Abstract
How parties compete by utilizing frames in the public debate is not well understood. A widespread understanding is that parties tend to compete by talking past each other, i.e. when they utilize dissimilar frames. This article relies on a unique data set of parties’ framing of four individual policy questions over two decades in Denmark to examine frame overlap, i.e. when parties adopt similar frames. Results show that parties utilize similar and dissimilar frames in the public debate. However, a high degree of frame similarity was generally found across parliamentary blocs (between the leading mainstream party from each bloc, their leading junior coalition partner, and the two blocs) and most often a very high degree of overlap was found within parliamentary blocs (between the leading mainstream party and its junior coalition partner). This result suggests a need to rethink thoroughly how to understand and study interaction among parties utilizing frames.
Subtle differences in how policy questions are framed can have a substantial impact on public opinion formation and public policy. The understandings or frames ascribed to policy questions influence the amendments that are possible to combat existing policy problems (e.g. Baumgartner et al., 2009; Riker, 1986). At the same time, numerous studies have shown that citizens are likely to form opinions based on the frames they receive (e.g. Chong and Druckman, 2007; Slothuus, 2010). Although a central aspect of party competition is to define the understanding or framing of policy questions, little attention has been devoted to this subject.
The limited literature on party frame competition focuses on how individual parties utilize different frames in the public debate (e.g. Kriesi and Hänggli, 2019; Helbling et al., 2010). Within this literature it is common to conclude that ideological constraints govern how parties utilize frames, which has lead us to expect that parties talk past each other in the public debate (Sniderman and Theriault, 2004). When citizens can choose between alternatives, it seems reasonable, and preferable, that they can make informed choices about who should govern the country. As Schattschneider (1960: 137–138) famously argued: “Above everything, the people are powerless if the political enterprise is not competitive”. Still, there are good reasons to expect that countries are governed by a certain level of stability reflected in at least some shared understandings of political matters. As Baumgartner et al. (2009: 176) formulate it: “In the policy process, there is virtually always an established, status quo frame that dominates discussion.” There are thus also good reasons to expect that parties not only talk past each other but at least to some extent share similar understandings. However, little attention has been devoted to the concept of frame overlap, i.e. when parties adopt similar frames (see similar concerns raised by Jerit, 2008 and van de Wardt, 2015).
The idea of overlap in parties’ communication is far from new. From the literature on issue competition, we know that parties prefer to talk about issues that are advantageous to them, e.g. “owned” issues on which they have a better reputation than their competitors (Petrocik, 1996; Lefevere et al., 2015). However, parties cannot simply pick and choose among issues and only talk about their favorites. Research has shown that they converge on similar issues (Green-Pedersen and Mortensen, 2010; Sigelman and Buell, 2004), but the question is why they should utilize similar frames.
One plausible explanation is that parties simply act strategically and employ a vote-seeking strategy. Parties’ drive to broaden their electoral base may lead them to tailor their frames to match those of the median voter to win office, especially leading mainstream parties competing for governing power. Consequently, parties may tend to echo each other’s understandings of policy questions rather than offer clear alternatives. Another potential explanation is that some understandings of reality hardly can be ignored within a given context, which leads to a certain level of stability in how policy questions are understood. An alternative idea is that a systemic logic governs the formation of party frame competition. Parties (and other elite actors) adopt certain frames because other parties (and elite actors) do so, while struggling over how to understand the policy question in the future. During this process, frame overlap emerges where parties emphasize one and/or a limited set of similar frames in the debate, while the vast majority of frames receive limited attention.
This article examines the extent to which political parties utilize similar frames. As we shall see, evidence on how parties compete is surprisingly sparse and focuses almost exclusively on individual party frame strategy rather than interaction among parties. Moreover, most studies examine a single question within a limited context, which makes it difficult to separate question- and context-specific dynamics from more general dynamics in party frame competition. This article relies on a unique and comprehensive data set covering parties’ framing of four individual policy questions—family reunification, the early retirement scheme, green taxes, and prostitution—in two broadsheet newspapers over two decades in Denmark to examine the extent of party frame overlap. Media data was used to systematically explore how parties utilize frames in the public debate, as there is little scholarly dispute that the news media is the most important arena for public debate in contemporary society (Ferree et al., 2002). News stories are important for parties, as they are the public’s primary source of common knowledge and information (Neuman et al., 1992). News media data is, thus, valuable for understanding and examining the concept of frame overlap.
Perspectives on party frame competition
When parties compete over frames, they emphasize certain understandings or interpretations of reality rather than others (Chong and Druckman, 2007; de Vreese, 2012; Entman, 1993). In this article, a frame refers conceptually to a “central organizing idea” about a selected part of reality (Gamson and Modigliani, 1989: 143). My analyses are motivated by two quite different perspectives—an ideological and a trespassing—on how the framing process in the public debate unfolds and what the outcome of this process is, e.g. whether we would expect frame dissimilarity or frame similarity in the public debate.
On the one hand, one of the most widely accepted understandings is that ideological beliefs constrain how parties utilize frames, as Hoeglinger (2012: 150) explains: “politicians are seriously constrained in their framing efforts by their ideological profile”. Or, as Sniderman and Theriault (2004: 141) put it: “political parties and candidates are not free to frame issues however they would wish. The party that puts itself forward as the guardian of order is constrained to make use of electoral appeals that are consistent with its commitment to order; the party that puts itself forward as the advocate of civil liberties is similarly constrained by its commitment to civil rights” (see also Sniderman, 2000). In a similar vein, some scholars apply a more competitive ideological perspective (Chong, 1996) and argue that parties may feel forced to react to a successful frame from their opponents by counter-framing; that is, they engage on the same frame dimension but with opposite perspectives (Chong and Druckman, 2013; Hänggli, 2020; Jerit, 2008). As an example, the Democrats may worry about innocent people on death row, while the Republicans may argue that wrongful convictions are overstated (Baumgartner et al., 2008). Both parties talk about the dimension of innocence (aggregated level) but utilize two different frames (disaggregated level) that offer two opposite perspectives on the innocence dimension.
On the other hand, a handful of scholars have recently begun to focus on the trespassing or the “riding the wave” perspective (Sides, 2006), which means that parties use a frame if it is “well supported and accepted by the general public” (Hänggli, 2020: 29; van de Wardt, 2015). The fundamental premise of this perspective is the existence of frame overlap in which parties naturally gravitate toward certain frames and away from others. That is when parties utilize similar frames, e.g. if both Democrats and Republicans worry about innocent people on death row.
Before offering my own analysis and reflections on frame similarity and dissimilarity, we need to assess the existing sparse evidence. Most empirical research in the field has focused on which frames individual parties emphasize in the debate and found that parties’ ideological beliefs are an important benchmark for understanding how they utilize frames. Examples are Helbling et al.’s (2010) study of the European integration issue in six Western European countries from 2004 to 2006, and Kriesi and Hänggli’s (2019) study of the unemployment issue in six European countries during 3 months in autumn 2010. Only a handful of empirical studies reflect on how often parties adopt similar frames, and even fewer provide a measure for the interaction among parties.
Hänggli (2020), however, found trespassing to be highly uncommon among rival parties across parliamentary blocs during three Swiss direct-democratic election campaigns in 2006 and 2008 (see also Hänggli and Kriesi, 2010), which may be less surprising as rival parties in such campaigns fight for clearly opposite policy implications. Van de Wardt (2015) found parties’ framing distance to be low (i.e., their frame overlap to be high) to all other parties on the European integration issue in party manifestos in the UK, the Netherlands, and Germany from 1987 to 2006. However, van de Wardt focused on the distance to all other parties, which one could imagine being smaller (and hence the overlap score higher), as he includes parties within the same parliamentary bloc sharing similar ideological beliefs. In comparison, Hänggli (2020) only examined rival parties across parliamentary blocs. Besides, van de Wardt examined rather aggregated frames (e.g. frame dimensions), which does not allow him to distinguish between a trespassing and a counter-framing strategy, which may further boost the level of overlap. Finally, most studies in the field examine a single policy question, like European integration or unemployment (Helbling et al., 2010; Kriesi and Hänggli, 2019; van de Wardt, 2015), and a limited context, such as an election campaign or a few months (Hänggli, 2020; Kriesi and Hänggli, 2019). This makes it difficult to separate policy question- and context-specific dynamics from more general dynamics in party frame competition. Although the results of these studies are quite informative, they do not offer a more general understanding of the extent to which parties utilize similar frames across policy questions over time.
Frame similarity or dissimilarity?
Why should parties with (more or less) different ideological beliefs utilize similar frames? One plausible explanation is that parties simply act strategically and employ a vote-seeking strategy. Given that research has shown that framing effectively influences public policy and public opinion formation (e.g. Baumgartner et al., 2009; Chong and Druckman, 2007), it is hardly controversial that parties seek to utilize frames strategically to achieve policy goals and gain votes (Strom, 1990). Although parties may prefer to utilize dissimilar frames constrained by their ideological belief (Helbing et al., 2010; Kriesi and Hänggli, 2019) to influence policy, they also want to maximize their vote share. As Downs (1957: 28) explained, “parties formulate policies to win elections, rather than win election in order to formulate policies.” Ultimately, parties compete over the median voter and which parliamentary bloc gains office. Especially mainstream parties (i.e. traditional mass parties dominating governing formation like the Social Democrats and the Liberals) have been argued to focus more on gaining electoral majority than niche parties (i.e. parties emerging around a particular issue like anti-immigrant right-wing parties) (Green-Pedersen et al., 2014). Parties’ drive to broaden their electoral base may lead them to tailor their frames to match those of the median voter, especially the leading mainstream parties competing for governing power. Consequently, they may tend to echo each other’s understandings of policy questions rather than offer clear alternatives.
Another potential explanation for the prevalence of frame similarity is that some understandings of reality hardly can be ignored within a given context. Parties do not frame within a vacuum, but in a greater social and political context that influence which frames that are constructed at a given point in time (Davis, 1990; Wolfsfeld 1997; Gamson and Modigliani 1989). The context in which parties utilize frames may make some frames more applicable than others, which may lead to a certain level of stability in how policy questions are understood (see also Baumgartner and Jones, 1993; Baumgartner and Mahoney, 2008; Jerit, 2008).
An alternative idea is that a systemic logic governs the formation of party frame competition. Green-Pedersen and Mortensen (2010) argue that the agenda is a structural phenomenon, and that relevant actors at any given time address the issues that are prominent, while they still compete over the future composition of the agenda. Similarly, the news media agenda is governed by a systemic logic. News criteria (e.g. Harcup and O’Neill, 2001) make media organizations follow each other, leading to homogeneity in the news production (e.g. Cook, 1998). The competition over frames may likewise be considered a structural phenomenon in the sense that parties (and other elite actors) adopt certain frames because other parties (and elite actors) do so, while struggling over how to understand the policy question in the future. During this process, frame overlap emerges where parties emphasize one and/or a limited set of similar frames in the debate, while the vast majority of frames receive limited attention.
This article asks the following research question: To which extent do political parties utilize similar frames? (RQ1). The article both examines overlap among parties within and across parliamentary blocs, as differences might appear as already hinted. On the one hand, it seems reasonable to expect that parties within rather than across parliamentary blocs share more similar ideological beliefs and therefore will rely more on similar frames. On the other hand, it could be argued that although parties within a parliamentary bloc share similar ideological views, they are each other’s closest allies as well as competitors (Downs, 1957), which requires them to stand out from one another. Therefore, they might use less similar frames than parties across parliamentary blocs who focus on capturing the median voter, for instance. The extent to which parties both within and across parliamentary blocs utilize similar frames will be examined.
Research design
The degree to which parties (within and across parliamentary blocs) utilize similar frames is examined based on a novel research design that covers the frames and the actors sponsoring them across four individual policy questions over two decades in Denmark. The paragraphs below elaborate on the design.
The second step was deductive coding of the news articles. The unit of analysis is frames, and the primary frame in each article was coded (e.g. Kellstedt, 2003), 4 which was identified as the first frame in the article to limit intersubjectivity and because it generally reflects the lead of the story. Journalists learn to write stories in the inverted pyramid, where the most important piece of information is placed at the top to quickly reveal the conclusion to the audience (Shoemaker and Reese, 1996). A policy question was frequently mentioned in a news article without a policy-specific frame, and the number of identified frames was thus n = 1115 for family reunification, n = 763 for the early retirement scheme, n = 1252 for green taxes, and n = 688 for prostitution. To illustrate the coding, the “human rights” frame pertinent to family reunification, which focused on humanitarian concerns and unification with one’s family as a human right, is utilized in the following news excerpt: “Farum’s mayor, Peter Brixtofte (V), attacks his party and denies that there is a need to tighten rules on family reunification. The state should not interfere in whom people can marry, he says” (Politiken, 2001).
The actor(s) sponsoring a frame was coded when a frame was explicitly present in a direct quote from the actor(s) or paraphrased. The coding of the actors included established political parties and their members (the red bloc consists of the Red-Green Alliance, the Socialist People’s Party, the Social Democrats, the Social Liberals, the Centre Democrats, and the Christian Democrats; the blue bloc consists of New Alliance, Liberal Alliance, the Conservatives, the Liberals, the Danish People’s Party, and the Danish Progress Party), another political party/politician, or another actor. 5 As mentioned, Danish politics is organized around blocs of left- and right-wing parties (Green-Pedersen and Thomsen, 2005) generally led by the Social Democrats (left/red) and the Liberals (right/blue), supported mainly by their main junior coalition partners, respectively the Social Liberals (left/red) and the Conservatives (right/blue). 6 The parties hold different ideologies and values—which should affect how they prefer to frame (Sniderman and Theriault, 2004)—reflected in different left-right scores: 3.8 for the Social Democrats, 4.9 for the Social Liberals, 7.2 for the Conservatives, and 7.3 for the Liberals (Döring and Manow, 2019).
Results
To examine the extent to which parties utilize similar frames (RQ1), the analysis focuses on overlap among parties across and within the same parliamentary bloc. A general overview is provided at the end.
To perform a robust test of frame overlap among parties across parliamentary blocs, frame overlap was examined between the leading mainstream party from each parliamentary bloc (the Social Democrats (red/left) and the Liberals (blue/right)), between their leading junior coalition party (the Social Liberals (red/left) and the Conservatives (blue/right)), and between the blocs in general (red/left and blue/right) for each of the for policy questions, respectively the early retirement scheme, family reunification, green taxes, and prostitution. Table 1 lists the frame overlap scores that express a mean calculated on an annual basis and using 3-year moving average to account for a small number of observations in some years. If we look at the frame overlap scores for how similarly rival parties across parliamentary blocs understand the policy questions, two patterns should be immediately clear.
Frame overlap across parliamentary blocs.
Note: The number of utilized frames from 1979/1983 to 1989 is calculated as an average to account for the lack of access to newspaper articles in Berlingske Tidende during this period.
First, frame overlap is a pronounced feature of how rival parties across parliamentary blocs interact. There is generally a high degree of frame similarity among rival parties both on an annual basis and by using 3-year moving average especially on prostitution (70%, 70%) followed by family reunification (59%, 59%), green taxes (56%, 55%), and the early retirement scheme (51%, 48%). In the competition for power, parties utilize not only dissimilar frames, but they also to a high extent share similar understandings of policy questions. Second, there are differences across (groups of) parties and policy questions. On average there is a very high degree of frame overlap in the debate about prostitution (70%, 70%), and a lower but still high degree of overlap in the debate about the early retirement scheme (51%, 48%). Most noticeable is the lower degree of overlap among the leading mainstream parties in the debate about the early retirement scheme (37%, 35%) compared to among their junior coalition partners (65%, 62%) and the two parliamentary blocs in general (50%, 48%). There might be strategic reasons why we found these variations across policy questions and across parties, which will be discussed later.
The variation in the degree of similarity does not change the general picture that despite ideological differences, rival parties across parliamentary blocs both talk past one another (utilizing dissimilar frames) as well as to a rather high degree utilize similar frames in the public debate.
Next, we examine the extent to which parties within the same parliamentary bloc utilize similar frames by focusing on overlap among the leading right-wing parties, the Liberals and the Conservatives, and the leading left-wing parties, the Social Democrats and the Social Liberals, on each of the four policy questions as illustrated in Table 2. The frame overlap scores express a mean calculated on an annual basis and using 3-year moving average to account for a small number of observations in some years. Two patterns should be immediately clear.
Frame overlap within parliamentary blocs.
Note: The number of utilized frames from 1979/1983 to 1989 is calculated as an average to account for the lack of access to newspaper articles in Berlingske Tidende during this period.
First, a very high degree of frame similarity was found especially in the debates about prostitution (75%, 74%), family reunification (70%, 69%), and green taxes (68%, 68%) followed by the debate about the early retirement scheme (62%, 64%). Coming from the same parliamentary bloc, these parties most often utilize similar frames to a very high extent, which seems to fit well with the idea that parties with similar ideological viewpoints (often) share similar understandings of policy questions. Second, there are some differences across the two blocs. The right-wing parties have a higher degree of frame overlap than the left-wing parties in the debates about the early retirement scheme (71% and 74% versus 54% and 55%), family reunification (79% and 78% versus 61% and 59%), and prostitution (80% and 78% versus 70% and 69%), while the left-wing parties have a higher overlap in the debate about green taxes than the right-wing parties (71% and 72% versus 66% and 65%). Although the parties within a parliamentary bloc share similar understandings of policy questions the extent to which they do so varies. There might be strategic as well as ideological reasons for these variations, as they might want to hold a more unique profile on some questions.
Nevertheless, the results show overall that parties within in the same parliamentary blocs often utilize similar frames to a very high extent.
Table 3 provides an overview of the results by listing the average frame overlap scores on an annual basis and using 3-year moving average among parties across and within parliamentary blocs, and a total score for parties across and within blocs on each of the four policy questions. Three patterns should be immediately clear. First, the results clearly show that both parties within and across the same parliamentary bloc not only talk past each other in the public debate but also share similar understandings to a high extent (64%, 63%). Second, parties within the same bloc tend to overlap slightly more (69%, 69%) than parties across blocs (often between 59%, 58%). Third, although the results are rather robust, some variations were found in the degree of overlap across policy questions and across parties. The extent of overlap was frequently very high in the prostitution debate and sometimes lower in the early retirement debate (especially for the leading mainstream parties across blocs, Table 1). The findings will be discussed below.
Overview of frame overlap among parties across and within parliamentary blocs.
Note: The number of utilized frames from 1979/1983 to 1989 is calculated as an average to account for the lack of access to newspaper articles in Berlingske Tidende during this period.
Discussion
A widespread understanding is that parties tend to compete over policy questions by talking past each other, i.e. when they utilize dissimilar frames (e.g. Sniderman and Theriault, 2004). This article examines an alternative idea, namely frame overlap, i.e. when parties utilize similar frames, by using a unique and comprehensive data set covering parties’ framing of family reunification, the early retirement scheme, green taxes, and prostitution in two broadsheet newspapers for two decades in Denmark. Most importantly the results show that parties are very similar, at least in communicative terms. Parties utilize both similar and dissimilar frames in the public debate. A high and very high degree of frame overlap was generally found among parties across different parliamentary blocs and among parties within the same bloc, respectively. Based on the results, this study suggests that we need to recognize that frame overlap is a much more stable element in party frame competition than is commonly assumed.
In a broader perspective, the findings have important implications for our understanding of the state of democracy and party behavior. Party communication cannot simply be described as conflict between parties offering different understandings of policy questions; it also involves more consensual elements than commonly assumed. Parties offer citizens alternatives (e.g. Schattschneider 1960) as well as shared status quo understandings of policy questions (e.g. Baumgartner et al., 2009). There may be multiple explanations for this, as pointed out previously. Parties may employ a vote-seeking strategy, be influenced by the context, or be governed by a systemic logic. Ultimately, the more consensual elements may be important for a well-functioning democracy, as sharing some common understandings of policy questions may lead to more stability and less drastic changes when another party wins office. Only few studies have, however, examined the effects of frame overlap on public opinion formation (see however Slothuus, 2010) or public policy (see however Baumgartner 2013). Some scholars argue for instance that adopting an opponent’s frame can be potentially dangerous (Sniderman and Theriault, 2004), as a party risks losing its credibility (Sides, 2006). Although it seems intuitively meaningful, the picture may also be reversed. Ignoring widespread concern for the economy following a financial crisis may have negative electoral consequences.
To further develop our understanding of party frame competition, the present analysis proposes at least four avenues for further research. First, although it was shown that frame overlap is generally (very) common, the scores were generally higher in the prostitution debate and lower in the early retirement scheme debate (especially for the leading mainstream parties across parliamentary blocs). Obviously, policy questions differ on a range of characteristics (Grossmann, 2013), and some specific characteristics may foster frame overlap more than others. Prostitution is a morality question on which party competition is less appealing in Denmark due to blurry conflict structures (Albæk et al., 2012) that may lead to more frame overlap. On welfare and labor market policies such as the early retirement scheme, party competition is predominantly structured by the classic left-right scale, as the policy contains some redistribution elements. On such questions, parties—especially the leading mainstream parties competing for governing power—might prefer to stick to their ideological profile, resulting in less overlap. More attention should be devoted to examining overlap patterns across different group of policy questions.
Second, another set of questions concerns the parties utilizing frames. As this article has already hinted, mainstream parties competing for office may be more vote-seeking than niche parties (e.g. Green-Pedersen et al., 2014) and therefore more willing to tailor their communication to capture the median voter. A next step is to design a study that in a more nuanced way examines how different party types may or may not influence parties’ utilization of frames. There is a range of interesting factors, such as parties’ government status, their frame-ownership status, and their status in the polls (see also van de Wardt, 2015). Likewise, it might be interesting to further examine who sponsors the frames within a party and whether (different) patterns can be detected. It for example seems interesting to examine whether party leaders tend to utilize certain frames while individual politicians (local or MPs) utilize other frames leading to more or less frame overlap.
Third, a set of questions concerns the role of the news media and hence the channels on which the public debate takes place. While I analyze parties’ utilization of frames in the public debate as captured in news articles, the analysis may not fully capture parties’ actual unmediated communication (e.g. in interviews, press releases). There are, however, good reasons to expect a high degree of overlap in media input and media output. First, studies from multiple European countries report strong issue and frame overlap between parties’ actual messages and those covered by the news media (e.g. Merz, 2017; Wirth et al., 2010). Second, journalists’ role may be more subtle when they (de-)select policy-specific frames sponsored by parties rather than utilize generic frames like the conflict frame that fits well with their institutional foundation (Bartholome et al., 2015; De Vreese, 2012). Third, parties’ communication is generally mediated, and when parties craft their messages, they anticipate the media selection bias (Elmelund-Præstekær et al., 2011). Nevertheless, it is valuable to study journalists’ role in advantaging or disadvantaging frame overlap in the future.
Fourth, this study was conducted in a Danish context, which is important when we discuss the findings and implications for the literature in a boarder perspective. Denmark can be described as a typical West-European country with its multiparty system characterized by minority governments that can rely on a bloc majority (with their support parties) (Green-Pedersen and Thomsen, 2005) similar to two-party systems with clear government alternatives. The institutional setting, however, seems to foster collaboration among parties especially within parliamentary blocs, but it also enables collaboration across blocs which may foster more similar communication profiles among parties within and across blocs. Two-party systems like the U.S. have a slightly different logic that fosters less interaction across blocs. It seems plausible that even in such settings, parties tailor their communication to the median voter and, hence, adopt each other’s frames, but the extent to which they do so may differ. Moreover, Denmark has, like other countries with a corporative media system, a medium to low level of political parallelism, i.e. overlap between the media system and the political system (Hallin and Mancini, 2004). The question is whether similar results would be found in countries with more political parallelism that may limit exchanges of understandings between at least parliamentary blocs.
This article theorizes and empirically shows that parties are very similar, at least in how they communicate. Generally, a high and a very high degree of frame overlap was found in the public debate among political parties across and within the same parliamentary bloc, respectively. This result suggests a need to rethink more thoroughly how we can understand and study interaction among parties when they utilize frames.
Supplemental material
Supplemental Material, sj-pdf-1-ppq-10.1177_13540688211028818 - To which extent do political parties utilize similar frames?
Supplemental Material, sj-pdf-1-ppq-10.1177_13540688211028818 for To which extent do political parties utilize similar frames? by Camilla Bjarnøe in Party Politics
Footnotes
Acknowledgement
The author thanks especially Christoffer Green-Pedersen, Regula Hänggli, Amber E. Boydstun, James F. Adams, Claes H. de Vreese, and Rens Vliegenthart for constructive criticism and suggestions of earlier versions of the article. Besides, an earlier version of this article has been presented at the Comparative Political Agenda Conference 2017, and I would like to thank the participants for constructive comments and suggestions.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
