Contestation and participation: Operationalizing and mapping democratic models for 80 electoral democracies, 1990–2009

Abstract

In recent decades, some attempts have been made to empirically assess different models of democracy in action by identifying and measuring various indicators. Lijphart’s seminal work, Patterns of Democracy, is one of the most comprehensive attempts to categorize models of democracy systematically. However, it has mainly focused on established democracies. There is no extensive operationalization of models of democracy which evaluates a broader range of countries, including new democracies in a systematic way. This article aims to fill the gap by operationalizing two persistent dimensions of democracy – contestation and participation – for 80 electoral democracies around the world between 1990 and 2009. Extracting empirical scores, we present a two-dimensional global map of democracy over two decades. We finally examine the associations between democratic models and indices of level and performance of democracy.

Appendices

Appendix A

Classification of democratic regimes

The classification of regime types by Cheibub (2007) is based on the formal constitutions and not necessarily on the real practice of democracy. Cheibub et al (2010, p. 82) argued that ‘in almost every instance where the formal rules do not seem to match practice at a first glance, we find examples of behavior that conform to the constitutional prerogatives of the president and/or the assembly’. We believe, however, that finding exceptional instances cannot overshadow the real and regular practices in many countries.

If not the constitution but the practice of democracy is the main focus of attention, as it is here, things will look somewhat differently. For instance, we are not convinced that the Austrian regime should be viewed as semi-presidential, rather than semi-parliamentary; it has a popularly-elected president, but this president has a ceremonial role only. Given this argumentation and considering the different definitions and classifications of semi-presidentialism in the literature (Duverger, 1980; Siaroff, 2003; Elgie, 2005), we reformulated and rearranged the criteria by Cheibub (2007, p. 35) and identified four types of political regime, as illustrated in Figure A1.

Figure A1

The criteria for classifying the democratic regimes in practice.

Countries in which the head of state is not popularly elected (‘No’ to the first question) are categorized under the ‘parliamentary’ regime here. Countries having an elected president with a ceremonial role in practice (No to the second question) are classified as ‘semi-parliamentary’ regimes. For answering the second question and assessing whether or not a president has ‘considerable power’, we use the ‘Presidential Power’ score introduced by Siaroff (2003), which extended for some more countries by Elgie and McMenamin (2008) and Qi (2010). If the score is lower than 5, then the system is semi-parliamentary (for example, Lithuania, Portugal, Croatia after 2000); otherwise the regime is semi-presidential (for example, France, Romania, Croatia before 2000) or full presidential (for example, United States, Latin America, South Korea) depending on the answer to the third question in Figure A1, regarding whether or not an executive branch responsible to the legislative (based on Siaroff (2003)).

According to our classification, Switzerland that has no popular elected executive, and Austria that has a ceremonial elected president are respectively considered parliamentary and semi-parliamentary democracies, while according to Cheibub (2007) they are presidential and semi-presidential regimes respectively. In Table C1 in the Appendix C, the regime type of each country, based on our formulation, is specified. Except for Croatia, Finland, Moldova and Poland, all democratic countries had the same regime type for the whole range of 1990–2009.

Figure A1

Appendix B

Varieties of referendums and the weighting rationale

Two main categories of referendums can be distinguished: top-down and bottom-up.

• Plebiscitary and mandatory (or compulsory) referendums are two types of top-down referendums. A plebiscite is defined here as a referendum called by the authorities, either the executive or legislative. A mandatory referendum is required ‘under circumstances defined in the constitution or in legislation’ (IDEA, 2008, p. 213).

• Optional (or abrogative or facultative or citizen-demanded) referendums and initiatives are two types of bottom-up referendums. The former is ‘initiated, or triggered, by a number of citizens referring to existing laws or political or legislative proposals’ (IDEA, 2008, p. 61). Initiatives are political proposals (for example, draft legislation) initiated by citizens. The proposal might be directly used for popular voting (direct initiative) or it could be revised by the legislative and then presented to the popular vote (indirect initiative which is also called counterproposal (Kaufmann et al, 2010)).

There is considerable inconsistency in the terminology of different types of referendums in the literature. For instance, in IDEA’s terminology, the ‘optional referendum’ is used in the meaning of plebiscite while in the terminology of Center for Research on Direct Democracy (C2D, 2012) it resembles the citizen-demanded referendum. In the terminology of IRI (Kaufmann et al, 2010), the term ‘optional referendum’ is not common and ‘citizen-initiated referendum’ is used instead. This could result in problematical operationalization

Some scholars distinguish and weigh referendums based on the main categories only (that is, top-down or bottom-up) (Peterlevitz, 2011). Some studies use only the number of referendums, no matter their types (Vanhanen, 2002; Altman, 2011; Kriesi and Bochsler, 2012). Another approach is to give a different weight to each type of referendum (Vatter, 2009; Altman, 2013). We follow this approach and weigh each type of referendum distinctively, based on its importance and contribution to direct democracy. It can be argued convincingly that bottom-up referendums with the citizens as initiators should have a higher weight than the top-down ones. Initiatives are seemed more participatory than optional (abrogative) referendums because in initiatives citizens themselves define, directly or indirectly, the issue and subject of the referendum. Plebiscites have more ‘governmental control’ than mandatory referendums (Setala, 2006; Vatter, 2009), which gives the latter higher participatory value than the former (see Table B1 for weights).

Table B1 Weight of different types, levels and effect of referendums

Normally only binding referendums are taken into account in the indices (Vatter, 2009; Peterlevitz, 2011). However, we believe that the non-binding referendum presents a form of participation that cannot be ignored. Setala (2006) asserted that in west-European democracies parliament binds itself to the result of advisory (non-binding) plebiscites. Therefore, we assign a weight for these referendums as well, albeit half the weight of a binding one. According to C2D (2012) database, all non-binding referendums have been plebiscites. However, we understood that there were also non-binding initiatives in some countries (for example, the Netherlands) (Van der Krieken, 2011).

Venhanen (2002), in his revised index of participation, includes the number of referendums as a supplement to voter turnout, assuming 5 points for each national referendum and 1 point for a local referendum. It seems an acceptable assumption, taking into account observation that in Switzerland – a country with a full-developed referendum culture – national referendums are five times as scarce as local ones. This ratio is reasonable if the practicalities (time, cost, facilities) in national and local referendums would also be taken into account. Moreover, the number of participants in a national referendum is also multifold than a local referendum. Thus a weighting factor of 0.20 for local referendums seems acceptable.

Table B1 summarizes the weighting coefficients for different type, level and effect of referendums.

As an example, Switzerland in 2000 had 1 mandatory referendum, 2 optional referendums and 13 initiatives at the national level, and 84 mandatory, 12 optional and 16 initiatives at the local level, all bindings. Thus the aggregated score of referendums for Switzerland in 2000 would be:

The aggregated score could be 0, when there is no referendum held in a country and it could be as high as in the above example for Switzerland. The REP score for Switzerland in 2000 is calculated as Log (28.4+1)=1.47, however, considering the highest score limit, its score would be 1.

We collected all binding and non-binding, national and local, and different types of referendums for all countries whose data are available in the extensive database of C2D (2012) in addition to other sources (Van der Krieken, 2011; Direct Democracy Navigator, 2012; Statistics Norway, 2012) for the range of 1990–2009.

Table B1

Appendix C

Table C1

Table C1 Countries’ information and scores for six indicators and final indices of integrative dimension of democracy (IDD) and participative dimension of democracy (PDD) for the range of 1990–2009