Rethinking Democracy:
How to Capture Democracy’s Empowering Nature
ABSTRACT
The core idea inspiring democracy is to empower people. To measure democracy in ways
that capture its empowering nature one needs to take into account rule of law as a state
quality making democracy effective. Based on this premise, we portray an index of
“effective democracy” and test its qualities against six alternative indicators of
democracy for some 180 states. We find that the index of effective democracy represents
best the empowering nature of democracy because it captures most clearly democracy’s
embedding in empowering conditions in the wider society. Specifically, effective
democracy is shown to be firmly embedded in (a) empowering socioeconomic conditions
that make people capable to practice democracy and (b) in empowering sociocultural
conditions that make people willing to practice democracy. In light of these findings,
people empowerment appears to be an entity of empowering societal conditions and
empowering regime characteristics, the latter of which are best depicted by the index of
effective democracy.
Key words: effective democracy – human empowerment – civil rights – rule of law.
Word count: 10,836
INTRODUCTION
The “global explosion” (Doorenspleet 2002) in the number of democracies that followed
the Third and Fourth Waves of Democratization (Huntington 1991; McFaul 2005) has
intensified scholars’ interest in the functioning and quality of the many new democracies
outside democracy’s traditional stronghold in the Western world (Adcock & Collier
2001; Bollen & Paxton 2000; Elkins 2000; Collier & Adcock 1999; Goertz 2006; Munck
& Verkuilen 2003). After initial enthusiasm scholars discovered pretty soon that the
quality of most of the newly emerged democracies falls far short of what is standard
among long established Western democracies. Since then researchers hold that a new
division between full democracy and partial democracy has become as important as the
old division between democracy and autocracy (Rose 2001; Ottaway 2003; Zakaria
2003). In describing democracies with deficient qualities various typologies have
emerged. Using terms like “illiberal” democracy, “deficient” democracy or “ineffective”
democracy (Diamond 2002; O’Donnell 2003; Merkel 2004), scholars attribute a
“diminishing adjective” to regimes in the hybrid zone between autocracy and fully
effective democracy (Collier & Levitsky 1997).
Prominent authors claim that state failure in enforcing rule of law and controlling
corruption is a major factor separating effective from ineffective democracies (O’Donnell
2004; Rose 2001; Warren 2006). Elaborating on this distinction, categorical approaches
that dichotomize effective against deficient democracies prevail (Merkel 2004). But
categorical approaches to state deficiencies have their own problems (Elkins 2001).
Implicitly they assume a bimodal distribution of deficiencies, such that given countries
either suffer or don’t suffer from a deficiency, even though in fact deficiencies might
1
differ by degree, establishing a continuum rather than a binary distinction (Bollen &
Paxton 2000). Yet, only one attempt has been made to use continuous data on state
deficiencies to produce a fine-graded index of “effective democracy” (Welzel, Inglehart
& Klingemann 2003:357). This index depreciates a given country’s level of democracy to
the extent that the state fails to establish rule of law, differentiating fully effective
democracies at one polar end from autocracies and completely ineffective democracies at
the opposite end, with many gradations in between (Inglehart & Welzel 2005:191-6).
However, the authors of the index do not demonstrate its conceptual strength in
ways satisfying the standards of proper concept formation, as outlined by Adcock and
Collier (2001) or Goertz (2006). Nor do they conduct a systematic validity test to
demonstrate the performance of the index in comparison with other established
democracy indices. Neither theoretically nor empirically have the merits of the concept of
effective democracy been laid out in sufficient clarity. Whether and in how far the index
of effective democracy is a better measure of democracy remains thus an open question.
This is a serious shortcoming in an era in which the rise of deficient democracies makes it
more pressing to identify effectively working democracies.
This article fills this gap, outlining the conceptual merits as well as the empirical
validity of the index of effective democracy, analyzing democracy data for some 180
states. The first part describes the theoretical rationales informing the concept of effective
democracy. The second part portrays the operationalization of the index of effective
democracy. The third part analyzes the empirical qualities of the index of effective
democracy in comparison with six alternative indices of democracy.
2
1. CONCEPTUALIZING EFFECTIVE DEMOCRACY
Adcock and Collier (2001) as well as Goertz (2006) outline best practice standards for
concept formation. Following these guidelines a proper conceptualization of the term
democracy should begin with the theorizing task, which is about to outline the “root
understanding” of what democracy means in a most general sense. Any further
elaboration of the concept must operate within the limits of the root understanding. The
task of definitional specification is to identify concrete subcomponents of the root
concept. The third task, operationalization, identifies the indicators that can be used to
measure the definitional components. In the fourth step, measuring, one gathers data on
the selected indicators and combines them in ways that are logically consistent with one’s
concept definition. Upon accomplishing this step, one arrives at a theory-grounded index
of democracy with empirical data ready to be analyzed. We explicate the concept and
measurement of effective democracy following these guidelines.
People Power as the Root Meaning of Democracy
In its literal meaning, “government by the people,” the ideal that ultimately inspires
democracy is to empower ordinary people to govern their lives (Arblaster 2002; Canovan
2006; Finer 1999; Holden 1992; Macpherson 1977; Philpott 1995; Sen 1999; Warren
2006). The literal meaning of democracy points to people empowerment as its root idea.
Academic definitions of democracy can differ from the literal meaning in that
they are more precise and elaborate. But they should be more precise and elaborate within
the semantic field of the literal meaning, not in establishing an alternative understanding.
Academic definitions of democracy should be elaborated in specification, not in
3
contradiction, to the term’s literal meaning. Otherwise, the term becomes a misnomer, in
which case its academically intended meaning will be difficult to communicate.
This is all the more true as the popular understanding of democracy largely
coincides with the term’s literal meaning. Recent survey data from mass publics around
the world show that ordinary people’s understanding of democracy emphasizes
everywhere the civil rights that empower ordinary people to govern their private and
public lives (Dalton and Shin 2006; Diamond 2003). People empowerment is the
common sense understanding of democracy.
In a historical perspective, too, people empowerment appears to be the root idea
that inspires people’s struggles for democracy. Modern democracy originates in the
liberal revolutions of the 18th century, which established partial democracy by such
rights-setting acts as the American Declaration of Independence in 1776 and the French
Declaration des Droits de L’Homme et des Citoyens in 1789 (Beetham 1999; Donnelly
2006; Finer 1999; Grayling 2007; O’Donnell 2003). These declarations entitled parts of
the public to practice personal and political freedoms, which was a decisive step into a
state in which significant shares of the population were empowered to govern their
private and public lives. This achievement established partial democracy because still a
majority of the adult population was excluded from suffrage. Yet, establishing partial
democracy encouraged further struggles of yet disempowered groups to become
empowered too, until universal suffrage gave birth to full democracies in a core of
Western nations early in the 20th century (Markoff 1996; McAdam, Tarrow & Tilly
2001).
4
Since then people’s struggles for democracy have continued and expanded: within
established democracies civil rights and equal opportunity movements have fought to
advance democracy’s empowering qualities (Tarrow 2003); beyond established
democracies, people power movements have pressured to replace autocracy with
democracy (Huntington 1991; Schock 2005; Thompson 2005). From the American
Revolution to the people power movements of today, movements for democracy have
invariantly been struggling for the civil rights that empower ordinary people (Ackerman
1991, 1998; Canovan 2006; Markoff 1996; Foweraker & Landman 1997; Karatnycky &
Ackerman 2005).
The literal meaning of the term democracy, its common sense understanding, and
the motives of the people struggling for democracy throughout its history, all suggest to
define the root meaning of democracy as “the empowerment of ordinary people to govern
their lives, in entitling them to follow their own preferences in their private lives and to
have their preferences count in public life.”
In its root meaning of people empowerment, democracy is not directly
observable. One can observe people empowerment only indirectly, by identifying the
institutional tools that are instrumental to people empowerment. Yet, one should not
reduce the definition of democracy to its observable institutional tools, abandoning the
idea inspiring them. It is still the idea of people power that informs us which institutional
tools are instrumental and in which order they are instrumental.
Keeping in mind that the features by which democracy becomes manifest are
instruments to empower people, prevents one from two types of misconceptions:
“electoral reductionism” and “unordered eclecticism.” Electoral reductionism is when
5
scholars limit the meaning of democracy to upholding regular competitive elections to
vote government into and out of office (Diamond 2001). Keeping in mind people
empowerment as the root meaning of democracy will prevent one from this reductionism
because it is evident that elections are just one out of many instruments to empower
people, next to a whole variety of political and personal rights that all help making the
people agents of their private and public lives.
Unordered eclecticism is when scholars define democracy by catalogues of
institutional features observable among existing democracies, without ordering these
features as to how instrumental they are to the core idea of people power. For instance,
personal and political rights are first-order instruments of democracy, for they are directly
instrumental to empower people, widening their decision making power in both their
private and pubic lives. Other institutional features of democracy, such as an independent
judiciary, exist to protect the citizens’ personal and political rights. Existing for this
purpose, such features are indirectly instrumental to empower people. They are secondorder instruments of democracy.
A concept of democracy that starts from democracy’s root meaning as people
empowerment does not base its empirical definition on just one instrument of democracy,
nor does it mix up first-order and second-order instruments. Instead, it focuses on firstorder instruments and encompasses them in as broad a scope as possible, avoiding both
reductionism and eclecticism.
Civil Rights as First-Order Instruments of Democracy
6
In reality democracies take on many institutional features. But some of these features are
more directly instrumental to the idea of people empowerment than others. The feature
most directly instrumental to the idea of people empowerment is civil rights: democratic
citizenship is eventually established by civil rights, as they entitle people to practice civil
freedoms, including personal freedoms to govern their private lives as well as political
freedoms to govern public life (Beetham 1999; O’Donnell 2003; Saward 2006; Williams
2006). In the form of personal rights, civil rights provide private freedom, that is,
freedom from external interventions into individual life decisions. Personal rights
empower people to act in a shielded sphere of autonomy into which no authority is
allowed to intervene. Personal rights include such freedoms as the freedom to choose
one’s religion, residence, sexual partner, occupation, and so on.
In state-organized societies people’s lives are affected in many ways by political
decisions. To govern one’s life in such a society does not only require a shielded sphere
of personal autonomy. Self-governance also requires the freedom to participate in the
political decisions that affect one. This public freedom is provided by political rights,
including the right to found and join political parties, to express political views freely in
public, to campaign for self-chosen political goals, and to run for public office, among
others.
Together, personal and political rights constitute civil rights. Since civil rights are
intended to empower people to govern their lives, they are democratic rights in the literal
meaning of the term. From the viewpoint of people empowerment, democratic rights are
the core definitional feature of democracy. Other institutional features of democracy such
as political pluralism or an independent judiciary, are derivates of a full set of operating
7
democratic rights. For instance, should the right of a citizen to have a free vote in
elections be operating, it must be tolerated that candidates compete for votes with
alternative policy programs. Necessarily this leads to political pluralism and competition.
Likewise, should the right of a citizen to be protected from abuses of executive power be
operative, there must be an independent judiciary that citizens can appeal to, to sue
executive authorities. We could continue with other examples but the major point is clear:
every institutional manifestation of democracy is derivative of people’s democratic
rights. The full set of known democratic rights is impossible to operate without a multiparty system, competitive elections, information pluralism and other features listed
elsewhere as institutional characteristics of democracy (Beetham 1999; O’Donnell 2003;
Sen 1999; Williams 2006).
Democracy’s Gradual Nature
Democracy is about people empowerment. But an absolute empowerment of people is, as
of now, only theoretically thinkable. It would be a state in which each resident’s private
autonomy would at any point in time be fully protected from collective interventions and
in which every collectively binding decision could at any point in time be submitted to a
popular vote in which every resident can participate with an equal vote. Empirically this
absolute level of democracy is unknown; we only know gradual approximations to it.
Likewise, an absolute disempowerment of the people is, as of now, only theoretically
thinkable. It would be a state in which an autocrat is at any moment in full control over
the lives of every resident. Again, in the empirical world we know only gradual
approximations to this absolute level of autocracy.
8
People empowerment is not a simple dichotomy such that people are either fully
empowered or entirely disempowered. Instead, people power constitutes a continuum
ranging from absolute autocracy at one extreme to perfect democracy at the other,
spanning a wide middle ground in which empowering and disempowering features
intermingle to create hybrid regimes. Because people power is by its nature a gradual
phenomenon, basing one’s understanding of democracy on people power unavoidably
establishes a gradual definition of democracy, not a dichotomous one (Bollen & Paxton
2000).
From this point of view, Sartori’s (1984) dictum that one has first to dichotomize
regimes as being either democratic or autocratic and only then to grade them within these
categories as to how autocratic and democratic they are, is mistaken. It overlooks that
autocracies and democracies differ in people empowerment and that this difference is a
continuum, so that one has to know first where on this continuum given regimes are
located before one can decide whether they are more autocratic or democratic. Grades as
to how autocratic or democratic given regimes are, are grades on one-and-the-same
underlying continuum: people power.
(Figure 1 about here)
To illustrate this point, Figure 1 assumes the ideal case of an interval scaled
people power index that ranges from zero, indicating the known minimum of people
power in the most autocratic regime, to 100, indicating the known maximum of people
power in the most democratic regime.1 The 50 percent mark constitutes a natural cutting
1
Alternatively, the endpoints of the continuum could demarcate an absolute, rather than an
empirical, minimum and maximum in people power. But since these absolute ends are
unknown and exist only in theory, it is impossible to assess where relative to them the
9
point: regimes below 50 percent disempower people more than they empower them and
hence are more autocratic than democratic. Regimes above 50 percent empower people
more than they disempower them and hence are more democratic than autocratic. The 25
and 75 percent marks also provide cut off points. The 75 percent mark divides
predominantly democratic regimes into those closer to the democratic maximum (above
75 percent) and those closer to the neutral point (below 75 percent). This allows
categorizing the former ones as rather “complete” democracies and the latter ones as
rather “incomplete” democracies. The 25 percent mark operates in similar fashion,
dividing the predominantly autocratic regimes into rather “incomplete” autocracies
(above 25 percent) and rather “complete” autocracies (below this mark).
Knowing where a regime is located on this people power continuum, one can
employ a dichotomous classification, separating rather autocratic regimes, between 0 and
50 points, from rather democratic ones, between 50 and 100 points. One can also use a
threefold classification, differentiating complete autocracies from 0 to 25 points, hybrid
regimes between 25 and 75 points, and complete democracies from 75 to 100 points. Or
one can decide for a four-fold classification, differentiating complete autocracies from 0
to 25 points, incomplete autocracies between 25 and 50 points, incomplete democracies
between 50 and 75 points, and complete democracies from 75 to 100 points. Whatever
classification one prefers, knowing where a regime is located on the people power
continuum precedes classification. This is Sartori’s logic reversed.
empirical minimum and maximum are located. Hence, it is preferable to define the full
range of people power by the empirically observed range rather than the theoretically
thinkable range of people power. As this range can change over time, the minimum of
zero should always demarcate the least people power ever observed, as much as the
maximum of 100 should always demarcate the most people power ever observed.
10
Defining democracy in terms of democratic rights illustrates this point easily.
Given a defined set of equally important democratic rights, democracy varies by degree
(1) due to the number of rights that are enacted and (2) due to the firmness by which they
are enacted. For instance, if one could list ten essential democratic rights and if each of
these rights could be classified as non-enacted, partially enacted, and fully enacted, one
could measure thirty different grades between the complete absence of any right and the
full presence of every right. It appears natural then to attribute a democracy-score of 0 to
a state that has not enacted a single democratic right as there is no people power at all
established in such a state. Likewise it appears natural to attribute a democracy-score of
100 to a state that has fully enacted every democratic right, for such a state establishes
people power to its known maximum, that is, hundred percent. Depending on how many
rights are enacted to what extent, given states fall somewhere on the wide middle-ground
between 0 and 100. Between these extremes the suggested marks at 25, 50, and 75 points
are easily identified, allowing one to separate complete autocracies, incomplete
autocracies, incomplete democracies and complete democracies. These categorizations
are substantively valid, as long as the 0 and 100 endpoints represent the minimum and
maximum of democratic rights in an absolute sense. Making these categorizations,
however, follows gradual measurement, not the other way round.
Rule of Law as the Effectiveness Factor of Democracy
There are features defining democracy and there are features making the defining features
effective. Democratic rights are the defining feature of democracy and for that matter
establish nominal democracy. But even if they are enacted to their full extent, with all
11
legal safeguards provided for, in the real practice of state power given democratic rights
might not be effectively respected. It needs other features than legally binding regulations
to make these regulations effectively respected. What matters are institutional practices.
When it comes to respect rights, the essential institutional practice is rule of law (Rose
2001; O’Donnell 2003). Consequently, there is a difference between democracy in
nominal terms and effective democracy. Democratic rights establish nominal democracy
but it needs rule of law to translate nominal democracy into effective democracy, such
that power holders respect legally enacted rights in their uses of state power.
Rule of law can be defined as government bound to the use of formal procedures
as defined by laws, excluding arbitrary and corrupt power practices that violate laws
(O’Donnell 2003; Warren 2006). Rule of law in this sense is not an exclusive attribute of
democracies. Government does not have to be democratic in order to be bound to the law.
Rule of law in this understanding does not separate democratic from autocratic
government. It separates arbitrary from principled government. Since the era of
“enlightened absolutism” in 18th century Prussia, the focus of rule of law is on principled
government, not democratic government (Dworkin 1986; Finnis 1980; Fuller 1964;
Holmes 2003).
In the republican tradition, rule of law is defined in ways that make it a
component of democracy (Shklar 1998; Stimson 2006). Such definitions focus on the
separation of powers, an independent judiciary, and the existence of civil rights. Indeed,
if one chooses to define rule of law in this way, it is hardly different from an operational
definition of democracy itself. The disadvantage of using the term rule of law in a
specifically democratic sense is that it cannot be used to distinguish law-bound from
12
lawless versions of autocracy. Nor can it be used to differentiate ineffective from
effective democracy when it is part of what defines democracy. For this reason we prefer
to define rule of law in the broader sense of rule bound to reliable formal procedures as
defined by laws. In this sense, rule of law is a universally applicable regime attribute that
is conceptually independent from democratic rights (Rawls 1971).
Effective Democracy as the Interaction of Democratic Rights and Rule of Law
Democratic rights and rule of law are conceptually distinct characteristics of political
regimes. Democratic rights are indicative of legitimate government; rule of law is
indicative of rational government. Still, the two features are inherently relevant to each
other.
Rule of law is relevant to democratic rights because these rights become
increasingly effective, in empowering the people, the more they are respected through
law-abiding institutional practices. Conversely, democratic rights are relevant to rule of
law: with an increasing number of rights the meaning of rule of law becomes more
substantive, for the domain of what falls under the rule of law widens with each
additional right to which the law applies.
Democratic rights and rule of law are inherently relevant to each other: as they
interact they produce effective democracy, spanning a continuum from not a single
democratic right being effective at all, at one end, to every democratic right being fully
effective, at the other end. Effective democracy in this sense is a matter of degree: it is
the extension of democratic rights insofar as rule of law sets them into real effect.
13
Rule of law has a varying potential to increase people empowerment, depending
on the given extent of democratic rights. To illustrate this point, let’s be simple: we
categorize regimes that possess more democratic rights than they miss as democracies
and regimes that miss more democratic rights than they possess as autocracies.2
Among autocracies, rule of law is a feature determining to what extent their
disempowering nature is alleviated, separating “unbound autocracy” when rule of law is
low from “bounded autocracy” when it is high. Among democracies, rule of law is a
feature determining to what extent their empowering potential is realized, separating
“ineffective democracy” when rule of law is low from “effective democracy” when it is
high. Thus, rule of law acts differently on the nature of autocracies and democracies: in
autocracies it works against their disempowering nature, in democracies it works in favor
of their empowering nature. The margin by which rule of law advances people power,
then, is contingent on the degree of democracy. In fact, this margin multiplies with the
extent of democratic rights enacted.
It should be intuitively clear that unbound autocracy empowers people the least
and effective democracy empowers them the most. It is less immediately clear, though,
whether people are less empowered in bounded autocracies or in ineffective democracies.
Intuitively, it might appear that ineffective democracy is the lesser evil, because to arrive
at effective democracy from there is less of a distance to bridge. It requires only a change
of the power holders’ ingrained power practices but not a change of basic regime
structures. By contrast, to get from bounded autocracy to effective democracy requires a
structural change of the regime. But if one thinks a little longer, a change of ingrained
2
This categorization presumes that there is a regime. When there are no democratic rights
and there is no regime, this is of course not autocracy but anarchy.
14
power practices might be less easily achieved than a change of legal regime structures.
For changing power practices is a cultural change that is less easily susceptible to human
engineering than the institutional remodeling of regime structures. Also, one can think of
extreme cases in which an almost complete absence of rule of law renders democracy so
ineffective that it leaves people just as disempowered as they are in some milder forms of
autocracy. An index of effective democracy should thus be constructed in a way that
allows for the possibility that—in extreme cases--an ineffective democracy scores as low
in effective people power as some autocracies.
2. MEASURING EFFECTIVE DEMOCRACY
The Logic of Interaction
A state might have enacted the full set of democratic rights known at a time, in which
case we consider it as 100-percent democratic. But if the state does not enforce rule of
law, the rights that define democracy are not set into real effect. More generally,
democratic rights are ineffective to the extent to which rule of law is absent. Even a fully
enacted set of democratic rights can be rendered entirely ineffective, if rule of law is
completely absent. In such a case the score for effective democracy should be at zero or
close to it, even if the regime is nominally democratic, reflecting the absence of effective
people power.
Vice versa, power practices in a given state might be entirely governed by the
law, in which case we assign the state a maximum score in rule of law. But the state
might have enacted few or no democratic rights, so no or few democratic rights are
effective. In this case, too, the score for effective democracy should be at zero or close to
15
it, again reflecting the absence of effective people power. Then, there are two reasons
why people are not effectively empowered: no democratic rights exist; or they exist but
absent rule of law renders them meaningless.
When two distinct features interact to generate a joint product, both of these
features are necessary, yet neither one is sufficient. The logic of necessary conditions
dictates that a low score in only one necessary condition suffices to make the joint score
equally low. With additive combinations this requirement fails because additive
combinations yield low joint scores only if both components are low. Addition models
the supplementation but not the interaction between components. Interaction can only be
modeled by multiplication. Hence, we model the interaction that produces effective
democracy by multiplying scores for democratic rights by scores for rule of law, as
shown in Figure 2. Multiplication is a mathematical expression of the necessary-butinsufficient condition logic.
(Figure 2 about here)
The multiplicative index construction implies that ineffective democracies can
end up with scores in effective democracy as low as autocracies. Does this make sense? It
does when effective people power is what is to be measured. And we think this definitely
is what is to be measured, as democracy is about people power. Under this premise, both
autocracies and ineffective democracies fail to empower the people. Ineffective
democracies fail to do so because lacking rule of law corrupts their empowering
potential. Autocracies fail to empower people because they don’t have an empowering
potential. For which of the two reasons people are not empowered, does not matter for a
measure of empowerment. What matters is the lack of empowerment.
16
Indicators and their Combination
Looking at indicators for democratic rights, the freedom ratings by Freedom House are an
obvious choice (Freedom House 2007). These freedom ratings include two concepts
roughly equivalent to the two sets of democratic rights we are interested in. Freedom
House’s “civil liberties” ratings mostly cover private freedoms and are thus roughly
equivalent to our concept of personal rights. The organization’s “political rights” ratings
cover public freedoms, coinciding with our own notion of political rights. The two
indicators correlate at r=.95 (N=190) and we combine them additively to obtain a scale of
democratic rights, transformed into a 0 to 100 range, yielding scores for nominal
democracy or nominal democratic rights. A score of 0 is indicative of regimes that
haven’t enacted a single democratic right; a score of 100 is indicative of regimes that
have enacted the fully known set of democratic rights at our time.
The most encompassing measure of rule of law is the World Bank’s “rule of law”
index (Kaufman, Kraay & Mastruzzi 2007). Measuring rule of law in terms of
government bound to laws, this index is based on data from more than a dozen different
sources, mostly based on expert ratings. Strongly overlapping with rule of law is another
indicator among the World Bank’s “good governance” scores, labeled “control of
corruption.” Corruption is an inverse indicator of rule of law, for it involves the rule of
cleptocraatic power practices that violate laws (Warren 2006). The “rule of law” and
“control of corruption” scores correlate at r=.95 (N=190) and so we average them to
obtain an overall scale of rule of law, transformed into a weighting scale with minimum 0
17
for the country with the lowest known level of rule of law and 1.0 for the highest known
level of rule of law.3
One might maintain that the Freedom House ratings do include information on
rule of law, making our efforts to weight the Freedom House ratings for rule of law data
superfluous, if not tautological. But this assumption is mistaken because the way in
which Freedom House includes rule of law information is as insufficient and as it is
inadequate. This can be substantiated in three points.
First, only two points on Freedom House’s 28-point checklist refer to rule of law,
giving this aspect the weight of a fourteenth in the index construction. This proportion is
as minor as it is arbitrary. Second, the combinatory logic is flawed. Rule of law is treated
as a supplementary feature that adds a minor quality on democratic rights when in fact it
is a substantiating feature that interacts with each single democratic right. The adequate
combination to measure effective democratic rights is to weight all democratic rights a
country has enacted for rule of law. Third, Freedom House does not include rule of law
information in a controlled way using standardized data. For these reasons Freedom
House is a good deal away from measuring effectively respected democratic rights. There
is a bias towards measuring just formally enacted rights, even though self-declared
intentions tell otherwise.
3
We are confronted with an unresolvable problem here: ascribing a value of 0 to the
lowest known level of rule of law assumes that this is the entire absence of rule of law,
the absolute 0-point. But this absolute 0-point is only theoretically thinkable and we don’t
really know how close the lowest observed level comes to it. In other words, the lowest
level is distant from 0 but we cannot know how far. Assigning a positive value to the
lowest level would thus be as arbitrary as leaving it at 0. We prefer the latter as it
provides us with a theoretically preferable scale range.
18
There is clear evidence for this claim, provided by Freedom House’s dichotomous
distinction of countries into “electoral democracies” and “non-democracies.” This
classification is conducted on purely formal grounds, looking at whether countries hold
competitive elections regularly. Into which of these two categories a country falls,
explains fully 77 percent of the variation in Freedom House’s fine-graded democratic
rights ratings. This is a very high percentage of explained variance for a binary predictor,
indicating a strong bias in the democratic rights ratings towards formalism rather than
effectiveness.
Still another indication that the Freedom House ratings are far from fully
incorporating rule of law information is this: Freedom House’s political rights ratings and
the World Bank’s rule of law scores have most of their variance (60 percent to be precise)
unshared. Thus, democratic rights data and rule of law data are sufficiently distinct from
each other to interact in meaningful ways that avoid tautology.
To obtain a measure of democratic rights as they are set into effect by rule of law,
we multiply the 0-100 scores for democratic rights by the 0-1.0 scores for the rule of law,
obtaining weighted percentages for effective democratic rights.
3. VALIDATING EFFECTIVE DEMOCRACY
Distributional Characteristics
If one cuts into half the rule of law and democratic rights scales, one obtains four
quadrants as shown in Figure 3. The split on the democratic rights scale divides regimes
into rather autocratic ones (below 50 points) and rather democratic ones (above 50
19
points). The split on the rule of scale divides regimes into rather unlawful ones (below .5)
and rather lawful ones (above .5).
(Figure 3 about here)
In combination, unlawfulness and autocracy constitute “unbounded autocracies”
in the lower left quadrant of Figure 3. Most examples of this regime type are found in
Sub-Saharan Africa and the Middle East but also in South and Central Asia. Lawfulness
and autocracy combine to create “bounded autocracies” in the upper left quadrant. Apart
from Singapore, an outstanding example of this regime type, bounded autocracies are
mostly found among the oil-exporting monarchies of the Middle East.
Unlawfulness and democracy combine to create “ineffective democracies” in the
lower right quadrant of Figure 3. This is where many of the younger democracies are
located, including most of the democracies in Latin America, Eastern Europe, and Africa
but also some long established democracies, provided their level of economic
development is low, as in India. Lawfulness and democracy merge into “effective
democracies” in the upper right quadrant of Figure 3. Here we find all old democracies in
rich societies but also a number of young democracies, provided they are economically
advanced, such as Taiwan, South Korea, Chile or Uruguay.4
Figure 3 reveals that countries with more democratic rights tend to score higher in
rule of law, indicating that expanding democratic rights brings with it a tendency to
advance rule of law (which is logical insofar as the idea of rights is inherently targeted at
rule of law). But this tendency is far from being unavoidable: most of the variance in rule
4
Still another distinctive group of states to be found among effective democracies are
small tropical island states, such as Saint Lucia, Saint Vincent, Barbados or the
Bahamas.
20
of law (56%) is unrelated to democratic rights. To be sure, most autocracies are found in
the lower half of the rule of law scale, making bounded autocracy an exceptional regime
type (covering 13 of the 71 autocracies worldwide). But democracies are split by half on
the rule of law scale, yielding an almost equal number of ineffective democracies (N=52)
and effective democracies (N=59). Only at the top level of nominal democratic rights is
rule of law so pronounced that effective democracies outnumber ineffective ones.
(Figure 4 about here)
Figure 4 shows how differences in rule of law translate nominal democratic rights
into effective democratic rights. The distribution is strongly curvilinear and highly
heteroskedastic. In the lower seven deciles of nominal democratic rights, even large
increases pay off only minimally in effective democratic rights, achieving 35 percent
points at best, as in the case of India. Only in the top three deciles of nominal democratic
rights, do increases tend to pay off greatly in effective democratic rights, making
achievements of up to 100 percent points possible. But these payoffs are far from being
sure, producing a wide range of effective democratic rights on top levels of nominal
democratic rights.
The payoff in effective democratic rights is predictable but minor at low to mid
levels of nominal democratic rights. By contrast, the payoff can be major but is hard to
predict at high levels of nominal rights. As it should be, nominal democratic rights
operate as a necessary but insufficient condition to yield effective democratic rights.
The scoring in effective democratic rights stretches the differentiation between
effective democracy and ineffective democracy. It stretches it so much even that quite a
number of ineffective democracies end up with scores in effective democratic rights as
21
low as or even lower than some autocracies. This is not to mean that ineffective
democracies have fewer democratic rights than autocracies. The fact that they are
categorized as democracies tells us exactly the contrary. But lacking rule of law corrupts
these rights so much in some cases that, in effective terms, people are left as
disempowered as in some autocracies.
Two extreme cases, Singapore and India, are illustrative. In terms of nominal
democratic rights, India falls on the democratic half of the scale, scoring at 75 percent
points. Singapore, by contrast, is on the upper edge of the autocratic half in nominal
democratic rights, scoring at 40 percent (so it is not a complete autocracy like North
Korea). But Singapore scores very high in rule of law (.90), making its 40 percent score
in democratic rights almost fully effective: the score for effective democratic rights is 36
points. India, for its part, scores pretty poor in rule of law (.45), rendering its 75 percent
score in democratic rights largely ineffective: the resulting score in effective democratic
rights is 34, two points below Singapore.
At first glance, it might seem strange that a nominally mostly democratic regime,
India, can end up with a score in effective democratic rights as low as a mostly autocratic
regime, Singapore. But when one focuses on people’s effective empowerment, it is
perfectly possible that lacking rule of law corrupts democratic rights to an extent that
leaves people as disempowered as in an autocracy. There is no doubt that in nominal
terms Indians have more democratic rights than Singaporeans. But it seems that
Singaporeans can use their fewer rights more effectively than the Indians can use their
wider rights. In terms of effective power, the two populations are similarly
22
(dis)empowered. Different combinations of regime characteristics can have the same
outcome in terms of effective people power.
(Figure 5 about here)
The boxplot in Figure 5 shows to what extent the four regime types generated in
Figure 3 differ in their scoring in effective democratic rights. As one would expect,
unbound autocracies score on the bottom of effective people power, with an average
score of 7.6. Ten points higher up follow the bounded autocracies at a mean of 17.4.
Another ten points higher up follow the ineffective democracies at a mean of 27.8. Thus,
ineffective democracies do on average perform significantly better in effective people
power than bounded autocracies. They overlap only in exceptional cases such as
Singapore and India. Apart from that, the most decisive feature of Figure 5 is how far off
the effective democracies are located from the other three regime types: their mean score
in effective people power is 69.3 (40 points above ineffective democracies). The
differences among the other three regime types, though clearly recognizable, are minor
compared to the empowerment gap that separates them from effective democracies.
In our measurement perspective, the empowerment gap between ineffective
democracy and autocracy is not nearly as large as the gap between effective democracy
and everything else. How valid a depiction of societal reality is this measurement
perspective?
How validly a measurement captures societal reality, is a matter of how indicative
the measurement is of other relevant aspects of societal reality that are not themselves
part of the measurement but are conceptually linked to it. This is the approach taken in
23
construct validity, often also called criterion validity or nomological validity (Adcock and
Collier 2001; Elkins 2000; Goertz 2006).
Democracy is about people power, so human empowerment is its conceptual link
to other aspects of reality. In the logic of criterion validity, then, effective democracy is a
valid measure of people power when it is indicative of other aspects of human
empowerment. And it is the more valid a measure of people power, the more indicative it
is of these other aspects of human empowerment. In this context, the framework of
human empowerment proposed by Welzel and Inglehart (2008) offers a suitable approach
to test the capacity of various democracy indices to measure democracy’s root meaning,
people power. In this scenario, among various democracy indices the one showing the
closest link to other aspects human empowerment is the most valid one in measuring
democracy’s empowering nature.
Democracy’s Embedding in an Empowering Environment
There is a strong tradition in democratic theory from Lipset (1959) to Dahl (1973) to
Putnam (1993) due to which democracy prevails as a function of basic societal
conditions. This tradition goes as far back as Aristotle (1984 [350 BC]) who claimed in
Book IV of Politics that democracies are to be found in relatively egalitarian, middleclass centered societies whose citizens emphasize civic virtues. In this understanding,
democracy is a societally embedded phenomenon (Merkel 2004). Since then scholars
have identified various socioeconomic and sociocultural conditions as embedding factors
of democracy, four of which are particularly prominent in the literature.
24
Among the socioeconomic conditions, a high level of economic development has
most widely been discussed as an embedding factor of democracy (Lipset 1959; Bollen
and Jackman 1985; Diamond 1992; Burkhart & Lewis-Beck 1994; Boix & Stokes 2003).
But distributional equality in basic human resources has also been emphasized as an
embedding factor of democracy (Acemoglu & Robinson 2006; Dahl 1971; Muller 1997;
Vanhanen 2003; Boix 2003). Among the sociocultural conditions, a civic culture that
emphasizes emancipative values has been claimed to be an embedding factor of
democracy (Almond & Verba 1963; Inlgehart 1997; Inglehart & Welzel 2005; Putnam
1993; Welzel Inglehart & Klingemann 2003; Welzel 2007). Closely related to this,
widespread civic engagement as the motor of a vibrant civil society has also been
championed as an embedding factor of democracy (Putnam 1993; Foweraker & Landman
1997; Anheier et al. 2004; Karatnycky & Ackerman 2005; Bernhard 2008).
These four embedding factors of democracy are not mutually exclusive. In fact, it
is likely that they are strongly intertwined and re-enforce each other in establishing an
entire embedding environment of democracy. Accordingly, Figure 6 integrates the four
embedding factors and their core, democracy, into one framework highlighting human
empowerment as the underlying theme.
(Figure 6 about here)
Effective democracy constitutes the center piece in this human empowerment
framework. Effective democracy is an empowering type of regime, entitling people to
practice personal and political rights. But an effective democratic regime is not the only
contribution to empower people. For people power does not only depend on legal
entitlements: people might have abundant entitlements but if they lack the resources that
25
make them capable to practice these entitlements or if they lack the values that make
them willing to practice them, people still lack empowerment. Thus, empowerment not
only has an institutional dimension that works on the level of entitlements; it also has a
socioeconomic dimension working on the level of capabilities as well as a sociocultural
dimension working on the level of values and habits. There is a socioeconomic and a
sociocultural empowerment context, in the center of which we place an empowering
regime, that is, democracy.
An empowering sociocultural context consists of certain values and habits. Values
have the potential to empower people motivation-wise. Values fulfill this potential when
they make people believe in their efficacy, in popular participation, in tolerating nonconformity and in trusting other people (Almond & Verba 1963; Putnam 1993). We use
Welzel’s (2007:403) index of emancipative values based on World Values Survey data to
measure the joint societal-level variation over four component attitudes, including beliefs
in personal efficacy, popular participation, tolerating non-conformity and trusting other
people. Societal-level variation over these attitudes reflects a common underlying
dimension indicating a cultural climate that emphasizes an ideal of people as efficacious,
participant, to-be tolerated and to-be-trusted individuals (see Internet-Appendix for
details).
Habits are behavioral manifestations of values and for this reason constitute
another aspect of culture. In an empowerment perspective, habits are important as an
indication of the extent to which people exert agency, express their concerns, and behave
assertively in relation to authorities. To the extent people do so, they have internalized
empowering habits. The frequency of people-initiated forms of civic engagement
26
indicates the societal radius of empowering habits in this sense. To measure civic
engagement we use World Values Survey data, calculating per country the percentage of
respondents reporting to have participated in such non-violent actions as petitions,
boycotts and demonstrations.
The socioeconomic context of democracy is manifest in a country’s level of
economic development. Higher levels of economic development increase people’s
participatory resources, such as their incomes and skills, making them more capable to
practice democratic rights. We measure economic development by per capita GDP in
purchasing power parities.
It is not only the aggregate stock of participatory resources that is important.
Resource dispersion is important as well, with more widely dispersed resources
empowering a wider circle of people. To cover this aspect we use Vanhanen’s (2003)
index of “power resources,” which measures the dispersal of material, intellectual, and
social resources (see Internet-Appendix for details).
Values empower people on the level of mentalities while resources empower
people on the level of capabilities. These empowering sociocultural and socioeconomic
conditions are closely intertwined, indicating that human empowerment is an experiental
entity of empowering mentalities and empowering capabilities. The factor analyses in
Table 1 support this claim, evidencing that all four aspects of an empowering societal
context represent just one over-arching factor of human empowerment.
(Table 1 about here)
Democracy in the narrow sense of a democratic regime is just an empowering set
of rights. But democracy in the broad sense of a democratic society is more than just an
27
empowering regime: it is an empowering regime within an empowering environment.
Democracy should be deeply anchored in an empowering social environment, for
democratic rights will not be widely practiced unless people are means-wise capable and
motivation-wise stimulated to practice them. This conception integrates socioeconomic
and sociocultural explanations of democracy into a coherent framework, emphasizing the
underlying theme of human empowerment.
Nominal democracy can be imposed by domestic elites or foreign powers fully
irrespective of empowering conditions in the wider society. Effective democracy,
however, should depend on exactly these empowering conditions. For if a public has the
resources that make it capable of practicing democracy and if it has the values that make
it willing to do so, it also has the power to raise popular pressure on elites so as to make
them and keep them effectively accountable to the people (Welzel 2006; 2007; Welzel &
Inglehart 2008). Hence, measures of effective democracy should show a closer
connection with empowering socioeconomic and sociocultural conditions than other
measures of democracy.
Our test question is which measure of democracy is mostly indicative of societal
conditions of empowerment other than democracy itself. This question involves no
causality assumption and can be solved by showing which measure of democracy is the
closest associated with societal conditions of empowerment. The direction of this
association is strictly speaking irrelevant. To make the point that democracy is indicative
of societal conditions of empowerment, it does not matter whether this is so because
democracy is a consequence, a product or just a correlate of empowering societal
conditions. Thus, the conclusiveness of our test is entirely unaffected by a particular
28
causal interpretation of the relationship between democracy and societal conditions of
empowerment. Yet, since Lipset (1959) most of the literature on societal prerequisites of
democracy sees democracy as the product rather than the precondition of societal
conditions of empowerment, assuming these conditions to predate democracy. To
represent this dominant assumption in the work on societal prerequisites of democracy,
we measure democracy over the most recent period in time and relate it to measures of
societal conditions of empowerment that predate it.
The Evidence
We have four indicators covering different aspects of an empowering societal
environment: economic development, distributional equality, emancipative values and
civic engagement. Figure 6 uses these four indicators separately and three different
combinations of them, each covering a broader set of empowering conditions, including
an empowering socioeconomic context (economic development and distributional
equality), an empowering sociocultural context (emancipative values and civic
engagement), as well as the entire empowering context (all four indicators). As the factor
analyses in Table 1 have shown, these conceptual summaries are empirically justified as
there is one underlying dimension of contextual empowerment, reflecting that
empowering socioeconomic conditions are intertwined with empowering sociocultural
conditions: empowering capabilities go together with empowering mentalities.
Figure 6 correlates each of the seven measures of empowering societal conditions
with seven different measures of democracy, including our index of effective democracy.
The other six democracy measures include (in descending order of the bars in Figure 6):
29
the Polity IV “autocracy-democracy” scores as of 2000-3; the “empowerment rights”
measure of Cingranelli and Richards as of 2000-4; the Freedom House democratic rights
ratings as of 2000-6; Vanhanen’s “index of democratization” as of 2001; the Economist
intelligence unit’s “democracy index” as of 2006; and the World Bank’s “voice and
accountability” measure as of 2000-6. The bottom bar in each of the seven sets of bars
represents our index of effective democracy as of 2000-6. The length of each bar on the
horizontal scale indicates how strongly the democracy measure represented by this bar
correlates with the respective aspect of contextual empowerment, as labeled on the
vertical axis. For instance, the length of the bottom bar in the bottom set of bars indicates
that our measure of effective democracy correlates at r=.93 with the measure of a
society’s entire empowering context.
(Figure 6 about here)
The pattern depicted in Figure 6 provides some clear lessons. Regardless which
measure of an empowering societal context one uses, all democracy measures are always
highly significantly and positively correlated with empowering conditions in wider
society. And for all measures of an empowering context, there is a similar pattern
considering with which indicator of democracy it correlates weakest and which it
correlates strongest. The Polity autocracy-democracy scores always show the weakest
correlation and the CIRI empowerment rights measure always shows the second-weakest
correlation with empowering contexts. The Freedom House democratic rights ratings
usually show the third-weakest correlation among the seven democracy indicators. As
regards the second-strongest correlate, the World Bank voice and accountability measure,
30
the Economist democracy index, and the Vanhanen index of democratization are usually
very close.
However, the World Bank and Economist democracy measures are unusual
contenders. These organizations stretch the boundary of democracy beyond democracy’s
definitional limitation to institutions, including features that are part of the societal
context of democracy but not part of democracy itself. For instance, the Economist
includes civic activity data from the WVS and this creates a tautological correlation with
civic engagement as one aspect of an empowering context. Compared to democracy
measures that remain within the limits of an institutional definition of democracy, the
World Bank and Economist democracy measures are unfairly superior competitors when
it comes to depict democracy’s association with empowering societal contexts. For the
completeness of the picture they are nevertheless included.
By contrast, the index of effective democracy remains within the limits of an
institutional definition of democracy and has no tautologically inbuilt correlation with
democracy’s empowering contexts. And yet it is the index of effective democracy that
shows the strongest correlation with each measure of an empowering context, among all
seven measures of democracy. It is particularly noteworthy how clearly the index of
effective democracy outperforms the Freedom House democratic rights ratings, which is
one of its components. Hence, our refinement of Freedom House data with rule of law
data does bring democracy’s embedding in empowering societal conditions clearer to the
fore, and it does quite strongly so.
These findings suggest that more than any other measure of democracy, the index
of effective democracy is indicative of empowering qualities in democracy’s wider social
31
context. In an empowerment perspective, then, the index of effective democracy appears
to be the most valid measure of democracy.
One possible interpretation of this finding is that effective democracy depends
more than other aspects of democracy on empowering conditions in wider society. The
regression analyses in Table 2 seem to confirm this suggestion. Variation in nominal
democracy is explained to 57 to 63 percent by variation in empowering societal
conditions (see Models 1 and 3), while variation in effective democracy is explained to
86 or 87 percent by variation in empowering societal conditions (see Models 5 and 7).
(Table 2 about here)
But what happens when one considers reverse causality, taking into account the
extent to which empowering societal conditions are themselves shaped by democracy of
the prior period? This is tested by including a lagged version of the dependent variable,
nominal democracy and effective democracy, measured at the beginning of the period
over which empowering societal conditions are measured. Doing so indicates how
strongly empowering societal conditions shape democracy of the subsequent period after
one has partialed out the extent to which these conditions are themselves the product of
democracy of the preceding period. To the extent to which empowering conditions
continue to show an effect on subsequent democracy, this effect is genuinely independent
and, as far as one can say, of causal quality.
With effective democracy, the impact of empowering societal conditions clearly
drops as one includes the lagged dependent variable (see Models 4 and 6). This is due to
the fact that empowering societal conditions are partly a product of effective democracy
of the preceding period. Still, this is not entirely but only partly so and hence empowering
32
societal conditions retain a highly significant and strongly positive impact on effective
democracy of the subsequent period, even controlling for effective democracy of the
prior period. For nominal democracy the pattern is different (see Models 2 and 4): after
controlling for the lagged dependent variable, the impact of empowering societal
conditions drops drastically and becomes almost insignificant. Evidently, nominal
democracy is much less a function of empowering societal conditions than is effective
democracy. This underlines once more that effective democracy is the more valid
measure of democracy’s core function: empowering people.
(Figure 8 about here)
Results of seven multivariate regressions, one for each of the seven democracy
measures as the dependent variable, confirm the previous result from a slightly different
angle of model specification. To take into account the extent to which empowering
conditions are themselves shaped by prior democracy, we do at this time not include the
lagged dependent variable but Gerring et al.’s (2005) “democracy stock” variable, as of
1995. This variable measures the number of years a country has spent under democracy5
and is supposed to capture any long-term effect of democracy on empowering societal
conditions.
In Figure 8, a post-2000 measure of each indicator of democracy is regressed on
our measure of the entire empowering context before 2000 and the democracy stock
accumulated until the earliest point of the period over which the empowering context is
measured, i.e., 1995. The bars in Figure 8 show for each of these seven regressions the
5
In fact, the measure is a bit more sophisticated, depreciating the count of each
democratic year by one percent for each year it fades to the past from the baseline of
1995. We are grateful to John Gerring for providing us the data on this index.
33
percent of variance in the respective democracy index explained by the empowering
societal context, after we have partialed out the extent to which the empowering context
is a product of long-term democracy. As one can see, whereas variation in nominal
democracy (see the bar for the Freedom House ratings) is explained to only 35 percent by
empowering societal conditions, variation in effective democracy is explained to 65
percent by empowering societal conditions, accounting for these conditions’ own
dependence on long-term democracy. Again, effective democracy depicts best
democracy’s embedding in empowering societal conditions. It appears to be the measure
most clearly capturing democracy’s root meaning: people power.
CONCLUSION
We argue that the core idea inspiring democracy is to empower people and that to
measure democracy in ways capturing its empowering nature one needs to take into
account rule of law as a state quality that makes democracy effective. By itself rule of
law does not make a country democratic but it does make nominal democracy effective.
Following these rationales we created an index of “effective democracy” in that we
weighed scores for nominal democracy by scores for rule of law, depreciating democracy
to the extent rule of law is lacking.
Inspecting the distributional features of the index of effective democracy, it
became obvious that lack of rule of law depreciates the scoring of many nominally
democratic countries so much that their effective democracy levels are found to be as low
as those of autocracies. Under recognition of democracy’s purpose to empower people,
we found this perfectly appropriate. For democracies that lack of rule law and hence fail
34
to set democratic rights into effect, do as little to empower people effectively as do some
milder versions of autocracies.
The internal logic of a measurement concept is one criterion of its quality.
Another one is a concept’s external validity, that is, its relation to other aspects of reality-aspects that are theoretically linked to the concept but not a definitional part of it.
Democracy is about people power, so empowerment is democracy’s theoretical link to
other aspects of reality. Other such aspects of reality, which are not themselves part of
democracy but empower people, include socioeconomic conditions that make people
capable to practice democratic freedoms, and sociocultural conditions that make them
willing to practice democratic freedoms. Measured against this human empowerment
framework, the externally most valid index of democracy is the one that shows the
closest association with empowering socioeconomic conditions and empowering
sociocultural conditions. This is at the same time the index measuring the empowering
nature--and thus the core--of democracy the best.
In all statistical tests with seven different measures of democracy, the index of
effective democracy turned always out to be the one that is most strongly associated with
empowering conditions in wider society among some 180 states. Further tests of the
direction of these associations suggest that effective democracy is depending on
empowering societal conditions rather than creating them. This result makes sense: as a
regime designed to empower people, democracy should depend on empowering societal
conditions—conditions needed to make people capable and willing to practice
democracy.
35
Yet, fully irrespective of the direction in the relation between democracy and
empowering contexts, effective democracy is the measure of democracy most valid in the
sense of depicting democracy’s empowering nature, that is, its relation to empowering
conditions in wider society. Finally, our findings indicate that human empowerment is an
entity and that democracy—or effective democracy at least—is an integral part of it.
REFERENCES
Acemoglu, D. & J.A. Robinson. 2006. Economic Origins of Dictatorship and
Democracy. New York: Cambridge University Press.
Ackerman, B. 1991. We the People I: Foundations. Cambridge: Harvard University
Press.
Ackerman, B. 1998. We the People II: Transformations. Cambridge: Harvard University
Press.
Adcock, R. & D. Collier. 2001. “Measurement Validity: A Shared Standard for
Qualitative and Quantitative Research.” American Political Science Review 95
(3): 529-45.
Anheier, H.K., M. Glasius & M. Kaldor. 2001. Global Civil Society 2001. Oxford:
Oxford University Press.
Almond, G.A. & S. Verba. 1963. The Civic Culture. Princeton: Princeton University
Press.
Arblaster, A. 2002. Democracy. London: Open University Press.
Beetham, D. 1999. Democracy and Human Rights. Cambridge: Polity Press.
36
Berlin, I. 1969. “Two Concepts of Liberty.” In H. Hardy (ed.), Isaiah Berlin – Liberty.
Oxford: Oxford University Press.
Bobbio, N. 1987. The Future of Democracy. Minneapolis: University of Minnesota Press.
Boix, C. 2003. Democracy and Redistribution. New York: Cambridge University Press.
Boix, C. & S.L. Stokes. 2003. “Endogenous Democratization.” World Politics 55 (July):
517-49.
Bollen, K.A. & R.W. Jackman. 1985. “Political Democracy and the Size Distribution of
Income.” American Sociological Review 50 (August): 438-57.
Bollen, K.A. & P.M. Paxton. 2000. “Subjective Measures of Liberal Democracy.”
Comparative Political Studies 33 (2): 58-86.
Burkhart, R.E. & M.S. Lewis-Beck. 1994. “Comparative Democracy: The Economic
Development Thesis.” American Political Science Review 88 (December): 903910.
Canovan, M.. 2006. “The People.” In J.S. Dryzek, B. Honig & A. Phillips (eds.), The
Oxford Handbook of Political Theory. Oxford: Oxford University Press, 349-362.
Collier, D. & S. Levitsky. 1997. “Democracy with Adjectives: Conceptual Innovation in
Comparative Research.” World Politics 49 (3): 430-51.
Collier, D. & R. Adcock. 1999. “Democracy and Dichotomies: A Pragmatic Approach to
Choices about Concepts.” Annual Review of Political Science 2: 537-65.
Dahl, R.A. 1971. Polyarchy: Participation and Oppostion. New Haven: Yale University
Press.
Donnelly, J. 2006. “Human Rights.” J.S. Dryzek, B. Honig & A. Phillips (eds.), The
Oxford Handbook of Political Theory. Oxford: Oxford University Press, 601-620.
37
Dworkin, A. 1986. Law’s Empire. Cambridge: Harvard University Press.
Elkins, Z. 2000. “Gradations of Democracy?” American Journal of Political Science 44
(April): 293-300.
Finer, S.E. 1999. The History of Government (3 vols.). Oxford: Oxford University Press.
Finnis, J. 1980. Natural Law and Natural Rights. Oxford: Oxford University Press.
Foweraker, J. & T. Landman. 1997. Citizenship Rights and Social Movements. Oxford:
Oxford University Press.
Freedom House (ed.) annually. Freedom in the World, Lanham: University Press of
America.
Fuller, L. 1964: The Morality of Law. New Haven: Yale University Press.
Goertz, G. 2006. Social Science Concepts. Princeton: Princeton University Press.
Held, D. 1996. Models of Democracy. Stanford: Stanford University Press.
Holden, B. 1992. Understanding Liberal Democracy. New York: Harvester Wheatsheaf.
Holmes, S. 2003. “Lineages of the Rule of Law.” In J. M. Maravall and A. Przeworski
(eds.), Democracy and the Rule of Law. New York: Cambridge University Press.
Inglehart, R. 1997. Modernization and Postmodernization. Princeton: Princeton
University Press.
Inglehart, R. & C. Welzel. 2005. Modernization, Cultural Change, and Democracy. New
York: Cambridge University Press.
Karatnycky, A. & P. Ackerman. 2005. How Freedom is Won. Washington D.C.: Freedom
House.
38
Kaufmann, D., A. Kraay & M. Mastruzzi. 2005. “Governance Matters IV.” World Bank
Policy Research Department Working Paper 3630, Washington, D.C.: World
Bank.
Lipset, S.M. 1959. “Some Social Requisites of Democracy.” American Political Science
Review 53 (March): 69-105.
Macpherson, C.B. 1977. The Life and Times of Liberal Democracy. Oxford: Oxford
University Press.
Markoff, J. 1996. Waves of Democracy. Thousands Oaks: Pine Forge Press.
Merkel, W. 2004. “Embedded and Defective Democracies.” Democratization 11(5):3358.
Munck, G.L. & J. Verkuilen. 2002. “Conceptualizing and Measuring Democracy.”
Comparative Political Studies 35 (1): 5-34.
O’Donnell, G. 2003. “Human Development, Human Rights, and Democracy.” In G.
O’Donnell, J. V. Cullell, and O. M. Iazzetta (eds.), The Quality of Democracy:
Theory and Applications. Notre Dame: University of Notre dame Press, 9-92.
Ottaway, M. 2003. Democracy Challenged. Washington, D.C.: Carnegie Endowment for
International Peace.
Philpott, D. 1995. “In Defense of Self-Determination.” Ethics 105: 352-85.
Putnam, R.D. 1993. Making Democracy Work. Princeton: Princeton University Press.
Rawls, J. 1971. A Theory of Justice. Cambridge: Harvard University Press.
Rose, R. 2001. “A Divergent Europe.” Journal of Democracy 12 (1): 93-106.
Sartori, G. 1984. “Guidelines for Concept Analysis.” In G. Sartori (ed.), Social Science
Concepts. Beverly Hills: Sage, 15-85.
39
Saward, M. 2006. “Democracy and Citizenship: Expanding Domains.” J.S. Dryzek, B.
Honig & A. Phillips (eds.), The Oxford Handbook of Political Theory. Oxford:
Oxford University Press, 400-422.
Sen, A. 1999. Development as Freedom. New York: Anchor Books.
Shklar, J N. 1998. “Political Theory and Rule of Law.” In S. Hoffman (ed.), Political
Thought and Political Thinkers. Chicago: Chicago University Press, 21-37.
Stimson, S.C. 2006. “Constitutionalism and the Rule of Law.” In J.S. Dryzek, B. Honig
& A. Phillips (eds.), The Oxford Handbook of Political Theory. Oxford: Oxford
University Press, 317-332.
Vanhanen, T. (ed.). 1997. Prospects of Democracy. London: Routledge.
Vanhanen, T. 2003. Democratization. London: Routledge.
Warren, M.E. 2006. “Democracy and the State.” J.S. Dryzek, B. Honig & A. Phillips
(eds.), The Oxford Handbook of Political Theory. Oxford: Oxford University
Press, 382-399.
Welzel, C., R. Inglehart & H.D. Klingemann. 2003. “The Theory of Human
Development.” European Journal of Political Research 42 (2):341-380.
Welzel, C. & R. Inglehart. 2006. “Democracy and Emancipative Values.” Studies in
Comparative International Development 41(3):74-94.
Welzel, C. 2006. “Democratization as an Emancipative Process.” European Journal of
Political Research 45(6):871-896.
Welzel, C. 2007. “Are Levels of Democracy Influenced by Mass Attitudes?”
International Political Science Review 28(4):397-424.
40
Williams, A. 2006. “Liberty, Equality, and Property.” J.S. Dryzek, B. Honig & A.
Phillips (eds.), The Oxford Handbook of Political Theory. Oxford: Oxford
University Press, 488-506.
41
Figure 1. People Power as the Continuum Differentiating Democratic and Autocratic Regimes
More Autocratic
More Democratic
Hybrid Zone
More
Completely
Autocratic
0
5
NONE
10
15
More
Incompletely
Autocratic
20
25
30
35
40
More
Incompletely
Democratic
45
50
P E O P L E
55
60
65
70
P O W E R
More
Completely
Democratic
75
80
85
90
95
100
FULL
42
Figure 2. Effective Democracy as the Interaction of Democratic Rights and Rule of Law
POLITICAL
REGIME
CITIZENS
Political
Rights
+
STATE
Personal
Rights
DEMOCRATIC RIGHTS
EFFECTIVE
Procedural
Regularity
*
+
Tamed
Corruption
RULE OF LAW
DEMOCRACY
43
Figure 3. Democratic Rights and Rule of Law
1.00
Bounded Autocracy
0.95
Effective Democracy
Singapore
Canada, UK
Iceland
NL, Switzerld.
NZ
Scandinavia
0.90
0.85
Baltic states
Czech R.
Hungary
Slovenia
Poland
Tropical Island states
0.80
Extent of Rule of Law 2000-6
0.75
UAE
0.70
Qatar Oman
Kuwait
Saudi Arabia
Seychelles
Maldives
Morocco
0.50
0.45
0.40
0.35
0.30
0.25
Turkmenist.
0.20
0.15
Turkey
Thailand
Namibia
Croatia
Trinidad
Israel
Uruguay
Italy
S. Africa
Kiribati
Suriname
Micronesia
Bulgaria
India
Lesotho
Egypt
Chile Spain
Costa Rica
Taiwan
Greece
S. Korea
y = 0.0045x + 0.2143
R2 = 0.4649
Tunisia
0.55
Antigua
Botswana
Malaysia
Jordan
Brunei
0.60
Japan
Bahrain
Bhutan
0.65
Australia
Germ. (W.)
Germ. (E.)
U.S.A.
Belgium
France
Bahamas
Belize
Romania Panama
Senegal
Brazil
Mexico
Marshall
Syria
Central Asian Republics
Peru
Macedonia
Argentina
Bosnia
China Iran
Gabon
Sao Tome
Colombia
Swaziland
Nepal
Algeria
Cuba
Yugoslavia
Dominica
Uganda Armenia
Vietnam
Guatemala
Benin
Zambia
Libya
Bolivia
Pakistan
Russia
Rwanda
Ecuador
Djibouti
Kenya
Laos Belarus
Georgia
Kyrgyzst.
Papua
Comoros
Indonesia
Chad
Solomon
Burundi
Venezuela
Cameroon
Congo
Sudan
Paraguay
Tajikist.
Guinea-B. Sierra L.
Uzbekist.
Cote Divoire
Eq. Guinea
CAR
N. Korea
Angola
Nigeria
Zimbabwe
Liberia
Burma
Haiti
Iraq
Eritrea
0.05
Gambia
Burkina
Mali
Fiji
Afghanistan
Somalia
0.10
Mauritania
Unbound Autocracy
Ineffective Democracy
0.00
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Extent of Democratic Rights 2000-6
Note: For data sources and variable construction see the Internet-Appendix at
http://www.__________.
44
Figure 4. Distributional Changes from Democratic Rights to Effective Democratic Rights
100
y = 0.0104x2 - 0.3808x + 0.8434
R2 = 0.8842
95
90
Extent of EFFECTIVE Democratic Rights 2000-6
85
Effective Democracy
80
75
70
65
60
55
S. Africa
50
45
40
Marshall Isl.
Singapore
India
35
30
Kuwait
25
Bounded Autocracy
20
15
Ineffective Democracy
10
5
Unbound Autocracy
0
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Extent of Democratic Rights 2000-6
45
Figure 5. Differentiating Regime Types on the Effective Democracy Scale
46
Figure 6. Effective Democracy in a Human Empowerment Framework
ECONOMIC DEVELOPMENT
EMPOWERING ECONOMY:
DISTRIBUTIVE EQUALITY
EMANCIPATIVE VALUES
EMPOWERING CULTURE:
CIVIC ENGAGEMENT
EMPOWERING REGIME:
EFFECTIVE DEMOCRACY
HUMAN EMPOWERMENT
(an empowering regime in an empowering environment)
47
Figure 7. The Embedding of Democracy in an Empowering Environment
Civic Engagement 199599
N = 70
Emancipative Values
1995-99
N = 70
Economic Development
1998
N = 155
Distributional Equality
1998
N = 101
Empowering Cultural
Context 1995-99
N = 65
Empowering Economic
Context 1998
N = 99
N = 53
Entire Empowering
Context 1995-99
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Correlation with Democracy Indices in 2000 and thereafter (Pearson's r )
Legend:
In order from top to bottom, bars represent various democracy indices as follows:
Bar 1: Polity Democracy Score 2000-3
Bar 2: CIRI Empowerment Rights Scale 2000-4
Bar 3: Freedom House Democratic Rights 2000-6
Bar 4: Vanhanen Democracy Index 2001
Bar 5: Economist Democracy Index 2006
Bar 6: World Bank “Voice and Accountability Index” 2000-6
Bar 7: Effective Democratic Rights 2000-6
Note: For data sources and variable construction see the Internet-Appendix at
http://www.__________.
48
Figure 8. The Predictive Power of an Empowering Societal Context for Various
Measures of Democracy, Controlling for Democracy’s Endurance
Notes : Results from seven OLS regressions, each with one
of the democracy indices (measured in 2000 and thereafter)
as the dependent variable and the entire empowering context
(measured over 1995-99) as well as the democracy stock
(measured until 1995) as independent variables. N = 68.
Polity Scores 2000-4
CIRI Empowerm. Rights 2000-4
FH Democratic Rights 2000-6
Economist Democracy Index 2006
WB Voice & Accountability 2000-6
Vanhanen Democracy Index 2001
Effective Democracy 2000-6
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
% Variance Explained by Entire Empowering Context
AFTER Controlling Democracy Stock
49
Table 1. The Unidimensionality of Empowering Societal Conditions
Loadings on first and single principal component:
Empowering Societal Conditions
Indicators:
Analyses 1: Socioeconomic
Empowerment
Analysis 2: Sociocultural
Empowerment
Analysis 3: Overall Societal
Empowerment
Economic Development 1998
.947
.948
Distributive Equality 1998
.947
.926
Emancipative Values 1995-99
.910
.950
Civic Engagement 1995-99
.910
.863
KMO-Index
.500
.500
.822
Explained Variance
89.7%
82.7%
85.1%
Number of Nations
161
74
74
Note: For data sources and variable construction see the Internet-Appendix at http://www.__________.
50
Table 2. The Dependence of Democracy on Empowering Conditions in Society
Dependent Variables:
Extent of Democratic Rights 2000-6
Predictors (mid
1990s):
M1
Socioeconomic
empowerment
.77***
(11.39)
M2
M3
.14*
(2.59)**
Overall societal
empowerment
M5
.93***
(24.42)
.80***
(9.45)
Lagged dependent
variable 1996
M4
Extent of Effective Democratic Rights 2000-6
.83***
(15.29)
M6
M7
.30***
(4.66)
.21*
(2.05)
.93***
(18.57)
.72***
(6.90)
M8
.68***
(10.55)
.35***
(4.23)
.64***
(7.65)
Adj. R2
.57
.87
.63
.81
.86
.94
.87
.94
N
99
99
53
53
99
99
53
53
Notes: Entries are standardized regression coefficients with T-ratios in parentheses. Significance levels: * p.05
p.01 *** p.001
**
Note: For data sources and variable construction see the Internet-Appendix at http://www.__________.
51