Breaking with stereotypes: how the Dutch press cover female politicians
leading traditionally male ministries
Student: Kelsey Bouwmeester Student ID: 10592962
Master Thesis
Graduate School of Communication
Master's program Communication Science: Political Communication University of Amsterdam
Supervisor: Judith Möller Date: 29-06-2018
Summary
Underrepresentation of females in political positions is often in part ascribed to a structural gender bias in the media: female politicians are found to be portrayed in different, often less favorable, ways in the media than their male colleagues. Consequently, citizens are discouraged from voting for women and women are deterred from running for high political posts. This disadvantageous media attention is said to be the result of media logic; a theory that describes how journalists, in addition to institutional and technological factors, are influenced by personal values and deeply rooted stereotypes when determining what stories to pass through the gates and how to cover them. By entering the ‘masculine’ world of politics, female politicians break with stereotypes commonly held by journalists, and consequently receive less (favorable) media attention. Following this rationale, female politicians working in traditionally masculine
workfields, like defense or economic affairs, break with two stereotypes and possibly open themselves up to even more disadvantageous media coverage. A content analysis of 900 Dutch newspaper articles published from 1993-2018 of 30 ministers in 4 (gendered) workfields tests this line of reasoning, but finds no significant connection between amount and type of press coverage in terms of visibility, tone and substantiality of media coverage and the sex and type of workfield of Dutch ministers.
1
Introduction
As of June 2016, still only 22.8 percent of all national parliamentarians around the world are women (UN Women, 2017). Some authors ascribe this underrepresentation of females in political positions in part to a structural gender bias in the media (Lawless, 2009); female
politicians are said to be portrayed in different, often less favorable, ways in the media than their male colleagues, which discourages citizens form voting for women and deter women from running for high political posts (Wasburn & Wasburn, 2011, p.1028).
This study researches this phenomenon in a Dutch context and adds an important new angle by researching the possible role of the gendered fields politicians work in. By
concentrating on newspaper coverage of male and female politicians in different workfields, focusing on education-culture and science, and public health- well-being and sports as stereotypical female workfields, and defense, and economic affairs as stereotypical male workfields, this research aims to answer the question: To what extent is the amount and type of press coverage in Dutch newspapers of male and female political leaders connected to their (gendered) workfields?
Extant research into a gender bias in the media shows that women candidates receive less attention than their male colleagues (Kahn & Goldenberg, 1991 ; Kahn, 1994; Lühiste &
Banducci, 2016; Vos, 2013), receive more negative coverage than males (Kahn & Goldenberg, 1991; Kahn, 1994, ; Heldman, Carroll & Olson, 2005; Miller, Peake & Boulton, 2010) and receive coverage that is less often about substantive issues and more often personalized, mentioning their sex, family roles and appearance (Robertson, Conley, Szymczynska &
Thompson, 2002; Wasburn & Wasburn, 2011; Aday & Devitt, 2001). Lastly, ''to the extent that their issue positions are discussed, they will concern topics defined in American political culture
2 as 'women's issues' such as abortion, childcare, education and the environment, rather than 'men's issues' such as the economy, national security and military affairs'' (Wasburn & Wasburn, 2011, p.1028).
These differences in media coverage are often explained by the theory of media logic which describes how journalists make editorial decisions based on their own values and deeply rooted stereotypes (Lühiste & Banducci, 2016, p.226). One of these stereotypes that is
commonly held by journalists is that politics is a masculine job and that ''what is means to be a ''women'' does not correspond well with expectations about what it means to be ''president'''' (Heldman et al., 2005, p.316). So, when women do run for high political posts, they break with that stereotype and their perceived role in society, with disadvantageous media attention as a consequence (Bystrom, Robertson, Banwart, 2001).
Interestingly, over the past few years, more and more women have become the political leaders of political departments that are culturally viewed as 'masculine', for example by being appointed minister of economic affairs (UN Women, 03-15-2017). By leading these departments, these women break with not one, but two stereotypes, possibly opening themselves up to even more disadvantageous media attention. This research aims to find out whether females working in stereotypical male fields indeed receive less (favorable) attention by the media as a
consequence of the field they work in.
This question is important because people are not able to experience all politicians face-to-face; they are dependent on the media for political information they need to base their voting decisions on. So, if it turns out that women in general, or women in particular workfields more specifically, are structurally portrayed less (favorable) in the media than their male colleagues, this can ''discourage citizens from voting for women'' and ''dissuade women from running for
3 public office'' (Wasburn & Wasburn, 2011, p.1028). This is not only problematic from a gender equality point of view, but this also possibly keeps societies form having highly capable leaders, as research suggests that women outperform their male counterparts on many of the measures for leadership effectiveness (Zenger & Folkman, 2012).
While previous academic research focuses on part of the problem by investigating the differences in media coverage based on the sex of politicians, while controlling for factors such as incumbency, seniority and party-affiliation, apart from one study by Fernandez-Garcia (2016), no attention has previously been given to the (gendered) workfields these politicians operate in. This is problematic, since research suggests that, when women overstep such traditional
boundaries, the media tend to react by covering them differently (Fernandez-Garcia, 2016, p.145), indicating that gendered workfields might play a role in the relationship between politicians and media.
By analyzing Dutch newspaper articles published from 1993 till 2018, this study tries to fill this gap and researches the differences in media coverage based on the sex of the politicians as well as the gendered fields they work in. This way, the aim is to find out whether female politicians in general indeed receive less media coverage in terms of quantity and quality, and if this is due to their sex or to the fields they work in.
This article continues with an overview of academic research in the fields of gender, political leadership and media, describing previous findings on the differences in media coverage in more detail. Next, the method and data used for this study are discussed, as well as the results. Lastly, the discussion describes the main findings and place this article into its context.
4
Politicians, Gender and the Media
The concept of media logic, defined as “the institutional, technological and sociological characteristics of the news media, including their format characteristics, production and
dissemination routines, norms and needs’’ (Strömbäck, 2011, p.373 in Cushion, Thomas, Kilby, Morani & Sambrook, 2016, p.473), describes the way in which media content is editorially shaped and structured by journalists. Following the theory of media logic, journalists are not “passive conduits of political information”. Rather, they make editorial decisions based on not only institutional and technological factors, but also on their own values and deeply rooted stereotypes (Lühiste & Banducci, 2016, p.226). These values are used by journalists to judge the newsworthiness of a story, and help them select what stories to write about and how to write them (Major & Coleman, 2008, p.318). So even when journalists strive for objective reporting, their editorial decisions are always influenced by the culture in which they live and by the ideas and stereotypes that are part of that culture (Heldman, Carroll, Olson, 2005, p.316).
A stereotype that is commonly held by journalists, is that what it means to be a woman does not correspond well with expectations about what is means to be in politics (Heldman et al., 2005, p.316). When women do run for high political posts, they break with that stereotype and their perceived role in society, with criticism from the media as a consequence (Bystrom, Robertson, Banwart, 2001, p.2002).
Moreover, according to the theory of gender ownership of issues ''female candidates are culturally viewed as better at handling ''compassion issues'', such as health care and education, while, as supposed protectors and breadwinners, male candidates are viewed as better equipped to handle national security and the economy'' (Herrnson et al., 2003). Women leading political departments in stereotypically male fields thus break with two stereotypes, possibly leading to
5 even more negative attention from the media in terms of visibility, tone and substantiality, than women working in stereotypically more feminine fields and males working in either fields (Fernandez-Garcia, 2016, p.145).
Visibility
One of the most researched issues is whether male and female politicians receive the same amount of media coverage, often called visibility. Media visibility is paramount for politicians because this ensures recognition by voters, and recognition is an important term for vote choice (Kahn, 1994, p.154).
Following media logic, it is likely that female politicians are less visible in the media because their position does not correspond well with commonly held views by journalists, who generally believe males to be the most important actors in the political sphere. Consequently, stories starring female politicians have less chances of making it past the gates (Major & Coleman, 2008). When leading stereotypical male fields, women place themselves in a context that is traditionally even more occupied by male actors, possible leading females working in these fields to have even less chances of making it into the news.
Pioneer research into the possible differences in media visibility of male and female politicians indeed shows that female politicians in the US indeed receive less coverage than their male colleagues (Kahn & Goldenberg, 1991; Kahn, 1994). However, more recent research shows no gender bias in the amount of coverage candidates receive (Devitt, 2002; Hayes & Lawless, 2015; Lavery, 2013) or even an advantage for female politicians when it comes to media visibility (Wasburn & Wasburn 2011; Bode & Hennings, 2012; Miller, Peake & Boulton, 2010; Jalalzai, 2006; Bystrom, Robertson & Banwart, 2001; Fowler & Lawless, 2009).
6 These different findings might be due to one of two things, or a combination of the two. Firstly, the differences could possibly be explained by the time period that was investigated, which would mean that journalists gave female politicians less media attention in the past but now give females equitable amounts of media attention or even more media attention than males.
Secondly, it could be explained by the type of race that was researched. Research on media visibility of candidates from US house elections found no gender bias in either direction (Hayes & Lawless, 2015; Lavery, 2013), whereas visibility research found an advantage for female candidates in (vice)presidential elections (Wasburn & Wasburn, 2011; Bode & Hennings, 2012; Miller et al., 2010). Research into media-visibility of male and female candidates in senatorial races is mixed; Kahn (1994) and Kahn and Goldenberg (1991) found a clear pattern disadvantaging women, whereas Jalalzai (2006) and Bystrom and colleagues (2001) found a pattern disadvantaging men. So, when it comes to senatorial races, time and type of race most likely both interact in the relationship between sex and visibility.
Research on visibility of male and female politicians outside the US paint a more univocal picture. Although numbers on the amount of differences between media coverage of male and female politicians differ, most researchers agree that, in Europe, female politicians receive less media coverage than their male colleagues (Vos, 2013; Midtbo, 2011 & Ross, Evans, Harrison, Shears & Wadia, 2013).
Contrary to US research, O'Neill, Savigny and Cann (2016) even find a trend that women female politicians in the UK are actually becoming less visible over the years (p.299). Which would mean that time interacts in the relationship between politics and media in a different way in a European context compared to the US context.
7 media visibility. Firstly, Sametko and Boomgaarden (2007) show that the two candidates from the 2005 Bundestag election in Germany, Merkel and Schröder, were roughly equally visible in the media. However, more generally, ''without Merkel's presence in the campaign, the share of female political actors would have been considerable lower'' (p.158), indicating that personality might play a role. Secondly, Dan and Iorgoveanu (2013) find that ''women dominated the tabloid outlets'' and ''men were more featured prominently in the broadsheets'' (p.208), indicating a role for media-type. Lastly, only Fernandez-Garcia (2016) researches the possibility of a role for type of workfield that interacts in the relationship between sex and visibility, and finds that ''women occupying high-prestige portfolios (like finance) obtain even greater visibility than their actual presence in the cabinet. A pattern that is not shown in medium (like health care) and low-prestige (like education) policy areas'' (p.151).
Based on these previous findings this research will investigate the following hypotheses:
H1a Male political leaders working in stereotypical male fields, and in second place male political leaders working in stereotypical female fields will receive the most coverage H2a Female political leaders working stereotypical female fields will receive less coverage H3a Female political leaders working in stereotypical male fields will receive the least coverage
Tone
To be successful in politics, politicians not only need to garner media coverage, they need this coverage to be favorable in tone (Kahn, 1994). However, following media logic theory, dominant male culture influences journalists’ views and consequently creates media content that
8 portraying them as not capable of leading political departments (Setzler & Yanus, 2017, p.767), resulting in negatively toned media coverage for females that do work in politics. When leading stereotypical male workfields, female politicians overstep two traditional boundaries, possibly leading the media to react by covering them even more negatively (Fernandez-Garcia, 2016).
When the tone of the media coverage is indeed more negative for female politicians in general, or for female politicians in stereotypical male workfields more specifically, this can have negative consequences for the electoral chances of these politicians which are not based on their competence but solely because their sex and perhaps the field they work in do not
correspond with commonly held views about political leadership.
Studies from 1991 (Kahn & Goldenberg) and 1994 (Kahn) find a disadvantage for female candidates running for senate in terms of the tone, while Smith (1997) finds that females running in senatorial and gubernatorial races in 1994 received mostly neutral coverage and males
received more negative coverage. He emphasizes the exceptionality of this finding and states that ''this finding may be due to nothing more than peculiarities of the 1994 campaigns'' (p.78). Interestingly however, more recent research from Robertson et al. (2002) again finds that women in senatorial and gubernatorial races were treated significantly more favorable than their male competitors.
Research into the tone of media coverage of in US (vice)presidential races finds however quite consistently that females receive significantly more negative coverage than males
(Heldman et al., 2005; Miller et al., 2010; Conroy et al., 2015). Indicating that, as with the quantity of coverage, findings about a gender bias in tone of coverage in US media is mixed and might be explainable by the time period and/or type of race that is researched.
9 pattern. While no significant differences in the tone of coverage of male and female politicians is found in Germany (Sametko & Boomgaarden, 2007, p.162), Lühiste and Banducci (2016) find that, on average in the EU, female politicians are evaluated more positively than the male
politicians, and Midtbo (2011) finds that women in Norway are more likely to be evaluated more negatively than male politicians. Lastly, Fernandez-Garcia (2016) finds that in Spain, media coverage of politicians is mostly neutral. However, ''female ministers nominated to a high-prestige portfolio get a lower negative tone than female ministers who receive medium and low-prestige portfolios'' (p.155), indicating again that media coverage might be conditional on the type of post the women holds.
Based on these previous findings and expectations following media logic, this research will investigate the following hypotheses:
H1b Male political leaders working in stereotypical male fields, and in second place male political leaders working in stereotypical female fields will receive most positive coverage H2b Female political leaders working stereotypical female fields will receive less positive coverage
H3b Female political leaders working in stereotypical male fields will receive the least positive coverage
Substantiality of coverage
Besides the visibility in, and tone of media coverage, research focuses on the substantialityof the media coverage politicians receive, here divided into the amount of issue-coverage, amount of personalized coverage and amount of references to the sex of politicians.
10 Following media logic theory, journalists are likely influenced by living in a society in which people tend to appreciate women for their caring capacities and their looks and tend to believe women’s levels of political knowledge to be inferior to men’s, especially when it comes to issues that are less directly relevant to women’s lives (Dolan, 2011). This possibly influences the way journalists interpret stories about female politicians (working in stereotypical male workfields), leading to less issue-coverage and more personalized coverage for these politicians.
Issue-coverage
Issue-coverage, defined as stories that describe political issues or public controversies (Slothuus & De Vreese, 2010), is important for politicians because ''voters think about policy
considerations, in addition to the candidates' viability when evaluating candidates for office'' (Kahn, 1994, p.154). If the press focusses more extensively on political issues for candidates of a certain sex and/or that are working in a certain field, voters may come to believe that these candidates are more knowledgeable about issues than their competitors, meaning an advantage for these politicians trying to gain or stay in office.
Research into the amount of issue coverage of male and female candidates in (vice)presidential races in the US does indeed show that women receive less issue-attention than their male colleagues (Heldman et al., 2005; Aday & Devitt, 2001; Wasburn & Wasburn, 2011). Studies researching issue-coverage in the media for male and female candidates running in US house elections show however no significant difference (Hayes et al., 2015; Lavery, 2013; Smith, 1997).
Moreover, research in the context of senatorial and gubernatorial races find either significantly less issue-attention for female politicians (Kahn & Goldenberg, 1991; Kahn, 1994;
11 Devitt, 2002, Dunaway et al, 2013), or the exact opposite (Jalalzai, 2006; Fowler & Lawless, 2009). These different findings are hard to explain, since all authors researched the same type of race and media, within the same country and used comparable operationalizations for the
measurement of issue-coverage. One possible explanation would again be the time period under investigation, however the results do not follow a logical time path. Studies with data from the 1980's show more issue-coverage for males (Kahn & Goldenberg, 1991; Kahn, 1994), studies with data from the 1990's show more issue-coverage for females (Jalalzai, 2006; Fowler & Lawless, 2009) and the one study with data from the 2000s show again more issue-coverage for males (Dunaway et al., 2013).
To my knowledge, only one article researched this question in a European context, and the findings are mixed: although O'Neill and colleagues (2016) find that a great deal of the coverage female politicians received was ''surprisingly straightforward, focusing on reports about policy'' (p.301), however, they also found male MPs appeared in more straightforwardly political stories, whilst female politicians appeared in more stories related to wider society or events (p.300).
By testing the following hypotheses, the extent to which the amount of issue-coverage is linked to the sex and type of workfield Dutch politicians work in will be researched:
H1c Male political leaders working in stereotypical male fields, and in second place male political leaders working in stereotypical female fields will receive most issue-coverage
H2c Female political leaders working stereotypical female fields will receive less issue-coverage H3c Female political leaders working in stereotypical male fields will receive the least issue- coverage
12
Personalized coverage
Personalized coverage is defined as stories that elaborate on politicians’ personal characteristics, non-political traits and private lives (Trimble, Wagner, Sampert, Rapheal & Gerrits, 2013). By describing the physical appearance, lifestyle and family(role) of a politician, a women's 'other'- status in the male dominated world of politics in emphasized (Vos, 2013, p.391). This type of coverage disadvantages women, because it suggests they are not fit for the (masculine) political job.
The lion share of literature shows that females running in different types of US races indeed receive more personalized coverage than their male competitors (Wasburn & Wasburn, 2011; Bode & Hennings, 2012; Aday & Devitt, 2011; Heldman et al., 2005; Robertson et al. 2002; Banwart et al, 2003; Bystrom et al., 2001; Devitt, 2002; Fowler & Lawless, 2009; Conroy et al., 2015).
Only Lavery (2013) finds no statistically significant gender difference in personalized coverage in the US, and Major and Colemann (2008) do not find significantly more mentions of females’ candidates appearance. Possibly, the divergent finding of Lavery (2013) is due to the media-type under investigation. Unlike the other researchers who focus on print media, Lavery researches television data. Secondly, the unique finding of Major and Colemann might be
explained by the context of their study. Unlike other studies that use data from different races and different states, they only researched newspapers form Louisiana, possibly hurting the
generalizability of their findings.
Evidence from non-US research seem less extreme. O'Neill et al (2016) argue that personalized stories for all politicians in the UK were on the rise, however, they found that ''the gap between women and men with regards to personal focus is widening'' (p.300), Garcia-Blanco
13 and Wahl-Jorgensen find that, in France, Italy, Spain and the UK, references to female
politicians' appearance were not as prominent, however these types of references were still present in the press coverage (2012, p.437), and Dan and Iorgoveanu (2013) find that, in Romania, a limited amount of remarks about the candidates' appearance were present, however female candidates still received a larger share than their male colleagues (p.222). Lastly, in Spain, women are more likely to have their marital status, family and appearance emphasized in the coverage (Fernandez-Garcia, 2016, p.153). However, she again emphasizes the importance of the prestige of the ministries, stating that ''male and female minister appointed to high-prestige portfolios obtain fewer mentions to their personal life than male and female cabinet members pointed to medium and low prestige policy areas'' (p.153).
Again, these findings seem to support the line of reasoning based on the media logic theory. Also, according to the work of Fernandez-Garcia (2016), workfield might play a role in the relationship between sex and personalized coverage. The following hypotheses will therefore be tested in this study:
H1d Male political leaders working in stereotypical male fields, and in second place male political leaders working in stereotypical female fields will receive least personalized coverage H2d Female political leaders working stereotypical female fields will receive more personalized coverage
H3d Female political leaders working in stereotypical male fields will receive most personalized coverage
14
Sex
Emphasizing the sex of female politicians in news coverage has a similar consequence as does personalized coverage: by describing a women politician as a woman, ''she is defined by what she is not, that is, she is not a 'typical' politician'' (Ross, 2004 in Vos, 2013, p.391). Again,
emphasizing she is a woman in a men's world and is therefore not fit for the political job, thereby hurting females’ electoral chances.
Almost all of the US-centered research concludes that significantly more mentions of sex are made in press coverage of female politicians than of male politicians (Robertson et al, 2002; Major & Colemann, 2008; Banwart et al. 2003; Heldman et al. 2005; Bystrom et al., 2001).
European-centered research finds similar results; in Romania, journalists emphasize candidate sex for the female candidates (Dan & Iorgoveanu, 2013), in Spain, newspapers are more likely to mention the sex of women ministers than of male ministers (Fernandez-Garcia, 2016), and in Germany, sex was explicitly emphasized in news stories (Sametko &
Boomgaarden, 2007). However, ''this was never the case when only Schröder (the male
candidate) was mentioned in a news story” (p.165, additions in brackets not from original quote). Only two articles find contradictory evidence. Hayes and Lawless (2015) find no proof that sex is related to the number of references to a candidates’ sex in US politics (pp.100-101), and Wagner (2011) find that, in Canada, male candidates were even twice as likely than female candidates to see local journalists highlight their sex (p.44). However, Wagner emphasizes that ''journalists typically gendered men in their individual profiles, where it is common to share personal and professional details about a candidate, and less often in news stories. In contrast, they included gender remarks for women candidates in news stories as often as they did in profiles, suggesting a women's gender plays a subtle role in her evaluation as a prospective
15 municipal politician'' (pp.44-45).
Overall, the findings seem to support the idea that female politicians receive more mentions to their sex than their male counterparts. To research this in a Dutch context, the following hypotheses will be tested:
H1e Male political leaders working in stereotypical male fields, and in second place male political leaders working in stereotypical female fields will receive less mentions of their sex H2e Female political leaders working stereotypical female fields will receive more mentions of their sex
H3e Female political leaders working in stereotypical male fields will receive most mentions of their sex
Lastly, since several authors found differences in the way male and female politicians were portrayed as a consequence of media type (Dan & Iorgoveanu, 2013; Semetko & Boomgaarden, 2007), and since findings seem to differ over time, the last two hypotheses are tested:
H4: Tabloid newspapers will advantage female politicians in their coverage, and quality newspapers will advantage male politicians in their coverage.
H5: The differences in the amount and type of coverage between female and male political leaders will decrease over time.
16
Method
Data collection
To research to what extent the amount and type of press coverage of male and female political leaders is connected to their (gendered) workfield, based on the theory of gender ownership of issues, newspaper coverage of ministers1 in the stereotypical male fields of economic affairs and
defense and in the stereotypical female fields of education-culture and science and public health-well-being and sports is gathered over a time period of 25 years2.
Within these past 25 years, nine different women and 21 different men occupied the four minister posts under investigation, resulting in a sample of 30 ministers in total3. For a detailed
description of ministers and their workfields that are included in the sample, see appendix A. Newspaper articles about the 30 ministers were gathered in two ways. Firstly, an
automatic content analysis was carried out to determine the total amount of articles published of these 30 ministers in 12 Dutch national newspapers, using the new database LexisNexis. The sample of 12 newspapers used for the search includes both 'tabloid' newspapers, like the Algemeen Dagblad and the Telegraaf, and 'quality' newspapers, like the Volkskrant, NRC
1
Ministers are selected as the way to define political leadership because ministers are clearly linked to a specific workfield, and they are considered to be the leaders of these fields; they come up with policies and are responsible for all events concerning their departments.
2 This time period is firstly selected because the engine used to search for relevant newspaper articles, LexisNexis,
only includes articles from 1980 – present. Secondly however, from 1980-1994, only one female occupied any of the four minister posts that are part of this study. Therefore, a more limited selection of 25 years is made.
3 For the final selection, two ministers are (partly) excluded from the sample. Firstly, minister Zalm has been
minister of economic affairs, however, he was only in office for 5 days. It seems unlikely that media coverage of this minister in such a short period of time is a good reflection of media coverage of minister in their (gendered)
workfields in general, and is therefore excluded from the sample. Secondly, minister Hoogervorst has also been a minister of economic affairs, however, during the same time, he was also the minister of finance. A random sample of the articles about minister Hoogervorst published during his time in office as minister of economic affairs and minister of finance show that, by far, most articles are about him as a minister of finance, and not about him and his work for the economic affairs department. Since this research focuses on the amount and type of press coverage of ministers in specific workfields, of which finance is excluded, minister Hoogervorst is not included in the sample for his work as minister of economic affairs
17 Handelsblad and Trouw (Boukes and Vliegenthart, 2017)4.
To carry out the automatic content analysis, a search was done starting with the word minister and ending with the last name of each of the 30 ministers under investigation5. For
every minister, the search period was aligned with the exact period the politician was in office as a minister of one of the four selected minister posts. The automatic content analysis brought about a total of 87079 articles.
Additionally, a manual content analysis was done to determine the visibility in coverage in more detail, and to determine the tone and the substantiality of the newspaper coverage. Using random selection, 30 articles out of the total amount of articles found by LexisNexis of each minister was selected to manually code using an elaborate codebook, resulting in a total sample of 900 articles6. Of each article, the title, newspapers it was published in, publication date,
minister and ministry he or she was working for was listed7. For a description of the measures
used to determine the amount and type of newspaper coverage of Dutch ministers, as well as descriptive information of these measures, see appendix B. The codebook with the exact questions and scales is included as appendix C.
4
For the purposes of this research, news coverage is researched in terms of newspaper articles, because, according to a recent report by the SCP, ''classic media is still consulted by a larger part of the Dutch public for (political) news than online media is'' (translated quote from report 'Nederlanders en nieuws', SCP, 2017, p.31).
5 For female ministers that carry both their maiden name and their married name, only their married name was used
for the search since this produced the most results, indicating that this way the least number of articles are missed.
6 Two ministers in the sample occupied two of the four minister posts that are of importance for this research. In
those two cases, the 30 coded articles of both ministers are divided between the two posts.
7 Because of the importance of analyzing the coverage of the ministers in a specific workfield, and because many of
the ministers in the sample have had minister posts in different workfields, all articles in the initial sample for a minister that indeed has had different minister posts, and from which the article did not specify for which ministry the minister was working, were later excluded from the sample, and replaced with articles that did specifically state the ministry the minister was working for at that time
18
Intercoder-reliability
To test the reliability of the data gathered by the manual content analysis, a second coder was trained to code a subsample of the articles. The second coder is a Master student of Political Communication studies at the University of Amsterdam and is a native Dutch speaker.After elaborate coder training and reaching agreement on the codebook, the second coder coded a subset of 77 articles (8,5%) of the sample. More detailed information about the procedure of coder training can be found in appendix D.
As shown in appendix B, results of a Krippendorff's Alpha test of the subsample of 77 articles shows scores between .38 and .1, with most of the measures falling well above the threshold generally accepted in social sciences8 (Lombard, Snyder-Duch & Bracken, 2002,
p.593).
Results Visibility
Table 1 shows the results for the visibility of male and female ministers in newspaper coverage. As shown, while male ministers are more visible in terms of the number of articles and the placement of the articles, female ministers are more visible in terms of frequency of references in a single article and the length of an average article. Additionally to showing no clear pattern, independent t-tests show that the differences between male and female ministers on all visibility
8 The variable measuring the amount of issue-coverage shows to be problematic, but is needed for an adequate
answer to the research question and is therefore not excluded from this study. However, any conclusions drawn from this measurement should be made with caution. Also, the Krippendorff's Alpha score for 'tone' is slightly less than eligible, however, given the widely known difficulty of measuring the tone of a text in social research, I find this score reasonable and will use this variable for analyses. Lastly, as shown, it was not possible to calculate the Krippendorff's Alpha for the variables physical appearance and family(ties), since there was too little variance in the measurements. However, the percentage agreement for both variables was 100 percent, as a result, the variables physical appearance and family(ties) are included in the analyses and deemed trustworthy.
19 measures are not significant.
Table 1
Newspaper visibility of male and female ministers
Visibility
Measures Male ministers Female ministers Independent T-test
Number of articles M=2.64, SD=1.46 M=1.94, SD=1.25 T(30)=1.30, p=0.203 Frequency of references M=1.27, SD=1.01 M=1.3, SD=1.07 T(989)=-0.42, p=0.678 Wordcount M=436.76, SD=327.30 M=478.03, SD=397.24 T(897)=-1.62, p=0.105 Page number M=7.78, SD=10.33 M=8.52, SD=7.62 T(769)=-0.98, p=0.326
Table 3, describing the differences in visibility for Dutch ministers based on both their sex and their gendered workfield shows again no clear pattern. Although male ministers working in stereotypical female fields are most visible in terms of number of articles, frequency of
references in a single article and page the articles are placed on, they are least visible in terms of the length of the articles. Also, whilst male ministers working in stereotypical female fields score lowest on the wordcount measure, the other three groups score lowest on the other three
measures.
Moreover, the differences between the four groups on the different measures are small. Firstly, all ministers, are on average mentioned two or three times in articles that describe them. Secondly, newspaper articles about both male and female ministers in either workfields are between 400 and 500 words in length and thirdly, articles about all ministers, except males working in stereotypical female fields, are placed on average on page 8 of a newspaper. The only
20 considerable difference was found between female ministers working in stereotypical male fields and male ministers working in stereotypical female fields in terms of the number of articles they are mentioned in: the former received almost half of the media attention than the latter. One-Way ANOVA tests showed that the differences between the four groups on all four measures of
visibility are however not significant. Hypotheses H1a, H2a and H3a are therefore not supported.
Tone
Results show that female ministers receive slightly less positive coverage (M=-0.29, SD=0.73) than their male colleagues (M=-0.27, SD=0.69). The difference is however very small and not significant t(898) = 0.28, p=0.780. Furthermore, both male and female ministers have received both very positive (+2) and very negative (-2) press coverage.
Table 3 shows that the average tone of all groups of ministers is between 0 and -1, meaning that the coverage of all ministers, despite their sex and workfield, is between neutral and slightly negative. Moreover, A One-Way ANOVA shows that the differences between the four groups is not significant. Hypotheses H1b, H2b and H3b are therefore not supported.
Substantiality of coverage
Issue-coverage
In 72 percent of the articles mentioning female ministers, the positions of the minister regarding substantial policies are at the center of the article (M=0.59, SD=0.71), for their male colleagues this is only the case in 65 percent of the articles (M=0.45, SD=0.80). An independent T-test shows that this difference is significant t(897)=-2.40, p=0.017, meaning that female ministers receive significantly more issue-attention in their newspaper coverage than do males.
21 Table 3 shows moreover that females in stereotypically male fields receive most issue-coverage, followed by females in stereotypically female fields and males working in
stereotypically female fields. Male ministers working in stereotypically male fields surprisingly receive least issue-coverage. The differences between these four groups are significant. PostHoc analyses using Turkey's HSD indicated that females working in stereotypically male workfields received significantly more issue-coverage than did male ministers working in stereotypically male workfields (p=0.038), but amount of issue-coverage did not differ significantly between the other groups.
The results do not support the hypotheses. In fact, the findings for issue-coverage support almost the exact opposite. Therefore, H1c, H2c and H3c are not accepted
Personalized coverage
In only four of the 900 articles researched, a mention is made to the family of a minister, and in only eight of the 900 articles, a reference was made to the appearance of a minister.
Unexpectedly, in nine out of these twelve articles, the minister of subject was a male. However, table 2 shows that both the differences between males and females in terms of references to family and appearance are not significant.
Three of the four times the family of a politician was mentioned, this was the case for a male minister working in a stereotypically female workfield, the other time it was for a male minister working in a stereotypically male workfield. Table 3 shows however that the differences between the four groups in terms of references to family are not significant.
Also, of these eight articles that mentioned the appearance of the ministers, two were made in articles describing a female minister in a stereotypical male field, two in articles
22 describing a male minister in a stereotypical female field, and four in articles describing a male minister in a stereotypical male field. Not one article described the way a female minister working in a stereotypical male field looked. A One-Way ANOVA showed that the differences between the four groups in terms of number of articles mentioning appearance was again not significant.
Taken together, these findings show that personalized coverage is actually scarce in Dutch newspaper coverage of ministers. When mentions to the family or appearance of ministers are made, this is slightly more often the case for male ministers than for female ministers. H1d, H2d and H3d are therefore not accepted.
Table 2
Amount of personalized coverage for male and female politicians
Personalized coverage
Measures Males Females Independent T-test
References to family M=-0.99, SD=0.08 M=-0.1, SD=0.00 t(898) = 1.31, p=0.190 References to appearance M=-0.99, SD=0.12 M=-0.99, SD=0.17 t(898) = -0.37, p=0.710 Mentions of sex
In 13.7% of coverage of female ministers, an explicit reference is made to their sex (N=37). For male ministers this is the case in 11.8% of their coverage (N=74). The differences between the male (M=-0.87, SD=0.38) and female ministers (M=-0.38, SD=0.46) is however not significant
23 t(898) = -1.25, p=0.212.
As shown in table 3, female ministers working in stereotypically male fields receive most coverage emphasizing their sex, next come male ministers working in stereotypically female fields, indicating that working in a field that does not stereotypically fit the sex of oneself produces most coverage explicitly referencing sex. Moreover, this effect is stronger for women than it is for men.
However, the differences with the other two groups are again small and not significant. Meaning that, although the results are mostly in the expected direction, H1e, H2e and H3e cannot be supported.
Type of newspaper & year of publication
Lastly, linear regression analyses were carried out to test for a potential interaction effect of type of newspaper (tabloid versus quality) and year of publication on the relationship between both sex and media coverage and sex, workfield and media coverage of Dutch ministers. No
significant results were found, meaning that neither the type of newspaper, nor the year the news article was published in, moderates the relationship between the sex of the minister and the field he or she is working in and all of the measures for amount and type of newspaper coverage. Hypotheses H4 and H5 are therefore not supported. For the exact test results see appendix E.
24 Table 3
Visibility, tone and substantiality of coverage of male and female politicians in stereotypical male and female workfields
Male ministers in male fields Male ministers in female fields Female ministers in male fields Female ministers in female fields One-Way ANOVA Visibility Number of articles M=2.41, SD=1.45 M=2.98, SD=1.48 M=1.55, SD=1.03 M=2.21, SD=1.40 [F(3, 28)=01.02, p=0.401] Frequency of references M=1.21, SD=1.08 M=1.35, SD=1.05 M=1.29, SD=1.0 M=1.31, SD=1.02 [F(3, 896)=0.98, p=0.401] Wordcount M=457.89, SD=311.02 M=408.58, SD=346.41 M=479.62, SD=381.97 M=477.02, SD=407.87 [F(3, 895)=-0.98, p=0.326] Page number M=8.59, SD=11.23 M=6.68, SD=8.87 M=8.25, SD=6.45 M=8.71, SD=8.36 [F(3, 767)=2.12, p=0.970] Tone Tone M=-0.26, SD=0.66 M=-0.28, SD=0.71 M=-0.20, SD=0.66 M=-0.35, SD=0.76 [F(3, 896)=0.99, p=0.398] Substantiality Issue-coverage M=0.39, SD=0.81 M=0.54, SD=0.78 M=0.62, SD=0.70 M=0.57, SD=0.72 [F(3, 895)=4.07, p=0.007] Appearance M=-0.99, SD=0.14 M=-0.99, SD=0.09 M=-0.96, SD=0.28 M=-0.1, SD=0.00 [F(3, 896)=1.82, p=0.142] Family(ties/roles) M=-0.1, SD=0.00 M=-0.99, SD=0.11 M=-0.1, SD=0.00 M=-0.1, SD=0.00 [F(3, 896)=1.38, p=0.247] Sex M=-0.89, SD=0.34 M=-0.84, SD=0.43 M=-0.80, SD=0.49 M=-0.85, SD=0.44 [F(3, 896)=1.46, p=0.223]
25
Discussion
Unlike most European-centered research (Vos, 2013; Midtbo, 2011; Ross et al., 2013), analysis of visibility of male and female ministers in Dutch newspapers show no significant advantage for ministers based on their sex. Also, unlike Fernandez-Garcia (2016), this study found no
significant effect for type of workfield on the relationship between sex and visibility, meaning that the amount of press coverage in Dutch newspapers for male and female political leaders is thus not connected to the gendered fields they work in.
Analyses of the tone and amount of personal coverage show moreover that coverage of Dutch political leaders is very balanced: the tone of coverage for all ministers is between neutral and slightly negative and the differences between the groups in terms of tone were very small and insignificant. Moreover, personalized coverage turned out to be almost absent from the coverage. Although this finding is striking, it is mostly in line with studies researching personal coverage in other European countries that find a limited amount of references to the appearance and family(life) of ministers (Garcia-Blanco & Wahl-Jorgensen, 2012; Dan & Iorgoveanu, 2013).
The only significant results in this study were found for the amount of issue-coverage. Contrary to my expectations, and contrary to findings from several studies (Heldman et al., 2005; Aday & Devitt, 2001; Wasburn & Wasburn 2011) female ministers received significantly more issue-attention than their male colleagues in general, and females working in stereotypically male workfields received significantly more issue-attention than did male ministers working in these fields. Conclusions based on these findings should however be made with caution, since the intercoder reliability of the measurement used to determine this variable was low
(Krippendorff's Alpha=.38).
26 expected; female ministers working in stereotypically male fields and male ministers working in stereotypically female fields receive most coverage emphasizing their sex. This indicates that working in a field that does not stereotypically fit the sex of the minister him- or herself produces most coverage explicitly mentioning sex. Although these findings are for the most part in the expected direction, they are not significant and did therefore not support the hypotheses.
Taken together, these results do not support the idea that the amount and type of press coverage in Dutch newspapers of male and female political leaders is connected to their
(gendered) workfields. Moreover, the results do not even support the well-researched theory that the amount and type of press coverage significantly differs between ministers based on their sex, thereby mostly disadvantaging female politicians (Kahn & Goldenberg, 1991; Kahn, 1994).
The most likely explanation for these divergent findings is the case selection; this study is based on data from the Netherlands of the past 25 years. Currently, the Netherlands ranks as 32nd country of the world when it comes to gender equality and has been at the top for many years prior (World Economic Forum, 2017). It could very well be that, due to being brought up in a comparatively gender equal society, Dutch journalist uphold high gender equality norms and are not that affected by gender-based stereotypes that are deeply rooted in less gender equal
societies, or societies that have not had gender-equality for a long time. As a result, it is likely that Dutch journalists do not see politics solely as a men’s job and therefore do not systematically differentiate in the amount and type of coverage they give male and female ministers, not even when these ministers overstep a second traditional boundary and start working in fields that stereotypically do not match their sex. Moreover, finding no significant interaction effects for the sex and type of workfields of ministers and both type of newspapers and year of publication strengthens the thought that Dutch journalists across the board are not strongly affected by
27 stereotypes that depict politics as a masculine world in which females have no place, and have not been affected by this over the past 25 years.
The theory of media logic, that describes how, in addition to technological and
institutional factors, journalists’ views influence the media coverage they create, still applies in this context, but likely works in the opposite direction. Beliefs about political competence of still color the way journalists judge the newsworthiness of stories, but result in an equal amount and type of news coverage for all politicians, despite their sex and field they work in, because Dutch journalists believe that these politicians are equitably capable.
This study contributes to the limited research into a possible gender bias in media coverage in Western-European countries and adds an important new angle by researching the possible role of the gendered fields political leaders work in. That being said, as with every research, this study suffers from some limitations. Firstly, this study only researched a sample of 900 articles, which comes down to about 1% of all newspaper articles describing the ministers of interest over the past 25 years. Although the selection has been completely random, a larger sample would have given more certainty about the generalizability of the findings to the rest of the coverage. Secondly, this research has focused solely on newspaper articles, and thus excludes political information about ministers coming from television and new media. Although
newspapers are an important source for political information for a large part of the Dutch public, television and new media do play a significant role in the interaction between politicians,
journalists and the public and influences the public in how they view politicians.
Future studies should focus on these issues, and research a possible gender bias in both traditional and new media from the same context to see whether the findings of this study are confined to newspaper coverage or can be found in media coverage in general. Secondly, it is
28 important to continue researching the role of workfields in media coverage of male and female politicians outside the context of the Netherlands, in an effort to determine the generalizability of these results to other countries. Possibly, journalists in countries with less gender-equal cultures are more affected by traditional stereotypes and do cover female politicians in stereotypical male workfields less (favorable). The best way to study this would be by comparing media coverage of male and female politicians in different workfields in both comparatively gender-equal and gender-unequal societies.
Additionally, studies researching the role of journalists' gender norms and the media coverage they create is of great importance for a deeper understanding of the relationship between media and politics and can possibly be an important part of trying to understand the differences in media coverage of male and female politicians in different workfields. Studies combining survey research, determining the stereotypes held by journalists, and content analysis of the news coverage created by these journalists will be meaningful and will help understand the weight of sociocultural factors as compared to institutional and technological ones in the process of creating media content.
Until then, this research has shown that we should be careful with making big claims about a worldwide structural gender bias in the media and understand that, at least in the context of the Netherlands, there seems to be no structural gender bias that dampens the chances of women for high political functions. From the point of view of gender equality, the findings of this study are reassuring, since they demonstrate no structural gender bias in Dutch media coverage that might lead to discourage citizens from voting for women and deter women for running for high political posts. Also, according to these findings, press coverage should be no impediment for females to lead departments that are traditionally more assigned to males.
29
Bibliography
Aday, S. & Devitt, J. (2001). Style over substance: Newspaper coverage of Elizabeth dole’s presidential bid. The Harvard International Journal of Press/Politics, 6(2), 52-73.
Banwart, M.C., Bystrom, D.G. & Robertson, T.A. (2003). From the primary to the general: A comparative analysis of candidate media coverage in mixed-gender 2000 races for governor and U.S. senate. American Behavioral Scientist, 46(5), 658-676.
Bode, L. & Hennings, V.M. (2012). Mixed signals? Gender and the media’s coverage of the 2008 vice presidential candidates. Politics & Policy, 40(2), 221-257.
Boukes, M. & Vliegenthart, R. (2017). A general pattern in the construction of economic
newsworthiness? Analyzing news factors in popular, quality, regional, and financial newspapers.
Journalism, 00(0), 1-22.
Bystrom, D.G., Robertson, T.A. & Banwart, M.C. (2001). Framing the fight: An analysis of media coverage of female and male candidates in primary races for governor and U.S. senate in 2000. American Behavioral Scientist, 44(12), 1999-2013.
Conroy, M., Oliver, S., Breckenbridge-Jackson, I. & Heldman, C. (2015). From ferraro to palin: sexism in coverage of vice presidential candidates in old and new media. Politics, Groups, and
Identities, 3(4), 573-591.
Cushion, S., Thomas, R., Kilby, A., Morani, M. & Sambrook, R. (2016). Interpreting the Media Logic behind Editorial Decisions: Television News Coverage of the 2015 U.K. General Election Campaign. The International Journal of Press/Politics, 21(4), 472-489
Dan, V. & Iorgoveanu, A. (2013). Still on the beaten path how gender impacted the coverage of male and female Romanian candidates for European office. The Harvard International Journal
30
of Press/Politics, 18(2), 208-233.
Devitt, J. (2002). Framing gender on the campaign trail: Female gubernatorial candidates and the press. Journalism & Mass Communication Quarterly, 79(2), 445-463.
Dolan, K. (2011). Do Women and Men know different things? Measuring Gender Differences in Political Knowledge. The Journal of Politics, 73(1), 97-107.
Dunaway, J., Lawrence, R.G., Rose, M. & Weber, C.R. (2013). Traits versus issues: How female candidates shape coverage of senate and gubernatorial races. Political Research Quarterly, 66(3), 715-726.
Fernandez-Garcia, N. (2016). Framing gender and women politicians representation: Print media coverage of Spanish women ministers. CECS-Publicacoes/eBooks, 141-160.
Fowler, L. L. & Lawless, J.L. (2009) Looking for sex in all the wrong places: Press coverage and the electoral fortunes of gubernatorial candidates. Perspectives on Politics, 7(3), 519-536.
Garcia-Blanco, I. & Wahl-Jorgensen, K. (2012). The discursive Construction of Women Politicians in the European Press. Feminist Media Studies, 12(3), 422-441.
Hayes, D. & Lawless, J.L. (2015), A non-gendered lens? Media, voters and female candidates in contemporary congressional elections. Perspectives on Politics, 13(1), 95-118.
Heldman, C., Carroll, S.J. & Olson, S. (2005). ‘She brought only a skirt’; Print media coverage of Elizabeth dole’s bid for the republican presidential nomination. Political Communication, 22, 315-335.
Herrnson, P. S., Lay, J. C. & Stokes, A. K. (2003). Women running ''as women'': candidate gender, campaign issues and voter-targeting strategies. The Journal of Politics, 65(1), 244-255 Jalalzai, F. (2006). Women Candidates and the Media: 1992-2000 Elections. Politics & Policy,
31
34(3), 606-633.
Kahn, K.F. (1994). The Distorted Mirror: Press Coverage of Women Candidates for Statewide Office. The Journal of Politics, 56(1), 154-173
Kahn, K.F. & Goldenberg E.N. (1991). Women Candidates in the News: An Examination of Gender Differences in U.S. Senate Campaign Coverage. The Public Opinion Quarterly, 55(2), 180-199
Lavery, L. (2013). Gender bias in the media? An examination of local television news coverage of male and female house candidates. Politics and Policy, 41, 877-910.
Lawless, J. L. (2009). Sexism and Gender Bias in Election 2008: A More Complex Path for Women in Politics. Politics & Gender 5(1), 70-80.
Lombard, M., Snyder-Duch, J & Bracken, C. C. (2002). Content Analysis in Mass
Communication: Assessment and Reporting of Intercoder Reliability. Human Communication
Research, 28(4), 587-604
Lühiste, M. & Banducci, S. (2016). Invisible women? Comparing candidates’ news coverage in
Europe. Politics & Gender, 12(2), 223-253.
Major, L. & Coleman, R. (2008). The intersection of race and gender in election coverage: what happens when the candidates don’t fit the stereotypes? Howard Journal of Communications, 19, 315-333.
Midtbo, T. (2011). Explaining media attention for Norwegian MPs: A new modelling approach.
Scandinavian Political Studies, 34(3), 226-249.
Miller, M.K., Peake, J.S. & Boulton, B.A. (2010). Testing the Saturday night live hypothesis: Fairness and bias in newspaper coverage of Hillary Clinton’s presidential campaign. Politics &
32 O'Neill, D., Savigny, H. & Cann, V. (2016) Women politicians in the UK press: not seen and not heard? Feminist Media Studies, 16(2), 293-307, DOI: 10.1080/14680777.2015.1092458
Robertson, T., Conley, A., Terry Robertson, Allison Conley, Szymczynska, K. & Thompson, A. (2002) Gender and the media: An investigation of gender, media, and politics in the 2000 election, Atlantic Journal of Communication, 10:1, 104-117
Ross, K., Evans, E., Harrison, L., Shears, M. & Wadia, K. (2013). The gender of news and news of gender. A study of sex, politics, and press coverage of the 2010 British general election. The
International Journal of Press/Politics, 18(1), 3-20.
Semetko, H. A. & Boomgaarden, H.G. (2007). Reporting Germany’s 2005 bundestag election campaign: Was gender an issue? The Harvard International Journal of Press/Politics, 12(4), 154-171.
Setzler, M. & Yanus, A. B. (2017). Evangelical Protestantism and Bias Against Female Political Leaders. Social Science Quarterly, 98(2), 766-778.
Slothuus, R., & De Vreese, C., H. (2010). Political Parties, Motivated Reasoning, and Issue Framing Effects. The Journal of Politics, 72(3), 630-645.
Smith, K. B. (1997) When all’s fair: Signs of parity in media coverage of female candidates.
Political Communication, 14(1), 71-82.
Sociaal en Cultureel Planbureau (2017). 'Nederlanders en nieuws'. Accessed at 13-05-2018 from file:///C:/Users/Gebruiker/Downloads/Nieuwsgaring%20-%20def.pdf
Trimble, L. (2007). Gender, political leadership and media visibility: Globe and mail coverage of conservative party of Canada’s leadership contests. Canadian Journal of Political Science, 40, 969-993.
33 Mediation in Newspaper Coverage of Canadian Party Leadership Contests, 1975-2012. The
International Journal of Press/Politics, 18(4), 462-481.
UN Women (July 2017), Facts and Figures: Leadership and Political Participation. Accessed at 06-24-2018 from
http://www.unwomen.org/en/what-we-do/leadership-and-political-participation/facts-and-figures
UN Women (03-15-2017). Press Release: NEW IPU and UN Women map shows women’s
representation in politics stagnating. Accessed at 06-24-2018 from
http://www.unwomen.org/en/news/stories/2017/3/press-release-new-ipu-and-un-women-map-shows-womens-representation-in-politics-stagnating
Vos, D. (2013). The vertical glass ceiling: Explaining female politicians’ underrepresentation in television news, Communications 38(4), 389–410
Wagner, A. (2011). On their best behavior? Newspaper journalists’ coverage of women municipal candidates in Alberta. Canadian Political Science Review, 5(1), 38-54.
Wasburn, P. C. & Wasburn, M. H. Media Coverage of Women in Politics: The curious case of Sarah Palin (2011). Media, Culture & Society, 33(7), 1027-1041
World Economic Forum (2017). The Global Gender Gap Report 2017. Accessed at 25-06-2018 from http://www3.weforum.org/docs/WEF_GGGR_2017.pdf
34
Appendix A – Distribution of male and female ministers in stereotypical male and female workfields
Table 4
Distribution of male and female ministers in stereotypical male and female workfields Female ministers in stereotypical female workfields Female ministers in stereotypical male workfields Male ministers in stereotypical female workfields Male ministers in stereotypical male workfields Minister van Engelshoven
Minister Bijleveld Minister Plasterk Minister Kamp
Minister Bussemaker Minister Hennis Minister Rouvoet Minister Hillen Minister Bijsterveldt Minister van Der
Hoeven
Minister Hermans Minister van Middelkoop Minister van der
Hoeven
Minister Jorritsma Minister Ritzen Minister Korthals
Minister Borst Minister de Jonge Minister de Grave
Minister Schippers Minister Klink Minister Voorhoeve
Minister Hoogervorst Minister Wijers Minister Bomhoff Minister Wiebes Minister de Geus Minister Verhagen
Minister Wijn Minister Brinkhorst Minister Heinsbroek
35
Appendix B – Table of measures and ways of measurement
Table 5
List of measures and their descriptive information
Measure Question Descriptive information
Minister Which minister is mentioned
in the article?
N=900, Krippendorff's Alpha =.1
Ministry What ministry is the minister
working for? N=900, Krippendorff's Alpha =.87 Visibility in coverage Frequency of references to minister
How often is the minister mentioned in an individual article?
N=900, M=1.38, SD=1.05, Krippendorff's Alpha =.90 Length of article What is the word count of the
article?
N=899, M=449.11,
SD=349.98, Krippendorff's Alpha =.99
Placement of article On what page in the newspaper is the article published?
N=771, M=8.00, SD=9.60, Krippendorff's Alpha =.1
Tone of coverage
Tone towards minister What is the tone of the article towards the minister?
N=900, M=-0.28, SD=0.70, Krippendorff's Alpha =.63
Substantiality of coverage
Issue-coverage Does the article mention any
policy plans of the minister, or is the position or opinion of the minister about a substantial policy issue mentioned?
N=899, M=0.49, SD=0.77, Krippendorff's Alpha =.38
Appearance Does the article refer to the
appearance of the minister in any way?
N=900, M=-0.99, SD=0.14, Krippendorff's Alpha = - Family(ties/roles) Does the article refer to the
family, family ties or family role of the minister in any way?
N=900, M=-1.00, SD=0.07, Krippendorff's Alpha =-
References to sex of minister Does the article explicitly refer to the sex of the minister?
N=900, M=-0.86, SD=0.41, Krippendorff's Alpha =.1 Type of newspaper What newspaper is the article
published in?
N=900, Krippendorff's Alpha =.88
Year of publication What date is the article published?
N=900, M=2007.27, SD=91.85, Krippendorff's Alpha =.1
36 The independent variables, minister and ministry, are measured by listing what minister is
mentioned in a single article, and the ministry that minister was working for according to the article. The independent variables: amount and type of coverage, are measured by the visibility of ministers in the newspaper articles, the tone of the coverage towards the ministers in these articles, and the substantiality of the coverage describing these ministers.
Visibility can be measured in multiple ways. Most often is it done by simply counting in how many of the articles or tv-items the politician is mentioned. However, some authors argue that the prominence of the candidate in the coverage should also be researched, since not always the entire article or tv-item is being read or watched by the public (Helman, Carroll & Olson, 2005, p.321). Prominence can be established by counting how often the politician is mentioned in one article or tv-item, determining whether the candidate was the only or central focus of the article or tv-item, whether the candidate was mentioned in the headline, on what page the article was published and/or of how many words or paragraphs the article consisted (Heldman et al., 2005; Miller, Peake & Boulton, 2010). This study therefore uses a multi-level measurement of visibility and codes how often a minister is mentioned in a single article, the length of these articles and listing on what page these newspaper articles are published on in addition to the automatic content analysis that found a total amount of articles mentioning a minister and divided this by the total of days the minister occupied the minister post. Secondly, following the Center for Politics and Communication, tone of media coverage is described in terms of
negativity, positivity, and neutrality towards the minister in a single article. Lastly, the substantiality of the newspaper coverage is measured by determining the amount of issue-coverage, the amount of personalized coverage and the amount of references made to the sex of the minister. More specifically, following Aday and Devitt (2001, p.58), issue-coverage is
37 measured by determining whether and if so, how often, an article mentions a politicians' position or record on public policy issues, personalized coverage is measured by counting the amount of times the reporter mentions the marital status, family members or children of the politician as well as any references to being childless, the family role a politician and physical appearance of the politician (Dan & Iorgoveanu, 2013, p.212) and lastly, differences in mentions of sex of male and female politicians are measured by counting the amount of explicit references of the sex of a minister in a given article.
Also, this study includes the variables of type of newspaper and year of publication and classifies the Nederlands Dagblad, Reformatorisch Dagblad, Algemeen Dagblad, De Telegraaf, Spits and Metro NL as tabloid newspapers and De Volkskrant, De Trouw, NRC Handelsblad, NRC Next, Het Financiele Dagblad and Het Parool as ‘quality’ newspapers.
38
Appendix C – Codebook
Codeboek
Eenheid van analyse:
Een volledig en individueel artikel uit een landelijke, Nederlandse krant waarin een van de ministers uit de onderstaande lijst genoemd wordt. Een artikel is niet relevant en wordt dus niet gecodeerd als het artikel gaat over een ex- of aanstaande minister, ook al staat de naam van deze minister op de lijst en is de persoon het onderwerp van het artikel, of over een politicus die op de lijst staat maar in het artikel wordt beschreven in een andere functie. Voorbeelden hiervan zijn: 'oud-minister', 'voormalig minister', 'lijsttrekker', VVD-politicus', 'aankomend minister voor volksgezondheid'.
39
Algemene informatie:
1. Noteer de titel van het artikel Open vraag: ….
2. Noteer de krant waarin het artikel is gepubliceerd
Als het artikel niet in een van de onderstaande kranten is gepubliceerd, of als het niet duidelijk is in welke krant het artikel is gepubliceerd, het artikel niet meenemen in het coderen. Let op: onderscheid tussen NRC Handelsblad, en NRC.NEXT.
Antwoordcategorieën:
1. De Volkskrant 2. Trouw
3. NRC Handelsblad 4. NRC.NEXT
5. Het Financiële dagblad 6. Nederlands Dagblad 7. Reformatorisch Dagblad 8. Algemeen Dagblad (AD) 9. Het Parool
10. De Telegraaf 11. Spits
12. Metro (NL)
40 3. Noteer de datum waarop het artikel is gepubliceerd
Open vraag: ddmmjjjj
4. Welke minister(s) wordt/worden er genoemd in het artikel?
Als er meer dan een minister van de lijst genoemd wordt in het artikel, dan het artikel
verschillende keren coderen, een keer voor elke minister op de lijst. Als je minister enkel in het bijschrijft genoemd wordt, en niet in de het artikel zelf, dan het artikel niet meenemen in het coderen.
Antwoordcategorieën:
1. Minister Van Engelshoven 2. Minister Bijleveld(-Schouten) 3. Minister De Jonge 4. Minister Wiebes 5. Minister Bussemaker 6. Minister Hennis(-Plasschaert) 7. Minister Schippers 8. Minister Kamp 9. Minister Hillen 10. Minister Verhagen 11. Minister Plasterk
12. Minister Van Middelkoop 13. Minister Klink
