Модуль 1. Политические аспекты масс-медийных продуктов
Government responses to media allegations: comparison between an established and a managed democracy
Gergana Dimova
(United Kingdom, University of Winchester)
The Importance of Studying Government Responses in the Media
Understanding government responses to public allegations is crucial because the executive spin game could pose a serious challenge to democracy (Allern/Sikorski, 2018). Blame avoidance could be a powerful tool to hold the population in check. Mikhail Khodorkovsky, one of Russia’s wealthiest oligarchs and a prisoner for 10 years, told a small audience in the British House of Commons on June 18, 2018 that Putin manages to keep the Russian in check because the president is so talented in spinning the interpretation of events. As an example, Khodorkovsky cited Putin’s blame game after the Malaysia Airlines flight MH17 was downed over Ukraine in 2014. According to Khodorkovsky, the presidential administration first denied that the plane was brought down, then it said that it was brought down but not by a missile; then it said that it was brought down by a missile but that the missile was not Russian; then it said that it was a Russian missile but the Russians did not launch it. It is important to understand that the stakes behind this verbal equilibristic are very high as they at least partially define how power is upheld.
Not only is studying the patterns of verbal executive responses important. Studying them in a comparative perspective, rather than within separate countries, is equally imperative. During the July 2018 Helsinki meeting, president Trump stood next to president Putin, and answered media questions about allegations that Putin had compromising material on Trump. This particular press conference provides a vivid example of the necessity to study executive blame games in relation to each other, and to include non-Western countries as well. In a wide ranging review of the literature on political scandals, Sikorski (Sikorski, 2018) criticises the fact that research stems predominantly from North America and Europe.
From the point of view of democratic research, the study of executive explanations is considered important as blame avoidance and blame acceptance constitute a crucial aspect of democratic accountability (March & Olsen, 1995; Norris, 2000; Strøm, et al 2003). The critical importance of government strategies has also been acknowledged by Aucoin and Heintzman (Aucoin, Heintzman 2000: 49-52) who argue that citizens’ confidence in the government is largely dependent on the way administrators explain and justify their actions. The growing significance of verbal responses has given rise to a body of literature on blame-avoidance strategies (Hood, 2014; Boin, et al 2009; Weaver, 1986) and crisis management (t’Hart, 2011). They boil down to two interrelated research sub-questions, which also constitute the backbone of the present article: (1) what are the different patterns of government responses in the media made by democratic and non-democratic incumbents? and (2) what factors affect these responses?
The first research question- what are the government responses in the media? - has suffered from an abundance of typologies. The multiple coding schemes of government blame avoidance strategies ranges from coding anticipatory versus reactive (Sulitzeanu-Kenan, 2006), presentational versus agency-related responses (Hood, 2016), or strategies related to managing blame versus managing information (McGraw, 1991, Coombs, 2009). These formulations of government responses have existed in parallel but not in conversation with each other. Thus, multiple coding schemes have made comparison across countries or across studies very difficult. In essence, they have obscured the “puzzle” of government responses. In the last two decades, however, the literature has forged a fragile but significant consensus to keep sight of the main inquiry, namely whether the responses that the incumbents provide make them more accountable to the public or not (Hood, 2016). The present article continues this trend of narrowing the criteria according the answerability dimension by reducing government responses to three main types: keep silent, accept blame and deny blame.
The second inquiry - what factors affect executive responses in the media? - has recently been pursued with the increasing awareness that government officials do not formulate their responses in a vacuum. Initial research has underscored the importance of “negativity bias,” a notion that incumbents prefer to keep silent rather than address an allegation (Weaver, 1986). It has also revolved around the question whether the nature of the allegation- policy failure or a moral misconduct- could open avenues for different spinning strategies (Boin, et al, 2010). This early research has been criticised for its implicit linearity, being too static and for ignoring contextual factors (Hinterleiter, 2017). These criticisms have been addressed by an interdisciplinary mix of literatures, such as situational crisis communication, contingency theory, public relations, blame-avoidance (Hood, 2014; Boin, et al, 2009; Weaver, 1986), crisis management (t’Hart, 2011) and even deliberative theory (Habermas, 2006). As the sheer number of theoretical approaches indicates, the contextual factors vary widely. They most generally range from the historical context to the institutional context to the personal context. These three types of factors are expected to shape the way incumbents make a case for themselves in the media.
Both the first and second strands of research on blame avoidance strategies suffer from a serious drawback: with just a few exceptions, they lack sufficient data-driven observations about the actual impact of various factors on the likelihood of government responses (Yang, 2012). The majority of studies limit their scope to theoretical speculations about categorising government responses (Bovens, et al, 1999) or the type of allegations (Mancini, 2018). A small but solid body of research ventures into quantifying and counting government responses (Jacobs & Schillemans, 2016). Even a smaller share of studies quantifies both the government responses and their determinants. Within this eclectic group, the number of causal variables that are quantified is not very encompassing. For example, Hood (Hood, et al, 2007; Hood, et al, 2009) quantify the effect of the level of critical coverage, the position of the accused minister and previous government responses on the current level of blame. In another statistical examination, Dowding and Dewan (Dowding & Dewan, 2005) measure the likelihood that an incumbent would resign given the level of critical coverage, the popularity of the government and current economic conditions.
Taking stock of these advances, the current article advances this vein of research by simultaneously (1) quantifying both the causal factors and government responses; (2) testing the impact of a number of contextual variables, which have never been tested, such as the type of allegation, the position of the accuser and the presence of an investigation; (3) relying on a unique original aggregate cross-country longitudinal data. The database of 692 accusations contains 1,890 articles published between 1995 and 2005 in Germany and Russia. The two main research questions are: how do government responses across an established and a managed democracy vary, and does the type of allegation (corruption or incompetence), the position of the accuser (journalist or opposition) and the presence of an investigation influence the government’s proclivity to deny or accept blame?
The article is organized as follows. The first section summarizes the current knowledge about account-giving in the media. The second part puts forward five hypotheses aimed to explain government responses in an established and a managed democracy. The third section clarifies why Russia and Germany are suitable case studies. The subsequent sections describe the methodology and the process of data collection. The sixth part presents the results. The final section concludes.
Literature Review: Government Account-Giving in the Media.
RQ 1: How Do Governments’ Responses Vary?
The first inquiry in this article compares the tactics that actors in democratic and non-democratic settings deploy in the face of media allegations. Paradoxically, this comparison is hindered by the prolific advances in categorisation made by communication research. As Hood (Hood, 2014: 609) sums up the problem, “The abundance of blame avoidance strategies obscures the main inquiry, namely how the government strategies affect accountability, namely whether they improve answerability and the flow of information.” Therefore, an important yardstick for comparing government responses across regime types is to find the categories that best capture the degree of answerability and the fullness of information provided by government members.
Various authors address the answerability/information criteria differently. Some communication scholars choose to disregard this criterion altogether and focus instead on the timing of the strategies, such as anticipatory versus reactive forms of blame avoidance (Sulitzeanu-Kenan, 2006). Others conflate the answerability and information dimensions of government responses. For example, the strategy of “assessing events” contains an element of information, such as “denial of a problem” and an element of answerability, such as “accusing the accuser” (Bovens, et al, 1999). Weaver’s typology (Weaver, 1986) also combines these two dimensions. His blame avoiding strategies could be achieved through limiting information by “agenda-limitation” and “redefining the issue” or through strategies limiting responsibility, such as “passing the buck” and “finding a scapegoat” (Weaver, 1986: 385).
An evolving body of research has sought to address the problem of answerability and information more precisely. McGraw (McGraw, 1991) draws a firm line between strategies aimed at curtailing answerability, such as excuses, and strategies aimed at limiting information or restating the principle for interpreting information, such as justifications. According to him, “Excuses deny some measure of responsibility for what is admittedly an offensive act. Justifications deny some measure of offensiveness in an act for which the individual admits responsibility” (McGraw 1991: 1136). Coombs’s typology of response strategies has similarly evolved to increasingly target the differentiation between responsibility and information. In his earlier writings, Coombs’ (Coombs, 1998: 170) categories tend to conflate responsibility and information. For example, “primary response strategies” include informational dimensions, such as “deny crisis” and answerability dimensions, such as “attack the accuser” and “find a scapegoat.”
In his more recent writings, Coombs (Coombs, 2010) clearly distinguishes between “managing information,” which is clearly an information-related strategy and “managing meaning,” which has to do with answerability as it includes “efforts to influence how people perceive... the organization involved in the crisis.” Hood (Hood, 2010) also devises a taxonomy, which gets at the above distinction. The diversionary tactics are geared to limiting information as they include “changing the subject,” while the “mitigating-perceptions-of-responsibility” approach is clearly targeted at responsibility. Importantly, Hood also includes the category of non-engagement.
The article rearranges the set of choices not according to the type of response, i.e. whether it is related to presentation, problem or responsibility, but according to the tendency to give evasive versus explanatory responses. Evasive responses include problem denial or responsibility denial. Explanatory responses encompass problem admission or responsibility admission. Non-engagement responses include passive presentational strategies and other types of passive strategies. In essence, the three categories suggested here present a rearrangement of the four categories introduced by Hood et al (Hood, 2016) with the specific goal of addressing the crucial issue of answerability and information. This methodological point is related to the theoretical point made above, which argues that the literature profiles government responses too narrowly, and that the multitude of narrow criteria for gauging responses is problematic because it hinders comparison across various categories. It obscures the main inquiry, mainly whether the government is accountable or not.
Table 1.
Reducing the Variety of Types of Government Responses
Explanatory strategies “improve answerability and the flow of information.” Specifically, they include the strategies of “defend position” and “confess, apologize.” These strategies generate additional information to the public, which levels the information asymmetries between the electorate and the government. Explanatory responses provide a fruitful ground for a public discussion of attributions of political responsibility, because they increase the level of information about apparent wrongdoings and provide explanatory responses.
Evasive strategies do not “improve answerability and the flow of information.” Their main purpose is to deflect attention from the allegation. Some evasive responses restrict the access to account-holding by de-legitimizing the accusers. When the government is constantly evasive, it becomes much harder for the public to attribute political blame because the details of the decision making process and the supposed wrong-doings remain unknown.
The first hypothesis speculates how government responses to media allegations vary. It is important to compare government responses across regime types. There is only speculation, but no empirical evidence, that governments in democratic countries are more likely to be verbally accountable in response to media allegations than governments in less democratic countries. This expectation is not hard to justify. Democratic incumbents in general are thought to believe in the value of public accountability and feel that their responses may impact their chances for re-election. Yet, we do not know whether this is the case for sure, what the difference in the proclivity to engage in explanations is, and to what extent avoiding giving explanations is done by failing to respond or by actively denying the charges.
The article offers a comparative quantitative approach which complements communicative research focusing on single-case studies (Yang, 2014: 155). As Schillemans (2013, 19-20) notes in his review of accountability studies, only 17% of such studies in major journals use quantitative data. Yang (Yang, 2012: 259; Yang, 2014: 155) laments the fact that most research focuses on single-case studies and that “there is a lack of methodological diversity and a dearth of quantitative designs.” In addition, there is a bias towards examining countries from, and mainly discussing, North America and Europe, with only 4% of studies originating in Asia, for example (Sikorski, 2018). Furthermore, political scandals are usually examined by political science and psychology journals, but rarely in communication outlets (Sikorski, 2018). Of course, some scholars venture beyond the usual confines. For example, Toepfl (Toepfl, 2011: 1302) states that “over the past two decades, most scholars have based their studies on ... [the] basic assumption that scandals can only occur in liberal democracies. Thus, this study wants to raise the question: What specific patterns of scandal communication can be observed in the semi-authoritarian political environment of contemporary Russia?”
The multiplicity of organising criteria and the lack of data prevent a clear formulation of the “puzzle”: how does the pattern of government responses vary across regime types? In short, the explanatory variable is unknown, although establishing it should be the first step in studying the spin game across the globe. To rectify this problem, we propose the following hypothesis:
H1: Government officials are more likely to give explanatory responses to media allegations in an established rather than in a managed democracy.
RQ 2: What Factors Impact Governments’ Verbal Responses?
The second inquiry pertains to the factors impacting the government responses in the aftermath of media allegations. The theoretical understanding in this regard has steadily evolved to consider more context-specific factors.
Primary Causal Variables Affecting Government Responses: “Volume of allegations” and “type of allegation”
Two primary causal variables affecting blame avoidance strategies are the ‘volume’ (Weaver, 1986) or ‘negativity’ of news coverage (Hood, 2010, Sulitzeanu-Kenan, 2010) and the “nature of the allegation.” The notion of negativity bias constitutes a central premise posits that the government’s chief priority is to avoid blame rather than claim credit. The rationale is that the public rewards success four times less than it punishes wrongdoing (Weaver, 1986).
The importance of the “nature of the allegation” as a factor influencing government responses has increased. In earlier writings, the nature of the allegation is deemed less consequential because incumbents’ strategies were thought to be entirely the verbal making of the incumbents themselves. Edelman (Edelman, 1988: 31) argued that “a policy failure, like all news developments, is a creation of the language used to depict it” and Boin et al (Boin, et al, 2010) view crisis responses are “framing contests.” In contrast, Brändström and Kuipers (Brändström & Kuipers, 2003) take a more evidence-based approach to allegations and suggest that the choice of strategies depends on the failures themselves. More specifically, government strategies are affected by considerations whether the fiasco violates crucial norms, relates to high-level officials, and whether the failure is on systematic or an individual level. Bolstering the importance of the nature of the crisis further, Bovens (Bovens, et al, 1999: 145) argues that the “observability” of the alleged wrongdoing is important. The authors distinguish between policy fiascos, such as natural disasters, which are observed, and political fiascos, which are construed. Furthermore, policy fiascos put the blame on the government as it was not misfortune but mismanagement that has brought on the bad consequences. Further research on the type of allegation distinguishes between the subjective and objective dimensions of the crisis, and crises with endogenous versus exogenous causes (Boin et al, 2009: 100). McConnell (McConnell, 2003) differentiates between the level of secrecy, the threat, the time horizons and the imminence of the crisis (sudden, creeping, chronic crisis). Djerf-Pierre et al (Djerf-Pierre, et al, 2013) discern between a moral scandal and a policy failure.
Context-specific Variables Affecting Government Responses
Early research on government responses has been criticised for being “widely acknowledged” to be “scattered and unconcentrated, and that, for the most part, it neglects both contextual factors and comparative research” (Hinterleitner & Sager, 2015: 140). More specifically, communication scholarship has been faulted for disregarding institutional constraints and interactional factors. It has also come under attack for following a linear logic of sequence. A few critics have lamented the absence of comparative research, which only exacerbates the above shortcomings. These criticisms have been addressed fully or partially by a second generation of research, which has fully embraced the value of contextualising the causal factors. This second wave of research has evolved by contextualising the actor-specific and environment-specific factors exerting an influence on government responses (Hinterleitner & Sager, 2015; Hinterleiter, 2017).
Zooming in on the contextual factors, we note that actor-specific factors take into account the leadership style of the actors (Boin, 2010) and the ordering of the agents’ preferences (Sulitzeanu-Kenan & Hood, 2005). Game –theoretical insights explain how the opportunity sets of government ministers weigh in the trade-off between being responsive in dismissing an alleged minister versus the benefits of appearing loyal to the prime-minister’s appointees (Belnski, et al: 2012). Some theorists have shifted attention from the government to the accuser. Hood (Hood, 2009) incorporates the identity of the accuser by pointing out that if the incumbent is alleged by the government’s own party or by a member of the government, then the allegation is perceived as a less legitimate and credible threat. In this vein of research, Dimova (Dimova, 2012; Dimova, 2013) discusses the costs and benefits of making accusations for alleged incumbents. The benefits could be symbolic or monetary, while the benefits are shaped by a myriad of factors, such as informal media practices, defamation laws, strength of journalistic associations, and relations with the Secret Services, which possess compromising materials.
Research on context-specific (rather than actor-specific) drivers of government responses has proliferated in several directions. Earlier writings highlight the impact of causality, responsibility or blameworthiness (Shaver & Drown, 1986). Further context-specific factors pertain to the networks within which actors are situated (Moynihan, 2012) and the institutional advantages some actors enjoy in dispersing blame (Brändström & Kuipers, 2003). Habermas's (Habermas, 2006) theory of communicative action adds to that research agenda by underscoring the interactive character of crisis responses. Situational crisis management contributes to context- specific factors by evaluating intensifying factors, such as a history of similar past crises or a negative prior reputation, because they are likely to “intensify” attributions of the organization’s crisis responsibility (Coombs, 2007). Insights from the field of public relations contextualises the causes of incumbents’ strategies by integrating the impact of public perceptions (Jin & Hong, 2010) and the type of the medium on which the allegations and the responses are delivered (Schutz, et al 2011; Liu & Fraustino, 2014). Contingency theory demonstrates how the timing of responses affects their public perception (Jin & Cameron, 2007).
Of particular interest to this article are the institutional conditions under which blame avoidance unfolds. The literature posits three types of connections between blame avoidance strategies and institutional factors. The first argument suggests that the institutional setting structures the way that blame is perceived by the public. This rationale dates back to Powell and Whitten’s (Powell & Whitten, 1993) insight that voters are not going to blame all governments equally for bad economic performance. The authors designed the “clarity of responsibility” index, which posits that voters will attribute responsibility differently, depending on whether there is a coalitional government, presidential regime, divided government, etc. This insight about the institutional context of attributing blame has been largely corroborated (Anderson, 2000; Lewis-Back, 2008). A second connection posits that the executives can avoid some of the blame for unpopular welfare policies if they frame these policies in a particular way, which is constrained by institutional factors. The third type of relationship between institutions and blame avoidance is posited by Brändström and Kuipers (Brändström & Kuipers, 2003), who suggest that the type of institutions provide various opportunities for government officials alleged in media scandals to disperse blame.
Formulating the Hypotheses
Out of more than 20 possible variables that could possible impact on government responses discussed above, the article focuses on comparing the relative explanatory power of the “nature of the allegation,” the “type of accuser” and “the presence of an investigation into the media allegation” as drivers of evasive versus explanatory responses. It is hypothesized that the relative significance of the nature of the accusation and the identity of the accuser would differ in Germany and Russia. In Germany, the nature of the accusation would be a more important factor impacting government responses because attacking journalists or members of the opposition is not a sustainable option. The strategy of counter-attacks is risky and costly. It could backfire because both the opposition and the media in an established democracy are credible institutions with long history. Thus, the only “wriggle room” for the incumbents is to make use of the “elasticity” and hidden character of misconduct allegations. The German office-holders would be more likely to be evasive when accused of misconduct rather than corruption because misconduct charges are less observable and harder to prove. We propose the following hypothesis:
H2: The “nature of the allegation” is a more impactful factor for government responses in an established democracy, while the “identity of the accuser” is more impactful in a managed democracy.
As argued above, the variation in the observability and consequences of various types of allegations open the allegation to politicisation and therefore to various responses (Bovens, et al, 1999). Building on these advances in theoretical research, which take into account the nature of the allegation, we distinguish between incompetence charges and misconduct charges. This distinction structures and solicits specific government responses by defining the room for blame avoidance. Incompetence charges are harder to deny than misconduct charges for three main reasons: (1) evidence of incompetence or a policy failure is easier to find; (2) the consequences of incompetence charges are more visible to the public than the consequences of corruption or personal misconduct and (3) the consequences of incompetence are easier to evaluate because they are less subject to moral judgement. For example, incompetence charges, such as failed government’s failure to stop a terrorist incident, are easier to prove than a minister’s marital infidelity. In another instance, the misconduct allegation that German chancellor Kohl accepted illegal party financing for the CDU is more liable to contention than the allegation that there was infected meat on the market shelves. Furthermore, the negative effects of illegal party financing are more open to moral judgement than the negative effects of eating infected meat. Based on these ruminations, we suggest the following hypothesis:
H3: Corruption allegations, rather than charges of incompetence, are more likely to elicit evasive responses in an established democracy.
Unlike in Germany, in Russia, the identity of the accuser would be a powerful causal factor. We suggest that the opposition and the media constitute two major categories of accusers. They have a different status and operate on the basis of a quite different political constituency, which is likely to be relevant to government responses. First, the media and the opposition differ in regard to the mechanisms they have at their disposal to elicit responses from the government. Their claims have various degrees of institutional “embeddedness.” The media are an informal accuser and the weight of their accusation hinges on the approval of the public and the propensity of other institutions to take the claim seriously. By contrast, the opposition has a formalised and institutionalised way to assert their claims in parliament. It can use parliamentary inquiry or parliamentary committees to follow up on their criticism. Consequently, the government may be more inclined to discard media allegations and take oppositional allegations seriously.
Second, the media and the opposition have different legitimacy as makers of the accusations. The opposition is elected, the media are not. The opposition is considered to be representative. By contrast, the criticisms of the media represent the views of the editorial staff or the owners of the media outlet. Furthermore, the media are not accountable, whereas the opposition is accountable to the electorate at the next elections. In addition, it is not clear what motivates the media and the opposition to criticise the government publicly. The opposition’s job is to watch for the public interest and monitor the government. The media, on the other hand, seek to make monetary profits as well. Because of their differing status in terms of legitimacy, accountability, representativeness and motivation, media and oppositional allegations are poised to create different leeway for government strategies. The following hypothesis is proposed:
H4: Allegations made by journalists rather than the opposition are more likely to elicit evasive responses in a managed democracy.
The article builds upon the literature arguing that institutional factors matter for the blame avoidance game. However, it adds a fourth rationale to the existing three connections between institutional settings and blame avoidance outlined above. It argues that the presence of an investigation affects the leeway for denying wrong-doing. The motivation behind it is that formal investigations have more authority and power to prove crimes, such as subpoena witnesses and demand written documentation than any other channel. Knowing that their misconduct or corruption is more likely to be found out through an official outlet, it would be harder for incumbents to deny the wrong-doing. Of course, the impact of the type of investigation on government responses would differ across regime types. In authoritarian regimes, the legislative body is largely impotent, which means that a parliamentary investigation is less likely to find any wrong-doing and thus the authorities would be less likely to be scared into admitting blame. The prosecutor general in Russia, on the other hand, is sometimes an extended arm of the executive, and the incumbents are more likely to be affected. The direction of the effect of a prosecutorial investigation on the likelihood of denying blame would differ, depending on which arm of the government the prosecutor favours and which arm is alleged in the behaviour. Conversely, in a democratic regime, the opposition is stronger, parliamentary investigations into media allegations would have more legitimacy and power, and consequently, the executive members would be less likely to deny the blame. These ruminations result in the following hypothesis:
H5: Incumbents are less likely to engage in evasive tactics if there is institutional investigation of the media allegation.
Selecting the Cases
Following the theoretical discussion about the importance of the nature of the allegation and the identity of the accuser as causal factors that matter for government responses, Germany and Russia emerge as suitable candidates because they exhibit a variety on these two independent variables. Germany and Russia exhibit variation on the independent variables because Germany is a well-established democracy with relatively free press, while Russia is a managed democracy, where the media freedom is more restricted. The German democratic landscape is dominated by political competition over issues rather than personalities. The political parties are centred on existing cleavages in society rather than on personalities (Beyme, 1996; Diamond & Plattner, 2002). By contrast, in Russia, the dominant political parties have been created by powerful players around the Kremlin (Oates, 2006, 72). The political opposition is weak (Oates, 2003).
It can be expected that politicians in democratic countries are more attuned to estimate the extent of politicization inherent in the type of allegation than rulers in a managed democracy. Politicians have different awareness of the importance and availability of evidence in criticisms in both systems. In democracies, the political discourse is more issue oriented (Mair, 2013). Attentiveness to issues means that the politicians place a higher premium on discussing actual policies rather than smearing personalities. That is why incumbents will likely take a fuller advantage of corruption and non-policy oriented accusations to avoid blame.
The Russian and German media systems differ substantially. The German system is characterised as a democratic corporatist model with high media circulation, high political pluralism, high professionalization, and relatively high state intervention (Hallin & Mancini, 2004). The Russian system best fits the polarised pluralism model with low professionalization, low development of mass press, high political pluralism and high state intervention (De Smaele, 2007). The period examined in this article includes a stage (1996-2000) when the Russian media were partially free but other media outlets were brought by oligarchs who used their media assets to wage vicious kompromat wars against each other and against Yeltsin’s ailing government (Oates, 2006: 112). It also includes a stage of media development (2000-2005), which started with Putin’s rise to power, when the government assumed control over national media (Dunn, 2011: 44). By contrast, the German media system is characterized by diversity of media sources. The ownership of media outlets is split roughly equally between private and public enterprises. There is a vibrant local press. Journalists enjoy editorial independence (Trappel, et al, 2007: 136).
Governments are more likely to attack journalists in the polarised corporatist model than the media in the democratic corporatist model for the following reasons: media outlets are more vulnerable because journalistic associations are weak; the state controls many outlets; alternative news sources are not financially independent enough (due to low circulation) to bolster the accusers claims against the government. Consequently, the government knows that if it attacks a media outlet, the chances that some other institution, such as the prosecutor, will take on the charge is small. In addition, the media sources are considered to convey the claims of the oligarchs owning them rather than the evidence contained in wrongdoing.
Data
This article seeks to measure and explain government account-giving strategies in response to allegations made in and by the media on the basis of an original database of 692 media allegations containing 1,890 articles that are critical of the governments in Russia and Germany. The content analysis encompasses the period from January 1995 to December 2005. While the data covers a limited and distant period of time, the data are valuable as being the first of its kind in terms of quantifying the intractable situation, where all the following conditions are upheld simultaneously: there is information related to a long list of contextual factors, there is a information for government responses, and the data uses the same coding to cover two very different countries. The available data enables the first of its kind statistical analysis between a large number of contextual factors and government responses. It can be replicated for more recent periods.
The German accusations were sampled manually from the newspaper Die Welt, which happens to be the only newspaper with an online archive at the time the content analysis was performed. Although Die Welt occupies the conservative end of the political spectrum, it is a suitable source because it expresses the views of the opposition for the better part of the period studied in the article. Arguably, the most numerous accusations can be found in oppositional newspapers. The information was double-checked against the contents of the more popular TV news program Die Tagesschau. The major stories from Die Welt were also present on national TV evening news broadcasts. Therefore, Die Welt stories containing accusations were relatively representative of a broader political discourse.
The Russian allegations were extrapolated from reading the daily press digests of Radio Free Europe Liberty (RFEL). RFEL was a suitable source of information because it sampled all available media sources. Single media sources in Russia do not produce enough criticisms. Additional sources were the Open Media Research Institute Daily Digest and the Jamestown Foundation Daily Monitor. Taken together, these three sources reach an expansive and representative sample of news outlets in Russia, the most important of which are: NTV, ITAR TASS, the newspapers Moskovsky Komsomolets, Nezavisimaya Gazeta, Izvestya the magazine Ogonek and radio Ekho Moskvy.
The unit of analysis is a media allegation of incompetence or malfeasance. One media allegation can be based on many articles. The characteristics of the data are described in Appendix 1.
The content analysis was conducted through manual coding by the authors and two PhD students. All coders used the same coding scheme. The authors read the full length of all articles of each daily issue for each outlet for the full duration of the examined period. After the coding was finished, a new coding was carried out to determine intercoder reliability. The two PhD students with knowledge of German and Russian read the news for 12 randomly picked months of the websites of Die Welt and on RFLE respectively. These 12 months selected for checking inter-coder reliability constitute 10% of all available data, as is common practice. The Krippendorff’s alpha for each indicator are listed in Appendix 2.
Measurement
Information about each accusation includes: the number of articles related to the accusation, the type of accusation, who made the allegation, which government minister was accused, whether there was an investigation of the allegation and how the government responded. A detailed description of each variable is found in Appendix 2.
Type of Allegation
The allegations were grouped into two main categories: incompetence and misconduct. Incompetence allegations include mishandling of natural disasters, fires, terrorist attacks, bad policy decisions. Misconduct charges include cases of corruption, marital infidelity, being gay, paedophilia, acts of illegality. The variable “type of allegation” equals 1 in case of misconduct allegations and 0 in case of incompetence allegations.
We have incorporated the logic used in the blame barometer, which is the most detailed coding method for grading the “media heat coming onto the government from the media” (Hood, et al, 2009). However, we have modified the blame barometer to be both more precise and more comprehensive. According to the blame barometer, the three most important indicators of the degree of media negativity contained in an allegation are: the volume of articles, the position of the accuser and the presence of various calls to investigate/sanction the government. Out method codes the volume of the articles and the position of the accuser as separate independent variables. The advantage of this process of essentially disaggregate the measure of media negativity (which is otherwise collapsed into one measure) is that it allows the investigator to determine which of the two actually causes the variation in government responses. In addition, our measure of a media allegation is much broader than the barometer’s “calls for investigations or resignation.” The latter constitute just a small part of all media criticism as it is entirely possible to criticise the government without necessarily demanding its resignation or an investigation. Our measures take stock of such additional factors.
Specifically, the allegations are coded according to three comprehensive criteria, such as: (1) explicit description of the act of accusing: use of verbs such as allege, accuse, charge, attack; (2) Indirect description of the act of accusing: Person A says the government’s policy tests his/her patience; Person B is angry at the government’s policy; Person C expresses dissatisfaction; People are gathered to protest against a policy; Person D threatens or initiates a lawsuit against the government; Person E calls for the government resignation; Person F calls for investigating the government.; (3) Media articles which were coded as accusations also include references to a policy or behaviour of an incumbent as morally wrong, unacceptable, harmful, illegal, corrupt, inappropriate, dishonest, untrustworthy. This is indicative list of all the words the coders found to express unfavourable attitude to the behaviour or the policies of the incumbents.
The identity of the accuser
The identity of the accuser consists of nine categories: government member, opposition, journalist, individual person, businessman, international organization, judicial representative, NGO, trade unions.
From all these accusers, the two most prominent and important ones are the media and the opposition. The variable the ‘media as an accuser’ is coded 1 if a journalist makes the accusation and 0 if any other person or organisation makes it. The variable ‘the opposition is an accuser’ is coded 1 if a member of the opposition makes the accusation and 0 if any other person or organisation makes it.
Government Responses
The responses that the incumbents gave to public allegations are classified in three broader categories of “no comment”, “evasive response” and “explanatory response”. In the regressions, only cases using evasive and explanatory responses were analysed. Government responses were coded as 1 if the government gave evasive responses, and were coded 0 if the government gave explanatory responses. Evasive responses include the categories of “deny charges”, “attack accuser verbally”, “file libel suit”, and “attack accuser non-verbally.” Explanatory responses include the categories of justifications, excuses, accepting the blame.
One example of an evasive strategy employed by Russian incumbents is the case where officials in Prime Minister Viktor Chernomyrdin's office persuaded a Moscow bank to freeze the credit of the popular weekly magazine Ogonek after it had reported on Chernomyrdin's lavish bear-hunting expedition in Yaroslav (RFE/RL, February 13, 1997). Another example of an aggressive Russian evasive strategy is to attack the journalist making the accusation psychologically. On September 16, 1997, the newspaper Nezavisimaya gazeta published a report suggesting that First Deputy Prime Minister Chubais's team fabricated reports of an alleged assassination plot against Chubais (RFE/RL, September 15, 1997). Attempting to intimidate the investigative journalist Alexander Khinshtein, the ministry attempted to force Khinshtein into in a psychiatric hospital.
Type of Investigation
This variable records all investigations preceding and following the emergence of allegations. The variables are dichotomous and equal 1 if the respective forum investigates the allegation, and 0 if it does not. The types of investigations are as follows: parliamentary questions after the accusation; parliamentary questions prior to allegation; lower court reviews the emerged allegation, the accusation is part of a prior lawsuit, the alleged incumbent is threatened with a suit, the allegation is investigated by an international commission, the allegation is investigated by internal governmental department, the allegation is investigated by a parliamentary committee, the allegation is investigated in a party meeting.
Logistic Regression Analysis
Logistic regression analysis is used to examine reported changes in government responses due to various factors. Logistic regression analysis is a suitable statistical procedure for our case as it predicts a categorical outcome variable, which in this case pertains to whether the incumbent gave an evasive or explanatory answer, from a set of categorical and continuous independent variables, which in our model are the type of accuser, type of accusation, government popularity and others. The logistic analysis identifies which predictors are statistically significant and the exponentiated coefficients of the variables provide an estimate of the changes in the odds that an outcome will occur. The data are fully compliant with the requirements of a logistic regression.
Evasive strategies do not “improve answerability and the flow of information.” Their main purpose is to deflect attention from the allegation. Some evasive responses restrict the access to account-holding by de-legitimizing the accusers. When the government is constantly evasive, it becomes much harder for the public to attribute political blame because the details of the decision making process and the supposed wrong-doings remain unknown.
The first hypothesis speculates how government responses to media allegations vary. It is important to compare government responses across regime types. There is only speculation, but no empirical evidence, that governments in democratic countries are more likely to be verbally accountable in response to media allegations than governments in less democratic countries. This expectation is not hard to justify. Democratic incumbents in general are thought to believe in the value of public accountability and feel that their responses may impact their chances for re-election. Yet, we do not know whether this is the case for sure, what the difference in the proclivity to engage in explanations is, and to what extent avoiding giving explanations is done by failing to respond or by actively denying the charges.
The article offers a comparative quantitative approach which complements communicative research focusing on single-case studies (Yang, 2014: 155). As Schillemans (2013, 19-20) notes in his review of accountability studies, only 17% of such studies in major journals use quantitative data. Yang (Yang, 2012: 259; Yang, 2014: 155) laments the fact that most research focuses on single-case studies and that “there is a lack of methodological diversity and a dearth of quantitative designs.” In addition, there is a bias towards examining countries from, and mainly discussing, North America and Europe, with only 4% of studies originating in Asia, for example (Sikorski, 2018). Furthermore, political scandals are usually examined by political science and psychology journals, but rarely in communication outlets (Sikorski, 2018). Of course, some scholars venture beyond the usual confines. For example, Toepfl (Toepfl, 2011: 1302) states that “over the past two decades, most scholars have based their studies on ... [the] basic assumption that scandals can only occur in liberal democracies. Thus, this study wants to raise the question: What specific patterns of scandal communication can be observed in the semi-authoritarian political environment of contemporary Russia?”
The multiplicity of organising criteria and the lack of data prevent a clear formulation of the “puzzle”: how does the pattern of government responses vary across regime types? In short, the explanatory variable is unknown, although establishing it should be the first step in studying the spin game across the globe. To rectify this problem, we propose the following hypothesis:
H1: Government officials are more likely to give explanatory responses to media allegations in an established rather than in a managed democracy.
RQ 2: What Factors Impact Governments’ Verbal Responses?
The second inquiry pertains to the factors impacting the government responses in the aftermath of media allegations. The theoretical understanding in this regard has steadily evolved to consider more context-specific factors.
Primary Causal Variables Affecting Government Responses: “Volume of allegations” and “type of allegation”
Two primary causal variables affecting blame avoidance strategies are the ‘volume’ (Weaver, 1986) or ‘negativity’ of news coverage (Hood, 2010, Sulitzeanu-Kenan, 2010) and the “nature of the allegation.” The notion of negativity bias constitutes a central premise posits that the government’s chief priority is to avoid blame rather than claim credit. The rationale is that the public rewards success four times less than it punishes wrongdoing (Weaver, 1986).
The importance of the “nature of the allegation” as a factor influencing government responses has increased. In earlier writings, the nature of the allegation is deemed less consequential because incumbents’ strategies were thought to be entirely the verbal making of the incumbents themselves. Edelman (Edelman, 1988: 31) argued that “a policy failure, like all news developments, is a creation of the language used to depict it” and Boin et al (Boin, et al, 2010) view crisis responses are “framing contests.” In contrast, Brändström and Kuipers (Brändström & Kuipers, 2003) take a more evidence-based approach to allegations and suggest that the choice of strategies depends on the failures themselves. More specifically, government strategies are affected by considerations whether the fiasco violates crucial norms, relates to high-level officials, and whether the failure is on systematic or an individual level. Bolstering the importance of the nature of the crisis further, Bovens (Bovens, et al, 1999: 145) argues that the “observability” of the alleged wrongdoing is important. The authors distinguish between policy fiascos, such as natural disasters, which are observed, and political fiascos, which are construed. Furthermore, policy fiascos put the blame on the government as it was not misfortune but mismanagement that has brought on the bad consequences. Further research on the type of allegation distinguishes between the subjective and objective dimensions of the crisis, and crises with endogenous versus exogenous causes (Boin et al, 2009: 100). McConnell (McConnell, 2003) differentiates between the level of secrecy, the threat, the time horizons and the imminence of the crisis (sudden, creeping, chronic crisis). Djerf-Pierre et al (Djerf-Pierre, et al, 2013) discern between a moral scandal and a policy failure.
Context-specific Variables Affecting Government Responses
Early research on government responses has been criticised for being “widely acknowledged” to be “scattered and unconcentrated, and that, for the most part, it neglects both contextual factors and comparative research” (Hinterleitner & Sager, 2015: 140). More specifically, communication scholarship has been faulted for disregarding institutional constraints and interactional factors. It has also come under attack for following a linear logic of sequence. A few critics have lamented the absence of comparative research, which only exacerbates the above shortcomings. These criticisms have been addressed fully or partially by a second generation of research, which has fully embraced the value of contextualising the causal factors. This second wave of research has evolved by contextualising the actor-specific and environment-specific factors exerting an influence on government responses (Hinterleitner & Sager, 2015; Hinterleiter, 2017).
Zooming in on the contextual factors, we note that actor-specific factors take into account the leadership style of the actors (Boin, 2010) and the ordering of the agents’ preferences (Sulitzeanu-Kenan & Hood, 2005). Game –theoretical insights explain how the opportunity sets of government ministers weigh in the trade-off between being responsive in dismissing an alleged minister versus the benefits of appearing loyal to the prime-minister’s appointees (Belnski, et al: 2012). Some theorists have shifted attention from the government to the accuser. Hood (Hood, 2009) incorporates the identity of the accuser by pointing out that if the incumbent is alleged by the government’s own party or by a member of the government, then the allegation is perceived as a less legitimate and credible threat. In this vein of research, Dimova (Dimova, 2012; Dimova, 2013) discusses the costs and benefits of making accusations for alleged incumbents. The benefits could be symbolic or monetary, while the benefits are shaped by a myriad of factors, such as informal media practices, defamation laws, strength of journalistic associations, and relations with the Secret Services, which possess compromising materials.
Research on context-specific (rather than actor-specific) drivers of government responses has proliferated in several directions. Earlier writings highlight the impact of causality, responsibility or blameworthiness (Shaver & Drown, 1986). Further context-specific factors pertain to the networks within which actors are situated (Moynihan, 2012) and the institutional advantages some actors enjoy in dispersing blame (Brändström & Kuipers, 2003). Habermas's (Habermas, 2006) theory of communicative action adds to that research agenda by underscoring the interactive character of crisis responses. Situational crisis management contributes to context- specific factors by evaluating intensifying factors, such as a history of similar past crises or a negative prior reputation, because they are likely to “intensify” attributions of the organization’s crisis responsibility (Coombs, 2007). Insights from the field of public relations contextualises the causes of incumbents’ strategies by integrating the impact of public perceptions (Jin & Hong, 2010) and the type of the medium on which the allegations and the responses are delivered (Schutz, et al 2011; Liu & Fraustino, 2014). Contingency theory demonstrates how the timing of responses affects their public perception (Jin & Cameron, 2007).
Of particular interest to this article are the institutional conditions under which blame avoidance unfolds. The literature posits three types of connections between blame avoidance strategies and institutional factors. The first argument suggests that the institutional setting structures the way that blame is perceived by the public. This rationale dates back to Powell and Whitten’s (Powell & Whitten, 1993) insight that voters are not going to blame all governments equally for bad economic performance. The authors designed the “clarity of responsibility” index, which posits that voters will attribute responsibility differently, depending on whether there is a coalitional government, presidential regime, divided government, etc. This insight about the institutional context of attributing blame has been largely corroborated (Anderson, 2000; Lewis-Back, 2008). A second connection posits that the executives can avoid some of the blame for unpopular welfare policies if they frame these policies in a particular way, which is constrained by institutional factors. The third type of relationship between institutions and blame avoidance is posited by Brändström and Kuipers (Brändström & Kuipers, 2003), who suggest that the type of institutions provide various opportunities for government officials alleged in media scandals to disperse blame.
Formulating the Hypotheses
Out of more than 20 possible variables that could possible impact on government responses discussed above, the article focuses on comparing the relative explanatory power of the “nature of the allegation,” the “type of accuser” and “the presence of an investigation into the media allegation” as drivers of evasive versus explanatory responses. It is hypothesized that the relative significance of the nature of the accusation and the identity of the accuser would differ in Germany and Russia. In Germany, the nature of the accusation would be a more important factor impacting government responses because attacking journalists or members of the opposition is not a sustainable option. The strategy of counter-attacks is risky and costly. It could backfire because both the opposition and the media in an established democracy are credible institutions with long history. Thus, the only “wriggle room” for the incumbents is to make use of the “elasticity” and hidden character of misconduct allegations. The German office-holders would be more likely to be evasive when accused of misconduct rather than corruption because misconduct charges are less observable and harder to prove. We propose the following hypothesis:
H2: The “nature of the allegation” is a more impactful factor for government responses in an established democracy, while the “identity of the accuser” is more impactful in a managed democracy.
As argued above, the variation in the observability and consequences of various types of allegations open the allegation to politicisation and therefore to various responses (Bovens, et al, 1999). Building on these advances in theoretical research, which take into account the nature of the allegation, we distinguish between incompetence charges and misconduct charges. This distinction structures and solicits specific government responses by defining the room for blame avoidance. Incompetence charges are harder to deny than misconduct charges for three main reasons: (1) evidence of incompetence or a policy failure is easier to find; (2) the consequences of incompetence charges are more visible to the public than the consequences of corruption or personal misconduct and (3) the consequences of incompetence are easier to evaluate because they are less subject to moral judgement. For example, incompetence charges, such as failed government’s failure to stop a terrorist incident, are easier to prove than a minister’s marital infidelity. In another instance, the misconduct allegation that German chancellor Kohl accepted illegal party financing for the CDU is more liable to contention than the allegation that there was infected meat on the market shelves. Furthermore, the negative effects of illegal party financing are more open to moral judgement than the negative effects of eating infected meat. Based on these ruminations, we suggest the following hypothesis:
H3: Corruption allegations, rather than charges of incompetence, are more likely to elicit evasive responses in an established democracy.
Unlike in Germany, in Russia, the identity of the accuser would be a powerful causal factor. We suggest that the opposition and the media constitute two major categories of accusers. They have a different status and operate on the basis of a quite different political constituency, which is likely to be relevant to government responses. First, the media and the opposition differ in regard to the mechanisms they have at their disposal to elicit responses from the government. Their claims have various degrees of institutional “embeddedness.” The media are an informal accuser and the weight of their accusation hinges on the approval of the public and the propensity of other institutions to take the claim seriously. By contrast, the opposition has a formalised and institutionalised way to assert their claims in parliament. It can use parliamentary inquiry or parliamentary committees to follow up on their criticism. Consequently, the government may be more inclined to discard media allegations and take oppositional allegations seriously.
Second, the media and the opposition have different legitimacy as makers of the accusations. The opposition is elected, the media are not. The opposition is considered to be representative. By contrast, the criticisms of the media represent the views of the editorial staff or the owners of the media outlet. Furthermore, the media are not accountable, whereas the opposition is accountable to the electorate at the next elections. In addition, it is not clear what motivates the media and the opposition to criticise the government publicly. The opposition’s job is to watch for the public interest and monitor the government. The media, on the other hand, seek to make monetary profits as well. Because of their differing status in terms of legitimacy, accountability, representativeness and motivation, media and oppositional allegations are poised to create different leeway for government strategies. The following hypothesis is proposed:
H4: Allegations made by journalists rather than the opposition are more likely to elicit evasive responses in a managed democracy.
The article builds upon the literature arguing that institutional factors matter for the blame avoidance game. However, it adds a fourth rationale to the existing three connections between institutional settings and blame avoidance outlined above. It argues that the presence of an investigation affects the leeway for denying wrong-doing. The motivation behind it is that formal investigations have more authority and power to prove crimes, such as subpoena witnesses and demand written documentation than any other channel. Knowing that their misconduct or corruption is more likely to be found out through an official outlet, it would be harder for incumbents to deny the wrong-doing. Of course, the impact of the type of investigation on government responses would differ across regime types. In authoritarian regimes, the legislative body is largely impotent, which means that a parliamentary investigation is less likely to find any wrong-doing and thus the authorities would be less likely to be scared into admitting blame. The prosecutor general in Russia, on the other hand, is sometimes an extended arm of the executive, and the incumbents are more likely to be affected. The direction of the effect of a prosecutorial investigation on the likelihood of denying blame would differ, depending on which arm of the government the prosecutor favours and which arm is alleged in the behaviour. Conversely, in a democratic regime, the opposition is stronger, parliamentary investigations into media allegations would have more legitimacy and power, and consequently, the executive members would be less likely to deny the blame. These ruminations result in the following hypothesis:
H5: Incumbents are less likely to engage in evasive tactics if there is institutional investigation of the media allegation.
Selecting the Cases
Following the theoretical discussion about the importance of the nature of the allegation and the identity of the accuser as causal factors that matter for government responses, Germany and Russia emerge as suitable candidates because they exhibit a variety on these two independent variables. Germany and Russia exhibit variation on the independent variables because Germany is a well-established democracy with relatively free press, while Russia is a managed democracy, where the media freedom is more restricted. The German democratic landscape is dominated by political competition over issues rather than personalities. The political parties are centred on existing cleavages in society rather than on personalities (Beyme, 1996; Diamond & Plattner, 2002). By contrast, in Russia, the dominant political parties have been created by powerful players around the Kremlin (Oates, 2006, 72). The political opposition is weak (Oates, 2003).
It can be expected that politicians in democratic countries are more attuned to estimate the extent of politicization inherent in the type of allegation than rulers in a managed democracy. Politicians have different awareness of the importance and availability of evidence in criticisms in both systems. In democracies, the political discourse is more issue oriented (Mair, 2013). Attentiveness to issues means that the politicians place a higher premium on discussing actual policies rather than smearing personalities. That is why incumbents will likely take a fuller advantage of corruption and non-policy oriented accusations to avoid blame.
The Russian and German media systems differ substantially. The German system is characterised as a democratic corporatist model with high media circulation, high political pluralism, high professionalization, and relatively high state intervention (Hallin & Mancini, 2004). The Russian system best fits the polarised pluralism model with low professionalization, low development of mass press, high political pluralism and high state intervention (De Smaele, 2007). The period examined in this article includes a stage (1996-2000) when the Russian media were partially free but other media outlets were brought by oligarchs who used their media assets to wage vicious kompromat wars against each other and against Yeltsin’s ailing government (Oates, 2006: 112). It also includes a stage of media development (2000-2005), which started with Putin’s rise to power, when the government assumed control over national media (Dunn, 2011: 44). By contrast, the German media system is characterized by diversity of media sources. The ownership of media outlets is split roughly equally between private and public enterprises. There is a vibrant local press. Journalists enjoy editorial independence (Trappel, et al, 2007: 136).
Governments are more likely to attack journalists in the polarised corporatist model than the media in the democratic corporatist model for the following reasons: media outlets are more vulnerable because journalistic associations are weak; the state controls many outlets; alternative news sources are not financially independent enough (due to low circulation) to bolster the accusers claims against the government. Consequently, the government knows that if it attacks a media outlet, the chances that some other institution, such as the prosecutor, will take on the charge is small. In addition, the media sources are considered to convey the claims of the oligarchs owning them rather than the evidence contained in wrongdoing.
Data
This article seeks to measure and explain government account-giving strategies in response to allegations made in and by the media on the basis of an original database of 692 media allegations containing 1,890 articles that are critical of the governments in Russia and Germany. The content analysis encompasses the period from January 1995 to December 2005. While the data covers a limited and distant period of time, the data are valuable as being the first of its kind in terms of quantifying the intractable situation, where all the following conditions are upheld simultaneously: there is information related to a long list of contextual factors, there is a information for government responses, and the data uses the same coding to cover two very different countries. The available data enables the first of its kind statistical analysis between a large number of contextual factors and government responses. It can be replicated for more recent periods.
The German accusations were sampled manually from the newspaper Die Welt, which happens to be the only newspaper with an online archive at the time the content analysis was performed. Although Die Welt occupies the conservative end of the political spectrum, it is a suitable source because it expresses the views of the opposition for the better part of the period studied in the article. Arguably, the most numerous accusations can be found in oppositional newspapers. The information was double-checked against the contents of the more popular TV news program Die Tagesschau. The major stories from Die Welt were also present on national TV evening news broadcasts. Therefore, Die Welt stories containing accusations were relatively representative of a broader political discourse.
The Russian allegations were extrapolated from reading the daily press digests of Radio Free Europe Liberty (RFEL). RFEL was a suitable source of information because it sampled all available media sources. Single media sources in Russia do not produce enough criticisms. Additional sources were the Open Media Research Institute Daily Digest and the Jamestown Foundation Daily Monitor. Taken together, these three sources reach an expansive and representative sample of news outlets in Russia, the most important of which are: NTV, ITAR TASS, the newspapers Moskovsky Komsomolets, Nezavisimaya Gazeta, Izvestya the magazine Ogonek and radio Ekho Moskvy.
The unit of analysis is a media allegation of incompetence or malfeasance. One media allegation can be based on many articles. The characteristics of the data are described in Appendix 1.
The content analysis was conducted through manual coding by the authors and two PhD students. All coders used the same coding scheme. The authors read the full length of all articles of each daily issue for each outlet for the full duration of the examined period. After the coding was finished, a new coding was carried out to determine intercoder reliability. The two PhD students with knowledge of German and Russian read the news for 12 randomly picked months of the websites of Die Welt and on RFLE respectively. These 12 months selected for checking inter-coder reliability constitute 10% of all available data, as is common practice. The Krippendorff’s alpha for each indicator are listed in Appendix 2.
Measurement
Information about each accusation includes: the number of articles related to the accusation, the type of accusation, who made the allegation, which government minister was accused, whether there was an investigation of the allegation and how the government responded. A detailed description of each variable is found in Appendix 2.
Type of Allegation
The allegations were grouped into two main categories: incompetence and misconduct. Incompetence allegations include mishandling of natural disasters, fires, terrorist attacks, bad policy decisions. Misconduct charges include cases of corruption, marital infidelity, being gay, paedophilia, acts of illegality. The variable “type of allegation” equals 1 in case of misconduct allegations and 0 in case of incompetence allegations.
We have incorporated the logic used in the blame barometer, which is the most detailed coding method for grading the “media heat coming onto the government from the media” (Hood, et al, 2009). However, we have modified the blame barometer to be both more precise and more comprehensive. According to the blame barometer, the three most important indicators of the degree of media negativity contained in an allegation are: the volume of articles, the position of the accuser and the presence of various calls to investigate/sanction the government. Out method codes the volume of the articles and the position of the accuser as separate independent variables. The advantage of this process of essentially disaggregate the measure of media negativity (which is otherwise collapsed into one measure) is that it allows the investigator to determine which of the two actually causes the variation in government responses. In addition, our measure of a media allegation is much broader than the barometer’s “calls for investigations or resignation.” The latter constitute just a small part of all media criticism as it is entirely possible to criticise the government without necessarily demanding its resignation or an investigation. Our measures take stock of such additional factors.
Specifically, the allegations are coded according to three comprehensive criteria, such as: (1) explicit description of the act of accusing: use of verbs such as allege, accuse, charge, attack; (2) Indirect description of the act of accusing: Person A says the government’s policy tests his/her patience; Person B is angry at the government’s policy; Person C expresses dissatisfaction; People are gathered to protest against a policy; Person D threatens or initiates a lawsuit against the government; Person E calls for the government resignation; Person F calls for investigating the government.; (3) Media articles which were coded as accusations also include references to a policy or behaviour of an incumbent as morally wrong, unacceptable, harmful, illegal, corrupt, inappropriate, dishonest, untrustworthy. This is indicative list of all the words the coders found to express unfavourable attitude to the behaviour or the policies of the incumbents.
The identity of the accuser
The identity of the accuser consists of nine categories: government member, opposition, journalist, individual person, businessman, international organization, judicial representative, NGO, trade unions.
From all these accusers, the two most prominent and important ones are the media and the opposition. The variable the ‘media as an accuser’ is coded 1 if a journalist makes the accusation and 0 if any other person or organisation makes it. The variable ‘the opposition is an accuser’ is coded 1 if a member of the opposition makes the accusation and 0 if any other person or organisation makes it.
Government Responses
The responses that the incumbents gave to public allegations are classified in three broader categories of “no comment”, “evasive response” and “explanatory response”. In the regressions, only cases using evasive and explanatory responses were analysed. Government responses were coded as 1 if the government gave evasive responses, and were coded 0 if the government gave explanatory responses. Evasive responses include the categories of “deny charges”, “attack accuser verbally”, “file libel suit”, and “attack accuser non-verbally.” Explanatory responses include the categories of justifications, excuses, accepting the blame.
One example of an evasive strategy employed by Russian incumbents is the case where officials in Prime Minister Viktor Chernomyrdin's office persuaded a Moscow bank to freeze the credit of the popular weekly magazine Ogonek after it had reported on Chernomyrdin's lavish bear-hunting expedition in Yaroslav (RFE/RL, February 13, 1997). Another example of an aggressive Russian evasive strategy is to attack the journalist making the accusation psychologically. On September 16, 1997, the newspaper Nezavisimaya gazeta published a report suggesting that First Deputy Prime Minister Chubais's team fabricated reports of an alleged assassination plot against Chubais (RFE/RL, September 15, 1997). Attempting to intimidate the investigative journalist Alexander Khinshtein, the ministry attempted to force Khinshtein into in a psychiatric hospital.
Type of Investigation
This variable records all investigations preceding and following the emergence of allegations. The variables are dichotomous and equal 1 if the respective forum investigates the allegation, and 0 if it does not. The types of investigations are as follows: parliamentary questions after the accusation; parliamentary questions prior to allegation; lower court reviews the emerged allegation, the accusation is part of a prior lawsuit, the alleged incumbent is threatened with a suit, the allegation is investigated by an international commission, the allegation is investigated by internal governmental department, the allegation is investigated by a parliamentary committee, the allegation is investigated in a party meeting.
Popularity
The popularity of the government, the opposition and the president in Russia’s case is included to control for the reputational effects and public perception effects inherent in the contextual environment. Government, oppositional and presidential popularity is measured as the first available public approval rating after the allegation emerges. It is a continuous variable.Position of the accused incumbent
The coding differentiates between the position of the prime minister versus the position of lower ranking ministers, government officials or the government as a whole. It is a dichotomous variable indicating 1 if the Prime minister is alleged, 0 if otherwise.Logistic Regression Analysis
Logistic regression analysis is used to examine reported changes in government responses due to various factors. Logistic regression analysis is a suitable statistical procedure for our case as it predicts a categorical outcome variable, which in this case pertains to whether the incumbent gave an evasive or explanatory answer, from a set of categorical and continuous independent variables, which in our model are the type of accuser, type of accusation, government popularity and others. The logistic analysis identifies which predictors are statistically significant and the exponentiated coefficients of the variables provide an estimate of the changes in the odds that an outcome will occur. The data are fully compliant with the requirements of a logistic regression.
Results
General Observations
In regard to the first causal variable, the type of allegation, Germany and Russia differ because incompetence allegations dominate in Germany (86%) whereas both types of allegations feature in fairly equal measures in Russia (56% vs. 44%) (Figure 1).
Figure 1: Distribution of incompetence and misconduct charges in Russia and Germany;
Source: Author database, N=692
Germany and Russia also show variation on the second causal factor, i.e. the identity of the accuser. In Germany, the most active critic of the government in the media is the opposition. In Russia, the most active critic of the government in the media is the media (Figure 2).
Figure 2: Distribution of the Types of Accusers in Russia and Germany; Source: Author database, N=692
The first hypothesis, which proposed that government officials are more likely to give explanatory responses to media allegations in an established rather than in a managed democracy, is supported. The findings show that the Russian government is indeed more likely to adopt evasive, sometimes even quite aggressive, responses to media allegations, whereas the German government is more likely to adopt defensive, explanatory strategies. Defensive strategies that are prevalent in Germany are supposedly more conducive to democratic accountability because they provide the public with useful information. By contrast, the Russian government is significantly less inclined to defend its positions and explain its actions in response to public allegations. We did found no examples where the Russian government confessed or apologized, as their German counterparts did on occasion.
The relation between the country, on the one hand, and the three categories of responses (evasive, explanatory and non-engagement) was significant, as Pearson X2 (2, N = 655) = 50.77, p < .0001. The association between the type of the country and the type of responses is strong as Cramér’s V equals .27 (Cohen 1998). Evasive strategies are more closely associated with Russia, while explanatory responses are more strongly associated with Germany. There is almost no probability that the association is accidental as Pearson X2(2, N = 655) = 1118.82, p < .00001.
Figure 3: Government Responses to Public Allegations in Germany and Russia (as a percentage of all allegations); Source: Author database, N=692
As Figure 3 demonstrates, non-engagement is the most frequent type of response in both countries. “No-comment” can often be a wise strategy, from the perspective of government communication, because comments prolong the life of a news story (Schillemans, 2012: 101). In fact, the German government is less likely to comment on accusations than the Russian government and is, in this respect, less responsive to criticisms. This can be due to the larger volume of accusations in Germany than in Russia.
Moving to the statistical analysis, a brief look at the overall findings lends credence to hypothesis two, namely that the “nature of the allegation” is a more impactful factor for government responses in an established democracy, while the “identity of the accuser” is more impactful in a managed democracy. While we would look more closely at the direction and magnitude of the relationships below, it is now important to note that the allegations made by journalists rather than the opposition are more likely to elicit evasive responses in a established democracy, which in this case is Germany. Unlike German strategies, Russian responses are affected by the identity of the accuser rather than by the nature of the allegation. The identity of the accuser, when the accuser is a journalist, is statistically significant while the nature of the allegation is not. This picture of varying causal patterns reveals that comparison across regime types is useful as it provides important keys to the variation of factors affecting government blame avoidance strategies.
Hypothesis 3, that corruption allegations, rather than charges of incompetence, are more likely to elicit evasive responses in an established democracy, in this case Germany, is supported. The government usually provides explanatory responses to
claims of incompetence while it uses evasive tactics in response to claims of misconduct. Table 2 shows how responses in Germany are predicated on the nature of the accusation, not on the accuser.
Table 2 reports on the model for government responses in Germany, listing the regression coefficient, standard error, odds ratio, and the ratio’s 95% confidence interval for each predictor. A test of the full model with all predictors was statistically significant at χ² (16, N = 166) = 21.59, p < .10, indicating that the predictors, as a set, relatively significantly distinguished between incumbents’ explanatory tactics versus those instances in which officials opted for evasive tactics. The model explained 24.6% (Nagelkerke R2) of the variance. Classification correctly predicted 69.9% of explanatory and evasive responses.
According to the Wald χ2 statistic, the following individual conditions predict a change in the odds of giving evasive versus explanatory answers. The statistically significant categorical predictor variables were as follows: members of parliament question the government in parliament, b = -.99, Wald χ2 (1) = 2.96, p < .10; being accused in the media of corruption compared to being accused in the media of incompetence, b = 1.04, Wald χ2 (1) = 3,21, p < .10; the allegation is reviewed in a lower court of justice, b = 4.004, Wald χ2 (1) = 4,92, p < .05; the number of articles covering the allegation, b = -.01, Wald χ2 (1) = 1.80, p < .10.
The findings confirm the conventional wisdom that the intensity of critical coverage matters for blame avoidance strategies. An increase of 1 article covering a certain allegation decreases the likelihood of an evasive response by 100%. Furthermore, making an allegation of corruption or misconduct rather than one of incompetence increases the chances of the government giving an evasive (rather than an explanatory) response 7.11 times. Why are the German incumbents more likely to deny charges of corruption than charges of incompetence? Corruption allegations, as argued above, are easier to rebuff because they harder to prove, observe and judge. Hence, claims of misconduct only rarely trigger substantive and explanatory responses. Rather, those claims are discredited and the German government seemed to be inclined to cast allegations of misconduct and corruption as attacks rather than to give information about them. For example, in 2001 the chancellor Gerhard Schröder was accused of taking part in money laundering through the privatisation of the Leuna refinery. This was alleged to have happened during Schröder’s mandate in Lower Saxony. The government’s response was to accuse the accuser by stating that the Leuna affair was misused to stage a campaign against the government. The government also gauged accusations as calumny and black smear (“Diffamierungen”) (Die Welt, July 27, 2001).
Table 3 reports on the model testing factors impacting on government responses in Russia, listing the regression coefficient, standard error, odds ratio, and the ratio’s 95% confidence interval for each predictor. A test of the full model with all predictors was statistically significant at χ2 (14, N = 122) = 26.42, p < .10, indicating that the predictors, as a set, significantly distinguished between evasive and explanatory responses to media allegations. The power of the model was relatively strong, with R2 = .43 (Nagelkerke), meaning that 43% of the variance was explained. Classification was relatively impressive, correctly predicting 88.52% of evasive and explanatory responses.
Table 2.
Logistic Regression: Predictive Model of Government Responses in Germany
Note 1: Odds ratio = exp(b). Model χ² (16, N = 166) = 21.59, p < .10. * p < .05. ** p < .01. ***
p < .001. Note 2: Condition: binary dependent variable=1 if government gives evasive responses; =0 if government gives explanatory responses.
p < .001. Note 2: Condition: binary dependent variable=1 if government gives evasive responses; =0 if government gives explanatory responses.
Table 3.
Logistic Regression: Predictive Model of Government Responses in Russia
According to the Wald χ2 statistic, the following individual conditions predict a change in the odds of giving explanatory versus giving evasive answers in Russia. The statistically significant categorical predictor variables were as follows: the accusation being made by a journalist, b = 2.05, Wald χ2 (1) = 2.82, p < .10; the number of articles covering the allegation, b = .06, Wald χ2 (1) = 2.29, p < .10; the accusation being reviewed by the prosecutor-general, b = -2.06, Wald χ2 (1) = 2.42, p < .10.
As shown in Table 3, once again, our findings are in line with mainstream research that has established that the intensity of critical coverage matters in established democracy. The article now shows that this finding applies to Russia’s managed democracy. As hypothesis four proposed, the identity of the accuser matters in Russia. If the accusation is made by a journalist (rather than anybody else, who is not a media representative, be it an incumbent, the opposition, or the public), the chances that the incumbent would give an evasive rather than an explanatory response increase by 4.29. Russian incumbents may be more likely to give evasive responses to journalistic claims because journalistic associations are weak and some media outlets have little credibility as they are thought to be the mouthpieces of wealthy tycoons. Another reason could be that the government selects the most outrageous or the most impotent critics to impart the impression that there are active critics of the government but that they are too outrageous or too unconvincing. It is much easier for the state to deny egregious accusations (that the state has enabled). For example, Zasursky (Zasursky, 1999) suggests that the state allows the nationalist Zhirinovsky to appear in the media because they make the state appear moderate in comparison. If the accuser is an oppositional member, there is no affect on the blame game. Attacking the opposition is less of a priority for Russian incumbents as the opposition poses little serious threats to the elites.
The findings offer some enlightening results regarding hypothesis 5, namely that “incumbents are less likely to deny blame if there is an institutional investigation into the media allegation.” The results reaffirm the purported connection between institutional design and government responses. Both in Germany and Russia, various types of investigations turn out to have a significant effect on blame avoidance strategies. The intriguing part of this result is that, while incumbents tend to take stock of various investigations, their strategies differ, depending on the source of investigation. Thus, parliamentary questions in Germany decrease the likelihood of denying blame by 35%, while investigations by lower courts tend to increase the likelihood of denying blame by 44.07 (!) This finding engenders the speculation that democratic governments are very sensitive to the legitimacy, authority and power of various investigations. They will engage in explanatory tactics only if the body investigating them is powerful enough to produce evidence and witnesses in due time. The most importance difference between the lower courts and parliamentary questions, however, is their public exposure. Thus, incumbents in a democracy would be most likely to switch from evasive to explanatory strategies, not only when there is proof of wrongdoing, but also when the public can witness it.
As for Russia, as expected, the Russian legislative body is too weak to exert any type of influence and change the blame game. A prosecutorial investigation is the only one that matters as it decreases the odds of evasive responses by 11%. If it is true that the prosecutor-general in Russia is mostly under the influence of the powerful president (Greenberg, 2009), then engaging in evasive tactics when investigated by the prosecutor would be a costly option.
Discussion and Implications
What have we learnt about practices of account-giving in two very different regime types? To begin with, we found that governments in various regime types have various proclivities to give account of their actions in the media. We were able to confirm the hypothesis that Russian governments tend to adopt more evasive strategies than the German incumbents. In general, German officials tend to explain their policies more often and to accept some blame for it to a greater extent. The findings also show that, irrespective of their political differences, the “volume of allegations” is an important consideration affecting responses both in Russia and Germany. The most common response to public allegations was ‘silence’ in both countries. Regardless of the regime type, incumbents hope that the media storm will die a natural death by ignoring the allegations.
Our statistical results extended theoretical knowledge of some of the contextual factors that were supposed to matter for government responses by testing the significance of two independent variables: the identity of the accuser and the nature of the allegation. In Germany, government strategies are mostly dependent on the type of accusation. Government responses are more evasive in cases of misconduct than in cases of incompetence. The account-giving and account-holding revolves around the issues, the evidence and its interpretation. The additional finding is that German ministers are more likely to explain themselves in the media, if there is a formal investigation of the government. By contrast, the process of account-giving in Russia is affected by the identity of the accuser rather than the type of accusation.
The article outlines two avenues for future research. The first implication is that communication research has been moving in the right direction to think critically about the way to expand the factors that could influence the blame game. If anything, it should perhaps think of even more novel avenues to expand these variables as the factors included in the present analysis are unprecedentedly comprehensive, yet the model for Germany explains only 24.6% of evasive and explanatory responses, and the model for Russia explains 43% of the data. Generally, the debate how to contextualise causal factors has been moving in two directions: expanding the list to account for the context in which the allegation is made, such as the interactions within the network in which the blame situation is embedded, the media “heat” and the reputational effects, and expanding the actor-specific factors, in which incumbents take decisions how to manage blame. Yet, if these two ways of formulating hypothetical determinants of the blame game do not fully explain the evidence, future research should think up additional criteria.
The second implication is that communication research should perhaps be more rigorous about the testing of potential causal factors impacting blame management. The dearth of data is indisputably problematic because, as pointed out, research relies mainly on case studies or aggregate data in single country studies. Aggregate data about media allegations, a full list of contextual factors, and government responses for a prolonged period of time across regime types are very rare. While it is understandable that situations of media accusations are intractable, quantifying them using a consistent methodology across all studies, is crucial. Such an approach presents the only opportunity to move to the next stage of communication research, i.e. linking the causal factors to the blame-avoidance tactics. Otherwise, the fragmentation in communication research produces two parallel universes, one debating the theoretical merits of causal factors and another- quantifying blame tactics.
Of all the contextual factors that matter for government tactics in the aftermath of media allegations, one particularly fruitful area of research would be the interaction between the institutional investigations and the incumbents’ decisions whether to accept or deny blame. The findings showed that Russian government officials were less likely to give evasive responses if the prosecutor general investigates them, while German government officials adapted their responses according to the source of investigation. They were more likely to give explanatory responses to questions posed during parliamentary inquiries, while they were more likely to shut down the allegations if the latter were viewed in a lower court. The article speculated that the public exposure of the various investigative bodies is an important factor in shaping blame avoidance strategies. Yet, future research should examine whether this speculation applies to a larger number of cases in democratic and non-democratic regimes.
2 Khodorkovsky’s statements were recorded in a written form by the author.
2. Anderson C.J. Economic voting and political context: a comparative perspective. Electoral Studies. – 2000. - 19(2). – P. 151 - 170.
3. Arnold R.D. Congress, the press, and political accountability. – Princeton: Princeton University Press, 2004. – 348 p.
4. Aucoin P., Heintzman R.The dialectics of accountability for performance in public management reform. // International Review of Administrative Sciences. – 2000. – 66. – P. 45-55.
5. Berlinski S., Dewan T. and Dowding K. Accounting for ministers: scandal and survival in British government 1945-2007. – Cambridge: Cambridge University Press, 2012. – 286 p.
6. Beyme Klaus von. Transition to democracy in Eastern Europe. - Palgrave, 1996. – 362 p.
7. Boin A., P.‘t Hart and McConnell A. Towards a theory of crisis exploitation: political and policy impacts of framing contests and blame games. // Journal of European Public Policy. – 2009. - 16 (1). - P. 81–106.
8. Boin A., P. 't Hart., McConell A., Preston T. Leadership style, crisis response and blame management: the case of hurricane Katrina. // Public Administration. – 2010. - 88 (3). – P. 706 - 725.
9. Bovens M. Analysing and assessing accountability: a conceptual framework. // European Law Journal. – 2007. - 13 (4). – P. 447 – 468.
10. Bovens M., Paul ’t Hart, Dekker S., Verheuvel G. The politics of blame avoidance: defensive tactics in a Dutch crime-fighting fiasco. In Anheier H.K. (ed.) When things go wrong: failures and breakdowns in organizational settings. - London: Sage, 1999. – P.123 – 147.
11. Bovens M., Schillemans T., Goodin R.E. Public accountability. In: The Oxford Handbook Public Accountability. - Oxford: Oxford University Press, 2014. – P. 1 -20.
12. Brandsma G.J., Schillemans T. The Accountability Cube: Measuring Accountability. // Journal of Public Administration Research and Theory. – 2013. – 23. – P. 953 - 975.
13. Brändström A., Kuipers S. From ‘Normal Incidents’ to Political Crises: Understanding the Selective Politicization of Policy Failures. // Government and Opposition. – 2003. – 38 (3). – P. 279 - 305.
14. Coombs W.T., Holladay S.J. Communication and attributions in a crisis: An experimental study in crisis communication. Journal of public relations research. – 1996. – 8 (4). – P. 279 - 295.
15. Coombs W.T. An Analytic framework for crisis situations: better responses from a better understanding of the situation. // Journal of Public Relations Research. – 1998. – 10 (3). - P. 177 - 191.
16. Coombs W.T. Impact of past crises on current crisis communication: Insights from situational crisis communication theory. // The Journal of Business Communication. – 2004. – 41 (3). – P. 265 - 289.
17. Cohen J. Statistical power analysis for the behavioural sciences. - Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1998. – 368 p.
18. Corbett E.P. Foreword. In H.R. Ryan (ed.) Oratorical Encounters: Selected Studies and Sources of Twentieth-Century Political Accusations and Apologies. - Westport, CT: Greenwood, 1988. – 276 p.
19. Dahl R. Democracy and Its Critics. - New Haven: Yale University Press, 1989. – 376 p.
20. Dawson J. Cultures of Democracy in Serbia and Bulgaria – How Ideas Shape Publics. - Ashgate: Farnham, 2014. – 382 p.
21. De Smaele H. In search for a label for the Russian media system. // https://www.welt.de/print-welt/article663372/Waigel-verteidigt-smaele.html
22. De Smaele H. Mass media and the information climate in Russia. // Europe-Asia Studies. – 2007. - 59(8). – P. 1299 - 1313.
23. Diamond L., Plattner M. Democracy after Communism. – Baltimore: Johns Hopkins University Press, 2002. – 367 p.
24. Die Welt. 2001. - July 27. Schmidt: Schweizer Leuna-Akten werden zu Kampagne mißbraucht. // http://www.welt.de/print-welt/article472425/Schmidt-Schweizer-Leuna-Akten-werden-zu-Kampagne-missbraucht.html
25. Die Welt. 1995. - October 28. Waigel verteidigt Bundeshaushalt. https://www.welt.de/print-welt/article663372/Waigel-verteidigt-Bundeshaushalt.html
26. Dimova G. Who Criticizes the government in the media? The symbolic power model. // Observatorio (OBS*). – 2012. – 6 (1). – P. 23 - 36.
27. Dixon R., Arndt R.C., Mullers M., Vakkuri J., Engnlom K., Hood C. A Lever for improvement or a magnet for blame? Press and political responses to international educational rankings in four EU countries. – 2013. - 91 (2). – P. 484 - 505.
28. Djerf-Pierre M., Ekström M., Johansson B. Policy failure or moral scandal? Political accountability, journalism and new public management. // Media, Culture and Society. – 2013. – November. - (35). – P. 960 - 976.
29. Dewan T., Dowding K. The corrective effect of ministerial resignations on government popularity. // American Journal of Political Science. – 2005. – 49 (1). – P. 46 - 56.
30. Dunn J. Where did it all go wrong? Russian television in the Putin era in the post-Soviet Russian media: conflicting signals. In Birgit Beumers, Stephen Hutchings and Natalia Rulyova (eds.) BASEES/Routledge Series on Russian and East European Studies, 2011. – P. 26 – 38.
31. Edelman M. Constructing the political spectacle. University of Chicago Press, 1988. – 376 p.
32. Esser F. Mediatization as a challenge: media logic versus political logic in democracy in the age of globalization and mediatization. – Palgrave: Macmillan, 2013. – P. 236 – 248.
33. Flinders M. Defending Politics: Why democracy matters in the 21st century. -Oxford University Press, 2012. – 368 p.
34. Goode P. The Decline of regionalism in Putin's Russia: boundary issues. // BASEES/Routledge Series on Russian and East European Studies, 2011. – 268 p.
35. Gorham M., Lunde I., Paulsen M. Digital Russia: the language, culture and politics of new media communication. // Routledge Contemporary Russia and Eastern Europe Series, 2014. – P. 63 – 78.
36. Ginsberg B., Shefter M. Politics by other means. - New York: Norton, 2002. – 386 p.
37. Greenberg J.D. The Kremlin's eye: the 21st century Prokuratura in the Russian authoritarian tradition. Stan. J. Int'l L., 2009. – Vol. 45. – Issue 1. – P. 27 – 38.
38. Habermas J. Political communication in media society: Does democracy still enjoy an epistemic dimension? The impact of normative theory on empirical research. // Communication theory. - 2006. – 16 (4). – P. 411 - 426.
39. Hallin D.C., Mancini P. Comparing media systems: Three models of media and politics. - Cambridge University Press, 2004. – 386 p.
40. Hinterleitner M. Reconciling perspectives on blame avoidance behaviour. // Political Studies Review. – 2017. – 15 (2). – P. 243 - 254.
41. Hinterleitner M., Sager F. Avoiding blame - a comprehensive framework and the Australian home insulation program fiasco. // Policy Studies Journal. - 2015. – 43 (1). – P. 139-161.
42. Hood C., Jennings W., Dixon R., Hogwood B., Beeston C. Testing times: exploring staged responses and the impact of blame management strategies in two examination fiasco cases. // European Journal of Political Research. – 2009. – 48 (6). – P. 695 - 722.
43. Hood C., Jennings W., Hogwood B., Beeston C. Fighting fires in testing times: exploring a staged response hypothesis for blame management in two exam fiasco cases. - European Journal of Political Research. – 2007. – 46 (3). – P. 395 - 422.
44. Hood C. 2010. The Blame Game: Spin, Bureaucracy, and Self-Preservation in Government. Princeton University Press.
45. Hood C. Accountability and blame avoidance. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – 286 p.
46. Hood C., Jennings W., Copeland P. Blame avoidance in comparative perspective: reactivity, staged retreat and efficacy. // Public Administration. – 2016. - 94 (2). – P. 542 - 562.
47. Iyengar S., McGrady J. Media Politics. - New York: Norton, 2007. – 283 p.
48. Jacobs S., Schillemans T. Media and Accountability. Typology and Exploration. // Policy & Politics. - 2016. – Vol. 3. – P. 26- 38.
49. Jin Y., Cameron G.T. The effects of threat type and duration on public relations practitioner's cognitive, affective, and conative responses in crisis situations. // Journal of Public Relations Research. – 2007. – 19 (3). – P. 255 - 281.
50. Jin Y., Hong S.Y. Explicating crisis coping in crisis communication. // Public Relations Review. – 2010. – 36 (4). – P. 352 - 360.
51. Keane J. The Life and Death of Democracy. New York: Simon and Schuster, 2009. – 286 p.
52. Keohane R. The Concept of accountability in world politics and the use of force. // Michigan Journal of International Law. – 2003. – 24. – P. 1130 - 1135.
53. Kriesi H. Democracy as a moving target in democracy in the age of globalization and mediatization. Kriesi, Hanspeter, Daniel Bochsler, Jörg Matthes and Professor Sandra Lavenex (eds.). – Palgrave: Macmillan, 2013. – 328 p.
54. Lewis-Beck M.S., Nadeau R., Elias A. Economics, party and the vote: causality issues and panel data. // American Journal of Political Science. – 2008. – 52 (1). – P. 84 - 95.
55. Liu B.F., Fraustino J.D. Beyond image repair: suggestions for crisis communication theory development. // Public Relations Review. – 2014. – 40 (3). – P. 543 - 546.
56. Maggetti M. The Media accountability of independent regulatory agencies. // European Political Science Review. - 2012. - 4 (3). – P. 385 - 408.
57. Mair P. Ruling the Void: the hollowing of Western democracy. - Verso, 2013. – 386 p.
58. McGraw K.M. Avoiding blame: an experimental investigation of political excuses and justifications. // British Journal of Political Science. – 1990. – 20 (1). – P. 119 - 131.
59. McGraw K. Managing blame: an experimental test of the effects of political accounts. // American Political Science Review. – 1991. - (85) 4. – P. 1133 - 1157.
60. McConnell A. Overview: crisis management, influences, responses and evaluation. // Parliamentary Affairs. – 2003. – 56. – P. 393 - 409.
61. Mancini P. Political scandals as a democratic challenge in “Assassination Campaigns”: corruption scandals and news media instrumentalization. // International Journal of Communication. – 2018. – 12. – P. 26 – 38.
62. March James G., Olson J. Democratic governance. - New York: The Free Press, 1995. – 368 p.
63. Markovits A., Silverstein M. The politics of scandal: power and process in liberal democracies. – New York: Holmes & Meier Publishers Inc., 1991. – 386 p.
64. Mazzoleni G., Schultz. W. Mediatization of politics: a challenge for democracy? // Political Communication. – 1999. – 16 (3). – P. 247 - 261.
65. McCubbins M., Schwartz T. Congressional oversight overlooked: police patrols versus fire alarms. // American Journal of Political Science. – 1984. – 28. – P. 16 - 79.
66. Meyer T. Media Democracy. How the media colonize politics. - Cambridge: Politics, 2002. – 396 p.
67. Moynihan D.P. Extra‐network organizational reputation and blame avoidance in networks: the hurricane Katrina example. // Governance. – 2012. – 25 (4). – P. 567 - 588.
68. Norris P. A Virtuous circle: political communications in post-industrial societies. – Cambridge: Cambridge University Press, 2000. – 286 p.
69. Norris P. Watchdog Journalism. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – P. 525 - 541.
70. Oates S. Television, democracy, and elections in Russia. // BASEES/Routledge Series on Russian and East European Studies. - 2006. – Vol. 16. – P. 173 – 186.
71. Oates S. Television, voters and the development of the ‘Broadcast Party’ in the 1999–2000 elections in Russia: their impact and legacy. Vicki Hesli and Bill Reisinger (eds.). - Cambridge and New York: Cambridge University Press, 2003. – 286 p.
72. Peruzzotti E., Smulovitz C. Enforcing the rule of law: social accountability in the New Latin American democracies. - Pittsburgh: University of Pittsburgh Press, 2007. – 348 p.
73. Petrov N., McFaul M. The Essence of Putin’s managed democracy. // http://carnegieendowment.org/2005/10/18/essence-of-putin-s-managed-democracy/2a3
74. Przeworski A., Stokes S.C., Manin B. Democracy, accountability, and representation. – Cambridge: Cambridge University Press, 1999. – 286 p.
75. RFE/RL, February 13, 1997. Bear hunt story causes trouble for Ogonek. // http://www.rferl.org/content/article/1141348.html
76. RFE/RL, September 15, 1997. ‘Nezavisimaya Gazeta’ Slams Chubais. // http://www.rferl.org/content/article/1141494.html
77. Ross F. Framing welfare reform in affluent societies: rendering restructuring more palatable? // Journal of Public Policy. - 2000. – 20 (2). – P. 169 - 193.
78. Sakwa R. The Putin redux: power and contradiction in contemporary Russia. New York: Routledge, 2014. – 268 p.
79. Schillemans T. Mediatization of public services. How public organizations adapt to the media. - Frankfurt: Peter Lang, 2012. – 328 p.
80. Schillemans T. The Public accountability review: a meta-analysis of public accountability research in six academic disciplines, 2013. // http://dspace.library.uu.nl/handle/1874/275784, last accessed on August 1, 2014.
81. Jacobs S., Schillemans T. Media and public accountability: typology and exploration. // Policy & Politics. – 2016. – 44 (1). – P. 23 - 40.
82. Schultz F., Utz S., Göritz A. Is the medium the message? Perceptions of and reactions to crisis communication via twitter, blogs and traditional media. // Public Relations Review. – 2011. – 37 (1). - P. 20 - 27.
83. Shaver K.G., Drown D. On causality, responsibility, and self-blame: a theoretical note. // Journal of Personality and Social Psychology. - 1986. – 50 (4). – P. 697.
84. Sikorski C. Political scandals as a democratic challenge the aftermath of political scandals: a meta-analysis. // International Journal of Communication. – 2018. – 12. - P. 25 – 32.
85. Strøm K., et al. Delegation and accountability in parliamentary democracies. - Oxford: Oxford University Press, 2003. – 386 p.
86. Sulitzeanu-Kenan R. Reflection in the shadow of blame: when do politicians appoint commissions of inquiry? // British Journal of Political Science. – 2010. – 40. – P. 613 – 634.
87. Sulitzeanu-Kenan R., Hood C. Blame avoidance with adjectives? // Motivation, Opportunity, Activity and Outcome. - 2005. – Vol. 16. – P. 63 – 78.
88. The Moscow Times, September 7, 2004. ‘Izvestia Editor Resigns Over Beslan Coverage’. // http://www.themoscowtimes.com/sitemap/free/2004/9/article/izvestia-editor-resigns-over-beslan-coverage/228529.html
89. Toepfl F. Managing public outrage: power, scandal, and new media in contemporary Russia. // New Media & Society. – 2011. – 13 (8). – P. 1301 - 1319.
90. Trappel J., Werner A. Meier and Leen d’Haenens and Jeanette Steemers and Barbara Thomass (eds.) Media in Europe Today. – New York: Intellect Books, 2007. – 286 p.
91. Vu T., Bruno D. Formal and informal mechanisms of accountability in local governance. Towards a new authoritarian governance model. // Business Systems Review. - 2013. - 2 (2). – P. 330 - 367.
92. Yang K. Qualitative Analysis. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – P. 159-176.
93. Yan J., Cameron G.T. The effects of threat type and duration on public relations practitioner's cognitive, affective, and conative responses in crisis situations. // Journal of Public Relations Research. – 2007. – 19 (3). – P. 255 - 281.
94. Yankova G. Political accountability and media scandals: a comparative exploration. // Political Science Journal. – 2006. – 1 (3). – P. 49 - 70.
95. Yankova-Dimova G. Media scandals: how political scandals arise. // Global Media Journal. – 2013. – Issue 3. – Vol. 1. – P. 12 – 18.
96. Vibert F. The Rise of the unelected: democracy and the new separation of powers. – Cambridge: Cambridge University Press, 2007. – 286 p.
97. Weaver R.K. The politics of blame avoidance. // Journal of Public Policy. – 1986. - Vol. 6 (4). – P. 371 – 398.
As shown in Table 3, once again, our findings are in line with mainstream research that has established that the intensity of critical coverage matters in established democracy. The article now shows that this finding applies to Russia’s managed democracy. As hypothesis four proposed, the identity of the accuser matters in Russia. If the accusation is made by a journalist (rather than anybody else, who is not a media representative, be it an incumbent, the opposition, or the public), the chances that the incumbent would give an evasive rather than an explanatory response increase by 4.29. Russian incumbents may be more likely to give evasive responses to journalistic claims because journalistic associations are weak and some media outlets have little credibility as they are thought to be the mouthpieces of wealthy tycoons. Another reason could be that the government selects the most outrageous or the most impotent critics to impart the impression that there are active critics of the government but that they are too outrageous or too unconvincing. It is much easier for the state to deny egregious accusations (that the state has enabled). For example, Zasursky (Zasursky, 1999) suggests that the state allows the nationalist Zhirinovsky to appear in the media because they make the state appear moderate in comparison. If the accuser is an oppositional member, there is no affect on the blame game. Attacking the opposition is less of a priority for Russian incumbents as the opposition poses little serious threats to the elites.
The findings offer some enlightening results regarding hypothesis 5, namely that “incumbents are less likely to deny blame if there is an institutional investigation into the media allegation.” The results reaffirm the purported connection between institutional design and government responses. Both in Germany and Russia, various types of investigations turn out to have a significant effect on blame avoidance strategies. The intriguing part of this result is that, while incumbents tend to take stock of various investigations, their strategies differ, depending on the source of investigation. Thus, parliamentary questions in Germany decrease the likelihood of denying blame by 35%, while investigations by lower courts tend to increase the likelihood of denying blame by 44.07 (!) This finding engenders the speculation that democratic governments are very sensitive to the legitimacy, authority and power of various investigations. They will engage in explanatory tactics only if the body investigating them is powerful enough to produce evidence and witnesses in due time. The most importance difference between the lower courts and parliamentary questions, however, is their public exposure. Thus, incumbents in a democracy would be most likely to switch from evasive to explanatory strategies, not only when there is proof of wrongdoing, but also when the public can witness it.
As for Russia, as expected, the Russian legislative body is too weak to exert any type of influence and change the blame game. A prosecutorial investigation is the only one that matters as it decreases the odds of evasive responses by 11%. If it is true that the prosecutor-general in Russia is mostly under the influence of the powerful president (Greenberg, 2009), then engaging in evasive tactics when investigated by the prosecutor would be a costly option.
Discussion and Implications
What have we learnt about practices of account-giving in two very different regime types? To begin with, we found that governments in various regime types have various proclivities to give account of their actions in the media. We were able to confirm the hypothesis that Russian governments tend to adopt more evasive strategies than the German incumbents. In general, German officials tend to explain their policies more often and to accept some blame for it to a greater extent. The findings also show that, irrespective of their political differences, the “volume of allegations” is an important consideration affecting responses both in Russia and Germany. The most common response to public allegations was ‘silence’ in both countries. Regardless of the regime type, incumbents hope that the media storm will die a natural death by ignoring the allegations.
Our statistical results extended theoretical knowledge of some of the contextual factors that were supposed to matter for government responses by testing the significance of two independent variables: the identity of the accuser and the nature of the allegation. In Germany, government strategies are mostly dependent on the type of accusation. Government responses are more evasive in cases of misconduct than in cases of incompetence. The account-giving and account-holding revolves around the issues, the evidence and its interpretation. The additional finding is that German ministers are more likely to explain themselves in the media, if there is a formal investigation of the government. By contrast, the process of account-giving in Russia is affected by the identity of the accuser rather than the type of accusation.
The article outlines two avenues for future research. The first implication is that communication research has been moving in the right direction to think critically about the way to expand the factors that could influence the blame game. If anything, it should perhaps think of even more novel avenues to expand these variables as the factors included in the present analysis are unprecedentedly comprehensive, yet the model for Germany explains only 24.6% of evasive and explanatory responses, and the model for Russia explains 43% of the data. Generally, the debate how to contextualise causal factors has been moving in two directions: expanding the list to account for the context in which the allegation is made, such as the interactions within the network in which the blame situation is embedded, the media “heat” and the reputational effects, and expanding the actor-specific factors, in which incumbents take decisions how to manage blame. Yet, if these two ways of formulating hypothetical determinants of the blame game do not fully explain the evidence, future research should think up additional criteria.
The second implication is that communication research should perhaps be more rigorous about the testing of potential causal factors impacting blame management. The dearth of data is indisputably problematic because, as pointed out, research relies mainly on case studies or aggregate data in single country studies. Aggregate data about media allegations, a full list of contextual factors, and government responses for a prolonged period of time across regime types are very rare. While it is understandable that situations of media accusations are intractable, quantifying them using a consistent methodology across all studies, is crucial. Such an approach presents the only opportunity to move to the next stage of communication research, i.e. linking the causal factors to the blame-avoidance tactics. Otherwise, the fragmentation in communication research produces two parallel universes, one debating the theoretical merits of causal factors and another- quantifying blame tactics.
Of all the contextual factors that matter for government tactics in the aftermath of media allegations, one particularly fruitful area of research would be the interaction between the institutional investigations and the incumbents’ decisions whether to accept or deny blame. The findings showed that Russian government officials were less likely to give evasive responses if the prosecutor general investigates them, while German government officials adapted their responses according to the source of investigation. They were more likely to give explanatory responses to questions posed during parliamentary inquiries, while they were more likely to shut down the allegations if the latter were viewed in a lower court. The article speculated that the public exposure of the various investigative bodies is an important factor in shaping blame avoidance strategies. Yet, future research should examine whether this speculation applies to a larger number of cases in democratic and non-democratic regimes.
Notes
The importance of political scandals for democracy is thoroughly investigated in the special subsection on “Political Scandals as a Democratic Challenge” in International Journal of Communication (2018).2 Khodorkovsky’s statements were recorded in a written form by the author.
3 Kompromat is Russian word for compromising material.
4 Ideally, the classification of government responses would take into account the veracity of the allegation. However, the literature has, to our knowledge, not integrated this factor, probably because it is usually unknown whether the alleged deed is true or false.References
1. Allern S. and von Sikorski C. Political scandals as a democratic challenge: from important revelations to provocations, trivialities, and neglect. // International Journal of Communication. - 2018. - 12. – P. 28 – 36.2. Anderson C.J. Economic voting and political context: a comparative perspective. Electoral Studies. – 2000. - 19(2). – P. 151 - 170.
3. Arnold R.D. Congress, the press, and political accountability. – Princeton: Princeton University Press, 2004. – 348 p.
4. Aucoin P., Heintzman R.The dialectics of accountability for performance in public management reform. // International Review of Administrative Sciences. – 2000. – 66. – P. 45-55.
5. Berlinski S., Dewan T. and Dowding K. Accounting for ministers: scandal and survival in British government 1945-2007. – Cambridge: Cambridge University Press, 2012. – 286 p.
6. Beyme Klaus von. Transition to democracy in Eastern Europe. - Palgrave, 1996. – 362 p.
7. Boin A., P.‘t Hart and McConnell A. Towards a theory of crisis exploitation: political and policy impacts of framing contests and blame games. // Journal of European Public Policy. – 2009. - 16 (1). - P. 81–106.
8. Boin A., P. 't Hart., McConell A., Preston T. Leadership style, crisis response and blame management: the case of hurricane Katrina. // Public Administration. – 2010. - 88 (3). – P. 706 - 725.
9. Bovens M. Analysing and assessing accountability: a conceptual framework. // European Law Journal. – 2007. - 13 (4). – P. 447 – 468.
10. Bovens M., Paul ’t Hart, Dekker S., Verheuvel G. The politics of blame avoidance: defensive tactics in a Dutch crime-fighting fiasco. In Anheier H.K. (ed.) When things go wrong: failures and breakdowns in organizational settings. - London: Sage, 1999. – P.123 – 147.
11. Bovens M., Schillemans T., Goodin R.E. Public accountability. In: The Oxford Handbook Public Accountability. - Oxford: Oxford University Press, 2014. – P. 1 -20.
12. Brandsma G.J., Schillemans T. The Accountability Cube: Measuring Accountability. // Journal of Public Administration Research and Theory. – 2013. – 23. – P. 953 - 975.
13. Brändström A., Kuipers S. From ‘Normal Incidents’ to Political Crises: Understanding the Selective Politicization of Policy Failures. // Government and Opposition. – 2003. – 38 (3). – P. 279 - 305.
14. Coombs W.T., Holladay S.J. Communication and attributions in a crisis: An experimental study in crisis communication. Journal of public relations research. – 1996. – 8 (4). – P. 279 - 295.
15. Coombs W.T. An Analytic framework for crisis situations: better responses from a better understanding of the situation. // Journal of Public Relations Research. – 1998. – 10 (3). - P. 177 - 191.
16. Coombs W.T. Impact of past crises on current crisis communication: Insights from situational crisis communication theory. // The Journal of Business Communication. – 2004. – 41 (3). – P. 265 - 289.
17. Cohen J. Statistical power analysis for the behavioural sciences. - Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1998. – 368 p.
18. Corbett E.P. Foreword. In H.R. Ryan (ed.) Oratorical Encounters: Selected Studies and Sources of Twentieth-Century Political Accusations and Apologies. - Westport, CT: Greenwood, 1988. – 276 p.
19. Dahl R. Democracy and Its Critics. - New Haven: Yale University Press, 1989. – 376 p.
20. Dawson J. Cultures of Democracy in Serbia and Bulgaria – How Ideas Shape Publics. - Ashgate: Farnham, 2014. – 382 p.
21. De Smaele H. In search for a label for the Russian media system. // https://www.welt.de/print-welt/article663372/Waigel-verteidigt-smaele.html
22. De Smaele H. Mass media and the information climate in Russia. // Europe-Asia Studies. – 2007. - 59(8). – P. 1299 - 1313.
23. Diamond L., Plattner M. Democracy after Communism. – Baltimore: Johns Hopkins University Press, 2002. – 367 p.
24. Die Welt. 2001. - July 27. Schmidt: Schweizer Leuna-Akten werden zu Kampagne mißbraucht. // http://www.welt.de/print-welt/article472425/Schmidt-Schweizer-Leuna-Akten-werden-zu-Kampagne-missbraucht.html
25. Die Welt. 1995. - October 28. Waigel verteidigt Bundeshaushalt. https://www.welt.de/print-welt/article663372/Waigel-verteidigt-Bundeshaushalt.html
26. Dimova G. Who Criticizes the government in the media? The symbolic power model. // Observatorio (OBS*). – 2012. – 6 (1). – P. 23 - 36.
27. Dixon R., Arndt R.C., Mullers M., Vakkuri J., Engnlom K., Hood C. A Lever for improvement or a magnet for blame? Press and political responses to international educational rankings in four EU countries. – 2013. - 91 (2). – P. 484 - 505.
28. Djerf-Pierre M., Ekström M., Johansson B. Policy failure or moral scandal? Political accountability, journalism and new public management. // Media, Culture and Society. – 2013. – November. - (35). – P. 960 - 976.
29. Dewan T., Dowding K. The corrective effect of ministerial resignations on government popularity. // American Journal of Political Science. – 2005. – 49 (1). – P. 46 - 56.
30. Dunn J. Where did it all go wrong? Russian television in the Putin era in the post-Soviet Russian media: conflicting signals. In Birgit Beumers, Stephen Hutchings and Natalia Rulyova (eds.) BASEES/Routledge Series on Russian and East European Studies, 2011. – P. 26 – 38.
31. Edelman M. Constructing the political spectacle. University of Chicago Press, 1988. – 376 p.
32. Esser F. Mediatization as a challenge: media logic versus political logic in democracy in the age of globalization and mediatization. – Palgrave: Macmillan, 2013. – P. 236 – 248.
33. Flinders M. Defending Politics: Why democracy matters in the 21st century. -Oxford University Press, 2012. – 368 p.
34. Goode P. The Decline of regionalism in Putin's Russia: boundary issues. // BASEES/Routledge Series on Russian and East European Studies, 2011. – 268 p.
35. Gorham M., Lunde I., Paulsen M. Digital Russia: the language, culture and politics of new media communication. // Routledge Contemporary Russia and Eastern Europe Series, 2014. – P. 63 – 78.
36. Ginsberg B., Shefter M. Politics by other means. - New York: Norton, 2002. – 386 p.
37. Greenberg J.D. The Kremlin's eye: the 21st century Prokuratura in the Russian authoritarian tradition. Stan. J. Int'l L., 2009. – Vol. 45. – Issue 1. – P. 27 – 38.
38. Habermas J. Political communication in media society: Does democracy still enjoy an epistemic dimension? The impact of normative theory on empirical research. // Communication theory. - 2006. – 16 (4). – P. 411 - 426.
39. Hallin D.C., Mancini P. Comparing media systems: Three models of media and politics. - Cambridge University Press, 2004. – 386 p.
40. Hinterleitner M. Reconciling perspectives on blame avoidance behaviour. // Political Studies Review. – 2017. – 15 (2). – P. 243 - 254.
41. Hinterleitner M., Sager F. Avoiding blame - a comprehensive framework and the Australian home insulation program fiasco. // Policy Studies Journal. - 2015. – 43 (1). – P. 139-161.
42. Hood C., Jennings W., Dixon R., Hogwood B., Beeston C. Testing times: exploring staged responses and the impact of blame management strategies in two examination fiasco cases. // European Journal of Political Research. – 2009. – 48 (6). – P. 695 - 722.
43. Hood C., Jennings W., Hogwood B., Beeston C. Fighting fires in testing times: exploring a staged response hypothesis for blame management in two exam fiasco cases. - European Journal of Political Research. – 2007. – 46 (3). – P. 395 - 422.
44. Hood C. 2010. The Blame Game: Spin, Bureaucracy, and Self-Preservation in Government. Princeton University Press.
45. Hood C. Accountability and blame avoidance. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – 286 p.
46. Hood C., Jennings W., Copeland P. Blame avoidance in comparative perspective: reactivity, staged retreat and efficacy. // Public Administration. – 2016. - 94 (2). – P. 542 - 562.
47. Iyengar S., McGrady J. Media Politics. - New York: Norton, 2007. – 283 p.
48. Jacobs S., Schillemans T. Media and Accountability. Typology and Exploration. // Policy & Politics. - 2016. – Vol. 3. – P. 26- 38.
49. Jin Y., Cameron G.T. The effects of threat type and duration on public relations practitioner's cognitive, affective, and conative responses in crisis situations. // Journal of Public Relations Research. – 2007. – 19 (3). – P. 255 - 281.
50. Jin Y., Hong S.Y. Explicating crisis coping in crisis communication. // Public Relations Review. – 2010. – 36 (4). – P. 352 - 360.
51. Keane J. The Life and Death of Democracy. New York: Simon and Schuster, 2009. – 286 p.
52. Keohane R. The Concept of accountability in world politics and the use of force. // Michigan Journal of International Law. – 2003. – 24. – P. 1130 - 1135.
53. Kriesi H. Democracy as a moving target in democracy in the age of globalization and mediatization. Kriesi, Hanspeter, Daniel Bochsler, Jörg Matthes and Professor Sandra Lavenex (eds.). – Palgrave: Macmillan, 2013. – 328 p.
54. Lewis-Beck M.S., Nadeau R., Elias A. Economics, party and the vote: causality issues and panel data. // American Journal of Political Science. – 2008. – 52 (1). – P. 84 - 95.
55. Liu B.F., Fraustino J.D. Beyond image repair: suggestions for crisis communication theory development. // Public Relations Review. – 2014. – 40 (3). – P. 543 - 546.
56. Maggetti M. The Media accountability of independent regulatory agencies. // European Political Science Review. - 2012. - 4 (3). – P. 385 - 408.
57. Mair P. Ruling the Void: the hollowing of Western democracy. - Verso, 2013. – 386 p.
58. McGraw K.M. Avoiding blame: an experimental investigation of political excuses and justifications. // British Journal of Political Science. – 1990. – 20 (1). – P. 119 - 131.
59. McGraw K. Managing blame: an experimental test of the effects of political accounts. // American Political Science Review. – 1991. - (85) 4. – P. 1133 - 1157.
60. McConnell A. Overview: crisis management, influences, responses and evaluation. // Parliamentary Affairs. – 2003. – 56. – P. 393 - 409.
61. Mancini P. Political scandals as a democratic challenge in “Assassination Campaigns”: corruption scandals and news media instrumentalization. // International Journal of Communication. – 2018. – 12. – P. 26 – 38.
62. March James G., Olson J. Democratic governance. - New York: The Free Press, 1995. – 368 p.
63. Markovits A., Silverstein M. The politics of scandal: power and process in liberal democracies. – New York: Holmes & Meier Publishers Inc., 1991. – 386 p.
64. Mazzoleni G., Schultz. W. Mediatization of politics: a challenge for democracy? // Political Communication. – 1999. – 16 (3). – P. 247 - 261.
65. McCubbins M., Schwartz T. Congressional oversight overlooked: police patrols versus fire alarms. // American Journal of Political Science. – 1984. – 28. – P. 16 - 79.
66. Meyer T. Media Democracy. How the media colonize politics. - Cambridge: Politics, 2002. – 396 p.
67. Moynihan D.P. Extra‐network organizational reputation and blame avoidance in networks: the hurricane Katrina example. // Governance. – 2012. – 25 (4). – P. 567 - 588.
68. Norris P. A Virtuous circle: political communications in post-industrial societies. – Cambridge: Cambridge University Press, 2000. – 286 p.
69. Norris P. Watchdog Journalism. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – P. 525 - 541.
70. Oates S. Television, democracy, and elections in Russia. // BASEES/Routledge Series on Russian and East European Studies. - 2006. – Vol. 16. – P. 173 – 186.
71. Oates S. Television, voters and the development of the ‘Broadcast Party’ in the 1999–2000 elections in Russia: their impact and legacy. Vicki Hesli and Bill Reisinger (eds.). - Cambridge and New York: Cambridge University Press, 2003. – 286 p.
72. Peruzzotti E., Smulovitz C. Enforcing the rule of law: social accountability in the New Latin American democracies. - Pittsburgh: University of Pittsburgh Press, 2007. – 348 p.
73. Petrov N., McFaul M. The Essence of Putin’s managed democracy. // http://carnegieendowment.org/2005/10/18/essence-of-putin-s-managed-democracy/2a3
74. Przeworski A., Stokes S.C., Manin B. Democracy, accountability, and representation. – Cambridge: Cambridge University Press, 1999. – 286 p.
75. RFE/RL, February 13, 1997. Bear hunt story causes trouble for Ogonek. // http://www.rferl.org/content/article/1141348.html
76. RFE/RL, September 15, 1997. ‘Nezavisimaya Gazeta’ Slams Chubais. // http://www.rferl.org/content/article/1141494.html
77. Ross F. Framing welfare reform in affluent societies: rendering restructuring more palatable? // Journal of Public Policy. - 2000. – 20 (2). – P. 169 - 193.
78. Sakwa R. The Putin redux: power and contradiction in contemporary Russia. New York: Routledge, 2014. – 268 p.
79. Schillemans T. Mediatization of public services. How public organizations adapt to the media. - Frankfurt: Peter Lang, 2012. – 328 p.
80. Schillemans T. The Public accountability review: a meta-analysis of public accountability research in six academic disciplines, 2013. // http://dspace.library.uu.nl/handle/1874/275784, last accessed on August 1, 2014.
81. Jacobs S., Schillemans T. Media and public accountability: typology and exploration. // Policy & Politics. – 2016. – 44 (1). – P. 23 - 40.
82. Schultz F., Utz S., Göritz A. Is the medium the message? Perceptions of and reactions to crisis communication via twitter, blogs and traditional media. // Public Relations Review. – 2011. – 37 (1). - P. 20 - 27.
83. Shaver K.G., Drown D. On causality, responsibility, and self-blame: a theoretical note. // Journal of Personality and Social Psychology. - 1986. – 50 (4). – P. 697.
84. Sikorski C. Political scandals as a democratic challenge the aftermath of political scandals: a meta-analysis. // International Journal of Communication. – 2018. – 12. - P. 25 – 32.
85. Strøm K., et al. Delegation and accountability in parliamentary democracies. - Oxford: Oxford University Press, 2003. – 386 p.
86. Sulitzeanu-Kenan R. Reflection in the shadow of blame: when do politicians appoint commissions of inquiry? // British Journal of Political Science. – 2010. – 40. – P. 613 – 634.
87. Sulitzeanu-Kenan R., Hood C. Blame avoidance with adjectives? // Motivation, Opportunity, Activity and Outcome. - 2005. – Vol. 16. – P. 63 – 78.
88. The Moscow Times, September 7, 2004. ‘Izvestia Editor Resigns Over Beslan Coverage’. // http://www.themoscowtimes.com/sitemap/free/2004/9/article/izvestia-editor-resigns-over-beslan-coverage/228529.html
89. Toepfl F. Managing public outrage: power, scandal, and new media in contemporary Russia. // New Media & Society. – 2011. – 13 (8). – P. 1301 - 1319.
90. Trappel J., Werner A. Meier and Leen d’Haenens and Jeanette Steemers and Barbara Thomass (eds.) Media in Europe Today. – New York: Intellect Books, 2007. – 286 p.
91. Vu T., Bruno D. Formal and informal mechanisms of accountability in local governance. Towards a new authoritarian governance model. // Business Systems Review. - 2013. - 2 (2). – P. 330 - 367.
92. Yang K. Qualitative Analysis. In The Oxford Handbook Public Accountability. Mark Bovens, Robert E. Goodin, Thomas Schillemans (eds.). - Oxford: Oxford University Press, 2014. – P. 159-176.
93. Yan J., Cameron G.T. The effects of threat type and duration on public relations practitioner's cognitive, affective, and conative responses in crisis situations. // Journal of Public Relations Research. – 2007. – 19 (3). – P. 255 - 281.
94. Yankova G. Political accountability and media scandals: a comparative exploration. // Political Science Journal. – 2006. – 1 (3). – P. 49 - 70.
95. Yankova-Dimova G. Media scandals: how political scandals arise. // Global Media Journal. – 2013. – Issue 3. – Vol. 1. – P. 12 – 18.
96. Vibert F. The Rise of the unelected: democracy and the new separation of powers. – Cambridge: Cambridge University Press, 2007. – 286 p.
97. Weaver R.K. The politics of blame avoidance. // Journal of Public Policy. – 1986. - Vol. 6 (4). – P. 371 – 398.
