Content uploaded by Abraham Diskin
Author content
All content in this area was uploaded by Abraham Diskin on Dec 23, 2013
Content may be subject to copyright.
http://ips.sagepub.com/
International Political Science Review
http://ips.sagepub.com/content/26/3/291
The online version of this article can be found at:
DOI: 10.1177/0192512105053787
2005 26: 291International Political Science Review
Abraham Diskin, Hanna Diskin and Reuven Y. Hazan
Why Democracies Collapse: The Reasons for Democratic Failure and Success
Published by:
http://www.sagepublications.com
On behalf of:
International Political Science Association (IPSA)
can be found at:International Political Science ReviewAdditional services and information for
http://ips.sagepub.com/cgi/alertsEmail Alerts:
http://ips.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://ips.sagepub.com/content/26/3/291.refs.htmlCitations:
What is This?
- Jul 12, 2005Version of Record >>
by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from by guest on October 11, 2013ips.sagepub.comDownloaded from
Why Democracies Collapse: The Reasons for
Democratic Failure and Success
ABRAHAM
DISKIN, HANNA DISKIN, AND REUVEN Y. HAZAN
A
BSTRACT. Most studies of democratic stability are based within either the
socioeconomic or the politico-institutional tradition, but usually not on
both. This article combines the two approaches. In all, 11 variables
associated with democratic stability are divided into four groups
(institutional, societal, mediating, and extraneous) and examined in 30
cases of democratic collapse and 32 cases of stable democracies. Five
variables prove to be the most influential on the fate of democracies.
When a country scores negatively on four of these five variables it is
almost doomed to collapse. Some of the variables prove to be correlated
in an opposite way to that which has been suggested in the literature.
Keywords: • Democracy • Economy • Government • Parties • Society
• Stability
The Controversy Over Democratic Stability
There are a number of comparative and empirical studies centered on democratic
stability. Most have focused on either the socioeconomic or the politico-
institutional tradition, but usually not on both. Moreover, the bulk of the research
addresses the transition to and consolidation of democracy, rather than its long-
term stability. An empirical investigation of variables from both research traditions
concerning their relevance to democratic stability has, thus far, been rare (Berg-
Schlosser and De Meur, 1994; Berg-Schlosser and Mitchell, 2000, 2002;
Gasiorowski and Power, 1998; Hadenius, 1994). This kind of research deserves
much praise for its interdisciplinary nature and focus, along with its distinctive
findings concerning the impact of each set of variables on the other, but there is
still room for expansion and development.
This article encompasses four groups of relevant independent variables to
assess why democratic regimes collapse. The first group is made up of institutional
International Political Science Review (2005), Vol 26, No. 3, 291–309
DOI: 10.1177/0192512105053787 © 2005 International Political Science Association
SAGE Publications (London, Thousand Oaks, CA and New Delhi)
variables, and addresses elements ranging from the type of regime to the
concentration of powers within it. The second group includes societal variables,
and focuses on factors ranging from the democratic historical background
through to social cleavages. The third group is comprised of mediating variables,
which are located between the macro-political and the macro-societal groups, and
exhibits variables ranging from the nature of the party system to the level of
government and coalition stability. The last group is actually a single variable,
foreign involvement, which has rarely received any attention in the discussion of
democratic stability, but which has proven to be essential. The 11 variables from
these four categories are elaborated below.
Institutional Variables
The four institutional variables adopted in this study include two that have
received substantial attention in the literature, and two that have not. The first
variable examined is federalism, which distinguishes between unitary governments
(Great Britain) on the one hand, and governments with federal (Germany) or
semi-federal (Cyprus pre-1974) features on the other. We hypothesize that federal
states are more prone to democratic collapse than unitary ones. The introduction of
federalism can lead to center–periphery struggles that could undermine
democracy, as was the case in the American Civil War. This hypothesis runs
counter to the small amount of literature that is available on the relationship
between this variable and democratic collapse. For example, Lijphart (1984a,
1999) includes federalism as one of the majority-restraining elements that
identifies his consensual model of democracy. Federalism is, therefore, conducive
to democratic stability, and not damaging (Lijphart, 1990). Diamond (1990),
Diamond et al. (1995), and Horowitz (1994) all focus on the usefulness of
federalism in alleviating ethnic and other social conflicts in order to achieve a
more stable democracy in pluralistic societies. Hadenius (1994), however, found
no significant quantitative relationship between federalism and the durability of
democracy in developing countries.
A second variable examined is presidentialism, which distinguishes between
parliamentary systems (Italy) and presidential or semi-presidential systems (the
USA and France, respectively). The bulk of the literature leads us to hypothesize
that presidential or semi-presidential regimes are more prone to democratic collapse than
parliamentary ones. The reason is the possibility of conflict between the executive
and legislative branches. This variable has produced a plethora of research,
particularly in the 1990s. For example, Stepan and Skach (1994) argue that
presidentialism can impede the consolidation of democracy. Yet, the clearest case
against presidentialism is presented by Linz (1990, 1994). Among the perils of
presidentialism for democratic stability Linz includes rigidity, zero-sum elections,
and dual legitimacies – all of which undermine the ability to produce com-
promises, which are a basic need for a democratic system. Horowitz (1990),
Mainwaring (1993), and Shugart and Carey (1992), among others, have defended
presidentialism. The debate is far from settled, yet the trend is in favor of
parliamentarism (Lijphart, 1991, 1995). Three recent quantitative analyses either
have found no evidence to support one regime type (Power and Gasiorowski,
1997) or have come down in favor of parliamentarism (Hadenius, 1994; Przeworski
et al., 1996).
A third institutional variable examined is proportionality, which distinguishes
292 International Political Science Review 26(3)
between cases in which electoral proportionality is low (Canada) and cases in
which the electoral results preserve proportionality (The Netherlands). The
literature points to the hypothesis that proportional electoral systems are more prone to
democratic collapse than those with less proportionality. The influence of proportionality,
however, can produce two offsetting results. On the one hand, high propor-
tionality can expand fragmentation, thereby increasing the fragility of the party
system. On the other hand, it can expand representation, thereby containing
conflicts and reducing their spillover outside the system. While the debate over
proportionality continues (Blais, 1991; Blais and Dion, 1990; Hadenius, 1994;
Lardeyret, 1991; Lijphart, 1994, 1999; Sartori, 1994), the literature has very few
proponents who advocate high proportionality. Even those scholars who favor
proportionality suggest a moderate version rather than an extreme one.
Therefore, it is the negative consequences of high proportionality that appear
more often in the literature.
The last institutional variable examined is constitutional weakness (instability),
which distinguishes between cases with stable constitutions (the USA), exhibited
by minor constitutional changes over time; cases with unstable, frequent, or major
constitutional changes (Thailand); and cases that lack any stable constitution
(Israel). We hypothesize that political systems with low constitutional stability are more
prone to democratic collapse than those with high constitutional stability. The reason is the
normative framework within which democracies function, that is, frequent
constitutional changes could be an indicator, not only a cause, of democratic
instability. Clear rules of the game, particularly in times of crisis, can prevent the
clashes that undermine democracy (Lutz, 1994). The issue of constitutional
stability has not been considered widely in the literature, but does have adherents
on both sides of the argument. Merkl (1993), for example, states that flexible
constitutions are preferable. Lijphart (1984a, 1990), on the other hand, posits that
rigid constitutions with judicial review are part of his consensual model, and
suggests this format for new democracies.
Societal Variables
The bulk of attention in the literature has been on two of the three societal
variables. The first societal variable is cleavages, which distinguishes between
countries with minor cleavages (Sweden) and those with deep or parallel social
cleavages (Lebanon). The literature points to the hypothesis that countries with deep
or parallel social cleavages, or both, are more prone to democratic collapse than those with low
or cross-cutting cleavages, or both. Lipset’s (1959, 1960, 1994) theory, whereby cross-
cutting cleavages make for a more stable democracy, is widely accepted in the
literature. Dahl (1971), Horowitz (1994), Lijphart (1977), and Powell (1982)
elaborate on the relationship between social cleavages and democratic perform-
ance. Yet, the literature mainly assesses which kinds of social cleavages, or lack
thereof, lead to a more stable democracy, whereas we are interested in inverting
this relationship to examine the influence of cleavages on the breakdown of
democracy.
The second variable is a malfunctioning economy, which distinguishes between
cases with minor economic problems (Norway) and those with significant
economic problems (Weimar Germany and a large number of democracies in the
developing world). The bulk of the literature leads us to hypothesize that countries
with weak or unstable economies are more prone to democratic collapse than those with stable
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 293
economies. Przeworski et al. (1996, 2000) examine economic performance (that is,
growth, absence of crises, and low inflation) and the stability of democracy. They
conclude that affluence (continued positive economic performance) contributes
to democratic stability. Powell (1982) examines the relationship between
economic inequality and violence. These are only some of the studies that
measure economic performance, not economic development, and its relationship
to democratic stability. Lipset (1959), for example, measures indices of democratic
development (wealth, industrialization, urbanization, and education) and claims
that the more economically developed a nation is, the more likely it is to sustain a
democratic regime. Lipset’s claim has been confirmed by numerous scholars who
have carried out multivariate analyses, such as Bollen (1979), Cutright (1963),
Cutright and Wiley (1969), Lipset et al. (1993), and Olsen (1968) among others.
The discussion among these scholars concerns the type of relationship, rather
than the existence of such a relationship. Is the relationship linear or not (Bollen
and Jackman, 1985)? Is the main variable economic development or income
inequality (Hadenius, 1994; Muller, 1988, 1995)?
A third societal variable, unfavorable history, is actually a cluster of variables: it
deals with the democratic background of the system on the basis of its historical
experience, its political culture, and the degree of development of its civil society.
We hypothesize that countries with undemocratic or mixed historical backgrounds (Peru
and Turkey, respectively) are more prone to democratic collapse than those with democratic
historical, cultural, and civil societal backgrounds (Switzerland). Almond and Verba’s
(1963) connection between their preferred mixed version of a “civic” culture and
the stability of a democratic regime is one of the most quoted references
regarding this variable, along with Lipset (1960) and, more recently, Diamond
(1993) and Putnam (1993). Huntington (1991) has also extrapolated a
relationship between culture and democratization. However, his work, as is true
for the bulk of the literature in this field, concerns the cultural prerequisites for
establishing democracy (that is, democratization) and not democratic stability or
collapse. Inglehart (1988) concludes that there is a connection between the
cultural character of the society and the tendency to adopt democratic
institutions. Diskin (2001) argues that “favorable history,” in the sense of the
development of civil society, as demonstrated by the role of the opposition
movement in Poland and the unique contribution of the Roman Catholic Church,
was a major factor behind the 1989 revolution in Eastern Europe.
A few scholars do assess the relevance of background variables for the survival
of democracy. Diamond et al. (1995), for example, find considerable evidence
that features of democratic culture are closely correlated with democratic stability.
Weiner (1987) observes that the British colonial model of tutelary democracy has
been more successful than other colonial models in sustaining democratic
institutions in newly independent countries. Weiner thus finds that the
institutional framework for democracy, once installed, helps create conditions for
its own persistence. Daalder (1966) argues that gradual democratic development,
particularly in relation to elite attitudes and economic growth, is a requirement
for democratic stability.
For operational purposes, reference was made also to the degree of domestic
violence and to previous democratic crises when measuring this variable in each of
the cases investigated.
294 International Political Science Review 26(3)
Mediating Variables
Three mediating variables between institutions and society were also assessed in
this study. The first of these is fragmentation, which deals with the degree of
fragmentation of the party system as represented by the number of parties in the
lower or single chamber of the parliament. The literature points to the hypothesis
that party systems with a high level of fragmentation (Weimar Germany) are more prone to
democratic collapse than systems with low fragmentation (the USA). The discussion of
fragmentation is largely responsible for defining the field of parties and party
systems. Since Duverger (1954) and Lipset (1960), the numerical counting of
parties and the resulting relationship with democratic stability has been a focus of
much of the literature. These scholars, among others such as Neumann (1956)
and Dahl (1971), perceived that the two-party system was a necessary condition for
political stability. Even those scholars who argue that fragmentation does not
necessarily lead to the breakdown of democracy (that is, that fragmented party
systems can be stable) advocate a moderate level of fragmentation rather than a
high one (Lijphart, 1977; Midlarsky, 1984; Rokkan, 1970; Sartori, 1976).
A second mediating variable is polarization, which distinguishes between party
systems with a low degree of polarization (Australia) and party systems in which
sizable anti-system political parties play a role (pre-Franco Spain). The bulk of the
literature leads us to hypothesize that highly polarized party systems are more prone to
democratic collapse than systems with low polarization. The reason is that polarization
produces extremist or anti-system parties which can undermine the legitimacy of the
regime, are usually excluded from power, and hinder the formation of governments
– thereby limiting the range of parties that have access to government. The end
result is democratic instability (Daalder, 1971; Hazan, 1995, 1997; Ieraci, 1992; Lane
and Ersson, 1987; Powell, 1986; Sani and Sartori, 1983; Sartori, 1976).
Another variable in the mediating category is governmental instability, which
refers to the degree of government stability based on the durability of government
coalitions or cabinets, or both. The literature leads to the hypothesis that unstable
governments, or governing coalitions (Fourth French Republic), are more prone to
democratic collapse than stable governments (Germany post-Second World War).
Moreover, governmental instability can also be an indication of an overall decline
in stability resulting from other factors, some of which make up other variables in
this study. The connection between government or coalition stability and regime
stability is usually examined in the framework of case studies, and the literature
concerning the level of government stability usually assesses the relationship in the
“opposite” direction, that is, the influence of regime attributes on government
stability (Dodd, 1976; Taylor and Herman, 1971). Moreover, most of the studies on
government and coalition stability are devoted to investigating particular
attributes of the coalition that lead to the stability or instability of the government
per se, rather than the democratic system in its entirety (Axelrod, 1970; Blondel,
1968; Browne et al., 1986; De Swaan, 1973; King et al., 1990; Laver, 1974; Laver
and Schofield, 1991; Laver and Shepsle, 1996; Sanders and Herman, 1977;
Warwick, 1994).
Extraneous Variable
An extraneous variable that does not fit into any of the three previous categories
in this article is foreign involvement, which deals with the involvement (or severe
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 295
threats) of foreign countries (or other elements) in domestic politics. The
explanatory power of this variable makes its inclusion imperative in any analysis of
democratic collapse. We hypothesize that countries experiencing serious levels of
involvement by foreign forces are more prone to democratic collapse than those with low
involvement. Foreign involvement has, in numerous cases, been the major factor in
bringing about an end to democratic rule. We excluded from our examination,
however, all cases of outright occupation, such as in France and Norway (1940),
Poland (1947), and Czechoslovakia and Hungary (1948). The literature has not
addressed this factor in a comparative, analytical manner. Obviously, there are
cases, such as western involvement in post-communist countries, in which foreign
influence is positive. Particular country or area studies have included a discussion
of the role of foreign forces in the collapse of democracy, but this variable has
rarely been included in those comparative studies that focus on democratic
stability (Fischer-Galati, 1992; Gasiorowski and Power, 1998).
Cases and Measurement
We examined 30 cases of collapsed democracies, compared to 32 cases of stable
democracies (countries with at least a generation of uninterrupted democratic
government). We could have included additional cases in both groups, and
omitted some others. The selection of cases depends on, among other factors, the
theoretical and operational definitions of the concept of democracy. This
definition, however, is not a simple matter (see, for instance, the opening chapter
of Przeworski et al., 2000). We made our final selection following numerous
suggestions by colleagues as well as from lists prepared by other scholars. Contrary
to other selections (for example, the Political Data Yearbook), we included in the
group of stable democracies a number of countries that are neither European nor
highly developed. On this matter, compare, for instance, Linz (1978) to Linz and
Stepan (1978). The group of collapsed democracies does not include cases in
which countries were far from being “full” democracies at the time of the collapse
(for example, Bolivia in 1964 and 1969 and Honduras in 1972) or where
democratic features continued to exist in spite of the “collapse” (for example,
India in 1975). When a country had several instances of collapse (for example,
Argentina and Turkey) only one case, usually the latest, was included in order to
reduce both theoretical and statistical problems. Nevertheless, despite the very
problematic nature of case selection, alternative selections examined did not
change the statistical results significantly.
We measured the values for the “collapsed” cases (1922–94) for the period
prior to collapse and values for the “stable” cases for 1998. The list of all 62 cases
appears in the Appendix. A tempting possibility could have been to examine only
those countries where democracy was restored, on the basis of “before and after.”
This, however, would have reduced the number of cases significantly. Przeworski et
al. (2000), among others, adopted an almost opposite approach. They examined
each country for each year as a separate case. Under certain conditions such a
selection is most desirable. When one is trying to compare “all” theories, though,
it does not seem very practical and it probably would not have produced
significantly different conclusions.
The cases in the category of collapsed democracies require a caveat. Some of
these cases did not reach the status of fully developed democracies, yet they were
included and received a low score on the democratic historical background
296 International Political Science Review 26(3)
variable. The reasons are threefold. First, some of the cases in the stable
democracies category were in a similar situation in their early development (that
is, they went through similar developmental phases and succeeded). Second, if
one were to assess the essence of democracy only after the consolidation phase was
complete, then a tautology could result. Such a definition of democracy would
create a regime type that simply does not collapse, unless it loses its sovereignty.
Third, defining democracy in an overly strict manner would produce few, if any,
countries that are, or were, “perfect” democracies because of possible contra-
dictions between different democratic properties (Arrow, 1964). This article,
therefore, assesses the collapse of democracy not only from a stable systemic
format, but also from the process of consolidation. Embryonic democracies are
thus included in a few cases, due to the “democratic experience” they have
undergone. That is, democratic competition took place in all of the cases included
in this study.
In short, this article adopts a “procedural minimum” definition of democracy,
in the tradition of Schumpeter (1947) and Dahl (1971). We focus on the
implementation of a small number of democratic procedural attributes at the level
of the national political regime (for example, competitive general elections).
These minimal procedures have been those most widely employed in research on
democracy and produce a viable standard for this regime type (Collier and
Levitsky, 1997). This definition is also in line with Sartori’s (1970) strategies
for achieving differentiation and avoiding “conceptual stretching,” while
simultaneously providing a more adequate, consistent, and clear basis for assessing
causal relationships.
Cases were coded at several levels and in a repeated manner, thereby allowing
for numerous checks. Researchers each received several cases and were asked to
code them according to the variables chosen. The cases allocated to the researchers
overlapped in a random order, so that at least three researchers assessed each case.
Discrepancies in the coding were reviewed by a research coordinator who was
exposed to all of the cases in the study, thus providing for uniformity in the coding
process. A panel of experts subsequently evaluated the results.
Rather than eliminate relevant variables that could not be quantified, we
decided at the final stage of the research to adopt binary measures. A theoretical
justification for this decision can be found in the approach suggesting that
opposite values of the same variables may result in the consolidation of democracy,
on the one hand, or in democratic collapse, on the other (see, for example, Linz
and Stepan, 1996). In other words, we assessed every case in terms of its possible
placement into the “pure” class for each variable, which is posited to contribute to
democratic stability (a “positive” coding). Those cases that did not fit, even if they
were a mixed or intermediate type, were placed in the opposing class,
membership in which, we argue, contributes to democratic collapse (a “negative”
coding).
Federalism, for example, is a basic test. If a country is not divided into regions
and if the upper house does not represent these regions with constitutionally
defined and safeguarded powers, the country is coded “positive” (unitary).
Presidentialism is almost as simple. If a country has a single election for a
legislative branch that is formally superior to the executive, it is coded “positive”
(parliamentary). Proportionality was measured using the indices prevalent in the
literature, such as Gallagher’s (1991) least-squares index, based on the available
election results prior to the collapse of democracy or, in the case of stable
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 297
democracies, on electoral results during the 1990s. Constitutional weakness was
measured not as a lack of rigidity within the constitution, but as actual, serious
changes in the constitution. Hence, one could have defined it as constitutional
instability. For example, substantial electoral reform or regime reform prior to the
collapse was coded as constitutional instability. At the same time, coders were
aware of the fact that at times the ability to reform institutions could reflect
democratic strength.
We assessed social cleavages according to the percentage of minorities, be they
ethnic, religious, national, or linguistic. If the minority measured 20 percent or
more, and the relevant literature elaborated a parallel nature for this cleavage, the
case was coded “negative.” A malfunctioning economy was measured based on the
economic status of the country as a whole, as well as according to economic gaps
within the country. We used measures such as GNP per capita, growth rate,
inflation, the Gini coefficient, and the Lorenz curve, whenever available, to code
the cases. A “positive” coding for an unfavorable history went to those cases with a
gradual movement toward democracy over decades, and its maintenance.
Conversely, cases that experienced an occasional absence of fully competitive
democracy, political violence, or previous major crises of their democratic regime
received a “negative” score.
We measured fragmentation of the party system according to the “effective”
number of parties (Laakso and Taagepera, 1979) in the last legislatures before the
collapse and the average of the legislatures in the 1990s for the stable
democracies. Polarization was measured based on the size and intensity of the
more extreme parties on a unidimensional scale. Those cases with both sizable
and extremist anti-system parties were coded “negative” (highly polarized),
whereas those cases that had either only one kind or none were coded “positive.”
We determined government stability according to the durability of the
government or governing coalition (Lijphart, 1984b). For example, a change in
the prime minister or in one of the coalition partners reduced the stability of the
government. Cases with frequent changes, both in coalition partners and in prime
ministers, were coded “negative.” Cases with low alternation were coded “positive,”
even if the governing coalition was normally a minority (Strøm, 1988), as, for
example, in Denmark.
The coding of foreign involvement was rather simple. The engagement of
Germany in Austria is an example of democratic collapse due to foreign
involvement and was coded “negative.” On the other hand, there were cases in
which foreign involvement was quite significant, yet the democratic regime
remained rather stable, for example, Finland.
Findings
The research findings are based on numerous assessments of the relationships
between and among the variables resulting from the coding of the cases. The
methods used include: simple breakdown tables that assess the different ratios
between the cases of democratic collapse versus those of democratic stability for
each variable; phi correlation coefficients and chi-square tests that exhibit the
extent of intercorrelation between any two pairs of dichotomous variables; and
logistic regression that shows how the dichotomous coding “produced” the
“expected” result, and indicates which variables were more important and the
extent to which they “predicted” why democracies collapsed.
298 International Political Science Review 26(3)
The phi correlation coefficient is the most appropriate for dichotomous
variables as it is equivalent to Cramer’s V, but indicates either a negative or a
positive correlation. The 55 correlation coefficients between the 11 independent
variables produce only 10 cases with a correlation coefficient of larger then 0.35. It
is interesting to note that five of the six highest phi scores are between only four of
the 11 variables: foreign involvement, malfunctioning economy, unfavorable
history, and social cleavages. Given this partial overlapping, the combined
predictive power of these variables (see below) is even more impressive.
Proportional elections and party-system fragmentation produced the highest phi
value (0.45). This is similar to Lijphart’s (1994: 76) Pearson coefficients of 0.29 in
57 proportional electoral systems and 0.45 in 70 electoral systems overall.
Table 1 provides an accumulation of statistics for each of the 11 variables. The
chi-square results show the relationship between each of the variables and
democratic collapse. This column produces several interesting findings. First, one
of the four institutional variables, federalism, produces insignificant results. A look
at the cases shows that the percentage of non-federal cases was practically
equivalent for both the stable and the collapsed democracies. Constitutional
weakness, on the other hand, is the strongest of these four variables vis-a-vis
democratic collapse, followed by presidentialism and proportionality. The
rationale for the first variable is quite apparent. The finding concerning
presidentialism is consistent with the literature. Indeed, of the stable democracies,
a majority are parliamentary (75.0 percent), while a minority of the collapsed
democracies are parliamentary (36.7 percent). This is correlated with the fact that
most European countries are parliamentary, while most countries in Latin
America are presidential. Proportionality, on the other hand, produced significant
results, but in the opposite direction to that expected. Of the collapsed
democracies, a majority had low proportionality (53.3 percent). Among the stable
democracies, the relationship is the opposite: only a minority of the cases have low
proportionality (34.4 percent). These findings run counter to the bulk of the
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 299
TABLE 1. Impact of Individual Variables
Logistic
Variable Chi-square Significance Phi prediction
Institutional
Federalism 0.19 0.667 +0.055 51.6%
Presidentialism 9.25 0.002 –0.387 69.4%
Proportionality 2.26 0.132 +0.191 59.7%
Constitutional weakness 11.49 0.001 –0.430 71.0%
Societal
Cleavages 17.42 0.000 –0.530 74.2%
Malfunctioning economy 19.21 0.000 –0.557 75.8%
Unfavorable history 24.35 0.000 –0.627 80.6%
Mediating
Fragmentation 3.11 0.078 +0.224 61.3%
Polarization 6.74 0.009 –0.330 66.1%
Governmental instability 14.54 0.000 –0.517 75.8%
Extraneous
Foreign involvement 25.93 0.000 –0.647 82.3%
literature devoted to this variable. That is, while some scholars defend propor-
tionality, stating that more proportional systems are not necessarily less stable than
less proportional systems, we are unaware of any study which argues that more
proportional electoral systems are more stable than those with low proportionality.
Nonetheless, this is the conclusion that our data produced.
In contrast to the mixed results for the four institutional variables, all three of
the societal variables provide robust and significant results. Indeed, they produce
three of the four highest results of all the variables. Social cleavages are related to
democratic collapse. Only a tiny minority of the democracies that collapsed had
low cleavages (6.7 percent), whereas a majority of the stable democracies possess
low cleavages (56.3 percent). Unfavorable history proved to be even more closely
connected to the collapse of democracy. Only a small minority of the collapsed
democracies had strong democratic historical backgrounds (10.0 percent),
compared to a majority of the stable democracies (71.9 percent). The
malfunctioning economy also produced impressive results. A look at the cases
shows that a majority of stable democracies also have stable economies (59.4
percent). The collapsed democracies, on the other hand, overwhelmingly lacked
stable economies (93.3 percent).
The three mediating variables also provided statistically significant results, yet
all were weaker than the societal variables, and one of them (as in the earlier case
of proportionality) proved to be counter-intuitive. The level of governmental
instability produced the highest score. Stable democracies also have stable govern-
ments (81.3 percent), while collapsed democracies lacked stable governments
(30.0 percent). Low party-system polarization is evident in a majority of the stable
democracies (59.4 percent), while it is absent in a large majority of the unstable
democracies (73.3 percent). Yet, it was party-system fragmentation that, like the
proportionality of the electoral system, produced results that ran counter to
the scholarly literature (see the phi scores in Table 1, which demonstrate the
“opposite/wrong” direction for both variables). A majority of the collapsed
democracies have low fragmentation (56.7 percent), while a substantial majority of
the stable democracies have medium or high fragmentation (65.6 percent). This
does not mean that particular cases, such as Poland in 1926 or Germany in 1933,
do not confirm the negative influence of high party-system fragmentation nor that
“atomization” does not have devastating effects.
The extraneous variable, foreign involvement, produced the most robust
findings, according to the chi-square and phi scores. Approximately nine out of
every 10 stable democracies incur a low level of involvement by foreign countries,
or other elements, in their domestic politics (87.5 percent), while in excess of
three out of four collapsed democracies were subject to relatively high foreign
involvement (76.7 percent).
In summation, the correlation of variables with democratic collapse and the
breakdown of each variable according to the 30 cases of collapsed democracies
and the 32 cases of stable democracies produced very interesting findings. The
strongest relationship to the collapse of democracies is not with the institutional
variables, but, rather, with the societal variables, the extraneous variable of foreign
involvement, and one mediating variable (governmental instability) which
together comprise the five strongest variables. Moreover, proportionality and
fragmentation tend to be related to the collapse of democracies in a manner
contrary to that usually suggested by the literature devoted to these topics.
When the dependent variable is binary, as in the cases examined in this article,
300 International Political Science Review 26(3)
the employment of logistic regression analysis seems most appropriate. It requires
fewer assumptions than discriminant analysis and it does not violate the
assumptions required by other multivariate techniques (Hosmer and Lemeshow,
1989). Interactive effects were also examined, but, with one exception, the results
do not add any substantial information to the concise models presented herein.
Logistic analysis fleshes out the relevance of these variables vis-a-vis the collapse
of democratic regimes in a more precise manner. The lowest possible predictive
power of an individual variable in our study is 51.6 percent. This is so because
when one predicts that “all” democracies will remain stable, the prediction will
prove correct in 32 of the 62 cases (51.6 percent). For there to be a contribution,
at least some of the cases must be predicted as a collapse. The more correct
predictions of stability and collapse that are produced by the logistic regression,
the better predictive power that particular variable has.
Thus, while many of the variables might be related to democratic collapse, most
do not contribute significantly to the correct prediction of which democracies will
collapse and which will not. The predictive power of five variables is less than 70
percent, while in only two cases (foreign involvement and unfavorable history) is it
greater than 80 percent. What remains to be seen is whether additional variables,
with or without independent predictive capabilities, can contribute to more robust
correct predictions overall.
A simple method to examine the cumulative contribution of variables is
demonstrated in Table 2. This table compares the number of negative scores of
the five most critical variables (cleavages, malfunctioning economy, unfavorable
history, governmental instability, and foreign involvement) in collapsed democracies
and in stable democracies. It is clear that the higher the number of negative values
for these variables, the lower the probability of a democracy remaining stable. Of
the 28 cases with two negative values or less, only one collapsed. On the other
hand, of the 27 cases with four or five negative scores, all but one collapsed.
An alternative way to examine the cumulative influence of the critical variables
is demonstrated in Tables 3–6. These tables represent, respectively, the best
logistic models for two, three, four, and five variables. The best two-variable logistic
model involves the two most influential variables: foreign involvement and
unfavorable history. The model correctly predicts 52 cases (83.9 percent),
compared with 51 or 50 correct predictions for each of the variables involved
when examined separately. In other words, even the best combination of two
variables does not increase our understanding of democratic instability.
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 301
TABLE 2. Collapsed and Stable Cases and the Number of Negative Scores on the Five Critical
Variables (Cleavages, Malfunctioning Economy, Unfavorable History, Governmental Instability,
and Foreign Involvement)
Number of negative scores Collapsed democracies Stable democracies Total
0077
101010
211011
3347
414115
512012
Total 30 32 62
The case with three-variable models is different. While four out of seven cases with
exactly three negative variables remained stable, the predictive power of the best
three-variable logistic model, which adds governmental instability to the two
variables employed in the previous model, is quite impressive. It correctly predicts
58 cases (93.5 percent). It should be noted that other three-variable models are
302 International Political Science Review 26(3)
TABLE 3. Best Two-Variable Logistic Model
Logistic Standard
Variable regression coefficient error Wald statistic Significance
Foreign involvement –3.049 0.866 12.39 0.000
Unfavorable history –3.046 0.909 11.22 0.001
Constant 9.361 2.288 16.75 0.000
Note: Dependent variable: Collapse
Number of cases: 62
Correctly predicted (overall): 83.9%
–2 log likelihood: 42.20
Nagelkerke R square: 0.675
TABLE 4. Best Three-Variable Logistic Model
Logistic Standard
Variable regression coefficient error Wald statistic Significance
Foreign involvement –3.399 1.042 10.63 0.001
Unfavorable history –3.544 1.188 8.896 0.003
Governmental instability –3.000 1.063 7.96 0.005
Constant 15.209 3.990 14.53 0.000
Note: Dependent variable: Collapse
Number of cases: 62
Correctly predicted (overall): 93.5%
–2 log likelihood: 30.99
Nagelkerke R square: 0.784
T
ABLE 5. Best Four-Variable Logistic Model
Logistic Standard
Variable regression coefficient error Wald statistic Significance
Foreign involvement –3.492 1.209 8.34 0.004
Unfavorable history –3.463 1.261 7.55 0.006
Governmental instability –3.293 1.236 7.10 0.008
Malfunctioning economy –2.826 1.340 4.45 0.035
Constant 20.631 5.69 13.142 0.000
Note: Dependent variable: Collapse
Number of cases: 62
Correctly predicted (overall): 95.2%
–2 log likelihood: 25.44
Nagelkerke R square: 0.831
much weaker. For example, we examined a model in which governmental
instability was replaced by malfunctioning economy. This model, which is the
second-best three-variable logistic model, correctly predicted only 54 cases (87.1
percent). Although malfunctioning economy is correlated with stability more than
governmental instability, its high correlation with the two strongest individual
independent variables reduces its impact in the combined model. When both
malfunctioning economy and governmental instability are used with foreign
involvement and unfavorable history, the best four-variable logistic model, the
predictive power increases to 95.2 percent (59 cases are correctly predicted and
only three are missed).
The five-variable model misses only two cases, with a predictive power of 96.8
percent. The missed cases are India in 1998 (with a very high probability of
collapse) and Uruguay in 1973 (with a negligible probability of collapse). Among
the stable countries, at 46.6 percent, Israel in 1998 is the case with the second
highest probability of collapse (see Diskin, 2003). The probability of collapse of
other cases, according to the five-variable model, is much lower. Among the
correctly predicted collapsed cases only Cyprus in 1974 had a very high probability
for remaining stable (49.4 percent).
Conclusion
This study attempts to transcend the politico-institutional and socioeconomic
theoretical approaches to democratic stability, employing a comparative and
empirical research design aimed at delineating which variables contribute to the
collapse of democratic regimes. In doing so, it presents a comprehensive
approach and generates findings that are both statistically significant and modify
previous theoretical studies. Based on these conclusions, it is easy to suggest
additional research designs and possible future investigative approaches. For
instance, it would be interesting to examine year-to-year changes in selected
democracies. A challenging research project conducted within such a framework
may help predict the potential stability of, for example, the new democracies.
This study investigated 11 variables: federalism, presidentialism, proportionality,
constitutional weakness, cleavages, a malfunctioning economy, unfavorable
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 303
T
ABLE 6. Best Five-Variable Logistic Model
Logistic Standard
Variable regression coefficient error Wald statistic Significance
Foreign involvement –3.189 1.245 6.56 0.010
Unfavorable history –2.948 1.290 5.22 0.022
Governmental instability –3.690 1.473 6.27 0.012
Malfunctioning economy –2.288 1.379 2.75 0.097
Cleavages –2.156 1.408 2.35 0.126
Constant 22.541 6.397 12.42 0.000
Note: Dependent variable: Collapse
Number of cases: 62
Correctly predicted (overall): 96.8%
–2 log likelihood: 22.71
Nagelkerke R square: 0.852
history, fragmentation, polarization, governmental instability, and foreign
involvement. All but one (federalism) produce results that show that they are
related to the collapse of democracy. Yet, two of the variables examined, the
fragmentation of the party system, and a related variable, the proportionality of
the electoral system, are correlated with the collapse of democracy in a direction
opposite to the one suggested by most of the literature.
Variables from all four categories examined (institutional, societal, mediating,
and extraneous) proved to be related to the collapse of democracy. It is evident,
however, that the most important cluster is composed of the societal variables.
It is important to emphasize that no single variable on its own is capable of
predicting democratic collapse. The findings show not only which variables are
related to the collapse of democracy, and in what hierarchical order, but that the
key to the demise of a democratic system is a combination of variables. The most
crucial variables are: cleavages, a malfunctioning economy, unfavorable history,
governmental instability, and foreign involvement. If four of these negative factors
appear simultaneously, the democratic regime is almost doomed to collapse.
Democracy is, therefore, neither fragile nor feeble, but, rather, a highly resilient
regime type. A single debilitating factor is highly unlikely to bring about its
collapse.
Appendix
List of Country Cases and their Probability to Collapse
Based on the Five-Variable Logistic Regression (Table 6)
Cases of democratic collapse, 1922–94 (%) Cases of democratic stability 1998 (%)
Argentina 1976 94.3 Australia 0.0
Austria 1933 99.8 Austria 0.5
Brazil 1964 99.8 Belgium 0.2
Chile 1973 95.5 Botswana 34.6
Colombia 1948 94.3 Canada 0.2
Costa Rica 1948 71.0 Costa Rica 5.8
Cyprus 1974 50.6 Denmark 0.0
Ecuador 1972 99.8 Finland 19.9
Estonia 1934 99.8 France 1.0
Fiji 1987 91.0 Germany 4.1
Gambia 1994 91.0 Greece 4.6
Germany 1933 99.8 Iceland 0.0
Ghana 1981 94.3 India 91.0
Greece 1967 97.6 Ireland 0.3
Grenada 1979 91.0 Israel 46.6
Guatemala 1954 91.0 Italy 16.4
Italy 1922 94.3 Jamaica 29.4
Latvia 1934 99.8 Japan 0.5
Lebanon 1975 91.0 Luxembourg 0.0
Lithuania 1926 99.8 Malta 0.3
Pakistan 1971 99.8 Mauritius 0.0
Peru 1992 91.0 Netherlands 8.1
Philippines 1972 91.0 New Zealand 2.1
Poland 1926 99.8 Norway 0.0
Portugal 1926 65.8 Portugal 4.6
Spain 1936 99.8 Spain 29.4
304 International Political Science Review 26(3)
Thailand 1991 94.3 Sri Lanka 2.1
Turkey 1980 99.8 Sweden 0.0
Uruguay 1973 2.1 Switzerland 0.2
Venezuela 1948 99.8 United Kingdom 0.2
USA 0.2
Venezuela 9.2
References
Almond, Gabriel and Verba, Sidney (1963). The Civic Culture. Princeton, NJ: Princeton
University Press.
Arrow, Kenneth (1964). Social Choice and Individual Values. New York: Wiley.
Axelrod, Robert (1970). Conflict of Interest. Chicago, IL: Markham.
Berg-Schlosser, Dirk and De Meur, Gisele (1994). “Conditions of Democracy in Interwar
Europe: A Boolean Test of Major Hypotheses,” Comparative Politics 26: 253–79.
Berg-Schlosser, Dirk and Mitchell, Jeremy, eds (2000). Conditions of Democracy in Europe,
1919–1939. Houndmills and New York: Palgrave Macmillan.
Berg-Schlosser, Dirk and Mitchell, Jeremy, eds (2002). Authoritarianism and Democracy in
Europe, 1919–1939. Houndmills and New York: Palgrave Macmillan.
Blais, Andre (1991). “The Debate over Electoral Systems,” International Political Science
Review 12: 239–60.
Blais, Andre and Dion, Stephane (1990). “Electoral Systems and the Consolidation of New
Democracies,” in Diane Ethier (ed.), Democratic Transition and Consolidation in Southern
Europe, Latin America and Southeast Asia. London: Macmillan.
Blondel, Jean (1968). “Party Systems and Patterns of Government in Western Democracies,”
Canadian Journal of Political Science 1: 180–203.
Bollen, Kenneth A. (1979). “Political Democracy and the Timing of Development,”
American Sociological Review 44: 572–87.
Bollen, Kenneth A. and Jackman, Robert W. (1985). “Political Democracy and Size
Distribution of Income,” American Sociological Review 50: 438–57.
Browne, E.C., Frendreis, J.P. and Gleiber, D.W. (1986). “The Process of Cabinet Dissolution:
An Exponential Model of Duration and Stability in Western Democracies,” American
Journal of Political Science 30: 628–50.
Collier, David and Levitsky, Steven (1997). “Democracy with Adjectives: Conceptual
Innovation in Comparative Research,” World Politics 49: 430–51.
Cutright, P. (1963). “National Political Development: Measurement and Analysis,” American
Sociological Review 28: 253–64.
Cutright, P. and Wiley, J.A. (1969). “Modernization and Political Representation:
1927–1966,” Studies in Comparative International Development 5: 23–44.
Daalder, Hans (1966). “Parties, Elites and Political Development in Western Europe,” in
Joseph LaPalombara and Myron Weiner (eds), Political Parties and Political Development.
Princeton, NJ: Princeton University Press.
Daalder, Hans (1971). “Cabinets and Party Systems in Ten Smaller European Democracies,”
Acta Politica 6: 282–303.
Dahl, Robert A. (1971). Polyarchy: Participation and Opposition. New Haven, CT: Yale
University Press.
De Swaan, Abram (1973). Coalition Theories and Government Formations: A Study of Formal
Theories of Cabinet Formation Applied to Nine European Parliaments After 1918. Amsterdam:
Elsevier.
Diamond, L. (1990). “Three Paradoxes of Democracy,” Journal of Democracy 1: 48–66.
Diamond, L., ed. (1993). Political Culture and Democracy in Developing Countries. Boulder, CO:
Lynne Rienner.
Diamond, L., Linz, J.J. and Lipset, S.M. (1995). “Introduction: What Makes for
Democracy?” in Larry Diamond, Juan J. Linz and Seymour Martin Lipset (eds), Politics in
Developing Countries. Boulder, CO: Lynne Rienner.
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 305
Diskin, Abraham (2003). The Last Days in Israel: Understanding the New Israeli Democracy.
London and Portland, OR: Frank Cass.
Diskin, Hanna (2001). The Seeds of Triumph: Church and State in Gomulka’s Poland. Budapest
and New York: Central European University Press.
Dodd, L.C. (1976). Coalitions in Parliamentary Government. Princeton, NJ: Princeton
University Press.
Duverger, Maurice (1954). Political Parties: Their Organization and Activity in the Modern State.
New York: Wiley.
Fischer-Galati, Stephen (1992). “Eastern Europe in the Twentieth Century: Old Wine in
New Bottles,” in Joseph Held (ed.), The Columbia History of Eastern Europe in the Twentieth
Century. New York: Columbia University Press.
Gallagher, Michael (1991). “Proportionality, Disproportionality and Electoral Systems,”
Electoral Studies 10: 33–51.
Gasiorowski, Mark J. and Power, Timothy J. (1998). “The Structural Determinants of
Democratic Consolidation: Evidence From the Third World,” Comparative Political Studies
31: 740–71.
Hadenius, Axel (1994). “The Duration of Democracy: Institutional vs. Socio-economic
Factors,” in David Beetham (ed.), Defining and Measuring Democracy. London: Sage.
Hazan, Reuven Y. (1995). “Center Parties and Systemic Polarization: An Exploration of
Recent Trends in Western Europe,” Journal of Theoretical Politics 7: 421–45.
Hazan, Reuven Y. (1997). Centre Parties: Competition and Polarization in European Parliamentary
Democracies. London: Pinter.
Horowitz, Donald L. (1990). “Comparing Democratic Systems,” Journal of Democracy 1: 73–9.
Horowitz, Donald L. (1994). “Democracy in Divided Societies,” in Larry Diamond and
Mark F. Plattner (eds), Nationalism, Ethnic Conflict, and Democracy. Baltimore, MD: Johns
Hopkins University Press.
Hosmer, David W. and Lemeshow, Stanley (1989). Applied Logistic Regression. New York:
Wiley.
Huntington, Samuel (1991). The Third Wave. Norman, OK: University of Oklahoma Press.
Ieraci, Giuseppe (1992). “Centre Parties and Anti-System Oppositions in Polarised
Systems,” West European Politics 15: 17–34.
Inglehart, Ronald (1988). “The Renaissance of Political Culture,” American Political Science
Review 82: 1203–30.
King, G., Alt, J.E., Burns, N.E. and Laver, M. (1990). “A Unified Model of Cabinet
Dissolution in Parliamentary Democracies,” American Journal of Political Science 34:
846–71.
Laakso, Markku and Taagepera, Rein (1979). “‘Effective’ Number of Parties: A Measure
with Application to West Europe,” Comparative Political Studies 12: 3–27.
Lane, Jan-Erik and Ersson, Svante (1987). “Multipartism,” in Manfred J. Holler (ed.), The
Logic of Multiparty Systems. Dordrecht: Martinus Nijhoff.
Lardeyret, Guy (1991). “The Problem with PR,” Journal of Democracy 2: 30–35.
Laver, Michael (1974). “Dynamic Factors in Government Coalition Formation,” European
Journal of Political Research 2: 259–70.
Laver, Michael and Schofield, Norman (1991). Multiparty Government. Oxford: Oxford
University Press.
Laver, Michael and Shepsle, Kenneth A. (1996). Making and Breaking Governments: Cabinets
and Legislatures in Parliamentary Democracies. New York: Cambridge University Press.
Lijphart, Arend (1977). Democracy in Plural Societies: A Comparative Exploration. New Haven,
CT: Yale University Press.
Lijphart, Arend (1984a). Democracies: Patterns of Majoritarian and Consensus Government in
Twenty-One Countries. New Haven, CT: Yale University Press.
Lijphart, Arend (1984b). “Measures of Cabinet Durability: A Conceptual and Empirical
Evaluation,” Comparative Political Studies 17: 265–79.
Lijphart, Arend (1990). “The Southern European Examples of Democratization: Six
Lessons for Latin America,” Government and Opposition 25: 68–84.
306 International Political Science Review 26(3)
Lijphart, Arend (1991). “Constitutional Choices for New Democracies,” Journal of Democracy
2: 72–84.
Lijphart, Arend (1994). Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies,
1945–1990. Oxford: Oxford University Press.
Lijphart, Arend (1995). “The Virtues of Parliamentarism: But which Kind of
Parliamentarism?” in H.E. Chehabi and A. Stepan (eds), Politics, Society, and Democracy:
Comparative Studies. Baltimore, MD: Johns Hopkins University Press.
Lijphart, Arend (1999). Patterns of Democracy: Government Forms and Performance in Thirty-Six
Countries. New Haven, CT: Yale University Press.
Linz, Juan J. (1978). The Breakdown of Democratic Regimes: Europe. Baltimore, MD: Johns
Hopkins University Press.
Linz, Juan J. (1990). “The Perils of Presidentialism,” Journal of Democracy 1: 51–69.
Linz, Juan J. (1994). “Presidential or Parliamentary Democracy: Does It Make a
Difference?” in Juan J. Linz and Arturo Valenzuela (eds), The Failure of Presidential
Democracy. Baltimore, MD: Johns Hopkins University Press.
Linz, Juan J. and Stepan, Alfred (1978). Latin America: The Breakdown of Democratic Regimes.
Baltimore, MD: Johns Hopkins University Press.
Linz, Juan J. and Stepan, Alfred (1996). Problems of Democratic Transition and Consolidation.
Baltimore, MD: Johns Hopkins University Press.
Lipset, S.M. (1959). “Some Social Requisites of Democracy: Economic Development and
Political Legitimacy,” American Political Science Review 53: 69–105.
Lipset, S.M. (1960). Political Man: The Social Bases of Politics. Baltimore, MD: Johns Hopkins
University Press.
Lipset, S.M. (1994). “The Social Requisites of Democracy Revisited,” American Sociological
Review 59: 1–22.
Lipset, S.M., Seong, K-R. and Torres, J.C. (1993). “A Comparative Analysis of the Social
Requisites of Democracy,” International Social Science Journal 45: 155–75.
Lutz, Donald S. (1994). “Toward a Theory of Constitutional Amendment,” American Political
Science Review 88: 365–80.
Mainwaring, Scott (1993). “Presidentialism, Multipartism, and Democracy: The Difficult
Combination,” Comparative Political Studies 26: 198–228.
Merkl, Peter H. (1993). “Which are Today’s Democracies?” International Social Science Journal
45: 257–70.
Midlarsky, Manus I. (1984). “Political Stability of Two-Party and Multiparty Systems:
Probabilistic Bases for the Comparison of Party Systems,” American Political Science Review
78: 929–51.
Muller, Edward N. (1988). “Democracy, Economic Development and Income Inequality,”
American Sociological Review 53: 50–68.
Muller, Edward N. (1995). “Economic Determinants of Democracy,” American Sociological
Review 60: 966–96.
Neumann, Sigmund, ed. (1956). Modern Political Parties: Approaches to Comparative Politics.
Chicago, IL: University of Chicago Press.
Olsen, M.E. (1968). “Multivariate Analysis of National Political Development,” American
Sociological Review 33: 699–712.
Powell, G. Bingham, Jr (1982). Contemporary Democracies: Participation, Stability, and Violence.
Cambridge, MA: Harvard University Press.
Powell, G. Bingham Jr (1986). “Extremist Parties and Political Turmoil: Two Puzzles,”
American Journal of Political Science 30: 357–78.
Power, Timothy J. and Gasiorowski, Mark J. (1997). “Institutional Design and Democratic
Consolidation in the Third World,” Comparative Political Studies 30: 123–55.
Przeworski, A., Alvarez, M., Cheibub, J.A. and Limongi, F. (1996). “What Makes
Democracies Endure?” Journal of Democracy 7: 39–55.
Przeworski, A., Alvarez, M., Cheibub, J.A. and Limongi, F. (2000). Democracy and
Development: Political Institutes and Well-Being in the World, 1950–1990. Cambridge:
Cambridge University Press.
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 307
Putnam, Robert (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton,
NJ: Princeton University Press.
Rokkan, Stein (1970). Citizens, Elections, Parties: Approaches to the Comparative Study of the
Processes of Development. Oslo: Norwegian University Press.
Sanders, D. and Herman, V. (1977). “The Stability and Survival of Governments in Western
Democracies,” Acta Politica 12: 346–77.
Sani, Giacomo and Sartori, Giovanni (1983). “Polarization, Fragmentation and
Competition in Western Democracies,” in Hans Daalder and Peter Mair (eds), Western
European Party Systems: Continuity and Change. London: Sage.
Sartori, Giovanni (1970). “Concept Misformation in Comparative Politics,” American Political
Science Review 64: 1033–53.
Sartori, Giovanni (1976). Parties and Party Systems: A Framework for Analysis. Cambridge:
Cambridge University Press.
Sartori, Giovanni (1994). Comparative Constitutional Engineering: An Inquiry into Structures,
Incentives and Outcomes. New York: New York University Press.
Schumpeter, Joseph (1947). Capitalism, Socialism and Democracy. New York: Harper.
Shugart, Matthew and Carey, John (1992). Presidents and Assemblies: Constitutional Design and
Electoral Dynamics. New York: Cambridge University Press.
Stepan, Alfred and Skach, Cindy (1994). “Presidentialism and Parliamentarism in
Comparative Perspective,” in Juan J. Linz and Arturo Valenzuela (eds), The Failure of
Presidential Democracy. Baltimore, MD: Johns Hopkins University Press.
Strøm, Kaare (1988). “Contending Models of Cabinet Stability,” American Political Science
Review 82: 923–30.
Taylor, M. and Herman, V.H. (1971). “Party Systems and Government Stability,” American
Political Science Review 65: 28–37.
Warwick, Paul (1994). Government Survival in Parliamentary Democracies. New York:
Cambridge University Press.
Weiner, Myron (1987). “Empirical Democratic Theory,” in M. Weiner and E. Ozbudun
(eds), Competitive Elections in Developing Countries. Durham, NC: Duke University Press.
Biographical Notes
ABRAHAM DISKIN is an Associate Professor in, and a former Chair of, the Depart-
ment of Political Science at The Hebrew University of Jerusalem. He is also the
former Chair of the Israeli Political Science Association. He has authored more
than 20 books and 100 articles published in journals such as Electoral Studies,
European Journal of Political Research, Journal of Conflict Resolution, and Journal of
Politics. ADDRESS: Department of Political Science, Hebrew University of Jerusalem,
91905 Jerusalem, Israel [email: mshanna@huji.ac.il].
HANNA DISKIN is the Director General of Maggie Publishers and lectures at the
Department of Political Science at The Hebrew University of Jerusalem. She has
also been affiliated with the Institute of Central/East European Studies and the
Centre for Research on Canadian–Russian Relations, Carleton University. Her
most recent book, The Seeds of Triumph, was given the Best Book in English award
by the Israeli Political Science Association in 2004. ADDRESS: Department of
Political Science, Hebrew University of Jerusalem, 91905 Jerusalem, Israel [email:
mshanna@huji.ac.il].
REUVEN Y. HAZAN is a Senior Lecturer in the Department of Political Science at
The Hebrew University of Jerusalem. His research interests include parties, party
systems, electoral systems, and legislative studies. He has published several books
308 International Political Science Review 26(3)
and many articles on these subjects, most recently “Does Cohesion Equal
Discipline? Towards A Conceptual Delineation” in the Journal of Legislative Studies
during 2003. ADDRESS: Department of Political Science, Hebrew University of
Jerusalem, 91905 Jerusalem, Israel [email: mshazan@huji.ac.il].
Acknowledgments. The authors would like to thank the Authority for Research and
Development at The Hebrew University of Jerusalem for its financial support; the many
colleagues who provided helpful theoretical and empirical comments and assisted in the
selection of cases, particularly Larry Diamond of Stanford University, along with the
anonymous referees of the International Political Science Review; our research assistants,
specifically Etti Ezra and Meira Hanson, for their valuable assistance; and Francine Hazan
for her editing.
DISKIN/DISKIN/HAZAN: Why Democracies Collapse 309
