Opinion Leadership in Rural America: How Local Influencers Shape Political Discourse and Partisanship

Politics in the Context of Rural America

Scholars have long understood that rural America differs from suburban and urban areas in numerous ways. In addition to demographic and population differences with suburban and urban areas (Albrecht 2022; Henderson 2021; Lyons and Utych 2023), there are also significant cultural, social, and economic differences (Albrecht 2022; Cramer 2016; Gimpel and Karnes 2006; Henderson 2021; McKee 2008). Increasingly, there have been political differences as well—a pattern known as the rural–urban divide (Brown and Mettler 2023; Cramer 2016; Gimpel et al. 2020; Scala and Johnson 2017).

It is important to recognize that rural areas across the United States are far from homogeneous. Significant variation exists in demographics, dominant industries, economic conditions, and cultural influences from one rural area to another. Yet despite these differences, residents of rural America have become increasingly aligned with the Republican Party since the late 1990s (Brown and Mettler 2023; Gimpel and Karnes 2006; Scala and Johnson 2017). According to the Pew Research Center (Nadeem 2024), the Republican Party held a 25-percentage-point advantage in voter identification among rural Americans in 2024—up sharply from just six points in 2000.

Understanding the dynamics of politics in rural areas is important because even though people in these areas make up only 19 percent of the U.S. population (United States Census Bureau 2017), states that are largely rural have an outsized influence in presidential elections through the Electoral College and in legislative deliberations through their Senate representation (Brown and Mettler 2023).

Rural Political Identity

Rural political identity refers to a distinct social identity shaped by place-based consciousness, group values, and perceived differences from urban communities (Cramer 2016; Jacobs and Shea 2023; Lyons and Utych 2023). Scholars have shown that rural identity involves more than simply living in a rural area. Instead, it arises from shared narratives of self-sufficiency, hard work, and traditions, as well as a belief that rural interests are neglected by urban elites and policymakers (Cramer 2016; Diamond 2021; Wuthnow 2017). Cramer’s (2016) concept of “rural consciousness” highlights how rural residents view themselves as fundamentally different from urbanites in values and lifestyles. It also revealed that many rural residents feel they do not receive their fair share of resources, despite evidence to the contrary.

Recent research demonstrates that subjective identification with being rural predicts political attitudes even beyond objective residence, suggesting a powerful psychological dimension to this identity (Lyons and Utych 2023). Rural identifiers are more likely to exhibit skepticism toward government, favor local governance, and express group resentment, contributing to contemporary rural–urban divides in attitudes and partisanship (Breitensteinet al. 2025). Social theorists further note that these identities are reproduced and reinforced through local narratives, cultural practices, and shared experiences of perceived disadvantage (Ching and Creed 1997; Hochschild 2016). In all, rural identity is a robust explanatory factor for political values and behaviors in the United States, providing crucial context for understanding rural political engagement and divides (Cramer 2016; Lyons and Utych 2023; Wuthnow 2017).

Political Engagement

Political engagement in the United States encompasses a spectrum of attitudes and behaviors that go beyond mere voting, ranging from party identification to deeper involvement in civic and political life.

Partisanship has long been considered the bedrock of American political behavior. Campbell et al. (1960) introduced the idea of party identification as a “psychological attachment” that powerfully predicts voting and participation, an idea later extended to consider the expressive and emotional dimensions of party identity (Huddy and Bankert 2017; Huddy, Mason, and Aarøe 2015). More recently, affective polarization has received scholarly attention, with citizens’ emotional reactions toward in- and out-parties shaping engagement even in the absence of strong ideological differences (Bankert, Huddy, and Rosema 2016; Iyengar, Sood, and Lelkes 2012; Mason 2018).

Another central form of engagement is “deep involvement” in politics, a phenomenon explored by Krupnikov and Ryan (2022). They argue that, contrary to common assumptions, the American electorate is divided not merely by party or ideology, but by levels of political involvement: the “deeply involved” actively seek, discuss, and share political information, often serving as key sources of influence within their communities. Deep involvement correlates with stronger expressions of partisanship and a tendency to display “rabble rousing” behavior—actively mobilizing others, seeking out conflict, and prioritizing political engagement as a central personal mission.

Scholarship also recognizes “negative partisanship”—hostility toward the opposing party—as an increasingly salient force in recent decades, fueling political participation driven more by animosity than by identification (Abramowitz and Webster 2016). These developments reflect a broader shift toward the personalization of engagement, social identity reinforcement, and the emotional intensification of political life (Mason 2018).

Contemporary research suggests that political engagement encompasses many overlapping forms. Together, these constructs underscore the centrality of social networks, emotional investments, and motivational differences in understanding not only why but also how Americans engage with politics—particularly in tightly bonded rural communities where the influence of hyper-involved actors is amplified.

Urban and rural residents often differ less in whether they engage politically than in how that engagement is expressed. Research suggests that rural residents are somewhat less likely to protest, likely because protest is more accessible in urban settings, but they are somewhat more likely to display campaign signs, which may reflect differences in housing form, yard space, and local norms of visibility (Lin and Lunz Trujillo 2023). More broadly, these findings suggest that place structures the channels and routines of participation, even when underlying political interest is similar.

Political Opinion Leadership and the Flow of Information

Building on the diverse ways individuals engage politically, opinion leadership represents a distinct and influential form of participation whereby highly engaged individuals act as key intermediaries shaping political information within their communities. In rural contexts, the influence of such figures may be especially pronounced, given close-knit social ties and the relative scarcity of alternative information channels (Wuthnow 2017).

The idea of individuals using political information they gather—usually from media—to shape the political opinions of others is not new. Lazarsfeld, Berlson, and Gaudet’s landmark work, The People's Choice: How the Voter Makes Up His Mind in a Presidential Campaign, Legacy Edition (1944), first described the limitations of media in directly influencing voting habits and illuminated the role of interpersonal communication. Later work expanded on its role, establishing a two-step flow theory, which proposed that mass media messages do not always directly influence the general public but rather reach opinion leaders first, who subsequently interpret and relay the information to others (Katz 1957; Katz and Lazarsfeld 1955).

This original foundational model has been extensively critiqued and expanded since its creation (Ognyanova 2017). Scholars criticized it for underestimating the direct effect of media, especially since the advent of the Internet and social media (Bennett and Manheim, 2006; Robinson 1976). Others found that the linear top-down model was oversimplified and that the flow of information was more complex, suggesting that information was also exchanged among opinion leaders and among the audience itself (Weimann 1982). Other work introduced the multidirectional flow, demonstrating the impact audiences can have on media coverage (Brosius and Weimann 1996).

Recent scholarship on network effects has illuminated how social connections within communities profoundly shape the dissemination and interpretation of political information. Rather than functioning as isolated nodes, individuals are embedded in dense networks where information, persuasion, and social pressure are intertwined (Huckfeldt and Sprague 1991; Rolfe 2014). These networks amplify both the reach and the credibility of locally salient narratives, as actors tend to trust and act upon information received through established social relationships (Sinclair 2012). As political information diffuses through rural networks, peer effects and collective discussion reinforce dominant attitudes, sometimes exerting greater influence than formal media sources (Ames et al. 2016; Nickerson 2008). In the rural context, physical proximity, repeated interaction, and overlapping social affiliations intensify these effects, encouraging conformity to local norms and further empowering opinion leaders. Empirical work also shows that such network effects foster a sense of shared identity and political efficacy, which may drive collective action and sustain partisan alignment even in the absence of robust institutional structures (Ball-Rokeach, Kim, and Matei 2001; Eubank et al. 2021).

Opinion leaders—those who shape the attitudes and beliefs of others (Ognyanova 2017)—can be found in a wide variety of contexts and with many different characteristics. One thing they have in common, however, is a high exposure to media (Katz 1957). In addition to media consumption patterns, Shah and Scheufele (2006) found that political opinion leadership is also characterized by political interest and personality strength. Opinion leaders are typically more politically involved, frequent information seekers, and active in both traditional and social media (Campus 2012; O’Cass and Pecotich 2005; Park 2013; Shah and Scheufele 2006; Winter and Neubaum 2016). Katz (1957) found that, in addition to individual characteristics, opinion leaders are also determined by their level of competence and their accessibility to access and mobilize their connections.

Research Questions

Existing scholarship suggests that rural partisanship cannot be understood through attitudinal predispositions alone. Scholarship on rural consciousness explains why many rural residents experience politics through place-based identity and perceived marginalization, while research on deep engagement and opinion leadership shows how highly involved actors can shape the political understandings of others. What remains less well specified is how these processes interact in rural communication environments characterized by limited media access, distrust of external institutions, repeated face-to-face interaction, and overlapping social networks. In such settings, opinion leaders may be especially consequential not because opinion leadership is unique to rural life, but because rural social structure can intensify the reach, credibility, and repetition of their interpretive work.

Despite the growing body of literature on opinion leadership, there remains a limited understanding of how these dynamics operate within the distinctive social and informational environments of rural communities. Rural opinion leaders are situated at the intersection of local culture, communication networks, and party politics, where their influence is shaped by the structure of community relationships and the limited availability of alternative information sources. This study aims to advance the field by investigating how opinion leaders interact with rural information environments and examining the processes by which they influence attitudes and reinforce partisan identities. With these aims in mind, the following research questions guide the inquiry: How do opinion leaders shape political attitudes and voting behavior in rural communities? What are the key features of rural information environments, and how do they influence the flow and interpretation of political information? And how does partisan alignment develop and persist in rural communities, particularly in contexts with widespread skepticism toward political parties?

Fieldwork

Fieldwork for this study was conducted in Garrett County, Texas, a sparsely populated county of 900 square miles in the state’s panhandle region. Agriculture—primarily cotton and cattle—is the county’s primary industry, and it lies within the economic and media influence of the distant regional urban centers of Amarillo (population 201,000) and Lubbock (population 260,000).

This county was chosen because it met the criteria for being rural according to the U.S. Census Bureau (Ratcliffe et al. 2016). The population was about 1,400 according to the 2020 U.S. Census (United States Census Bureau 2025). About 65 percent of the population is located in one of two pseudonymously named towns, Bralley, the county seat, and the somewhat smaller Red Canyon. The rest of the county’s residents live in the countryside among the numerous farms and ranches.

According to the United States Census Bureau (2025), the county was about 50 percent female, with slightly more than 25 percent of the population over the age of 65. About 70 percent of residents were white, 25 percent were Hispanic or Latino, and 3 percent were African American. Of the population, about 85 percent had at least a high school diploma, and less than 20 percent had a bachelor’s degree or higher. The median household income in 2020 was about $38,000, with almost 15 percent of the population living in poverty.

Analysis

After data were collected, I conducted an inductive thematic analysis as described by Braun and Clarke (2006, 2022), which emphasized the emergence of research findings from raw data in the form of themes. This began by importing interview transcripts and field notes into the qualitative analysis software ATLAS.ti and taking a first pass over the data over a few weeks. Next, I began a close reading, utilizing the software tools to highlight sections of data and give them a code based on their topic. Codes were both broad and specific and were descriptive of what was being said or observed (e.g., “Donald Trump,” “farming,” or “economy”). Other codes described emotions or attitudes (e.g., “anger,” “resentment,” “contentment”). I took a total of three passes of coding, with a few days between each.

Initial codes were then grouped based on conceptual similarities. This grouping of the codes moved them beyond simple descriptions and applied a deeper level of analysis, recognizing relationships using conceptual categories (Mihas 2023; Thomas 2006). From there, Braun and Clarke’s (2006) process required reviewing the themes by challenging them to make sure they were sound and supported by the data.

Isolation and the Information Ecology

As Roy said, “When you live in the middle of nowhere, everything is far away.” Access to goods and services that people in urban or suburban areas might take for granted required county residents to travel long distances, which they tended to measure in travel time rather than miles. Something as simple as going to a supermarket or a high school basketball game required advanced planning and a significant investment of time and money. There were a few local businesses, including gas stations, a few restaurants, bank branches, and a Dollar General in the county, but anything else that residents needed required a trip out of town. Medical care and other basic services required a drive of an hour or more. Students at the county’s schools traveled long distances to compete in regular-season athletic events—one Bralley High School football opponent during the 2022 season was a town 3 hours north near the border with Oklahoma.

In addition to being isolated by geography, participants also lived in a media environment that left many feeling like they were sometimes isolated from what was going on in the rest of the world. Like most Americans, Garrett County residents consumed news media in traditional forms such as television, radio, and newspapers, as well as digital forms like websites, social media, and podcasts. Their consumption habits, however, might have differed from those of their urban counterparts due to both access issues in their rural environment and their personal choices.

Garrett County was within the coverage area of a Gannett-owned daily newspaper in one of the Panhandle region’s large cities, but the paper was not widely read because it could only be acquired by subscription and arrived by mail a few days after publication. A local newspaper focused exclusively on local news. Radio was also a source of news, and a handful of stations served the county, but signal quality and reception varied widely by location and other conditions. Traditional television consumption was low because most of the county lacked a traditional cable provider, over-the-air signals were too weak for reception, and satellite television was not particularly appealing to many because of issues with installation and service so far from a major population area. Jack thought accessing television content was just too much of a hassle:

Our TV hadn’t been on in nearly three weeks, and we don’t watch the TV because we lost our [service] . . . And so, you got all the news stations and now we’re on a Roku thing and I don’t have a six-year-old to run it, if that makes sense. It’s like, “Put it down, give me a book.” We don’t miss it.

Like other areas of rural America (Federal Communications Commission 2024), Garrett County lagged behind the rest of the country in broadband internet access, although much of the county’s population could access the internet with sufficient speeds to accommodate regular browsing and streaming. The internet seemed to be the primary source of news and information, especially political information, for most participants, who described mostly passive consumption from news websites (like FoxNews.com) through their mobile devices in small amounts throughout the day.

Most people I encountered would only sheepishly admit to using social media as a source of information. Participants largely viewed social media skeptically, and some saw it as a threat. Facebook was the primary social media platform used by participants, with X and Instagram used to a lesser extent. With few exceptions, descriptions of their Facebook use (some even volunteered to show me their Facebook feeds) revealed that their feeds and activity were mainly of a personal nature and were used to connect with friends and family and to keep track of things happening in their community. Most, but not all, participants said they were skeptical of political content on social media and approached it with caution. “People can and do say anything on Facebook, and none of it may be true,” Roy said.

Even for those who intentionally maintained access to news media, trust in news organizations was low, regardless of their perceived partisan bent. Much of the antipathy stemmed from their view that media outlets were only out to influence them and, as Martha said, would “say anything just to get us to be quiet.” Their disdain was most intense for what they described as liberal news outlets, which was consistent with research about conservatives’ lack of trust in news media (Nadler and Taussig 2022). Interestingly, not all of their ire was focused on so-called liberal news outlets. Those they considered conservative were not spared from their assessment that they, as rural dwellers, were perceived as gullible rubes by those in urban areas where news content was produced. Whether it was MSNBC or Fox News, participants felt like media outlets in urban areas thought they were less intelligent and blamed them for the rise in misinformation. Martha elaborated on this feeling:

Yeah, we’re just a bunch of dumb hicks out here who can’t tell up from down. People in California say that if any old idiot who says he is a Republican tells us something we are just going to believe it and do what he says, okay. So, maybe everything isn’t true. I don’t really know. But neither do they. But we’re the only ones that can be hoodwinked? I don’t think so.

With their general lack of access to media and their distrust or disdain for the media they could access, I was curious about their knowledge of current events. Indeed, some participants expressed concern about their isolation and worried that they and their fellow citizens were disconnected from what is happening in the wider world. “I mean, gosh, we’re way out here, so far from where things are going on and maybe living in our own little bubble,” Terri said. During my 2 months in the field, however, I found that almost all participants, regardless of their news media consumption habits, were well informed about national current events. Often, by the time I arrived in the county in the mornings, they knew more about what was happening in the world that day than I did, but were unable to tell me where they got their information. I began asking participants to recount their day so that I could see where they gained news information. Sometimes this led a participant to recall seeing something on their phone (via social media or an alert from the Apple News app, which most did not know how to turn off), even though they had not deliberately sought it out. Occasionally, they had read it in a newspaper, heard it on their car radio, or learned it through some other incidental exposure to news media. But more often than not, they heard it from someone else.

A stop at one of the local spots unofficially designated for the consumption of copious amounts of coffee and conversation was built into many people’s daily routines. Informal groups gathered at the gas station in Bralley and the feed store in Red Canyon starting as early as 7:30 a.m. Others met at various times throughout the day at an implement store and a bank. These meetings each had regulars, although attendance shifted, and people also spent time at each group. At each place, information of all sorts was exchanged. I heard conversations about everything from the cryptocurrency markets to high school football scores to speculation about the length of the drought to developments about the war in Ukraine. Politics was a frequent topic of discussion. Chuck described how these groups worked:

Yeah, anywhere you go . . . they’re going to talk politics. I mean, it’s an every-morning deal at the coffee shop, at the filling station, or [the tire shop], either one. There’s going to be politics. Most of the time, they’ll talk about anything.

Other opportunities for information exchange were woven into day-to-day life, where, because of a lack of anonymity and the tight social fabric, there was a social element to even the most basic of routines. Picking up a child at school involved seeing most of one’s neighbors. A stop at the local post office to pick up mail (the town was too small for door-to-door mail delivery) usually involved conversation and sharing of information.

Outsized Role of Opinion Leaders

The ethnographic data collected in Garrett County made it clear that a small group of highly engaged individuals played a pivotal role in shaping political knowledge, discourse, and participation. By tracing news from one person to another as best I could, and through direct observation, it became apparent that much of the community’s knowledge of national political news originated from opinion leaders.

In this study, opinion leaders were identified through a combination of repeated participant attribution, direct observation, and self-description rather than through a formal network instrument. Individuals were treated as opinion leaders when multiple forms of evidence converged: they were repeatedly named by others as people who “knew what was going on” or regularly introduced political information into conversation; they were directly observed initiating or steering political discussion across multiple settings; and interview data indicated that they consumed political news more intentionally or more frequently than many other residents. This approach does not claim to measure influence with precision, but it does allow for the identification of actors who occupied visibly central roles in the community’s everyday circulation and interpretation of political information.

Generally, these people self-reported that they still watched or streamed television news content when others in the community did not, they read newspapers regularly, spent time reading news online, and had a deep interest in politics. They were interested in consuming news media and were happy to share it with others, who were usually happy to listen.

This type of information flow was in line with Katz and Lazarsfeld’s (1955) model of a two-step flow of communication, and therefore not surprising. Opinion leaders consumed news and spread it to those in their networks. It was possible, however, that Garrett County’s rural nature—its size, isolation, and community interdependence—magnified the impact opinion leaders had on others, especially in a multi-step flow of communication. When information spread from opinion leaders, it was shared over and over at coffee meetings, post-office conversations, church gatherings, and the myriad other social interactions of small-town life. Thus, it did not take long for it to be fully disseminated in such a small community. Furthermore, the isolation from news media that many participants described experiencing may have made them more likely to listen to news information shared by news distributors since they knew they were less likely to hear news on their own, as illustrated by Carla’s description of her news consumption:

Well, you know, I don’t spend a lot of time watching [the news] these days. There hasn’t been a newspaper here since Moses. I sometimes see things online. My husband listens to the radio, mostly when he’s on the tractor, and he tells me things. And we hear things too. Someone always seems to know what’s going on so I can keep up pretty good that way.

Most of the people whom I could identify as being the key news distributors were people who could be considered “deeply engaged” in politics (Krupnikov and Ryan 2022). They were interested in politics, saw themselves as key political actors in their networks, felt a duty to engage politically and inform their fellow citizens, and were generally comfortable with any conflict that arose because of their engagement. To be clear, not all of the people who were news distributors were necessarily deeply engaged in politics, but if one were to create a Venn diagram of those who were deeply engaged and those who served as the community’s news distributors, there would be a significant overlap.

Furthermore, those I observed or interviewed who were both deeply engaged in politics and news distributors were all enthusiastic supporters of Donald Trump. Some, like Ronald, were formally active in the Republican Party at the local or state level. “I’ve been a Republican for so long, before it was the thing to be around here,” Ronald said of his party involvement. Others, like Dale, had no formal involvement other than voting. All seemed to see their activities spreading information as part of their duty—both to their community and to Donald Trump. Dale expressed a sense of responsibility for what he was doing:

Nothing is going to stop me from telling the people who don’t pay a lick of attention what is really going on. Quite frankly, some people just think the way they do because they don’t know what is going on anywhere else. I mean, if they knew what was happening to our country out there and even what Trump has gone through for this country, fighting for all of us, if they really knew . . . maybe they would fight a little harder.

At times, the framing presented by opinion leaders traveled in addition with the information—or misinformation—they shared. Early one morning at the feed store in Red Canyon, Carl, an opinion leader and self-proclaimed “news junkie,” relayed information to the five men and one woman gathered for coffee about the negotiations for the release of WNBA player Brittney Griner, who was then detained in Russia. Carl’s information was gleaned, he said, from watching Fox News overnight and “reading about it on the net,” and mostly factual, with the addition of his opinion that “she was probably doing more than just sneaking in some hash. She was smuggling coke for sure. You know they all do it.” He also said, “Biden would cave. Such a weak f***er. He’ll bend over backward just to get a coke addict back.” Later that evening, I was present for a conversation before the start of the local high school football game, where the Griner situation was brought up by someone who had been present for Carl’s morning briefing. He referred to Griner as “that basketball player who snuck coke into Russia” and said Biden “was planning to cave.” While it is possible that this information came from elsewhere, the specific misinformation about the drug and the framing of Biden’s weakness may not have been a coincidence.

While the data collected suggest that much of the community’s access to information came through deeply engaged opinion leaders, it was clear that people were not blindly following the opinions of others. In fact, the small size of their community meant they had an extreme familiarity with each other’s strengths, weaknesses, backgrounds, and faults, which gave them context to judge the information they received from one another. Albert explained it to me with the following hypothetical example (paraphrased for clarity):

If Fred comes to the bank coffee group one morning with news about an upcoming Environmental Protection Agency (EPA) regulation that will negatively impact a farmer’s water rights, the other members of the group will not automatically assume that everything Fred tells them is true. Instead, they will recall that one time, twenty years ago, Fred had a citation from the EPA for illegal dumping. They may also know a thing or two about Fred’s source of information. Most likely, they also keep an eye out on agriculture regulatory developments themselves. They will filter the information through their past experiences with Fred and their knowledge of his biases, the EPA, and the agriculture industry, and, with further investigation, form their own opinion.

Chuck, who was a regular at the gas station in the mornings, described how he would sometimes look up news articles on his phone when people mentioned a current event because, “You just can’t believe everything you hear down there.”

But even for the most skeptical residents of Garrett County, like Chuck and Albert, there was still a significant risk that they were influenced by opinion leaders because of what the opinion leaders chose not to share. People like George, who estimated that he got about 80 percent of his news information from other people because he no longer watched the news on television, were often only exposed to news that was selected and shared by opinion leaders.

Rural Opinion Leaders Reinforce Partisanship

Like much of rural Texas, Garrett County voted heavily Republican, but that wasn’t always the case. In fact, the Republican candidate William Henry Harrison failed to collect even one of the county’s 152 votes during its very first election in 1892. With few exceptions, the county remained staunchly loyal to the Democratic Party at the local, state, and national levels for almost a century, until Garrett County began mirroring the national trend of Southern white conservatives gradually realigning their political affiliation to the Republican Party (Knuckey 2001). Since Bill Clinton became the last Democratic presidential candidate to win Garrett County in 1992, the percentage of the vote cast for the Democratic presidential candidate fell precipitously with every election. In the 2020 presidential election, Joe Biden’s percentage of the vote total just barely reached 10 percent. The partisan realignment from the Democratic Party to the Republican Party was also strong and swift in other federal, state, and local elections.

Partisanship, as described in the literature, is clearly a factor in politics in Garrett County. At first glance, it would be easy to agree with local Republican Party leader Ronald’s assessment that “Garrett County is dyed in the wool Republican.” But it became clear during observation and interviews that although Republican candidates consistently earned 90 percent or more of the vote in county elections, that did not necessarily reflect enthusiasm or even loyalty to the party. In fact, when asking participants to reflect on specific recent elections, there were many instances—mostly in federal elections—where participants mentioned they would have preferred a choice other than the two parties. Many residents believed strongly that the Democratic Party did not represent them, but that didn’t automatically mean they thought Republican candidates did. “I mean, a politician is a politician. [Republicans] don’t care about us [rural people] any more than the Dems,” Albert, a father in his 40s, said.

Albert’s thoughts were common. When describing their vote choice, participants were more likely to frame it in the context of the rural–urban divide than a partisan divide. While “San Francisco” or “The East Coast” were frequently used as metonyms for liberals or Democrats in conversation, Texas politics, which has been dominated by the Republican Party since the mid-1990s, was similarly defined along an urban/rural divide. To many participants, Interstate 35, which runs roughly north and south through Texas from Dallas to Laredo and connects major urban areas like Austin and San Antonio, served as a sort of dividing line in Texas, separating a rural “us” in the state’s west from the urban “them” to the east. All participants said a political figure’s position on rural issues was just as important to them as their party.

As an example, during a discussion at a coffee gathering at the bank in Bralley, a group of older male participants expressed their dislike of Texas Republican Senator Ted Cruz. One described him as “a bit arrogant, a bit of a jerk, and kind of slick.” Another thought he didn’t stand up for agriculture producers because he was from the “wrong side of I-35” and “just looking out for Houston.” Other Republican political figures faced similar vitriol.

While it was tempting to consider this as a possible sign of the softening influence of the Republican Party in rural areas, that was not the case. Although many residents voiced a deep reluctance to enthusiastically embrace the party, the cumulative effect of opinion leadership within key community networks ultimately stabilized and reinforced the region’s Republican alignment.

Opinion leaders in Garrett County functioned as the gatekeepers of political narratives. Through their selection, sharing, synthesis, and interpretation of news, they provided accessible explanations for choosing the Republican Party—and Donald Trump—while emphasizing practical outcomes and the perceived priorities of “rural people” over rigid ideological or partisan arguments. Regular gatherings in local banks, coffee groups, and feed stores highlight how these figures employ stories and analogies (e.g., invoking cultural or religious symbolism) to generate group consensus around the pragmatic logic of voting Republican as a defensive move against perceived “urban” values and the national Democratic platform.

Notably, these leaders seldom extolled the virtues of the Republican Party or its local representatives. Instead, the collective narrative—steered by opinion leaders—focused on preventing the worst outcomes associated with Democratic rule, such as hostile policies toward agriculture, energy, or gun rights. This negative partisanship became a rallying point: voting Republican was framed not as enthusiastic support, but as a necessary safeguard for the rural way of life.

Through repeated discussion, correction of misperceptions, and agenda-setting in local conversations, opinion leaders ensured Republican partisanship remained the default electoral option. During observed conversations, dissenting voices that question this alignment were often outnumbered or gently redirected back to the perceived necessity of Republican voting in a context of rural threat and marginalization. During a coffee meeting of women at the farm implement store, Anita, speaking of the January 6, 2021, attack on the U.S. Capitol, said Trump “just really broke our trust. We just don’t do that. And we don’t lie. It really just kind of stings your heart. I guess I thought he was going to be different, you know?” But she was chided by Rose, a known opinion leader, who reminded her that “Trump was better than the alternative.” Ronald, an opinion leader and local Republican Party official, told one person that Trump was “. . . just a modern King David. Is the guy perfect? No. But God uses imperfect people all the time and he is using Donald Trump to save America from evil, and from itself.”

Even those dissatisfied with both parties described votes for Republicans as a pragmatic, almost habitual choice, reinforced by the opinions and arguments of the most engaged individuals in the community. Political opinion leaders in Garrett County operated less as celebrants of Republican ideology and more as pragmatic gatekeepers who, through everyday acts of interpretation, caution, and consensus-building, stabilized regional Republican loyalty—often in spite of broad skepticism and reluctance among ordinary voters.

Theoretical Implications

Classical two-step flow theory posits a largely uniform information pathway starting with information shared by media, which is consumed by opinion leaders, who then share it with a general audience (Katz and Lazarsfeld 1955; Lazarsfeld et al. 1944). This study reveals that the efficacy of this model is contingent on environmental scarcity. In rural contexts with limited media access and low trust in mainstream outlets, opinion leaders’ influence is not merely channeled through networks but amplified through multi-step reinforcement. Rather than information flowing linearly through one or two intermediaries, rural opinion leaders’ messages circulate repeatedly through overlapping social spaces—coffee shops, post offices, churches, and school events—creating a cycle of reinforcement. When George reported receiving 80 percent of his news directly from other people, and when most participants could not recall where they initially learned major news events, this revealed the depth of multi-step circulation. Importantly, in information-dense urban environments with more media access, competing media sources, and higher anonymity, the flow may remain largely two-step; in information-scarce rural environments, it becomes a multi-step amplification process where each repetition strengthens community consensus. This means opinion leaders in rural communities function less like neutral media filters and more like narrative gatekeepers whose interpretive frame becomes the community’s default.

Political communication scholarship has traditionally portrayed opinion leaders either as ideological persuaders or information distributors (O’Cass and Pecotich 2005; Shah and Scheufele 2006). This study identifies a distinct category: pragmatic gatekeepers who curate political information through community self-defense rather than ideological conviction or partisan enthusiasm. In Garrett County, opinion leaders did not necessarily advocate for formal Republican policy platforms or celebrate the Republican Party as an institution. Instead, they framed Republican alignment as essential to defending rural livelihoods, cultural autonomy, and political voice against perceived urban threats. Rose’s comment that “Trump was better than the alternative” despite acknowledging his moral failings, or Albert’s frustration that “politicians don’t care about us any more than the Dems,” revealed opinion leaders functioning as pragmatic strategists, not zealous advocates. This distinction matters theoretically because opinion leaders’ power lies not in ideological transmission but in narrative framing that links group identity to political necessity. By demonstrating that opinion leaders stabilized Republican partisan alignment despite widespread skepticism—and even reluctance—toward the party itself, this research reveals that opinion leadership may function differently in identity-defensive contexts (such as rural areas experiencing cultural marginalization) than in contexts with positive ideological attachments. This extends scholarship on affective polarization and negative partisanship (Mason 2018) by showing a specific mechanism through which opinion leaders maintain partisan loyalty through threat perception rather than positive party identification.

Existing scholarship identifies opinion leader characteristics as media consumption, political interest, and perceived competence (O’Cass and Pecotich 2005; Park 2013; Shah and Scheufele 2006). This study contributes an undertheorized dimension: interpersonal validation through deep social proximity. In rural communities with extreme familiarity and virtual impossibility of anonymity, community members evaluate opinion leaders’ claims through intimate knowledge of their backgrounds, past behaviors, and known biases. When Albert recalled how coffee group members would contextually evaluate Fred’s claims by remembering “one time twenty years ago Fred had a citation from the EPA for illegal dumping,” he was describing an interpersonal accountability mechanism specific to low-anonymity environments. This deep familiarity simultaneously strengthens and constrains opinion leaders’ influence. Their messages carry weight precisely because everyone knows their sources and biases; simultaneously, this knowledge prevents blind acceptance. The authority derives from relational credibility developed over years of sustained social interaction. This finding suggests opinion leadership operates along a spectrum contingent on community structure. In high-anonymity environments (urban areas, online communities), opinion leaders rely on visible institutional credibility; in low-anonymity environments (rural communities), they function through relational credibility earned through sustained presence and personal accountability. This enriches theories of information trust in an era of declining institutional confidence.

While the two-step and multi-step flow theory explain how information circulates and affective polarization scholarship explains why emotions intensify engagement, both offer limited insight into community-specific identity mechanisms that sustain partisan alignment despite skepticism toward parties themselves. This research demonstrates that opinion leaders functioned as custodians of rural identity narratives, continuously linking political choice to place-based consciousness and collective interests. Throughout fieldwork, opinion leaders repeatedly framed Republican alignment as essential to defending rural livelihoods and cultural autonomy. By repeatedly invoking rural identity—emphasizing rural–urban divides, invoking Interstate 35 as a symbolic boundary between “us” and “them”—opinion leaders transformed partisan choice from individual political preference into collective identity performance. This extends rural political consciousness scholarship (Cramer 2016; Lyons and Utych 2023) by revealing that place-based identity does not automatically translate into stable partisan behavior; rather, opinion leaders actively maintain, perform, and mobilize rural identity to sustain partisan loyalty. Their role was not merely informational but identity-performative: through everyday discourse in community gathering spaces, they continuously reinforced what it means to be rural and why rural identity necessarily entails a specific political alignment. Understanding partisan stability in rural contexts requires examining not only individual attitudes and media consumption but also the relational identity work through which community opinion leaders construct collective narratives.

Collectively, these contributions suggest a revised understanding of opinion leadership calibrated to rural contexts. Rather than viewing opinion leaders as neutral information intermediaries or partisan advocates, this study reveals them as identity custodians and pragmatic strategists operating within distinctive information ecologies characterized by scarcity, interdependence, and bounded social worlds. Their influence operates through four synergistic mechanisms: amplification (repetition across overlapping networks), pragmatic defense (framing political choice as necessity), relational credibility (interpersonal accountability), and identity performance (affirming collective belonging).

Limitations and Future Directions

As with all qualitative work, this study’s findings are context-specific. Ethnographic research in a single rural county offers deep insight into one community’s communicative life but limits generalizability. Garrett County’s demographic profile and agricultural economy may differ from those of other rural regions; future comparative studies could examine how opinion leadership varies across different economic and cultural settings.

Snowball sampling through gatekeepers for the interview portion of data collection may have created selection bias toward more politically engaged or visible community members. Furthermore, the small sample size for interviews and limited demographic diversity may not have captured the full range of community perspectives. Because the study relied, in part, on self-reported media use and observed interactions, some underreporting or recall bias is possible. Combining ethnographic methods with network or survey data could more precisely map how information circulates through rural communities.

The fieldwork also coincided with the 2022 midterm elections, when political salience was unusually high. Longitudinal research across non-election periods could help distinguish stable communication patterns from campaign-driven dynamics. As rural broadband expands, future research should explore how greater digital access may reshape local opinion leadership and media trust.