The political pulpit: issue framing and conservative partisanship in evangelical protestant pastors’ sermons before and after the 2008 election of Barack Obama

Abstract The intersection between religion and politics in the United States remains prevalent in the 21st century. Especially in regard to many Protestant Christian denominations, the relationship between religion and political partisanship seems to continually intensify. Through a content analysis of online sermon transcripts, this study focuses on Protestant pastors’ interpretations of numerous political issues in the 21st century U.S. both before and after the 2008 election of President Obama. Drawing from framing theory, the majority of pastors primarily framed their sermons in three main ways: (1) increased discussions of political conservatism after the 2008 election, (2) maintained abortion as the most popular social issue discussed, and (3) increased discussions of literalist interpretations of the U.S. Constitution following the 2008 election. The study’s results and implications for various avenues of future research are discussed at length.

, none of the aforementioned works examine the presence or absence of politics preached from the pulpit before and after the historic 2008 election of Barack Obama. Such framing can influence political movements, voting behavior, public opinion and public policy (Layman 1997;Brooks 2002), providing a clear warrant for the present investigation. For issues entering local, state and national electionswhether presently or in the near futureframing, if executed effectively by pastors, may also influence values (Regnerus, Smith, and Sikkink 1998). Regarding the power dynamics of issue framing and, subsequently, the consequences of such dynamics, Sniderman and Theriault explained that 'whoever controls the framing of an issue controls the political outcome ' (2004, 148).
In this study, the researchers conducted a content analysis of 105 Protestant sermon transcripts posted on www.sermoncentral.com, an online social network of pastors, from 23 May 2001 to 11 May 2014, to gain an understanding of how various political issues are framed by pastors from the pulpit. This study addresses the frequency at which pastors mentioned specific social, economic, and political issues both before and after the 2008 presidential election of Barack Obama to determine whether his election resulted in a demonstrable shift in the types of content offered within sermons specifically focused on political issues. By examining pastoral rhetoric before and after the 2008 presidential election, comparisons can be drawn regarding the issues in which pastors found most salient. Although this sermon sample does not provide for broad generalizations regarding Protestant Christianity in the United States, the findings are significant in showing how the largest and most popular sermon website (i.e. over 150,000 sermons and over 250,000 church leaders who use the site every week) serves as a forum for modern political discussions and debates.

Issue framing
The current study operates under issue framing (Sniderman and Theriault 2004), which when defined in the context of citizenry-more specifically churchgoers-frequently is oriented to Lippmann's (1922) foundational piece, Public Opinion. Lippmann (1922) describes the world of public opinion as complex, with many actors challenging the attention of the citizen group. Significant literature has followed to expand on Lippmann's (1922) explanation of how politics attracts citizen attention, including oft-used imagery set forth through framing. A controversy or compelling story, for example, may capture the fascinated attention of a community or broader citizenry (Gamson and Modigliani 1987). This theoretical development qualifies an assumption for why churchgoers are receptive and attentive to framing during sermons. Such an assumption is bolstered when considering how pastors typically deliver 'a central organizing idea or storyline that provides meaning to an unfolding strip of events, weaving connection among them. The frame suggests what the controversy is about, the essence of the issue' (Gamson and Modigliani 1987, 143).
Utilizing broader conceptions of framing from Goffman (1974), Entman's (1993) articulation of framing can be incorporated to understand the potential impact of content analyzing pastoral sermons to detail politicization in church communities.
According to Entman, 'To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation ' (1993, 52). When considering a problem, definition, or moral evaluation of political issues provided by pastors in sermons can mark an event or occurrence (Bateson 1972). From this theoretical position, this study analyzes sermons before and after the historic 2008 election of Barack Obama.
Religion in the modern U.S.
Many of the political issues coded in the current study are relatively modern when considering the U.S. political landscape and public policy. One cannot address theories attempting to explain religious groups' battle with modernity without addressing the concept of secularization. Weber (1930) is credited with first introducing the concept, which gained most of its popularity in the late 1950 s. Secularization has since been debated by scholars across multiple disciplines (Swatos and Christiano 1999). Within Weber's (1930) conception of the term, secularization was the rational alternative to believing in the mystical and unknown. Secularization theory explains that as society becomes more modern, the prominence of religious beliefs and practices will continually lessen (Wallace 1966;Swatos and Christiano 1999). This concept is relevant to church communities who, for the purposes of this study, could be considered non-secular. While such concepts are continually debated, the concept has been employed in influential studies addressing modern religious life and group identity associated with churches (Seul 1999;Bruce 2001).
The current study's sermon sample includes several religious denominations; the distinctions between them may offer insight on the variations of political intensity preached from the pulpit. More notably, community-specific information and demographic profiling may help detail why there are trends and distinctions in sermons. This study does not explore causal effects of sermon variance or analyze latent variables of pastors' partisanship beyond frequency at which a pastor mentions a particular issue, yet religious denomination may contribute to trends and political influences in church rhetorical leadership. Consequently, the study focuses on one broad branch of Christianity, Protestantism, to define the sample drawn from sermoncentral.com. The characteristics of this branch are articulated in the next section.

Understanding protestantism
As a broad term, Protestantism is a branch of Christianity including several denominations present in the United States and abroad. Institutionally, Protestantism differs from Catholicism in that Protestants do not give authority to the Pope. In this consideration, authority is only attributed to God, and the Bible is often cited or exemplified as the primary means by which authority is allocated (McGrath and Marks 2004;Melton 2005).
Arguably one of the most influential religious group identifications in United States Christianity lies within the overarching term of 'Protestant', specifically within the subset labeled as 'Evangelical Protestantism' (Bebbington 1989;Hankins 2008;Wellman 2008). Beliefs and characteristics of Evangelical Protestants can be summarized in four distinct parts: (1) biblical inerrancy; (2) the emphasis of conversion, or accepting Christ; (3) missionary deeds of activism and evangelism; and (4) the belief in Christ's sacrificial death to save his followers (Bebbington 1989). This religious subculture's perceived mission of outreach is evident in many conservative Protestant denominations' choice to evangelize to the secular world ). This is also described as bringing people to know Jesus Christ as their personal Lord and Savior (Hankins 2008;Wellman 2008). Through this evangelizing, opportunities for interaction and potential conflicts with out-groups arise. This is demonstrated by targeting individuals' decisions to accept or deny Christ, whereby this societal engagement further strengthens the religious subculture's strength, in-group identity, and in-group affiliation .
Although various scholars have defined Protestantism with relatable qualities, there are many differences among the denominations within Protestant Christianity. Nonetheless, the term 'Protestant' in its broad interpretation (see Melton 2005) is used in the present study of 105 sermons delivered by pastors under the term's definition. Furthermore, only Evangelical Protestant sermons are included in the analysis of this study to narrow the scope and focus on the frames prevalent within this subset of Christianity.

Religion and politics in the U.S.
American history is rich with narratives where religion and politics interlock despite constitutional provisions expressing separation (Burleigh 2007). Many argue the dynamic of religious compatibility with government and its closeness has increased over the past 50 years (Greenawalt 1988;Shields 2009;Bean 2014). One contributing factor is the politicization of Evangelical activism, which entered into the American socio-political scene in the late 20th century (Fetner 2008;Hankins 2008). This political movement involved a combination of efforts by political leaders, religious leaders and socially conservative activists. The movement proved to other fundamentalist believers participating in political campaigns that they could acquire support from their church communities and their ideological constituents if/when entering the political arena (Fetner 2008). This socio-political activism enables many conservative Christians to spread religious messages to millions.
If Christians lose sight of political and social activist goals, many within the Evangelical movement believe that such deviation is a sign that one or many have been deceived by Satan and that they are actually doing more harm than good (Shields 2009). Therefore, despite many changes in public perceptions on social issues over recent decades (see Sherkat et al. 2011), the Christian Right (also referred to as the Moral Majority or the Religious Right) has remained relatively active for fear of repercussions of God's wrath if they believe faith is wavering on core conservative issues like abortion and gay marriage (Fetner 2008;Froese, Bader, and Smith 2008;Barton 2012). The strength and impact of highly conservative Christianity can exert a strong influence on the political rhetoric circulated and communicated to the public, especially during political election campaigns (Domke and Coe 2010;Gonzalez 2012). Thus, when religious 'theology becomes a political commodity' (Gonzalez 2012, 571), political leaders tailor their words to appeal to their constituents. This study proposes this intertwining relationship between religion and politics is as profound and relevant today as it has been at any other point in United States history, underscoring the warrant for the present study.

Research questions and hypotheses
This study is focused on issue frames potentially influencing church attendees' feelings, opinions, viewpoints, beliefs or behaviors surrounding the 2008 presidential election that are measurable and potentially predictable in their outcome (see Gross and Goldman 2003). Despite the many topics presented before and after the election of Barack Obama during the study's sample period, the framing identifier was narrowed to a single broad topic: politics. Political sub-categories are then provided to mark the specific and trending issues preached from the pulpit, also considering the values and attributes associated with the issues framed by pastors (Sniderman and Theriault 2004). Such values and attributes are what help define political partisanship, including conservatism.
The current study explores the political intensity of conservatism preached by Evangelical Protestant pastors, which considers the political dynamics associated with the sample period. When determining what pertains to conservatism, the researchers refer to the 2008 Republican Party platform (NRLC 2008). The first research hypothesis also considers how political factors can influence pastoral leadership and their roles within church walls and within local communities (Irwin and Roller 2000). Various issues impacting the surrounding environment of each church community (e.g. social, economic and historical) may shape how church leaders address their congregations (Carroll 2006). Overall, however, American Protestant Christianity is characterized in the literature with a general commonality-political conservatism (Brooks 2002;Layman 1997;PRC 2012bPRC , 2016Shields 2009;Wald, Owen, and Hill 1988).
Political conservatism is a unifying value-set for church communities across America (Wuthnow and Evans 2002), and can have a potentially significant impact on society due to micro-level and macro-level political activism by pastors and other members in church leadership positions (Guth et al. 1997). As such, church leadership often assumes community roles by inspiring congregational participation and activism within local (Martinson and Wilkening 1987), state (Green, Guth, and Hill 1993), national (Leege and Kellstedt 1993) and international political affairs (Buss 2003).
Despite general trends in conservative activism, reports indicated that American Protestants had various perceptions of Barack Obama following the 2008 Presidential election (Gonzalez 2012). Nevertheless, prior research suggests that conservative issue framing of pastors may impact the immediate and broader social environment where pastors preach politically conservative viewpoints (Brooks 2002;Layman 1997;PRC 2012bPRC , 2016. The following pair of two-pronged research questions are offered with the former related to the issues pastors will make more salient and the latter pertaining to whether these issues shifted demonstrably after the presidential election of Barack Obama. The first research question addresses the specific issues that may detail the level of politicization. RQ 1a :What issues are Evangelical Protestant pastors most likely to address within self-categorized political sources (sermons)?
The second research question addresses the potential shift in issue framing regarding the historical event considered in this study. The researchers inquire about issues framed before and after the election of Barack Obama to detail any potential variance in politicization as it relates to the historical event.
RQ 1b :To what extent did pastoral issue focus shift within sermons after the election of Barack Obama when compared to before his election?
The third research question addresses the frequency of issues mentioned in political sermons, which may better define the intensity at which pastors politicize the pulpit.
RQ 2a :What issues are Evangelical Protestant pastors most likely to address within references of self-categorized political sermons?
In consideration of the previous research questions, the fourth research question addresses issue framing shifts as it relates to political references preached by pastors before and after the historical event considered in this study.
RQ 2b : To what extent did pastoral issue focus shift within references after the election of Barack Obama when compared to before his election?
Confirming the oft-fused connections between religion and politics, the 2008 Republican Party's platform has been argued to be significantly influenced by pastoral rhetoric and the politicization of the church (Djupe and Neiheisel 2008;Conger 2010). Given that the Religious Right tends to focus on social issues more than economic issues (Fetner 2008), one finds that the 2008 Republican platform states, 'At its core, abortion is a fundamental assault on the sanctity of innocent human life. Women deserve better than abortion. Every effort should be made to work with women considering abortion to enable and empower them to choose life' (NRLC 2008, 53). Moreover, legalized abortion and abortion laws are driving factors within both Christian activism and dialogue (Fetner 2008;Manning 1999;Noonan 1970;Shields 2009). As Hoffmann and Johnson (2005) describe, abortion remains one of the most, if not the most, contentious socio-political issues in which Protestant Christians oppose in the 21st century United States. Barton (2012) notes the focus on core social issues with policy ramifications, citing abortion and gay marriage as two key touchstones. As such, abortion is coupled with the public debate over gay marriage to posit the following hypothesis: H 1 : Abortion and gay marriage will have the highest frequency of mentions compared to other social issues from the sample of sermons.
Heavily debated political issues like abortion prompt advocacy battles directed at influencing legislative outcomes in the United States Congress (Hertzke 1988). The assumption of political advocacy, particularly from American religious communities, is that legislation will impact society (Hertzke 1988). Several studies highlight this profound influence of legislative decisions and public policy on the lives of American citizens (Cochran et al. 2011;Peters 2012). As previously stated, numerous forms of the social and cultural life greatly shape dialogue and debate in the political realm (Habermas 2006;Shields 2009). Reciprocally, dialogue and debate in the political realm may shape society (Adler, Hoegeman, and West 2014;Sinclair 2012). As indicated by the sermon samples in this study, Protestant pastors are potentially politicizing their churches through issue framing from the pulpit. Hertzke (1988) suggests issues like abortion have inspired advocacy groups including politicized conservative Christians.
The literature offers many historical examples where Protestantism has influenced public policy (Hammond 1992;Mooney and Lee 1995;Adam 2003). Influence of public opinion and legislative policy in the United States is potentially one of the most impacting outcomes of issue framing by pastors. Therefore, an analysis of the frequency of legal and legislative mentions is relevant for understanding its potential impact on American society. Furthermore, the current analysis focuses on sermon rhetoric before and after the 2008 presidential election; whereby, future studies may draw correlations of public opinion and legislative development with sermon rhetoric. Consistent with this study's prediction that conservative issue framing will increase after Obama's 2008 election, the researchers anticipate legal and legislative mentions will also increase. Thus, the following hypothesis is proposed: H 2 : Legal and legislative references will increase following the 2008 presidential election.

Methods
For this study, sermon content is sourced from sermoncentral.com, an aggregate social site facilitating profile connections and engagement. With a similar concept to Facebook, the site has less real-time features (e.g. newsfeeds, etc.) and is specific to pastors. Nevertheless, pastors connect by the thousands through this social platform. The sermons this study analyzes are posted on Sermon Central's website from various pastors, or 'contributors,' across the country, with the sermons becoming self-reports because they are posted onto the website by the pastors who preached them. Anyone wanting to post a sermon on the website must become a member of Sermon Central; however, the sermons are publicly available to anyone who visits the website. The website boasts over 152,000 sermons. The longitudinal range of sermons spanned from 23 May 2001 to 11 May 2014, with the 2008 election of Barack Obama as a key variable for testing the study's hypotheses. To narrow the study's focus, the search term 'politics' was entered. After the researchers input the search term, 122 sermons become readily available, each citing the term as one of the keywords in the sermon. A total of 17 sermons were from pastors not matching the operational definition of Protestant, resulting in a final sample size of 105 sermons from 74 pastors.

Database and coding
To code the 105 sampled sermons, each was uploaded to QSR NVivo 10 (QSR International 2014), a software program storing large data sets for researchers to conduct various types of analyses. Four 'nodes' were created, each representing major categories detailing the study's focus: (1) 'Political Features,' (2) 'Social Issues,' (3) 'Satan,' and (4) 'Before or After 2008 Election.' From two of the four nodes, 'child nodes' were created specifying the subthemes. In certain instances, tertiary categories beyond child nodes were created as well.
For the 'Political Features' node, the child nodes are: Political Ideology, Political Elections, Laws and Legislation, Politicians, and Barack Obama. The 'Political Ideology' child node includes any references to politics, broadly-defined, and any references to political, economic, and/or philosophical ideologies (e.g. conservatism, liberalism, communism, socialism, and capitalism). The 'Political Elections' child node includes any references of United States political elections (e.g. primaries, mid-terms, and presidential), and any references to voting. The 'Laws and Legislation' child node contains any mentions of United States laws or court decisions (e.g. Supreme Court rulings and the United States Constitution). The 'Politicians' child node includes any mentions of political leaders, whether past or present, in the United States. An additional subcategory under the child node 'Politicians' is added, entitled 'Barack Obama,' to count the number of times the 44th President of the United States is mentioned by name.
The second major node is 'Social Issues,' including the following child nodes: Abortion, Criminal Justice, Economics, Gay Marriage, Health Care, and Immigration. The 'Criminal Justice' child node includes any references to the criminal justice system in the United States (e.g. crime, crime-rates, and the death penalty), and the 'Economics' child node includes any references to economic entities in the United States or abroad (e.g. economic systems, foreign trade, inflation, recessions, taxes, and welfare). Whenever a pastor discussed any of the other aforementioned social issues in their sermon, the text is highlighted and included under the matching subcategory. Along with analyzing which political entities mentioned in the sermons most frequently occur, the social issues node is important in analyzing which social issues gain the most attention from the sample of pastors.
The third node, 'Satan,' includes any references to Satan, the devil, and/or Lucifer within the sermon passages. Lastly, the final node, 'Before or After 2008 Election,' simply records the date in which each sermon was posted online to determine if the date posted is before or after the election on 4 November 2008.
To test for intercoder reliability for the content analysis, two coders compared findings on approximately 25% (n ¼ 26) of the sampled sermons, calculating agreement using Cohen's (1960) kappa formula, achieving reliability scores for sources (K ¼ .91), references (K ¼ .89), and pastoral demographics, (K ¼ .98). The overall intercoder reliability of the twenty-six sermons using Cohen's kappa exceeded .93.
Chi-square analysis of the data was incorporated to determine statistically-significant differences between the nodes and child nodes within the study. To do this, the percent of sermons and the percent of references, both before and after the 2008 election, were used as expected frequencies for each node and child node. Since thirty-seven sermons (35%) were delivered prior to the 2008 election, and sixty-eight sermons (65%) were delivered after the election, the expected frequency for before the election was thirty-five percent and the expected frequency for after the election was sixty-five percent. Likewise, since 981 of the total 2750 references (i.e. 36%) emerged from sermons delivered prior to the 2008 election, and 1769 of the total 2750 references (i.e. 64%) emerged from sermons delivered after the 2008 election, the expected frequency for before the election was thirty-six percent and the expected frequency for after the election was sixty-four percent.

Sermon demographics
From the sermon sample, 37 sermons (35%) were delivered prior to the November 4, 2008 election, and 68 sermons (65%) were delivered after the election. This significant difference in sermon frequencies may suggest that the election of President Obama had a major impact on the amount of sermons delivered after the 2008 election. Table 1 breaks down the sermon sample by year.
Regarding self-labeled pastoral denomination, Baptists comprised the majority of sermons from the sample (n ¼ 30), followed by the Church of Christ (n ¼ 16), Evangelicals (n ¼ 16), and Presbyterians (n ¼ 12), all of which could be classified under the 'Evangelical Protestant' label Wellman 2008). In reference to pastor gender, 73 of the 74 (i.e. 99%) sampled pastors within the database were male, roughly matching gender divisions for Protestant pastors. Based on the pastor photos provided by 56 of the pastors, the racial dynamics of the sample pastors included: White/Caucasian (n ¼ 52), African American (n ¼ 1), and Other (n ¼ 3).

Content analysis results
Two important terms to note from the QSR NVivo 10 (QSR International 2014) software are sources and references. 'Sources' refers to the amount of documents analyzed. In other words, the term 'sources' is synonymous with sermons. In the case of this study, there are 105 sermons (or 105 sources). In contrast, the term 'references' is the amount of mentions of a node or child node within a particular sermon (or source). For example, Barack Obama could be mentioned three times in one sermon. Therefore, for this hypothetical scenario, the source amount would be one and the reference amount would be three. In total, there are 2750 references from the sampled sermons. (2) a Among the 34 sermons from 2008, 20 were posted prior to the Nov. 4th election and 14 were posted following the election.
Research Questions 1a and 1b pertained specifically to sources, with the results calculated in Table 2 in relationship to sample size, where expected frequencies for sources were 35% before the 2008 election, and 65% after the 2008 election. Table 2 includes the frequencies and statistically-significant chi-square results for each node and child node pertaining to sources before and after the 2008 election.
As Table 2 illuminates, the top three categories mentioned within the sources involved political ideology (n ¼ 105), elections (n ¼ 90), and politicians (n ¼ 72). These areas were mentioned significantly more than any social issue, answering Research Question 1a in that the focus of the sermons tended to be more on the mechanisms of elections and people running within them than on any singular social issue.
Regarding differences in sources before the election of Obama when compared to after his election, the sole area in which a difference persisted was in the mentions of Obama himself, as he was significantly more likely to be mentioned after his election than before it (v 2 ¼ 4.47, df ¼ 1, p < .05). More tellingly, Obama was never mentioned in any of the sermons delivered before the 2008 election. He was, however, mentioned thirty-eight times throughout eighteen sermons delivered after the 2008 election, providing one noteworthy dialogue shift as an answer to Research Question 1b.
Research Questions 2a and 2b pertained specifically to references, with expected frequencies of 36% for before the 2008 election, and 64% after the 2008 election. The frequencies and statistically-significant chi-square results for each node and child node pertaining to references from both before and after the 2008 election are shown below in Table 3.
Research Question 2a pertained to the most-referenced topics within the sermons. Again, the top three categories mentioned within the references were political ideology (n ¼ 499), elections (n ¼ 234), and politicians (n ¼ 275), providing an answer for RQ2a.
However, four topic-based differences emerged when addressing references before Obama's election when compared to after it. First, Obama himself was referenced more after his election than before (v 2 ¼ 6.93, df ¼ 1, p < .05). Second, political ideology was higher before his election than after it (v 2 ¼ 4.07, df ¼ 1, p < .05). Third, abortion was more likely to be mentioned as a key issue after Obama's election (v 2 ¼ 3.96, df ¼ 1, p < .05). Finally, immigration was a salient issue before the election, but received little attention after Obama's election (v 2 ¼ 6.77, df ¼ 1, p < .05). These four differences provide a multi-leveled response to Research Question 2b.
Hypothesis 1 predicted a greater focus on the two largest social issues of the social conservative movement: abortion and gay marriage. While abortion was more prominent in references after Obama's election, these two issues were not significantly more likely to be mentioned than other issues, providing little support for H 1 .
When taking the results from the research questions and hypotheses in aggregate, one could aptly conclude that the overall focus of the sermons shifted to one man, Obama, beyond broader foci on issues and the laws and legislation that could be passed to advance those aims. A second level of qualitative analysis was employed to understand whether President Obama's 2008 election victory led to enhanced politically-conservative sermon rhetoric, with the following examples provided as illustration.
In April 2010, Pastor Grant Sisson of Countryside Christian Church in Shreveport, Louisiana proclaimed to his church congregation, 'We live in an increasingly anti-Christian world. Our own President has openly stated to the world that we are no longer a Christian nation.' Extending this theme of an 'anti-Christian' nation and world, and also coincidentally in April of 2010, Pastor Mike Fogerson of First Baptist Church of Chester in Chester, Illinois warned his church congregation of the dangers of secularism by stating, 'To date, not one president would have been popular enough to be seen as a unifier/leader of a new global order, to form a United States of Europe, until the election of the 44th President, Barack Hussein Obama.' Both Pastors Sisson's and Fogerson's statements were made in sermons addressing the perceived prophetic signs of the end of the world and the rise of the Anti-Christ as mentioned throughout the Bible (e.g. Matthew 24:21-22, 2 Peter 3:3-4, and Revelation 3:3, 13:16-17, 16:13) (see IBS 2001). The aforementioned pastors continued to preach President Obama by addressing him as a non-Christian, a socialist and a Marxist. These broad ideological accusations are evidence of politically conservative rhetoric following the 2008 election.
The one issue that did appear to be tied to the heightened polarized rhetoric surrounding Obama was abortion, with the following exemplars of the type of focus abortion received.
in Palmyra, Illinois exclaimed that 'we have to recognize that an unborn fetus is a living human being … The baby living in his/her mother is as distinct and unique a human being as you are from me.' Lastly, in July 2012, Pastor Freddy Fritz of Tampa Bay Presbyterian Church in Tampa, Florida compared abortion to historic gladiators when he stated, 'We have our own modern crude and barbaric killing for pleasureabortion. Although there are some medically necessary abortions-such as when the life of the mother is at stake-the fact is that the vast majority of abortions are for the pleasure of the individuals who do not want that child.' The aforementioned quotes illustrate the potential widespread popularity among Protestant Christian pastors of discussing the issue of abortion from the pulpit.

Discussion
The literature provides many examples of how public opinion and legal outcomes, like Supreme Court cases and legislation, are interrelated (Hall 2014;Gibson and Nelson 2015). However, few empirical studies provide specific instances where public opinion and legal outcomes are influenced from churches in the United States. This study addresses a sample of 105 pastoral sermons and analyzes issue framing from the pulpit around a historic event, the election of Barack Obama. The results indicate that Obama was not a focus of the pastors at all until after his election, a possible function of denial that he could gain election, particularly in Southern states in which Obama did not perform well on Election Day.
With quantitative evidence of church politicization, the implications of this study may contribute to the policy debate regarding the separation of church and state. Moreover, this study's results are important for recognizing the significance of Obama's election as the 44th President of the United States in 2008 relative to issue framing in the church. Considering future studies, this analysis provides an opportunity for scholars to advance research on framing effects and the role of religion and politics in American society.
From a macro-level standpoint, this study shows that pastoral framing could have significant legal implications. For example, Senator Lyndon B. Johnson passed legislation through congress in 1954, known as the Johnson Amendment, which prohibited tax-exempt organizations, such as churches, from endorsing or opposing political candidates, parties, and issues (Mayer 2010;PRC 2012b). The Johnson Amendment is regularly violated to the point that over 100 sermons not only discuss political matters, but these sermons are also made publicly available online. Additionally, the Internal Revenue Service (IRS) has never taken legal actions against a church or organized religious organization for allegedly violating the Johnson Amendment's provisions (Mayer 2010). This is due to fears of violating one's First Amendment rights which states: Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances (U.S. Bill of Rights 1789).
Thus, political ideological endorsements from the pulpit, as well as explicit, partisan biases toward political candidates and/or parties, have been almost completely unchallenged (Mayer 2010;PRC 2012b).
Since 2008, the Alliance Defending Freedom group (organized by many conservative Protestant leaders and activists) has fought against the Johnson Amendment by launching the annual 'Pulpit Freedom Sunday' to encourage U.S. pastors to publicly discuss political issues (PRC 2012b). However, many pastors support the Johnson Amendment, arguing that it is the best way for churches to avoid mixing partisanship and religion (PRC 2012b). Thus, this issue and legislation remains pertinent among religious and secular (or nonreligious) communities throughout the United States. Whether pastors adhere to, or disobey, the Johnson Amendment, the possible framing effects could include attitude changes on contentious socio-political issues, political ideologies, political partisanship, and voting behaviors.
Considering the broad effects of issue framing from the pulpit, like voting behavior, a potential limitation of this study could be claimed within a perceived lack of generalizability. After all, this study focuses on a small sample of sermons relative to the thousands of sermons delivered each week. Nevertheless, the power of the results, at a minimum, indicate a convoluted relationship between the separation of politics and religion. However, since the study includes sermons from various Evangelical Protestant pastors, the amount of sermons and pastors is not enough to generalize to the broader branch of Protestant Christianity in the United States. Therefore, more analysis is recommended.
Another limitation pertains to the sample pool of pastors that comprise the sermon sample. The sampled pastors cannot be fully reflective of Protestant pastors at large in the United States because these pastors have a specific characteristic, engaging in online pastoral conversations made publicly available. Consequently, the online connectivity through social media could influence sermon rhetoric from shared information among pastors. Another potential distinguisher of online users is that a pastor engaging on the Internet may have an interest in growing their presence online; whereby their sermon rhetoric may appeal to less local-specific demographics. Pastors who engage the Internet to post their sermons are reasonably expected to differ from non-users or less engaged users of the Internet.
Thus, future studies should consider broadening the scope of the study by using a larger sample size to include sermons from church denominations and subsets that are more generalizable to the broader branches of Christianity (i.e. Catholicism and Protestantism) in the United States. Additionally, the study scope could be expanded upon by using more comprehensive keyword search terms. Furthermore, the study scope could be broadened by examining sermons from other religions (e.g. Buddhism, Hinduism, Islam, and Judaism), secular philosophies (e.g. Atheism and Agnosticism), and from multimedia platforms outside of the Sermon Central website. Therefore, more analysis is recommended.

Conclusions
Issue framing studies are a result of significant development stemming from Goffman's (1974) frame analysis suggesting that people, whether alone or in groups, find meaning once they have perceived, identified and labelled an event. This foundational understanding of framing is integral to imagine the potential impact of this study, enabling future research on how church attendees interpret their experiences while listening to politicized sermons. While this study does not apply survey method in search of framing effects on attitude, beliefs and behavior in the church environment, it does provide the necessary data qualifying secondary use for such a research endeavor. More specifically, this study provides an analysis of when, and by whom, Evangelical Protestant church sermons are politicized. Perhaps the greatest potential impact of this study is the result of a targeted observation of the politicization of churches from the pulpit; researchers could more credibly assume pastors explicitly express political viewpoints, opinions, sentiments, perspectives, narratives, and identity constructs to a potentially impressionable community (Lippmann 1922).
Within the United States, there is an active debate over separation of church and state, which remains relevant in today's socio-political and religious contexts (Habermas 2006). Furthermore, the scope of this issue spans across arguably all religious and secular citizens throughout the nation (Boston 2007). As the results of this study reveal, the debate over separation of church and state must consider issue framing from the pulpit. Moreover, the results of this study reveal that heavily debated issues like abortion are framed in the church, which may have implications on society at micro-levels and macro-levels (Guth et al. 1997). Thus, this study's analysis helps to define the processes and potential outcomes of politicization of the church, and it also provides the basis from which future studies may advance the scholarly research on framing effects pertaining to religious communities and political behaviors.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributors
Ethan C. Stokes is an Assistant Professor of Public Relations at the University of Alabama. His research focuses on the relationship between politics, religion, and the mass media in the United States, as well as political public relations, political communication, public policy, and international relations.