Regime Type and Political Destabilization in Cross-National Perspective

Abstract

In this article, we re-analyze the hypothesis that the relationship between the type of political regime and its political instability forms an inverted U shape. Following this logic, consistent democracies and autocracies are more stable regimes, whereas intermediate regimes (anocracies) display the lowest levels of political stability. We re-test this hypothesis using a data set that has not been previously used for this purpose, finding sufficient evidence to support the hypothesis pertaining to the aforementioned U-shaped relationship. Our analysis is specifically focused on the symmetry of this U shape, whereby our findings suggest that the U-shaped relationship between regime types and sociopolitical destabilization is typically characterized by an asymmetry, with consistently authoritarian regimes being generally less stable than consolidated democracies. We also find that the character of this asymmetry can change with time. In particular, our re-analysis suggests that U-shaped relationship experienced significant changes after the end of the Cold War. Before the end of the Cold War (1946-1991), the asymmetry of inverted U-shaped relationship was much less pronounced—though during this period consistent authoritarian regimes were already less stable than consolidated democracies, this very difference was only marginally significant. In the period that follows the end of the Cold War (1992-2014), this asymmetry underwent a substantial change: Consolidated democracies became significantly more stable, whereas consolidated autocracies became significantly more unstable. As a result, the asymmetry of the U-shaped relationship has become much more pronounced. The article discusses a number of factors that could account for this change.

Keywords

political regimes sociopolitical destabilization autocracy democracy intermediate regimes the Cold War destabilization indices

Introduction

The inverted U-shaped relationship between the level of political instability and the type of regime was discovered in the 1970s. This direction of research in the field was triggered by systematic data accumulation on the internal conflicts around the world. Thus, Gurr (1974) maintains that semi-democracies are the type of regime most prone to destabilization. His observation was later examined statistically in a number of publications using cross-national data (see the review below). This line of research resulted in the theory of an inverted U-shaped relationship between the regime type and the risks of sociopolitical destabilization. According to the theory, consistent democracies and autocracies are more stable regimes, whereas intermediate regimes (anocracies) display the lowest levels of political stability.

There are several definitions of political instability. The notion can include such phenomena as civil wars, ethnic wars, interstate wars, drastic regime changes to democratization or autocratization,¹ as well as mass demonstrations, political strikes, violent and non-violent protests, and so forth (Regan & Norton, 2005).

Early quantitative studies of instability and regime type were centered on civil wars (Ellingsen & Gleditsch, 1997; Francisco, 1995; Muller & Weede, 1990). They were based on data for a limited number of countries and rather short periods of time. Authors of the studies empirically showed the existence of a statistically significant relationship in the predicted direction: Hybrid regimes turned out to be the most vulnerable to the risk of a civil war. More comprehensive research (Ellingsen, 2000; Hegre, Ellingsen, Gates, & Gledish, 2001) not only confirmed this result but also established the presence of a relationship between regime change in a country’s recent past and political instability. Hegre et al. (2001) argue that young democracies or young autocracies (i.e., countries with recent experience of democratization or autocratization) are much more vulnerable to the emergence of civil war than those regimes that have not changed their type in the recent past. However, the presence of the relation between the start of the civil war and the regime type proved to be statistically significant even when controlled for the impact of the recent regime change (Fearon & Laitin, 2003). The relation between democracy and the onset of civil war was explored in detail in several scientific works (K. S. Gleditsch & Hegre, 2014; N. P. Gleditsch, Hegre, & Strand, 2009). Partial democracies were found to be the most prone to civil wars, but the authors do not exclusively attribute this fact to political instability as the U curve is robust when controlled for this variable. At the same time, fewer conflicts in democracies can be explained by the fact that they tend to be high-income countries with stable institutions. If these factors are controlled, authors find no linear relationship between democracy and the risk of conflict.

Research into the impact of democratization and autocratization on the countries’ participation in war² (Mansfield & Snyder, 1995) also revealed that democratizing countries are more than 60% frequently involved in wars than countries with no regime change. Moreover, the countries undergoing democratization had a higher risk of being involved in a war than all other states (including the autocratizing ones).

Some other works (Cederman, Hug, & Krebs, 2010) confirm the specific impact of democratization on this extreme type of political instability.³ However, there are also papers (e.g., K. S. Gleditsch, 2002) contending that the direction of the regime change does not influence the risk of civil war.

When democratization and autocratization are viewed as independent variables, one of the key additional factors influencing the level of sociopolitical destabilization is political factionalism (Goldstone et al., 2010; Marshall & Cole, 2008, 2012) as presented in the Polity database (Marshall & Jaggers, 2016).

The relation not only between civil wars but also ethnic civil wars and regime type is investigated and supported in a number of other papers (Buhaug, 2006; Sambanis, 2001).

The theoretic explanation of the causes underlying the weakness of intermediate regimes is twofold. On one hand, there is the theory explaining the emergence of political violence as a whole, namely, the relative deprivation theory focused on economic factors. On the other hand, there is the theory of rational agents or opportunities centered around political drivers. According to the first theory, based on Davies (1962, 1969) and Gurr (1970), rebellion and civil war can be caused by expectations raised by modernization left unsatisfied, motivating people to take to the streets in protest. Such a situation is most frequently observed in intermediate regimes undergoing intensive modernization. According to the second theory, the instability of intermediate regimes is based on the weakness of their central power, which gives the rebels greater hope for profiting more from a rebellion, rather than from inaction or legal political struggle (Muller & Weede, 1990). Some authors (Gates, Hegre, Jones, & Strand, 2006; Regan & Norton, 2005) combine the two theories by explaining the higher destabilization risk of the intermediate regimes by the fact that they are too weak to suppress all possibilities of a rebellion, and not democratic enough to prevent disappointment. As for the more particular factors stipulating the instability of a political regime, they include a recent regime change, drastic change of regime type by a large number of points toward democratization or autocratization, and autocratization itself as compared with democratization (Regan & Bell, 2009).

An accurate definition for the type of a political regime is one of the primary questions when investigating the relation between regime types and instability. Generally, there are two approaches to the problem.

The first approach maintains that there exist strictly defined parameters allowing to draw a line between autocracy and democracy; thus, the sole reason for our inability to define the type of a certain regime is the limitations in our knowledge about it (Przeworski, Alvarez, Cheibub, & Limongi, 2000). This very perspective narrows our view on regimes to a dichotomy, whereby intermediate or hybrid regimes are absent.

On the contrary, the second viewpoint states that “pure” democracies and autocracies are highly uncommon, and the majority of regimes lie between these two extreme points (Dahl, 1971). Furthermore, theories may differ in the number of criteria used for differentiating between “pure” and “hybrid” regimes.

Relatively early papers (Muller & Weede, 1990) used regime type as defined in Freedom House for their measure of democracy, whereas the indicator of instability (namely civil war) was taken from the third issue of World Handbook of Political and Social Indicators (Taylor & Jodice, 1983). However, this research only included data for 4 years, 1973 to 1977. Nevertheless, even this material supported the presence of the inverted U-shaped relationship between the regime type and sociopolitical destabilization. Freedom House data have been criticized for being both biased and methodologically flawed. Vreeland (2008) pointed out that the Freedom House Index cannot be taken as an independent measure of political instability as the questionnaire underlying one of its components, namely, the Index of Civil Freedoms, has a direct reference to the presence or absence of a civil war or a rebellion (the question is as follows: Is there freedom from war and insurgencies?).

Quite frequently just one scale criterion is used for defining the various types, most often the “Democracy Index”⁴ or “Polity Index” in the Polity database (Hegre et al., 2001; Jaggers & Gurr, 1995). The use of the Democracy Index allows one to identify intermediate types of regimes (anocracies), but does not allow to understand what factors influence the regime stability (Hegre, 2014).

The usage of the Democracy Index (Polity) as an indicator of the level of democracy has been rather heavily criticized. Thus, Vreeland (2008) considers the presence of the inverted U-shaped relationship to be rather a result of the specific features of coding the intermediate regimes (anocracies) in Polity, when countries affected by civil wars are coded as intermediate regimes. To demonstrate this viewpoint, Vreeland excludes from the aggregation the indicators related, in his opinion, to civil war (namely, PARCOMP and PARREG), and, using data by Fearon and Laitin (2003) and Hegre et al. (2001) shows that the “regime type” parameter turns insignificant in this case. Other works (Treier & Jackman, 2008) criticize the way the democracy indicators are coded in Polity, the interdependence of the five indicators, the non-homogeneity of the ordinal scale, and other inaccuracies of coding which, as the authors show, question the possibility of using the Polity indicators without serious modifications.

Vreeland’s conclusions have been tested in several models. Thus, when parameters directly related to instability were added to Goldstone’s model, this did not result in either increased significance of the “regime type” variable or increased accuracy of the model (Goldstone et al., 2010). This means that “regime type” variable is not an indicator of political violence. In intermediate regimes, the variable indicates a special type of inter-elite relations characterized by extreme polarization.

Albeit there is a possibility of using the Democracy Index (Polity) as the indicator of democracy level, some researchers point out the necessity of taking into account Polity’s categorical scale rather than the ordinal one, all the while indicating a very high weight of one component out of five (namely, Xconst) in regime classification (Ellingsen & Gleditsch, 1997).

Intermediate regimes are highly heterogeneous and can differ greatly from one another; these differences are lost when the regimes are ranked using a single indicator. Certain authors prefer to use several components of the Democracy Index, or even some other indicators included in Polity, to operationalize the notion of democracy. Thus, some authors (e.g., Gates et al., 2006) use only two or three indicators of five, or use a completely different indicator, such as the openness of executive recruitment (EXREC; Goldstone et al., 2010; Ulfelder & Lustik, 2007).

The presence of inverted U-shaped relationship has been acknowledged by a number of studies including those undertaken by Russian researchers. Their research focused on the quantitative and qualitative analysis of the Arab Spring events has demonstrated in various ways (including the application of formal mathematical models) that inconsistently authoritarian regimes turned to be the most unstable; what is more in the multiple regression analyses performed by these authors, the intermediate type of the political regime has consistently turned out to be one of the strongest destabilization predictors (Grinin & Korotayev, 2012a, 2012b, 2013, 2014a, 2014b; Korotayev, Issaev, Malkov, & Shishkina, 2013, 2014; Malkov, Yu, Korotaev, Isaev, & Kuzminova, 2013; Tsirel, 2012a, 2012b). In addition, they have demonstrated that the intermediate type of political regime was a significant predictor of political destabilization during the revolutionary wave of 2013-2014 (Korotayev, Issaev, & Vasiliev, 2015; Korotayev, Issaev, & Zinkina, 2015).

In this article, we re-test the presence of inverted U-shaped relationship between the regime type and the degree of sociopolitical destabilization of respective states using a data set that has not been used earlier for this purpose, and find that the U-shaped relationship hypothesis holds true. We also test U shape for symmetry, and find that the U shape is highly asymmetrical.

Let us note that there are certain theoretical grounds to expect that the character of this asymmetry should change with time, especially after the end of the Cold War.

First and foremost, after the end of the Cold War and the collapse of the Union of Soviet Socialist Republics (USSR), one should expect a visible decrease in the level of sociopolitical instability in the democratic states of the World System core (note that the majority of the World System core countries are coded as full democracies, and, vice versa, the majority of such countries belong to the World System core).

Note that after the end of the Cold War, leftist movements became much less active in the core countries. Before 1991, it was largely the leftist organizations who drove people to the streets to achieve, say, improved working conditions, who organized political strikes and other disruptive activities along the lines of the same logic. Besides, in the 1990s, the Western countries (most of which are consolidated democracies) started to actively shift the industrial production away to the global semi-periphery and periphery. This resulted in lower protest activity driven by a decrease in the number of industrial workers, who used to be the most active participants of the leftist protest actions (the famous strikes of the British miners could not but cease due to a near disappearance of miners in this country). A notable reason underlying the lower level of sociopolitical instability in Western democracies was their aging population in the 1990s and the 2000s (see, for example, Goldstone, Kaufmann, & Toft, 2012). It is widely known that young people are most attracted to radical action, and a pronounced increase in the absolute number and relative proportion of younger population is an impetus to the rise in the level of sociopolitical destabilization within a society (see, for example, Fuller, 2004; Goldstone, 1991, 2002; Heinsohn, 2003; Korotayev, 2014; Korotayev, Malkov, & Grinin, 2014; Korotayev & Zinkina, 2011; Korotayev et al., 2011; Mesquida & Weiner, 1999; Moller, 1968). Thus, one would expect that the decrease of the proportion of youth in the adult population would tend to decrease the general degree of sociopolitical instability.

Second, we can assume that the end of the Cold War and the collapse of the USSR led to a remarkable increase in the instability of authoritarian regimes. Let us recall the words allegedly ascribed to Franklin D. Roosevelt: “Somoza may be a son of a bitch, but he’s our son of a bitch.” This phrase, though supposedly pronounced initially about 1939, really caught on after 1945, when the Cold War climaxed (see, for example, Schmitz, 1999). This phrase reflects an important circumstance. During the Cold War, it was not only the Societ Bloc who was not interested in the collapse of pro-Soviet authoritarian regimes. The opposite bloc headed by the United States was also afraid of the collapse of pro-U.S. authoritarian regimes, because such a collapse could pave the way to power for pro-Soviet forces. As a result, both the USSR and the United States served as warrantors of the stability of authoritarian regimes in their spheres of influence.

The collapse of the USSR and the end of the Cold War changed the situation dramatically. The Soviet sphere of influence no longer had a pillar of stability for authoritarian regimes. Meanwhile, the U.S. sphere of interest no longer feared that a collapse of an authoritarian regime would transfer power to pro-communist forces. As a result, for example, the end of the Soviet support for the USSR-client regimes in Afghanistan or Ethiopia resulted in their collapse as early as 1992, whereas the stop of the U.S. support to the military dictatorship, for example, in Guatemela resulted in its failure in the 1990s.

What is more, the United States started using various democratization programs to strengthen its influence (see, for example, Dalacoura, 2005). “Democracy promotion has been an explicit doctrine of U.S. foreign policy since the end of the cold war. Between 1990 and 2003 resources for democracy programs increased by over 500 percent” (Finkel, Pérez-Liñán, & Seligson, 2007⁵). Apart from that, as Finkel et al. (2007) showed, the post-cold war U.S. democracy promotion programs were quite an effective tool of destabilization of authoritarian regimes.

Thus, there are significant grounds to expect that the end of the Cold War brought about a remarkable decrease of the level of sociopolitical destabilization in states classified as full democracies by Goldstone et al. (2010), and a remarkable increase of this indicator in the states classified as full autocracies.

In the final part of the tests section of this article, we test this hypothesis.

Materials and Method

To test the relationship between the type of political regime and instability, we use Polity IV data for the regime type (Marshall & Jaggers, 2016) and A. Banks’s Cross National Time Series (CNTS) database (Banks & Wilson, 2016) as the data source for the dependent variable (level of instability). The CNTS database was chosen mainly for two reasons: First, it contains the different types of domestic instability that the research is focused on, both violent and non-violent instability. Second, unlike other databases, for example, Nonviolent and Violent Campaigns and Outcomes (NAVCO), CNTS is updated yearly and contains instability data for a sufficiently extensive period of time. The version used for the purposes of this research was published in 2015 and contains political destabilization data spanning from 1919 up to 2014.

The CNTS database is a result of data compilation and systematization started by Arthur Banks (Banks & Wilson, 2016) in 1968 in the State University of New York–Binghamton. The work was based on generalizing the archive of data from The Statesman’s Yearbooks, published since 1864. It also contains approximately 200 indicators for more than 200 countries. The database contains yearly values of indicators starting from 1815 excluding the periods of World Wars I and II (1914-1918 and 1939-1945).

CNTS database is structured by sections, such as territory and population, technology, economic and electoral data, internal conflicts, energy use, industry, military expenditures, international trade, urbanization, education, employment, legislative activity.

In our article, we take a close look at the data describing internal conflicts (domestic). This section includes data starting from 1919 based on the analysis of events in eight various subcategories, which are used to compile the general index of sociopolitical destabilization (domestic9). In building the general index, the compilers of CNTS database give each category a certain weight (see Table 1).

Table 1.

Weights of Subcategories Used in Compiling the Index of Sociopolitical Destabilization.

Subcategory	Variable name	Weight in the index of sociopolitical destabilization (domestic9)
Assassinations	domestic1	25
General strikes	domestic2	20
Guerrilla warfare	domestic3	100
Government crises	domestic4	20
Purges	domestic5	20
Riots	domestic6	25
Revolutions	domestic7	150
Anti-government demonstrations	domestic8	10

To calculate the index of sociopolitical destabilization (Weighted Conflict Measure, domestic9), the numerical values of each subcategory are multiplied by their corresponding weights, the results of the multiplications are summed up, then the sum is multiplied by 100 and divided by 8 (see Equation 1).

d o m e s t i c 9 = \frac{\begin{array}{l} 25 d o m e s t i c 1 + 20 d o m e s t i c 2 + 100 d o m e s t i c 3 + 20 d o m e s t i c 4 + \\ 20 d o m e s t i c 5 + 25 d o m e s t i c 6 + 150 d o m e s t i c 7 + 10 d o m e s t i c 8 \end{array}}{8} \times 100 .

Our definition of the independent variable—the type of political regime—is based on the work by Goldstone and his colleagues (2010). They use Polity’s scale for the openness of executive recruitment (EXREC) as a measure of contestation and Polity’s scale of the competitiveness of political participation (PARCOMP) to capture variation in the degree and forms of inclusiveness. This measure is free from the shortcomings of linear regime type measures and uses two dimensions that have been long considered as the most important characteristics of the modern forms of government (Dahl, 1971). The classification includes five categories ranging from full autocracies to full democracies. The three intermediate categories fall into partial autocracies and partial democracies, further subcategorized as follows: partial autocracies, which hold competitive elections for national office but repress or tightly control participation or allow substantial political participation but fail to subject the office of chief executive to truly competitive elections; partial democracies, systems in which the chief executive is chosen through competitive elections and political competition is not effectively repressed, but either elections are not fully free and fair or political participation is not fully open and well institutionalized. Among partial democracies, Goldstone et al. further distinguish between those that exhibit factionalism, as coded on the PARCOMP variable, and those that do not.⁶

We also use Freedom House (2016) database for a robustness test to check whether our results hold when using slightly different definitions of regime types. Depending on the degree to which regime is democratic, all political regimes found in Freedom House database are divided into “free,” “partly free,” and “not free,” where “free” states generally correspond to democracies, “not free” states do to autocracies, and “partly free” states correspond to “intermediate regimes” (Puddington et al., 2015). Their full methodology can be viewed at https://freedomhouse.org/report/freedom-world-2015/methodology.

Tests

To test the presence of the U-curve relationship on CNTS and Polity IV data, we use a shortened version of the scale proposed by Goldstone et al. and combine three intermediate regimes into one.

A one-way ANOVA shows that the effect of regime type is significant, F(2,974) = 31.24, p < .001. Post hoc analyses using the Scheffé post hoc criterion for significance indicate that the level of destabilization is significantly higher in intermediate regimes (anocracies; M = 1,409, SD = 5,775) than in other types of regimes (p < .001); democracies (M = 540, SD = 1,457) are significantly more stable than anocracies and autocracies (p < .008 for autocracies and p < .001 for intermediate regimes), autocracies (M = 933, SD = 2,806) occupy an intermediate position being significantly more stable than anocracies and significantly less stable than democracies (p < .001).⁷ Thus, a direct test of the hypothesis has verified the presence of the inverted U-shaped relationship in question (see Tables 2 and 3).

Table 2.

Mean Values of the Index of Sociopolitical Destabilization for Intermediate Regimes as Compared With Countries With Autocratic and Democratic Regime Types, 1946-2014.

Regime types	n	M	SD	F value	p
Autocracies	2,736	934	2,806	31.274	.001
Intermediate regimes	4,861	1,409	5,775
Democracies	2,144	541	1,457

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

Table 3.

Results of Scheffé’s Post Hoc Comparison of the Index of Sociopolitical Destabilization for Intermediate Regimes as Compared With Countries With Autocratic and Democratic Types of Political Regimes, 1946-2014.

Regime types	M difference	SE	p
Autocracies
Intermediate regimes	−475.35	105.05	<.001
Democracies	393.09	126.78	.008
Intermediate regimes
Autocracies	475.35	105.05	<.001
Democracies	868.44	113.95	<.001
Democracies
Autocracies	−393.09	126.78	.008
Intermediate regimes	−868.44	113.95	<.001

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

What appears to be the most important point is that the analysis reveals a pronounced asymmetry of the inverted U-shaped relationship, in that the authoritarian regimes turn out to be much more prone to sociopolitical destabilization than the democratic ones.

Thus, already the very first series of tests reveals a marked asymmetry of the inverted U-shaped relationship.

Separate testing of the relevant correlations for the periods before and after the end of the Cold War produces the following results:

In the period up to the end of the Cold War, the inverted U-shaped relationship is very pronounced, F(2,491) = 56.41, p < .001. The mean level of political destabilization in intermediate regimes (M = 1,451, SD = 2,171) is significantly higher than in democracies and autocracies. Statistically significant differences from intermediate regimes in the level of instability are observed both for autocratic and for democratic regimes (p < .001). The mean level of sociopolitical instability in autocracies is somewhat higher than that in democracies, but this difference is only marginally significant (p = .064; see Tables 4 and 5).

For the period after the end of the Cold War, our empirical tests support both our hypotheses, indicating the increased stability of democratic regimes as well as increased instability of autocratic polities. A decrease in the level of sociopolitical instability in consolidated democratic polities is particularly strong (by 48%) and significant, t(1825) = 4.748, p < .001 (see Table 6, row 1).

Furthermore, the hypothesis of an increase in the level of sociopolitical instability in autocratic regimes after the end of the Cold War also finds empirical support. Indeed, the mean yearly destabilization index here grew by 27%, and this growth is statistically significant (p = .023; see Table 6, row 2). As for the intermediate polities, the level of sociopolitical destabilization generally remained virtually the same after the end of the Cold War and did not demonstrate any statistically significant change (see Table 6, row 3).

As a result, during the period after the Cold War, the level of instability in autocracies has approached the level of instability in intermediate regimes to such an extent that the difference between these two types of regimes with respect to the average annual level of sociopolitical destabilization has become statistically insignificant (see Tables 7 and 8). The difference between autocratic and democratic regimes is significant at one tail (p = .05), whereas intermediate regimes demonstrate an unequivocally significantly higher level of sociopolitical instability than democracies do (p = .002).

Table 4.

Mean Values of the Index of Sociopolitical Destabilization for the Main Regime Types in 1946-1991.

Regime types	n	M	SD	F value	p
Autocracies	2,104	891	1,995	56.416	.001
Intermediate regimes	1,777	1,451	2,172
Democracies	1,033	711	1,755

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

Note. For example, n = 2,104 for autocracies for the 1946-1991 period means that for this period, we have 2,104 yearly observations of autocratic regimes.

Table 5.

Results of Scheffé’s Post Hoc Comparison of the Index of Sociopolitical Destabilization for the Main Regime Types in 1946-1991.

Regime types	M difference	SE	p
Autocracies
Intermediate regimes	−560.42	64.89	<.001
Democracies	179.51	76.52	.064
Intermediate regimes
Autocracies	560.42	64.89	<.001
Democracies	739.93	78.80	<.001
Democracies
Autocracies	−179.51	76.52	.064
Intermediate regimes	−739.93	78.80	<.001

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

Table 6.

Mean Values of the Index of Sociopolitical Destabilization for Different Types of Political Regimes Before and After the End of the Cold War.

	Period	N	M	SD	T	p	Comments
1	1946-1991	1,033	711	1,755	4.748	.001	Mean values of the index of sociopolitical destabilization for democratic polities
1	1992-2014	794	371	1,137	4.748	.001
2	1946-1991	2,104	891	1,995	−2.273	.023	Mean values of the index of sociopolitical destabilization for autocratic polities
2	1992-2014	406	1,246	5,563	−2.273	.023
3	1946-1991	1,777	1,451	2,172	0.175	.6	Mean values of the index of sociopolitical destabilization for intermediate polities
3	1992-2014	2,401	1,350	7,875	0.175	.6

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

Table 7.

Mean Values of the Index of Sociopolitical Destabilization According to the Regime Type in 1992-2014.

Regime types	n	M	SD	F value	p
Autocracies	406	1,246	5,562	6.416	.002
Intermediate regimes	2,401	1,350	7,875
Democracies	794	371	1,136

Source. Marshall and Jaggers (2016) and Banks and Wilson (2016; domestic9).

Table 8.

Results of Scheffé’s Post Hoc Comparison of the Index of Sociopolitical Destabilization According to the Type of Political Regime in 1992-2014.

	M difference	SE	p
Autocracies
Intermediate regimes	−103.98	360.52	.959
Democracies	875.32	409.91	.102
Intermediate regimes
Autocracies	103.98	360.52	.959
Democracies	979.30	275.04	.002
Democracies
Autocracies	−875.32	409.91	.102
Intermediate regimes	−979.30	275.04	.002

We also use the data of Freedom House (2016) to test whether our hypotheses hold up with a different source of data on regime types. All conclusions find empirical support using Freedom House data: Autocracies (“not free” states) tend to be less stable than democracies (“free” states; p < .001) and more stable than intermediate regimes (“partly free” states; p = .003) if we consider the whole period spanning from 1973 (the year when Freedom House started publishing its database) to 2014 (see Table 9, row 1, and Table 10, row 1). An almost “ideal” U curve with no significant difference between autocracies and democracies (p = .223) exists before the end of the Cold War (1973-1991; see Table 9, row 2, and Table 10, row 2). After the Cold War, autocracies become so prone to political instability that the index of sociopolitical destabilization in autocracies becomes almost equal to the one of intermediate regimes (see Table 9, row 3, and Table 10, row 3).

Table 9.

Mean Value of the Index of Sociopolitical Destabilization for the Main Types of Political Regime in 1973-2014 Based on Freedom House Data.

	Time period	Regime type	n	M	SD	F value	p
1	1973-2014	Autocracies	2,066	912	1,754	106.597	.001
		Intermediate regimes	2,115	1,088	2,019
		Democracies	2,696	422	1,184
2	1973-1991	Autocracies	1,096	808	1,466	18.363	.001
		Intermediate regimes	916	1,128	1,896
		Democracies	985	682	1,596
3	1992-2014	Autocracies	970	1,030	2,024	105.773	.001
		Intermediate regimes	1,199	1,058	2,109
		Democracies	1,711	273	825

Table 10.

Results of Scheffé’s Post Hoc Comparison of the Index of Sociopolitical Destabilization for the Main Types of Political Regime in 1973-2014 Based on Freedom House Data.

	Time period	Regime type	M difference	SE	p
1	1973-2014	Autocracies
		Intermediate regimes	−176.39	51.08	.003
		Democracies	489.44	48.28	<.001
		Intermediate regimes
		Autocracies	176.39	51.08	.003
		Democracies	665.83	47.96	<.001
		Democracies
		Autocracies	−489.44	48.28	<.001
		Intermediate regimes	−665.83	47.96	<.001
2	1973-1991	Autocracies
		Intermediate regimes	−320.51	73.86	<.001
		Democracies	125.50	72.43	.223
		Intermediate regimes
		Autocracies	320.51	73.86	<.001
		Democracies	446.01	75.73	<.001
		Democracies
		Autocracies	−125.50	72.43	.223
		Intermediate regimes	−446.01	75.73	<.001
3	1992-2014	Autocracies
		Intermediate regimes	−28.15	70.94	.924
		Democracies	756.71	66.02	<.001
		Intermediate regimes
		Autocracies	28.15	70.94	.924
		Democracies	784.87	61.87	<.001
		Democracies
		Autocracies	−756.71	66.02	<.001
		Intermediate regimes	−784.87	61.87	<.001

Discussion

Thus, CNTS and Freedom House data seemingly attest to the disappearance of the inverted U-shaped relationship after the end of the Cold War. However, we argue such an assumption to be an oversimplification, because a more detailed analysis presented below suggests that the inverted U-shaped relationship has not disappeared altogether—it has rather experienced a certain reconfiguration.

To get a more nuanced vision of the investigated relationship, one has to return to the full classification by Goldstone et al. As mentioned earlier, in (Goldstone et al., 2010) polities are classified based on two dimensions EXREC (openness of executive recruitment) and PARCOMP (competitiveness of political participation) of Polity IV into five types ranging from full autocracies to full democracies. The resulting index can be viewed as what we call “Adjusted Democracy Index” (ADI) for the purposes of the study.

Let us now view the correlation between the ADI and the level of sociopolitical instability for the period before the end of the Cold War (1946-1991; see Figure 1).

Figure 1.

Correlation between the ADI and the mean values of the destabilization index for the period before the end of the Cold War (1946-1991).

First, we see that for the period before the end of the Cold War, the consistent authoritarian regimes (M = 891, SD = 1,995) are almost as stable as consolidated democracies (M = 711, SD = 1,755; p = .235). Partial autocracies (M = 1,162, SD = 1,882), though more unstable than consolidated regimes, demonstrate a visibly lower degree of instability than partial democracies with factionalism (M = 1,953, SD = 2,460; p < .001) and only marginally lower instability than partial democracies without factionalism (M = 1,502, SD = 2,297; p = .088). The inverted U curve demonstrates a rather slight asymmetry in this period (see Figure 2).

Figure 2.

Correlation between the ADI and the mean values of the destabilization index for the period before the end of the Cold War (1946-1991), ANOVA.

After the end of the Cold War, the situation changes (see Figure 3).

Figure 3.

Correlation between the ADI and the mean values of the destabilization index in the periods before (1946-1991) and after (1992-2014) the end of the Cold War.

Partial democracies with factionalism remain the most unstable regime during this period as well. At the same time, there is a notable decline in the level of instability for consolidated democratic polities, t(1825) = 4.748, p < .001, and partial democracies, t(1399) = 5.648, p < .001. The decline in destabilization in democratic regimes is accompanied by a significant increase in the mean values of sociopolitical destabilization in the full authoritarian regimes (t = −2.273, p = .023).

After the end of the Cold War, the inverted U-shaped relationship changes and the most changes happen at both ends of the scale with fully consolidated regimes. The U-curve relationship demonstrates a more pronounced asymmetry whereby full autocracies (M = 1,246, SD = 5,563) become as unstable as partial autocracies, though still significantly more stable than the least stable regime type, that is, partial democracies with factionalism (M = 2,485, SD = 14,206; p = .091) and democracies remarkably gain in stability (see Figure 4).

Figure 4.

Correlation between the ADI and mean values of the destabilization index for the period after the end of the Cold War (1992-2014), ANOVA.

Summary and Conclusion

Thus, our empirical tests using the CNTS and Polity IV databases have generally confirmed the presence of the inverted U-shaped relationship between type of political regime and the level of sociopolitical destabilization therein. In addition, our re-analysis has revealed a number of important findings:

Inverted U-shaped relationship between regime type and the level of sociopolitical destabilization typically is characterized by an asymmetry.

The character of this asymmetry can change with time.

The character of the inverted U-shaped relationship experienced significant changes after the end of the Cold War.

The partial democracies with factionalism were the most unstable regimes both before and after the Cold War. However, in the period before the end of the Cold War (1973-1991), consolidated democracies were only slightly more stable than full autocracies, whereas partial democracies both with and without factionalism were less stable than partial autocracies.

After the end of the Cold War, the instability of full autocracies increased significantly, whereas both full democracies and partial democracies without factionalism became substantially more stable. As a result, in this period, the level of sociopolitical instability in both full and partial autocracies is significantly higher than in both full and partial democracies without factionalism, whereby the difference between the full autocracies and the full democracies becomes unequivocally pronounced and undeniably significant.

We suggest that the autocracies become more prone to the destabilization as the relationship between democratic and autocratic countries changes: After the Cold War, democratic powers were no more inclined to ally with autocracies and support the status quo as the common enemy had disappeared. Not only the level of instability changes with time, the number of regimes undergo changes as well: Autocracies become less numerous and democracies and intermediate regimes grow in number. Future research can be focused on the evolution of autocratic regimes and understanding the influence that the end of the Cold War had on the democratization of autocratic regimes. Analyzing the data, we have also noticed that democracies and autocracies differ not only in “quantity” of instability but also in the “quality” of destabilizing events, that is, each regime has types of instability that are proper to it. Democracies tend to be more prone to demonstrations and political strikes, whereas autocracies are more often shaken by guerrilla warfare, terrorist attacks, and coups; so, the difference in destabilization types for different regimes also constitutes a promising path for future research.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article is an output of a research project implemented as part of the Basic Research Program at the National Research University Higher School of Economics (HSE) in 2016 with support by Russian Science Foundation (Project No. 14-11-00634).

Notes

Author Biographies

Elena Slinko, PhD, is head of Research Centre of the International News Agency “Russia Today.” Her interests include the comparative studies of the value systems and their roles in providing integration and political stability, as well as forecasting of instability processes.

Stanislav Bilyuga is a PhD student at the Faculty of Global Studies of the M. V. Lomonosov Moscow State University and a junior researcher at the Center for Long-Term Forecasting and Strategic Planning of the Moscow State University. His interests include the modeling of global trends and dynamics and forecasts of destabilization processes.

Julia Zinkina is a research professor at the Laboratory for Sociopolitical Destabilization Risks Monitoring, National Research University Higher School of Economics. She is an author of more than 100 scholarly publications, including articles such as “Kondratieff Waves in Global Invention Activity” (Technological Forecasting & Social Change, 2011, 78, pp. 1280-1284); “The Origins of Dragon-Kings and Their Occurrence in Society” (Physica A, 2012, 391(21), pp. 5215-5229); “Urbanization Dynamics in Egypt: Factors, Trends, Perspectives” (Arab Studies Quarterly, 2013, 35(1), pp. 20-38); “Measuring Globalization: Existing Methods and Their Implications for Teaching Global Studies and Forecasting” (Campus-Wide Information Systems, 2013, 30(5), pp. 11-25); “Projecting Mozambique’s Demographic Futures” (Journal of Futures Studies, December 2014, 19(2), pp. 21-40); “Explosive Population Growth in Tropical Africa: Crucial Omission in Development Forecasts” (World Futures, 2014, 70(2), pp. 120-139); and “On the Structure of the Present-Day Convergence” (Campus-Wide Information Systems, 2014, 31(2-3), pp. 139-152). Her sphere of academic interests includes political demography, global history, and global networks, as well as social processes in Tropical Africa and Middle East.

Andrey Korotayev is the head of the Laboratory for Monitoring of the Risks of Sociopolitical Destabilization of the National Research University Higher School of Economics, as well as a senior research professor at the Institute of Oriental Studies of the Russian Academy of Sciences. Together with Askar Akaev, he is a coordinator of the Russian Academy of Sciences Program “System Analysis and Mathematical Modeling of World Dynamics.” His main research interests include cross-cultural research, quantitative analysis of sociopolitical destabilization and its factors, as well as mathematical modeling of the World System development. He is an author of more than 200 academic articles and 20 monographs, including the recently published Great Divergence and Great Convergence. A Global Perspective (New York, NY: Springer, 2015; with Leonid Grinin). His monograph “Introduction to Social Macrodynamics” (Moscow: URSS, 2006; in collaboration with Artemy Malkov and Daria Khaltourina) has brought him an award of the Russian Science Support Foundation in “The Best Economists of the Russian Academy of Sciences” nomination. In 2012, he was awarded with the Gold Kondratieff Medal by the International N. D. Kondratieff Foundation.

References

Banks

A. S.

Wilson

K. A.

(2016). Cross-national time-series data archive. Databanks international. Jerusalem, Israel. Available from http://www.databanksinternational.com

Buhaug

(2006). Relative capability and rebel objective in civil war. Journal of Peace Research, 43, 691-708.

Cederman

L.-E.

Hug

Krebs

L. F.

(2010). Democratization and civil war. Journal of Peace Research, 47, 377-394.

Dahl

R. A.

(1971). Polyarchy: Participation and opposition. New Haven, CT: Yale University Press.

Dalacoura

(2005). US democracy promotion in the Arab Middle East since 11 September 2001: A critique. International Affairs, 81, 963-979.

Davies

(1962). Toward a theory of revolution. American Sociological Review, 27(1), 5-19.

Davies

(1969). Toward a theory of revolution. In McLaughlin

(Ed.), Studies in social movements. A social psychological perspective (pp. 85-108). New York, NY: Free Press.

Ellingsen

(2000). Colorful community or ethnic witches’ brew? Multiethnicity and domestic conflict during and after the Cold War. Journal of Conflict Resolution, 44, 228-249.

Ellingsen

Gleditsch

N. P.

(1997). Democracy and armed conflict in the third world. In Volder

Smith

(Eds.), Causes of conflict in the third world (pp. 69-81). Oslo, Norway: North/South Coalition and International Peace Research Institute.

10.

Fearon

Laitin

(2003). Ethnicity, insurgency, and civil war. American Political Science Review, 97, 75-90.

11.

Finkel

S. E.

Pérez-Liñán

Seligson

M. A.

(2007). The effects of US foreign assistance on democracy building, 1990-2003. World Politics, 59, 404-439.

12.

Francisco

R. A.

(1995). The relationship between coercion and protest: An empirical evaluation in three coercive states. Journal of Conflict Resolution, 39, 263-282.

13.

Freedom House. (2016). Individual country ratings and status, FIW 1973-2015. Retrieved from https://freedomhouse.org/sites/default/files/Individual%20Country%20Ratings%20and%20Status%2C%201973-2015%20%28FINAL%29.xls

14.

Fuller

G. E.

(2004). The youth crisis in middle eastern society. Clinton Township, MI: Institute for Social Policy and Understanding.

15.

Gates

Hegre

Jones

M. P.

Strand

(2006). Institutional consistency and political instability. American Journal of Political Science, 50, 893-908.

16.

Gleditsch

K. S.

(2002). All international politics is local: The diffusion of conflict, integration, and democratization. Ann Arbor: University of Michigan Press.

17.

Gleditsch

K. S.

Hegre

(2014). Regime type and political transition in civil war. In DeRouen

Newman

(Eds.), Routledge handbook of civil war (pp. 145-156). London, England: Routledge.

18.

Gleditsch

N. P.

Hegre

Strand

(2009). Democracy and civil war. In Midlarsky

(Ed.), Handbook of war studies III (pp. 155-192). Ann Arbor: University of Michigan Press.

19.

Goldstone

J. A.

(1991). Revolution and rebellion in the early modern world. Berkeley: University of California Press.

20.

Goldstone

J. A.

(2002). Population and security: How demographic change can lead to violent conflict. Journal of International Affair, 56, 3-21.

21.

Goldstone

J. A.

Bates

R. H.

Epstein

D. L.

Gurr

T. R.

Marshall

M. G.

Lustik

M. B.

. . . Ulfelder

(2010). A global model for forecasting political instability. American Journal of Political Science, 54, 190-208.

22.

Goldstone

J. A.

Kaufmann

E. P.

Toft

M. D.

(Eds.). (2012). Political demography. Oxford, UK: Oxford University Press.

23.

Grinin

Korotayev

(2012a). Does “Arab Spring” mean the beginning of World System reconfiguration? World Futures, 68, 471-505.

24.

Grinin

Korotayev

(2012b). Tsikly, krizisy, lovushki sovremennoj Mir-Sistemy. Issledovanie kondrat’evskikh, zhyuglyarovskikh i vekovykh tsiklov, global’nykh krizisov, mal’tuzianskikh i postmal’tuzianskikh lovushek [Cycles, crises, traps of the modern World System. Study of the Kondratieff, Juglar and secular cycles, global crises, Malthusian and post-Malthusian traps]. Moscow, Russia: LKI/URSS.

25.

Grinin

Korotayev

(2013). Demokratiya i revolyutsiya [Democracy and revolution]. Istoriya i sovremennost’, 2(18), 15-35.

26.

Grinin

Korotayev

(2014a). Revolution and democracy in the context of the globalization. In Kiss

(Ed.), The dialectics of modernity—Recognizing globalization. Studies on the theoretical perspectives of globalization (pp. 119-140). Budapest, Hungary: Arisztotelész Kiadó.

27.

Grinin

L. E.

Korotayev

A. V.

(2014b). Revolyutsiya vs. demokratiya [Revolution vs. democracy]. Polis, 3, 139-158.

28.

Gurr

T. R.

(1970). Why men rebel. Princeton: Princeton University Press.

29.

Gurr

T. R.

(1974). Persistence and change in political systems, 1800-1971. American Political Science, 68, 1482-1504.

30.

Hegre

(2014). Democracy and armed conflict. Journal of Peace Research, 51, 159-172.

31.

Hegre

Ellingsen

Gates

Gledish

N. P.

(2001). Towards a democratic civil peace? Democracy, political change, and civil war, 1816-1992. American Political Science, 95, 33-48.

32.

Heinsohn

(2003). Söhne und Weltmacht. Terror im Aufstieg und Fall der Nationen [Sons and world power. Terror in the Rise and Fall of Nations]. Zürich, Switzerland: Orell Füssli Verlag.

33.

Jaggers

Gurr

T. R.

(1995). Tracking democracy’s third wave with the polity III data. Journal of Peace Research, 32, 469-482.

34.

Korotayev

(2014). Technological growth and sociopolitical destabilization: A trap at the escape from the trap? In Mandal

Asheulova

Kirdina

S. G.

(Eds.), Socio economic and technological innovation: Mechanisms and institutions (pp. 113-134). New Delhi, India: Narosa Publishing House.

35.

Korotayev

Issaev

Zinkina

(2015). Center-periphery dissonance as a possible factor of the revolutionary wave of 2013–2014: A cross-national analysis. Cross-Cultural Research, 49, 461-488.

36.

Korotayev

Malkov

Grinin

(2014). A trap at the escape from the trap? Some demographic structural factors of political instability in modernizing social systems. History & Mathematics, 4, 201-267.

37.

Korotayev

Zinkina

(2011). Egyptian Revolution: A Demographic Structural Analysis. Entelequia: Revista Interdisciplinar, 13, 139-169.

38.

Korotayev

Zinkina

Kobzeva

Bogevolnov

Khaltourina

Malkov

(2011). A trap at the escape from the trap? Demographic-structural factors of political instability in modern Africa and West Asia. Cliodynamics, 2, 276-303.

39.

Korotayev

A. V.

Issaev

L. M.

Malkov

S. Y.

Shishkina

A. R.

(2013). Developing the methods of estimation and forecasting the Arab Spring. Central European Journal of International and Security Studies, 7(4), 28-58.

40.

Korotayev

A. V.

Issaev

L. M.

Malkov

S. Y.

Shishkina

A. R.

(2014). The Arab Spring: A quantitative analysis. Arab Studies Quarterly, 36, 149-169.

41.

Korotayev

A. V.

Issaev

L. M.

Vasiliev

A. M.

(2015). Kolichestvennyj analiz revolyutsionnoj volny 2013-2014 gg [Quantitative analysis of the revolutionary wave of 2013-2014]. Sotsiologicheskie issledovaniya, 8, 119-127.

42.

Malkov

S. Y.

Korotaev

A. V.

Isaev

L. M.

Kuzminova

E. V.

(2013). O metodike otsenki tekuschego sostoyaniya i prognoza sotsial’noj nestabil’nosti: Opyt kolichestvennogo analiza sobytij Arabskoj vesny [On the methods of assessing the current state and forecasting social instability: A quantitative analysis of the Arab Spring events]. Polis (Politicheskie issledovaniya), 4, 137-162.

43.

Mansfield

Snyder

(1995). Democratization and the danger of war. International Security, 20(1), 5-38.

44.

Marshall

M. G.

Cole

B. R.

(2008, 29 August). A macro-comparative analysis of the problem of factionalism in emerging democracies. Paper presented at the 2008 annual meeting of the American Political Science Association, Boston, MA.

45.

Marshall

M.G.

Cole

B.R.

(2012, 22 June). Societal systems analysis and the problem of factionalism in emerging democracies. Paper presented at the 2012 Meeting of the International Studies and British International Studies Associations, Edinburgh, UK.

46.

Marshall

M. G.

Jaggers

(2016). Polity IV project: Political regime characteristics and transitions. Retrieved from http://www.systemicpeace.org/polity/polity4.htm

47.

Mesquida

C. G.

Weiner

N. I.

(1999). Male age composition and severity of conflicts. Politics and the Life Sciences, 18, 113-117.

48.

Moller

(1968). Youth as a force in the modern world. Comparative Studies in Society and History, 10, 238-260.

49.

Muller

E. N.

Weede

(1990). Cross-national variations in political violence: A rational action approach. Journal of Conflict Resolution, 34, 624-651.

50.

Przeworski

Alvarez

M. E.

Cheibub

J. A.

Limongi

(2000). Democracy and development. Cambridge, UK: Cambridge University Press.

51.

Puddington

Aghekyan

Coogan

Dunham

Nelson

Repucci

. . . Tucker

(2015). Discarding democracy: Return to the iron fist. Freedom in the world. Washington, DC: Freedom House.

52.

Regan

P. M.

Bell

(2009). Changing lanes or stuck in the middle: Why are anocracies more prone to civil wars? Political Research Quarterly, 63, 747-759.

53.

Regan

P. M.

Norton

(2005). Greed, grievance, and mobilization in Civil Wars. Journal of Conflict Resolution, 49, 319-336.

54.

Sambanis

(2001). Do ethnic and nonethnic civil wars have the same causes? A theoretical and empirical enquiry (part 1). Journal of Conflict Resolution, 45, 259-282.

55.

Schmitz

(1999). Thank God They’re On Our Side: The United States & Right-Wing Dictatorships. Chapel Hill, NC: University of North Carolina Press.

56.

Taylor

Jodice

(1983). World handbook of political and social indicators (3rd ed.). New Haven, CT: Yale University Press.

57.

Treier

Jackman

(2008). Democracy as a latent variable. American Journal of Political Science, 52, 201-217.

58.

Tsirel

S. V.

(2012a). Revolyutsii, volny revolyutsij i Arabskaya vesna [Revolutions, waves of revolutions and the Arab Spring]. In Korotayev

Zinkina

Khodunov

(Eds.), Sistemnyj monitoring global’nykh i regional’nykh riskov [System monitoring of global and regional risks] (pp. 128-161). Moscow, Russia: LIBROKOM/URSS.

59.

Tsirel

S. V.

(2012b). Usloviya vozniknoveniya revolyutsionnykh situatsij v arabskikh stranakh [Conditions of emergence of revolutionary situations in the Arab world]. In Korotayev

Zinkina

Khodunov

(Eds.), Sistemnyj monitoring global’nykh i regional’nykh riskov [System monitoring of global and regional risks] (pp. 162-173). Moscow, Russia: LIBROKOM/URSS.

60.

Ulfelder

Lustik

(2007). Modeling transitions to and from democracy. Democratization, 14, 351-387.

61.

Vreeland

J. R.

(2008). The effect of political regime on Civil War. Journal of Conflict Resolution, 52, 401-425.