The 2019 European Elections: something old, something new, something borrowed, and something green

In the aftermath of a European Parliament (EP) election, there are normally two prominent aspects that receive attention by scholars and experts: the turnout rate and whether the Second Order Election (SOE) model proposed by Reif and Schmitt (1980) still applies. That model is based on the idea that, because EP elections do not themselves provide enough stimulus as to replace the concerns normally present at national elections, the outcomes of EP elections in any participating country manifest themselves as a sort of distorted mirror of national (Parliamentary) elections in that country. The mirror is distorted because those national concerns are modified, not so much by the concerns arising from the European context in which EP elections are held as simply by the fact that EP elections are not national elections. In particular, at EP elections, national executive power is not at stake. The same party or parties will rule in each country after an EP election as ruled there before.

In the aftermath of a European Parliament (EP) election, there are normally two prominent aspects that receive attention by scholars and experts: the turnout rate and whether the Second Order Election (SOE) model proposed by Reif and Schmitt (1980) still applies. That model is based on the idea that, because EP elections do not themselves provide enough stimulus as to replace the concerns normally present at national elections, the outcomes of EP elections in any participating country manifest themselves as a sort of distorted mirror of national (Parliamentary) elections in that country. The mirror is distorted because those national concerns are modified, not so much by the concerns arising from the European context in which EP elections are held as simply by the fact that EP elections are not national elections. In particular, at EP elections, national executive power is not at stake. The same party or parties will rule in each country after an EP election as ruled there before. Some features of national election outcomes are due to concerns of voters regarding who will govern their country and those features are absent at EP elections. In particular, (a) turnout is expected to be lower at EP elections because those elections do not engage this critical voter concern; and (b) absent this concern, voters have no reason to consider how likely it is that each party will receive enough seats in the Parliament to be able to influence government policy. In national (Parliamentary) elections, voting for a party that is most unlikely to be able to translate its policies into law is often considered to be a 'wasted vote' and, rather than waste their vote, some citizens choose to use their vote to help decide which of the better-placed parties does win (Franklin et al., 1994). At EP elections, some of the voters who do vote are thus likely to choose a different party than the one they would have chosen had the election been a national election: often a smaller party that is not likely to attain government office (and hence is also unlikely to be a member of the current government).
Other aspects of EP elections that sometimes receive attention by commentators are suggested by a so-called Structural Model proposed by Franklin (2011). This model builds on the SOE model by suggesting that certain features of that model vary in their impact depending on three specific features of the EP election in each country: (a) whether it is the first EP election ever held in that country or notsuch elections see higher turnout; (b) what proportion of that country's electoral cycle (the expected gap between Parliamentary elections in that country) has elapsed at the time of the EP election concerned -EP elections occurring late in the cycle see higher turnout; and (c) whether compulsory voting is enforced in that countryagain leading to higher turnout. Few countries switch from compulsory voting to non-compulsory voting but, as additional countries join the EU that do not enforce compulsory voting, so the composition of the EU changes from (at the earliest EP elections) one in which close to half the countries enforced compulsory voting to one (in 2014 and 2019) in which only some 15% of member countries did so, lowering the overall impact of compulsory voting considerably. Franklin (2011) was concerned uniquely with effects of these features of an EP election on turnout but more recently (De Sio et al., 2019) the structural model has also been applied to effects on party choice, which have been found to be stronger in the immediate aftermath of a national Parliamentary election but to fade as the next national election approaches.
The 2019 elections were no exception to the general rule that such elections draw the attention of commentators and, immediately after those elections, we ourselves took a close look at the features of those elections that we have highlighted (De Sio et al., 2019). As already mentioned, we found that the SOE model did apply when it came to party choice but that the new direction of turnout change seen after 2014 (after which turnout increased sharply) was not readily explained by the structural model.
Indeed, turnout was the most obviously distinctive feature of these elections. The 2019 EP elections were the first to show a clear break in what had been a fairly steady decreasing trend in terms of participation. As Figure 1 shows, it was possible to observe an increase in participation for the first time ever, and also the first breach of a 50% threshold in the quarter-century since 1994. 1 Still, there is a sense in which the structural model did actually predict this increase, since the 2014 turnout level was considerably lower than predicted by that model and so an upward correction was not unexpected (although the correction was more than twice as great as called for to remedy the anomalously low turnout seen in 2014). And, if we see the 2019 EP election turnout as corrective, then it follows that the structural model was still relevant in 2019. Moreover, seen in that way, the structural model was able to correctly explain some 77% of the variance in EP election turnout over the course of all EP elections to date (De Sio et al., 2019). This very high explanatory power (corresponding to a correlation between predicted and actual turnout of close to 0.9) was based on a very parsimonious model involving only For a detailed overview of the turnout rate country by country, see Table A1. [2] National Rally (Rassemblement National or RN) used to be called National Front (FN) until 2018.
[3] The classification made in the PopuList database for the present populist countries in Croatia is valid until the year 2015. Therefore, the data regarding Croatia are not updated for 2019.
[5] The classification made in the PopuList database for the existent populist parties in Slovakia is updated only until the year 2006. Therefore, the data reported here adhere to this old classification of populist parties in Slovakia.
[6] The PopuList database is up-to-date. Nevertheless, for the critical situation of UK at the moment, by relying on the thinking and theories of many authors, I regard the Brexit Party (BP) as a populist one, even if this was not classified so by the original PopuList database.
[7] Not a party, but a single independent politician.
[8] Not officially associated with the Green group in the EP.
three structural factors (first elections boost, electoral cycle, and composition effect of compulsory voting). Along with the explanatory power of SOE and structural models, at least another two features of the 2019 EP elections captured the attention of commentators: the success in several countries of the green candidates and the failure of populist parties to achieve an electoral breakthrough. Table 1 shows an overview of both, along with the other prominent feature of the 2019 EP elections, that is, turnout, by offering a country-based overview and a comparison with the previous EP election.
The green wave, illustrated in Figure 2, seems to suggest that these might have been the first real European elections, in the sense of displaying an overarching issue whose effects were observable across virtually the entire European Union.
Finally, as shown in Figure 3, the performance of populist parties exhibits a more fragmented dynamic, raising many questions regarding possible reactions to Brexit and dynamics pertaining to the Second Order Election Model and its structural extensions.
The papers in this special issue try to tackle these questions by addressing different aspects of the 2019 European Elections. In their paper, Second-order Effects or Ideational Rifts? Explaining Outcomes of European Elections in an Era of Populist Politics, Talving and Ehin (2020) test the SOE model, both the traditional one, and a sui-generis one of their own design, in which party characteristics (such as populism, extremism, and euroscepticism) are included. With regard to the traditional second-order model, they find no evidence of an incumbency or electoral cycle effect, but party size still matters. Zooming on parties, they show how party ideologies also matter, and how the issue of European integration is still widely discussed in a national political arena. Angelucci et al. (2020), in their paper Europe matters… upon closer investigation: A novel approach to analysing individual-level determinants of vote choice across first-and second-order elections, applied to 2019 Italy, investigate whether there is evidence showing that a European dimension characterised the 2019 elections, or whether the SOE model still applies. They base their findings on data collected in Italy, which permits consideration of a large number of issues.  Similarly to Ehin and Talving (2020), they also find that there are still second-order effects, but that there is also a distinct emerging important of European-related issues.
The congealing of a new cleavage? The evolution of the demarcation bloc in Europe (1979-2019) (Emanuele et al., 2020) employs all European elections, including that of 2019, in order to investigate the evolution of the demarcation cleavage, which pits the 'losers' and 'winners' of globalisation against each other. The authors find that politicisation of the demarcation cleavage is growing over time, that the cleavage is in a mobilisation stage, and that, in the Italian 2019 EP election, this cleavage was a relevant dimension of competition.
Finally, Martill (2020) addresses one of the most interesting country cases, that of the UK, facing what should have been its last EP election. This Research Note offers a detailed overview of the dynamics at play in the UK, vis-à-vis its upcoming exit from the EU.
Funding. The research received no grants from public, commercial or non-profit funding agency.