In the current media landscape, few political news outlets receive universally favorable evaluations from Americans. The erosion of public agreement about trusted information providers sits at the root of selective exposure as well as partisan disagreements over factual evidence and policy (De Benedictis-Kessner et al. Reference De Benedictis-Kessner, Baum, Berinsky and Yamamoto2019; Flynn, Nyhan, and Reifler Reference Flynn, Nyhan and Reifler2017; Stroud Reference Stroud2011). Despite these developments, local news outlets still enjoy credibility with a broad segment of the public (Fioroni Reference Fioroni2022) and can ameliorate some aspects of polarization while informing voters (Andrews, Kim, and Hyun Kim Reference Andrews, Kim and Kim2023; Darr, Hitt, and Dunaway Reference Darr, Hitt and Dunaway2021; Hayes and Lawless Reference Hayes and Lawless2015). Given the normative importance of local news, the dearth of other widely trusted information sources, and the influence of source cues on how people respond to political news (Druckman Reference Druckman2001; Ladd Reference Ladd2012; Miller and Krosnick Reference Miller and Krosnick2000), it is important to understand factors that affect the reputations of local news organizations.
We used a survey experiment to assess how evaluations of local media are affected by criticism from politicians. One possibility is that the distinctive ties between local news sources and the communities that they cover make their reputations resilient to elite criticism, in contrast to attacks on national news outlets (Ladd Reference Ladd2012). Alternatively, the generally positive reputations of local outlets may not be able to withstand elite criticism, with their current good standing instead owed to politicians largely targeting national—not local—news in their attacks. We focus on criticism from Republican politicians, given an original content analysis showing that they are more likely than Democrats to attack the media and qualitative evidence that local Republican elites have begun to insulate themselves in a friendlier, right-wing media ecosystem (Kurtzleben Reference Kurtzleben2022).
Our results demonstrate that exposure to an attack on a nearby local newspaper by a Republican politician dramatically reduces the public’s trust and intent to use that news source. While Democrats and Independents update their views of local media in line with Republican politicians’ critiques, this messaging is particularly effective at changing how Republicans view local news. After seeing the attack, overall trust and partisan divisions in views of local newspapers resemble those for national media—a demonstration of the powerful influence of politicians on media-source reputations in a new context. These findings also show that elite criticism can contribute to the increasingly imperiled demand for local news (Hopkins and Gorton Reference Hopkins and Gorton2024; McCrain and Peterson Reference McCrain and Peterson2023). Finally, the experiment establishes the fragility of the local news trust advantage, showing that it is attributable to the absence of critical elite rhetoric rather than durable local news reputations. This finding highlights the importance of factors that curb elite criticism of local news. Our study directly addresses one factor in that we find that politicians face reputational costs for attacking local media (see online appendix table C2). More speculatively, the benefits of local news coverage for many politicians—such as claiming credit for their political accomplishments (Grimmer, Messing, and Westwood Reference Grimmer, Messing and Westwood2012)—suggest another motivation for maintaining collegial relations with local news. Broadly, the results underscore the importance of how politicians weigh these factors for the local news trust advantage.
[T]he experiment establishes the fragility of the local news trust advantage, showing that it is attributable to the absence of critical elite rhetoric rather than durable local news reputations.
THE DISTINCTIVE NATURE OF LOCAL NEWS TRUST
Since its high point in the 1970s, Americans’ trust in the institutional media has declined (Ladd Reference Ladd2012). Attacks on the news media by politicians and pundits are a crucial part of this trend (for an overview, see Archer Reference Archer, Kristin and Goethals2020). After encountering criticism of news organizations from a communicator they trust, the public evaluates the media more negatively and perceives greater news coverage bias (Ladd Reference Ladd2010; Smith Reference Smith2010). Republican politicians and conservative pundits criticize the media most frequently—as epitomized by Donald Trump’s repeated characterization of mainstream media as “fake news”—which has contributed to the low levels of trust among Republicans (Domke et al. Reference Domke, Watts, Shah and Fan1999; Ladd Reference Ladd2012; Meeks Reference Meeks2020).
Despite the downward trajectory in views of national media, local news remains trusted by the public. A 2022 Morning Consult poll of US adults found that 73% reported a “very” or “somewhat” favorable view of local news, vastly exceeding evaluations that the New York Times (44%), Fox News (47%), and even CBS News (50%) received.Footnote 1 Although the survey revealed a partisan gap in local news credibility—10 percentage points higher among Democrats than Republicans—this difference is smaller than partisan divides of 36 percentage points for the New York Times, 34 points for Fox News, and 18 points for CBS News (Shevnock Reference Shevnock2022).
A potential explanation for the public’s continued trust in local news is that attacks targeting such outlets may be viewed as inappropriate. In particular, local outlets focus on subnational issues that do not map neatly onto national divides (Darr, Hitt, and Dunaway Reference Darr, Hitt and Dunaway2021) and offer ideologically moderate political coverage compared to national media (Hassell, Stiles, and Reuning Reference Hassell, Stiles and Reuning2022). Even if elite attacks resonate, local media have close ties to the communities that they cover and represent important aspects of local identity (Matthews Reference Matthews2022; but see Hopkins and Gorton Reference Hopkins and Gorton2024). This could make people—particularly those most familiar with a local news outlet—rather forgiving when politicians criticize local news. Indeed, whereas previous research has found limited reputational consequences for politicians who attack the media (Egelhofer et al. Reference Egelhofer, Boyer, Lecheler and Aaldering2022), the continued popularity of local news makes it plausible that politicians may face backlash for attacking local outlets.
Alternatively, local news may have largely avoided the criticism that erodes trust in other media. That is, if they were targeted by politicians, the reputations of local news organizations may be no more durable than those of other news outlets. Previous studies have suggested this dynamic in their focus on elite messages about national news (Ladd Reference Ladd2010; Meeks Reference Meeks2020; but see Smith Reference Smith2010) and critiques of the media in general—a concept the public rarely associates with local news (Ladd Reference Ladd2012, ch. 4).
We buttress the claim that local news is criticized less frequently than other media with an original content analysis (see online appendix B) of how Republican and Democratic politicians discussed news sources in congressional e-newsletters from 2009 to 2022 (Cormack Reference Cormack2017). Similar to research that focused on presidents and presidential candidates (Meeks Reference Meeks2020; Watts et al. Reference Watts, Domke, Shah and Fan1999), we first found that Republican legislators attack the media more frequently (19% of media mentions) than their Democratic counterparts (3%). Further, negative media mentions almost always refer to the general concept of “the media” or, for Republicans, a national news outlet (e.g., the New York Times)—not local news. These findings show that Republican politicians beyond presidents are also more likely to attack the news media than Democratic politicians. This also raises the possibility that if attacked to the same degree as national media, local news reputations may not prove to be any more durable.
Given Republicans’ greater tendency to attack the news media, we next assessed the public’s response to GOP politicians’ criticism of local news to determine whether such outlets can withstand elite attacks. This consideration offers a valuable opportunity to assess elite influence on how the public views the media in a setting where such messaging may be ineffective. Further, unlike national news, we consider elite media attacks that need not contend with extensive public pretreatment to similar messages. This assessment also has forward-looking relevance. Although the supply of elite criticism of local news has been low relative to national news thus far, there are indications of a turn against local news in recent elections. Republican candidates have circumvented local news organizations at press conferences and during their campaigns (Kurtzleben Reference Kurtzleben2022). They also have criticized local news in a manner similar to national media—for example, in former Congressman Devin Nunes’s long-running feud and defamation lawsuit against the Fresno Bee (Ulloa Reference Ulloa2018) (see online appendix B for additional examples). Finally, Trump’s harsh rhetoric about journalists (Meeks Reference Meeks2020) provides strong incentives for conservative politicians to attack local news outlets.
A SURVEY EXPERIMENT STUDYING GOP CRITICISM OF LOCAL NEWS
We used a survey experiment to examine the effects of elite attacks on local news among 3,653 survey respondents: 1,521 from Prolific and 2,132 from CloudResearch Prime Panels (Archer and Peterson Reference Archer and Peterson2024).Footnote 2 The Prolific study was quota sampled to census benchmarks for age, sex, and race/ethnicity; the CloudResearch sample used quotas to ensure equal numbers of Democrats and Republicans. Our expectations and analytic decisions were established in a pre-analysis plan (see https://aspredicted.org/h9dx7.pdf). More detail about the sample is in online appendix A.
All respondents read a press release by a Republican politician from their state. Again, we focused on Republican elites because they attack the news media more than Democrats, as reflected in our content analysis and in previous literature (Watts et al. Reference Watts, Domke, Shah and Fan1999). Further, the local news trust advantage is attributable largely to Republicans trusting these outlets more than other types of media. Our experiment tested whether this trust can endure elite rhetoric that has reduced Republicans’ trust in other media.
Respondents were randomly assigned into a treatment or a control condition. Those in the control condition saw a press release in which the politician discussed the importance of registering to vote. In the treatment condition, the politician attacked a local media outlet and argued that it published biased coverage, ignored important campaign issues, and endorsed candidates who were bad for the communities they represented. (Examples of real-world messages that informed this treatment wording are in online appendix B.) To make the scenario plausible, the politician and local news outlet varied based on the state in which the respondents resided. The local news source was a large newspaper from the respondent’s state.Footnote 3 The Republican politicians, generally current officeholders or recent candidates for statewide office, also were from their state. They were selected based on how they discussed local or national media in previous communications so that it was realistic for them to attack the local newspaper. For example, respondents from Wisconsin in the treatment condition saw an attack on the Milwaukee Journal Sentinel by Senator Ron Johnson, an outlet that he had previously attacked during his 2022 Senate race.Footnote 4 (See online appendix A for a list of the news sources and politicians selected for each state.)
We used this design to test how Republican attacks on local news affect evaluations of local news sources and the politician making the attack. Our first hypothesis is that these attacks will harm the reputations of and future intent to use the local news outlet (Hypothesis 1). Further, we anticipate that the effects should be stronger for Republicans than for Democrats and Independents due to shared partisanship with the politician making the attack (Hypothesis 1a). This expectation drew from prior research on elite criticism of national, institutional media (Ladd Reference Ladd2010) that uncovers larger effects among copartisan leaders and supporters. We also anticipate that the effects should be weaker for those most familiar with the local news outlet (Hypothesis 1b). Prior experience with a local news outlet should “soften the blow” of an elite attack because an established relationship with a local news source might make individuals less receptive to attacks relative to those who lack such relationships with the outlet. Moving beyond news outlet reputations, the attack should affect views of the politician targeting the local news outlet—again with heterogeneous consequences by political party. Given Republicans’ greater general wariness of the press, they should reward politicians who attack the media more than their Democratic and Independent counterparts, who instead should react negatively to criticism of a trusted local news source (Hypothesis 2).Footnote 5
EFFECTS OF AN ATTACK ON LOCAL NEWS
We first consider the attack’s effects on the local news outlet. The first outcome variable is the newspaper’s reputation, measured using a seven-item index that consists of a rating of the news outlet’s ideology, a feeling thermometer focused on the local news source, and scales asking how appropriate various terms (e.g., “trustworthy”) were for describing the local newspaper (adapted from Egelhofer et al. Reference Egelhofer, Boyer, Lecheler and Aaldering2022). The second outcome variable is intent to use the local newspaper, measured using an index of the respondent’s interest in subscribing to the newspaper, their reported likelihood of reading the newspaper in the next week, and their willingness to select the local newspaper when prompted to search for political information about their state. We combined these items into a scale using principal components analysis and standardized the resulting index to have a mean of zero and a standard deviation of 1. To increase the precision of the effect estimates, we also controlled for a pretreatment measure relevant to the outcome under consideration (i.e., either newspaper reputation or use) when estimating the treatment effects.
The first two columns of table 1 display estimates of the effects of the attack on these outcomes. The attack had a negative effect on the local outlet’s reputation, reducing it by 0.3 standard deviations relative to the control group. Although the effect size is smaller when moving from source trust to source use, the attack also reduced future intent to use the news outlet by a 0.1 standard deviation. In both cases, these effects were statistically significant.
Effects of an Attack on a Local Newspaper
The third and fourth columns in table 1 examine whether the effects of the attack vary by party. The statistically significant interaction between the attack treatment and Republican partisanship for local newspaper trust in the third column shows that this heterogeneity exists. Among Democrats and Independents, the attack reduced trust in the local newspaper by a 0.1 standard deviation (95% confidence interval [-0.2, -0.1]). Among Republicans, the marginal effect of the attack is far larger, producing a 0.8-standard-deviation decline in the newspaper’s reputation (95% confidence interval [-0.9, -0.7]). This demonstrates that their initially positive views of local news can be dramatically reduced by exposure to criticism. Notably, when we examined only the feeling-thermometer component of the index, the attack generated a 16-point decrease among Republicans—an effect size similar to that found by previous research among copartisan followers of a politician attacking the media writ large (Ladd Reference Ladd2010). This suggests that local media reputations are as vulnerable to attacks as other news outlets.
The fourth column in table 1 also reveals partisan heterogeneity when news use is the dependent variable; however, as in the main effects, the magnitude of the treatment effect is smaller when moving from source reputation to source use. Democrats and Independents do not adjust their future intent to use the local newspaper based on the attack, as indicated by the small and insignificant coefficient on the treatment indicator (0.01, 95% confidence interval [-0.1, 0.1]). The negative, statistically significant interaction among Republicans indicates that the attack was more effective for this group. Ultimately, the marginal effect of the attack on intent to use local news among Republicans also was negative and statistically significant (-0.2 standard deviations, 95% confidence interval [-0.3, -0.1]). This indicates that elite rhetoric can expand the partisan divide in preferred information sources about local politics.
To provide additional context for the magnitude of these effects, we used post-treatment feeling thermometers where respondents rated the local newspaper and also evaluated CNN, a national news source with lower trust and a more divisive, crystallized reputation. In the experiment’s control group, local newspapers received an average evaluation of 56 points on the feeling thermometer—more favorable than CNN’s average rating of 50. However, in the attack treatment condition, ratings of the local newspaper decreased to 49 points and were no longer distinguishable from CNN. In addition to overall trust, figure 1 displays the partisan divide in views of these sources. It shows a 31-point difference in how Republicans and Democrats view CNN, indicative of sharp divisions toward national news outlets. In the control group, the partisan divide in views of local newspapers was only 13 points, but this increased to 26 points in the treatment condition. In terms of levels of overall trust and the partisan divide in how they are viewed, an attack from a Republican politician makes views of local news resemble assessments of national media.Footnote 6
Partisan Divide in Views of News Media
In online appendix C, we examine the consequences of attacking a local news outlet for Republican politicians themselves. The results suggest that Republican politicians face a reputational cost for critiquing local news because it reduces their support by a 0.1 standard deviation on an approval index. In support of our expectations, this response was concentrated among Democrats and Independents. However, against expectations, rank-and-file Republicans neither reward nor punish Republican elites for attacking local news (see online appendix table C2).
HETEROGENEOUS EFFECTS BY FAMILIARITY
To assess the boundaries of the effects for media trust and use, we examined familiarity with the news outlet as a moderator. As outlined in Hypothesis 1b, we expected that prior experience with a local news outlet would “soften the blow” of an elite attack. We constructed a measure of familiarity based on an index of the strength of a respondent’s pretreatment views of a news source and the pretreatment measure of local news use. This variable was standardized to have a mean of zero and a standard deviation of 1. We estimated regression models with trust and use intent as the outcomes and interacted the attack treatment condition with news-source familiarity. Figure 2 displays the marginal effect of the attack across levels of news-source familiarity. At the bottom of figures 2a and 2b the distribution of news-source familiarity is shown (see online appendix table G2 for the corresponding regression table).
Effects of Attack by Local News Source Familiarity
(a) Dependent Variable: News Source Trust (b) Dependent Variable: News Source Use
We found evidence consistent with our expectations. For media trust, the attack on the newspaper is less effective among those who have high levels of familiarity with local news compared to those with less familiarity. However, whether at the minimum (-0.4-standard-deviation effect, 95% confidence interval [-0.5, -0.3]) or maximum (-0.2-standard-deviation effect, 95% confidence interval [-0.4,-0.1]) level of news-source familiarity, the attack reduces trust in the local news source.
The results for news use were not consistent with our expectations because the attack was more effective at reducing future intent to use the news organization among those who had a high level of familiarity. Here, the attack did not produce a detectable decline in intention to use the news source for those who were at low levels of news source awareness; the attack only began to produce a detectable decline in intention to use the local news outlet for those with higher levels of awareness (above -0.3, approximately the 40th percentile of the awareness variable). This was an unexpected pattern that did not align with the notion that this group would be less responsive to elite criticism than others.
To summarize, this examination of familiarity as a moderator provided little indication that the unique relationship between a local news organization and its community can offset the reputational costs of Republican elite criticism. This also suggests that the treatment effect is not simply capturing respondents reporting negative evaluations of an unfamiliar newspaper. Instead, even those who seemingly had a relationship with the news outlet downgraded their view of it in response to critical elite rhetoric. Similar results held when we considered those who reside in the physical circulation area of the newspaper (see online appendix D). Thus, familiarity and physical proximity are unable to insulate local news outlets from the negative effects of a Republican politician’s attack.Footnote 7
DISCUSSION AND CONCLUSION
Although local news has struggled economically, one bright spot has been its continued credibility among a broad swath of the American public. In the words of a former CBS News executive, local news “can’t be demonized as fake news. If there’s a traffic light broken at Elm and Maple, people know it, and there are no alternative facts. Americans are having trouble finding common ground, but in a local market, they have it” (Grynbaum et al. Reference Grynbaum, Koblin, Mullen and Roberston2024). Our study suggests that such common ground is easily undermined: trust in local news rapidly dissipates when it is attacked by Republican politicians—a finding with important implications for political communication.
First and foremost, this experiment reveals the fragile nature of broad public trust in information providers. Local news organizations have close ties to the areas they cover and they have retained public trust longer than other media; yet, elite rhetoric is still effective at reducing their perceived credibility. The effectiveness of Republican attacks on news outlets with local ties and moderate coverage profiles suggests limits on these source-level attributes as shields against elite criticism. Our findings instead center the presence or absence of elite criticism as the critical factor shaping public consensus about news credibility.
Local news organizations have close ties to the areas they cover and they have retained public trust longer than other media; yet, elite rhetoric is still effective at reducing their perceived credibility. The effectiveness of Republican attacks on news outlets with local ties and moderate coverage profiles suggests limits on these source-level attributes as shields against elite criticism.
A limitation of our study is that we cannot directly speak to the effects of a Democrat criticizing local news—a growing possibility as new trends in local television ownership have tilted some outlets rightward (Martin and McCrain Reference Martin and McCrain2019). Our findings shed light on what might happen in this case, however. Given that we found that even groups that do not share a party label with the attacking politician downgrade their views of the news outlet, one expectation is that Democratic criticism also would lower levels of trust in local news among the general public. As in previous research on attacks by politicians on national news outlets (Ladd Reference Ladd2012), our study also uncovered more significant effects among rank-and-file voters who share the same partisan affiliation as the politician delivering the attack. This suggests that attacks on local news by Democratic politicians would resonate more with Democratic partisans.
While this is the most straightforward extrapolation of our findings to a setting in which Democrats attack local news, future work on this topic is needed. It is possible that a Democratic attack on local media could resonate strongly with Republican voters’ preexisting anti-media sentiment and not resonate with Democratic voters who generally trust the media. In that case, the reputational costs for a Democratic politician also could be larger among Democrats than the reputational costs for a Republican politician observed in this study. Democratic criticism could also backfire and increase views of local news among Republicans due to negative partisanship against the opposing politician (Abramowitz and Webster Reference Abramowitz and Webster2018). This did not occur among out-partisans in our study but it has appeared in studies of elite attacks on national media and their effect on news consumption (Archer Reference Archer2023).Footnote 8
Ultimately, our findings offer a cautionary tale by suggesting that the trust advantage local news organizations currently experience will disappear quickly if negative interactions between Republican politicians and local news outlets continue to escalate (Kurtzleben Reference Kurtzleben2022) (see also online appendix B). Thus far, we suspect that these attacks are not more prevalent because they impose modest reputational costs on the politicians who make them (see online appendix table C2). This reputational penalty is concentrated among those who do not share the politician’s party label, which may deter politicians who want to appeal to general-election constituencies from attacking local news. Other factors not incorporated into our study—such as the dependence of many politicians on local media sources to convey their accomplishments to constituents—also may help to check the emergence of elite criticism of local news. Future work should examine how politicians navigate these tradeoffs. As our findings show, how elites discuss local news has crucial implications for one of the few remaining categories of trusted news in the United States.
[O]ur findings offer a cautionary tale by suggesting that the trust advantage local news organizations currently experience will disappear quickly if negative interactions between Republican politicians and local news outlets continue to escalate.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit http://doi.org/10.1017/S1049096524000775.
ACKNOWLEDGMENTS
The authors thank Dustin Carnahan and attendees at the 2023 Annual Meeting of the Midwest Political Science Association and the Political Behavior Workshop at Aarhus University for helpful comments. We also thank EunHye Grace Cho, Sho Izumisawa, and Kashif Naqvi for research assistance.
DATA AVAILABILITY STATEMENT
Research documentation and data that support the findings of this study are openly available at the PS: Political Science & Politics Harvard Dataverse at https://doi.org/10.7910/DVN/JSICOQ.
CONFLICTS OF INTEREST
The authors declare that there are no ethical issues or conflicts of interest in this research.
