Personalization of political newspaper coverage: a longitudinal study in the Dutch context since 1950

Ellis Aizenberg, Wouter van Atteveldt, Chantal van Son, Franz-Xaver Geiger VU University,

This study analyses whether personalization in Dutch political newspaper coverage has increased since 1950. In spite of the assumption that personalization increased over time in The , earlier studies on this phenomenon in the Dutch context led to a scattered image. Through automatic and manual content analyses and regression analyses this study shows that personalization did increase in The Netherlands during the last century, the changes toward that increase however, occurred earlier on than expected at first. This study also shows that the focus of reporting on politics is increasingly put on the politician as an individual, the coverage in which these politicians are mentioned however became more substantive and politically relevant.

Keywords: Personalization, content analysis, political news coverage, individualization, privatization

Introduction When personalization occurs a focus is put on politicians and party leaders as individuals. The context of the news coverage in which they are mentioned becomes more private as their love lives, upbringing, hobbies and characteristics of personal nature seem increasingly thoroughly discussed. An article published in 1984 in the Dutch newspaper De Telegraaf forms a good example here, where a horse race betting event, which is attended by several ministers accompanied by their wives and girlfriends is carefully discussed1. Nowadays personalization is a much-discussed phenomenon in the field of political communication. It can simply be seen as: ‘a process in which the political weight of the individual actor in the political process increases

1 Ererondje (17 juli 1984). De Telegraaf, pp. 4

1 over time, while the centrality of the political group declines’ (Rahat & Sheafer, 2007: p. 65). The process of personalization is part of a larger shift, in which the role of the media seems to change from being societally responsible to being profitable where are increasingly concerned with what the public is interested in rather than what the publics’ interest should be (Underwood, 2001; Hanitzsch, 2007; Takens, 2012). A report published in 2003 concludes that the press increasingly focuses on individual politicians and therefore turns away from substantive reporting (RMO, 2003). According to this study this process results in improperly informed citizens that are lacking political engagement causing an unhealthy democracy. Some scholars share this view and argue that the personalization of political newspaper coverage can cause damage to a healthy democracy as it puts focus on individual politicians and thus turns the attention away from parties, programs, policies and substantive issues (Van Aelst, 2012; Van Santen, 2012; Langer 2007). This process clashes with the normative requirements that a healthy democracy demands: it can damage the public debate and it could inhibit citizens from casting a well-informed vote (Takens, 2013; Van Aelst et al., 2012), which could form a problem according to some models of democracy and their associated requirements. In spite of the assumption that personalization is a process that increases over time (Rahat & Sheafer, 2007; Hart, 1992), empirical findings illustrate a scattered image as empirical studies in several Western democracies showed no increase of personalization (Takens, 2012; Vliegenthart 2011; Wilke & Reinemann, 2001: Sigelman & Bullock, 1991), a few studies did show an increase of personalization (Wattenberg, 1998; Dalton et al., 2000) or presidentalization (Takens, 2012) and some studies show mixed results (e.g. Kriesi, 2011). Van Santen (2012) shows through a qualitative content analysis that personalization could be seen as a historical continuity in Dutch television. It is not a surprising fact that studies led to different results in different countries as trends that can lead to personalization do not per se occur (simultaneously) in every country. It is however rather interesting that scholars found differences in results when it comes to countries that are much alike and even found mixed results within one country: The Netherlands (Vliegenthart, 2011; Kriesi, 2011; Takens, 2012; Van Santen, 2012). According to Kepplinger (2002) and Vliegenthart (2011) these contradictory findings can be explained by the timeframes that were analyzed in these studies for two reasons. Firstly, as the fundamental changes toward that trend of personalization, such as the process of depillarization and the emerging competitive media market, occurred earlier on. Secondly, the larger shift of which personalization seems to be part has been identified as a slow process and can therefore be best analyzed over several decades. Due to the scattered image that exists of personalized political newspaper coverage in The Netherlands that can be explained by the relatively short timeframes that were analyzed

2 and the threat that personalization can be to a healthy democracy it is of importance to study personalization in the Dutch context while applying a broad timeframe. The research question of this paper is stated as follows:

RQ: Did the amount of personalized political newspaper coverage increase since 1950 in The Netherlands?

In this paper I will first shed light on the theory of personalization, where conceptualizations, explaining factors such as depillarization and the emerging competitive media market and existing literature on personalization will be discussed, which will lead to the hypotheses that are formulated. Then the newspaper data and methods of automatic and manual content analysis will be discussed. Subsequently the results of this study on the development of personalization since 1950 will be presented, interpreted and be put into their context together with the theoretical and practical implications of these findings. Following this paper will be concluded with a discussion including limitations and suggestions for future research.

Theory

Personalized political newspaper coverage In the literature there does not seem to be a broad consensus on how personalization should be conceptualized. Personalization is often described as the relative attention for politicians compared to the total amount of attention for political parties in newspaper coverage (Wattenberg, 1998; Vliegenthart et al., 2011). What that attention comprises however is unclear (Van Santen, 2012) and therefore this conceptualization seems incomprehensive. In 2007 Rahat and Sheafer propose a more refined concept of personalization in which they distinguish between institutional personalization, media personalization and behavioral personalization. The first form claims that an emphasis is put on the individual politician by institutions and mechanisms. The second form relates to how politics is presented by the media. The authors refer to a change over time where a focus is placed on individual politicians and less attention is given to organizations and parties. The third type of personalization refers to politicians that increasingly participate in individual political activities rather than political party activities (Rahat & Sheafer, 2007). In the second form of personalization Rahat and Sheafer (2007) further differentiate between (media) personalization and privatization. Where personalization is related to activities and traits of the individual politician that are politically relevant and where privatization refers to activities and traits that are personal (Rahat & Sheafer, 2007). Langer (2007) also distinguishes between traits of individual politicians that are politically relevant and traits that are related to the personal life of a politician.

3 In 2012 Van Aelst and his colleagues propose a concept of personalized newspaper coverage in which they build on the following definition of media personalization: ‘a change in the presentation of politics in the media, as expressed in a heightened focus on individual politicians and a diminished focus on parties, organizations, and institutions’ (Rahat & Sheafer, 2007: p. 69). This concept will be followed in this paper. Van Aelst and colleagues differentiate between two types of personalization: individualization and privatization.

Individualization Individualization refers to a change over time where there is a heightened focus on politicians as individuals in newspaper coverage rather than a focus on political parties. The first concept consists of two sub dimensions: general visibility and concentrated visibility. The first sub dimension, general visibility, relates to a shift in focus on individual politicians rather than parties. The second sub dimension, concentrated visibility, refers to a shift in focus on political leaders (Van Aelst et al., 2012). This dimension can also be referred to as presidentialization (Van Santen, 2012; Takens, 2012; Langer, 2007; Kriesi, 2011). Langer (2007) further differentiates between the presidentialization of power and presidentialization of presentation where the former relates to a ‘shift in the distribution of power towards leaders’ and the latter refers to ‘the associated increase in leaders’ overall mediated visibility’ (Langer, 2007; p. 373).

Privatization The second form of personalization, which is privatization, relates to a change over time where individual politicians are put in a more private context rather than in a context that is politically relevant and substantive. Privatization also consists of two sub dimensions where the authors distinguish between personal characteristics and personal life of the politician (Van Aelst et al., 2012). The former has also been identified by Langer (2007). She refers to this type of personalization as politicisation of private persona. The first concept, personal characteristics, can best be described as ‘a change from features regarding their professional competence and performance to features concerning personality traits related to their personal life’ (Adam & Maier, 2010: p. 216). The second concept, personal life, involves a shift toward a focus in media coverage on the personal life of the politician, where there is an increased focus on news about the family, upbringing, love life and leisure time of the politician (Van Aelst et al., 2012).

The rise of personalization: explaining factors The applied definition of personalization of Rahat and Sheafer (2007) refers to a change in the presentation of politics in the media over time, that is: there is simply more of it today (Hart, 1992: p. 68). Therefore it can be argued that a shift in focus on individual politicians in newspaper coverage has taken place. There are two important interlinked factors that can

4 explain the rise of personalized political newspaper coverage: the process of depillarization in The Netherlands and the changing media landscape. First of all, the traditional ties between the press, voters and political parties have weakened. During the age of pillarization in the Netherlands between 1917 and 1967 (Lijphart, 1992), close ties used to exist between newspapers, political parties and its voters. As Hallin and Mancini (2004; p. 53) put it: ‘different subcommunities – Protestant, Catholic, Socialist, and Liberal – developed their own educational, cultural, social and political institutions – ranging from sports clubs to trade unions and political parties’. Lijphart (1990) describes the concept of pillarization as four closed groups with separate political parties, pressure groups and media outlets. There was no discourse between the four groups; the elites of the groups however did liaise with members from other groups. After 1960 when the process of depillarization took place these ties began to fade and a more general bond between newspapers, parties and the public started to arise (Van Kempen, 2007; Hallin & Mancini, 2004; Dalton et al., 2000). Lijphart (1990) argues that developments as secularization, ‘de-ideologising’ and the loosening of ties between pillarized organizations, such as political parties, pressure groups and media outlets are all part of the process of depillarization. In the recent decades political scholars argue that the press has become an autonomous institution, which is independent and not tied to political parties (Cook, 1998; Schudson, 2002; Hallin & Mancini, 2004). A result of this development is the institutionalization of the media (Takens, 2012: Cook, 1998; Rahat & Sheafer, 2007). During the pillarization era, the news media used to function as instruments of parties to reach potential voters, this concept can be referred to as the instrumentalization of the media (Hallin & Mancini, 2004). When these ties began to weaken the press became a professional institution of its own, in service of the public interest. However due to the second factor: the changing media landscape, the press started to determine its coverage based on the preferences of consumers rather than the interest of the public good (Brants & Van Praag, 2006; Takens, 2012; Van Aelst et al., 2012, McManus, 2009). Two aspects of the changing media landscape are of particular importance: firstly the growing role of television (Hart, 1992; Schudson, 2002; Van Aelst et al., 2012; Mazzoleni & Schulz, 1999; Adam & Maier, 2010) and secondly the emerging competitive media market (McManus, 2009; Takens, 2012). The former is of importance as television has the tendency to give more attention to personalities compared to parties because of its visual nature (Schudson, 2002). As Eide (1997: p. 179) puts it: ‘politicians in a way become human beings, while the voters become customers’. The latter could occur due to declining readership numbers (Bird & Dardenne, 2009) and declining numbers of paid journalists (McManus, 2009). In to survive the competitive media market the press has to attract the attention of the public at all costs and is therefore addressing its readers, more than ever, as consumers rather than citizens where the

5 press, among other things, increasingly focuses on personalities (Gans, 2009; Kriesi, 2011) as they ‘are available to serve as objects of general identification’ (Galtung & Ruge, 1965: p. 29) to the readers. It is also time and cost saving as it is easier to capture a news story about a person than an event (Galtung & Ruge, 1965). As a result of the interconnecting factors above, not only a heightened focus on individual politicians in newspaper coverage emerged, but also parties, NGO’s and politicians had to adapt their strategies in order to generate coverage (Schudson, 2002; Meyer, 2002). As Cook (1989: p. 168) puts it: generating news coverage is a ‘crucial component of making ’. This development might in turn contribute even more to the personalization of political newspaper coverage.

Effects of personalization on democracy: a normative perspective The much-discussed topic and possible rise of personalization also brings along a normative discussion among scholars on personalization and its possible effects. In a healthy democracy, citizens, journalists and politicians are all expected to fulfill a certain role. It can be argued that the relationship between these three entities is somewhat symbiotic. To participate in a democracy, citizens need to be able to cast a well-informed vote. To do so the public needs information provided by journalists about current societal issues (Takens, 2013; Strömbäck, 2006). The press, in turn, needs citizens to subscribe to their newspapers and to pay heed to their coverage in order to attract and maintain advertisers. So as to attract readers and compete with other media outlets, journalists have to include scoops and newsworthy facts in their news stories. When reporting on politics, the press needs scoops that have to be provided by politicians. Politicians also benefit from this close relationship with the press, as they can reach their potential voters through these media outlets in order to generate votes (Takens, 2013; Strömbäck, 2006). The phenomenon of personalization can clash with the normative requirements that a healthy democracy demands. It would not however cause a problem for all models of democracy. Strömbäck (2006) proposes four models of democracy by describing their central mechanisms for securing the primacy of the common good and core normative expectations from the public. Personalization can clash with the following three models of democracy and their central mechanisms and expectations from the public: a competitive democracy, a participatory democracy and a deliberative model of democracy. A competitive democracy is characterized by competitive elections and demands a public that has clear opinions of societal issues. A participatory model of democracy is marked by citizen participation in public life and political parties and therefore requires its citizens to be politically interested and to engage in public and political life. Within a deliberative model of democracy discussions among all sections of the public and their representatives should be held. This model of democracy hence expects its

6 public to participate in discussions on societal issues. These three models thus demand a public that at least has knowledge of societal issues in order to meet the requirements given by these models of democracy (Strömbäck, 2006). Personalization would not necessarily cause a problem for a procedural model of democracy as it is characterized by free and fair elections and it ‘does not put any normative demands on citizens that they should vote, should consume news journalism, should participate in public life, or should be well-informed’ (Strömbäck, 2006: p. 334). According to Oegema and Kleinnijenhuis (2000), a focus on individual politicians in newspaper coverage does not necessarily imply that there is less attention given to substantive political issues: news articles that focus on politicians even included more substantive societal issues than news articles that focus on political parties. Van Aelst et al. (2012) and Van Santen (2012) also argue that a simple heightened focus on individual politicians in newspaper coverage is not necessarily a threat to democracy. However when privatization occurs, a concept that refers to media presenting individual politicians no longer in a political context but rather in a private context, there is more cause for concern. Not only will this damage the public debate based on substantive political issues, it can also cause cynicism among the public (Hart, 1992; Boukes & Boomgaarden, 2014, Jebril, Albaek & De Vreese 2013; Van Aelst et al., 2012; Schudson, 2002). Cynicism would especially clash with participatory and deliberative models of democracy, as these models demand a public that is politically interested as well as participating in public life and political discussions.

A scattered image of personalization Although an increase of personalization is assumed over time in the literature (Rahat & Sheafer, 2007; Langer, 2007) this assumption has not always been supported by empirical studies (Kriesi, 2011). The empirical evidence of the personalization thesis is somewhat contradictory and therefore a gap between assumptions in the literature and empirical findings exists. Takens (2012) studied personalization in the context of Dutch election campaigns. She conducted a semantic network analysis of two news broadcasts and four national newspapers between 1998 and 2010. Takens showed that the news has become less personalized. She found however, an increase of presidentialization. Vliegenhart and his colleagues (2011) performed an automatic content analysis to study personalization in the Dutch and British context. They analyzed quality and tabloid- newspapers in the period of 1991 until 2007 and found small differences between levels of personalization in the different newspapers. However, they found no empirical evidence for an increase of personalized newspaper coverage in Britain and The Netherlands. In 2001 Wilke and Reinemann showed that there was no clear trend towards an increase of personalization through their content analysis of election coverage of four German newspapers between 1949 and 1998. Sigelman and Bullock (1991) also showed no trend

7 towards an increase of personalization through their content analysis of campaign coverage of five newspapers in the United States between 1888 and 1988. Wattenberg (1998) examined the personalization thesis through the content of US election coverage in two newspapers and three weekly newsmagazines between 1952 and 1980. He found that political candidates were mentioned more often than parties throughout the whole timeframe: “the ratio increased from about two to one in the 1950s to roughly five to one by 1980” (Wattenberg, 1998: p. 93). Dalton and his colleagues (2000) show through their comparative study that personalization increased in election media coverage in the US, Austria, and the UK in the period of 1952 until 1997. Kriesi (2011) found no trend of an increasing level of personalization through his content analysis in election coverage of quality press and tabloids in France, , the United Kingdom, and Switzerland between 1970 and 2000. He did however find an increasing trend of personalization in the Netherlands. Van Santen (2012) showed through her qualitative content analysis of televised portraits of Dutch politicians between 1961 and 2006 that personalization was already present in the 1960s. Van Santen also finds that political televised portraits always primarily contain politically relevant ideas, activities and goals which are embedded in the political and societal affiliations of the concerning politician. She therefore concludes that personalization has never interfered with the attention for substantive political coverage (Van Santen, 2012). Three important factors could account for these contradictory findings in the literature. The first factor is related to the differences between countries when it comes to their media systems and the associated characteristics. Hallin and Mancini (2004) differentiate between three models of media systems and their associated characteristics: the Mediterranean or Polarized Pluralist Model, the Northern European or Democratic Corporatist Model and the North Atlantic or Liberal Model. The models are based on geographical region and an important element of the political system that is crucial to understanding the elements that mark the relationship between politics and the media in that specific model. The countries that are grouped under these categories share four important characteristics, that is: newspaper industry, political parallelism, professionalization and the role of the state in the concerning media system (Hallin & Mancini, 2004). Especially professionalization and the role of the state are of particular interest here. The level of professionalization of the media in a specific country relates to our first interlinked factor that could explain the rise of personalized political newspaper coverage: the instutionalization of the media. When the press of a certain country is characterized by weaker professionalization and thus by instrumentalization (e.g. Italy), it is plausible that the media are tied to political parties and are therefore less likely to act in service of the public interest. However when the media system of another country is characterized by

8 strong professionalization (e.g. Austria and Canada) it is more likely that the press is serving the public interest. The role of the state relates to our second interlinked factor that could account for the rise of personalization: the changing media landscape. When the media system of a specific country is characterized by strong state intervention (e.g. Italy) and is therefore not market-dominated, it is less likely that the press is struggling to attract its readers’ attention in order to survive in the emerging competitive media market. It is therefore also less likely that the press addresses its readers as consumers instead of citizens. When the media system of a certain country is, however characterized by market-domination (e.g. Canada and the U.S.) it is more likely that the press treats its readers as consumers. Thus for the press operating in countries with media systems characterized by institutionalization and market-domination it is more likely to address readers as consumers than for the press operating in countries with media systems characterized by instrumentalization and strong state intervention. Therefore it is also more likely for the press, operating in countries such as the former, to act according to a certain media logic. Based on the above reasoning it is possible that studies, which have been conducted in different countries, could show differences in results of levels of personalization. The second factor is linked to the timeframe of the analyses of the personalization thesis. Vliegenthart and colleagues (2011) argue that the fundamental changes that resulted in a tendency toward reporting according to the values of media logic occurred earlier on and therefore have not been found in the timeframe that was analyzed. It is very likely therefore, that some of the illustrated studies above did not show a clear trend toward personalized newspaper coverage. Kepplinger (2002) also states that mediatization, and the associated processes such as media logic, is a slow process and can therefore be best analyzed over several decades. He notes however that there seems to be a lack of studies in which media logic and its characteristics like the personalization of newspaper coverage are systematically analyzed over time. The third factor concerns conceptual clarity and the operationalization of personalization. In the illustrated studies above, different concepts and operationalizations of personalization are applied and it is therefore presumable that this leads to different results. Some studies for example measured the count of references to politicians as individuals (Wattenberg, 1998; Kriesi, 2011; Vliegenthart et al., 2011; Takens, 2012) and other studies also analyzed the context in which these politicians occurred (Wilke & Reinemann, 2001; Langer, 2007). Van Aelst and his colleagues (2012: p. 204) express their concern regarding this ‘lack of conceptual clarity and common operationalizations’ and therefore propose a model in which personalization is one concept with two distinct sub dimensions: individualization and privatization.

9 To conclude: different findings on the personalization thesis between countries can be explained by differences in media systems and the associated characteristics and by the differences in trends that can lead to personalization in different countries, it is however quite interesting that studies on personalization within one country led to different results as for example in the Dutch case. These differences can possibly be explained by the application of different timeframes and conceptualizations of the concept of personalization.

This study: personalization in The Netherlands since 1950 Although an increase of personalization is assumed in The Netherlands over time, the empirical evidence for such a positive trend toward however is quite thin. In order to overcome the contradictory findings of other studies and to attempt resolving the scattered image that exists on the personalization thesis in The Netherlands, this study will analyse a broader timeframe and apply a concept of personalization that exists of two distinct sub dimensions (Van Aelst et al., 2012). First, this study will analyse a broader timeframe (1950-2011) than applied in other personalization studies (1970-2011) in The Netherlands as the developments toward an increase of personalization plausibly occurred earlier on. Furthermore the larger development from which personalization seems to be part of has been identified as a slow process that needs to be analysed over several decades (Kepplinger, 2002; Vliegenthart, 2011). Second, in this paper the concept of personalization will be treated as one concept that consists of two distinct sub dimensions for two reasons. Firstly, because of a lack of studies which measured both types of personalization within one study, especially in The Netherlands. Secondly, both analyses on individualization and privatization can lead to different results as one concept might have increased over time, whilst the other concept may have decreased over time. The results on these sub dimensions however still need to put in the context of one concept: personalization. The two distinct dimensions are: individualization and privatization. Individualization is described as a process where a change over time is assumed toward a heightened focus on politicians in newspaper coverage compared to a focus on political parties. The first sub variable of individualization; general visibility is referred to as a shift in focus on politicians rather than parties in political newspaper coverage (Van Aelst et al., 2012).

H1: The relative attention in Dutch political newspaper coverage for individual politicians has increased since 1950 compared to the attention for political parties.

The second sub variable of individualization; concentrated visibility refers to a shift in focus on political leaders in news coverage (Van Aelst et al., 2012).

10 H2: The relative attention in Dutch political newspaper coverage for party leaders has increased since 1950 compared to the attention for political parties.

Privatization relates to a change over time where individual politicians are put in a more private context rather than in a context that is politically relevant and substantive. Privatization consists of two sub dimensions: personal characteristics and personal life of the politician. The concept of personal characteristics refers to an increase of characteristics mentioned related to the personal life of politicians in news coverage (Van Aelst et al., 2012).

H3: The relative attention in Dutch political newspaper coverage for characteristics of politicians in a personal context has increased since 1950 compared to the attention for characteristics of politicians in a political context.

The second sub variable of privatization, personal life news involves an increase of media coverage on the personal life of the politician (Van Aelst et al., 2012).

H4: The relative attention in Dutch political newspaper coverage for the personal life of the politician has increased since 1950 compared to the attention for substantive political topics and debates.

Method This study analyses whether personalization in political newspaper coverage has increased since 1950 in the Dutch context through longitudinal automatic and manual content analysis. Both sub dimensions of personalization; individualization and privatization were measured.

News data selection and filtering 4.625.650 articles from the popular daily newspaper de Telegraaf between 1945 and 2011 were retrieved from the archive of the Dutch Royal Library (Koninklijke Bibliotheek) and Lexis Nexis. All articles were stored into AmCAT (Van Atteveldt, 2008), a web-based database and coding program. The dataset was filtered for domestic political news. For the filtering of domestic political news the program R was used. R is an open sourced programming language and software environment for statistical computing and graphics. In order to filter for domestic political news, two codebooks were created in AmCAT. One codebook was created to filter for political news. This codebook contained three categories: individual politicians, party leaders and political parties. In the first category all the members of parliament and ministers since 1950 were included. In the second category all party leaders, from parties that had one or more seats in

11 parliament since 1950, were included. In the third category all parties, that had one or more seats in parliament since 1950, were included. Furthermore a query that included politically related words was added to the codebook to ensure the inclusion of political news articles that did not include mentions of politicians, party leaders or parties. Examples of words that were included are ‘minister’, ‘parliament’ and ‘ministry of the interior’. The full query can be found in codebook I in Appendix I. Another codebook was created to filter for domestic news. This codebook contained two categories: domestic and foreign. In the first category names of Dutch cities and towns were included. In the second category names of foreign countries and capital cities were included. Codebook II can be found in Appendix I. All articles that contained at least one political term were included. Articles that contained a foreign keyword were excluded unless they also contained a domestic keyword. This resulted in a dataset of 1.710.800 newspaper articles. The dataset does not contain an equal amount of articles in every time period. There are more articles included from the period of 1950 until 1994, which were retrieved from the archive of the Dutch Royal Library, than from the period between 1999 and 2011, which were retrieved from Lexis Nexis. There are no articles present in the dataset from the period between 1994 and 1999. In order to factor this in, all variables were calculated as a ratio as described in Table 1..

Operationalization of personalization variables The conceptualization and operationalization of personalization of Rahat and Sheafer (2007) and Van Aelst et al., (2012) are applied. Table 1. shows how personalization of political newspaper coverage is operationalized.

12 Table 1. Operationalization of personalization and its (sub)dimensions

Variable Dimension Sub dimension Explanation Method Calculation/ approach

The relative attention for Automatic content N individual politicians/ (N individual politicians analysis individual politicians + N General compared to the total amount political parties) visibility Individualization of attention for political

parties in political newspaper coverage. The relative attention for Automatic content N party leaders/ (N individual party leaders compared to analysis politicians + N political Concentrated Personalization the total amount of attention parties) visibility for individual politicians and

political parties in political newspaper coverage. The relative attention for Manual content Three categories: 1. Private characteristics of the analysis characteristics 2. Political politician in a personal characteristics 3. Both private context compared to the and political characteristics Privatization attention for characteristics Characteristics in a political context. N Of characteristics of the politician mentioned in personal

context/ (N Of characteristics mentioned in personal context + N Of characteristics mentioned in political context) The relative attention for the Manual content Four categories: 1. Coverage personal life of the politician analysis of the family 2. Past life and compared to the attention for upbringing 3. Leisure time 4. political topics/debates Love life. Indicators can be Personal life of combined to a variable. the politician

N Of articles about the personal life/ (N Of articles about the personal life + N Of articles about issue news)

13 The variable personalization consists of two dimensions, namely individualization and privatization. Individualization further differentiates between general visibility and concentrated visibility. All four variables were calculated as a ratio. General visibility and concentrated visibility are operationalized as follows:

General visibility The relative attention for individual politicians compared to the total amount of attention for political parties in political newspaper coverage.

Concentrated visibility The relative attention for party leaders compared to the total amount of attention for all individual politicians and political parties in political newspaper coverage.

Privatization further differentiates between the personal characteristics of the politician and the personal life of the politician. Personal characteristics of the politician and personal life of the politician are operationalized as follows:

Characteristics of the politician The relative attention for characteristics of the politician in a personal context compared to the attention for characteristics in a political context.

Personal life of the politician The relative attention for the personal life of the politician compared to the attention for political topics/debates.

Analysis

Individualization In order to create the variables, general and concentrated visibility, individual politicians, party leaders and political parties were counted through an automatic content analysis. The programmes R and AmCAT (Van Atteveldt, 2008) were used to query for the individual politicians, party leaders and political parties. The search terms that were used to query for these three categories can be found in the first codebook, which is enclosed in Appendix I.

Privatization To identify the two sub variables of privatization, the domestic political news dataset was manually coded using the coding instructions given by Van Aelst and his colleagues (2012). 50

14 articles per decade were coded (300 articles in total). The first sub variable of privatization, characteristics of a politician, was coded on sentence level. The following set of characteristics was included: appearance, rhetorical skills, morality, competence and leadership. Subsequently all characteristics were evaluated by their context of presentation: political context, personal context or both political and personal context. The political context refers to all activities and statements inside the political arena (e.g. during a campaign or in parliament) and the personal context refers to all activities and statements outside the political arena (e.g. related to the love life or family activities) (Van Aelst et al., 2012). Additionally the politicians were coded for gender and whether or not the characteristic was mentioned during a year in which elections were held. The coding schema can be found in Table 4. In Appendix II. The second sub variable of privatization, personal life of the politician, was coded on article level. When the article contained news on family life (e.g. family relationships), past life and upbringing (e.g. biographical information), leisure time (e.g. hobbies) or love life (e.g. affairs and marriage), it was annotated as news about the personal life of the politician. In order to calculate the relative attention for the personal life of the politician compared to the attention for political topics in Dutch political news, issue news was annotated as well. Articles about substantive issues that have been discussed in parliament were coded as issue news. The articles were also coded for whether or not the article was published during a year in which elections were held. The full coding schema can be found in Table 5. in Appendix II. To test the inter coder reliability for the measures presented above, Krippendorf’s α was calculated. Two coders independently coded 60 articles of the domestic political news data. Krippendorf’s α amounted to .81 for the evaluation of (ir)relevant articles, .78 for personal life news, .65 for issue news, .68 for the evaluation of (ir)relevant sentences and 1 for the context variable, indicating acceptable to good reliability.

Results In this chapter the results of this study will be presented. The main aim of this study is to test whether personalization and its dimensions individualization and privatization have increased over the decades since 1950.

Changes in personalization over time: individualization

General visibility General visibility is described as a shift in focus on politicians rather than parties in political newspaper coverage (Van Aelst et al., 2012). Figure 1. shows the degree of general visibility over time, displaying the attention share for individual politicians in political newspaper coverage. The relative attention for individual politicians has increased since 1950. The trend shows an increase in the period of 1966 until 1984. After 1984 the trend displays a decrease of general

15 visibility with a high peak starting in 2002. A simple correlation test indicates a significant positive relationship between time and general visibility between 1950 and 1994, r(43) = 0.72, p < .001 and a non-significant positive relationship between time and general visibility between 1999 and 2011, r(11) = 0.36, p = .256. The overall trend of general visibility in the period of 1950 until 2011 is positive, r(56) = 0.64, p < .001, indicating a significant positive relationship between time and general visibility. The results presented in Figure 1. thus support H1: the relative attention in Dutch political newspaper coverage for individual politicians has increased since 1950.

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0 1940 1950 1960 1970 1980 1990 2000 2010 2020

Figure 1. Degree of general visibility in political newspaper coverage of De Telegraaf aggregated by year (1950-2011).

To test the effect of election year on general visibility a regression analysis was conducted: a medium effect regression model (R2 = .47) showed that year (b= .003, t (55) = 6,63, p < .001) and election year (b = -.037, t (55) = -2,25, p < .05) both have a significant effect on concentrated visibility. Where year has a slight positive effect and election year a slight negative effect on concentrated visibility.

Concentrated visibility Concentrated visibility refers to a shift in focus on political leaders in news coverage (Van Aelst et al., 2012). Figure 2. shows the degree of concentrated visibility over time, displaying the attention share of party leaders in political newspaper coverage. The relative attention for party leaders has increased since 1950. The trend shows an overall increase starting in 1950 with higher peaks in 1957 and 1994. The graph also shows an upward trend starting in 1999. A correlation test between time and concentrated visibility indicates a significant positive

16 relationship between 1950 and 1994, r(43) = 0.88, p <.001 and a non-significant positive relationship between 1999 and 2011, r(11) = 0.45, p = .45. The overall trend of concentrated visibility in the period of 1950 until 2011 is positive, r(56) = 0.92, p <.001, indicating a significant positive relationship between time and concentrated visibility. The results presented in Figure 2. thus support H2: the relative attention in Dutch political newspaper coverage for party leaders has increased since 1950.

0.16

0.14

0.12

0.1

0.08

0.06

0.04

0.02

0 1940 1950 1960 1970 1980 1990 2000 2010 2020

Figure 2. Degree of concentrated visibility in political newspaper coverage of De Telegraaf aggregated by year (1950-2011).

Furthermore a large effect regression model (R2 = .83) showed that there is a slight positive yet significant effect of year on concentrated visibility (b=0.002, t (55) = 16,94, p < .001). Election year on the other hand was found not to have a significant effect on concentrated visibility (b = 0.001, t (55) = 0.27 p = .79).

Changes in personalization over time: privatization

Characteristics of the politician Before the results of the characteristics, referred to as an increase of characteristics mentioned related to the personal life of politicians in news coverage (Van Aelst et al., 2012), are discussed, I first present the number of measured characteristics, on which the results are based. The third figure displays the total number per category of the measured characteristics morality, competence, appearance, credibility and rhetorical skills mentioned in Dutch news coverage per decade. Figure 3. shows an overall growth of mentions of characteristics. The characteristic

17 competence is clearly most mentioned during the analysed timeframe and shows a positive trend over time. The other four characteristics are mentioned less often but also show a slight positive trend with the exception of the characteristic rhetorical skills, which shows no growth over time.

35

30

25

20

15

10

5

0 1940 1950 1960 1970 1980 1990 2000 2010 Morality Competence Appearance Credibility Rhetorical skills

Figure 3. Total number per category of characteristics mentioned in Dutch political newspaper coverage per decade (1950-2000).

The relative attention for characteristics mentioned in a private context has decreased since 1950. The results of the analysis show an overall significant negative trend of characteristics mentioned in a private context since 1950, ( r(4) = -0.89, p = < .05). These reults thus do not support H3: the relative attention in Dutch political newspaper coverage for characteristics of politicians in a personal context has not increased since 1950. In the early decades of the timeframe under study, a higher score of characteristics mentioned in a private context was measured. Characteristics such as appearance and competence are commonly discussed in a private context. Sidney van den Bergh for example who was Minister of Defence at that time is described as a ‘real morning person’ who is not very active at night2 while , Member of Parliament then, is being evaluated by appearance characteristics that are not politically relevant such as his blue eyes and blonde

2 Ochtendmens (19 november 1966). De Telegraaf, pp. 5 3 Vertrouwen (17 december 1970). De Telegraaf, pp. 5 4 Contact (26 maart 1969). De Telegraaf, pp. 19

18 hair3. Former Prime Minister and Member of Parliament at the time, is described as a ‘true loyal friend’4. From 1980 onwards characteristics of politicians are less often described in a private context and whenever their characteristics are discussed in a more private context the evaluations seem to fit a context that is both private and political. MP Felix Rottenberg for example is evaluated as cheerful and enthusiastic and therefore seems very suitable as a for the labour party and Prime Minister who is being praised by his electorate for his self-mockery and authenticity5. The above-mentioned characteristics can be considered as private but are in this case politically relevant as well. As illustrated in Figure 4., which shows the count of characteristics of politicians mentioned in a private or political context, the private context seems less important over the years while the political context of characteristics increases. The private context of characteristics thus did not only decrease, the political context became more important. To summarize: the number of characteristics mentioned in political newspaper coverage increased over time, especially competence, and the context of these characteristics became increasingly politically relevant.

60

50

40

30

20

10

0 1940 1950 1960 1970 1980 1990 2000 2010

Private Political

Figure 4. Count of characteristics of politicians mentioned in a private or political context.

Personal life news Personal life news involves a shift toward a focus in media coverage on the personal life of the politician. A correlation test shows a non-significant relationship between time and personal life

5 Balkenende (1 oktober 2006). De Telegraaf, pp. 7

19 news, r(4)= - 01, p = 0.99. The results thus do not support H4: the relative attention in Dutch political newspaper coverage for the personal life of the politician has not increased since 1950. The trend of personalization found in this study shows an interesting peak starting roughly in the sixties, which corresponds somewhat to the findings presented above on general and concentrated visibility where a growth in both trends starting in the sixties is visible. This peak can be explained by several articles about politicians participating in socialite activities such as high-class gatherings and horse race betting. These articles contain details about their marriage, children, love affairs as well as their upper-class homes6. The amount of articles describing politicians attending such elite events diminishes from 1980 onwards. In the decades later on in the analysed timeframe, above-mentioned articles are still published, most of these articles however seem to describe personal life events which are directly linked to the political life of the politicians in concern. An article about a financial scandal in which former Secretary of Economical and Financial Affairs Albert Jan Evenhuis is involved, forms a good example here. In this article several controversial private loans caused him to resign7, where the loans are of personal matter but become political because of their controversial nature. Another example of this phenomenon could be an article about Paul Rosenmöller, party leader of the green party at the time, in which the author evaluates him as not much of an environment activist as his wife and children flew to their villa in Nice by aeroplane8. As can be seen in Figure 5., which shows the count of articles including either news on the personal lives of politicians, issue news or both, personal life news did slightly increase over the years. The trend of issue news, however seems steeper in a positive manner. The graph therefore indicates that whilst news with a focus on the personal lives of the politicians increased over the years, news with a focus on issue news increased even more.

6 Tegenslagen (30 september 1972). De Telegraaf, pp. 33, Symbool (22 december 1975). De Telegraaf, pp. 3, Ererondje (17 juli 1984). De Telegraaf, pp. 4 7 Persprijs voor Evenhuis (VVD) (21 februari 1990). De Telegraaf, pp. 7. 8 Stan Huygens Journaal: Rosenmöller blijft vliegen (27 oktober 2000). De Telegraaf, pp. 6

20 40

35

30

25

20

15

10

5

0 1940 1950 1960 1970 1980 1990 2000 2010

Personal life news Issue news

Figure 5. Count of articles including either news on the personal lives of politicians, issue news or both.

Conclusion This study tested whether personalization and its sub dimensions individualization and privatization have increased over the decades since 1950 in The Netherlands. This chapter will answer this research question and discuss the different hypotheses and their theoretical and practical implications.

The analysis shows that the relative attention in Dutch political newspaper coverage for individual politicians has increased since 1950 (Hypothesis 1). An overall positive trend shows a clear increase starting in the mid-sixties. This result is in line with the findings of Kriesi (2011), who found an increase of personalization in The Netherlands and also applied a broad timeframe (1970-2000). The findings supplement other studies that applied less broad timeframes, which revealed a decrease or continuity (Vliegenthart, 2011; Takens, 2012; Van Santen 2012) of personalization in the Dutch context. It is therefore rather interesting that the results support the common explanation for a decrease or continuity of personalization found in The Netherlands, which is: the idea that a certain change in which the media started to report according to the values of media logic occurred earlier on in the analysed timeframe due to fundamental changes and the slow process of mediatization and its associated phenomena (Kepplinger, 2002; Vliegenthart 2011; Takens, 2012). In addition a regression analysis illustrates that individual politicians have become more visible in political newspaper coverage

21 over the years. During election years, however they seem relatively less visible in Dutch political newspaper coverage. The results also show an increase in the relative attention in Dutch political newspaper coverage for party leaders since 1950 (Hypothesis 2), which seems in line with the findings of Takens (2012). Here too, the overall positive trend shows clear growth from roughly 1960 onwards and is therefore in line with the belief that fundamental changes which plausibly caused personalization, occurred at the beginning of the analysed timeframe (Kepplinger, 2002; Vliegenthart 2011). Further analysis points out that party leaders are evenly visible during election and non-election years but became more visible in media coverage over time. While individual politicians seem less visible during elections, party leaders thus seem evenly visible during election years and non-election years. A plausible explanation for this finding could be that during election times party leaders are seen as most important as it is ‘essentially the leader and not the party who competes for a popular mandate’ (Poguntke & Webb, 2005: p. 9; Takens, 2012). While both individual politicians and party leaders became more visible over time the results point out that the context of their characteristics mentioned became less private over the years (Hypothesis 3). The analysis also illustrates that the context of these characteristics at stake did not only became less private, the number of characteristics of politicians mentioned increased and the political context of these characteristics became more important. This idea does not seem in line with Langer (2007) who found an increase of emphasis on leaders’ personal qualities in a similar analysed timeframe in Britain. Differences in trends that can lead to personalization and different media systems in Britain and The Netherlands could plausibly account for these contradictory findings. Another explanation could be that Langer focused solely on party leaders whilst this study focused on politicians in general. The results related to news about the personal life of politicians indicate that there is no straightforward trend: there is no relationship between time and personal life news. It can therefore be stated that personal life news has not increased over the years in the Dutch context (Hypothesis 4). The trend illustrates a peak from 1960 onwards of news about the private life of the representatives. This peak is filled with articles about politicians attending elite events and the associated scandals and gossip. The results also illustrate a clear drop after this period including several published articles on politicians participating in high-class events. This drop does not seem to be caused by a decrease of news on the personal life of politicians, it seems to be caused by a steep increase of issue news. These results do not seem in line with Langers’ (2007) overall findings on references to leaders’ personal lives. In contrast to these findings, Langer found a steep increase of news on the personal life of the politician in the last two decades.

22 To conclude and answer the research question of this study then:

RQ: Did the amount of personalized political newspaper coverage increase since 1950 in The Netherlands?

Individual politicians and party leaders became more visible over the years, the context of their characteristics mentioned becomes less private and there is no clear trend of news about the personal life of the politician in Dutch political newspaper coverage with a drop of personal life news from 1970 onwards. It thus can be stated that individualization increased over the years, the findings on privatization however are less straightforward: the context of characteristics of politicians mentioned became less private and personal life news showed no clear trend over the years but became less important during the later decades of the timeframe. The answer to the research question of this study is therefore as follows: personalized newspaper coverage has increased over the past decades in The Netherlands as the focus of the news is increasingly put on the politician as an individual, the focus of the coverage around these individual politicians mentioned became less centered around their privately related skills and their personal life as politically relevant skills, especially competence, and substantive issues relatively became more important over the years.

Theoretical and practical implications The overall increase of personalization in The Netherlands over time, found in this study, seems in line with Kriesi’s (2011) study. As mentioned above, the results presented in this study, also support the theory and common explanation for earlier findings of no significant increase of personalization in the Dutch context in the later decades of the last century and the beginning of the current one (Vliegenthart, 2011; Takens, 2012), that is: fundamental changes, which resulted in a tendency towards reporting according to the values of media logic, occurred plausibly in earlier decades the of the last century after World War II (Kepplinger, 2002; Vliegenthart, 2011; Takens, 2012). These above-mentioned interlinked changes are: first, the process of depillarization which refers to secularization, ‘de-ideologising’ and the loosening of ties between pillarized organizations, such as political parties, pressure groups and media outlets (Van Kempen, 2007; Hallin & Mancini, 2004; Dalton et al., 2000; Lijphart 1990). Second, the institutionalization of the media (Takens, 2012: Cook, 1998; Rahat & Sheafer, 2007) that can be regarded as a development caused by the depillarization process as the press became an autonomous institution not working closely with pillarized organizations but in service of the public and its interest. Third, due to another important development: the changing media landscape the press started to see its public as consumers instead of citizens (Schudson, 2002; Van Aelst et al., 2012; Mazzoleni & Schulz, 1999; Adam & Maier, 2010; McManus, 2009; Takens,

23 2012). These factors did not only plausibly result in journalists reporting according to the values of media logic, it forced political actors into changing their coverage-generating strategies (Schudson, 2002; Meyer, 2002). According to the findings of this study these interlinked factors do seem to be part of a plausible explanation that needs to be taken into serious consideration. This study therefore contributes to a less scattered image when it comes to the phenomenon of personalized newspaper coverage in the Dutch context: personalization did increase in The Netherlands, the changes toward that increase however, occurred earlier on than expected at first. The findings of this study also bring along some normative implications as it is argued that the phenomenon of personalization can clash with some models of democracy and the associated requirements as it could possibly put focus on personalities rather than substantive issues. Although it is up for discussion which democratic model, proposed by Strömbäck (2006), fits The Netherlands best, it can definitely be stated that the democratic system at least expects its citizens to cast a well-informed vote based on current societal issues. This study showed that personalization has increased over the years and therefore it seems likely that citizens are being increasingly hindered in their potential effort to cast a well-informed vote. A simple increase of individualization however does not always imply that there is less attention for substantive issues and is therefore not necessarily a threat to the democratic system of The Netherlands (Oegema & Kleinnijenhuis, 2000; Van Aelst, 2012; Van Santen, 2012). Although the increase of individualization does not inevitably cause a threat to democracy in the Dutch context, a potential increase of privatization in political newspaper coverage can be a cause for concern as privatization, could not only harm the public discourse it is also said to stimulate cynicism amongst the public. Cynicism could hinder citizens in being or becoming interested in politics and engaging in public life and discussions (Hart, 1992; Schudson, 2002; Van Aelst, 2012), which would not be beneficial to the democratic system in The Netherlands. The results of this study however illustrate that the context of characteristics of politicians mentioned in a private angle decreased. Not only because it slightly decreased itself, it was caused by an increase of characteristics mentioned in a politically relevant context. This study also showed that personal life news follows no clear trend over the years and shows a drop in the later decades caused by an increase of news on substantive issues. The findings of this study thus support the idea that a simple increase of a focus on the politician as an individual in political reporting is not necessarily a threat to democracy as this study shows that the overall reporting on politics became more centered around politicians as individuals, the focus of the context around reporting on these politicians however became more politically relevant and more substantive over the years.

24 Discussion Apart from the contributions that this study brings along, it has some limitations as well, which will be discussed in this chapter. This chapter will also shed light on suggestions for future research on the topic of personalization.

Limitations of this study This study has three important limitations that need to be discussed. Firstly, automatic content analysis was applied in order to filter for domestic political news coverage and for the creation of the two sub variables of individualization. The method of automatic content analysis has some major advantages as this method is incredibly time saving and is almost inevitable when working with big data. This method however is accompanied by some problems when it comes to validity, as it might be possible that relevant articles were not included in the filtered set and irrelevant articles were included in the filtered set. In addition, for the creation of the two sub variables of individualization key word search was applied. Therefore references to individual politicians or parties such as ‘he’, ’she’ or ‘party’ were not taken into account and not included in the sub variables of individualization. This limitation must be taken into account when interpreting the results. Second, due to availability, this study only measured personalization and its sub variables in one medium. This factor must be taken into account when generalizing the results. It can however also be argued that De Telegraaf is a newspaper with a serious tone but it contains lots of gossip and sports related articles as well. De Telegraaf is therefore one of the least serious-toned newspapers of the serious-toned quality newspapers in The Netherlands. As the results of this study indicate that the reporting on politics in the Dutch context became increasingly politically relevant and more substantive over the years it could be stated that this might be the case for the other Dutch quality newspapers as well. The third limitation can be ascribed to an availability problem as well as this study had to deal with missing data between the period of 1994 and 1999. This limitation did not lead to major problems but the data between this period of time could have been a contribution to the insights that have been gained as a result of this study.

Suggestions future research on the topic of personalization The above-mentioned problem of validity caused by the method of automatic content analysis could be diminished in future research. This specific problem calls for an interdisciplinary approach in cooperation with the field of computational linguistics, in which a method could be developed where so called ‘co-references’ can be recognized automatically by co-reference resolution. Co-reference ‘occurs when two or more expressions refer to the same person or

25 thing’9. When this method is applied it could contribute to a better insight in the phenomenon of personalization in political news coverage. It is also suggested to look into possibilities of automatic recognition when it comes to the identifying of the sub variables of privatization, as it is currently only possible to study a limited number of articles, as the identifying of these variables needs to be done manually. This challenge calls for an interdisciplinary approach as well, in which the possibilities of machine learning for solving this problem should be studied carefully. For a better understanding of the phenomenon of personalization in The Netherlands it is of great importance that future studies look into the other quality Dutch news outlets, such as other newspapers and television broadcasts, as well. This will not only contribute to a better insight of the process of personalization in the Dutch context it would also be more safely to generalize the results to The Netherlands as a unique case. Additionally it is suggested to look into other variables and indicators of media logic such as negativity and horse race news in a broad timeframe as applied in this study as well. This could be a better foundation to form a more complete notion of the phenomenon of media logic in The Netherlands. Furthermore it is suggested to study the possible effects that are related to the shift that is found in political newspaper reporting in The Netherlands. It would be beneficial to the understanding of the phenomenon of personalization to analyse the effects of this shift by looking at voter turnout and the dynamics of the relationship between political newspaper reporting and politics since 1950. This could be analysed by comparing the results of this study to minutes of meetings and debates held in parliament and reports of numbers concerning voter turnout. Questions at stake would be related to the shift that is found and the potential negative or positive effect on voter turnout as well as whether it was the press that initially forced politicians to become more visible, compared to the party or if this development was initiated by politics that caused the press to focus increasingly on politicians as individuals.

References Altheide, David L. and Robert P. Snow (1979) Media Logic. Beverly Hills: Sage. Bird, S.E. & Dardenne, R.W. (2009). Rethinking News and Myth as Storytelling. In K. Wahl- Jorgensen and T. Hanitzsch, The handbook of Journlism Studies, pp. 205-217. New York: Routledge. Boukes, M. & Boomgaarden, G. (2014). Soft news with hard consequences? Introducing a nuanced measure of soft versus hard news exposure and its relationship with political cynicism. Communication Research, 1-31. Doi: 10.1177/0093650214537520.

9 Co-reference, consulted from: http://en.wikipedia.org/wiki/Coreference

26 Brants, Kees and Philip Van Praag (2006) Signs of Media Logic. Half a Century of Political Communication in the Netherlands. Javnost-The Public 13(1), 25-40. Cook, T.J. (1998) Governing with the news: The news media as a political institution. London: University of Chicago Press. Dalton, Russell J., I. McAllister and M.P. Wattenberg (2000) ‘The Consequences of Partisan Dealignment’, in J. Russell Dalton and M.P. Wattenberg (eds) Parties Without Partisans: Polit- ical Change in Advanced Industrial Democracies, pp. 37–63. Oxford: Oxford University Press. Eide M. 1997. A new kind of newspaper? Understanding a popularization process. Media Culture Soc. 19:173–82. Gans, H. J. (2009). Can popularization help the news media? In B. Zelizer (Ed.), The changing faces of journalism: Tabloidization, technology and truthiness (pp. 17-28). Abingdon, UK: Routledge. Hallin, D. C. & P. Mancini (2004). Comparing Media Systems. Cambridge: University Press. Hanitzsch, T. (2007), Deconstructing Journalism Culture: Toward a Universal Theory. Communication Theory, 17: 367–385. doi: 10.1111/j.1468-2885.2007.00303.x Hart, R. (1992). Seducing America. How Television Charms the Modern Voter. London: SAGE. Jebril, N., Albaek, E., De Vreese, C.H. (2013), Infotainment, cynicism and democracy: The effects of privatization vs personalization in the news. European Journal of Communication. Doi: 10.1177/0267323112468683. Kepplinger, H.M. (2002). Mediatization of politics: Theory and data. Journal of Communication, 52 (4), 972-986. doi: 10.1111/j.1460-2466.2002.tb02584.x Kriesi, H.P. (2011). Personalization of national election campaigns. Party Politics, 18 (6), 825- 844. doi: 10.1177/1354068810389643 Langer, A. I. (2007). A Historical Exploration of the Personalisation of Politics in the Print Media: The British Prime Ministers (1945-1999). Parliamentary Affairs , 60(3), 371-387. doi: 10.1093/pa/gsm028. Lijphart, A. (1990). Verzuiling, pacificatie en kentering in de Nederlandse Politiek. : Becht. Mazzoleni, G., Schulz, W. (1999). “Mediatization” of politics: A challenge for democracy? Political Communication, 16, 247–262. McManus, J.H. (2009). The Commercialization of the news. In K. Wahl-Jorgensen and T. Hanitzsch, The handbook of Journlism Studies, pp. 205-217. New York: Routledge. Meyer, Thomas (2002). Media Democracy: How the Media Colonize Politics. Cambridge: Polity Press.

27 Oegema, D., & Kleinnijenhuis, J. (2000). Personalization! in Television NewsιΑ 13-Wave Servey Study to Assess Effects of Text and Footage. Communications, 25(1), 43 Rahat, G. & Sheafer, T. (2007) The Personalization(s) of Politics: 1949-2003. Political Communication 24(1), 65-80. RMO (2003) Medialogica: Over het Krachtenveld tussen Burgers, Media en Politiek. 's Gravenhage: SDU. Schudson, M. (2002). The news media as political institutions. Annual Review of Political Science, 5, 249-269. doi: 10.1146/annurev.polisci.5.111201.115816 Silke, A. & Maier, M. (2010) Personalization of Politics: A Critical Review and Agenda for Research, in pp. 213-258 Salmon, C.T. (Ed.), Communication Yearbook. New York: Routledge. Sigelman, L., & Bullock, D. (1991). Candidates, issues, horse races, and hoopla: Presidential campaign coverage, 1888–1988. American Politics Quarterly, 19, 5–32. Strömbäck, J. (2006). In search of a standard: Four models of democracy and their normative implications for journalism. Journalism Studies, 6(3), 331-345. doi: 10.1080/14616700500131950 Strömbäck, J. (2008). Four phases of mediatization: An analysis of the mediatization of politics. The International Journal of Press/Politics, 13(3), 228-246. doi: 10.1177/1940161208319097 Takens, J. (2012). Media logic and electoral democracy. (Dissertation, Vrije Universiteit, Amsterdam). Underwood, D. (2001). Reporting and the push for market-oriented journalism: Media organizations as businesses, in pp. 99-116 Bennett, W.L. & Entman, R.M. (Eds.), Mediated Politics: Communication in the Future of Democracy. Cambridge: Cambridge University Press. Van Aelst, Peter, Tamir Sheafer and James Stanyer (2012) The Personalization of Mediated Political Communication: A Review of Concepts, Operationalizations and Key Findings. Journalism 13(2), 203-220. Van Kempen, H. (2007) Media-Party Parallelism and Its Effects: A Cross-National Comparative Study. Political Communication, 24(3), 303-320. 10.1080/10584600701471674 Van Santen, R. (2012) Popularization and personalization: a historical and cultural analysis of 50 years of Dutch political television journalism. (Dissertation, Universiteit van Amsterdam). Vliegenthart et al., (2011). Changes in political news coverage: Personalization, conflict and negativity in British and Dutch newspapers. In Brants, K. & Voltmer, K. (eds.), Political communication in postmodern democracy: Challenging the primacy of politics (p. 92-110). Great Britain: Palgrave Macmillan.

28 Wattenberg, M.P. (1998) The Decline of American Political Parties, 1952–1992 Cambridge: Harvard University Press. Wilke, J. & Reinemann, C. (2001) ‘Do the Candidates Matter? Long-Term Trends of Campaign Coverage – A Study of the German Press since 1949’. European Journal of Communication 16(3), 291–314.

Appendix I

Codebook I IndividualPoliticians # “Piet Aalberse” “” “Jacques Aarden” “Gijs van Ardenne” “Harry Aarts” “” “” “Ine Aasted-Madsen-van Stiphout” “Ron Abel” “Hette Abma” “Jan Achttienribbe-Buijs” “” “” “” “Liesbeth Aiking-van Wageningen” “Hans van den Akker” “Sjouke Akkerman” “” “Leo Albering” “Wim Albers” “Ton Alblas” “” “” “Rendert Algera” “ Marius van Amelsvoort” “Mieke Andela-Baur” “” “Jan Andriessen” “Kees van den Anker” “” “Hendrik Jan Ankersmit” “Thanasis Apostolou” “” “-van der Hoeven” “” “” “Gerard van As” “Anneke Assen” “Wim Assmann” “” “Marijke Augusteijn-Esser” “” “ Naïma Azough” “Jan van Baal” “” “Ton van Baars” “Ies Baart” “Ada Baas-Jansen” “Martine Baay-Timmerman” “Joannes Herman Bachg” “Fons Baeten” “Marcus Bakker” “” “Bert Bakker” “Ben Bakker” “Ernst Bakker” “Eric Balemans” “Jan Peter Balkenende” “Nel Barendregt” “” “Jacques Baruch” “” “-de Bruijn” “Cas van Beek” “” “” “Jan Beekmans” “Pieter Beelaerts van Blokland” “Pol de Beer” “” “” “Jacqueline Beijlen-Geerts” “Marten Beinema” “Leo de Bekker” “Judith Belinfante” “Jhim van Bemmel” “Jan van Bennekom” “” “Ybeltje Berckmoes-Duindam” “Ed Berg” “Koos van den Berg” “Joop van den Berg” “” “Jan Berger” “Rob van den Bergh” “Harry van den Bergh” “” “Cees Berkhouwer” “-Jansen” “Ino van den Besselaar” “Marianne Besselink” “Klaas Beuker” “Bouke Beumer” “” “Pieter Jan Biesheuvel” “Arend Biewenga” “Anke Bijleveld-Schouten” “” “Marja Bijsterveldt-Vliegenthart” “Eddy Bilder” “Suzanne Bischoff van Heemskerck” “” “ ” “Jan Dirk Blaauw” “Pieter Blaisse” “Elly Blanksma- van den Heuvel” “Piet Blauw” “Anke Blerck-Woerdman” “Liesbeth Bloemen” “” “Jaap Blom” “Luuk Blom” “” “Pieter Bode” “” “Gerard Boekhoven jr” “Jan Boelhouwer” “” “” “Joep de Boer“ ““ “Mieke Boers-Wijnberg“ “” “Kees Boertien” “” “Eppo Bolhuis” “” “” “Jan Bommer” “Desiree Bonis” “Vic Bonke” “” “Arie de Boo” “Leo Boogaard sr” “Rie de Boois” “Gijs Boot” “Fred Borgman” “Bets Borm-Luijkx” “Henk van den Born” “Cor Borst” “-Eilers” “Corstiaan Bos” “” “Bob van den Bos” “” “Clemens Bosman” “Andre Bosman” “” “Samira Bouchibti” “” “” “Bruno Braakhuis” “Reinier Braams” “Johan van de Brake” “” “Arie van den Brand” “Bertus Brandsen” “Cor Brandsma” “Gerda Brautigam” “Mauk de Brauw” “Wiel Bremen” “Cees Bremmer” “Wien van den Brink” “Theo Brinkel” “Laurens-Jan Brinkhorst” “” “” “Jeanette ten Broecke Hoekstra” “” “” “Bert Broekhuis” “” “Philip Brood” “Rinus Broos” “” “Ina Brouwer” “ Rintje van der Brug” “Jan Bruggeman” “Hans Bruggeman jr” “Sieuwert Bruins Slot” “Hanke Bruins” “Hubert Bruls” “Ben van Buel” “Siem Buijs” “Jan Buikema” “” “Vincent van der Burg” “Mieke van der Burg” “” “” “Piet Burggraaf” “” “Flip Buurmeijer” “Michael Rudolph Hendrik Calmeyer” “” “” “Frits Castricum” “Yasemin Cegerek“ “Metin Celik” “Wim

29 du Chatinier” “Oussama Cherribi” “Dick de Cloe” “” “Dolf Coppes” “Johan Cornelissen” “Pam Cornelissen” “Dien Cornelissen” “Clemens Cornielje” “Coskun Coruz” “Han Corver” “Grace Cotterell” “Henk Couprie” “Jean Couzy” “Ernst Cramer” “Ferd Crone” “Frits Daams” “” “Gerrit van Dam” “” “” “Nancy Dankers” “” “Huub Dassen” “Coen Deering” “” “” “” “Fred Dekker” “Suzan Dekker” “Staf Depla” “Norma Dettmeijer Labberton” “Ineke Dezentje Hamming” “” “Isaac Diepenhorst” “Tony van Dijck” “Klaas van Dijk” “” “Jan Jacob van Dijk” “Asje van Dijk” “” “” “Leen van Dijke” “Marjo van Dijken” “” “Stef Dijkman” “” “Gerda Dijksman” “” “Minne Dijkstra” “” “” “” “Carla Dik-Faber” “ ” “” “” “Rie Dirx” “Cor van Dis” “Cor van Dis jr” “” “Jaap van der Doef” “Hans van den Doel” “Theo van den Doel” “Alie Doelman-Pel” “Anneke van Dok-van Weele” “” “Sjef van Dongen” “Leendert Antonie Donker” “Piet-Hein Donner” “” “Kris Douma” “Frits Dragstra” “Willem Drees” “Wim Drees jr” “Herman Drenth” “” “Willem Droesen” “Niesco Dubbelboer” “Desiree Duijkers” “Dirk Duinker” “” “” “Adri Duivesteijn” “Willy Dusarduijn” “Loek Duyn” “Anthony Duynstee” “” “Cor Eberhard” “” “” “Kees Egas” “Eske van Egerschot” “Jan van Eibergen” “Henk van Eijsden” “” “Doeke Eisma” “Klaasje Eisses-Timmerman” “” “Pieter Elfferich” “” “Andre Elissen” “Joop van Elsen” “Wim Elsthout” “Jan Emmens” “Ben Engelbertink” “” “Henk Engelsman” “” “Meiny Epema-Brugman” “Marius Ernsting” “Broos van Erp” “Andree van Es” “Evelien Eshuis” “Nihat Eski” “Berry Esselink” “Theo van Eupen” “” “Albert-Jan Evenhuis” “Hanske Evenhuis-van Essen” “Huib Eversdijk” “Sytze Faber” “Jaap-Jelle Feenstra” “Willem van der Feltz” “Johannes Fens” “Hubert Fermina” “Kathleen Ferrier” “Wim van Fessem” “Thea Fierens” “Ferdinand Fievez” “” “Jan Fokkema” “Jeanne Fortanier-de Wit” “Huub Franssen” “” “Ton Frinking” ”Leon Frissen” “” “Bas de Gaay Fortman” “Dzsingisz Gabor” “Til Gardeniers-Berendsen” “” “Nico Geelkerken” “Vivien van Geen” “” “Wil van Gelder” “Jan Geluk” “” “Harmen Gerbrandij” “Karen Gerbrands” “Pieter Gerbrandy” “Arda Gerkens” “Gerrit Gerritse” “” “” “” “Aart Geurtsen” “Wout van der Gevel” “Lambert Giebels” “” “” “Frans Gijzels” “Chantal Gill'ard” “Nel Ginjaar-Maas” “Francine Giskes” “” “Frans Goedhart” “Marinus van der Goes van Naters” “Pier van Gorkum” “Henk Gortzak” “Wouter Gortzak” “Anneke Goudsmit” “Bob Goudzwaard” “Theo de Graaf” ““ “Arie de Graaf“ “Theo de Graaf“ ““ “Jan de Graaf” “” “” ““ “” ““ “Jacobus Groen” “Egbert Jan Groenink” “Louise Groenman” “Berthe Groensmit-van der Kallen” “Paul de Groot” “” “Hans Grosheide” “” “Hans Gualtherie van Weezel” “Frans der Gun” “Sultan Gunal-Gezer” “Henk de Haan” “Bert Haars” “Johannes de Haas” “Ineke Haas-Berger” “” “” “Jose Hageman” “Kees ten Hagen” “Jan Haken” “” “” “Henk de Hamer” “Mariette Hamer” “” “Evert Harmsen” “Ab Harrewijn” “Wouter van Harselaar” “Henk Hartmeijer” “Floor den Hartog” “Maarten Haverkamp” “Ted Hazekamp” “” “Kees Hazenbosch” “Annemieke van Heel-Kasteel” “Dilia van der Heem-Wagemakers” “Mathieu Heemelaar” “Sari van Heemskerck Pillis-Duvekot” “” “Peter van Heemst” “” “Bob Heeringa” “Enneus Heerma” “” “Ton Heerts” “Frans-Jozef van der Heijden” “Jan Heijmans” “” “Arie van der Hek” “” “Martien van Helvoort” “Jo Hendriks” “Theo Hendriks” “Ben Hennekam” “Jeanine Hennis-Plasschaert” “Mat Herben” “” “Corrie Hermann” “Huub Hermans” “” “Max Hermans” “Ad Hermes” “Ben Hermsen” “” “Took Heroma-Meilink” “Willem Herrebrugh” “Jos Hessels” “Jos Hessing” “Godelieve van Heteren” “Ruud van Heugten” “Chris van den Heuvel” “Nico van den Heuvel” “” “” “” “Rik Hindriks” “” “” “Annet van der Hoek” “Jan Hoekema” “Henk Hoekstra” “

30 “” “Pieter Hofstra” “Theodorus Hogendorp” “Michiel Holtackers” “Victor Honig van den Bossche” “” “Jan Hoogcarspel” “” “” “” “Theo Hooij” “” “Jan ten Hoopen” “Jo Horn” “Louis Horsmans” “Hamid Houda” “Jan van Houwelingen” “” “Hans Huibers” “Coos Huijsen” “Matthijs Huizing” “-Heringa” “Michel van Hulten” “Rein Hummel” “Dolf Hutschemaekers” “Servaas Huys” “Joost van Iersel” “Joseph IJsselmuiden” “Sef Imkamp” “” “Wijnie Jabaaij” “” “Huub Jacobse” “” “Tjalle Jager” “Rikus Jager” “” “Hans Janmaat” “Minouche Janmaat-Abee” “Henk Jans” “Jan Hendrik Jansen” “Leo Jansen” “” “Martinus Janssen” “Paul Janssen” “” “Jim Janssen van Raaij” “Hans Jeekel” “Dick Jense” “” “Theo Joekes” “Cisca Joldersma” “” “Jacques de Jong” “Winny de Jong” “Arie de Jong” “Gerrit de Jong” “Mechteld de Jong” “Leon de Jong” “Aat de Jonge” “Pieter Jongeling” “Corien Jonker” “-Lebbink” “Joke Jorritsma-van Oosten” “Gerlof Jukema” “Paul Jungbluth” “Erik Jurgens” “Jacques de Kadt” “Paul Kalma” “” “Flip de Kam” “Kees Kammeraad” “” “Margreet Kamp” “” “Annelien Kappeyne van de Coppello” “” “” “” “Wim Keja” “” “Joke Kersten” “” “Annie Kessel“ “Klaas Keuning” “Sierk Keuning” “Heine Keuning” “Willem Keur” “Mike Keyzer” “Garmt Kieft” “Henk Kikkert” “Eefje Klaassens-Postema” “” “” “Cors Kleijwegt” “Ger Klein” “Norbert Klein” “Jan-Hendrik Klein Molekamp” “Cor Kleisterlee jr” “Reinette Klever” “” “” “Marga Klompe” “” “Jopie Knol” “Henk Knol” “” “Jan Knot” “David Kodde” “Alis Koekkoek” “Hendrik Koekoek” ““ “Theo Koersen” “Helmer Koetje” “Jo van Koeverden” “Gert Koffeman” “” “Truus Kok” “Ton de Kok” “” “Erik Kolfschoten” “Kees Kolthoff” “Hans Kombrink” “” “Jan de Koning” “Jan de Koning” “Marijn de Koning” “Jan Koningh” “Martin Konings” “Bouke van der Kooij” “” “” “Myra Koomen” “Jo Koopman” “” “” “Wim de Kort” “” “Virginie Korte-van Hemel” “Leonardus Kortenhorst” “Roland Kortenhorst” “Jules Kortenhorst” “Willem Kortenoeven” “Simon Korteweg” “” “” “” “Lucy Kortram” “Fatma Koser Kaya” “” “” “Gerard Koudijs” “Jeltien Kraaijeveld-Wouters” “Guus Krahe” “Jan Krajenbrink” “Johannes Kramer” “Ferdinand Kranenburg” “Margot Kraneveldt-van der Veen” “Wim Kremer” “Anneke Krijnen” “” “Tjeerd Krol” “” “” “Hubert Kronenburg” “Toon Krosse” “Annie Krouwel-Vlam” “Joanneke Kruijsen” “” “Kees van Kuijen” “Arie Kuijper” “” “Ada Kuiper-Struyk” “” “Willem de Kwaadsteniet” “Karel van Laak” “Dick Laan” “” “Rein Laan jr” “Saskia Laaper-ter Steege” “” “Cees Laban” “Ineke Lambers-Hacquebard” “Jan Lamberts” “Ursie Lambrechts” “” “Siepie Langedijk-de Jong” “Marianne Langkamp” “Frouwkje Laning- Boersema” “Henk Lankhorst” “Peter Lankhorst” “Ad Lansink” “” “Ali Lazrak” “Rene Leegte” “” “” “Bertus Leerkes” “Meindert Leerling” “” “Andre de Leeuw” “Johan de Leeuw” “Herman van Leeuwen” “Hannie van Leeuwen “ “Hans van Leeuwen” “Pieter Leffertstra” “Bob Leibbrandt” “Gerard van Leijenhorst” “Frans Leijnse” “” “Bram van der Lek” “Trees Lemaire” “William Lemaire” “Paul Lempens” “Arie Lems” “Janmarc Lenards” “Marijke van Lente-Huiskamp” “” “Johanneke Liemburg” “Kees van Lienden” “Theo van Lier” “John Lilipaly” “” “Rene van der Linden” “Patricia Linhard” “” “Rie Lips-Odinot” “Erik van Lith” “” “” “Jan Lonink” “Dirk de Loor” “Andre van der Louw” “” “Anton Lucas” “Anne-Wil Lucas-Smeerdijk” “Eric Lucassen” “Anne-Marie Lucassen- Stauttener” “Ruud Luchtenveld” “Ton Luckers-Bergmans” “Fons Luijben” “” “” “Jos Maenen” “Jan Maenen” “” “” “” “” “” “Jan Masman” “Cor van Mastrigt” “Jan Mastwijk” “Wim Mateman” “Cees van Meel” “” “Cees Meeuwis” “Durk van der Mei” “Wim Meijer” “Jose de Meijer” “Han Meijer” “Wim Meijer” “Els Meijer” “Peter Meijer” “Theo Meijer” “Hendrik Meijerink”

31 “Fre Meis” ““ “Jur Mellema” “” “Chel Mertens” “Jaap Metz” “Jan Meulink” “Ron Meyer” “Bert Middel” “Ruud van Middelkoop” “” “Marinus Mieras” “” “Adriaan van Mierlo” “Gerrit Mik” “Jacques de Milliano” “” “” “Chris Mol” “Hillie Molenaar” “Henk Molleman” “Joep Mommersteeg” “” “Riekus de Mooij Azn” “Corry Moolhuysen-Fase” “Frans Moor” “Harry Moorman” “Perjan Moors” “Corrie Moret-de Jong” “” “Aart Mosterd” “Gerard van Muiden” “” “” “Nel Mulder-van Dam” “Wiel Mulders” “Ina Muller-van Ast” “Karel Nagel” “” “Gerard Nederhorst” “” “Helma Nepperus” “Frans de Neree tot Babberich” “” “Atzo Nicolaï” “Jacques Niederer” “Kees van Nierop” “Frits Niessen” “” “-Wijbenga” ““ “Ruud Nijhof “ “Ad Nijhuis” “Jan Nijland” “” “Jet Nijpels-Hezemans” “” “Agnes Nolte” “Jan van Noord” “Oene Noordenbos” “Saskia Noorman-den Uyl” “Jacobus van de Noort” “Govert Nooteboom” “Harrij Notenboom” “Toon Nuijens” “” “” “Marjet Ockels” “Tara Oedayraj Singh Varma” “Ad Oele” “Niny van Oerle-van der Horst” “” “Bert Oldenbanning” “Annie van Ommeren- Averink” “” “David van Ooijen OP” “Ria Oomen-Ruijten” “Astrid Oosenbrug” “” “Gert Jan Oplaat” “Fadime Orgu” “” “-Martijn” “” “Gerrit-Jan van Otterloo” “” “Greetje den Ouden-Dekkers” “Gonny van Oudenallen” “Rob Oudkerk” “” “Gerritjan van Oven” “Selcuk Ozturk” “Nevin Ozutok” “Annemiek Padt-Jansen” “Frits Palm” “Ab te Pas” “Wim Passtoors” “Marleen de Pater- van der Meer” “Connie Patijn” “” “” “Walter Paulis” “” “Cor van der Peijl” “Rinus Peijnenburg” “Harry Peschar” “Jan Peters” “Willem Peters” “” “Hein Pieper” “Tom Pitstra” “” “Kees van der Ploeg” “Sake van der Ploeg” “Ad Ploeg” “” “Ans Ploeg-Ploeg” “Minke van der Ploeg-Posthumus” “Hugo Polderman” “Stan Poppe” “Remi Poppe” “Henk Pors jr” “Frits Portheine” “Siep Posthumus” “” “Wilfried de Pree” “” “Monique Quint-Maagdenberg” “” “Mohamed Rabbae” “Nirmala Rambocus” “Ram Ramlal” “Francisca Ravestein” “” “Jan Reehorst” “Lambertus Reestman” “Peter Rehwinkel” “Jakob Reitsma” “Patricia Remak” “” “Len Rempt-Halmmans de Jongh” “Frits Reuter” “” “Arie van Rhijn” “Hessel Rienks” “Nicky van ’t Riet” “” “Theo Rietkerk” “Karel van Rijckevorsel” “Leni van Rijn-Vellekoop” “Jan Rijpstra” “Govert Ritmeester” “Lia Roefs” “Peter Roels” “” “Dirk Roemers” “Hein Roethof” “Guikje Roethof” “” “” “Nathalie de Rooij” “” “” “Eugenius Roolvink” “” “Trix de Roos-Consemulder” “Corrie de Roos-Oudegeest” “Riet Roosen-van Pelt” “” “Paul Rosenmoller” “Clemence Ross-van Dorp” “Henk van Rossum” “Andre Rouvoet” “” “Evan Rozenblad” “” “Marianne Ruigrok-Verreijt” “Gustave Ruijs de Beerenbrouck” “Jan van Ruiten” “Jo de Ruiter” “Piet de Ruiter” “” “Nellien de Ruiter” “Piet de Ruiter” “Jacqueline Rutgers” “” “” “Geert Ruygers” “Nora Salomons” “” “” “Everardus van der Sande” “Piet van der Sanden” “Usman Santi” “” “Ge Schaapman” “Afke Schaart” “” “Sijbrand Schagen” “Theo van Schaik” “Maarten Schakel” “Herman Schaper” “Haje Schartman” “Olga Scheltema-de Nie” “Johan Scheps” “Wim Schermerhorn” “Janneke Schermers” “Jaap Scherpenhuizen” “Anton van Schijndel” “Tineke Schilthuis” “Jan Schilthuis” “Arthie Schimmel” “Jan Schinkelshoek” “Wim van het Schip” “Adri Schipper” “” “Johan Schlingemann” “Jan Schmal” “” “Herman Schoemaker” “Gerrit Schoenmakers” “Lubbertus Scholten” “” “Jan-Nico Scholten” “Fred Schonewille” “Eegje Schoo“ “Leo Schoots” “Jan Schouten” “Ad Schouten” “” “” “Arie Jan Schouwenaar” “Annie Schreijer-Pierik” “Gijs Schreuders” “Wim Sijt” “Dirk Sitemaker” “Wybrand Sitemaker” “Melanie Schultz van Haegen-Maas Geesteranus” “Fedde Schurer” “Hein Schuring” “Anoushka Schut-Welkzijn” “” “Wil Schuurman” “Gert-Jan Segers” “Har Seijben” “Johan Sens” “Jos Serrarens” “” “James She” “” “Hannie Singer-Dekker” “Leoni Sipkes” “” “Harm van Sleen” “” “Willem Sloots” “Gerard Slotemaker de Bruïne” “Wietze van der Sluis” “” “” “Pauline Smeets” “Margreeth Smilde” “Jose Smits” “Ries

32 Smits” “” “” “Harry Smulders” “Janneke Snijder-Hazelhoff” “-Downer” “Louis van Son” “Marian Soutendijk-van Appeldoorn” “Andre Mensert Spaanderman” “Fred van der Spek” “” “Bonno Spieker” “” “John Spinks” “Ernst van Splunter” “Laurette Spoelman” “” “” “Henk Staneke” “Antoon Stapelkamp” “” “Hugo van der Steenhoven” “Dick Stellingwerf” “” “” “Mieke Sterk” “” “” “Piet Stoffelen” “Joke Stoffels-van Haaften” “Jacobus Stokman” “Benno Stokvis” “Piet Straub” “” “Theo Stroeken” “Nico Stufkens” “” “Hannie Stuurman” “” “Ton de Swart” “Willie Swildens-Rozendaal” “Zsolt Szabo” “Grace Tanamal” “” “Sjeng Tans” “” “Rob Tazelaar” “” “Haty Tegelaar-Boonacker” “” “Uke Tellegen-Veldstra” “Bas van den Tempel” “Corry Tendeloo” “” “Jan Terpstra” “” “Gerrit Terpstra” “Frans-Joseph van Thiel”“” “” “” “Arnold Tilanus” “Hendrik Tilanus” “Anja Timmer” “” “Varina Tjon-A-Ten” “” “” “” “Evelien Tonkens” “” “Rob van den Toorn” “Rene Toussaint” “” “Maarten van Traa” “Max Tripels” “Danny Tuijnman” “Jan Tuin” “Cor Tuinenburg” “Elida Tuinstra”“Klaas Tuinstra” “Cathy Ubels-Veen” “Be Udink” “Thijs Udo” “” “” “Auke van ” “” “Koos van der Vaart” “Gerrit Valk” “João Varela” “Els Veder-Smit” “Pieter van der Veen Binne” “Nico van der Veen” “” “Eeke van der Veen” “” “Jelleke Veenendaal” “Phia van Veenendaal-van Meggelen” “Pieter ter Veer” “Antoon Veerman” “Elske ter Veld” “Willem van der Velden” “Henk Veldhoen” “” “Jan te Veldhuis” “Jan Veldhuizen” “” “Jakob Vellenga” “” “Jan van de Ven” “” “Albert Venverloo” “” “Hendrikus Verberk” “Bart Verbrugh” “Nellie Verbugt” “” “” “Ton Verdijk” “” “Fenna Vergeer-Mudde” “” “” “Bernard Verhoeven” “” “” “Elias Verkerk” “Will Verkerk” “Nico Verlaan” “Arend Vermaat” “” “Anne Vermeer” “Evert Vermeer” “” “Willem Vermooten” “Machteld Versnel-Schmitz” “Josephine Verspaget” “Thieleman Versteeg” “Eef Verwoert” “Antoinette Vietsch” “Netty de Vink” “Pieter Vis” “Eddy Visch” “Erik Visser” “Martin Visser” “Arno Visser” “Piet de Visser” “” “Marry Visser-van Doorn” “” “” “Frans van Vliet” “Stefanie van Vliet” “” “Thijs van Vlijmen” “” “” “Gijsbertus Vonk” “Joop Voogd” “Poulus Voogd” “Jan de Voogd” “Joel Voordewind” “” “Berend-Jan baron van Voorst tot Voorst” “Arend Voortman” “” “Koos Vorrink” “” “Otto Vos” “Henk Vos” “” “Hella Voute-Droste” “” “” “Jan de Vreeze” “Thom Vreugdenhil” “Yvonne Vriens-Auerbach” “Klaas de Vries” “Jan de Vries” “Bibi de Vries” “Monique de Vries” “” “” “” “” “” “Nicolien van Vroonhoven-Kok” “” “Anna de Waal” “Harry Waalkens” “Harm-Evert Waalkens” “Jules de Waart” “Gerben Wagenaar” “Marja Wagenaar” “Aad Wagenaar” “Hans Wagner” “Tjebbe Walburg” “” “Jan van Walsem” “Henk Waltmans” “Jan-Willem van Waning” “” “Martien van der Weijden” “Steef Weijers” “Toon Weijters” “” “” “”“Ans van der Werf-Terpstra” “Paul Wessels” “Theo Westerhout” “” “Frits van de Wetering” “”“Hans Wiebenga” “Jan-Kees Wiebenga” “Hans Wiegel” “Esme Wiegman-van Meppelen Scheppink” “Ep Wieldraaijer” “Wiebe Wierda” “Ko Wierenga” “Harm Wiersma” “Frans Wijffels” “Peter van Wijmen” “” “Harry Wijnschenk” “Wil Wilbers” “” “Joan Willems” “Wilbert Willems” “Ans Willemse-van der Ploeg” “Jan Wilmans” “Cees van Wingerden” “Bart van Winsen” “Herman Wisselink” “” “” “” “Tineke Witteveen-Hevinga” “” “Andre de Wolf“ “Joop Wolff” “Gerrit-Jan Wolffensperger” “Aleid Wolfsen” “Frans Wolters” “Thijs Woltgens” “Eisso Woltjer” “Gerard ter Woorst” “Joop Worrell” “Bas van ’t Wout” “Christine Wttewaall van Stoetwegen” “Marijke Wuthrich-van der Vlist” “” “” “Keklik Yucel” “Cornelis van der Zaal” “” “Harrie van der Zanden” “Pieter Zandt” “Nancy Zeelenberg” “” “Roelof Zegering Hadders” “” “Piet Zelissen” “Fierous Zeroual”

33 “” “Jan van Zijl” “” “Rinse Zijlstra” “Kees Zijlstra” “Marten Zijlstra” “” “Cor Zonneveld” “Marjet van Zuijlen” “Milos Zvonar” “Johan Zwanikken”

PoliticalParties # “KVP” “Anti-Revolutionaire Partij” “CHU” “CDA” “SGP” “GPV” “RPF” “ChristenUnie” “KNP” “RKPN” “LPF” “Leefbaar Nederland” “PVV” “PvdV” “Boerenpartij” “DS’70” “Nederlandse Middenstands Partij” “Centrumpartij” “Centrumdemocraten” “D66” “PvdD” “PvdA” “CPN” “Pacifistisch Socialistische Partij” “PPR” “EVP” “GroenLinks” “Socialistische Partij” “Algemeen Ouderen Verbond” “Unie 55+” “50 Plus”

PartyLeaders # “Laurentius Nicolaas Deckers” “” “Carl Romme” “Wim de Kort” “Norbert Schmelzer” “Frans Andriessen” “Hans de Boer” “Jan de Koning” “Antoon Veerman” “Anton Roosjen” “Wiert Berghuis” “Anton Roosjen” “Jan Schouten” “Hendrik Willem Tilanus” “Henk Beernink” “Jur Mellema” “Christine Wttewaall van Stoetwegen” “Arnold Tilanus” “Jur Mellema” “Roelof Kruisinga” “Jur Mellema” “Arnold Tilanus” “Roelof Kruisinga” “Gerrit Hendrik Kersten” “Pieter Zandt” “Cor van Dis sr” “Hette Abma” “Henk van Rossum” “Bas van der Vlies” “Kees van der Staaij” “Piet Jongeling” “Bart Verbrugh” “Gert Schutte” “Meindert Leerling” “Leen van Dijke” “Charles Welter” “Klaas Beuker” “Mat Herben” “Harry Wijnschenk” “Gerard van As” “Fred Teeven” “Geert Wilders” “Steven Bierema” “Hendrik Koekoek” “Willem Drees jr” “Jan Berger” “Rudolf Nijhof” “Ab te Pas” “Hans Janmaat” “ Hans van Mierlo” “Jan Terlouw” “” “Maarten Engwirda” “Hans van Mierlo” “Els Borst” “Thom de Graaf” “Boris Dittrich” “Lousewies van der Laan” “Alexander Pechtold” “Marianne Thieme” “Gerben Wagenaar” “Henk Gortzak” “Paul de Groot” “Marcus Bakker” “Ina Brouwer” “Cathy Ubels-Veen” “Jet Nijpels-Hezemans” “Willibrord Verkerk” “Bertus Leerkes” “Marinus van der Goes van Naters”“Leendert Antonie Donker” “Jaap Burger” “Anne Vondeling” “Gerard Nederhorst” “Joop den Uyl” “Ed van Thijn” “Wim Meijer” “Wim Kok” “Thijs Woltgens” “Jacques Wallage” “Ad Melkert” “Jeltje van Nieuwenhoven” “Wouter Bos” “Jacques Tichelaar” “Mariette Hamer” “Job Cohen” “Jeroen Dijsselbloem” “Diederik Samsom” “Dries van Agt” “Wim Aantjes” “Ruud Lubbers” “Bert de Vries” “Elco Brinkman” “Enneus Heerma” “Jaap de Hoop Scheffer” “Jan Peter Balkenende” “Maxime Verhagen” “Pieter van Geel” “Sybrand van Haersma Buma” “Ria Beckers” “Peter Lankhorst” “Paul Rosenmoller” “Marijke Vos” “Femke Halsema” “Jolande Sap” “Bram van Ojik” “Remi Poppe” “Hans van Hooft sr” “Jan Marijnissen” “Emile Roemer” “Andree van Es” “Fred van der Spek” “Bram van der Lek” “Hans Wiebenga” “Henk Lankhorst” “Nico van der Veen” “Ria Beckers” “Bas de Gaay Fortman” “Jacques Aarden” “” “Andre Rouvoet” “Arie Slob” “Henk Krol” “Norbert Klein”

PoliticalTerms # Minister kamerlid staatssecretaris VVD'er PVV'er PvdA'er CDA'er CU-er PvdD'er LPF'er RKPN'er RPF'er GPV'er SGP'er EVP'er PPR'er CPN'er VVD-lid PVV-lid PvdA-lid CDA-lid PvdD-lid D66-lid LPF-lid RKPN-lid RPF-lid GPV-lid SGP-lid EVP-lid PPR-lid CPN-lid Groenlinks-lid "50 plus lid" "Unie 55+ lid" Centrumdemocraten-lid "Nederlandse Middenstands Partij-lid" DS'70-lid Boerenpartij-lid minister-president premier partijleider lijsttrekker *lijsttrekker vicepremier viceminister-president vice-premier oud-premier oud-minister *kamerlid fractievoorzitter politiek? *premier *minister* "Tweede Kamer" "Eerste Kamer" *Kamerl* fractie* bewind* kabinet* parlement* “binnenlandse zaken” “buitenlandse zaken” *democrat* libera* regering* politic* overheid* conservatief conservatieve? progressief progressieve? kamervragen wethouder wets* Staten-Generaal verzorgingsstaat “Sociale Zaken en Werkgelegenheid” “Raad van State” staatssecretaris defensie volksgezondheid

Codebook II Binnenland # Nederland* Holland* Zierikzee Zevenbergen Zevenaar Zaltbommel Zaandam Woudrichem Workum Woerden Winschoten Winkel Wilsum Willemstad Wijk bij Duurstede Wieringen Westwoud Westkapelle Wessem Weert Wageningen Waalwijk Vollenhove Vlissingen Vianen Venlo Veere Valkenburg Tiel Thorn Tholen Texel Terneuzen Terborg Susteren Steenwijk Steenbergen Stavoren Staverden Spanbroek Sneek Sluis Sloten Sittard Sint-Oedenrode Sint-Maartensdijk Sint Anna ter

34 Muiden Schoonhoven Schellinkhout Schagen Roermond Rijssen Rhenen Reimerswaal Ravenstein Purmerend Philippine Oudewater Oss Ootmarsum Oosterhout Oostburg Ommen Oldenzaal Oisterwijk Nijkerk Nieuwstadt Nieuwpoort Stede Niedorp Naarden Muiden Montfort Montfoort Monnickendam Middelburg Meppel Megen Medemblik Maassluis Maasbommel Lochem Leerdam Langedijk Laag-Keppel Kortgene Klundert Kessel Kampen IJzendijke IJsselstein IJlst Hulst Huissen Hoorn Hoogwoud Hindeloopen Heusden Heukelum ’s-Hertogenbosch Hem Helmond 's-Heerenberg Heenvliet Hattem Hasselt Harlingen Harderwijk Hardenberg Hagestein Haastrecht Haarlem Groenlo ’s-Gravenzande ’s-Gravenhage Grave Gramsbergen Grafhorst Gouda Gorinchem Goor Goes Goedereede Gennep voor Genemuiden Gendt Geervliet Geertruidenberg Franeker Enschede Enkhuizen Elburg Eemnes-Buiten Eemnes-Binnen Eembrugge Edam Echt Domburg Dokkum Doetinchem Doesburg Diepenheim Deventer Delden Culemborg Coevorden Buren Bunschoten Brouwershaven Bronkhorst Stede Broek Brielle Bredevoort voor Borculo Bolsward Biervliet Beverwijk Berlikum Batenburg BarsingerhornBaarn Axel Austerlitz Assen Asperen Arnemuiden Appingedam Amsterdam Ammerstol Amersfoort Ameide Almelo Abbekerk Aardenburg

Buitenland # ijsland ierland hongarije honduras haiti guyana "guinee bissau" guinee guatemala griekenland grenada georgie gambia gabon frankrijk finland filipijnen fiji ethiopie estland eritrea "equatoriaal guinea" el salvador egypte ecuador duitsland "dominicaanse republiek" dominica denemarken cyprus cuba "costa rica" "congo kinshasa" "congo brazzaville" comoren colombia china chili "centraal afrikaanse republiek" canada cambodja burundi "burkina faso" bulgarije brunei brazilie botswana "bosnie en herzegovina" bolivia bhutan benin belize belgie barbados bangladesh bahrein bahamas azerbeidzjan australie Armenie argentinie "antigua en barbuda" angola andorra algerije albanie soedanese oegandese zwitserse zweedse sudanese afrikaanse zimbabwaanse zambiaanse russische vietnamese amerikaanse britse venezuelaanse vaticaanse vanuatuaanse uruguayaanse ugandese tuvaluaanse turkmenistaanse tunesische tsjechische tsjaadse trinidadiaanse tongalese togolese thaise tanzaniaanse tadzjikistaanse syrische swazilandse spaanse somalische slowaakse sloveense singaporese leonese seychellese servische senegalese saudische samoaanse salomonse rwandese roemeense qatarese portugese poolse peruaanse paraguayaanse panamese palauaanse pakistaanse timorese oostenrijkse omaanseoezbekistaanse oekraiense noorse koreaanse nigeriaanse nigerese nicaraguaanse nepalese naurese warschau wenen kiev oslo podgorica chisinau valletta skopje vilnius vaduz rig zagreb pristina rome reykjavik dublin boedapest athene parijs helsinki tallinn berlijn kopenhagen sofia sarajevo brussel amerika mongolische monacese moldavische micronesiaanse mexicaanse mauritanische "new york" chicago harare lusaka tunis ndjamena lome dodoma mogadishu khartoem freetown kigali kampala abuja niamey nouakchott rabat lilongwe antananarivo tripoli monrovia maseru nairobi yaounde praia yamoussoukro bissau conakry accra banjul libreville asmara guinea cairo djibouti kinshasa brazzaville moroni bujumbura ouagadougou gaborone porto luanda algiers zwitserland zweden "zuid " "zuid korea" "zuid afrika" zimbabwe "wit rusland" vietnam "verenigde staten" "verenigde arabische emiraten" "verenigd koninkrijk" venezuela vaticaanstad vanuatu uruguay uganda tuvalu turkmenistan tunesie tsjechie tsjaad "trinidad en tobago" tonga togo thailand tanzania tadzjikistan syrie swaziland sudan sri lanka spanje somalie slowakije slovenie singapore "sierra leone" seychellen servie senegal "saudi arabie" "sao tome en principe" san marino samoa salomonseilanden "saint vincent en de grenadines" "saint lucia" "saint kitts en nevis" rwanda rusland roemenie qatar portugal polen peru paraguay "papoea nieuw guinea" panama palau "oost timor" oostenrijk oman oezbekistan oekraine noorwegen "noord korea" nigeria niger "nieuw zeeland" nicaragua nepal nauru namibie myanmar mozambique mongolie namibische myanmarese mozambiquese montenegrijnse mauritiaanse marshalleilandse maltese malinese maleisische maldiviaanse malawische madagaskarse macedonische luxemburgse litouwse liechtensteinse libische liberiaanse libanese letlandse lesothaanse laotiaanse kroatische koeweitse kiribatische

35 kirgizische keniaanse kazachstaanse kameroense kaapverdische jordanese jemenese japanse jamaicaanse italiaanse israelische iraanse irakese indonesie indiase ijslandse ierse hongaarse hondurese haitiaanse guyanese guinese guatemalaanse griekse grenadaanse ghanese georgische gambiaanse gabonese franse finse filipijnse fijische ethiopische estlandse eritrese salvadoraanse egyptische ecuadoriaanse duitse dominicaanse djiboutaanse deense cypriotische cubaanse congolese comorese colombiaanse chinese chileense canadese cambodjaanse burundese burkinabeze bulgaarse bruneise braziliaanse botswaanse bosnische boliviaanse bhutanese beninse belgische barbadese bengaalse bahreinse bahamaanse azerbeidzjaanse australische armeense argentijnse angolese andorrese algerijnse albanese afghaanse tirana korea hanoi dhabi asjchabad ankara bangkok taipei doesjanbe damascus arabie doha islamabad timor masqat tasjkent kathmandu naypyidaw ulaanbaatar "kuala lumpur" male beiroet vientiane isjkek astana amman sanaa tokio jeruzalem teheran bagdad jakarta delhi tbilisi manilla nicosia peking phnom thimphu dhaka manamah bakoe jerevan kabul bern stockholm londen praag madrid bratislava ljubljana belgrado moskou boekarest lissabon monaco moldavie micronesia mexico mauritius mauritanie marshalleilanden malta maleisie maldiven madagaskar macedonie luxemburg litouwen liechtenstein libie liberia libanon letland lesotho laos kroatie koeweit kiribati kirgizie kenia kazachstan kameroen kaapverdie jordanie jemen jamaica ivoorkust italie israel iran irak indonesische india

Appendix II

Table 4. Coding schema private context on sentence level Label Value Sex of politician Male/Female Politician Concerned politician from codebook Context Political/Private/Both private and political Characteristics Competence/ Credibility/ Rhetorical skills/ Appearance/ Morality.

Table 5. Coding schema personal life news on article level Label Value Other Yes/no Horse Race News Yes/no Issue News Yes/no Personal Life News Yes/no

36