Comparative historical analysis (CHA) provides a unique macro-perspective that has allowed political scientists to address major changes in political development over time. The foundational works of CHA addressed the origins of macro-phenomena associated with modernity via explanations based on large structures and processes embedded in states, classes, and the international political economy.Footnote 1 A more recent wave of comparative historical scholarship, using an approach called historical institutionalism (HI), has incorporated methodological individualism into older structural accounts of change.Footnote 2
Historical institutionalism moved CHA away from purely structural explanations and posited that there were critical junctures at which actors choose between alternative historical paths. At such junctures, actors possess autonomy from the constraints of structure and can shift political development in new directions. The advantage of such an approach over traditional CHA is that by incorporating actor preferences into causal explanations one acquires the ability to explain why similar structural conditions sometimes lead to different outcomes. After critical junctures, new rules become institutionalized and actor autonomy contracts due to constraints posed by new structures. This process, known as path dependence, comes to mark a new stable institutional equilibrium. Through a variety of mechanisms, such choices become locked in and actors have difficulty affecting anything other than evolutionary or incremental change.
Despite the continued vitality of CHA, and especially its HI variant, it suffers from a certain stability bias in explaining macro-level phenomena such as regime change. Given that large swaths of the globe periodically suffer from periods of chronic instability and weak rule (witness the situation in the contemporary Middle East or Africa, as well as historical epochs in other regions), it is essential for practitioners of comparative historical analysis to develop a set of tools to understand chronic instability. I argue that HI despite present shortcomings offers just such a possibility with proper modification.
HI uses the dual mechanism of critical junctures and path dependence to model regime change.Footnote 3 Junctures only become “critical” once they exert long-term effects, i.e., they must generate path dependence. But what of junctures that fail to do so? To understand the puzzle of chronic regime instability that motivates this inquiry, we must consider the possibility that when junctures do not establish path dependence they can punctuate periods within sequences of sustained instability. Each new juncture offers the possibility of a new equilibrium, but when it fails to do so, it cannot be understood as “critical.” A string of failures to create path dependence thus provides us with a way to explain chronic instability within the conceptual structures of historical institutionalism.
My purpose here is to explore and theorize about the possibility that institutional settlements made at a series of junctures do not generate sets of institutions that are capable of managing the inherent logic of social cleavage and conflict in society. Actors face serial episodes where they have strong autonomy, but may find that stable institutional outcomes elude them. This condition can be a function of “defective” institutional choice or of a “lose-lose” situation where there is no possible compromise between actors that will yield a stable, effective set of institutions.
The history of political regimes abounds in states that have failed to lock in stable rule over extended periods. Classic examples include Germany from the fall of the Empire in 1918 to reestablishment of stable rule in a divided country in 1949, nineteenth century Spain, or France from the Revolution of 1789 to the establishment of stability in the Third Republic by the passing of the Constitutional Laws of 1875.Footnote 4 In the postwar period and beyond, Asian countries such as South Korea, Thailand, Bangladesh, Turkey, and Pakistan have experienced periods of frequent alternation between military and civilian rule. Similar problems have plagued Argentina and Suriname in South America or Nigeria in Africa, just to name a few examples on other continents. How are we to make sense of such a large number of cases where regime path dependence fails in a chronic fashion over sustained periods? Such cases necessitate a reconsideration of the relationship between structural constraint and actor volition in explaining regime outcomes.
I make a case for adapting the analytic tools of HI to account for this wide range of regime outcomes that seem to fall outside of standard HI explanation. This theory is not a return to the structural determinism of older varieties of CHA, nor does it seek to overturn HI, but to expand the bounds of an actor-centric structural comparative historical analysis to make sense of a large set of important but under-theorized cases. It proceeds in four steps. The first section examines how HI punctuates political development via the sequencing of critical junctures and path dependence. I argue that HI, as well as earlier structural forms of CHA, has a predisposition to seek out and explain situations of stable institutional equilibrium that leaves situations of chronic instability poorly explained. The second section discusses chronic instability and how to understand it in a framework that brings together the outcomes of conflict over institutions between actors and the larger macro-structural environment. The third section outlines a series of causal mechanisms that can help to produce sequences of instability. The final substantive section employs the mechanisms identified in one such episode—regime instability in Germany from collapse of the Empire in 1918 through the founding of the Federal Republic in 1949. In the conclusion, I return to the limitations of HI using path dependence to understand chronic instability, but show how the tools of HI can be adapted to address the problem satisfactorily.
Historical Institutionalism and Developmental Trajectories
HI essentially sees two kinds of institutional change, the disjunctive and the evolutionary. The stability bias in HI is more common to earlier work in the tradition that was concerned with disjunctive change. The recent trend in conceptualizing and explaining evolutionary change is more flexible and less susceptible to charges of being teleological. It has been more commonly deployed to explain the development of meso-level institutions (e.g., policy regimes, welfare states, or systems of interest articulation).Footnote 5 Though the evolutionary tendency in HI has sought to address issues of stability bias, it has not addressed patterns of chronic disjunctive change. Steven Levitsky and Maria Victoria Murillo also point out that both the punctuated equilibrium and evolutionary forms of HI are ill-suited to what they call “serial replacement” of institutions.Footnote 6 Even more nuanced HI explanations of regime change like Giovanni Capoccia and Daniel Ziblatt’s account of the emergence of democracy in Western Europe look at a series of institutional equilibria that over time lead to the construction of democratic regimes in an evolutionary fashion.Footnote 7
Still, with reconceptualization and redeployment the tools of HI are up to the challenge of explaining chronic instability. Wolfgang Streeck and Kathleen Thelen’s typology of the forms of institutional change is useful in characterizing this important difference. While their adaptation of HI incorporates and explains incremental change as “reproduction by adaptation” and “gradual transformation,” this account is concerned in creating an HI framework to explain chronic “breakdown and replacement.”Footnote 8 Similarly, James Mahoney and Thelen, in their framing of a “theory of gradual institutional change” acknowledge more disjunctive forms of change, but their innovation is confined to indentifying and explaining evolutionary paths of institutional change.Footnote 9 The focus here is on a different phenomenon, chronic instability that is conceptualized in this framework as multiple, frequent, and connected episodes of disjunctive change. While the phenomenon under consideration is radically different, the concepts developed both in the earlier and recent generation of HI provide a means to explain it with some adaptation.
The sequencing of disjunctive change in HI corresponds to the model of punctuated equilibrium adapted by Stephen Krasner from evolutionary biology in his discussion of the development of the state.Footnote 10 He argues that institutions are subject to periods of stability until they break down under external pressure. Such external shocks are followed in turn by the creation of new institutions. Thus punctuated equilibrium as a concept captures this pattern where periods of routinized stability follow unexpected abrupt changes in institutional structures.
Kathleen Thelen and Sven Steinmo criticize Krasner for being too mechanistic. In periods of stability, institutional structures seem to have the upper hand over actor volition in determining political outcomes. Once institutions break down, however, actor volition becomes determinant in adopting a new set of institutions. They see the punctuated equilibrium model as lacking in the dynamism necessary “to capture the interplay between the two variables over time” (i.e., preferences and structures).Footnote 11 The combination of critical junctures and path dependence captures this dynamic. Critical junctures capture inflection points in institutional development, when actor preferences are weightier in the determination of outcomes, and path dependence captures periods of stability when institutional lock-in constrains the autonomy afforded to actors in terms of both the ends they can pursue and the means by which they pursue them.Footnote 12
Critical junctures are concentrated periods where actors have a high degree of independence from structural and environmental constraints in pursuing their preferences in institutional design. Though multiple options may be available to actors at each juncture, once a particular set of decisions are made, possibilities are foreclosed and paths of development are locked in.Footnote 13 The logic of critical juncture/path dependence sequences can be productively explored in terms of temporality, causality, agency, and contingency.
In terms of temporality, critical junctures are short lived. Capoccia and Ziblatt use the metaphor of a window that only remains open for so long.Footnote 14 The question of just how long these windows remain open is a difficult question that has not yielded easy answers.Footnote 15 Capoccia and Daniel Kelemen have clarified the issue by suggesting that specification of the unit of analysis (e.g., individual institution, institutional subsystem, or system of institutions) is important and in arguing that critical junctures need to be considerably shorter than the periods of path dependence that follow.Footnote 16 Temporal definition of junctures in the study of regime change is somewhat easier to identify than in other contexts because interregna and forms of provisional rule punctuate disjunctive forms of regime change. Even where regime change is evolutionary, major moments of reform are relatively easy to identify. Whereas critical junctures are measured in weeks, months, and perhaps a small number of years, path dependence must be durable. It is reasonable to think of those regimes whose duration can be measured in decades rather than years as outside the scope of chronic instability.
Causality in historical institutional accounts is linked to contingent decisions made during critical junctures leading to new configurations of institutions characterized by path dependence. For Mahoney such outcomes are deterministic. In contrast, Capoccia and Kelemen think of the outcomes of critical junctures probabilistically. The difference between the two seems attributable to differences in the conception of critical junctures with which they work. Capoccia and Kelemen conceptualize them as junctures in terms of patterns of institutional development, whereas Mahoney does not limit himself to institutional patterns but also in terms of broader “event chains.” Both of these formulations leave room for failure to achieve lock-in, though this goes beyond the scope of both articles (see below for elaboration).
Agency and contingency are linked in critical junctures. As noted, they are periods in which the constraints of structure have weakened and political actors have enhanced autonomy to restructure, overturn, and replace critical systems or sub-systems. Contingency means that more than one outcome exists thus precluding structural determination. This is where agency comes into play. Paths are “chosen” by actors but not cleanly. Different actors have different values and interests, and their “choice” is a product of deliberation, negotiation, conflict, or other means of deciding between potential alternatives.Footnote 17
Critical junctures are “critical” because they generate path dependence. This means that contingency collapses and paths previously open at critical junctures are no longer “available.”Footnote 18 The literature on path dependence gives several reasons why decisions taken at critical junctures yield stable institutional outcomes that come to constrain actors. The economic literature on path dependence talks about efficiencies and sunk costs that are hard to recover if attempts are made to change paths.Footnote 19 The sociological literature pays a great deal of attention to the role of norms in the process of institutional reproduction.Footnote 20 Finally, some explanations of institutional lock-in are power political. Stability is understood in rational terms as a self-enforcing outcome in which the costs of defection for any player outweigh the costs of continued cooperation/compliance.Footnote 21 Some of these accounts move beyond contingent behavioral compliance and talk about how actors embed themselves in the system. Footnote 22
Having discussed how historical institutionalism has explained regime change and stability with the dual mechanism of critical junctures and path dependence, I turn to the problem of chronic instability and how to adapt the tools of HI to explain this type of regime history.
Explaining Regime Instability within a Historical Institutionalist Framework
I address a hole in HI’s ability to account for regime outcomes. Thus, the unit of analysis discussed here is regime. The vast majority of historical institutionalist work examines the establishment of stability, focusing on one critical juncture and the set of path-dependent institutions that emerge from it. Similarly, the earlier CHA literature also had this sort of stability focus, pinpointing key developments that led to long-term changes in political trajectories.Footnote 23 By the very nature of the project, they select cases that produce durable patterns of political change. Thus, a return to a more classically structural form of CHA does not address the problem either. This account focuses on junctures that do not produce path dependence but prolonged periods of regime instability and alternation, something not addressed by either HI or the older structural form of CHA.
Numerous countries have swung back and forth between short periods of democracy and dictatorship over an extended period. Douglas Chalmers described such sequences in Latin America prior to the recent period of regime stability:
The enduring quality of Latin American politics in this century may not be a particular form of regime, but rather the fact of change and the quality of politics in any regime which has only a short history. … Most Latin American countries have … experienced several types of political institutions, with shifting combinations of democratic and authoritarian, federal and centralized, populist and conservative characteristics. These regimes have usually replaced each other dramatically and rapidly in the midst of crisis and confrontation.Footnote 24
Samuel Huntington also identified a similar pattern of regime change:
In the cyclical pattern countries alternated back and forth between democratic and authoritarian systems. … Under a democratic regime radicalism, corruption, and disorder reach unacceptable levels and the military overthrow it. … In due course, however, the coalition supporting the military regime unravels, the military fails to deal effectively with the country’s economic problems … and … the military withdraw from or are pushed out of office. … The alternation of democracy and authoritarianism is the country’s political system.Footnote 25
This pattern where countries experience frequent and rapid regime change is quite familiar to those who have studied Latin America, but would not be alien to those who study many other developing countries (e.g., Turkey, Nigeria, South Korea, Thailand, Bangladesh, and Pakistan, to name but a few). Recent work by Levitsky and Murillo returns to this theme, but like Huntington and Chalmers, the explanation for it is seen as the product of structural or behavioral legacies.Footnote 26 This pattern is recognizable to those who study historical political development in Europe. There was high regime turnover in France from the revolution of 1789 until the stabilization of rule in the Third Republic with the passage of the Constitutional Laws of 1875 and the resolution of the crisis of 1877.Footnote 27 Later in this piece I will discuss a similar pattern of regime cycling in Germany from 1918–1949.
It is important to distinguish chronic instability from simple regime breakdown and replacement. Single episodes of regime change are a common enough occurrence and the tools of conventional HI explain it well. What is of interest here are cases in which path dependence fails to emerge over multiple junctures and yields a prolonged period of regime instability. To incorporate regime instability of this type into an HI framework, it is necessary that its models of causality admit such a possibility. Both approaches, the probabilistic and the causal, would seem to leave open a possibility. In Capoccia and Kelemen’s probabilistic model, this is not a particularly thorny problem. The probabilistic nature of causality means that a failure to create institutional lock-in through contingent choice is a possibility. Instability of this nature would necessitate a series of sequential ineffectual choices. How often such patterns of instability emerge would be a function of the probabilities inherent in making such choices in sequence.
In Mahoney’s causal model, deterministic institutional lock-in does not leave room for long periods of regime instability. However, his discussion of “reactive sequences” holds out the possibility of junctures that do not produce stable institutions. Such sequences entail a chain of causally connected events in which each successive event is “dependent” on the prior. What distinguishes this from mere structural determinism is that it is based on a series of junctures that involve contingency. Such a chain of junctures is marked by what Mahoney calls “inherent sequentiality.”Footnote 28
Mahoney laments that the exploration of inherent sequentiality has not moved beyond description. The next step is to provide causal explanations for such sequences.Footnote 29 I will next explore a series of causal mechanisms that can explain chronic sequences of regime instability.
Causal Mechanisms Giving Rise to Chronic Instability
The following discussion outlines three causal mechanisms with the potential to disrupt institutional lock-in. By mechanism I understand, “a delimited class of events that change relations among specified sets of elements in identical or closely similar ways over a variety of situations.”Footnote 30 They provide explanations for why some junctures do not become critical. Non-critical junctures include missed opportunities that prolong path dependence, whereas others produce change but not lock-in. Daniel Ziblatt captures the former in his discussion of the politics of voting reform in Prussia.Footnote 31 My focus will be on the latter. The first mechanism explored is the well-known problem of external shocks that produce unanticipated changes in structure that may undermine institutions. Second, institutions may fail endogenously because of a mismatch between the institutions chosen and the structure of social conflicts that they were designed to regulate. Third, actors may behave in ways radically different from those anticipated by the framers of institutions also undermining them.
External shocks can disrupt institutionalization before lock-in can take place. The discussion here will move beyond the structural parsimony of Krasner, and take up Thelen and Steinmo’s reservations concerning the absence of an account of actors and their preferences. If we want explanations of structural impossibility to be convincing, we need to show how the impact of structural change confounds the ability of actors to solve the fundamental conflicts that structural change poses.
Structure can also subvert institutionalization endogenously. Thelen captures this in her discussion of the tension between the “logic of change” and “mechanisms of reproduction” of institutions.Footnote 32 Specifically, we need to be cognizant of the reassertion of structural constraints as politics move from the extraordinary to the routine. The reemergence of antecedent structural constraints may make institutional choices problematic or reversible if the conditions that enabled them at the beginning evolve in ways inimical to their lock-in. I explore a specific subset of such problems, so-called “confounding conditions,” where the structural factors that contribute to the opening of junctures also block institutional lock-in. Such confounding conditions would seem to be of particular interest to the problem of chronic instability.
Finally, we need to pay attention to the dynamic development of actors and their standing in relation to each other. The fates of the actors themselves are subject to the vagaries of change. Path-dependent accounts sometimes make assumptions about the fixedness of actors and their orientations following critical junctures. However, as polities move out of junctures, the nature and balance of power between actors may change, or actors may come to understand their interests differently than they did during the juncture. In an age of mass politics, especially competitive politics, the volition of the masses and whom they support are also subject to change.Footnote 33 Both elite actors and their supporters may change their preferences in response to new conditions or supporters may change allegiance from one set of political actors to others. This realm of possibilities makes it critical to integrate coalitional politics into our understanding of institutional stability.Footnote 34 Changes in coalitional politics triggered by a change in the mix of actors in the political field have the potential to make the decisions taken in critical junctures untenable when they undermine the asymmetric power relations that make lock-in possible.
Exogenous Structural Disruption
One of the most common ways regime stability is undermined—external shocks—can also impede path dependence. If a shock quickly follows a juncture, it may block institutional lock-in. The kinds of events that constitute potentially regime destabilizing shocks have been well identified in the literature on regime change and may include international pressure, change in the international political economy, economic crisis, or defeat in war. It is important to note that the use of shocks here goes beyond Krasner’s punctuated equilibrium.Footnote 35 They do not only produce disequilibrating effects in established systems, but potentially disrupt institutional lock-in.
It is important to note that shocks like defeat in war and major economic crises do not inevitably lead to change. As Juan Linz pointed out in his classic discussion of democratic breakdown, regimes sometimes survive crisis and reequilibrate.Footnote 36 Several examples I discuss here include instances of shocks that regimes survived. In the discussion of Theda Skocpol’s analysis of the Russian revolution, defeat in war in 1905 did not spell the end of Tsardom, but did in 1917. Similarly, the hyperinflation experienced by Weimar Germany in the 1920s did not lead to its demise, but its eventual stabilization when the German People’s Party (DVP) was enticed to take a constructive role in the ruling coalition. And in the discussion of Capoccia’s work on how coalitions of democratic parties in Belgium, Czechoslovakia, and Finland, despite facing crisis conditions similar to Weimar Germany, were able to defend democracy by coordinated action. All these examples attest to the fact that though shocks have the potential to destabilize regimes, actors have some leeway to adapt to the new conditions provoked by shocks and preserve them.
Many works of CHA document how external shocks have produced, or perhaps better, contributed to regime change. One such example of widespread regime failure is the breakdown of competitive regimes in South America in the 1960s and 1970s. Guillermo O’Donnell’s theory of bureaucratic authoritarianism attributes this widespread adoption of dictatorship to the end of the easy phase of import substitution industrialization (ISI). As economic growth began to slow, a conflictual politics of austerity emerged, and militaries overthrew civilian leaders. This change was also predicated on a shift of actor configuration and preferences in response to economic conditions. Specifically, O’Donnell pinpoints tension between elite and middle sector interests as growth began to slow, and the creation of a new coup coalition that backed a technocratic project of export based growth in response to the saturation of domestic markets.Footnote 37
A similar logic is detectable in one of the best natural examples of controlled comparison available, Skocpol’s account of Russia in the early twentieth century. The same external shock, defeat in war, destabilized Tsarism in 1905 and 1917. It survived the first revolution but not the second. The difference lay that in 1905 the main two bulwarks of the autocratic regime, the aristocracy and the military, came through the revolutionary events intact. The Tsar was able to withdraw the army from the Far East following the Treaty of Portsmouth and reassert his control. In 1917, the kernel of the nobility, embedded in the traditional officer corps, was decimated by the war and the army, which had begun to disintegrate under the harsher conditions of total war, completely fell apart when its soldier-conscripts abandoned the front lines when peasants began to seize noble land holdings spontaneously. It is the differing fortunes of these two bedrocks of the regime that explains the ability of the Tsarist regime to persist in 1905 and why it crumbled in 1917. Footnote 38
The outcomes of global conflagrations may also affect the stability of the international system and its effects in cementing domestic order. Adam Tooze argues that the effects of World War I in sweeping away the Euro-centric imperial order did not lead to its replacement with a stable form of liberal hegemony. The lack of constraint by a postwar international order in flux opened up the possibility of domestic challenges to the abortive liberalism of the interwar period. The experiments with militarist, fascist, and communist regimes by states facing external challenges in the international system created models for emulation and worked to delegitimize and overthrow regimes inspired by Wilsonian liberalism.Footnote 39 The impact of such shifts in the international order has been shown to be an important determinant in clusters of regime change.Footnote 40 Such effects are less immediate than the impact of loss in war, but are nevertheless consequential.
Two of the examples mentioned are quite useful in illustrating the point that structural factors also play out at the level of actors and thus validates Thelen and Steinmo’s criticism of the underdevelopment of the actor side of Krasner’s insight on the destabilizing nature of external shocks.Footnote 41 Shocks are destabilizing when they undermine the coalitional basis for institutions. Bureaucratic authoritarianism is predicated not only on the end of ISI, but the creation of a coalition supporting the overthrow of competitive institutions. In Skocpol’s account of Russia the fate of groups providing the bulwark of the state’s ability to control lower class unrest explains the success or failure of revolution. Structure is critical in explaining these outcomes, but its impact is mediated by purposive action. I next turn to the question of the endogenous impact of structure following junctures.
Tension between the Logic of Change and Mechanisms of Reproduction: Confounding Conditions
In understanding the junctures that make up sequences of chronic instability Adam Przeworski’s discussion of democratic transitions provides insight. He argues that the durability of new democratic institutions is a product of the fit between contingent institutional choice and political conflict structures.Footnote 42 For purposes of understanding chronic instability we can apply this proposition to encompass regime change generally. In any regime-changing juncture where actors choose a set of institutions incapable of managing their political conflicts, or where no such settlement is possible, this will lead to a failure of institutional lock-in and will not produce path dependence.
Recent work on critical junctures argues that the degree of actor autonomy from structures is more relative, that antecedent conditions may lead to variation in outcomes produced by similar critical juncturesFootnote 43 or that structural conditions promote specific configurations of institutional lock-in.Footnote 44 This new line of argument opens the possibility that the windows of choice afforded by junctures and the actors who exercise them are not fully free of the weight of initial conditions. The expansion of the impact of structure on the outcomes of junctures opens another possibility not explored in this innovative literature—that the persistence of antecedent structural conditions might block institutional lock-in in specific contexts. A subset of the conditions that give rise to junctures may become confounding conditions that prevent new institutions from achieving lock-in.
Dan Slater and Erica Simmons argue that there is less autonomy from structural conditions in critical junctures than conventionally argued. They frame the concept of “critical antecedents” that “shape the choices and changes that emerge during critical junctures.”Footnote 45 They define them “as factors or conditions preceding a critical juncture that combine with causal forces during a critical juncture to produce divergence in outcomes.”Footnote 46 They illustrate such divergence in outcomes using Anna Grzymala-Busse’s account of the political fortunes of communist successor parties in East-Central Europe. Whereas Poland, Hungary, the Czech Republic, and Slovakia had a high degree of similarity in the conditions preceding and in the sequencing of democratization, the latter two did not see an early revival in the fortunes of the communist successor party. Grzymala-Busse explains this as a product of higher levels of political and cultural capital possessed by the successor parties in Poland and Hungary. These different levels of capital serve as conditioning causes that lead to divergent outcomes.Footnote 47
Building on Slater and Simons, Hillel Soiffer distinguishes between two types of causal conditions that are essential to understanding how critical junctures operate. “Permissive conditions” ease the constraints of structure, facilitating the emergence of junctures. He distinguishes them from “productive conditions” which, given the opening of a critical juncture, “produce the outcomes or range of outcomes that are then reproduced over time after the permissive conditions disappear.”Footnote 48 Soiffer illustrates these concepts using Brian Downing’s account of how the increased costs of warfare in Europe in the early modern period had ramifications for the emergence of modern autocracy. These costs posed a challenge to constitutional constraints on monarchic power. This was a permissive condition opening a critical juncture for changes in the way state power was organized. Where self-defense was predicated on the raising of new revenues from domestic sources, this required greater extraction by the state, and promoted absolutism (e.g., France, Prussia). Some states did not have to resort to absolutism because of alternative means for providing for defense—foreign rents, the availability of alliances, geographic barriers, and reliance on emerging modern capital markets (e.g., the Netherlands, Sweden, and England). These constituted productive conditions for the maintenance of limited government.Footnote 49
Soiffer, along with Slater and Simons, argues that subsets of conditions that precede critical junctures play important conditioning roles in the outcomes that follow. Both do so predicated on the notion that small variations in the complex of conditions out of which critical junctures grow can have important ramifications for what sort of institutional lock-in they produce. Where this account needs to go beyond their insights is in showing how the conditions that open junctures can prevent institutional lock-in, specifically how such conditions can be confounding.
What would a confounding condition, a factor that both facilitated the opening of a juncture and prevented institutional lock-in, look like in the context of chronic instability? Within the comparative historical literature, Gregory Luebbert’s account of the fate of democracy in interwar Western Europe, particularly his analysis of how democratizing episodes collapse into fascism, provides an example of this sort of confounding condition.Footnote 50 At the center of Luebbert’s theory is a critical juncture opened by pressures for the incorporation of labor into competitive political systems. When the incorporation of labor and its regulation by the market failed in the previous era of liberal rule, states faced the dilemma of incorporating labor via political means. One solution—social democracy—deepened democracy while the second—fascism—undermined it. Class-based party alliances determined the outcome. In cases where family farmers allied themselves with urban middle-class parties, the result was fascism and where they allied with working-class parties the result was social democracy. The key determinant of this outcome was whether working-class parties had organized landless agricultural labor prior to the juncture, thus putting them in conflict with family farmers. Whereas the rise of organized labor and its attempt to secure political representation was central to the emergence of democracy, wider and more successful organizing on its part also helped to undermine the democratic openings that they sparked. This illustrates the logic of a confounding condition. Where pressures to incorporate subaltern groups into competitive electoral systems by social-democratic and labor parties were broader, they also created conditions that undermined democracy in the subsequent period by threatening family farmers.
The connection between redistributive politics and democratization is rife with potential for the emergence of confounding conditions, because regime change has distributional consequences that are not always foreseeable. Actors often base support for a new set of institutions on distributional assumptions. For instance, one set of actors may expect the system of distribution to remain relatively unchanged while another might expect democracy to address questions of inequality (e.g., income redistribution or land reform). If the distributional consequences of the new institutions are different from the assumptions that actors initially held they may withdraw their support because of unexpected policy outcomes. Such possibilities are inherent in recent work on inequality and regime change, though not explicitly articulated. Distributionist theories of democracy see pressures for redistribution as a prime cause of democratization and democratic breakdown.Footnote 51 If this is the case, pressure for greater equality may undermine dictatorship but if it goes too far, it holds the possibility of undermining democracy. Initial support for or toleration of regime change will not always translate into continued support for the system if new patterns of distribution are inimical to upper class interests. Further, if the distributional consequences of a new set of institutions change the balance of power between political forces, it may disrupt the kind of power asymmetry necessary to maintain institutional lock-in. For example it would be much harder to understand Pinochet’s ability to put together a winning coup-coalition in Chile in 1973 had not Allende’s policies lost the toleration of the Christian-Democrats.Footnote 52 This linkage between actors and structures brings us to the next important mechanism—the changing power and orientation of actors.
The Mutability of Actors, Changing Preferences, and the Balance of Power
Regime altering junctures change the basis of coalitional politics and the balance of power between actors. Institutional framers may not anticipate the future salience of critical issues. They may build institutions for an extinct world or an imagined world that has no chance of fruition. Changes in the structure of opportunities may give rise to new actors or doom existing ones. Actors may grow dissatisfied by the substantive results of regime change and this may be hazardous to the lock-in of new institutions.
In my earlier work on institution-building in new democracies I raised the issue of the strength of what I termed the “initial institution framing coalition” on whether new democratic institutions are durable. Thelen has also examined the weakening of coalitional support for policy initiatives in her work on policy shift.Footnote 53 Weakening of the framing coalition, I argued, had the potential to undermine the democratic system and to promote its failure. I noted five different ways this might come about: (1) the entry into the system of “actors excluded from the institutional decision-making process,” (2) realignment of politics so that actors from the institutional-framing coalition came into conflict, (3) defection of actors from the institution-framing coalition, (4) defection of social forces that had supported the institution framing coalition to other parties or through withdrawal from politics, and (5) the strengthening of anti-system parties.Footnote 54
Given that some change in the configuration of actors and their relative strength is inevitable, why is it disruptive of regime stability in some cases but not others? Certain kinds of change in actor configuration, such as the entry of a range of new actors, enfeebling of actors central to the initial institutional settlement, or defection of constituencies from one set of actors to another, all have the potential to disrupt an institutional equilibrium. When the actors in the system do not find new ways to cooperate under the existing rules, the result can be fatal to the regime.
It is essential to demarcate why changing actor configurations and preferences lead to disruptive as opposed evolutionary change. Here the character and mix of relevant actors seems most important. The question of undermining coalitions of support for institutional settlements is key. Linz raised this in his influential study of democratic breakdown where he highlighted the role of semi-loyal and anti-system actors.Footnote 55 Mahoney and Thelen’s most recent evolutionary work in HI updates this in ways that focus such distinctions on change in institutions. It frames a typology of political agents who are likely to foster institutional change. Two in particular—insurrectionaries, who primarily seek to overturn institutional settlements outside the rules of the game, and subversives, who seek to undermine institutional settlements while playing within the rules, are key in undermining institutional settlements. Their typology enriches our understanding by adding to the mix opportunistic actors, who may engage in bouts of anti-system behavior to further their own interest, and “parasitic symbionts” who weaken institutions by “exploiting them for private gain” at the expense of their efficacy.Footnote 56 When the interplay between institutions and actor interests promotes these forms of behavior, the prospect of gradual change diminishes and the prospect of regime breakdown increases.
Capoccia’s work on the survival of democracies in interwar Europe is instructive in understanding the importance of maintaining coalitions of institutional support in the face of anti-system opposition. Weimar Germany works as a control in his design. It represents a case where democracy collapsed in the face of a triple onslaught by three actors who rejected democracy: the communists, monarchist reactionaries, and the Nazis. Capoccia contrasts this with several other countries that faced structural conditions similar to those that undermined Weimar democracy—economic crisis, highly polarized social cleavages, and an array of aggressive anti-system actors. He shows how actors in Belgium, Czechoslovakia, and Finland were able to defend democracy by adopting a series of policies that buttressed the government and maintained cooperation between the political forces of the center-left and the center-right.Footnote 57
One of the lessons we can draw from Capoccia’s account is that similar configurations of conditions that threatened democracy produced different outcomes based on how actors responded to them. This raises the possibility that the configuration of actors in the polity and their volition holds the potential to mitigate or exacerbate harsh conditions that threaten regime failure. If we are to understand why structure is sometimes determinant, it is important to move beyond the kinds of claims made by older works of CHA on how structural preconditions doom outcomes or of contemporary studies of regime survival using large-n global research designs. Explanations of regime failure such as labor repressive systems of accumulation, insufficient levels of development, or high levels of inequality also need to travel down to the level of actor coalitions.Footnote 58 Structural explanations of regime failure gain full credibility when we can show the ways in which coalitions of regime change are not sustainable as coalitions of institutional reproduction. It requires a deeper level of investigation that situates structurally-induced failure at the level of the actors, thus integrating structure and actor volition.
One way to demonstrate this logic is through exploration of similar sets of conditions that give rise to opposing outcomes. One such fruitful comparison is the constitutional framing process in two multinational interwar Eastern European cases, Poland and Czechoslovakia. In both cases, the elite of one ethnic group played the dominant role in framing the constitutional settlement that created democracy. In Czechoslovakia, the constitution was framed largely by Czech politicians to the exclusion of the Slovaks, Germans, Ruthenians, and Hungarians. Early on the system seemed deadlocked and unlikely to succeed. In response an extraconstitutional institution, a council of the leaders of the five largest Czech and Slovak parties (the “Pětka”) managed to cooperate effectively by “agreeing to agree” for the sake of stable government.Footnote 59 While the exclusion of important ethnic constituencies was clearly detrimental to the chances of establishing stable democracy in both cases, in Czechoslovakia the creation of a second informal institutional settlement by actors made it the region’s only stable democracy.
This contrasts with the failure of interwar Polish democracy. Despite its middle-low level of economic development for interwar Europe, a complex ethnic and religious patchwork, and the absence of self-government for an extensive period of partition (1795 to 1918), Poland reestablished itself as a democratic republic following a protracted period of bargaining involving sharp conflict between the Polish right against the center and left. Due to the unsettled nature of the boundaries of the state, its minority populations remained largely on the sideline during this process.
The first regular election in 1922 created a distribution of parties into four camps. The strongest was the Polish right with almost 40 percent of the seats in parliament. The remainder was almost equally distributed between the center, the left, and the national minorities (about 20 percent each). The structure of cleavages was untenable for the formation of stable governments. In the period from the first election to the democratic breakdown in 1926 (3.5 years) there were five governments. Most of them were of a center-right orientation, but lacked a stable basis of cooperation due to tensions between landlord and peasant elements over land reform. A center-left coalition purely on a Polish-national basis was not tenable because it needed to rely on allies from the minority bloc to constitute a majority. This was met with chauvinistic attacks from the right that made such coalitions untenable for the lynchpin of such a formation, the centrist “Piast” Peasant Party. The constitutional settlement lasted less than six years, when the dominant figure in Polish politics, Józef Piłsudski, made a coup d’état in reaction to the merry-go-round of ineffective parliamentary governments.Footnote 60
The example of short-lived democratic government in interwar Poland helps to understand better how inhospitable structure undermines mechanisms of institutional reproduction. Tensions between large and small landholders and between Poles and ethnic minorities made the formation of stable governments impossible. In this case the asymmetries of power that allowed for the framing of institutions did not last and see to its maintenance. It produced a democratic constitution and one round of elections when the forces of the right, center, and center-left found short-term agreement. However, this institution-framing coalition fell apart when normal politics divided them and the participation of the ethnic minorities changed the balance between them. The lesson of such scenarios of changing actor configuration is that the response of the actors themselves can turn them into fatal regime-ending crises as in Poland or into opportunities to stabilize the system as in Czechoslovakia.
To this point, I have discussed the problem of chronic instability, and have now outlined a set of mechanisms that explain individual instances of failed lock-in following a juncture. Chronic instability however, requires the repetition of sequential junctures that fail to produce institutional lock-in. The next step is to see if the mechanisms outlined here are productive in explaining such an instance of chronic instability.
A Case of Chronic Instability: Germany from Empire to Cold War
The sketch that follows is not meant as a novel explanation of the failure of Germany to sustain democracy in the first half of the twentieth century but rather of its sustained inability to maintain regime stability. The material presented here is a “plausibility probe,”Footnote 61 intended to show that the causal mechanisms outlined earlier are useful in analyzing a well-known sequence of chronic regime instability. It is an exercise in congruence testing that connects the presence of the causal mechanisms specified in my theory to a series of failures to establish path dependence. On its own it cannot conclusively validate the theory but can increase confidence that further testing is warranted and worthwhile.Footnote 62 If analysis of the case of chronic instability in Germany shows that these mechanisms do not provide a credible explanation for a sequence of unstable institutional outcomes, then this would raise doubts ex ante about the theory’s validity.
Prior to its collapse the German Empire provided a period of stability (1871–1918). The state was strongly governed under law in a bureaucratic fashion. It combined elements of monarchy and a modern competitive regime, falling short of standard criteria for democracy. While there were universal male suffrage, competitive elections, and a developed party system, the head of government, the Chancellor, was appointed by the Kaiser and served at his pleasure.
The episode of chronic instability began with the defeat of the Kaiserreich at the end of World War I. In the ensuing revolutionary interregnum partisans of democracy triumphed over the revolutionary left and monarchial reactionaries. The ensuing Weimar Republic (1919–1933) was unstable, conflict ridden, and short-lived. Hampered by strong anti-system forces on the right and left throughout its existence, it was incapable of solving a severe crisis of governance and with the full onset of the Great Depression it succumbed when its elected president handed power to the National Socialists in 1933.
The ensuing Nazi dictatorship (1933–1945) was also short-lived. Its demise was less a product of the contradictions of its internal organization than of its provocation of a global conflagration that led to its demise. This was followed by a period of quadripartite allied occupation (1945–1949) and the founding of two durable states. The Federal Republic of Germany created a successful democratic regime that has persisted until the present day. The German Democratic Republic, a Soviet-type regime, lasted for over forty years until German unification in 1990. The unified state has persisted under the basic law written for the Federal Republic.
The period from 1918 to 1949 has all the hallmarks of chronic instability. It is marked by a revolution, two short-lived unsuccessful regimes, and a period of extended military occupation. It is bracketed on both sides by durable regimes. It is a sustained period of thirty years marked by three different junctures in which new institutions were chosen (1918–1919, 1930–1933, and 1945–1949) which provoked three changes in regime. Stability was restored only by the third. This string of clustered regime changes and their linked character also meet the criteria for inherent sequentiality. The failure of each regime led to subsequent regime choices until the restoration of stable rule in 1949. My final task here will be to establish the plausibility of the causal mechanisms outlined in the previous section to see if they can shed light on this string of failed institutional lock-ins. As will become clear in the discussion that follows, more than one of these causal mechanisms may be at play in any failed juncture. Obviously if more than one is at play, this makes lock-in and regime stability more difficult to establish. The discussion will be organized in terms of the mechanisms rather than through chronological consideration of each juncture.
The first causal mechanism discussed earlier was exogenous structural disruption, which impedes lock-in by bringing about a change in the set of structural conditions under which institutions must persist. Here I examine their role in preventing the institutional lock-in that path dependence requires. In the period of German history under discussion there were at least four events of this nature—the defeat in both World Wars, the hyperinflation of June 1921–January 1924, and the Great Depression. Three of the four played an important role in triggering regime-changing junctures.
The first regime-transforming juncture was caused by the collapse of the German economy and war effort in November 1918 in the face of extensive urban protests and mutinies in the armed forces. This led to the rapid abandonment of the monarchy, and a revolutionary struggle for power between the adherents of a democratic republic and a council form of rule like that of the fledgling Soviet Union. From 1919 to 1921 the adherents of the former, including the main faction of the Social Democratic Party, the left liberals, and the Catholic Centre party took control, promulgated a democratic constitution, and founded the Weimar Republic (with help from traditional elements in the Army).
The total defeat of the Nazi regime in the Second World War was also a decisive external shock that shattered the state and left society in a condition of profound disorganization. Military occupation ensued and competitive political life was rebuilt in the Western zones of occupation that would become the Federal Republic. This juncture would prove to be critical. It produced path-dependent institutional lock-in and restored regime stability to Germany. Curiously, the occupation and division of Germany promoted this institutional lock-in and will be addressed further in the discussion of coalition politics that follows.
As for the two economic shocks that destabilized Weimar, the hyperinflation of the early 1920s did not bring the system down, though it did make for a rocky start to the life of the republic and served to weaken its prospects for survival. The ensuing period from 1924 to 1928 was a less stressful time when the republic was governed by a series of center-right governments, made possible by the inclusion of the previously recalcitrant German People’s Party into the government.Footnote 63 This points out that not all shocks lead to junctures and in this case it proved not to be fatal to democracy. The depression effectively destroyed this, making governance by parliamentary majority impossible and creating another juncture in which the Nazis were able to seize power. The analysis of the impossibility of rule posed by the Great Depression will be addressed further in the discussion of coalition politics that follows. This is in line with my argument that the impact of structural determinants is more plausible if its effect can be demonstrated at the level of political actors.
The second causal mechanism outlined earlier, the tension between the logic of change and mechanisms of reproduction, also helps to explain this extended period of regime instability. In this instance the revolutionary nature of change in the transition from the Empire to the Weimar Republic can be seen as a confounding condition. It constituted a permissive condition for the foundation of democracy because the collapse of the monarchy in November 1918 strengthened the hand of those forces favoring a republic. Because the political forces that backed the old regime were utterly disorganized and demoralized by the collapse, the army, which retained a high degree of organizational integrity, lacked a viable political option that favored a defense of the status quo ante. Given the outright call for revolution by the far left, the army’s best option was to support the Weimar coalition, which backed a democratic constitutional state, rule of law, and protection of private property, while providing enhanced social protections for labor. Having the backing of the army in the battles for the streets (Spartacist Uprising in January 1919, the Bavarian Soviet Republic of 1918–1919, the Ruhr Uprising of March—April 1920) clearly enhanced the probability of a victory by Weimar democracy over the radical left.
Subsequently the revolutionary nature of change worked against institutional lock-in and path dependence. Revolutionary and counterrevolutionary violence had a strongly negative effect on German politics, promoting the polarization of the polity during Weimar.Footnote 64 Further, the welfare concessions that the bourgeois democrats made to the social democrats in order to secure the protection of private property were seen by some elements of capital as onerous once the threat of open revolution had passed.Footnote 65 This was one of the reasons for the abandonment of liberal parties over the course of the republic. While protection of private property played a strongly positive role in the foundation of a democratic republic, the protection of the interests of labor threatened the interests of capital and led to its alienation from the republic once property was no longer threatened. Thus the revolutionary nature of change facilitated the creation of a democratic republic and yet at the same time worked to undermine it once in place.
A second example of a confounding condition is the role of charismatic leadership in the replacement of Weimar by Nazi dictatorship and its role in the subsequent failure of the Third Reich. As we know from Max Weber, crises promote the emergence of charismatic leaders. This is because profound crises challenge and even overturn normative orders, and leaders who have success in providing outcomes seen as mitigating crisis command unprecedented levels of personal allegiance. Charismatic leaders face the additional hurdle of providing ongoing evidence of their charismatic gift.Footnote 66 Hitler continued to provide this evidence via an aggressive, expansionist foreign policy that worked in the short-run until it ran aground by provoking the other major world powers. From this perspective Hitler’s charismatic authority, at first within the context of the crisis of Weimar, facilitated the establishment of Nazi dictatorship, but triggered an external shock that spelled the end of the regime.
In order to substantiate the plausibility of the third causal mechanism discussed earlier—the mutability of actors, their orientations, and relative power—it is necessary to switch the object of analysis to how changing structural conditions affect coalitional politics. Some background on Imperial Germany is necessary. In classic CHA accounts strong attention is paid to the dual class structure of German society. The German aristocracy was able to continue to play an important role in the running of the state under the monarchy and to forge an alliance with elements in the bourgeoisie, especially those linked to heavy industry. This alliance of iron and rye (the staple crop grown on the estates of the Junkers) was cemented with a domestic program of military buildup, tariffs to protect both heavy industry and large-estate agriculture, and the control of the urban and rural working classes.Footnote 67
The pressure to create a democratic republic came from three parties whose strength had been growing in the late imperial period—the Social Democrats representing unionized labor, the Centre Party representing the Catholic minority in a cross-class formation, and the German Democrats representing the liberal middle classes allied in the “Weimar coalition.” In the heady days of the Revolution, when the forces supporting the monarchy were discredited by the failure of the war effort, and the republicans were the only political force that stood in the way of the revolutionary left, the coalition enjoyed widespread support. In the constituent National Assembly they controlled over 75 percent of the mandates. This allowed for the creation of a republic and the promulgation of a democratic constitution.
Once the situation stabilized, the dual nature of German society, split between modern and feudal interests, reemerged. The problem of hyperinflation and the reduction of the threat of a takeover from the revolutionary left quickly cut into support for the coalition. For the rest of the life of the republic, the three parties of the Weimar coalition never took more than fifty percent of the vote. This led to eleven governments in the first four years of the republic. When the economy stabilized in 1924 there were several years of less contentious rule under a series of mostly bourgeois cabinets, based on coalitions of the Centre and liberals with the DVP.
Government fragility was exacerbated by the institutional settlement formulated in the Weimar Constitution, notably a highly proportional system of representation and an ineffective and unresponsive semi-presidential system of government.Footnote 68 With the onset of the Great Depression it became increasingly difficult to form any sort of parliamentary coalition as the anti-system parties of the authoritarian left and right, the communists and Nazis, garnered increasing support. After the election of 1930 there was no obvious coalition of parties who could assemble a majority. Indeed, after the fall of the Mueller government in early 1930, there was not a government that ruled based on a parliamentary majority.Footnote 69 Instead rule was by decree, based on presidential cabinets, which were possible for short periods under Weimar’s semi-presidential system. The system ceased to be democratic, though it continued to hold elections and was governed by law. By the elections of 1932 all hope of restoring democratic equilibrium was lost when the combined strength of the Nazis and Communists in the parliament exceeded fifty percent, thus dooming any prospect of forming a parliamentary government. A little more than a decade after the founding of the republic, the combined strength of the parties that formed the initial institution-framing coalition had fallen to only one-third of the mandates in the legislature. When the feeble reactionary president Hindenburg handed the chancellorship to Hitler in 1933, the Nazis used their position in the state to dismantle what remained of the democratic constitution, ban the other political parties, and establish dictatorship.
Further confirmation of the centrality of coalitional politics is provided by the postwar critical juncture that led to stable democratic rule. Given the devastation of the German society and the economy following World War II and the partition of the country into East and West, the structural conditions for the creation of democracy and institutional lock-in did not seem auspicious. The end of chronic instability in Germany came with a similar set of pro-democratic actors, but this time they were able to manage the context effectively and create a regime that has become a model of stability. Why was a democratic coalition of political parties able to accomplish this only following a disastrous military defeat and economic devastation? First, the defeat discredited the authoritarian right in both its radical Nazi and reactionary militarist versions. Occupation truncated the forms that acceptable German political expression could take. In addition, the creation of a Soviet-type dictatorship in the German Democratic Republic had a demonstration effect. It forced anti-communist populations within the Federal Republic to accept democracy as the best available defense against communism.Footnote 70
Second, it changed the territorial and political cultural basis on which the Germans would try to create a democratic state. The Federal Republic, founded in the region of Germany west of the Elbe, did not have a Junker problem and did not have to cope with the problem posed by the residual power of the aristocracy. These areas now lay under Soviet control in East Germany, or outside of Germany altogether. Furthermore the west had a far less polarized political party structure during the Weimar period. There both the Catholic Centre Party and the Social Democratic Party had more effectively resisted the inroads of radicalism in the 1920s and 1930s. Even in Weimar these regions of the country had a more moderate party spectrum.Footnote 71 Thus the postwar division and occupation of Germany worked to promote the installation of democracy in the western zones of occupation, as well as its institutional lock-in.
The dominant parties in the institution-framing coalition, the Social Democrats and Christian Democrats (with their more conservative allies in Bavaria), were able to compromise and write a highly functional and democratic basic law. These two parties have remained dominant since 1949. They have consistently captured seventy percent of the legislature.Footnote 72 Every government since then was formed by a chancellor from one of these parties, in coalition either with smaller parties or in grand coalitions between them. The initial institution-forming coalition has remained strong and the regime it created has been highly stable, proving strong enough to integrate the states of the former German Democratic Republic in 1990.
The idea of examining structural factors through the prism of coalitional politics in this case helps to understand the impact of structure in a more concrete political fashion than the purely structural theories used in older forms of CHA. The adaptation of the framework used here helps us to understand the causal impacts of structure in a more fine-grained manner that elucidates the reasons for regime failure or institutional persistence through lock-in. This however does not mean that we can ignore structural factors. The failure of the Weimar Republic when looked at from the perspective of coalitional politics seems no less doomed than when observed from the negative economic impact of the depression. It helps to understand why its failure was all but ordained by structure given the configuration of political forces at that time. This stands in contrast to the last juncture discussed, which proved critical and led to path dependence. Despite inhospitable conditions to democracy, the actors were able to find a workable institutional solution.
Beyond Equilibrium: Adapting HI to Contend with Regime Instability
We live in an era where there is great uncertainty about the future of democracy. Freedom House began 2015 with a report entitled “Discarding Democracy: Return to the Iron Fist.” Every year since 2006 more countries have experienced deterioration in their Freedom House rankings than have experienced improvements.Footnote 73 In the period since the Cold War a substantial number of states concentrated in Africa, the Middle East, and in Southern Asia have experienced sustained periods of chronic instability, marked by turmoil, violence, and rapid regime change. This pattern is similar to that experienced by other countries and regions in earlier epochs. Such events are a central problem to our discipline and it is incumbent upon us to develop the tools to analyze and make sense of them.
HI has enriched our understanding of the process of institutional change and stability. However, there are areas where the work to be done in this regard is just beginning. This discussion has identified a relatively common phenomenon, chronic instability, which has been under-theorized and analyzed by both HI and the older forms of CHA out of which it developed. Specifically, the linking of critical junctures and path dependence, which has been fruitful in understanding stable outcomes, requires modification to take account of chronic instability. The assumptions concerning the degree of autonomy of actors from structure in the critical juncture/path dependence model of institutional change and equilibrium does not seem to hold so neatly in understanding cases of chronic instability.
I have pinpointed three sets of causal mechanisms—exogenous shocks, the mutability of coalitional politics, and the disjuncture between mechanisms of change and institutional reproduction—that can explain the failure of a new set of institutions to lock in following junctures. The investigation of these mechanisms has shown that the kinds of structural tensions that open critical junctures may be more intrinsic to the success or failure of path dependence than suggested by previous theories. Here the distinction drawn between Soiffer’s productive conditions and the notion of confounding conditions explored here is quite useful. At the same time this account has highlighted that the kinds of explanations offered by focusing on structure can be improved by looking at the impact of structural factors on the coalitions of actors that support institutions or work to undermine them. The degree of autonomy that actors enjoy may be less fixed from case to case and we need to make assessments of this to verify our assumptions. Careful attention to the politics of institutional coalitions is useful to understanding both why structural conditions are fatal to regimes, and also for understanding why reequilibration is possible in some crisis scenarios.
At a preliminary level, the causal mechanisms outlined exhibit explanatory traction given their utility in explaining the episode of chronic instability explored here—Germany in the first half of the twentieth century. Clearly more work in analyzing sequences of regime instability is necessary to further test and elaborate these mechanisms. It is also likely that exploring other cases of chronic instability will uncover new ways in which the causal mechanisms outlined here interact in producing instability. This can only expand the tools at our disposal to make sense of these episodes, which are common and troubling enough to warrant a better explanatory theory. Further application of this theory may offer insights that may help to improve HI accounts of institutional change, whether or not they conform to the critical juncture/path dependency model.
In a recent account that attempts to critically evaluate the causal insights embedded in HI, Peter Hall argues that the theoretical lenses through which we view the world create assumptions about what we should see.Footnote 74 This account has uncovered such a delimiting lens, that assumptions about how path-dependent equilibria ensue from critical junctures have constrained HI’s ability to think productively about chronic regime instability. Given that chronic instability is a recurrent and troubling problem, this might be seen as a fundamental shortcoming. However, I have also shown that when that assumption is relaxed, HI still possesses a powerful set of tools up to the challenge of explaining chronic instability. Specifically, it can address junctures that do not produce path-dependent institutional lock-in by thinking about how actors embedded in structures fail to create coalitions of support for institutional solutions to the problem of durable order.
