Congruence of national and Europarty manifestoes in the EU elections 2019

This paper seeks to answer the question how congruent the policy positions of the four main Europarties were compared with their respective national member parties, as well as how they influenced each other in the drafting process, using the comparative example of German parties in the EU election 2019. It finds that the programmatic congruence of Euro- and their German member parties is small, with only insignificant coherences in issue salience and policy prioritization. In fact, some parties are even revealed to have higher programmatic congruence with other German parties than with their own Europarty. Nevertheless, cases of up- and downloading of policy positions as well as general political reconciliation processes are reported for all parties.


Introduction
Although European party federations (Europarties) have been formally recognized as 'expressing the will of citizens of the Union' in the Maastricht Treaty (Art. 10 TEU), people still do not have the ability to directly vote for them in the European elections. Instead, citizens can only vote for their national member parties (Gaffney 1996, p. 17). Hence, if it is true that Europarties shall express the will of their voters, it is therefore necessary that their policy positions are substantially aligned with those of their national member parties (Sigalas and Pollak 2012, p. 28;Hix and Lord 1997, p. 204 f). But are they?
This paper thus seeks to answer the question how congruent the policy positions of the four main Europarties were compared with their respective national member parties, using the comparative example of German parties in the most recent EU election 2019. It will empirically examine the programmatic role of the Europarties by assessing and comparing electoral manifestoes, as well as analysing conducted interviews with party officials. Thereby, it seeks to contribute to the general understanding of the programmatic relevance and impact of Europarties, as well as their relationship with and influence on their national members.

Theory
Overall, the field of European party structures is still principally under-researched. Consequently, only few studies have ever been conducted on examining the coherence of Europarties and their national members. Moreover, nearly all of those have only assessed very broad dimensions of coherence, namely left/right (Camia and Caramani 2012;Mair and Thomassen 2010) or pro-/contra-EU alignment (Marks et al. 2000;2002), and mostly also just concentrated on the horizontal relationship between either national parties (Ladrech 2002), Europarties (Gabel and Hix 2002), or MEPs and their parliamentary group (Hix et al 2005), instead of the vertical relationship with their respective Europarties. Only Sigalas et al. (2010; have so far focused on a vertical perspective of policy coherence in the case of the 2009 EU elections. This paper builds up on this existing groundwork, by further elaborating the respective theories and for the first time empirically testing them at the example of the 2019 EU elections. It thereby seeks to contribute to the academic debate about the Europeanization of party structures.

Relevance of party accountability and programmatic congruence
According to Camia and Caramani (2012, p. 49), 'converging and eventually cohesive ideologies within party families cutting across national and territorial identities represent an important step in the integration of Europe from an electoral perspective.' Albeit, at the moment, the interest in and salience of Europarties is rather low among citizens. To put in the words of Hix and Lord (1997, p. 15), they are 'invisible elements in the Union's political infrastructure'. This is closely connected to the 'second-order' character of European elections (Reif and Schmitt 1980): Accordingly, voters do not base their electoral decision on EU issues, but rather on national considerations (Hobolt and Tilley 2014, p. 131). In fact, this is not least caused by the institutional framework of EU elections, which for now only lets national parties appear on the ballot paper. Thus, particularly Europarties (and their manifestoes) have so far played an insignificant role in public perception (Adams et al. 2011(Adams et al. p. 2011. Nevertheless, according to Mair and Thomassen (2010, p. 29), even 'when voters vote on the basis of national issue dimensions, the aggregation of these national systems of political representation at the European level can still prove effective in representing the will of the European people' as insinuated by the EU Treaty. However, this would only be the case, 'when (and if) the party systems across the member states of the Union are congruent and when voters across Europe vote for similar reasons for similar parties belonging to a particular party family.' Therefore, this is what this paper seeks to assess.

The programmatic role of Europarties
Today, all Europarties adopt European manifestoes prior to the EU elections, which 'are a reasonably accurate statement of the positions that the European political elites take on issues on the EU agenda' (Gabel and Hix 2002, p. 937). Julie Smith (1996, p. 279) even regards it as the Europarties' 'most significant role' to 'co-ordinate transnational party manifestos'. Accordingly, they 'have relatively sophisticated systems […] to coordinate the positions of the national parties on the main issues on the EU agenda': The European manifestoes are usually drafted in 'special working groups' (Hix and Lord 1997, p. 209) consisting of 'senior party officials', and subsequently 'signed at the highest political level, by the national party leaders' (Gabel and Hix 2002, p. 937).
However, at the same time, most of their member parties have not renounced adopting their own national manifestoes for the EU elections, as well; and Europarties are technically not able to force them to actually consider and adopt their common positions in those (Sigalas and Pollak, 2012, p. 29). Not least, because it is still those national parties, which appear on the ballot papers, the crucial question thus arises, how much the policy positions of their national manifestoes actually correspond with those in the European manifestoes of their Europarty. After all, according to Sigalas and Pollak (2012, p. 28), 'a degree of ideological cohesion which can lead to agreements on common political positions at the public level is absolutely necessary for a political association to act as a party'. Consequently, by examining their programmatic coherence, one can also evaluate how much of a 'party' different Europarties actually already are. Considering that they do have divergent ideological approaches towards European integration in general, it will be interesting to see whether this is also reflected in different levels of party integration.
According to Simon Hix (1996, p. 321), theories on the impact of European party structures 'range from a "minimalist" view, which sees the [Europarties] as no more than transnational interest groups, to a "maximalist" view, which regards them as European-level political parties.' Hence, while for instance Smith (1996, p. 279) states that national parties have accepted those common European manifestoes only 'with varying degrees of enthusiasm', Hix and Lord (1997, p. 209) on the other hand noted that 'there is a growing coherence in the policy platforms pursued by each Euro-party'. This paper will thus empirically assess these contradictory positions. Therefore, based on those, it will present three possible levels of integration in the following section and formulate respective hypotheses for every model.

The maximalist model: federal democracy (downloading)
Under a maximalist paradigm, Europarties would be regarded as 'real' parties, centrally setting the policy on EU issues for all of their national members (Sigalas and Pollak 2012, p. 24-26). Sigalas et al (2010, p. 4) call this case 'federal democracy', which insinuates a parallel to fully-fledged democratic states, in which manifestoes for national elections are normally solely adopted on federal level. Europeanization theory would refer to the underlying political process as (supranational/neo-functional) top-down policy 'downloading' (Wong 2011, p. 158).
Consequently, if national parties even publish their own manifesto at all, one would expect full programmatic coherence, with national parties aligning to the previously adopted European standpoints (Sigalas et al 2010, p. 4). This leads to the following hypotheses: • (H max 1) The European manifestoes are adopted before the national ones. • (H max 2) European policy positions are downloaded by national parties. • (H max 3a) The policy statements within national manifestoes are congruent with their European manifesto. • (H max 3b) The issue salience within national manifestoes is congruent with their European manifesto. • (H max 4) If programmatic disagreements emerge, national parties are expected to adjust their position to the European one.

The medium model: delegated democracy (uploading)
On the other side, the medium model, which Sigalas et al (2010, p. 3) refer to as 'delegated democracy', sees Europarties just as coordinative platforms of national parties (Hix and Lord, 1997, p. 63f). Accordingly, positions in the European manifestoes would originate from national interests that were advocated and accepted by the party leaders on European level. Europeanization theory would refer to this political process as (intergovernmental) bottom-up policy 'uploading' (Wong, 2011, p. 158). Consequently, one would expect European manifestoes to be far less detailed than their national counterparts, and only to include those positions that all national party leaders could unanimously agree on. They would therefore only reflect lowest common denominator policies and have mainly the function of outlining the general policy priorities (Hix and Lord 1997, p. 13). However, although European manifestoes are less extensive, the member parties are at the same time expected to adopt at least those positions that could be agreed on European level, and especially 'not to contradict them', as well as to adhere to the general policy prioritization (Smith 1996, p. 281). This leads to the following hypotheses: • (H med 1) European manifestoes include less words and statements than national manifestoes. • (H med 2) National parties upload their positions to the European manifesto. • (H med 3) National parties adopt all European statements. • (H med 4) Policy statements in European manifestoes are never contradictory to a related statement in a national manifesto. • (H med 5) The general policy prioritization within national manifestoes is congruent with their European manifesto.

The minimalist model: regional democracy (crossloading)
The last model however sees no relevant programmatic role of Europarties whatsoever. In such a 'regional' democracy (Sigalas et al. 2010, p. 4), Europarties only function as lose networks of independent national parties, but have no significant impact on their specific policies. Therefore, any policy coherence would not result from active coordination, but rather from ideological vicinity (Hix et al. 2005, p. 213). Hence, the European manifestoes 'count for little' (Smith 1996, p. 281) and are 'more symbolic than statements of intent' (p. 279). Also Sara Hobolt (2014, p. 1531) seconds: 'While Euro-parties produce electoral manifestos, the extent to which the national parties use these manifestos in their own campaigning has traditionally been minimal. Instead, European election campaigns have tended to focus on domestic political matters and be dominated by national political actors.' Thus, one would expect: • (H min 1) There is no significant influence of Europarties on national manifestoes.
Nevertheless, also in regional democracies, there might be a programmatic cue for parties; and that is the perceived national interests. Accordingly, parties would rather orientate themselves by other national parties than by their Europarties. Europeanization theory would refer to such a political process as (realist) horizontal policy 'crossloading' (Wong 2011, p. 158). This consideration is based on studies like the one of Dorussen and Nanou (2006), who explore how parties from the same member state unite on European issues, if national interests are at stake. Although previous studies have inter alia already examined the respective voting behaviour of MEPs and found that left/right ideology (and therefore party) alignment is hereby more important than national coherence (Hix et al. 2005;Marks et al. 2002), none has hitherto compared the programmatic coherence of manifestoes among national parties to their Europarties. Hence, the hypothesis for such a process is: • (H min 2) National manifestoes are programmatically closer to other national parties than to their Europarty.

Methodology
In order to test the hypothesis stated above, this study analysed the electoral manifestoes for the 2019 EU elections of the four main Europarties-the Liberals (ALDE), Christian-Democrats (EPP), Greens (EGP) and Socialists (PES)-and their respective German member parties (FDP, CDU/CSU, 1 GRÜNE and SPD). I decided to focus on those four political groups, which Gabel and Hix (2002, p. 935) describe as the 'four main party families', because they 'have dominated politics in the EU' since the 1970s (p. 937) and are the only Europarties that 'possess all the elements of a transnational party federation' (Hix 1996, p. 308). Germany as comparative counterpart was chosen, because (other than e.g. France or Italy) it has exactly one corresponding, major national party for every main Europarty, which simplifies the analysis. Moreover, since it is the largest and therefore most powerful member state, it is not only particularly important to look at, but it can also be assumed that if any member party has a decisive impact on the European manifesto, it should be the German.
In total, over 500 pages of manifestoes with roughly 110,000 words were categorized by hand into over 1470 distinct statements (codes). Each statement represents one specific policy pledge or political claim (not only the mention of an area). 2 Every new statement was categorized with one, and only one, code. The entire main body text of the manifestoes was coded following the MECE principle; other contents, including headlines, pictures or table of contents, were excluded. I hereby followed established methodology for content analysis of manifestoes (Gabel and Hix 2002).
After that, the statements were controlled for any doublings and subsequently categorized into one of the 49 issue families defined by Sigalas et al (2010, p. 14). Finally, these issue families were further grouped into eight broad policy fields, namely (1)  Moreover, I conducted interviews and questionnaires with high-ranking officials from every party, who were involved in the drafting processes of the manifestoes. The names of the individuals have been anonymized and the answers translated from German.

Results and analysis
Hereinafter, the results of the collected data will be presented and analysed by reference to the hypotheses formulated in chapter 2.

The maximalist model: federal democracy (downloading)
• (H max 1) The European manifestoes are adopted before the national ones.
All Europarties had set up working groups for drafting the manifestoes in mid-2018 and, according to all party officials that were interviewed, there was an active programmatic exchange with the national parties. Three of four Europarties (ALDE, EPP and PES) also formally adopted their manifestoes before their German members. However, the European Greens (EGP) passed their manifesto only after the German GRÜNE (see Table 1).   Therefore, this hypothesis is accepted for the Liberals, Christian-Democrats and Socialists, while for the Greens it must be rejected. This indicates that the Green European manifesto could not have significantly pointed the way for the national manifesto and the latter was formulated independently. This contradicts the maximalist model, which would expect an already adopted European manifesto to guide the programmatic process of its member parties.
Both Green party officials stated though that at least the draft European manifesto was nevertheless considered on national level; particularly through close and constant personal channels. This might also explain why both manifestoes bear the same title: 'Time to renew the promise of Europe' (EGP 2018, p. 1). However, both officials admitted at the same time that the EGP drafting process started much later than the one of the GRÜNE, which hampered the coordination process.
• (H max 2) European policy positions are downloaded by national parties.
The maximalist model further predicts that the EU policy positions in national manifestoes are mostly downloaded from European level. And indeed, most interviewed party officials could offhand name several examples of policy positions that were in fact borrowed from their respective Europarty: The ALDE/FDP official for instance called the reference to European universities in the national manifesto 'a greeting to Macron', since the French President had previously proposed those in his Sorbonne speech. The EGP/GRÜNE official mentioned European night trains as an idea that originated from the EGP. And the CSU/CDU official could even name a whole range of concrete policy proposals that were taken from the EPP manifesto, such as '10,000 additional personnel for Frontex, an EU cyber-brigade within two years', and the idea of a 'masterplan against cancer, which […] was 1:1 adopted by CDU/ CSU'. Therefore, a priori, this confirms the hypothesis. However, the mere fact that some policies were downloaded does not validate the maximalist paradigm, which expects the whole national manifesto to be programmatically coherent with its European counterpart. And in fact, most party officials stated at the same time that these examples were not the general rule (see also H med 2 and H min 1). Consequently, a closer look at the actual statements of both manifestoes is needed: The policy statements within national manifestoes are congruent with their European manifesto.
To test the second hypothesis, hence the congruence of statements, following Niwattanakul et al (2013) and Sanger and Warin (2019), I used the Jaccard Coefficient, which is defined by: with E = statements in the European manifesto and N = statements in the national manifesto. Moreover, it pertains that J ∈ [0, 1] , where 0 means no and 1 total congruence of policy statements (i.e. political positions). For the examined manifestoes, this led to the following results: Those quantitative data now indicate significant divergences between the European and national manifestoes. Although the national parties did have between 23.51 (SPD) and 36.21 (CDU/CSU) per cent of all statements in common with their Europarty, the vast majority of statements were different. As shown in Table 1, this mainly results from a large amount of statements that were only included in the national manifesto. Thus, on average, only 28.93 per cent of all statements were congruent among Euro-and national parties. Furthermore, European statements only constituted between 24.95 (SPD) and 51.54 (CDU/CSU) per cent of all statements within national manifestoes (on average 34.65 per cent). Therefore, the Christian-Democrats have the most and the Socialists the least congruent manifestoes, but overall, no party came even close to the predicted full programmatic coherence.
Note that the fact that if a statement was marked 'European' or 'national only', this does not necessarily indicate that both party levels have different positions on the topic, but only that the statement appeared in just one manifesto. Nevertheless, the small Jaccard Coefficients suggest that the national parties did not fully align to the previously adopted European standpoints, as the maximalist model suggested, but in fact independently formulated their own national manifestoes with unique statements.
• (H max 3b) The issue salience within national manifestoes is congruent with their European manifesto.
However, it might be the case that, although the national parties included additional statements, at least the salience of each issue family was congruent in both manifestoes. According to Sigalas and Pollak (2012, p. 32), 'when the national manifestos focus largely on the same thematic families', 'high programmatic convergence' can be assumed. Following Hix (1999, p. 82), the issue salience was therefore 'measured by calculating the percentage [of words] of each document dedicated to all the' statements of that issue family and subsequently compared (see Table 2). 3 3 I also calculated the issue saliences by using the number of codes per issue family for comparison. However, they did not lead to significantly different results. The results show that also the issue saliences of the national manifestoes differ significantly from the European manifestoes. Even among the top-13 issue families, 4 one can already recognize major differences in priorities (*/**): For example, while PES dedicates 13.95 per cent of their manifesto to social affairs (which even represents the largest issue salience of all), it only constitutes 4.22 per cent of the SPD manifesto; thus, just a third of space and even less than 'EU democracy'. Similarly, while the EGP spends only 4.46 per cent on foreign affairs, the German GRÜNE devote 10.02 per cent to it.
Among all 49 issue families, FDP and GRÜNE deviate on average more than 1.5 and SPD 2.33 percentage points from their European manifesto, 5 making the latter once more the most incongruent party, with also the most significant deviations among the top-13. CDU/CSU on the other side again shows the highest congruence, but still at a low level. Moreover, the conducted Pearson's Chi-squared tests for the number of words in each issue family show highly significant independences (p < 0.001) for all parties and Χ 2 -statistics far above the critical value of 65.17 (for α = 0.05). Therefore, the issue saliences cannot be regarded as congruent and hence, also this hypothesis must be rejected.
• (H max 4) If programmatic disagreements emerge, national parties are expected to adjust their position to the European one.
Although SPD, FDP and GRÜNE officials stated that they did internally discuss to alter a position that conflicted with the European manifesto, none of all interviewed party officials was aware of any case in which their national party actually changed its manifesto in order to avoid programmatic disagreements. Instead, SPD, FDP and GRÜNE stated that they actively worked towards changing the policy position in the European manifesto. Hence, also having in mind the results above, a top-down maximalist process must be discarded for all parties. Therefore, the following section will turn towards the medium model: The medium model: delegated democracy (uploading) • (H med 1) European manifestoes include less words and statements than national manifestoes.
As evident from Table 1, national manifestoes did indeed include both more words and statements. In addition, most of the interviewed party officials stated that the European manifesto is 'generally shorter, handier, and less specific' (SPD Official), because it supposedly only serves the purpose of outlining the general policy priorities of the group, while national manifestoes must be much more detailed (EGP/ GRÜNE Official). The EGP/GRÜNE official reasoned that this results from the fact that on European level, much more interests from different national backgrounds must be coordinated and reconciled. 5 For comparison: The average issue salience is 2.04 per cent. 4 i.e. all issue families that appeared among the top-3 of any party + those that appeared among the top-10 of more than 3 parties. Together, the top-13 issue families represent more than 50% of all manifestoes.
Notably though, only the CDU/CSU official stated that, albeit being slightly shorter, the EPP manifesto was 'more profound and detailed', while on the contrary the national manifesto had the objective to rather only reflect the 'general guidelines'.
• (H med 2) National parties upload their positions to the European manifesto.
Although most national parties adopted their manifestoes after the Europarties, all party officials confirmed that they influenced the European manifesto in the drafting process by uploading own positions. In fact, they could even state concrete national policy proposals, which made their way into the European manifesto: Accordingly, for instance, the GRÜNE could insert an immediate ban of toxic pesticides and discriminating algorithms, as well as a digital regulating authority (GRÜNE Official), the PES statements on sustainability reportedly came from the German SPD (SPD Official), and the FDP claimed to have set the general narrative of the ALDE manifesto (ALDE/FDP Official). Only the EPP/CDU and CSU/CDU officials again stated that the process rather ran the other way round.

• (H med 3) National parties adopt all European statements.
According to Table 1, German parties adopted between 54.91 (CDU/CSU) and 80.28 (SPD) per cent of the statements in the European manifesto. Interestingly, although the Christian-Democrats seemed to be most congruent when it comes to overall statements and issue salience, in the adoption rate of European statements they thus turn out to be least compliant. That can be explained by the little amount of own (national only) statements. Hence, a higher congruence rate does not necessarily point towards a higher adoption rate and thus actual compliance with European policy positions.
Although the mentioned adoption rates are far from all statements, and the hypothesis can thus not be fully supported, the relatively high numbers anyways suggest a significant influence of the European manifestoes. Sigalas and Pollak (2012, p. 32) even regard any adoption rate above 50 per cent still as 'high programmatic convergence'. Accordingly, even though the medium model implies the adoption of all lowest common denominators (i.e. all European statements), the results should not be misinterpreted to discard any interdependence.
Nevertheless, all Europarty manifestoes include some even quite concrete statements that were disregarded on national level: For example, the EGP (2018, p. 4) urges for an 'increase [of] marine protected areas to 20% of our seas', PES (2019, p. 1) for 'a ban on zero-hours contracts', ALDE (2018, p. 8) wants to 'abolish geoblocking', and EPP (2019, p. 13) to cut '1.000 outdated rules and regulations'. Hence, the question remains, why some (like those) European statements were not adopted by their member parties, although they did already just constitute a lowest common denominator.
Several reasons are thinkable: First, it could be explained with mere disregard: This option would support the minimalist model and will be considered for H min 1. However, second, national parties might also feel that some concrete positions in the European manifesto are just not of interest for their national electorate, although they do not principally oppose them: For instance, the EGP/GRÜNE official mentioned that Nordic member parties had lobbied to include some specific positions concerning the arctic in the EGP manifesto, which were very important for them, but of no significant interest for the GRÜNE. Hence, they consented it on European level, but did not adopt it in their national manifesto anyways, because they felt that this was not relevant for them.
And third, it could also constitute a form of constructive abstention: Accordingly, and analogous to Article 31 TEU, member parties might choose not to block an issue on European level in order to resolve a deadlock, but then leave it aside in their national manifesto. In fact, some interviewed party officials did admit that their national party had different opinions on a few policy positions in the European manifesto, but just kept silent about it in order to avoid conflicts: For instance, the SPD representative stated that 'the PES was considerably clearer and more pointed concerning climate change and a digital tax, while the SPD leadership was more chary in this regard, and thereby more in line with the German Ministry of Finance. 6 ' However, according to him, there were also internal factions that pushed for a more progressive standpoint. Thus, they decided not to bring up those controversial issues, 'in order to avoid visible contradictions with the PES manifesto.' (SPD Official) This would be in line with the finding of Hellström and Blomgren (2016, p. 265) that 'political parties downplay issues in response to internal divisions among their supporters.' Another notable example was mentioned by the EGP/GRÜNE official, who stated that the GRÜNE actively decided to refrain from putting forward too many amendments at the EGP congress, in order to avoid the impression of pomposity, since they are the largest member party. Moreover, an explicit case of constructive abstention can in fact even be found in the EGP manifesto, where a small note indicates: 'Alternattiva Demokratika, Malta, registers its objection to all references to abortion' (EGP 2019, p. 12). Consequently, at least parts of the not-adopted European statements may be explained by such considerations.
• (H med 4) Policy statements in European manifestoes are never contradictory to a related statement in a national manifesto.
The previous section revealed that the national parties did not adopt every statement of the European manifesto. Nevertheless, as was also argued regarding H max 3a, this alone does not necessarily mean active opposition to and thus (public) incoherence with the Europarty. However, this would be the case, if the national party not only keeps silent about an issue, but also openly contradicts it. Looking at the concrete statements though, such examples are in fact very rare. The few slight differences, for instance a more open wording about the accession of the Western-Balkans in the EPP (2019, p. 7) manifesto compared to the CDU/CSU (2019, p. 21), or different concepts for youth voluntary services among the EGP (2018, p. 8) and the GRÜNE (2018, p. 172), are not notable enough to constitute real contradictions. Therefore, this hypothesis cannot be rejected.
• (H med 5) The general policy prioritization within national manifestoes is congruent with their European manifesto.
As predicted by the medium model, and confirmed by several party officials (see H med 1), the European manifesto has the most important aim to determine the general policy prioritization of the group, rather than stating detailed policy measures. Thus, one would expect the manifestoes to be at least congruent when it comes to the prioritization of broad policy fields. Therefore, the percentage of words per policy field 7 was measured and ranked accordingly, in order to assess the level of coherence.
As Table 3 shows though, even among those general policy fields, there is again no significant coherence in salience, and not even in ranking, within party groups. The very few cases of equal ranks (*), together with the strongly significant results of the conducted Chi-squared tests instead once more indicate high levels of independence. Even when looking at the order in which issues were presented in the manifestoes, there was little resemblance: The only parallel was that both green parties stated ecological and the Social-Democrats social issues first, respectively (which comes with no surprise). 8 Therefore, this hypothesis must be rejected.
Overall, it can be concluded for the medium model that, other than expected, national parties did not adopt all European statements and even the general policy prioritization differed significantly from their Europarties. Hence, also the related theory of a 'delegated democracy' cannot fully be supported, although national parties did in fact both up-and download some positions to/from European level.

The minimalist model: regional democracy (crossloading)
• (H min 1) There is no significant influence of Europarties on national manifestoes.
As predicted by the minimalist model, the measurable impact of the European manifestoes thus seems to be low. Also most party officials stated in the questionnaire that they perceived their influence on the national manifestoes as 'small'. In fact, the GRÜNE official even openly stated that the European manifesto 'did not actually alter the programmatic statements' of the GRÜNE manifesto, 'also because the national manifesto was adopted earlier', but at the same time stressed that anyways, 'there were no significant controversies'. Insofar, the hypothesis cannot be rejected for ALDE/FDP, EGP/GRÜNE, and PES/SPD, albeit, as shown above, some (minor) cases of influence were reported.
Only the representatives from CDU/CSU by contrast replied that the EPP manifesto had a 'large' (EPP/CDU Official) or even 'very large' (CSU/CDU Official) programmatic impact. In fact, the CSU/CDU official stated that 'in the overall programmatic process, the EPP […] was even more decisive than the own party leadership'. 'Although it did not literally write the CDU/CSU manifesto', he said, 'the overall programmatic lead resided in the EPP.' • (H min 2) National manifestoes are programmatically closer to other national parties than to their Europarty. However, the 'regional democracy' model not only predicts an insignificant impact of the European manifestoes, but also a higher influence of other national parties, manifested in programmatic 'crossloading'. To test this hypothesis, I compared the number of same statements from the European manifesto with other national manifestoes: Interestingly, as Table 4 shows, for every national party, there are indeed examples of higher congruence (i.e. more statements in common) with another German party than with its own Europarty (*): For example, while the SPD only adopted 114 statements from the PES manifesto (which represent 24.9 per cent of all its 457 national statements), it had 299 statements in common with the CDU 9 (65.4 per cent). Furthermore, I also compared the number of statements that a national party only had in common with at least one other German party (hence excluding policy positions that were also part of the European manifesto). Even in this case, two parties, FDP and SPD, had still more 'national only' statements in common with other German parties, than same statements with their own Europarty.
However, note that these data do not directly imply a lack of coherence with the own Europarty, since, as the results for H med 3 showed, the national parties did in fact adopt a high percentage of European statements (they were just less in absolute numbers in some cases). Furthermore, it was outside the scope of this paper to (1) compare those results with national parties in other member states, and (2) to run the same tests as conducted with the European manifestoes above with the other national manifestoes, so it cannot give an indication about attributes like congruence in issue salience, which would be needed though in order to eliminate disturbances, such as the length of national manifestoes.
Nevertheless, the results do again heavily question both the maximum and medium model, since those would predict the Europarties to play a much more significant (or even predominant) role compared to other national parties. On contrary, the higher coherence among some national parties do suggest that perceived common national interests should not be disregarded when it comes to EU elections; once more supporting the theory of 'second-order elections' (Reif and Schmitt, 1980). However, to thoroughly assess those effects would be the task of possible further research.

Discussion
As was shown by the foregoing analysis, the answer to the initial research question, how congruent the policy positions of the four main Europarties were compared with their respective German member parties, and the overall programmatic role of Europarties in this respect, is not that clear-cut. On the one hand, national parties did neither adopt all statements, nor align to the issue salience or even policy prioritization of the European manifestoes. On the other hand, they did not entirely ignore them as well, but both up-and downloaded a significant amount of policy positions to/from the European level.
Consequently, none of the introduced models is able to thoroughly explain the impact of Europarties on the programmatic process regarding the EU elections 2019. Albeit, this was also never their aspiration, since they only represent idealistic models. However, what could indeed be shown is that the maximalist model (federal democracy) is the least appropriate theory to explain the interaction between Euroand national parties, since all its hypotheses had to be clearly discarded. Accordingly, for most German parties, there was no indication that their Europarty decisively guides their whole national manifesto; if at all, then rather vice versa.
Only with the Christian-Democrats, there were findings that point to a more maximalist approach: With the highest Jaccard Coefficient (i.e. share of same statements), the highest percentage of European statements in the national manifesto, the highest congruence in issue salience (i.e. lowest deviation), and also matching statements by both interviewed party officials, the CDU/CSU seems to be the most integrated party that was examined. Surely, at the same time, it had the lowest adoption rate of European statements and, such as the other parties, no congruence in issue salience or policy prioritization as well as higher statement convergence with some other German parties-so it can still not be regarded as an example for a full 'federal democracy'. Nevertheless, the question remains, how the former findings can be explained.
One possible answer to that is delivered by the interviewed CSU/CDU official: According to him, the close reconciliation can mainly be explained by the fact that Manfred Weber was the Spitzenkandidat of both CDU/CSU and EPP, and thus 'played a decisive role' in the whole process. Hence, he stated, the whole programmatic lead resided in Weber, who 'was given free rein' by the party leadership. In fact, the EPP reportedly even posted own personnel to the party bureau of CDU/ CSU. Hence, for the 2019 EU elections, the programmatic process of the Christian-Democrats was apparently indeed more 'downloading' than 'uploading'. This was as well confirmed by the CSU/CDU official, who stated that conversely, the EPP manifesto was drafted mostly autonomously on European level, with only a personal link via the office of Manfred Weber. However at the same time, according to him, this was an exceptional situation, not only due to Weber, but also because for the first time, CDU and CSU rallied behind one common electoral manifesto; so already on national level, a reconciliation process was necessary.
Overall, although the results suggest a generally rather low programmatic congruence, they should nevertheless be interpreted relative to the level of EU democracy as a whole. At the current stage of European integration, and given the still non-state character of the EU as well as the decentral election of the European Parliament, we can and should not expect Europarties to already function as fully-fledged federal parties. In fact, the high (albeit not full) adoption rates of European statements of consistently above 50 per cent do indicate a certain programmatic influence of Europarties. Moreover, all interviewed party officials also stated that the influence of their Europarties on the national level has increased compared to the last elections. Although this could not be quantitatively verified for now, it suggests growing relevance of Europarties; and would be another field for further research.

Limitations
As explained in the initial methodology section, this paper focused on the main four Europarties and their German members. Hence, it can specifically not make assertions about other, smaller Euro-and member parties. It may well be that national parties in smaller member states download more positions from the European manifesto. Moreover, possible national differences in, for example, party structures, traditions, and participation in government, as well as political history (such as length of EU membership) might distort the expected results for parties in other member states (Rashkova 2014).
Although the presented findings therefore cannot be unconditionally generalized to all parties in Europe, the paper nevertheless contributes to a general understanding of European party structures and their programmatic processes; particularly in relation to the largest and most powerful member state Germany. Thus, inter alia the maximalist model could be exposed as an unlikely explanation of the European party system as a whole, since it would require all member parties to have a highly coherent programme; so incongruence with one (and in this case even the largest) member state / party already dismantles this theory. In parts, this also applies to the medium model. Hence, only the minimalist model remains largely unchallenged.
Furthermore, it is important to stress that this paper did not seek to analyse the principal ideological coherence of Europarties and their national members, as other studies have done before. Instead, it solely looked at concrete policy proposals and claims. Therefore, it would be misplaced to draw conclusions about the general ideological proximity of parties from the presented results.

Conclusion
As argued in the introduction, Europarties are referred to as 'expressing the will of citizens of the Union' in the EU Treaty (Art. 10 TEU), and they also often depict themselves accordingly; most recently particularly concerning the Spitzenkandidaten for Commission President (Hobolt 2014). However, the findings of this paper suggest that the national parties, which actually appear on the ballot paper, do not necessarily share the same positions on EU policies. In fact, no significant congruence of policy positions of the four main Europarties with their German member parties could be detected for the EU elections 2019, although the Europarties did play a role in the drafting process of manifestoes and national parties did indeed upand download a significant amount of policy proposals to/from the European level.
Hence, the assessment of Hix and Lord (1997, p. 197) that 'real parties at the European level are still some way off' seems to be true until today. If Europarties seek to become 'real' parties at one point in time though, which can actually claim to represent the people's will, the programmatic coherence with their national member parties therefore has to be improved. Until then, EU citizens will presumably continue disregarding their European manifestoes and instead predominantly relying on national cues. Eventually, such as in other fields, this stands and falls with the willingness of national parties to transfer significant policy-making competences to the European level; by either voluntarily relinquishing their power to adopt own electoral manifestoes, or by institutional changes, such as the idea of transnational lists, through which Europarties might be directly elected in the future. Right now though, such further steps of political integration are still far-off.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.