Differences Based on Individual- and Organizational-level Factors in Experiences of External Interference among Finnish Journalists

ABSTRACT

This article explores the degree to which journalists in Finland experience different types of external interference and how they perceive the implications of interference. For this study, external interference is defined as all active and invasive methods that external actors use to influence journalists and interfere in the journalistic processes to influence editorial content. By using Finland as a case example, this article provides new empirical evidence on how external interference manifests in the contemporary journalistic environment in a democratic Western country with strong safeguards for press autonomy. Based on the statistical analysis of survey responses from 875 Finnish journalists, the results indicate that individual-level factors of age and gender have only a marginal relation to the prevalence of external interference. Of analyzed organizational-level factors—employment type, occupational position, and media outlet used for reporting—the latter two were most significant. This article offers three important empirical contributions: (1) it highlights the existence of editorial defense shield as journalistic practice; (2) it illustrates the complex relationship between gender and external interference; and (3) it demonstrates how journalists in national and regional newspapers are more prone to interference than their colleagues in other media outlets.

KEYWORDS:

• Journalism (profession)
• autonomy
• interference
• survey
• harassment
• intimidation

Introduction

Journalistic autonomy can be understood as the professional aspiration of journalists to protect their work from any attempts of manipulation or interference (Kunelius 2003, 23). As the societal environment is changing due to technological, social, political, and cultural developments, these transitions are reflected in the journalistic work and professional autonomy of journalists (Reich and Hanitzsch 2013; Waisbord 2013; Nygren, Dobek-Ostrowska, and Anikina 2015). The techniques of manipulation and interference that journalists face are also constantly shifting, and new methods are frequently introduced (Luostarinen 1994, 73). Empirical research systematically assessing new challenges and threats to journalistic autonomy is therefore needed to identify these shifts and develop effective responses.

In the contemporary communication environment, the authority, position, and power of traditional media actors is restructuring (Chadwick 2017), and journalistic work is transitioning from monologue to more of a dialogue with audiences (Nygren, Dobek-Ostrowska, and Anikina 2015, 79). Especially online, journalists are more visible and accessible than ever and find their choices, credibility, and integrity more often publicly contested (Löfgren Nilsson and Örnebring 2016, 881). Working in online environment exposes journalists to new detrimental phenomena like crowdsourced harassment (e.g., Binns 2017), hate speech (e.g., Obermaier, Hofbauer, and Reinemann 2018), cyberstalking and state-sponsored trolling campaigns (e.g., Luque Martinez 2015) and gendered harassment (e.g., Adams 2018; Chen et al. 2018).

Multiple organizations compile statistics on extreme intimidation of journalists, but systematic investigations or statistical monitoring of any of the less severe forms of interference and manipulation have been rare (Parker 2015, 3). The lack of previous research is especially true in the context of democratic and stable Western societies, where external pressure and threats have rarely been studied or discussed (Löfgren Nilsson and Örnebring 2016) and there has long existed a tendency to take journalistic autonomy for granted (Reich and Hanitzsch 2013, 133–134).

This research represents an exploratory attempt to empirically study external interference and its perceived implications among Finnish journalists. Nordic countries generally rank high in measures of media freedom, and Finland topped the World Press Freedom Index seven consecutive times, from 2009 to 2016 (Reporters Without Borders [RSF] 2016). Systemic factors supporting media autonomy and freedom in Finland include strong position of public service broadcast media, high newspaper circulation, high levels of professionalism, institutional self-regulation with independent press council and low political parallelism in media (Nord 2008). These are reflected in the journalistic culture, as Finnish journalists report a high degree of perceived professional autonomy and seem to regard themselves distant from direct political, government, and business influences (Pöyhtäri, Ahva, and Väliverronen 2014). In comparison, Finnish journalists feel less economic pressure than journalists in non-Nordic countries where press freedom is high, indicating that the influence of commercialization and growing competitiveness are moderated by the traditions and structural particularities of the Nordic media system (Ahva et al. 2017, 607). However, new concerns have arisen in Finland due to perceived increase in anti-media rhetoric, and harassment, intimidation, and public defamation targeting journalists (see e.g., Haapalainen 2016). These concerns highlight new challenges and threats to journalistic autonomy brought on by the contemporary communication environment.

By using Finland as a case example, this research seeks to demonstrate how external interference manifests in the contemporary journalistic environment within a democratic Western country that has strong legal, cultural, and institutional safeguards for press autonomy. Expanding on previous research (Hiltunen 2019), this article uses statistical analysis to explore differences in experiences of external interference and its implications based on individual- and organizational-level factors. This approach provides valuable new empirical insights to the contemporary journalistic environment and pinpoints developments that deserve further attention.

The research questions are as follows:

1. What kind of differences, based on individual-level factors (age or gender) and organizational-level factors (employment type, medium used for reporting or occupational position), can be observed in experiences of external interference among Finnish journalists?

2. What kind of differences, based on individual-level factors (age or gender) and organizational-level factors (employment type, medium used for reporting or occupational position), can be observed in perceived implications of external interference on journalistic work and the journalistic profession?

By answering these research questions, this article explores the connections that can be identified between experiences of external interference and its perceived implications among different groups and seeks to recognize factors that contribute to these connections. This article contributes to an ongoing discussion surrounding journalistic autonomy in the contemporary communication environment and threats to press freedom in Europe and Western democracies.

The article is structured as follows: We begin by outlining the theoretical framework of the study and the concept of external interference, and present previous research conducted on the subject. This segment is followed by an introduction to our methodology, the research sample, and an analysis of our findings.

Theoretical Framework: External Interference

Journalistic autonomy acts as a fundamental building block of the shared professional ideology of modern journalism that, besides autonomy, consists of public service, objectivity, immediacy, and ethics (Deuze 2005). Of these elements, objectivity especially is considered problematic and there are ongoing debates regarding the prospect of objectivity and whether it should be replaced by ideals such as “neutrality”, “fairness”, “professional detachment” or “impartiality” (Deuze 2005, 448; Boudana 2011). However, both objectivity and the proposed alternatives share similar underlying notion that highlights the importance of independent decision-making based on autonomous journalistic deliberation. Autonomy can, therefore, be understood as a necessary prerequisite that makes ethical consideration, independent reporting, and serving shared public interests instead of private ones possible, thus enabling features that separate journalism from other forms of mass communication (Kunelius 2003, 23–25).

The professional ideology of journalism and its ability to create a collective public forum for diverse interests makes it an attractive target for interference (Luostarinen 1994, 28–29). The audience expectations for journalism are radically different compared, for example, to advertising or political communication that are fundamentally recognized as partisan and persuasive forms of communication (Kunelius 2003, 23–25). Therefore, the benefits of being able to steer or influence journalism to one’s benefit or strategically silence it can be significant (Luostarinen 1994, 53–56).

Studies typically separate two analytical dimensions of journalistic autonomy—external and internal. The external dimension is related to societal and political autonomy of journalism and how protected journalists are from coercive forces external to journalistic organizations. The internal dimension highlights the extent to which journalists can make their decisions free from management pressures and other forces inside media outlets (Reich and Hanitzsch 2013, 135; Nygren, Dobek-Ostrowska, and Anikina 2015, 80–81).

This study focuses on the external dimension of autonomy and explores methods that external actors use while trying to transgress the boundaries of journalistic autonomy. For this purpose, external interference is defined broadly as all active and invasive methods actors external to journalistic organizations use to interfere in the journalistic process and/or influence journalists and/or editorial content (see also Luostarinen 1994; Shoemaker and Reese 1996). Mixed interference, defined as situations in which external interference is transmitted internally through media organization’s chain of command, is also included within the scope of the study (Hemánus 1983, 192). Typical example of mixed interference would be a situation in which an editor after being subjected to external pressure orders reporters to stop investigating certain case or topic (Goyanes and Rodríguez-Castro 2018; see also Kuutti 1995, 248). Methods of external interference can also be used pre-emptively to induce chilling effect and self-censorship among journalists with the aim of influencing journalism in the future (Nerone 1994; Clark and Grech 2017).

The definition of external interference used in this study emphasizes the individual experience and perspective of the journalist, thus rendering the social phenomenon of external interference accessible to the researcher (c.f. Reich and Hanitzsch 2013, 136). By using this type of a broad definition of external interference, it is possible to simultaneously study low-intensity interference alongside more intrusive and aggressive methods. This allows for a more balanced overall picture of different aspects of the phenomenon in the contemporary journalistic environment. This definition, however, excludes standard supply-side public relations (PR) activities, such as press releases and conferences, information subsidies, and other similar agenda-building efforts. While one can argue that these practices can have significant impact on journalistic content, standard PR-activities do not actively aim to limit or violate external autonomy of journalism and are therefore not included in the scope of this study.

Previous Research

Urbániková (2019, 4–5) states that studies of journalistic autonomy have typically examined general perceptions of freedom instead of the occurrence and incidence of concrete types of interference or its implications for journalistic work. This is especially true for stable and democratic Western countries, as very few studies have explored the prevalence or effects of external interference in this context. In the following, we will highlight the key findings of previous studies on external interference in the Western context.

Parker (2015) studied 286 working journalists from Australia, Canada, New Zealand, the United Kingdom, and the United States, and found that 63 percent (60 percent of women and 68 percent of men) reported experiencing occupational intimidation, defined as “behaviors that pressure or threaten individuals with harm or sanctions in order to discourage the carrying out of certain job duties”. No connections were discovered between individual-level factors like age, gender or ethnic background, and occupational intimidation, but male journalists reported more physical intimidation, suggesting that gender might be a risk factor for high-intensity intimidation behaviors. Employment type did not affect occupational intimidation. Type of story covered was a strong risk factor, as journalists covering politics, war/international affairs, human rights, and investigative reporting reported increased levels of occupational intimidation. Experiences of occupational intimidation were negatively related to job performance. Based on the findings, Parker concludes that, among journalists, occupational intimidation is mainly instrumental aggression used to achieve some desired objectives or goals related to journalism, and therefore is not primarily motivated by the demographic characteristics of journalists.

Clark and Grech (2017) studied experiences of unwarranted interference among 940 journalists in Council of Europe member states and Belarus. Clark and Grech found that respondents from Western European EU and non-EU countries reported a high prevalence of threats of violence, psychological violence, cyber bullying and sexual harassment during the past three years. Although this region had the second-lowest levels of self-censorship and other direct effects on journalistic work, a considerable share of journalists admitted to, for example, toning down critical stories, and being selective about reporting because of interference. Among all respondents, male journalists were significantly more likely to be threatened with force, intimidated by police, and experience physical assault, while female journalists were more likely to experience sexual harassment and/or violence. In response to interference, male journalists were more likely than females to report in a less controversial manner, abandon sensitive, critical stories and be selective about what items to report. Veteran journalists with over ten years of work experience were more likely to report physical assault, threats with force and intimidation by police during the reference period than their less-experienced colleagues.

A study comprising almost 1500 Swedish journalists (Löfgren Nilsson and Örnebring 2016) found that 80 percent of male journalists and 72 percent of female journalists had received abusive comments, and 31 percent had received threats during the 12-month reference period. No gender-based differences were found in the frequency of the threats. Journalists working in tabloids and metropolitan morning dailies were more likely to receive threats, with 71 percent of journalists working in the former and 48 percent in the latter having received at least one threat during the reference period. Freelancers and journalists working in different kinds of magazines were less likely to receive threats, with 20 percent having been threatened. The same patterns applied to abusive comments. The occupational position was a significant factor, as columnists and op-ed writers, followed closely by managers, reported more threats and abusive comments. The results indicate that high-profile journalists with increased visibility are more likely to be verbally abused and threatened.

Studying Norwegian journalists, Landsverk-Hagen (2015) found that close to half (43 percent of men and 44 percent of women) of the 1341 studied journalists reported having experienced online harassment, defamation or verbal abuse, and a quarter (27 percent of men and 23 percent of women) had received threats during the past five years. Regarding self-censorship and chilling effects, 44 percent of Norwegian journalists answered that harassment and threats do not affect their journalism in any way, while 20 percent reported that they would think twice before publishing because of harassment, and 14 percent reported they had felt directly or indirectly silenced by reactions from the public.

Studying contemporary journalism in US and Germany, Revers (2017) observed journalists using resistance tactic conceptualized as editorial defense shield against external pressure. Journalists with strong confidence in their editors’ and organization’s professional integrity and support were able to utilize this as an asset when faced with external interference. When threatened by external actors, editors sprung to their defense and bore the brunt of the consequences, shielding journalists working for them from external pressure. Similar practice in the context of Finnish journalism has been previously discussed by Kuutti (1995, 246).

As the previous research on the relations between individual- and organizational-level factors and external interference is inconclusive, this article seeks to provide new empirical evidence strengthening the understanding of these connections and their impact on the working conditions of journalists.

Methodology

The detailed account of our study design, comprehensive breakdown of the survey sample, complete survey questionnaire, and data that support the findings are openly available in the institutional repository of Tampere University (Hiltunen and Suuronen 2019). In this article, we examine differences in experiences of external interference in four thematic categories: non-physical interference (questions 10–20 in the questionnaire), physical interference (questions 21–26), institutional interference (questions 27–29), and economic interference (questions 30–33) (ibid., 83–84). Journalists’ perceptions of how external interference may affect their work or the journalistic profession in general (questions 42–56), were measured using a standard Likert scale (ibid., 86).

The survey was conducted using an anonymous online self-report questionnaire in Finnish, consisting of 56 closed multiple-choice questions and four optional fields for open-ended comments. The survey method was chosen because incidents of low-intensity interference and harassment, in particular, often go unreported and unrecorded, and self-report surveys are useful tools to uncover previously undocumented personal experiences and perceptions (Clark and Grech 2017, 26).

The data were collected between 13 and 26 March 2017. The reference period in the questionnaire was the past three years (2014–2017), which was considered sufficiently long to provide a consistent overview while mitigating possible reliability issues. As the Union of Journalists in Finland and three Finnish editors’ associations technically administered the survey, the definition of journalist used in this study was consistent with the union’s or editors’ association membership requirements.¹ Pöyhtäri, Ahva, and Väliverronen (2014) estimated that working journalists in Finland number roughly 8000, indicating that the total of 8275 survey recipients included close to the whole national population.

A total of 875 responses were received, representing a participation rate of 10.6 percent. Moreover, 353 respondents (40 percent) provided one or more text comments to supplement their responses. The response rate was objectively low, even for a web-based survey (cf. Cook, Heath, and Thompson 2000), which should be kept in mind when examining the findings. However, Krosnick (1999) states that surveys with low response rates can yield useful data, highlighting the importance of sample composition.

The representative sample constructed for the Worlds of Journalism (WJS) study (Pöyhtäri, Ahva, and Väliverronen 2014) and the Union of Journalists in Finland membership statistics were used as reference points to assess the survey sample. This comparison indicated that the survey sample matched the proportions of the estimated target population reasonably well, with regard to factors like gender, age, occupational position, employment type, and medium used for reporting. The comparison is illustrated in Table 1.

Table 1. Survey sample in comparison.

	Survey sample	WJS sample (Finland)	Union of Journalists in Finland members^a
Gender distribution (Female/Male)	57/43%	55/45%	57/43%
Dominant age group (36–55 years)	54%	61%	55%
Salaried employees	81%	82%	81%
Freelancers or entrepreneurs	15%	17%	18%
Working for newspapers or magazines	65%	69%	67%
Working for broadcasting (including public broadcasting)	23%	23%	32%
Position as reporter, special reporter, or visual journalist	68%	72%	76%
Managerial position (Managing editor, producer, editor-in-chief)	28%	24%	15%^b

^aBased on the detailed membership statistics from 2015. To determine the age structure, employment type, medium used for reporting and occupational position, a limited sample was used omitting members (e.g., technical personnel, etc.) whose professional description was deemed not relevant to the scope of the study.

^bJournalists who exercise significant power over terms of employment cannot be accepted as union members, which largely explains the low share of journalists with managerial positions in union membership.

While the proportional similarities do not make the sample representative of the population of Finnish journalists, they increase confidence in that no one group dominated the sample or skewed the results significantly. The sample contains substantial internal variety, including a significant number of responses from typically hard-to-reach groups, such as editors-in-chief. This diversity supports the research aim of exploring variations in experiences and perceived implications based on different individual- and organizational-level factors.

Nonetheless, the survey sample has several limitations due to the non-probability, self-selection sampling method, and the low overall response rate. All surveys are prone to biases derived from issue salience and nonresponse, which can lead to more input from respondents with a personal connection to, or extreme opinions on, the subject (Cook, Heath, and Thompson 2000). The cover letter explicitly encouraged participation, even if the recipient had no personal experience of any external interference, to minimize such effects. It is still possible that the amount of external interference experienced by journalists in the survey sample might be higher than in the whole population, reducing the ability to generalize from the sample to the whole population (Sivo et al. 2006).

Data collected through self-report surveys are always prone to perceptual bias and individual differences when interpreting questions and deciding what to report. To provide more consistent data and to combat biases, the survey questions regarding experiences of external interference included descriptions of methods of interference (Hiltunen and Suuronen 2019, 80–87). These questions were developed and formulated based on the pre-survey interviews, consultations and pilot test of the survey (see ibid. 8–9). We are confident that this helped to alleviate the problems related to individual differences in interpretation and the possible ambiguity regarding to the concept, as the survey questionnaire provided the respondents clearly defined framework of methods considered external interference in the context of this study.

Additionally, it should be noted that the frequency of certain elements of journalistic work logically increases the likelihood of journalists encountering certain types of interference. For example, a journalist regularly reporting from the field is more likely to experience physical interference than one that rarely leaves her or his desk. This was not controlled in the study. To improve the internal validity of the survey respondents were given the option to answer “Don’t know/No opinion” when the question was not applicable or relevant to their work as a journalist.

Because of these limitations and the lack of previous empirical research on the subject, the findings should be considered as indicative and illustrative rather than representative and definitive.

To identify differences in reported experiences of external interference based on individual- (age or gender) and organizational- (employment type, medium used for reporting or occupational position) level factors, we conducted extensive cross-tabulation with chi-square tests of the complete survey data. For the purposes of this analysis, responses of “once in six months”, “once in three months,” “once a month” and “once a week or more frequently” were combined to form a category of “every six months or more often” and responses of “Don’t know / No opinion” were removed. The perceived implications of external interference measured using standard Likert scale were analyzed using non-parametric versions of the Student's t-test and Analysis of variance (ANOVA) tests due to non-normally distributed data and a lack of appropriate interval or ratio scale variables. Mann–Whitney U test was used to identify differences between two groups (gender) and Kruskal–Wallis H test to identify differences between groups of three or more (age, employment type, medium used for reporting, and occupational position). The most common critical value of p < 0.05 was used (with a 95 percent confidence level) as a statistical significance threshold for all tests.

Findings

The findings are presented in two sections: (1) experiences of external interference, and (2) perceived implications of external interference. The first section, exploring experiences, is further divided into four thematic subsections: (1) non-physical interference (2) physical interference, (3) institutional interference, and (4) economic interference. Selected tables are included in this article to demonstrate key findings of the analysis. As the percentages in the tables are rounded to the nearest whole, the total may not always add up to 100 percent. Direct extracts from open survey comments translated by the authors are used to illustrate how respondents reflected on the survey themes.

Experiences of External Interference

Non-physical Interference

Non-physical methods of external interference included verbal pressure and abuse, intimidation, defamation, and other non-physical methods of interference directed toward journalists. Respondents most commonly reported experiences of non-physical interference.

Statistically significant difference between age groups in non-physical interference can be observed on organized feedback campaigns. However, these findings do not suggest straightforward connection between age and this method of interference, as the second-youngest age group of 36–45 years had the largest percentage of respondents having experienced organized feedback campaigns (30 percent had experienced this at least once during observation period), but respondents in the two oldest age groups (46–55 and 56 years or over) reported experiencing them most frequently (9 percent of respondents in these groups experienced them every six months or more frequently).

No statistically significant differences were observed between genders in methods or frequency of non-physical interference. Staff journalists reported more incidences where their editors or employers had been contacted and pressured than freelancers and entrepreneurs. In addition, staff journalists experienced more organized feedback campaigns and mediated verbal abuse compared to freelancers and entrepreneurs. These differences are likely explained with staff journalists having closer ties to their working community and their superiors and more fixed and prominent position in particular media outlet, which typically increases their visibility and accessibility, both of which are indicated as risk factors for receiving abusive comments (Löfgren Nilsson and Örnebring 2016, 884).

Journalists in national and regional newspapers generally experience highest levels of non-physical interference while journalists in magazines experience them less than others. These results are presented in Table 2. When it comes to occupational position, editors-in-chief and special reporters reported highest levels of non-physical interference. These are illustrated in Table 3.

Table 2. Key differences in non-physical forms of external interference (medium used for reporting).

Question	Medium used for reporting	Never	Once a year or less	Every six months or more often	Pearson's Chi-squared
Contacting and pressuring the editor, managing editor, or owner of a media outlet	Magazine (n = 142)	51%	41%	8%	χ² = 33,592 p < 0,001 df = 8 V = 0,153
	Regional newspaper (n = 134)	31%	38%	31%
	Local or semi-local newspaper (n = 182)	41%	40%	19%
	National newspaper (n = 84)	40%	33%	26%
	YLE (n = 177)	51%	31%	18%
Threats of negative occupational consequences (e.g., loss of work or journalistic credibility, hampering of future work)	Magazine (n = 147)	82%	14%	5%	χ² = 30,707 p < 0,001 df = 8 V = 0,143
	Regional newspaper (n = 138)	58%	30%	12%
	Local or semi-local newspaper (n = 189)	70%	26%	4%
	National newspaper (n = 88)	61%	25%	14%
	YLE (n = 184)	72%	19%	9%
Face-to-face verbal abuse (e.g., insults, name-calling, and other verbal expressions of hate)	Magazine (n = 148)	77%	19%	4%	χ² = 25,783 p = 0,001 df = 8 V = 0.131
	Regional newspaper (n = 142)	56%	30%	13%
	Local or semi-local newspaper (n = 189)	54%	38%	8%
	National newspaper (n = 89)	63%	30%	7%
	YLE (n = 186)	62%	29%	9%
Mediated verbal abuse (e.g., insults, name-calling, or other verbal expressions of hate through phone calls, letters, email, online comments, social media, and websites)	Magazine (n = 148)	55%	30%	15%	χ² = 59,834 p < 0,001 df = 8 V = 0,2
	Regional newspaper (n = 139)	25%	34%	41%
	Local or semi-local newspaper (n = 188)	38%	37%	26%
	National newspaper (n = 89)	22%	25%	53%
	YLE (n = 185)	39%	30%	31%
Systematic or unusually large volumes of feedback (e.g., organized feedback campaigns)	Magazine (n = 148)	81%	15%	4%	χ² = 45,841 p < 0,001 df = 8 V = 0,176
	Regional newspaper (n = 139)	71%	20%	9%
	Local or semi-local newspaper (n = 188)	84%	13%	3%
	National newspaper (n = 86)	56%	21%	23%
	YLE (n = 183)	70%	18%	12%

Notes: The effect sizes are reported as Cramér's V calculations (V).

Table 3. Key differences in non-physical forms of external interference (occupational position).

Question	Occupational position	Never	Once a year or less	Every six months or more often	Pearson's Chi-squared
Contacting and pressuring the editor, managing editor, or owner of a media outlet	Reporter (n = 466)	45%	39%	16%	χ² = 14.042 p = 0.029 df = 6 V =0,094
	Special reporter (n = 92)	39%	35%	26%
	Manager (n = 153)	42%	40%	18%
	Editor-in-chief (n = 86)	31%	40%	29%
Threats of negative occupational consequences (e.g., loss of work or journalistic credibility, hampering of future work)	Reporter (n = 488)	70%	24%	6%	χ² = 18,752 p = 0,005 df = 6 V = 0,106
	Special reporter (n = 97)	67%	22%	11%
	Manager (n = 158)	71%	23%	6%
	Editor-in-chief (n = 84)	57%	25%	18%
Face-to-face verbal abuse (e.g., insults, name-calling, and other verbal expressions of hate)	Reporter (n = 489)	64%	29%	7%	χ² = 25,003 p < 0,001 df = 6 V = 0,122
	Special reporter (n = 100)	66%	24%	10%
	Manager (n = 159)	67%	27%	6%
	Editor-in-chief (n = 86)	41%	41%	19%
Mediated verbal abuse (e.g., insults, name-calling, or other verbal expressions of hate through phone calls, letters, email, online comments, social media, and websites)	Reporter (n = 484)	40%	34%	26%	χ² = 20,371 p = 0,002 df = 6 V = 0.111
	Special reporter (n = 100)	38%	27%	35%
	Manager (n = 158)	42%	29%	29%
	Editor-in-chief (n = 86)	23%	29%	48%
Systematic or unusually large volumes of feedback (e.g., organized feedback campaigns)	Reporter (n = 483)	78%	15%	7%	χ² =20,729 p = 0,002 df = 6 V = 0,112
	Special reporter (n = 99)	72%	14%	14%
	Manager (n = 156)	75%	17%	8%
	Editor-in-chief (n = 84)	58%	31%	11%

Notes: The effect sizes are reported as Cramér's V calculations (V).

Differences in Table 3 highlight the significance of occupational position. Through their occupational role, editors-in-chief are expected to provide an editorial defense shield (Revers 2017, 162–165) that protects reporters working for them from external interference and therefore must often get involved if other journalists in the media outlet encounter interference (see also Kuutti 1995, 246). As public figures and wielders of the highest journalistic authority in media outlets, editors-in-chief represent their media through their person in public. They are typically more visible than other journalists, suggesting a higher risk of verbal abuse and intimidation.

Additionally, special reporters were more likely to experience non-physical interference than reporters. Typically, special reporters focus more often than reporters on investigative journalism, controversial topics, or stories that might cast some actors in a negative light. These are indicated as risk factors for occupational intimidation (Kuutti 1995; Parker 2015). Also, special reporters often have increased personal visibility due to their status. These factors may explain the elevated levels of non-physical interference among special reporters.

Physical Interference

In the survey, forms of physical interference included violence, monitoring, or physically interfering with the performance of journalistic work and breaking or tampering with work-related equipment.

Compared to other types of interference, forms of physical interference were reported the least as the vast majority of respondents reported not having experienced such interference at all. Therefore, no differences could be observed between different groups or the low number of observations rendered statistical analyses unreliable.

The only consistent factor producing multiple differences in physical interference was the medium used for reporting. Journalists working in national and regional newspapers reported generally more experiences of unwarranted denial of entry or removal from the scene, monitoring and following while conducting journalistic work, disruptions of work, and physical violence. However, we were not able to reliably confirm if these observations are statistically significant due to the low number of observations.

Institutional Interference

Institutional interference included methods that operate through the legal system or media self-regulation.

Statistically significant difference between age groups was found in experiences of having been threatened with or sued for damages or compensation. More respondents in the second-oldest age group (46–55 years) reported having experienced this compared to other age groups.

No statistically significant differences based on gender were observed in institutional interference. Staff journalists experienced institutional interference more than freelancers and entrepreneurs did. These differences are likely due to their more fixed employee position, as methods of institutional interference typically target media outlets rather than individual journalists.

Reporting media outlet had a significant effect on institutional interference, with journalists working in national and regional papers generally experiencing it the most and journalists working in magazines the least.

The position in the occupational hierarchy of media outlet was strongly connected to experiences of institutional interference. As representatives of media outlets, editors-in-chief experienced more institutional interference than other journalists did. Due to their occupational role, institutional interference directed at journalists working for them has typically considered editors’ responsibility, providing possible explanation for these findings.

Economic Interference

Economic interference included attempts to influence journalism with positive or negative economic sanctions.

Age seems to be connected with receiving offers of economically valuable benefits or gifts, as the youngest age group of under 36 years had the largest share of respondents (34 percent) having experienced this and most respondents (16 percent) experiencing this every six months or more often. No statistically significant differences were observed between genders in economic interference. Freelancers were less likely to be threatened with the loss of subscribers or audiences than staff journalists were, likely because of their looser affiliation with specific media outlets.

Journalists working in regional and local or semi-local newspapers had the biggest share of respondents having been threatened with loss or subscribers or sponsors and advertizers. Also, respondents from national newspapers reported higher numbers compared to journalists working in magazines and in Finland’s national public broadcasting company Yleisradio Oy (YLE). National newspapers had the highest share of journalists that reported having been offered economically valuable benefits or gifts, while YLE had the lowest share of journalists reporting this.

Editors-in-chief and to lesser extent managing editors reported more threats of negative economic sanctions compared to reporters and special reporters. Through their professional role, editors-in-chief typically have responsibilities related to financial matters of media outlets, which makes targeting them with this type of economic interference understandable. When it comes to offers of economically valuable benefits or gifts, more editors-in-chief compared to other journalists reported experiencing these. However, reporters and visual journalists encountered these offers most frequently (with 11 percent experiencing them every six months or more often).

Perceived Implications of External Interference

When perceived implications of external interference were analyzed by age groups, several questions produced statistically significant differences. The older age groups were generally more confident that external interference does not affect their journalistic work and reported less passive self-censorship and less concern about the effects of external interference on the credibility of journalism in Finland. Considering that differences between age groups in prevalence of external interference were very limited and sectional, the results indicate that experienced journalists have stronger confidence on their professional integrity and that they perceive external interference less a threat than their younger colleagues.

While no significant differences were observed in prevalence or methods of external interference between genders, female journalists held consistently more negative perceived implications of interference. These are shown in Table 4.

Table 4. Key differences in perceived implications of external interference (gender).

Question	Gender	Strongly disagree	Somewhat disagree	Neither agree nor disagree	Somewhat agree	Strongly agree	Mann-Whitney U
The amount of external interference I encounter in my work has increased in the last three years.	Male (n = 346)	25%	25%	22%	21%	8%	U = 65914
							p < 0.001
	Female (n = 449)	20%	18%	18%	33%	10%	z = −3.76
							r = 0.133
I have consciously developed methods and strategies to ward off external interference.	Male (n = 338)	18%	16%	24%	30%	12%	U = 69717
							p = 0.01
	Female (n = 460)	14%	12%	22%	42%	10%	z = −2.585
							r = 0.092
External interference does not affect my journalistic work in any way.	Male (n = 361)	4%	20%	10%	30%	36%	U = 101375
							p < 0.001
	Female (n = 478)	4%	27%	16%	32%	21%	z = 4.499
							r = 0.155
I am confident that my editor or employer will support me from external interference.	Male (n = 365)	5%	8%	3%	32%	52%	U = 98311
							p = 0.004
	Female (n = 487)	7%	11%	7%	32%	44%	z = 2.858
							r = 0.098
External interference increases the mental strain of my work.	Male (n = 357)	21%	18%	17%	31%	13%	U = 74549
							p = 0.002
	Female (n = 475)	15%	17%	15%	35%	19%	z = −3.071
							r = 0.106
Advertizers and sponsors are able to influence the journalism that my media outlet produces.	Male (n = 347)	28%	30%	10%	24%	8%	U = 71520
							p = 0.011
	Female (n = 459)	24%	23%	14%	27%	12%	z = −2.551
							r = 0.09
I prefer not to report about certain topics or present certain viewpoints due to external interference.	Male (n = 360)	51%	24%	9%	11%	4%	U = 78581
							p = 0.009
	Female (n = 484)	42%	26%	13%	15%	4%	z = −2.595
							r = 0.089
I have altered or removed something from my journalism pieces as I feared external interference.	Male (n = 366)	60%	22%	6%	10%	2%	U = 76601
							p < 0.001
	Female (n = 485)	46%	27%	11%	14%	2%	z = −3.732
							r = 0.128
My media outlet does not hand over control of journalistic decisions to external actors under any circumstances.	Male (n = 350)	5%	11%	7%	21%	55%	U = 90119
							p = 0.001
	Female (n = 458)	4%	14%	10%	31%	41%	z = 3.24
							r = 0.114
I am worried about the effects of external interference on the credibility of journalism in Finland.	Male (n = 366)	8%	19%	8%	42%	23%	U = 76892
							p < 0.001
	Female (n = 482)	4%	12%	9%	46%	29%	z = −3.3947
							r = 0.117

Notes: The effect sizes are calculated as correlation coefficients (r).

Multiple explanations can be offered for these findings. With history as a male-dominated industry, the journalistic profession has been associated with masculine virtues, requiring thick skin, toughness, and refusal to submit to external authority (Ross and de Bruin 2004). Male journalists may be especially reluctant to report negative emotions and effects caused by external interference or admit to self-censorship (see Binns 2017), considering this professional ethos.

In addition, external interference experienced by female journalists may have qualitative differences, not quantitatively accounted for by this study. Survey comments support this interpretation, especially regarding online verbal abuse and threats:

People who send threatening messages […] can be described as active online racists and misogynists.

(Special reporter, Female)

[…] At the worst, there were online discussions about raping me.

(Reporter, Female)

Threats are usually verbal. Insinuations that your life expectancy or health will bear the consequences or calling me a whore and spouting inappropriate tits-pussy-ass-slurs.

(Managing editor, Female)

Studies have identified gendered differences with female journalists more often targeted with sexist abuse and threats of sexualized violence, indicating the pervasiveness of sexist and misogynist discourse (Löfgren Nilsson and Örnebring 2016; Adams 2018; Chen et al. 2018). These qualitative differences can partially explain why Binns (2017) found that female journalists targeted with online abuse reported stronger emotional reactions, feeling upset, intimidated, or angry more often. Obermaier, Hofbauer, and Reinemann (2018, 515) found that German female journalists exposed to hate speech tended to use both emotionally focused, and problem focused coping strategies more than their male colleagues did, pointing out to an increased need to address hate speech on a psychological and social level.

Employment type seems to have an effect on perceived implications of interference, as journalists working as freelancers or entrepreneurs expressed consistently less confidence in their superiors and their media outlets ability to resist external interference. One possible explanation for this is that their looser connection with journalistic organizations and specific media outlets does not provide similar conditions for building reciprocal trust as staff journalists have.

Editors-in-chief and journalists in managing positions expressed high trust on their superior’s support against external interference and had strong confidence in their media outlets ability to ward off interference and its effects to journalism. Differences concerning this are shown in Table 5.

Table 5. Key differences in perceived implications of external interference (occupational position).

Question	Occupational position		Strongly disagree	Somewhat disagree	Neither agree or disagree	Somewhat agree	Strongly agree	Kruskal-Wallis H
I have consciously developed methods and strategies to ward off external interference.	Reporter	(n = 449)	16%	14%	21%	39%	10%	H = 10,84 p = 0,013 df = 3 E^2  = 0,012
	Special reporter	(n = 90)	14%	18%	23%	31%	13%
	Manager	(n = 149)	16%	14%	28%	34%	8%
	Editor-in-chief	(n = 82)	11%	5%	23%	40%	21%
I am confident that my editor or employer will support me from external interference.	Reporter	(n = 480)	6%	11%	5%	36%	42%	H = 14,124 p = 0,003 df = 3 E^2  = 0,016
	Special reporter	(n = 100)	9%	10%	6%	32%	43%
	Manager	(n = 154)	6%	8%	3%	23%	60%
	Editor-in-chief	(n = 85)	6%	6%	11%	20%	58%
External interference increases the mental strain of my work.	Reporter	(n = 469)	17%	17%	17%	34%	15%	H = 9,6505 p = 0,022 df = 3 E^2  = 0,011
	Special reporter	(n = 94)	15%	20%	16%	24%	24%
	Manager	(n = 152)	21%	20%	14%	31%	13%
	Editor-in-chief	(n = 85)	9%	13%	12%	49%	16%
The audience has a right to know about all incidents of external interference; therefore, they should always be made public.	Reporter	(n = 460)	3%	21%	17%	38%	22%	H = 14,631 p = 0,002 df = 3 E^2  = 0,017
	Special reporter	(n = 98)	4%	24%	11%	36%	24%
	Manager	(n = 147)	4%	23%	20%	30%	23%
	Editor-in-chief	(n = 83)	6%	34%	23%	27%	11%
Advertizers and sponsors are able to influence the journalism that my media outlet produces.	Reporter	(n = 447)	23%	24%	13%	28%	11%	H = 8,304 p = 0,04 df = 3 E^2  = 0,01
	Special reporter	(n = 90)	28%	26%	10%	27%	10%
	Manager	(n = 153)	30%	32%	8%	20%	9%
	Editor-in-chief	(n = 84)	30%	25%	18%	19%	8%
Politicians are able to influence the journalism that my media outlet produces.	Reporter	(n = 454)	26%	35%	11%	22%	6%	H = 30,345 p < 0,001 df = 3 E^2  = 0,035
	Special reporter	(n = 93)	29%	27%	8%	28%	9%
	Manager	(n = 149)	39%	34%	7%	15%	5%
	Editor-in-chief	(n = 85)	52%	31%	5%	9%	4%
My managing editor, editor, or supervisor gives in to external interference more easily than I do.	Reporter	(n = 400)	22%	18%	20%	30%	10%	H = 16,458 p = 0,001 df = 3 E^2  = 0,019
	Special reporter	(n = 81)	16%	25%	16%	28%	15%
	Manager	(n = 142)	32%	20%	15%	22%	10%
	Editor-in-chief	(n = 72)	43%	12%	21%	17%	7%
I prefer not to report about certain topics or present certain viewpoints due to external interference.	Reporter	(n = 472)	42%	24%	14%	15%	5%	H = 15,338 p = 0,002 df = 3 E^2  = 0,018
	Special reporter	(n = 99)	54%	28%	5%	12%	1%
	Manager	(n = 155)	50%	24%	12%	11%	3%
	Editor-in-chief	(n = 83)	52%	35%	4%	7%	2%
I have altered or removed something from my journalism pieces, as I feared external interference.	Reporter	(n = 475)	48%	25%	11%	13%	3%	H = 7,859 p = 0,049 df = 3 E^2  = 0,009
	Special reporter	(n = 99)	60%	21%	5%	13%	1%
	Manager	(n = 156)	57%	24%	7%	10%	1%
	Editor-in-chief	(n = 86)	49%	34%	9%	8%	0%
My media outlet does not hand over control of journalistic decisions to external actors under any circumstances.	Reporter	(n = 449)	5%	15%	9%	29%	41%	H = 42,321 p < 0,001 df = 3 E^2  = 0,048
	Special reporter	(n = 93)	4%	16%	11%	31%	38%
	Manager	(n = 150)	7%	9%	9%	21%	55%
	Editor-in-chief	(n = 85)	0%	2%	5%	16%	76%

Notes: The effect sizes are calculated as epsilon-squared estimates (E²).

These observations can be reflected against occupational ideals of journalists with managerial responsibilities. As there exists strong professional ideal of editors as safeguards of journalistic autonomy and providers of editorial defense shield, editors-in-chief, managing editors, and producers may be especially hesitant to admit external interference having any effect on them (Revers 2017, 162–165). Additionally, as indicated in previous sections, journalists in managing position may have to deal with external interference more frequently than reporters do and can, therefore, be better accustomed to it.

However, the findings indicate a discontinuity in the confidence level within the organizational command chain. Editors-in-chief expressed strong confidence in their publishers to support them from external interference and managing editors and producers had equally strong confidence on support from editors-in-chief. This strong confidence was not so widely shared by reporters and special reporters. The relationship between editors and reporters can be prone to tension and reporters may, for example, feel that their editors are reluctant or afraid to publish stories that may potentially harm media outlets’ relations with advertizers or other external actors (Kuutti 1995, 248).

Journalists’ perceptions of how external actors can influence journalism in their media outlets are illustrated in Table 6.

Table 6. Key differences in perceived implications of external interference (medium used for reporting).

Question	Medium used for reporting		Strongly disagree	Somewhat disagree	Neither agree or disagree	Somewhat agree	Strongly agree	Kruskal-Wallis H
Advertizers and sponsors are able to influence the journalism that my media outlet produces.	Magazine	(n = 140)	26%	28%	15%	22%	9%	H = 61,882 p < 0,001 df = 4 E^2  = 0,071
	Regional newspaper	(n = 134)	12%	28%	13%	37%	10%
	Local/semi-local newspaper	(n = 186)	15%	28%	13%	30%	13%
	National newspaper	(n = 83)	20%	37%	11%	22%	10%
	YLE	(n = 150)	51%	20%	9%	14%	5%
Politicians are able to influence the journalism that my media outlet produces.	Magazine	(n = 138)	51%	28%	8%	12%	1%	H = 33,104 p < 0,001 df = 4 E^2  = 0,038
	Regional newspaper	(n = 132)	23%	40%	12%	23%	2%
	Local/semi-local newspaper	(n = 186)	30%	35%	10%	22%	4%
	National newspaper	(n = 82)	26%	38%	10%	22%	5%
	YLE	(n = 166)	25%	32%	10%	23%	10%

Notes: The effect sizes are calculated as epsilon-squared estimates (E²).

Journalists working in local, semi-local and regional newspapers felt that advertizers and sponsors could influence the journalism of their media outlet more than their colleagues in other media outlets. Kuutti (1995, 249) states that the status and size of a media outlet can have a significant effect on their ability to resist external interference. Local, semi-local, and regional papers are often dependent on a smaller number of companies for advertising than national journalism outlets, and therefore, can be more vulnerable to economic pressure. Several survey comments support this interpretation:

Especially in a local newspaper, managing editors and editors-in-chief are under great pressure from advertisers, and it has effects on journalistic work. Even to the extent, where some advertisers are totally off limits.

(Producer, Female, Multiple employers)

The biggest threat to journalism is inside the newsroom: weak and incoherent editorial leadership, that bows to advertisers’ wishes […] Because of this, some topics, and even some segments of journalism pieces are censored.

(Reporter, Male, Regional newspaper)

In predominantly tax-funded broadcaster YLE, respondents saw their journalism significantly less influenced by advertizers or sponsors. However, the opposite was true when examining how respondents perceived the influence of politicians on journalism. Survey comments below illustrate these concerns:

Preventive censorship has been used, and topics have been blacklisted, which has not happened before. This reflects […] how concern over funding has effected journalism.

(Producer, YLE)

There has been a tendency in YLE news production to avoid news topics that can cause friction regarding the funding. Especially news stories concerning members of the business or political elites have been edited and softened.

(Special reporter, YLE)

An integral part of the management of YLE’s news- and current affairs has been the practice of killing off disconcerting exposés concerning those high in the power structures.

(Reporter, YLE)

Waisbord (2013) conceptualizes the relationship between public broadcasting and political power as essentially problematic while journalism operating under marked conditions has a similar uneasy relationship with market power. Ahva et al. (2017, 607) discovered that this phenomenon can be observed in all the Nordic countries: Journalists working for media in public ownership feel less commercial influence but more political pressure than journalists working for other media.

Conclusions and Discussion

This research indicates that individual-level factors (age or gender) have only marginal relevance on the prevalence of external interference that journalists encounter in their work in Finland. These findings are in line with Parker’s (2015) conclusions that did not establish a connection between occupational intimidation and the demographic characteristics of journalists. Concerning organizational-level factors, the differences between staff journalists and freelancers were limited and sectional. Occupational position and media outlet used for reporting produced significant differences between respondents, suggesting that these two organizational-level factors are most determining when examining the prevalence of external interference in large scale.

The findings concerning the significance of occupational position support the existence of editorial defense shield as practice. Similar to descriptions by Revers (2017, 162) and Kuutti (1995, 246), the survey comments indicated that editors-in-chief have to frequently get involved when reporters working for them encounter external interference. This practice can partly explain why they reported experiencing most types of external interference more than journalists in non-managing positions did. Editors-in-chief and managers also shared high confidence in their superiors and the view that interference does not affect journalism produced by their media outlet. This fits in the occupational perception of editors and managers as safeguards of journalistic autonomy. However, the discontinuity in the confidence levels between journalists with managing responsibilities and reporters and special reporters suggest that this relationship can be tense, and the trust to editorial defense shield might falter at least occasionally (see Kuutti 1995, 246–248; Revers 2017, 163–164).

Similar to Löfgren Nilsson and Örnebring (2016, 884) and Landsverk-Hagen (2015), this study does not support the expectation that female journalists are more exposed to intimidation, harassment, or verbal abuse than their male colleagues are. Nonetheless, female journalists held consistently more negative perceived implications of interference and reported more mental strain and less confidence in their superiors and media outlets. These differences, however, could be better explained by social structures or other aspects since only minor observable differences were made in the actual experiences of external interference. As this example suggests, the relationship between experiences of interference and perceived implications is not linear, but often complex and mediated by a plethora of factors on different levels.

Respondents in national and regional newspapers experienced external interference generally the most, while respondents in magazines encountered it the least. The differences between media outlets can be at least partially attributed to their status and visibility. Newspapers with significant regional or national visibility are typically considered more prestigious and socially influential than smaller newspapers and magazines and therefore, more publicly contested (Kuutti 1995, 249). The increased amounts of mediated verbal abuse, public defamation, and intimidation in these media outlets partly support the interpretation by Löfgren Nilsson and Örnebring (2016) concerning the visibility of journalists as a risk factor. However, YLE seems to be an outlier in both respects, boasting strong status and national and regional visibility, but with respondents experiencing less external interference. This observation might be partly explained by differences in media production, as it is possible that a larger segment of journalists employed by YLE work in less prominent supporting positions and do not directly engage in daily news production.

Several limitations can be pointed out. While this study focuses on the external autonomy of journalism, it provides a very limited perspective on how the material and structural conditions affect journalists’ aspirations for professional autonomy. Therefore, future research should strive to combine these findings with broader analysis of structural conditions of journalism in order to synthesize more nuanced overall picture of how the journalistic autonomy is transforming in the contemporary communication environment.

Excluding the comments, the survey did not distinguish between different sources, contexts, or locales of interference. Survey data based on self-reporting is susceptible to social desirability bias. External interference can be considered a sensitive topic, and journalists might assess its effects consciously or unconsciously to be less pronounced than they materialize in their daily work, due to professional ideals and identity. Additionally, the lack of longitudinal data makes it impossible to identify any possible shifts or changes.

One additional limitation of the study is that we were unable to account for the influence of the topic that the respondent most commonly reports on, due to highly skewed response distributions. While the respondents were asked the most common topic they covered in the questionnaire, over 60 percent of the respondents indicated working with either local (29 percent) or current national affairs (18 percent) or selecting the “Other” category (17 percent). This skewed distribution in favor of general or undefined topics prevented us from making meaningful comparisons.

High degree of journalistic autonomy and relatively low levels of economic pressure have been identified as distinctive features of the Nordic journalistic culture (Ahva et al. 2017, 607). While Finnish journalists have previously perceived themselves autonomous and distant from direct external influences (Pöyhtäri, Ahva, and Väliverronen 2014), our findings demonstrate that this does not indicate the absence of external interference in the context of Finnish journalism. Our analysis provides more nuanced empirical account on how external interference imposes diverse challenges for journalists working in different media outlets and occupational positions. Simultaneously, the findings highlight new pressures and threats often stemming from the audience, like verbal abuse, harassment and public defamation. At least to a degree, these pressures are exacerbated by the contemporary communication environment, where online visibility and presence of journalists have increased manifold. Based on our findings, we echo Löfgren Nilsson and Örnebring (2016, 889) while stating that dimension of external interference and its effects should be acknowledged and taken into account when studying journalistic autonomy also in the context of stable and democratic Western societies. Furthermore, our findings pose a methodological challenge on how to implement observed individual- and organizational-level variation and its combined effects into future studies.

Acknowledgements

The authors would like to thank Associate Professor Jenny Wiik from Malmö University, Assistant Professor Marína Urbániková from Masaryk University and the three anonymous reviewers provided by Journalism Practice for insightful feedback that greatly assisted the research article.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Media Industry Research Foundation of Finland under Grants 201610183 and 201710214, the University of Tampere, and the Union of Journalists in Finland.

Notes

1 “Your work involves essential journalistic features and is professional in nature. […] a significant portion of your earnings derive from such work that has essential journalistic qualities” (The Union of Journalists in Finland). All editors’ associations’ membership criteria included a managing position in a media outlet.