DEMOCRATIZATION 2021, VOL. 28, NO. 6, 1075–1094 https://doi.org/10.1080/13510347.2021.1881487 ### Between elites and opposition: legislatures’ strength in authoritarian regimes Felix Wiebrecht Department of Government and Public Administration, The Chinese University of Hong Kong, Sha Tin, Hong Kong ARTICLE HISTORY Received 19 July 2020; Accepted 7 January 2021 KEYWORDS authoritarianism; legislatures; legislative strength; co-optation; power sharing; institutions Introduction Virtually all modern authoritarian regimes have legislatures and recent research has convincingly argued that these are often more than mere “rubber stamps”. Instead, they also contribute to regimes’ durability and economic growth. [1] However, just as different kinds of authoritarian regimes diverge significantly from one another so do their legislatures. [2] Country-specific evidence suggests that some of these legislatures are appropriately classified as “rubber stamps”, while others are gradually gaining more power. [3] There is considerable variation in the number of powers that authoritarian legislatures have at their disposal and this heterogeneity is also greater among authoritarian regimes than amongst democracies. [4] The differences in legislative strength across authoritarian regimes remain puzzling and deserve more scholarly attention. CONTACT Felix Wiebrecht [felix.wiebrecht@link.cuhk.edu.hk](mailto:felix.wiebrecht@link.cuhk.edu.hk) [Supplemental data for this article can be accessed https://doi.org/10.1080/13510347.2021.1881487](https://doi.org/10.1080/13510347.2021.1881487) © 2021 Informa UK Limited, trading as Taylor & Francis Group 1076 F. WIEBRECHT This article aims at filling this gap and investigates the relationships between legislative strength and important institutional characteristics of authoritarian regimes. In accordance with both co-optation and power-sharing theories of authoritarian institutions, the results highlight that three factors are particularly important in predicting legislative strength. These are a regime’s level of personalism, the presence of an opposition in parliament, and the existence of elections. However, beyond these factors, it appears that stronger legislatures are not generally associated with less authoritarian control over legislative and electoral processes. These empirical findings come with two key qualifications. For one, there are important differences between de facto and de jure powers and secondly, the empirics also highlight important historical differences. While more competitive elements are negatively associated with legislative strength before 1990, this relationship cannot be established after the Cold War. I provide empirical evidence for the effects of factors related to both power-sharing and co-optation theory through a large-N cross-country study on authoritarian regimes from 1946 to 2010. These results also carry practical implications with them. For one, they highlight the vulnerability of authoritarian legislatures and suggest that legislatures have been used as “bargaining chips” by dictators in exchange for more contestation, particularly before 1990. The findings also paint a bleak picture for voters in authoritarian regimes as supporting the opposition at the ballot box does not generally correspond to stronger legislatures. This article contributes to our understanding of authoritarian regimes in several ways. First, it provides an original exploratory empirical analysis of legislative strength in authoritarian regimes and thus, adds to earlier in-depth case studies of legislatures’ power. [5] By doing so, it highlights the applicability and suitability of previously devised indices of legislative strength that allow for a comparative study of factors that predict parliaments’ power. [6] Secondly, this study highlights that authoritarian legislatures are shaped through different mechanisms that include elite-elite interactions, elite-society interactions, and intra-parliamentary processes. Thirdly, this article adds a historical dimension to our understanding of authoritarian institutions and draws attention to significant differences between legislatures’ de jure and de facto powers in authoritarian regimes. In the following, this article will introduce the dominant theoretical approaches to legislatures in authoritarian regimes before highlighting the conceptualization of legislative strength. Subsequently, the article will introduce the empirical approach and the findings. Legislatures in authoritarian regimes Initially, legislatures in authoritarian regimes have been decried as window-dressing institutions and being merely of a ceremonial nature. [7] In this perspective, they mainly provide a democratic façade but do not contribute significantly to policy- or decision-making in dictatorships. However, more recent research has revealed how legislatures enhance dictators’ rule in more substantive terms and also prolong authoritarian rule. [8] The two main mechanisms put forward concern rulers’ strategies of coopting opposition forces as well as sharing power among the elites. [9] In addition, authoritarian legislatures fulfil further tasks for dictators such as information collection and transmission as well as being a useful scapegoat. [10] DEMOCRATIZATION 1077 The co-optation approach holds that legislatures are created once dictators feel the need to compromise with the opposition that poses a growing threat to their rule. [11] Thus, dictators bargain over policies and offer concessions to the opposition in order to stay in power. This approach holds that institutions are important since they make the dictators’ claim to concessions credible. [12] However, this view is based on the precondition that authoritarian legislatures are indeed meaningful actors in policy-making processes. Without actual legislative power, they could hardly function as co-optation mechanism and the opposition would continue to aim at overthrowing the regime. [13] A dictator concerned with the opposition’s threat to his rule will have to establish a legislature that possesses some legislative strength in order to convince the opposition to work through parliamentary procedures. Thus, some of the dictator’s powers will have to be delegated to the legislature. While the co-optation approach is primarily focused on opposition forces, an alternative stance sees authoritarian institutions as necessary to structure relations among the elites. [14] Intra-elite conflicts are the more common threat for dictators and thus, they need to devise credible strategies for power sharing. [15] Legislatures provide a device through which dictator and elites can observe each other’s intentions and reduce the information asymmetry between them. [16] The institutions are created in order to safeguard a certain balance of power among the regime insiders and allow for some intra-elite bargaining. [17] Power sharing may come in different forms such as financial remuneration, natural resources, or influence over personnel decision and policies. [18] Recent contributions show that authoritarian legislatures can, in fact, be characterized by elite conflicts over divergent policy preferences. [19] Thus, a credible strategy of power sharing also entails giving elites some influence over decision-making and aligning policies with their preferences may be an incentive for elites to remain loyal to the dictator. In order to function as an effective and convincing power-sharing mechanism then, legislatures in authoritarian regimes need to have some influence on policy making and it is not enough for the dictator to establish a rubber stamp legislature. Instead, it has to be one that also serves as a constraint for the dictator and divides power between executive and legislature in order to qualify as a power-sharing tool. [20] The next section will conceptualize legislative strength and highlight why it is important to analyse it, particularly in authoritarian regimes. Legislative strength The conventional view of legislatures in authoritarian regimes as rubber stamp parliaments only applies to some of them. Other parliaments such as the Grand National Assembly of Turkey for instance have been powerful even throughout the country’s authoritarian periods. In yet other cases, the legislature was the highest organ of state power constitutionally speaking but did not come close to this position in real terms as in the Communist dictatorships of Central and Eastern Europe. Previous studies have operationalized these differences between authoritarian legislatures in relatively one-dimensional approaches. Gandhi, for instance, mainly distinguishes between rubber stamp and strong legislatures that can serve the purpose of co-optation, while Wright draws a distinction between binding and non-binding legislatures. [21] Yet, for our understanding of legislatures as well as authoritarianism more broadly it is imperative to analyse the institutions’ strength in more detail. The existing 1078 F. WIEBRECHT literature has predominantly focused on dictators’ rationale to establish institutions such as parties and legislatures and their advantages to authoritarian rule. [22] However, the number of authoritarian regimes with legislatures has increased steadily to the point where virtually all of them have a legislature. Secondly, authoritarian rulers that come to power often have to work with an existing institutional framework rather than imposing completely new institutions. Thus, contemporary dictators are rarely faced with the decision of whether to create a legislature or not. However, all dictators face the task of shaping legislatures. For the institutional configuration of an authoritarian regime, it is also crucial to understand what powers institutions possess. Autocrats have to give some power to legislatures in particular, either to protect themselves against an elite rebellion or to appease an opposition. [23] However, seen from the dictators’ point of view, legislatures should also not be overly powerful because that could severely constrain them and ultimately threaten autocratic rule. Legislative strength then is the result of dictators’ delicate balancing acts of trying to retain control while having to cede some power to the legislature. For the wider study of authoritarian regimes, it is important to understand how dictators manage this act. Legislative strength highlights under what circumstances the authoritarian ruler is willing to concede how much power to the legislature. Legislatures can be seen as “bargaining chips” and depending on threats emanating from opposition or elites, rulers may respond by giving the legislature more authority, for instance, in response to opposition gains or mass mobilizations. Alternatively, legislative strength can be seen as an outcome of power struggles in authoritarian regimes. Recent studies based on legislatures’ historical development suggest that legislatures become the primary institution for intra-elite struggle if the ruling party is weak and that their development is contingent on whether and how the ruler interferes in legislative processes. [24] Further empirical analyses will help to illustrate these mechanisms. Finally, legislative strength may also have implications for the study of authoritarian regimes as an independent variable. Authoritarian legislatures generally are said to contribute to regime durability, economic growth, and increasing citizen grievances. [25] It will be worthwhile to assess the effects of legislatures with varying strengths on these outcomes. For regime survival, in particular, conflicting expectations exist for strong legislatures. On one hand, fulfilling their function as co-optation or power-sharing mechanism is contingent on their strength. However, stronger legislatures potentially also have a democratizing consequence. [26] In addition, stronger parliaments also lead to legislative defeats for authoritarian rulers and, as a result, likely also shape a regime’s public policies. [27] In a similar vein, legislative strength also “affects distributive outcomes” but with regressive rather than progressive implications. [28] Analytically, scholars have made significant progress on the question of how to operationalize legislative strength cross-nationally in recent years. Legislative strength here refers to the accumulation (or absence) of different powers of legislatures vis-à-vis the executive. The Parliamentary Powers Index (PPI) was a first step to create an index of different powers to compare legislatures across regimes. [29] Subsequently, both the Institutions and Elections Project (IAEP) as well as the Varieties of Democracy (VDem) Project have further improved the data availability. Finally, based on V-Dem data, Wilson and Woldense have constructed an index of legislative powers that is being used in this study as well. [30] DEMOCRATIZATION 1079 This index contains de facto and de jure powers and can thus compose a good balance between constitutionally safeguarded powers and expected powers. It is beneficial to include de facto powers in the index here even if they are based on hypothetical scenarios since constitutions may not unambiguously express all powers conferred upon dictators and parliaments. [31] However, in authoritarian regimes the danger looms large that constitutional prescriptions and practice do not match up as in the case of parliaments in the Soviet Bloc that were often weak in practice but the highest state organs according to their respective constitutions. In addition, the strengthening and weakening of de facto and de jure powers may be results of distinct underlying mechanisms. Thus, additional analyses distinguishing these two kinds of powers are also essential. Table 1 shows all 16 indicators included in the index of legislative strength. While Wilson and Woldense’s measurement has improved scholars’ ability to compare legislatures across regimes, it does not come without potential caveats. Regarding the selection of indicators, they have opted for items that were largely also part of prior data collection efforts such as the PPI without “avoiding an overly ” inclusive approach that mixes information that may only be marginally relevant . [32] The index is mainly aimed at estimating the relative strength of the legislature vis-àvis the executive as this relationship is the most important dimension when assessing legislative strength. [33] Thus, it is crucial to include these factors at the cost of neglecting others such as the relation to the head of government and the cabinet. The V-Dem data is particularly useful insofar as it specifies “upper and lower bounds on the point estimates calculated based on intercoder disagreement and measurement error” and thus, helps to mitigate measurement error concerns. [34] In addition to these points, this article would like to make another relevant potential caveat explicit. The legislative powers included in the index are modelled after “idealtype” legislatures that can only be found in full democracies. However, authoritarian legislatures often function differently and “serve distinctively authoritarian ends”. [35] Authoritarian institutions may perform similar tasks, but they do so in order to contribute to authoritarian rule rather than to provide a democratic supplement. Examples of this include the public oversight mechanisms in Vietnam ultimately aimed at evaluating government ministries not controlled by the Communist Party. [36] Oversight mechanisms are also essential features of parliaments in full democracies but they fulfil different functions in authoritarian regimes. On the other hand, authoritarian legislatures may have distinct functions and powers that are not captured by an index based on democratic ideal types. The Table 1. Legislative strength index. De jure powers De facto powers Can introduce bills in all policy areas Opposition has oversight Has to approve wars Legislates in practice Has to approve treaties Regularly questions executive branch Can grant amnesties Can investigate executive independently Can amend constitution Controls financial resources for internal operations Approval required to pass legislation Appoints Head of State (HOS) Can remove HOS HOS cannot dissolve legislature HOS cannot propose legislation HOS cannot veto 1080 F. WIEBRECHT Standing Committee of the Chinese National People’s Congress for instance has the power of judicial interpretation including that of Hong Kong’s Basic Law. The separation of powers would prohibit parliaments in democracies to pass and interpret the very same laws. Similarly, the People’s Chamber in the German Democratic Republic had the formal power to decide on whether government or administration acted unconstitutionally – a power usually reserved to constitutional courts in democracies. Nevertheless, the abovementioned index still offers a valuable measurement also relevant to authoritarian regimes because it captures the most important dimension of the legislature’s relation with the executive. Although legislatures are essential for almost any political regime today, broadbased comparisons on their strength and powers are relatively limited. Across regimes the level of democracy is generally a strong predictor of legislative strength [37] but as Figure 1 shows the Polity score is only weakly correlated (r = 0.23) with the amount of legislative powers in authoritarian regimes. This highlights the need for more substantive explanations as to why authoritarian legislatures show significant differences. The next section will highlight a number of relevant factors following the abovementioned theoretical approaches to legislatures in authoritarian regimes. Theoretical expectations First, the role of an opposition in parliament is important. As mentioned above, legislatures established to co-opt opposition forces need to have certain powers granted to them by the dictator that will persuade the opposition to work through the formal institutions. Being part of a legislature that has no substantive powers, that is, a rubber stamp institution, is not attractive for the opposition since they ultimately want to change policies. [38] Therefore, following co-optation theory, authoritarian regimes with an opposition in parliament can be expected to have stronger legislatures because dictators aim to incorporate regime adversaries into the regimes’ formal institutions. Figure 1. Legislative powers and Polity score. DEMOCRATIZATION 1081 However, the mere existence of an opposition may not illustrate the full picture since in a number of authoritarian regimes the autonomy of the opposition is far from guaranteed. In contrast, authoritarian regimes may also have an interest in permitting and even promoting a controlled opposition since it will add legitimacy to nominally democratic procedures. [39] Thus, the co-optation mechanism may only work as predicted when opposition parties are also autonomous from the authoritarian elite. Only if they are independent from the regime, they can be a threat to the authoritarian rule and expect concessions from the dictator. If the opposition is not autonomous, the dictator has numerous other ways of influencing them. Thus, not only the existence of opposition parties but also their independent status may be decisive in predicting legislative strength. In addition to the opposition, elections are potentially important factors associated with legislative strength. Giving voters the choice among several candidates opens up the opportunity for opposition forces to be selected as members of parliament. [40] As a result of elections, the dictator may recognize the full strength of the opposition and see himself in a position in which he needs to make concessions and give parliament more power. Alternatively, elections are also used by authoritarian regimes in order to devise a mechanism of “controlled” elite competition. [41] This can be an efficient system for the regime to identify the most competent and loyal candidates. In turn, if the regime wants to incentivize loyal and competent candidates to join parliament, the institution itself also needs to have some power that can attract candidates. Thus, if elections are used as a mechanism to decide upon the composition of the parliament, the legislature can be expected to be stronger according to both approaches. In order to disentangle these mechanisms, the nature of the elections will be assessed as well. For the opposition to enter parliaments in larger numbers the elections have to be open and competitive. Yet, some authoritarian regimes hold elections but do not give voters the chance to choose one party over another. Thus, elections that are more open to multiple parties can also be expected to allow for more opposition and in turn, lead to a stronger legislature according to the co-optation mechanism. Alternatively, authoritarian regimes may also perceive competitive elections as a threat. If the opposition can acquire a significant number of seats in a powerful parliament this could constrain authoritarian rule substantially and lay the basis for its end. Thus, multi-party elections could also be negatively associated with legislative strength since an authoritarian regime may only allow these when the legislature is weak. In other words, legislatures could be used as “bargaining chips” to lower the stakes in competitive elections. These opposing mechanisms will be explored by testing the relationship between legislatures’ strength and elections’ competitiveness. In addition to the existence of elections and opposition, it is necessary to assess the latter’s strength and the threat it can present to the authoritarian ruler. An autonomous but weak opposition does not pose a great menace to the dictator’s rule and is, therefore, not likely to force him to make concessions. Conversely, the stronger the opposition, the more likely the dictator will try to appease them and grant more powers to the legislature in order to hold onto power according to co-optation theory. However, the opposite mechanism is also conceivable, namely that as a result of the opposition’s strength the dictator may curtail the powers that the legislature has. In this way, he can minimize their impact on authoritarian rule and continue to govern without constraints. 1082 F. WIEBRECHT An additional relevant factor for opening up political structures in authoritarian regimes are candidate nomination procedures. In closed regimes, the candidate selection is one potential way of ensuring that the legislature can serve as an efficient powersharing mechanism. If the central party leadership has control over the candidate selection, they can safeguard that elite members from different factions can enter the parliament. Thus, a certain balance of power between them can be upheld and they can utilize the legislature to resolve elite conflicts. Malesky and Schuler, for instance, show the effects of differential processes of candidate selection in Vietnam. [42] If the process of candidate nomination and selection is managed locally candidates that are more critical of the regime can gain access to parliament and demand concessions, for instance, in the form of constraining the dictator. The need for power sharing is inherently expressed in a regime’s level of personalism. Strong personal leaders have less interest in power-sharing arrangements because they already amassed far-reaching powers. In Svolik’s words, these are established autocrats that “can no longer be credibly threatened by an allies’ rebellion” [43] as they have already successfully managed to eliminate regime insiders’ ability to constrain their rule. Essentially, a personalist authoritarian regime is one in which the dictator has eradicated any formal or informal constraints on his power. [44] That is also why they are relatively more threatened by popular uprisings than by “insider challenges” such as coups. [45] A higher level of personalism is thus, expected to be a key factor associated with legislative weakness. In addition, party cohesion may be expected to play a role for legislative strength because it describes individual delegates’ autonomy to push for legislative powers and share power through legislative procedures. In order to facilitate power sharing within a legislature, members of the elite require a certain degree of autonomy. If elite members are bound too strictly by norms and rules there can be no efficient competition for influence and policies among them. [46] Thus, a lower party cohesion may be expected to lead to stronger legislatures because only then can individual elite delegates utilize the legislature as a power-sharing instrument. [47] However, on the other hand, less party cohesion also signifies less control of the authoritarian regime over legislative processes. This uncertainty may not be acceptable to authoritarian leaders and they may only be willing to cede powers to the legislature when they have more control over it. Due to these conflicting potential mechanisms the relationship between party cohesion and legislative strength also needs to be examined. The electoral systems in authoritarian regimes are further mechanisms affecting the autonomy of delegates and, consequently, their ability to engage in elite bargaining through the legislature. Majoritarian electoral systems tend to give the individual members of parliament more autonomy than proportional representation in both democracies and authoritarian regimes. [48] With more individual autonomy legislators tend to engage in more intra-elite bargaining and ultimately have an interest in a stronger legislature that they can use to advance their interests. [49] As a result, majoritarian systems are also expected to coincide with stronger legislatures. Methodology These relationships are tested using a large-N cross-sectional research design. This study makes use of the Autocratic Regime Features Dataset (GWF) by Geddes et al. that covers autocratic country-years for the period 1946–2010. [50] Since some variables DEMOCRATIZATION 1083 suffer from missingness I make use of a set of imputed datasets for the analyses. For the dependent variable, I follow Wilson and Woldense’s approach in constructing an index of legislative strength. [51] All the above-mentioned indicators are accessible through the V-Dem data [52] and they have been combined in an index counting the number of legislative powers with equal weighting. The descriptive statistics (in Tables 1 and 2 in appendix) show that authoritarian legislatures on average possess 5 of these 16 powers. The independent variables have been drawn from a variety of data sources but most of them are taken from the widely used V-Dem Dataset such as the autonomy of the opposition, the existence of elections, elections’ openness for multiparty competition, the electoral system as well as the candidate selection procedures. As per V-Dem’s methodological approach, most of these variables have either been coded by project managers and research assistants or through a survey of country experts. The electoral system has been transformed into an ordinal variable differentiating between majoritarian, mixed, and proportional systems. The V-Dem variable on candidate selection captures a variety of different selection procedures that may have different implications. For the purposes of this study, it has been recoded into a binary variable assessing whether candidates are selected by the central party leadership or not. To assess the existence of different political parties in authoritarian legislatures I am relying on the Political Institutions and Political Events (PIPE) dataset. [53] Their measurement originally distinguishes between parliaments with no, one, or more than one political party. However, it has been recoded into a binary variable that captures whether a legislature houses an opposition or not. This aspect is of more importance for the purposes of this study than the question of whether any party is present in the parliament. While a direct measure of the opposition vote and seat share is available in the Database of Political Institutions, it suffers from a severe missingness problem. Instead, I have made use of the vote share of the authoritarian incumbents as found in the VDem dataset. Thus, this study takes the inverse approach of estimating the strength of the opposition through the results of the incumbents. To assess the degree to which an authoritarian regime is personalized I utilize the measure of personalism by Geddes et al. [54] Contrary to previous studies that have combined personalist dictatorships into a distinctive regime type, this study follows the more recent approach to see personalism as a characteristic of any regime. [55] Military regimes and single- or dominant-party regimes can also display high degrees of personalism such as China under Mao or Libya under Gaddafi. Importantly, the eight indicators used to establish this personalism index do not cover dictators’ relations with the legislature and thus, there is no risk of having both dependent and independent variables captured by the same measurement. Authoritarian regimes differ significantly in a number of defining characteristics such as openness, repression, and guarantee of personal and collective freedoms. Thus, the most important control variable is the level of democracy. The Polity scores have been utilized for all regimes in the given time period to control for the level of democracy. In addition, the models control for countries’ GDP per capita, whether an authoritarian regime was a democracy prior to the observed regimecase, the regime’s natural resources, and the duration of the autocratic spell. The latter one as well as the GDP per capita are both logged. Data on whether the regime was previously democratic, and the autocratic spell duration were taken 1084 F. WIEBRECHT from GWF while the data on natural resources and GDP per capita were taken from Haber and Menaldo. [56] To address the issue of autocorrelation of the standard errors and the slow-moving character of the variables used in this analysis, I follow Beck and Katz’ suggestion of fitting a dynamic model for time-series-cross-sectional data. Thus, all models include the lagged dependent variable (LDV) at t − 1. Empirical results The analyses were conducted in two separate steps. While the first models show the results for all authoritarian regimes with legislatures, the second step of the analysis focuses on electoral dynamics and only includes dictatorships that also hold elections. All following models were estimated using OLS regression models. The results for all legislatures in authoritarian regimes are shown in Table 2. The dependent variable in all models is the legislature’s score in the abovementioned index in any given country-year. Model 1 uses the existence of an opposition in the legislature as one of the explanatory variables, while Model 2 examines the autonomy of opposition parties. Both variables are included in Model 3. Model 4 adds the effects of elections on legislative strength. In accordance with the theoretical expectations, these models highlight three main aspects that are associated with the strength of legislatures in authoritarian regimes. For one, they illustrate that the existence of an opposition in authoritarian parliaments has a statistically significant positive effect on legislative strength. This finding corresponds to the effect of co-optation in dictatorships. Only legislatures that have some influence over policy making can function as credible co-optation mechanism for the opposition. Secondly, the existence of elections is also an important predictor of Table 2. Analysis of all authoritarian regimes with legislatures, 1946–2010. Model 1 Model 2 Model 3 Model 4 Opposition in Legislature 0.1810*** 0.2093*** 0.1760*** (0.0414) (0.0432) (0.0414) Autonomy of opposition −0.0218 −0.0474* (0.0187) (0.0193) Personalism −.2088** −0.2283** −0.2046** −0.2478*** (0.0694) (0.0697) (0.0697) (0.0709) Party Cohesion 0.0305 [†] 0.0272 0.0256 0.0333 [†] (0.0171) (0.0174) (0.0174) (0.0172) Elections 0.1520** (0.0582) Polity 0.0167*** 0.0274*** 0.0246*** 0.0158*** (0.0044) (0.0055) (0.0055) (0.0044) GDP Log 0.1072* 0.1431** 0.1260* 0.0796 (0.0514) (0.0522) (0.0522) (0.0525) Prior Democracy −0.0317 −0.0258 −0.0346 −0.0292 (0.0473) (0.0476) (0.0475) (0.0473) Duration Log 0.0142 −0.0003 −0.0001 −0.0051 (0.0488) (0.0495) (0.0493) (0.0493) Oil Income 7.59e [−][6] 2.16e [−][6] 3.46e [−][6] 8.89e [−][6] (0.000011) (0.000011) (0.000011) (0.000011) Legislative Powers t − 1 0.9085*** 0.9106*** 0.9089*** 0.9036*** (0.0081) (0.0081) (0.0081) (0.0083) N 3173 3154 3154 3173 Note: [†] p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001, Standard errors in parentheses. DEMOCRATIZATION 1085 legislative strength illustrated by the significant positive coefficient in Model 4. Legislatures in authoritarian regimes that hold elections have on average 0.15 powers more than in regimes that do not hold elections. As expected, personalism has a consistently strong and highly significant negative effect throughout all models. The more powerful the dictator is in relative terms compared to other members of the elite, the less need he sees for a powerful legislature. In extreme cases, these dictators have essentially eliminated every threat from their inner circle and do not need to use legislatures as part of a power-sharing strategy. The coefficients for party discipline are positive throughout all four models but fail to reach high levels of significance. Model 2 and 3 also show that the autonomy of opposition parties does not contribute to legislative strength and instead is even associated with weaker legislatures. Turning to the control variables, a country’s Polity score is also a significant predictor of the strength of its parliament although with a weaker magnitude than across all regime types. [57] A higher GDP per capita also leads to stronger legislatures in most models which – is in line with modernization theory. According to the results of Models 1 4, greater income from oil is not significantly linked to either stronger or weaker legislatures. Prior research found no evidence that natural resources are necessarily a curse and lead to authoritarian regimes. [58] In turn, they may also not necessarily lead to weaker legislatures as the findings of this study suggest. The results can also not establish with certainty whether authoritarian regimes that follow democracies tend to have weaker legislatures. The second step of the analysis sheds more light on the dynamics that elections bring into the functioning of legislatures in authoritarian regimes as shown in Table 3. The existence of an opposition in the legislature remains a strong predictor of legislative strength and significant throughout all models. The level of personalism retains its significance in Models 5 and 6 but is only significant on the p < 0.1 level in the last two models. The results also show that vote and seat shares for the authoritarian incumbents fail to have a significant effect on legislatures’ strength. Contrary to cooptation theory, more open elections cannot be associated with stronger legislatures. In fact, the opposite appears to be the case. Elections in authoritarian regimes that are more open to multiple parties are linked to weaker legislatures. Regarding the candidate selection processes, the electoral systems as well as the levels of party cohesion, no significant relationships to legislative strength can be established. Overall, these results highlight that the level of personalism, the presence of an opposition in parliament, and the existence of elections are important predictors of legislative strength. However, beyond these factors, it appears that stronger legislatures are not generally associated with less authoritarian control over legislative and electoral processes. This is mainly captured by the negative effect of more open and competitive – elections (Models 5 8). Yet, the absence of significant results for party cohesion, seat and vote shares, as well as candidate selection procedures, point towards the same conclusion. In the following, further analyses will explore differences between de facto and de jure powers and historical differences in order to highlight the contexts under which these findings remain valid. De jure vs. De facto powers In authoritarian regimes, de jure legislative powers may not necessarily reflect the actual legislative powers. The parliament of the German Democratic Republic, for 1086 F. WIEBRECHT Table 3. Analysis of authoritarian regimes with legislatures and elections, 1946–2010. Model 5 Model 6 Model 7 Model 8 Opposition in Legislature 0.2100*** 0.2768*** 0.2960*** 0.2887*** (0.0484) (0.0548) (0.0564) (0.0566) Seat Share −0.0003 (0.0012) Vote Share 0.0008 0.0012 0.0012 (0.0014) (0.0015) (0.0015) Multiparty Election −0.1053*** −0.0905** −0.0940** −0.0877** (0.0227) (0.0278) (0.0284) (0.0287) Electoral System −0.0148 −0.0207 (0.0313) (0.0315) Candidate Selection 0.0957 (0.0618) Personalism −0.2270** −0.1695* −0.1572 [†] −0.1683 [†] (0.0784) (0.0853) (0.0880) (0.0883) Party Cohesion 0.0135 −0.0092 −0.0202 −0.0229 (0.0229) (0.0258) (0.0276) (0.0276) Oil Income 0.000018 2.89e [−][7] −1.09e [−][6] −2.93e [−][6] (0.000038) (0.000046) (0.000046) (0.000046) Polity 0.0356*** 0.0311*** 0.0321*** 0.0351*** (0.0062) (0.0068) (0.0069) (0.0072) GDP Log 0.0833 0.1743* 0.1587* 0.1616* (0.0643) (0.0750) (0.0763) (0.0763) Prior Democracy 0.0226 −0.0225 −0.0208 −0.0109 (0.0529) (0.0577) (0.0600) (0.0604) Duration Log −0.0128 −0.0231 −0.0172 0.0036 (0.0.0592) (0.0670) (0.0705) (0.0718) Legislative Powers (t − 1) 0.8986*** 0.8822*** 0.8852*** 0.8822*** (0.0095) (0.0110) (0.0113) (0.0115) N 2439 2094 1958 1958 Note: [†] p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001, Standard errors in parentheses. instance, was denoted as the highest state organ in the constitution but in practice, its powers were very limited. Not only are de facto and de jure powers not necessarily consistent in authoritarian regimes but their strengthening and weakening may also follow different logics that warrant additional analyses. This is particularly important since de facto powers appear slightly more volatile than de jure powers. A count of the changes in legislative powers in the dataset reveals that de facto powers have changed in 437 country-years, while de jure powers have only changed in 397 instances. Figure 2 highlights Model 8 introduced above complemented with the autonomy of the opposition in order to highlight the complete set of significant variables. The de facto and de jure powers at t − 1 are included in the models as respective lagged variables but not shown in Figure 2. Figure 2 highlights that the existence of an opposition is conducive to more de jure and de facto powers whereas the oppositions’ autonomy has negative effects on de jure powers. Furthermore, multi-party elections are detrimental to de jure powers on a p < 0.1 level but do not have a significant relationship to de facto powers. The opposite is the case for the level of personalism as more personalist dictators are associated with less de facto legislative powers on a p < 0.1 level. In addition, the higher the party cohesion, the less de facto powers an authoritarian legislature has. The polity score is also a stronger indicator for de facto powers rather than de jure ones. The regime duration has differential effects on legislatures. While legislatures tend to have significantly more de facto power in long-lasting regimes, they also appear to have significantly less de DEMOCRATIZATION 1087 Figure 2. Analysis of all authoritarian regimes with legislatures and elections, de jure and de facto powers. jure powers. This points towards the importance of legislatures in enhancing and prolonging authoritarian rule. [59] Although most of the coefficients fail to reach significance in relation to both de facto and de jure legislative powers they nevertheless reveal some dynamics at play. Some competitive or even “liberalizing” elements are associated with less de jure powers such as the autonomy of the opposition and the openness of elections. This stands in contrast to variables that highlight more authoritarian control but are linked to less de facto legislative powers. These include the level of personalism as well as the level of party cohesion. Thus, there is some evidence to suggest that stronger de facto legislatures are to be found in authoritarian regimes with at least some political competition. In addition to the low significance levels, the analysis of de facto legislative powers comes with another potential caveat. When deciding on the de facto legislative powers V-Dem coders may potentially be influenced by the legislatures’ open- and competitiveness, for instance, characterized by the strength and nature of the opposition. However, the analysis here shows that some variables related to the opposition do not yield positive results. In particular, the autonomy of an opposition does not generate a significant result for de facto powers. The vote and seat shares of the authoritarian incumbent also remain insignificant highlighting that the oppositions’ mere electoral strength is also not linked to legislative strength. Thus, the potential coding caveat does not appear to pose a general concern in this analysis. In addition to de facto and de jure powers, the following section will also analyse historical differences. 1088 F. WIEBRECHT Historical differences The nature of authoritarian regimes has changed significantly from the totalitarian model of the inter-war dictatorships in Europe to today’s more common form of competitive authoritarianism. [60] The end of the Cold War was a major turning point in recent history and has led to the demise of many authoritarian regimes but enabled new forms of competitive authoritarianism to flourish. These differences are also noteworthy when it comes to predicting legislative strength in authoritarian regimes. Authoritarian regimes in the post-Cold War era are more open in that they tolerate more competition in the formal political arena. Yet, they remain authoritarian regimes because “the playing field is heavily skewed in favor of incumbents”. [61] This development is for instance embodied through the increased use of legislatures and elections that are also open to opposition. The average seat shares of authoritarian incumbents as shown in Figure 3 highlight this development. These changes are also similar when communist regimes are excluded (Figure 3 in appendix). The historical differences are shown in Figure 4 below highlighting the results of the LDV model. First, the magnitude of the effect that an opposition has on legislative strength is smaller in the post-1990 era. This is due to the fact that an increasing number of authoritarian regimes have opened their legislatures to opposition actors. With more authoritarian legislatures including opposition parties, it is to be expected that their presence loses some of its explanatory power. However, in the post-Cold War era, the autonomy of the opposition appears to be associated with stronger legislatures while the opposite was the case before 1990. In addition, the level of personalism is significant (p < 0.1) in both models but more so before 1990. The central Figure 3. Average seat share of authoritarian incumbents, 1946–2010. DEMOCRATIZATION 1089 Figure 4. Analysis of authoritarian regimes with elections, before and after 1990. candidate selection is associated with stronger legislatures until 1990 but not after that. Competitive multi-party elections are only associated with weaker legislatures before 1990 but are not significant after the Cold War. While a higher vote share for the incumbents is conducive and significant for legislative strength before 1990, it is not significant after that anymore. Overall, the main factors of personalism and the existence of an opposition remain significant in both historical periods. However, beyond these factors more authoritarian control over legislative and electoral processes appears to go hand in hand with more legislative strength before 1990. A centralized candidate selection and less competitive elections are signs of this connection for instance. However, after 1990 some of these aspects lose their significance such as elections’ competitive character or are even detrimental to legislative strength. A more autonomous opposition, for instance, tends to be associated with stronger legislatures after 1990. Thus, the baseline findings appear to mainly describe pre-1990 authoritarian regimes but do not necessarily relate to more modern authoritarian regimes. Robustness tests In order to validate the reliability of the above findings, this study has also conducted further robustness tests. First, tests for multicollinearity appear unproblematic in all analyses. Secondly, as pointed out above, some of the variables had to be imputed in order to account for a high level of missingness. The results for the non-imputed dataset are shown in the appendix (Tables 3 and 4) and are generally in line with the results presented here. 1090 F. WIEBRECHT In addition, some control variables such as the oil income, the regime duration, and whether the regimes succeed democracies are far from being significant and show small effect magnitudes. In order to avoid the potential danger of model overspecification, I have also conducted the same analyses without these variables (Tables 5 and 6 in appendix). The results for the variables of interest remain largely unchanged both in terms of significance and effect magnitudes. Furthermore, some variables have been replaced with alternative measurements. Further analyses have utilized a country’s total resource income rather than the oil income (Tables 7 and 8 in appendix). In an additional robustness test, the latent measure of personalism has been replaced by Geddes’ (1999) original typology to see whether personalist regimes are also linked to weaker legislatures. While the magnitudes are slightly weaker for personalist regimes the overall effects remain robust (Tables 9 and 10 in appendix). In addition, the historical differences could be driven by the collapse of many communist regimes with similar institutional makeup around the same time. However, excluding them from the analysis does not yield substantively different results (Figures 1 and 2 in appendix). The trend of declining seat shares of the authoritarian incumbents is also apparent when excluding communist regimes (Figure 3 in appendix). Discussion and conclusion This article has offered an empirical analysis of the factors that are linked to legislatures’ strength in authoritarian regimes. It provides empirical evidence for the effects of factors related to both power-sharing and co-optation theory through a large-N cross-country study on authoritarian regimes from 1946 to 2010. The results show that three factors are critical for legislative strength, namely the level of personalism and the existence of an opposition and elections. However, beyond these factors, it appears that authoritarian regimes that have more control over legislative and electoral processes are associated with more powerful legislatures. Yet, these findings seem to apply primarily to the time period before 1990. For regimes after that and for de facto legislative powers stronger legislatures tend to coexist with more competitive mechanisms and less authoritarian control over electoral and legislative processes. The results highlight how different institutions go hand in hand in authoritarian regimes and suggest that legislatures have been used as essential “bargaining chips” for dictators in the past. When a dictator had relatively little control over electoral and legislative outcomes, legislatures also tended to be weaker. In other words, dictators could afford to have less control over legislative processes when the stakes, that is, the strength of the legislature, were lower. For de facto powers and also the historical period after the end of the Cold War the reverse appears to be the case. These findings are closer in line with expectations from the general literature, namely that more competitive processes strengthen political institutions. [62] Consequently, as a dictator has less control over elections and legislative outcomes the legislature is generally also stronger. These results also echo some recent findings from Africa that highlight less party cohesion and less dictatorial meddling in legislative processes as important factors for legislative development in authoritarian regimes. [63] These findings demonstrate the changing nature of parliaments in authoritarian regimes in the post-Cold War era. After the Cold War, legislatures have become a DEMOCRATIZATION 1091 place for more contestation between incumbents and opposition for instance illustrated by the decline of the authoritarian incumbents’ seat and vote shares. Thus, some mechanisms primarily related to power sharing appear to be less important in the post-1990 authoritarian legislature. For instance, personalism and the central candidate selection have a smaller effect magnitude or lose their significance. Before 1990, authoritarian legislatures were less competitive and many of them were only filled with members of the ruling elite. However, in more recent years legislatures include more members of the opposition as well and these factors tend to have less impact on legislatures. There are several avenues for future research to build upon the findings of this article. First of all, scholars should disentangle some of the relationships found in this article and highlight deeper causal mechanisms between the different variables. Secondly, as prior case studies have pointed out that legislative strength also has effects in itself future research should investigate these effects on a more comparative basis. Several research questions on the consequences of differences in legislative strength may be worth pursuing. For instance, do stronger legislatures generally strengthen dictatorial rule or do they, in contrast, contribute to democratization? Prior research has highlighted that legislatures encourage citizens to voice their grievances in authoritarian regimes and thus, may set in motion democratizing processes. [64] In addition, it may be worthwhile to analyse the redistributive consequences of stronger legislatures across regimes and policy outcomes in authoritarian regimes more broadly. Notes 1. Gandhi, Political Institutions; Wright, “Do Authoritarian Institutions Constrain.” 2. Geddes, “Democratization After Twenty Years.” 3. Adiputri, “Empowerment of Parliament”; Egreteau, “Towards Legislative Institutionalisation.” 4. Wilson and Woldense, “Contested or Established?” 5. Bonvecchi and Simison, “Legislative Institutions and Performance.” 6. Wilson and Woldense, “Contested or established.” 7. Brancati, “Democratic Authoritarianism.” 8. Gandhi and Przeworski, “Cooperation, Cooptation and Rebellion.” 9. Gandhi, Political Institutions; Svolik, Politics of Authoritarian Rule. 10. Truex, Making Autocracy Work; Manion, Information for Autocrats; Williamson and Magaloni, “Legislatures and Policymaking.” 11. Gandhi, Political Institutions. 12. Wright and Escribá-Folch, “Authoritarian Institutions.” 13. Bonvecchi and Simison, “Legislative Institutions and Performance”; Koeker, “Why Dictators Veto.” 14. Svolik, “Power Sharing.” 15. Grundholm, “Taking it Personal”; Magaloni, “Credible Power-sharing.” 16. Svolik, “Politics of Authoritarian Rule.” 17. Boix and Svolik, “Foundations of Limited Authoritarian.” 18. Truex, “The Returns to Office”; Lust-Okar, “Elections under authoritarianism.” 19. Noble, “Authoritarian Amendments”; Lü, Liu, and Li, “Policy Coalition Building.” 20. Bonvecchi and Simison, “Legislative Institutions and Performance.” 21. Gandhi, Political Institutions; Wright, “Do Authoritarian Institutions Constrain.” 22. Brownlee, Authoritarianism; Gandhi, Political Institutions. 23. Opalo, Legislative Development in Africa. 24. Collord, “Political Economy of Institutions”; Opalo, Legislative Development in Africa. 25. Gandhi, Political Institutions; Wright, “Do Authoritarian Institutions Constrain”; Woo and Conrad, “Differential Effects.” 1092 F. WIEBRECHT 26. Opalo, Legislative Development in Africa. 27. Bonvecchi and Simison, “Legislative Institutions and Performance.” 28. Collord, “Political Economy of Institutions,” 14. 29. Fish and Kroenig, Handbook of National Legislatures. 30. Wilson and Woldense, “Contested or Established?” 31. Ibid. 32. Ibid., 590. 33. Bonvecchi and Simison, “Legislative Institutions and Performance.” 34. Wilson and Woldense, “Contested or Established,” 590. 35. Gandhi, Noble, and Svolik, “Legislatures and Legislative Politics,” 3. 36. Schuler, “Position Taking.” 37. Fish, “Stronger Legislatures, Stronger Democracies.” 38. Gandhi, Political Institutions. 39. Albrecht, “Opposition Support Authoritarianism.” 40. Malesky and Schuler, “Nodding or Needling.” 41. Blaydes, Elections and Distributive Politics; Koehler, “Authoritarian Elections in Egypt.” 42. Malesky and Schuler, “Nodding or Needling.” 43. Svolik, Politics of Authoritarian Rule, 55. 44. Geddes, Wright, and Frantz, How Dictatorships Work. 45. Grundholm, “Taking it Personal”; Roessler, “The Enemy Within.” 46. Opalo, Legislative Development in Africa. 47. Collord, “Political Economy of Institutions.” 48. Stratmann and Baur, “Plurality Rule, Proportional Representation”; Krol, “Amending Legislatures.” 49. Collord, “Political Economy of Institutions.” 50. Geddes, Wright, and Frantz, How Dictatorships Work. 51. Wilson and Woldense, “Contested or Established?” 52. Coppedge et al., V-Dem Codebook v8. 53. Przeworski, Political Institutions and Political Events. 54. Geddes, Wright, and Frantz, How Dictatorships Work. 55. Geddes, “Democratization After Twenty Years”; Grundholm, “Taking it Personal.” 56. Haber and Menaldo, “Natural Resources.” 57. Fish and Kroenig, “Handbook of National Legislatures.” 58. Haber and Menaldo, “Natural Resources.” 59. Gandhi, Political Institutions. 60. Howard and Roessler, “Liberalizing Electoral Outcomes”; Levitsky and Way, Competitive Authoritarianism. 61. Levitsky and Way, Competitive Authoritarianism, 5. 62. Mayhew, Congress; Ferejohn, “Incumbent Performance.” 63. Collord, “Political Economy of Institutions”; Opalo, Legislative Development in Africa. 64. Woo and Conrad, “Differential Effects of ‘Democratic’ Institutions.” Acknowledgements The author wishes to thank Chen Xi, Pierre F. Landry, James F. Downes, Zheng Ruoting and two anonymous reviewers for comments and suggestions for revising the manuscript. Disclosure statement No potential conflict of interest was reported by the author(s). Notes on contributor Felix Wiebrecht is a PhD Candidate in the Department of Government and Public Administration at The Chinese University of Hong Kong. His research interests include authoritarianism, legislatures, and consultative institutions. DEMOCRATIZATION 1093 ORCID Felix Wiebrecht [http://orcid.org/0000-0002-9159-5024](http://orcid.org/0000-0002-9159-5024) Bibliography Adiputri, Ratih. “The Empowerment of Parliament in the Transition from an Authoritarian to a Democratic Regime: Indonesian Experiences and Problems.” Parliaments, Estates, and Representation 38 (2018): 49–62. Albrecht, Holger. “How can Opposition Support Authoritarianism? Lessons from Egypt.” Democratization 12, no. 3 (2005): 378–397. Beck, Nathaniel, and Jonathan N. Katz. “Modeling Dynamics in Time-series-cross-Section Political Economy Data.” Annual Review of Political Science 14 (2011): 331–352. Blaydes, Lisa. Elections and Distributive Politics in Mubarak’s Egypt. Cambridge: Cambridge University Press, 2012. Boix, Carles, and Milan Svolik. “The Foundation of Limited Authoritarian Government: Institutions, Commitment, and Power-sharing in Dictatorships.” Journal of Politics 75, no. 2 (2013): 300–316. Bonvecchi, Alejandro, and Emilia Simison. “Legislative Institutions and Performance in Authoritarian Regimes.” Comparative Politics 49, no. 4 (2017): 521–539. Brancati, Dawn. “Democratic Authoritarianism: Origins and Effects.” Annual Review of Political Science 17 (2014): 313–326. Brownlee, Jason. Authoritarianism in an Age of Democratization. Cambridge: Cambridge University Press, 2007. Collord, Michaela. “The Political Economy of Institutions in Africa: Comparing Authoritarian Parties and Parliaments in Tanzania and Uganda.” PhD diss., University of Oxford, 2019. Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Svend-Erik Skaaning, Jan Teorell, et al. V-Dem Country-Year Dataset v8. Varieties of Democracy (V-Dem) Project, 2018. Egreteau, Renaud. “Towards Legislative Institutionalisation? Emerging Patterns of Routinisation in Myanmar’s Parliament.” Journal of Current Southeast Asian Affairs 38, no. 3 (2020): 265–285. Ferejohn, John. “Incumbent Performance and Electoral Control.” Public Choice 50 (1986): 5–25. Fish, M. Steven. “Stronger Legislatures, Stronger Democracies.” Journal of Democracy 17 (2006): 5–20. Fish, M. Steven, and Matthew Kroenig. The Handbook of National Legislatures: A Global Survey. New York: Cambridge University Press, 2009. Gandhi, Jennifer. Political Institutions under Dictatorship. New York: Cambridge University Press, 2008. Gandhi, Jennifer, Ben Noble, and Milan Svolik. “Legislatures and Legislative Politics Without Democracy.” Comparative Political Studies 53, no. 9 (2020): 1359–1379. Gandhi, Jennifer, and Adam Przeworski. “Cooperation, Cooptation, and Rebellion under Dictatorships.” Economics & Politics 18 (2006): 1–26. Geddes, Barbara. “What do We Know about Democratization After Twenty Years?” Annual Review of Political Science 2 (1999): 115–144. Geddes, Barbara, Joseph Wright, and Erica Frantz. How Dictatorships Work: Power, Personalization, and Collapse. Cambridge: Cambridge University Press, 2018. Grundholm, Alexander T. “Taking it Personal? Investigating Regime Personalization as an Autocratic Survival Strategy.” Democratization 27, no. 5 (2020): 797–815. Haber, Stephen, and Victor Menaldo. “Do Natural Resources Fuel Authoritarianism? A Reappraisal of the Resource Curse.” American Political Science Review 105 (2011): 1–26. Howard, Marc M., and Philip G Roessler. “Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes.” American Journal of Political Science 50, no. 2 (2006): 365–381. Koehler, Kevin. “Authoritarian Elections in Egypt: Formal Institutions and Informal Mechanisms of Rule.” Democratization 15, no. 5 (2008): 974–990. Koeker, Philipp. “Why Dictators Veto: Legislation, Legitimation and Control in Kazakhstan and Russia.” Democratization 27, no. 2 (2020): 204–223. Krol, Gerrit. “Amending Legislatures in Authoritarian Regimes: Power Sharing in Post-Soviet [Eurasia.” Democratization (2020). doi:10.1080/13510347.2020.1830065.](https://doi.org/10.1080/13510347.2020.1830065) Levitsky, Steven, and Lucan Way. Competitive Authoritarianism: Hybrid Regimes After the Cold War. New York: Cambridge University Press, 2010. 1094 F. WIEBRECHT Lü, Xiaobo, Mingxing Liu, and Feiyue Li. “Policy Coalition Building in an Authoritarian Legislature: Evidence from China’s National Assemblies (1983–2007).” Comparative Political Studies 53, no. 9 (2020): 1380–1416. Lust-Okar, Ellen. “Elections under Authoritarianism: Preliminary Lessons from Jordan.” Democratization 13, no. 3 (2006): 456–471. Magaloni, Beatriz. “Credible Power-Sharing and the Longevity of Authoritarian Rule.” Comparative Political Studies 41, no. 4–5 (2008): 715–741. Malesky, Edmund, and Paul Schuler. “Nodding or Needling: Analyzing Delegate Responsiveness in an Authoritarian Parliament.” American Political Science Review 104, no. 3 (2010): 482–502. Manion, Melanie. Information for Autocrats: Representation in Chinese Local Congresses. Cambridge: Cambridge University Press, 2015. Mayhew, David. Congress: The Electoral Connection. New Haven, CT: Yale University Press, 1974. Noble, Ben. “Authoritarian Amendments: Legislative Institutions as Intraexecutive Constraints in Post-Soviet Russia.” Comparative Political Studies 53, no. 9 (2020): 1417–1454. Opalo, Kennedy Ochieng. Legislative Development in Africa. Cambridge: Cambridge University Press, 2019. Przeworski, Adam. Political Institutions and Political Events (PIPE) Data Set. 2013. Roessler, Philip. “The Enemy Within: Personal Rule, Coups, and Civil War in Africa.” World Politics 63 (2011): 300–346. Schuler, Paul. “Position Taking or Position Ducking? A Theory of Public Debate in Single-party Legislatures.” Comparative Political Studies 53, no. 9 (2020): 1493–1524. Stratmann, Thomas, and Martin Baur. “Plurality Rule, Proportional Representation and the German Bundestag. How Incentives to Pork-Barrel Differ Across Electoral Systems.” American Journal of Political Science 46, no. 3 (2002): 506–514. Svolik, Milan W. The Politics of Authoritarian Rule. Cambridge: Cambridge University Press, 2012. Svolik, Milan W. “Power Sharing and Leadership Dynamics in Authoritarian Regimes.” American Journal of Political Science 53, no. 2 (2009): 477–494. Truex, Rory. Making Autocracy Work: Representation and Responsiveness in Modern China. New York: Cambridge University Press, 2016. Truex, Rory. “The Returns to Office in a ‘Rubber Stamp’ Parliament.” American Political Science Review 108, no. 2 (2014): 235–251. Wiliamson, Scott, and Beatriz Magaloni. “Legislatures and Policy Making in Authoritarian Regimes.” Comparative Political Studies 53, no. 9 (2020): 1525–1543. Wilson, Matthew Charles, and Josef Woldense. “Contested or Established? A Comparison of Legislative Powers Across Regimes.” Democratization 26, no. 4 (2019): 585–605. Woo, Ae sil, and Courtenay R Conrad. “The Differential Effects of ‘Democratic’ Institutions on Dissent in Dictatorships.” The Journal of Politics 81, no. 2 (2019): 456–470. Wright, Joseph. “Do Authoritarian Political Institutions Constrain? How Legislatures Affect Economic Growth and Investment.” American Journal of Political Science 52, no. 2 (2008): 322– 343. Wright, Joseph, and Abel Escribà-Folch. “Authoritarian Institutions and Regime Survival: Transitions ” – to Democracy and Subsequent Autocracy. British Journal of Political Science 42, no. 2 (2012): 283 309.