For some time, at least in the United States, trust in the press has been weak. According to the General Social Survey, fewer than 20% of Americans have reported “a great deal of confidence” in the press in every year the survey has been run since 1988. Worse, as illustrated in the graph below, the increase in people reporting they have “hardly any” confidence (the gray line) in the press has sharply increased over the same time:
Other public opinion outlets have also documented this decline (see e.g., here and here), and have further noted general dissatisfaction with the media and its reporting of news events. New research by Hanitzsch, Van Dalen, and Stenidl provides some evidence that this decline may be linked to a growing sentiment of anti-elitism in a number of countries.
Abstract (Emphasis added):
Despite signs of declining press trust in many western countries, we know little about trends in press trust across the world. Based on comparative survey data from the World Values Survey (WVS) and European Values Study (EVS), this study looks into national levels of trust in the press and identifies factors that drive differences across societies and individuals as well as over time. Findings indicate that the widely noted decline in media trust is not a universal trend; it is true for only about half of the studied countries, with the United States experiencing the largest and most dramatic drop in trust in the press. Political trust has emerged as key factor for our understanding of trust in the press. We found robust evidence for what we called the trust nexus — the idea that trust in the news media is strongly linked to the way publics look at political institutions. The link between press trust and political trust was considerably stronger in politically polarized societies. Furthermore, our analysis indicates that the relation between press trust and political trust is becoming stronger over time. We reason that the strong connection between media and political trust may be driven by a growing public sentiment against elite groups.
Hanitzsch et al. begin with a brief review of recent scholarship documenting a decline of public trust in the media (see e.g., Gronke & Cooke, 2007; Jones, 2004; Muller, 2013), particularly in developed democracies (Gronke & Cooke, 2007). They suggest a variety of reasons for this decline, including “a persistent pattern of negativity and cynicism in the news” (Hanitzsch et al., 2018, p. 4), media scandals such as plagiarism and fabrication, and antipathy fostered by political actors.
Knowledge of how trust in the press develops over time has been conducted in the United States and is thus limited. Of the scholarship available, much of it appears rooted in research on trust in public institutions and in theories of public engagement. Hanitzsch et al. (2018, p. 5) elaborate:
Three theoretical assumptions about trust inform most definitions: First, theorists generally emphasize that trust is based on past experiences that lead to expectations about (and the assessment of) how another person or institution will perform in the future (Misztal, 1996; Vanacker and Belmas, 2009). Second, this process involves risk and uncertainty because outcomes or intentions of actors are not fully known, which makes trust essential specially where verification is most difficult. Third, trust reduces social complexity by generalizing expectations of future behavior (Luhmann, 1979). Thus, trust in the media is a “psychological state comprising the intentions to accept vulnerability based on positive expectations” of a trustee’s future actions, which cannot be controlled by the truster (Mayer et al., 1995; Rousseau et al., 1998: 395). Regarding media trust, audiences are taking risks when they decide to trust the media because they are not able to verify news content on their own (Tsfati and Cohen, 2005), and they do not know whether journalists and news media adhere to any professional norms (Tsfati and Cappella, 2003). For the purposes of this study, we therefore define media trust as a form of institutional trust; it is the willingness of the audience to be vulnerable to news content based on the expectation that the media will perform in a satisfactory manner.
The scholarship on institutional trust is dominated by two schools of thought: Institutional theories and cultural theories. The former contend that trust is endogenous and a result of performance, and would explain a decline in media trust by pointing to declining performance of the media. The latter emphasizes exogenous factors and views institutional trust as an extension of social trust, learned early in development, and then projected later onto public institutions. This view would point to a general decline in trust of others as the primary explanation for reduced trust in the media.
More recent work from outside the United States suggests “media trust is strongly influenced by factors emanating from the media system, political system, and political culture (Ariely, 2015; Muller, 2013; Tsfati and Ariely, 2014)” (Hanitzsch et al., 2018, p. 4). These studies did not allow for comparisons across different types of societies (e.g., Western vs. Non-Western; democratic vs. authoritarian).
Thus, Hanitzsch et al. (2018) drew on longitudinal data from the World Values Survey (WVS) and the European Values Survey (EVS). They also shifted their focus to two additional explanations for media trust, ideological polarization and political trust. Ideological polarization was considered distinct from ideological extremity. Ideological extremity occurs when people take more extreme political positions and is measurable at the individual level, whereas “ideological polarization is a process that takes place on the aggregate level” (Hanitzsch et al., 2018, p. 6). Increasing ideological extremity and ideological polarization may increase the potential for hostile media effects (Vallone et al., 1985), thus hypothesis one was: “People have less trust in the press (a) when they hold more extreme attitudes, (b) when they live in societies with high levels of ideological polarization, and (c) as societies become more polarized” (Hanitzsch et al., 2018, p. 7).
Hypothesis two was based on what Hanitzsch et al. (2018) referred to as the trust nexus: “We argue that media trust is not a phenomenon isolated from public perceptions of other public institutions. Rather, the erosion of trust in the media is broadly connected to a public disenchantment with and widespread sense of disdain for social institutions more generally but for political institutions most particularly” (p. 7). Ariely (2015) provides the only empirical data for the basis of hypothesis two, namely that the relationship between media trust and political trust in thirty-two European countries was more or less pronounced depending on media autonomy, journalistic professionalism, and party/press parallelism. Thus, a set of sub-hypotheses made up hypothesis two:
- People have more trust in the press when they are more confident in political institutions (hypothesis 2a).
- The relation between political trust and trust in the press is stronger in ideologically polarized societies (hypothesis 2b).
- Over time, public trust in the press shrinks as political trust decreases (hypothesis 2c).
- The relation between trust in the press and political trust becomes stronger over time (hypothesis 2d).
Analyses were based on longitudinal data from the WVS (1981-2014) and the EVS (1981-2008). Measures for trust in the press were identified at the individual-level and the societal-level. At the individual level the WVS asks “I am going to name a number of organizations. For each, could you tell me how much confidence you have in them.” Answer options were none at all, not very much, quite a lot, and a great deal. “The press” was one of the organizations asked about. At the societal level trust in the press was operationalized as the percentage of the population in each country who indicated to have “quite a lot” or a “great deal” of trust in the press. Political trust was measured with three items, employing the same scale, that asked about trust in parliament, the government, and in political parties.
Ideological extremism was accounted for by “looking at the extent to which people have extreme positions on four questions about redistribution and the role of the government in the economy … Following Lindquist and Östling (2010), we calculated a combined ideological extremism score on the individual level for the four questions, based on the distance of the respondents’ answers from the center of the scale. Respondents choosing the middle positions (5 or 6) received a score of 0 (i.e., low extremism), 4 and 7 were recoded to 1, 3 and 8 to 2, 2 and 9 to 3, and 1 and 10 to 4 (i.e., high extremism)” (Hanitzsch et al., 2018, p. 9).
Social trust was assessed with responses to the question: “Generally speaking, would you say that most people can be trusted or that you need to be very careful when dealing with people” (There were only two answer options: most people can be trusted or can’t be too careful). In addition to social trust, newspaper exposure, interest in politics, education, gender, and level of democratic development were included as controls in the final model predicting press trust.
To assess each of the hypotheses four sets of analyses were conducted. Forty-five countries were included in a comparison of trust in the press over time. Next, a multilevel model was built to test hypotheses about predictors of trust in the press at the individual and societal levels. Data from fifty-two countries were used in this multilevel model. Third, changes in press trust over time were again assessed. Whether differences in trust between waves 5 (2005-2009) and 6 (2010-2014) of the WVS were impacted by changes in ideological polarization and political trust was investigated. This analysis was repeated for a comparison of press trust between wave 6 and wave 4 (1999-2004). Finally, the relationship between press trust and political trust was assessed using the whole longitudinal WVS data set. It should be noted that while 6 waves of data collection for the WVS have been conducted, each of these waves constitutes a panel study, thus the data precludes investigation of changes over time within individuals.
Clear differences in trust in the press between countries, along socio-political lines, emerged. “Australia, the United States, and New Zealand – representing the Anglo-Saxon world – were among the five countries with the lowest trust levels; the four countries with highest rust levels were all from Asia (Japan, China, India, and the Philippines)” (Hanitzsch et al., 2018, p. 11). In other words, the decline of trust in the press is not universal. Fourteen countries saw significant increases, led by Japan, while twenty-four saw significant decreases. In the United States, trust in the press decreased by more than half from 1990 to 2011.
Analyses further revealed that:
- “People tend to have more trust in the press when they are interested in politics, when they are regularly exposed to press contents, and when they trust other people. The effects of education, age, and gender are, while significant due to the large sample size, rather negligible” (Hanitzsch et al., 2018, p. 11).
- Ideological polarization was not a significant predictor of trust in the press, thus hypothesis 1a was not supported.
- A weak effect emerged for ideological extremity, but this effect became negligible once other predictors were also included in the model, thus hypothesis 1b was not supported.
- The effect of political trust on trust in the press dwarfed all other effects, thus supporting hypothesis 2a.
- The relationship between trust in the press and political trust was more pronounced in more strongly polarized societies, thus supporting hypothesis 2b.
- Changes in polarization over time were not associated with changes in trust in the press, thus hypothesis 1c was not supported.
- Changes in political trust were related to changes in trust in the press over time, thus supporting hypothesis 2c. This association held when controlling for the effects of democratic change and economic development.
- Finally, over time the relation between political trust and trust in the press has become stronger, thus supporting hypothesis 2d.
Although the decline of trust in the press is not universal the current findings have identified that this decline is substantial in a number of countries, and it is strongest in the United States. Australia and New Zealand, two other Anglo-Saxon, English-speaking countries, have also experienced a substantial decline in trust in the press. Muller (2013) previously suggested that Anglo-Saxon Western media systems tend to breed distrust in the press.
One reason for this, we believe, is the relatively long history of a two-party system in many of these countries, in which two major political ideologies struggle over political dominance (e.g., Democrats and Republicans in the United States, Labor and the Coalition in Australia, and National and Labor in New Zealand). Two-party systems are arguably more vulnerable to ideological polarization than ideologically more diverse political systems (Hanitzsch et al., 2018, p. 16).
More generally, it appears that changes in trust in the press are linked to volatility in a society’s political environment. Trust in the press has eroded in countries that went through political upheaval and uncertainty (e.g., Egypt). Decline in trust in the press was also found in many post-Communist societies (e.g., Poland, Slovenia).
Most importantly, in regards to the proposed trust nexus:
Looking at countries individually (after controlling for all other individual-level variables), standardized effect sizes of political trust vary between a low of .22 (the Netherlands and Argentina) and a high of .71 (Uzbekistan) and .61 (Azerbaijan). Effect sizes were greater than .30 for forty-six out of the fifty-three societies. Regression coefficients were larger than .50 for Algeria, Azerbaijan, Colombia, Kyrgyzstan, Malaysia, Russia, Singapore, South Africa, South Korea, Uzbekistan, and the United States. Our study therefore found substantive support for what we tentatively term the trust nexus, according to which the erosion of trust in the press is connected to a broad public disenchantment with and widespread sense of disdain for social institutions, most specifically for political institutions (Hanitzsch et al., 2018, p. 17-18).
Hanitzsch et al. (2018) close by suggesting that the major force behind this trust nexus and the notable decline of trust in the press in a number of countries is a growing anti-elitist sentiment which has created fertile ground for populist political movements. Such movements typically rely on antagonistic rhetoric (e.g., “the people” vs. “the elite”) that casts the press as members of an elitist institution. Thus, the decline of trust in the press may be a warning signal of other substantial issues of trust between a society’s institutions and its people.
Why These Findings are Important
These findings suggest political systems that lack ideological diversity are at risk for widespread declines in trust of long-standing societal institutions, including, but not limited to, the press. At first blush, these findings do not seem directly relevant to Heterodox Academy and its core issue of viewpoint diversity in the academy. Yet, like the press, American universities comprise a long-standing social institution which has also experienced a decline in public trust and confidence (see also here and here). The overall decline in political trust evident at the societal level may also have an impact on trust in universities. Increasing ideological homogeneity among the professorate (Bonica, Chilton, Rozema, & Sen, 2017; Duarte, Crawford, Stern, Haidt, Jussim, & Tetlock, 2015; Honeycutt & Freberg, 2016; Langbert, Quain, & Klein, 2016) may contribute to societal level declines in political trust and, thus, fuel the anti-elite sentiment hinted at by Hanitzsch et al.
Hanitzsch, T., Van Dalen, A., & Steindl, N. (2018). Caught in the nexus: A comparative and longitudinal analysis of public trust in the press. The International Journal of Press/Politics, 23, 3-23.