Abstract
The Eurovision Song Contest is not only the largest song contest worldwide but also probably the world’s largest election for a non-political office. In this article, we are interested in the voting behaviour of Eurovision viewers. Do they vote sincerely, strategically according to rational choice assumptions (i.e. for the song they believe will be the likely winner) or for another song? Using data from a large-scale survey carried out in Europe, we find interesting voting patterns with regard to these questions. Roughly one-fourth of the survey participants would vote for either their preferred song or for the song they think will win. However, the percentage of strategic voters is lower (11%). In contrast, many individuals (i.e. 36% of participants) would vote for another song, one that is neither their preferred song, the likely winner, nor a rational choice. The reasoning behind these remaining votes may include neighbourhood voting, ethnic voting, and voting for one’s favourite European country.
Keywords
Today [the Eurovision Song Contest] is a magical place where (mostly) European countries come together to battle one another using insanely elaborate costumes (and nudity), glitter, and fireworks instead of bombs and guns.
Founded in 1956 as an attempt by the European Broadcasting Union (EBU) to bring Europeans together after the trauma of World War II, the Eurovision Song Contest (ESC) is much more than the largest song contest in Europe; it is one of the most widely watched non-sporting events in the world. The 2017 edition of the ESC attracted 200 million viewers worldwide across television and Internet streaming media. Aside from the fact that the contest has become a European institution whose winners can go on to become national heroes of sorts, it also is a competition with strong political repercussions on at least four dimensions. First, like most major sporting events, the ESC provides economic, marketing, and other opportunities to not only host cities and participating countries but also to performers, producers, and broadcasters (Akin, 2013; Andersson and Niedomysl, 2010; Bolin, 2006; Kressley, 1978). Notably, Southern and Eastern European countries select the ESC’s venue and performers carefully to construct an image that is hardly authentic, but nonetheless attractive to tourists (Amegger and Herz, 2016; Baker, 2008).
Second, the ESC raises an opportunity to engage in reflection and public discussion of identities at the national level (Christensen and Christensen, 2008). The ESC contributes to nation-building and the formation of national identities, particularly in countries at Europe’s peripheral sphere, which were formerly part of the Soviet bloc and have (re)gained independent statehood relatively recently (Iglesias, 2015; Johnson, 2014; Miazhevich, 2012; Mitrović, 2010). 1 Third, despite the fact that political songs are officially not allowed, some countries such as Russia and Ukraine have been remarkably ingenious in promoting their political agenda through their choice of representatives (Cassiday, 2014; Johnson, 2014; Miazhevich, 2010). 2 Finally, the contest has been instrumental in promoting the lesbian, gay, bisexual, and transgender (LGBT) community and LGBT politics more generally (Baker, 2016; Cassiday, 2014; Lemish, 2004). 3
In addition to the aforementioned reasons, the ESC has also been important for scholars of voting behaviour. While we assume that many ESC viewers will indeed vote for the song they like the most, we are also interested in whether some fans may be driven by other considerations. In this article, we are interested in whether those who vote in the ESC do so according to which considerations. Do Eurovision watchers cast a sincere ballot? If they do not vote sincerely, do they follow the logic of rational choice by voting for the most likely winner among their preferred songs? Or do ESC televoters jump on the bandwagon by jettisoning their personal preferences and supporting the country whose song they think is most likely to win? Or rather, do voters cast their ballot according to other considerations? The existing literature on Eurovision voting patterns, which has looked at phenomena as geographical biases in voting, patriotic voting, and voting based on the song’s language, or order in the competition, suggests that ESC viewers’ voting behaviour is not as simple as it may seem at first glance (Blangiardo and Baio, 2014; Gatherer, 2004, 2007). In this article, we aim to determine the degree to which there are discernible voting patterns in ESC voting.
For this study, we developed an original questionnaire and fielded it online during the week of the 2016 ESC (see Supplementary Appendix). Our individual-level analyses of more than 500 study subjects reveal four key findings with regards to the voting behaviour in the ESC. First, most respondents do not vote for their preferred song; sincere voting occurs only in 26% of the cases in our sample. Second, another 26% of respondents vote according to the logic of bandwagoning, choosing to vote for the song they think has the highest chance of winning (despite the fact that this was not their preferred song). Third, strategic voting is relatively infrequent; only 11% of the vote choices follow rational choice assumptions of choosing the most likely winner among one’s preferred songs. Finally, most individuals in our survey vote for a song other than their preferred song, the likely winner, or the strategic choice.
This article proceeds as follows. After providing some background information about Eurovision, we go on to succinctly summarize the current state of the Eurovision voting literature. In the following section, we present some theories of voting behaviour and derive some expectations about the results of our survey data. Afterwards, we introduce our survey and discuss our sampling procedure before presenting our analysis and discussing the results. Finally, we summarize the main findings of this study and situate them in a broader perspective.
Background information on the ESC
Created in 1956 to generate some pan-European content for the nascent EBU (Merziger and Nathaus, 2009), the ESC has long entertained viewers tuning in from all corners of Europe and the world. While all songs entering the competition must have a vocal component, there is no restriction on the type of music. Thus, songs performed have ranged from romantic ballads to hard-rock anthems. The most recent winners have included Salvador Sobral in 2017 with a smooth Portuguese-language ballad, Susana Jamaladinova in 2016 with a haunting and mournful ode, and Måns Zelmerlöw in 2015 with a catchy pop anthem. Over the course of the competition’s history, performers such as ABBA and Céline Dion have taken to the stage to launch very successful international careers after Eurovision.
In addition to being a musical spectacle, many academics have commented on the ESC as a site of European nation formation, identity building, and culture formation (Bohlman, 2004; Sassatelli, 2002; Tragaki, 2013). While only seven countries took part in the first edition of the competition, the number of entries has steadily increased, with the 2017 edition counting 42 entries. The fall of the USSR and the dissolution of Yugoslavia precipitated a rise in the number of independent countries on the European periphery, thereby significantly increasing the number of participating countries since the 1990s. This eastward expansion has not been without consequence. Lee and Bideleux (2009) point to the ESC as a way of changing mental maps of Europe, as the competition contributes to the debate over what constitutes ‘Europe’. This question arises time and time again as countries that may not be traditionally seen as ‘European’ have also been participants, such as Israel, Morocco, Armenia, Azerbaijan, and Australia.
The general format of the ESC follows that of Italy’s Sanremo Music Festival (Fricker and Gluhovic, 2013). In short, the competition is held annually, with each participating country being invited to submit one original song to be performed on the night of the competition. After all the songs have been performed, each country then casts votes for other countries’ songs (but they are not allowed to vote for their own song). The song that receives the most points is the winner. More relevant for scholars of electoral behaviour, the voting system used in the ESC is unique, with each participating country awarding a set number of points to 10 countries. A song seen as a particular country’s ‘favourite’ will be awarded 12 points, with the second-favourite song given 10 points. From there, 8 points through 1 point are given to the remaining ‘favourites’. Those not in the Top 10 get 0 points from that country.
The voting system has been modified several times since the contest’s inception. 4 Early days of the competition used national juries to award points. Televoting was introduced in 1997 as a test in five countries and then encouraged in all participating countries the following year. From 2009, the competition has moved to a mandatory hybrid voting system, combining elements from both a popular vote and a jury vote to calculate the final score. Until 2015, the jury and the televote counted equally in determining a country’s attribution of points. The voting rules slightly changed again starting with the 2016 competition, with the jury and televotes being tabulated separately. This slight modification effectively doubles the total number of points that can be awarded to countries. Nonetheless, we believe that studying televoting patterns in the ESC offers an interesting insight into ordinary citizens’ voting behaviour in a setting where the outcome is (largely) non-political in nature. 5
The Eurovision voting literature
On a certain level, the distribution of who chooses to vote in the ESC resembles that of who chooses to vote in elections for political office. There seem to be three types of individuals. First, there is a majority of uninterested citizens not interested in Eurovision; they do not watch the contest and will most likely not vote. Second, there are rather passive viewers. These individuals watch the ESC from the ‘safe distance of ironic whimsy’ (Coleman, 2008: 132). This second group may or may not vote in the contest. The third group consists of avid fans. These enthusiasts connect in fan clubs or broader communities, with social media facilitating these connections (Highfield et al., 2013). This third group is almost guaranteed to vote.
Actual turnout figures in the ESC are hard to determine. For one, the EBU has not made turnout figures and actual vote numbers public. Equally important is the fact that during the 20-minute voting window following all performances in the final, individuals can vote as many times as they like. Hence, it is difficult to determine how many individuals participate exactly. For sure, turnout for the ESC is lower than turnout in national or subnational elections. However, some research indicates that it can reach up to 10% of the population, which rivals turnout in European Parliament elections in some constituencies in the United Kingdom (Tormey, 2015). Despite our lack of knowledge about the actual percentage of voters, we deem it safe to assume that several million people vote across all participating countries. Yet, in this study, we are less interested in who votes in the ESC, and more interested in how they vote.
A relatively large literature, which uses aggregate voting results by country, has tried to tackle this question, coming up with several explanations for how individuals vote. These studies suggest that some individuals might not necessarily vote for the song they like the most or the song they think is most likely to win. Rather, they might vote following other strategic criteria that may reflect geographic, cultural, linguistic, historical, and political proximities (Clerides and Stengos, 2012; Gatherer, 2004, 2006; Ginsburgh and Noury, 2008; Saavedra et al., 2007; Spierdijk and Vellekoop, 2009; Yair, 1995).
Several studies identify so-called ‘voting blocs’ in the ESC. For example, Dekker (2007: 55) uses data from the 2005 edition of the competition to identify five voting blocs: Eastern, Nordic, Balkan, Eastern Mediterranean, and Western. He observes that scoring within friendship blocs is higher than between blocs. In the same vein, but using more politicized names, Gatherer (2004, 2007) speaks of the ‘Viking Empire’ and the ‘Warsaw Pact’ as two strong voting blocs. Recent research illustrates that two blocs have been particularly strong in the 2010s: the former Soviet and former Yugoslav states (Blangiardo and Baio, 2014; Clerides and Stengos, 2012).
Similar to bloc voting, Eurovision voters seem to have a strong tendency to vote for neighbouring countries (Dekker, 2007; Orgaz et al., 2011; Spierdijk and Vellekoop, 2009). For example, Fenn et al. (2006) find that Cyprus and Greece have almost always exchanged votes; that is, they have reciprocally given each other the most votes. Voting blocs and the tendency of individuals to vote for neighbouring countries could be due to the fact that adjacent countries share a common language, religious background, or similar cultural traits. Because voters know neighbouring countries, they might also have some familiarity or positive affect towards them. More subtly, it might also be that they have comparable tastes, including similar musical tastes. Yet, Charron (2013) points out that the extent of bloc voting and neighbouring voting differs from one country to another. He demonstrates that countries with less impartial political institutions, characterized by norms of patronage and favouritism, vote more for their allies regardless of their performance than countries with more impartial institutions, reflecting strong norms of meritocracy.
Another cultural effect is diaspora voting. Using the example of Turkish emigrants, Orgaz et al. (2011) highlight that Turkey receives disproportionally strong support from countries with large Turkish populations such as Germany. Dekker (2007) calls this phenomenon ‘vote donation’. Finally, the language of the country’s song performance also influences voting behaviour. For example, Spierdijk and Vellekoop (2009: 423–424) find that language plays a role in explaining voting behaviour; songs sung in English represent almost half of all winning songs. Blangiardo and Baio (2014) confirm this finding in their 2014 study. 6
In addition to cultural patterns in ESC voting, other scholars have identified general behavioural patterns. Bruine de Bruin (2005) and Haan et al. (2005) find not only that host countries systematically receive a larger number of votes but also that the performance order is positively correlated with vote shares; the later countries perform on the night of the competition, the more points they receive. These behavioural or mechanical effects have become stronger since the introduction of televoting in the late 1990s (Clerides and Stengos, 2012). In the post-1998 era, exposure matters too; in ESC finals, TV viewers are more generous with contestants they have already seen in the semi-finals (Verrier, 2012).
This short literature review illustrates that voting behaviour in the ESC is multifaceted, at least partially driven by cultural and linguistic components (e.g. cultural voting blocs and neighbourhood voting), but also by some discernible behavioural patterns. In this article, we want to determine if and to what degree ESC voting follows the three regular voting mechanisms: sincere voting, strategic voting, and bandwagoning. We also juxtapose these voting patterns to what we label ‘other voting’, which is voting behaviour that does not follow any of the three aforementioned patterns. In the next section, we apply sincere voting, strategic voting, bandwagoning, and other voting to the ESC and discuss how any of these patterns could manifest themselves.
Voting patterns and Eurovision
The first type of vote that we examine is sincere voting. Sincere voting means that a person votes for their preferred choice regardless of whether or not that option has a chance of winning. Applied to the ESC, sincere voting simply means voting for one’s preferred song. We expect sincere voting to be widespread. Under the current voting system introduced in 2016, there is a district magnitude of 10 for each country (i.e. 10 songs get points from each country). This implies that by voting sincerely, individuals are unlikely to waste their votes; nearly half of the participating songs get points from each country. We further assume that the rate of sincere voting should be quite high because a song is a cultural representation of a country. ESC viewers’ vote choice might be impacted by the song’s language, content, and style, which, in turn, might precondition individuals to like certain songs and dislike others. In addition, music is about emotion. If an individual likes a song, she likely feels strongly drawn towards that song, which, in turn, might make her likely to vote for it as her favourite. This applies even more so, considering that the stakes of these elections are relatively low.
The second type of voting is strategic voting. Following Duverger (1951), the theory and empirics of strategic voting were first developed in the case of single-member district plurality elections with more than two candidates (McKelvey and Ordeshook, 1972). In such a first-past-the-post setting, some people may decide not to vote for their preferred candidate because they believe that they have no chance of winning, and instead support their second-preferred candidate. This is what is called a ‘strategic vote’: a vote for a candidate that is not the preferred one, based on expectations about the possible outcomes of the election with the aim to making one’s vote count (Blais et al., 2001, 2009). 7 From this perspective, one simple way to determine whether a given voter votes strategically is to show that (1) she does not vote for her preferred option, (2) she does so because she perceives her preferred option to have little or no chance of winning and (3) she votes for her preferred option among those she perceives to be viable. (see Blais et al., 2009). 8
What would strategic voting entail in the case of the Eurovision Song Contest? Very simply, it would mean that individuals would vote for a song that is not their most preferred song because that song is perceived to be an unlikely winner. Rather, the strategic Eurovision voter would vote for her favourite among the subset of songs deemed most likely to win. This means that she votes partly on the basis of personal preference and partly on the basis of expectations about various songs’ chances of winning. One obvious implication is that the strategic voter must establish which songs are viable and then supports the song that she prefers among the group of viable songs. This raises the question of what constitutes a ‘viable’ song. The (simple) answer is that it is any song that the person believes has some chance of winning or gaining points. This perception may be completely wrong, but what matters is whether she makes her vote choice on the basis of that perception.
But what does winning mean exactly in the case of Eurovision? First and foremost, it means obtaining the highest number of points overall across all participating countries. A song is viable if and only if it has some chance of winning the most points overall. But we must also keep in mind that each person’s vote is first counted in her own country and so the ESC voter should consider also whether a given song is viable in her own country, that is, whether it has some chance of being awarded at least 1 point by her own country’s jury. This, in turn, requires that a given song must be among the 10 most popular songs in that country. In short, a viable song is one that is among the most popular in a given country and across all participating countries. The strategic voter should consider only viable songs and vote for her preferred option within that subset.
There is a peculiarity of Eurovision that gives rise to another potential strategic consideration: an individual cannot vote for her own country’s song. This means that some people cannot vote for her preferred option. If an individual living in France has the French song as her favourite, she is unable to vote for her first preference. Theoretically, this means that that individual should vote for her second-preferred option, which, in turn, is the first preference among those she can choose from. However, in case that an individual’s second preference is the German song and that she also believes that there is a close race overall between the French and the German songs, then a vote for the German song may contribute to the defeat of the French song, which is her favourite. In such cases where an individual does not have the possibility of voting for her first preference, she may strategically decide to support a song that is viable in their own country, but not overall, in order to increase her preferred song’s chances of winning the contest overall. A strategic voter, then, is someone who votes non-sincerely, that is, supports an option that is not her most preferred, in order to increase the possibility that her vote will make a difference. When making her vote choices, she takes into account how other voters are likely to vote.
The third type of voting is bandwagoning. A bandwagon occurs when someone votes for the candidate she expects to win, irrespective of her preferences (see McAllister and Studlar, 1991; Singh et al. 2016). While the sincere voter votes only on the basis of her preferences, the strategic voter takes into account both her own preferences and expectations about the outcome of the election (voting for the candidate she prefers among those that she perceives to be viable). A bandwagoner’s motivation is to be on the winning side and is thus completely focused on supporting the candidate with the best chance of winning (see Aldrich et al., 2016).
The fourth type of voter is one we label as ‘other’. Such a voter is a person who does not vote according to any of the above criteria. That is, she votes for neither her preferred option, the candidate who she thinks has the best chance of winning, nor her preferred option among the pool of candidates that are most likely to win. Rather, a voter who is in the ‘other’ category may use criteria such as language affinity, ethnic voting, or neighbourhood voting to determine her vote choice. Despite the fact that these criteria should already matter when individuals pick their preferred song, we nevertheless hypothesize that, contrary to voting in an election for elected office, the group of ‘other’ voters should be relatively large in ESC voting.
To sum up, we have four types of voters: (1) a sincere voter who votes solely based on preferences; (2) a strategic voter who decides based on a combination of preferences and expectations over outcomes; (3) a bandwagoner, who makes decisions solely on expectations, that is, the voter votes for the expected winner, irrespective of her own preferences; and (4) an ‘other’ voter for whom neither expectations nor preferences matter.
Research design
Questionnaire design
To ascertain the degree of sincere voting, strategic voting, bandwagoning, and ‘other’ voting, we devised a short questionnaire consisting of 10 questions asking participants about their preferences in the 2016 ESC (see Supplementary Appendix). To determine whether individuals vote according to any of the three previously mentioned concepts in the voting literature, we need to know individuals’ preferences, their evaluation of the likely winner(s), and their vote choice. The first and second question asks individuals about their three most preferred songs, and who they perceive to be the three most likely winners, respectively. After presenting some betting odds as an ‘unbiased’ assessment of the songs’ chances of winning, we then ask individuals for which song they would vote for under the current voting system (i.e. the 10 songs with the most votes from a country, get 12, 10, 8, 7, 6, 5, 4, 3, 2, and 1 point, respectively, from that country). Because we initially thought the type of electoral system might moderate an individual’s vote choice (particularly if voters vote strategically), the next question asks if individuals would vote for another song if the existing (rather proportional) system was replaced by a winner-takes-all formula, whereby the song that receives the most votes in a particular country receives all of that country’s 58 votes. The remaining questions are contextual; they ask survey subjects about their preferred electoral system, their interest in the ESC, their knowledge about Eurovision, and some demographic information.
Sampling procedures
Because of the format of the ESC with semi-finals before the final competition, we only had a short period in which to field our survey. In the interest of putting together a survey that is as realistic as possible to actual ESC voting in the final, our survey went online on 12 May 2016 just after the results of the second semi-final were announced. We left it in the field until televoting for the final ended on 14 May 2016 at approximately 23:30 Central European Time (CET). 9 In total, we had four surveys in the field: one survey for Germany, France, the United Kingdom, and Europe at large. We translated the survey for Germany and France in the country’s native language, while the survey targeting the United Kingdom and Europe at large went to the field in its original English version. We recruited participants using a snowball sampling technique; we sent the survey link to all people we knew in the three respective countries and in Europe at large, asked them to participate, and to further distribute the survey. We also used Facebook to distribute our survey via various Eurovision clubs and fan pages, even purchasing targeted ads on Facebook. Furthermore, we created a specific Twitter account to publicize our survey. To encourage participation, we offered a monetary prize to one randomly chosen participant.
We are well aware that this sampling procedure does not provide a random or representative sample of Eurovision viewers or voters. Rather, what we have is a convenience sample. However, this is not detrimental for the purpose of our study. It is not our goal to predict the outcome of Eurovision correctly (despite the fact that our results are not that far away from the actual results, see Table 2). Rather, we aim to figure out what criteria individuals use when they vote in Eurovision. Despite our convenience sampling technique, we were able to collect data from a diverse sample of individuals. In general, participants were rather young (i.e. the average participant age was 31 years), 10 highly educated (80% of participants had some kind of university-level education), as well as interested in and knowledgeable about the ESC. 11 In terms of gender, women were slightly overrepresented. Our survey sample consists of 684 participants, out of which 464 individuals answered all questions. Out of the 601 participants that revealed their residency, 211 are from Germany, 139 are from the United Kingdom, 79 are from France, and 161 are from other countries. Altogether, we believe that this diverse sample allows us to adequately address our research question.
Data analysis and results
Is there sincere voting, strategic voting, bandwagoning, or ‘other’ voting in the ESC? We should point out at the outset that to fully address this question, we would need to have full information about participants’ preference rankings and the perceived chances of winning for all the entries. The information we have is about the three songs that each person likes the most and the three that they thought had the best chances of winning. That information is obviously incomplete, 12 and so all we can do is investigate patterns that are consistent with sincere voting, bandwagoning, strategic voting, and voting that does not fit any of the three aforementioned categories.
Table 1 summarizes the frequencies of the four types of voting. Sincere voting, which simply entails voting for one’s preferred option, is lower than expected. Under the existing voting system, 26% would vote for their preferred song. This may seem surprising, given the reduced risk of wasting one’s vote (and thus increased incentives to vote sincerely) under the new voting rules for the 2016 competition.
Sincere voting, strategic voting, bandwagoning, and other voting.
What about strategic voting? The clearest scenario of a strategic vote is a non-sincere vote for the second- or third-preferred option that is perceived to have better chances of winning than the most preferred candidate. Rather few votes satisfy these conditions. First, relatively few individuals vote for their second or third choice (57 and 52 individuals, respectively). Second, these votes could be deemed to be strategic only if their second or third choice is viable, according to their own assessment. However, only a minority of 51 individuals report that they would vote for a second or third choice that they perceive to be the likely winner. This implies that overall roughly 1 out of 10 (11%) participants voted strategically.
What about bandwagoning? Bandwagoning implies voting for the most likely winner regardless of one’s preference, so we take a non-sincere vote for the candidate with the best chance of winning to be consistent with bandwagoning. Looking at the survey data, there is strong prime facie evidence of bandwagoning. In total, 264 (57%) of our respondents indicate that they would vote for the candidate that they perceive to have the best chances of winning. However, this convergence between individuals’ vote choice and their assessment of the likely winner is only consistent with bandwagoning, if individuals either do not report an intention to vote for their preferred song or for any song that they rank as their second or third preference. Considering the first scenario, we have some sincere voting. For 35% of all individuals who voted for the likely winner, the likely winner is indeed their preferred candidate. These voters are sincere voters. For another 19% within this group, the likely winner is either their second or third favourite. These voters are strategic voters. Yet, 46% of them (26% of the total sample) would vote for a song that is not one of their three preferred songs. These cases are clear instances of bandwagoning.
Two additional pieces of evidence further point to bandwagoning rather than strategic voting. First, what matters for the bandwagon voter is which candidate is most likely to win, while the strategic voter cares more about which candidates are viable. It is telling that while 57% of the participants would vote for the song they think has the best chance of winning, only 10.9% would vote for the candidate that they believe has the second-highest chances of winning and 3.4% would vote for the entry they consider to be the third-strongest candidate. The fact that it is only the top candidate that people rally to is more consistent with bandwagoning than with strategic voting. 13 Second, individuals could strategically vote for a song with low chances of winning to boost the chances of their preferred song. Yet, Table 2, which gives aggregate figures of vote choice across all four surveys, does not provide evidence for this proposition either. Under the second type of strategic voting, many votes, in particular in countries with a chance of winning, such as France, should have gone to countries with very little chances of winning, such as Armenia. However, this is not the case. Individuals tended to vote for the song that they believe will win, rather than voting for a weak song to increase the chances of their preferred song winning.
Vote percentage according to the existing voting system. a
The figures indicate shares of votes attributed by respondents.
The votes for France from France and for the United Kingdom from the United Kingdom are in parentheses because these votes should not count as participants in Eurovision are not allowed to vote for their country of residence. In the specific surveys in France and the United Kingdom, participants could not vote for their home country. However, in the survey featuring Europe at large, participants could technically vote for their home country. The votes for the United Kingdom from the United Kingdom and for France from France come from this survey.
Finally, slightly more than one-third of the respondents vote for a song which, according to our classification, falls in the ‘other’ category. These individuals expressed a vote choice for none of their favourite songs and for none of the likely winners. Although this finding may sound counterintuitive, it does not come as a surprise in light of the existing studies on aggregated voting behaviour in the ESC. Similar to our research, recent studies find little evidence of strategic voting at the aggregate level (Clerides and Stengos, 2012). Other works also suggest that some of the voting patterns consist of ‘unexplainable noise’ (Spierdijk and Vellekoop, 2009: 405). In our study, this ‘other’ category is likely to be made up of individuals whose vote choice is driven by other factors identified in earlier research: geography, culture, language, history, or politics (e.g. Gatherer, 2004, 2006; Ginsburgh and Noury, 2008; Saavedra et al., 2007; Yair, 1995). 14
Conclusion
Examining the extent of sincere voting, strategic voting, bandwagoning, and ‘other’ voting in the ESC, this article contributes to the narrow literature on voting in this particular contest. We have shown that there is little strategic voting, some sincere voting (26% of our survey participants expressed their intention to vote for their preferred song), and some bandwagoning behaviour (26% of individuals would vote for the song they think has the highest chance of winning). Unfortunately, we cannot answer the underlying question of what makes individuals believe that a song can win. Finally, we find that the largest category of voters is in the category ‘others’. What makes individuals vote for other than their preferred song or any likely winner? Is it affinity towards a certain country? Is it intuition or impressions gleaned from friends? Placing the results of this article in context with the broader Eurovision literature, it might be that an individual’s tendency to vote for songs other than their favourites or the likely winners may be influenced by considerations such as bloc voting, neighbourhood voting, or positive affect towards a certain country. Future research should continue to look into these possibilities.
This article also contributes to the larger literature on voting. Survey respondents were presented with Eurovision betting odds to help them make an informed vote choice; however, our analysis shows that this publicly available data may not influence individuals’ assessment of who is most likely to win, or even their vote choice. As with Eurovision, it is possible that in elections for political office, many individuals vote for the candidate or party who they think has a high chance of winning, even if these perceptions might not correspond to reality. This study also suggests that the role of the electoral system in individuals’ decision to vote strategically might not be as strong as theory predicts. In our study, 84% of the individuals would not change their vote under a new system.
While pushing forward our understanding of the Eurovision voter, this article also asks for renewed research into voting patterns in the ESC. Future studies should include a fuller set of questions about which songs they prefer and which songs have the best chances of winning. The data that we have provided here strongly suggest the presence of a variety of voting patterns: sincere, strategic, bandwagon, and ‘other’ voting. Further research is required to determine what kinds of individuals are more prone to engage in these types of voting (and what contexts can mediate this relationship). These future studies should also include motivational questions such as where the voter has family ties or a country where he/she regularly goes on vacation to further unravel this puzzle of ‘other’ voting in the ESC.
Footnotes
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Supplementary Information
Additional supplementary information may be found with the online version of this article.
Notes
Author biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
