Religious rhetoric and political communication

Existing literature in the field focuses on the use of the “God talk” and “God strategy” by political elites (Knoll and Bolin Reference Knoll, Jo Bolin, Paul, Mark and Ted2020). “God talk”—that is, the use of religious rhetoric (Calfano and Djupe Reference Calfano and Djupe2009; Coe and Domke Reference Coe and Domke2006; Djupe and Calfano Reference Djupe and Calfano2013)—is used by both Republicans and Democrats alike (Balmer Reference Balmer2008; Chapp and Coe Reference Chapp and Coe2019). More specifically, “God strategy” refers to the use of religious rhetoric motivated by both sincere religious beliefs and strategic political goals (Coe and Chapp Reference Coe and Chapp2017; Domke and Coe Reference Domke and Coe2010). Most research on the use of “God talk” and “God strategy” by political elites focuses on the executive branch and the use of religious rhetoric by the president, as religious language is a consistent aspect of presidential speeches, used to signal to certain societal factions (Coe and Domke Reference Coe and Domke2006). In a content analysis of hundreds of presidential addresses, Hughes (Reference Hughes2019) found that presidents from Roosevelt to Trump often used religious language strategically, with its use increasing over time, culminating in Trump’s presidency, which featured the highest frequency of religious references. Similarly, Domke and Coe (Reference Domke and Coe2010) found that since Reagan, presidents have increasingly invoked God and faith, making direct references to God a more consistent and prominent part of their speeches. Audience played an important role in this rhetoric: presidents who had a higher proportion of mainline and evangelical Protestants in their electoral coalition were more likely to employ religious rhetoric in their major addresses (Kradel Reference Kradel2008).

A growing body of literature has also studied how members of Congress use religious language. In their analysis of 1.5 million tweets from members of Congress, Bramlett and Burge (Reference Bramlett and Burge2021) found that both Republican and Democratic members used a “religious code” on Twitter to indicate their own identity and activate their constituents’ religious identities, with Republicans using the code more extensively and with Judeo-Christian specific terms. Analyses of congressional websites and online biographies found that Republicans, ideological conservatives, Southerners, and legislators with more religious constituents were more likely to include religious references (Gade et al. Reference Gade, Dreier, Wilkerson and Washington2021). Legislators also used such language strategically, using religious rhetoric to co-opt issues that belong to the other party, such as Republicans discussing healthcare or Democrats addressing the economy and defense (Hughes Reference Hughes2019, Sides Reference Sides2006). Legislators’ religiosity may also impact their actions beyond just their rhetoric, influencing their voting behavior (Fastnow, Grant, and Rudolph Reference Fastnow, Grant and Rudolph1999; Guth 2014; Richardson and Fox Reference Richardson and Wightman Fox1972; Schecter Reference Schecter2001; Yamane and Oldmixon Reference Yamane and Oldmixon2011; McTague and Pearson-Merkowitz Reference McTague and Pearson-Merkowitz2013).

Political elites’ decision to use religious language has two main impacts. First, religious rhetoric may affect how voters view legislators. As electoral incentives heavily shape legislators’ priorities (Mayhew Reference Mayhew1974), legislators are motivated to craft a compelling “home style” that resonates with their constituents (Fenno Reference Fenno1978). This dynamic underscores the relationship between electoral imperatives and legislative behavior, helping to explain how legislators’ view of the partisan or religious makeup and preferences of their constituency could influence and alter their own actions. Because “members of the House of Representatives represent distinct and often religiously homogeneous districts” (Knoll and Bolin Reference Knoll, Jo Bolin, Paul, Mark and Ted2020 1126), candidates can use this language as a signal to certain voters. Such language may draw on religious stereotypes: Evangelicals are seen as conservative, competent, and trustworthy (McDermott Reference McDermott2009), attracting more Republican votes (Campbell, Green, and Layman Reference Campbell, Green and Layman2011). Religious language can enhance perceptions of trustworthiness and morality, as religious candidates, regardless of partisanship, are perceived as more trustworthy (Clifford and Gaskins Reference Clifford and Gaskins2016). However, overtly religious messages can alienate outside groups and can be more polarizing than subtle cues (Albertson Reference Albertson2015; McLaughlin and Wise Reference McLaughlin and Wise2014). For this reason, candidates may choose to utilize “multi-vocal appeals” that only in-group religious members can pick up on, also called “Dogwhistles” (Albertson Reference Albertson2015). Still, other studies note that religious cues may matter less to voters than secular information, such as ideology and race (Weber and Thornton Reference Weber and Thornton2012, Simas and Ozer Reference Simas and Ozer2017, McLaughlin and Thompson Reference McLaughlin and Thompson2016). Ultimately, legislators may strategically use religious rhetoric, adopting their presentational style to emphasize values their constituents support to influence their constituents’ view of them and the work they do (Grimmer Reference Grimmer2013).

Second, the use of religious rhetoric may have societal consequences. When legislators or candidates use religious language, they can inadvertently influence the religious socialization of the public and embed norms that political leaders are Christian (Coe and Domke Reference Coe and Domke2006). This can strengthen the civic religion seen in America today, further tying patriotism to religious faith. For example, the addition of “Under God” to the Pledge of Allegiance during the Cold War highlights America’s unique culture with shared values and customs based on the relationship between civic and religious life.

This civic religion in America can impact who is seen as a viable candidate for public office. In 2016, 62% of respondents to a Pew poll agreed that it was important that a president have strong religious beliefs (Smith Reference Smith2017), a sentiment that was still held by nearly half of Americans four years later (Bruce Reference Bruce2020). Voters are also much less likely to vote for a secular candidate (Castle et al. Reference Castle, Layman, Campbell and Green2017). The more candidates and legislators use religious language, the deeper entrenched civic religion becomes in America, and the more difficult it becomes for non-religious candidates to be elected into office. This dynamic can be further exacerbated by polarization when one party becomes closely entwined with religion. Since the election of Trump in 2016, more conservative Americans are identified as evangelical, even though many rarely, if ever, attend church and have no attachment to Protestant Christianity (Burge Reference Burge2021). This disparity suggests that for many, evangelical identity is now more of a political marker than a religious one, further entrenching the association between Republicanism and Christianity. As a result, religious rhetoric not only influences American elections but also shapes broader political culture and representative dynamics (Chapp Reference Chapp2012).

Leadership in congress

The role of the Speaker of the House is established in Article I, Section II of the United States Constitution. Since its inception, the position has come to perform both partisan and nonpartisan responsibilities, particularly in guiding the House to pass legislation (Green Reference Green2010). Today, the Speaker of the House is also responsible for “establishing rule-making precedents, evaluating points of order related to floor procedures, referring legislation to committees, and influencing legislators’ committee assignments” (Green Reference Green2010, 4). Like other members of Congress, the Speaker of the House is motivated by (re)election, in addition to ensuring their desired policy is passed and maintaining internal influence (Fenno Reference Fenno1973). However, while seeking re-election to the legislature, the Speaker of the House also seeks re-election to their position—“an office of great honor and influence” (Galloway Reference Galloway1955, 346).

The Speaker wields significant power over the legislative process by manipulating the legislative agenda, crafting bills, and strategically referring legislation to committees (Green Reference Green2010). One of the Speaker’s most potent tools is floor advocacy, where they speak on the House floor to advocate for specific legislative outcomes. This method, combined with other leadership tactics, leverages the Speaker’s authority and credibility to sway votes and influence legislative results (Green Reference Green2010). Additionally, the Speaker can affect outcomes by influencing committee assignments, which can alter the trajectory of legislation and the priorities of individual representatives (Green and Harris Reference Green and Harris2019). Scholars often measure the impact of party leadership by analyzing roll-call votes to determine how party alignment and leadership directives shape legislative behavior (Snyder and Groseclose Reference Snyder and Groseclose2000). Through these mechanisms, the Speaker of the House not only directs the flow of legislation but also molds the legislative behavior and decisions of House members, ensuring alignment with broader party goals and strategic objectives (Green Reference Green2010).

In addition to formal powers, the Speaker’s role as a symbolic leader is significant. Like other party leaders, the Speaker has the potential to influence the rhetoric of their followers. Party leaders, including presidents, often frame issues in ways beneficial to their party, influencing how rank-and-file members communicate with their constituents (Lee Reference Lee2016, Cavari Reference Cavari2017). Parties or political elites can have “brands,” which can set them apart as trustworthy, reputable, or holding a certain identity (Brubaker Reference Brubaker2020, Lees-Marshment Reference Lees-Marshment2014). Party members may adopt a political leader’s brand if they believe it will be electorally beneficial. For example, during the 2020 election many Republican candidates were more likely to embody the “Trump Political Brand” on Twitter, mimicking Trump’s communication style by things such as tweeting aggressively or tweeting in all-caps (Horan et al. Reference Horan, Brubaker, King and Meinhold2024). In the case of religious rhetoric, the Speaker’s personal religiosity and use of “God talk” (Coe and Domke Reference Coe and Domke2006) may serve as a signal to members that such language is both permissible and advantageous. This creates a “permission structure”—that is, the implicit and explicit cues that leaders provide, which encourage or dissuade members from adopting particular strategies or behaviors (Holland and Bohan Reference Holland and Bohan2013). For example, a Speaker’s public invocation of religious themes can encourage like-minded legislators to incorporate similar rhetoric, particularly when such language aligns with the Speaker’s ideological commitments and the broader priorities of their party.

Despite these insights, the causal mechanisms through which the Speaker influences rhetorical behavior remain under-explored. Much of the existing literature focuses on the Speaker’s procedural and policy influence, while research on their impact on rhetorical norms is still in its infancy. Green (Reference Green2010) notes that studies on the Speaker’s broader influence are relatively nascent, with significant gaps in understanding how leadership shapes communication styles within Congress. This gap is particularly evident in studies of religious rhetoric, where the relationship between leadership signaling and member behavior has yet to be fully theorized or empirically tested.

Case selection

The election of Mike Johnson as Speaker of the House in October 2023 provides a unique case for studying the intersection of religious rhetoric and political leadership in Congress. Johnson is the first Protestant Speaker in 16 years (Scanlon Reference Scanlon2023; Pinho Reference Pinho2023) and is vocal about how his religious beliefs are central to his political identity. When first running for Congress in 2016, he stated, “I’m a committed Christian and my faith informs everything I do” (Hensley Reference Hensley2023). Whitehead and Perry (Reference Whitehead and Perry2023) note that while Johnson has not explicitly identified himself as a Christian nationalist, his rhetoric and policy positions align with a broader Christian nationalist ethos, a framework advocating for the fusion of a particular Christian worldview with American civic life and governance.

Johnson’s ties to Christian nationalism have drawn scrutiny, particularly from members of the Congressional Freethought Caucus (CFC). In a report titled Speaker Johnson: Christian Nationalism in the Speaker’s Office?, the CFC asserted that “Speaker Johnson is deeply connected in political practice and philosophy to Christian Nationalism, more so than any other Speaker in American history” (Walker Reference Walker2024). The report further argues that Johnson has spent decades working to “deny, reject, and undermine the constitutional separation of church and state, including trafficking in fake histories about our nation’s founding.” This characterization was echoed by Rep. Jamie Raskin (D-MD), who stated on X that Johnson’s leadership “is what theocracy looks like” (2023). Republican legislators have also recognized how religious the speaker is; five months before Johnson assumed his position as Speaker, Rick Allen (R-GA) shared on the floor, “As my friend Mike Johnson explained to me, what our Founders did was simply this: They left God at the top, they got rid of the king, and put the people in charge” (169 Cong. Rec H1799, 2023).

Beyond his religious rhetoric, Johnson also enjoys the full endorsement of Donald Trump, further solidifying his position within the Republican Party’s ideological landscape. Ahead of the House vote to elect a Speaker in December 2024, then-President-elect Trump posted on Truth Social: “Speaker Mike Johnson is a good, hard-working, religious man…Mike has my Complete and Total Endorsement. MAGA!” (Brooks Reference Brooks2024). This endorsement, coupled with Johnson’s longstanding emphasis on religious values in governance, solidifies his distinct position among modern Speakers of the House. Even though past Speakers have been religious, few have utilized religion in a way Johnson has. Of the three past speakers, only two mentioned religion in their inaugural speeches (excluding the expected “God bless America”): Nancy Pelosi mentioned the importance of being “good stewards of God’s creation” (2019), while Paul Ryan asked legislators to pray for each other (2015). Representatives clearly saw this distinction; only two legislators wrote to their constituents about Pelosi or Ryan’s faith after they were elected. By contrast, 31 Republican representatives mentioned the Speaker Johnson’s religiosity in their newsletters to their constituents.Footnote ¹

While the influence of party leaders is not unique, Johnson’s ideological positioning, his explicit religious framing of policy—particularly as indicated in his inaugural speech as Speaker—and the discourse surrounding his speakership set him apart. As Speaker, he holds a critical leadership position within his party, and his ascension was rife with controversy and debate (Vazquez et al. Reference Vazquez, Mariana Alfaro, Theodoric Meyer, Goodwin and Wang2023), making him a particularly interesting figure in the eyes of the nation. The extent to which his leadership has shaped religious rhetoric in Congress presents a compelling case for examination, particularly given that party leadership has influenced legislators’ behavior in the past.

Hypotheses

We have demonstrated that congressional leaders can shape legislators’ behavior and that religious language serves various strategic purposes in legislative discourse. Given this, there are three key reasons to expect an openly religious Speaker to influence Republican representatives’ use of religious rhetoric. First, Johnson signals to legislators that religious language is not only acceptable but encouraged. As the U.S. grows increasingly secular, representatives may hesitate to use explicitly religious rhetoric for fear of alienating constituents. However, Johnson’s approach reassures them that such language has a place in legislative discourse. Representative Keith Self (R-TX) echoed this sentiment, writing to his constituents, “It is apparent that Johnson’s faith is a guiding force in his life and it was inspiring to hear him mention this on the house floor.” After quoting Johnson’s opening statement on God and the Bible, Self concludes, “I can’t tell you how encouraging it was to hear these words” (Self Reference Self2023). Still, given the Republican Party’s longstanding ties to religion, this alone may be the weakest of the three explanations.

Second, Johnson’s election as Speaker of the House signaled not only his own views on the role of religion in politics but also the priorities of part of the modern Republican Party. By invoking the relationship between religion and politics in his opening speech, Johnson signals to Republican legislators that religion remains a priority of the Republican Party. Prior research shows that party leaders shape legislative priorities, and members are quick to fall into line (Cox and McCubbins Reference Cox and McCubbins1993; Grimmer Reference Grimmer2013). This pattern is evident in how many Republican legislators adopted Trump’s distinctive communication style on Twitter, suggesting that members are highly responsive to leadership cues (Horan et al. Reference Horan, Brubaker, King and Meinhold2024).

Finally, since Johnson’s rise was driven by the MAGA-aligned faction of the Republican Party, Republican representatives may see religious rhetoric not just as a reflection of Johnson’s personal beliefs, but as a necessary signal of allegiance to this emerging power center. According to former Representative Matt Gaetz (R-FL), “The swamp is on the run, MAGA is ascendant, and if you don’t think that moving from Kevin McCarthy to MAGA Mike Johnson shows the ascendance of this movement, and where the power of the Republican party truly lies, then you’re not paying attention” (Gaetz Reference Gaetz2023).

Conversely, as Johnson’s leadership further entwines religion with the Republican Party, Democratic legislators are likely to distance themselves from religious rhetoric. Johnson’s ties to Christian nationalism have been widely criticized by Democratic lawmakers (Walker Reference Walker2024), including Representative Jared Huffman (D-CA), who wrote to his constituents, “Speaker Johnson embodies one of the most serious threats to democracy of our time: an outright rejection of the separation of church and state and documented ties to extreme Christian nationalism… he’s more interested bending public policy to reflect his personal religious views, which is a slippery slope to theocracy and governing under one imposed religion” (Huffman Reference Huffman2023). Given these criticisms, Democratic legislators may be even more inclined to distance themselves from religious rhetoric, reinforcing a partisan divide in religious discourse. Even if some religious Democrats feel compelled to offer an alternative religious perspective, the broader partisan dynamics may lead most Democratic legislators to reduce their religious rhetoric.

We expect that representatives will adapt their rhetoric when talking to their fellow legislators on the House floor and in newsletters to their constituents to reflect these trends. Thus, from these expectations, our first hypothesis with its sub-hypotheses follows:

We run two analyses to test this hypothesis. First, we use a DiD model to compare religious rhetoric used by Republican representatives to Democratic representatives from two months before Johnson became Speaker on October 25, 2023 to two months afterward. Democrats act as a control group to account for other nonpartisan factors that may have caused an increase in religious speech among both periods during the period of study. Hypothesis 1A would be correct if there is an increase in the number of religious House speeches or religious terms used in speeches, and Hypothesis 1B would be correct if there is an increase of religious newsletters or religious terms used in newsletters given by Republican representatives compared to Democratic representatives during this period. To ensure that this difference is not due to simply the holiday season, our second test is a DiDiD model. This model uses the same comparisons as our (DiD) model but further compares it to the same groups and time period in 2022. For Hypothesis 1A or 1B to be correct, we would expect to see a positive increase in the number of religious speeches or newsletters, or the number of religious terms used in speeches or newsletters, by Republican representatives in 2023 compared to Democrats and the use of religious rhetoric by both parties in 2022.

Based on the literature, we also note that elected officials use different rhetoric when addressing different audiences (Mayhew Reference Mayhew1974; Neumann Reference Neumann2019). Thus, we would expect legislators to use different language when addressing their colleagues on the floor of Congress compared to how they write to their constituents in a newsletter. These newsletters are sent out by nearly every member of Congress, intended to showcase their achievements and policy positions to their constituents. An analysis of legislators’ e-newsletters found that both Republican and Democratic congresspeople were more likely to show a more extreme ideology in their newsletters compared to their behavioral ideology (Cormack Reference Cormack2016). We expect this to apply especially to religious language, as it is found to potentially have a stronger impact on voters than non-religious language, particularly when looking at partisan effects (Castle et al. Reference Castle, Layman, Campbell and Green2017). As religious language is viewed as more partisan, we expect Republican representatives to use it at higher rates when addressing their constituents in newsletters after Johnson assumes office compared to how often they use it on the House floor. As Johnson’s election as Speaker may represent a new direction for the Republican Party, legislators may believe their voters may also value more extreme ideological beliefs and policies, leading legislators to change their language to reflect such. Our second hypothesis follows:

To test this hypothesis, we specify the distinctions between religious references given in House speeches and newsletters in our two tests for our first hypothesis. If our second hypothesis is correct, we expect to find a different impact of Johnson on speeches and newsletters, and this difference should indicate that representatives use more religious language in their newsletters than in their speeches.

Data

We pull from two data sources: speeches made on the House floor, in which legislators addressed each other; and representatives’ e-newsletters, in which legislators addressed their constituents.

In order to analyze speeches made in the House, we obtained all speeches made in the House between August 25, 2022 and December 31, 2023. Most of these data were provided courtesy of the Polarization Research Lab, and we supplemented any missing dates with our own collection of floor speeches using the Congressional Record (Congressional Record).Footnote ² These speeches include anything said by representatives included in the Congressional Record. Each speech was counted as a segment in which a congressperson spoke; for instance, if a floor debate occurred and a congressperson was briefly interrupted, that speech is counted as two separate speeches, with the interrupting congressperson’s comments also counting as one speech. We dropped speeches that were less than 350 characters long to exclude short statements, such as legislators asking how much time they had left to speak, and speeches where legislators gave personal explanations as to why they missed a vote.

We used DCinbox (DCinbox), a source which collects congress members’ e-newsletters, to gather representatives’ e-newsletters. These newsletters are sent out by nearly every member of Congress. Lindsey Cormack, who created this resource, found in a 2012 CCES question that 19% of respondents had signed up to receive their representative’s newsletter and 14% had signed up for their Senator’s newsletter (Cormack Reference Cormack2016). These subscribers tended to be older, wealthier, more politically active, and more extreme.

To test our hypotheses, our analysis is limited to two months before and after Johnson became Speaker, August 25th through December 25th. We compare the speeches from this period in 2023 with the speeches from this same period in 2022. This resulted in 5,951 speeches in 2023 and 4,009 speeches in 2022. We used all representatives’ e-newsletters collected by DCInbox sent during the same time frame as our data on congressional speeches. This gave us 4,798 newsletters in 2023 and 4,079 in 2022.

In searching for religious rhetoric, we split chosen religious terms into six separate dictionaries. As we are looking for trends over time, we opted for a broader, limited dictionary as opposed to attempting to find every instance of religious speech. These dictionaries, with their included terms,Footnote ³ are listed below:

• God: god, jesus, christ, savior, our/the lord, holy spirit

• Blessings: bless, faith, worship, pray, divine, salvation, sin, heaven, holy, spiritual

• Scripture: scripture, bible, biblical, psalm, ten commandments

• Religion: religion, christian, gospel

• Political: judeo, religious freedom, religious liberty, freedom of religion, christian nation

• Dogwhistle: pornography, gender ideology, gender agenda, pro-family, traditional family, traditional marriage, unborn, pro-life, protect life, abortion agenda

We analyzed these dictionaries separately and, with the exception of terms in the Dogwhistles dictionary,Footnote ⁴ added together as a single dictionary, ensuring that each phrase was only counted once, which we labeled “Religious Terms.”Footnote ⁵

Descriptive analysis

We begin with a descriptive analysis of the use of religious language by representatives on the House floor during 2023. Table 1 reports the total number and percentage of speeches given in the House which included at least one religious word, as well as how many representatives who spoke on the floor at least once during 2023 used these religious terms. These data are further broken up into religious topic and party.

Table 1. Number and percentage of speeches and representatives who used religious terms in House speeches in 2023

${{\rm{\;}}^{\rm{*}}}$ Excluding Dogwhistle terms

Table 1 shows that Republicans used religious references at a higher rate than Democrats on the House floor: 18% of Republican speeches included at least one religious term, compared with 14% of Democratic speeches. Furthermore, 201 Republican representatives, or 89% of Republican representatives who spoke on the House floor in 2023, gave at least one speech which included a religious term, compared to 179 or 83% of Democratic representatives. Both of these differences are statistically significant ( $p \lt 0.01$ and $p \lt 0.1$ , respectively). When it came to specific religious topics, both Republicans and Democrats spoke most often with “Blessings” terms followed by “God” terms, though Republicans had a higher percentage of speeches that used these terms compared to Democrats. A t-test comparing the amount of religious words used by representatives indicates that Republican legislators used more religious words on average (16.8 compared to Democrats’ 11.0), a difference which is statistically significant at the 5% level.

To better showcase the partisan differences in religious usage, Figure 1 shows the average number of religious terms per speech each month by Republicans and Democrats. Figure 1 illustrates that Democrats use religious language rather consistently through 2023, with a rare high in August.Footnote ⁶ Republicans’ use of religious language is much more variable, primarily due to celebrating religious events and holidays. For example, the date with the most religious rhetoric is December 11, 2023 with 348 religious references due to the House celebrating Bible Week, during which various legislators spoke on the Bible and its influence on America (Flick Reference Flick2023). April had the highest average religious terms per speech due to the celebration of Faith Month, a holiday began in 2022 by Concerned Women for America to celebrate how faith benefits society (McPhee Reference McPhee2023). October also has high average religious terms for Republicans, largely due to Representative Jefferson Van Drew (R-NJ) giving 37 speeches before October 25th, all of which recognized individuals from his district and ended with “God bless [individual], and God bless the United States of America.”

Figure 1. Average number of religious terms in House speeches, by party.

To better understand how representatives may discuss religion with their constituents, we turn our analysis to the use of religious rhetoric in congressional e-newsletters. Figure 2 shows the average religious terms used per newsletter each month in 2023 by party. This figure stands in stark contrast to our earlier findings on House speeches. When it comes to newsletters, Republicans consistently invoke religious speech more than Democrats at a rate of three to one or higher. This indicates that representatives discuss religion in different terms depending on their audience: when addressing their colleagues, Republicans may not be as pressured to use religious terms, while Democrats feel more comfortable using them, and such pressure reverses when addressing their constituents. Further data on the number and percentage of newsletters and representatives who used religious terms in their newsletters in 2023 can be found in Table 1 in the Appendix.

Figure 2. Average number of religious terms in newsletters, by party.

Does Mike Johnson’s assumption of the office of the Speaker of the House influence the use of religious rhetoric in either of these situations? We turn to statistical analyses to answer this question.