 22 measures relate to each other and form distinguishable, meaningful
18
The number of fully independent judiciaries decreases from its peak of 83 in 1993 to its nadir of 68 for
2000, 2001, and 2003. The number of somewhat independent judiciaries decreases from its peak of 59 in
1992 to its low point of 38 in 2004.
30
dimensions. A straightforward, orthogonally-rotated (Varimax) factor analysis of the
structure of the 22 measures is reported in Table 3.19
19
Unless otherwise noted, all factor analyses reported in this paper were accomplished using the FACTOR
program in SPSS 15.0. The program was instructed to produce a principal components solution for each
set of indicators, using Kaiser’s Criterion to extract only factors with eigenvalues greater than 1.0. Each
principal components solution was then rotated orthogonally using the Varimax criterion.
31
Table 3: Varimax Factor Analysis of 22 Constitutional Provisions of Rights and
Liberties, Judicial Independence, and Emergency Rule, All Country Years, 19762005*
Freedom of Speech
Freedom of Association
Freedom of Assembly
Freedom of Press
Freedom of Religion
Right to Strike
Right of Habeas Corpus
Right to a Public Trial
Right to Fair Trial
Ban on Torture
Guaranteed Terms
Finality of Decisions
Exclusive Authority
Ban Against Exceptional/Military Courts
1
.852
.823
.801
.590
.764
Varimax Factor
2
3
4
.170 .101 .170
.145 .152 .191
.154 .191 .212
.229 -.057 -.162
.106 .158 .086
.117 .496
.347 .020
.418 -.037
.383 -.227
.360 .179
.233 .337
.087 .706
.112 .762
.129 .176
.001
.498
.442
.383
.254
.711
.161
.229
.158
.338
.463
.053
.360
.543
.132
.086
.074
-.073
5
.025
.127
.010
.337
.055
.360
.183
-.010
-.373
.059
-.040
-.001
.088
.652
.002 .026 .732 -.003 .056
Fiscal Autonomy
.195 .648 .487 .032 .031
Separation of Powers
.061 .268 .599 .139 .074
Enumerated Qualifications
.136 .753 .158 .126 .159
Judicial Review
.269 .352 .496 .010 .119
Hierarchical System
.170 -.056 .033 .244 .579
Legislative Declaration - State of Emergency
.164 .158 .142 .569 -.377
Duration Limits - State of Emergency
Cannot Dissolve Legislature - State of Emergency .138 .589 -.190 .167 -.365
.032 .192 -.022 .670 .147
Non-Derogable Rights - State of Emergency
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
*Shading indicates that a variable’s highest principal component or Varimax loading was
on this factor.
The highest factor loading for each individual constitutional provision is
highlighted in Table 3. The highlights make it easy to see that the first factor measures
32
mostly the common variance among five personal freedoms, freedom of speech,
association, assembly, press, and religion, though fair trial also shares its highest (and
necessarily modest) loading with factor one and factor three. The right to strike and the
guarantee against torture or cruel and inhuman punishment load on other and different
factors (two and four, respectively). Public trial and the right to habeas corpus, the other
two judiciary-related measures among the rights and liberties indicators, also have their
highest (but, again, modest) loadings on factor three. We shall return to them shortly.
While they do not have the factors entirely to themselves, eight of the nine
judicial independence measures clearly define factors two and three. Finality of
Decisions, Exclusive Authority, Separations of Powers, and Judicial Review principally
define factor two.20
Factor three is also defined most clearly by judicial independence indicators:
Guaranteed Terms, Fiscal Autonomy, Enumerated Qualifications, and, at a lower level,
Hierarchical System. Factor three also contains the highest loading for the rights to
habeas corpus, public trial, and fair trial. By comparison with the judicial independence
provisions that primarily define factor three, the loadings for habeas corpus, public trial,
and fair trial are moderate. In addition, these measures also exhibit modest loadings on
factors other than factor three. Nevertheless, while we shall continue to treat them as
measures of protected rights, rather than indicators of formal judicial independence, their
loadings on factor three makes clear that they are related to judicial independence.
To explore the relationship between our behavioral measure of judicial
independence and the 22 constitutional provisions, we replicated the factor analysis
20
At a lower level, the right to strike and the prohibition against dissolving the legislature during a state of
emergency also have their highest (but modest) loadings on factor two.
33
reported in Table 3 adding our measure of behavioral judicial independence. Adding this
particular 23rd variable reduced the number of includable country-years to those available
for 1981-2006 and yielded a six factor solution, distinguished mostly by the fact that the
behavioral independence measure loaded on the sixth factor, virtually alone. For
validation, we replicated the analysis adding Cingranelli and Richards’ preliminary
judicial independence indicator. This reduced the number of country-years further, to
those with data available for 1981-2002. But it also produced a six factor solution with
both our and their behavioral measures of judicial independence loaded strongly on a
single factor, but few other changes relevant to our analysis in this paper.21
The Structure of Formal Judicial Independence
Having explored the structure of all our constitutional provisions, we turn our
attention to the analysis of the nine formal judicial independence measures. A factor
analysis of the nine measures produced the principal components results reported in
Table 4. These show that seven of the nine indicators are a good fit with the first
principal component. Only Fiscal Autonomy and Ban on Exceptional/Military Courts
load on the first principal component at less than .600. The former loads moderately on
the first component at .466 and more strongly on the second at .633. The latter loads on
neither component, even to a moderate degree.
Rotating the principal components solution produces the Varimax factor results
also shown in Table 4. This analysis confirms our earlier observation that eight of the
judicial independence indicators load clearly onto two factors (four each). But they also
confirm that the Ban on Exceptional/Military Courts measure, though it seems a very
We do not report these analyses to save space and because Cingranelli and Richards’ data are not yet
complete and publicly available. We can provide them to anyone who would like to see them.
21
34
logical and important measure of judicial autonomy or independence, does not belong in
the same set as the other eight measures of judicial independence.22
Table 4: Principal Components and Varimax Factor Analyses of Nine Formal
Judicial Independence Measures*
Guaranteed Terms
Finality of Decisions
Exclusive Authority
Fiscal Autonomy
Principal
Component
1
2
.761
.326
.656
-.393
1
.383
.757
2
.734
.111
.742
.466
.833
.601
.721
.649
.338
.815
-.039
.699
.216
.828
.349
.298
.175
.785
.460
.685
.126
.599
.170
-.379
.633
-.086
.394
-.426
.243
-.057
Varimax
Factor
Separation of Powers
Enumerated Qualifications
Judicial Review
Hierarchical System
Ban on Exceptional/Military
Courts
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
*Shading indicates that a variable’s highest principal component or Varimax loading was
on this factor.
That the ban on exceptional or military courts does not load on the same factor as
the other eight indicators of judicial independence does not mean that it is unimportant,
or that it is an inappropriate measure of judicial autonomy or independence. For
example, in other research we have conducted on the determinants of personal integrity
abuse (state terror) (Tate, Keith and Poe 2007), this indicator has been consistently shown
to decrease state terror. It may be that it warrants separate treatment as an indicator of
judicial autonomy. Therefore, we are reluctant to downplay the importance of the
22
Along with Fiscal Autonomy, this provision is the least likely to be adopted across the entire twenty-five
year period. Further, some constitutions actually explicitly provide for these exceptional courts. Whether
these peculiarities account for the fact that this measure does not play well with the others is uncertain.
35
provision. Nevertheless, our purpose here is to explore the structure of the formal
indicators that do form coherent dimensions of judicial independence. Thus we have
repeated the analyses of Table 4 omitting the ban on exceptional courts indicator (see
Table 5).
Table 5: Principal Component and Varimax Factor Analyses of Eight Formal
Judicial Independence Measures*
Guaranteed Terms
Finality of Decisions
Exclusive Authority
Fiscal Autonomy
Separation of Powers
Enumerated Qualifications
Judicial Review
1st Principal
Component
(Formal
Judicial
Independence)
.766
.670
.742
.470
.835
.599
.713
Varimax
Factor
1
2
(Judicial Decision
Making)
(Organizational)
.375
.741
.774
.125
.813
.190
-.050
.784
.695
.473
.200
.689
.817
.141
.653
.343
.605
Hierarchical System
Extraction Method: Principal Component Analysis; second principal component
extracted is omitted.
Rotation Method: Varimax with Kaiser Normalization.
*Shading indicates that a variable’s highest principal component or Varimax loading was
on this factor.
Formal and Behavioral Independence
The first principal component from the factor analysis of the eight formal judicial
independence measures presented in Table 5 clearly demonstrates that they tap a common
underlying dimension that we can fairly call formal judicial independence. Thus we feel
very comfortable in creating a single measure of formal judicial independence using the
factor scores on the first principal component in Table 5.
36
In addition, Table 5 allows us also to identify and compute scores for two clear
subcomponents of overall formal judicial independence. Varimax Factor 1 in Table 5 is
clearly defined by Judges Decisions Final, Exclusive Authority, Separation of Powers,
and Judicial Review. Varimax Factor 2 is equally clearly defined by Guaranteed Terms,
Fiscal Autonomy, Enumerated Qualifications, and Hierarchical System. Naming factors
is always an exercise in art and discretion, of course, but we would argue that the analysis
reported in Table 5 reveals the existence of two subcomponents of judicial independence
having to do with judicial decision making matters (factor 1) and organizational or
structural matters (factor 2). The organizational and judicial decision making
subcomponents of formal judicial independence reflect many of the a priori
classifications given in the definitions and literature we reviewed in the conceptual
section of this article.
An immediate question that arises, of course, is how do the measures of formal
judicial independence relate to the behavioral measure we have developed? A quick
answer can be discerned from the bivariate correlations among the various judicial
independence measures reported in Table 6.
37
Table 6: Bivariate Correlations for Judicial Independence Indicators
Formal
Judicial
Independence
.061**
Judicial
Decision
Making
Subcomponent
-.051**
Behavioral Judicial Independence
(N = 4138)
.761**
Formal Judicial Independence
(N = 4911)
**Pearson correlation coefficient is significant at the 0.01 level (2-tailed).
Note: Omits correlations determined by definition or statistical procedure.
Organizational
Subcomponent
.155**
.649**
It is not surprising that the factor scores for the formal judicial independence
subcomponents are highly correlated the overall formal judicial independence measure
based on the first principal component: that is a statistical near necessity. What is
unanticipated are the weak correlations between our behavioral and formal measure of
judicial independence. Even more surprising is the small negative correlation between the
judicial decision making subcomponent and behavioral judicial independence.
While somewhat unexpected, a gap between promises and reality is hardly news,
and this one parallels the reported gap between promises and behavior of human rights
treaties signatories (e.g. Keith 2002, Hathaway 2004, Hafner-Burton and Tsutsui 2005).
To understand this gap between promise and reality, we may need to turn to the issue of
why states adopt certain provisions of law, and the body of research referred to as the
compliance literature in international relations and international law seems like a rich
theoretical resource (e.g. Risse, Ropp, and Sikkink 1999; Simmons 2000; Hathaway
2004; Hafner-Burton and Tsutsui 2005).
We decided to look another way at the relationship among the various judicial
independence measures and to plot them all over time on a common graph (Figure 3).
38
(Behavioral Independence is plotted against the right vertical axis; the three formal
independence factor scores are plotted against the left vertical axis.)
Figure 3: Behavioral and Formal Judicial Independence World Wide, 1981-2006:
Formal and Behavioral Scores
Figure 3 shows, as we have already seen, that the behavioral measure declines
over time, but we can clearly see that the overall and subcomponent formal measures
increase over time, with a sharp increase around the time of the democratization and
constitution writing wave of the 1990s, followed by relative stability after the late 1990s.
Thus as constitutional promises of judicial independence have gone up globally, the
achievement of judicial independence at the behavioral level has gone down.
39
Conclusion
Our Research
Our analyses document that judicial independence can be usefully conceptualized
and operationalized across nations and time (using eight measures of constitutional
provisions) as a single, unidimensional phenomenon. They also introduce and report on
the characteristics of a new behavioral indicator of judicial independence that promises to
allow scholars to analyze that elusive phenomenon, actual judicial independence, with
some replicability and validity across the globe for the last quarter century.
Structurally, formal judicial independence can also be validly viewed as
containing two core subcomponents, organizational and judicial decision making. The
organizational component is tapped by provisions for guaranteed terms of judicial office,
enumerated qualifications for judges, fiscal autonomy, and a hierarchical judicial system.
The judicial decision making component is captured by provisions for separation of
powers, for the judiciary to have finality within its decision making, for exclusive judicial
authority, and judicial or constitutional review. A constitutional ban against exceptional
courts, while arguably as important to judicial independence and even more important for
human rights, was not structurally linked to the principle components of judicial
independence here.
Behavioral judicial independence proved to be reliably codeable from field-based
assessments of the overall actual state of judicial independence provided by U.S. State
Department officers as a part of the process of assembling the annual reports that form
the Department’s Country Reports on Human Rights Practices. Validity checking of our
40
indicator against a similar but not identical indicator coded by Cingranelli and Richards
produced encouraging results.
Temporally, our analyses demonstrated a strong over time increase in formal
judicial independence, particularly with the end of the cold war and the new wave of
democratizing states. However, our analyses also demonstrated a simultaneous decrease
in behavioral judicial independence, creating a significant, clear gap between the formal
promises of judicial independence and achieved judicial independence.
While disappointing to those of us who would hope to see judicial independence
advancing both formally and behaviorally around the world, our findings actually parallel
those of studies of promises made in constitutions and international human rights treaties
and actual state human rights practices (see, for example, in regard to constitutional
promises, Boli-Bennett 1976; Pritchard 1986; Blasi and Cingranelli 1996; Davenport
1996; Keith 2002; Keith, Tate, and Poe 2008; in regard to treaties, see Keith 1999;
Hathaway 2004; Hafner-Burton and Tsutsui 2005). The degree of the gap and the trend
are quite similar.
The Way Forward
Judicial independence is likely a slowly developing norm, especially in newly
democratizing states. To understand its development, we will need to explore how it and
its various components and operationalizations vary across both space (regions, types of
nations) and time more fully in future analyses. It may be that the organizational
components of judicial independence are more transparent, more quickly and easily
achieved, than the customs of judicial independence related to judicial decision making,
41
which may take substantially longer to consolidate, not only in the newly democratizing
states of the Eastern Europe but also in post-colonial states globally.
Clearly there are many intervening factors related to both the domestic and the
international context that we will need to explore as we move toward fully specifying a
cross-national model of judicial independence. Countries may lack sufficient resources
to create the necessary infrastructure to establish an efficient, accessible judiciary or to
fund the judiciary and its staff at a level appropriate to deter corruption. Domestic
instability, particularly organized violence or civil wars, may result in states of
emergency or similar contexts in which the judiciary’s independence is constrained by
other actors or by the judges themselves.
There may also be a general lack of political will to implement constitutional
provisions for judicial independence, both on the part of the executive and legislative
branches, as well as the upper levels of the judiciary who may be reluctant to relinquish
their influence over lower court judges. Constitutional provisions may indeed represent
hollow promises, based on the regime’s short term self interest, rather than on true
commitments reflecting strong norms of constitutionalism and democratic governance.
For example, regimes may adopt such provisions to signal commitment to international
lending institutions, foreign investors, or bilateral aid organizations when they have little
intention of following through, once they have obtained the benefit they seek, particularly
if the cost of not complying is low, at least in the short term.
To understand this gap between promise and reality, we may need to turn to the
issue of why states adopt certain provisions of law, and the body of research referred to
as the compliance literature in international relations and international law seems like a
42
rich theoretical resource to pursue our similar question in regard to domestic institutions.
Thus while we are very pleased with our measures of judicial independence at this stage
of our research, it is clear that our pursuit of valid and replicable measures of judicial
independence will lead us well beyond the questions of conceptualization and
operationalization.
43
References
Ackermann, L.W. H. 1989. “Constitutional Protection of Human Rights: Judicial Review.”
Columbia Human Rights Law Review 21: 59-71.
Andrews, William G. 1964. Constitutions and Constitutionalism. New York: D. Van
Nostrand Company.
Becker, Theodore. 1970. Comparative Judicial Politics. Lanham, MD: University Press
of America.
Blasi, Gerard J. and David L. Cingranelli, 1996. “Do Constitutions and Institutions Help
Protect Human Rights?” In Cingranelli, ed. Human Rights and Developing Countries,
Greenwich: JAI.
Blaustein, Albert P. 1993 and continuing. Constitutions of the World. Nashville:
Carmichael and Carmichael.
Boli-Bennett, John. 1976. "The Expansion of National States. 1870-1970”. Ph.D.
dissertation. Stanford University.
Cingranelli, David L., and David L. Richards. 1999. "Measuring the Pattern, Level, and
Sequence of Government Respect for Human Rights” International Studies
Quarterly 43: 407-417.
Clark, David. 1975. “Judicial Protection of the Constitution in Latin America.” Hastings
Constitutional Law Quarterly 2: 405-442.
Cross, Frank. 1999. "The Relevance of Law in Human Rights Protection," International
Review of Law and Economics 87-98.
44
Cumaraswamy, Dato' Param. 195. Report of the Special Rapporteur on the independence
of judges and lawyers,. Submitted pursuant to Commission on Human Rights
resolution 1995/36.
Davenport, Christian, 1996. "’Constitutional Promises' and Repressive Reality: A
Cross-National Time-Series Investigation of Why Political and Civil Liberties are
Suppressed,’ Journal of Politics 58(3): 627-654.
Davenport, Christian. 1996. “‘Constitutional Promises’ and Repressive Reality: A CrossNational Time-Series Investigation of Why Political and Civil Liberties are
Suppressed,” Journal of Politics 58:627-54.
De Tocqueville, Alexis. 1966. Democracy in America, ed. J.P. Mayer. New York: Harper
and Row.
Elkins, Zachary, and John Sides. 2007. "Can Institutions Build Unity in Multiethnic
States?" American Political Science Review, 101: 693-708.
Elkins, Zachary, Tom Ginsburg, and James Melton. 2007. "The Structure and Fate of
Occupation Constitutions." William and Mary Law Review, 49: 1139-1178.
Elster, Jon, 1993. “Introduction” in Jon Elster and Rune Slagstad, eds. Constitutionalism
and Democracy. Cambridge: Cambridge University Press.
Ferejohn, John, Frances Rosenbluth, and Charles Shipan. 2004. “Comparative Judicial
Politics.” Manuscript available at: http://www.yale.edu/leitner/comparative judicial politics.pdf
accessed May 1, 2008
Finer, S. E., Vernon Bogdanor, and Bernard Rudden. 1995. Comparing Constitutions.
Oxford University Press.
45
Fiss, Owen. 1993. “The Limits of Judicial Independence.” University of Miami InterAmerican Law Review 25: 58-76
Flanz, Gisbert H. 1997 and continuing. Constitutions of Countries of the World. New York:
Oceana Publications.
Frühling, Hugo E. 1993. "Human Rights in Constitutional Order and in Political Practice in
Latin America." In Douglas Greenberg, et al., eds. Constitutionalism and
Democracy: Transition in the Contemporary World. New York: Oxford University
Press.
Gibney, Mark and Matthew Dalton. 1996. "The Political Terror Scale." In Human Rights
and Developing Countries, eds. Stuart S. Nagel and David Louis Cingranelli.
Greenwich, CT: JAI Press., inc
Ginsburg, Tom, Zachary Elkins, and Svitlana Chernykh. 2008. "Commitment and
Diffusion: How and Why National Constitutions Incorporate International Law."
University of Illinois Law Review, 201-238.
Guarnieri, Carlo and Patrizia Pederzoli. 2002. The Power of Judges: A Comparative Study
of Courts and Democracy. Oxford: Oxford University Press.
Hafner-Burton, Emilie and Kiyoteru Tsutsui. 2005. “Human Rights in a Globalizing
World: The Paradox of Empty Promises.”American Journal of Sociology 110(5):13731411.
Hathaway, Oona. 2004. “The Promise and Limits of the International Law of Torture.” In
Stanley Levinson, ed. Torture: A Collection. Oxford University Press.
Hathaway, Oona. 2004. “The Promises and Limits of the International Law of Torture” in
Sanford Levinson, ed., Torture: A Collection. Oxford: Oxford University Press.
46
Johnson, Kenneth. 1976. “Scholarly Images of Latin American Political Democracy in
1975.” Latin America Research Review 129-40.
Keith, Linda Camp, and Steven C. Poe. 2004. "Are Constitutional State of Emergency
Clauses Effective? An Empirical Exploration." Human Rights Quarterly 26(4):
1071-1097.
Keith, Linda Camp, C. Neal Tate, and Steven C. Poe. 2008. “Is the Law a Parchment
Barrier to Human Rights Abuse?” Unpublished manuscript
Keith, Linda Camp. 1999. “The United Nations International Covenant on Civil and
Political Rights: Does It Make a Difference in Human Rights Behavior?” Journal of
Peace Research 36:95-118.
Keith, Linda Camp. 2002. "Constitutional Provisions for Individual Human Rights (19761996): Are They More than Mere 'Window Dressing?'" Political Research
Quarterly (March) 55: 111-143.
Keith, Linda Camp. 2002. "Constitutional Provisions for Individual Human Rights (19761996): Are They More than Mere 'Window Dressing?'" Political Research Quarterly
(March) 55: 111-143.
King, Justice L.J. King. 1982. “Minimum Standards of Judicial Independence.” The
Australian Law School 58: 340-345.
Kornhauser, Lewis. 2002. “Is Judicial Independence a Useful Concept?” In Judicial
Independence at the Crossroads: An Interdisciplinary Approach, Stephen B.
Burbank & Barry Friedman eds. Thousand Oaks, CA: Sage Publications.
La Porta, Rafael, Florencio López-de-Silanes, Cristian Pop-Eleches, and Andrei Shleifer.
2004. “Judicial Checks and Balances.” Journal of Political Economy 112:445-470.
47
Larkins, Christopher M.
1996.
“Judicial Independence and Democratization: A
Theoretical and Conceptual Analysis.” American Journal of Comparative Law
44: 605-626.
Moderne, F. 1990. ‘Human Rights and Postcolonial Constitutions in Sub-Saharan
Africa’, in L. Henkin and A.J. Rosenthal (eds.), Constitutionalism and Rights: the
Influence of the United States Constitution Abroad. New York, Columbia
University Press.
Mutua, M. 2001. ‘Justice Under Siege: The Rule of Law and Judicial Subservience in
Kenya’, Human Rights Quarterly, Vol. 23, pp. 96-118.
Prillaman, William. 2000. The Judiciary and Democratic in Latin America. Westport, CT:
Praeger.
Pritchard, Kathleen. 1986. “Comparative Human Rights: An Integrative Explanation.”
Policy Studies Journal 15 (1): 110-128.
Ramseyer, J. Mark and Eric B. Rasmussen. 2003. Measuring Judicial Independence: The
Political Economy of Judging in Japan. Chicago: University of Chicago Press.
Ramseyer, J. Mark. 1994. “The Puzzling (In)Dependence of Courts: A Comparative
Approach.” Journal of Legal Studies 23: 721-753.
Risse, Thomas, Steven C. Ropp, and Kathryn Sikkink, eds. 1999. The Power of Human
Rights; International Norms and Domestic Change. Cambridge: Cambridge
University Press.
Rosenn, Keith S. 1987. “The Protection of Judicial Independence in Latin America.”
University of Miami Inter-American Law Review 19 (1):3-35.
48
Russell, Peter. 2001. “Towards a General Theory of Judicial Independence,” in Judicial
Independence In The Age Of Democracy, Critical Perspectives From Around The
World, Peter H. Russell & David M. O’Brien eds.
Shapiro, Martin. 1981. Courts: Comparative and Political Analysis. Chicago: University
of Chicago Press.
Siegan, Bernard. 1994. Drafting a Constitution for a Nation or Republic Emerging into
Freedom. Fairfax, VA: George Mason University Press.
Simmons, Beth A. 2000. "International Law and State Behavior: Commitment and
Compliance in International Monetary Affairs." American Political Science Review
94: 819-36.
Stotzky, Irwin P. 1993. "The Tradition of Constitutional Adjudication." In Transitions to
Democracy in Latin America: The Role of the Judiciary, ed. Irwin P. Stotzky.
Boulder, CO: Westview.
United Nations High Commissioner for Human Rights. 1985. Basic Principles on the
Independence of the Judiciary. Adopted by the Seventh United Nations Congress
on the Prevention of Crime and the Treatment of Offenders held at Milan from 26
August to 6 September 1985 and endorsed by General Assembly resolutions 40/32
of 29 November 1985 and 40/146 of 13 December 1985.
USAID. 2002. Guidance for Promoting Judicial Independence and Impartiality. Revised
edition. Washington DC: Office of Democracy and Governance, Bureau for
Democracy, Conflict, and Humanitarian Assistance, U.S. Agency for International
Development. Accessed May 3, 2008 at
http://www.usaid.gov/our_work/democracy_and_governance/publications/pdfs/pnacm007.pdf.
49
Appendix:
Examples of Judicial Independence Language in the State Department Human
Rights Reports with Ratings Assigned Sample Countries
The following excerpts provide examples of some of the language the State
Department uses in its assessment of judicial independence, generally in its first
summative evaluation of the Country. Sometimes, additional support for the assessment
follows in this section but not always. Additional information is sometimes provided in
the reports’ Right to Public and Fair Trial Section.
Benin: coded as (1) somewhat independent
The Government generally respected the constitutional provision for an
independent judiciary; however, the executive has important powers in regard to the
judiciary, and the judiciary was inefficient and susceptible to corruption at all levels.
The Constitution provides for an independent judiciary, and the Government
generally respected this provision in practice; however, the executive branch has
important powers with regard to the judiciary, and the judiciary remained inefficient in
some respects and susceptible to corruption at all levels
Botswana: coded as (2) fully independent
The Government generally respected the constitutional provisions for an
independent judiciary. The judicial system did not provide timely fair trials due to a
serious and increasing backlog of cases.
The Constitution provides for an independent judiciary, and the Government
generally respected this provision in practice.
50
Marshall Islands: coded as (0) not independent
The Constitution provides for an independent judiciary; however, judges are
appointed by the Cabinet, and past governments have attempted to influence the
judiciary.
The Government generally respected the human rights of its citizens, and the law
and the judiciary provide effective means of dealing with individual instances of abuse.
Actions to improve the Attorney General's office and the independence of the judiciary
improved the human rights situation.
Solomon Islands: coded as (1) somewhat independent
The Constitution provides for an independent judiciary; however, prior to
RAMSI's arrival, the judiciary was hampered by police ineffectiveness, lack of resources,
and threats against judges and prosecutors.
The Government did not encourage any judicial or independent investigation of
human rights abuses that occurred during the conflict, which contributed to a climate of
impunity. The judicial system functioned poorly during the first half of the year due to
the ongoing violence and a lack of resources.
Bulgaria: coded as (0) not independent
The Constitution provides for an independent judiciary; however, the judiciary
suffered from corruption and wide-ranging systemic problems.
The judiciary continued to struggle with wide-ranging systematic problems and
suffered from serious corruption.
Austria: coded (2) as independent
The judiciary is independent.
51