My principal concern with this some of this weeks' readings, specifically the Boix and Stokes piece and the Przeworski and Limongi piece, is that they conflate correlation with theory, and stylised facts with empirical evdience. Throughout these two pieces, which I gather are influential in the literature, the focus is on the extent to which development 'causes' democracy. I do not think that this is necessarily a helpful rhetoric in which to conduct a debate on democratisation. I think it is over-simplistic. Of course, political scientists must always simplfy in order to carry out their work with any degree of generality, but in this instance I think the costs of simplication outweigh the gains.
The specific circumstances surrounding regime change are heterogeneous across countries; peaceful and violent struggles, abrupt and gradual processes, mass revolutions and insider coups. There is no reason to believe that any of these separate process, thought they may have similar outcomes, have similar antecedents. In fact, the factors that give rise to one form of regime transition may be completely different, even contrary to, the factors that give rise to another. Further, the process of 'economic development' is totally heterogeneous across countries, not simply in terms of the heterogeneity of economic activity, but in the extent to which it is influenced by geography, geology, demographics, culture and so on. The measure of GDP per capita is an enormously blunt instrument with which to capture a country's economic activity.
Thus, if we are to say that Y is a function of X, Y being a regime transition probability and X being economic development, our analyses will naturally be confounded by the fact that Y can, in fact, be any number of different things, and X can too be any number of different things, which may be related to the various 'Y' in a huge number of different ways. In the language of mathematics, we deal not with a function, but with a correspondence. I think this point, while hinted at in the Jones and Geddes pieces, is entirely missed by Boix, Stores, Przeworski and Limongi .
Pamela Paxton's insistence that operationalizations of 'democracy' should match most contemporary normative definitions, which include really universal suffrage (both genders, all ethnic groups, etc.), is understandable. In one sense, it is outrageous for current social science to trumpet democratization in societies at times where women, and often significant other portions of the adult native-born population, could not formally participate. But I disagree with her--it is not helpful to the purpose of pinpointing the why and the how of democratization processes to condition those on total suffrage.
It's true that formal legal inclusion of all adult groups may be realized mainly as a result of social movements promoting such changes, and determined by political cost/benefit calculations on the part of powerful decision makers, as is the case with other major political changes. But the actual inclusion (and in some cases, the formal inclusion too) of both genders and all races/ethnicities--when the latter are not territorially divided--is more dependent on the development of society's moral order as it refers to the worth and capacity of all adults as equal persons. While there are individuals that 'break the mold' and push the boundaries of what is allowed and expected of them based on their social position, gender and/or ethnicity, the manifestations of these society-wide beliefs on their capacity and worth in family, social and community structures determines whether the vast majority of once-excluded individuals may actually participate. (This is to some extent true for any type of marginalized group, but historical views on the capacity of women, and often of non-white races, to think 'like full men' and wield full moral capacity have changed as moral orders of human beings have been revised.)
Essentially, I argue, the processes normally studied as part of democratization, have to do with satisfying the demands various constituencies within a political community--and historically, women have been simply subsumed into those constituencies to which their husbands or fathers belonged. Thus, processes to include these constituencies was perceived to include them, with or without their votes. The process of changing women from passive constituents was really a change of views (sometimes gradually) of their capacities as human persons that is part of a long historical development--still occurring in some places. This characteristic is to some extent shared in decisions and processes that gradually included minority ethnic groups in many societies.
Paxton's article--while protesting that gender inclusion changes the assumptions and conclusions of democratization research--does not go far enough in asking how the operationalizations she suggests might actually be helpful to the research community, and how they might also add confusion to the dynamics being studied.
Where do political regimes come from? While many factors contribute to the emergence and consolidation of democracies and autocracies, at the end of the day, at the core of disputes over political regimes are conflicts over the distribution of assets (e.g. income) between different social groups. We know political regimes have important distributional consequences (e.g. the labor share is higher in democracies than under dictatorships). This is why social groups have preferences over regime types, and do “things” (creating parties, implementation of electoral reforms, revolutions) to bring about them. Thus, it is concerning that with the exception of Boix and Stokes, the issue of income distribution has remained outside the theoretical framework in the studies of democratization reviewed this week. Przeworski and Limongi hint at the issue in their “intuitive story” on democratic consolidation, but in their account of transitions, income distribution plays no role: democracies just happen. Cheibub relegates income distribution to a footnote (p.139) when trying to understand why presidential regimes are so unstable (the fact that presidentialism is the norm in Latin America, one of the most unequal regions of the world, could provide some clues to this issue) . Finally, Geddes discussion on the interests of non democratic actors suggests that the military functions as an autonomous political force. However, this ignores the stylized fact that military interventions in many countries of Latin America during the XXth century were backed by social groups interested in halting redistributive pressures.
Geddes' concern with the authoritarian structure prior to states' democratization stems from the fact that the process of democratization varies greatly across countries. Her characterization of authoritarian regimes into three types (personalist, single-party and military) aides in determining the underlying factors of each types’ downfall and transition into democracy. The transition from authoritarian rule to democracy should not be viewed as a single shift, without accounting for the forces driving the authoritarian regime, for regime changes can result in different allocation of resources, depending on the prior skeletal institutions leftover from the authoritarian regime type.
The central idea that different authoritarian regimes may face different pressures leading to democratization is ignored by Przeworski and Limongi, as well as Boix and Stokes. Their concern with endogenous versus exogenous democratization provides very limited (if any) breathing room for variation among states’ transition to democracy. This is troubling and limiting, for the transition to democracy does not occur from a single step, as their dichotomous models propose.
In a more positive light, Boix and Stokes’ use of what they believe were omitted variables by Przeworski and Limongi (capital mobility, international forces, income quality, etc.) suggests that the unbundling of potential causes of democratization have in fact an impact on the transition process. Economic development encompasses various attributes, most of which are proxied by GDP per capita, a variable which is not enlightening in and of itself. The association between economic growth and democratization via GDP per capita neglects institutional developments and the evolution of socio-economic indicators (such as education and income equality, as suggested by Boix and Stokes). One issue neglected by Boix and Stokes in their use of alternate development variables supporting democracy is that it is possible for regimes to provide income growth with an equitable distribution, education and health care, such as the case of Singapore, while remaining authoritarian. Geddes offers a possible explanation for this specific outlier, as single party authoritarian regimes often have institutional structure allowing for greater participation and popular influence, while retaining their political power. Additionally, if their interest is to hold office, then by providing institutions fostering investment, supplying conditions for economic stability and improving living standards, the People’s Action Party is able to remain in power without contestation. This would then suggest that perhaps states don’t require democracy per se in order to grow, but merely certain institutions that have been associated with democratic regimes.
Finally, in addressing the issue of various transitions to democracy and authoritarian regime types, Paxton’s recommendation of degrees of democratic transition could be helpful in understanding the transformation of each state. While it may be more difficult to determine the various “shades” of democracy, perhaps it is more realistic to view regimes in such light (returning to the Dahl’s degrees of liberalization and inclusiveness) . Graded measures of democracy would not only be more accurate, but also provide more insight into what institutional changes and developments are needed in order to achieve full democracy.
Within the abundant literature of regime transition, Geddes’ work (1999) provides a fresh perspective to understand the stability/instability of authoritarian regimes. By classifying authoritarian regimes into personalist, military and single-party three types, she attempts to explain the microfoundations of their performance differences in terms of democratization transition. Although Geddes’ article makes undeniable contribution to the study of authoritarism which is still deficient in existing literature of regime transition, I feel that her argument suffers from quite a few weaknesses.
First, Gedde makes different assumptions of individual behaviors under the three organizations: military, single party and cliques, without solid reasoning and evidences. Why military officials value the integration and autonomy of their organization more than anything else while their counterparts in the single political party put individual interests, say pursuing and maintaining office, above the interest of their organization? When internal cleavage jeopardizes integration of the organization, members of the organization no matter it is military or single party, face similar dilemma, either to cooperate or negotiate with the opposition to resume organizational integration, or “put down their heads and wait to see who wins the power struggle”. In Gedde’s piece, there are not sufficient empirical evidences or theories to support that military officials are more likely to adopt the former strategy than party officials do. The inconsistent assumptions of the interests of military and party make her explanations look very “ad hoc” and lack prediction power, not to mention the ambiguous interpretations of clique members’ interests.
Second, based on statistic records Gedde concludes that military regime is the most vulnerable to external shocks such as military threat, economic downturns while personalist regime is the most likely to be destabilized by internal threats including rebellion and assassination. In contrast, single-party regime is much more resilient than other authoritarian regimes. If we assume that all of the observation is accurate, what on earth is the causality behind these phenomena? Why can single party survive in crisis by internalizing threats with cooperating dissidents while military or autocrats are unable to achieve so by negotiation? The regularity behind this contrast deserves further research efforts.
Last, although Gedde reserves some room for the amalgams of the three pure types of authoritarian regimes in the classification, she puts less weight on the hybrid type which I believe is the most common case in reality. Moreover, authoritarian regimes might interchange among the three pure types, depending on specific circumstances. Will the shift between military regime to personalist regime, or personalist regime to single party regime affect the progress of democratic transition? Is there any regularity of the authoritarism evolution? It seems that questions we have about authoritarism evolution much exceed the questions we can solve now.
I was struck this week by the varying definitions and classifications adopted by the different authors in defining political regimes. Linz dedicates his article to a discussion of this issue expanding conventional typologies to make careful distinctions between different types of non-democratic regimes. In a somewhat contradictory approach (that is perhaps reflective of the period of writing) he then happily bundles the quite diverse forms of democracy into a single ‘democratic’ category. In direct contrast Cheibub’s research – which admittedly focuses on democratic regimes - takes considerable pains to distinguish between the finely gradated versions of parliamentary and presidential forms of democracy. Przeworksi and Limongi as well as Boix and Stokes adopt a more stylized approach, viewing regime type as a dichotomous choice between dictatorship and democracy. Whilst one can understand the forces driving the various definitions – both precise and stylized – it would appear that research of this type would benefit from a standardization or at least common definition of these various terms and typologies.
I would like to also specifically comment on the articles by Boix and Stokes and Przeworksi and Limongi that examined the endogenous and exogenous links between development and democracy. I echo Pierce’s concerns regarding the over-simplification of the argument presented in these articles. Not only has the analysis tended to generalize, and lump together the varying routes to democratization, but they seem to make grossly simplifying assumptions regarding the dependent and independent variables – equating ‘development’ to a simple measure of income/capita, and democracy to a broad category that could potentially encompass a wide range of regime types. Whilst useful in examining crude influences, this sort of analysis risks overlooking the contribution of other key factors to development (e.g. technology, human capital) and the variety of nuanced routes to, and types of democratic outcomes that are possible.
Barbara Geddes addresses, in part, the interesting question of how and if case studies of single countries (or regions)can be integrated with generalizations drawn from large-N analyses. While case studies have generated a great deal of data, she argues that there is not yet a great theoretical understanding of democratization. Instead, there are a number of contradictory theories. Geddes uses the data "hidden" in case studies to find a pattern that unravels one particular contradiction---why do some theories emphasize splits within the regime as a cause of the inititation of democratization, while others do not find factions and splits to be as important? In class, we have frequently discussed the value of case studies vs. the value of large-N studies. The best case studies give us carefully outlined and detailed causal stories, while analyses which look at a large set of cases can give us parsimonious and generalizable explanations. If we do not attempt to use the broader theories to sort through the numerous cases studies, however, we are left with conflicting theories (and unused data) that does not tell help us explore the questions we really want to answer.
In "Hit or Miss" Jones and Olken demonstrate the impact of political assassination on institutional change and war. They find political assassination of autocratic leaders increases the probability of transition to democracy, while political assassination of democratic leaders has no effect on the probability of transition to autocracy. Regarding war, Jones and Olken find that successful assassinations can intensify small-scale conflicts but can hasten the cessation of large-scale conflicts. Jones and Olken note how individual leaders and historical contingency interact in ways that have major repercussions on institutional change.
Jones and Olken’s provocative findings raise a number of questions. First, why are democracies more resilient than autocracies when it comes to assassination and institutional change? In short, what are them causal mechanisms underlying Jones and Olken’s findings? Could it be that in democracies the political process is important to maintaining legitimacy whereas in autocracies it is the leaders? Second, are Jones and Olken’s findings inconsistent with modernization theory? What is the relative importance of contingency and economic factors in transitions to democracy? Perhaps in line with Przeworski, assassination is simply an exogenous factor that can lead to democratization, but that underlying economic factors are what allow democracy to be sustained once it is established. Were this to be the case, it would be important to verify that there exists no correlation between the underlying economic factors (or any other factors) that make democracy more durable and the incidence of political assassination. Third, it is interesting (if not also a bit irresponsible) to contemplate the policy implications of Jones and Olken’s argument regarding the positive effects of political assassination on democracy promotion. Would the effects observed by Jones and Olken hold if assassinations were carried out by foreign states?
The methodological approaches utilized by Cheibub and Geddes are particularly appealing since they combine plausible and elegant micro-level behavioral models with macro-level analyses of striking empirical findings. Cheibub concludes that there are relatively few circumstances where presidential systems degenerate into irresolvable deadlock (as his coalition-buidling model suggests), so there is no reason to think that presidential systems are more likely to become authoritarian. But why does the data say that they in fact are more likely? His empirical analysis suggests that various historical coincidences, combined with a degree of institutional stickiness, made presidential systems more common in places where military interventions were also more common. Furthermore, the recurrence of military dictatorships observed by Cheibub seems plausible given what Geddes identifies as that regime type's tendency to collapse more easily and more orderly than other authoritarian regimes. In the absence of other exogenous pressures (e.g., post-WWII occupations in Europe), this ensures the survival of military institutions (often unreformed) and allows for future intervention. The prescription, as described by Cheibub, are policies focusing on diminishing the role of the military or changing its incentives for intervention. In this respect, Geddes's discussion of the Battle of the Sexes game should be illustrative for policy makers.
Because of their attention to the micro-level behavioral processes, Cheibub and Geddes provide material that is both informative and applicable across a variety of cases. The disproportionate reliance on macro-level statistical analysis by Przeworski and Limongi is less informative (even if it is "agenda setting"), given the methodological weaknesses discussed by Boix and Stokes. This is a more general observation: it seems that, given the nature of this type of research, studies without micro-level process modeling will be prone to the types of back-and-forth methodological argumentation as seen last week with AJR and Glaeser et al. and this week with Przeworski and Limongi and Boix and Stokes.
The readings this week attempt to answer the question of where regimes come from and when are democratic regimes more likely to be stable. Cheibub argues that when transitions to democracy occur (for whatever reasons, which he claims are rooted in each countries' historical particularities) they are more likely to suffer authoritarian reversals if they were preceded by military dictatorships (versus civilian ones). He thus discards claims about the inherent instability of presidential systems. Przeworski and Limongi claim that transitions to democracy occur for reasons unrelated to 'modernization' and development (measured by GDP), but that level of development is crucial in determining whether a country will remain a democracy. Their stories are plausible in that history shows that the proximate reasons for the breakdown of authoritarian regimes seem many times unrelated to development and instead connected to things as loss of legitimacy of a military rule, loss in war, death of a dictator and international influences. Yet Boix and Stokes dispute the findings of P & L (through critiques of robustness and omitted variable) and find instead that development does lead to a higher likelihood of transition to democracy. They complement their findings with explanations of why this may be so (a clear step forward from earlier very fuzzy accounts modernization theory). Boix describes how development decreases income inequality, and lower inequality is conducive to democratization as it lowers expected demands from redistribution for elites. What i wonder, then, is if development reduces inequality and leads to democracy, what is the political calculus of authoritarian regimes: A) foster development, increase the pie and then inevitably have to share it?, or B) keep the pie smaller but keep all of it? their answer seems to be A, but this would exclude that possibility that authoritarian regimes have little or no preference for power for its own sake. Is this the case? And, of course, their account leaves out the cases where income equality serves to temper demands for democracy, the classic example being singapore. Also, if development leads to greater equality, what purpose would democratization serve for those demanding it? If development itself is an equalizer, then demands for democracy may be rooted in other factors, preference for democracy per se, etc (and this of course brings us back to historical particularities).
The assassination piece is interesting in the sense that it links the individual leadership to institutional change. This relationship is significant especially in authoritarian countries, since leadership succession is institutionalized in democracies. I think the immediate influence of a successful assassination is the problem of succession, the mechanisms of which are various among countries. And as Linz and Stepen point out, there is a great variety among different authoritarian regimes, so are the roles of their leadership. There should thus be discrepant effects on institutional change. For example, I would suppose assassination should have the least effect on institutions in the single-party authoritarianism. Yet, from Jones and Olken’s work, we cannot tell these differences. I don’t know if this would work methodologically, but if we categorize the sample cases according to Geddes’ or Stepen and Linz’s typologies, we should find different effects of assassination on institutional change. It is probably most relevant in personalistic or Sultanist regimes. Through this practice, we might as well identify the levels of institutionalization of succession among different authoritarianism.
In the context of longstanding debate about the effectiveness of parliamentary versus presidential democracies as a function of institutional features or exogenous conditions, Cheibub offers an explanation for the instability of presidential democracies that relies on the nexus between militarism and presidentialism: democratic regimes following military dictatorships are more likely to fail and presidentialism arises more often from military dictatorships than parliamentarism. While this theoretical link seems plausible, the argument neglects the intrinsic institutional structure of parliamentarism that has been shown to be more conducive to a stable democracy than presidentialism (Linz, The Perils of Presidentialism,1990) and this is likely to account for part of the results. Furthermore, given Huntington's waves of democratization, it might be useful to control for the end of the Cold War in order to examine the extent to which differences in the emergence of democracies, especially as a result of international democratization efforts post-CW affect the results. Finally, the question arises how robust Cheibub's results are given his coding criteria in view of Paxter's discussion of the impacts of coding rules for regimes on results.
(In my excitement I neglected to post my comment, as with Emily, I apologize for the length)
Geddes is quite correct when she observes that the usefulness of any typology is dependent on the question one is trying to answer. She develops a very simple, clean explanation as to why and how different regimes collapse and crafts a typology that neatly splits extent regimes along her main independent variable of interest. The advantage of such an approach is that she is then able to conduct very nice statistical tests in order to test her theories. Unfortunately, she simultaneously shows the pitfalls of using an overly simple coding scheme in that she forced to relegate a sizeable number of regimes to a hybrid “mixed” category, thus potentially muddying the explanatory power and usefulness of her typology.
Linz and Stepan, on the other hand, develop a much messier typology that classifies regimes along multiple dimensions: Pluralism, ideology, mobilization, and leadership. In doing so, they create a rich typology that manages to capture regimes neatly into each category without any real residual and that also makes it easy to create structured, descriptive case studies. Unfortunately, however, this typology makes it difficult to test theories statistically because each regime type varies along all of the relevant dimensions with few overlaps. Moreover, it is unlikely that the categories are independent of each other: ideology and mobilization are likely so highly correlated as to break down standard statistical assumptions.
Is there a middle ground, however? An ideal typology would need to be relatively simple and have orthogonal elements to insure statistical testability, while at the same time being nuanced enough to avoid potentially bias-inducing mixed categories. Given this, do you all agree with Geddes that attempting to create uniform typologies useable across multiple studies is a lost cause?
“Since democracies base their claim to obedience on the procedural foundations of democratic citizenship, as well as performance, they have a layer of insulation against weak performance not available to most post-totalitarian or authoritarian regimes” (Linz, 49)
What happen when notion of citizenship needs to be discussed (immigration concern) or performance is weaken by an exogenous shock (financial crisis for example: Quid of Iceland?). It does not seem obvious to me that the notion of citizenship is all clear in the democratic context, especially in a globalizing economic context.
Economic Development and Regimes’ Type
Geddes states that single-party regimes are more resilient than “authoritarian/military” regimes (definition is different from Linz) to exogenous shocks such as economic crisis (see p139-140). So why the proportion of dictatorships let by military is higher when there are economic crisis (associated with war scenarios in the long run) as described in Cheibub (fig. 6.1, p149)?
There is something contradictory in the readings in general. It is said that economic development increases the likelihood of democratic politics but also that during economic crisis, authoritarian regimes are more likely to break down (Geddes, 140). So when are authoritarian regimes developing, under what conditions? This seems contradictory to me and I am wondering if it’s not because of “by country” approach. Each country in the statistical analysis that are used (from my understanding) is equal at 1. So when the Soviet block collapsed, there is a large sample of countries whose behavior is similar. But is it the representation of the relationship between authoritarian regimes and economic crisis? It’s not because several cases are similar that they demonstrate this relation, it may be because of their own relation between each other that they behave the same way and not because of the exogenous factor on each one of these countries.
Democracy and Informality
Following the Linz’ model: could we have institutional pluralism –ie. second economy and parallel culture- in a democratic regime?
“post-totalitarian leaders tend to be more bureaucratic and state technocratic than charismatic” (Linz, 47). Are African “democracies” more of post-totalitarian regimes then considering the leaderships and the importance of informal economy and culture?
The problem I have with the classifications that are used (democracy vs. authoritarian regimes) is that the main assumption is that democracy is based on a narrow definition of private property, hence capitalism. Giving the fact that capital accumulation is the reference point to economic development, democracy and economic development will go well together (Przeworski). But values outside private property and capital accumulation and in general values that are not formalized within this context are not captured. The notion of modernization is associated to westernization by Said, Mamdani, Chatterjee for example. Przeworski show that there are links between modernization (economic development) to democracy (at least in the survival of a democratic regime vis-à-vis its economic development). Are we defining a western specificity? What is the point of comparing regimes’ behavior if we use only a western referential?
I agree with several of the other comments that the dichotomous relationship that Przeworski & Limongi and Boix & Stokes (perhaps to a lesser extent) establish between economic development and democratization is problematic for reasons outlined above. Be that as it may, it is not the focus of my comment. Rather I want to introduce a further consideration; a variation over the theme of contextual importance that emerges from some of the comments and constitutes the brunt of criticism in some of them.
A contrasting picture emerges from consideration of the aforementioned works. There is an endogenous effect, but it is unclear how strong it is across different periods of time. It had a clear impact on the chances of democratization in the century leading up to 1950, as Boix & Stokes show, but the effect is much smaller after that, whereas the effect of economic development on democratic stability becomes much more pronounced, as Przeworski & Limongi demonstrate. Importantly, Boix & Stokes point out that a thorough reading of Preworski et al’s book shows that even they find a small, but statistically significant, effect of development on the probability of democratic transitions. However, the picture that emerges from consideration of the entire period from 1850 to 1990 is that on average the relative effect of development is stronger in maintaining the stability of democratic regimes and less in inducing transitions.
I think this leads us to a fundamental question. If the effect of given levels of economic development ostensibly has what appear to be significantly different effects, in terms of strength and causality, on democratizations in different historical periods – that is, differences across time and type of authoritarian regimes – could this imply that processes of democratization or transitions to democracy under such varying circumstances indeed are different? Should they, then, be treated as such?
The evidence at least suggests, as Barbara Geddes points out (in “What Causes Democratisation”, Oxford Handbook of Comparative Politics, 2006) that processes of democratization may be different in different historical periods. This surely matters for our understanding of the topic and we should therefore not uncritically believe that one explanation is adequate or that the same explanation is appropriate for different periods and their distinctive contexts.
She puts it this way: “We might make progress faster, both empirically and theoretically, if we identified clear domains for our arguments about the causes of democratization rather than assuming that just because we cover many processes of democratization with one word we should also uncritically model it as one process regardless of what we know about historical and other differences.” (See p. 336 in the same volume). In short, as Robert Franzese has put it, context matters.
In Presidentialism, Parliamentarism and Democracy, Jose Cheibub attempts to refute the Linzian claim that the intrinsic features of presidential systems cause them to have less longevity than parliamentary systems. Instead, Cheibub argues that it is not presidentialism itself that is determining; rather, presidential systems are more fragile because they typically emerged in nations with a strong military presence, which “placed democracy of any kind at risk.” Do we find Cheibub’s claim convincing? Indeed, Cheibub offers no sound empirical test of his claim that “initial institutions …[are adopted] for reasons hat are unrelated to the ones that lead to the occurrence of military dictatorships” (147). Cheibub’s analysis is missing several causal links. While Cheibub describes the military/presidential nexus as an “historical accident,” he fails to explain what may influence a country’s propensity for militarization. Without that analysis, how can he claims that the reasons for militarism are not the same as for institutional development? Further, what causes a militaristic dictatorship to form rather than a civil dictatorship? We can also ask whether Cheibub’s work suffers from the same type of sample selection problem that Stokes and Boix point out in Przeworski and Limongi’s work. As noted in Chapter 4, Cheibub’s dataset is confined to the period 1946-2002. If we took the dataset farther back in time would the results change? Further, how sound is Cheibub’s theory about the importance of the Cold War in fostering the kind of militarism that leads to the breakdown of democratic regimes when most of his dataset consists of the time period of the Cold War? Would his analysis be different if he studied different waves of the Cold War, such as early Cold War, height of Cold War and post Cold War? Indeed, could we make the argument that the international environment at the time of regime formation rather than the presence or absence of a military regime may be more important in effecting the stability of presidential regimes? By confining his analysis primarily to Cold War cases, Cheibub is not differentiating between the importance of each variable. Finally, what can we learn from the Geddes article that may influence our understanding of why military regimes may cause subsequent democracies to be more unstable? Geddes argues that most military regimes end in negotiation as there are many internal disagreements and splits. It could be the case that this “symptom” of militarism rather than militarism itself plays a more important causal role in Cheibub’s analysis.
In Presidentialism, Parliamentarism, and Democracy, Cheibub argues that the high mortality rates of presidential democracies are not due to inherent flaws in their institutional framework as suggested by other scholars. Contrary to arguments proposed by other scholars which claim that parliamentary democracies are more stable because they promote party discipline, create relationships of interdependence between various branches of government, and facilitate inter-branch cooperation, Cheibub argues that the instability present in presidential regimes are due to the environment in which these regimes are created. Specifically, he argues that many presidential regimes were formed in countries where military regimes reigned supreme and that due to the stickiness of these past regimes with their authoritarian tendencies, any form of democracy (presidential or parliamentary) would be unstable.
I agree that much of the instability observed in presidential democracies can be attributed to institutional legacies from past regimes. This claim correctly takes into account that once new regimes structures are created, the old institutions very rarely completely disappear. His argument forces an analysis of relevant historical and macro level developments within specific states, which I believe is extremely important.
However, I have several questions. First, what impact does a federal or unitary government structure have on Cheibub’s claims? For example, is there any variation between federal/unitary parliamentary governments or federal/unitary presidential regimes? Also, what is the impact between successive shifts between parliamentarism and presidentialism for regime stability? If one were able to find a case of a state that switched from one form to another, it would provide an interesting opportunity to test (within one environment) the implications of each system on regime stability. It seems that many of the cases that he discusses involve a country transitioning to democracy and then adopting a particular form (presidentialism or parliamentarism) and sticking with it. Although, he does note that when a switch occurs, it is more likely to be from parliamentarism to presidentialism, rather than the reverse. Why is this the case?
Over the last two or three decades, comparative politics transitioned from a focus on regime stability to an emphasis of the study of democratic institutions and their attributes. Linz and Stepan argue that such a shift is necessary because earlier typologies can no longer document the variation in regime characteristics. According to their calculation, 90 percent of non-democratic regimes qualify as authoritarian states. Beginning with Linz and Stepan, an increasing number of studies and indices (Freedom House) also ranks countries on their democratization level. Democracy replaced stability as the gadget of political science.
It remains to be seen whether the paradigm shift towards the study of democracies, dictatorships and the shades of pluralism in-between is not a normative projection of Western liberal values. The expansion of democratization scholarship has gone hand in hand with the establishment of consolidated democracy assistance programs by the United States and other Western donors in the mid 90s. Many of these programs are directly linked to the scholarship on democratization and its triggers. From a policy point a view, it might be premature to use evidence from the social sciences to design and implement foreign aid policy. Boix and Stokes point out that Juan Lopez, a Cuban-American academic, used the exposition of modernization theory by Przeworski and Limongi (PL) to argue for continued economic sanctions to the communist regime in the Carribean. Lopez does not mention that the results from the original study by PL can be hardly applied to single cases because of omitted variable bias, case selection problems and the lack of reliable robustness tests.
To what extent should policymakers tap into current debates on political regimes? Are democracies indeed more stable than other consolidated regime types (e.g. corporatist states)? What recent studies of regime stability can we think of?
My principal concern with this some of this weeks' readings, specifically the Boix and Stokes piece and the Przeworski and Limongi piece, is that they conflate correlation with theory, and stylised facts with empirical evdience. Throughout these two pieces, which I gather are influential in the literature, the focus is on the extent to which development 'causes' democracy. I do not think that this is necessarily a helpful rhetoric in which to conduct a debate on democratisation. I think it is over-simplistic. Of course, political scientists must always simplfy in order to carry out their work with any degree of generality, but in this instance I think the costs of simplication outweigh the gains.
ReplyDeleteThe specific circumstances surrounding regime change are heterogeneous across countries; peaceful and violent struggles, abrupt and gradual processes, mass revolutions and insider coups. There is no reason to believe that any of these separate process, thought they may have similar outcomes, have similar antecedents. In fact, the factors that give rise to one form of regime transition may be completely different, even contrary to, the factors that give rise to another. Further, the process of 'economic development' is totally heterogeneous across countries, not simply in terms of the heterogeneity of economic activity, but in the extent to which it is influenced by geography, geology, demographics, culture and so on. The measure of GDP per capita is an enormously blunt instrument with which to capture a country's economic activity.
Thus, if we are to say that Y is a function of X, Y being a regime transition probability and X being economic development, our analyses will naturally be confounded by the fact that Y can, in fact, be any number of different things, and X can too be any number of different things, which may be related to the various 'Y' in a huge number of different ways. In the language of mathematics, we deal not with a function, but with a correspondence. I think this point, while hinted at in the Jones and Geddes pieces, is entirely missed by Boix, Stores, Przeworski and Limongi .
Pierce O Reilly
(I apologize for the length!)
ReplyDeletePamela Paxton's insistence that operationalizations of 'democracy' should match most contemporary normative definitions, which include really universal suffrage (both genders, all ethnic groups, etc.), is understandable. In one sense, it is outrageous for current social science to trumpet democratization in societies at times where women, and often significant other portions of the adult native-born population, could not formally participate. But I disagree with her--it is not helpful to the purpose of pinpointing the why and the how of democratization processes to condition those on total suffrage.
It's true that formal legal inclusion of all adult groups may be realized mainly as a result of social movements promoting such changes, and determined by political cost/benefit calculations on the part of powerful decision makers, as is the case with other major political changes. But the actual inclusion (and in some cases, the formal inclusion too) of both genders and all races/ethnicities--when the latter are not territorially divided--is more dependent on the development of society's moral order as it refers to the worth and capacity of all adults as equal persons. While there are individuals that 'break the mold' and push the boundaries of what is allowed and expected of them based on their social position, gender and/or ethnicity, the manifestations of these society-wide beliefs on their capacity and worth in family, social and community structures determines whether the vast majority of once-excluded individuals may actually participate. (This is to some extent true for any type of marginalized group, but historical views on the capacity of women, and often of non-white races, to think 'like full men' and wield full moral capacity have changed as moral orders of human beings have been revised.)
Essentially, I argue, the processes normally studied as part of democratization, have to do with satisfying the demands various constituencies within a political community--and historically, women have been simply subsumed into those constituencies to which their husbands or fathers belonged. Thus, processes to include these constituencies was perceived to include them, with or without their votes. The process of changing women from passive constituents was really a change of views (sometimes gradually) of their capacities as human persons that is part of a long historical development--still occurring in some places. This characteristic is to some extent shared in decisions and processes that gradually included minority ethnic groups in many societies.
Paxton's article--while protesting that gender inclusion changes the assumptions and conclusions of democratization research--does not go far enough in asking how the operationalizations she suggests might actually be helpful to the research community, and how they might also add confusion to the dynamics being studied.
This comment has been removed by the author.
ReplyDeleteWhere do political regimes come from? While many factors contribute to the emergence and consolidation of democracies and autocracies, at the end of the day, at the core of disputes over political regimes are conflicts over the distribution of assets (e.g. income) between different social groups. We know political regimes have important distributional consequences (e.g. the labor share is higher in democracies than under dictatorships). This is why social groups have preferences over regime types, and do “things” (creating parties, implementation of electoral reforms, revolutions) to bring about them. Thus, it is concerning that with the exception of Boix and Stokes, the issue of income distribution has remained outside the theoretical framework in the studies of democratization reviewed this week.
ReplyDeletePrzeworski and Limongi hint at the issue in their “intuitive story” on democratic consolidation, but in their account of transitions, income distribution plays no role: democracies just happen. Cheibub relegates income distribution to a footnote (p.139) when trying to understand why presidential regimes are so unstable (the fact that presidentialism is the norm in Latin America, one of the most unequal regions of the world, could provide some clues to this issue) . Finally, Geddes discussion on the interests of non democratic actors suggests that the military functions as an autonomous political force. However, this ignores the stylized fact that military interventions in many countries of Latin America during the XXth century were backed by social groups interested in halting redistributive pressures.
Geddes' concern with the authoritarian structure prior to states' democratization stems from the fact that the process of democratization varies greatly across countries. Her characterization of authoritarian regimes into three types (personalist, single-party and military) aides in determining the underlying factors of each types’ downfall and transition into democracy. The transition from authoritarian rule to democracy should not be viewed as a single shift, without accounting for the forces driving the authoritarian regime, for regime changes can result in different allocation of resources, depending on the prior skeletal institutions leftover from the authoritarian regime type.
ReplyDeleteThe central idea that different authoritarian regimes may face different pressures leading to democratization is ignored by Przeworski and Limongi, as well as Boix and Stokes. Their concern with endogenous versus exogenous democratization provides very limited (if any) breathing room for variation among states’ transition to democracy. This is troubling and limiting, for the transition to democracy does not occur from a single step, as their dichotomous models propose.
In a more positive light, Boix and Stokes’ use of what they believe were omitted variables by Przeworski and Limongi (capital mobility, international forces, income quality, etc.) suggests that the unbundling of potential causes of democratization have in fact an impact on the transition process. Economic development encompasses various attributes, most of which are proxied by GDP per capita, a variable which is not enlightening in and of itself. The association between economic growth and democratization via GDP per capita neglects institutional developments and the evolution of socio-economic indicators (such as education and income equality, as suggested by Boix and Stokes). One issue neglected by Boix and Stokes in their use of alternate development variables supporting democracy is that it is possible for regimes to provide income growth with an equitable distribution, education and health care, such as the case of Singapore, while remaining authoritarian. Geddes offers a possible explanation for this specific outlier, as single party authoritarian regimes often have institutional structure allowing for greater participation and popular influence, while retaining their political power. Additionally, if their interest is to hold office, then by providing institutions fostering investment, supplying conditions for economic stability and improving living standards, the People’s Action Party is able to remain in power without contestation. This would then suggest that perhaps states don’t require democracy per se in order to grow, but merely certain institutions that have been associated with democratic regimes.
Finally, in addressing the issue of various transitions to democracy and authoritarian regime types, Paxton’s recommendation of degrees of democratic transition could be helpful in understanding the transformation of each state. While it may be more difficult to determine the various “shades” of democracy, perhaps it is more realistic to view regimes in such light (returning to the Dahl’s degrees of liberalization and inclusiveness) . Graded measures of democracy would not only be more accurate, but also provide more insight into what institutional changes and developments are needed in order to achieve full democracy.
My Response Paper for the week can be found on Courseworks.
ReplyDeleteWithin the abundant literature of regime transition, Geddes’ work (1999) provides a fresh perspective to understand the stability/instability of authoritarian regimes. By classifying authoritarian regimes into personalist, military and single-party three types, she attempts to explain the microfoundations of their performance differences in terms of democratization transition. Although Geddes’ article makes undeniable contribution to the study of authoritarism which is still deficient in existing literature of regime transition, I feel that her argument suffers from quite a few weaknesses.
ReplyDeleteFirst, Gedde makes different assumptions of individual behaviors under the three organizations: military, single party and cliques, without solid reasoning and evidences. Why military officials value the integration and autonomy of their organization more than anything else while their counterparts in the single political party put individual interests, say pursuing and maintaining office, above the interest of their organization? When internal cleavage jeopardizes integration of the organization, members of the organization no matter it is military or single party, face similar dilemma, either to cooperate or negotiate with the opposition to resume organizational integration, or “put down their heads and wait to see who wins the power struggle”. In Gedde’s piece, there are not sufficient empirical evidences or theories to support that military officials are more likely to adopt the former strategy than party officials do. The inconsistent assumptions of the interests of military and party make her explanations look very “ad hoc” and lack prediction power, not to mention the ambiguous interpretations of clique members’ interests.
Second, based on statistic records Gedde concludes that military regime is the most vulnerable to external shocks such as military threat, economic downturns while personalist regime is the most likely to be destabilized by internal threats including rebellion and assassination. In contrast, single-party regime is much more resilient than other authoritarian regimes. If we assume that all of the observation is accurate, what on earth is the causality behind these phenomena? Why can single party survive in crisis by internalizing threats with cooperating dissidents while military or autocrats are unable to achieve so by negotiation? The regularity behind this contrast deserves further research efforts.
Last, although Gedde reserves some room for the amalgams of the three pure types of authoritarian regimes in the classification, she puts less weight on the hybrid type which I believe is the most common case in reality. Moreover, authoritarian regimes might interchange among the three pure types, depending on specific circumstances. Will the shift between military regime to personalist regime, or personalist regime to single party regime affect the progress of democratic transition? Is there any regularity of the authoritarism evolution? It seems that questions we have about authoritarism evolution much exceed the questions we can solve now.
I was struck this week by the varying definitions and classifications adopted by the different authors in defining political regimes. Linz dedicates his article to a discussion of this issue expanding conventional typologies to make careful distinctions between different types of non-democratic regimes. In a somewhat contradictory approach (that is perhaps reflective of the period of writing) he then happily bundles the quite diverse forms of democracy into a single ‘democratic’ category. In direct contrast Cheibub’s research – which admittedly focuses on democratic regimes - takes considerable pains to distinguish between the finely gradated versions of parliamentary and presidential forms of democracy. Przeworksi and Limongi as well as Boix and Stokes adopt a more stylized approach, viewing regime type as a dichotomous choice between dictatorship and democracy. Whilst one can understand the forces driving the various definitions – both precise and stylized – it would appear that research of this type would benefit from a standardization or at least common definition of these various terms and typologies.
ReplyDeleteI would like to also specifically comment on the articles by Boix and Stokes and Przeworksi and Limongi that examined the endogenous and exogenous links between development and democracy. I echo Pierce’s concerns regarding the over-simplification of the argument presented in these articles. Not only has the analysis tended to generalize, and lump together the varying routes to democratization, but they seem to make grossly simplifying assumptions regarding the dependent and independent variables – equating ‘development’ to a simple measure of income/capita, and democracy to a broad category that could potentially encompass a wide range of regime types. Whilst useful in examining crude influences, this sort of analysis risks overlooking the contribution of other key factors to development (e.g. technology, human capital) and the variety of nuanced routes to, and types of democratic outcomes that are possible.
Barbara Geddes addresses, in part, the interesting question of how and if case studies of single countries (or regions)can be integrated with generalizations drawn from large-N analyses. While case studies have generated a great deal of data, she argues that there is not yet a great theoretical understanding of democratization. Instead, there are a number of contradictory theories. Geddes uses the data "hidden" in case studies to find a pattern that unravels one particular contradiction---why do some theories emphasize splits within the regime as a cause of the inititation of democratization, while others do not find factions and splits to be as important?
ReplyDeleteIn class, we have frequently discussed the value of case studies vs. the value of large-N studies. The best case studies give us carefully outlined and detailed causal stories, while analyses which look at a large set of cases can give us parsimonious and generalizable explanations. If we do not attempt to use the broader theories to sort through the numerous cases studies, however, we are left with conflicting theories (and unused data) that does not tell help us explore the questions we really want to answer.
In "Hit or Miss" Jones and Olken demonstrate the impact of political assassination on institutional change and war. They find political assassination of autocratic leaders increases the probability of transition to democracy, while political assassination of democratic leaders has no effect on the probability of transition to autocracy. Regarding war, Jones and Olken find that successful assassinations can intensify small-scale conflicts but can hasten the cessation of large-scale conflicts. Jones and Olken note how individual leaders and historical contingency interact in ways that have major repercussions on institutional change.
ReplyDeleteJones and Olken’s provocative findings raise a number of questions. First, why are democracies more resilient than autocracies when it comes to assassination and institutional change? In short, what are them causal mechanisms underlying Jones and Olken’s findings? Could it be that in democracies the political process is important to maintaining legitimacy whereas in autocracies it is the leaders? Second, are Jones and Olken’s findings inconsistent with modernization theory? What is the relative importance of contingency and economic factors in transitions to democracy? Perhaps in line with Przeworski, assassination is simply an exogenous factor that can lead to democratization, but that underlying economic factors are what allow democracy to be sustained once it is established. Were this to be the case, it would be important to verify that there exists no correlation between the underlying economic factors (or any other factors) that make democracy more durable and the incidence of political assassination. Third, it is interesting (if not also a bit irresponsible) to contemplate the policy implications of Jones and Olken’s argument regarding the positive effects of political assassination on democracy promotion. Would the effects observed by Jones and Olken hold if assassinations were carried out by foreign states?
-Simon Collard-Wexler
The methodological approaches utilized by Cheibub and Geddes are particularly appealing since they combine plausible and elegant micro-level behavioral models with macro-level analyses of striking empirical findings. Cheibub concludes that there are relatively few circumstances where presidential systems degenerate into irresolvable deadlock (as his coalition-buidling model suggests), so there is no reason to think that presidential systems are more likely to become authoritarian. But why does the data say that they in fact are more likely? His empirical analysis suggests that various historical coincidences, combined with a degree of institutional stickiness, made presidential systems more common in places where military interventions were also more common. Furthermore, the recurrence of military dictatorships observed by Cheibub seems plausible given what Geddes identifies as that regime type's tendency to collapse more easily and more orderly than other authoritarian regimes. In the absence of other exogenous pressures (e.g., post-WWII occupations in Europe), this ensures the survival of military institutions (often unreformed) and allows for future intervention. The prescription, as described by Cheibub, are policies focusing on diminishing the role of the military or changing its incentives for intervention. In this respect, Geddes's discussion of the Battle of the Sexes game should be illustrative for policy makers.
ReplyDeleteBecause of their attention to the micro-level behavioral processes, Cheibub and Geddes provide material that is both informative and applicable across a variety of cases. The disproportionate reliance on macro-level statistical analysis by Przeworski and Limongi is less informative (even if it is "agenda setting"), given the methodological weaknesses discussed by Boix and Stokes. This is a more general observation: it seems that, given the nature of this type of research, studies without micro-level process modeling will be prone to the types of back-and-forth methodological argumentation as seen last week with AJR and Glaeser et al. and this week with Przeworski and Limongi and Boix and Stokes.
The readings this week attempt to answer the question of where regimes come from and when are democratic regimes more likely to be stable. Cheibub argues that when transitions to democracy occur (for whatever reasons, which he claims are rooted in each countries' historical particularities) they are more likely to suffer authoritarian reversals if they were preceded by military dictatorships (versus civilian ones). He thus discards claims about the inherent instability of presidential systems. Przeworski and Limongi claim that transitions to democracy occur for reasons unrelated to 'modernization' and development (measured by GDP), but that level of development is crucial in determining whether a country will remain a democracy. Their stories are plausible in that history shows that the proximate reasons for the breakdown of authoritarian regimes seem many times unrelated to development and instead connected to things as loss of legitimacy of a military rule, loss in war, death of a dictator and international influences. Yet Boix and Stokes dispute the findings of P & L (through critiques of robustness and omitted variable) and find instead that development does lead to a higher likelihood of transition to democracy. They complement their findings with explanations of why this may be so (a clear step forward from earlier very fuzzy accounts modernization theory). Boix describes how development decreases income inequality, and lower inequality is conducive to democratization as it lowers expected demands from redistribution for elites. What i wonder, then, is if development reduces inequality and leads to democracy, what is the political calculus of authoritarian regimes: A) foster development, increase the pie and then inevitably have to share it?, or B) keep the pie smaller but keep all of it? their answer seems to be A, but this would exclude that possibility that authoritarian regimes have little or no preference for power for its own sake. Is this the case? And, of course, their account leaves out the cases where income equality serves to temper demands for democracy, the classic example being singapore. Also, if development leads to greater equality, what purpose would democratization serve for those demanding it? If development itself is an equalizer, then demands for democracy may be rooted in other factors, preference for democracy per se, etc (and this of course brings us back to historical particularities).
ReplyDeleteThe assassination piece is interesting in the sense that it links the individual leadership to institutional change. This relationship is significant especially in authoritarian countries, since leadership succession is institutionalized in democracies. I think the immediate influence of a successful assassination is the problem of succession, the mechanisms of which are various among countries. And as Linz and Stepen point out, there is a great variety among different authoritarian regimes, so are the roles of their leadership. There should thus be discrepant effects on institutional change. For example, I would suppose assassination should have the least effect on institutions in the single-party authoritarianism. Yet, from Jones and Olken’s work, we cannot tell these differences.
ReplyDeleteI don’t know if this would work methodologically, but if we categorize the sample cases according to Geddes’ or Stepen and Linz’s typologies, we should find different effects of assassination on institutional change. It is probably most relevant in personalistic or Sultanist regimes. Through this practice, we might as well identify the levels of institutionalization of succession among different authoritarianism.
In the context of longstanding debate about the effectiveness of parliamentary versus presidential democracies as a function of institutional features or exogenous conditions, Cheibub offers an explanation for the instability of presidential democracies that relies on the nexus between militarism and presidentialism: democratic regimes following military dictatorships are more likely to fail and presidentialism arises more often from military dictatorships than parliamentarism. While this theoretical link seems plausible, the argument neglects the intrinsic institutional structure of parliamentarism that has been shown to be more conducive to a stable democracy than presidentialism (Linz, The Perils of Presidentialism,1990) and this is likely to account for part of the results. Furthermore, given Huntington's waves of democratization, it might be useful to control for the end of the Cold War in order to examine the extent to which differences in the emergence of democracies, especially as a result of international democratization efforts post-CW affect the results. Finally, the question arises how robust Cheibub's results are given his coding criteria in view of Paxter's discussion of the impacts of coding rules for regimes on results.
ReplyDeleteThis comment has been removed by the author.
ReplyDelete(In my excitement I neglected to post my comment, as with Emily, I apologize for the length)
ReplyDeleteGeddes is quite correct when she observes that the usefulness of any typology is dependent on the question one is trying to answer. She develops a very simple, clean explanation as to why and how different regimes collapse and crafts a typology that neatly splits extent regimes along her main independent variable of interest. The advantage of such an approach is that she is then able to conduct very nice statistical tests in order to test her theories. Unfortunately, she simultaneously shows the pitfalls of using an overly simple coding scheme in that she forced to relegate a sizeable number of regimes to a hybrid “mixed” category, thus potentially muddying the explanatory power and usefulness of her typology.
Linz and Stepan, on the other hand, develop a much messier typology that classifies regimes along multiple dimensions: Pluralism, ideology, mobilization, and leadership. In doing so, they create a rich typology that manages to capture regimes neatly into each category without any real residual and that also makes it easy to create structured, descriptive case studies. Unfortunately, however, this typology makes it difficult to test theories statistically because each regime type varies along all of the relevant dimensions with few overlaps. Moreover, it is unlikely that the categories are independent of each other: ideology and mobilization are likely so highly correlated as to break down standard statistical assumptions.
Is there a middle ground, however? An ideal typology would need to be relatively simple and have orthogonal elements to insure statistical testability, while at the same time being nuanced enough to avoid potentially bias-inducing mixed categories. Given this, do you all agree with Geddes that attempting to create uniform typologies useable across multiple studies is a lost cause?
Julie Touber – Comment Week 4
ReplyDeleteIdeology (Linz)
“Since democracies base their claim to obedience on the procedural foundations of democratic citizenship, as well as performance, they have a layer of insulation against weak performance not available to most post-totalitarian or authoritarian regimes” (Linz, 49)
What happen when notion of citizenship needs to be discussed (immigration concern) or performance is weaken by an exogenous shock (financial crisis for example: Quid of Iceland?). It does not seem obvious to me that the notion of citizenship is all clear in the democratic context, especially in a globalizing economic context.
Economic Development and Regimes’ Type
Geddes states that single-party regimes are more resilient than “authoritarian/military” regimes (definition is different from Linz) to exogenous shocks such as economic crisis (see p139-140). So why the proportion of dictatorships let by military is higher when there are economic crisis (associated with war scenarios in the long run) as described in Cheibub (fig. 6.1, p149)?
There is something contradictory in the readings in general. It is said that economic development increases the likelihood of democratic politics but also that during economic crisis, authoritarian regimes are more likely to break down (Geddes, 140). So when are authoritarian regimes developing, under what conditions? This seems contradictory to me and I am wondering if it’s not because of “by country” approach. Each country in the statistical analysis that are used (from my understanding) is equal at 1. So when the Soviet block collapsed, there is a large sample of countries whose behavior is similar. But is it the representation of the relationship between authoritarian regimes and economic crisis? It’s not because several cases are similar that they demonstrate this relation, it may be because of their own relation between each other that they behave the same way and not because of the exogenous factor on each one of these countries.
Democracy and Informality
Following the Linz’ model: could we have institutional pluralism –ie. second economy and parallel culture- in a democratic regime?
“post-totalitarian leaders tend to be more bureaucratic and state technocratic than charismatic” (Linz, 47). Are African “democracies” more of post-totalitarian regimes then considering the leaderships and the importance of informal economy and culture?
The problem I have with the classifications that are used (democracy vs. authoritarian regimes) is that the main assumption is that democracy is based on a narrow definition of private property, hence capitalism. Giving the fact that capital accumulation is the reference point to economic development, democracy and economic development will go well together (Przeworski). But values outside private property and capital accumulation and in general values that are not formalized within this context are not captured. The notion of modernization is associated to westernization by Said, Mamdani, Chatterjee for example. Przeworski show that there are links between modernization (economic development) to democracy (at least in the survival of a democratic regime vis-à-vis its economic development). Are we defining a western specificity? What is the point of comparing regimes’ behavior if we use only a western referential?
I agree with several of the other comments that the dichotomous relationship that Przeworski & Limongi and Boix & Stokes (perhaps to a lesser extent) establish between economic development and democratization is problematic for reasons outlined above. Be that as it may, it is not the focus of my comment. Rather I want to introduce a further consideration; a variation over the theme of contextual importance that emerges from some of the comments and constitutes the brunt of criticism in some of them.
ReplyDeleteA contrasting picture emerges from consideration of the aforementioned works. There is an endogenous effect, but it is unclear how strong it is across different periods of time. It had a clear impact on the chances of democratization in the century leading up to 1950, as Boix & Stokes show, but the effect is much smaller after that, whereas the effect of economic development on democratic stability becomes much more pronounced, as Przeworski & Limongi demonstrate. Importantly, Boix & Stokes point out that a thorough reading of Preworski et al’s book shows that even they find a small, but statistically significant, effect of development on the probability of democratic transitions. However, the picture that emerges from consideration of the entire period from 1850 to 1990 is that on average the relative effect of development is stronger in maintaining the stability of democratic regimes and less in inducing transitions.
I think this leads us to a fundamental question. If the effect of given levels of economic development ostensibly has what appear to be significantly different effects, in terms of strength and causality, on democratizations in different historical periods – that is, differences across time and type of authoritarian regimes – could this imply that processes of democratization or transitions to democracy under such varying circumstances indeed are different? Should they, then, be treated as such?
The evidence at least suggests, as Barbara Geddes points out (in “What Causes Democratisation”, Oxford Handbook of Comparative Politics, 2006) that processes of democratization may be different in different historical periods. This surely matters for our understanding of the topic and we should therefore not uncritically believe that one explanation is adequate or that the same explanation is appropriate for different periods and their distinctive contexts.
She puts it this way: “We might make progress faster, both empirically and theoretically, if we identified clear domains for our arguments about the causes of democratization rather than assuming that just because we cover many processes of democratization with one word we should also uncritically model it as one process regardless of what we know about historical and other differences.” (See p. 336 in the same volume). In short, as Robert Franzese has put it, context matters.
In Presidentialism, Parliamentarism and Democracy, Jose Cheibub attempts to refute the Linzian claim that the intrinsic features of presidential systems cause them to have less longevity than parliamentary systems. Instead, Cheibub argues that it is not presidentialism itself that is determining; rather, presidential systems are more fragile because they typically emerged in nations with a strong military presence, which “placed democracy of any kind at risk.” Do we find Cheibub’s claim convincing? Indeed, Cheibub offers no sound empirical test of his claim that “initial institutions …[are adopted] for reasons hat are unrelated to the ones that lead to the occurrence of military dictatorships” (147). Cheibub’s analysis is missing several causal links. While Cheibub describes the military/presidential nexus as an “historical accident,” he fails to explain what may influence a country’s propensity for militarization. Without that analysis, how can he claims that the reasons for militarism are not the same as for institutional development? Further, what causes a militaristic dictatorship to form rather than a civil dictatorship? We can also ask whether Cheibub’s work suffers from the same type of sample selection problem that Stokes and Boix point out in Przeworski and Limongi’s work. As noted in Chapter 4, Cheibub’s dataset is confined to the period 1946-2002. If we took the dataset farther back in time would the results change? Further, how sound is Cheibub’s theory about the importance of the Cold War in fostering the kind of militarism that leads to the breakdown of democratic regimes when most of his dataset consists of the time period of the Cold War? Would his analysis be different if he studied different waves of the Cold War, such as early Cold War, height of Cold War and post Cold War? Indeed, could we make the argument that the international environment at the time of regime formation rather than the presence or absence of a military regime may be more important in effecting the stability of presidential regimes? By confining his analysis primarily to Cold War cases, Cheibub is not differentiating between the importance of each variable. Finally, what can we learn from the Geddes article that may influence our understanding of why military regimes may cause subsequent democracies to be more unstable? Geddes argues that most military regimes end in negotiation as there are many internal disagreements and splits. It could be the case that this “symptom” of militarism rather than militarism itself plays a more important causal role in Cheibub’s analysis.
ReplyDeleteAmy Semet
In Presidentialism, Parliamentarism, and Democracy, Cheibub argues that the high mortality rates of presidential democracies are not due to inherent flaws in their institutional framework as suggested by other scholars. Contrary to arguments proposed by other scholars which claim that parliamentary democracies are more stable because they promote party discipline, create relationships of interdependence between various branches of government, and facilitate inter-branch cooperation, Cheibub argues that the instability present in presidential regimes are due to the environment in which these regimes are created. Specifically, he argues that many presidential regimes were formed in countries where military regimes reigned supreme and that due to the stickiness of these past regimes with their authoritarian tendencies, any form of democracy (presidential or parliamentary) would be unstable.
ReplyDeleteI agree that much of the instability observed in presidential democracies can be attributed to institutional legacies from past regimes. This claim correctly takes into account that once new regimes structures are created, the old institutions very rarely completely disappear. His argument forces an analysis of relevant historical and macro level developments within specific states, which I believe is extremely important.
However, I have several questions. First, what impact does a federal or unitary government structure have on Cheibub’s claims? For example, is there any variation between federal/unitary parliamentary governments or federal/unitary presidential regimes? Also, what is the impact between successive shifts between parliamentarism and presidentialism for regime stability? If one were able to find a case of a state that switched from one form to another, it would provide an interesting opportunity to test (within one environment) the implications of each system on regime stability. It seems that many of the cases that he discusses involve a country transitioning to democracy and then adopting a particular form (presidentialism or parliamentarism) and sticking with it. Although, he does note that when a switch occurs, it is more likely to be from parliamentarism to presidentialism, rather than the reverse. Why is this the case?
Over the last two or three decades, comparative politics transitioned from a focus on regime stability to an emphasis of the study of democratic institutions and their attributes. Linz and Stepan argue that such a shift is necessary because earlier typologies can no longer document the variation in regime characteristics. According to their calculation, 90 percent of non-democratic regimes qualify as authoritarian states. Beginning with Linz and Stepan, an increasing number of studies and indices (Freedom House) also ranks countries on their democratization level. Democracy replaced stability as the gadget of political science.
ReplyDeleteIt remains to be seen whether the paradigm shift towards the study of democracies, dictatorships and the shades of pluralism in-between is not a normative projection of Western liberal values. The expansion of democratization scholarship has gone hand in hand with the establishment of consolidated democracy assistance programs by the United States and other Western donors in the mid 90s. Many of these programs are directly linked to the scholarship on democratization and its triggers. From a policy point a view, it might be premature to use evidence from the social sciences to design and implement foreign aid policy. Boix and Stokes point out that Juan Lopez, a Cuban-American academic, used the exposition of modernization theory by Przeworski and Limongi (PL) to argue for continued economic sanctions to the communist regime in the Carribean. Lopez does not mention that the results from the original study by PL can be hardly applied to single cases because of omitted variable bias, case selection problems and the lack of reliable robustness tests.
To what extent should policymakers tap into current debates on political regimes? Are democracies indeed more stable than other consolidated regime types (e.g. corporatist states)? What recent studies of regime stability can we think of?