Where the Grass is Greener: Comparing Green Party Success in National Parliamentary Elections and the 2019 European Parliamentary Election

Why do Green parties perform better in European Parliament (EP) elections than in national parliament elections? Even in countries that use proportionally representative voting systems for both national and EP elections, many Green parties gained more than twice the proportion of EP seats in 2019 compared to the previous national election. Using national and EP election results, European Social Survey data, and Eurobarometer polling from the 2019 EP election, I test competing theories of Green party success. As EP voter issues become more salient and more voters believe that their vote matters to EU policy, I find that the “second-order election” effect is less relevant. Surprisingly, I also find that the Greens are the only small party family that consistently gained more representation in the EP. Different voter priorities and party issue framing, for both environmental and pro-EU issues, boosted the Green vote in the 2019 EP election.


Introduction
The most recent European Parliament (EP) election, in May 2019, saw a dramatic surge in the number of seats won by Green Parties across Europe. Winning nearly 10% of votes, the Greens/European Free Alliance group became the fourth-largest party group in the EP (Graham-Harrison, 2019). The Greens' increased representation gives them the potential to act as kingmakers within the EP and demand climate action, such as the European Green Deal-a plan to make the EU climate neutral by 2050-that President of the European Commission Ursula von der Leyen promised in the EP in July 2019 (Tagliapietra, 2019). The rise of European Green Parties should be understood in the context of a shift away from mainstream big-tent parties towards smaller niche parties throughout Europe. The Greens, who prioritize the post-materialist values of environmentalism, anti-militarism, anti-discrimination, and social justice, are on the opposite end of the ideological spectrum from the right-wing populist parties which also increased their seats in the 2019 EP election. Yet both party groups champion ideological platforms focused on niche issues-environmentalism or immigration-to which traditional mainstream parties have failed to respond (Berman, 2019). Understanding why niche issue parties are growing at the expense of mainstream parties is crucial to understanding the party realignment currently taking place within Europe's democracies. While this dynamic has generated significant interest at the national level, it is comparatively understudied on the supranational EU level.
Why do European Green parties do better in European Parliament elections than in national parliamentary elections? In the 2019 EP election, many Green parties won twice the proportion of seats that they won in the most recent preceding national election. This disparity between Green voting at the national and European levels indicates that party choice is not fixed for voters. Party behavior, different institutional structures, and the perceived importance of elections matter in predicting vote switching. Several theories have been put forward to explain the different electoral outcomes in national and EP elections, including EP elections being perceived as unimportant compared to national politics, voters engaging in "sincere" rather than tactical voting, different framings of issues between elections, and Green voting as a signal of pro-EU preferences. This paper examines each of these theories in turn, using data from electoral results, the European Social Survey, Eurobarometer polling, and party manifestos, to determine the extent to which they hold true. Ultimately, I find that Green success in EP elections is the result of changes in behavior by both voters and parties. Voters with a "sincere" preference for the Greens are more likely to vote Green in EP elections even if they tactically vote for a larger mainstream party in national elections. Green parties highlight different issues in national and EP campaigns, focusing more on environmentalism and pro-EU integration policies for EP campaigns.

Cases
The phenomenon of greater Green representation in the EP occurs across EU member states, although it is most common in northern and western Europe, perhaps because climate change is a lower priority in eastern Europe (Ashdown, 2019). This presents an interesting puzzle: given the characterization of Green voters as ideologically motivated, we might expect them to vote for the party with which they most closely align-the Greens -in all elections. Additionally, higher effective electoral thresholds for EP elections in many countries, due to the small number of EP seats available, should result in the underrepresentation of small parties like the Greens. https://scholarship.claremont.edu/urceu/vol2021/iss1/10 One possible explanation of this puzzle is that different electoral institutions for national and EP elections result in different outcomes. For example, the French Green Party won 13 out of 79 seats elected from a single national constituency in the 2019 EP election but won only 1 of 577 single-member district seats in the 2017 National Assembly election. To fairly compare national and EP elections for this paper, I focus on the 16 EU member states that use proportionally representative electoral institutions for both national and EP elections and which have a Green Party that won seats in either of the most recent national legislature or EP elections. I define Green Parties as parties that are full members of the European Green Party (EGP). Although non-EGP members sit with the Group of the Greens/ European Free Alliance in the EP, I consider that only EGP members are ideologically similar enough to compare on the national and European levels. These 16 member states, the seats held by their Green parties in national parliament and EP elections, and the effective electoral thresholds for these elections, can be found in Table 1. Denmark, Finland, Sweden, Germany, and Ireland have much higher Green representation in the EP than in national legislatures. Luxembourg, the Netherlands, Bulgaria, and Spain also have higher EP representation. Austria, Belgium, Poland, and Portugal show little or no difference between EP and national Green representation, while Cyprus, Hungary, and Latvia have more Green representation in the national legislature. However, in Poland, Portugal, Cyprus, Hungary, and Latvia the proportion of Green votes in the national elections would be insufficient to clear the higher effective electoral threshold for EP elections. In Austria, the Die Grünen party actually received the same vote share (14%) in the EP election and the national election, even though it won a larger proportion of seats in the national legislature (Eddy, 2019) (European Parliament, 2019. To test my hypotheses, I use data from the five member states with the most significant difference between national and EP representation: Denmark, Finland, Germany, Ireland, and Sweden. These cases illustrate the puzzle of Green representation especially well because their Green parties all have over 50% more representation in the EP compared to the lower house of the national legislature. Additionally, these member states represent the EU's diversity of proportionally representative electoral systems: all elect representatives from multi-member districts but use slightly different systems. Denmark, Finland, and Sweden use a party list system, with Denmark and Sweden also allocating additional "adjustment seats" to ensure national proportionality (Valmyndigheten Sweden, 2018;Folketinget Denmark, n.d.). Germany uses a mixed member proportional formula where voters select individual candidates as well as parties to ensure proportionality (A.K., 2013), and Ireland elects representatives using the Single Transferable Vote (Lijphart, 1999).

Literature Review
The growing success of Green parties across Europe in recent years has prompted investigation into how and why they are gaining votes. Green parties generally do well in societies with post-materialist conflicts caused by high levels of wealth or the presence of a tangible environmental dispute (Grant & Tilley, 2019). Green party voters not only share a set of attitudes, but also several specific social characteristics: they are typically young, highly https://scholarship.claremont.edu/urceu/vol2021/iss1/10 educated, urban, and less attached to traditional churches (Dolezal, 2010). The increase of potential Green voters in post-materialist societies, combined with a rise in tangible environmental issues, may explain the rise in overall support for Green parties. However, these studies fail to explain the discrepancy between Green success at the national and European levels.
One explanation for why EP elections have different outcomes compared to national elections is that EP elections are "second-order elections." Reif and Schmitt (1980) propose that second-order elections are viewed as non-decisive, where the electorate perceives there to be "less at stake." Their theory claims that when there is more at stake, i.e., in first-order national elections, voters will strategically opt for a more powerful party that supports the general direction of their political views rather than a niche party that will have minimal influence in government. The second-order dimension of EP elections provides opportunities for small and new parties because they gain voters whose true preference lies with the small or new party but who support a larger established party in the first-order election. Government parties are also likely to lose votes in EP elections because support for the government has often declined since the national election, or voters may vote for the opposition to apply pressure on the governing party (Reif & Schmitt, 1980). This leads to my first hypothesis:

H1: Voters are less likely to engage in strategic voting in EP elections so small parties will gain vote share at the expense of larger parties, especially the ruling party or coalition.
A more nuanced version of the second-order election theory holds that there are different motivations for vote switching: "sincere voting"-changing behavior due to ideological differences on the European level, and "protest voting"-changing behavior to send a signal of disapproval to a party or government on the national level. Examining the 2004 EP election, Hobolt and Spoon (2012) found that the magnitude effect of sincere and protest motivations were remarkably similar on the likelihood of vote switching or abstaining. Because "sincere voting" is not accounted for in H1, this leads to my second hypothesis:

H2: More voters have a "sincere" preference for Green parties than actually vote for Green parties in first-order national elections.
A third explanation for vote switching to Green Parties is that voters' policy preferences for the European Parliament motivate voting. Carrubba and Timpone (2005), in their study of the 1994 EP election, established that vote switching to Green parties was consistent with the second-order election theory. However, they also found that "at least some of the electorate is demonstrating a tendency to cast votes because of how the EP may influence policy outcomes in the future" (Carrubba & Timpone, 2005). They reconcile these competing explanations by suggesting that the EP is becoming an increasingly important legislative body, so elections are no longer considered exclusively second-order. Instead, voters consider their policy preferences, which may differ between the national and European levels of government. If voters are motivated by EP policy outcomes, we should expect Green issues to have greater salience on the EU level, or Greens to run on different campaign issues in national and EP elections. The Greens' central issue of environmentalism may be more salient to voters on the supranational level, especially because the EU now plays an important role in setting environmental policy for its member states. For example, the EU has the power to fine member states for failure to meet their binding 2030 emissions targets (O'Sullivan, 2019). This leads to my third hypothesis: Where the Grass is Greener

H3: Voters prioritize different issues in the EP as compared to their national legislatures.
The increasing politicization of the question of European integration and even the EU's existence may also motivate Green votes in EP elections as a signaling mechanism of voters' pro-EU ideology. In the context of EU polarization, Green parties have differentiated themselves from other parties which offer "piecemeal reforms" by "making a strong case for full supranational democracy and for moving away from national vetoes" (Sorace, 2019). Hobolt and Spoon (2012) observed that the increased politicization of the question of the EU influenced vote choice. However, the effect of EU politicization on vote switching is disputed. Schmitt & Toygür's (2016) analysis of the 2014 EP election indicated that EU policy was not significant to the election results, and "the politicization of European Union politics did not really discourage the predictions of the second-order model" (Schmitt & Toygür, 2016). This dispute is especially relevant to more recent EP elections, when the issue of European integration and pooled sovereignty was highly politicized in many countries. In the leadup to the 2019 election, Frans Timmermans, the first vice president of the European Commission, predicted that the vote would be a battle for "the very soul of Europe" (Schaart, 2018). This leads to my fourth hypothesis: H4: Increased political polarization over the future of the EU motivates Green voting in EP elections.

Data and Methods
To test the second-order election hypothesis, I compare the proportions of seats won by all parties which ran candidates in the 2019 EP election and the previous national election held between the 2014 and 2019 EP elections. If the second-order election theory holds true, I expect representation to increase for all small parties, not only the Greens. I also expect representation of large parties, especially those in government, to decrease in the EP election.
To test the "sincere" vote switching (Hobolt & Spoon, 2012) hypothesis, I examine European Social Survey data on "party voted for in the last election" and "which party feel closer to" (ESS Round 7, 2014; ESS Round 8). This data is available for Sweden, Finland, Germany, and Ireland from 2016, and for Denmark from 2014. I also compare the reported percentage of votes for Green parties to the actual vote share Greens received in the previous election. If voters only express their "sincere" preferences in second-order elections, more respondents should feel closer to the Greens than voted for them in national elections.
To test the "different issue preferences" hypothesis, I analyze campaign materials from Green parties in national and EP elections in Ireland and Finland. Because Ireland and Finland are "new destination" countries that were not affected by the 2015 migrant crisis to the same extent as Germany, Sweden, and Denmark, migration should not skew issue importance in their elections. I examine which issues were emphasized by Green parties in the national and EP elections and analyze Eurobarometer data collected during the 2019 EP election to determine which issues were most important to voters.
To test the hypothesis that voters signal their support for the EU by voting Green, I use data from the ESS on belief that "European unification should go further" as a proxy to measure support for the EU and movement towards supranational democracy. If this hypothesis holds true, I expect respondents who want greater European integration to be more likely to vote Green due to Green parties' explicit pro-EU stance. I do not include https://scholarship.claremont.edu/urceu/vol2021/iss1/10 data on Denmark for this hypothesis because the most recent ESS data is from 2014 and would therefore correspond to the 2014 EP election. I also use Eurobarometer survey data from the 2019 EP election to identify what factors motivated voters to vote in the election.

Second-Order Elections
Across the five countries, there was no strong trend of vote switching in favor of small parties in the EP election. In Denmark (Table 2), the large center-left Socialdemokratiet party won less representation in the EP than in the Folketing, and the right-wing populist Dansk Folkeparti (Danish People's Party) suffered a dramatic loss. The large conservative Venstre party, however, increased its representation. The Green SF party and the socialliberal Radikale Venstre parties both achieved over three times their national representation in the EP. Although these parties are both relatively small, other small parties did not experience a difference in representation between the national and European levels. The most significant similarity is that both the SF and Radikale Venstre parties are left-wing and pro-EU. However, the Alternativet party is also left-wing and pro-EU and did not win any EP seats. In Finland (Table 3), the 2019 EP and national elections occurred within six weeks of each other. The Social Democratic Party and right-wing populist Perussuomalaiset party both lost representation in the EP election, while the center-right Kansallinen Kokoomus party, Green Vihreät -De Gröna, and social-liberal Svenska Folkpartiet all increased their representation in the EP compared to the national parliament. Some large parties won comparatively less representation in the EP, but other large, medium, and small parties were more successful in the 2019 EP election than they had been in the previous national election.
Where the Grass is Greener In Sweden (Table 4), most parties won about the same proportion of seats in the 2018 Riksdag and 2019 EP elections, excluding the Green Miljöpartiet de Gröna. There was some variation in small party representation, but there were both increases and decreases in representation between elections. In Germany (Table 5), most large parties lost representation in the 2019 EP election compared to their seats won in the 2017 Bundestag election, except for the Green Bündnis 90/Die Grünen party. Several small parties that did not clear the five percent national electoral threshold in the 2017 Bundestag election won seats in the 2019 EP election, but this is likely a result of the lower effective electoral threshold rather than widespread vote switching. In Ireland (Table 6), the governing Fine Gael party actually gained a higher proportion of seats than it won in the 2016 Dáil election. The Green Party also dramatically increased its representation, as did the small Independents 4 Change party. However, all other small and medium parties decreased their proportion of seats between the 2016 Dáil election and 2019 EP election.   These data challege Reif and Schmitt's second-order election theory. Instead of consistent vote switching in favor of small parties, vote switching benefited and hurt both small and large parties. Green parties are the only party family which experienced an increase in representation in the EP election compared to the previous national election in all five countries. Government parties tended to win fewer seats in EP elections (Ireland is the ex-Where the Grass is Greener ception) but this change was usually small. These data clearly demonstrate that the Green Party vote switching phenomenon is unique to the Greens and not a feature of all small parties.

Sincere Voting
In all five cases, as shown in Table 7, ESS respondents were more likely to select their national Green Party as the party they "feel closer to" than as the party they voted for in the last election (ESS Round 7, 2014;ESS Round 8, 2016). However, the percentage of respondents who reported voting for the Greens in the last election was generally higher than the actual vote share the Greens received in those elections. In Finland, the Greens won 8.5% of votes in the 2015 general election, in Sweden they won 6.9% of votes in the 2014 general election, in Germany they won 8.4% in the 2013 general election, and in Ireland they won 2.7% in the 2016 general election. In Denmark, for which the latest ESS data available is from 2014, the Greens won 5.6% in the 2013 local elections and 9.2% in the 2011 general election. This disparity between reported voting and actual voting may be a result of the ESS's small sample size (between 1,500 and 2,900 respondents for each of the countries). Because the ESS question does not specify to which "last election" it refers, there could also be variation in responses referring to national, local, or even European elections. The voting disparity could also be a result of respondents' tendency to overreport normative behavior (Brenner & DeLamater, 2016). Because voters "feel closer to" the Green Party, they likely view voting for the Greens as normatively good, and therefore may report voting for the Greens even if they voted for a different party in the last election. Hobolt and Spoon's (2012) theory that voters do not always vote according to their "sincere" preferences in first-order national elections is supported by these data. Given that Green Parties' actual vote share was less than the ESS reported voting, the disparity between "sincere" party preference and actual voting is even larger than suggested by the ESS data. Green Parties have a distinct socio-cultural constituency of young, highly educated voters who work as social-cultural specialists or are students, are predominantly urban, and are less attached to Christian churches (Dolezal, 2010). This constituency's structural position within post-materialist European societies informs their environmental, libertarian, and proimmigration attitudes, and therefore their strong affinity with the Green party. However, https://scholarship.claremont.edu/urceu/vol2021/iss1/10 even socio-culturally Green voters may not support Green Parties in national elections when the party is viewed as having betrayed its values in government, as was the case with the Irish Green Party in 2011 (Bale, 2013). Green parties' small size and perceived insignificance in governments may also influence Green voters to strategically vote for larger mainstream parties (Hobolt & Spoon, 2012).

Issue Voting
My third hypothesis is that voters prioritize different issues in the EP as compared to their national legislatures. For this analysis, I compare the 2019 EP election manifestos of the Irish and Finnish Green parties with those parties' manifestos for the previous national elections. The results are summarized in Table 8 below. For the 2016 national election, the Irish Green Party's platform was focused on domestic economic reform and increased spending on social services. The key issues highlighted were the development of economic sectors, ensuring affordable housing, developing public transport, developing tourism, and environmental protection (Irish Green Party, 2016). The 2016 election manifesto notably focuses on economic issues instead of explicitly environmental issues. In contrast, the Irish Green Party's election manifesto for the 2019 EP elections frames the party as explicitly pro-EU. The first line of the "Our Vision" section of the manifesto states "The Green Party believes in Europe" (Irish Green Party, 2016). The manifesto directly refers to the challenges of "authoritarianism, racism, terrorism and the threat of further economic uncertainty" which weaken the EU. The party's slogan for the election is "It is time to renew the promise of Europe." This explicitly pro-EU language dominates the first pages of the manifesto, sending a strong signal that a vote for the Green Party is primarily a vote in favor of the EU. After this section, the first three issues highlighted are "the right to inherit a livable planet," focusing on carbon, energy, and waste; "transport," including a ban on fossil-fuel cars; and "protecting nature on land and at sea," including forestry and fishing. The first economic issue comes on the 8 th page of the manifesto, and even then is primarily centered on promoting a circular and green economy .
In contrast to the Irish Green Party, Vihreät -De Gröna's national election manifesto focuses on more traditionally "green" issues. Its top priorities are listed as "stopping climate change," "building the best education system in the world," "preventing poverty and inequality," and "creating work and safety" (Vihreät -De Gröna, 2019a). The primary focus is on environmental protection, combined with appeals to the post-materialist values (Fagerholm, 2016) of Green voters with education and equality policies. This manifesto is cer-Where the Grass is Greener tainly in line with the expected political preferences of the young, highly educated, urban, and non-religious (Dolezal, 2010) Green voter base. Like the Irish Green Party, Vihreät -De Gröna's 2019 EP election manifesto primarily focuses on the party's pro-EU position. It emphasizes that "Finland needs Europe" and that "Europe is needed more than ever" in the world (Vihreät -De Gröna, 2019b). The manifesto also frames the EU as a "promoter of climate change prevention [and] environmental protection" which can "[show] the way in building a sustainable future for the world." Like the 2019 national election manifesto, it appeals to the values of Green voters by referencing the need to combat the divisions of "disputes, closed borders, isolationism, hatred and prejudice" threatening the EU (Vihreät -De Gröna, 2019b). A vote for Vihreät -De Gröna is a vote both in favor of continued EU integration and expansion of power, and for environmental policies on the EU level. Interestingly, the EP election manifesto does not mention any of the economic or welfare issues discussed in the national elections manifesto.
In summary, different issues are emphasized in the national and European election manifestos of the Irish and Finnish Green parties. The national manifestos tend to focus on economic and social issues such as the local economy, employment, and education. The EP manifestos tend to focus on the more abstract and aspirational goals of renewing European unity and environmental protection. Because Green voters are ideological voters, the comparative ideological orthodoxy of the EP election manifestos may offer greater appeal. According to a Eurobarometer report commissioned by the European Parliament to survey voters after the 2019 EP election, "combating climate change and protecting the environment" was the top issue in the 2019 EP election for 37% of respondents across all EU countries and the most mentioned issue by respondents in Denmark, Sweden, Germany, and Finland, among other countries. When asked why they voted for their party of choice in the election, the most common answer was "their proposals on European issues were closest to your ideas or values" chosen by 43% of respondents, closely followed by "their proposals on national issues were closest to your ideas or values" chosen by 42% of respondents (Zalc, Becuwe, & Buruian, 2019). These data demonstrate that voters differentiate between parties' proposals on national and European issues, and slightly prioritized their European issue preferences when voting in the EP election.
Another important indicator of issue voting in this Eurobarometer report was voters' belief that "their voice counts in the EU." Contrary to the assumption of the second-order election model that EP elections are viewed as insignificant, the high rate of belief among European voters that their voice counts indicates that issues are important to voters because they believe they can influence the policy of the EP. This is especially true for Green voters: according to the report, "citizens for whom climate change and the future of the EU are decisive are more likely to believe they can make things change by voting" (Zalc et al., 2019). Green party issues on the European level are clearly important to these voters, and the Greens' stronger ideological appeal in its EP election manifestos could explain why Green parties win more representation in the EP context.

Pro-EU Voting
During the 2019 EP elections campaign, Green parties across Europe positioned themselves as the party of choice for pro-EU voters. As demonstrated in the Irish and Finnish Green Party manifestos, the parties made their support for the EU and continued European integration explicitly clear in their campaign materials (Irish Green Party, 2019; Vihreät -De Gröna, 2019b). The Greens/European Free Alliance group within the EP has https://scholarship.claremont.edu/urceu/vol2021/iss1/10 also been one of the most active groups in putting forward proposals to increase participatory democracy in the EU and move away from national vetoes towards a full supranational democracy (Sorace, 2019). Importantly, the pro-integration position reflects the priorities of Green parties' core electorate, as Green party voters tend to support the EU more than average voters in their countries. Using the belief that "European unification should go further" as a proxy for support for the EU and European integration, Table 9 demonstrates that respondents who feel closest to their national Green party were far more likely than average respondents to support further European unification. This disparity was especially pronounced in Ireland, Finland, and Germany, where there were differences of over 10 points. In a post-Brexit Europe facing the rise of far-right Eurosceptic parties in many member states, the future of the EU was a hugely salient issue in the 2019 EP elections and may have been especially important to Green voters. 25% of all voters in the election ranked their support for the EU among their top three reasons for voting, an 11 point increase from the 2014 EP election (Zalc et al., 2019). "The way that the EU should be working in the future" was listed among the top two issues which motivated 36% of voters to vote in the 2019 election, making it the fourth most salient issue in the election, closely behind economy and growth (44%), combating climate change (37%), and promoting human rights and democracy (37%) (Zalc et al., 2019). Further, 27% of voters in the 2019 EP elections said that Brexit, the most prominent manifestation of Eurosceptic sentiment, influenced their decision to vote in the election. In contrast, only 17% of non-voters said that Brexit influenced their decision, indicating that the media coverage and discussions around Brexit encouraged people to vote rather than to abstain (Zalc et al., 2019). Respondents who finished their education at age 20 or above, were employed in white-collar professions, or who believed that their "voice counts" in the EU were more likely to say that Brexit influenced their decision. Because these demographics are overrepresented in the Green voter base, it is likely that Green voters were more influenced by Brexit than average voters.

Conclusion
The success of European Green parties in the 2019 EP election reflects the increased salience of issues that must be addressed at a transnational level: environmental policy, climate change, and EU governance. In this election, ideologically aligned Green voters voted according to their "sincere" preference rather than engaging in tactical voting for larger mainstream parties as they may have done in previous national elections. Green parties were also more explicit about their environmental and pro-EU policies in their EP election Where the Grass is Greener manifestos. Finally, Green parties were able to mobilize voters against the threat of Euroscepticism in response to Brexit and the growth of far-right politics.
The second-order election dynamic, which has traditionally characterized EP elections, did not explain Green success in 2019 nor did it consistently explain electoral outcomes for other parties. Small parties did not always do better in EP elections as compared to national elections, and the losses predicted for large governing parties were small or nonexistent in several cases. Perhaps the validity of the other hypotheses examined in this paper can shed light on the new dynamics of EP elections. The increase in voters' belief that their voice counts in the EU reflects a growing perception that EP elections matter and should not be treated as second-order. The salience of specifically "European issues" as a motivation for party choice in the 2019 EP election indicates that EP elections are not simply a reflection of national politics. Future EP elections will determine whether the firstorder election dynamics of issue salience observed in the 2019 election are the beginning of a larger trend or were a one-off effect of heightened EU polarization. However, the Greens' outperformance of their national representation certainly suggests that environmental and niche party politics are acquiring new relevance in the EU.