What Do We Know about Hybrid Regimes after Two Decades of Scholarship?

In two decades of scholarship on hybrid regimes two significant advancements have been made. First, scholars have emphasized that the hybrid regimes that emerged in the post-Cold War era should not be treated as diminished sub-types of democracy, and second, regime type is a multi-dimensional concept. This review essay further contends that losing the lexicon of hybridity and focusing on a single dimension of regime type—flawed electoral competition—has prevented an examination of extra-electoral factors that are necessary for understanding how regimes are differently hybrid, why there is such immense variation in the outcome of elections and why these regimes are constantly in flux. Therefore, a key recommendation emerging from this review of the scholarship is that to achieve a more thorough, multi-dimensional assessment of hybrid regimes, further research ought to be driven by nested research designs in which qualitative and quantitative approaches can be used to advance mid-range theory building.


Introduction
As the Third Wave of Democracy ended in the 1990s, a plethora of regimes emerged in the non-western world that were qualitatively different from each other, but also from Western democracies. These were hybrid regimes that occupied the "grey zone" between liberal democracies on the one hand and closed authoritarian regimes on the other (Carothers, 2002). The main challenge that scholars of comparative politics faced was how to define and classify these hybrid regimes without falling prey to concept stretching. This spawned a vast literature, which attempted to unpack this category of regimes. Nearly two decades later it is important to evaluate how this research agenda has evolved and if we have enhanced our understanding of this regime type.
Early work on hybrid regimes focused on conceptualizing these regimes because it was necessary to distinguish the boundaries among different regime typesauthoritarian, hybrid and democracy (Merkel, 2004;Puhle, 2005). Scholars have since then established that hybrid regimes should try to avoid the teleological bias of earlier studies that categorized hybrid regimes as diminished sub-types of either democracy or authoritarianism (Bogaards, 2009;Gilbert & Mohseni, 2011;Morlino, 2009). We have now also learnt that hybrid regimes are not transitional phases but in fact political regimes that manifest a combination of both authoritarian and democratic tendencies that ought to be examined in comparison to each other and not against the standards of democracy.
A thriving set of literature that treats hybrid regimes as being a sub-type of authoritarianism due to flawed electoral competition (Levitsky & Way, 2010;Schedler, 2002Schedler, , 2006 tends to examine these regimes along only one dimension of electoral competitiveness. This review makes the case that if we are to understand the political consequences of elections on regime type, it is imperative to treat hybrid regimes as multidimensional concepts. However, to conduct meaningful multi-dimensional analysis, this article proposes nested research designs that entail both qualitative and quanti-tative approaches. Qualitative research should be driven by single-n case studies, or paired comparisons based on in-depth field research with the intention to advance our contextual knowledge of these regimes and facilitate mid-range theory-building. This qualitative approach should be complemented by large-n statistical analysis that tests the strength of the independent variable gleaned from the case study.
This article proceeds in the following way: section 2 sheds light on the conceptual confusion that has persisted among scholars over the question, what are hybrid regimes and makes the case that these regimes are not transitional states. The next section examines the importance of treating regime type as a multi-dimensional concept. Section 4 suggests ways to advance causal research on hybrid regimes.

What Are Hybrid Regimes?
Despite the extensive theorization of hybrid regimes and numerous attempts to bring some clarity to the blurred lines among different political regimes, it is difficult to find consensus among scholars over what hybrid regimes actually are. This unfortunately has hampered the "accumulation of knowledge" on what a hybrid regime is (Cassani, 2014, p. 548). The inconsistency in the variety of approaches used to define hybrid regimes is proof that scholars are not in conversation with each other or building on each other's work to advance the research agenda.

Diminished Subtypes
A diminished subtype stems from a root concept, where the attributes of the latter are not fully shared by the former (Collier & Mahon, 1993, p. 848). If we can visualize hybrid regimes to comprise the "sprawling middle of a political continuum between democracy and nondemocracy" (Bunce & Wolchik, 2008), the graded nature of the spectrum sees "democracy as an institutional quality that is principally a matter of degree" (Wahman, Teorell, & Hadenius, 2013, p. 21). It follows therefore that as one moves away from the democratic end of the spectrum essential qualities of the regime are lost, making the democratic regime a diminished version of itself. This led scholars to proliferate typologies of democracy, a trend described as "democracy with adjectives" (Collier & Levitsky, 1997) or the "terminological babel of democratization studies" (Armony & Schamis, 2005, p. 113).
Some of the most popular terms were delegative democracy (O'Donnell, 1994), semi-democracy (Diamond, Linz, & Lipset, 1995), illiberal democracy (Zakaria, 1997), and pseudo-democracy (Diamond, 2002). Most recently, German scholars like W. Merkel, (2004) built on the concept of defective democracy or incomplete democracy. The central pre-occupation of these scholars was to understand why these diminished forms (or hybrid regimes) were unable to transition to democracy because, as Linz (2000, p. 34) points out, there was a hope that these imperfect democracies would amend themselves.
On the other end of the spectrum, scholars of authoritarianism have also been complicit in viewing hybrid regimes as a "corruption of the preceding regime" (Morlino, 2009, p. 280). The only redeeming merit according to A. Cassani (2014, p. 544) was that these scholars "stressed the attributes that these regimes possessed than what they lacked". M. Ottaway (2003) coined the term semi-authoritarianism to describe a regime that displayed characteristics of both democracy and authoritarianism-these were not failed democracies, but regimes that wanted to maintain their ambiguous character. The co-existence of elections as the predominant form of elite succession and dictatorial control led A. Schedler (2002, p. 36) to identify electoral authoritarianism-a regime in which leaders "hold elections and tolerate some pluralism and interparty competition but violate democratic norms so severely and systematically that it makes no sense to call them democracies, however qualified". The distinguishing feature of multi-party elections with the absence of democracy led to an even more precise typology including the hegemonic electoral authoritarian regimes (see Magaloni, 2006, on Mexico), in which the leader's party routinely wins; competitive authoritarianism (Levitsky & Way, 2002), in which opposition parties can win substantial majorities in elections; or the closed authoritarian regime, where no opposition parties are allowed to exist.
The implication of defining hybrid regimes as diminished subtypes of either democracy or authoritarianism was that empirical research would be extremely challenging because the boundaries between the mixed regimes and their root concept were blurred (Bogaards, 2009). Furthermore, the definition of a hybrid regime would vary depending on how scholars understood the root concept. His solution to this problem was a double-root strategy in which the "root concepts are defined in relation to each other and cases are classified with a view to both" (2009, p. 410). As a result, Bogaards argued that hybrid regimes ought to be conceived as a residual category that fits neither democracy nor authoritarianism but as a regime type unto itself.

A Residual Category
The proposal to treat hybrid regimes as a residual category and to study them on their own terms instead of being anchored to either democracy or authoritarianism resulted in the proposal of fresh typologies that have attempted to advance the comparative analysis of political regimes (Gilbert & Mohseni, 2011;Wigell, 2008). Scholars also attempted to create intermediate types between democracy and authoritarianism. For example, the trichotomous scheme advocated by Mainwaring, Brinks and Perez-Linan (2001) organized regimes into democracy, semi-democracy and authoritarianism. Other labels for this intermediate regime type included mixed (Bunce & Wolchik, 2008) or simply hybrid (Ekman, 2009;Karl, 1995). Although this approach allowed for greater differentiation, its analytical utility is limited. The regimes that fit this residual category are so qualitatively different from each other that except for the one commonality that they are neither democratic nor authoritarian, it is very hard to actually compare them systematically.

An Authoritarian Regime
Those who favor a more dichotomous approach to categorization treat hybrid regimes as overt instances of authoritarianism. For these scholars there is no overlap between regime types. This allows for more parsimonious categorization by creating mutually exclusive categories that classify regimes as being democratic or authoritarian (Cheibub, Gandhi, & Vreeland, 2010;Sartori, 1987). The presence of political institutions that have been integral to democratization in the West contribute to the hybridity of the regime, such as political parties (Gandhi, 2008;Greene, 2009;Magaloni & Kricheli, 2010); elections (Brownlee, 2007;Gandhi & Lust-Okar, 2009) and the legislature (Gandhi & Przeworski, 2007;Malesky & Schuler, 2010). However, Cassani (2014) notes that these scholars downplay the mixed nature of the regime, because the presence of democratic institutions does not fundamentally alter the identity of the authoritarian regime itself.
Another way to conceptualize the hybrid regime is to categorize it as an authoritarian sub-type, a trend that is suggested by B. Magaloni's (2006) hegemonic party regime or the limited multi-party regime (Hadenius & Teorell, 2007) which corresponds to Schedler's conception of the electoral authoritarian regime.

A "Transitional Situation"
Partly the reason for why there is a conceptual divergence among scholars on how seriously to view hybrid regimes as a regime type, and not in relation to other regimes is due to the inherent instability stemming from the ruling elite competing over the resources of the state and the effort expended to implement policies aimed at their self-preservation. The fluid nature of politics is assumed to be symptomatic of the regime being in a transitional "situation", suggesting that the transition will either be completed once democracy consolidates or the regime could backslide and renew forms of autocratic control (Armony & Schamis, 2005). But there is also a third possibility, where a regime could stabilize in this uncertain state and persist as a hybrid regime. This is because the coalition of individual or collective actors that maintain the regime can use the ambiguity of the regime to achieve their preferred political goals and, therefore, do not have the incentive to aim for an ideal regime type. Morlino (2009) argues that if the co-existence of authoritarian and democratic features continues to persist for more than ten years, then one can plausibly argue that the main actors in the regime have found an adequate means for their perpetuation, or that a central power keeps the regime in its characteristic "state of ambiguity and uncertainty" (p. 286).
This section has shown that although the conceptualization of hybrid regimes led to a vast literature, unfortunately confusion over what hybrid regimes are still lingers. The disagreement on whether hybrid regimes are diminished subtypes, residual category, transitional phase or an outright case of authoritarianism has important implications for empirical work. Cassani's analysis of eight studies that attempt to identify political regimes found that they disagreed on which regimes to call hybrid (2014). Morse (2012) also observes that there is significant divergence among scholars studying the phenomenon of electoral authoritarianism in that there is no consensus over the extent of electoral violations that would make a regime non-democratic. For example, Egypt and Singapore, which do not have competitive elections would be considered authoritarian by Levitsky and Way (2010), yet Schedler would deem them to be electoral authoritarian and therefore hybrid. The selection bias stemming from this conceptual divergence can make it difficult to evaluate causal research. Moreover, the misidentification of regime type can have important policy implications for those in the field of democracy promotion.
The conceptual confusion over what constitutes a hybrid regime has also led scholars to move away from the lexicon of hybridity towards either electoral democracy (Diamond, 2002) or electoral authoritarian regimes (Lindberg, 2009;Schedler, 2006), with a greater empirical focus on dimensions or components of these regimes through continuous measures suing databases such as V-Dem, Polity IV, Freedom House, etc. This shift has encouraged the recognition of how regimes may be "differently democratic" or "differently authoritarian" or even "differently hybrid".
However, I would urge caution before we discard the term "hybrid" for two reasons: first, political regimes are not inherently stable. Therefore, we need to move beyond the teleological assumptions embedded in the "diminished subtypes" approaches. Contemporary politics shows that democracy itself may not be a stable category. After the election of Trump in the United States, following the election of similar leaders in established democracies like Hungary, Poland and Venezuela, not to mention less established democracies such as Russia and Turkey, there is increasing concern over democratic backslid-ing. Democracies themselves can be "diminished" (see Mounk, 2016Mounk, , 2018, which challenges the need to measure "hybrid" against stable categories of democracy or dictatorship. Perhaps hybrid is the norm?
Second, hybrid regimes are also not a transitional state that is expected to quickly amend itself and become either a democracy or an authoritarian regime. In fact, what we have found is that hybrid regimes are often quite durable and need to be understood for what they truly are and that the terminology used to refer to such regimes should reflect this.

The Importance of Multi-Dimensionality
The most common underlying dimension for classifying regimes is electoral competition (Howard & Roessler, 2006;Levitsky & Way, 2010;Lindberg, 2009;Schedler, 2006). However, hinging the classification of hybrid regimes on the basis of electoral competition on a unidimensional spectrum anchored by liberal democracy on the one end, and closed authoritarianism on the other, can prove to be problematic. L. Morgenbesser (2014) explains that the role of elections in democracies, and the meaning attached to them, is not the same as in an authoritarian regime (Gandhi & Lust-Okar, 2009). In the former, the assumption is that elections are a democratic institution for the purposes of legitimate elite succession. But elections in authoritarian regimes can also serve as instruments of elite-management, distribution of patronage and signaling legitimacy in non-democratic regimes (Brownlee, 2007;Gandhi & Lust-Okar, 2009;Magaloni, 2006). For Morgenbesser (2014) this has resulted in concept stretching because analysts have failed to recognize that the role performed by elections-an institution that makes the regime hybrid-will vary depending on the root concept used and also the political context. Regime analysts only pay attention to the "quality" elections for regime analysts and not to "quality and meaning" (p. 25).
The overwhelming emphasis on elections also overlooks a range of more fundamental dimensions that are critical to the analysis of political regimes. It ignores the reality that hybrid regimes can be different from each other ways besides the competitiveness of an election. Munck and Snyder (2004) explain that the political consequences of elections depend on interaction with key extra-electoral factors, such as who rules, how do they rule, why do they rule and how much do they rule. These are all factors that Morse refers to as "actor capacity" (2012, p. 173).
Similarly, Gilbert and Mohseni (2011) have also found that by redefining of the electoral regime they discovered additional regime dimensions-competitiveness and competition-that are important for classification. They explain that "democracies are competitive regimes with fair competition, whereas authoritarian regimes are uncompetitive regimes with unfair competitions. Hybrid regimes occupy the conceptual void of competitive regimes with unfair competition" (Gilbert & Mohseni, 2011, p. 280). While competitiveness is crucial for distinguishing democratic and hybrid regimes from authoritarian ones, only the quality of competition is operationalized in distinguishing democracies from non-democracies (including authoritarian and hybrid regimes). Beyond elections, another key element that establishes this boundary between democracy and nondemocracies is tutelary interference, when unelected bodies such as the military, religious authorities or a monarch constrain the agency of elected leaders or veto national legislation.
Multi-dimensional conceptualizations of regime types are not new to political science. For example, R. Dahl (1971) dropped the use of the word democracy and introduced polyarchy-a regime type defined by the intersection of two dimensions, contestation (the right to compete over desired policies) and participation (the freedom to participate in the political process). Merkel (2004) emphasized three dimensions of democracy including vertical legitimacy, horizontal accountability plus rule of law and effective government. These dimensions were further broken down into sub-criteria, which, if violated, would result in four types of defective democracy-exclusive, illiberal, delegative, and tutelary democracy. Inspired by Merkel, Wigell (2008) also starts with the root concept liberal democracy, emphasizing that the goal of democracy is popular government, and the goal of liberalism is limited government, leading to the construction of a typology along two dimensionselectoralism and constitutionalism. Yet the problem he faced is that by assuming that flawed elections and the absence of rule of law are the only ways a regime can be hybrid, he misses out on cases like Pakistan, that are in fact hybrid due to the presence of reserved domains of power (Adeney, 2015). This is why he is compelled to add additional attributes to his minimal criteria of electoralism and constitutionalism. Although, these studies make valiant efforts to move beyond the simplistic unidimensional conceptions of hybrid regimes, they still retain the tendency to view regimes through the "prism of democracy" (Munck & Snyder, 2004, p. 1) and are therefore limited in grasping the full range of variation in regime type globally.
The greatest stride made in regime classification has been made by L. Gilbert & P. Mohseni (2011) who proposed a multi-dimensional conception of hybridity utilizing a configurative approach (Geddes, 1999;Linz, 2000) for the categorization of hybrid regimes. This approach is particularly suitable because it can capture the complexity posed by hybrid regimes, by combining multiple attributes (competitiveness, tutelary interference and civil liberties) as the defining characteristics of the regime. These dimensions cannot be combined to form a single continuum. Each attribute is viewed dichotomously because it emphasizes "differences in kind rather than degree" (p. 282). Gilbert & Mohseni advocate strongly for this approach for three reasons: First, because it enables the comparison or measurement of hybrid regimes in re-lation to other regime types. Second, it allows regimes to be hybrid not just because of flawed competition but may be due to other factors such as the presence of reserved domains of power. Third, in comparison to other typologies of hybrid regimes, the configurative approach provides the most comprehensive list of hybrid regimes that have existed from 1990-2009, grouped under the categories of illiberal hybrid regime, tutelary illiberal hybrid regime and tutelary liberal hybrid regime. As noted by Gilbert & Mohseni (2011), their three-dimensional categorization shows which countries can be grouped together to facilitate comparison and advance the research agenda.
The configurative approach to regime classification that pays heed to multi-dimensionality is a complex categorization, but that is also the reason for its completeness. The challenge of gainfully adopting this approach is that one needs to understand the unique political context of each case before categorizing it, thereby requiring scholars to undertake in-depth single-n case studies. An excellent example is K. Adeney's analysis of Pakistan's hybrid regime (2015). She argues that Pakistan has meaningful, multi-party elections, increasing civil autonomy from the tutelary control of the military compared to the 1990s or early 2000s because of which there would be temptation to view Pakistan as a transitional democracy. However, Pakistan ranks very low on civil liberties and the indirect intervention by the military in politics and foreign policy make it a hybrid regime. By acknowledging regime heterogeneity, and scoring Pakistan's hybrid regime on a three-dimensional continuum, Adeney demonstrates the utility of the configurative approach, which is to pinpoint precisely what factors are preventing Pakistan from crossing the threshold of democracy.

Advancing Causal Research
To recap, two decades of scholarship on hybrid regimes has advanced our understanding of political regimes in two important ways. First, it is unfruitful to make the teleological assumption that the hybrid regimes that emerged in the post-Cold War era as being diminished sub-types of democracy or authoritarianism, which anchor the grey zone that these regimes occupied. Further, it is unrealistic to expect these regimes to necessarily democratize as liberalization occurred and elections were held.
Yet, the response to this lesson has been to view hybrid regimes as a type of authoritarian regime, advanced by Schedler's conception of the electoral authoritarianism. This has been an important conceptual shift for two reasons. First, scholars have begun to study hybrid regimes relative to one another instead of examining how they fall short of meeting the prerequisites of a democracy. Second, instead of democratization, the central occupation of scholars is to understand authoritarian durability and to consider factors that perpetuate hybridity or catalyze democratization. However, an over-emphasis on only a single dimension of regime typeelectoral competition-has prevented an examination of extra-electoral factors that are necessary for understanding how regimes are differently hybrid or why there is such immense variation in the outcome of elections.
Therefore, the second lesson learnt from the hybrid regimes literature is to adopt multi-dimensional assessment of regimes, as showcased by Gilbert and Mohsenis' refreshing hybrid regime classification described in section 3. However, complex, multi-dimensional categorizations necessitate that further research on hybrid regimes ought to be driven by single-n case studies, or paired comparisons based on in-depth field research with the intention to advance our contextual knowledge of these regimes and facilitate midrange theory-building.
Practically, the challenge of a single-n, case-driven research design is one of access and observation. The institutions and norms of such regimes are dismissed as being unstable and therefore difficult to examine (Loyle, 2016). Often hybrid regimes are unsafe and politicallycharged environments, where research travel is viewed as suspicious activity, and getting access to political elite and authentic evidence very difficult. Methodologically, the challenge is best described by Morse (2012, p. 163): "research cannot be too distant from actual cases, leading to conceptual ambiguity, nor too close to specific cases, thus failing to generate comparative leverage". It is not enough to just undertake the configurative approach, populate it with country cases, perhaps also with quantitative indicators because this will only tell us about how politics really works in Malaysia, Egypt, Singapore or elsewhere. It is necessary for the knowledge gleaned from the individual cases to help bridge research agendas and generate new avenues for causal research. Perhaps one way to do this is to adopt a nested research design (see Howard & Roessler, 2006, p. 366), which includes both "quantitative and qualitative methods, with the goal of providing a more valid, reliable, and powerful causal explanation than could be achieved with either method alone". They use large-n statistical analysis to test the strength of their independent variable, and follow-up with a single-n case study to demonstrate how the independent variable matters.
Since this article advocates a multi-dimensional approach, the remaining task is to make a case and provide the rationale for a dimension that is often under-studied in hybrid regimes-elite recruitment and selection. In hybrid regimes that are manifestly multi-party systems, investigating the recruitment of the party elite to elected public office is relevant to understanding the distribution of power among the coalition of collective and individual actors who are decisive in maintaining the regime. In other hybrid regimes, we should leverage existing knowledge on non-democratic regimes to identify the relevant political elite, which would be the selectorate (military junta, political party, family/tribe) and the ruler (military dictator, civilian ruler or monarch). I argue that the question of political recruitment and selection is impor-tant because it determines who gains power, it empowers the recruiters and defines the relationship between the rulers and the ruled by guiding and affecting the behaviour of the political leadership.
There are two reasons for choosing this dimension over all others: First, the definition of a regime entails both behavioural and institutional dimensions, because of which examining the actions of political actors is integral to the understanding of a regime. S.-E. Skaaning's definition of a political regime highlights precisely why elite recruitment and selection matters to an analysis of regime type. A political regime is the "institutionalized set of fundamental formal and informal rules identifying the power holders (character of the possessor(s) of ultimate decisional sovereignty) and it also regulates the appointments to main political posts (extension and character of political rights) as well as the vertical limitations (extension and character of civil liberties) and horizontal limitations on the exercise of political power (extension and character of division of powers-control and autonomy)" (2006, p. 15).
This definition is relevant for the ensuing discussion for three reasons: 1) It accepts that institutions are an important contextual factor shaping and limiting the actions of political actors, while simultaneously acknowledging that the institutional setting is often constructed by the actors themselves. 2) This definition acknowledges that very often the rules and procedures defining a regime may not always be formal and officiallysanctioned, which behoves scholars to also examine the informal aspects of how power is distributed in society.
3) This definition does not just focus on the relationship between the rulers and the ruled (vertical limitations), but also the relationships that might exist among the various power-holders (horizontal limitations). The latter set of relationships entail a constant renegotiation among elites that may cause shifts in the balance of power underpinning a regime.
Political elites and their capacity to perpetuate themselves in positions of power determines the nature of the regime. Examining recruitment and selection of the political elite entails examining not just the resources (time, money, support networks) available to these actors, but also their motivations to pursue political careers and how much autonomy they have in these positions. This is particularly crucial in making this dimension a more superior one because, although Gilbert and Mohseni's dimension of tutelary influence captures the freedom with which leaders can rule, it does not take into account the incentive structures confronting political actors and explain their motivations.
Second, examining recruitment and selection provides a replicable and valid framework that can make comparative analysis of hybrid regimes possible. Any study of political recruitment takes for granted that the elite seek to perpetuate themselves, their goals, and their policies. In authoritarian regimes elites maintain themselves through arbitrary decisions that do not need to be justified ideologically as we might expect in democratic regimes, where multiple elites compete for control over policymaking processes by mobilizing and seeking support from the electorate. The political elite are interested in having a hand in their own succession because as policymakers they also have a stake in the future. A regime is held together at the foundations by stable coalitions of interests made possible by consistent policies. As policies are made by the elite, the perpetuation of the latter is a prerequisite for regime maintenance. Methods of elite recruitment and succession therefore partly define the nature of the regime. When elites are willing to expose themselves to electoral competition and are willing to let citizens determine "who shall rule" in a free election, the regime can be defined as democratic (Huntington, 1996;Schumpeter, 1950). However, when opposite conditions prevail, the regime must be defined as oligarchic at best or authoritarian at worst. If one treats elite recruitment as the independent variable explaining political regimes, the strategies employed by the elite to access power and perpetuate themselves in power is worth exploring (Eulau & Czudonowski, 1976).

Looking Ahead
The suggestions made in this article with respect to advancing the research agenda on hybrid regimes are certainly not exhaustive. My suggestion to examine the dimension of recruitment and selection is by no means the only dimension worth exploring, although I would argue it is a robust, valid and replicable starting point. However, more research on other important dimensions of political regimes must also be undertaken. For example, J. Ekman (2009), instead of focusing on horizontal accountability among political elites, examines vertical accountability between the ruler and ruled. He measures the participation of citizens using three variables: confidence in political parties, turnout and confidence in elections and public support for democracy. Another avenue of research would be to reflect on hybrid regimes in their international context (see Tansey, 2013), in relation to democracy promotion, authoritarian diffusion and various forms of globalization. Thus far, patterns of continuity and change in hybrid regimes have only been studied domestically and have not taken into account the influence of global politics.