BACKGROUND PAPER GOVERNANCE and THE LAW From de-jure to de-facto: Mapping Dimensions and Sequences of Accountability Staffan I. Lindberg V-Dem Institute, Department of Political Science, University of Gothenburg Anna Lührmann V-Dem Institute, Department of Political Science, University of Gothenburg Valeriya Mechkova V-Dem Institute, Department of Political Science, University of Gothenburg Disclaimer This background paper was prepared for the World Development Report 2017 Governance and the Law. It is made available here to communicate the results of the Bank’s work to the development community with the least possible delay. The manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally-edited texts. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Background Paper for the World Development Report 2017 From de-jure to de-facto: Mapping Dimensions and Sequences of Accountability Valeriya Mechkova V-Dem Institute, Department of Political Science, University of Gothenburg Anna Lührmann V-Dem Institute, Department of Political Science, University of Gothenburg Staffan I. Lindberg V-Dem Institute, Department of Political Science, University of Gothenburg 1 Abstract Accountability is one of the cornerstones of good governance. Establishing accountable institutions is a top priority on the international development agenda. Yet, scholars and democracy practitioners know little about how accountability mechanisms develop and thus can be supported by international and national actors. The present study tackles the questions of how, and in what order accountability mechanisms develop. We consider not only vertical and horizontal, but also diagonal accountability mechanisms (active civil society organizations and independent media) in both their de-jure and the de-facto dimensions. By utilizing novel sequencing methods, we study their sequential relationships in 173 countries from 1900 to the present with data from the new V-Dem dataset. Considering the long-term dimensions of institution building, this study indicates that most aspects of de-facto vertical accountability precede other forms of accountability. Effective institutions of horizontal accountability – such as vigorous parliaments and independent high courts – evolve rather late in the sequence and build on progress in many other areas. 2 Contents Introduction ........................................................................................................................................................ 4 Part 1: Concept and Measurement of Accountability ........................................................... 6 Conceptualizing Accountability .............................................................................................................. 6 Operationalization of Accountability.................................................................................................... 9 Part 2: Theorizing Sequences of Accountability Mechanisms ......................................... 15 The Implementation Gap: De-jure vs. De-facto Accountability ............................................... 16 Closing the Gap: Sequences in Accountability Mechanisms ..................................................... 19 Summary and Hypotheses ..................................................................................................................... 23 Part 3. Empirical Analysis ............................................................................................................ 25 Methodology................................................................................................................................................ 25 Global Trends .............................................................................................................................................. 27 Closing the gap between de-jure and de-facto ............................................................................... 30 Disaggregated sequencing analysis.................................................................................................... 32 Regional and Time Trends ..................................................................................................................... 38 Conclusions ...................................................................................................................................................... 44 References ............................................................................................................................................. 46 Appendix. ............................................................................................................................................... 46 3 Introduction Accountability is one of the cornerstones of good governance and, not surprisingly, it has grown to be a ‘fashionable’ topic to study (Lindberg 2013:203). Strengthening government accountability is ostensibly one of the most efficient ways to combat critical threats to the rule of law, such as corruption and clientilism, in developing and democratizing countries (WB 2005:11). Holding leaders accountable to the people is one of the fundamental pillars of representative democracy; hence, the concept of accountability also occupies a central place in the most widely used definitions of democracy (e.g. Dahl 1971, 1989; Schmitter et al. 1991). Establishing accountable institutions is moreover a top priority on the international development agenda, and is singled out under Goal 16 as a central target in the newly endorsed Sustainable Development Goals (UN Resolution: A/Res/70/1). Yet, scholars and practitioners know very little about how the complex and interrelated series of accountability mechanisms evolve. Arguably, solving that puzzle requires a systematic study of the patterns of long-term historical development across many countries. The present study presents such an approach by investigating the question: Are there generalizable sequences of accountability building? More specifically, the study focuses on whether certain types of accountability mechanisms are preconditions for others. If so, is the improvement of specific mechanisms more likely to lead to the enhancement of others? For accountability to be effective de-jure regulations need to be in place, but oversight, extraction of justification, and imposition of sanctions must also be effected in practice. In this paper, we map the sub-dimensions of de-jure and de-facto vertical, horizontal and diagonal accountability, and investigate the empirical relationships between them in terms of sequential developments. In this way, we tackle the questions of when and how some types and components of accountability relate to others. We argue that historical patterns of accountability building show a common trend in which it takes substantial time until de-jure accountability mechanisms evolve and start working efficiently in practice. In addition, we argue that developing high levels of de-facto accountability of any type reinforces positive changes in other dimensions of accountability too. Furthermore, we reason that the majority of the countries have set up high levels of 4 some or all de-facto institutions of vertical accountability, while the institutions of de-facto horizontal accountability that directly oversee and constrain the ruling elites develop last. Finally, we suggest that active civil society organizations and independent media (diagonal accountability) serve a key function in pushing for the implementation of the checks and balances between institutions in practice and thus we expect diagonal accountability to develop after vertical and before horizontal accountability. We test our results in different specifications and disaggregate the analyzed sample by regions and time. Based on intuitions found in the existing literature1 about the relationship between different types of accountability, and using the new V-Dem data (Coppedge et al. 2015a, b) as well as a novel methodology developed by Lindenfors et al. (2015) building on established approaches in evolutionary biology, this paper offers a distinctive depiction of sequences between a large number of variables, which has been not been possible until now. Providing an answer to how different aspects of accountability develop in sequence is critical from a policy perspective since this can inform strategies on what types of interventions should be supported before others in a particular country. For example, not only can one locate where country X is in the sequence of accountability, but one can also identify exactly which aspects this country needs to develop further in order to reach the next level in the sequence. 1 See for example, Ackerman (2003), Anderson (2002), Barton (2006), Considine (2002), Finn (1993), Fisse and Braithwaite (1993), Foweraker and Krznaric (2002), Goetz and Jenkins (2005), Hobolt and Klemmensen (2008), Hunhold (2001), Keefer (2007), Kelly (2003), Klijn and Koppenjan (1997), Maskin and Tirole (2004), Mulgan (2003), O’Loughlin (1990), Oliver (1991), Olukoyun (2004), Painter-Morland (2006, 2007), Philips and Berman (2007), Radin and Romzek (1996), Rakner and Gloppen (2003), Rosenau (1992), Schmitter (2004), Shenkin and Coulson (2007), Stirton and Lodge (2001), Stone (1995), Thomas (1998), Tsai (2007), Wang (2015a), Walker (2002), Weber (1999), Woods and Narlikar (2001), Wrede (2006). 5 Part 1: Concept and Measurement of Accountability Conceptualizing Accountability Accountability is defined most broadly as the processes through which decisions and actions are being called to ‘account’ (Jones 1992:73) by an external authority (Mulgan 2000:555) to demand information and explanations for the decisions made (Schedler 1999:14), and if needed, impose sanctions (Schedler 1999:14, Mulgan 2000:555) to the responsible individuals or institutions. In this process, public officials and institutions are supposed to be subject to continuous oversight and scrutiny both by other government institutions and citizens, in order to ensure that the former are meeting the public interest of the community (World Bank 2005:11). In the political realm, sub-types of accountability are typically organized in regards to the spatial direction of the accountability relationships (Lindberg 2013:212). In this categorization accountability is assumed to be a classical concept,2 where different types of accountability are nominal categories – capturing not more or less, only different components of accountability (Lindberg 2009:7). Vertical accountability focuses on mechanisms outside the state institutions (Schedler 1999, O’Donnell 1998) and concerns the relationship between the citizens and their elected representatives (Fox 2015, Mainwaring 2003). One of the most influential theories of democracy, put forward by authors like Schattschneider (1942) and Schumpeter (1950) among others, highlights elections as the main mechanism by which people exert control over politicians and political decisions. Through voting for competing candidates running on the tickets of various parties, citizens can select political leaders and punish them for improper behavior. This establishes a mechanism that induces responsiveness and accountability based on incentives for leaders who want to gain and keep power (Skaaning 2 The other types being radial and family resemblance (ibid, p.7). 6 2015:5). Faced with the uncertainty of re-election, incumbents should have to meet voters’ expectations in terms of policy- and decision-making in order to get re-elected (Olson 2000). In modern political systems, political parties serve an important mediating function in the process of elections and in establishing conditions for vertical accountability (Key 1949; Ranney 1962; Schattschneider 1942). Increasingly, political systems with strong parties are seen as being better equipped to implement accountability by, for example, resisting clientelism and corruption practices or by guarantying the provision of desired outcomes such as more public goods (Hicken et al. 2014; Hicken, Keefer 2013; Kitschelt 2000; Kitschelt et al. 2006; Pierskalla et al. 2011; Simmons et. al. 2014, Bernhard et al. 2015). Thus, stronger and more organized parties can impose constraints on leaders’ behavior more successfully, to the effect that leaders fulfill the party program and prevent illicit activities that would hurt the party’s reputation (Gehlbach et al. 2011; Svolik 2012). Parties are also instrumental in minimizing policy volatility by committing to a coherent line of action (Bernhard et al. 2015:8-10), which is essential for accountability as citizens will have better understanding of the policies politicians will pursue. Finally, parties also serve a coordinating function for different groups and interests in a society by adopting ideologies and policies intended to embrace a broad section of the society (Jankowski 1988). Thus, as parties represent a wide segment of the citizenry, their role as accountability actors becomes more relevant also between elections. By contrast, the oversight exercised between different institutions in a political system is commonly termed horizontal accountability. Typical institutions of horizontal accountability include legislative committees, which question governments about their actions, and have the power to hold them accountable by initiating a vote of no-confidence. The horizontal accountability mechanism thus emphasizes the separation of power in a state. This should ensure that checks between institutions prevent abuse of power by demanding information, questioning officials and possibly, punishing improper behavior (Rose-Ackerman 1996). In order to establish horizontal accountability, varying state agencies should not only have the legal power but also the will to oversee the actions of other agencies and impose sanctions (O’Donnel 1998: 117). This sub-type of accountability is thus the consequence of checks and balances between networks of institutions, including the 7 executive, legislative, and judicial branches, as well as various oversight agencies, ombudsmen, prosecutor generals, etc. (ibid: 119). Finally, diagonal accountability seeks to engage citizens directly in the exercise of accountability mechanisms (Pelizzo et al. 2013:8, Goetz et al. 2010: 364).3 Some authors use the term social accountability to describe the type of accountability that centers on civic action, in which citizens, civil society organizations (CSOs) and the media participate directly in the realization of accountable leaders and state institutions beyond the participation in elections (Malena et al. 2004:3). By empowering citizens and actively involving them in the monitoring of government performance, diagonal accountability mechanisms can enhance government transparency and exert sanction power via “naming and shaming”, thus potentially serving as powerful tools against corruption and ensuring that government agencies serve the interest of the people in a better way (World Bank 2006: 7). Peruzzotti and Smulowitz (2006:4, 10) also underline the growth of diagonal accountability mechanisms as a form of alternative political control, using informal tools (social mobilization and media exposure) to activate institutional tools (e.g. legal oversight by controlling agencies). Diagonal accountability mechanisms include a broad range of actions and mechanisms that citizens, civil society organizations (CSOs), and an independent media can use to hold the government and public officials accountable, such as “public demonstrations, protests, investigative journalism, and public interest lawsuits” (Malena et 3 Some scholars view what we call here diagonal accountability as part of vertical accountability (Fox 2015, p.347). Others argue that diagonal accountability should not be separated from the other horizontal or vertical accountability, because the success of actions undertaken by civil society organizations and media ultimately would depend on whether the formal institutions respond to them (Mainwaring 2003). However, formal institutions of accountability do not have the opportunity and the time to continuously oversee the day-to-day activity of the whole state apparatus (McCubbins et al. 1984, Grimes 2013). This is where citizens and media can serve an important “fire alarm” accountability function, monitoring government offices and reporting observed irregularities through their daily lives and professional activities (Grimes 2013:382). This combination of formal and informal institutions working together can result in the full-scale exercise of accountability. In addition, the purpose of this study is to investigate as comprehensively as possible the effects of different accountability mechanisms independently, which is why we have decided to separate accountability in three sub-types. 8 al 2004:3). In addition, when important policy is being considered, independent public deliberations among a wide range of groups, such as participatory public policy-making or budget discussions, constitute other means of diagonal accountability. It is important to underline that in order to ensure effective accountability of any kind, oversight mechanisms need to be working in practice and not just exist as inconsequential de-jure regulations and institutions. Therefore, an informative measure of political accountability needs to go beyond formal provisions and measure to what extent the formal means is employed in practice. If the two aspects diverge there is an “implementation gap”. By mapping separately de-jure and de-facto aspects of each of the three mechanisms of accountability across space and over time, we are able to investigate the implementation gap between de-jure versus de-facto accountability. Operationalization of Accountability Based on the brief review above, Figure 1 maps the conceptualization of accountability and identifies a measurement scheme with a combination of factual and evaluative indicators, using data from Varieties of Democracy (V-Dem) and the Comparative Constitutional Project (CCP).4 At the highest and conceptually most abstract level, we model accountability as consisting of three mechanisms: vertical, horizontal and diagonal. At the second and more disaggregated level, we distinguish between the de-jure and de-facto components, creating two separate measures for each mechanism: The first appraises whether the formal institutions that make accountability possible are in place (e.g. constitutional provisions for multi-party elections; legal provisions for extensive suffrage; statutes providing for formal independence of the judiciary; and so on). The second level depicts to what extent accountability mechanisms are realized in practice (e.g. free and fair elections without vote buying or systematic fraud; effective oversight by the judiciary; and so on). Most indicators come from the V-Dem dataset that provides more than 350 distinct indicators for 173 polities between 1900 and 2012. Data for 76 countries have been updated 4 The exact formulations of the questions are presented in Table A.1 in the Appendix. 9 up until 2015 and for 37 countries until 2014. The non-factual data are produced by aggregating expert coding, with a Bayesian ordinal IRT-model drawing on the over 2,600 country experts V-Dem works with to code the evaluative indicators. The raw ratings are aggregated to country-year point estimates with accompanying measures of uncertainty (highest posterior density) in the custom-designed Bayesian ordinal item-response theory model (for details, see Pemstein et al. 2015, Coppedge et al. 2015c). This unique dataset provides the set of nuanced and detailed indicators we need for the present proposed investigation, in sufficiently long time series. 5 5 We use the country-year point estimates from v6.2 of the V-Dem data set obtained with the V-Dem measurement model (see Pemstein et al. 2016). The measurement model aggregates the ratings provided by multiple country experts and, taking disagreement and measurement error into account, produces a probability distribution over country-year scores on a standardized interval scale (Coppedge et al 2016: 33, see V-Dem Methodology document). As the sequencing models we use require ordinal variables, we use the transformed ordinal version of the variables. To arrive at that point, we take the most likely ordinal value on the original codebook scale into which a country-year would fall, given the average coder’s usage of that scale. More specifically, we assign each country-year a value that corresponds to its integerized median ordinal highest posterior probability category over Measurement Model output (Coppedge et al 2016: 33). 10 Figure 1. Conceptualizing accountability and its sub-types 11 Vertical accountability entails directly or indirectly selecting, and possibly removing, the head of the executive, as well as the members of the legislature, through regular and competitive elections with universal suffrage where political parties are free to field candidates. Four indicators are used to capture the de-jure aspects of this type of accountability. First, we measure whether regularly scheduled elections for parliament and for the chief executive are on course or if there is an interruption in the electoral regime due to a coup, military defeat or similar event. Second, we consider whether it is legally possible for parties not affiliated with the government to form. Thirdly, we identify if laws allow for multiple parties to register for the elections, and finally, if they were held under universal suffrage.6 With the four de-jure indicators we form a “lexical” index, arguing that the indicators present a series of necessary and sufficient conditions arrayed on an ordinal scale. Each successive level is comprised of an additional condition, which defines the scale in a cumulative fashion7 (see Gerring et al. 2014, Skaaning et al. 2015) where the second variable only matters if the first is fulfilled, and so on. Following the operationalization by Skaaning and colleagues (2015), we order the indicators as follows: regular elections are held, parties are free to form, elections are multi-party, and there is virtual universal suffrage. Second, to capture the de-facto aspect of vertical accountability, we aggregate seven separate indicators. The first measures the extent to which electoral contestation is truly multi-party in practice and, second, the degree to which freedom to form political parties is unrestricted. Furthermore, we use measures of to what extent the electoral management body (EMB) has the autonomy from the government to apply election laws impartially; an indicator of to what extent elections are free and fair and not marred by fundamental flaws and irregularities, and fifth, to what extent vote-buying is present. In order to capture the mediating role of political parties, we include indicators on the extent of programmatic profiling of political parties (versus clientelistic linkages), gauging to what extent voters are 6 We define virtual universal suffrage to be achieved when 98% of the population is enfranchised - as is common practice in the field (Skaaning 2015). 7 “Condition A is necessary and sufficient for L1, conditions A&B are necessary and sufficient for L2, and so forth. In achieving these desiderata, four criteria must be satisfied: (a) binary values for each condition, (b) unidimensionality, (c) qualitative differences, and (d) centrality or dependence” (Skaaning et al 2015: 6). 12 de-facto presented with distinct platforms to use for evaluation or are bought off to disregard policy evaluation and thus policy accountability. Finally, we include a measure of the extent to which opposition parties are autonomous of the ruling regime, in order to determine the level of actual choice present for voters. The index is constructed by running the above indicators in Bayesian factor analysis (BFA) that takes uncertainty estimates into account. We ran one thousand BFAs and took the average for the resulting country-year estimates and measures of uncertainty for the index point estimates. For details, see the Appendix8 and Pemstein et al. (2015). In order to measure the horizontal mechanisms for accountability, we use indicators that measure the relationships between the different branches of government. To capture de-jure horizontal accountability, we construct an index by taking the average of the following four binary indicators.9 First, we account for whether a parliament that has the formal authority to legislate exists, and, second, whether it has the formal power to question members of the executive. Third, the index includes a measure of if the constitution explicitly declares central judicial organs independent and, fourth, whether there are provisions for an ombudsman, general or public prosecutor. To measure for the de-facto situation of horizontal accountability, we create an index (using the BFA procedure described above) by aggregating indicators gauging the situation in practice: How likely it is that the legislature would question the executive branch to explain its policies or testify, and would conduct an investigation and issue an unfavorable decision or report if the executive were engaged in unconstitutional or unethical activity? Second, how likely is it that other bodies such as a comptroller general, general prosecutor, or ombudsman would conduct an investigation and issue an unfavorable decision or report if the executive were engaged in unconstitutional or unethical activity? Third, we take into account how likely it is that judges are disciplined if they are found responsible for serious 8 Table A.4 in the Appendix reports the uniqueness scores for all variables constituting the indices for vertical, horizontal and diagonal accountability. 9 Taking the mean is standard practice when there is no strong theory available to guide the aggregation, which is the situation we face in this case. 13 misconduct, and, finally, how likely it is that judicial rulings from the high and lower courts merely reflect the government wishes in cases that are important to the ruling elite. We argue that media is one critical source of information empowering citizens to make informed political choices, during and between elections, and that a wide civic engagement in the political life is also a key instrument of what we term here diagonal accountability (Voltmer 2009: 139). To describe the de-jure institutional framework for diagonal accountability, we therefore combine two indicators reflecting whether the constitution provides for freedom of assembly, as well as an indicator of whether freedom of speech and freedom of the press is guaranteed in the constitution. For the de-facto index of diagonal accountability, we aggregate six indicators of media freedom and civic engagement using the BFA procedure described above. We include indicators measuring the extent to which governments censor major media; whether print/broadcast media outlets regularly voice critique of the government; and the extent to which media represent a wide range of political perspectives. Three indicators are included to describe the participatory environment for civic action: the extent to which civil society organizations (CSOs) are free to organize, associate, strike, express themselves, and to criticize the government without fear of government sanctions or harassment; an indicator of how wide and how independent are public deliberations when important policy changes are being considered; and finally, to what degree participation in CSOs is voluntary and there is wide popular involvement.10 In sum, for this study we utilize 35 detailed indicators of specific aspects of accountability from the V-Dem and CCP datasets. Based on these data, we create six original mid-level indices that capture de-jure and de-facto dimensions of vertical, diagonal and horizontal accountability. 10 We create ordinal versions of the resulting indices in order to be able to use them in the sequencing analysis (Lindberg 2016). 14 Part 2: Theorizing Sequences of Accountability Mechanisms The historical experience of institution-building has demonstrated that there is no single recipe for success. Creating an accountable and efficient government is a complex process requiring considerable time and idiosyncratic processes in order to develop a fully functioning mechanism of interdependent institutions. Nevertheless, a number of scholars argue that prior developments shape political outcomes to a significant degree (e.g. Mahoney 2001, Yashar 1997). Scholars, particularly in the tradition of historical institutionalism, emphasize that studying gradualism and sequencing is a worth-while endeavor (Wilson 2015: 72). In particular, discussions continue about the successful order for setting up practices and institutions in different areas (Carothers 2007, Fukuyama 2011, Mansfield et al. 2007). However, most contributions do not go much beyond how such sequential relationships look at the bivariate level. For example, some scholars argue that “it is dangerous to push states to democratize” if a necessary set of pre-conditions are not in place (Mansfield et al. 2007:6- 7). According to this analysis, a premature “out-of sequence” push to hold competitive elections in culturally diverse societies without reasonably effective institutions, is likely to fail and even lead to violence (e.g. Iraq and Lebanon). Contrary to that view, Carothers (2007: 20-21) reasons that having established the principles of rule of law, stable political institutions and accountability mechanisms are much more likely to develop “as part and parcel” of the process of democratization rather than separate from it. Similarly, for example, Howard and Roessler (2006) and Lindberg (2006) argue that countries that hold repeated elections, even if they are marked by fraud, are more successful in democratizing than countries that delay holding elections. Debating that conclusion, others like Gandhi and Lust- Okar (2009) suggest that elections have a stabilizing and legitimizing function for dictatorships. Another such bivariate argument about sequence is that a powerful legislature must develop first in order to curb executive corruption (Fish 2005a, 2005b). A case study in Ukraine shows that even in a system characterized by otherwise underdeveloped democratic practices, the opposition parties in a legislature can effectively hold the executive 15 accountable through legislative requests (Herron et al. 2015:132). On the other hand, scholars have found that conflicts between the legislature and the executive may ultimately lead to democratic breakdown (Stepan et al. 1993). Similarly, establishing a robust civil society is generally viewed as a precondition for the subsequent fall of authoritarian regimes as well as for building a resilient democracy (Bernhard 1993). Carothers (2007: 20) points out that the development of strong grass root movements (e.g. Solidarity in Poland, the African National Congress in South Africa) have often been necessary conditions, and taken place sequentially prior to democratic change in many countries. Other authors argue that putting in place institutions that channel the interests of the society to the government often precedes authoritarian stability (Gandhi 2008, Magaloni 2008, Berman 1997). Nevertheless, due to a lack of comparable data, scholars have been unable to study the interactive relationship of several institutions in a large number of countries and over substantial periods of time. For example, do powerful CSOs or the holding of elections systematically reinforce or weaken the emergence of a functioning rule of law, more accountable executives and legislatures or critical media? Current scholarship has little to contribute on how these aspects interact in sequential terms. It seems that historically genuine examples of “correct” sequencing and unproblematic political development are practically nonexistent (Berman 2007:16). However, the present paper seeks to start answering questions like whether eliminating arbitrary power in the executive branch begins with a stronger legislature, or alternatively from a competent judiciary and a vigorous civil society. The Implementation Gap: De-jure vs. De-facto Accountability It has become conventional wisdom that there is an important difference between the introduction of de-jure institutions of accountability and their de-facto implementation (Besley 2008: 37, Snyder 2006: 219). The Global Commission on Electoral Integrity (2012: 12) chaired by Kofi Annan, voiced the critique that “many authoritarian governments (…) 16 seek to wrap themselves in the veneer of democratic legitimacy.” To this end, regimes introduce pro-forma institutions but skilfully strip them of their power to constraint the executive (Gandhi et al. 2009). For example, in the twenty first century, almost all countries in the world hold some form of elections. However, this surge in elections does not necessarily imply a surge in vertical accountability. In the twentieth century, limitations in the quality of elections often came in the form of formal exclusion – for example, many countries did not allow women to vote until relatively late in the overall democratization process, while others placed significant de-jure barriers to forming a political party or running for elections. Nowadays, most countries fulfill the formal standards of de-jure vertical accountability – including full suffrage and freedom to stand for elections. Nevertheless, more than half of the current elections in the world violate the basic principles of freedom and fairness (Hafner-Burton et al. 2013:152). This includes obvious manipulation on Election Day – from stuffing ballot boxes to the fabrication of voting results, but also more subtle activities such as restriction of media freedom and opposition activities (Schedler 2002). Empirical studies have found that the formal independence of election management bodies (EMBs) does not protect elections from manipulation (Birch 2008). Such de-facto practices undermine the potential for vertical accountability inherent in the formal institutions pertaining to both procedural certainty and ex ante uncertainty of outcomes, which are the two key preconditions for vertical accountability (Przeworski 1986: 56–57). Similarly, although parliaments with constitutionally guaranteed rights of oversight have been established in most countries of the world, scholars have illustrated the limited extent to which they can effectively exercise their prerogatives in terms of scrutinizing the executive and holding them to account (Salih 2005 Ed., Rakner and van de Walle 2009, Vliet 2014). Even today many authoritarian regimes use legislatures as part of their co-optation and hence stabilization strategy (Gandhi 2008). For example, in many African countries, the division between legislative and executive powers necessary for horizontal accountability is undermined in practice due to limits in the implementation of existing constitutional frameworks (such as impeachment procedure and long presidential term limits), one-party- dominance, and concurrent elections for presidency and parliament (Cranenburgh 2009: 17 64). For example, Namibia and Zimbabwe’s presidents successfully removed constitutional limits on their terms as president. Finally, media is widely recognized as a key instrument for holding governments accountable. Yet, journalists are in practice often severely restricted even if freedom of expression and freedom of the press are guaranteed formally in the constitution. Russia is a contemporary example of non-coercive capture of key media outlet by the government effectively obstructing free and critical reporting (Besley et al. 2002: 720, citing Freedom House). Although the Russian constitution formally guarantees freedom of the media, in practice state-controlled companies own all national broadcasters, and journalists tend to be submissive to the regime. Based on prior research we therefore expect a time lag between the formal introduction of accountability-related institutions and the development of de-facto accountability, at least in the contemporary era. The time-lag is typically attributed to three factors. First, some authors stress that societies and states need a certain level of capacity and development for the de-facto implementation of accountability mechanisms (e.g. Fukuyama 2011). Second, profound institutional change is thought to take time due to the stickiness of historically developed processes and patterns (e.g. Katznelson and Weingast 2005). Third, accountability is about constraining the exercise of power. Naturally, those in power have a strategic interest in undermining the de-facto effectiveness of institutions of accountability while oppositional forces wish to have stronger means to expose the illicit practices and failing decisions of the ruling elites. Many contemporary rulers have decided this struggle in their favor by introducing de-jure institutions but manipulating them in order to limit the constraints on their rule (Schedler 2013). Even alternations in power often produce disappointing results in terms of advancing the rule of law, impartiality, and corruption because new political elites generally find themselves with the same vested interests as the old ruling group (Lindberg 2003). 18 Closing the Gap: Sequences in Accountability Mechanisms In order to understand how de-facto accountability evolves, we move beyond explaining the de-jure introduction of accountability-related institutions. As discussed above – such institutions need to be effective in practice in order to constrain the actions of the ruling elites. For example, legislatures were already introduced in many European countries before the 20th century, but it was not until much later that they actually developed the power to effectively oversee government actions. In Germany – for instance – the emperor had already established a national parliament in 1871, but full de-facto parliamentary oversight did not evolve until after World War II.11 Therefore, the evolution of de-facto accountability is not inevitable, but results from a political struggle among elites and between elites and citizens. In this process – so we theorize – citizens and opposition elites create a demand for more de- facto accountability, whereas ruling elites tend to try to keep their power unconstrained. The three sub-types of accountability constrain the power of the ruling elites differently. As discussed earlier, based on Schedler (1999) and others, we conceptualize accountability to have two main dimensions: availability of information and power to enforce sanctions. We argue that some of the sub-types of accountability are more effective in one of the dimensions than others (Table 1). Table 1: Effectiveness of the three sub-types of de-facto accountability in the information and enforcement dimension of accountability Type of accountability Vertical Diagonal Horizontal Dimension of accountability Information Low High High Enforcement High Low High 11 As captured by the V-Dem data, for 1949 Germany receives for the first time the highest score for the indicator Legislature investigates executive in practice (v2lginvstp), which answers: “if the executive were engaged in unconstitutional, illegal, or unethical activity, how likely is it that a legislative body would conduct an investigation that would result in a decision or report that is unfavorable to the executive ”. 19 Vertical accountability between citizens and elites can be an effective enforcement tool but is also often considered a “long route” to implement accountability (World Bank 2004). Ruling elites have multiple instruments to deceive and manipulate the electorate (Schedler 2002). In addition, elections provide opportunity for exercising accountability only in periodic intervals. Furthermore, citizens often lack accurate information to correctly assess the performance of the government and their voting behavior is not only based on past government performance. For instance, studies have shown that allegations of corruption do not necessarily hurt re-election prospects (Chang et al 2010, Costaz-Perez 2010, Dimock 1992). Such problems undermine the effectiveness of elections as a direct sanctioning mechanism for unethical behavior. Conversely, the strength of effective diagonal accountability mechanisms lies in the realm of the uncovering and provision of information. For instance, media are the main source of information for many citizens and hence a vital link between the government and citizens. Without sufficient information, accountability cannot be implemented in practice (Besley 2008: 37; Schedler 1999:21). Thus, investigative journalists and watch-dog NGO’s are key for uncovering unethical behavior. Furthermore, civil society associations can oppose “the power of the central body”, thus performing important “external power functions” (Lipset 1956:80). Nevertheless, their power to enforce sanctions is limited as they are dependent upon the institutions reacting to their messages and campaigns. However, it is more difficult for ruling elites to evade fully effective and independent horizontal oversight mechanisms. This notion builds on authors like Laver and Shepsle (1999) who argue that in most democracies, the parliament is the key intermediary institution to probe the ruling government. Self-confident, independent and capacitated legislatures, high courts, and other oversight bodies have institutional incentives and the power to monitor the actions of the executive on a day-to-basis. They cannot easily be deceived and have ways of accessing information not available to ordinary citizens. For example, in many countries such as Germany, independent courts of auditors have the right to thoroughly scrutinize records of public expenditure. Their regular reports are important tools for parliamentarians and journalists to hold the government to account. At the same 20 time, powerful parliaments – for instance through votes of non-confidence – and high courts have the power to directly sanction ruling elites. Hence, we argue that for ruling elites it is less costly to improve institutions that are further away from the center of power – such as elections – than to allow institutions that can constrain the effectiveness of their day-to-day actions. Thus, we reason that full de-facto vertical accountability is achieved first in the sequence of accountability building, whereas full de-facto horizontal accountability comes last. Furthermore, we expect that progress in one type of accountability can push countries to achieve higher levels in other types of accountability. There is evidence that introducing de- jure accountability-related institutions creates demands for more de-facto accountability. The struggle between ruling elites and citizens intensifies when the opposition actors come to possess sufficient de-facto organizational power to start demanding political recognition and inclusion (Acemoglu and Robinson 2006, Boix 2003). When the leaders are unable or unwilling to repress demands – when the costs of repression are greater than the costs of toleration, to use Dahl’s (1971) phrase – they increase co-optation and inclusion of non-elites in the political processes (Acemoglu and Robinson 2006). For example, several studies demonstrate that the introduction of de-jure competitive elections tends to push countries towards greater respect for civil liberties (Lindberg 2009 Ed.). Improvements in the freedom and fairness of elections can increase the willingness and ability of opposition parties to insist that the legislature actually investigates the executive and holds it accountable. Furthermore, repeated participation in elections can change citizens’ expectations towards political regimes, encouraging them to demand democratic procedures and broader participation (Gandhi et al. 2009: 415). If a sufficient number of actors believe that vertical accountability is the new game in town, the formal introduction of accountability institutions may turn into a “self-fulfilling prophecy” (Lindberg 2009: 335). In the process of elections, political parties serve an important vertical accountability function by aggregating the interests of citizens. As an institution capable of connecting popular groups to the national government, therefore, the general structure of political parties reinforces a system based on vertical accountability (Wilson 2015: 71). Likewise, if 21 opposition parties are allowed to contest elections, they are likely to campaign for less media censorship and more effective parliamentary oversight mechanisms. Therefore, we expect that opening space in one of the aspects of accountability would create opportunities for improvement in other aspects too. Furthermore, in addition to vertical accountability, enhancing the role of media and civil society is likely to re-enforce the transition from de-jure to de-facto accountable governments. In conditions of serious restrictions on the work of the media by the government, elected politicians are more likely to engage in illegal activities as they assume that they are less likely to be publicly exposed, while at the same time incompetent or corrupt politicians are less likely to be identified and thereafter punished in elections or otherwise (Besley and Prat 2002: 721). Civil society organizations can provide an institutional structure that integrates the interests of mass-based groups (Wilson 2015:71) and can also play a vital role in enforcing accountability. A robust civil society is generally viewed as a critical institution for building a resilient democracy (Bernhard 1993) and is increasingly seen as a critical component of a broader effort to hold governments accountable beyond elections (Besley 2006; Johnston 2005b; Peruzzotti and Smulovitz 2006). Associations can have important internal functions in communicating new opinions to the population and serving as an arena in which citizens become knowledgeable about politics, thus increasing the political awareness and involvement of their members (Lipset et al 1956). Putnam adds that the performance of institutions is “shaped by the social context within which they operate” (Putnam 1993:8). His argument is that civil society contributes to the effectiveness of democratic mechanisms by enhancing social collaboration, articulating and aggregating the interests of groups (Putnam et al 1993). Thus a vibrant civil society is key to effectively raise awareness on important societal issues and demand specific actions from the ruling governments. Thus, we expect a working system of diagonal accountability mechanisms to create demand for improved horizontal accountability. For instance, a well-informed, organized and active citizenry is likely to push for better governance (Grimes 2013: 381). One successful example of a civil society organization using the media to improve horizontal accountability is the recent campaign for reforms to the judicial system in Argentina. The non-profit organization 22 Asociación Por Los Derechos Civiles (ADC) led a coalition of NGOs in a campaign that resulted in the president providing for parliamentary and public hearings for Supreme Court of Justice nominees (Fisher 2013: 238). Similarly, NGOs spearheaded judicial reforms at provincial level in Argentina. Several NGOs held a preparation course for judicial nominees for the provincial government of Santiago del Estero in 2008. As result, within one year, the selection of judges was removed from political control and now takes place under the auspices of the Council of Magistrates, and is monitored by NGOs (Fisher 2013:239). However, in some contexts civil society organizations and media are not given the opportunity to exercise the aforementioned key functions. For example, recent studies have found that civil society may push for better quality of government only when conditions such as political competition, or a minimum level of government transparency and press freedom, are already in place (Grimes 2013). To sum up, we expect that full- de-facto horizontal accountability develops relatively late in the sequence. Such institutions have the power to effectively question and sanction the actions of ruling elites and hence, enhancing their effectiveness is not in the elites’ self- interest. Conversely, concessions for improvement of diagonal and vertical accountability are less costly at first sight for ruling elites. However, once improved, such institutions are also likely to create the demand for effective horizontal oversight. Ultimately, only if the principal has the power to enforce their will and sanction the agent, will effective accountability relationships evolve. For such an endeavor, de-facto horizontal accountability is indispensable. Summary and Hypotheses Changing formal institutions should be easier than changing de-facto practices. Furthermore, in many countries transition processes get stalled and high levels of de-facto accountability are not achieved even though de-facto institutions of accountability have been introduced. Therefore, 23 H1: After the formal introduction of institutions of accountability, it takes substantial time until de-facto accountability evolves. This implementation gap often remains for a long time. The emergence of de-facto accountability is not inevitable but subject to power struggles within the elites and masses. Ruling elites are likely to make initial concessions in areas of de-facto accountability that are less effective in directly constraining their actions. Citizens often lack accurate information about the behavior of elected officials, limiting their ability to effectively sanction leaders. Tools of diagonal and horizontal accountability – such as powerful legislatures and “watchdog“ CSOs – are more effective in constraining elites. Hence, the emergence of vertical accountability is less threatening to ruling elites and therefore more likely to evolve before de-facto horizontal and diagonal mechanisms. At the same time, vertical accountability and gradual improvements in diagonal and horizontal accountability create the demand and incentives for de-facto horizontal and diagonal accountability to evolve. Such demand-mechanisms can evolve at the level of elites as well as citizens. Thus, H2: High levels of de-facto accountability of any type need progress in the other dimensions of accountability. H3: High levels of de-facto institutions of vertical accountability develop before other institutions of accountability. Finally, in the realm of horizontal accountability, institutions with the capacity to directly oversee and constrain ruling elites are most costly for them. Therefore, elites may be ready to de-jure introduce parliaments and high courts, but can be expected to be reluctant to allow such institutions to become fully effective. Only if an engaged civil society, critical media and active opposition parties develop a strong demand for them, are elites likely to make substantive concessions in this realm. Hence, 24 H4: Institutions of horizontal accountability that directly oversee and constraint the degree of freedom of ruling elites become de-facto fully effective relatively late in the sequence. Part 3: Empirical Analysis Methodology To identify patterns of accountability development, we use a set of novel analytical approaches only now being introduced to the field of political science. To depict failing and successful sequences of democracy, the V-Dem project has developed methods adapted from evolutionary biology. The methods are more fully detailed in Lindenfors et al. (2015). The first application of such methods found that women’s relative equality in the area of civil liberties is close to being a necessary condition for successful democratization (Wang et al. 2015). In brief, in this paper we use the following two sequencing methods 1) graphical investigation of changes: revealing the exact pathways for how variables change in relation to one another; and 2) dependency analysis, exploring whether the values of one variable are systematically conditional on certain values of other variables. The latter method is inspired by “the contingent states test”, developed to investigate dependencies in biological evolution (Sillén-Tullberg 1993).12 It is suitable for analyzing the conditional relationships between different accountability types and mechanisms. If and when one can establish absolute dependencies, this is evidence of sequential contingencies in the data that are not easily inferred from regular regression statistics. For each value of 12 In particular, Sillén-Tullberg (1993) investigates whether the warning colors of butterfly larvae developed in solitary species (which would indicate that the selection happened on the individual level and that the individuals colored with colors that warn that they taste bad, survive when birds try to eat them), or whether the warning colors of butterfly larvae developed in species that lived in groups (which would indicate that the selection happened on the genetic level – that the individuals with warning colors die when birds try to eat them but their relatives, which have the same warning colors, get an advantage). By using phylogenetic models, the paper reconstructs the evolution of both solitary and group living species with warning colors and concludes that in the majority of cases individuals that live in groups had already developed warning colors, which means that the evolution happened on the individual level. 25 one variable, we scan the dataset for the lowest value in the other variables. We then exclude the lowest 5% of observations following the convention regarding similar methodological applications, in order to reduce the risk that outliers drive our results (Lindenfors 2015:10). If high values in Variable A always correspond to a certain minimal value of Variable B, then it can be inferred that the high values of Variable A are likely to be conditional on this minimal value of Variable B. Conversely, if for the highest value of Variable B, the corresponding value of Variable A is its minimum, then this shows that Variable B is not contingent on Variable A. For an analysis of sequential relationships between a larger number of variables, dependency tables can be constructed for all possible combinations of variables, and then summarized. An example of how several such bivariate dependency tables can be summarized is found below in Table 2. For each variable, we summarize the minimum values in other variables and report them as number of “Contingency conditions”. In this illustration, the maximum sum of thresholds, or contingency conditions, for a variable reaching its highest state, is 20 (five other variables, and each variable’s maximum level is four, for the highest state). The illustrative results would indicate that Variable B comes first in attaining its maximum value in a sequence. It can reach its highest state unconditional on any other variables. Table 2. Example for a contingency conditions table Variable # Contingency conditions A 16 F 14 C 6 D 5 E 3 B 0 Variables E, D, and C constitute a middle group with some conditions required for them to reach their highest states. The low number of dependencies indicates that the variables, on which their highest states are conditional, are relatively low states on other variables. Finally, variables F and A are the “late-comers” that have only been observed at their highest states after a greater number of other variables reach their highest, or close to their highest, 26 states. Together, this indicates a rough sequence that can be instructive for analysis of direct policy relevance. For our study, the dependencies of the highest indicator states are of particular interest, because we are interested in what these conditional relationships look like for developing the de-facto accountability mechanisms. If one were, for example, interested rather in the onset of such developments, one should look at the number of dependencies for different variables reaching the first, or perhaps the second level, which would indicate “early moves” rather than “final push”. Global Trends For analyses of sequential relationships, we construct dependency tables, which investigate all possible combinations and the number of dependencies between the indicators we have identified, as mechanisms of the three types of accountability. A low number of dependencies for a variable suggest that there are few contingency conditions for it to assume the highest level. Conversely, a high number of dependencies for a variable indicate that it cannot develop to “perfect” (the highest level) before many other variables have reached high levels. At the very least, the analyses below tell us that this has not happened in the history of states since 1900, across some 17,500 country-years. This is arguably rather strong evidence that it is unlikely to ever happen. Thus, we can present evidence on which mechanisms develop first, in the middle, and last in the processes of building accountability. Table 3 below presents the resulting aggregate summary of all bivariate analyses of the indices for the de-jure and de-facto measures of accountability, using the dependency analysis approach. The table displays the number of contingency conditions for each of the variables reaching their highest state (the top category). Table 3. Number of conditions required to reach the highest state De- Contingency jure/De- Type of accountability conditions facto (max=18) De-facto Horizontal Accountability 10 27 De-facto Vertical Accountability 10 De-facto Diagonal Accountability 9 De-jure Vertical Accountability 3 De-jure Diagonal Accountability 2 De-jure Horizontal Accountability 2 When interpreting the results one should not draw any strong conclusions from small differences in the number of dependencies and contingency conditions presented in such a table, but we could draw inferences on sequence mechanisms from large differences (Lindenfors 2015: 24). The de-jure measures for accountability require significantly fewer conditions before reaching their highest score, while the three de-facto measures require the attainment of high levels on more variables. These results support the theoretical expectations, outlined in Hypothesis 1, that there is often a considerable time gap between the achievement of de-jure and de-facto accountability. Figure 2 illustrates the development of the de-jure vertical accountability (x-axis) and its de- facto implementation (y-axis) further. The size of the dots on Figure 2 varies with the number of countries that have reached the specific combination of de-jure and de-facto accountability. For example, the relatively small dots to the left of the diagonal line indicate that only a few cases have reached high levels of de-facto vertical accountability before achieving the highest level of de-jure accountability. This indicates that most countries held multi-party elections under universal suffrage – hence scored a 4 on our de-jure index - before achieving full de-facto vertical accountability. The arrows on Figure 2 illustrate the pathways of countries moving from one combination of de-jure and de-facto vertical accountability to another. Thick arrows indicate that such pathway occurred more often than pathways with thin arrows. It is noteworthy, that there is no strong direct connection between the lowest and highest states of de-facto vertical accountability. Conversely, even after achieving full de-jure vertical accountability, de-facto accountability remains limited in many countries – for instance when elections are marred by fraud and unfair competition, political parties are not autonomous from the ruling regime and/or do not have programmatic platforms. 28 Figure 2. Vertical Accountability: De-facto versus De-jure These results are in consonance with previous findings in the literature that regimes typically install de-jure institutions of vertical accountability but limit their effectiveness in practice for a significant period. A cursory reading of political history is enough to corroborate that this is often an effect of deliberate maneuvering on behalf of rulers. The well-documented brokerage systems used to buy votes in places such as Argentina, Brazil, and Mexico is one example (Stokes et al 2013). Use of violence and intimidation in Nigeria to curb opposition parties’ efforts in Nigeria (e.g. Bratton 2008, Collier and Vicente 2013), various forms of irregularities to tilt the elections and evade elections monitors in Ghana (Ichino and Schündlen 2012), as well as the systematic efforts to exclude black voters in the United States, are other examples of deliberate measures by ruling elites to undermine de facto vertical accountability. Yet, even if a ruling coalition makes sincere efforts, it takes time and continuous reform, including institutions like an autonomous and impartial electoral management body, before de-facto vertical accountability is fully realized. These results corroborate the intuition about accountability and the way it develops. At the same time, the analyses at this level of aggregation of indicators to indices do not allow us to say anything with much specificity. 29 Closing the gap between de-jure and de-facto Hence, we explore how much time it has taken countries to close the gap between de-facto and de-jure accountability. An implementation gap between the de-jure and the de-facto situation occurs when the formal laws and institutions for exercise of vertical, diagonal, and horizontal accountability are in place but their actual functioning is impaired. Based on hypothesis 1, we expect a significant time lag between the introduction of de-jure accountability mechanisms and their realization in practice. In order to investigate the implementation lag we calculate the average time between the achievement of the highest levels of de-jure accountability and the highest levels of de-facto accountability. In most countries, political regimes broke down on one or more times during the studied time period as a consequence of coups, civil war, and similar events. Hence, these countries have had more than one “spell” of de-jure accountability, understood as a non- interrupted period with the maximum score on our de-jure accountability index. In order to simplify the analysis, we focus on the first time de-facto accountability was reached in the most recent de-jure spell from 1900 to 2012. For vertical accountability we use the highest values on our de-jure and de-facto indices described above. For horizontal accountability, we focus the analysis on legislatures and calculate the time from the introduction of a legislature (de-jure) until the legislature investigates the executive in practice.13 Table 4 shows that it took a considerable amount of time after the de-jure institutions were in place until de-facto accountability was achieved for both vertical and horizontal accountability.14 In many countries this gap is yet to be closed. This supports the first 13 This corresponds to a value of 4 on the V-Dem indicator v2lginvstp (Legislature investigates executive in practice, Coppedge et al 2016). We do not use the indices for de-jure and de-facto horizontal and diagonal accountability, because the de-jure indicators coming from the CCP data set for the de-jure horizontal and diagonal accountability are missing for many observations, particularly in the beginning of the twentieth century. This does not seriously impede the sequencing analysis, but would limit our ability to calculate the time gaps. 14 The detailed calculations of the time gap are presented in bar charts in the Appendix (Figure A.1 and A.2). 30 hypothesis. However, some countries that introduced elections early in the last century are an exception. 15 Those countries often achieved high levels of de-facto vertical accountability before granting universal suffrage - which in our theoretical framework is a necessary condition for de-jure vertical accountability. For example, Sweden achieved full de-facto vertical accountability in 1912 but women got the right to vote only in 1921. Second, the vertical accountability gap seems to close faster than the horizontal gap. This result is line with our expectation that the “sharp” tools of horizontal accountability are always found last in the sequences we analyzed (Hypothesis 4). Yet, the magnitude of difference is noteworthy. It took on average 11 years for the vertical accountability gap to close, while for parliaments it took on average 36 years to develop fully effective oversight mechanisms. The results indicate that closing the horizontal accountability gap could be a considerably thornier and lengthier process than closing the same gap for vertical accountability. This is also indicated by the fact that for a larger share of countries the gap between de-jure and de-facto horizontal accountability in the realm of legislatures is yet to close. In 2012, 82 percent of countries still face an implementation gap with regards to effective parliamentary oversight, while the corresponding figure for vertical accountability is only about half, at 47 percent. This suggests that it might be easier for countries to close the implementation gap between de-jure and de-facto accountability in the realm of vertical rather than horizontal accountability. Table 4. Time for closing the implementation gap between de-jure and de-facto accountability Vertical Accountability Horizontal Accountability (Legislature) Average number of years16 11 31 Share of countries without gap 47% 82% closure in 201217 16 Only includes the first time de-facto accountability was achieved in the most recent de-jure spell. Median vertical accountability: 3 years; Median horizontal accountability: 21 years. 17 Of countries that achieved in 2012 the maximum value on the de-jure vertical accountability index or had a National Legislature in place respectively. 31 Disaggregated sequencing analysis The following analysis disaggregates the indices of accountability and investigates how the 35 individual indicators of various mechanisms of accountability relate to one another. Table 5 presents the aggregate summary of 595 bivariate analyses following the dependency analysis approach outlined above, displaying the sum of contingency conditions for each of the variables reaching their highest state (the top category) with a 5% cut-off point as discussed earlier. For selected indicators detailed numbers of contingency conditions can be found in Table A.3 in the Appendix. The first thing to note is that almost all de-jure indicators have very few dependencies. This reflects the fact that several countries had achieved most aspects of de-jure accountability before making much progress on realizing any de facto aspects. These results are in-line with what we observed in the analysis of the aggregated indices reported above (Table 3). The only exception is the formal establishment of an ombudsman office, which comes relatively late in the sequence. The indicators of de-facto accountability have in general a greater number of dependencies. This finding supports the second hypothesis. Fully achieving any individual aspect of de-facto accountability seems to require progress in several other areas of accountability, because the different types of accountability reinforce each other. Table 5. Number of conditions required for reaching the highest state on each de-facto accountability indicator De Contingency Type of facto/De Indicator name conditions accountability jure (max 127) Legislature De facto Horizontal investigates in 62 practice De facto Horizontal Executive oversight 57 17 Of countries that achieved in 2012 the maximum value on the de-jure vertical accountability index or had a National Legislature in place respectively. 32 High court De facto Horizontal 54 independence De facto Diagonal Engaged society 48 Gov. censorship De facto Diagonal 48 Media De facto Vertical EMB autonomy 47 De facto Diagonal CSO entry and exit 42 Freedom of De facto Diagonal 38 discussion Print or broadcast De facto Diagonal 37 media critical De facto Diagonal CSO repression 37 Print or broadcast De facto Diagonal 34 media perspectives CSO participatory De facto Diagonal 30 environment Opposition parties De facto Vertical 27 autonomy De facto Vertical Election free and fair 25 De facto barriers to De facto Vertical 25 parties Lower court De facto Horizontal 24 independence De jure Horizontal Ombudsman 23 Legislature controls De facto Horizontal 17 resources De facto elections De facto Vertical 17 multi-party De facto Horizontal Judicial accountability 13 De facto Vertical Party linkages 10 De jure Vertical De jure party ban 6 De jure Vertical De jure multi party 6 Executive electoral De jure Vertical 5 regime index De facto Vertical Election vote buying 5 De jure Diagonal Freedom of the press 4 De jure Horizontal Legislature exists 3 Electoral Regime De jure Vertical 3 Index Legislative electoral De jure Vertical 3 regime index Attorney general, De jure Horizontal 3 Prosecutor Freedom of De jure Diagonal 3 expression Legislature questions De jure Horizontal 2 executive De jure Horizontal Judicial independence 2 De jure Diagonal Freedom of assembly 2 33 Share of population De jure Vertical 1 with suffrage For example, virtually no country in the world has fully ended CSO repression before achieving at least medium levels of freedom of discussion, freedom and fairness of elections and media that is able to routinely criticize the government.18 Furthermore, the expectation was that vertical accountability would evolve first in the accountability sequence (Hypotheses 3). The findings in Table 5 partially support this. Most indicators of de-facto vertical accountability require fewer contingencies than indicators of de-facto horizontal and diagonal accountability. The sequence pattern demonstrates that improving vertical accountability by diminishing vote buying in elections is something that can be achieved very early in the order of things. Conversely, some other mechanisms of accountability found high up in the sequence-ladder are dependent on countries solving the problem of vote buying before these other mechanisms can function properly. Similarly, the evidence in Table 5 reveals that getting elections to be multi-party de-facto and transforming party linkages from clientelistic to programmatic can also be achieved very early in the sequence of developing the three types of accountability mechanisms. However, there are two exceptions to the rule that mechanisms of vertical accountability develop first. Three of the indicators capturing aspects of horizontal accountability have fewer dependencies than key indicators of vertical accountability such as free and fair elections. These are lower court independence, financial independence of the legislature, and accountability of the judiciary. These results are intuitive. These three horizontal mechanisms belong to the “less threatening” mechanisms for an incumbent executive, compared to both the sharp real possibility of being removed if elections are fully free and fair, and to other much “sharper” tools for holding the executive to account (see below). 18 See Table A.3 that documents the specific contingencies for selected individual indicators. Contingency tables for the remaining indicators are available upon request. 34 Second, the “last holdout” for incumbents in the area of vertical mechanisms is manipulating elections through influencing election management. A fully independent electoral management body (EMB) requires more contingency conditions than most indicators of diagonal accountability, as well as the other mechanisms of vertical accountability. This result suggests that regimes give up this opportunity for controlling the outcome of an election rather late in the sequence, which seems plausible because the work of the EMB is largely done behind closed doors. The indicators of de-facto diagonal accountability cluster together in the upper half of the contingency table, which indicates that they require more conditions than most indicators of vertical accountability. In the realm of diagonal accountability, participation in CSOs seems to have the lowest requisites whereas common and unconstrained public deliberations have the highest. Thus, the results go in line with the theoretical expectations that the mechanisms of diagonal accountability serve as “watchdogs” and push for implementing accountability. Yet, all the mechanisms of diagonal accountability have fewer contingencies than the three mechanisms of horizontal accountability that develop last. These three variables have the greatest number of contingency conditions of all. They are specific mechanisms of horizontal accountability that directly oversee and constrain the actions of ruling elites. The indicators capture if it is likely that a) the legislature and b) other bodies (such as ombudsman, comptroller general or prosecutor) would conduct an investigation of the executive that would lead to an unfavorable report or decision for the executive, if the latter engaged in illegal activities; and c) the likelihood that the high court would rule independently on cases salient to the government, regardless of the government’s position. Table 5 provides evidence that no country has scored high on these three indicators without achieving significant progress in many other mechanisms of accountability. The legislature and other bodies can effectively hold the executive to account, and the high court can issue rulings independently only in an environment in which politicians are subject to regular and free and fair elections, citizens are free to organize themselves and express their political will through political parties and independent CSOs, and the media is able to scrutinize the work 35 of the government.19 This finding supports our fourth hypothesis that effective institutions of de-facto horizontal accountability develop last in the accountability sequence. These general findings based on 115 years of data from 173 countries also tally well with country experiences. Consider, for example, Ghana that after eleven years of civilian-cum- military authoritarian regime, returned to an electoral regime in the fall of 1992. The new constitution essentially installed all the relevant de-jure provisions for accountability immediately. However, the first presidential poll in 1992 was split between the National Democratic Congress (NDC, 58%) and other candidates (42%). Most of the losing vote went to the Danquah-Busia legatee of the New Patriotic Party (NPP), which refused to accept the outcome and withdrew from the subsequent first parliamentary contest (Morrison 1999). The 1992 elections were somewhat free and fair, by all accounts largely free of vote buying, and barriers for parties to form and participate were low. But there was evidence of irregularities and questions about the autonomy of the EMB. 20 Despite these problems, after the election, the legislature headed by a well-known liberal, Justice D. F. Annan, asserted its independence in control over its own resources. With the 1996 elections, opposition party autonomy was beyond doubt, and lower courts were clearly independent of the regime even though the ruling NDC and its leader President Jerry J. Rawlings remained in power. A fully independent and critical media that would openly challenge the sitting government did not develop until around the third elections in 2000. The opposition party NPP then won both a narrow majority in parliament, and the presidential office. Despite this electoral turnover, the legislature was still not fully capable of exercising executive oversight and conducting real investigations of illicit behavior by the executive. The new President John A. Kufour and his government even actively sought and managed to minimize the reach of the legislature’s oversight power and continued doing so 19 See Table A.3 which documents the specific contingencies for selected individual indicators. Contingency tables for the remaining indicators are available upon request. 20 There has been some debate about the actual level of fraud in the 1992 elections, but current evidence suggests that the irregularities could not have altered the outcome. For further details on these two elections, and the last one in 2000, see Boahen 1995; Green 1998; Gyimah-Boadi 1999, 2001; Jeffries 1998; Lindberg 2003; Lyons 1997; Ninsin 1998; Nugent 2001; Sandbrook & Oelbaum 1999; Oquaye 1995. 36 into the party’s second term from 2005 to 2008. The most important explanation for this circumvention of the legislature is to be found in the strategy of the President John Kufour to coopt members of the legislature in order to reduce high political competition (Lindberg 2009). As illustrated by the history of Ghana, many governments across the world resist full de-facto horizontal accountability for as long as they can, even if they came to power in free and fair elections. This finding also has important policy implications. International actors may seek to enhance the horizontal mechanisms of accountability that can directly constrain the regime, such as the legislature’s de-facto power and ability to investigate actions of the executive. However, our results suggest that such interventions are very unlikely to be fully successful unless a series of other mechanisms of accountability are in place and working in practice. In fact, the data tell us that virtually no country has ever managed to achieve full executive oversight through legislatures or other bodies, and judicial independence without fully free and fair elections and autonomous opposition parties, and developed civil society and media in the period from 1900 to 2012.21 The graphical presentation of the development of the three types of accountability shows similar results to the contingency analysis. Figures A.3, A.4 and A.5 in the Appendix compare the development of vertical, horizontal and diagonal accountability. The figures demonstrate that there is a large number of country-years when vertical accountability takes on higher values than horizontal (Figure A.3), while vertical accountability is higher than diagonal less frequently (Figure A.4). Finally, mechanisms of social accountability also display higher values than horizontal (Figure A.5). 21 Table A.3 documents the specific contingencies for selected individual indicators. Contingency tables for the remaining indicators are available upon request. 37 Regional and Time Trends The results presented so far are based on comprehensive analyses incorporating all 173 countries in the V-Dem data set from 1900 to 2012.22 In this section we disaggregate by exploring the patterns of accountability development by regions and by time. Regional Trends We split the sample by world regions in order to investigate regional trends using the same methods described above.23 Table 6 lists the de-facto accountability indicators sorted in descending order with the indicators with the highest number of dependencies at the top of the list, and the lowest at the bottom.24 The table contains a lot of information, but the key findings from the sequence of variables in the global sample hold across regions. The variables that necessitate the lowest number of conditions tend to be associated with vertical accountability (indicators displayed in red in the table); many diagonal accountability indicators (displayed in green) are concentrated in the middle of the table, and for most regions the indicators that come at the latest stage of development reflect horizontal accountability (blue indicators). Some exceptions to this pattern in Table 6 can also be found in the global sample: e.g. establishing autonomous EMB comes relatively late in time, while in some regions progress in terms of horizontal accountability, like financial independence of the legislature and judicial accountability, comes before reaching high levels on any other mechanisms of accountability. While the exact ordering sometimes varies a little, the indicators at the bottom, the middle, and at the top in the three types of accountability are the same as in the global analysis for most regions. There are a number of interesting differences in the progress of accountability mechanisms across regions. First, in some regions no country has reached the highest level on all 22 For 113 countries data is also available for 2013 and 2014, and for 76 countries 2015 is covered. 23 To divide the countries, we have used a politico-geographical classification scheme (e_regionpol) from the V- Dem data set v6 (taken from QoG (2013)). Due to the low number of countries and cases, we dropped the Pacific region (excluding Australia and New Zealand). 24 Table A.2 in the Appendix documents the full table with number of contingencies for each region. 38 accountability indicators. These are indicated by being crossed-out in Table 6. For example, no government has yet fully given up on media censorship or enabled the legislature to effectively investigate in practice in the MENA region (here including Turkey and Israel). Second, the pattern of development of vertical accountability seems to differ across regions. In most regions, vote buying is eradicated relatively early. However, in Western countries as well as in the Caribbean, vote buying persists longer than other deficits in vertical accountability – with the exception of EMB autonomy, which has been fully realized relatively late in the sequence everywhere. EMB autonomy comes particularly late in the sequence in Sub-Saharan Africa and South Asia, indicating that there governments have kept a backdoor for electoral manipulation open longer than other instruments for limiting accountability. Finally, free and fair elections are achieved rather late in the MENA region (if at all) and unlike in other regions, countries from the Caribbean have not developed programmatic relationships between political parties and citizens early in the sequence. Thus, interventions to help make the EMB fully autonomous should be synchronized with efforts to strengthen the other mechanisms of vertical accountability too. On the other hand, vote buying is something that can be addressed early in most regions of the world where weak mechanisms of accountability is an issue, and regardless of the state of other mechanisms being in place or not. 39 Table 6. Sequencing of accountability mechanisms by region Eastern Europe Sub- Western and Central Saharan Europe & South-East Asia Latin America MENA Africa USA East Asia Asia South Asia Caribbean Judicial Executive Gov. censorship Judicial Executive High court Judicial Executive Engaged accountability oversight Media accountability oversight independence accountability oversight society Legislature Legislature Executive EMB Engaged Lower court Executive EMB Executive investigates in investigates in oversight autonomy society independence oversight autonomy oversight practice practice Legislature Gov. censorship Lower court Executive Engaged Executive Gov. censorship High court investigates in Party linkages Media independence oversight society oversight Media independence practice Legislature Gov. Legislature Gov. Lower court High court Lower court Engaged society investigates censorship Engaged society investigates in censorship independence independence independence in practice Media practice Media Legislature Gov. Legislature Judicial CSO entry and Judicial investigates in Engaged society CSO repression censorship EMB autonomy investigates accountability exit accountability practice Media in practice High court CSO entry and High court Freedom of CSO entry and CSO entry and Freedom of CSO entry EMB independence exit independence discussion exit exit discussion and exit autonomy Gov. Executive High court High court Engaged society EMB autonomy EMB autonomy CSO repression Engaged society censorship oversight independence independence Media Legislature Gov. censorship Free and fair High court Gov. censorship Engaged EMB autonomy EMB autonomy EMB autonomy investigates in Media elections independence Media society practice Legislature CSO participatory Freedom of CSO participatory CSO entry and Media critical investigates in Media critical Media critical Media critical environment discussion environment exit practice Print or Print or Print or broadcast Freedom of broadcast CSO Media critical broadcast media CSO entry and exit Media critical CSO repression media discussion media repression perspectives perspectives perspectives 40 Print or CSO Print or Freedom of broadcast Lower court Freedom of Elections vote- CSO entry and exit participatory Media critical broadcast media discussion media independence discussion buying environment perspectives perspectives Print or broadcast Print or Legislature CSO Elections vote- CSO CSO entry and media Media critical Media critical broadcast media controls repression buying repression exit perspectives perspectives resources CSO CSO CSO Legislature Opposition Lower court Freedom of Free and fair Party linkages participatory participatory participatory controls parties autonomy independence discussion elections environment environment environment resources Print or broadcast Opposition Opposition Opposition CSO Freedom of Barriers to media parties CSO repression Party linkages parties parties participatory discussion parties de-facto perspectives autonomy autonomy autonomy environment Barriers to CSO Barriers to parties Barriers to Barriers to Party Lower court CSO repression CSO repression parties de- participatory de-facto parties de-facto parties de-facto linkages independence facto environment Opposition CSO Multiparty Free and fair Opposition parties Free and fair Free and fair parties Party linkages Party linkages participatory elections de- elections autonomy elections elections autonomy environment facto Print or Legislature Opposition Opposition Legislature Multiparty Free and fair Lower court broadcast Party linkages controls parties parties controls elections de-facto elections independence media resources autonomy autonomy resources perspectives Legislature Multiparty Legislature Multiparty Barriers to parties Multiparty Free and fair High court Free and fair controls elections de- controls elections de- de-facto elections de-facto elections independence elections resources facto resources facto Multiparty Barriers to Elections vote- Judicial Lower court Barriers to Judicial Freedom of elections de- Party linkages parties de- buying accountability independence parties de-facto accountability discussion facto facto Legislature Legislature Multiparty Opposition Elections vote- Legislature Free and fair Elections vote- Elections controls controls elections de- parties buying controls resources elections buying vote-buying resources resources facto autonomy Multiparty Multiparty Barriers to Judicial Elections vote- Elections Elections vote- Judicial Party linkages elections de- elections de- parties de- accountability buying vote-buying buying accountability facto facto facto Note: Indicators sorted in descending order: highest number of dependencies on top. Vertical accountability indicators are in red, horizontal in blue, diagonal in green. Indicators that are stricken out have not reached the highest level in a single country in the respective region by 2012. 41 There are also some interesting differences across regions with regards to horizontal accountability. Notably, no country from Eastern Europe and Central Asia, Sub-Saharan Africa and South-East Asia has reached full judicial accountability – a measure of whether judges are held accountable for possible illegal actions - before making substantial progress in many other aspects of accountability. This is one instance where the disaggregated, regional analysis is very useful. Because of the fact that in a minority of regions (e.g. Western Europe) judicial accountability developed to a high degree early, the global analysis “hides” that in most of the regions it is actually a mechanism of accountability that comes very late in the sequence. Similarly, lower court independence was developed relatively late in the sequence in regions in the world covering a substantial number of countries (Eastern Europe and Central Asia, Latin America, East and South Asia), but in other regions it had a relatively low number of contingent conditions. While the present analysis cannot provide an answer to why these regional differences occur, it will be important to note these exceptions to the global pattern if and when the analyses here are used to make inferences about programs and interventions. Time Trends Since the fall of the Iron Curtain, the number of electoral authoritarian regimes, i.e. regimes which allow de-jure multiparty elections to take place but limit their freedom and fairness, has surged. During the first and second wave of democratization this practice was not as widespread. It seems plausible that this trend is reflected in different sequencing patterns. Therefore, we split the sample into two parts: One including all countries in 1988 or earlier and one with all countries after 1988. Based on the methods described above, Table 7 lists the de-facto accountability indicators sorted in descending order with the indicators with the highest number of dependencies at the top of the list, and the lowest at the bottom. Most key findings from the general patterns are similar to the results described for the global sample, as well as for the regional analyses. In particular, it is noteworthy that the three mechanisms of horizontal accountability that directly oversee and constrain the degree of freedom of ruling elites are on top of the contingency table for both before and after the fall of the Iron Curtain. These mechanisms of horizontal accountability require most other mechanisms to be relatively highly developed de facto, in order for them to be realized in full. The result is that the reluctance of ruling elites to give-in on these issues is time-invariant and strong enough to make it possible for incumbents to “hang on” to limiting these key mechanisms until the very end. This reinforces the conclusion that interventions seeking to strengthen accountability on these three mechanisms should take into account the developments in the other relevant areas. Table 7. Sequencing of accountability mechanisms by time 1988 or earlier After 1988 Contingency Contingency Indicator name Indicator name conditions conditions Legislature investigates in practice 63 Lower court independence 62 Executive oversight 61 Executive oversight 59 Engaged society 58 High court independence 57 High court independence 52 Legislature investigates in practice 55 Gov. censorship Media 50 Party linkages 52 EMB autonomy 48 Gov. censorship Media 51 CSO entry and exit 47 Engaged society 50 Freedom of discussion 37 EMB autonomy 49 Print or broadcast media critical 34 Election free and fair 44 Print or broadcast media 32 Print or broadcast media critical 43 perspectives CSO repression 30 CSO entry and exit 42 CSO participatory environment 30 Freedom of discussion 42 Opposition parties autonomy 23 CSO repression 41 De facto barriers to parties 20 De facto barriers to parties 37 Print or broadcast media Lower court independence 19 36 perspectives Election free and fair 14 CSO participatory environment 35 Judicial accountability 13 Legislature controls resources 35 Legislature controls resources 11 Opposition parties autonomy 35 De facto elections multi-party 11 De facto elections multi-party 26 Party linkages 6 Judicial accountability 22 Election vote buying 4 Election vote buying 8 Second, however, there are instructive differences between the two samples in terms of some vertical accountability mechanisms. Before 1988, two important indicators of de- facto vertical accountability - free and fair elections and programmatic party linkages – are at a similar spot in the sequence as multi-party elections. Such aspects often developed hand-in-hand. However, after the fall of the Iron Curtain the development of electoral freedom and fairness and non-clientelistic party linkages seems to require considerable 43 more progress in other aspects of accountability than the mere holding of multi-party elections. This could be linked to the emergence of a larger number of electoral autocracies in the latter period, which only improve the quality of elections – if at all – after internal as well external pressure (Lindberg 2006, Schedler 2013). Third, in the period after 1988 lower court independence developed last in the sequence, whereas for the earlier time period it can be found in the middle of the contingency table. This finding suggests that countries that developed accountability after the fall of the Iron Curtain had to struggle with a legacy of weak low courts. Conclusions Accountable institutions are a prerequisite for good governance. Vertical accountability ties ruling elites directly to the will of the people. Institutions in the realm of horizontal accountability – parliaments, courts and other bodies – constrain and oversee the day-to- day actions of ruling elites. In addition, for effective government oversight a vibrant civil society and critical media need to create diagonal accountability, which has only recently become a focal point for study. Furthermore, a major implementation gap often exists between the introduction of de-jure institutions and their de-facto effectiveness. Our research suggests that it took countries on average 11 years to close the implementation gap in the realm of vertical accountability, and even longer – 31 years – in the realm of horizontal accountability. For many countries the implementation gap still persists. Therefore, it is crucial to better understand how governments become more accountable de-facto. This paper sheds light on this issue in several ways. Using novel sequencing methods, we present new evidence on how accountability has evolved in 173 countries from 1900 until the present. Our findings uncover the following empirical trends. In general, high levels of any type of de-facto accountability can only be achieved if some progress in other areas has been made. Hence, different accountability mechanisms reinforce each other, because progress in one area creates space and incentives to demand improvements in other areas. For international and national actors who want to support the development of accountability, this finding implies that they should use any available space for enhancing the capacities of civil society groups, parliaments, citizens, 44 media and political parties in order to enable them to increase their pressure for more accountable governance. In particular, our findings show that high levels of de-facto accountability in the realm of vertical accountability can evolve before other types of accountability. Ruling elites seem to be more likely to make initial concessions in this area of accountability, because it is less effective in directly constraining their actions. At the same time, we find that effective horizontal accountability is contingent on progress in vertical and diagonal accountability. As a matter of fact, without fully free and fair elections, autonomous opposition parties and a developed civil society and media, virtually no country in the world has ever achieved effective government oversight through independent high courts, vigorous parliaments or other institutions. This suggests that international efforts to improve freedom and fairness of elections and the situation of opposition parties can have positive repercussions for other areas of accountability as well. The judiciary and legislature have stronger incentives to oversee the actions of the executive if members of parliament are held accountable through free elections and functioning political parties, and if, at the same time, the media scrutinizes the work of judges and legislators, and civil society organizations push for implementing the checks and balances between institutions. In sum, the novel sequencing methods utilized in this paper make an important contribution to our understanding of how high-levels of accountability evolve. Finally, our next step is to look more closely at the development of a few typical cases. For future research, it would be relevant to shed more light on the ambiguous middle ground of the accountability spectrum, where one step forward is often followed by two steps backwards. 45 References Abbott A. Sequence Analysis: New Methods for Old Ideas. Annual Review of Sociology. 1995; (21):93-113. Acemoglu DR, JA. Economic backwardness in political perspective. American Political Science Review. 2006:115-131 Ackerman J. 2003. Co-Governance for Accountability: Beyond ‘Exit’ and ‘Voice’. World Development 323:447-463. Anderson L. The Authoritarian Executive? Horizontal and Vertical Accountability in Nicaragua’, Latin American Politics and Society. 2002; (482):141-169. Barton AD. Public Sector Accountability and Commerical-in-Confidence Outsourcing Contracts. Accounting, Auditing & Accountability Journal. 2006; 192:256-271 Berman S. Civil Society and the Collapse of the Weimar Republic. World Politics. 1997; 493:401- 429 Berman S. How Democracies Emerge: Lessons from Europe. Journal of Democracy. 2007; 18:28- 41. Bernhard M. Civil Society and Democratic Transition in East Central Europe, Political Science Quarterly. 1993; 108:307-326. Bernhard M, Bizzarro F, Coppedge M, Gerring J, Hicken A, Knutsen CH, Lindberg SI, Skaaning S. Party Strength and Economic Growth. Varieties of Democracy Institute: Working Paper No. 10. 2015 Besley T. Principled Agents? The Political Economy of Good Government. Oxford University Press; 2006. Besley T, Burgess T, Prat A. Mass media and political accountability. World Bank; 2002. Birch S. Electoral institutions and popular confidence in electoral processes: A cross-national analysis. Electoral Studies. 2008; 27:305–320. Boahen A. A note on the Ghanaian elections, African Affairs. 1995; 94:277-80. Bollen K. and Jackman R. ‘‘Democracy, Stability, and Dichotomies.’’ American Sociological Review. 1989; 54:612-621. Boix C. Democracy and redistribution. Cambridge University Press, 2003. Bratton M. Vote buying and violence in Nigerian election campaigns. Electoral Studies. 2008;27:621-632. Carothers T. The Sequencing Fallacy. Journal of Democracy. 2007;18:13-27. Carothers T. The End of the Transition Paradigm. Journal Of Democracy. 2002;13(1)5-21. Chang E, Golden M, Hill S. Legislative malfeasance and political accountability. World Politics. 2010;622:177–220. Collier P, Pedro CV. Votes and Violence: Evidence from a Field Experiment in Nigeria. The Economic Journal. 2013;124:327-355. Costas-Perez E, Sole-Olle A, Sorribas-Navarro P. Corruption scandals, voter information, and accountability. European Journal of Political Economy. 2012;284:469-484. Considine M. The End of the Line?: Accountable Governance in the Age of Networks, Partnerships, and Joined-up Services. Governance. 2002;151:21-40. Coppedge M, Reinicke W. Measuring polyarchy. Studies in Comparative International Development. 1990;25(1):51-72. Coppedge M, Gerring J, Lindberg SI, Skaaning S, Teorell J, Altman D, Andersson F, Bernhard M, Fish SM, Glynn A, Hicken A, Knutsen CH, Marquardt KL, McMann K, Mechkova V, Miri F, Paxton P, Pernes J, Pemstein D, Staton J, Stepanova N, Tzelgov E, Wang Y, Zimmerman B. V- 46 Dem [Country-Year/Country-Date] Dataset v6. Varieties of Democracy V-Dem Project. 2016a. Coppedge M, Gerring J, Lindberg SI, Skaaning S, Teorell J, Altman D, Andersson F, Bernhard M, Fish SM, Glynn A, Hicken A, Knutsen CH, Marquardt KL, McMann K, Mechkova V, Miri F, Paxton P, Pernes J, Pemstein D, Staton J, Stepanova N, Tzelgov E, Wang Y, Zimmerman B. V- Dem Codebook v6. Varieties of Democracy V-Dem Project. 2016b. Cranenburgh van O. Restraining executive power in Africa: horizontal accountability in Africa’s hybrid regimes. South African Journal of International Affairs. 2009;16(1)49-68. Dahl RA. Polyarchy: Participation and Opposition. New Haven: Yale University Press; 1971. Dahl RA. Democracy and Its Critics. New Haven: Yale University Press; 1989. Diamond L. Democracy in Latin America: Degrees, Illusions, and Directions for Consolidation. In Tom Farer, ed., Beyond Sovereignty: Collectively Defending Democracy in the Americas. Baltimore: Johns Hopkins University Press.1996;52–104. Dimock M, Jacobson G. Checks and choices: The House bank scandal’s impact on voters in 1992. Journal of Politics. 1995;574:1143–59. Elkins Z, Ginsburg T, Melton J. Characteristics of National Constitutions. Comparative Constitutions Project. Version 2.0. April 18, 2014. Finn P. Public Trust and Public Accountability. Australian Quarterly. 1993;65: 50-59. Fish MS. Democracy Derailed in Russia: The Failure of Open Politics. Cambridge: Cambridge University Press. 2005b. Fish MS. Stronger Legislatures, Stronger Democracies. Journal of Democracy, 2005a:17:5-20. Fisher J. Importing Democracy. The role of NGOs in South Africa, Tajikistan & Argentina. Kettering Foundation Press; 2013. Fisse B, Braithwaite J. Corporations, Crime, and Accountability. Cambridge: Cambridge University Press; 1993. Foweraker J, Krznaric R. The Uneven Performance of Third Wave Democracies: Electoral Politics and the Imperfect Rule of Law in Latin America’, Latin American Politics and Society. 2002;443:29-60. Fox JA. Social Accountability: What Does the Evidence Really Say? World Development. 2015;72:346–361. Fukuyama F. Is There a Proper Sequence in Democratic Transitions? Current History. 2011 110739: 308-310. Gandhi J. Political institutions under dictatorship. Cambridge: Cambridge University Press; 2008. Gandhi J, Lust-Okar E. Elections Under Authoritarianism. Elections Under Authoritarianism. Annual Review of Political Science. 2009:12(1)403–422. Gehlbach S, Keefer P. Investment without Democracy: Ruling-party Institutionalization and Credible Commitment in Autocracies. Journal of Comparative Economics. 2011;39:123–39. Gerring J, Skaaning S, Pemstein D. A concept-driven approach to measurement: The Lexical Scale. Unpublished manuscript, Boston University, MA. 2014. Global Commission on Elections, Democracy and Security Deepening Democracy. A Strategy for Improving the Integrity of Elections Worldwide. International IDEA; Kofi Annan Foundation; 2012. Goetz AM, Jenkins R. Reinventing Accountability: Making Democracy Work for Human Development. London: Palgrave-Macmillan; 2005. Green D. Ghana: structural adjustment and state reformation in L. Villalón & P.A. Huxtable, eds. The African State at a Critical Juncture. Boulder, CO: Lynne Rienner. 1998;185-211. 47 Grimes M. The Contingencies of Societal Accountability: Examining the Link Between Civil Society and Good Government. Studies in comparative international development. 2013;484:380-402. Gyimah-Boadi E. Ghana: the challenges of consolidating democracy, in R. Joseph, ed., State, Conflict and Democracy in Africa. Boulder CO: Lynne Rienner, 1999;409-27. Gyimah-Boadi E. A peaceful turnover in Ghana, Journal of Democracy. 2001;12(1):103-4. Hafner-Burton EM, Hyde SD, Jablonski RS. When do governments resort to election violence?. British Journal of Political Science 44.01. 2014:149-179. Herron ES, Nazar B. Horizontal accountability during political transition: The use of deputy requests in Ukraine, 2002–2006. Party Politics. 2015;211:131–142. Hicken A, Kuhonta EM. Party system institutionalization in Asia: Democracies, autocracies, and the shadows of the past. Cambridge University Press; 2014. Hobolt SB, Klemmensen R. Government Responsiveness and Political Competition in Comparative Perspective. Comparative Political Studies. 2008;41(3)309-337. Howard MM, Roessler PG. Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes. American Journal of Political Science, 2006;50:365–381. Hunhold C. Corporatism, Pluralism and Democracy: Toward a Deliberative Theory of Bureaucratic Accountability, Governance. 2001;142(4): 151-167. Ichino N, Schündeln M. Deterring or Displacing Electoral Irregularities? Spillover Effects of Observers in a Randomized Field Experiment in Ghana. Journal of Politics. 2012;74:292– 307. Jankowski R. Preference Aggregation in Political Parties and Interest Groups: A Synthesis of Corporatist and Encompassing Organization Theory. American Journal of Political Science.1988;321:105–125. Jeffries R. The Ghanaian elections of 1996: towards the consolidation of democracy?, African Affairs. 1998;97:189-208. Jones GW. The search for local accountability. in S. Leach editor, Strengthening local government in the 1990s. London: Longman. 1992: 49-78 Johnston M. Civil society and corruption: mobilizing for reform. University Press of America; 2005. Katznelson I; Weingast BR. Preferences and situations. Points of intersection between historical and rational choice institutionalism. New York: Russell Sage Foundation; 2005. Keefer P. Collective Action, Political Parties and the Political Economy of Development. Washington DC: The World Bank, Development Research Group, Macroeconomics and Growth Team; 2013. Kelly JM. Counting on the Past or Investing in the Future? Journal of Politics. 2003; 653(8)864- 880. Key VO, Jr. Southern Politics in State and Nation. Cambridge: Harvard University Press;1949. Klijn E, H. and J. Koppenjan. Eds. Managing Complex Networks: Strategies for the Public Sector. London: Sage; 1997. Kitschelt H. Linkages between Citizens and Politicians in Democratic Polities. Comparative Political Studies. 2000;336(7):845-879. Kitschelt H, Wilkinson S.eds. Patrons or Policies? Patterns of Democratic Accountability and Political Competition. Cambridge: Cambridge University Press; 2006. Laver M, Shepsle KA. Accountability in Parliamentary Democracy. Democracy, accountability, and representation 1999: 279. Lindberg SI. It’s our time to ‘chop’: do elections in Africa feed neopatrimonialism rather than counter-act it?, Democratization. 2003;10(2):121-40. 48 Lindberg SI. Democracy and Elections in Africa. Baltimore: Johns Hopkins University Press; 2006. Lindberg SI. Ed. Democratization by Elections - A New Mode of Transition. Baltimore: John Hopkins University Press; 2009. Lindberg SI, Zhou Y. Cooptation Despite Democratization in Ghana. Chapter 5 in Barkan, Joel ed. Legislative Power in Emerging African Democracies. Boulder, C.O.: Lynne Rienner, 2009;147-176. Lindberg SI. Mapping accountability: core concept and subtypes. International Review of Administrative Sciences. 2013;792:202-206. Lindberg SI. Ordinal Versions of V-Dem’s Indices: For Classification, Description, Sequencing Analysis and Other Purposes. Geopolitics, History and International Relations. 2016;82: 76–111. Lindenfors P, Jansson F, Wang Y, Lindberg SI. Investigating Sequences in Ordinal Data: A New Approach with Adapted Evolutionary Models Varieties of Democracy Institute: Working Paper No. 18. 2015. Lipset SM, Trow MA, Coleman JS. Union Democracy: The Internal Politics of the International Typographical Union. Glencoe, Ill.: Free Press;1956. Lyons T. A major step forward, Journal of Democracy. 1997;8(2):65-77. Magaloni B. Credible power-sharing and the longevity of authoritarian rule. Comparative Political Studies; 2008. Mahoney J, Dietrich R. eds. Comparative Historical Analysis in the Social Sciences Cambridge University Press; 2003. Mahoney J. The Legacies of Liberalism: Path Dependence and Political Regimes in Central America. Baltimore, MD: The Johns Hopkins University Press; 2001. Mainwaring S, Welna C. Democratic Accountability in Latin America. Oxford University Press; 2003. Malena C, Forster R, Singh J. Social Accountability. An Introduction to the Concept and Emerging Practice. World Bank Social Development Papers, Participation and Civic Engagement Paper. 2004:No. 76. Mansfield ED, Snyder J. The Sequencing Fallacy. Journal of Democracy. 2007;18:6-9. Maskin E, Tirole J. The Politician and the Judge: Accountability in Government. The American Economic Review. 2004;944(9):1034-1054. McCubbins MD, Schwartz T. Congressional Oversight Overlooked: Police Patrols versus Fire Alarms. American Journal of Political Science 28.1. 1984:165-79. Morrison MKC. The modulated return of Ghana to civilian rule. Africa Contemporary Record. New York: Holmes & Meier; 1999. Mulgan R. Accountability: An Ever-Expanding Concept? Public Administration. 2000;783:555- 573. Mulgan R. Holding Power to Account: Accountability in Modern Democracies. Basingstoke: Palgrave; 2003. North DC, Wallis JJ, Weingast BR. Violence and social orders: a conceptual framework for interpreting recorded human history. Cambridge University Press; 2009. Ninsin K. A. Elections, democracy and elite consensus, in K.A. Ninsin, ed. Ghana: transition to democracy. Dakar: CODESRIA. 1998;211-30. Nugent P. ‘Winners, losers and also rans: money, moral authority and voting patterns in the Ghana 2000 election’, African Affairs. 2001;100:405–28. O'Donnell GA.; Schmitter, Philippe C.; Whitehead, Laurence Eds. Transitions from Authoritarian Rule: Comparative Perspectives. Baltimore and London: Johns Hopkins Univ. Press; 1986. 49 O’Donnell G. Horizontal Accountability in New Democracies. Journal of Democracy. 1998;93:112-126. O’Loughlin MG. Bureaucratic Accountability and How to Measure It?. Administration and Society. 1990;223(11):275-302. Oliver D. Government in the United Kingdom. Milton Keynes: Open University Press; 1991. Olson M. Power and Prosperity. New York: Basic Books; 2000. Olukoyun A. Media Accountability and Democracy in Nigeria, 1999-2003. African Studies Review. 2004;47(3): 69-90. Oquaye M. The Ghanaian Elections of 1992: A Dissenting View, African Affairs. 1995; 94: 259-75. Painter-Morland M. Defining Accountability in a Network Society. Business Ethics Quarterly. 2007;173:515-534. Pemstein D, Marquardt KL, Tzelgov E, Wang Y, Miri F. The V-Dem Measurement Model: Latent Variable Analysis for Cross-National and Cross-Temporal Expert-Coded Data. Varieties of Democracy Institute: Working Paper No. 21. 2015. Peruzzotti E, Smulowitz C. eds. Enforcing the Rule of Law: Social accountability in the new Latin American democracies. Pittsburg: University of Pittsburg Press; 2006. Pelizzo R, Stapenhurst F. Parliamentary oversight tools: A comparative analysis. Vol. 45. Routledge; 2013. Philips R, Berman SL. 2007. Introduction to the Special Section on Accountability, Business Ethics Quarterly 173: 449-452. Putnam RD, Leonardi R, Nanetti RY. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press; 1993. Pierskalla J, Fernandez M. Partisan Strategies and Political Economic Performance. Do Modes of Democratic Accountability Affect Economic Fortunes? Department of Political Science, Duke University; 2011. Przeworski A. Some Problems in the Study of the Transition to Democracy. Transition to Democracy. In: O'Donnell, Guillermo A.; Schmitter, Philippe C.; Whitehead, Laurence Eds. 1986: 47–63. Radin BA, Romzek BS. Accountability Expectations in an Intergovernmental Arena: The National Rural Development Partnership. Journal of Federalism. 1996;26(2):59-81. Rakner L, van de Walle N. Opposition parties and incumbent presidents: the new dynamics of electoral competition in sub-Saharan Africa’ in Staffan Lindberg ed., Democratization by Elections: A new mode of transition? Johns Hopkins University Press, Baltimore, MD; 2009. Rakner L, Gloppen S. Tax reform and democratic accountability in Sub-Saharan Africa, in Nicolas van de Walle et al. eds.: Beyond Structural Adjustment: The Institutional Context of African Development. London and New York: Pagrave MacMillan; 2003. Ranney A. The Doctrine of Responsible Party Government: Its Origins and Present State. Urbana: Univ. Ill. Press; 1962. Rose-Ackerman S. Democracy and ‘grand’ corruption. International Social Science Journal, 1996;48149:365-380. Rosneau JN. Governance, Order, and Change in World Politics, in Rosenau, James N. and E-O Czempiel, eds., Governance without Government: Order and Change in World Politics Cambridge: Cambridge University Press; 1992. Salih MMA. ed. African Parliaments: Between governance and government. Palgrave Macmillan, New York, NY; 2005. Sandbrook R, Oelbaum J. Reforming the Political Kingdom: governance and development in Ghana’s fourth republic. Accra: Centre for Democracy and Development; 1999. Schattschneider EE. Party government. New York, NY: Rinehart; 1942. 50 Schedler A. Conceptualizing accountability. In: Schedler, L. Diamond and M. F. Plattner eds. The Self-restraining State: Power and Accountability in New Democracies. Boulder, CO: Lynne Rienner. 1999;13-28. Schedler A. Judgment and Measurement in Political Science. Perspectives on Politics. 2012;101:21-36. Schedler A. Elections Without Democracy. The Menu of Manipulation. Elections Without Democracy. Journal Of Democracy. 2002;13(2)36–50. Schedler A. The Politics of Uncertainty. Sustaining and Subverting Electoral Authoritarianism. Politics of Uncertainty. 1st ed, Oxford, New York, NY: Oxford University Press; 2013. Schmitter P. The Ambiguous Virtues of Accountability. Journal of Democracy. 2004;154:47-59 Schmitter P, Karl TL. What democracy is ... and is not. Journal of Democracy. 1991;31:75-88. Schumpeter JA. Capitalism, socialism and democracy. New York, NY:Harper & Row; 1950. Shenkin M, Coulson AB. Accountability Through Activism: Learning from Bourdieu, Accounting, Auditing & Accountability Journal. 2007;202:297-317. Sillen-Tullberg B. The effect of biased inclusion of taxa on the correlation between discrete characters in phylogenetic trees. Evolution. 1993;1182-1191. Skaaning S, Gerring J, Bartusevičius H. A Lexical Index of Electoral Democracy. Comparative Political Studies. 2015;1–35. Simmons J, Hicken A, Kollman K, Noorudin I. Dividing the Spoils: Party Nationalization, Credibility, and Foreign Direct Investment. Unpublished Manuscript; 2014. Snyder R. Beyond Electoral Authoritarianism. The Spectrum of Nondemocratic Regimes. Beyond Electoral Authoritarianism. In: Schedler, Andreas Ed. 2006: Electoral Authoritarianism. The Dynamics of Unfree Competition. Electoral Authoritarianism, Boulder: Lynne Rienner Publishers. 2006; 219–233. Stapenhurst R, O’Brien M. Accountability in governance. World Bank Paper 4. http://siteresources.worldbank.org/PUBLICSECTORANDGOVERNANCE/Resources/Accou ntabilityGovernance.pdf. Accessed 01 July 2016. Stepan A, Skach C. Constitutional Frameworks and Democratic Consolidation: Parliamentarianism versus Presidentialism. World Politics. 1993;46(1)1-22. Stirton L, Lodge M. Transparency Mechanisms: Building Publicness into Public Services. Journal of Law and Society. 2001;284: 471-489. Stokes SC, Dunning TT, Nazareno M, Brusco V. Brokers, Voters, and Clientelism: The Puzzle of Distributive Politics. Cambridge: Cambridge University Press; 2013. Stone B. Administrative Accountability in the ‘Westminister’ democracies: Towards a New Conceptual Framework. Governance. 1995;8:505-26. Svolik M. The Politics of Authoritarian Rule. Cambridge: Cambridge University Press; 2012. Thomas PG. The changing nature of accountability, in Peters, B. G. and D. Savoie eds. Taking Stock: Assessing Public Sector Reforms. Montreal: McGill-Queen’s University Press; 1998. Tsai LL. Accountability Without Democracy: Solidarity Groups and Public Goods Provision in Rural China. Cambridge: Cambridge University Press; 2007. United Nations Resolution A/Res/70/1. Adopted by the General Assembly on 25 September 2015 Transforming our world: the 2030 Agenda for Sustainable Development. Van Cranenburgh O. Restraining executive power in Africa: horizontal accountability in Africa's hybrid regimes. South African Journal of International Affairs. 2009;16.1:49-68. Van Vliet M. Weak Legislatures, Failing MPs, and the Collapse of Democracy in Mali. African Affairs. 2014;113/450, 45–66. Voltmer K. The media, government accountability and citizen engagement. In book: Public Sentinel: New Media and the Governance Agenda, Chapter: The media, government 51 accountability and citizen engagement., Publisher: The World Bank, Editors: Pippa Norris, 137-159. Walker P. Understanding Accountability: Theoretical models and their Implications for Social Service Organizations, Social Policy and Administration. 2002;361:62-75. Wang Y, Lindenfors P, Sundström A, Jansson F, Lindberg SI. No Democratic Transition Without Women’s Rights: A Global Sequence Analysis 1900-2012 Varieties of Democracy Institute: Working Paper No. 12. 2015b. Wang Y, Mechkova V, Andersson F. Does Democracy or Good Governance Enhance Health? New Empirical Evidence 1900-2012. Varieties of Democracy Institute: Working Paper No. 11 2015a. Weber EP. The Question of Accountability in historical Perspective: From Jackson to Contemporary Grassroots Ecosystem Management. Administration and Society. 1999;314:451-494. Wilson MC. Castling The King: Institutional Sequencing And Regime Change. Unpublished PhD thesis. The Pennsylvania State University; 2015. Woods N, Narlikar A. Governance and the Limits of Accountability: the WTO, the IMF, and the World Bank. ISSJ 170/2001. UNESCO. World Bank. World Development Report 2004: Making Services Work for Human Poor People. Washington DC; 2004. World Bank Institute.. Social Accountability in the Public Sector. Washington DC: WBI Working Paper No.33641; 2005. Wrede M. Uniformity Requirement and Political Accountability. Journal of Economics. 2006;89(2):95-113. Yashar DJ. Demanding Democracy: Reform and Reaction in Costa Rica and Guatemala, 1870s- 1950s. Stanford, CA: Stanford University Press; 1997. 52 Appendix 53 Note: Year= de-jure/ de-facto gap closed for the first time in the most recent spell; Spell= non-interrupted periods with de facto vertical accountability in place; * = no de-facto vertical accountability in 2012 54 Figure A.2. Implementation gap between de-jure and de- facto horizontal accountability (years, last spell) Switzerland 1900 Norway 1900 Netherlands 1946 Lithuania 1990 Germany 1949 Denmark 1945 Uruguay 1985 Haiti 1991 * Estonia 1992 Sweden 1911 Slovenia 2005 Madagascar 1992 * Ukraine 2007 * Moldova 2010 Liberia 2004 * Median France 1969 Average Venezuela 1990 * Austria 1987 Finland 1946 Japan 1947 United States 1948 Taiwan 2000 * Bulgaria 1990 * Mauritius 2000 Belgium 1996 * Albania 1998 Costa Rica 1994 United Kingdom 1981 Tanzania 2010 * New Zealand 1997 0 20 40 60 80 100 120 Note: Year= de-jure/ de-facto gap closed for the first time in the most recent spell; Spell= non-interrupted periods with a sitting Parliament; *= no de-facto horizontal accountability in 2012. 55 Table A.1 Variable names and question text. Variable name and tag Aggregation/Note Clarification/full question text Source Vertical v2x_acver accountability De jure vertical v2x_acverju Lexical index accountability At this time, are regularly scheduled national Electoral regime (v2x_elecreg) elections on course, as stipulated by election law V-Dem or well-established precedent? First two categories Is it legally possible for multiple parties to run in Elections multiparty (v2elmulpar) V-Dem of v2elmulpar elections? First two categories Barriers to parties (v2psbars) Are there legal barriers to forming a party? V-Dem of v2psbars What percentage (%) of adult citizens (as defined Percentage of population with 98% is 1, less is 0 by statute) has the legal right to vote in national V-Dem suffrage (v2elsuffrage) elections? De facto vertical v2x_acverfa BFA accountability Elections multiparty (v2elmulpar) Categories 2, 3, 4 Are multiparty elections being held in practice? V-Dem Barriers include legal requirements such as Barriers to parties (v2psbars) Categories 2, 3, 4 requirements for membership or financial V-Dem deposits, as well as harassment. Taking all aspects of the pre-election period, election day, and the post-election process into EMB autonomy (v2elembaut) V-Dem account, would you consider this national election to be free and fair? A party-constituent linkage refers to the sort of “good” that the party offers in exchange for Party linkages (v2psprlnks) V-Dem political support and participation in party activities. An opposition party is any party that is not part of Opposition parties autonomy the government, i.e., that has no control over the V-Dem (v2psoppaut) executive. Taking all aspects of the pre-election period, election day, and the post-election process into Free and fair elections (v2elfrfair) V-Dem account, would you consider this national election to be free and fair? Vote and turnout buying refers to the distribution of money or gifts to individuals, families, or small groups in order to influence their decision to Election vote buying (v2elvotbuy) V-Dem vote/not vote or whom to vote for. It does not include legislation targeted at specific constituencies, i.e., “porkbarrel” legislation. Horizontal v2x_achor accountability De jure horizontal v2x_achorju mean accountability Is there a legislature in place? Advisory bodies that do not have the formal authority to legislate– Legislature bicameral (v2lgbicam) as stipulated by statute, legislative rules, the V-Dem constitution, or common law precedent–are not considered legislatures. 57 Does the legislature have the power to interpellate members of the executive branch, or similarly, is CCP (INTEXEC) CCP the executive responsible for reporting its activities to the legislature on a regular basis? Does the constitution contain an explicit CCP (JUDIND) declaration regarding the independence of the CCP central judicial organ(s)? Does the constitution provide for an ombudsman, CCP (ATGEN, OMBUDS) CCP attorney general or public prosecutor? De facto horizontal v2x_achorfa BFA accountability If the executive were engaged in unconstitutional, illegal, or unethical activity, how likely is it that a Legislature investigates in legislative body would conduct an investigation V-Dem practice (v2lginvstp) that would result in a decision or report that is unfavorable to the executive? In practice, does the legislature control the Legislature controls resources resources that finance its own internal operations V-Dem (v2lgfunds) and the perquisites of its members? When the high/lower court in the judicial system is ruling in cases that are salient to the High court/lower court government, how often would you say that it independence (v2juhcind, V-Dem makes decisions that merely reflect government v2juncind) wishes regardless of its sincere view of the legal record? When judges are found responsible for serious Judicial accountability (v2juaccnt) misconduct, how often are they removed from V-Dem their posts or otherwise disciplined? If executive branch officials were engaged in unconstitutional, illegal, or unethical activity, how likely is it that a body other than the legislature, Executive oversight (v2lgotovst) V-Dem such as a comptroller general, general prosecutor, or ombudsman, would question or investigate them and issue an unfavorable decision or report? 58 Social v2x_acsoc accountability De jure social v2x_acsocju mean accountability Does the constitution provide for freedom of CCP (ASSEM) CCP assembly? Does the constitution provide for freedom of CCP ([EXPRESS]) CCP expression or speech? Does the constitution provide for freedom of the CCP (PRESS) CCP press? De facto social v2x_acsocfa BFA accountability Indirect forms of censorship might include politically motivated awarding of broadcast frequencies, withdrawal of financial support, Government censorship effort - influence over printing facilities and distribution V-Dem Media (v2mecenefm) networks, selected distribution of advertising, onerous registration requirements, prohibitive tariffs, and bribery. Print/broadcast media critical Of the major print and broadcast outlets, how V-Dem (v2mecrit) many routinely criticize the government? Print/broadcast media Do the major print and broadcast media represent V-Dem perspectives (v2merange) a wide range of political perspectives? Does the government attempt to repress civil CSO repression (v2csreprss) V-Dem society organizations (CSOs)? Is participation in civil society organizations CSO participatory environment (CSOs) voluntary and is there wide popular V-Dem (v2csprtcpt) involvement? When important policy changes are being Engaged society (v2dlengage) considered, how wide and how independent are V-Dem public deliberations? CSO entry and exit (v2cseeorgs) V-Dem 59 This indicator specifies the extent to which citizens are able to engage in private discussions, particularly on political issues, in private homes and public spaces (restaurants, public transportation, sports events, work etc.) without Freedom of discussion fear of harassment by other members of the polity V-Dem (v2xcl_disc) or the public authorities. We are interested in restrictions by the government and its agents but also cultural restrictions or customary laws that are enforced by other members of the polity, sometimes in informal ways. 60 Table A.2 Sequence analysis of accountability mechanisms by region, detailed. Eastern Contingency Contingency Contingency Contingency Contingency Sub-Saharan Europe and conditions Latin America conditions MENA conditions conditions Western Europé conditions Africa Central Asia (max 127) (max 127) (max 127) (max 127) (max 127) Highest Judicial Executive Gov. censorship Judicial Executive 75 75 score not 66 70 accountability oversight Media accountability oversight reached Legislature Legislature Highest Executive 70 investigates in 66 investigates in score not EMB autonomy 50 Ombudsman 70 oversight practice practice reached Gov. censorship Lower court Executive 66 60 64 Engaged society 48 Engaged society 66 Media independence oversight Legislature Legislature Lower court High court 62 54 Engaged society 57 investigates in 42 investigates in 63 independence independence practice practice Legislature Freedom of Gov. censorship Gov. censorship investigates in 61 Engaged society 52 46 40 61 expression Media Media practice High court CSO entry and High court Freedom of Judicial 61 49 46 39 61 independence exit independence discussion accountability Executive CSO entry and Engaged society 59 Ombudsman 49 EMB autonomy 46 38 58 oversight exit Free and fair High court High court EMB autonomy 56 EMB autonomy 48 45 36 58 elections independence independence CSO Gov. censorship CSO participatory CSO entry and participatory 53 40 43 33 EMB autonomy 58 Media environment exit environment Freedom of CSO entry and Media critical 50 40 42 Media critical 29 Media critical 57 discussion exit Print or Print or broadcast Print or broadcast CSO entry and 47 broadcast media 36 Ombudsman 42 media 27 media 57 exit perspectives perspectives perspectives Civil society re Print or CSO participatory Freedom of Lower court 41 34 37 Freedom of 26 57 broadcast environment discussion independence expression 61 media perspectives Opposition Elections vote- parties 41 Media critical 29 Media critical 36 Ombudsman 21 57 buying autonomy Freedom of Lower court CSO participatory Judicial 41 Party linkages 27 35 17 56 discussion independence environment independence Print or broadcast Barriers to Opposition Freedom of Ombudsman 36 27 media 31 16 56 parties de facto parties autonomy discussion perspectives Civil society re Civil society re Freedom of Barriers to parties Barriers to parties 35 Freedom of 25 24 16 Freedom of 55 expression de facto de facto expression expression Free and fair Opposition Opposition CSO participatory 31 21 22 Party linkages 14 55 elections parties autonomy parties autonomy environment Multi-party Legislature Free and fair elections de 29 17 Party linkages 19 controls 13 Party linkages 55 elections facto resources Legislature Barriers to Multi-party Multi-party Opposition 27 controls 15 17 12 52 parties de facto elections de facto elections de facto parties autonomy resources Executive Multi-party Judicial Lower court Free and fair 11 15 11 11 52 electoral regime elections de facto accountability independence elections Multi-party Elections vote- Executive Barriers to parties elections de 10 14 Legislature exists 10 8 49 buying electoral regime de facto jure Attorney Legislature Judicial Freedom of Freedom of general, 10 12 controls 10 7 49 accountability expression expression Prosecutor resources Legislature Elections vote- Barriers to Barriers to parties Attorney general, 9 9 9 6 controls 48 buying parties de jure de jure Prosecutor resources Legislature Barriers to Freedom of Elections vote- Multi-party 8 7 8 5 investigates 47 parties de jure expression buying elections de jure executive Judicial Attorney general, Multi-party Freedom of 8 7 7 5 Suffrage 44 independence Prosecutor elections de jure assembly Legislature Multi-party Barriers to parties controls 7 6 Electoral regime 5 Legislature exists 4 40 elections de jure de jure resources 62 Legislative Legislative Free and fair Multi-party Party linkages 7 5 5 4 40 electoral regime electoral regime elections elections de facto Freedom of Attorney general, Elections vote- Multi-party 7 Electoral regime 5 5 4 37 expression Prosecutor buying elections de jure Legislature Freedom of the Freedom of Barriers to parties investigates 6 5 5 4 Legislature exists 34 press expression de jure executive Executive Executive Suffrage 5 4 4 Electoral regime 3 Electoral regime 33 electoral regime electoral regime Freedom of the Legislative Legislative 5 Legislature exists 4 Suffrage 3 3 33 press electoral regime electoral regime Legislature Freedom of Judicial Freedom of the 5 Suffrage 4 investigates 3 3 11 assembly independence press executive Legislature Judicial Freedom of the Freedom of the Freedom of 4 4 3 3 11 exists independence press press assembly Legislature Electoral Freedom of Freedom of Executive 4 4 3 investigates 2 9 regime assembly assembly electoral regime executive Legislature Legislative Judicial Attorney general, 4 investigates 2 1 Suffrage 1 9 electoral regime independence Prosecutor executive 63 Contingency Contingency Contingency Contingency East Asia conditions South-East Asia conditions South Asia conditions Caribbean conditions (max 127) (max 127) (max 127) (max 127) Highest Highest Highest High court Judicial Highest score score not score not Executive oversight Engaged society score not independence accountability not reached reached reached reached Highest Highest Highest Lower court Executive score not score not EMB autonomy 61 Executive oversight score not independence oversight reached reached reached Highest Highest Executive Gov. censorship High court Freedom of score not 70 60 score not oversight Media independence expression reached reached Highest Legislature Highest Lower court Freedom of Ombudsman score not investigates in 62 58 score not independence assembly reached practice reached Highest Legislature Freedom of score not CSO entry and exit 47 investigates in 51 Party linkages 76 expression reached practice Highest Freedom of Freedom of Gov. censorship score not 46 CSO entry and exit 48 71 expression discussion Media reached Highest Freedom of Gov. censorship Judicial score not Engaged society 45 43 70 assembly Media accountability reached Engaged society 77 EMB autonomy 44 Engaged society 43 EMB autonomy 57 High court EMB autonomy 70 Media critical 40 Ombudsman 41 51 independence Civil society re Legislature CSO entry and exit 68 Freedom of 39 Media critical 36 investigates in 49 expression practice Civil society re Print or broadcast Print or broadcast Freedom of 66 media 35 34 Media critical 47 media perspectives expression perspectives Civil society re Gov. censorship Legislature Freedom of 65 29 34 Freedom of 47 Media controls resources discussion expression 64 Legislature Civil society re CSO participatory Elections vote- investigates in 65 25 Freedom of 33 47 environment buying practice expression Freedom of Opposition parties CSO participatory 64 23 32 CSO entry and exit 46 discussion autonomy environment Barriers to parties Opposition parties Legislature Media critical 63 23 32 45 de facto autonomy controls resources Print or broadcast Freedom of CSO participatory media 62 Legislature exists 21 32 44 expression environment perspectives Free and fair Free and fair Lower court 61 20 Party linkages 30 44 elections elections independence Lower court Multi-party Attorney general, Party linkages 60 18 28 40 independence elections de facto Prosecutor Barriers to parties Executive electoral Legislature controls Free and fair 57 17 27 38 de facto regime resources elections CSO participatory High court Free and fair Freedom of 54 15 27 36 environment independence elections expression Opposition parties Barriers to parties Print or broadcast 51 Party linkages 15 24 35 autonomy de facto media perspectives Legislature Elections vote- Elections vote- Multi-party 41 15 22 35 controls resources buying buying elections de facto Judicial Executive electoral Freedom of 40 Ombudsman 15 16 34 accountability regime discussion Executive electoral Barriers to parties Judicial 35 9 14 Ombudsman 31 regime de jure accountability Attorney general, Barriers to parties Opposition parties Legislature exists 33 6 14 29 Prosecutor de jure autonomy Multi-party Multi-party 33 Electoral regime 5 13 Legislature exists 25 elections de facto elections de jure Multi-party Legislative Barriers to parties 27 5 Legislature exists 12 24 elections de jure electoral regime de facto Barriers to parties Multi-party Attorney general, Multi-party 19 5 9 19 de jure elections de facto Prosecutor elections de jure Elections vote- Judicial Freedom of the Barriers to parties 11 5 9 17 buying independence press de jure 65 Attorney general, Freedom of Freedom of 10 5 9 Electoral regime 13 Prosecutor expression assembly Legislature Legislature Legislative investigates 10 investigates 4 Suffrage 7 13 electoral regime executive executive Judicial Freedom of the 9 4 Electoral regime 7 Suffrage 12 independence press Freedom of Legislative electoral Executive electoral Suffrage 6 4 7 12 assembly regime regime Judicial Judicial Electoral regime 6 Suffrage 1 6 10 independence independence Legislature Legislature Legislative Multi-party 6 1 investigates 5 investigates 6 electoral regime elections de jure executive executive 66 Table A.3 Detailed contingency levels to reach the highest state on selected indicators. Legislature investigates in practice Level High court independence Level Executive oversight Level CSO repression Level Opposition parties autonomy 4 Opposition parties autonomy 4 Opposition parties autonomy 4 CSO entry and exit 3 Opposition parties Election free and fair 4 Election free and fair 4 Election free and fair 4 autonomy 3 Print or broadcast media Gov. censorship Media 3 Gov. censorship Media 3 perspectives 3 Freedom of discussion 3 Print or broadcast media perspectives 3 CSO repression 3 CSO repression 3 Gov. censorship Media 2 Print or broadcast media CSO repression 3 CSO entry and exit 3 Engaged society 3 critical 2 Print or broadcast media Engaged society 3 Lower court independence 3 CSO entry and exit 3 perspectives 2 Legislature investigates in CSO entry and exit 3 EMB autonomy 3 practice 3 Engaged society 2 High court independence 3 Election vote buying 3 EMB autonomy 3 Election free and fair 2 Lower court independence 3 De facto barriers to parties 3 Party linkages 3 De facto barriers to parties 2 De facto multi-party Executive oversight 3 De facto multi-party elections 3 De facto barriers to parties 3 elections 2 EMB autonomy 3 Freedom of discussion 3 De facto multi-party elections 3 Legislature exists 1 Print or broadcast media Party linkages 3 critical 2 Freedom of discussion 3 Electoral Regime Index 1 Print or broadcast media Legislative electoral regime De facto barriers to parties 3 perspectives 2 Gov. censorship Media 2 index 1 Print or broadcast media CSO participatory De facto multi-party elections 3 CSO participatory environment 2 critical 2 environment 1 CSO participatory Legislature investigates in Freedom of discussion 3 Engaged society 2 environment 2 practice 1 Print or broadcast media Legislature investigates in critical 2 practice 2 Election vote buying 2 High court independence 1 CSO participatory environment 2 Executive oversight 2 Legislature exists 1 Lower court independence 1 Judicial accountability 2 Legislature exists 1 Electoral Regime Index 1 Judicial accountability 1 67 Legislative electoral regime Election vote buying 2 Electoral Regime Index 1 index 1 Executive oversight 1 Legislative electoral regime Legislature exists 1 index 1 Legislature controls resources 1 EMB autonomy 1 Electoral Regime Index 1 Judicial accountability 1 High court independence 1 Party linkages 1 Legislative electoral regime index 1 Party linkages 1 Lower court independence 1 Election vote buying 1 Legislature controls resources 1 De jure barriers to parties 1 Judicial accountability 1 De jure barriers to parties 1 De jure barriers to parties 1 De jure multi-party elections 1 De jure barriers to parties 1 De jure multi-party elections 1 Executive electoral regime Executive electoral regime De jure multi-party elections 1 index 0 De jure multi-party elections 1 index 0 Legislature investigates Legislature controls executive by law 1 Legislature controls resources 0 Freedom of expression 1 resources 0 Executive electoral regime Attorney general/Prosecutor Share of population with Attorney general/Prosecutor index 0 general 0 suffrage 1 general 0 Attorney general/Prosecutor Legislature investigates Executive electoral regime Legislature investigates general 0 executive by law 0 index 0 executive by law 0 Attorney general/Prosecutor Judicial independence by law 0 Judicial independence by law 0 general 0 Judicial independence by law 0 Legislature investigates Ombudsman 0 Ombudsman 0 executive by law 0 Ombudsman 0 Freedom of expression 0 Freedom of expression 0 Judicial independence by law 0 Freedom of expression 0 Freedom of the press 0 Freedom of the press 0 Ombudsman 0 Freedom of the press 0 Freedom of assembly 0 Freedom of assembly 0 Freedom of the press 0 Freedom of assembly 0 Share of population with Share of population with Share of population with suffrage 0 suffrage 0 Freedom of assembly 0 suffrage 0 Contingency Conditions 62 54 57 37 68 Figure A.3 Development of De-facto Vertical and Horizontal Accountability Figure A.4 Development of De-facto Vertical and Social Accountability Figure A.5 Development of De-Facto Social and Horizontal Accountability 70 Table A.4. Uniqueness scores for variables constituting accountability indices Index name Variable name Uniqueness_Mean Uniqueness_SD Legislature investigates in practice De-facto Horizontal Accountability (v2lginvstp) 0.450763607 0.038192895 Legislature controls resources De-facto Horizontal Accountability (v2lgfunds) 0.674933712 0.028211217 High court court independence De-facto Horizontal Accountability (v2juhcind 0.4225201 0.035438714 Lower court independence De-facto Horizontal Accountability (v2juhcind 0.436616424 0.036065532 De-facto Horizontal Accountability Judicial accountability (v2juaccnt) 0.733202937 0.025124283 De-facto Horizontal Accountability Executive oversight (v2lgotovst) 0.505242725 0.036215195 Government censorship effort - De-facto Social Accountability Media (v2mecenefm) 0.325917812 0.013330779 Print/broadcast media critical De-facto Social Accountability (v2mecrit) 0.269576119 0.011950645 Print/broadcast media perspectives De-facto Social Accountability (v2merange) 0.291978976 0.013836275 De-facto Social Accountability CSO repression (v2csreprss) 0.304771888 0.012474715 CSO participatory environment De-facto Social Accountability (v2csprtcpt) 0.422154037 0.014945371 De-facto Social Accountability Engaged society (v2dlengage) 0.390589027 0.016186377 De-facto Social Accountability CSO entry and exit (v2cseeorgs) 0.260895503 0.010346011 De-facto Social Accountability Freedom of discussion (v2xcl_disc) 0.321944655 0.011879748 De-facto Vertical Accountability Elections multiparty (v2elmulpar) 0.213527399 0.013881491 De-facto Vertical Accountability Barriers to parties (v2psbars) 0.62667262 0.021009623 De-facto Vertical Accountability EMB autonomy (v2elembaut) 0.178986361 0.014415409 De-facto Vertical Accountability Party linkages (v2psprlnks) 0.855352814 0.019153117 Opposition parties autonomy De-facto Vertical Accountability (v2psoppaut) 0.6466125 0.021472596 De-facto Vertical Accountability Free and fair elections (v2elfrfair) 0.163615072 0.016193179 De-facto Vertical Accountability Election vote buying (v2elvotbuy) 0.539584443 0.027312169 71