Matthew A. Baum and Philip B.K. Potter, War and Democratic Constraint: How the Public Influences Foreign Policy (Princeton University Press, 2015)
Courtenay R. Conrad and Emily Hencken Ritter, Contentious Compliance: Dissent and Repression under International Human Rights Law (Oxford University Press, 2019).
Seva Gunitsky, Aftershocks: Great Powers and Domestic Reforms in the Twentieth Century (Princeton University Press, 2017)
Helen V. Milner and Dustin Tingley, Sailing the Water's Edge: The Domestic Politics of American Foreign Policy (Princeton University Press, 2015)
Megumi Naoi, Building Legislative Coalitions for Free Trade in Asia: Globalization as Legislation (Cambridge University Press, 2015)
Jessica L.P. Weeks, Dictators at War and Peace (Cornell University Press, 2014)
Jessica Chen Weiss, Powerful Patriots: Nationalist Protest in China's Foreign Relations (Oxford University Press, 2014)
Across several generations of scholarship in international relations (IR), whether a country was a democracy or not stood as the dominant way to understand the role of domestic politics in international affairs. Many scholars concluded that accountability to domestic audiences made democracies’ international behavior distinctive from that of autocracies, whose leaders enjoyed a much freer hand.Footnote 1 Overall, much of the foundational literature on domestic politics in IR—going back to Kant—assumed that democratic leaders were more constrained than autocrats when conducting foreign policy.Footnote 2
But several strands of recent research in IR challenge this view of regime type. The most direct challenges arise from more nuanced studies of domestic institutions themselves, both within democracies and long-neglected authoritarian regimes.Footnote 3 Additionally, a wave of scholarship has renewed the study of leaders and individuals as distinct actors in international relations who can challenge domestic political constraints.Footnote 4 Finally, there has been increased attention to international-level factors that shape how regime type matters in IR over time.Footnote 5
One striking takeaway from this research is that democratic and autocratic leaders can often achieve similar levels of domestic constraint—which in many issue areas results in similar international outcomes—and can do so through mechanisms traditionally associated with the opposite regime type. For example, recent scholarship on authoritarian regimes in IR highlights how authoritarian leaders can be constrained by domestic audiences or even mass publics, while scholarship on democracies in IR now emphasizes how leaders can escape domestic audience constraint.Footnote 6 Recent works also directly challenge a consensus in much post-Cold War scholarship on domestic politics and international relations: that democracies are distinctive in their foreign policy, and on average achieve distinct international outcomes. One manifestation of democratic distinctiveness is a “democratic advantage” across multiple areas of international politics, including war initiation, war fighting, coercive diplomacy, finance, free trade, economic growth, credible commitments, foreign investment, and human rights.Footnote 7 Across a variety of important issue areas in IR, recent research can leave the impression that the line between democracies and autocracies is blurry, and some scholars have used these and related findings to question the continuing relevance of regime type to international behavior.
What should scholars conclude from all this? If regime types can achieve similar international outcomes, is it that regime type does not matter, that domestic institutions vary within regime type, that different regime types reach similar outcomes via different processes, or that leaders aim to escape or encourage domestic constraints by strategically adopting features of another regime type? Despite the individual and collective progress made by the reviewed works, we are left with a basic puzzle: if authoritarian regimes are able to achieve advantageous international outcomes—such as generating audience costs, winning wars, or making credible commitments—by activating domestic constraints, then why don't they do it more often? Similarly, if democratic leaders can escape domestic audience constraints, how large and durable is the democratic advantage? In short, does regime type still matter in IR?
Our answer, which stands in contrast to several of the works we review, is that regime type is still a crucial differentiator in international affairs even if, as recent studies suggest, democratic and autocratic leaders can sometimes be similarly constrained by domestic audiences and thus achieve similar international outcomes. For clarity, we focus on one component of regime type that has been particularly salient in IR—the degree to which domestic audiences constrain leader action.Footnote 8 Regime type—or, more specifically, domestic political institutions that govern the interactions between leaders, elites, and the mass public—provide all governments with a default level of domestic audience constraint that is generally higher in democracies, but leaders maintain some agency within these institutions and can deliberately increase their exposure to or insulation from this constraint. Thus we argue that there are both structural, institutionally based and strategic, leader-driven determinants of domestic audience constraint. Few scholars examine both sources, but we argue that both are essential to fully understand domestic audience constraints across many issue areas in IR.
Inspired by the books we review, we propose a new theoretical framework, which we call a “malleable constraints” approach, aimed at recapturing regime type in IR. The framework brings together several ongoing research agendas that should be in greater dialogue with one another—particularly research on democracies and autocracies in IR, two concentrated areas of research that have had surprisingly little interaction. Our approach emphasizes not only variation in the default levels of constraint across the regime type spectrum (domestic structure), but also the boundaries set by those institutions, within which leaders can move strategically (leader agency), as well as the ways in which the international environment can influence both dimensions (international pressure). Institutions define default levels of audience constraint that are typically greater for democracies than autocracies, and which may or may not be internationally advantageous in specific settings. Within these boundaries, leaders have what we call “agency space,” within which they can voluntarily incur “agency costs” to become more exposed to or insulated from domestic audience constraints. Agency space allows for the possibility that autocrats can become temporarily more constrained and democrats less constrained without changing regime type, and shifts the focus to how agency costs manifest, and how risks associated with utilizing agency space vary across the regime type spectrum. We speculate that the risk of triggering institutional or regime change in the process of trying to adjust constraints is higher for autocracies because democracies retain an advantage in institutional plasticity and in flexibility for leaders to adjust the size, composition, or information levels of their audience.
This framework reconciles many competing claims in recent scholarship, including the puzzle of why autocracies can but only rarely increase constraint. Just because autocracies can engage audience constraints and democracies can escape them does not mean that they can do so with equal ease, frequency, or risk. Regime type shapes not only the potential constraints imposed on leaders by domestic audiences for particular international actions, but also the costs for taking advantage of agency space within the boundaries set by domestic political institutions.
Domestic Politics and International Relations: Beyond the Democracy “Dummy”
For decades, a dominant view of domestic politics in international relations was that they were largely irrelevant. One of Kenneth Waltz's critiques of “reductionist” domestic-level explanations was that they could not explain how dissimilar states behave similarly under the pressures of the international environment.Footnote 9 Although scholarship on domestic institutions and international relations began long before Waltz, his perspective, as well as the end of the Cold War, spurred an “institutionalist” response as scholars opened up the black box of the state and identified distinctive international behavior by democracies.
Four features characterized this post-Cold War wave of scholarship on regime type in IR. First, much of the first generation implicitly assumed a simple and often dichotomous relationship between regime type and domestic audience constraint, with regime type serving as a proxy for whether a leader was exposed to or insulated from domestic audiences. The upshot—even when first-generation scholars used continuous measures of democracy—was often a blunt distinction between democracies and autocracies, that is, a real or proverbial “democracy dummy.”Footnote 10
Second, and relatedly, first-generation arguments about democracies’ distinctiveness usually focused on their greater degree of domestic institutional constraint. Such arguments suggested, either implicitly or explicitly, that autocracies were much less constrained (if at all), and therefore had more flexibility in international affairs.Footnote 11 One important manifestation of democratic distinctiveness arguments (but by no means the only one) is the idea that greater domestic constraint leads to a “democratic advantage.” Security scholars, for example, have explored a democratic advantage in crisis bargaining and war.Footnote 12 International Political Economy (IPE) has been centrally concerned with explaining variation in openness to the international economy.Footnote 13 By this openness yardstick, scholars concluded that democracies are distinctive, and sometimes have an advantage in maintaining sovereign credit, attracting foreign direct investment (FDI), liberalizing international trade, and respecting central bank independence.Footnote 14 Many arguments rely on democratic institutions or the preferences of democratic audiences as constraining forces. For example, Schultz and Weingast argue that the “democratic advantage” in sovereign borrowing stems from democratic mechanisms to sanction leaders who default, whereas nondemocratic leaders have more “discretion” to default “unilaterally.”Footnote 15 Democracies have similarly been seen as more likely to respect liberal norms, comply with international treaties, and respect human rights.Footnote 16 Of course, scholars have also suggested that democracies can suffer disadvantages: for example, in so-called small wars, either because the public will not support the effort to win, or because it is shielded from the costs of war, sometimes leading to aggression, poor strategy, or prolonged war.Footnote 17
Third, first-generation scholarship contemplated some strategic manipulation of audience constraint, but focused primarily on the democratic side of the spectrum.Footnote 18 In the context of coercive diplomacy, the concept of “audience costs” was an important special case of this kind of strategic manipulation of domestic constraint, with leaders attempting to tie their hands by increasing domestic constraint.Footnote 19 IPE scholars often considered the opposite: insulation of policymaking from domestic audiences, for example, through delegation of trade policy to the executive or bureaucracy.Footnote 20
A final feature of first-generation regime type scholarship is that with some exceptions, many tended to treat the system level as background structure, ignoring interactions between the international level and regime type.Footnote 21 Even in IPE, which often focuses on institutions or domestic cleavages defined in relation to the international economy,Footnote 22 some have argued that the preference aggregation approach of “open economy politics” (OEP) examines individuals and institutions at the expense of systemic factors.Footnote 23
The more recent, second generation of research on domestic politics and IR has challenged the sharp dividing line between democracies and autocracies, although thus far there has been little dialogue between those working on the autocratic and democratic sides of the spectrum. A particularly important development is that rather than treating autocracies as a residual category for all nondemocracies, this research disaggregates institutions and constraints within regime types, showing gradations of constraint between types of authoritarian regimes and among democracies. As a result, a slew of new findings not only suggests that there are important differences within and between the traditional categories of democracy and autocracy, but also shows that under some conditions, democracies and certain authoritarian regimes can achieve similar international outcomes via similar levels of constraint.Footnote 24 These second-generation scholars therefore return to some of the insights of Waltz's neorealism—namely, the similarity of outcomes across dissimilar units—but they use a more nuanced form of “reductionism.”
At the same time, some recent research advocates for greater attention to how international-level factors can moderate the effect of regime type, but in far richer ways than Waltz's sparse characterization, or even those of his neoliberal institutionalist critics.Footnote 25 Recent scholarship on international organizations, for example, has examined how domestic accountability structures can be influenced by international actors, leading to both real and superficial changes in areas like election quality or respect for human rights.Footnote 26 Similarly, recent IPE scholarship underscores that international factors like global capital liquidity, pressure from the International Monetary Fund on domestic spending choices, or the diffusion of practices like bilateral investment treaties (BITs) can change how regime type matters.Footnote 27
These second-generation studies map broadly into five categories. First, some studies show that autocracies can sometimes achieve similar levels of domestic audience constraint, and ultimately, similar international outcomes to those typically attributed to democracies. Work in this vein emphasizes domestic constraint and accountability in authoritarian regimes. This Autocracies in IR (AIR) category is exemplified by Weeks's Dictators at War and Peace and Weiss's Powerful Patriots, which explore authoritarian crisis and conflict behavior. In a second category, some studies show that democracies’ domestic constraints vary or are more limited, that some democracies or democratic leaders can gain flexibility by evading or maneuvering around audience constraints, and that some democracies can behave quite similarly to autocracies. This Democracies in IR (DIR) category is exemplified by Baum and Potter's War and Democratic Constraint, which disaggregates types of democracies by their media access and party structures. Milner and Tingley's Sailing the Water's Edge, which argues that US presidents sometimes choose foreign policy instruments to avoid domestic constraints, also fits in this category, as does Naoi's argument in Building Legislative Coalitions for Free Trade in Asia. Contrary to arguments that democratic leaders are beholden to protectionist interest groups, Naoi argues that they can instead build pro-liberalization coalitions through regular legislative politics and side payments to legislators who might otherwise push for protectionism.Footnote 28 Other studies go further, arguing that democratic constraint regularly leads to undesirable international outcomes.Footnote 29
In the third category of scholarship, international-level variables influence regime type and/or incentives to use domestic politics for foreign advantage. We label this category SIR after Peter Gourevitch's seminal 1978 article, “The Second-Image Reversed.” Within recent SIR scholarship, Gunitsky's Aftershocks, for example, shows how “hegemonic shocks” can lead states to adopt particular regime characteristics or face coercion to do so, in ways that can moderate democratic distinctiveness at the system level. Conrad and Ritter's Contentious Compliance examines how international law changes the calculus of leaders—across regime type—about whether to repress or otherwise insulate from domestic audiences.
In this article, we focus on the first three categories, but we note the contributions of two other relevant categories. A fourth cluster of studies explain similarity in policy choices or international outcomes using state characteristics that cut across regime type, such as civil-military relations.Footnote 30 Fifth are studies that question the importance of regime type on primarily empirical grounds, such as the posited democratic advantage in generating audience costs,Footnote 31 or the democratic disadvantage in counterinsurgency wars.Footnote 32
Structural and Strategic Sources of Domestic Audience Constraint
Across the AIR, DIR, and SIR literatures, second-generation scholarship finds greater variation in constraint, within and across regime types, than first-generation scholarship. Our discussion of domestic audience constraint follows many scholars who have used a broad concept of “domestic audiences” to indicate actors or groups who can observe a government or leader's actions and have the potential to impose constraints across a range of issue areas. This concept is broader than the usage associated with “audience costs” in two ways.Footnote 33 First, it includes many different types of domestic actors. As even Fearon highlighted, “relevant domestic audiences have included kings, rival ministers, opposition politicians, Senate committees, politburos, and, since the mid-nineteenth-century, mass publics informed by mass media.”Footnote 34 Second, domestic audience constraint applies to a more general set of issues than the narrower meaning of “audience costs,” that is, backing down from a threat in crisis bargaining and coercive diplomacy. For example, in Dictators at War and Peace, Weeks uses a broader concept of “audience” to explore autocratic elite constraints across several aspects of war.Footnote 35 Our use of “domestic audience” echoes E.E. Schattschneider's argument that the outcome of a political conflict depends on “the extent to which the audience becomes involved in it,” and thus politicians seek to manage audience participation.Footnote 36 Additionally, our framework for domestic audience constraint is relevant to arguments about how different regimes respond to domestic audiences when making foreign policy, as well as arguments about whether different regime types tend to achieve different international outcomes, one example of which is a “democratic advantage” on a given issue.
Theoretically and empirically, second-generation research on both autocracies and democracies delves deeply into various parts of a basic accountability chain necessary for domestic audience constraint in any regime type. Accountability requires that a leader face a domestic audience, and that the members of this audience have the power to impose costs on leaders, preferences that may deviate from those of the leader, information to judge what leaders are doing, and enough concern about the issue for it to be politically relevant.Footnote 37 Audience constraint can occur ex ante, through the costs of adjusting policy to suit an audience in ways that move a leader further from her ideal point or that may degrade a policy's effectiveness, or ex post, through accountability and possibly punishment for a leader's decisions.
We argue that much second-generation scholarship explores one of two distinct sources of constraint: structural or strategic, as Table 1 illustrates.Footnote 38 Structural sources of constraint are relatively durable domestic institutional forces that set a default level and range of constraint associated with a given regime type. Strategic sources of constraint include actions leaders take, within this institutionally defined range, to strategically alter the size, composition, information, or political activation of relevant domestic audiences.
Table 1. Categorizing reviewed books
The books we review here, put in dialogue with one another, show that both structural and strategic factors affect various steps along the chain of accountability. For example, audience size and composition are important elements of constraint and can range from the extremely constricted, in personalist regimes, to very large and heterogeneous, in consolidated democracies.Footnote 39 Audience size and composition depend in part on the structure of domestic institutions, but leaders can also work strategically to change the coalition, or to exclude or circumvent segments of an existing audience. Similarly, as Baum and Potter argue, structural factors like the number of opposition parties and the diversity and robustness of a free press affect how well-informed an audience is about its leader's international behavior and whether that behavior deviates from audience preferences. But leaders within a single country can strategically tighten or loosen information flow, which can, in turn, affect audience information levels, or even effective audience size, if participants are cut out of the loop or are rationally ignorant.Footnote 40
Table 1 organizes the books we review according to (1) their focus on democracies, autocracies, or the international level (in the arrows) and (2) structural versus strategic sources of constraint. On the AIR side of Table 1, although both Weeks and Weiss focus on how autocratic leaders can be constrained by domestic audiences, they differ on the structural/strategic dimension. In the upper-left quadrant, Weeks explores how structural variation across autocratic audience constraints in authoritarian regimes shapes conflict behavior. She argues that two dimensions—the degree to which authority is centralized in the hands of a personalist dictator, and whether ruling elites are civilian or military—shape both the preferences and power of domestic audiences that can constrain authoritarian leaders. In contrast, in the lower-left quadrant, Weiss looks at variation in audience constraint over time within a single authoritarian regime—China—where ruling elites selectively allow nationalist, antiforeign protests to proceed. She argues that these protests, which run the risk of getting out of hand and threatening the regime, can increase the credibility of Chinese threats in China's disputes with other countries, such as Japan and the United States.
On the DIR side of Table 1, the upper-right quadrant includes studies that disaggregate structural democratic constraints, exemplified by Baum and Potter's book, which arrays democracies according to the number of parties and access to media.Footnote 41 The lower-right includes arguments about how democratic leaders within single countries can strategically manage their domestic audience constraint to achieve preferred policies. This quadrant includes Naoi's argument about how some democratic leaders in Asia use side payments to legislators to smooth the path of trade liberalization by evading protectionist audience constraints, keeping trade in the legislative arena rather than insulated from electoral constraints as in prior work on protectionism. Also in this quadrant, Milner and Tingley's book argues that US presidents strategically choose foreign policy tools based on the distributional and ideological nature of the issue or policy instrument. One implication is that presidents may seek to militarize policy, since military tools are less distributional and ideological and thus carry lower domestic audience constraint.
The SIR category, depicted in arrows, also suggests a strategic/structural divide at the international level. Gunitsky's book examines how long-term structural changes that shift international incentives for states to adopt particular institutional forms (like democracy) affect the audiences that constrain leaders. But as Conrad and Ritter argue, international institutions like international human rights treaties can alter whether leaders have strategic incentives to engage in human rights abuses as a way to insulate themselves from domestic pressure.Footnote 42
These books and related work provide more nuanced accounts of how domestic audience constraints vary within regime type. But particularly when they make comparisons across regime type, many suggest similarities in constraint between democracies and autocracies. For example, Weeks states that “surprisingly, many autocratic leaders face a realistic possibility of punishment by a civilian domestic audience; they confront many of the same domestic pressures as democratic leaders, only in a different guise.”Footnote 43 Weiss concludes, “if antiforeign protests are credible signals in international bargaining, then it is not necessarily the case that democracies have the advantage in utilizing domestic politics as diplomatic leverage. If both regime types utilize public opinion to reveal information and communicate credibly, other factors must explain the democratic peace.”Footnote 44 Baum and Potter argue that “with information comes democratic constraint. Without it, democracies are in some important regards functionally equivalent to autocracies.”Footnote 45
But as our framework makes clear, observed similarities in domestic constraints or similarity in international outcomes can come via different paths and by paying very different costs. The authors just quoted each trace how democratic or autocratic leaders incur domestic costs associated with their international policy choices. These costs can sometimes be hard to observe but are crucial for assessing whether regime type still matters.
The important findings of second-generation scholarship leave unresolved puzzles, especially on the autocratic side. Although the second generation clearly shows that autocrats can generate constraint, scholars have made strides in studying how autocracies invest in repression, and while the record is imperfect, democracies tend to respect human rights more.Footnote 46 Autocracies regularly forgo international benefits to maintain their existing levels of repression, as when Myanmar refused humanitarian aid following Cyclone Nargis or when Robert Mugabe continued repressive tactics and election manipulation in Zimbabwe, even at the cost of significant reductions in Western foreign aid and expulsion from international organizations. Why do autocracies not take advantage of constraints for international gain more regularly? Is it possible to reconcile the findings of the first and second generations?
Malleable Constraints and Agency Space: Accountability Across Regime Type
We develop a “malleable constraints” framework that integrates the durable structural differences between democracies and autocracies with the shorter-term malleability of audience constraints. As discussed, we assume that accountability requires a domestic audience with the power to impose costs on leaders, preferences that may deviate from those of the leader, relevant information, and enough concern about the issue for it to be politically salient. Changes across this chain affect accountability.
We argue that democratic distinctiveness may stem not only from higher default levels of audience constraints in democracies but also from greater flexibility and lower risks to take advantage of the range for strategic action defined by institutions—a range we term agency space. Thus, similarity in constraints or outcomes across regime types may sometimes be a result of short-term leader agency to intervene in the chain of accountability and is not necessarily evidence that regime type does not matter. The framework thus helps reconcile first- and second-generation scholarship.
Integrating Second-Generation Insights
As a heuristic, Figure 1 depicts the progress and limitations of second-generation scholarship and serves as a jumping-off point for our “malleable constraints” approach. The figure shows the overall degree of domestic audience constraint (i.e., exposure to or insulation from domestic audiences) on the y axis as a function of institutional or structural arrangements at each point on the regime type spectrum, represented on the x axis. The endpoints of the x axis can be thought of as full autocracy on the far left, and full democracy on the far right. The end points of the y axis represent an executive who is fully exposed to a domestic audience (e.g., with extreme values across the accountability chain), or fully insulated (e.g., a leader who is completely immune from any domestic audience constraint).
Figure 1. Strategic and structural sources of audience constraint
The dotted line shows how these institutional arrangements define the default level of domestic audience constraint which, as in the first generation, is higher for democracies, but as the second generation has shown, contains more variation in levels of domestic constraint within the broader regime type categories (i.e., along the x axis). The two surrounding lines represent boundaries of agency space for a given regime type. These boundaries are also a function of domestic political institutions. Thus, for any default position on the dotted line there is a range of agency space defined by the maximum potential audience exposure (top line) and insulation (bottom line) beyond which leaders cannot move without changing regime type. For simplicity, the lines have the same slope. Democracies are characterized by higher default levels of constraint and relatively more agency space in the “exposure” range (where y > 0). Autocracies are characterized by relatively more agency space in the “insulation” range (where y < 0). The gray zone in Figure 1 also captures another feature of much second-generation scholarship: the possibility that leaders of different regime types can experience similar levels of domestic audience constraint, which can, in turn, lead to similar international outcomes.
To understand how a leader is constrained by domestic audiences at a given time, one needs to know three things: (1) which country, that is, the particular set of relatively fixed and nonmalleable set of institutions along the x axis that define the default level of constraint (the dotted line) and the limits on agency space (the solid lines); (2) whether the leader has decided to adjust those constraints strategically within this agency space, that is, move up or down (the vectors); and (3) what the costs/risks of vertical movement entail (not depicted graphically but discussed later). Some leaders may rarely move from their default constraints, remaining on or near the dotted line. Others may move up or down, increasing or decreasing domestic audience constraints and paying some agency cost, all without changing regime type. Moving beyond the top or bottom lines triggers institutional change.
Figure 1 highlights several contributions of second-generation scholarship. The regime type disaggregation that has been a hallmark of this work can be illustrated by comparing different points along the x axis. On the left side of the x axis are authoritarian regimes. Personalist dictatorships are on the far left (a); “constrained” authoritarian regimes, where, as Weeks argues, elites can exert accountability over leaders, are to the right within the authoritarian category (b).Footnote 47 In the center, hybrid regime types combine some institutional features of both democracy and autocracy.Footnote 48 On the right side, comparisons between forms of democracies—such as differences in party or media structures, as Baum and Potter describe, or the representativeness of the electoral system—can be represented by moving along the x axis (c and d).
There have been large gains in AIR scholarship in developing structural arguments about how domestic audiences constrain the international behavior of different types of autocratic leaders, in essence comparing different points along the x axis within the broader “autocracy” category. Scholars of conflict have been at the forefront of this effort (perhaps because of an empirical democratic bias in many IPE issues),Footnote 49 although IPE is now tackling variation in types of institutionalized authoritarian constraints. Drawing on trends in comparative politics,Footnote 50 a common finding across several issue areas is that civilian dictatorships (usually single-party) often achieve similar outcomes to democracies because civilian dictatorships face a domestic audience with a clear and relatively fixed ability to impose constraints.Footnote 51 Many arguments rest on audience size and leader and audience preferences, given an autocratic leader's need to assuage or avoid punishment by party elites.Footnote 52 Further along the accountability chain, several scholars have examined information and transparency in autocracies. Some suggest that autocracies with a relatively large circle of elites that could threaten the leader have more incentives to disclose information about their policies to keep those elites on board, and that institutions like legislatures and parties within authoritarian regimes help facilitate transparency and alleviate monitoring and commitment problems.Footnote 53 On the DIR side, party structure and other institutional audience features can differentiate democracies in international conflict.Footnote 54
Although regime type provides all leaders a default level of constraint, they can strategically generate more or less audience constraint in the short term if they are willing to bear the costs. Such strategic movements reflect another contribution of second-generation scholarship: deeper understanding of strategic sources of constraint, not only in democracies, but also in autocracies. In Figure 1, more negative values on the y axis indicate that a leader is more insulated from domestic audiences. Downward movement from the default could, for example, reflect democratic leaders keeping information hidden or secret from relevant domestic audiences.Footnote 55 Actions in the range y > 0 are traditionally associated with democracies, like allowing public protests, opposition party criticism, or going public to activate domestic audiences. Within the range of agency space associated with y < 0, insulating actions include suppressing information, intimidation of political opponents, and surveillance.
Scholars in both IR and comparative politics increasingly see tactics like information disclosure or even the expansion of political participation as tools of authoritarian control or resilience.Footnote 56 Autocrats can increase their exposure to domestic audiences (vertical movement in Figure 1), as in Weiss's argument that China strategically allows nationalist protest to try to tie its own hands in international bargaining. Such exposure can change the size of the audience, its distribution of preferences by allowing more hawkish voices to participate, and the degree to which the relevant audience is well informed. Autocrats can also strategically use emigration policy to select a more loyal population.Footnote 57 Among democracies, second-generation DIR scholarship inverts prior contributions in security and IPE. The security literature has recently emphasized how democratic leaders can insulate themselves from domestic audiences and take advantage of their room for maneuver,Footnote 58 while the IPE literature is grappling with popular resistance and backlash to insulating institutions as well as how politicians address or even try to activate anger at these institutions.Footnote 59 Both trends help scholars understand “regular” DIR politics in new ways, in part by recognizing the room for maneuver within agency space. Naoi's book is one example, illustrating how trade policy in Asia is made through open, everyday legislative politics as long as side payments can grease the wheels in favor of liberalization, allowing leaders to avoid protectionist domestic constraints. Similarly, audience costs in the coercive diplomacy setting can be understood as one manifestation of a larger phenomenon of democratic leaders taking advantage of agency space, deliberately increasing or decreasing their exposure to domestic audiences for international gain.
The vectors in Figure 1 illustrate four cases of strategic movement within agency space in which leaders pay a cost and/or bear risk to move up or down. Leaders of more autocratic countries, like personalist dictatorships (a), have further to travel to reach the same level of constraint than more constrained autocracies (b), a difference implied by Weeks's argument, with Weiss's argument encapsulated by b’s vertical movement. Likewise, following Baum and Potter, democracies with fewer parties and media access (c) are less constrained than those with more parties and robust media (d), who would, by implication, have to do more to insulate from relevant domestic audiences (i.e., move downward). Agency costs can manifest ex ante or ex post. Some leaders may face small ex ante agency costs (for example, personalist dictators who face no institutional checks on loosening constraints or changing policy), but large ex post risks and potential costs (e.g., the much larger risk of a coup once a personalist dictator temporarily increases the circle of decision making). Although the vectors are useful in depicting agency costs—for example, for a given regime type, agency costs are larger for moving further from the default level of constraint—the costs may manifest very differently across regimes, and so Figure 1 does not fully depict these costs and risks, a topic we discuss later.
Comparing Across Regime Type: Similar Constraint Via Dissimilar Mechanisms
This discussion illustrates an important overarching contribution of second-generation scholarship: the chain of accountability for international choices is relevant across the full regime type spectrum, even for the most authoritarian states. This point may seem obvious to those who are familiar with this research, but it is easy to forget how much these findings—exemplified in the books we review—have shifted views about regime type. As Gourevitch summarized the conventional wisdom in 2002, first-generation scholarship concluded that “public accountability in a democracy limits the range of likely behaviors that happen in an autocracy, where the rulers have fewer immediate constraints” and goes on to say that “[authoritarian] leaders appear to have substantial discretion, despite the constraints of a selectorate. They can shift institutions and their selectorate with greater ease than in a democracy, which leaves a more fluid game.”Footnote 60
The gray zone in Figure 1 also illustrates an important aspect of second-generation scholarship, which suggests that all leaders can reach at least some level of domestic audience constraint that is also accessible to leaders of the opposite regime type. The solid lines indicate limits on y axis movement such that the maximum amount of exposure available to the most autocratic regime and the maximum amount of audience insulation available to the most democratic regime define an area of overlap accessible to all leaders. Domestic constraint is reachable even for personalist dictators who are highly insulated from domestic audiences by default, as the far left side of Figure 1 shows. For example, in her study of dictatorships and military effectiveness, Talmadge shows that in the face of extreme external threat, even personalists like Saddam Hussein can loosen their coup-proofing constraints, become more exposed to domestic audiences (analogous to vector a), and improve military practices dramatically.Footnote 61
Overall, Figure 1 makes clear the nature of the second generation's progress—not only in comparisons between different points on the x axis (structural sources of constraint), but also in vertical movement (strategic sources of constraint), both of which can lead to temporary similarity in constraint across regime type, and in turn, similar international outcomes. When we make more specific comparisons between regime types that are not at the extremes, there are larger zones of potential overlap (larger gray zone), as in Figure 2.1 where “constrained” autocracies (b) have access to more democracy-like exposure, or Figure 2.2 where an electoral democracy (c), for example, has the option to activate more autocracy-like insulation.
Figure 2. Regime type-specific comparisons
But the figures make clear that just because leaders can temporarily reach a similar level of constraint does not mean that regime type is irrelevant. The structural and strategic determinants of audience constraint might vary with issues, occur under only some circumstances, or be limited for certain regime types. Talmadge's argument illustrates the latter point: although personalist dictators can increase audience constraint, they can do so only to a limited extent (as indicated by the upper bound at point a), at potentially high ex post costs. As a result, personalists like Saddam Hussein are willing to loosen coup-proofing constraints on their militaries only when the enemy is at the “palace gates,” that is, the danger from external enemies trumps internal threats from coups.Footnote 62 Once Saddam reluctantly eased domestic constraints on the military, Iraq's army achieved significant gains, such as the ability to conduct complex, combined-arms operations, in part by becoming more exposed to domestic audience constraint.Footnote 63 This strategic adjustment of constraints did not make Iraq more democratic, nor did Iraq achieve the level of military effectiveness of democracies like the United States or Israel (since as Figure 1 shows, personalists can access only limited space in the “exposure” zone). Analogously, democratic leaders do not automatically change regime type when they withhold or suppress information, or even take more extreme measures like the suspension of habeas corpus during the US Civil War, anticommunist purges during the Cold War, or post-9/11 infringements on civil liberties. Instead, democratic leaders willing to pay the associated agency costs can temporarily reach a level of audience constraint that is more insulated than their default constraint and approaches the most exposed a dictator can ever reasonably achieve.
Overall, our malleable constraints framework departs from first-generation literature, which tended to assume that democracies were more constrained than autocracies, or saw autocratic constraints as smaller and relatively fixed.Footnote 64 But it also departs from the implication of second-generation arguments that similarity in international outcomes undermines the democratic advantage or diminishes the relevance of regime type in IR.
Alternative Sources of Democratic Distinctiveness and an Agenda for Future Research
Although Figure 1 provides important context for understanding the continued importance of regime type, especially the greater agency space in the exposure zone, we contend that several other features of democracies and autocracies, not shown on Figure 1 and not yet fully developed or integrated in the literature, are necessary to fully understand regime type in IR. Crucial to our revisionist view of democratic distinctiveness is the flexibility afforded to democratic leaders to utilize agency space. We posit that this flexibility can stem from at least three sources: the nature of agency costs, the resilience or “plasticity” of domestic institutions and the larger set of options democratic leaders have to alter audience features, and international pressure. Some of our conjectures may lead to countervailing predictions or require future research to explore how they affect international outcomes. We highlight the continued distinctiveness of democracy and urge caution in concluding that autocratic leaders can regularly or easily achieve democracy-like levels of audience constraint.
The Nature of Agency Costs
We already noted that second-generation scholarship either explicitly or implicitly describes the differing costs that regimes pay to approximate similar levels of audience constraint. It is not only the magnitude of agency costs but, critically, how these costs manifest, that determine when and whether leaders are likely to utilize agency space. Leaders may pay ex ante costs to take advantage of agency space (for example, the costs of implementing increased repression), or risk paying ex post costs (for example, removal from office).
There are fundamental differences in the nature of costs that autocratic and democratic leaders pay. Ex post, the highest risk to a democrat is that she is voted out of office. The highest risk to an autocrat is that she is tortured and killed.Footnote 65 Although scholars of authoritarian regimes and conflict disagree about whether fear of violent removal is central to autocrats’ decisions for war,Footnote 66 the nature of ex post costs that autocrats pay is clearly different. In the IPE realm, Ballard-Rosa argues that autocratic leaders have an “urban bias” because they fear the threat of protest from their concentrated urban populations, sometimes leading autocrats to default on sovereign loans rather than risk such protest.Footnote 67 Similarly, while Miller and Peters argue that autocrats can benefit from letting some citizens emigrate, allowing emigration to democracies carries the risk that migrants will bring democratic ideas home.Footnote 68
The likelihood of paying these costs also differs across regimes, for both structural and strategic reasons. Personalists, for example, coup proof precisely to keep the probability of violent removal low.Footnote 69 Ex ante costs are low for autocrats who want to increase their exposure to domestic audiences, but they may struggle to make those constraints credible given potentially violent removal or the risk of more lasting changes in domestic political institutions. As Weiss argues, the risk of instability stemming from nationalist protests can enhance credibility, so regimes like China must push the limits of agency space to tie hands credibly.
Technological change may be an important and productive area of future research on how agency costs manifest.Footnote 70 From research on human rights, internal security, and the role of technology in authoritarian regimes, we already know more about how authoritarian leaders engage in preventive repression and information control.Footnote 71 But these developments suggest that the ex ante costs of repression may be low for regimes with a robust repressive apparatus, and lower still when technology reduces the marginal cost of preventive repression.Footnote 72 These factors may have complex effects on authoritarian regimes’ international behavior: technologies that help countries maintain domestic control might allow leaders to turn their attention to international disputes,Footnote 73 but they may also make it more difficult to generate credible strategic increases in authoritarian audience constraint.
Plasticity of Democratic Institutions and Audience Flexibility
We suggest that democratic leaders can take advantage of agency space more readily than authoritarian governments, and can do so more frequently without altering their institutions, because democratic institutions are more structurally plastic—they can withstand and adapt to challenges—and allow for greater flexibility in managing domestic audiences. This greater plasticity of democratic institutions helps us resolve the puzzle of why, if autocrats can increase audience exposure, they do not do so more often. Moving too far or too frequently might generate unintended institutional change, that is, unwanted x axis movement—much like a rubber band that is stretched too many times and ultimately snaps. As Weiss notes in the Chinese context, “the management of nationalist protests may be a tactical asset in the short run but a strategic liability in the long run. Domestically, the government may find it increasingly difficult to preserve domestic stability as the cycle of nationalist mobilization repeats.”Footnote 74
There are several reasons democratic institutions have greater plasticity or flexibility. Democracies include mechanisms for peaceful and periodic transitions in power. Democratic leaders who increase their exposure risk losing an election, but authoritarian regimes tend to have more brittle power transitions where the regime itself is at risk.Footnote 75 Democratic institutions are also more plastic in the sense that they can accommodate leaders who try to challenge or alter constraints—that is, they can bend the rules without breaking them.Footnote 76 A related but separate source of flexibility is that democracies have more coalition-building options within a given set of institutions than their autocratic counterparts. As recent research has shown, autocratic leaders make significant tradeoffs to ensure their tools of autocratic control can address their central threats.Footnote 77 But having made these tradeoffs, autocrats cannot simply turn to alternative audiences without serious risk—any shift could destabilize these arrangements. By contrast, democratic leaders may have more potential coalitions available and can strategically alter domestic audience constraints—from both elite and public audiences—at lower risk.
This difference applies to both leadership turnover and to incumbent leaders’ foreign policies. With peaceful transitions and the absence of violence in politics, democratic institutions can accommodate a new leader with a different coalition. Mattes, Leeds, and Matsumura find that democracies see more changes in the societal sources of leader support, and that such changes can occur even “without irregular leader transitions or large institutional changes.”Footnote 78 For incumbents, autocratic leaders may find fewer options to alter audience constraint at acceptable risk, while a democratic leader seeking to alter foreign policy could activate preference cleavages or public opinion on a particular issue, or alternatively, craft a coalition that insulates or diffuses the issue politically. Recent studies of public opinion illuminate how democratic leaders can navigate heterogeneous public audiences.Footnote 79 Rather than engaging public opinion for its own sake as a hand-tying mechanism, several recent studies have shown that democratic leaders can take advantage of preference cleavages by activating or avoiding segments of the public, increasing their room for maneuver.Footnote 80 In the security realm, research on secrecy shows how democratic leaders can also avoid traditional audience costs or engage in “secret wars” to insulate themselves from hawkish publics.Footnote 81 On trade policy, Guisinger shows that for decades US politicians took advantage of demographic and other trends in trade preferences to dampen protectionism's political impact.Footnote 82
Democratic leaders may also have greater flexibility to manage audience information and issue salience. Although democracies allow freer flow of information (with important structural variation, as Baum and Potter detail), they also have opportunities to take advantage of variation in how much audiences know and care about an issue. For example, Guisinger finds that citizens largely do not know how their elected representatives vote on trade and rank it relatively low in importance, undermining an important link in the accountability chain.Footnote 83 In India, there is wide variation in domestic issue salience, yielding what Narang and Staniland call different “accountability environments.”Footnote 84 Saunders argues that leaders strategically manage the cues that reach the public in the first place by bargaining with elite cue givers or managing information within elite circles, effectively reducing the size of the selectorate in democracies.Footnote 85 Similarly, Milner and Tingley suggest that security policies are easier to insulate from domestic constraints than distributional, economic issues that may be more immediate to segments of the domestic audience. All these strategies can increase democratic leaders’ room to maneuver when making foreign policy; and even if suppressed information emerges, democratic leaders face lower political costs than autocrats. By contrast, autocratic leaders bear greater risk when they manipulate information: for example, Hollyer, Rosendorff, and Vreeland argue that autocracies risk mass protest and thus regime instability when they disclose information. Some autocratic leaders might find it useful to manipulate this risk by using the threat of mass protest to unify elites—but for our purposes, it is the difference in risk that is crucial.Footnote 86
If our conjecture is right that democracies, relative to autocracies, have greater institutional plasticity and greater scope to manage audience composition and/or information, it is still an open question whether this greater flexibility translates into an international-level advantage or disadvantage on a given issue. For example, in terms of credibility, if democracies can seek alternative coalitions or untie their hands, the democratic advantage in audience constraints may be smaller or less durable than prior research suggested, even if the default level of democratic constraint is higher. Large-scale shifts in societal preferences—like the increased anti-war sentiment after the Vietnam and Iraq Wars—may also reduce democratic leaders’ potential flexibility in audience composition if their preferences diverge from those of the society. But democratic institutions may also smooth out the effects of leadership or coalitional change on international outcomes: for example, Leeds, Mattes, and Vogel find that in contrast to autocracies, democracies’ alliance commitments are not affected by changes in societal coalitions (although time will tell whether more direct attacks on trade and security agreements, such as those of President Trump, have longer-term effects).Footnote 87
International Pressure
We also argue that international pressure can shape how leaders manage domestic constraints, and that most second-generation DIR and AIR arguments would benefit from more consideration of how international conditions (SIR) shape the way leaders engage with domestic audiences.Footnote 88 Of course, the notion that international factors can influence domestic politics is not new.Footnote 89 Yet recent literature, including the SIR books reviewed, suggests an important role for international factors, which we argue can shape the boundaries of agency space, leaders’ incentives to take advantage of agency space, and the nature of agency costs.
For simplicity, Figure 1 depicts a snapshot of domestic audience constraints without varying international pressure, yet changes in international pressure can change the parameters of Figure 1. In a given issue area, one way that international factors shape the role of regime type is by changing the boundaries of agency space. Such shifts can expand agency space, compress it to zero for some or all leaders, or moderate or enhance regime type differences. Gunitsky's book underscores that the pro-democracy bent of the post-Cold War era is only one manifestation of how international dimensions can systematically alter domestic political institutions. As he argues, great powers’ backing, or even simply incentives to emulate their institutional form, makes particular domestic political forms more attractive globally.Footnote 90 His book reminds us that the recent international emphasis on democracy and democratization is just that—a recent emphasis.
In addition to shifting the boundaries of agency space, these types of international or system-level trends also change incentives for leaders to take advantage of agency space. When the international environment rewards democracy, as in the 1990s, agency space in the “exposure zone” might grow. In a pro-democracy era, leaders may also have greater incentive to utilize that zone of exposure and a disincentive to move “down” if repressive strategies are penalized by international actors. For example, nondemocratic leaders might superficially adopt some formal trappings of democracy and avoid more overt authoritarian strategies for managing domestic audiences.Footnote 91 “Pseudo-democratic” states may allow opposition party challenges, liberalize laws regarding women's rights, permit critical reporting of the government, or tolerate increases in anti-government protest.Footnote 92 These processes, we argue, can be strategic uses of agency space to gain international benefits rather than genuine reforms. Crucially, leaders who use this strategy risk unintended structural (x axis) change. International-level effects might sometimes also be countervailing: for example, the rise of China could increase autocratic institutions’ attractiveness for other states, as Gunitsky implies. But increased financial integration may lead authoritarian elites to diversify their assets and reduce their opposition to democratization, as Freeman and Quinn suggest.Footnote 93
While great powers can deliberately exert international pressure on domestic political institutions, changes in incentives to use agency space and/or the boundaries on agency space can also follow from external trends or shocks to the system. For example, global trends in investment and capital flows can condition the extent to which regime type differences matter. Ballard-Rosa, Mosley, and Wellhausen find that when global capital liquidity is high, investors will tolerate the higher risks associated with authoritarian regimes, reducing the “democratic advantage” in sovereign debt.Footnote 94 In terms of shocks, the 9/11 terrorist attacks resulted in foreign policy changes in which the US prioritized allies in the war on terror, turning a blind eye to repression that they would have condemned or penalized in the 1990s.Footnote 95
International factors also condition AIR/DIR comparisons by influencing how agency costs manifest, though this is least developed in the SIR literature. The international diffusion of specific strategies, technologies, or practices can make some domestic behaviors more internationally acceptable or desirable. More generally, as in Conrad and Ritter's Contentious Compliance, research on human rights and transnational activism suggests that treaties or pressure at the international level can activate domestic mechanisms that change the nature of costs leaders would pay to continue their prior policies.Footnote 96 In terms of international institutions and IPE, an interesting avenue for future research is to explore whether democracies can better absorb the costs of breaking or renegotiating international agreements. For example, as investor-state dispute settlement (ISDS) and bilateral investment treaties (BITs) have become more politically controversial in an era of backlash against globalization, both democracies and autocracies have worked around the problem, sometimes breaking agreements but protecting investment.Footnote 97 If autocracies gained greater benefit from these agreements in the first place—as substitutes for the credibility of democratic institutions—perhaps the same international outcome carries very different costs for autocrats.Footnote 98
These three factors often work together. Consider the rise of China and a concurrent decrease in US support for democracy. Even if China does not seek to promote autocracy,Footnote 99 the rise of an authoritarian great power might provide alternative international benefits or more permissive conditions for autocrats, increase the amount of repressive agency space, or decrease the costs for utilizing repressive agency space. Autocrats may also see fewer benefits to signing international agreements that could constrain their ability to manage domestic audiences. China already lends to the developing world with fewer conditions than Western lenders.Footnote 100 In parallel, the Trump-led US shift away from supporting democratic allies and toward supporting authoritarian adversaries has emboldened formerly pseudo-democratic leaders to act more overtly authoritarian, though the lasting effects of this change are not yet clear.
Conclusion: Regime Type Is Here to Stay
Our framework—and the books and related work that inspired it—lead us to conclude that regime type is still a crucial concept in IR, and that the overarching picture from second-generation scholarship has reaffirmed, rather than rejected, its importance. The pendulum that swung toward domestic politics and particularly the distinctiveness of democracies after the Cold War has not swung back as far as some suggest. Regime type provides important structural constraints and bounds on state leaders and the degree to which political elites can strategically manipulate those constraints. Such manipulation may have important payoffs in particular issue areas, but crucially, it can also backfire. The costs and risks associated with strategic manipulation of domestic audiences suggest many avenues for future research. Beyond this research agenda, our review yields several important takeaways about the direction of research on regime type and international relations. First, while scholarship on what we call structural and strategic sources of leader-audience accountability is not new, few scholars address both. The connection between structural constraints and strategic behavior is important because it affects how much leaders can temporarily increase or decrease accountability before triggering more lasting institutional changes, and is essential to understanding regime type differences. It also influences the degree to which autocrats and democrats can move strategically against type, as when autocrats increase constraints and democrats increase insulation from domestic accountability. Even if an authoritarian state can sometimes generate democracy-like outcomes in the international arena, there is a ceiling on how constrained its leaders can be without causing instability or regime change.
Second, our focus on the strategic sources of audience constraint highlights the limits of what better measures can do for understanding how regime type affects international behavior, despite improved measurement of both democracy and autocracy.Footnote 101 The potential for states to behave in ways that lead to similar outcomes, and the difficulty in detecting and measuring the effects of strategic use of agency space, underscore that new and better measures of political institutions are not enough to assess democratic or autocratic distinctiveness on a given IR dependent variable. Leader incentives to emulate features of other regime types—incentives which themselves depend on international-level factors—may bias measures and represent a thorny problem for cross-national empirical work. Additionally, state leaders may seek to confound not only direct measures of regime type, but also data on economic performance, for both domestic and international audiences.Footnote 102 As new insights from comparative authoritarianism move into IPE, this problem may be even more challenging because bias in self-reported indicators may be systematically related to regime type in ways that are difficult to document.
Third, what of the democratic advantage? Our reading is that the tone of research on autocracies in IR is often too optimistic, while the tone of research on democracies is sometimes too gloomy. Even when autocrats decide to pay costs and reach a level of audience constraint accessible to democrats, it may not be a particularly large shift in international policy. For example, Vreeland argues that multiparty autocracies—which in our framework, have a higher default level of constraint than other autocracies—join institutions like the UN Convention Against Torture (CAT) precisely because it is a “relatively cheap concession” to interest groups that have some voice in domestic politics, and given the preferences of Western powers and NGOs, might bring some international benefits.Footnote 103 But these autocratic CAT joiners continue to torture post-ratification.
Likewise, some recent research on how democracies fare in the international arena can seem overly pessimistic. Some pessimism is natural given that many findings point either to similar outcomes across regime type or a smaller democratic advantage. And the news is not all good for democracies. For example, Bastiaens and Rudra show that democracies are less well-equipped to deal with the “revenue shock” resulting from a loss of trade taxes under globalization.Footnote 104 As we discuss here, democratic distinctiveness does not always translate into democratic advantage, or even preferable foreign policy choices. But other assessments would benefit from more clarity on exactly how costs manifest for democratic leaders to take advantage of agency space. When do democratic institutions provide leaders with flexibility to take advantage of agency space in different ways? How do democratic institutions withstand repeated attempts to manipulate constraint?Footnote 105
Finally, renewed attention to system-level effects on domestic institutions suggests that domestic politics and international politics remain inseparable. These effects include the “hegemonic shocks” Gunitsky highlights, which directly affect the costs and benefits of certain domestic institutions; international pressure that incentivizes states to adopt democratic forms; and international market forces like capital abundance or scarcity that change the degree to which regime type matters. Democracy may no longer pay quite so much as it did in the post-Cold War era of US hegemony. But if this era ushers in more similarity in international behavior across regime type, we should not conclude that similarity in domestic accountability necessarily reflects reduced influence of domestic politics. The works reviewed here remind us that regime type will remain a rich area of scholarship in IR for the foreseeable future. We paint a picture of democratic flexibility, with democratic leaders taking advantage of agency space more readily or with lower risks—leading to foreign policy outcomes that are sometimes, though not always, beneficial.
Acknowledgments
For comments on prior drafts of this essay, we thank seminar participants at University of California, Berkeley; Stanford University; University of California, Los Angeles; University of California, Santa Barbara; Brown University, and the University of California, San Diego. For detailed comments and conversations we give special thanks to Nick Anderson, Sarah Binder, Rob Blair, Austin Carson, Stephen Chaudoin, Jeff Colgan, Dan Drezner, Henry Farrell, Peter Gourevitch, Alexandra Guisinger, Seva Gunitsky, Emilie Hafner-Burton, Nate Jensen, Miles Kahler, Robert Keohane, David Lake, Yon Lupu, Nikolay Marinov, Kate McNamara, Vipin Narang, Abe Newman, Irfan Nooruddin, Joseph Nye, Emily Ritter, Nita Rudra, Rachel Schoner, Ken Schultz, Susan Sell, Paul Staniland, Caitlin Talmadge, Jennifer Tobin, Michael Tomz, Jim Vreeland, Jessica Weiss, Rachel Wellhausen, and many others. For excellent research assistance we thank Andrew Blinkinsop. This essay arose partly out of discussions at the workshop on “Reconsidering Regime Type in IR,” Yale University, April 2015, and we gratefully acknowledge the participants at this conference. We thank Jonathan Baron and Lauren Pinson for excellent assistance at the conference. Note that we chose not to review books of our colleagues and coauthors. All errors and omissions are our own.
Funding
We gratefully acknowledge conference funding from the Edward J. and Dorothy Clarke Kempf Fund from the MacMillan Center, Yale University; The Institute for Security and Conflict Studies at the Elliott School of International Affairs, George Washington University; and the Leitner Program in International and Comparative Political Economy, Yale University.
