CAPÍTULO II 2 MARCO TEÓRICO
2.3 Bases teóricas
2.3.2 Componentes del humedal
This thesis argues that the study of sub-state elections, and indeed of elections in general, requires a different approach. If scholars continue to see sub-state politics as a function of statewide politics, they are closing themselves off to a world of information that can contribute to our understanding of political behaviour at every level. An evolving body of work has developed in recent years that ad- vocates for a multilevel understanding of elections and electoral behaviour (Schakel and Romanova,
2018). In this next section I discuss electoral research that can help us understand the dynamics of
elections in multi-level systems.
1.3.1 Elections in multi-level systems: a two-way street
An essential first step in understanding the dynamics of electoral politics in multilevel systems is to acknowledge the interchange that exists between different levels within a nation-state, and in partic- ular the interchange of behaviours that occurs between elections. Unlike the SOE, which sees elec- toral behaviour at sub-state elections as a function of statewide issues and attitudes, I argue that we should instead view elections in multi-level systems as a two-way street. This means that scholars of elections should account not only for how statewide factors might influence sub-state politics, but how sub-state factors can have a substantial impact on statewide, and indeed international, political events. To do this, we can draw upon a wide range of evidence from political science. For example, inherent within the theory of electoral cycles is the idea that past elections, and the relative timing of them, have an impact on the electoral outcomes of any other given election.Schmitt and van der Eijk(2008) identify different relationships, or ‘interdependencies’, between elections in multilevel systems:
1. ‘interdependencies between previous [and future elections]’ and;
2. ‘interdependencies between elections at different levels of government.’
In other words, the focus of each election is necessarily defined by the preceding one; the win- ning party or parties set the policy agenda until the next election at which they must defend these policies against an opposition. By the same logic, each election sets the parameters within which the following election takes place. If it is true that election results at different levels are interdependent of one another, then the motivations and behaviours of voters and politicians are necessarily inter- dependent also, given that they determine electoral results. Therefore, when examining an election
in a multi-level system of governance, it is necessary to consider how factors at other levels may in- fluence the outcome. In this sense elections at a particular level in multi-level systems should not be viewed as isolated events affected only by factors measured at the level of the nation-state. After all, when government terms of office run to full term (or near full term) it is very likely that any election that precedes another will be one that occurs at a different level, excluding the small number of areas where elections at multiple levels are held concurrently (Belgium and Spain, for example). Its likely to be the case then that one of the largest and temporally relevant influences of voter and elite be- haviour will be a recent election that occurred at a different level of government. Once again, we can turn to the literature of electoral cycles for support. This body of work describes a trend observed across a variety of states whereby the fortunes of political parties are influenced in part by where the poll falls in the political business cycle (Tufte,1975;Soldatos,1994). By this reading, it is the tempo- ral distance between elections, rather than whether they are first or second-order elections, which plays the largest role in determining how one election impacts upon another: i.e an election held 12 months prior to another election is likely to play a larger role than than an election held 24 months prior (Schakel and Dandoy,2014)
Contemporary political events across Europe illustrate how sub-state issues and elections have had substantial impact on statewide politics. In the UK, general election campaigns in Scotland and Wales – and indeed areas beyond their borders – are often framed by the previous sub-state election. The 2011 devolved election victory of the SNP, and subsequent Scottish Independence Referendum of 2014, dominated Conservative campaign messaging across the rest of Great Britain at the 2015 statewide general election (see Figure1.2for illustrative example) (see Cowley and Kavanagh,2018).
In Spain, the recent electoral success of right-wing populist party VOX can be viewed as a reac- tionary response to, among other things, the Catalan independence movement. Santiago Abascal, the party’s leader, has vowed to uphold “the unity of Spain and centralization of the state” (Hedg- coe,2018), whilstFernandez-Albertos(2019) has argued that “the Catalan issue is what has most
Figure 1.2:Conserva ve campaign poster, 2015 General Elec onSource:M&C Saatchi, London
motivated their support so far”. There are also recent examples of how sub-state factors have af- fected not only statewide political events, but international agreements between multiple nation- states. The Canada-Europe Trade Agreement was twice seriously delayed by sub-state interests; first by dairy farmers in Quebec (Johnson et al.,2013), and then by the sub-state Walloon Parliament in Belgium (BBC,2016). These examples present compelling evidence for a detailed and rigorous exam- ination of two-way interdependencies that exist between elections at different levels of government.
Such evidence of interactions between tiers of government also calls into question the idea that the relationship between statewide and sub-state politics is a one-way relationship. Elections at different levels of government can be used to ‘frame’ future elections. This is particularly impor- tant when assessing media coverage of multilevel politics. Prior elections are likely to be used in the media as ‘referendums’ (Simon,1989) or ‘barometers’ (Anderson and Ward,1996) that offer a real test of public opinion.Mattila(2003) has argued that ‘politicians, journalists and potential voters show increased interest in voting because [European Parliament] elections serve as markers of party strength in the upcoming [state-wide] elections.’ (pp. 456). It is also important to consider that such influence does not necessarily follow the normal logic of causation, where earlier events are causes and later ones are consequences.Oppenhuis et al.(1996) show that the effect of a statewide
election on a European Parliament election that precedes it, is different in character from one that follows it. Political actors are likely to anticipate future political events like elections and modify their behaviour as a result. I would suggest, then, that the way in which the media cover and frame politics in multilevel systems can therefore have a substantial impact on the political behaviour of citizens.
Another way that elections can be seen as interdependent is through the formation of voting habits. Habit is often cited as a proposed factor that can explain voting behaviour and turnout in particular: people are more likely to vote in the current election if they voted in previous elections (Gerber et al.,2003). Yet it is possible that other aspects of political behaviour can manifest them- selves as habits. For example,Shachar(2003) has presented evidence that party choice can also be viewed as a habit. In multilevel systems, this habit can by disrupted if voters cast their ballots for different parties at different levels of government.
The literature identifies two ways in which multilevel systems facilitate this differential voting. The first is that voters choose the party they vote for based on level-specific factors (Trystan et al.,
2003;Hough and Jeffery,2003;Wyn Jones and Scully,2006;Schakel and Dandoy,2014;Schmitt and Teperoglou,2017). This theory of electoral behaviour has been referred to as ‘multi-level vot- ing’, and challenges the SOE understanding of electoral behaviour as a reaction to statewide factors (Trystan et al.,2003).
Second, multilevel institutions can also disrupt the voting habits of citizens by creating incen- tives to cast votes for different parties at different elections. This is particularly relevant in multilevel systems where different electoral systems are used at different levels of government. Electoral sys- tems are frequently cited as having a considerable impact on the behaviours of voters and political parties (Duverger,1954;Sartori,1976;Cox,1997;Farrell,2001). How a system translates votes into representation shapes how parties coordinate organization and campaign efforts and how voters measure the costs and benefits of voting. When voters do face incentives to vote differently at dif-
ferent levels of government, this can change their future electoral behaviour.Shachar(2003) argues that the act of voting for a party different to the one a voter normally votes for, increases the proba- bility of that voter defecting to another party at subsequent elections. This can create a ‘spill-over’ of knowledge and behaviours from one election to the next, and between levels. Voters will procure an understanding of which parties are competitors and potential challengers; ways of maximizing the influence of their vote; and the political issues that are prevalent in their arena. Future voting will be built upon this knowledge, with knowledge from more recent elections having a larger influencing role. This notion of path dependency – that electoral choices are made utilizing past knowledge and decisions – has been examined in electoral research to a limited extent (see Campbell,1966;Reif and Schmitt,1980), but has not yet been analysed to the same extent in multi-level systems.
1.3.2 Summary
Elections at one level cannot be viewed in isolation, and instead should be viewed as points on an electoral continuum. Voters use knowledge from multiple levels of politics when formulating their electoral choices, and are often unaware of doing so, as elites and parties coordinate between lev- els and voters seek quick sources of information. This means that the behaviours and attitudes of voters are not constant, and are shaped by the level of government most salient at the time and the outcomes of previous elections at any level. Elections in multilevel systems must be viewed in this context. Recent political events in Spain, the UK and Belgium have highlighted that scholars can no longer view sub-state factors as a function of statewide issues. Instead, our analysis of elections in general need to account for both of these factors.