Classifying Political Regimes: A Six-Fold Measure of Democracies and Dictatorships

Jos Antonio Cheibub, Yale University Jennifer Gandhi, Emory University

October 2004

ABSTRACT

The centrality of political regimes for the research agenda of political scientists is undeniable.

Considerable effort has been dedicated to empirically testing propositions about the conditions

under which political regimes emerge and survive, and their consequences for a broad set of

outcomes, notably their impact in promoting economic development and international peace.

Several measures of political regimes have appeared in the recent years, which differ significantly

in theoretical scope, interpretability, and repeatability. In this paper we present a six-fold

classification of political regimes which originates in a dichotomous distinction between

democracies and dictatorships. This classification covers 1999 countries between 1946 and 2002,

and is based on the institution that is capable of removing the government from office. We

compare our classification with existing alternatives, and show that the choice of measure is not

inconsequential. For this reason, decisions about which measure to employ in analysis should be

chosen in terms of whether they serve to address important research questions, can be interpreted

meaningfully, and are reproducible.

1

INTRODUCTION

A perusal of the past few years of the main political science publications demonstrates the

renewed centrality of political regimes to the disciplines research agenda. Considerable effort has

been dedicated to empirically testing propositions about the conditions under which political

regimes emerge and survive, and their consequences for a broad set of outcomes, but notably their

impact in promoting economic development and international peace. Part of this effort was

possible due to the proliferation of measures of political regimes covering a large number of

countries over a relatively long period of time.

In the wake of such development, a debate has emerged over the most appropriate way to

measure political regimes. Disagreement exists over what exactly should be measured and how it

should be measured: What is the notion of democracy that underlies existing measures? Should a

measure be continuous or categorical? If categorical, should it be dichotomous or polychotomous?

Should the input into the measures be exclusively observable events or should subjective judgment

be allowed in generating them? Important as these debates are, they have not been of much

consequence since most scholars seem to believe that, in the end, measures of democracy are

interchangeable. They correlate with each other and, in fact, tend to generate similar results when

used against one another in robustness checks of empirical findings (e.g., Fearon and Laitin

2003:85).

We disagree with this view. We believe that existing measures of political regimes are

significantly different in terms of both their theoretical grounding and operationalization and, for

this reason, should not be treated as interchangeable. In our view, we should take the differences

across measures more seriously and evaluate them in terms of whether they (1) serve to address

important research questions, (2) can be interpreted meaningfully, and (3) are reproducible.

2

The objective of this paper is to present a six-fold classification of political regimes which

distinguishes three types of democracy (parliamentary, mixed and presidential) and three types of

dictatorship (civilian, military and monarchic). Democracies and dictatorships, we believe, are

qualitatively different and hence are not easily classified along one single dimension. Although at

one level we use the same factor in distinguishing different types of democracies and dictatorships

the institution that is capable of removing the government the symmetry of this factor across

the two basic kinds of regime is limited. In democracies we focus exclusively on the legislature

and its prerogative to remove the government from office: exclusive in parliamentary regimes,

shared with the president in mixed regimes, and absent in presidential regimes. In dictatorships, in

turn, we consider the different types of institutions that can remove the government as the defining

trait of the regime: kin and family structures in monarchies, the armed forces in military regimes,

and the party in civilian regimes. Although the six-fold classification of regimes that we introduce

here represents only one among several possible ways to classify democracies and dictatorships, it

is valuable because it can be used to address empirically several of the issues currently on the

agenda of political scientists, is theoretically grounded and, most importantly, is reproducible.

Since the six-fold measure of political regimes we present here is conditioned on a

dichotomous classification of regimes as democracies and dictatorships, we start by summarizing

the rules that generate such a classification (section 1). We then proceed to address each of the

debates that have emerged around the classification of political regimes, arguing that the charges

commonly made against a dichotomous, minimalist measure of democracy are not valid. We argue

against the substantive view of democracy that underlies the alternatives to our dichotomous

measure (section 2), show that these alternatives are based on vague and arbitrary operational rules

(section 3), and that since their distributions are bimodal, they do not add any more information

3

than what is contained in the classification of regimes into democracies and dictatorships (section

4). For these reasons, existing measures of democracy are not interchangeable and the choice of

measure should be guided by its theoretical and empirical underpinnings. We then present the

reasons and the rules for classifying democracies as parliamentary, mixed and presidential (section

5), and dictatorships as monarchic, military and civilian (section 6). The appendix contains the list

of countries with their regimes and the dates they existed.

1. DEMOCRACIES AND DICTATORSHIPS

The six-fold regime classification we present in this paper is rooted on the dichotomous

classification of regimes as democracy and dictatorship introduced in Alvarez et al (1996) and

Przeworski et al (2000), extended to cover more countries (58 more) over a longer period of time

(from 1946 to 2002). Given that there has not been any change in the rules for classifying these

regimes, here we simply summarize these rules and compare the current with the previous

classification. The changes that did occur were entirely due to the fact that new information about

specific cases was made available.

Democracies are regimes in which governmental offices are filled as a consequence of

contested elections. For a regime to be democratic, both the chief executive office and the

legislative body must be filled by elections.1 Contestation occurs when there exists an opposition

that has some chance of winning office as a consequence of elections. This implies that elections

are ex ante uncertain and repeatable, and that outcomes are ex post irreversible. Operationally, a

regime was classified as a democracy if it met the requirements stipulated in all of the following

1 That not all offices need to be filled by elections is uncontroversial. Collier and Adcock (1999:549), in turn,

believe that having only one of those offices filled by elections should be sufficient to qualify a regime as at least

partially democratic. Yet, is a system in which the president is elected in contested elections but the laws are made

by a legislative body composed of her appointees at all democratic?

4

four rules: (1) the chief executive must have been elected; (2) the legislature must have been

elected; (3) there has been at least two parties or lists competing in the elections; and (4) an

alternation in power under identical electoral rules must have taken place.

The implementation of the first two rules is straightforward since it is simple to observe

whether the relevant offices are filled as a result of elections. The third rule, although slightly more

complex, is also straightforward: for a contested election to take place, voters must have at least

two alternatives to choose from. Hence, elections in which a single party competes, or in which

voters are presented with a single list, do not qualify as contested elections.2 The implementation

of the last rule is more complicated since it requires that we make one assumption and one decision

about what kind of error we are willing to accept. It does not, however, require any subjective

judgment on the part of the analyst and hence does not compromise the classifications

reproducibility.

An alternation in power takes place when the individual occupying the chief executive

office is replaced through elections that were organized under the same rules as the ones that

brought him or her to office. For obvious reasons, the alternation only becomes relevant in the

cases where the first three rules apply. The implementation of this rule, however, is complicated by

the fact that, given the occurrence of elections in which two or more parties compete, it is difficult

to distinguish cases where incumbents never lose power because they are popular, from those

where incumbents only hold elections because they know they will not lose them. Since there is

nothing in any conception of democracy that precludes the emergence of a highly popular

2 Not all elections are contested; of the 1457 legislative elections and 489 presidential elections that occurred in the

1946-96 period, 728 and 268, respectively, were under regimes that are here classified as dictatorships. Data on

elections from Golder (2004).

5

incumbent who is time and again returned to office by very pleased voters, the first case should be

coded as a democracy. And since incumbents who are ready to call off elections at the moment

they anticipate a defeat violate the ex ante uncertainty and repeatability conditions for contested

elections, the second case should be coded as a dictatorship. Yet, these two cases are

observationally equivalent.

Part of the problem can be addressed if we assume that current actions are revealing of

what incumbents would have done at different points in time. Consider, for examples, the cases of

Malaysia and Japan. Between independence in 1957 and 1969 there were three multiparty

elections in Malaysia. The incumbent party won an absolute majority in the first two, but not in the

third one. As a result, the government declared a state of emergency, closed parliament, issued a

harsh internal-security law and rewrote the constitution in such a way that it never lost an election

after parliament was reopened and elections resumed in 1971. We code Malaysia as a dictatorship

under the assumption that the incumbents actions in 1969-71 demonstrated their predisposition of

holding elections only to the extent in which they were assured of winning. In Japan, the Liberal

Democratic Party was in office continually until 1993, when, after an electoral defeat, it yielded

power to the opposition. We code Japan as a democracy under the assumption that the LDP would

have yielded power had it lost elections prior to 1993.

Yet, even if assumptions are made, in some cases we cannot know what type of incumbent

is in office. The best example is, of course, Botswana, where seven multi-party elections were held

since independence in 1966 under conditions that most analysts consider to be free and fair (no

constraints on the opposition, little visible repression, no apparent fraud), and the incumbent has

won each of them by a very high margin of victory. Had the Botswana Democratic Party lost one

of these elections and allowed a different party to form a government, or had it closed parliament

6

and changed the electoral rules, we would be able to identify the regime either as a democracy or a

dictatorship. As it is, we simply do not know and, until one of these events happens, we need to

accept that we are not capable of coding Botswana with the rules we now have. We can exclude all

cases such as Botswana from the data set, we can call them democracies, or we can call them

dictatorships. Whatever we do, there will be some systematic error due to the fact that we cannot

tell the cases apart. But we can, at least, control for this error.3

One of the consequences of this rule is that the uncertainty inherent in cases such as

Botswana may be resolved as history unfolds. In the original classification (Przeworski et al.

2000), the four rules unambiguously classified 92% of the country-years between 1950 and 1990.

In the current extension, the proportion is exactly the same: there are 8% of country-years between

1946 and 2002 that we classify as dictatorships on the grounds that they fail the alternation rule

only. Some of the cases, however, have changed, either because new countries have emerged and

events were not sufficient to allow us to determine their regime type (e.g., Georgia, Kyrgyzstan,

Tajikistan) or because history provided the information we needed and we had to revise our

original coding (e.g., Mexico, Senegal).

The issue that arises with countries such as Mexico under Vicente Fox and Senegal under

Abdoulay Wade, where the opposition won after a long period of incumbent victory in multiparty

elections, is to determine when exactly the transition occurred. That the new government should be

classified as a democracy according to our rules is not problematic. But should the government

that allowed the alternation to take place be also classified as a democracy? If yes, what about the

3 We code cases such as Botswana as a dictatorship, but identify them through a variable called Type II, so that

users of our classification may either recode them as democracies or remove them from the analysis. Note that these

are not intermediate cases; we simply cannot tell whether they are a democracy or a dictatorship given what we

can know or what can be known about the country.

7

government prior to that one, the previous one, and so on? Specifically, does the fact that Fox took

office in Mexico in 2000 require that we recode the regime as a democracy all the way back to the

1920s, when the PRI first came to office?

We address this issue by focusing on the rules under which the incumbent was elected. If

the opposition wins under rules that are identical to the ones that led to the victory of the

incumbent, then we consider the incumbent democratic: the years under that persons rule meet all

four rules for classifying a regime as a democracy. This is done with all previous governments up

to the point where the electoral rules were changed. The rules that matter are the broad electoral

rules who votes, how votes are counted, and who counts the votes. Thus, in the case of Mexico,

we date the transition to democracy to 2000, when Fox, the candidate of one of the opposition

parties, was sworn into the presidency. The electoral rules were changed under the Zedillo

presidency (1994-2000) when, in 1996, an accord between the ruling PRI and the two opposition

parties (PAN and PRD) ended the PRIs control of the Federal Electoral Institute. Similarly, we

date transition to democracy in Senegal to 2000, when the incumbent Abdou Diouf, of the Socialist

Party, lost to Abdoulaye Wade, of the Democratic Party. Dioufs last victory had been in the 1993

election, prior to the creation of the independent National Observatory of Elections in 1997.

With these rules, thus, we classify democracies and dictatorships in 199 countries between 1946

and 2002, for a total of 7,880 country-years: 4,607 (58.5%) dictatorships and 3,373 (41.5%)

democracies.

In addition to this classification (henceforth referred to as DD) there are two other

8

measures of political regimes that are widely used:4 the Freedom House (FH) measure of political

and civil liberties, which covers all countries of the world between 1973 and 2002 and the Polity

IV (POLITY) measure of political regime characteristics, which covers all countries between 1800

and 2000. In FH each country is given an independent score from the highest (1) to the lowest (7)

level of political and civil liberty. In POLITY, there are two conceptually independent scales of

autocracy and democracy, each ranging from 0 to 10, which frequently are then subtracted to

generate a political regime measure that ranges from -10 (least democratic) to 10 (most

democratic).

Although these measures are similar in the sense that they cover a large number of

countries for a relatively large number of years, they are different in at least three important ways:

(1) the conception of democracy that underlies each of them; (2) the nature of the data used to

assess political regimes; and (3) the type of measurement they perform. We turn now to a

discussion of each of these aspects as a way to assess the value of these three measures.

2. CONCEPTION OF DEMOCRACY

Measures of political regimes differ as to whether they adopt a strictly procedural,

minimalist view of democracy as opposed to a more substantive one. In the first case, democracy

depends exclusively on the presence of certain institutions, with no reference to the kinds of

outcomes that are generated by their operation. Thus, underlying DD is the notion that democracy

is a regime in which those who govern are selected through contested elections and, once

identified, the occurrence of contested elections is necessary and sufficient to characterize a regime

4 Several other measures have been proposed but they are not widely used, mostly because they are available for

only a few years (Bollen 1980 and Coppedge & Reinicke 1991), are not sufficiently defined (Arat 1991 and

Gasiorowski 1996), or include inappropriate indicators (Vanhanen 2000).

9

as democratic.

In substantive conceptions of democracy, institutions are seen as necessary but not

sufficient to characterize a political regime. Although it may be that no democracy exists that does

not have contested elections, not all regimes that are based on contested elections may be called

democratic. What matters is that, through these elections, something else happens: the public good

is achieved, citizen preferences are represented, governments become accountable, citizen

participation in political life is maximized, economic equality is enhanced, rationality is

implemented, economic conditions improve, and so on. Those who use FH, therefore, believe that

the measure of freedom it offers can be used to indicate democracy. Polity IV is variously

used to indicate democratic regimes conceived in terms of accountability, fairness, freedom or

some other attribute.

Thus, one point of debate with respect to measures of political regimes is whether a

minimalist conception of democracy, such as the one underlying DD, is sufficient to characterize

political regimes. There are many researchers who believe it is not. For example, for Diamond

(1999) such a view of democracy merely identifies electoral democracy, which is distinct and

inferior to liberal democracy. For Mainwaring, Brinks and Prez-Lin (2001), it is subminimal

since it leaves out participation, guarantees of civil and political liberties, the governments ability

to really govern, and civilian control of the military. For ODonnell (1995), it fails to distinguish

between delegative and liberal democracies. For Levitsky and Way (2002), it fails to identify a

distinct regime type, competitive authoritarianism. For Coppedge (2002:36), it fails to handle

emerging questions about the quality of democracy. Finally, for Weeden (2004:4) a measure of

political regime based on a minimalist conception of democracy is limited since it tends to

obscure concerns of central importance to substantive representation, such as how democratic

10

rulers should act once elected, and what their duties and obligations as rulers are, or should be.

A measure of democracy based on a minimalist conception, however, is compatible with

most of the theoretical issues that animate empirical research on political regimes. For instance,

democracies are considered to undermine economic development because governments may

become hostages of voters short term interests (DeSchwinitz 1964, ODonnell 1973); they are less

likely to go to war because of the domestic costs (paid at election time) of such a decision (Fearon

1994, Schultz 1999); and, they manage to avoid catastrophes such as famines because governments

cannot avoid challenges they will face when elections are held (Sen 2000). In addition,

macroeconomic performance is thought to suffer because of governments attempts to manipulate

the economy for electoral purposes (Nordhaus 1975, Tufte 1978); or, alternatively, long term

economic performance will be better because they must satisfy the voters who will be able to

remove them from office in periodic elections (Paldan 1991, Powell and Whitten 1993, Wilkin et

al. 1997). Market-oriented reforms, in turn, will not be attempted or implemented consistently

because governments fear voters reaction to them (Przeworski 1991, Haggard and Kaufman 1995)

or, on the contrary, they will be attempted and implemented consistently because governments will

be rewarded in future elections (Hellman 1998). In addition, income inequality will be reduced as

voters will choose governments in elections that will correct through the political system the

allocation generated by the economic market (Meltzer and Richard 1981, Przeworski 1990). In all

of these areas of research, and in many more, the mechanism that links political regimes to

outcomes is the occurrence of contested elections.

Substantive conceptions of democracy, on the other hand, generate measures that are not

amenable to the empirical investigation of at least some of these issues. If democracy is defined as

the regime where rulers are accountable to the ruled, then the issue of whether governments under

11

democracy do indeed act in voters best interests becomes redundant. If besides political equality

democracy also requires economic equality, the finding that income distribution is more egalitarian

under democracy only corroborates what is true by definition. Moreover, including more

dimensions along which to classify political regimes makes it harder to specify the causal

mechanisms that link regime and the outcomes of interest. When a study employing FH reports a

positive effect of democracy on economic growth we may reasonably wonder what is it among the

over 20 dimensions that enter into it that is driving the observed relationship. Similarly, how

should we substantively interpret the hundreds of possible response patterns that the components

of the POLITY measure generate?

Expanding the notion of democracy to include more than contested elections may also blur

the boundaries between political regimes and other political entities, and lead to the inclusion of

attributes that are, as Munck and Verkuilen (2002:54) put it, aspects of the state as opposed to a

regime type. Civilian control of the military, national autonomy with respect to the international

system, bureaucratic responsiveness to executive and legislative authorities, are attributes that vary

across political systems, irrespective of the rules they follow to choose who makes decisions for

the country.

Finally, substantive notions of democracy also tend, albeit marginally, to undermine the

legitimacy of regimes where governments, although chosen on the basis of contested elections, are

not representative of new cultural, social and economic relations. We may hear echoes of the old

lefts disdain for the mere formalism of bourgeois democracy when we hear that some

democracies, invariably the new ones in developing countries, are not worthy of the label since

all that they do is to choose governments by contested elections. The only difference is that the

capitalist democracies that the old left found wanting with respect to some ideal form of

12

democracy the absolute equality that would be brought about by communism are now taken to

be the model existing regimes in developing countries should strive to achieve. On these grounds,

even India, in spite of over 50 years of periodic contested elections with alternation in power

between political parties representing very distinctive social forces, is a case of a democracy that

needs to be qualified (Zakaria 1997, Diamond 2002).

A minimalist definition of democracy is compatible with a variety of specific ways in

which social and political life is organized. It does not attach any weight to the way governments

are formed, political parties compete, candidates selected, voters vote, and votes counted; to the

way justice is organized and dispensed, to how much or in what ways the state intervenes in the

economy or to whether private property is upheld. It recognizes that all governments are

constrained in their actions, be it by those who hold guns or by those who own capital, domestic or

international. All that a minimalist definition of democracy requires is that citizens be given

periodically the opportunity to choose their leaders in electoral contests; that they be presented

with more than one alternative; and that those who win become, indeed, the countrys leaders.

Disagreement about conceptions of democracy probably cannot and certainly does not need

to be resolved at the philosophical level. Measures are nothing but ways to make specific concepts

operational and hence should not be evaluated in terms of the intrinsic values of the concept they

are meant to operationalize. What matters is that the concept be clearly formulated, the measure

clearly related to it, and that its use does not preclude the study of the theoretical questions we care

about. On these points, DD is very clear.

3. WHAT KIND OF DATA?

At the limit, all measures are subjective in the sense that people have different ways to look

at the world and may choose to use different rules for generating a measure. The issue is whether

13

existing measures are reproducible, that is, whether knowledge of the rules and the relevant facts is

sufficient to unambiguously lead different people to produce identical readings on specific cases.

We take as axiomatic that this is something of value in any measure.5 Yet, neither FH nor POLITY

is reproducible in this sense.

Part of the reason they are not has to do with the fact that the rules on which they are based

are insufficiently defined. FH requires answer to the following questions, with no clear attempt to

define the relevant terms or qualifiers: Are there fair electoral laws, equal campaigning

opportunities, fair polling, and honest tabulation of ballots? Are voters able to endow their

representatives with real power? Do minorities have reasonable self-determination, self-

government, autonomy, or participation through informal consensus in the decision-making

process? Are the people free from domination by the military, foreign powers, totalitarian parties,

religious hierarchies, economic oligarchies, or any other powerful group? Are there free and

independent media? Are there free trade unions and other professional organizations, and is there

effective collective bargaining? Is there personal autonomy? Is there equality of opportunity?

Similarly, the Polity IV democracy scale requires one to decide whether constraints on the chief

executive in any given country are close to parity, whether the executive faces substantial

limitations, or whether the constraints are located in one of two possible intermediate categories.

5 To our knowledge, Mainwaring, Brinks and Prez-Lin (2001) are the only ones to explicitly defend the use of

subjective judgments in classifying political regimes. They state that their classification of Latin American countries

is predicated on the belief that this enterprise demands some subjective judgments about the nature of political

regimes. By subjective we do not mean arbitrary, but rather informed judgment based on knowledge of the cases

and guided by explicit coding rules (p.37). Later on they say that regimes should be classified according to

observables, but social scientists must make judgments about whether an infringement is sufficiently serious as to

regard a regime as less than democratic (p.42).

14

Another reason these measures are not reproducible has to do with the difficulty in obtaining the

information they require. Whereas DD requires one to know whether an election occurred, how

many parties competed, who won and who took office, FH requires knowledge of that plus of how

the electoral campaign was conducted, the content of party platforms and campaign financing laws

and practices, the operation of the justice system, the structure of labor representation, collective

bargaining, and more. POLITY requires less, since it focuses on the processes leading to the

selection of the executive and the participation of politically active members of the political

system. Yet, a good deal must be known about how these processes take place in order for one to

be able to adjudicate a specific case among the several values of each of the component

dimensions. Having been involved with trying to discover merely the basic facts about elections in

the different countries of the world since 1946, we can attest that the data required by FH and

POLITY are hard, if not impossible, to obtain; and that, as a consequence, much of the coding

must be made on the basis of inferences, extensions, guesses and perhaps even prejudice.

These are, of course, not irremediable problems. Those responsible for generating the

measures may explicitly attempt to define the terms and specify the cases to make them less vague

and subjective.6 Moreover, the amount of information available to researchers has increased

exponentially and there is no reason for us not to believe that, someday, we will know everything

that needs to be known about each of the cases we want to code. Yet, even when rules are clearly

6 POLITY has become increasingly concerned with this and, in an attempt to make the coding process less

subjective, has provided lengthy descriptions of specific cases in order to illustrate what goes into each of them. It is

interesting, however, that these are ostensible definitions of each of the cases, rather than ex ante rules for coding

regimes. POLITY also releases to the public the component variables of the final measure thus allowing researchers

to aggregate them in different ways and to recode cases one may find dubious. FH releases neither the rules it uses

to make specific decisions nor the components that go into its measure.

15

defined and facts are known, the possibility of systematic error is not eliminated remember that,

as discussed in section 1, about 8% of the country-years in DD cannot be unambiguously coded as

democracy or dictatorships because history has not yet provided the information we need to make

a decision. Thus, the lack of clear rules and the unavailability of necessary information just

compound the amount of error that may be going into each of these measures. And since the

measures are irreproducible across different coders, we cannot even assess the amount of

systematic error that goes into each of them.

4. TYPE OF MEASUREMENT

Whereas DD classifies political regimes as democracy or dictatorship, FH and POLITY

offer polychotomous classifications. Among the debates that have engaged those who study

political regimes empirically, this is probably the one that has generated the highest level of

controversy, with the majority of views coming down against the use of a dichotomous regime

classification.

The main argument against the dichotomous classification of political regimes is that it is

limited in terms of the information it conveys. Collier and Adcock (1999:538), for instance, claim

that the use of a dichotomous regime classification is puzzling since it places regime differences

at what is traditionally viewed as the lowest level of measurement. Bollen, having postulated

the inherently continuous nature of the concept of political democracy (1989:612), claims that

we unnecessarily compromise the concept by considering it a dichotomous phenomenon

(1991:9). This, he proceeds, leads to a crude lumping of countries into the same category when in

reality they have very different degrees of political democracy (1991:9).

This criticism is, in part, based on a misunderstanding since it assumes that the

dichotomous nature of DD is simply the product of a decision to measure at that level a

16

phenomenon that is otherwise continuous. But the use of a dichotomous classification of political

regimes is not driven by the desire to simplify the measurement process. It does not imply

imposing a cut-off point of any sort over an underlying, latent distribution of political regimes.7

Rather, it is based on the notion that political regimes can be directly observed and that we can

distinguish two main types, depending on whether or not the government is chosen through

contested elections. Thus, the matter is not whether one should adopt a continuous or a categorical

measure of democracy, observable across all political regimes. The issue is whether there is a

natural zero-point that divides democracies and non-democracies. Even those who, like the present

authors, develop and use categorical measures of democracy will agree that, given some

appropriate yardstick, democratic regimes may differ as to how democratic they are, and that some

measure to assess their degree of democratic-ness may make sense. Note, however, that this refers

to democratic regimes, as opposed to non-democratic regimes. It assumes that some regimes fail

whatever requirement there is for them to be called democratic.

The belief that democracy is an attribute that can and should be measured over the

spectrum of cases leads to assertions that cannot be easily supported, such as that there are positive

levels of democracy in places like Albania under Hoxa, North Korea and Chile under Pinochet.

But even if it is granted that it does not make sense to speak of democracy in these cases, the

argument could still be made that a measure with more categories such as FH and POLITY is to be

preferred since it will convey more information than a measure with only two categories. This,

however, is not necessarily true. The informational content of different measures depends on the

7 Elkins (2000), for example, assumes that the true distribution of political regimes is continuous. This being the

case, it is only natural that he discovers that more categories will be better than fewer categories at capturing the

true value of democracy.

17

way they are conceptualized and observed, at least as much as it depends on the level of

measurement. Given the degree of subjectivity and vagueness that goes into FH and POLITY, it is

not at all clear what information they are actually conveying. What exactly is being conveyed

when we say that, according to POLITY, the level of democracy in 1965 Singapore was -2, or that

it was -6 in Burma under the military junta?

Moreover, there is a fair degree of arbitrariness in the construction of these measures. In

FH, for instance, coders assign raw points ranging from 0 to 4 for each of the 8 questions in the

political liberties and 14 questions in the civil liberties checklists, for a maximum of 32 and 56

points respectively. Countries are then distributed into one of the seven categories that make up the

final political and civil liberties scales according to the number of raw points they received. For

example, a country with 28-32 raw political rights points is placed in category 1 of the political

rights scale; with 23-27 points, it is placed in category 2; and so on. No reason is ever given for

attributing a maximum of 4 points in each category, rather than 5, 6 or a million; or for why the

lower bound of category 1 is 28 and not 27, 25 or 10 raw points. The aggregation method of each

measures components is also fairly arbitrary. It gives equal weight to each of the components,

with no justification and no consideration of the possibility that the different dimensions may have

different impact on the nature of the regime.8

8 Gleditsch and Ward (1997) have shown that most of the variation in the overall index is accounted for by changes

in one dimension Chief Executive Constraint. Treier and Jackman (2003) adopt a more adequate aggregation

model on the Polity components and show that there is considerable error in the latent levels of democracy

underlying the Polity scores, (p. 24) and that, because this error is heteroskedastic, ignoring it is likely to lead to

inferential error. Gandhi and Vreeland (2004), in turn, report that the inverted-U relationship between democracy

and risk of civil war, identified by Hegre et al. (2001) on the basis of Politys overall score, only holds for one of its

four components.

18

This degree of arbitrariness is extended to the practice of creating regime types on the

basis of the FH and POLITY measures. Many of the questions that political scientists care about

have to do with being in or out of a given state such as the political regime, and not with the

incremental changes over a gradation. The entire transitions literature is predicated on the notion

that one can identify the point at which a political regime stopped being a dictatorship and became

a democracy. We ask questions about the supply of public goods under different types of

democratic regimes, about the propensity of different types of regimes to go to war, about the

choice of regime political elites make under different conditions, and so on. To study these

questions empirically we need measures that identify these states. Since scale measures or the

categories of the existing multinomial measures do not represent any of the states that are

theoretically identified, researchers are required to collapse regimes into democratic and non-

democratic so that they can then study what brings these states about and the consequences of

being in them. Thus, in their study of the democratic peace, for example, Rousseau et al. (1996)

consider that a democracy is any regime that received a score of 17 or higher in their transformed

POLITY scale.9 Epstein et al. (2003), who argue that partial democracies emerge from their

analysis as the key to understanding democratic transitions (p.21), define them as regimes that

score between 1 and 7 in the POLITY measure (whereas full autocracies are those that score

between -10 and 0 and full democracies those that score between 8 and 10). Starr and Lindborg

(2003) categorize political regimes into democratic, non-democratic, and transitional

according to the Freedom Houses classification of countries as free, non-free, and partially

free. These, in turn, are the countries whose average political and civil liberty scores were,

9 The Polity scale is normally built so that it will range from -10 to 10; Rousseau et al. (1996) transform it so that its

range is 0 to 20.

19

respectively, between 1 and 2.5, 3 and 5.5, and 5.5 and 7. This practice can hardly be innocuous.

Deciding when a case belongs to one type or another, when it is democratic or not, is entirely

arbitrary, justified not by theory, but by the empirical salience, or perhaps even the convenience, of

specific cut-off points.10

Finally, although FH and POLITY are polychotomous, their distributions are actually

bimodal, with a high concentration of cases in their low and high ends: 56% of the cases are

classified in the three lowest and highest categories of FHs 13-point scale; 73% of the cases have

scores that are 7 and lower or 7 and higher in the 21-point POLITY scale. Thus, in spite of a

larger number of categories, these measures add little if any additional information to a

dichotomous classification of political regimes. It is not surprising, therefore, that the correlation

among the three measures is high: the correlation between POLITY and FH is 0.91; POLITY

predicts correctly 94% of the cases classified as democracies by DD and 95% of those classified as

dictatorships; FH predicts 88% and 95%, respectively. However, once observations at the extremes

of the distribution are excluded, the correlations among existing measures become significantly

weaker: it is 0.75 between POLITY and FH; POLITY predicts 70% of the democracies in DD

and 83% of dictatorships, whereas FH predicts 67% of democracies and 81% of dictatorships.

This fact is not without consequences for empirical research. For example, the inversed-U shaped

relationship between political regimes and civil war identified by Hegre et al. (2001) disappears

when we exclude the cases scoring less than -8 and more than 8 in the Polity measure. The

negative impact of Islam on democracy identified by Fish (2002) survives weakened when the

extremes of the FH measure are trimmed, but not at all when the extremes of the POLITY measure

10 Zinn (2004) reports that studies of the onset of civil war are not robust to the adoption of different cut-off points

of the POLITY measure.

20

are removed from the analysis.

Thus it is the uncontroversial cases that drive the high correlation among different

measures of political regimes and, perhaps, are driving the empirical patterns identified in studies

of political regimes. No measure will produce very different readings for the regimes in, say,

England, Sweden, North Korea or Iraq. The difficulty will appear with cases such as Mexico,

Botswana, Malaysia, Peru, Guatemala, and scores of other countries that populate the middle of

the distribution of existing measures and for which the rules, as stated, do not apply clearly. What

is it that these countries represent in terms of political regimes? What does it mean to be located in

the middle range of the POLITY or FH distributions? Given that once we get to these cases no

consensus seems to exist across measures, the choice of measure must be made on conceptual

grounds.

We thus firmly believe that the use of a dichotomous measure to gauge political regimes

does not necessarily imply over-simplification, a lack of information or any of the pitfalls that are

commonly attributed to it. To the contrary, in our view, such a measure is valuable since it is

theoretically motivated, can be meaningfully interpreted and is reproducible. It is, therefore, an

adequate base on which to generate further differentiation among political regimes, to which we

now turn.

5. PARLIAMENTARY, MIXED AND PRESIDENTIAL DEMOCRACIES

Democracies may be distinguished in several ways: their form of government, type of

electoral system, structure of interest representation, styles of policy-making, nature of party

system, judicial organization, and so on. Each of these distinctions raises important issues about

the way democracies operate and the consequences they have; and each of them has been

addressed, in one way or another, in the political science literature.

21

There is no hierarchical order among the several possible ways of differentiating democracies. In

this section we focus on the rules that specify who may dismiss the government and thus classify

democratic regimes into parliamentary, mixed and presidential. A similar focus will orient the

classification of dictatorships in section 6.

Classifications of forms of democratic government abound in the literature. There seems to

be a general consensus that there are two pure types of systems parliamentary and presidential

and one system that combines features of both commonly called mixed, semi-presidential, or

parliamentary-presidential systems. Our classification is not different in that it also groups

democracies into these three categories. Existing classifications, however, use redundant and/or

insufficient criteria and thus are unable to unambiguously place all cases into one of the three

categories. What is distinctive about the classification we offer here is that it provides a clear set of

operational criteria to classify democratic regimes according to their form of government.

Conceptually, the form of government in a democracy depends on the relationship between

the government, the assembly and, where they exist, elected presidents. The main issue is whether

the government can be removed by the assembly in the course of its constitutional term in office.

Systems in which governments cannot be removed by the assembly are presidential. Systems in

which they can are either parliamentary (when only the assembly is allowed to remove the

government) or mixed (when both the assembly and the elected president can remove the

government). The mechanism of removal by the legislature is the vote of no-confidence initiated

by the legislature, or the vote of confidence, initiated by the government itself (Huber 1996). The

mechanism of removal by the elected president may be direct, such as when she/he can unilaterally

replace the government, partially or completely, or indirect, such as when the president dissolves

the assembly, calls early elections, and thus causes the government to fall.

22

There are several other important aspects related to the nature and/or operation of the

government in democracies, some of which have been made into defining features of democratic

forms of government. They include the nature of the executive power, thought to be collective or

collegial in parliamentarism and individual in presidentialism (Verney 1992, Lijphart 1999:118);

the separation of heads of state and government under parliamentarism and their fusion under

presidentialism (Verney 1992); the indirect election of government in parliamentarism and popular

election in presidentialism (Lijphart 1999:117); the existence of a president with constitutionally

granted lawmaking authority in presidential regimes or with considerable powers in mixed

regimes (Shugart & Carey 1992:19, 24).

These features of democracies, however, are not sufficient to distinguish forms of

governments. Uruguay is a presidential democracy that had at some point (1952-1967), a collective

executive; Israel popularly elects its prime minister and yet cannot be considered a presidential

democracy; in Bolivia the president is, under some circumstances, elected by the assembly and yet

it is not parliamentary; Venezuela prior to the 1999 constitution had a president with no

constitutionally mandated powers and yet was fully recognized as a presidential democracy.

Some might suggest that these constitute anomalous, intermediate cases, hybrid regimes that fall

into neither the parliamentary nor the presidential categories (Mainwaring 1993, Lijphart 1999).

Yet, unless we are able to provide a positive criterion for identifying these regimes, this does not

seem to be a satisfactory solution as it simply creates a residual category that lumps together very

heterogeneous cases. Thus we need to produce a set of criteria that unambiguously classify

democratic regimes according to their form of government.

Systems in which governments, in order to exist, must enjoy the support of a legislative

majority are classified as parliamentary; systems in which governments do not need the support of

23

the legislative majority in order to exist are classified as presidential; and systems in which, in

order to exist, governments depend both on majorities in the legislative assembly and on elected

presidents are classified as mixed. Operationally, the following three questions provide a sequence

of steps summarized in Figure 1 that unambiguously identify presidential, parliamentary and

mixed democracies:

*** Figure 1 here ***

1. Is there an independently (either directly or indirectly) elected president? This is a necessary,

but not a sufficient condition for a presidential regime. In some countries, such as Italy or

Germany, presidents are independently elected, and yet no one claims they are presidential. What

this question allows us to do is to identify countries that, due to the absence of an independently

elected president, cannot be presidential or mixed. They must be, therefore, parliamentary.

2. Is the government responsible to the assembly? Assembly responsibility means that a legislative

majority has the constitutional power to remove the government from office. Given the existence

of an independently elected head of government, the issue becomes whether the assembly plays

any role in the existence of the government. Since assembly responsibility is a necessary condition

for the existence of either a parliamentary or a mixed system, cases where the president is

independently elected and there is no assembly responsibility, cannot be either. They are,

therefore, presidential.11

11 Formally, assemblies may affect both the formation and the survival of governments, and whether it does one or

the other, or both, has been made one of the dimensions along which democratic regimes are classified (Shugart &

Carey 1992, Mainwaring 1993). Yet, the crucial aspect for assembly responsibility is survival and not formation of

the government. Theoretically the former subsumes the latter: an assembly that is deprived of the right to elect the

government but which can pass a vote of no confidence can do so immediately following the constitution of the

government, thus effectively preventing it from coming into existence. Conversely, an assembly that is allowed to

24

3. Is the government responsible to the president? Government responsibility to the president can

be direct, as when the president can dismiss the government either in its entirety or a minister at a

time (such as in Portugal under the 1976 constitution). It can also be indirect, as when the president

dismisses the government by dissolving the assembly (such as in Portugal under the 1982

constitution). This distinction may or may not be relevant for a number of factors,12 but it is not

sufficient to make them different types of regime. In both cases the government depends on the

support of a legislative majority and an independently elected president in order to stay in office.

Thus, given assembly responsibility and an independently elected president, either case is

sufficient to characterize the regime as mixed. Cases in which the president cannot dismiss the

government and dissolve the assembly are classified as parliamentary democracies.

Our classification is entirely based on the rules prescribed in the countrys constitution.

This decision is justified, in part, by the fact that we are dealing with a set of countries that have

been classified as democratic on other grounds. For this reason, it makes sense to take the

constitution as the document that effectively stipulates the way in which governments are formed

and survive in power. In the vast majority of cases this leads to clear and uncontroversial decisions

elect the government may, like in Switzerland and Bolivia, be barred from removing it from office. As Strom

(2000:265) has pointed out, in the real world parliamentarism rarely means that the legislature actually elects the

executive. What matters, he continues, is that the cabinet must be tolerated by the parliamentary majority, not that

the latter actually plays any direct role in the selection of the former (2000:265). Note, also, that the nature of the

executive collective or not is immaterial for the classification of forms of democratic regimes. Thus,

Switzerland, where legislatures elect a collective government which cannot be removed before the end of its term, is

classified as a presidential regime: the assembly does not affect the survival of the government.

12 We believe this is the basis for Shugart & Careys (1992) definition of premier-presidential and president-

parliamentary democracies although it is unclear in their detailed discussion of these systems (pages 55-75)

whether the distinction really matters.

25

since the rules of government formation are well defined and political practice conforms to the

constitutional provisions. In a few cases, however, there will be ambiguity, mostly because some

of the scenarios prescribed by the constitution have never materialized, but also because of

misconceptions induced by the language adopted in the constitution (or in the translation the

authors had to rely on in order to do this work).

The best example of the latter issue comes from South Africa, where the head of state and

government are one and the same person, who is named the President. However, according to the

1996 constitution (as well as the interim 1994 constitution), this president is subject to a vote of

non confidence by a majority of the National Assembly, which, if approved, requires his/her

resignation and the formation of a new government. The fact that votes of non-confidence have

been far from likely in South Africa has nothing to do with the form of government and, we

believe, everything to do with the fact that parliament has been dominated by a party holding about

two-thirds of the seats since competitive elections were held in 1994. Had such a large majority not

existed, the relation between the government and the parliament in South Africa would have been

considerably different, with issues of government survival due to legislative action at the forefront.

Regarding the former issue constitutional scenarios that do not materialize the major

uncertainty emerges with respect to mixed democracies where the room for ambiguity is the

largest, and the feeling that a mixed regime is a pure form of parliamentarism or presidentialism in

disguise is the strongest. In Iceland, for example, the directly elected president is commonly

perceived as a figurehead and symbol of unity rather than a political leader (Kristinsson

1999:87). Hence, as Kristinsson puts it (1999:86), it is customary in Iceland to regard the form of

government as a parliamentary one, essentially similar to the Danish one, despite the different

ways heads of states come into office. Yet, the Icelandic constitution is ambiguous with regard to

26

the powers of the president. At the same time that the president may dissolve parliament (article

24) and appoint and discharge ministers, including the prime minister (article 15), the constitution

also states that ministers execute the power of the president (article 13), thus providing the grounds

for a view of a passive presidency. On the opposite extreme, while many African countries have

adopted French-style, that is, mixed constitutions, there is a strong sense that real power lies

with the president (Bratton and van de Walle 1997, Carlson 1999).

Note, however, that it should matter whether the rules in a country allow for behavior that

is proscribed in another. In almost every instance where the formal rules do not seem to match

practice at a first glance, we find examples of behavior that conform to the constitutional

prerogatives of the president and/or the assembly. Thus, in Iceland, for example, the presidents

constitutional prerogative of choosing the formateur was crucial for bringing to power the coalition

between the Social Democratic Party and the Independence Party that governed between 1959 and

1971. Similarly, the head of states decision to form a non-partisan government after two

legislative elections and successive failed attempts at government formation by different parties

played an important role in the formation of future governments in Iceland (Kristinsson 1999:93-

94)

With respect to the mixed democracies that may look more like presidential ones, examples

of government changes due to confidence votes, or threats of confidence votes, abound: on April

11, 1995, the prime minister of Central African Republic, Jean-Luc Mandaba resigned upon the

filing of a no-confidence motion signed by a majority of National Assembly members; on May 19,

1993, the prime minister of the Comoro Islands resigned after losing a vote of no-confidence in the

legislature; on June 18, 1995, the Comoran president dissolved the assembly to forestall a vote of

no-confidence in the government; the Congolese government of Stphane Maurice Bongho-

27

Nouarra fell on November 14, 1992, as a result of a vote of no-confidence approved by the

assembly; the government of prime minister Rosny Smarth in Haiti survived a vote of no-

confidence on March 27, 1997; in Madagascar, the government of prime minister Emmanuel

Rakotovahiny fell on May 17, 1996, after a motion of no confidence was approved by 109 to 15

votes; in Niger, the government of Souley Abdoulaye resigned on October 16, 1994 after losing a

non-confidence vote in the assembly; faced with the choice of appointing a prime minister

supported by opposition parties or dissolving the National Assembly, Nigers president Mahamane

Ousmane chose to dissolve the assembly and call elections for January 1995; the government of

Abdirizak Hadji Husseing in Somalia lost a vote of non-confidence on July 13, 1964, after which it

resigned. In other countries, such as Albania, Armenia, Brazil (in 1962), Senegal, Sri Lanka and

Taiwan, there is evidence that political practice was clearly guided by the possibility that the

legislature could pass a vote of no confidence in the government.

We thus take constitutional provisions regarding government formation seriously in

classifying parliamentary, presidential and mixed democracies. Given that this classification is

conditional on the fact that the regime is a democracy, that is, on the fact that governments come to

power as a result of contested elections, we distinguish the different ways in which the link

between elections and governments is specified.

6. MONARCHIC, MILITARY, AND CIVILIAN DICTATORSHIPS

The way in which governments are removed from power drives the distinction between

democracies and dictatorships and distinguishes among types of democracies. The method of

removal is no less important for dictatorships. Yet in dictatorships, we know that there is no one

institution, such as elections or lottery, which determines the removal and succession of

authoritarian leaders. Dictatorial regimes, in fact, frequently succumb to internal disputes over

28

leadership succession.

We also know, however, that members of the ruling elite constitute the first major threat to

dictators.13 Dictators, in fact, are frequently deposed by a fellow member of the regime. As a

member of the ruling elite, the usurper is in a privileged position to gain the guns and support he

needs to successfully depose the incumbent. So to mitigate the threat posed by elites, dictators

frequently establish inner sanctums where real decisions are made and potential rivals are kept

under close scrutiny.

We distinguish dictatorships according to the characteristics of these inner sanctums.

Monarchs rely on family and kin networks along with consultative councils; military rulers confine

key potential rivals from the armed forces within juntas; and, civilian dictators usually create a

smaller body within a regime party a political bureau to co-opt potential rivals. Because

decision-making power lies within these small institutions, they generally indicate how power is

organized within the regime, to which forces dictators are responsible, and who may be likely to

remove them.

What is noteworthy of monarchs, both traditionally and currently, is their reliance on their

family and kin networks to come to power and maintain it. Hereditary succession is the rule among

monarchs, but primogeniture is not. As a consequence, family members can play a crucial role in

deliberating on succession to the throne and, by extension, on other important matters. In Kuwait,

for example, succession alternates between two branches of the Sabah family, but the most basic

rule of the succession is that family elects the ruler by consensus, based on the perception by

family leaders of their own best interests (Herb 1999: 80). In Oman, the next-in-line must be a

13 As opposed to threats stemming from within civil society, such as unions, professional associations, religious

institutions, and other larger social groups.

29

male descendant from the al-Said family, but must also be chosen by a family council. Saudi

succession became resolved by a more consensual process after Faysal established the Higher

Committee of Princes as an advisory council to the king on issues of succession. The Committee's

composition was designed to rally the entire family, and the Committee was even given the

authority to supervise the succession in the event of Faysal's death (Bligh 1984: 88).

Military rulers, in turn, come to power through coup dtats and see themselves as

guardians of the national interest, who come to the rescue of a nation at the brink of disaster

wrought by corrupt and myopic civilian politicians. They justify their role as neutral arbiters on

the basis of their membership within the armed forces, an institution that is supposed to stand

above politics. Hence the popularity of titles such as National Redemption Council, Committee

of National Restoration, or Council for National Salvation, to designate the ruling military

junta.

Once in power, military dictators harness the organizational apparatus of the armed forces

to consolidate their rule. Since the armed forces already control the territory through their

monopoly of violence, it takes just a small step to use this apparatus to serve the regime. Rule

typically takes the form of a junta. When coups are led by generals on behalf of the institutional

military, their juntas typically are small and include heads of the various service branches. When

power is seized by lower-ranked members of the military, juntas tend to be larger given the new

rulers need to attract members to their cause. Finer (1988: 260) reports that in contrast to most

Latin American juntas, normally composed of the three or four heads of the service branches,

juntas outside of the region were usually organized by middle-ranking officers and averaged 11

members. The incorporation of the military also takes other forms. Thus, under General Suharto, a

fifth of the Indonesian parliamentary seats were reserved for member of the armed forces, and a

30

soldier was stationed in each village serving as the military representative (Brooker 1995). In

Argentina under the Processo, before it was considered by members of the junta, legislation was

reviewed by various subcommittees within each service branch, as well as by the Legislative

Action Committee, composed by members from each branch. (Fontana 1987).

Internecine fighting among the branches of the armed forces is common. In Chile under

Pinochet, for example, each service branch had its own intelligence service that spied not only on

the population, but on members of other branches (Barros 2002). As a matter of fact, this internal

fighting is one of the biggest constraints on military rulers while in power. The consequence of the

delegation of power and distribution of spoils to the militarys branches may very well serve to

strengthen other members of the military who then overthrow the current dictator. In Argentina,

General Videla was deposed by General Viola, his fellow junta member, who was, in turn, ousted

by General Galtieri. These palace coups occurred in spite of the Argentina generals' careful attempt

to regulate succession. Similarly, in Nigeria, Major General Muhammad Buhari was deposed by

senior members of Supreme Military Council which then decided to install Major General Ibrahim

Babangida as the head of government.

Unlike monarchs and military dictators, civilian rulers do not have a ready-made

organization on which to rely. Most civilian dictators do not have sufficient family and kin

networks to establish permanent dynastic succession, and cannot appeal to the armed forces in the

same way that a military dictator can. To counteract their precarious position and to have an

organization through which they may govern, civilian dictators usually have a regime party, the

necessity of which was most forcefully recognized by Lenin (1921). A party is an instrument by

which the dictatorship can penetrate and control the society (Huntington and Moore 1970).

Members of a single party mobilize popular support and supervise behaviors of people unwilling

31

to identify themselves with the dictator. In exchange, the party offers individuals willing to

collaborate with the regime a vehicle for advancing their careers within a stable system of

patronage. The party also extends access and legitimacy to particular groups in making demands

on the government, and usually a smaller body within the regime party is used to co-opt rivals.

It seems, thus, that we may distinguish types of dictatorships in terms of the nature of their

executive office. Regimes in which the executive comes to and maintains power on the basis of

family and kin networks are classified as monarchies. Regimes in which the executive relies on the

armed forces to come to and stay in power are military. All other dictatorships, many of which

characterized by the presence of a regime party, are civilian. Figure 2 summarizes the steps

necessary for unambiguously identifying each type of dictatorship.

*** Figure 2 here ***

Who Rules? The first step is to identify who is the effective ruler. In democracies, this

identification is easy: it is the president in presidential democracies and the prime minister in

parliamentary and mixed democracies. In dictatorships, identification is frequently unproblematic:

usually the ruler is the president, the king, and, less frequently, the prime minister, the head of the

military junta, the leader of the ruling party or the executioner of the state of emergency. But

sometimes the nominal ruler is not the effective head of the government. In most communist states

the general secretary of the communist party is usually the effective head of government even

though the chairman of the Council of State, or president, is the head of state. In other cases, such

as in Somozas Nicaragua, an minence grise lurks behind the scenes as elections duly occur and

presidents change according to constitutional rules.

Does the head of government bear the title of king and have a hereditary successor and/or

predecessor? The ruler is a monarch if he, first, bears the title of king or emir, and, second,

32

takes power or is replaced by rules of hereditary succession. Most monarchs are identified by the

first rule. The second rule is for slightly more complicated cases in which the title of king has

been taken more recently. In two instances during the post-war period, a member of the armed

forces seized power and declared himself king. If he succeeded in passing power to a family

member, as did Reza Khan to his son, Mohammad Reza Pahlavi, in Iran, both members are

considered to be monarchs. If, however, the ruler fails in his succession plans, he is not considered

to be a monarch. Jean-Bedel Bokassa of the Central African Republic falls into the latter category.

A colonel in the army, he seized power in 1966, declared himself Emperor, and planned to have

his son succeed him. His dynastic plans collapsed, however, once he was deposed in 1979.

This rule highlights an important point about modern-day monarchs. In considering

whether a ruler is a rightful successor, we look only at whether the ruler belongs to the current

family in power. We do not determine whether that family or individual has historically well-

founded claims to the throne since contemporary monarchs rule in countries that often were carved

by colonial powers without reference to historical claims or social considerations. British colonial

authorities created the Transjordan state, for example, and installed Abdullah, a member of the

Hashemite family, on its throne. Because he was succeeded by a family member, both Abdullah

and his successors are considered monarchs.

Is the head of government a current or past member of the armed forces? The effective head of

government is a military ruler if he is or was a member of the institutionalized military prior to

taking power.14 Even if retired from service, the shedding of his uniform does not eliminate his

14 Leaders who belonged to the armed forces during World War II, but then left, are an exception. Because almost

all able-bodied men at the time either volunteered or were drafted, membership in the military during only this

period does not count towards ones type. This exception mostly affects those communist rulers of eastern Europe

who fought in World War II.

33

military status. Attempts to appear more palatable to voters who are more accustomed to civilian

rule do not erase these rulers connections and access to the armed forces.

Not included as military dictators are those rulers who come to power as heads of guerilla

movements. Successful insurgency leaders, such as Castro in Cuba, Musaveni in Uganda, and

Kagame in Rwanda, are considered to be civilian rulers. One might object that heads of guerilla

movements, often like military rulers, come to power using violence. In addition, once in power,

these rulers often give themselves military titles or become heads of the armed forces themselves.

Yet there are three good reasons not to consider those involved in guerilla movements as military

leaders. First, not all leaders who originated from guerilla movements were involved in fighting.

Many of them were members of the civilian, political arm of the successful movement and have no

more experience in warfare than the average civilian on the street. In addition, some guerilla

leaders, once they take power, never assume a formal military role. Even though Castro wears

fatigues, the leadership of the Cuban armed forces belongs to Raoul, his brother. Finally, and most

importantly, having never been a member of the armed forces, these leaders do not answer to that

institution. And since the constraints and support offered by the armed forces to one of their

members in power is the main reason for distinguishing military from non-military leaders,

guerilla leaders do not fall into this category.

Is the head neither monarchic nor military? As discussed above, civilian leaders often create a

regime party through which they govern. Yet, unlike kin networks with monarchs and the armed

forces with military rulers, the party does not define the civilian ruler. The diversity of modes of

government is what characterizes them and, for this reason, we think it is best to leave them as a

residual category. Thus, if dictators do not qualify as either monarchs or military rulers, they are

civilian.

34

Our reliance on a single aspect of dictatorships the nature of their executive office

makes our classification unique. Existing classifications of authoritarian regimes are based on

multiple criteria, rely on subjective judgments, and/or are defined, at least in part, by the behavior

of leaders. Braton and van de Walles (1997), for example, identify five regime types, none of

which they call a democracy. Although these regimes may exhaust the range of political

arrangements predominant in postcolonial Africa (p.77), it is not clear they exhaust the space of

possible dictatorships. Moreover, their precise location in each of the two dimensions they

consider competition and participation is not clearly defined, as it depends on their analysis of

the characteristics of the leaders, the nature of the decision-making process, the level of pluralism

among the elites, the policy implementation process, among other factors. Geddes (1999: 20), in

turn, defines some dictatorships as personalist if the leader, who usually came to power as an

officer in a military coup or as the leader of a single-party government, had consolidated control

over policy and recruitment in his own hands, in the process marginalizing other officers

influence and/or reducing the influence and functions of the party. Leaving aside the problems of

observability this definition entails, it is easy to see that it precludes the study of a number of

questions that may be of interest to researchers (since it is based on the behavior to be explained).

Because our coding entails no assumptions about dictators behavior, it allows for the

testing of hypotheses related to the institutional structure of dictatorships. For example, we can

determine if, in fact, military dictators are more likely to allocate greater resources to the armed

forces. We can investigate whether the survival of dictators is due to their economic resources or

their monarchical structure. We can examine whether civilian dictators may be more likely to build

personalist coalitions by spending on private goods since they do not have a ready-made

organization with which to rule. Obviously the classification we offer here will not account for all

35

of the variation we observe in outcomes generated in authoritarian regimes. But it a base upon

which further work can be built. Suppose for example, that military dictators vary quite

significantly in their duration in power, with some surviving for no more than one or two years

while others hold on to power for quite a long time. We, then, can determine whether the

difference is due to the conditions under which they govern or some other objective trait about the

military dictators themselves, such as their rank. But making such further distinctions is impossible

if we do not first objectively identify the set of military dictators and determine the degree to

which membership in the armed forces is an explanatory factor. Similarly, if some civilian

dictators are better able to consolidate power, we can determine whether other factors, such as their

previous political experience or their control of a nationalist movement, explain their success. The

answer will not be found by ex ante collapsing these leaders into a single category with a new

label.

CONCLUSION

The empirical testing of hypotheses related to the emergence and survival of regimes, and

their consequences for policies and outcomes, remains the focus of much of the disciplines

research agenda. We join an increasingly large number of scholars who advocates careful choice of

measurement as a function of the research questions under investigation, as well as conceptual and

operational clarity of the instrument to be used. (e.g., Gleditsch and Ward 1997, Reiter and

Tillman 2002, Treier and Jackman 2003). We also offer an alternative to existing measures of

political regimes.

We have, thus, a six-fold classification of political regimes that emphasizes the institutions

capable of removing the government from power. Among democracies, we distinguish

parliamentary (only the legislature can remove the government), presidential (only the president

36

can remove the government) and mixed (both the legislature and the president can remove the

government) regimes. In dictatorships we distinguish monarchic (family and kin networks remove

the government), military (the armed forces remove the government) and civilian (a residual

category often characterized by the presence of a political party as the institution capable of

determining the fate of existing governments) regimes.

This classification is not the only possible one, as both democracies and dictatorships may

and have been distinguished in many different ways. Yet, it does posses the same attributes of the

classification of regimes as democracies and dictatorships on which it is rooted. Because it is based

on observational data, it is reproducible, a characteristic that is not present in any of the existing

alternative measures of political regimes. Moreover, the regime classification presented here, again

contrary to other measures, can be meaningfully interpreted as it is based on attributes of political

regimes that are identifiable and recognized as important by most researchers. Finally, as

suggested throughout the paper, this classification is useful in addressing a number of important

research questions that have been shaping the agenda of contemporary political scientists.

37

BIBLIOGRAPHY

Alvarez, Michael, Jos Antonio Cheibub, Fernando Limongi and Adam Przeworski. 1996.

"Classifying Political Regimes." Studies in Comparative International Development 31(2):3-36.

Arat, Zehra. 1991. Democracy and Human Rights in Developing Countries. Boulder, CO: Lynne

Rienner.

Barros, Robert. 2002. Constitutionalism and Dictatorship: Pinochet, the Junta, and the 1980

Constitution. Cambridge: Cambridge University Press.

Bligh, Alexander. 1984. From Prince to King: Royal Succession in the House of Saud in the

Twentieth Century. New York: New York University Press.

Bollen, Kenneth A. 1980. Issues in the Comparative Measurement of Political Democracy.

American Sociological Review 45(3):370-390.

Bollen, Kenneth A. 1989. Democracy, Stability, and Dichotomies. American Sociological

Review 54:612-621.

Bollen, Kenneth A. 1991. Political Democracy: Conceptual and Measurement Traps in Alex

Inkeles, ed. On Measuring Democracy: Its Consequences and Concomitants. New Brunswick:

Transaction Books.

Bratton, Michael and Nicolas van deWalle. 1997. Democratic Experiments in Africa: Regime

Transitions in Comparative Perspective. Cambridge: Cambridge University Press.

Brooker, Paul. 1995. Twentieth-Century Dictatorships: the Ideological One Party States. New

York: New York University Press.

Carlson, Rolf. 1999. Presidentialism in Africa: Explaining Institutional Choice. Ph. D.

Dissertation, Department of Political Science, University of Chicago.

38

Collier, David and Robert Adcock. 1999. Democracy and Dichotomies: A Pragmatic Approach to

Choices about Concepts. Annual Review of Political Science 2:537-65.

Coppedge, Michael. 2002. Democracy and Dimensions: Comments on Munck and Verkuilen.

Comparative Political Studies 35(1):35-39.

Coppedge, Michael and Wolfgang H. Reinicke. 1991. Measuring Polyarchy in Alex Inkeles, ed.

On Measuring Democracy: Its Consequences and Concomitants. New Brunswick, NJ:

Transaction. Pp.47-68.

De Schwinitz, Jr., Karl. 1964. Industrialization and Democracy: Economic Necessities and

Political Possibilities. Glencoe: The Free Press.

Diamond, Larry. 1999. Developing Democracy: Toward Consolidation. Baltimore: Johns Hopkins

University Press.

Diamond, Larry. 2002. Thinking About Hybrid Regimes. Journal of Democracy 13, 2: 21-35.

Elkins, Zachary. 2000. Gradations of Democracy? Empirical Tests of Alternative

Conceptualizations. American Journal of Political Science 44(2):293-300.

Epstein, David, Robert Bates, Jack Goldstone, Ida Kristensen, and Sharyn OHalloran. 2003.

Democratic Transitions. Manuscript. Department of Political Science, Columbia University.

Fearon, James D. 1994. Domestic political audiences and the escalation of international disputes.

American Political Science Review 88 (3): 577-592.

Fearon, James D. and David D. Laitin. 2003. Ethnicity, Insurgency, and Civil War. American

Political Science Review 97 (1): 75-90.

Fish, M. Steven Fish. 2002. Islam and Authoritarianism. World Politics 55(1).4-37

Fontana, Andres Miguel. 1987. "Political Decision-Making By a Military Corporation: Argentina

1976-1983." Dissertation. Department of Political Science, University of Texas -- Austin.

39

Gandhi, Jennifer and James Vreeland. 2004. Political Institutions and Civil War: Unpacking

Anocracy. Emory University, manuscript.

Gasiorowski, Mark J. 1996. An Overview of the Political Regime Change Dataset. Comparative

Political Studies 29(4):469-83.

Geddes, Barbara. 1999. "What Do We Know About Democratization After Twenty Years?"

Annual Review of Political Science 2: 115-144.

Gleditsch, Kristian S. and Michael D. Ward. 1997. Double Take: A Reexamination of Democracy

and Autocracy in Modern Polities. Journal of Conflict Resolution 41(3):361-383.

Golder, Matthew. 2004. Democratic Electoral Systems Around the World, 1946-2000"

Forthcoming in Electoral Studies.

Haggard, Stephan and Robert Kaufman. 1995. The Political Economy of Democratic Transitions.

Princeton: Princeton University Press.

Hegre, Havard, Tanja Ellingsen, Scott Gates, and Nils Petter Gleditsch. 2001. Toward a

Democratic Civil Peace? Democracy, Political Change, and Civil War, 1816-1992. American

Political Science Review 95: 33-48.

Hellman, Joel S. 1998. Winners Take All: The Politics of Partial Reform in Postcommunist

Transitions. World Politics 50(2):203-234, February.

Herb, Michael. 1999. All in the Family: Absolutism, Revolution, and Democracy in the Middle

Eastern Monarchies. Albany, NY: State University of New York Press.

Huber, John D. 1996. The Vote of Confidence in Parliamentary Democracies. American

Political Science Review 90:269-82

Huntington, Samuel and Clement Moore. 1970. Authoritarian Politics in Modern Society: the

Dynamics of Established One-Party Systems. New York: Basic Books.

40

Kristinsson, Gunnar Helgi. 1999. Iceland. In Robert Elgie, ed. Semi-Presidentialism in Europe.

Oxford: Oxford University Press.

Lenin, Vladimir. 1921. "Preliminary Draft Resolution Of The Tenth Congress Of The R.C.P. On

The Syndicalist And Anarchist Deviation In Our Party." In The Tenth Congress of the R.C.P.

Verbatim Report, March 8-16, 1921. Moscow.

Levitsky, Steven and Lucan A. Way. 2002. The Rise of Competitive Authoritarianism. Journal

of Democracy 13(2):51-65.

Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-

Six Countries. New Haven: Yale University Press.

Mainwaring, Scott. 1993. Presidentialism, Multipartism, and Democracy: The Difficult

Combination. Comparative Political Studies 26 (2):198-228.

Mainwaring, Scott, Daniel Brinks and Anbal Prez-Lin. 2001. Classifying Political Regimes in

Latin America, 1945-1999. Studies in Comparative International Dev