Issue Competition Between Niche Parties and Mainstream Parties in the News PDF

FIELD OF GREENS 1 Field of Greens: Issue Competition Between Niche Parties and Mainstream Parties in the News Rachid Azrout, University of Amsterdam...

FIELD OF GREENS 1 Field of Greens: Issue Competition Between Niche Parties and Mainstream Parties in the News Rachid Azrout, University of Amsterdam Joost van Spanje, Royal Holloway, University of London Aurelia Ananda, Royal Holloway, University of London Abstract Niche parties have emerged in many democracies worldwide. Various aspects of these parties have been studied, including the role of mainstream parties in their electoral success. Key to that success, arguably, is news media attention. Is the media attention that niche parties receive affected by mainstream parties as well? In this paper, we combine news value theory with party competition theory to argue that other parties influence niche party visibility. Focusing on green parties in 11 countries between 1992 and 2021, we find evidence that the salience of green policies in mainstream parties’ manifestoes enhance green party visibility in newspapers if these parties take an adversarial position. This positive effect turns negative as the mainstream party becomes greener, which suggests that it steals a niche party’s thunder. The insight that mainstream parties can influence media attention to niche parties opens up new lines of research on the emergence of niche parties. FIELD OF GREENS 2 Field of Greens: Issue Competition Between Niche Parties and Mainstream Parties in the News In Western Europe, two new party types have emerged since World War II that mobilize on immigration and environmental concerns, respectively. Taken together with a few other party types such as communists and regionalists, these parties are typically called “niche parties” (e.g., Adams et al., 2006; Van de Wardt, 2015). Various aspects of niche parties have been studied, including effects of mainstream parties on their electoral success (e.g., Meguid, 2005; Abou-Chadi, 2016). In this success, news media arguably have a role to play. Media attention is even more important for niche parties than for mainstream parties, as the latter enjoy a longstanding reputation and support base regardless of their visibility. At the same time, for many niche parties it is more difficult to create media attention. This is because they tend to have fewer established relationships with journalists and less power over them, as they are often smaller opposition parties. News media will typically find little incentive to include niche party voices in their news stories, and even less so when more powerful mainstream actors provide similar soundbites. In this paper, we combine Meguid’s (2005) party competition framework with the notion of balanced reporting and with news value theory, the standard theory that explains what makes the news (Galtung and Ruge, 1965; Harcup and O’Neill, 2001; 2017). Based on these three ideas, we make the novel argument that mainstream parties influence niche party visibility, and we empirically test it. In doing so, this paper ‘brings in’ the media. News media do not often take center stage in the niche party literature. This is remarkable, as media likely shape all these parties’ public image. Vice versa, the literature on news media seldom takes mainstream versus niche party rivalry into account. FIELD OF GREENS 3 Our argument brings along the insight that mainstream parties may try to make the plight of a niche party even more daunting – or try to help the party. In analogy to Meguid’s (2005) theoretical framework, we argue that the default option for mainstream parties is to dismiss a niche party’s concerns, keeping its media visibility low. A party can deviate from that default by giving attention to the niche party’s core policy issue to hurt a mainstream rival. Given news values such as conflict combined with an internalized duty of balanced reporting, media may play along. If dismissal becomes untenable, mainstream parties may either adopt an accommodative tactic or adversarial one. This increases niche party visibility. Which niche parties should we study? The niche parties we select must be relatively homogenous, must vary considerably in media visibility, and must be present in several countries. It makes sense, then, to focus on successful niche parties. Of the two most successful niche party types, greens have the advantage that their core policy issue is unambiguously and abundantly clear. Moreover, environmental policy issues generally are valence issues, so that measuring their voter salience is less problematic than for, e.g., immigration. Furthermore, studying greens makes our contribution to the literature greater, because they are less often studied than right-wing niche parties in Western Europe.1 We test our hypotheses by comparing media visibility of green parties in 11 European countries. To do this, we collect two media datasets containing over 1.1 million articles on politics published in national and regional newspapers between 1992 and 2021. We combine this with data from the Comparative Manifesto Project to examine how mainstream parties respond to green issues through their policy platforms. We find evidence for the strategies 1 There is a growing literature on green parties and a few exceptions are studies by Carter (2006, 2013, and 2018), Kitschelt (1988), Rohrschneider (1993), Benoit and Laver (2006), Dalton (2009), Rüdig (2012), Rüdig and Sajuria (2020), and Abou-Chadi (2016), among others. FIELD OF GREENS 4 employed by mainstream parties affecting the visibility of green parties. This said, this effect is conditional on mainstream party ideology. Theory Issue competition between niche and mainstream parties Issue competition is defined as competition between political parties in which parties emphasize some policy issues over others (Petrocik 1996; Riker, 1993). In particular, parties compete to bring attention to issues that bring them the most reputational advantage. They may try to establish ownership over a particular issue (Petrocik, 1996; Budge and Farlie, 1983). They do this by building a strong association between the party and the issue as well as the perception that they are the most competent at addressing problems related to the issue (Walgrave et al., 2012). In contemporary Western Europe, for example, education and healthcare are typically associated with left-wing parties, while economic growth and military defense are typically associated with right-wing parties. Newly emerging issues, such as climate change and immigration have changed existing political dynamics and have allowed space for non-conventional actors to burst onto the political scene. The literature refers to the original mobilizers of these political ideas as niche parties, which relates to both the nature of the issue itself and its advocates. Meguid (2005: 347-348), for example, defined niche parties as those that “reject the traditional class-based orientation of politics” and “differentiate themselves by limiting their issue appeals”, while Adams et al. (2006: 513) categorized niche parties as parties that present a non-centrist ideology, commonly embraced by mainstream parties. Niche parties are also often ascribed labels such as ‘single-issue’ (Mudde, 1999), ‘anti-establishment’ (Abedi, 2004), or ‘challenger’ (De Vries and Hobolt, 2020). Green and far right parties are typically said to fall under this rubric, sometimes complemented with communists or regionalists. As advocates of niche FIELD OF GREENS 5 issues, niche parties are therefore considered to be their natural owners: environmentalism is associated with green parties, Euroscepticism and anti-immigration are associated with radical right parties, privacy and information freedom with Pirate Parties, and so on. Yet, issue ownership is fluid and parties can—and do—lose ownership over certain issues (Bélanger and Meguid, 2008; Green and Hobolt, 2008; Green-Pedersen and Mortensen, 2015; Hobolt and de Vries, 2015). Since mainstream parties tend to be vote maximizing, their main goal is to control the political agenda and bring media attention to issues that they have already owned. When media attention is elsewhere, parties are constrained both by their opposition and media logic in determining their issue competition strategy. Mainstream parties may risk issue trespassing, even at the risk of being accused of pandering, flip-flopping, or causing internal friction (Adams et al., 2006; Abou-Chadi et al., 2020). The success of this strategy depends on a few factors. Walgrave et al. (2009) argued that parties will only succeed in stealing owned issues when the original owner does not directly reclaim the issue. In the case of niche issues, niche parties are unlikely to let their ownership lapse. Mainstream parties must also consider that the focus on stealing non- owned issues leaves little room to highlight their own owned issues (Kaplan et al., 2006). Trespassing risks accentuating the niche competitor’s strength rather than their own and may alienate segments of their voters as a consequence. Niche parties, on the other hand, have less to win by changing their stance or scope of issues. Unlike mainstream parties, niche parties aim to build long term support and so rely on policy stability to seek votes (Kitschelt, 1994). In fact, Adams et al. (2006) find that when niche parties attempt to move closer to the median voter or broaden their policy programs, their national vote shares drop relative to their support in the previous election. Compromising their ideology and policy objectives may cause internal rifts and demobilization of the already stretched-thin party activists, which, if widely publicized, raises FIELD OF GREENS 6 questions on the party’s competence and reliability (Adams et al., 2006; Clark, 2005). It is more beneficial, then, for niche parties to reinforce their niche issue, both to signal their ownership over the issue in order to differentiate themselves from the catch-all mainstream party position and to defend themselves from issue trespassing by mainstream parties (Abou- Chadi et al., 2020). Meguid’s (2005) important work conceptualizes mainstream party responses toward niche parties as three distinct strategies. First, a mainstream party may choose to ignore or dismiss the niche issue entirely. Green-Pedersen (2012: 125) argues that mainstream parties may adopt a dismissive strategy if the issue negatively affects the coalition formation or if they offer little electoral benefit. For the niche party, the dismissal by mainstream parties reduces the salience of the party’s core issue, which in turn reduces the party’s electoral support (Meguid, 2005). Second, mainstream parties may opt for an adversarial strategy, where they take an issue position that is diametrically opposed to the niche party. Typically, the strategy is employed by mainstream parties on the opposite side of the political spectrum. The aim is to accentuate the policy difference of both sides, and to exploit the potential vote split between the niche party and the mainstream party to which they are most ideologically proximate. Niche parties additionally benefit from this, as a direct adversarial response also spotlights their issue position and increases their relevance in the political debate. Finally, mainstream parties can undermine the policy distinctiveness of niche parties by bringing attention to the issue and taking a similar issue position. Such an accommodative strategy is most appealing when niche parties are viewed as a larger electoral threat than their mainstream opposition, when an issue is highly salient, and when the niche party is ideologically proximate. The aim is to lure voters interested in niche issues, especially strategic voters, because mainstream parties have the advantage of policymaking and governance experience. While niche issues are explicitly not their owned issues, the success FIELD OF GREENS 7 of issue trespassing increases when they are able to reframe the issue to fit with their already existing policy program and offer a concrete solution via public policy (Adams et al., 2006; Spoon et al., 2014). News values and party competition News values have been used extensively to explain why certain issues and political actors receive more media attention than others (e.g., Hopmann et al., 2011; Tresch, 2009, Van Dalen, 2012). Galtung and Ruge (1965) identified 12 news factors—originally used to examine why certain foreign events are reported—such as the level of meaningfulness and intensity of an event, references to elite actors, unexpectedness, and presence of negativity. The authors argue that the more an event satisfies the criteria, the more likely it will be considered newsworthy. Furthermore, these factors will often be accentuated when writing the news story, as they are considered to be more valuable. Harcup and O’Neill (2001) revisited the original taxonomy and tailor it to fit the evaluation of general news. Through refinement and combination, they proposed that stories containing power elites, celebrities, entertainment, surprise, good and bad news, high magnitude and relevance, follow-up, and newspaper agenda to be the most applicable and applied news values by journalists. Since then, other studies have further refined these criteria, maintaining most values, such as power elite, negativity/conflict, and relevance, and added more practical aspects of journalist, such as cost of pursuing news stories and availability of source or materials (e.g., Brighton and Foy, 2007; Bednarek and Caple, 2017; Schultz, 2007). In terms of news values, mainstream parties hold particular advantages over niche parties. To see this, consider the process of creating a news story. Daily journalistic work is often fast-paced and reactionary, which reflects the importance of news values as a shorthand FIELD OF GREENS 8 (Harcup and O’Neill, 2001). In a study involving an experiment on political journalists, Helfer and Van Aelst (2016) found that the power status of the actor, unexpectedness, and magnitude of political actions are the most pronounced influences on journalistic decision making. Naturally, mainstream parties score highly on some of these factors as they tend to be on the top end of political power hierarchy and, being in positions of power, tend to produce actions that directly influence many citizens (e.g., Green-Pederson et al., 2015; Tresch, 2009). ‘Official’ government sources are allocated more space because they are typically perceived as a legitimate source of information through their direct involvement with policymaking, easily accessible, and relatively inexpensive to cover—though this space is similarly extended to main opposition parties (Bennett, 1990; Hopmann et al., 2012; Strömbäck and Van Aelst, 2013). From the media and audience perspective, relying on official sources lends greater credibility to the story itself. Mainstream parties are already recognizable to consumers, which to the media means that news stories in which they are featured would garner the most interest from audience. This in turn grants them habitual access to journalists (Molotch and Lester, 1974; Strömbäck and Van Aelst, 2013). Wolfsfeld (1997) argued that this pattern leads to elite actors in the system becoming more powerful, which he termed the ‘principle of cumulative inequality’. In their review of studies on variation in attention towards politicians, Vos (2014) confirmed that 17 out of 18 studies found significant positive relationships between the power position of politicians and their media visibility. Niche parties are up against established news routines, and against news values that generally do not work in their favor. They are rarely in positions of power; when they are, they tend to be part of a coalition which often forces them to compromise on some of their more ‘radical’ positions (Carter, 2013). Their usual absence from policymaking means that they have lower levels of issue competence attributed to them, further marginalizing them as FIELD OF GREENS 9 viable alternatives. Issues advocated by niche parties also tend to struggle to gain media attention and only do so to the extent that the issue is salient among news consumers. This reliance on issue salience limits niche parties’ capacity to stand out and they often must rely on disruptive access to the media to “make news”, either by breaking newsmaking routines or by generating surprise or shock (Molotch and Lester, 1974; Haselmayer et al., 2017; Helfer and Van Aelst, 2016). Jacobs and Van Spanje (2020), for example, found that outrageous statements from leaders of anti-immigration parties—which resulted in their prosecution for hate speech—led to higher short-term visibility of the parties and their leaders. A risky strategy, that may lead to loss of legitimacy in the eyes of news consumers. Unlike their mainstream counterparts, non-mainstream parties are often viewed with greater skepticism by the media. Van Spanje and Azrout (2022) found that new parties are trivialized more often than mainstream parties in the news. As such, some niche parties resort to ‘mainstreaming’ to shed their ‘other’ reputation by moderating their core position, expanding their issue agenda, breaking with controversial party members, and toning down their public communication strategy (Akkerman et al., 2016: 10). This has proven to work for France’s Front National under Marine Le Pen’s leadership. In addition to rebranding into Rassemblement National, Le Pen actively sought to “detoxify the party’s reputation” by distancing the party from its extreme nationalist image and rhetoric and adopting a broader social agenda (Ivaldi, 2016: 226; Surel, 2019). We focus on two mechanisms that conditions niche party visibility in the media. The first is based on the journalistic norm of balanced reporting. Although difficult to conceptualize (see Hopmann et al., 2011), balancing the news is seen as “[diminishing] the role of news values as the primary basis for story selection” (Semetko, 1996, p. 51). During non-election times, the “always get an official reaction” rule and focus on enacted policies and accountability often lead to greater attention on mainstream parties (Bennett, 1996; FIELD OF GREENS 10 Green-Pederson & Mortensen, 2010). During election period, however, the focus shifts from accountability to the candidates’ policies and future plans, which balances coverage of all parties (e.g., Van Aelst and De Swert, 2009; Sheafer and Wolsfeld, 2004). This provides more opportunities for niche parties to steal attention from mainstream parties and highlight their issues. The second mechanism is the media’s preference for conflict, one of the most prominent determinants of news coverage (Harcup and O’Neill, 2001; Walter and Vliegenthart, 2010). Recent studies found that political news reports tend to be more negative, which assigns higher values for negative or critical party communication (Hansen and Pedersen, 2008; Elmelund-Præstekær, 2010; Elmelund-Præstekær and Molgaard Svenson, 2014). Politicians are more likely to respond to critical stories, especially when they are attacked (Iyengar, 2011: 176; Elmelund-Præstekær, 2011). This is especially true in media systems characterized by high “infotainment” tendencies as negativity and conflict fulfill the demands for unexpectedness and excitement (Jerit, 2004: 563). News stories containing conflict are also more impactful because these are cognitively more “vivid” and contain more information on issues and actors (Bohle, 1986: 796; Soroka, 2014; Galtung and Ruge, 1965). For journalists personally, seeking opinions from opposing parties fulfils the “objectivity routine” and protects them from accusations of bias (Shoemaker and Reese, 1996: 107-109). Although journalists typically follow the trail of power—preferring sound bites from influential actors who are likely to affect the political outcome (Althaus, 2003)—niche parties can also benefit from conflict by bringing unexpected views and new ideas into policy debates. Journalists assign high news values to novelty and expertise, and conflict can often extend this platform to less powerful actors (Van Dalen, 2012). FIELD OF GREENS 11 Competing on green issues in the news Environmentalism developed from social movements in the 1960s and 1970s, which spurred the foundation of green parties across Europe during the 1980s and early 1990s (Carter, 2013). Some scholars argued that green issues and environmental protection policies cut across traditional left/right partisan lines, in particular because they appear in conflict with policy priorities that underpin these lines, such as economic growth and competitiveness (Knutsen, 1987; Dalton, 2008). Yet, as the environmental impact of deforestation, fossil fuels, and climate change has become a concern for more and more voters in recent decades, the pressure has increased for mainstream parties to respond to green issues and to depoliticize their actions in order to face these environmental challenges. Spoon et al. (2014: 372) found that mainstream parties “raise the stakes” as green issues become salient among voters and green parties fare better in elections. This is especially true for mainstream left parties, to whom green parties are a larger electoral threat than for mainstream right parties. On average, mainstream left parties are also more receptive to the environmental agenda than right parties though both sides have shown more responsiveness over the years (Rohrschneider, 1993; Carter, 2013). What is more revealing is how these parties address green issues in their broad policy platforms. Large scale environmental protection actions, for example, entail a more “fundamental restructuring of markets and more severe regulation of behavior than addressing any other environmental problem”— something that a mainstream party may consider to be diametrically opposed to their interests (Farstad, 2017: 699). As such, a party may instead pay lip-service to environmental protection by employing more cost-friendly approaches, such as banning single-use plastics, without having to ‘sacrifice’ other policies. In this case, green policies are still included in the party manifesto but not in a meaningfully effective way. Green parties may instead benefit from mainstream FIELD OF GREENS 12 parties’ lip-service environmentalism because they offer more comprehensive policies on the issue. Thus, mainstream parties must find a way to balance the adoption of green policies with the cost of reducing the impact of their owned policies. Regardless of degree of policy adoption, mainstream parties tend to take away media attention from green parties. News media are interested in political actors with greater policy impact but also prefer to construct news items that contain opposing points of view. Here, issue competition strategies take effect in two ways. First, when a mainstream party rejects (or, more likely, minimizes) green issues, news values favor green parties as the issue owner when constructing an article on green issues. However, we acknowledge that most mainstream parties—at least those in our sample—currently do not reject green issues outright. As such, the most common approach is the second competition strategy: mainstream parties pay greater attention to green issues and attempt to accommodate them. What we are specifically interested here is the extent of this accommodation. When a mainstream party chooses to claim green issues as their own, the media respond by elevating the mainstream party as a primary advocate because it is inherently a more valuable resource in terms of public interest and impact on policymaking (Walgrave et al., 2009; Helfer and Van Aelst, 2016). In sum, the salience other parties attach to green issues legitimzes a green party while offering it the opportunity to create media attention. It may do so on its own green issue, on which journalists are hypothesized to pit it against other parties given the news value of conflict and their internalized duty for balanced reporting. However, as mainstream parties become just as green as the green party, they may tend to crowd such a party out. Journalists are expected to replace the greens with mainstream parties who have gone green in such both-sides reporting. In accordance with this, we formulate the following hypotheses: FIELD OF GREENS 13 H1 Green party visibility increases as green policies are more salient among mainstream parties. H2 The saliency effect (H1) decreases as mainstream parties use more accommodative tactics. Empirical approach Media data To test our hypotheses, we collected media data comprised of newspaper articles on the subject of politics published in national and regional daily, weekly, and Sunday newspapers based in Western European countries.2 We accessed the newspaper articles through the Nexis Uni database. Newspapers were selected based on the geography and frequency of circulation with the aim to include the largest audience of politics articles as possible. In terms of geography, priority was given to newspapers with national circulation, followed by newspapers primarily published in the capital city, and finally, newspapers published in the top three largest cities in larger countries like Germany. For example, we included Berlin-based Der Tagesspiegel as well as Munich-based Süddeutsche Zeitung. We were, however, limited by the availability of the Nexis Uni database. For example, the French publication Le Monde was omitted by unavailability despite being one of the largest newspaper publications in the country. In total, we have 102 national and regional newspapers in our sample, although the variation between years depends on their inclusion in the Nexis Uni database. 2 For this study we excluded publications that focus primarily on non-general or non-politics topics, such as sports. FIELD OF GREENS 14 We searched specifically for articles discussing politics at national or provincial/state level while excluding local politics. Politics articles are operationalized as articles containing political issues (e.g., election, governance), policies (e.g., education, environmentalism), and political actors (e.g., prime ministers, political parties). To prevent inconsistencies in data collection, we used the root term politi* due to the similarity of the root when translated in all languages relevant to this study. In addition, we also included articles with politics subject labels. Again, we used politi* as the subject category. With the above mentioned parameters, we constructed two datasets. For Dataset 1 we collected articles within a period of three months leading up to each general election between 1992 and 2021. We focus on campaign periods in this dataset because the effects of news values is reduced during elections as journalists take on the responsibility to provide more equal coverage of all participating parties (Haselmayer et al., 2017; Green-Pedersen et al., 2015). Theoretically, this should result in greater coverage for niche parties such as greens, especially when the top election agenda points include environmental protection. Campaign periods are also windows in which mainstream parties are most likely to respond to changes in the electoral dynamics and to niche parties’ potential electoral threat. The first dataset contains over 700,000 articles from nine countries. To test our hypotheses, we aggregated these articles to 65 observations of country*election years. Dataset 2 is centred on the most important problem (MIP) survey item to gauge how public salience of green issues affect media coverage of the topic and subsequent interest in political actors mobilizing the issue. The data collection period follows that of Eurobarometer’s fieldwork beginning from the second half of the 2002 survey, where MIP FIELD OF GREENS 15 began to be included consistently in every survey, up to 2021.3 Eurobarometer is conducted twice a year. As a result, depending on data availability, our sample contains between 10 to 38 survey periods. More specifically, we collected articles from national and large regional newspapers published in the two week period after the end of the Eurobarometer fieldwork date. In total, we collected nearly 400,000 articles from ten countries, which we aggregated to a total of 318 country*survey observations. The data complements our election dataset by testing our hypotheses during non-election times, when media coverage are less balanced and mainstream parties have fewer incentives to discuss non-owned issues in the media (Van Aelst and De Swert, 2009; Abou-Chadi et al., 2020). Table 1 below shows the different countries sampled in each dataset as well as the total number of articles we collected for each country (please see Appendix A for a full list of newspapers and the number of articles collected per country per year). In total, we collected over 1.1 million articles. Of these, nearly 23% specifically discussed environmental issues, mainly driven by increased salience of climate change and international cooperation on the issue in recent years. Dependent variable We synthesized our variables by creating country-specific dictionaries that contain keywords relating to the green party, including the party’s name in the country’s language and its abbreviation. This includes references to ‘greens’ or ‘green party’ as a shorthand for the green party as well as coalitions that include them. We recorded mentions of 27 3 The most important problem item is only included consistently beginning from the second survey period in 2002. Due to Covid-19, the survey was only fielded once in 2020. FIELD OF GREENS 16 ideologically left green parties, including parties that dissolved and established partway through our data collection period, as well as the coalitions in which they are involved.4 Table 1. Descriptive statistics of newspaper articles and green party visibility Country Years Periods Articles Media data 1 Austria 2006 – 2019 5 11,212 Denmark 1998 – 2015 6 32,219 France 1993 – 2017 6 42,205 Germany 1994 – 2021 8 135,041 Ireland 1992 – 2020 7 31,998 Italy 1992 – 2018 8 60,135 Netherlands 1994 – 2021 9 129, 674 Spain 2000 – 2019 8 117,066 United Kingdom 1992 – 2019 8 143,017 Total 65 702,477 Media data 2 Austria 2004 – 2021 34 10,261 Denmark 2002 – 2017 29 31,843 France 2002 – 2021 38 61,759 Germany 2002 – 2021 38 99,892 Ireland 2002 – 2021 38 34,118 Italy 2002 – 2021 38 43,418 Luxembourg 2016 – 2021 10 2,462 Netherlands 2002 – 2021 38 57,949 Portugal 2012 – 2021 17 2,453 United Kingdom 2002 – 2021 38 55,041 Total 318 399,196 There were two aspects of green party appearances in the news that we focused on. We calculate Presence and Mentions. The first is the operationalized as the number of articles in which green parties are present proportional to the overall number of articles published in each observation. Presence represents the inclusion of green party voices in the media. The second variable, Mentions, captures the number of mentions of green parties or green party members in each article proportional to the total number of mentions of all 4 A full list of political parties, including a timeline of merger, splits, and dissolution, can be found in Appendix C. FIELD OF GREENS 17 parties.5 Approaching visibility by looking at mentions, we are able to observe which party is emphasized in the news—in other words, whether mainstream parties are able to “drown out” green parties by dominating the media narrative. On average, green parties appear in 6% of the articles we sampled and mentioned 13% of the time in Dataset 1. In Dataset 2, green parties are present in 7% of articles and received 15% of mentions. Independent variables Our independent variables are salience and relative positioning of green policies in mainstream parties’ manifesto. We used manifesto as a proxy for mainstream parties’ issue competition strategy, as the addition of specific policies in mainstream parties’ manifesto signals a serious response to said issues (Brouard et al., 2018). Party manifestoes are linked to the legislative agenda—increasing attention to a particular policy also increases the likelihood that a law corresponding to the issue will be adopted once the party becomes part of the government. As such, green policies outlined in the manifesto would also be integrated into the greater campaign communication agenda to convey their issue stance to the media and the public. We used data from the Comparative Manifesto Project to estimate our independent variable. CMP coded policy preferences of over 1,000 parties since 1945 in over 50 countries in five continents and is updated twice a year. Changes to manifestos tend to coincide with every election year, as parties emphasize and de-emphasize certain policies in their issue agenda in accordance with their vote maximizing strategy. 5 We have also examined effects by left and right party. However, there is no theoretical reason why the one or the other would have a different impact on the dynamics concerned. FIELD OF GREENS 18 To estimate the extent to which green policies are salient in mainstream party manifestoes, we used the CMP variable environmental protection (501), which covers statements related to “protecting the environment fighting climate change, general preservation of natural resources, preservation of countryside or forests, protection of national parks, and animal rights.”6 In such statements, topics like nuclear energy are not included, likely because supporters of green issues are divided on the subject and most green parties are reluctant to support nuclear power as an alternative to fossil fuels (Carter, 2013). We used the logged percentage of statements falling into this category as a measure of salience of green policies in the manifesto. We applied Meguid’s (2005) framework of issue competition to estimate the level of accommodation of green issues within a party manifesto. To do this, we also consider the economic aspect of support for green issues based on recommendations from Meguid (2005) and Lowe et al. (2011: 137). Parties more accommodating toward green issues combine support for environmental protection with support for anti-growth economy and sustainability (416).7 Parties less accommodating toward green issues is estimated using support for economic growth (410) as a contrast to green issue support.8 Lowe et al.’s (2011) argued that policy issues like pro-environmentalism are two-sided and require paired categories to be estimated together to gauge the relative balance, or the proportional changes on a symmetrical pro or against. For example, pro-growth policy stances often indicate “thinly veiled 6 Based on CMP coding scheme version 5. 7 CMP describes the variable as statements that include “favorable mentions of anti-growth politics” and “call for sustainable economic development”. 8 CMP describes the variables as statements that include “paradigm of economic growth”, “general need to encourage or facilitate greater production” and the “need for the government to take measures to aid economic growth”. FIELD OF GREENS 19 antienvironmental messages”, which reflects the unwillingness of some mainstream parties to make way for more radical environmental policies (Lowe et al., 2011: 137). This estimation provides a better representation of party stance on a particular issue. We estimate the relative positioning of green policies as: 𝐺 +.5 𝜃 𝑃 = 𝑙𝑜𝑔 𝐵 +.5 where 𝐺 indicates combined percentage share of support for green issues (501 and 416) and 𝐵 indicates percentage share of ‘opposition’ to green issues (410), as explained above. This resulted in a continuous scale measuring the level of accommodation of green policies in mainstream party manifesto. The focus on accommodation is motivated by the fact that mainstream parties are increasingly more receptive to the inclusion of green policies in their policy agenda. Controls To test the robustness of our estimations, we included three controls. Our first control is issue salience, measured using the Most Important Problem (MIP) item from Eurobarometer. Specifically, we are interested in the share of respondents who state that environmental issues, such as pollution and climate change, are the most important problem facing the country at the moment. The assumption is that as green issue salience increases, we should also see greater relevance, and therefore higher news values, of green parties by extension. Our second set of control is the green party electoral performance, taken from the CMP database and from official national election statistics. Performance is measured as the percentage vote share of green parties from the previous election and we expect that an increase of vote share would have a positive effect on green parties’ media attention. Finally, FIELD OF GREENS 20 our third control is newspaper bias, which estimates the percentage of articles in each election year published by left-leaning newspapers.9 We expect that left-leaning newspapers would provide greater coverage of green issues and more balanced coverage for green parties. Appendix B shows the descriptive statistics for both datasets. Results and analysis To test our hypotheses, we conduct OLS regression analysis to estimate the effects of salience (H1) and its interaction with accommodation (H2) on green party visibility. To increase the confidence in our findings, we use two versions of our dependent variable and we base our analyses on two datasets. We thus present four series of models, each regressing media visibility of greens on salience (Models 1), salience and accommodation (Models 2) and their interaction (Models 3) plus our three controls. As the observations are not entirely independent within country and within time points, we use clustered robust standard errors by country and by year. See Table 2 for the results based on our first dataset, which focuses on election campaign periods. As can be seen from Table 2, there is no empirical support for H1 based on Dataset 1. There is no indication of a salience effect in either of the two ‘Models 1.’ Turning to H2, the interaction between salience and accommodation is significantly negative, as expected (see both of the ‘Models 3’). However, we have to plot the effect of salience on green party visibility against empirically observed values of accommodation to assess its significance across these values. For illustrations of the salience effect across values of accommodation, see Figures 1 and 2. 9 We use expert analyses from Media Bias/Fact Check and Media Landscapes.  FIELD OF GREENS 21 Table 2. OLS regression of the effect of green policy salience and positioning on green party visibility (Dataset 1) DV = Presence DV = Mentions Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Salience -0.001 0.002 0.019* 0.003 0.004 0.034* (0.005) (0.004) (0.008) (0.009) (0.009) (0.014) Accommodation -0.010 0.012 -0.004 0.032* (0.007) (0.010) (0.015) (0.015) Salience * Accommodation -0.019* -0.033* (0.008) (0.015) Newspaper bias 0.010 0.007 0.016 0.117*** 0.116** 0.132*** (0.017) (0.017) (0.018) (0.030) (0.032) (0.029) MIP 0.002 0.002 0.002† 0.004*** 0.004*** 0.004*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Green vote share (lagged) 0.005** 0.005** 0.005** 0.018*** 0.018*** 0.019*** (0.002) (0.002) (0.002) (0.003) (0.003) (0.003) Constant 0.023† 0.030* 0.007 -0.046* -0.043 -0.083** (0.011) (0.014) (0.017) (0.023) (0.029) (0.026) F 4.90** 4.28** 4.21** 50.04*** 40.40*** 35.68*** R2.26.28.33.63.63.66 Note: Clustered standard errors by country and year. N = 65. p <.001, p <.01, p <.05, † p <.1. *** ** * Salience and accommodation are logged variables. Figure 1 and 2 suggest that there is a salience effect but is depends on accommodation. If accommodation is low, indicating an adversarial position, the salience effect is positive. This means that, in that case, more salience of green issues in mainstream party manifestoes leads to more visibility of green parties in the media. If accommodation is high, i.e., mainstream parties being green, more salience in mainstream manifestoes comes with less green party visibility. This suggests that mainstream parties are able to steal a green party’s thunder. However, only in terms of presence (on the right-hand side of Figure 1) the effect becomes negative according to conventional levels of statistical significance, not relative to other parties’ mentions (on the right-hand side of Figure 2). FIELD OF GREENS 22 Figure 1. The salience effect on the presence of green party visibility as accommodation values change in Dataset 1 (election campaign dataset). Figure 2. The salience effect on the mentions of green party visibility as accommodation values change in Dataset 1 (election campaign dataset). We repeat these analyses based on Dataset 2, geared towards non-election periods. See Table 3. When looking at the findings based on Dataset 2, a few differences become clear. First, the expected salience effect pops up on the presence of green parties in the newspapers, relating to presence (H1), although only marginally significant. Second, only  FIELD OF GREENS 23 one of the interaction effects is significant (H3). This effect also pertains to the values, not the relative mentions. We also see this when plotting the effects, See Figures 3 and 4. Table 3. OLS regression of the effect of green policy salience and positioning on green party visibility (Dataset2) DV = Presence DV = Mentions Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Salience 0.003† 0.003† 0.025 *** -0.006 -0.008 * -0.003 (0.002) (0.001) (0.004) (0.004) (0.004) (0.011) Accommodation 0.003 0.025*** 0.017* 0.022* (0.004) (0.004) (0.007) (0.009) Salience * Accommodation -0.021*** -0.005 (0.004) (0.008) Newspaper bias -0.003 -0.003 0.002 0.101*** 0.104*** 0.105*** (0.007) (0.007) (0.007) (0.013) (0.014) (0.013) MIP 0.001** 0.001* 0.001** 0.003*** 0.003*** 0.003*** (0.000) (0.00) (0.000) (0.000) (0.000) (0.001) Green vote share (lagged) 0.005*** 0.006*** 0.006*** 0.013*** 0.014*** 0.014*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Constant 0.019*** 0.016** -0.010 0.016† -0.001 -0.007 (0.004) (0.006) (0.006) (0.009) (0.013) (0.015) F 25.00*** 20.16*** 22.00*** 68.84*** 55.16*** 46.13*** R2.26.27.32.52.53.53 Note: Clustered standard errors by country and year. N = 318. *** p <.001, ** p <.01, * p <.05, † p <.1. Salience and accommodation are logged variables. Figure 3. The salience effect on the presence of green party visibility as accommodation values change in Dataset 2 (non-election dataset). FIELD OF GREENS 24 Figure 4. The salience effect on the mentions of green party visibility as accommodation values change in Dataset 2 (non-election dataset). Discussion In this paper, we have looked at niche party visibility in news media. This complements earlier work on niche parties, this time focusing on green parties. To explain green party visibility, we marry the literature of party competition to the literature on news values. We contribute to the literature by theorizing about effects that mainstream party strategy might have on green party visibility. Based on over 1.1 million articles in newspapers in 11 countries, we compiled two datasets, one focused on election periods and one on all kinds of periods. Using two different variables on the basis of two different datasets we tested two hypotheses: that salience of green issues in mainstream party manifestoes leads to more green party visibility (H1) and that this salience effect is weaker the greener these mainstream parties become (H2). Whereas we found marginal support for that general salience effect in one out of four analyses, the interaction effect was supported in three out of four. FIELD OF GREENS 25 In particular, we found no significant interaction effect on the mentions of green parties relative to mentions of other parties in Dataset 2, which was based on non-election time periods. Is the combination of the specific operationalization of our dependent variable in non-election times a special case that sets it apart from the rest? We already referred to previous research that found that in non-election times journalists tend to focus more on enacted policies and accountability (Bennett, 1996; Green-Pederson & Mortensen, 2010). This leads to a situation where government officials, who are usually not members of a green party, are often asked for formal reactions on issues that are directly related to their portfolio. So although we find that also in no—election times green parties are more present in the media when mainstream parties do attach salience to green issues but do not accommodate the green party position, this effect is likely dampened when we focus on the mentions relative to other parties by the need to include the government parties. But in election times, i.e., Dataset 1, as the focus shifts from accountability to future plans, the formal reaction from government officials is not necessary anymore, leading the effects to be more pronounced with both operationalizations. This study comes with quite a few limitations. First, the scope of the research is not unlimited, but restricted to about three decades in 11 countries and only a few parties in each of these countries. Related to this, we have investigated newspapers, even though the media landscape is obviously much richer, and has become more diverse across the period under study. To our defence, the scope of our research is wider than that of many comparable studies. In addition, we have not examined changes in effects over the time span of the investigation. Similarly, the use of party manifestoes is not beyond criticism. This said, using them as an indicator of party salience is arguably less problematic than using them in other ways. The measurement of the control variables can also easily be FIELD OF GREENS 26 criticized, including the voter salience that we distill from mass surveys. Yet, the Eurobarometer is a high-quality that is available for cross-national and cross-time. Notwithstanding these and other limitations of our research, we can speculate about the implications of these findings. It looks like mainstream party strategy has effects on niche party media visibility similar to that on niche party electoral performance (see Meguid, 2005). These similar, or perhaps parallel, effects on party visibility in the media and electoral performance lead to the question on the causal ordering of the two. Do mainstream strategies affect niche party visibility, and does visibility affect niche party electoral performance? Or do the mainstream party strategies affect the standing in the polls and perceptions of success of niche parties among journalists, which in turn affects media attention towards the niche party? In our analyses explaining media attention, we control for prior electoral performance. However, this past performance of one electoral cycle ago does not capture the dynamics of the present cycle, leaving the causal question partly unresolved. Irrespective of such causal order, these effects suggest that mainstream parties can reduce niche parties’ media and electoral success. Following from this, one could argue this may be problematic for a healthy democracy. A main reason for this is that niche parties are key to introducing new political ideas. However, the question is to what extent new ideas are being stopped in this way. Our findings indicate that by ignoring its core issue mainstream parties could reduce the success of a niche party, but our models show consistently that a niche party tends to benefit from salience of the issue among the general public. No matter the mainstream party response, salience drives media success for niche parties. When mainstream parties challenge the niche party position, this leads to additional visibility for niche parties. When they accommodate, the niche party may be unsuccessful but the political idea nonetheless receives much coverage. In that case, the niche party can FIELD OF GREENS 27 be deemed successful, not so much in achieving media success for itself but in achieving media success for their idea. Finally, what directions should future studies take? We believe that we have opened a new avenue for future research. The role of news media in niche party success or failure has been understudied. Moreover, among these questions related to media, the question of how mainstream party strategy affects media visibility of niche parties has, to the best of our knowledge, never been tackled. We have taken a few modest steps onto this new path by using two dependent variables based on two datasets we assembled ourselves. Besides widening the scope of such exercises (taking into account more media outlet types, more parties, more countries and more time points), next steps should go beyond visibility to also include evaluative tone and news media framing of niche parties. And also, voter reactions to such changes in news media coverage of niche parties must be examined. This is because, in the end, the fate of niche parties is determined not by media, but by voters.