Eur Polit Sci (2019) 18:84–96

https://doi.org/10.1057/s41304-018-0149-8

RESEARCH DATASET

Classifying political regimes 1800–2016: a typology

and a new dataset

Carsten Anckar1 · Cecilia Fredriksson2

Published online: 14 March 2018

© European Consortium for Political Research 2018

Abstract  In the present text, we introduce a classification scheme where we

decompose democracies and autocracies into several categories. Based on this clas-
sification scheme, we create a global dataset covering the time period 1800–2016. In
the dataset, we make yearly observations for all countries that have been independ-
ent at any point in time since the Second World War. Regarding democracies, we
first distinguish between republics and monarchies. We then split the category of
republics into presidential, semi-presidential, and parliamentary systems. Within the
category of monarchies, almost all systems are parliamentary, but a few countries
are conferred to the category semi-monarchies. Authors differ markedly in terms
of how autocratic regimes should be classified. In the present dataset, we classify
autocratic countries into the following main categories: absolute monarchy, military
rule, party-based rule, personalist rule, and oligarchy. Within the categories party-
based rule and oligarchy, we also identify a number of subcategories.

Keywords  Autocracy · Democracy · Global dataset · Political regimes

Electronic supplementary material  The online version of this article (https​://doi.org/10.1057/

s4130​4-018-0149-8) contains supplementary material, which is available to authorized users.

* Carsten Anckar
carsten.anckar@abo.fi
Cecilia Fredriksson
cecilia.fredriksson@phdstudent.hhs.se
1
Department of Political Science, Åbo Akademi University, Fänriksgatan 3, 20500 Turku,
Finland
2
Department of Management and Organization, Stockholm School of Economics, Saltmätargatan
15‑17, 113 59 Stockholm, Sweden
Classifying political regimes 1800–2016: a typology and a new… 85

Introduction

During the last two or three decades, we have witnessed a growing interest in his-
torical comparative analysis, and a number of new global datasets on political
regimes have emerged (e.g., Alvarez et  al. 1996; Cheibub et  al. 2010; Boix et  al.
2013; Coppedge et  al. 2017). Another trend in recent years is that authoritarian
regimes have received a growing amount of attention at the same time as there are
indications that less research efforts are devoted to democracies and democratization
(Goode and Ahram 2016: 823–824). These trends have revitalized the question how
political regimes in general, and authoritarian ones in particular, should be defined
and classified.
Although the dominating strategy of categorizing political regimes still appears
to be to separate democracies from non-democracies, either qualitatively or quanti-
tatively, a number of authors have chosen to account for regime varieties within the
respective categories. However, it is striking that whereas a number of scholars (e.g.,
Marshall et al. 2014; Vanhanen 2013; Boix et  al. 2013) have assessed the democ-
racy–autocracy dimension for the latest two centuries, there is still, to our knowl-
edge, no dataset in which democratic and autocratic regimes have been subclassified
according to regime types for the same period of time.
The ambition with the present database is to provide researchers with a longi-
tudinal dataset on democratic and autocratic regime characteristics which extends
further back in time than any comparable database. It contains yearly observations
for all countries that have been independent at any point in time between 1946 and
2016. The fact that the dataset provides a subclassification of authoritarian regimes
is of particular importance; whereas the overwhelming majority of the countries
were authoritarian before the end of the First World War, existing databases (Geddes
et al. 2014; Cheibub et al. 2010; Hadenius and Teorell 2007; Kailitz 2013) only pro-
vide classifications of authoritarian regimes for the last 70 years or so.1

Defining and classifying political regimes

The authors of the three most cited datasets on political regimes differ widely in
their conceptualization of regimes. In the present work, our point of reference is
Hadenius and Teorell who presented a first version of their dataset a decade ago
(Hadenius and Teorell 2007) and a slightly revised and updated one 6  years later
(Wahman et al. 2013). Their categorization of regimes is based “on the institutions
on which …elites rely in order to regulate the access to and maintenance of public
authority” (Wahman et  al. 2013: 21). An important consequence of applying this

1
  The extensive V-dem dataset (Coppedge et al. 2017) is a highly welcomed contribution as it contains
data on a wide range of regime variables from the year 1900 onward. This dataset is currently being com-
plemented (historical V-dem), with data for the time period 1800–1920 (Knutsen et al. 2016). However,
neither of these databases provides a qualitative categorization of the regimes of the countries (although
some of the variables contained in the datasets are extremely valuable as a source for making such clas-
sifications).
 86 C. Anckar, C. Fredriksson

definition is that institutions and not leadership are crucial for determining when
a regime change has occurred. The focus on institutions means that “[t]he identity
or longevity of dictators is left out of the picture, as are the ideological stance of
the regime and its source of legitimation…” (Wahman et  al. 2013: 20). Geddes
et al. (2014: 314), in their dataset, chose a different strategy, where emphasis is on
“the rules that identify the group from which leaders can come and determine who
influences leadership choice and policy.” An important practical consequence of
these different approaches is that whereas Geddes et al. sometimes identify several
regimes during a period where the regime type remained the same (Geddes et  al.
2014: 315, 323), a regime change in Hadenius and Teorell’s dataset always presup-
poses a change from one regime category to another.
Following Geddes et al. (2014) as well as Cheibub et al. (2010), we consider the
holding of contested elections to be the central criterion when separating democracies
from autocracies. In line with a number of authors we also stipulate that the criterion
of competition must be complemented with a criterion of participation (e.g., Dahl
1971; Vanhanen 1990; Munck and Verkuilen 2002; Geddes et al. 2014). As pointed
out by Munck and Verkuilen (2002: 11), the inclusion of this criterion is particularly
important in datasets which extend far back in time since “a key feature of the experi-
ence with democratization in the nineteenth and early twentieth centuries as opposed
to the late twentieth century … [was] the gradual expansion of the right to vote.”
Since it is our ambition to subclassify democratic and autocratic regimes, the
two categories must be separated using a qualitative scale. To do so, we rely on
the dataset by Boix et al. (2013), who make use of a dichotomous qualitative scale,
classifying 213 countries as either democracies or autocracies on a yearly basis for
the time period 1800–2010. Focusing on elections, Boix et  al. separate democra-
cies from autocracies. Following Dahl (1971), their measure also encompasses two
dimensions of elections, competition, and participation.
Since the dataset by Boix et  al. does not cover the period 2011–2016, we add
these years following the classification of Boix et  al. as much as possible. The
V-dem electoral democracy index (v2x_polyarchy) corresponds very well with Boix
et al.’s conception of democracy. We use this source as the most important one for
our classifications. In addition, we compare values with the scores countries have
received on the political rights dimension in Freedom House (www.freed​omhou​
se.org) and also on the Polity 2 scale in the Polity IV-dataset (Marshall et al. 2014).
In Fig. 1, we provide an illustration of how the regimes are classified in the data-
set. We base the classifications on the situation prevailing on December 31 each
year. Within the category of democracies, we make a primary distinction between
constitutional monarchies and republics. We thereafter split up the category of
republics into presidential, semi-presidential, and parliamentary systems. Within the
category of constitutional monarchies, almost all systems are parliamentary, but we
confer a few countries to a category labeled semi-monarchic systems.
When classifying authoritarian systems into categories, our point of departure is
the four broad categories Geddes et  al. operate with, namely monarchy, military,
dominant party, and personalist. However, we differ in our distinction of, the cri-
teria of inclusion in the respective categories from Geddes et al., and we also treat
hybrid regimes differently. We treat Oligarchies as a category of its own, and we
Classifying political regimes 1800–2016: a typology and a new… 87

Regime

Democracy Autocracy

Monarchy (Absolute)
Military rule
Party-based
Monarchy (Const.) Republic single-party
multi-party
Parliamentary Presidential Personalist
Semi-monarchy Semi-presidential Oligarchy
Parliamentary monarchic oligarchy
other oligarchy

Fig. 1  Structure of the database

also identify a separate subcategory of this regime type, labeled monarchic oligar-
chies. In addition, we use the term party-based regime (which is the one Geddes
et al. use in their codebook) instead of dominant party and split up this category into
single-party and multi-party authoritarian regimes.
The distribution of the regimes is presented in Table  1. Among the democratic
countries, we note that the vast majority of the cases are republics, and within this
category, the parliamentary form of government is almost as popular as the presi-
dential one. The semi-presidential form of government occurs less frequently,
although it has become increasingly popular during the last three decades. Among
monarchies almost all cases are, naturally, parliamentary systems, whereas semi-
monarchies are quite rare. Among autocracies, the category of monarchies contains
the largest number of cases, followed by military regimes. The number of party-
based regimes and personalist regimes is approximately the same, whereas oligar-
chies occur less frequently.

Democracies

Presidentialism

Most authors agree that at least three criteria should be met in order for a system to
be presidential, namely that the president (or rather the chief executive) is elected
by popular vote; the government cannot be dismissed by a parliamentary vote of no
confidence; and the president appoints and directs the government (e.g., Shugart and
Carey 1992: 19; Sartori 1997: 83–84).
In the dataset, we apply this minimal definition of presidentialism. However, for
the sake of parsimony, we ease the criterion of popular election of the president
to some extent. There are, for instance, countries like the USA and Bolivia, where
 88 C. Anckar, C. Fredriksson

Table 1  Regime types in the

world 1800–2016 Democracies 6175
 Constitutional monarchies 2193
  Parliamentary 1962
  Semi-monarchy 231
 Republics 3959
  Parliamentary 1439
  Presidential 1639
  Semi-presidential 881
 Hybrid 23
  Parliamentary (Andorra) 23
Autocracies 11,467
  Absolute monarchy 3068
  Military rule 2904
  Party-based rule 2109
   Single-party rule 1280
   Multi-party rule 829
  Personalist rule 2014
  Oligarchic systems 1372
   Monarchic oligarchy 797
   Other oligarchy 575
Occupation, civil war, otherwise unclear 552
Total 18,194

presidents occasionally have been elected by parliament and currently Micronesia,

Suriname, and Switzerland have presidents elected by parliament, while they oth-
erwise function in accordance with the principles of presidentialism. In line with,
for instance, Cheibub et al. (2010), we consider these countries presidential. Since
some authors likely prefer a stricter application of the criterion of popularly elected
presidents, the database includes information also on whether the president has been
popularly elected or not.

Parliamentarism

The central feature of parliamentarism is that the government is dependent on

the legislature for its survival. We also add another important defining character-
istic, suggested by Budge et al. (1997: 238), namely that there is “no popularly
elected president with real political powers.” This criterion is important since it
separates parliamentary systems from semi-presidential ones. Here, however, we
substitute “popularly elected president” with “head of state.” The crucial issue is
whether or not relations between the executive and the legislature are affected by
a powerful “third actor,” not by how this actor is titled or how he or she comes
Classifying political regimes 1800–2016: a typology and a new… 89

to power. (There are, for instance, countries where the monarch has more or less
the same position as a president in a semi-presidential system.)

Semi‑presidentialism

A minimal definition of semi-presidentialism stipulates that executive powers

are shared by a president and a prime minister, who is responsible to parlia-
ment. In addition, most authors add the requirement that the president is popu-
larly elected (e.g., Duverger 1980: 166; Shugart and Carey 1992: 23–27; Sartori
1997: 131; Shugart 2005). The question how much power a president must pos-
sess in relation to the prime minster has been a subject of controversy among
authors throughout the last three to four decades. One strategy is to disregard
the power dimension altogether and define as semi-presidentialism all systems
where there is a popularly elected president alongside a prime minister and gov-
ernment responsible to the legislature (Elgie 1999: 13; Schleiter and Morgan-
Jones 2010: 1418; Cheibub et  al. 2010: 81). Although such a strategy has the
advantage of making classifications easier, it has the disadvantage of blurring
the distinction between parliamentary systems and semi-presidential systems,
since, in the end, the crucial criterion of classification boils down to how a
(powerless) head of state is selected.
We therefore argue that a definition of semi-presidentialism must also include
the criterion that executive power is shared by the prime minister and the head
of state. In order to separate semi-presidential systems from parliamentary sys-
tems, we use Siaroff’s (2003) measure of presidential powers. However, not all
powers are of relevance when separating semi-presidentialism from parliamenta-
rism. The crucial issue is that the president is in possession of important explicit
executive powers, at least one of the following Siaroff’s criteria should be met:
(1) the president has the power to chair cabinet meetings, (2) is in charge of for-
eign policy, (3) has a central role in government formation (and/or dissolution),
or (4) [has the] ability … to dissolve the legislature at will.
Admittedly, the power to dissolve the legislature does not fall explicitly within
the executive sphere, but it is a very powerful tool for a president to indirectly
affect government formation and termination (e.g., Shugart 2005: 334–335; Neto
and Lobo 2009). As in the case of presidentialism, we have eased the criterion
that the president must be popularly elected for the sake of parsimony (compare
with Anckar 1999; O’Neil 1993). However, the vast majority of countries where
the president is in possession of executive powers also have popularly elected
presidents. Since some authors may wish to include the popular election of the
president as a necessary criterion of semi-presidentialism, the database also pro-
vides information on whether the president was popularly elected or not. The
inclusion of this additional variable makes it possible to apply a wider definition
of semi-presidentialism, which disregards the criterion that the president must
share executive powers with the president.
 90 C. Anckar, C. Fredriksson

Semi‑monarchy

Executive power sharing mechanisms are of much less relevance in constitutional

monarchies than in republics, and it is consequently not surprising that this issue has
received very little scholarly attention. However, some countries have been classified
as democracies by most of the authoritative sources despite the fact that a hereditary
monarch has, or has had, more or less the same position as a president in a semi-
presidential system. We choose to label such systems semi-monarchic systems. In
all, the number of semi-monarchic systems has been quite limited in the world. Gen-
erally, they have existed for short periods of time in former absolute monarchies
immediately after the countries in question have surpassed the threshold of democ-
racy, for instance in Italy 1919–1921 and Sweden 1911–1916. However, there are
also countries that have experienced longer periods of semi-monarchic rule, notably
Liechtenstein, where the regime form continues to be in use.

Autocracies

Absolute monarchy

There tends to be little disagreement among authors regarding the defining char-
acteristics of a monarchy. Hadenius and Teorell (2007: 146) regard as monarchies
“those regimes in which a person of royal descent has inherited the position of head
of state in accordance with accepted practice or the constitution.” In the present
work, we shall apply the definition by Hadenius and Teorell. Admittedly, it is not
always clear what we mean by “accepted practice.” It happens that rulers are ousted
from office by members of their own family, something that happened in Qatar in
1995 and in Saudi Arabia in 1975, when king Faisal was assassinated by his nephew
Faisal bin Musaid (Cheibub et al. 2010: 85). Therefore, we apply a generous defi-
nition of “accepted practice” and accept instances where the monarch is ousted of
office and replaced by family members as hereditary succession. At the same time,
it is important to emphasize that “accepted practice” does not mean that all cases
where power is passed from father to son count as monarchies (Hadenius and Teor-
ell 2007: 146). For instance, in Syria, Bashar al Assad succeeded his father Hafez al-
Assad as president in 2000, but he did not inherit the position of president; rather, he
was approved for the position in a public referendum according to the constitutional
provisions of the country.
Furthermore, we concur with the view of Hadenius and Teorell that a monarchy
does not emerge simply because a political leader proclaims him- or herself king,
queen, emperor etc. Rather, the crucial question is whether or not the monarch was
installed by himself/herself or by other actors. In Europe, during the time period in
question, parliaments generally installed new monarchical dynasties. In such cases
we regard the regime as a monarchy despite the fact that the new regent did not
inherit his or her position. The same strategy applies to countries where a monarch
was (s)elected by parliament or third parties at the time the country received its
independence
Classifying political regimes 1800–2016: a typology and a new… 91

However, we do not consider a person who puts himself in power, for instance
as a consequence of an act of treason, a rebellion or a civil war, as a monarch. The
fact that a hereditary monarchy has been formally proclaimed by the regime does
not alter this conclusion. We regard the notion of “accepted practice” fulfilled only
when the new ruler has been succeeded by a member of the same family, and con-
sider the regime a monarchy from that point in time.

Military rule

Following Nordlinger (1977: 2), Hadenius and Teorell define military regimes as
instances “in which military officers are major or predominant actors by virtue of
their actual or threatened use of force,” and where “the armed forces may exercise
political power either directly or indirectly by controlling civilian leaders” (Hadenius
and Teorell 2007: 146). This definition is more inclusive than the one proposed by
Geddes et al. (2014: 319), who define military rule as “rule by an officer constrained
by other officers” and less inclusive than the definition used by Cheibub et al. (2010:
88) who stipulate that “[t]he effective head of government is a military ruler if he is
or was a member of the institutionalized military prior to taking power.”
The only drawback with the definition used by Hadenius and Teorell is a practi-
cal one. It is often rather difficult to determine how much control the military should
exercise in order for the regime to be regarded as a military regime. Another issue of
concern is the inherent notion of the military regime as an exceptional form of gov-
ernment, which often means that in situations where the military is reluctant to give
up its powers, it tends to transform itself to a civilian government. The tricky ques-
tion is to determine when a military regime is turned into something else. The fact
that the military establishes a party which prevails in a fraud election does not turn a
military regime into a one-party regime and neither does the fact that a general puts
on a suit and assumes the office of president turn the system into personalist rule.
These problems are of course aggravated by the fact that the dataset covers more
than 200 years. We use the simple rule that a regime is considered a  military regime
if the country has been uninterruptedly ruled by the same person who came to power
in a military coup. In addition, the country must have remained autocratic during
this period. However, this is not a rule without exceptions. Especially in cases where
a military leader has transformed himself or herself into a civilian leader and stayed
in power for an extensive period of time, it is likely that his or her power base has
shifted from the military to another institution.

Party‑based rule

In line with Geddes (2003: 51), we define party-based rule as systems where “access
to political office and control over policy are dominated by one party, though other
parties may exist and compete as minor players in elections.” Most countries that
do not meet the requirements of democracy have political parties and conduct elec-
tions on a regular basis. However, this category of countries includes a wide vari-
ety of political systems. In the most extreme cases, such as the Soviet Union and
 92 C. Anckar, C. Fredriksson

currently Laos, only one party is legally allowed. At the other side of the spectrum,
we find countries where real opposition parties are allowed to exist and participate
in elections, but the dominant party makes sure it has the support of a majority of
the members in parliament.
Hadenius and Teorell (2007) originally made a distinction between four catego-
ries of authoritarian regimes with elected assemblies. However, they subsequently
found it more reasonable to operate with only three, namely no-party, one-party,
and multi-party regimes (Wahman et  al. 2013). As it turns out, no-party regimes
are extremely difficult to find. One-party regimes are “authoritarian regimes with
only one legal party (formally or de facto)” (Wahman et al. 2013: 26). However, the
category also includes systems where “satellite parties,” i.e., parties “that are inde-
pendent in name but [that] do not take oppositional positions” exist. In multi-party
regimes “at least a minimal level of competition is allowed and some opposition
candidates (although not necessarily all) are allowed to participate in national elec-
tions” (Wahman et  al. 2013: 27). Thereby “[multi-party regimes should] normally
have more than one party represented in the parliament” (although this might not
be the case if the opposition chooses to boycott the elections) (Wahman et al. 2013:
27).
We follow this line of reasoning and split the category of party-based rule into
the subcategories single-party systems and multi-party authoritarian systems. This
makes it possible for users of the dataset to either adopt a broad definition of one-
party rule or not.

Personalist rule

Basing their argument on Brooker (2000), Hadenius and Teorell (2007: 149) dismiss
the notion of personalist regimes. Instead, they consider personalism to be a charac-
teristic which, to varying extents, denotes all authoritarian systems. Geddes et al., on
their part, argue that personalist regimes are “autocracies in which discretion over
policy and personnel are concentrated in the hands of one man, military or civilian”
(Geddes et al. 2014: 319). This definition resembles the one proposed by Hunting-
ton (1991: 111) who defines personal dictatorship as a system where “the individual
leader is the source of authority and [where] power depends on access to, closeness
to, dependence on, and support from the leader.” In line with Geddes et al. we find
it relevant to consider personalism a regime type of its own. Hadenius and Teorell
base their classification on three modes of power access and maintenance: heredi-
tary succession in monarchies, military force in military regimes, and popular elec-
tions in party-based regimes. However, this classification disregards systems where
the power holder lacks a strong institutional backing and secures his or her position
on personal power.
Here, we define personalist regimes according to Linz’ (2000: 151–155) defini-
tion of sultanism [which is based on Weber’s (1968) definition]. According to Linz,
sultanism is characterized by a fusion of the personal and public sphere, where “[t]
he ruler exercises his power without restraint at his own discretion and above all
unencumbered by rules or by any commitment to an ideology or value system”
Classifying political regimes 1800–2016: a typology and a new… 93

(Linz 2000: 151–152). The lack of an ideology is of particular importance when

separating personalist regimes from party-based regimes. Although ruthless leaders
in one-party regimes can behave much in the same respect as leaders in personalist
regimes, their power is exercised with reference to an ideological totality. In a sul-
tanistic society, on the other hand, if anything resembling an ideology exists, it has
emanated from, and is consistent with, the person in power. In other words, as Linz
and Stepan (1996: 53) argue, “there could be questions raised as to whether Stalin’s
practices and statements were consistent with Marxism–Leninism, but there would
be no reason for anyone to debate whether Trujillo’s statements were consistent with
Trujilloism.”
Another distinctive feature of sultanism (and personalist rule) is a low degree of
institutionalization and organization. Public office holders receive their positions
from the ruler in person, and since their careers are completely dependent on the
capriciousness of the ruler, they are loyal to the ruler in person, not the office (Linz
2000: 152–153). Furthermore, the persons who receive the highest positions in the
regime “are neither “disciples” nor old fighters of a movement party or conspira-
torial group,” but rather “men who would not enjoy any prestige or esteem in the
society on their own account … [and] whose power is derived exclusively from the
ruler” (Linz 2000: 151–152), quotation marks in original.

Oligarchy

In line with Geddes et al. (2014: 318), we classify systems which meet the criteria
of democracy with regard to public contestation (i.e., with free and fair elections and
with an executive responsible either to the legislature or to the people) but where
only a minority (here  less than half of the adult male population) has the right to
vote as oligarchies. However, in contrast to Geddes et al., we do not find it relevant
to collapse oligarchies into the party-based category since the countries classified
as oligarchies are not necessarily dominated by one single party (although this was
indeed the case in South Africa from 1948 to 1994, with the Nationalist Party in
power). In many oligarchies, two (or sometimes more) parties compete for power
in fair elections and election results determine the composition of government. Fur-
thermore, a handful of other types of political systems also qualify as oligarchies.
Currently, for instance, in Iran power effectively resides in the hand of the religious
clergy.
Within the category of oligarchies, we also extract a separate regime category
labeled monarchic oligarchy. Especially for countries where the process of democ-
ratization is slow and extends over many decades, it makes sense to pay regard to
the relationship between, on the one hand, the monarch and the legislature and the
government on the other. Britain during the time period 1800–1884 constitutes a
good example. Although it was far from always clear how the powers between the
monarch and the prime ministers were divided, it was equally evident that the pow-
ers of the monarch were significantly greater than those of a purely ceremonial head
of state. The political system of Britain, therefore, was a special variant of oligar-
chy. The term monarchic oligarchy is consequently used in order to describe systems
 94 C. Anckar, C. Fredriksson

where a hereditary monarch shares power with a parliament and government in a

country where elections are free and fair, but where franchise is restricted to such an
extent that the country does not qualify as a democracy.

Conclusion

Despite the fact that there has been a growing interest in conducting global com-
parative research with historical data, there is still a shortage on historical data on
regime characteristics. Particularly with regard to autocratic systems, this is prob-
lematic, since the share of autocratic countries tends to increase, the farther back
in time we go. In order to remedy this shortcoming, the present article introduces
a new global dataset, which subclassifies democratic and autocratic regimes into a
number of categories for the period 1800–2016. As there is far from any agreement
on how authoritarian regimes should be classified, we have, to a very large extent,
based the categorization on existing pioneering works in the field. In addition, the
database contains a number of variables and subcategories of regimes which allow
authors to apply different definitions of controversial regime types like semi-presi-
dentialism, one-party rule, and oligarchy.
The database is likely to advance research on a number of topics in political sci-
ence. For instance, in peace and conflict studies, the democratic peace argument
has been widely debated. However, whereas authors have paid a lot of attention to
how democratic institutions affect the propensity for external and internal conflicts,
they have directed much less attention toward the question whether the propensity
for conflicts varies between different types of authoritarian regimes. So, far large
N-studies regarding the association between different authoritarian regime types
and conflicts have been rare, and either covered, at maximum, only the period after
the Second World War (e.g., Peceny et al. 2002; Ishiyama et al. 2008; Fjelde 2010;
Conrad and Souva 2011) or, in cases where the time periods under study have been
longer, paid attention to regime similarity in terms of degree of autocracy (or democ-
racy) rather than on qualitative differences in regime types (e.g., Bennett 2006). The
present dataset, however, makes it possible to conduct such analyses very far back in
time, thus making it possible to detect time—(as well as region)—specific patterns
of variable relations.
The dataset can also make a significant contribution to research concerning ques-
tions such as whether certain regime types are more stable than others and whether
certain regime types are more likely to be succeeded by a specific regime type than
others. The empirical evidence at hand suggests that military regimes tend to be
the most unstable authoritarian regime type (Geddes 1999, 2003: 78; Smith 2005;
Brownlee 2009; Magaloni 2008), whereas one-party regimes and monarchies are the
most stable authoritarian regime types (Geddes 1999, 2003; Smith 2005: 78; Had-
enius and Teorell 2007; Brownlee 2009; Magaloni 2008). However, these studies
cover, at most, the period after the Second World War. The database presented here
makes it possible to expand the time period extensively, which is important since a
phenomenon like regime breakdowns tends to occur infrequently.
Classifying political regimes 1800–2016: a typology and a new… 95

References
Alvarez, M., J.A. Cheibub, F. Limongi, and A. Przeworski. 1996. Classifying Political Regimes. Stud-
ies in Comparative International Development 31(2): 3–36.
Anckar, D. 1999. Finland inför millenniumskiftet: en regimbestämning. Statsvetenskaplig tidskrift
102(3): 241–261.
Bennett, D.S. 2006. Toward a Continuous Specification of the Democracy–Autocracy Connection.
International Studies Quarterly 50(2): 313–338.
Boix, C., M. Miller, and S. Rosato. 2013. A Complete Data Set of Political Regimes. 1800–2007.
Comparative Political Studies 46: 1523–1554.
Brooker, P. 2000. Non-democratic Regimes: Theory, Government and Politics. New York: St.
Martin’s.
Brownlee, J. 2009. Portents of Pluralism: How Hybrid Regimes Affect Democratic Transitions. American
Journal of Political Science 53(3): 515–532.
Budge, I., K. Newton, et al. 1997. The Politics of the New Europe: Atlantic to Urals. London: Longman.
Cheibub, J.A., J. Gandhi, and J.R. Vreeland. 2010. Democracy and Dictatorship Revisited. Public Choice
143(1): 67–101.
Conrad, J., and M. Souva. 2011. Regime Similarity and Rivalry. International Interactions 37(1): 1–28.
Coppedge, M., J. Gerring, S.I. Lindberg, S.-E. Skaaning, J. Teorell, et al. 2017. V-Dem Codebook v7.
Varieties of Democracy (v-Dem) Project. https​://www.v-dem.net/en/refer​ence/versi​on-7-may-2017.
Accessed 8 Jan 2018.
Dahl, R. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University Press.
Duverger, M. 1980. A New Political System Model: Semi-presidential Government. European Journal of
Political Research 8(2): 165–187.
Elgie, R. 1999. The Politics of Semi-Presidentialism. In Semi-Presidentialism in Europe, ed. R. Elgie,
1–21. Oxford: Oxford University Press.
Fjelde, H. 2010. Generals, Dictators, and Kings. Authoritarian Regimes and Civil Conflict 1973–2004.
Conflict Management and Peace Science 27(3): 195–218.
Freedom House. http://www.freed​omhou​se.org. Accessed 12 Dec 2017.
Geddes, B. 1999. What Do We Know About Democratization After Twenty Years? Annual Review of
Political Science 2: 115–144.
Geddes, B. 2003. Paradigms and Sand Castles: Theory Building and Research Design in Comparative
Politics. Ann Arbor: The University of Michigan Press.
Geddes, B., J. Wright, and E. Frantz. 2014. Autocratic Regimes and Transitions. Perspectives on Politics
12(2): 313–331.
Goodey, J.P., and A.I. Ahram. 2016. Special Issue Editors’ Introduction. Observing Autocracies from the
Ground Floor. Social Science Quarterly 97(4): 823–833.
Hadenius, A., and J. Teorell. 2007. Pathways from Authoritarianism. Journal of Democracy 18(1):
143–156.
Huntington, S.P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman: Uni-
versity of Oklahoma Press.
Ishiyama, J., R. Conway, and K. Haggans. 2008. Is There a Monadic Authoritarian Peace: Authoritarian
Regimes, Democratic Transition Types and the First Use of Violent Force. African Journal of Politi-
cal Science and International Relations 2(3): 31–37.
Kailitz, S. 2013. Classifying Political Regimes Revisited: Legitimation and Durability. Democratization
20: 39–60.
Knutsen, C.H., J. Møller, and S.-E. Skaaning. 2016. Going Historical: Measuring Democraticness Before
the Age of Mass Democracy. International Political Science Review 37(5): 679–689.
Linz, J. 2000. Totalitarian and Authoritarian Regimes. Boulder, CO: Lynne Rienner.
Linz, J., and A. Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe,
South America, and Post-Communist Europe. Baltimore: Johns Hopkins University Press.
Magaloni, B. 2008. Credible Power-Sharing and the Longevity of Authoritarian Rule. Comparative Polit-
ical Studies 41(4/5): 715–741.
Marshall, M.G., T.R. Gurr, and K. Jaggers. 2014. Polity IV Project: Political Regime Characteristics and
Transitions, 1800–2013. Arlington: Center for Systemic Peace, George Mason University.
Munck, G., and J. Verkuilen. 2002. Conceptualizing and Measuring Democracy: Evaluating Alternative
Indices. Comparative Political Studies 35(1): 5–34.
 96 C. Anckar, C. Fredriksson

Neto, O.A., and M.C. Lobo. 2009. Portugal’s Semi-presidentialism (Re)considered: An Assessment of
the President’s Role in the Policy Process, 1976–2006. European Journal of Political Research
48(2): 234–255.
Nordlinger, E. 1977. Soldiers in Politics: Military Coups and Government. Englewood Cliffs, NJ:
Prentice-Hall.
O’Neil, P. 1993. Presidential Power in Post-communist Europe: The Hungarian Case in Comparative Per-
spective. Journal of Communist Studies 9(3): 177–201.
Peceny, M., C. Beer, and S. Sanchez-Terry. 2002. Dictatorial Peace? American Political Science Review
96(1): 15–26.
Sartori, G. 1997. Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and
Outcomes. 2nd ed. London: Macmillan.
Schleiter, P., and E. Morgan-Jones. 2010. Who’s in Charge? Presidents, Assemblies, and the Political
Control of Semipresidential Cabinets. Comparative Political Studies 43(11): 1415–1441.
Shugart, M.S. 2005. Semi-presidential Systems: Dual Executive and Mixed Authority Patterns. French
Politics 3: 323–351.
Shugart, M.S., and J.M. Carey. 1992. Presidents and Assemblies: Constitutional Design and Electoral
Dynamics. New York: Cambridge University Press.
Siaroff, A. 2003. Comparative Presidencies: The Inadequacy of the Presidential, Semi-presidential and
Parliamentary Distinction. European Journal of Political Research 42(3): 287–312.
Smith, B. 2005. Life of the Party: The Origins of Regime Breakdown and Persistence under Single-Party
Rule. World Politics 57(3): 421–451.
Vanhanen, T. 1990. The Process of Democratization: A Comparative Study of 147 States, 1980–1988.
New York: Crane Russak.
Vanhanen, T. 2013. Measures of Democracy, 1810–2012. FSD1289, Version 6.0. Tampere: Finnish
Social Science Data Archive.
Wahman, M., A. Hadenius, and J. Teorell. 2013. Authoritarian Regime Types Revisited: Updated Data in
Comparative Perspective. Contemporary Politics 19(1): 19–34.
Weber, M. (1968) Economy and Society, 3 vols (Edited and trans: Roth, G., and Wittich, C.S.). New
York: Bedminster Press.

Carsten Anckar  is Professor of political science (comparative politics) at Åbo Akademi University,
Finland. He is the author of Determinants of the Death Penalty (Routledge, 2004) and Religion and
Democracy (Routledge, 2011).

Cecilia Fredriksson  is a Ph.D. student in Business Administration at Stockholm School of Economics

in Sweden. She is also affiliated with Stockholm Center for Organizational Research (Score).
 Reproduced with permission of copyright owner.
Further reproduction prohibited without permission.