Studying the Tango: An Analysis of Parliamentary Questions and Press Coverage in the Netherlands (PDF)

STUDYING THE TANGO An analysis of parliamentary questions and press coverage in the Netherlands Peter Van Aelst and Rens Vliegenthart The relationship between politicians and journalists is often portrayed as an intimate dance, characterized by intFerdependence. The p...

STUDYING THE TANGO An analysis of parliamentary questions and press coverage in the Netherlands Peter Van Aelst and Rens Vliegenthart The relationship between politicians and journalists is often portrayed as an intimate dance, characterized by intFerdependence. The political agenda-setting approach has been used successfully to improve our knowledge of certain aspects of this power relationship. This study builds on this line of research, but in more depth by conducting a detailed and reciprocal analysis of the relationship between press coverage and parliamentary questions in the Netherlands (19952010). Our macro-level approach shows that the effect of media on written questions is stronger than the reverse. A more detailed micro-analysis of media coverage preceding and following oral questions does indicate that most of the questions can be traced back to coverage in the days before, but that they receive less media attention afterwards. This might initially indicate that the media are leading the dance with parliamentarians, at least when it comes to questioning behaviour. In many instances, however, media are not creating the news that MPs rely on, but rather are transmitters of information that originates from other political and non- political sources. KEYWORDS parliamentary questions; political agenda-setting; press coverage; The Netherlands Introduction The relationship between politicians and journalists is often portrayed as a dance that requires both partners to cooperate. The dance metaphor raises a scientifically interesting question: who is the leading partner? This straightforward question has not led to a clear-cut answer*and one could even conclude that such an answer does not exist. While a growing number of political scientists and media scholars are convinced that the news media can be considered as a dominant player in the political arena (Edwards 2001; Strömback and Nord 2006), others believe that parties and politicians still remain in charge (Bennett, Lawrence, and Livingston 2007; Dalton et al. 1998). One of the reasons for these conflicting findings might be that the relationship between media and politics is characterized by interdependence (Neveu and Kuhn 2002), reciprocity (Kepplinger 2007) and contingency (Walgrave and Van Aelst 2006). This means that the influence almost always works both ways (Bennett and Livingston 2003) and that which partner is dominant over the other depends on the context and focus. This is more likely to be true as news- making and policy-making have become so intertwined that it is hardly possible to tell them apart (Cook 2006). Therefore, we believe that understanding the close and complex relationship between journalists and politicians requires a focused approach on specific aspects. In this study we will try to disentangle the reciprocal nature of the relationship between Journalism Studies, 2014 Vol. 15, No. 4, 392410, http://dx.doi.org/10.1080/1461670X.2013.831228 # 2013 Taylor & Francis  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 393 politicians and journalists by a detailed analysis of written and oral parliamentary questions and press coverage in the Netherlands. These questions may have limited direct political consequences (Frankin and Norton 1993; Wiberg 1994), but they offer MPs the opportunity to put certain issues forward in an attempt to influence higher political agendas and the legislative process. To attract public attention and put more pressure on the government it is important that the press devotes attention to an MP’s initiative. Trying to link up the question with the news of the day, news coverage often serves as a source of inspiration for these questions (Green-Pedersen and Stubager 2010). At the same time, parliamentary questions offer a steady diet of events that may influence the media agenda. Parliamentary questions can reveal factual information about different issues in society or in some instances create conflict between political actors. Of course not all questions will be considered newsworthy. Not only because the ‘‘carrying capacity’’ of the media is too limited to report everything politicians do and say (Hilgartner and Bosk 1988), but also because the media are mainly concerned by telling political stories that interest the public at large, or at least their particular audience (Wolfsfeld 2011). The fact that parliamentary questions often have limited direct impact on policy and do not always involve high-status political actors strengthens the gatekeeping role of the media in this particular case (Shoemaker and Reese 1991). There are several ways to dismantle ‘‘the dance’’ between parliamentary questions and press coverage. This study combines two different approaches. First, it presents an overview of all written parliamentary question and press coverage in five newspapers for a period of 15 years (19952010). We look at the number of explicit references to news media in the written questions (media 0 politics) and to explicit references to those questions in the press (politics 0 media). In addition to this macro-analysis, this study uses a more detailed micro-analysis to gain a precise insight into the process of mutual dependency and influence. For every oral parliamentary question asked in the years 1997 and 2009 it will be determined whether it referred explicitly to previous media coverage or considered an issue that figured prominently in the previous days (media 0 politics) and/or whether it received media attention in the next days (politics 0 media). Since we will analyse newspaper coverage and parliamentary questions over a longer period, we can test the so-called mediatization thesis that assumes an increased influence for media coverage on politics. The study departs from the existing political agenda-setting tradition that studies when and how the media agenda influences the political agenda and vice versa. However, the paper goes beyond agenda-setting by adding a micro-approach to more fully disentangle the reciprocity of the relationship between press coverage and parliament. Therefore, specific attention will be given to the question whether the media serve as an independent actor influencing parliamentary initiatives, or whether they can rather be characterized as a transmission channel for other (more prominent) political actors. Theory: Improving the Agenda-setting Approach Both in communication science and media studies, on the one hand, and political science, on the other hand, agenda-setting has become one of the dominant paradigms. The same concept, however, means quite different things in both domains. In communication science, agenda-setting is mainly a theory about media-effects: media 394 PETER VAN AELST AND RENS VLIEGENTHART coverage of issues influences the issue priorities of the public and, indirectly, their voting preferences. Since the study of McCombs and Shaw (1972), the popularity of the agenda- setting approach among media scholars has grown steadily and has been one of the most cited media-effects concepts in the past decades (Bennett and Iyengar 2008). In political science, the agenda-setting approach deals mainly with the (limited) attention of policy makers for (a wide range of) political issues and focuses on the policy-making process. Inspired by the work of leading scholars on US politics (Baumgartner and Jones 1993; Kingdon 1984), agenda-setting has become a central concept in political science, and public policy in particular. A more recent stream of research has tried to combine both traditions and focused on the effect of mass media coverage on the political agenda (for an overview, see Walgrave and Van Aelst 2006). For these scholars the central question is to what extent mass media coverage affects the political agenda. Most of this work relies on a time-series design testing to what extent the actual behaviour of political actors regarding specific issues is preceded in time by media coverage about the same issues (Van Aelst and Walgrave 2011). A majority of these studies have concluded that ‘‘the media matter’’, but at the same time stressed the conditionality of the media’s influence on the political agenda. First, the influence of the mass media varies considerably across issues with a larger agenda-setting role for the media on foreign policy, for example (e.g. Soroka 2003; Wood and Peake 1998). A second crucial distinction is that between symbolic and substantial political agendas. Most studies finding relatively strong media effects focused on symbolic agendas such as parliamentary debates or politicians’ public speeches and less so on real policy measures with tangible consequences such as the allocation of resources (e.g. Edwards and Wood 1999; Pritchard and Berkowitz 1993). Recently, Green- Pedersen and Stubager (2010) and Vliegenthart and Walgrave (2011b) found party characteristics in multi-party systems to be a third set of contingent factors. They showed that the political influence of the media depends on the parties’ position (opposition versus government) and the own issue agendas of the parties. Although, political agenda-setting studies mainly treat the news media agenda as an independent variable potentially influencing a certain political agenda, a few studies explicitly incorporated reciprocal effects. In general they agree that the influence works both ways, with some showing the dominance of the media-to-politics dynamic (Van Noije, Oegema, and Kleinnijenhuis 2008; Walgrave, Soroka, and Nuytemans 2007), while others proved that the politics-to-media effects are much stronger (Brandenburg 2002; Jones and Wolfe 2010). From a more theoretical perspective there also seem to be diverging views on how to study the reciprocal nature of the relationship. Some authors always start from the intentions and initiatives of political actors. For instance, Sellers (2010) analysed the complete cycle of agenda-setting and press coverage in US Congress. According to his four- stage model, politicians first create a message and next promote it. The crucial third step is when the message receives coverage and in the final stage the coverage feeds back to influence politicians’ political communication and policy debates. The four stages constitute what Sellers calls a ‘‘cycle of spin’’ where both journalists and politicians react to each other, but the process is initiated by the latter (see also Entman 2003). Following the same line of reasoning, Wolfsfeld (2011) argues that the relationship between media and politics can be best understood as a PoliticsMediaPolitics cycle: political actors take certain initiatives, the media react to these words and actions, which in turn influence the behaviour of the political actors. In contrast, others argue, much in line with the concept of  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 395 mediatization, that political actors already in their initial behaviour take into account how the media will react (Strömbäck 2008). This is what Davis (2007) labelled as an anticipatory effect of political actors on future media coverage. In that case the PoliticsMediaPolitics cycle of Wolfsfeld becomes a MediaPoliticsMedia (Politics) cycle. In sum, political agenda-setting studies that scrutinized the influence of the media on the political agenda have improved our understanding of the power relationship between media and politics. At the same time these studies have shown certain shortcomings or at least have raised doubts about the extent to which they really tap ‘‘net media influence’’ (Vliegenthart and Walgrave 2011b). Or as Reese (1991, 316) put it: ‘‘Can the media be said to be powerful if they simply serve as a conduit for the powers of others?’’ Is the political agenda influenced by the media agenda, or just by the coverage of the (earlier) words and actions of political actors that are discussed in the media coverage? For example, if MPs react in parliament to an interview with the Prime Minister in a newspaper, can we consider this as ‘‘real’’ media influence? Or is this rather an example of how some powerful politicians use the media to promote their agenda? Van Noije, Oegema, and Kleinnijenhuis (2008, 478) admit that the study of the media agenda is complex, already simply ‘‘because it embraces quotations and paraphrases from politicians’’. The macro-level perspective, focusing on the interaction between issue-attention on the media agenda and the political agenda, has been able to provide a general picture of the mutual influence politics and media have on each other. The question remains whether even the most sophisticated of those agenda-setting studies can fully disentangle the intimate relationship between media and politics and especially the process by which this mutual influence comes about. We believe this puzzle requires an additional micro- perspective. Such a perspective is mostly found in case studies drawing on micro data, detailed content analyses or in-depth interviews (Davis 2007). This study employs a micro- analysis that remains closer to the original agenda-setting approach (for a similar approach see Thesen 2013). By analysing each oral parliamentary question separately and the detailed media coverage that both precedes and follows this initiative we hope to get a better insight into how the dynamic works. Before discussing our combination of methods more in detail we formulate several research hypotheses. Hypotheses: A Reciprocal and Contingent Relationship Following our theoretical discussion, our hypotheses start from two general expectations: reciprocity and contingency. First, the relationship between media and politics is characterized by reciprocity: both journalists and MPs influence each other’s work. Following the recent work of Sellers (2010), we argue that agenda-setting by the media and the strategic efforts of politicians to influence the media agenda should be studied together as both processes form an integrated whole. This leads to our first general hypothesis: H1: Newspaper coverage and parliamentary questions influence each other. Previous studies have shown contradictory results on who has the upper hand in this reciprocal relationship. Or put differently, it has remained unclear on who leads the tango. Therefore, we do not formulate a hypothesis about which influence is greater. Second, we expect that this general pattern is contingent. Here we focus on differences across different types of questions, government and opposition parties and 396 PETER VAN AELST AND RENS VLIEGENTHART differences over time. Scholars have argued that the agenda-setting influence of media on politics is especially prevalent when symbolic political agendas are considered. We focus on oral and written parliamentary questions, which are both considered to be part of the symbolic agenda (see further below). However, oral questions can be considered to carry an even higher symbolic value than written questions: they are formulated only just before the question hour takes place, they get an immediate response from a member of government and in the Netherlands, for several years they have been broadcast live on public television (see further below). Therefore, our second hypothesis is: H2: Oral questions refer more explicitly to media coverage than written questions. Previous studies have shown that opposition and government parties behave differently in relation to the media (Thesen 2013). As Vliegenthart and Walgrave (2011a) demonstrate, media coverage has a stronger influence on opposition parties than on government parties. They argue that this is mainly because of the fact that opposition parties are not restricted by any agreements made in government and can without any constraints look in media coverage for ammunition to attack the government. However, the questions asked by MPs from government parties might carry a higher relevance for journalists, since the chances are higher that these questions will actually result in policy changes or endanger the coherence among coalition parties. Therefore, our third set of hypotheses is as follows: H3a: The oral parliamentary questions of opposition parties are more inspired by media coverage than questions of government parties. H3b: The oral parliamentary questions of government parties result in more newspaper coverage than questions of opposition parties. Scholars have argued that the influence of media on the political process has increased over time. For instance, Van Noije, Oegema, and Kleinnijenhuis (2008) show that the influence of the media on the parliamentary agenda in Britain (19882003) and the Netherlands (19952003) has grown significantly. This trend can be seen as part of a larger process that is often labelled as the mediatization of politics (Gianpietro and Wilfried, 1999; Strömbäck 2008). For our study this would mean that media coverage has become a more important source of inspiration over time. We formulate the following hypotheses: H4a: The influence of newspaper coverage on written parliamentary questions increases over time (19952010). H4b: The oral parliamentary questions in 2009 are more inspired by media coverage than the oral parliamentary questions in 1997. Finally, we pose two research questions. The first one addresses the impact of various outlets: the literature on media’s influence on politics contains several studies that focus on the question which medium is more important and which is less so. The findings are mixed, offering no conclusive answer to questions such as whether television is more influential than newspapers or not, or whether quality newspapers are more influential than popular newspapers (Kleinnijenhuis 2003). The same is true for the reversed relationship, for which very few studies exist. Therefore, our first research question is: RQ1: Which media outlets have the strongest impact on parliamentary questions and which media outlets are most strongly influenced by those questions?  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 397 The second research question is specifically focusing on the oral questions. Here, we are interested to know more about the coverage that inspires those questions. Is it journalists themselves that come up with the relevant information (i.e. does the media act as independent actor?) or is it rather that political sources that are used in media coverage that provide the input for questions (i.e. the media merely as a transmitter of information). RQ2: Does the media coverage that inspires oral parliamentary questions mainly rely on political sources? Parliamentary Questions in the Netherlands Having the opportunity to ‘‘question’’ the government is common in most parliamentary democracies. Wiberg (1995) showed that in 18 Western countries MPs are able to ask both written and oral questions. Parliamentary questions have become a natural part of the parliamentary functioning, but the specific formal and informal rules about both types of questioning differ widely across countries (Rozenberg et al. 2011; Russo and Wiberg 2010). Therefore, we briefly explain the characteristics and peculiarities of the Dutch system. The member of the Dutch parliament (Tweede Kamer) obtained the individual right to ask questions in 1906 (Visscher 2006). The right to ask oral questions resulted in the 1960s in a weekly question hour, from 1984 onwards scheduled on Tuesday afternoon to allow MPs to react on recent events that happened that week, and during the weekend in particular. To secure the up-to-date character of the question hour, MPs can suggest questions to the chair of the Chamber until a few hours before the debate. This rule clearly favours questions that are closely related to current issues in the media. In countries such as Germany questions should be submitted well in advance, which lowers the chance that the question deals with a recent event in the news. On average two or three questions can be asked which results in a short debate with the (junior) cabinet minister and other members of the Chamber. When at the beginning of the 1990s the enthusiasm to ask questions among MPs eroded, the decision was made to broadcast the weekly question hour live on public television. Oral questions regained their popularity among MPs and more requests for questions are submitted than there is room for during the one-hour meeting (Visscher 2006). Recently, the Second Chamber decided to allow more questions (up to six) and lower the speaking time of the politicians involved. In sum, the national question hour in the Netherlands is deliberately designed to link parliamentary debate to the broader public debate in the media. Questions should relate to current issues and the question hour is broadcast live on television. Since the number of oral questions is limited and therefore particularly for ‘‘ordinary’’ MPs often not available, written questions offer a viable alternative. Contrary to oral questions, cabinet members are allowed to take quite some time to answer them. Usually a written response is provided within a few weeks, but occasionally it takes several months before a question is answered. There is little doubt that MPs have increasingly made use of this opportunity. Although the popularity of written questions fluctuated heavily over time, there is a clear increase of its use. At the beginning of the 1960s on average 300 questions were asked, in the parliamentary year 20082009 almost 4000 questions were asked (see Figure 1). The reasons for this spectacular increase are diffuse, but the introduction of new (more populist) parties in the Second Chamber that posed a large amount of written questions seems to be part of the explanation (van Holsteyn and 398 PETER VAN AELST AND RENS VLIEGENTHART 4500 40.00% 4000 35.00% 3500 30.00% 3000 25.00% 2500 20.00% 2000 15.00% 1500 10.00% 1000 500 5.00% 0 0.00% 1995–1996 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 2006–2007 2007–2008 2008–2009 2009–2010 number of parliamentary questions percentage questions using any of news sources FIGURE 1 Number of written parliamentary questions and references to news sources Post 2009).1 It is indeed true that the left-wing Socialist Party (SP) and the populist-right Freedom Party (PVV) ask by far the most questions. Written questions can be based on pressing issues of the day, but do not necessarily have to be. They can be argued to be more diverse than oral questions, though also here, inspiration comes often from news coverage and in practice serves the same purposes. An analysis of a sample of 500 written questions per year for the period 19952011 using the issue classifications from the Comparative Agenda Setting project2 shows that a wide variety of topics is covered by written questions, with no category getting more than 15 per cent of the total attention*justice/crime is the most used topic with 14.5 per cent. A similar analysis for the oral questions asked in 1997 and 2009 shows a similar pattern. Also here, attention is spread over a large number of topics, the most salient ones (financial and international affairs) cover both slightly over 10 per cent of the questions. Both written and oral parliamentary questions have been used in political agenda- setting studies. In a recent contribution on questioning behaviour in Belgium and Denmark, Vliegenthart and Walgrave (2011a) spell out the functions these questions have in current Western democracies. First, it is the main tool for (opposition) MPs to control the government and also one of the main ways in which confrontation between opposition and government takes place. Second, it reveals the broader political agenda of a political party. Third, parties try to influence policy-making by asking questions, but as argued, due to the inconsequential and symbolic nature of parliamentary questions, this influence is only indirect and might be rather limited (see also Russo and Wiberg 2010). Research Design We collected all written parliamentary questions and the answers provided by government for the parliamentary years 19951996 until 20092010.3 A total of 34,943  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 399 questions were asked during that period. We use this dataset to investigate the overall use of the main national newspapers (NRC Handelsblad, de Telegraaf, de Volkskrant, Trouw and Algemeen Dagblad) and television news broadcasts (the public NOS journaal and the commercial RTL Nieuws) as a source of information in those questions and to investigate whether the use of those sources increased over time. To study the reverse relationship we looked for references to parliamentary questions in the Dutch press. Using the digital LexisNexis database, we counted for each parliamentary year the number of articles referring to parliamentary questions. We did this for each of the five national newspapers mentioned above.4 For the micro-analysis of oral questions we followed a different approach. We collected all oral questions posed in 1997 (N 80) and 2007 (N112). For each question we coded the actors involved (MP, cabinet minister), the issue focus, whether the question (explicit or implicit) referred to previous media coverage and the actual source on which the question was based. Next, for each question the coverage in seven (national) newspapers (NRC Handelsblad/NRC Next, de Telegraaf, de Volkskrant, Trouw, Algemeen Dagblad, Financieel Dagblad and Het Parool) in the week before (N 718 newspaper articles) and the two days after the question (N 295) were collected via Lexis Nexis5 and subsequently analysed. All articles (up to 15) that were about exactly the same topic as the question, not just the same issue, were coded. For example, if a question concerned headscarves for civil servants only press coverage on headscarves was taken into account and not related articles on immigration. The analysis of news coverage after the question includes the presence or absence of the MP asking the question and the (junior) Minister answering. Results Macro-analysis: The Influence of Media Coverage on Written Questions  (1995 2010) First, we look at the influence of media on written parliamentary questions over time. Figure 1 presents an overview of the total number of questions asked and the percentage that is using any of the five national newspapers and two television broadcasts. In the whole research period, the percentage of questions that refers explicitly6 to any of the seven news sources is roughly one in three. This observation is in line with our first hypothesis that news coverage influences parliamentary questions. The expectation that the influence of the media over time (H4a) would increase finds no confirmation. On the contrary, the most recent year of the research period has the lowest number of references to the most important media sources (only 24 per cent). Overall, we see a considerable increase in questions asked from 1689 in the parliamentary year 1995 1996 to 3854 questions in 20082009, but relatively few of these questions refer to the seven most important news sources. Of course MPs can refer to other media outlets as the main source, but we deem this to be unlikely, as other sources play a much smaller role.7 However, it might be that MPs refer less explicitly to media coverage, but still take it into account. This will be tested for oral questions in our micro-analysis (see below). Figure 2 provides a more detailed picture of the use of individual sources. Two things stand out: first, throughout the research period newspapers are clearly a more important source than television broadcasts. Second, there has been a considerable shift 400 PETER VAN AELST AND RENS VLIEGENTHART 14.00% 12.00% 10.00% 8.00% 6.00% 4.00% 2.00% 0.00% 1995–1996 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 2006–2007 2007–2008 2008–2009 2009–2010 telegraaf volkskrant trouw nrc ad nos-journaal rtl-nieuws FIGURE 2 Use of different news sources in written parliamentary questions in the sources that are being used. Most notably, in the last years the popular newspaper de Telegraaf has taken over from the quality paper de Volkskrant as the main source of information. This is mainly due to the populist-right Freedom Party of Geert Wilders. This party entered parliament in 2006 and parliamentarians from this party extensively use de Telegraaf as their source of information. Four out of ten of their questions are based directly on their news coverage. The popularity of de Telegraaf, the most widely read Dutch newspaper with readership that resembles the Dutch population (Bakker and Scholten 2009), shows that politicians want to link their questions with issues that address a large and diverse audience. As a partial answer to our first research question (RQ1), newspapers are more often used as sources for written questions than television news, while the popular newspaper de Telegraaf becomes increasingly important. Macro-analysis: The Influence of Parliamentary Questions on Newspaper Coverage (1995 2010)  We now turn to the reversed relationship and look at the references to parliamentary questions in the written press. Figure 3 demonstrates the number of articles that refer to parliamentary questions in four national newspapers (de Telegraaf is excluded here, since it is not available for the whole period), as well as the ratio between the number of those newspaper articles and number of written parliamentary questions. It shows that there is no linear trend in the number of articles that refer to parliamentary questions: it decreases between 1995 and 2001, then rises between 2002 and 2004 and remains stable for the last few years. This development can, at least partly, be understood as a consequence of the changing political situation in the Netherlands. After 2001, with the rise of Pim Fortuyn’s populist party entering parliament, the country witnessed a highly turbulent and politically uncertain time, with early elections being held. It became (again) relevant for journalists to consider the activities of parliamentarians, because they  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 401 800 0.45 700 0.40 0.35 600 0.30 500 0.25 400 0.20 300 0.15 200 0.10 100 0.05 0 0.00 1995–1996 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 2006–2007 2007–2008 2008–2009 2009–2010 number of newspaper articles (left axis) ratio: number of articles divided by number of written questions (right axis) FIGURE 3 References to parliamentary questions in four newspapers (19952010) might be a source of potential conflict. However, given the fact that the number of parliamentary questions has increased enormously in the last decade-and-a-half, the ratio between articles and questions has decreased steadily between 1995 and 2010. Now, for each newspaper article referring to a question (either written or oral), between six and seven written questions have been submitted. Overall, the results offer only moderate support for our first hypothesis concerning the reciprocal relationship between press coverage and parliamentary questions. Yes, journalists refer explicitly to parliamentary questions, but only infrequently. Figure 4 provides a comparison between the separate newspapers. Here, it is similarity rather than difference that stands out: all newspapers follow the same pattern as described in the previous paragraph. Some differences exist (e.g. the large peak of NRC Handelsblad in 20032004), but they are relatively minor. There is a clear difference between the quality newspapers NRC Handelsblad, de Volkskrant and Trouw, on the one hand, and the popular newspapers Algemeen Dagblad and de Telegraaf, on the other hand: the first group refers more often to parliamentary questions. Especially NRC Handelsblad stands out in using questions as a source. In that respect, one can state that this source is most strongly influenced, while Algemeen Dagblad and de Telegraaf are the least influenced (RQ1). Micro-analysis: The Influence of Press Coverage on Oral Questions (1997 and 2009) In our micro-analyses we focus on all oral questions in 1997 and 2009. Hypothesis 2 states that we expect that MPs would refer more to media coverage in oral than written questions. At first sight Table 1 shows that the media coverage is indeed more important 402 PETER VAN AELST AND RENS VLIEGENTHART 250 200 150 100 50 0 1995–1996 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 2006–2007 2007–2008 2008–2009 2009–2010 telegraaf volkskrant trouw nrc ad FIGURE 4 References to parliamentary questions in five newspapers (19952010) for oral questions. In both years a majority of oral questions refer to media coverage in general or to a specific news item by one medium. The list of media the oral questions refer to in both years is quite diverse. Only one trend (similar to that for written questions), stands out: the newspaper Telegraaf was hardly ever used as a source in 1997, but became the most popular source in 2009. If we only take references into account to the same seven news outlets coded for the written questions, the percentage of references to media is lower for oral questions. Therefore, H2 cannot be confirmed. However, the picture changes quite drastically if we rely not on the media sources the MPs refer to in their question, but rather on the actual press coverage that precedes the oral question. In both years about 8 out of 10 topics were reported in at least one national newspaper before the actual oral question was asked in parliament. Furthermore, it might well be that the remaining 20 per cent of questions were also influenced by media coverage as we only analysed national newspapers and did not include other types of newspapers8 and audiovisual media. So it seems that even if MPs do not refer to a specific medium, or to the media in general, media coverage almost always inspired or influenced the question. The expectation that the oral questions of opposition MPs are more inspired TABLE 1 Relationship between oral parliamentary questions and media coverage 1997 2009 Total Percentage of oral questions referring to media coverage Explicit reference to one medium 35.0 40.2 38.0 General reference to media 18.8 16.1 17.2 No reference 46.3 43.8 44.8 Percentage of oral questions that ‘‘follows’’ from preceding press coverage 77.5 79.5 78.6 Percentage of oral questions that ‘‘is followed by’’ press coverage 58.8 43.8 50.0 N 80 112 192  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 403 TABLE 2 Percentage of newspaper articles that mention the MP and/or cabinet minister? 1997 2009 Total MP 42.1 30.5 36.9 Cabinet minister 92.7 80.2 87.1 N 164 131 295 Only articles that report on the topic of an oral question are taken into consideration. by media coverage is not confirmed (H3a). MPs of both government and opposition parties rely to a similar degree on previous press coverage for their selection of questions. Micro-analysis: The Influence of Oral Questions on Press Coverage (1997 and 2009) The opposite causal relationship about the influence of oral parliamentary questions on newspaper coverage is less clearly present: only half of oral parliamentary questions also lead to newspaper coverage in the following days. The number of oral questions that receive subsequent newspaper coverage in 2009 is lower than in 1997.9 This might be partly explained by the higher average number of oral question in 2009 that increased competition for media attention. It also reflects a more general trend of decreasing newsworthiness of traditional parliamentary work and debates. A tendency that can be witnessed in several Western democracies (Jones and Wolfe 2010; Negrine 1999). Furthermore, a majority of the press coverage following the parliamentary question hour does not refer to the MP that asked the question (or not even to the debate), but only to the answer of the (junior) cabinet minister. Table 2 shows that less than half of the newspaper articles that report on the actual topic of the oral question refer to the MP that asked the question. Again, the media attention for these MPs goes down over time, and in 2009 has become less than a third of the articles covering the topic of the question. If oral questions are reported upon, the coverage mentions almost always the government official that replies to the question. We will elaborate on this finding in the conclusion. In general, MPs of government parties and opposition parties are equally (un)successful in getting their questions reported in the news. This is in contrast with our expectation (H3b) that government MPs would be more successful in gaining media attention. Furthermore, MPs of parties in government are mentioned even less (20 per cent) in the articles that follow coverage of oral questions compared to opposition MPs (41 per cent). Finally, we address our second research question by further investigating the role of the media as sources for oral questions. Do the media function as the actual source of these parliamentary initiatives or are they just the channel that MPs use to get information? To be considered as an actual source, the medium should have ‘‘created’’ the coverage that leads to the questions themselves. This can be done in several ways, ranging from a media-sponsored opinion poll, forms of investigative journalism, or an editorial that is picked up by an MP. In all these cases the media initiate the coverage rather than reflecting the initiatives of others. Table 3 shows that the media can be considered as the ‘‘real initiator’’ for barely one in five questions. Mostly it concerns active 404 PETER VAN AELST AND RENS VLIEGENTHART TABLE 3 The actual source of the oral parliamentary question (%) 1997 2009 Total Media 16.3 19.8 18.3 Politician 37.5 21.6 28.3 Government institution 32.5 27.0 29.3 Interest group 6.3 17.1 13.6 Foreign actor 6.3 8.1 7.3 N 80 112 192 research of journalists that is used by an MP to raise critical questions about government policy. Although there is a small increase over time (from 16 to 20 per cent), it is not the media but national politicians and (semi-independent) government institutions which remain the main source for oral questions. When other politicians are the source, it mostly concerns questions about a statement the cabinet minister made in the media or public arena. In the case of government institutions, questions deal more with policy documents distributed by the government or a specific department. Note that we defined government broadly including semi-independent institutions such as universities and the national bank. Comparing both years, there is significant increase of external national sources at the expense of politicians. The number of oral questions that rely on activities or information of interest groups has increased from 6 to 17 per cent. Besides the traditional socio-economic interest groups (e.g. trade unions and organizations of employers), this source category also consists of less formal action groups of, for instance, consumers. Although the number of cases is too low to make sweeping generalizations, their growing importance seems to indicate that these organizations have become more successful in bringing their concerns on to the parliamentary agenda. Conclusion and Discussion Our analysis of the questioning behaviour of Dutch parliamentarians and the role of media coverage in the past 15 years provides a rich, yet nuanced picture. In general, we find our main hypothesis confirmed: media coverage leads to parliamentary questions and these parliamentary initiatives in turn lead to more media coverage. Both the macro- analysis and the micro-analysis confirm, in different ways, that the relationship is highly reciprocal. Both types of questions refer frequently (3040 percent) to the explicit role of news media as a source of inspiration. But the implicit influence of the media is even stronger: the micro-analysis illustrates that up to 8 out of 10 oral questions are preceded by national press coverage on the same topic. The reverse relationship also found support. About half of oral questions are indeed covered in the national press, mostly without mentioning the initiating role of the MP. Also the written questions appear in the press, but because of their growing number the chance for a written question to receive media attention has declined over the years. Comparing the media to politics and the politics to media influence, it seems that in our study the first is clearly stronger. In the terminology of Wolfsfeld (2011) it seems that there is rather a MediaPoliticsMedia cycle instead of a PoliticsMediaPolitics cycle.  AN ANALYSIS OF PARLIAMENTARY QUESTIONS 405 Media coverage inspires MPs’ initiatives, who hope in turn to get some attention in the media. Although parliamentarians engage in frequent interaction with the media, they are seldom able to set the media agenda independently. Or put differently, it seems that journalists are the leading partner in this dance, at least most of the time. There are, however, two important elements that provide nuance to this conclusion. First, this study deals with a symbolic political agenda that has little direct impact on actual policy-making. In line with previous research, the media influence is most outspoken on these lower agendas (see also Bonafont and Baumgartner 2013; van Noije, Oegema, and Kleinnijenhuis 2008). This means that on more substantial aspects of parliamentary work, such as law- making, the media might serve less as a source of inspiration than is the case for parliamentary questions. Second, our micro-analysis shows that the media are mainly a channel through which politicians and other societal actors communicate, rather than an actor that influences parliamentarians autonomously. This does not imply that the media are irrelevant, on the contrary, it shows that politicians but also other actors such as interest groups are eager to get their message in the media. In this way they can influence the general public, but perhaps more importantly, the agenda of other political players. The role of journalists as gatekeepers (Shoemaker and Reese 1991), deciding who gets a voice in the media realm and who is left out is essential. However, this role should be clearly distinguished from an agenda-setting role that suggests that media can independently determine the issues in the news and the parliamentary agenda (see also Wolfsfeld and Sheafer 2006). In sum, this means that parliamentarians will perceive the cycle to be MediaPoliticsMedia, but that in a broader perspective the relationship is closer to a PoliticsMediaPolitics cycle as the media agenda is highly influenced by (other) political actors in the first place. The fact that parliamentarians have growing difficulties in getting attention for the questions they ask might have two different reasons that might reinforce each other. On the one hand, it seems that the newspapers devote less attention to traditional parliamentary work and, on the other hand, the number of parliamentary questions in the Netherlands has increased considerably. This means that the competition over the limited media space for parliamentary work has increased considerably, resulting in less attention for individual questions and the topics they address. This increases the inter-MP competition for media attention. As a result we see that MPs rely on issues that are already in the news, with a growing preference for those media outlets like de Telegraaf that reach a large audience. In that way MPs try to connect their parliamentary work with the public debate and issues the broader public is concerned about. This strategy seems to work as about half of oral questions do lead to media coverage. This means that MPs can influence the media agenda and promote their issues in the news (Green-Pedersen 2010). It does not, however, guarantee coverage for the individual MP. The finding that mainly the response of the cabinet minister is newsworthy is in line with extant research that argues that journalists favour sources with more institutional power (Bennett 1990; van Dalen 2011). That, in particular, MPs of government parties are seldom mentioned in the news following a question might be a consequence of the fact that the cabinet ministers (being the more prominent politicians of sometimes the same party) have already received a lot of attention (see also Van Aelst et al. 2008). We can conclude that our approach combining macro- and micro-analyses is fruitful in unravelling at least partly the complex relationship between media and politics. At the