Moving beyond input legitimacy: When do democratic innovations affect policy making?

Pogrebinschi, Thamy; Ryan, Matt Article — Published Version Moving beyond input legitimacy: When do democratic innovations affect policy making? European Journal of Political Research Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Pogrebinschi, Thamy; Ryan, Matt (2018) : Moving beyond input legitimacy: When do democratic innovations affect policy making?, European Journal of Political Research, ISSN 1475-6765, Wiley-Blackwell, Hoboken, NJ, Vol. 57, Iss. 1, pp. 135-152, https://doi.org/10.1111/1475-6765.12219 This Version is available at: https://hdl.handle.net/10419/172506 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ European Journal of Political Research 57: 135–152, 2018 135 doi: 10.1111/1475-6765.12219 Moving beyond input legitimacy: When do democratic innovations affect policy making? THAMY POGREBINSCHI1&MATTRYAN 2 1WZB Berlin Social Science Centre, Germany; 2University of Southampton, UK Abstract. This article makes three key contributions to debates surrounding the effectiveness of democratic innovation, deliberation and participation in representative political systems. In the first instance, it argues that more attention should be paid to the role that participation actually plays in governance. The literature on democratic institutional design often neglects concern about the effects of innovative institutional designs on more traditional representative fora, at the expense of concerns about their internal procedures. Second, the article argues that despite limitations, replicable systematic comparison of the effects of institutional design is both necessary and possible even at the level of national governance. A comparative analysis of 31 cases of National Public Policy Conferences (NPPCs) in Brazil is presented. Finally, the article shows that popular deliberative assemblies that vary in their familiarity and their policy area of interest, and that organise their structure and sequence deliberation in different ways can be associated with differential effects on both option analysis and option selection stages of the policy process, respectively. Keywords: democratic innovations; deliberation; participation; effectiveness; national public policy conferences Introduction In building the now extensive body of scholarship on deliberative democracy, democratic theorists seem to have taken quite seriously Jürgen Habermas’ claim that ‘the central element of the democratic process resides in the procedure of deliberative politics’ (Habermas 1996: 296). In recent years, research on democratic innovations, the new institutions ‘specifically designed to increase and deepen citizen participation in the political decision-making process’ (Smith 2009: 1), has consolidated in an academic field of its own. Considerable scholarly attention has focused on the notion of the mini-public; institutions ‘small enough to be genuinely deliberative, and representative enough to be genuinely democratic’ (Goodin & Dryzek 2006: 220). In the search for the ideal conditions to ensure the equality of participation or maximise the quality of deliberation, scholars have paid less attention to the outcomes of institutional innovations,without which no procedure can really be claimed to improve democracy. In this article, by uncovering alternative conjunctions of factors that link innovative institutional design and lawmaking, we hope to recall that Habermas’search for the ideal procedure aimed at ‘the production of legitimate law through deliberative politics’ (Habermas 1996: 318). Deliberative theory has its foundations in Habermas’ procedural concept of democracy, therefore assuming that legitimate decisions are the outcomes of the public deliberation of citizens. However, the presumption that quality of deliberation yields more democratic outcomes may be as misleading as the belief that democratic innovations, simply because C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 136 THAMY POGREBINSCHI & MATT RYAN they involve more citizens participating as equals,produce legitimate decisions.Normatively, good deliberation and equal participation may be necessary institutional design features to bring about more legitimacy in a world of increasing public distrust and political disaffection. But,empirically,this is not sufficient if democratic innovations are expected to solve the crisis of representation. In order to tackle democratic deficits, democratic innovations must also be effective. Despite considerable research on the institutional design of democratic innovations, there are few studies on the outcomes engendered by the different designs, and almost no evidence of how institutional conditions constrain or enable those outcomes. Most research on the institutional design of democratic innovations has been concerned with issues related to the equality of participation and the quality of deliberation. The question as to when either or both are conducive to effective democratic outcomes has been almost neglected. We test how different institutional design options may lead (or not) to effectiveness, measured by policy responsiveness (translation of citizens’ preferences into policy making). In order to do so, we will look into one of the purported most effective democratic innovations known to date: the National Public Policy Conferences (NPPCs) in Brazil. NPPCs stand out among democratic innovations in their claim to have impacted on the macro-democracy level (Pogrebinschi & Samuels 2014). Employing qualitative comparative analysis (QCA), we explore how various combinations of NPPCs’ institutional design features associate with agenda-setting and the decision making of Brazil’s legislature, resulting (or not) in lawmaking. Our findings suggest that institutional design features such as high or low volumes of participation, high or low dispersion of deliberative fora, institutionalisation, the policy area at stake and strong or weak civil society influence can be related in different ways to NPPC’s outputs and outcomes. In particular, we show that, contrary to what some critics of participatory democracy suppose, mass participation can indeed yield laws at the macro-democracy level. A second relevant finding is that a low number of instances within a deliberative process is a necessary institutional design feature for effectiveness in certain circumstances, confirming expectations of deliberative democrats regarding restrictive requirements of good quality deliberation. We also found that the policy area that democratic innovations concern themselves with contributes to explaining their effectiveness. If decisions authored by citizens ensure democratic legitimacy, only their transformation into actual policy and law can ensure effectiveness. Without effectiveness citizens cannot be subject to the decisions they have authored, and democracy would remain a normative ideal. Procedure x outcomes: The limits of democratic theory and case studies With notable exceptions (Fung 2006; Papadopoulos & Warin 2007; Smith 2009), few scholars have seriously considered how effectiveness should be taken into account in the institutional engineering of democratic innovations. Fung’s (2006) ‘democracy cube’, for instance, is a geometrical illustration of how different institutional design options are more or less suited to produce particular outcomes. Drawing on case studies, Fung (2006: 73) claims that democratic innovations seeking to enhance effectiveness are likely to involve C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research MOVING BEYOND INPUT LEGITIMACY 137 a small number of citizens, who would be willing to engage intensively due to a deep interest in the problems at hand. These mini-publics would also suggest that the subject of deliberation, recurrence and monitoring are other institutional design features that may generate more information to officials, and therefore increase public action (Fung 2003: 349). Scholarship recognises that mini-publics tend to be unable to impact the macro political system, and therefore rarely determine public policy. Goodin and Dryzek (2006) surveyed diverse mini-publics to ascertain how they may affect political decision making through the public sphere and representative institutions. They point to the British Columbia Citizen’s Assembly as a case of a mini-public ‘actually making policy’, although rejection of its recommendations by referendum implied no eventual policy change at all. The Danish Consensus Conferences and Texan Deliberative Polls are given as examples of mini-publics’ recommendations ‘taken up in the policy process’, although this presumption relies on secondary reporting and not on direct congruence analysis of policy or legislation.Almost no information is given on exactly how institutional design features relate to policy outcomes, when they do exist. With few exceptions, studies on democratic innovations focus on outcomes internal to the procedure and participants, and not on policy responsiveness. Many case studies have provided valuable qualitative evidence on how innovations have effects on matters like citizens’ empowerment (Baiocchi 2001), personal skills and civic competences (Grönlund et al. 2010) or better-informed decisions (Fishkin 2009). These works contribute enormously to empirical knowledge of how institutional innovations produce desirable consequences associated with a deepening of democracy. But epistemic and ethical outputs internal to innovations say nothing about their real outcomes and impacts on policy making. Studies that do address political outputs are usually circumscribed to policy impact at the local level. For the most part these studies seek to assess effectiveness by measuring whether innovations have solved specific local problems, such as water pollution (Geissel 2009) or the delivery of community policing (Fung 2006). Touchton and Wampler (2014) show how participatory budgets increase health care spending and decrease infant mortality rates in Brazilian municipalities. However, studies that aim at a more systemic approach, such as Parkinson’s (2006) analysis of citizens’ juries’ involvement in health policy making in the United Kingdom, do not yet contain any substantive comparative assessment of effectiveness. Comparing institutional designs is crucial to assessing their effectiveness.Ryan and Smith (2012) call for a systematic comparative turn in the study of democratic innovations. A subfield that is constituted by ‘innovation’ suffers more than most from a methodological bias to favour case studies. We should necessarily compare institutions that appear innovative both with one another and with more traditional institutions of governance in order to understand their effect within political systems. Deliberative democrats have rightly focused on how legitimacy can be drawn from procedurally correct decisions, but have failed to grasp that ‘government by the people’ (input legitimacy) is only meaningful as a ‘government for the people’ (output legitimacy), one which makes sure that ‘the policies adopted will generally represent effective solutions to common problems of the governed’ (Scharpf 2003). Participatory democrats for their C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research 138 THAMY POGREBINSCHI & MATT RYAN part have focused their concern on how failing representative institutions can be bypassed by innovative institutional designs to the detriment of considering their complementarity. If designs that achieve equal participation and good deliberation cannot ever engender effectiveness, there is little point in calling such innovations democratic. Designing effectiveness: Institutional variation and outcomes The few attempts to deal with the relation between institutional design and effectiveness of democratic innovations seem to suffer three main limits. First, they rely on secondary narrative accounts and do not directly and systematically investigate relations between process’ characteristics and outcomes achieved. Second, they are mostly restricted to case studies at the local level, lacking wider systematic comparisons. Third, they focus on outputs that take place within the process neglecting political outputs and outcomes at the macrolevel. We show how different combinations of institutional design features (or their absence) may lead to more (or less) effectiveness, measured by responsive policy making. Our assumption is that an effective democratic innovation is one that will impact on policy either directly (when authorised by delegation or devolution) or indirectly through the activity of authorised institutions like the legislature or the public administration. Such impact can be gauged at different stages of the policy cycle. We assume that to be considered effective, a democratic innovation must directly or indirectly affect at least one of these stages. We expect that effectiveness is determined by different combinations of institutional design features (or their absence) and that relationships between input conditions and outcomes are likely to be multiple and conjunctural with a variety of designs having alternative effects in different circumstances. We test this general expectation by looking at the NPPCs and their impact on the Brazilian legislature. The NPPCs consist of a series of simultaneous and subsequent instances of deliberative mini-publics that link the local, the state and the national levels through chains of delegation involving civil society organisations (CSOs) and government representatives, as well as ordinary citizens and private stakeholders (Avritzer & Souza 2013). Although they may not offer the cocooned spaces envisaged by some mini-publics such as citizen’s juries, typically NPPCs will alternate facilitated deliberation in small working groups where recommendations are prioritised, with larger plenary discussions that take place at the closure of each round. The policy impact of NPPCs has been maintained in several studies (Pogrebinschi 2013; Souza & Pires 2013; Pogrebinschi & Samuels 2014), but none have engaged in systematic comparative tests. Given that NPPCs are innovations specifically designed to provide inputs to policy making, we examine whether their purported outcomes can be explained by their institutional design features. NPPCs are a national-level democratic innovation. Their impact is expected mostly at this level, either in the shape of policies formulated and enacted by the national legislature or designed and implemented by the federal public administration. The federal executive implements the NPPCs, and thus some direct impact is expected in the federal administration. As we want to test whether deliberation is translated into lawmaking both at the agenda-setting and decision-making stages, we assume that effectiveness is indicated C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research MOVING BEYOND INPUT LEGITIMACY 139 when policies drafted as bills or enacted as laws by the national legislature are congruent with, and therefore responsive to, NPPCs’ recommendations. We look at how NPPCs influence both problem definition and policy formulation in the legislature, assessing how proposed and enacted legislation responds to the NPPC’s recommendations. We will therefore consider the diverse political outputs and outcomes of the NPPCs. In order to systematically compare NPPCs and evaluate and elucidate the diverse conjunctions of the presence and absence of key conditions that link procedure to outcomes, we undertake a qualitative comparative analysis (QCA). Fuzzy-set QCA, as introduced by Ragin (2000), provides a structured method for conducting set-theoretic analysis in the social sciences, particularly where researchers are dealing with more than a handful of cases. While social scientists should be interested in correlation coefficients, they are also interested in necessity and sufficiency. Much of the causal interpretation and theorising that is traditional to arguments proffered by scholars of democratic institutional design has been instinctively set-theoretic, underpinned by implication rather than covariation hypotheses (see Thiem et al. 2016a). Hypotheses often consider conditions like mass participation, deliberation or civil society demand to be required or almost always required in order to produce or negate a given democratic outcome; and/or combinations of these conditions that always or almost always suffice to produce an outcome. Necessity and sufficiency are logical properties of set relations. QCA has only recently been introduced to the subfield of democratic innovations research in the work of Ryan and Smith (2012). While QCA requires good theory and case-knowledge on the part of users to reduce risks of measurement error and model misspecification (see Thiem et al. 2016b for a review of debates),QCA tools like truth tables and Boolean reduction are extremely useful for identifying parsimonious descriptions of combinatorial set-relations across cases (Berg-Schlosser & Cronqvist 2005). The method holds particular promise for research in this field then, as it matures beyond small populations of distinct innovative case studies to cumulating knowledge across multiple comparable cases. How outcomes relate to procedure: A qualitative comparative analysis of the NPPCs Although the NPPCs have taken place under a variety of governments, they evolved after Brazil’s re-democratisation and significantly under the Workers’ Party’s (Partido dos Trabalhadores, PT) many years in office. The election of President Luiz Inácio Lula da Silva in 2003 prompted a vivid shift in the frequency and breadth of NPPCs. Lula introduced 28 new policy areas, raising the total number of issues deliberated through NPPCs to 40. Lula’s eight years in office also sees a palpable increase in the number of policy recommendations arising from NPPCs (Pogrebinschi & Samuels 2014). With this intensification and variance in impact of policy recommendations, an opportunity for systematic comparison and a robust contribution to the general study of democratic institutional innovation presents itself. We aim at demonstrating that certain institutional design features of the NPPCs are necessary or sufficient for policy making in the legislature.We discuss our cases in the light of cross-case findings in the final sections.Before moving to the analysis, the following sections make transparent the decisions involved in constructing a comparison of cases of NPPCs. C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research 140 THAMY POGREBINSCHI & MATT RYAN Cases and data Our analysis focuses on the NPPCs that have occurred between 2003 and 2010, which is the period corresponding to Lula’s two terms in office. As we are interested in the institutional design features that led NPPC’s recommendations to be taken up by federal legislators, it is useful to hold the political environment relatively constant in the first instance.1 Our data have been compiled using two distinct datasets.For constructing the influencing conditions, we obtained most data from figures compiled by the Institute for Applied Economic Research in Brazil (IPEA).2Despite our numerous efforts to improve the IPEA dataset on institutional design features of NPPCs, data on the five conditions of interest that we test in the models presented in this article remained missing in some of the 60 cases that take place in this time period. List-wise deletion of cases with missing data has produced a final population of cases for analysis of 31. Errors leading to missing data records do not appear to be systematic (see Online Appendix B for analysis). For constructing the outcomes, we have used Pogrebinschi’s (2010) dataset, which comprises the final recommendations from all NPPCs held between 1988 and 2010, as well as all legislative acts (bills introduced, laws and constitutional amendments enacted by the legislature) congruent with recommendations. A congruent legislative act is one that matches the preferences contained in the NPPCs’ final recommendations. For example, when the first NPPC on Youth, held in 2008, recommended that the youth be considered a subject entitled to rights, in 2010 the legislature enacted a constitutional amendment stating precisely that – this legislative act is considered to be congruent.3By counting congruent legislative acts, the dataset provides a proxy for policy responsiveness. Our data includes the congruent legislative acts from between January 2003 and October 2010. We note that this leaves less than 200 days for responses for some NPPCs that finished in early 2010. The dataset also leaves aside bills proposed and laws passed between October and December 2010. To accommodate this and ensure comparability, we exclude from our analysis cases whose final recommendations were produced after October 2009. In the next section, we introduce our explanatory conditions and some relevant expectations. We then present two models investigating relations between conditions based on two separate outcome sets calibrated from the data on legislative responses to NPPC policy recommendations. In Online Appendix A we explain in detail how we calibrated our variables to represent sets of the hypothesised conditions as we understand them. Outcome conditions The two outcome conditions we analyse refer to different stages of the policy cycle, as well as to two moments of the legislative process. Policy outputs. The first outcome condition seeks to explain when democratic innovations are effective in influencing the definition of policy in the legislature. If democratic innovations cannot yield authoritative final decisions, they can at least begin to sway the legislative process in the first instance by influencing agendas. We assume here that the NPPCs have an impact on the agenda-setting stage of policy making.We calculated legislative bills that are congruent with recommendations of a given NPPC as a proportion of the total C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research MOVING BEYOND INPUT LEGITIMACY 141 number of recommendations made by that NPPC. We then make a qualitative distinction between proportions of response that signify policy output and those that do not, as explained in Online Appendix A. Policy outcomes. The second outcome of interest is the set of NPPCs that informed the formulation of policy in the legislature. Our assumption is that the impact of some NPPCs goes beyond agenda-setting, and reaches the decision making stage of policy making. We aim to verify under what conditions NPPCs’ recommendations are enacted as laws and constitutional amendments, providing not only policy outputs, but also concrete outcomes. Influencing conditions Institutionalisation. We expect that NPPCs institutionalised over time may build networks and path dependence in channels of communication with the legislature. Policy makers may place greater trust in processes that have witnessed repeated incidence and built up learning and familiarity. Therefore, we expect that when the kinds of legitimacy afforded by mass participation of citizens and strong roles for civil society are combined with a buildup of bureaucratic know-how and trust in the process over time, this allows legislators the confidence to propose laws that were conceived in democratic innovations (E2). However, NPPCs that have taken place a number of times may present a lower demand for new legislation in their specific policy area, and aim more at monitoring already enacted legislation. In addition, NPPCs that take place for the first or second time may deal with a relatively new policy area (like environment and minority rights), and therefore present more opportunities and demand for legislative action. We have to investigate alternate causation here. In particular, when NPPCs are supported by civil society actors to motivate large numbers of participants in newly opened policy areas that focus on redistribution and recognition we expect to see large numbers of recommendations passed into law by legislators (E3). Redistributive issue. This set comprises NPPCs that are organised toward a policy area that deals with redistributive and recognition issues. We assume that variance in the subjects and stakes on which participation takes place may play a role in the policy process (Fung 2003). We also know that different democratic innovations are seen to be more or less relevant and/or authoritative with regard to social redistribution or more thematic policy concerns (Smith 2009). Following this literature, we expect that designs with high numbers of participants that focus on redistributive policies will be associated with both outputs and outcomes to an extent that other conditions lose their explanatory relevance (E4). We also expect that designs that focus on broader thematic policy areas that employ a small number of deliberative instances are associated with policy outputs and outcomes in a similar way (E5). Civil society control of organising committee. This set comprises NPPCs where members of civil society make up the organising committee. The committee plays an important C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research 142 THAMY POGREBINSCHI & MATT RYAN role in making decisions concerning the NPPC’s rules and procedures, and can therefore significantly influence its process. We are particularly interested to ascertain if the presence or absence of a meaningful majority of civil society actors in the organisation of an NPPC is an INUS4condition in different paths to effecting legislative activity that is congruent with policy recommendations. It is possible that, depending on the degree to which other factors like high citizen participation or a high number of deliberative instances are present or absent, legislators are determined (or not) to act given the strong presence of civil society in the organising commission. Mass participation. This set contains NPPCs that involve a critical mass of participants in deliberations across all levels of the process. In our dataset, the numbers of those who participate in an NPPC ranges from 10,000 to 524,461. Many theorists have argued that the combination of mass participation and deliberation is unfeasible (Fishkin 2009; Przeworski 2010), while others have argued that it may be sequenced in a democratic process in useful ways (Smith 2009). Mass participation can signal to the legislature that there is a tangible constituency of support for change in a policy area. We have a strong expectation that high levels of participation may lend legitimacy to the democratic process and make it harder to ignore NPPC policy recommendations. In other words, there may be certain quantities of participation that when reached, influence decision making and therefore increase its democratic quality. This quality of mass participation may be necessary or sufficient to effect policy action. Decentralised institutional design This set represents the degree to which municipalities and other subsidiaries are involved in the first stage of the process. While each NPPC should involve a sequence of deliberations scaling up from local, to state, and then to national level, NPPCs are different in the extent to which this is mandated or encouraged in the organisational rules. Some policy areas have geographic/thematic niches. The sample comprises an NPPC on indigenous people’s health, for example, which, due to the nature of the policy it addresses, does not have its first stages held in the municipalities, but rather within each of the different indigenous groups existent in Brazil. We have therefore used this number of ‘subsidiaries’ to measure the extent of participation in the first stage of the NPPC process. We aim at investigating whether a large spread of municipal/subsidiary first stages can signal more support for certain policies and therefore prompt higher legislative response. Similarly NPPCs can be designed with less spread to leverage more concentrated deliberation among affected populations. Table 1 summarises our expectations regarding set-relations among conditions, while Table 2 provides the fuzzy-set membership scores in each condition (set) for each case.Cases ascribed a score of ‘0’ have no membership in a set (did not display the condition). Cases with a score of ‘1’ have full membership (condition observed). Values between ‘1’ and ‘0’ represent various partial memberships in the set (condition is present to a degree). Analysis of necessary conditions We first tested for necessary conditions using the inclusion algorithm as developed in the work of Bol and Luppi (2013) and using the QCA package for R (Thiem & Dusa C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research MOVING BEYOND INPUT LEGITIMACY 149 committee, NPPCs on redistributive policy areas still impact political decision making when they are well institutionalised (S7). This solution has low coverage (raw =0.84, unique =0.058). The one good example we have of this type of case is the eleventh edition of the NPPC on Human Rights held in 2008. This is a particular case, because after many controversial and conflictual previous editions, this NPPC aimed specifically at revising the existing national policy on human rights. The National Policy on Human Rights 3, enacted by the executive, was the main result of the NPPC, almost paraphrasing many of its recommendations.As several recommendations explicitly demanded supporting legislative activity for implementation, the legislature was pressed into action. Given its specific aim of drafting the new human rights plan, this NPPC had no municipal stages, explaining its low decentralisation. The legislature (the Human Rights Commission of the Chamber of Deputies) organised this NPPC together with the federal executive, which may have reduced the influence of civil society in the organising committee, but, on the other hand, may have reduced barriers to the support needed for recommendations to become legislation. This case shows how participatory processes held repeatedly over time may experience institutionalisation within the representative policy-making routines themselves, and therefore dispense with massive numbers of participants and civil society control of a process as legitimisers, as well as no longer requiring a largely decentralised deliberative process in order to be effective. Conclusion The search for the normative legitimacy promised by deliberative democracy has led democratic theorists to produce a considerable amount of work with the intention of designing a procedure that could reproduce, as close as possible, Habermas’ ideal speech situation. The belief that equal participation and good deliberation would ensure more legitimate and therefore more democratic decisions seem to have obfuscated an essential part of this equation: policy outcomes. While in the normative world of democratic theory a well-designed procedure is sufficient to provide legitimacy to decisions, in the empirical world of real existing – and representative – democracies, an ideal institutional design would not bring much democratic quality if it were not to have an effect on legislation or legislative agendas. This study is very much among the first of its kind in that it tries to associate the consequences of the institutional and environmental circumstances of democratic innovations with their effect on legislative activity. One important limitation of this study is we cannot, at the same moment, assess to what extent the legislative agendas we discuss may have been shaped by circumstances external to the NPPCs. With the available data we cannot give surety across cases as to whether participatory processes are only used to legitimate decisions made elsewhere. We hope that our findings can inspire research strategies that, for example, seek testimonies from legislators as to what motivated their responses. However, the level of congruence between recommendations of NPPCs and legislative actions, at the very least, suggests that this is less the zero-sum game it is sometimes made out to be by radical participatory democrats. We also think that this research provides an important bridge between the work on how democratic innovations aim to achieve outcomes and the wider literature covering what conditions affect agendas C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research 150 THAMY POGREBINSCHI & MATT RYAN and decisions in public policy. In other words, our work would be complemented nicely by further research that was able to compare relevant legislative effects among democratic innovations and other institutions/phenomena which are known to achieve such effects. Despite these limitations, the work contains a number of important findings. The results show that, contrary to what critics of participatory democracy suppose, mass participation designs can indeed yield lawmaking at the macro-democracy level. In addition to that, we showed that a compacted deliberative process is also an institutional design feature that can lead to effectiveness in certain circumstances, confirming deliberative democrats’ expectations regarding the restrictive requirements of a good quality deliberation. We have also found that NPPCs that open up new non-redistributive policy areas to participation can be effective in agenda-setting when civil society takes a back seat in process organisation and there is mass participation in a decentralised institutional design. Nonredistributive NPPCs are also effective where they have more streamlined designs, but once institutionalised, may require a significant participatory mandate to achieve high levels of legislative response. On the other hand, we found that democratic innovations concerned with redistributive policies very often affect political decisions and see their recommendations passed into law. This happens when mass participation is combined with CSO support in less expansive designs. Nevertheless, the latter two conditions are irrelevant when innovations are a trailblazer in the policy area. Finally, institutionalised NPPCs rely on a great number of participants and dispersed local instances of deliberation for legitimacy in the eyes of legislators. When it comes to designing democratic institutions, effectiveness should be as important as legitimacy. Otherwise the potential to solve democratic deficits and offer a way out of the crisis of representation is very limited. Acknowledgements We are greatly indebted to Tiago Ventura, who worked uncountable hours on the preparation of the raw data we use in this article, and also provided us with good insights in the process of data calibration and missing data analysis. We also thank Talita Tanscheit and Alessandro Amorim for helping us to complete the missing data searches. An earlier version of this article was presented at the American Political Science Association Annual Meeting, the European Consortium for Political Research General Conference and the Political Studies Association Annual Conference. We are grateful to the discussants and audiences for their comments.We also thank David Samuels,Graham Smith,Brian Wampler and the anonymous reviewers of this journal for their valuable criticism. Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Appendix A: Calibration of set membership values from raw data Appendix B: Analysis of missing cases C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research MOVING BEYOND INPUT LEGITIMACY 151 Appendix C: Truth Tables Appendix D: Raw Data Matrix (60 cases including those with missing values) Appendix E: Replication code for R (also available as .R file) Appendix F: Raw_data_file (31 cases following listwise deletion) Notes 1. The data for the original 60 cases is reproduced in the Online Appendix. Although there were 74 NPPCs under Lula, we adopt Pogrebinschi’s (2010) classification, and look only to cases national in scale, deliberative in scope and normative for producing recommendations. NPPCs were also successful in contributing to policy production under previous regimes, but by reducing our scope we provide an important control for endogeneity problems linked to variance in the political environment. 2. www.ipea.gov.br/participacao/2012-06-14-1811-50 3. The method used to collect the data is described in Pogrebinschi (2010) and Pogrebinschi and Samuels (2014). See also http://led.iesp.uerj.br/index.php/pesquisas/dados 4. Independent necessary part of an unnecessary but sufficient combination of conditions. 5. Data and replication code including details of robustness tests are provided in the Online Appendix. 6. A sufficient but unnecessary part of a factor that is insufficient but necessary for an outcome. 7. Core conditions indicate the most parsimonious possible form the solution can take if certain counterfactual assumptions about logical remainders (combinations of conditions for which we have no empirical cases) are accepted. References Avritzer, L. & Souza, C. (eds) (2013). Conferências Nacionais: Atores, dinâmicas participativas e efetividade. Brasília: IPEA. Baiocchi, G. (2001). Participation, activism and politics: The Porto Alegre experiment and deliberative democratic theory. Politics and Society 29(1): 43–72. Berg-Schlosser, D. & Cronqvist, L. (2005). Macro-quantitative vs macro-qualitative research in the social sciences: An example from empirical democratic theory employing new software. Historical Social Research 30(4): 154–175. Bol, D. & Luppi, F. (2013). Confronting theories based on necessary relations: Making the best of QCA possibilities. Political Research Quarterly 60(1): 205–210. Fishkin, J.S. (2009). When the people speak: Deliberative democracy and public consultation. Oxford: Oxford University Press. Fung, A. (2003). Recipes for public spheres: Eight institutional design choices and their consequences. Journal of Political Philosophy 11(3): 338–367. Fung, A. (2006). Varieties of participation in complex governance. Public Administration Review 66(s1): 66–75. Geissel, B. (2009). How to improve the quality of democracy? Experiences with participative innovations at the local level in Germany. German Politics and Society 27(4): 51–71. Goodin,R.E.& Dryzek,J.S.(2006).Deliberative impacts:The macro-political uptake of mini-publics.Politics and Society 34(2): 219–244. Grönlund, K., Setälä, M. & Herne, K. (2010). Deliberation and civic virtue: Lessons from a citizen deliberation experiment. European Political Science Review 1: 95–117. Habermas, J. (1996). Between facts and norms: Contributions to a discourse theory of law and democracy. Cambridge, MA: MIT Press. Papadopoulos, Y. & Warin, P. (2007). Are innovative, participatory and deliberative procedures in policy making democratic and effective? European Journal of Political Research 46: 445–472. Parkinson, J. (2006). Deliberating in the real world: Problems of legitimacy in deliberative democracy. Oxford: Oxford University Press. C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research 152 THAMY POGREBINSCHI & MATT RYAN Pogrebinschi, T. (2010). Entre Representação e Participação: As Conferências Nacionais e o Experimentalismo Democrático Brasileiro. In Série Pensando o Direito (27). Brasília: Ministério da Justiça. Pogrebinschi,T.(2013).The squared circle of participatory democracy:Scaling up deliberation to the national level. Critical Policy Studies 7(3): 219–241. Pogrebinschi, T. & Samuels, D. (2014). The impact of participatory democracy: Evidence from Brazil’s National Public Policy Conferences. Comparative Politics 46(3): 313–332. Przeworski,A.(2010).Democracy and the limits of self-government.Cambridge:Cambridge University Press. Ragin, C.C. (2000): Fuzzy-set social science. Chicago, IL: University of Chicago Press. Ryan M. & Smith, G. (2012). Towards a comparative analysis of democratic innovations: Lessons from a small-N fsQCA of participatory budgeting. Revista Internacional de Sociología 70(extra 2): 89–120. Samuels, D. (2013). Brazil: Democracy under Lula and the PT. In J.I. Domínguez & M. Shifter (eds), Constructing democratic governance in Latin America, 4th edn. Baltimore, MD: Johns Hopkins University Press. Scharpf, F.W. (2003). Problem-solving effectiveness and democratic accountability in the EU. Working Paper 03/1. Cologne: Max Planck Institute for the Study of Societies. Smith, G. (2009). Democratic innovations: Designing institutions for citizen participation. Cambridge: Cambridge University Press. Souza, C. & Pires, R. (2013). Conferências Nacionais como Interfaces Socioestatais: seus usos e papéis na perspectiva de gestores federais. In L. Avritzer & C. Souza (eds), Conferências Nacionais: Atores, dinâmicas participativas e efetividade. Brasília: IPEA. Thiem, A. & Dusa, A. (2014). Qualitative comparative analysis with R: A user’s guide. London: Springer. Thiem, A., Baumgartner, M. & Bol, D. (2016a). Still lost in translation! A correction of three misunderstandings between configurational comparativists and regressional analysts. Comparative Political Studies 49(6): 742–774. Thiem, A., Spöhel, R. & Dusa, A. (2016b). Enhancing sensitivity diagnostics for qualitative comparative analysis: A combinatorial approach. Political Analysis 24(1): 104–120. Touchton, M. & Wampler, B. (2014). Improving social well-being through new democratic institutions. Comparative Political Studies 47(10): 1442–1469. Address for correspondence: Thamy Pogrebinschi, WZB Berlin Social Science Centre, Reichpietschufer 50, D-10785 Berlin, Germany. Email: [email protected] C2017 The Authors.European Journal of Political Research published by John Wiley & Sons Ltd.on behalf of European Consortium for Political Research