Causes, treatments and attributing responsibility
Framing the gender pay gap issue in Finnish newspapers 1988–2017
Riku Mattila
Student number: 11368217
Master’s Thesis
Graduate School of Communication / University of Amsterdam / Master’s programme
Communication Science
Track: Political Communication
Supervised by Dr. Penny Sheets Thibaut
Abstract
There’s an ongoing debate over the causes and right treatments of the gender pay gap which is still a persistent feature of every industrialized nations’ labour market. My study focuses on how this complex issue is explained and framed in the Finnish newspapers Helsingin Sanomat and
Kauppalehti, between 1988 and 2017. First, I ask which sources have had access to define the issue and suggest treatments for it. Second, I examine which causal and treatment attributions have arisen on the agenda. Third, I consider how the sources have framed responsibility – is the gender pay gap caused by individuals or society – and who’s responsible for the treatment of this problem? Fourth, I make a comparison between two newspapers and how the news has changed over time. This content analysis is conducted on 730 arguments in 385 news stories, addressing causes and treatments for the gender pay gap. My results show that the public discussion around the gender pay issue has given access to many groups – with the exception of disregarding employees and citizens’ voice. My results indicate that some interest groups frame the issue of gender pay gap in a narrow way that serves their own mission. Due to the different standpoints, sources have not always been framing and solving the same problem. My research gives suggestions to journalists' work, to urge them to frame the right problems and, most of all, to challenge public policy makers more accurately with the treatments around this complex issue. Instead of asking how one will solve the whole gender pay gap issue, journalists should ask: which problem are you actually solving?
Introduction
In 1987, the Act on Equality between Women and Men was adopted in Finland. One of the main goals of this act was to promote the equality of women in working life, especially in equal pay (Ministry of Social Affairs and Health, 2015). 30 years after the act, Finland ranks high in the Global Gender Gap Index (Leopold, Ratcheva & Zahidi, 2016) but it still has work to do to fully close the gender pay gap, which is a persistent feature of every industrialized nations’ labour market (Blau & Kahn, 2003). The gender pay gap is a complex issue, and there’s an ongoing debate over the cause of this continued gap, whether it’s caused by occupational segregation (Petersen & Morgan, 2005), gender-differentiated perceptions when making career choices (Correll, 2001), gendered processes determining which occupations pay well (Kilbourne et al., 1994), individual characteristics (Rubery, Grimshaw & Figuiredo, 2005), employment discrimination (Lips, 2013) or shorter and discontinuous work lives (Blau & Kahn, 2007), among other considerations. However, economists believe that discrimination plays at least a partial role in the pay difference (Arulampalam, Booth & Bryan, 2007: Blau & Kahn, 2007). Lips (2013) argues that the role of discrimination is downplayed by attributing the issue mainly to the choices individuals make.
How has the media explained gender pay gap to the audience? The idea of news framing is important in how issues are understood and negotiated in society. When we watch or read the news, or consume social media, we develop opinions on diverse issues (Nelson & Oxley, 1999). Framing helps us and guides us to do so because it highlights and selects only certain aspects (Zhang et al., 2016), especially when the issue is complex. The gender pay gap is a frequently studied issue in fields such as economics, sociology and gender studies, but it hasn’t been examined in the news framing context domestically or internationally.
This study focuses on who has had an access to define problems and suggest treatments for the gender pay gap in the news – and from which perspective they have done it. A content analysis is conducted on 730 arguments in 385 news stories, addressing causes and treatments for the gender pay gap in two Finnish newspapers, Helsingin Sanomat and Kauppalehti, between 1988 and 2017. The time frame of 30 years lets me also make a further analysis of how the definitions have changed over time. When we understand the problem definitions and treatment recommendations that the sources and journalists have addressed for the issue, we can make further conclusions about the public discussion of gender pay gap in the Finnish context. The concerns of the pay gap have stayed in the news throughout the decades and the policy makers have repeatedly awaked for the same issue, for example in the early 1990s (Tuominen, 1990), early 2000 (Pohjanpalo, 2003), and again in 2014 (Savolainen, 2014).
Theoretical background
Sources in the news
Sources are indispensable for journalists. They provide vital information, facts and attributed statements, so they make it possible to approach different news events through a personal (and often expert) perspective. Besides the person sources, journalists also use documents, reports, and other formal sources. According to Berkowitz and Beach (1993), most of the information relevant to news events and issues come from sources, which makes the interaction between journalists and sources an “important force shaping the news” (p. 4). The third actor in this equation is, undoubtedly, the audience. It’s worth noticing that the dependence between journalists and sources is mutual.
Journalists need sources, but sources often need the access to the public (Franklin & Carlson, 2010). Which sources are consulted by journalists has significant implications not only for a particular article but for society more broadly. Franklin and Carlson (2010) offer three approaches to understand the social power of sources. First, the one who becomes a news source is also the one who has the power to speak publicly. This raises a question about access: who has access to be a source? To “privilege the voices of the powerful and marginalize those of the powerless” (Cottle, 2000, p. 427) has been an explicit critique of critical theory about media. If the media favours already-powerful voices, they are more or less consciously maintaining the social, political, economical and cultural power structures, which leads them to, for example, legitimize social inequality (Cottle, 2000). Second, sources have the power to define the world and make choices about which
interpretations should be heard in the public discussion (Franklin & Carlson, 2010), often or at least sometimes based on their own motivations and interests. Third, sources have the power to respond to and counter other sources, which is rarely possible for ‘ordinary citizens’. Therefore, in examining how an issue is framed in news, it is equally important to examine who is allowed to promote these frames.
Getting a chance to act as a source is not only about supplying information. The function of being a source is also “symbolic in supplying legitimacy and authority” (Carslon & Franklin, 2010 p. 4). If journalists use certain elite sources, it may narrow down the standpoints in the news. The impact of this unequal access is in how the range of voices shapes the public opinion and understandings of the world (Carlson & Franklin, 2010). Brown et al. (1987) found that front page news relies heavily on government sources, thus, men in executive positions. Hardin and Whiteside (2010, p. 326) suggest that feminists should offer more “alternative storylines” to media environment. Keyser, Rayemaeckers and Paulussen (2011) add, that in the past decades, more and more interest groups have gained access to media debates, disregarding the ordinary citizen.
Analyzing sources is a central concept on behalf of this framing study because one of my aims is to find which voices have been heard in the news about gender pay gap. However, the census of the types of sources does not indicate with the possible impact the stories have had, but rather
suggest who has had an access to contribute in the public discourse and evaluate the issue. Without enough literature to pose a specific hypothesis about what sorts of sources will be featured, I pose my first research question:
RQ1: Which sources have addressed causal and treatment attributions for the gender pay gap most frequently?
News framing
News plays an essential role in facilitating democracy (Gurevitch & Blumler, 1990). First, media should guard against the abuses of power holders. Second, they should offer diverse opinion sources for the public from the opposing sides, and third, offer meaningful political choices, including the access to participate in the public discussion. Gurevitch and Blumler (1990, p. 27) argue that “the press in a democratic society can be seen as performing an indispensable, bridging function in
democratic politics”. McCombs (2014) takes notice of how majority of the concerns and issues on the public agenda are structured by journalists’ reports and through their daily selection of what are the most important issues. The pictures of our reality are constructed by the information provided by the news media and when the public’s attention to a certain issue is caught, people may think it more worthwhile to have an opinion of it (McCombs, 2014). Scholars, such as Berelson (1952) and
Habermas (1991) see the participation in the public discussion (by electorates) as a core act for a good citizen, which requires relevant information produced by the media.
The concepts of framing and agenda-setting are closely related to each other. Agenda-setting theory sees the power of the media through a power to tell audience about some news events while obscuring the others, and also, what to think about the news (Ghanem, 1997). It shapes the public opinion by suggesting in which topics the audience should even have an opinion. It also gives a voice for some sources, and show some opinions and aspects over another. This is when the concept of framing and its various definitions come around. If agenda-setting looks more on story selection and how does the media influence in issue importance, framing focuses more on the ways those issues are presented and problems formulated for the audience. (Ghanem, 1997.)
According to de Vreese (2005), a news frame provides a way to understand an issue, and it’s a part of journalistic norms and political arguments. Entman (1993, p. 52), defines framing as “to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described”. Gamson and Modigliani (1989) argue that frames are making sense and providing meanings of relevant issues and events. Issues, such as the gender pay gap or poverty, and their causes and solutions, are then contested in the news media (Zhang et al. 2016) and the end result influences citizens’ political attitudes and opinions (Nelson, Oxley & Clawson, 1999). Lawrence (2010), highlights that frames often have political purposes and
political actors target their preferred frames to the public discussion. According to Tuchman (1978, p. 156) frames “narrow the available political alternatives”. Also journalists contribute frames to the news (Gamson & Modigliani, 1989; Nelson & Oxley, 1997), especially in the domestic policy context (Lawrence, 2010).
In this study, I follow framing from three aspects. First, I focus which (political) actors have targeted their preferred frames to the public discussion. Second, I find out which problem definitions and treatment recommendations are selected in the news texts to describe gender pay gap. In the result of this analysis, I will make further implications of how different political alternatives are presented – or possibly limited – in the gender pay gap issue.
Although frames can influence citizens’ political opinions, this study is not examining how strongly the audience is influenced. However, it’s notable that frames are not omnipotent. The information processing and interpretation are always influenced by pre-existing aims that the person has (Scheufele, 1999), not only by frames. The receivers may process the information actively from many sources, or talk to others to understand the issue more, or selectively scan and seek only the information relevant to them (Kosicki & McLeod, 1990). A thorough research about in which
circumstances the audience is influenced by the news frames they’ve been given has done for example by Iyengar (1987, 1990, 1991), Nelson, Oxley, and Clawson (1997) and Lecheler and de Vreese (2012). Iyengar found that a relationship between media frames and audience frames was strongly contingent upon the issue under study. In the experiments by Nelson, Oxley, and Clawson (1997), framing influenced in political attitudes and opinions about welfare policy. Lecheler and de Vreese (2012) found, that people with higher levels of political knowledge were affected more by framing effects.
In conclusion, news frames are not synonyms with media effects. They may affect learning or evaluation of issues (De Vreese, 2005), activate knowledge or create contexts (Cappella & Jamieson, 1997), guide public discourse (Lawrence, 2010) or make judgements and choices (Iyengar, 1996) but the total effect is harder to evaluate. This background of news framing creates context for my study about the gender pay gap issue. Instead of examining the media effects, this study approaches who has had an access to define problems and suggest treatments in the media, and from which perspective they’ve done it. Without enough literature to pose a specific hypothesis about what sorts of frames will be featured, I pose my second and third research question:
RQ2: Which causal and treatment attributions are most frequently mentioned in the news of the gender pay gap?
RQ3: How have the most frequently addressed causal and treatment attributions changed over time?
Responsibility framing
“Responsibility framing” is commonly studied in analyses of the news. In the news media, issues, such as social problems – including their causes and solutions – are ideally presented through various arguments and opinions. As Entman (1993) argues, to frame issues is to define problems, identify causes, and suggest treatments while obscuring other elements and narrowing the political alternatives. The central idea of responsibility framing is the suggestion that someone – typically either the government, society more broadly, or individuals – is responsible for causing a specific problem and/or for addressing it (de Vreese, 2005). Attributions of responsibility can be categorized into causal and treatment responsibilities (Kim, Carvalho & Davis, 2010). Causal responsibility concerns the origin of a problem, and treatment responsibility concerns who or what has the ability to create a solution (Iyengar, 1996). How the media addresses issue responsibilities may shape public opinion about who should be held accountable for causing and solving problems (Zhang et. al, 2016). Furthermore, if social norms favour more collectivistic or individualistic culture, it may influence such responsibility framing. According to Zhang et al. (2016), more collectivistic societies frame responsibilities at the societal level and individualistic cultures attribute responsibilities to personal factors.
A study by Semetko and Valkenburg (2000) found that a ‘responsibility frame’ was the most commonly used in a frame analysis of television and print news about European politics. It was especially evident in more serious news outlets, which matches with this study, too. Their findings suggest that the way in which responsibility is framed in the news is influenced by the political culture and social context in which the news is produced. For example, in the Netherlands, where the social welfare state is strong, the government is expected to provide answers to social problems (Semetko & Valkenburg, 2000). This study approaches gender pay gap from the Finnish perspective, where the welfare system is one of the strongest in the world.
Other studies on different social issues and their responsibility attributions have received various outcomes. Iyengar (1996) conducted an experiment to study how people attribute
responsibility in poverty, terrorism, crime and racial inequality. The results showed that the attribution of responsibility was dependent on a type of coverage in poverty and terrorism issues. If the news coverage was thematic, placing the issue in general context, the societal attributions of responsibility were dominant. If the framing was episodic, reporting more about specific cases, then individualistic attributions of responsibility were dominant. In his earlier experimental research, Iyengar (1991) argued that episodic television coverage encouraged people to hold the individuals responsible for their fate. On the contrary, experimental studies on crime and racial inequality yielded more complex results (Iyengar, 1996). Kim (2015), compared articles about responsibility framing and found that media’s focus between individual and societal responsibility varied across different issues. A study by Zhang et al. (2016), analyzed news about depression from three decades and found that individual-level attributions of responsibility were dominant, but that societal attributions of responsibility
increased over time. Due to highly mixed results in previous literature, it is difficult to pose specific hypotheses for this study. To examine the responsibility frames in the news, I pose my fourth and fifth research question:
RQ4: How have the sources attributed responsibility for the gender pay gap issue? RQ5: How have the attributions of responsibility changed over time?
Framing and gender-related issues
Some feminist media researchers have raised concerns that journalists’ use of frames reinforce gender-related myths in their reporting of issues, and such frames impact social policy decisions (Hardin & Whiteside, 2010). Since the 1970s, studies analyzing media coverage of women have been produced by several feminist researchers (Hardin & Whiteside, 2010). Tuchman (1979), who is frequently cited in feminist media research, has argued that images of women in mass media have “some sort of detrimental impact upon both individual consciousness and collective social life” (p. 530). Williams, Manvell & Bornstein (2006), found that the news narrative of mothers “opting-out’’ of employment in order to take care of children, has had negative impacts on public policy, on employer behaviour, and on younger women. Graff (2007, p.1), argues that “if journalism repeatedly frames the wrong problem, then the folks who make public policy may very well deliver the wrong solution”.
Media hegemony theory is one key notion, which ties framing process to the considerations of power (Hardin & Whiteside, 2010). Carragee and Roef (2004) see that frames produce hegemonic meanings, which have tendency to favour political elites, whereas the social movements, such as feminism, have to challenge these hegemonic values. This view also challenges some
conceptualizations about framing. Frames are not just meanings – evaluating, contextualizing, informing, valuing or guiding – but they are always signs about hegemony and power (Carragee & Roef, 2004).
Gender pay gap is a gendered issue and the news stories in my content analysis are expected to cover a variety of topics around gender and work, such as male and female dominated professions, gender-differentiated perceptions of career or assumed individual characteristics based on gender. I do not expect that attributions in individual level are dominant, such as a way to ‘blame’ women for their problems – but I expect these attributions to be present in the analysis, to some extent. Therefore, it is relevant to discuss about the topics and frames concerning the individual responsibility after the results are further analyzed.
Finnish news landscape
Before turning to the methods, a brief word on the Finnish news landscape – and in particular on the two news sources under examination. Helsingin Sanomat (HS) is the largest daily newspaper
in Finland and its role in the Finnish public discussion is significant (Lounasmeri, 2010). It is seen as a desired arena for influential societal actors, who aim to get their message out in public to legitimate their views and actions (Lounasmeri, 2010). As the “largest daily newspaper and elite media operating in the capital city”, HS has traditionally had close relations to those in power, thus, the voice of political elites is frequently visible (Harjuniemi, Herkman & Ojala 2015, p. 11) and often, the most distinct (Lounasmeri, 2010; Rahkonen, 2006). In their research about Finnish elites’ media use, Kunelius, Noppari and Reunanen (2010), found out that HS is “clearly the number 1 media” among the elites. It’s also seen as an agenda-setter for other Finnish media (Lounasmeri, 2010). HS is owned by Sanoma Oyj, one of the two influential commercial media corporations (which operate alongside YLE, the public broadcasting company).
Finland’s leading business paper, Kauppalehti (KL), is owned by the other strong media corporation, Alma Media. KL reaches the Finnish business and political elites, but it has considerably fewer readers (HS 697 000/day, KL 136 000/day) than HS. The distinction between the leading daily and the leading business paper relates to the genre and the approach of news stories. In their research on Euro crisis coverage in Finnish newspapers, Harjuniemi, Herkman, and Ojala (2015) found that HS typically approached the crisis as a societal and political question, while KL emphasized an
economical approach, and how the crisis affected economically significant actors, such as businesses. KL, specialised in economics, business, and finance, could be labelled as more ‘economistic’ (Arrese & Vara, 2015). According to Arrese and Vara (2015), the economic media supports the logic of interpretation of current events from a pro-business angle, and produces their reports for a more specialised audience, while almost completely exclude the ordinary citizen.
With this in mind, HS and Kauppalehti are both targeted to elite audiences but they
differentiate with their approach of topics. Concerning the differences (or similarities) of gender pay gap news coverage between Helsingin Sanomat and Kauppalehti, I pose my sixth and last research question:
RQ6: How does the gender pay gap coverage differ between Helsingin Sanomat and Kauppalehti?
Methods
To investigate the news coverage of the gender pay gap in Finland, I analyzed 730 arguments about the gender gap which occurred in 385 news stories between 1988 and 2017 in two newspapers. 522 arguments were published in Helsingin Sanomat (HS), the largest daily newspaper, and 208 arguments in Kauppalehti (KL), the largest daily business newspaper. Stories were searched from the newspapers’ own digital news archives. HS has digitalized news from the year 1990 onwards, while KL archive starts in 1988. I chose the time frame beginning from 1988, the year after the Act on Equality was enacted in 1987, to see how the public discussion has evolved after the renewed law.
The relatively long time frame is suitable with the gender pay gap issue because it keeps arising on the media agenda continuously, for example each year as the Statistical Institution of Finland
(Tilastokeskus) publishes the pay statistics. To select newspaper articles, I searched with the keyword ‘palkkaero nai’ (pay gap wom) from HS archive (behind a paywall) which gave me 689 news articles. From this total, I chose every other article. The KL archive (behind a paywall) was more complicated and I needed to do searches with multiple keywords. Keywords were ‘palkka nais’ (pay women),
‘naisen euro’ (women’s euro), ‘palkkaero’ (pay gap), ‘naisen and palkka’ (women’s and pay), ‘palkka and naiset’ (pay and women), ‘palkka and tasa-arvo (pay and equality), and ‘palkkasyrjintä’ (pay and discrimination). These searches gave me in total of 219 articles.
My coding procedure consisted of two phases. In the first phase, each time a writer or a source addresses either a causal attribution for gender pay gap or a treatment attribution for fixing it, this whole argument was coded as a single unit of data collection. A maximum of five arguments were coded per news story. In total, 283 news articles in HS and 102 news articles in KL included one or more causal and/or treatment attributions for the issue. The rest of the articles (269) were excluded because any causal or treatment attributions weren’t mentioned. In the end, there were a total of 730 arguments (HS 522, KL 218), which went through a further coding procedure (Appendix 1). First, I coded if the argument is either causal or treatment attribution for the issue. Second, I coded if either individual or the society is responsible for the cause or the treatment, according to the source. Then I coded the type of a source if it’s the Central Trade Union, a politician or a researcher, among other definitions. These three variables, 1) causal and treatment attributions for the gender pay gap, 2) the attribution of responsibility, and 3) type of a source – play a key role in my analysis about the gender pay gap in the news.
To investigate the causal and treatment attributions further, I needed a second coding phase. In this phase, I coded each causal and treatment attribution (n = 730) to causal and treatment
categories, which were identified and grouped after the first coding round (Appendix 1). For example,
job segregation, discrimination and pay policies were named as causal categories, and equality payments and company responsibility were named as treatment categories. Because some of the
arguments included more than one category, I set a maximum of three categories per argument. The intercoder reliability was tested during the first coding phase with 51 units (7% of total) and five variables: 1) causal or treatment attribution, 2) individual or societal responsibility, 3)
source type, 4) source gender, and 5) source location. Krippendorff’s Alpha was between 0.654 and
0.865, and all but one variable (source location) was above 0.7 (Appendix 2). After the second coding phase, I also did the intercoder test with the causal and treatment categories. 51 randomly selected units (7% of total) were tested and for the first causal and treatment category Krippendorff’s Alpha was 0.779. However, if the causal or treatment attribution included more than one category, the Krippendorff’s alpha was a bit lower, 0.696 for the second category and 0.587 for the last category.
However, only approximately 2 percent of attributions included the third category, so it has few impact for the whole results. (Appendix 2.)
Results
This section focuses on the main findings regarding my research questions. First, I start with the causal and treatment attributions which were identified during two coding phases. To clarify, I coded full sentences and arguments from the news stories in the first phase and then grouped each argument into one or more group which were named example as segregation or discrimination. Table 1 answers to my second research question about which causal and treatment attributions are most frequently addressed in the news of the gender pay gap?
As presented in Table 1, the quantity of causal and treatment attributions was almost even. Segregation as the most frequently addressed cause includes two central themes: occupational segregation between professions (e.g. nurse and engineer) and job level (e.g. executive and expert). It’s worth noticing, that segregation is only fourth of the treatment categories. Three most frequently addressed treatment attributions are policymaker responsibility, equality payments, and company responsibility. The mutual characteristic for all three is the emphasis in the actions made by someone else than an individual, whether it’s a politician, the Central trade union or a company. The list of most frequently addressed causal attributions, instead, gives more emphasis for individual side. For example, the cause category individual characteristics might include arguments about ‘women being too modest’ or ‘men giving more emphasis to money and status’.
Why are the causal and treatment attributions somewhat imbalanced can be partly explained by the different preferences of sources. When I compared column proportions between sources and attributions, I found few significant differences. While the Central Organizations of Trade Unions, politicians and ministries were more likely to emphasis treatments for the issue (p < .05), the researchers were more likely to address causes (p < .05). (Appendix 3.)
Table 1
Which causal (n = 421) and treatment attributions (n=480) are most frequently mentioned in the news of the gender pay gap?
Causal attributions % Treatment attributions %
Segregation 26.8 Policymaker responsibility 23.4
Discrimination 20.0 Equality payments 17.9
Individual characteristics 16.6 Company responsibility 16.2
Pay policy 11.2 Deconstructing segregation 11.1
Career boundaries 9.0 Family policies 10.2
Family and house 6.9 Change in attitudes 6.6
Structures and attitudes 6.9 Individual actions 6.2
Other 2.6 New job evaluation system 6.0
Other 2.6
Total 100.0% 100.0%
The sources and the gender pay gap
In Table 2, I present results for my first research question about which sources have
addressed causal and treatment attributions for the gender pay gap most frequently? The source type was coded after each argument, which addresses either causal or treatment attribution for the gender pay gap.
As seen in Table 2, the researchers, the Central Organization of Trade Unions, and journalists were the three most frequently used sources. Journalists high position is related to the editorial section of the news. In both newspapers, especially in Helsingin Sanomat, gender pay gap issue was often covered in the editorial page and deliberated by journalists. Researchers consist of academic researchers, governmental research institutes, and private research institutes. When comparing the Central Trade Unions, the employers' organizations, and the government, whose are all negotiators in national income policy agreements, the Central Trade Unions STTK, SAK and Akava have brought the issue most diligently on the agenda. Employees and citizens, again, are rarely heard in the news.
Table 2
Which sources have addressed causal and treatment attributions for the gender pay gap most frequently? (n=730)
Source %
Researcher or research institute 27.1 The Central Organization of Trade Unions
or a Trade union 20.8
Journalist 13.2
Politician or Political party 8.9
Employer’s organization or Employer 7.8
Ministry, Government or EU 6.3
Equality official 6.0
Employee or Citizen 4.8
Other 5.1
Total 100.0%
To implicate the information from Tables 1 and 2, I did further research if some of the sources have differences between their most preferred causal and treatment attributions. One significant difference was found when I compared column proportions between sources and causal attributions (Appendix 4). I found out that the employer's organizations and employers are more likely to address individual characteristics as causes than the trade unions and the researchers (p < 0.05). When I compared column proportions between sources and treatment attributions, I found three significant cases. Trade unions attributed equality payments more likely than journalists and researchers (P < 0.05), politicians addressed family policies more likely than trade unions (p < 0.05), and the researchers addressed deconstructing segregation more likely than the trade unions (p < 0.05). (Appendix 5.)
The responsibility attributions of the gender pay gap
In Table 3, I examine my fourth research question about how have the sources attributed responsibility for the gender pay gap? After each causal and treatment attribution, I coded if the source has attributed the responsibility either for society, individual or for both. In some cases, the source hasn’t attributed the responsibility clearly to any group.
Table 3
How have the sources attributed responsibility for the gender pay gap issue? A source mentions a cause (n = 322)
(A) Society is responsible of the cause (B) Individual is responsible of the cause (C)
Both are responsible of the cause
(D)
Unclear who is responsible of the cause
42.5% 22.7% 6.8% 28.0%
A( ,000) A( ,002) A, B, C( ,000)
A source mentions a treatment (n = 383) (A)
Society is responsible for the treatment
(B)
Individual is responsible for the treatment
(C)
Both are responsible for the treatment
(D)
Unclear who is responsible for the treatment
87.7% 7.8% 3.9% 0.5%
B, C, D( ,000) D( ,000) D( ,000)
*Significance level for upper-case letters (A, B, C, D): ,05.
According to the sources, society is more responsible for causing the gender pay gap and, especially, it’s responsible for the treatments to fix the issue. However, in one-fourth of causal attributions, the responsibility is attributed to individuals. These cases relate, for example, to segregation (e.g. occupational segregation) and individual characteristics (e.g. men work more). Causes which have unclear responsibility attribution relate most often to segregation. In these cases, the source has usually addressed segregation as a cause, without going deeper into the topic or separately mentioned if it’s the society influencing to individuals or individuals responsible for their own decisions.
Helsingin Sanomat and Kauppalehti
I compared two Finnish newspapers in this analysis. Table 4 presents the findings concerning to similarities and differences between Helsingin Sanomat and Kauppalehti. First, I compared which were the most frequently discussed causal and treatment attributions in both newspapers. Second, which were the most frequently used sources, and third, is there a difference how the responsibility has been attributed in both newspapers.
Table 4
How does the gender pay gap coverage differ between Helsingin Sanomat and Kauppalehti? (n = 730)
Helsingin Sanomat (HS) Kauppalehti (KL)
Most frequently discussed causal attributions n = 389 Segregation (71) Discrimination (57) Individual characteristics (43) Segregation (42) Discrimination (27) Individual characteristics (27) Most frequently discussed treatment attributions n = 470 Policymaker responsibility (88) Equality payments (67) Company responsibility (53) Company responsibility (23) Policymaker responsibility (22) Family policies (20)*** Most frequently used sources HS (n = 522) KL (n = 208)
Researcher or research institute (137) Central trade union/trade union (107) Journalist (68)
Politician (57)*** Equality official (35)
Researcher or research institute (61) Central trade union/trade union (45) Journalist (28)
Employer’s org or Employer (23)*** Ministry, Government or EU (19)*** The responsibility attributions – causes Society responsible (47.9%)*** Unidentifiable (27.2%) Individual responsible (18.3%) Both responsible (6.6%) Society responsible (32.1%) Individual responsible (31.2%)*** Unidentifiable (29.4%) Both responsible (7.3%) The responsibility attributions – treatments Society responsible (87.9%) Individual responsible (8.3%) Both responsible (3.1%) Unidentifiable (0.7%) Society responsible (87.2%) Individual responsible (6.4%) Both responsible (6.4%) Unidentifiable (0.0%) *** = pairwise difference, significant at the .05 level.
As presented in Table 4, the most frequently discussed causal attributes don’t differ significantly between HS and KL. Family policies is the only treatment, which is more likely to be attributed in KL (p < .05). When I compared column proportions between the newspapers and most frequently used sources, I found out that politicians were more likely to be present in HS – and again – Employer’s organizations and the highest political decision-makers (ministry, government, EU) in the KL. What comes to treatments and attributing responsibility, the results don’t differ. However, KL is more likely to attribute responsibility of the causes for individuals than HS (p < .05) but the results are the other
way around with the societal responsibility attribution (p < .05).
What has changed between 1988 and 2017?
My third and fifth research question address the changes in the causal and treatment categories, and changes in the responsibility attributions over time. My data consists of causal and treatment attributions of the gender pay gap in 30 years’ time period which I have divided in three 10-year-period. These time periods include 174, 319, and 237 causal and treatment attributions.
Table 5 focuses on my third research question about how have the most frequently addressed causal and treatment attributions changed over time? When I compared the column proportions, I found out that Pay policy was more likely to be addressed as cause before 1998 than after (p < .05). Segregation, again, was more likely to be addressed between 1998 and 2007 than before (p < .05). Discrimination is the only one in the three-most-frequent attributions throughout the decades.
Table 5
How have the most frequently addressed causal and treatment attributions changed over time?
(A) 1988-1997 (B) 1998-2007 (C) 2008-2017 Causal attributions 1. Pay policy (B, C)*** 2. Discrimination 3. Segregation 1. Segregation (A)** 2. Discrimination 3. Individual characteristics 1. Segregation 2. Individual characteristics 3. Discrimination Treatment attributions 1. Equality payments (C)*** 2. Policymaker responsibility 3. Job evaluation (C)** 1. Policymaker responsibility 2. Equality payments (C)*** 3. Company responsibility 4. Family policies (A)***
1. Company responsibility 2. Policymaker responsibility 3. Deconstructing segregation 4. Family policies (A)***
*** Significant at the .05 level. ** Significant at the .1 level.
What comes to treatment attributions, family policies are more likely to be addressed as a treatment 1998 onwards (p < .05). In addition, the equality payments were less likely to be addressed as
treatment between 2008 and 2017 than in the earlier decades (p < .05) when it was the most frequently attributed treatment. Especially in the 90s, the Central Trade Unions (and single trade unions)
presented many models how to increase the wage level in women-dominated professions, usually those in the public sector.
Table 6
How have the attributions of responsibility changed over time? (A) 1988-1997 (B) 1998-2007 (C) 2008-2017 Causes Society is responsible
Individual is responsible
Unclear who's responsible
Both are responsible
60.0% (BC)*** 16.7% 16.7% 6.7% 40.9% 25.5% 27.0% 4.8% 36.0% 22.4% 34.4% (A)*** 8.4% (A) 1988-1997 (B) 1998-2007 (C) 2008-2017 Treatments Society is responsible
Individual is responsible
Unclear who's responsible
Both are responsible
86.4% 11.8% (B)** 0 1.8% 90.4% 4.8% 1.2% 3.6% 85.0% 8.4% 0 6.5% *** Significant at the .05 level.
** Significant at the .1 level.
Table 6 answers to my fifth research question about how have the attributions of responsibility changed over time? The comparison of column proportions between three 10-year time periods and responsibility attributions exposes that societal responsibility of causes was more likely to be addressed between 1988 and 1997 than later (p < .05). The unclear responsibility attribution, again, differs significantly in the last 10-year-period compared to the first period (p < .05). The increased amount of segregation category attributed as a cause also increases the amount of unclear
responsibility attributions. What comes to treatments and responsibility attributions, there has been only few changes over time.
Discussion
This study has examined how two leading newspapers in Finland have covered the issue of gender pay gap and maintained public discussion around the issue between 1988 and 2017. The topic has gained plenty of publicity during the whole analysis period. A content analysis was concluded by
730 causal and treatment attributions about the gender pay issue. It is complex for decision-makers to fix, complex for journalists to explain, and, arguably, for citizens to internalize. In the light of my results, if the issue is examined through a pay gap between women and men dominated professions; or through individual choices; or through pure discrimination; or even through a simple mathematical calculation, it changes significantly how treatments are – and how they should be – positioned. One challenge in this issue is that sources are talking about distinctly independent challenges under the same topic. Raising salaries of nurses, changing family policies, or deconstructing segregation in the job market are not related to each other. The first one is aiming to raise the valuation of a specific profession, the second one changing how parents could balance the costs of parenting, and third one is a bigger attitudinal change in the whole society. The only similar argument for most sources in this study was the mutual understanding of the aim of getting rid of the gap. Some promised to fix it in ten years, some in five years, some said even 50 years won’t be enough. Regardless, no one has yet succeeded.
My first research question approached gender pay gap from the perspective of news sources. As Franklin and Carlson (2010) argue, the one who becomes a news source is the one who has power to speak publicly, define the issue and make choices about which interpretations should be heard in the discussion. Based on my content analysis, the public discussion around the gender pay issue has given access to many groups, from decision-makers to interdisciplinary researchers and equality officials. However, two arguments should be discussed further.
First, the share of company representatives, employees and citizens in sources is considerably small. The issue is rarely personified through real cases and people, but usually generalized as third-party talk about ‘public sector women’, ‘export business men’, ‘women leaders’ or ‘paper industry men’. Even many of the articles in the opinion section are written by experts, not citizens. As Keyser, Rayemaeckers and Paulussen (2011) have argued, more and more interest groups have gained access to media debates in the past decades but they still lack the voice of the ordinary citizen. Where are the voices of the discriminated; or successful female leaders; or ‘modern’ couples who share childcare responsibilities? As Cottle (2000) says, if the media favours already-powerful voices, they are more or less maintaining the social, political and cultural power structures. Job segregation was the most frequently addressed cause for the gender pay gap in this study. How can society do something for job segregation if it keeps maintaining those power structures?
Regarding my second argument, Franklin and Carlson (2010) argue that sources often or at least sometimes define the issues based on their own motivations and interests. The Central
Organizations of Trade Unions, such as STTK, SAK and Akava are responsible for their members. Politicians are responsible for the voters. Employer’s organizations are responsible for companies. My analysis examined that the causal and treatment attributions were somewhat imbalanced. While the researchers addressed that deconstructing segregation is the right treatment for the issue, the Central Trade Unions and trade unions addressed equality payments for lower-wage professions as the right
treatment. Employer’s organizations, again, were more likely to address individual characteristics as a cause for the gender pay gap. This comes back to the challenge that I mentioned in the beginning of this section – the sources are talking about distinctly independent challenges under the same topic. However, my analysis shows that journalists were also participating actively in the discussion in the editorial page. Around this issue – and concerning to the lack of citizens’ voices – I keep journalists’ active participation important because the definitions could be otherwise interpreted mainly by sources whose frames have political purposes, as Lawrence (2010) has highlighted. Still, journalists should challenge the decision-makers even more in the future. If they claim publicly that they are solving the gender pay gap issue, journalist should ask: which problem are you actually solving? In my third research question, I asked how have the most frequently addressed causal and treatment attributions changed over time? In the light of my results, discrimination has stayed in the list of most frequently mentioned causes over the decades and segregation have become the most addressed cause after 1998. Before 1998, the sources addressed most frequently that current pay policy is causing the issue, and also, that equality payments are the best treatment for fixing it. This is an example of a news frame which has probably had political purposes, as Lawrence (2010) has argued. Pay policy and equality payments are both related with Finnish national income policy agreements (known as ‘tupo’), which are tripartite agreements between Finnish trade unions, employers' organizations, and the government. These negotiations cover national and economical issues, including salaries. Agreements are usually made for a two-year period. So-called “equality payments”, such as minor percent increases to lower-wage industry salaries, were sometimes added to the agreements, aiming to narrow the gender pay gap. However, the unsuccessful efforts to integrate these payments in reality also led to arguments against pay policy agreements and how they maintain the gender pay gap. In the light of this historical background, framing the gender pay gap issue could have been part of a negotiation strategy for the Central Organizations of Trade Unions. My results show that addressing equality payments as a treatment for the gender pay gap dropped after 2008. According to Parviainen (2008), the year 2007 symbolises a change for Finnish national income policy agreements because it was the first time after 40 years when employer’s organizations refused to participate in the negotiations. One argument for them was that competition in the global markets requires more flexibility in the job market and the current negotiation model doesn’t support this (Parviainen 2007).
Almost 90 percent of sources attributed responsibility for the treatments to fix the gender pay gap issue for the society – either in a larger societal scale or in a company responsibility level. This result was same in both newspapers. In Finnish culture, social norms favour more a collectivistic than individualistic culture, which may have influenced such outcomes in responsibility framing, as Zhang et al. (2016) have previously found. However, society’s role as causing the issue have declined after 1997. Between 2008 and 2017 the share of unclear responsibility attributions has raised significantly. Partly due to the fact that segregation is frequently addressed as a cause, the share of “unclear”
responsibility attributions is bigger. The sources are not clearly attributing who or what causes segregation. It’s probably hard to address but it’s also problematic not to address it. How the media addresses issue responsibilities may shape public opinion about who should be held accountable for causing and solving problems (Zhang et. al, 2016) but if it doesn’t hold anyone accountable, I wonder if anything will change.
In my sixth research question, I compared if the gender pay gap coverage differ between Helsingin Sanomat and Kauppalehti. My results show that they have some minor differences in the treatment attributions, used sources, and attributions of responsibility, but overall their coverage and standpoints follow each other. This might follow the earlier findings made by Lounasmeri (2010) who has argued that Helsingin Sanomat is an agenda-setter for other Finnish media, too. When the gender pay gap issue has been on the agenda, both newspapers have mostly used similar sources.
There are some limitations to this study. First, I limited the data gathering only to newspapers and online news. For example, Iyengar (1991) has made experimental studies with television coverage with different results. Second, I gathered the data only from the news texts, addressed by the sources. According to Entman (1993) frames are located in the images, the keywords and also, in the culture. Some elements are also more highlighted than others but my study addressed all data equally. Third, I didn’t study how people have processed the information, only how the sources have framed the issue. Iyengar (1987, 1989, 1991) has previously found that a relationship between media frames and audience frames is strongly contingent upon the issue under study. Consideration of how different demographics build frames and attribute responsibility upon the gender gap news coverage could be an interesting approach in the future research. Especially, how does the audience frame the question of segregation – do they think it is just an outcome of our individual choices and divergent
preferences – or is it a result of discriminative, stereotyped, and biased gender norms in our society? Multiple research papers and reports were also mentioned in my analysis. Many of them considered the questions of the actual pay gap. Why do the women earn less? Most of the time, the answer was the same: men and women do different jobs inside the company, the industry or in a national context. To understand the issue more broadly requires more questions than this. What actions could some specific industries or companies do to decrease discrimination and segregation? What actions have they done so far? What actions should they do next? I think that journalists’ role is significant. As Graff (2017, p. 1) argued: “if journalism repeatedly frames the wrong problem, then the folks who make public policy may very well deliver the wrong solution” – or not solution at all.
References
Arulampalam, W., Booth, A. L., & Bryan, M. L. (2007). Is there a glass ceiling over Europe? Exploring the gender pay gap across the wage distribution. ILR Review, 60(2), 163-186.
Berelson, B. (1952). Democratic theory and public opinion. Public Opinion Quarterly, 313-330.
Berkowitz, D., & Beach, D. W. (1993). News sources and news context: The effect of routine news, conflict and proximity. Journalism Quarterly, 70(1), 4-12.
Blau, F. D., & Kahn, L. M. (2003). Understanding international differences in the gender pay gap. Journal of Labor economics, 21(1), 106-144.
Blau, F. D., & Kahn, L. M. (2007). The gender pay gap have women gone as far as they can?. The Academy of Management Perspectives, 21(1), 7-23.
Brown, J. D., Bybee, C. R., Wearden, S. T., & Straughan, D. M. (1987). Invisible power: Newspaper news sources and the limits of diversity. Journalism Quarterly, 64(1), 45-54.
Cappella, J. N., & Jamieson, K. H. (1997). Spiral of cynicism: The press and the public good. Oxford University Press.
Carragee, K. M., & Roefs, W. (2004). The neglect of power in recent framing research. Journal of communication, 54(2), 214-233.
Correll, S. J. (2001). Gender and the career choice process: The role of biased self-assessments. American journal of Sociology, 106(6), 1691-1730.
Cottle, S. (2000). Rethinking news access. Journalism Studies, 1(3), 427-448.
De Keyser, J., Raeymaeckers, K., & Paulussen, S. (2011). “Are Citizens Becoming Sources?. Journalists, sources and credibility: new perspectives, 139-151.
De Vreese, C. H. (2005). News framing: Theory and typology. Information Design Journal & Document Design, 13(1).
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of communication, 43(4), 51-58.
Franklin, B., & Carlson, M. (Eds.). (2010). Journalists, sources, and credibility: New perspectives. Routledge.
Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American journal of sociology, 95(1), 1-37.
Ghanem, S. (1997). Filling in the tapestry: The second level of agenda setting. Communication and democracy: Exploring the intellectual frontiers in agenda-setting theory, 3-14.
Graff, E. J. (2007). The opt-out myth. Columbia Journalism Review, 45(6), 51. Retrieved 25th January, 2018 from https://archives.cjr.org/essay/the_optout_myth.php.
Gurevitch, M., & Blumler, J. G. (1990). Political communication systems and democratic values. Democracy and the mass media, 269-289.
Habermas, J. (1991). The structural transformation of the public sphere: An inquiry into a category of bourgeois society. MIT press.
Harjuniemi, T., Herkman, J., & Ojala, M. (2015). Eurokriisin politisoituminen suomalaisissa sanomalehdissä. Media & viestintä, 38(1).
Iyengar, S. (1987). Television news and citizens' explanations of national affairs. American Political Science Review, 81(3), 815-831.
Iyengar, S. (1990). Framing responsibility for political issues: The case of poverty. Political behavior, 12(1), 19-40.
Iyengar, S. (1991). Is anyone responsible?: How television frames political issues. University of Chicago Press.
Iyengar, S. (1996). Framing responsibility for political issues. The Annals of the American Academy of Political and Social Science, 546(1), 59-70.
Kilbourne, B. S., England, P., Farkas, G., Beron, K., & Weir, D. (1994). Returns to skill,
compensating differentials, and gender bias: Effects of occupational characteristics on the wages of white women and men. American Journal of Sociology, 100(3), 689-719.
Kim, S. H., Carvalho, J. P., & Davis, A. C. (2010). Talking about poverty: News framing of who is responsible for causing and fixing the problem. Journalism & Mass Communication Quarterly, 87(3-4), 563-581.
Kim, S. H. (2015). Who is responsible for a social problem? News framing and attribution of responsibility. Journalism & Mass Communication Quarterly, 92(3), 554-558.
Kosicki, G. M., & McLeod, J. M. (1990). Learning from political news: Effects of media images and information-processing strategies. Mass communication and political information processing, 69-83.
Kunelius, R., Noppari, E., & Reunanen, E. (2010). Media vallan verkoissa.
Lawrence, R. G. (2010). Researching political news framing: Established ground and new horizons. Doing news framing analysis: Empirical and theoretical perspectives, 265-285.
Lecheler, S., & de Vreese, C. H. (2012). News framing and public opinion: A mediation analysis of framing effects on political attitudes. Journalism & Mass Communication Quarterly, 89(2), 185-204.
Leopold, T. A., Ratcheva, V., & Zahidi, S. (2016). The global gender gap report 2016. World Economic Forum. Retrieved 25th January, 2018 from http://reports.weforum.org/global-gender-gap-report-2016/the-global-gender-gap-report-2016/.
Lips, H. M. (2013). Acknowledging discrimination as a key to the gender pay gap. Sex roles, 68(3-4), 223-230.
Lounasmeri, L. (2010). Kansallisen konsensuskulttuurin jäljillä: Globalisaatioajan Suomi Helsingin Sanomissa.
McCombs, M. (2014). Setting the agenda: Mass media and public opinion. John Wiley & Sons.
Ministry of Social Affairs and Health. (2015). The Act on Equality between Women and Men 2015. Retrieved 25th January, 2018 from
https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/75131/Act_on%20Equality_between_wome n_and_men_2015_FINAL.pdf?sequence=1.
Nelson, T. E., Clawson, R. A., & Oxley, Z. M. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(3), 567-583.
Nelson, T. E., & Oxley, Z. M. (1999). Issue framing effects on belief importance and opinion. The journal of politics, 61(4), 1040-1067.
Petersen, T., & Morgan, L. A. (1995). Separate and unequal: Occupation-establishment sex segregation and the gender wage gap. American Journal of Sociology, 101(2), 329-365.
Pohjanpalo, O. (2003). Palkkatasa-arvoa ei vaadita suureen ääneen. Helsingin Sanomat. Retrieved 25th January, 2018 from https://www.hs.fi/kotimaa/art-2000004125465.html.
Rahkonen, J. (2006). Journalismi taistelukenttänä. Suomen Nato-jäsenyydestä käyty julkinen keskustelu 2003-2004. Tampere University Press.
Rubery, J., Grimshaw, D., & Figueiredo, H. (2005). How to close the gender pay gap in Europe: towards the gender mainstreaming of pay policy. Industrial Relations Journal, 36(3), 184-213.
Savolainen, J. (2014). Sdp:n Rinne vaatii naisille parempia palkkoja. Helsingin Sanomat. Retrieved 25th January, 2018 from https://www.hs.fi/politiikka/art-2000002774152.html.
Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of communication, 49(1), 103-122.
Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. Journal of communication, 50(2), 93-109.
Tuchman, G. (1978). Making news: A study in the construction of reality.
Tuchman, G. (1979). Women's depiction by the mass media. Signs: Journal of Women in Culture and Society, 4(3), 528-542.
Tuominen, E. L. (1990). Palkkaerot poistettakoon vihdoinkin. Helsingin Sanomat. Retrieved 25th January, 2018 from https://www.hs.fi/mielipide/art-2000003018290.html.
Zhang, Y., Jin, Y., Stewart, S., & Porter, J. (2016). Framing responsibility for depression: how US news media attribute causal and problem-solving responsibilities when covering a major public health problem. Journal of Applied Communication Research, 44(2), 118-135.
Whiteside, E., & Hardin, M. (2010). Public relations and sports: Work force demographics in the intersection of two gendered industries. Journal of Sports Media, 5(1), 21-52.
Williams, J. C., Manvell, J., & Bornstein, S. (2006). " Opt Out" or Pushed Out?: How the Press Covers Work/Family Conflict. The Center for WorkLife Law.
Appendix
Gender pay gap codebook
The aim of this codebook is to identify, which causal and treatment attributions (1-2) have been suggested for the gender pay gap issue; to what extent are individuals or societal actors responsible of the causal or treatment attribution (3); and which sources have suggested these particular causes and solutions (4-7) for the issue. After these steps, the basic information of each news story (8-15), such as the date, headline and news section, will be coded. After the first coding round (1-15), each causal and treatment attribution will be coded into the topic categories. As a research material, I use news stories published in Helsingin Sanomat and Kauppalehti between 1988-2017.
Coding phase 1, Part 1:
1. Causal and treatment attributions for gender pay gap in Helsingin Sanomat and Kauppalehti 1998-2017 (pay_gap_in_news)
Each time the gender pay gap is addressed in the news story, and a writer or a source addresses either a causal attribution for the issue or a solution for fixing it, a new unit is coded as the row in the excel file. Maximum of five causes and solutions are coded per news story. If the story contains more, the first five will be coded. Some of the news stories don’t necessarily include any cause or solution, while some could have one, or even more than five. If a same source addresses more than one separate argument, it counts as separate rows in excel. But if a source presents multiple arguments in same paragraph, it is counted as one.
Each cause or solution could be suggested either by 1) the writer; journalist, columnist or a writer of an opinion text, or by 2) the source, interviewed or referenced by a journalist. Thus, the cause or solution might be presented as 1) the opinion by the writer; 2) the direct or indirect quote; or 3) the interpretation of a writer of something said or written by a source. In this codebook, I will call opinions, quotes and interpretations collectively as ‘arguments’.
E.g.
1 Cause suggested by the journalist, opinion: “The cause for the pay gap is extremely simple: women do different work than men. Women could, if they want, apply more to tech industry, where the salaries are higher than in the health care.”
2 Solution suggested by the source, indirect quote: “EU commissioner of Gender Equality promises new action plan considering to gender pay gap issue.”
3 Cause suggested by a journalist, interpretation/opinion: “According to Richard D. Mandell, a US historian, women’s absence in sports has influenced in their absence of boards of directors and high salary jobs. (...) it looks that women’s smaller earnings in Finnish national football team are a continuity for this hegemonic use of power.”
Only the causes and solutions, which distinctly concern the gender pay gap issue will be coded. Some of the stories cover gender pay gap as the main topic, but it might be covered as side topic, for example, in the news story about gender equality in general. In stories like this, the coder has to pay attention if the argument concerns to gender pay gap or not. If it’s not obviously about the pay gap, the coder should not code it. In addition, the arguments might concern the overall issue or the issue in some specific industry (public health care, tech, banking sector, sports, etc.) Both should be coded.
Due to a large spectrum of causes and solutions over time, all of them are first written in the codebook by copying the text from the news, or describing them clearly. When a coder writes either cause or solution to the codebook, two back up questions should be asked: 1) Does the following argument contain either cause or solution for gender pay gap? 2) Can any reader understand what the description of either cause or solution addresses? Ps. the causes and solutions should be written to codebook in Finnish, but the following examples are in English.
E.g.
This is accurately coded: “The salaries of women dominated professions, such as nurses and nannies, can’t reach the same level with private sector salaries due to the differences in cost efficiency” This is an incomplete coding of the same attribution: “Differences in cost efficiency.”
Why? Certain sectors are mentioned. In the end of the coding, it is good to know if some sectors are more discussed than others.
This is good: “There are problems in women’s career paths (considering to pay), and also men should face the facts. Less than 6 percent of men use their justification for parental leave. This not: “The uneven use of parental leaves.”
Why? The first one addresses that men are responsible for fixing the issue.
2. Cause or solution? (cau_sol)
Once the argument to be coded has been identified through the first step, it should also be coded according to the following coding scheme. First, the coder chooses if the argument addresses a causal or a treatment attribution for gender pay gap. Code either:
1 = Cause
If the argument of a source or the writer addresses a cause for the gender pay gap. 2 = Solution
If the argument of a source or the writer addresses a solution or treatment for the gender pay gap. 3 = Both
If the argument of a source addresses both, a cause and a solution for the gender pay gap. 99 = Undetermined
If unclear, choose this.
3. Individual or societal responsibility? (cau_sol)
The aim of this code is to find out who is responsible for the causal or treatment attribution for the gender pay gap, according to the writer or the source. Code either:
1 = Individual
If the source or the writer suggest that the gender pay gap issue is to some extent caused by individuals or to some extent individuals are responsible for the treatment.
2 = Societal
If the source or the writer suggest that gender pay gap issue is to some extent caused by societal injustice or that to some extent society-at-large or some societal actor is responsible for the treatment.
3 = Both
If a source addresses that both are to some extent responsible, choose this. 99 = Undetermined
If the responsibility attribution is unclear, choose this*.
*during the first practise round of intercoder pretest, we noticed that causes which consider to segregation (e.g. men and women work in different professions and/or levels), the responsibility is
rarely pointed to anyone. Here, the sources might use expressions, such as “men and women are selected”, “men and women have been divided” or “men and women have ended up” in different positions (not by their own choice, per se). These expressions are unclear. However, if the source clearly says that “men and women apply to different sectors”, it can be coded as individual responsibility.
4. Source in the news story (src_type)
This coding unit describes which types of sources have had the chance to present their arguments in the news. Choose the right code to describe the source of the argument. If the origin of the source is not clearly mentioned in the story, the coder should try to google more information about it.
1 = The Central Organisation of Finnish Trade Unions (työntekijöiden keskusjarjesto) or its representative (e.g. STTK, Akava, SAK)
2 = Employers’ organization (työnantajajärjestö) or its representative (e.g. EK, KT, VTML) 3 = Trade union (ammattiliitto) or its representative
4 = Politician/Political party 5 = Academic researcher
6 = Non-governmental or private research institute/company or its representative
7 = Governmental research institute, association, governmental working group (“tyoryhma”) or its representative
8 = Company/Association/State-owned company or its representative 9 = Employee
10 = Citizen (when a profession is not mentioned) 11 = Journalist** (e.g. if a column or editorial)
12 = Equality official (e.g. The Ombudsman for Equality, the EU Commissioner for Gender Equality, Minister of Equality)
13 = Other newspaper
14 = Ministry/Government/EU Commission/Parliament 15 = Workplace steward (luottamusmies)
16 = Writer/Actor/Culture person 99 = Unidentifiable
**When to code a journalist?
Very often, journalists rely on sources when they write a news story and present facts, because journalists are rarely the best experts of the issues they write about. However, in some text types, such as in editorials and columns, the journalist’s own voice is more likely to be present. In these cases, journalist might present arguments, which are not clearly based on sources, or, the sources are not
mentioned. When the coder recognizes a case, either causal or treatment attribution for gender pay gap, which is addressed per se by the journalist, without a clear reference to a source, the journalists unique code should be coded.
5. Location of the source (src_location)
The location of the source is primarily determined by the person. If the person is Finnish or represents a Finnish company, the coder marks Finland (1) as the location. If the source is not a person, but for example a company, a trade union or a research institute in general, the location is determined by the company headquarters. If the location is not mentioned in the news story, the coder should try to google more information of it. Otherwise, code ‘unidentifiable’.
1 = Finland 2 = Foreign
99 = Unidentifiable
6. Source gender (src_gender)
If the name of the source clearly belongs to a man or a woman, or the coder knows this person, or there’s a photo of a source, the gender is easy to code. If not, the coder should try to google more information. Otherwise, code ‘unidentifiable’.
1 = Man 2 = Woman
99 = Unidentifiable
7. Source name (src_name)
Copy the name of the source from the news story to a codebook. Write the name first (if a person) and add the organization second.
Coding phase 1, Part 2:
8. Link for the news story (link)
Copy the unique URL address from the browser to the Excel sheet
9. News source (newspaper)
Code the news source from which the news story was derived. 1 = Helsingin sanomat
10. Date (date)
Copy the date that the news story was published. If the date is not in the following form, coder should write it herself. DD/MM/YYYY
11. Headline (headline) Copy the title of the news story.
12. Number for the news story (number)
Each news story will be assigned a unique number. This will be determined by the order in which the story was analyzed (e.g. first news story analyzed will be given a “1”). There will be one to five units per each news story. All units found from the same news story will get the same number.
13. Journalists gender (jrnlst_gender)
If the name of the journalist belongs clearly to a man or a woman, or the coder knows this person, or there’s a photo of this person, the gender is easy to code. If not, the coder should try to google more information. Otherwise, code ‘unidentifiable’.
1 = Man 2 = Woman
3 = Both (e.g. more than one journalist) 99 = Unidentifiable
14. Section (section)
Copy the section of the newspaper from which the news story was derived. (domestic, economy, politics, sport, etc.). If the section is not mentioned, write ‘news’.
15. Column (column)
Code ‘yes’ if the news story is a column. 0 = no
