ERC Consolidator Grant 2016 - Research proposal [Part B2] project.pdf · on-going “standard” (inquiring into questions of democratization using standard approaches) V-Dem research

Lindberg Part B2 FASDEM

1

ERC Consolidator Grant 2016 - Research proposal [Part B2]

Failing and Successful Sequences of Democratization FASDEM

Part B2: The scientific proposal (max. 15 pages) The study of democracy and democratization lies at the center of political science and is increasingly important in economics, sociology, and history. In the post-Cold War world, democracy has also become a central foreign policy objective for many countries, and is often a critical condition for the distribution of international development assistance. The transition to democracy and its consolidation remain key issues in global development today. Yet, uncertainty persists over why some countries become and remain democratic and others do not (e.g. Acemoglu and Robinson 2006, Boix 2003, Houle 2009, Inglehart et al 2005, Lindberg 2009, Lipset 1959, Paxton 2000, Przeworski et al 2000, Teorell 2010). The recent literature provides a range of specific hypotheses that, while accompanied by relatively sophisticated empirical strategies for testing, reach few firm conclusions.

Most critically perhaps, questions about causality remain unresolved (e.g. Acemoglu, et al 2008; Bueno de Mesquita et al. 2003, Glaeser et al. 2007, Schedler 2013, Seawright 2010, Simpser 2013, Stokes et al. 2013; Weitz-Shapiro 2013). How to establish causal relationships for observational data, when controlled or natural experiments are not options, is one of the thorniest issues in social science. The challenge is multiplied when we are confronted with chains of relationships between many variables.

FASDEM, if funded, will capitalize on a set of novel analytical approaches and methods adapted from modeling in evolutionary biology by a research team under the leadership of the PI in a related, ongoing project. Critically, it will also take advantage of the new, unique V-Dem dataset with 15 million data (Coppedge 2015b,c), that is the first disaggregated democracy (and among the first in national-level comparative social science) data to include score-specific reliability estimates.The new methods, that together can establish descriptively the sequences between sets of hundreds of ordinal variables, make the current proposed project possible. FASDEM will also take an additional step, developing upon the latest statistical methodologies of establishing causal identification in observational data, and use these to test each step of such manifest sequences.

Hence, FASDEM promises to revolutionize our understanding of both the failing trajectories of democracy, and the successful pathways, and tell us why this is the case. These findings would open up a new world of the largely unexplored area of endogenous sequences of democratization, by addressing two key questions:

1. Which are the failing versus successful sequences of democratization? 2. What are the determining causal relationships in these sequences?

Answering these questions would for example allow us to provide advice to policymakers, national and international actors about what they, at the particular juncture a country is at, should focus their support on: empowering the media, strengthening the autonomy of the judiciary, building political parties, or strengthening elections to be free and fair? Which is the better strategy? FASDEM is designed to provide a first set of answers based on the first systematic and comprehensive effort using a variety of cutting-edge methods and the new and unique V-Dem database. Its academic urgency and social relevance can hardly be overstated given the dearth of systematic, comprehensive research in this field. The international community spends billions of Euros every year on democracy support with very little sound scientific evidence as to its adequacy. Critical for the viability of this proposal is the new Varieties of Democracy (V-Dem) dataset and infrastructure. I, the applicant of the current proposal, is one of four PIs for the V-Dem project, and Director of the V-Dem Institute, have led the development of this new resource over the past 6 years. I also lead the on-going “standard” (inquiring into questions of democratization using standard approaches) V-Dem research program. The unique V-Dem dataset includes some 350 indicators, 34 component-indices, and five main indices of varieties of democracy from 1900 to the present for 173countries – about 15 million data points on democracy. Using the set of “breakthrough” methods for which I led the development during the past two years, the proposed FASDEM project will make a radical departure from the crude and

This proposal version was submitted by Staffan I. LINDBERG on 01/02/2016 10:42:24 Brussels Local Time. Issued by the Participant Portal Submission Service.


2

“correlational” paradigm in democratization studies in order to detail and explain sequences of democratization (aka chains of endogenous causal effects) for the first time. The Prerequisite: State of the Art of Measuring Democracy One of the obstacles to advancement in the field of democratization studies is the absence of a wide-ranging database that tracks multifarious aspects of countries’ institutional histories. Several democracy-measurement ventures cover only democracies (e.g. Lijphart 1999) or particular regions (e.g. Munck 2009, Lindberg 2006), or just one aspect such as elections (e.g. NELDA). Existing democracy indices that seek to measure democracy tend to be problematic in several respects. This includes the BNR index developed by Bernhard, Nordstrom & Reenock (2001); the Bertelsmann Transformation Index (“BTI”) directed by the Bertelsmann Stiftung (various years); the Democracy Barometer developed by Wolfgang Merkel & associates (Bühlmann, Merkel, Müller & Weßels 2012); the BMR index developed by Boix, Miller & Rosato (2013), the Contestation and Inclusiveness indices developed by Coppedge, Alvarez, & Maldonado (2008); the Political Rights, Civil Liberty, Nations in Transit, and Countries at the Crossroads indices, all sponsored by Freedom House (freedomhouse.org); the Economist Intelligence Unit (2010) index (“EIU”); the Unified Democracy Scores (“UDS”) developed by Pemstein, Meserve & Melton (2010); the Polity2 index from the Polity IV database (Marshall, Gurr & Jaggers 2014); the democracy-dictatorship (“DD”) index developed by Adam Przeworski & colleagues (Alvarez, Cheibub, Limongi & Przeworski 1996; Cheibub, Gandhi & Vreeland 2010); the Lexical index of electoral democracy developed by Skaaning, Gerring & Bartusevičius (2015); the Competition and Participation indices developed by Tatu Vanhanen (2000); and the Voice and Accountability index developed as part of the Worldwide Governance Indicators (“WGI”) (Kaufmann, Kraay & Mastruzzi 2010). To summarize the issues with this range of alternatives broadly for the sake of brevity: i) they focus primarily on the “electoral” dimension of democracy; ii) they are comprised of components not measured independently of each other; iii) except for Polity IV, they do not extend back far in time for a global sample of countries; iv) they are not transparent in design and replicable, hence the accuracy of the data is either unknown or unreported; v) they do not address potential problems of measurement error and vi) they cannot make fine distinctions across polities or through time in a reliable fashion (Treier 2008; Pemstein et al. 2010). These complex methodological issues are discussed at length in Coppedge et al 2015a, d (see also Coppedge et al. 2011) and in other studies (e.g. Beetham 1994; Bollen et al 2000; Bowman et al 2005; Munck et al 2002; Treier 2008). Just to mention some thing, existing indicators of democracy are typically highly aggregated, bundling a great many potential causal factors together. Even when they purport to measure sub-components of a concept, it is uncertain whether these sub-components are measured independently (e.g., the various components tracked by Polity IV, the civil liberties and political rights indices compiled by Freedom House). A counterfactual understanding of causality is thus difficult to apply and it is unclear what the potential causal mechanisms might be, even when a strong x → y relationship is uncovered. Problems of causal heterogeneity associated with a highly aggregated causal factor make it difficult to confirm or discredit a hypothesis, even when the correlational evidence is strong. In contrast, V-Dem is a wide-ranging database consisting of a series of measures of varying ideas of what democracy is or ought to be (electoral-, liberal-, participatory-, deliberative-, and egalitarian democracy ideals), a wide variety of 34 meso-level indices of different components of such ideals of democracy, and about 350 specific indicators (for details, see Coppedge et al. 2015b). In addition to disaggregation, several features of the V-Dem project deserve emphasis:

• Historical data extending back to 1900 (eventually to 1789) • Multiple, independent coders for each (non-factual) question • Inter-coder reliability tests, incorporated into a Bayesian measurement model • Confidence bounds for all point estimates associated with non-factual questions • Multiple indices reflecting varying theories of democracy • Transparent aggregation procedures • All data freely available, including original coder-level judgments (exclusive of any personal

identifying information) Of particular interest is attention to estimating error. Although V-Dem has an elaborate protocol for careful selection, experts exhibit varying levels of reliability and bias, and may not interpret questions consistently,



3

which necessitates using a measurement model (Bollen & Paxton 2000, Clinton & Lapinski 2006, Clinton & Lewis 2008, Jackman 2004, Treier & Jackman 2008, Pemstein, Meserve & Melton 2010). To combine expert ratings for a particular country-indicator-year to generate a single “best estimate” for each question, V-Dem employs methods inspired by psychometric and educational testing (see, e.g., Lord & Novick 1968, Jonson & Albert 1999, Junker 1999, Patz & Junker 1999). These measurement models have straightforward bases: patterns of cross-rater (dis)agreement to estimate variations in reliability and systematic bias. In turn, these techniques make use of the bias and reliability estimates to adjust estimates of the latent—that is, only indirectly observed—concept (e.g., executive respect for the constitution, judicial independence, or property rights) in question. Specifically, V-Dem fits ordinal IRT models (see Johnson & Albert 1999 for a technical description). These models achieve three goals. First, they work by treating coders’ ordinal ratings as imperfect reflections of interval-level latent concepts. Second, IRT models allow for the possibility that coders have different thresholds for their ratings (e.g. one coder’s somewhat might fall above another coder’s almost on the latent scale), estimate those thresholds from patterns in the data, and adjust latent trait estimates accordingly to correct for this potential source of bias. Finally, IRT models assume that coder reliability varies, produce estimates of rater precision, and use these estimates—in combination with the amount of available data and the extent to which coders agree—to quantify confidence in reported scores. Using measures accompanied by estimates of confidence is an important step in guarding against the potential for bias and forces both academics and practitioners to be more honest in their empirical assessments of the validity of theoretical arguments and the effectiveness of applied policies. The V-Dem data is the first disaggregated democracy (and among the first in national-level comparative social science) data to include score-specific reliability estimates. These estimates will also enable FASDEM to focus on only the most reliable data or to weight analyses to correct, for instance, for heteroscedasticity when our indicators are the dependent variables and reliability is correlated with one or more independent variables. I am currently the Director of the V-Dem Institute that is the organizational headquarters for the V-Dem data infrastructure; PI for the on-going (2014-2018) standard research program; as well as PI for the funding associated with both, totalling €9mn. V-Dem is a collaboration across a large team of academics, including leading exerts in the areas: democratization, executives, and Europe (J. Teorell/Lund U); measurement of democracy, diffusion and Latin America (M. Coppedge/U of Notre Dame); democracy and development, qualitative methods, and Western democracies (J. Gerring/Boston U); direct democracy, and Central Americas (D. Altman/UPC, Chile); gender and democracy (P. Paxton/U of Texas); civil liberties, and Europe (SE. Skaaning/Aarhus U); the judiciary and Latin America (J. Staton/Emroy U); civil society, historical methods, and Eastern Europe (M. Bernard/U of Florida); sub-national democracy and former Soviet Union (K. McMann/Case Western); political parties, and Asia (A. Hicken/U of Michigan); legislatures, and post-communist countries (S. Fish/UC Berkeley); statistical modeling and causal inference (A. Glynn/Emroy U); sequencing and evolutionary models (P. Lindenfors/Stockholm U); and experimental research and democratization (B. Zimmerman/UNC Chapel Hill). In addition to these scholars, the PI of the current proposal supervises and has supervised several postdocs contributing to the V-Dem Institute: Eitan Tzelgov (2013-15, PhD Penn State U); Yi-ting Wang (2013-15, PhD Duke U); Brigitte Zimmerman (2014-15, PhD UCSD); Carolien van Ham (2014-15, PhD EUI); Anna Lührmann (2015-present, PhD Humbolt U); Kyle Marquardt (2015-present, PhD U Winsconsin); and Rachel Sigman (2015-present, PhD Syracruse U). As part of the global research team managed by me and my team at the V-Dem Institute, there are 37 academics who function as Regional Managers, 164 who are Country Coordinators, and over 2,600 academic and other experts that as Country Experts supply the country-ratings. At the V-Dem Institute, there are two program managers, two analysts, a data officer, three postdocs, five PhD students, and most of the time several interns. Further information can be found on the website www.v-dem.net. Going Beyond the State of the Art on Democratization The study of democratization has been enriched with both compelling theorists and skilled area specialists (e.g. Diamond et al 1990, Linz et al 1996, O’Donnell et al 1986, Schedler 2001) providing numerous insights based on both country-specific and comparative work. Quantitative democratization researchers started compiling datasets in the 1960s and have diligently tested hypotheses ever since (e.g. Bollen 1993, Hannan et al 1981, Inglehart et al 2005, Jackman 1973, Norris 2008, Lipset 1959, Przeworski et al 2000). Yet, few propositions have been rejected and most remain inconclusive. Existing work typically tells us



4

which factors are associated with democracy, but rarely why or under what conditions (e.g. Acemoglu et al 2005, Boix 2003, Coppedge 2012, Haerpfer et al 2009, Przeworski 1991, Teorell 2010). Democratization studies is also awash with typologies depicting semi-authoritarian regimes (e.g. Gandhi 2010; Geddes 1999; Levitsky et al 2002); democratic regimes (e.g., Lijphart 1999); or on the entire range of regimes (e.g. Collier et al 1997). Democratization studies may be divided into two main categories: external and endogenous. First, there are studies focused on societal or international-system factors that are external to the political system such as geography, modernization, colonialism, inequality, and societal/class conflict. Beginning with Lipset’s landmark (1959) study, a long debate has ensued about whether, or to what extent, modernization affects democratization and democratic consolidation (Acemoglu et al 2009; Boix 2003; Bollen 1983; Bollen et al 1985; Burkhart et al 1994; Huntington 1991; Przeworski et al 1997; Teorell 2010). Analyses have alternated between corroborating that modernization only facilitates democratic stability, not transitions (Przeworski et al. 2000); that it facilitates neither transitions nor stability (Acemoglu et al. 2009); or that it facilitates both (Boix 2003). Yet, we still have few robust results about causes of democratization in general. Regardless, such historical/structural factors are typically less relevant for informing policy. We cannot rewrite history and change past colonial powers or geographic location. Economic development may be mallable but only over the very long term. Second, there are studies focused on endogenous regime factors, including institutions from elections and electoral systems to the role of autonomous judiciary; on actors such as leaders, civil society, the media; etc. These are also typically factors that are more malleable. Yet, we know in fact very little about endogenous sequences of changes (aka chains of causal relationships among components) that precipitate failing and successful democratization outcomes. To say that these endogenous processes of democratization are poorly understood is not to say that they are understudied. This field, which will be used a point of departure, may be summarized as the following genres. One genre offers a generic or ideal-type description of the process by which countries democratize. Rustow’s (1970) classic account identified four stages: (i) national unity (ii) prolonged political struggle, (iii) deliberate accords, and (iv) habituation to democratic rules (see critical discussion in Carothers 2002). A second genre suggests particular sequences as pre-conditions for successful democratization to occur: Classical liberal theory (e.g. Berlin 2002) anticipates that democracies are more likely to endure if individual rights and institutional checks and balances precede the granting of mass suffrage; Dahl (1971) proposes that democracy is more likely to survive where competition is established before participation expands; O’Donnell et al (1986) argue that the process of democratization is more likely to succeed if political liberalization leads to a bargaining between moderate actors or both sides precedes “founding” elections; Linz and Valenzuela (1994), Mainwaring (1993), and Sartori (1987) argue that the configuration of executive power and systems of party representation are important for establishing democratic stability; Berman (1997) finds egalitarian aspects to be essential to the establishment of durable democracy in Europe; and Mansfield et al (2007) argue that if elections are held prior to the establishment of an impartial state apparatus, conflict may ensue that hardens cleavages and serves as an enduring barrier to democratic consolidation. A third genre explores alternate paths to democratization. For example, Dahl (1971) proposes that there are three possible routes: (1) liberalization precedes inclusiveness, (2) inclusiveness precedes liberalization, and (3) the simultaneous achievement of these two institutional developments. A more complex typology is proposed by Linz and Stepan (1996: 57-60), who argue that there are six types of democratic transitions, each with different consequences for democratic consolidation (see also Karl 1990; Munck et al 1997.) At the level of “components of democracy”, some studies claim that there are critical institutional arenas for democratization. Howard & Roessler (2006) and Lindberg (2006) argue that repeated elections – even if not entirely free and fair – are instrumental to spur and sustain processes of democratization, while others have argued that elections can be a constituent and stabilizing component of dictatorship (Gandhi and Lust-Okar 2009). A substantial literature is devoted to civil society and its relationship to democracy in individual countries and regions (Ekiert et al 1999, Putnam 1993, Bernhard et al 2007; Howard 2003) generally viewing the mobilization of civil society as critical to the breakdown of authoritarianism (O’Donnell et al 1986, Przeworski 1991, Bernhard 1993, Bunce et al 2010), alternatively that institutions that convey the



5

interests of society to the regime may promote authoritarian stability (Gandhi 2008, Magaloni 2008). Relatedly, studies have analysed political parties as linkage institutions, connecting civil society to government and it has been argued that where parties are poorly institutionalized democratic instability is likely to arise (Mainwaring et al 1995; Roberts et al 1999; Hicken et al 2011). There is also an abundance of theorizing about this question (Mainwaring 1999; Mair 1997; Randall et al 2002). However, existing scholarship provides poor guidance on how the development of parties and party systems interacts with, in sequential terms, other critical institutions such as the judiciary and civil society. It has been argued that the failure to establish legal constraints on political leaders weakens democratic regimes (Acemoglu et al 2001, North et al 2001, Gibler et al 2011). However, since judges who carry out this review are not self-appointed and judicial orders are not self-enforcing, it is not even clear that courts with constitutional jurisdiction constrain executives in practice. In order to understand whether and when they will, we must understand both the incentives judges have for autonomous behavior and the process by which their decisions are enforced, all which may involve other critical institutions and a complex chain of changes in order to be successful (Helmke et al 2011). Despite other advances, recent scholarship on legislatures (Desposato 2006, Fish 2006) has not examined how the executive-legislative balance, judiciary, civil society, and the party system combine to affect democratization. FASDEM promises to examine the sequential relationships between the above critical institutions of democracy. Because researchers only have had highly aggregated indices of democracy to draw upon it has never been possible to test propositions about more specific relationships in a systematic and comprehensive fashion. The empirical basis for conclusions about endogenous interrelationships is generally limited to selected countries (e.g., Cappocia et al 2010). The few exceptions to this have looked at fairly obtrusive characteristics such as the timing of suffrage extension (Dix 1994), the prior performance of regimes as measured by authoritarian legacies (Linz et al 1996; Geddes 2003; Gandhi et al 2006), the age of the regime (Bernhard et al 2003), the number of government turnovers (Gasiorowski et al 1998), or the number of past regime breakdowns (Przeworski et al 2000). FASDEM will be able to test existing claims with more extensive, detailed, and better data. In addition, both the universe of quantitative studies and the vast number of contextual accounts suggest that the factors involved in democratization are likely to be multiple and historical accounts typically highlight the contingencies of democratization. This also means that the volume of existing studies is remarkably inconclusive. Scrutiny of the literature offers support for both sides of virtually every argument. There is a need to provide a rigorous academic approach to sequences of democratization, which will enable sound policy advice. FASDEM will enable delineation and testing of long series involving many variables, also capitalizing on V-Dem’s multidimensional understanding of democracy. The project offers an opportunity to test existing theories of failing and successful sequences of democratization in the most rigorous fashion possible with a new set of methodologies available from the on-going V-Dem research lead by me. Going even further beyond…… Perhaps even more significant for the understanding of democratization, unexplored and under-theorized causal chains can be investigated since the in-depth and highly differentiated V-Dem indicators capture a range of attributes of democracy for more than a century. Thus V-Dem data will make it possible to search for endogenous democratic sequences not necessarily contemplated by current theory – and do so with regards to long chains of sequential relationships between many factors. This is a form of descriptive, basic research whose importance should not be underestimated. Description has led to ground-breaking advances across many sciences, including evolutionary biology from which we borrow and adapt methodological approaches. Simply put, we do not know the answers yet to relatively simple questions like: When a country transitions from autocracy to democracy (or vice versa), which elements come first? Which are there common patterns, a finite set of sequences for sequences that are failing to lead to democracy, and those that result in democratization? With a large set of indicators measured over many years, it would become possible for the first time to explore transition sequences.1 Are there, perhaps, also different waves for each variety of democracy? FASDEM will use the V-Dem indices of seven varieties of democracy: electoral, liberal, majoritarian, 1 Sequencing is explored by Schneider and Schmitter (2004b) with a smaller set of indicators and a shorter stretch of time. See also McFaul (2005) and Møller and Skaaning (2010).



6

consensual, participatory, deliberative, and egalitarian. It is quite possible, maybe even probable, that there are varying paths, or sequences as causal chains, to each of them. The several dozen component indices and some 350 distinct indicators offer even more opportunities for sophisticated tests of existing theories of democratization in the most rigorous fashion possible, but also the possibility to discover entirely new sequences. Hence, the vast number of indicators and components collected by the V-Dem project, coupled with the new emerging methodologies proposed, afford an opportunity to disentangle the causal mechanisms of failing and successful sequences in democratization. For policy purposes this is also instrumental: Which components of democracy are most exogenous (affecting other components) and least endogenous (dependent on other components) and therefore the ideal targets for democracy promotion at different stages? The present proposal is for creating a new line of inquiry of cutting-edge research in the form of the FASDEM project to be led by me as PI for a group of scholars (some drawn from the existing V-Dem team, and some to be recruited from the outside). Objective: Revolutionizing the Field – Failing and Successful Sequences of Democratization The FASDEM project will thus seek to address the shortcomings of the existing analyses of democratization, with a clear focus on endogenous processes that we can now investigate systematically for the first time. The objective of FASDEM is to investigate the unexplored area of “endogenous” patterns and causes of democratization, to answer two questions:

1. Which are the failing versus successful sequences of democratization? 2. What are the determining causal relationships in these sequences?

I therefore propose to consolidate and expand the team that I have started to build specifically designed to tackle these two questions. If given funding for this proposal, the analyses promise to yield substantively high-value pay-offs for the knowledge of sequences in democratization with both academic and direct policy relevance.

Section b. METHODOLOGY

To accomplish the suggested aims of research in FASDEM, I suggest collaborating with not only with distinguished fellow political scientists, but also with specialists in development of statistical analysis and causal inference, as well as in evolutionary approaches and Bayesian dynamic systems modelling. I also suggest recruiting two assistant professors on tenure-track positions to add additional specialized competence to the team, as well as a two-year postdoc with advanced methodological skills. FASDEM therefore brings together not only experts in democracy and democratization with area expertise but also a set of specialized experts for analyses in this emerging field. I believe that FASDEM’s contributions in this area will be ground-breaking. Because we are charting partly unknown territories, I propose to engage fully with both two main approaches over the first two years, and then assess their relative contributions and pursue the most promising avenues further in the final three years. Project 1: Adapted Evolutionary Approaches To Depict Failing and Successful Sequences of Democracy Researchers: Staffan I. Lindberg (Prof., UGOT, 25%, 5yrs), Jan Teorell (Prof., Lund U, 15%, unsalaried, 5yrs), Ellen Lust (Prof., UGOT, 10%, unsalaried, 5yrs), Patrik Lindenfors (Assoc. Prof., Centre for the Study of Cultural Evolution, Stockholm U., 25%, 1 year), Assist. Prof. (tb. recruited at UGOT, 4yrs) , PostDoc (100% tb. recruited, 2yrs) In this project I propose to utilize methods we have borrowed from evolutionary biology, adapted and developed further for the use of V-Dem-type data and the study of democratization. Within the on-going V-Dem research program, and additional resources from my position as Wallenberg Academy Fellow, I had some limited funding available to create a subgroup of researchers and assistants to explore if it would be possible to use evolutionary methods in the study of democratization. It proved more challenging than we had originally anticipated but we have now, just at the time of writing, managed to develop a set of methodologies that works and opens up exactly the kind of avenues that the text above envisions. The methods are more fully detailed in a working paper (Lindenfors et al. 2015). The first tentative application



7

found that women’s relative equality in the area of civil liberties is close to a necessary condition for successful democratization, also presented in a working paper (Wang et al. 2015). In brief, the methods include frequency counts, a graphical representation of observed changes that can be compared to expected changes, dependency analyses adapted from evolutionary biology (Sillén-Tullberg 1993) and in some ways similar to the logic found in Qualitative Comparative Analysis (Ragin 1987, Rihoux and Ragin 2009), as well as Bayesian dynamical systems modeling (Spaiser et al. 2014; Ranganathan et al. 2014). Analyses based on the proposed approaches below can, in principle, be conducted for data measured at any level (interval, ordinal, binary) but, in practice, require ordinal or binary variables in order to be interpretable, only the dynamical systems analyses require continuous data.

A. Frequency Analysis Taking two variables at a time, we first investigate whether one of them in general tends to be larger than the other (indicating that its values are ahead of the other) in a frequency table. We combine all annual observations when the values of both variables are unchanged into one “event”. We then calculate the percentage of observations where one variable is greater than the other and where they are both of the same magnitude. For example, in Table 1.1, Variable A is higher than Variable B in 14 cases, both are equal in 13 cases, while Variable B is never larger than Variable A.

Table 1.1 Example of a frequency table of observed combinations of values of two variables. Variable A

0 1 2 3

Variable B

0 4 3 1 0 1 0 3 5 1 2 0 0 3 4 3 0 0 0 3

We can then do the same for Variable A and Variable C, for example:

Table 1.2 Another combination of variables . Variable A

0 1 2 3

Variable C

0 0 0 0 0 1 3 3 0 0 2 2 4 1 0 3 5 4 2 3

Finally, we can (with any number of variables) construct a relative frequency table to systematize the number of combinations.

Table 1.3 Example of relative frequencies table . x= B or C when A > x when A = x when A < x B 52% 48% 0% C 0% 26% 74%

Variable A is thus always larger than or equal to Variable B, Variable C is always larger than or equal to Variable A. Assuming that we have a complete universe, this would constitute strong descriptive evidence that there exists a sequence Variable C => Variable A => Variable B. B. Graphical Investigation of Changes To further investigate the exact pathways for how variables change, we use a graphical approach (e.g. Figure 1). The question of interest is again if one variable tends to be larger (or smaller) than the other after a change. Partly, the general thinking is similar – if Variable A tends to be larger than Variable B then changes ending above the diagonal will be more common than changes ending below the diagonal (Figure 1). Partly, however, the situation is different because a multitude of other potential paths are possible: variables may go in both directions; or variables may be unrelated but one variable has a skewed distribution, etc. Thus, as a second type of analysis, it is important to visually inspect the movements using a graphical method – because of this, we term the approach graphical rather than statistical.



8

The figure reveals reform paths that are more common than expected by utilizing observed data indicated by the circles and the thicknesses of the lines. Almost all “states, and the movements are below the line. C. Dependency Analysis To explore whether certain values of one variable are systematically conditional on certain values of other variables in the existing data, we have developed a method here termed dependency analysis. Inspired primarily by “the contingent states test” – an established method developed to investigate dependencies in biological evolution (Sillén-Tullberg 1993) – the method checks for absolute dependencies in the data. If and when one can establish absolute dependencies, this is evidence of necessary conditions in the data that are not contingent on inferences from regression statistics. For each value of one variable, we scan the dataset for the lowest value in the other variables. If higher values in Variable A always correspond to higher “lowest values” in Variable B, then it can be inferred that certain values of Variable A are likely to be conditional on certain values of Variable B. If, simultaneously, for each value of Variable B, the corresponding “lowest value” in Variable A is its minimum, then this shows that Variable B is not restricted by Variable A. These two observations in combination indicate a dependency between the two variables that exists only in one direction. Table 2 shows an example of such a procedure. The left table (a) indicates that higher states (2 and 3) in Variable A occur only together with higher values in Variable B (2 and 3, respectively). This means, for example, that Variable A is never observed to reach value 1, before Variable B has reached value 2. Thus, Variable B must necessarily reach value 2, before Variable A can “start moving”. The right table (b) indicates no such dependency since Variable A can be 0 at any level of Variable B. Thus, variable A is dependent on changes of Variable B having taken place at several stages, while in the opposite direction there is no such dependency. In this case, we would conclude that improvements in Variable B are a necessary condition for improvements in Variable A.

Table 2: Example of dependency tables

(a) Variable A

Lowest value of Variable B (b)

Variable B

Lowest value of Variable A

0 0 0 0 1 2 1 0 2 2 2 0 3 3 3 0

For an analysis of sequential relationships between a larger number of variables, dependency tables can be constructed for all possible combinations of variables, and be summarized. One measure is to look at the lowest value in Variable B when A attains its maximum value, here 3. In this example, A never reaches it highest state without B being at its highest value 3. An example of how several such bivariate dependency tables can be summarized is found below in Table 3. For each focal variable, we summarize these threshold values and report them as “#Necessary conditions”.

Figure 1. Graph showing the observed - expected frequencies in the Electoral component index and Alternative source information. From these results it can be inferred that the Electoral component index lags or changes simultaneously as Alternative source information.



9

Table 3. Example of combining dependency by reporting, for each variable, the number of conditions (sum of the lowest values for all other variables) required to reach highest state.

Variable # Necessary conditions % of max B 0 0% E 4 20% D 5 25% C 6 30% F 14 70% A 16 80%

In this illustration, the maximum sum of thresholds, or necessary 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. 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 reaching their highest, or close to highest, states. Together, this indicates a rough sequence that can be instructive for analysis of direct policy relevance. This example of looking at the highest states’ dependencies, is of particular interest when one is analyzing, for example, what these conditional relationships look like for achieving democratization, understood as becoming fully democratic. Then it is natural to focus on the highest states of variables. If one were, for example, interested rather in the onset of transitions, one should probably look at the number of dependencies for different variables reaching the first, or perhaps the second level on the ordinal scales, which would indicate “early moves” rather than “final push”. Table 4 below exemplifies the resulting type of aggregate summary of some 2,000 individual analyses following the dependency analysis approach outlined above, over 22 variables included in the V-Dem indices for electoral and liberal democracy. Here we present the number of necessary conditions for each of these 22 variables reaching their highest state (the top category) from our first explorative test.

Table 4. Number of conditions required to the reach highest state (Category 5)

# Necessary conditions (max = 188)*

% of max

Share with suffrage 28 15% Property rights for men 33 18% Freedom of movement for men 46 24% Property rights for women 47 25% Election voter registry 57 30% Government intimidation 60 32% Freedom of movement for women 63 34% Party ban 62 33% Barriers to parties 79 42% Opposition parties autonomy 104 55% Media self-censorship 109 58% CSO repression 121 64% CSO entry and exit 121 64% Media bias 125 66% Legislature opposition parties 124 66% EMB autonomy 134 71% Compliance with judiciary 142 76% Access to justice for men 148 79% Access to justice for women 152 81% Harassment of journalists 152 81% High court independence 158 84% Executive oversight 160 85% *For this example, we have analyzed 22 variables. The sum is 188 as some of the variables had less than 4 states.



10

One should naturally not draw any strong conclusions from small differences but we can draw pretty strong inferences about sequences from large differences. For example, several indicators of civil liberties have very few dependencies and tend to develop “in full” first. Other variables related to rule of law and oversight of the executive have very high numbers in terms of dependencies. This indicates that many other variables must reach high levels before the executive oversight, high court independence, or government media censorship can become really democratic. The table above is illustrative only but nonetheless illustrates the kind of nuanced and detailed sequences that can be established empirically with the method proposed here. D. Bayesian Nonlinear Dynamical Systems In order to study nonlinear dynamics in the interaction between variables, we also employ a newly developed Bayesian dynamical systems approach that models the probable reform direction of countries depending on state combinations. This method identifies the best nonlinear functions that capture the interactions between two or more variables. The method gives a pair of differential equations, modeling how the values in each of the two variables involved affect the direction of each. The dynamical system with the highest Bayes’ factor when allowing for up to four terms for the variables Freedom of expression (x), Alternative source information (y) and Electoral component index (z) is:

!" = −0.069! + 0.024 !! + 0.015!! + 0.032!

! ! (1)

!" = (0.06 − 0.12! + 0.055!")! !" = 0.099 − 0.44! + 0.31!! ! + 0.05!"

There is an interaction between all the variables. The rate of change of y is linear to the value of z, and x is close to it, with an added cubic term. The rate that is amplified is positive for y unless y is much larger than x, that is, alternative source information cannot remain at high values when there is little freedom of expression. The equation for dx is a bit more complicated, but is positive as long as not both x and y are large. This means that in order for freedom of expression to increase when already high, the electoral component index also needs to be high. Finally, z can grow on its own merits up to 0.28, after which the variable becomes self-limiting. Thus, for the electoral component index to grow beyond a third, both freedom of expression and alternative source information are needed. From this, we can infer which is the most likely trajectory a country will follow, given any starting point. The resulting dynamical system can be illustrated by a phase portrait, where the modeled trajectories are depicted with arrows. The method is described in detail in two papers by Spaiser et al. (2014) and Ranganathan et al. (2014b). The phase portrait in Figure 2 shows that freedom of expression is generally larger than the electoral component index and predicts a trajectory in which the former will attain values larger than the latter. However, the sequence of events is also here that the variables grow together, and that the Electoral component index grows fastest in the beginning.

Figure 2. Phase portrait showing the expected direction of change given different values of Freedom of expression and Electoral component index. Longer arrows indicate more rapid change. The figure also includes the actual trajectories of six arbitrarily chosen countries.



11

We believe the combination of the approaches above has great potential for the analysis of many types of pressing issues that social science confronts, not the least of which is the sequences of democratization. It is our goal that the combination of employing the “breakthrough” methods described above and the unique V-Dem dataset will yield entirely new and potentially revolutionizing knowledge about the dynamics of democratization. Project 2: Identifying and Testing Causal Identification of Chains in Sequences Researchers: Staffan I. Lindberg (Prof., UGOT, 25%, 5yrs), Ellen Lust (Prof. UGOT, 10%, unsalaried, 4 yrs), Adam Glynn (Assoc. Prof., Emroy U, 20% at UGOT, 3 years), Matthew Wilson (Assist. Prof., West Virginia U, 50% at UGOT, 2 yrs), Kyle Marquardt (PostDoc, UGOT, 25%, salaried via WAF-grant, 2yrs), PostDoc (100% tb. recruited, 2yrs) With the team assembled above and with the descriptive patterns laid out by Project 1, we would engage in adapting and applying a set of techniques that have rarely been used in the study of democratization. When the long, multi-variable sequences are established, the question remains to what extent such relationships are robust and which, if any, can be identified as causal. The methods we will work with in Project 2 are specifically designed to address that issue. A. Sequencing Algorithms Sequencing algorithms were first developed by sociologists to test developmental patterns (Abbott et al 2000; Elzinga 2003). Conveniently, they are also well suited to test developmental patterns within and across polities. However, the vast number of indicators makes sequence analysis challenging. To analyze any set of sequences one must first define a metric, i.e., a measure of the distance (lack of similarity) between two sequences. To analyze trajectories of countries we can draw on work on sequence analysis of family-life trajectories of individuals (Bras, Liefbroer et al 2010) but refine the distance measure to account for the range of V-Dem indicators. Throughout, we are mindful of the possibility that a number of quite different sequences may be identifiable, and that these may change over time. The quality of the V-Dem data lends itself to advanced techniques for analyzing complex political processes. One approach is to represent discrete changes in political developments in countries across time as sequences, and to evaluate the extent to which particular patterns of development support or undermine democracy. A number of studies have debated whether democratic governments are more likely to consolidate if democratic prerequisites were installed in the country in a particular pattern (Rustow 1970, Carothers 2007, Mansfield and Snyder 2007, Fukuyama 2010). However, such studies were limited in their ability to visualize and to compare sequences. In reality, many of the existing approaches to dealing with time dependence do not adequately account for the effect of order and sequence on the distribution of errors. Instead, they are better suited for uncovering mechanisms that occur within path dependent processes, in which the events in a path matter, but necessarily their order (Page 2006). Nevertheless, scholars argue that political outcomes are determined by prior events (Mahoney 2001, Yashar 1997). There are also renewed calls for scholars to distinguish among different forms of time dependence (Grzymala-Busse 2011). The notion of path dependence asserts that actors select among viable options, and with each additional choice, some outcomes become more likely than others (Levi 1997, Page 2006, Pierson 2000). Some have claimed, for example, that a country’s former experience with authoritarianism influences the success or failure of democratic transitions and democratic consolidation (Brownlee 2007, Casper 1995). By representing attributes of democracy as sequences, we can utilize tools from other disciplines to revisit outstanding debates regarding critical junctures, institutional legacies, and path dependence more generally. The sequence analysis we envision here is an analytic approach that involves calculating a measure of similarity, or distance, between pairs of sequences. The approach has its origins in biology and computer science, but has gained newfound attention in the social sciences by scholars applying the technique to categorical data. In sociology, sequencing algorithms have been used to identify patterns of employment and life changes among cohort members (Assave et al. 2007; Blanchard 2005; Scherer 2001). Wilson (2014) – who will contribute to FASDEM if funded – provides an overview of the method’s applicability to study regime change in political science. The core lies in the algorithms used for distance comparison. There are several common metrics by which to compute sequence similarity, of which optimal matching is the most common (Gabadinho et al. 2010). Optimal matching generates edit distances that are the minimal cost, in terms of insertions, deletions and



12

substitutions, for transforming one sequence into another. This edit distance has first been proposed by Levenshtein (1966) and has been popularized in the social sciences by Abbott (Abbott and Forrest, 1986; Abbott, 2001). The optimal matching algorithm compares two sequences state by state and determines the set of operations that would align them at the lowest cost. The total number of most efficient operations by which to make two sequences match perfectly signifies their dissimilarity, or the distance between them. Pairwise comparison produces an n x n matrix that is symmetric around a diagonal of zeros, representing the comparison of each sequence with every other sequence, as well as itself. Among the many applications of the distances calculated by the OM algorithm, clustering methods can be used to aggregate the sequences into a reduced number of groups. Cluster analysis is aimed at organizing the data into subsets, or "clusters", such that those within each cluster are more closely related to one another than objects assigned to different clusters. As nominal level data, the clusters created through hierarchical clustering can be further analyzed through sub-sequence analysis or can be compared using standard measures of association. An example of this can be found in Casper and Wilson (2015), who apply the method to explain bargaining between actors during national crises. Another application of the method is to create theoretical sequences—for example, what a “successful” pattern of democratization might look like—and to calculate its similarity to observed patterns of development in each country. There are also a number of alternative uses of sequence information that have not yet been explored. More immediately, however, techniques for visualizing and tabulating sequences that are common to Sequence Analysis are capable of greatly enhancing inferences and guiding the specification of empirical models of democratization and democratic breakdown. In using V-Dem indicators to analyze changes in patterns of democracy over time, the proposed collaboration will expand and build upon current practices in sequence analysis using algorithms. As an example, in Castling the King: Institutional Sequencing and Regime Change, Wilson (2015) developed a solution for comparing sequences in a sample with censored data. Future developments to evaluate sequences of V-Dem data will also expand on existing methods by developing new ways of visualizing sequence dissimilarity. During the course of the project, changes to current practices and assumptions associated with the method can be made to make this type of sequence analysis more amenable to political science research. B. Dynamic Treatment Regimes The previously discussed methods allow us to suggest solutions to the sequencing problem, but they do not directly address the policy sequencing question: what policies should be adopted, in what order, and when. To directly address this sequencing question, we will utilize the literature on optimal dynamic treatment regimes (see Murphy (2003), with discussion). To fix ideas, consider a simplified version of the aforementioned motivating question: should policy makers focus first on empowering the media or first on making elections free and fair. The problem from the policymaker’s stand point is that while the V-Dem data provide credible measurements of media empowerment and election quality, there is no policy that directly empowers the media or directly makes elections free and fair. These particular V-Dem variables measure the de facto situation, but there are also V-Dem variables that measure de jure policies, such as government censorship and EMB autonomy, that represent potential policy levers. In order to precisely examine the sequencing question, we must consider dynamic treatment regimes involving both the de jure and de facto variables. For example, we might compare the following three regimes: 1) simultaneously eliminate government censorship and establish EMB autonomy; 2) eliminate government censorship, once this results in a sufficient level of media empowerment, establish EMB autonomy; 3) establish EMB autonomy, once this results in a sufficient level of free and fair elections, eliminate government censorship. Note the dynamic interplay between the de facto and the de jure variables in the second and third treatment regimes. This dynamic interplay represents the core of the sequencing question, but it complicates attempts to find the optimal regime. Fortunately, the literature on optimal dynamic regimes (cited above) is now well developed, and we plan to use these approaches to address this important question. C. Vector AutoRegression Vector autoregression, VAR (Sims 1980), is the canonical econometric approach to simultaneously modelling the dynamics of multiple dependent economic time series. For example, economists use VAR to model inflation, unemployment, and real interest rates. VAR models provide a framework for principled description of multiple time series and forecasting. With additional strong assumptions, they provide a



13

framework for causal inference. The V-Dem data consist of multiple dependent indicators of democracy over time and modelling the dynamics of these indicators will address the aforementioned theoretical questions in the democratization literature. VAR can also be extended to accommodate multiple countries in a single model (e.g., see Kalaitzidakis et al 2000). Together these methodological approaches would present the study of democratization with a range of new ways to produce ground breaking results identifying causality with observational data, in the links of the sequence-chains. Relevance The academic importance of developing and testing adequate methodologies for moving us forward with the state of knowledge regarding the details and complexities of sequences of endogenous causal chains of democratization is indisputable. For over half a century social scientists have struggled with inadequate data sets, computational and statistical limitations. In addition, learning of methodological approaches from the natural sciences to social science has been limited. We are now poised to overcome these impediments. Now over 60 years old, the field of democratization studies remains fraught with inconsistent and partial results and in need of a systematic and comprehensive approach using the new data and tools available and detailed here. FASDEM promises to transform our knowledge of democracy. Finally, the direct policy relevance can hardly be overstated. Individual countries and multilateral organizations spend billions of Euros every year in support of democracy, human rights, and good governance. Meanwhile, we know very little about if they are designed to generate the best effects possible in each case: We cannot say whether to first invest in strengthening the media, or try to help build stronger political parties, or if actors should contribute to empowering the judiciary, or something else. Case-experts, comparativists, and practitioners naturally have intuitions about this and they may be right or wrong. FASDEM will, at the minimum, provide a set of answers, unprecedented in their basis on rigorous and systematic science. Knowing more about how to expand democracy is one of the most important tasks we have. If funded for FASDEM, I also plan to apply for funding to advance this research agenda further both in terms of methods-developments and substantial areas of research. The latter could involve such as sequences leading to civil war, better governance in terms of health improvements, and improving the qualities of democracies beyond the transitions. This would also contribute to further expanding the V-Dem Institute at UGOT further making it move towards the long-term vision of building a “CERN for Democracy”: a world-leading institute for measurement and data collection as well as research on democracy and autocracy across time and space.

Section c. RESOURCES (INCLUDING PROJECT COSTS)

I, Staffan I. Lindberg was selected a Wallenberg Academy Fellow and a member of the Young Academy, Royal Academy of Sweden for terms 2014-2019, and is one of the PI’s for V-Dem and the Director of the V-Dem Institute. These distinctions and positions come with collaborating with a network of distinguished scholars across all fields in a range of academic activities, and resources to support FASDEM should it be funded. The project will be hosted by the V-Dem Institute, Department of Political Science, University of Gothenburg. The V-Dem Institute was established in 2012 by me and received a six-year “center of excellence” program grant from the Bank of Sweden Tercentenary Fund in 2013. In 2014, we also won one of the most exclusive grants from the Swedish Research Council: Recruitment of a scholar tenurable as Full Professor at a Top-10 university: a €15.5 million grant to recruit Ellen Lust from Yale University. That grant includes three postdocs per year for ten years, and eight fully funded PhD students mainly to establish her “Governance and Local Development” program, but part of this extra personnel can be allocated to support FASDEM should this application be approved. The V-Dem Institute with its two fulltime project managers, three fulltime postdocs, two fulltime analysts, and the data manager working under my leadership, would be an infrastructure that the FASDEM project can benefit from by not having to build a organizational structure. The Department of Political Science at University of Gothenburg is one of the most foremost political



14

science departments in the Nordic countries. A recent evaluation conducted by a large team of internationally recognized scholars rated our department excellent to outstanding across all areas. The Swedish Research Council’s evaluation in 2002 already recognized the department as best in the country. It consists of some 60+ faculty, 20 postdocs and researchers, and additionally 25 PhD students providing a rich environment for FASDEM to flourish in. Budgeted Costs: The applicant and PI will spend 50% of fulltime on the project over 5 years. As Professor of Political Science at University of Gothenburg he has 40% of fulltime (FTE) paid for by the Department for research activities. Half of that time will be spent on continuing managing the V-Dem data-collection and regular research project. To spend 50% of my time on FASDEM, I will only have to be funded at 30% FTE by FASDEM, over the 60 months. He will spend the remaining 30% of his FTE on teaching (10%) and leading other research projects (20%) in the regular V-Dem Research Program funded by Riksbankens Jubileumsfond and the Wallenberg Academy Fellowship program. Professors Lust and Teorell will participate at 20 and 10% of FTE respectively using already paid-for time. Assoc. Professor Lindenfors is budgeted for 25% of FTE for the first year, where after his contribution should be done. Professor Glynn is budgeted for working 20% of FTE over three years anticipating that causal identification will have to stretch out over a longer period. PhD (2015, Winsconsin U) Kyle Marquardt is a specialist in Bayesian statistics currently at the V-Dem Institute and will be financed out of Lindberg’s WAF-grant and the costs are therefore not in the budget below. The budget includes funds to recruit a specialized Assistant Professors (tenure-track) and two PostDoc’s to buttress the necessary manpower to each of the two parts of the project. They will be placed in the V-Dem Institute, UGOT. In addition, the project will include two PhD students but they will be funded by UGOT and the Lust/VR grant and costs are therefore not included in the budget below. The budget also includes a research assistant for the PI over four years of the program. Equipment includes funds to provide high-performance computers for the new hires. Travel includes funds for presenting the results at academic conferences and public meetings (5/year year). Publications include fees and related costs to make the findings available not only to academics but also for the policy community with particular focus on European ministries of foreign affairs, donor agencies, and international NGOs in the area of democracy-support. Audit costs are standard fees for University of Gothenburg. Finally, the budget has a line under C1 for subcontracting of programming a web-based module for online analysis following the new methods in Project 1. This is to make these potentially revolutionizing new methods freely available as a public good for both academics without the requisite modelling skills, and for students as well as practitioners across the EU. This will be subcontracted to a well-referenced programming facility. In addition to the budgeted costs above and the co-funding indicated in the text, UGOT commits to add 25% to the total sum, or €484 660, to FASDEM as co-funding should it be approved and funded at the requested level.



16

References Abbott, A. and J. Forrest 1986. “Optimal Matching Methods for Historical Sequences.” Abbott, Andrew, Angela Tsay. 2000. “Sequence Analysis and Optimal Matching Methods in Sociology.”

Sociological Methods and Research 29(1): 3–33. Abbott, Andrew. 2001. Time Matters: On Theory and Methods. Chicago: Chicago Press. Acemoglu, D et al 2005 Economic Origins of Dictatorship and Democracy. Cambridge UP Acemoglu, D et al 2009 “Reevaluating the Modernization Hypothesis.” J of Monet Econ 56(8) Acemoglu, Daron and James A. Robinson. 2001. “A Theory of Political Transitions.” American Economic

Review 91(4): 938-963. Acemoglu, Daron and James A. Robinson. 2006. Political Origins of Dictatorship and Democracy. New

York: Cambridge University Press. Alvarez, Michael, Jose A. Cheibub, Fernando Limongi, Adam Przeworski. 1996. “Classifying Political

Regimes.” Studies in Comparative International Development 31(2): 3–36. Bächtiger, André, ed. 2005. Special issue: “Empirical Approaches to Deliberative Democracy.” Acta Politica

40(1–2). Berlin, Isaiah. 2002. Liberty. Oxford: Oxford University Press. Beetham, David, ed. 1994. Defining and Measuring Democracy. London: Sage. Berman, S 1997 “Civil Society and the Collapse of the Weimar Republic” World Pol 49 Bernhard, M 1993 “Civil Society and Democratic Transition in East Central Europe” Pol Sc Quart 108 Bernhard, Michael, Christopher Reenock, and Timothy Nordstrom. 2004. “The Legacy of Western Overseas

Colonialism on Democratic Survival.” International Studies Quarterly 48(3), 225-250. Bernhard, Michael, Timothy Nordstrom, Christopher Reenock. 2001. “Economic Performance, Institutional

Intermediation, and Democratic Survival.” Journal of Politics 63(3): 775–803. Bertelsmann Stiftung. [various years]. Bertelsmann Transformation Index [2003–]: Towards Democracy and

a Market Economy. Washington, DC: Brookings Institution Press. Blanchard, Philippe. 2005. “Multi-dimensional biographies. Explaining disengagement through sequence

analysis”, Congrès de l’European consortium for political research (Ecpr), Budapest, 2005 Boix, Carles, Michael Miller, Sebastian Rosato. 2013. “A Complete Dataset of Political Regimes, 1800-

2007.” Comparative Political Studies 46(12), 1523-1554. Boix, Carles. 2003. Democracy and Redistribution. New York: Cambridge University Press. Bollen, K.A. 1983. "World System Position, Dependency and Political Democracy." American Sociological

Review 48:468-79. Bollen, K.A. and Robert Jackman. 1985. "Regression Diagnostics: An Expository Treatment of Outliers and

Influential Cases." Sociological Methods and Research 13:510-42. Bollen, Kenneth A. 1993. “Liberal Democracy: Validity and Method Factors in Cross-National Measures.”

American Journal of Political Science 37(4): 1207–1230. Bollen, Kenneth A. et al 2000 “Subjective Measures of Liberal Democracy.” Comp Pol St 33(1) Bollen, Kenneth A., Pamela Paxton. 2000. “Subjective Measures of Liberal Democracy.” Comparative

Political Studies 33(1): 58–86. Bowman, Kirk, Fabrice Lehoucq, James Mahoney. 2005. “Measuring Political Democracy: Case Expertise,

Data Adequacy, and Central America.” Comparative Political Studies 38(8): 939–970. Bras, H. et al 2010 “Standardization of Pathways to Adulthood?” Demography 47(4) Brownlee, Jason. 2007. Authoritarianism in an Age of Democratization. Cambridge: Cambridge Bueno de Mesquita, Bruce; Alastair Smith; Randolph M. Siverson; James D. Morrow. 2003. The Logic of

Political Survival. Cambridge: MIT Press. Bühlmann, Marc, Wolfgang Merkel, Lisa Müller, Bernhard Weßels. 2012. “The Democracy Barometer: A

New Instrument to Measure the Quality of Democracy and Its Potential for Comparative Research.” European Political Science 11(4): 519-536.

Bunce, V 2010 Defeating Authoritarian Leaders in Postcommunist Countries. Cambridge UP Burkhart, RE et al 1994 “Comparative Democracy” Am Pol Sci Rev 88 Capoccia, Giovanni, Daniel Ziblatt. 2010. “The Historical Turn in Democratization Studies: A New



17

Research Agenda for Europe and Beyond.” Comparative Political Studies 43(8-9): 931-968. Carothers, Thomas. 2007. The “Sequencing” Fallacy. Journal of Democracy, 18:13-27. Casper, Gretchen, and Matthew Wilson. 2015. “Using Sequences to Model Crises.” Political Science

Research and Methods 3(02): 381–97. Casper, Gretchen. 1995. Fragile Democracies: The Legacies of Authoritarian Rule. Pittsburgh: The Central America. Baltimore: John Hopkins University Press. Cheibub, Jose Antonio, Jennifer Gandhi, James Raymond Vreeland. 2010. “Democracy and Dictatorship

Revisited.” Public Choice 143(1–2): 67–101. Clinton, Joshua D., David Lewis. 2008. “Expert Opinion, Agency Characteristics, and Agency Preferences.”

Political Analysis 16(1): 3–20. Clinton, Joshua D., John S. Lapinski. 2006. “Measuring Legislative Accomplishment, 1877-1994.” American

Journal of Political Science 50(1): 232–249. Collier, David, Steven Levitsky. 1997. “Democracy with Adjectives: Conceptual Innovation in Comparative

Research.” World Politics 49(3): 430–451. Coppedge, M 2012 Approaching Democracy. Cambridge UP Coppedge, Michael, Angel Alvarez, Claudia Maldonado. 2008. “Two Persistent Dimensions of Democracy:

Contestation and Inclusiveness.” Journal of Politics 70(3): 335–350. Coppedge, Michael, John Gerring with David Altman, Michael Bernhard, Steven Fish, Allen Hicken,

Matthew Kroenig, Staffan I. Lindberg, Kelly McMann, Pamela Paxton, Holli A. Semetko, Svend-Erik Skaaning, Jeffrey Staton, Jan Teorell. 2011. “Conceptualizing and Measuring Democracy: A New Approach.” Perspectives on Politics 9(1): 247–267.

Coppedge, Michael, John Gerring, Staffan I. Lindberg, Svend-Erik Skaaning, and Jan Teorell. 2015a. V-Dem Comparisons and Contrasts with Other Measurement Projects. Varieties of Democracy (V-Dem) Project.

Coppedge, Michael, John Gerring, Staffan I. Lindberg, Svend-Erik Skaaning, Jan Teorell, David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Carl Henrik Knutsen, Kyle Marquardt, Kelly McMann, Farhad Miri, Pamela Paxton, Daniel Pemstein, Jeffrey Staton, Eitan Tzelgov, Yi-ting Wang, and Brigitte Zimmerman. 2015b. V-Dem [Country-Year/Country-Date] Dataset v5. Varieties of Democracy (V-Dem) Project

Coppedge, Michael, John Gerring, Staffan I. Lindberg, Svend-Erik Skaaning, Jan Teorell, with David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Carl Henrik Knutsen, Kelly McMann, Pamela Paxton, Daniel Pemstein, Jeffrey Staton, Brigitte Zimmerman, Frida Andersson , Valeriya Mechkova, and Farhad Miri. 2015c. V-Dem Codebook v5. Varieties of Democracy (V-Dem) Project.

Coppedge, Michael, John Gerring, Staffan I. Lindberg, Svend-Erik Skaaning, Jan Teorell, Frida Andersson, Kyle Marquardt, Valeriya Mechkova, Farhad Miri, Daniel Pemstein, Josefine Pernes, Natalia Stepanova, Eitan Tzelgov, and Yi-ting Wang. 2015d. V-Dem Methodology v5. Varieties of Democracy (V-Dem) Project.

Coppedge, Michael, Wolfgang H. Reinicke. 1990. “Measuring Polyarchy.” Studies in Comparative International Development 25(1): 51–72.

Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University Press. Democratization.” In Advances in Sequence Analysis: Methods, Theories and Applications, Desposado, Scott. 2006. “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of

Deputies.” American Journal of Political Science 50, 1: 62-80. Diamond, L et al 1990 Democracy in Developing Countries, Lynne Rienner Dissertation, Penn State University. Dix, RH 1994 “History and Democracy Revisited” Comp Pol 27(1) Ekiert, Grzegorz and Jan Kubik. 1999. Rebellious Civil Society. Ann Arbor: University of Michigan Press. Elzinga, C.H. (2003) Sequence Similarity - A Non-Aligning Technique, Sociological Methods & Research,

31(4), 3-29. Fish, MS 2006 “Stronger Legislatures, Stronger Democracies” J of Demo 17(1) Fukuyama, Francis. 2011. “Is There a Proper Sequence in Democratic Transitions?" Current History Gabadinho, Alexis, Gilbert Ritschard, Matthias Studer, and Nicolas S. Muller. 2010. “Mining sequence data

in R with the TraMineR package: A user’s guide.” <http://mephisto.unige.ch/traminer/>.



18

Gandhi, J 2010 Political Institutions Under Dictatorship Cambridge UP Gandhi, J et al 2006 “Cooperation, Cooptation, and Rebellion under Dictatorships” Econ & Pol 18: 1–26 Gandhi, J et al 2009 “Elections under Authoritarianism” An Rev o Pol Sci 12: 403-22 Gasiorowski, MJ et al 1998. "The Structural Determinants of Democratic Consolidation" Comp Pol St 31(6) Geddes, B 1999 "What Do We Know about Democratization after Twenty Years?" An Rev o Pol Sci 2: 115-

44. Geddes, Barbara. 2003. Paradigms and Sand Castles: Theory Building and Research Design in Comparative

Politics. Ann Arbor, MI: University of Michigan Press. Gibler, D et al 2011 "Testing the Effects of Independent Judiciaries on the Likelihood of Democratic

Backsliding" Am J of Pol Sc 55(3) Glaeser, Edward L., Giacomo Ponzetto, Andrei Shleifer. 2007. “Why Does Democracy Need Education?”

Journal of Economic Growth 12(2): 77-99. Grzymala-Busse, Anna. 2011. “Time Will Tell? Temporality and the Analysis of Causal Guatemala, 1870-1950s. Stanford: Stanford University Press. Helmke, G et al 2011 “The Puzzling Judicial Politics of Latin America” Courts in Latin America, ed. G

Helmke & J Rios-Figueroa. Cambridge UP Hicken, A et al 2011 "Shadows from the Past" Comp Pol St 44(5) Houle, Christian. 2009. “Inequality and democracy: Why inequality harms consolidation but does not affect

democratization.” World Politics 61(4): 589-622. Howard, M et al 2006 “Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes.” Am J of

Pol Sc 50(2) Howard, MM 2003 The Weakness of Civil Society in Post-Communist Europe. Cambridge UP Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman, OK:

University of Oklahoma Press. Inglehart, Ronald and Welzel, Christian. 2005. Modernization, Cultural Change and Democracy: The

Human Development Sequence. Cambridge: Cambridge University Press. Jackman, Simon. 2004. “What Do We Learn from Graduate Admissions Committees? A Multiple Rater,

Latent Variable Model, with Incomplete Discrete and Continuous Indicators.” Political Analysis 12 (4): 400–424.

Johnson, Valen E., James H. Albert. 1999. Ordinal Data Modeling. New York: Springer. Journal of Interdisciplinary History, 16:471-494. Junker, Brian 1999. Some Statistical Models and Computational Methods that may be Useful for

Cognitively-Relevant Assessment. http://www.stat.cmu.edu/~brian/nrc/cfa/documents/final.pdf Karl, TL 1990 “Dilemmas of Democratization in Latin America” Comp Pol 1-21 Kalaitzidakis, Pantelis, & Mamuneas, T.P. & Savvides, A. & Stengos, T. 2001. " Measures of Human

Capital and Nonlinearities in Economic Growth," Journal of Economic Growth, Springer, vol. 6(3), pages 229-54

Kaufmann, Daniel, Aart Kraay, Massimo Mastruzzi. 2010. The Worldwide Governance Indicators: Methodology and Analytical Issues. World Bank Policy Research Working Paper No. 5430.

Levenshtein, V. 1966. “Binary Codes Capable of Correcting Deletions, Insertions, and Reversals.” Levi, Maragaret. 1997. Consent, Dissent, and Patriotism. New York: Cambridge University Press. Levitsky, S et al 2002 “The Rise of Competitive Authoritarianism” J o Demo 13(2): 51-66. Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six

Countries. New Haven: Yale University Press. Lindberg, S ed 2009 Democratization by Elections. Johns Hopkins UP Lindberg, Staffan I. 2006. Democracy and Elections in Africa. Baltimore: Johns Hopkins University Press. Lindenfors, Patrik, Fredrik Jansson, Yi-ting Wang, and Staffan I. Lindberg. 2015. “Investigating Sequences

in Ordinal Data: A New Approach with Adapted Evolutionary Models* University of Gothenburg, Varieties of Democracy Institute: V-Dem Working Papers Series No. 18.

Linz, J. et al 1996 Problems of Democratic Transition and Consolidation. Johns Hopkins UP Linz, JJ et al eds 1994 The Failure of Presidential Democracy. Johns Hopkins UP



19

Lipset, SM 1959 "Some Social Requisites of Democracy" Am Pol Sc Rev 53(1): 69-105 Lord, Frederic M., and Melvin Novick. 1968. Statistical Theories of Mental Test Scores. Reading, MA:

Addison-Wesley. Magaloni, B 2008 “Credible Power-Sharing and the Longevity of Authoritarian Rule” Comp Pol St 41 Mainwaring, S 1993 “Presidentialism, Multipartism, and Democracy” Comp Pol St 26(2): 198–228 Mainwaring, S et al 1995 Building Democratic Institutions. Stanford UP Mair, Peter. 1997. Party System Change: Approaches and Interpretations. Oxford: Clarendon Press. Mansfield, E et al “Exchange: The Sequencing ‘Fallacy,’” J of Demo 18:7 Mansfield, Edward D. and Jack Snyder. 2007. The Sequencing “Fallacy.” Journal of Democracy, Marshall, Monty G., Ted Gurr, Keith Jaggers. 2014. Polity IV Project: Political Regime Characteristics and

Transitions, 1800–2006. (http://home.bi.no/a0110709/PolityIV_manual.pdf), accessed April 9, 2015. McFaul, Michael. 2005. “Transitions from Postcommunism.” Journal of Democracy 16(3): 5–19. Møller, Jørgen, Svend-Erik Skaaning. 2010. “Marshall Revisited: The Sequence of Citizenship Rights in the

Twenty-First Century.” Government and Opposition 45(4): 457-483. Munck, G et al 1997 “Modes of Transition and Democratization” Comp Pol, 343-362. Munck, Gerardo L. 2009. Measuring Democracy: A Bridge between Scholarship and Politics. Baltimore:

John Hopkins University Press. Munck, Gerardo L., Jay Verkuilen. 2002. “Conceptualizing and Measuring Democracy: Alternative Indices.”

Comparative Political Studies 35(1): 5–34. Murphy, Susan A. 2003. “Optimal Dynamic Treatment Regimes” Journal of the Royal Statistical Society –

Series B (Statistical Methodology), Vol 65 (2): 331-355. Norris. P 2008 Driving Democracy. Cambridge UP North, D et al 2000 Order, Disorder, and Economic Change. Yale UP O’Donnell, G et al 1986 Transitions from Authoritarian Rule. Johns Hopkins UP Page, Scott E. 2006. “Path Dependence”. Quarterly Journal of Political Science, 1: 87-115. Patz, Richard J., and Brian W. Junker. 1999. “A Straightforward Approach to Markov Chain Monte Carlo

Methods for Item Response Models.” Journal of Educational and Behavioral Statistics 24: 146-178. Paxton, Pamela. 2000. “Women’s Suffrage in the Measurement of Democracy: Problems of

Operationalization.” Studies in Comparative International Development 35(3): 92–110. Pemstein, Daniel, Stephen Meserve, James Melton. 2010. “Democratic Compromise: A Latent Variable

Analysis of Ten Measures of Regime Type.” Political Analysis 18(4): 426–449. Pierson, Paul. 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton: Princeton

University Press. Przeworski, A 1991 Democracy and the Market. Cambridge UP Przeworski, A et al 2000 Democracy and Development. Cambridge UP Przeworski, Adam, Fernando Limongi. 1997. “Modernization: Theories and Facts.” World Politics 49(1)

155-183. Putnam, R 1993 Making Democracy Work. Princeton UP Ragin, C. C. (1987). The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies.

University of California Press. Randall, V 2002 “Party Institutionalization in New Democracies.” Party Pol 8(1) Ranganathan, S., Spaiser, V., Mann, R. P., & Sumpter, D. J. (2014). “Bayesian Dynamical Systems

Modelling in the Social Sciences”. PloS one, 9(1), e86468. Rihoux, B., & Ragin, C. C. (Eds.). (2009). Configurational comparative methods: Qualitative comparative

analysis (QCA) and related techniques. Sage. Roberts, K et al 1999 “Party Systems and Electoral Volatility in Latin America” Am Pol Sc Rev 93(3) Rustow, Dankwart. 1970. Transitions to Democracy. Comparative Politics, 2:337-63. Sartori, G 1987 The Theory of Democracy Revisited. Chatham House. Schedler, Andreas. 2013. The Politics of Uncertainty: Sustaining and Subverting Electoral Authoritarianism.

Oxford: Oxford University Press.



20

Scherer, Stefani. 2001. “Early Career Patterns: A Comparison of Great Britain and West Germany.” Schneider, Carsten Q., Philippe C. Schmitter. 2004b. “Liberalization, Transition and Consolidation:

Measuring the Components of Democratization.” Democratization 11(5): 59–90. Seawright, Jason. 2010. “Regression-Based Inference: A Case Study in Failed Causal Assessment.” In Henry

E. Brady and David Collier, eds. Rethinking Social Inquiry: Diverse Tools, Shared Standards, 2d ed. (Lanham, MD: Rowman & Littlefield), 247-71.

Sillen-Tullberg, Birgitta. "The effect of biased inclusion of taxa on the correlation between discrete characters in phylogenetic trees." Evolution (1993): 1182-1191.

Simpser, Alberto. 2013. Why Governments and Parties Manipulate Elections: Theory, Practice, and Implications. Cambridge: Cambridge University Press

Skaaning, Svend-Erik, John Gerring, Henrikas Bartusevičius. 2015. “A Lexical Index of Electoral Democracy.” Comparative Political Studies 48(12): 1491-1525.

Soviet Physics Doklady, 10: 707-710. Spaiser, V., Ranganathan, S., Mann, R. P., & Sumpter, D. J. (2014). “The Dynamics of Democracy,

Development and Cultural Values”. PloS one, 9(6), e97856 Stokes, Susan C., Thad Dunning, Marcelo Nazareno, Valeria Brusco. 2013. Brokers, Voters, and

Clientelism: The Puzzle of Distributive Politics. New York: Cambridge University Press. Teorell, J 2010 Determinants of Democratization. Cambridge UP Treier, Shawn, and Simon Jackman. 2008. “Democracy as a Latent Variable.” American Journal of Political

Science 52(1): 201–217. Vanhanen, Tatu. 2000. “A New Dataset for Measuring Democracy, 1810–1998.” Journal of Peace Research

37(2): 251–265. Wang, Yi-ting, Patrik Lindenfors, Aksel Sundström, Fredrik Jansson, and Staffan I. Lindberg. 2015. “No

Democratic Transition Without Women’s Rights: A Global Sequence Analysis 1900–2012”. Varieties of Democracy Institute: Working Paper Series, No. 12.

Weitz-Shapiro, Rebecca. 2013. Curbing Clientelism in Argentina: Politics, Poverty, and Social Policy. Cambridge: Cambridge University Press.

Wilson, Matthew Charles. 2014. “Governance Built Step-by-Step: Analysing Sequences to Explain Democratization.” In Advances in Sequence Analysis: Theory, Method, Applications. eds. Philippe Blanchard, Felix Bühlmann, and Jacques-Antoine Gauthier. New York: Springer International Publishing, 213–27.

Wilson, Matthew. 2015. “Castling the King: Institutional Sequencing and Regime Change.” Ph.D. Penn State University: Department of Political Science


ERC Consolidator Grant 2016 - Research proposal [Part B2] project.pdf · on-going “standard” (inquiring into questions of democratization using standard approaches) V-Dem research

Documents