Politicization of Attitudes Toward the FBI Under the Trump Administration

Rachel Hertzberg

Abstract: President Donald Trump’s administration placed the Federal Bureau of Investigation (FBI) in the spotlight of American politics. The uncommon attention focused on the Bureau raises questions about how the public perceives an agency meant to be insulated from politics. This article examines whether partisanship influences public opinion of the FBI. I hypothesize that President Trump’s critical comments contributed to a partisan divide in attitudes about the FBI, a notion I test via an Amazon MTurk survey experiment administered in January of 2020. My findings suggest that exposure to partisan messages influences the expression of emotions toward the FBI. My conclusion that the Bureau is viewed by many Americans within the framework of party politics has concerning implications for its functionality in today’s society.

I. Introduction

In recent years, the FBI has received considerable media attention due to its investigation of Russian interference in the 2016 presidential election. The FBI opened this investigation in response to the hacking of a Democratic National Committee (DNC) server by Russian intelligence. After this hacking, WikiLeaks released compromising information obtained about Hillary Clinton in an apparent attempt to undermine her campaign and increase support for Donald Trump (Zurcher 2017). As part of the investigation, the FBI assessed whether the Trump campaign played any part in the Russian effort to undermine candidate Clinton. Political elites, including media pundits and elected officials, have openly criticized the FBI for this controversial investigation, as some consider it an attempt to delegitimize the Trump presidency. This type of discourse is new to the American political landscape and can have important implications for the relationship between the public and the FBI. Further, these relational changes can have even broader consequences for the Bureau’s ability to conduct effective investigations.

This study examines whether exposure to partisan cues influences attitudes about the FBI, and whether these attitudes differ between Republicans and Democrats. Recent studies have indicated that there is a partisan divide in opinions of the FBI (Brenan and Ander 2020; Penn State Social Science Research Institute 2018; Howland 2018). However, these studies fall short of establishing a causal relationship regarding exposure to partisan cues and the development of opinions about the FBI. In this study, I build upon prior empirical work with a survey experiment to assess potential causal mechanisms shaping attitudes toward the FBI. By exposing subjects to partisan critiques of the Bureau, I examine my hypothesis that there is a correlation between elite framing and public perception of this agency.

Study results suggest that increasing political polarization and criticism of the Bureau by high-profile politicians have contributed to shifts in opinion of the FBI. This piece therefore contributes to our understanding of how political elites play a role in the activation of partisan identities and the formation of political attitudes, with implications for the Bureau’s efficacy and role in American society.

II. Literature Review

A. Background

The FBI’s official mission statement is simply to “protect the American people and uphold the Constitution of the United States” (FBI n.d.) While falling under the purview of the Attorney General and the Department of Justice, it is meant to fulfill its duties in an impartial manner, regardless of who is president. However, in the current political environment, some Americans have argued that recent investigations by the Bureau have had partisan motivations. Such arguments became more prevalent during and after 2016, when investigations involving presidential candidate Hillary Clinton, as well as those concerning members of the Trump administration, were criticized as politically biased. Complaints often noted how these investigations may have directly impacted the outcome of the presidential election or the perceived legitimacy of the eventual winner. This growing tension between the FBI and political leaders is a recent development.

B. Framing 

A variety of news media outlets have disseminated criticism of the FBI. Zaller (1992) describes the transmission of news the following way: “Information that reaches the public is never a full record of important events and developments in the world. It is, rather, a highly selective and stereotyped view of what has taken place” (7).

These “stereotypes” are synonymous with the commonly used term “frames.” Frames help the public interpret events that, on their face, seem indecipherable (Zaller 1992). Mass media and political elites have the power to set frames of reference, which can signal how an audience should interpret and discuss complex political events. 

Based on this definition, party identification can be categorized as a type of frame. Individuals use partisanship as a frame to evaluate events happening around them. Druckman (2001) further analyzes political framing and its influence on public opinion. He defines the framing effect as the way a particular interpretation of an event, communicated by party elites, helps shape the public’s views. The framing effect is pertinent to this study because when party leaders speak about the FBI, they generate a frame through which individuals interpret the significance and implications of news involving the Bureau. By identifying that frame, I can discern its effect on the citizenry and its attitudes toward the FBI (Druckman 2001, 228).

Druckman (2001) also highlights certain characteristics that influence individuals’ susceptibility to framing effects. Critically, he finds that better-informed citizens are less likely to be swayed by external frames transmitted by party leaders, suggesting that the political knowledge of my experiment’s respondents should play a role in their receptiveness to the rhetoric they encounter.

The power of frames is not just a product of their content or citizens’ political knowledge, however. Messengers matter, as Druckman (2001) notes that the “persuasiveness” of a frame can hinge on who “sponsors” it (244). In my research, one highly salient sponsor of a frame for evaluating the FBI is the former president. Donald Trump holds a prominent place in the minds of Americans. Regardless of whether one views Mr. Trump in a positive or negative light, the frame he provides is difficult to ignore while developing opinions of the Bureau.

C. Political Elites and Voter Attitudes

In their groundbreaking book The American Voter, Campbell et al. (1960) explain why partisanship may influence evaluations of topics in public life. According to the authors, only a small percentage of the voting population displays “ideologue” levels of conceptualization. In other words, the majority of voters do not interpret political events through a consistent ideological lens. This information is pertinent when evaluating political attitudes. If most individuals do not apply ideological principles when consuming news, the way information is framed by well-known political elites may have a greater influence on the formation of individuals’ attitudes (Campbell et al. 1960, 218-223). 

Converse’s ([1964] 2006) work further supports these findings. He observes differences in belief systems held by elite political actors relative to those held by the wider public. A “belief system” is the organization of ideas and attitudes in which elements are linked by some type of constraint or functional interdependence (Converse [1964] 2006, 3). Average citizens often lack the systematic belief systems of the political elites who disseminate political messaging, creating an asymmetry that can be exploited for political gain (Converse [1964] 2006, 4). 

D. Party Identification

Party identification is defined as a lasting sense of psychological attachment to a political party. It is commonly measured on a continuum because it varies in direction and level of intensity. Partisanship is often a reflection, albeit an imperfect one, of the ideological platform that political parties and their leaders promote.

Parties serve to provide their supporters with information that individuals do not have the time or ability to experience directly (Campbell et al. 1960). Zaller (1992) adds to this theory by arguing that individuals are more likely to be influenced by elites from the same political party. Partisan elites signal how their supporters should categorize new information such that the conclusions drawn are consistent with their party principles. In other words, partisanship alters the digestion of information.

Van Bavel and Pereira (2018) further investigate this phenomenon, questioning why political affiliations alter perception of facts. They conclude that party identification incentivizes the distortion of the truth because exposure to information contradicting partisan beliefs can produce a state of extreme discomfort known as cognitive dissonance. People often try to avoid the tension spurred by encountering new information that appears to conflict with their existing beliefs. By applying partisan frames to the world around them, individuals can maintain ideological consistency. Party leaders regularly manipulate information to provide this consistency and help members of their base avoid feelings of dissonance. Of course, such manipulation may result in the misrepresentation or distortion of facts. This is important to account for when considering how political actors leverage party identification in their rhetoric about the FBI.

E. Shift in Perception of the FBI

Gallup conducted a survey in 2017 that measured positive and negative views of government agencies. The FBI ranked second highest among all agencies the last time this topic was probed in 2014 but fell to fifth place three years later (Brenan and Ander 2018). Despite receiving the same score in the 2014 and 2017 polls, the FBI’s relative ranking dropped significantly in 2017 because other agencies’ ratings increased. A more granular analysis of these trends reveals that the Bureau’s favorability fell 13 percentage points among Republicans but grew by 9 percentage points among Democrats (Bur 2018). 

These changes among party identifiers contributed to a 20-point partisan gap in ratings of the FBI by 2017, a result with no parallel even among similar agencies like the CIA, which enjoyed higher degrees of roughly uniform support across party lines (Bur 2018). This partisan division could be connected to the involvement of the FBI during the Trump presidency. President Trump has publicly criticized the FBI’s investigations into presidential candidate Hillary Clinton as inadequate, while simultaneously decrying the FBI’s “witch hunt” against him. As a result, Republicans and Democrats may have come to perceive the FBI’s action or inaction on particular issues as evidence that it favors one political party or the other, prompting party members to evaluate the Bureau’s performance through a partisan lens.

In recent surveys, measures of trust have been used to assess whether partisanship affects views of government agencies, particularly the FBI. For example, the McCourtney Institute for Democracy fielded the Mood of the Nation Poll in 2018, which indicated that party identification strongly correlated with individuals’ trust in the FBI. Given the political context, it was predictable that Republicans tended to have less trust in the Bureau than Democrats (Penn State Social Science Research Institute 2018). A 2018 Marist poll found similar partisan divides after comparing levels of trust in the FBI to measures of trust in President Trump. Forty-nine percent of Republicans believed the FBI was biased against Trump (Howland 2018). By contrast, only 10% of Democrats believed that the FBI exhibited bias against the then-president. Additionally, when asked who they would believe if President Trump and the FBI disagreed, only 28% of Republican respondents sided with the Bureau, compared to 94% of Democrats. These findings illustrate that Republicans and Democrats differed in their perceptions of the FBI and its interactions with President Trump (Marist Poll 2018). 

While such data are important, they only demonstrate a correlation between partisanship and attitudes toward the FBI. These studies cannot identify whether partisanship was the main driver of divergent views. Using an experimental survey, I can more directly test the hypothesis that partisan cues frame and ultimately predict views of the Bureau.

III. Hypotheses

This study tests the following formal hypotheses:

1. When exposed to President Trump’s criticism of the FBI’s Russia investigation, Republicans will display the greatest reduction in support for the FBI relative to the control group. They will view the Russia investigation as an attack on the legitimacy of Trump’s presidency and the Republican Party. Therefore, their partisan identity will be activated in accordance with partisan framing to generate negative political attitudes toward the FBI. 

2. When exposed to President Trump’s criticism of the FBI’s Russia investigation, Democrats will register the greatest increase in support of the FBI relative to the control group. They will view the investigation as a check on the power of the Republican Party and as a result, view the FBI more favorably.

3. When exposed to President Trump’s criticism of the FBI’s Russia investigation, less politically knowledgeable respondents will demonstrate the greatest drop in support for the FBI relative to the control group. I believe respondents with low levels of political knowledge will lack significant awareness of the Russia investigation. Thus, the criticism will generate negative sentiment toward the FBI because respondents will rely on the frame expressed by the critic for their own personal evaluations.

4. Politically knowledgeable respondents’ positions will be least affected by exposure to messaging, regardless of the content or the source of criticism. This is because politically informed citizens have already encountered information on President Trump, then-candidate Joe Biden, and the Russia investigation. As a result, they have already formed their positions on these people and topics, and do not need to defer to partisan heuristics or source frames when expressing their feelings toward the FBI.

5. President Trump’s criticism of the FBI Russia investigation will have a greater impact overall on respondents compared to criticisms of the FBI for misallocation of resources because the former issue is much more salient to partisans. 

An alternate hypothesis is that a respondent’s conceptualization of the importance of the law has a greater influence than partisanship on the development of political attitudes toward the FBI. If a respondent believes that strict enforcement of the law is important, they may disproportionately rate the FBI positively, regardless of the treatment. Conversely, if a respondent does not hold the law as a high priority, they may disproportionately express negative sentiments toward the FBI, regardless of treatment. If this is the case, the manipulation of type and source of criticism will not generate variance within the dependent variables. I account for this alternative explanation by including three questions that measure the respondents’ regard for the law. 

IV. Research Design

My main dependent variables assess political attitudes about the FBI. Specifically, I evaluate emotional responses toward the Bureau and perceptions of political bias. My main independent variables are type and source of political criticism of the FBI, which have been controlled to produce four different treatment groups, along with a nonpartisan control group. My experimental survey was divided into three parts and distributed to 542 respondents using Amazon MTurk in January of 2020. Figure 1 outlines my approach. 

Figure 1:

To develop the treatment, I produced original news articles in the style of BBC Breaking News. I picked this style because its format is only three to four paragraphs long, increasing the probability that the respondent would read the entire article. I also placed the primary signaling sentences within the header and first lines of the article to ensure the source and type of criticism were properly transmitted in case respondents failed to read the entire article. In addition, BBC News was selected as the media outlet because it is generally perceived to be freer of political influences than domestic news outlets.

 I selected Donald Trump and Joe Biden as sources of the criticism because each is well-known to the American public, held/holds high levels of political power in the United States government, and represents either of the major political parties. For type of criticism, I selected the FBI Russia investigation and allegations of misallocation of resources by the Bureau. As already discussed, the Russia investigation is relevant because it has been heavily publicized and pushed the FBI into the center of American political discourse. Conversely, the alleged misallocation of resources does not trigger any popular partisan frames nor threaten the legitimacy of either political party. For the control treatment, the type of criticism is devoid of politicized language, and its source and topic of criticism are both unknown to the reader. 

Respondents completed a pre-treatment survey asking them to identify their party identification and self-reported partisan strength. Political knowledge was also assessed via general questions about politics (i.e., to name the House minority leader). Political knowledge is an important variable to account for as it may be a moderator of my treatments. The pre-treatment also included a cluster of questions about obligations to follow the law and the effectiveness of the FBI to facilitate an evaluation of my alternate hypothesis. 

Following this first section of the survey, the respondents were randomly assigned to one of five different treatment groups, displayed in Table 1. Note that in contrast with Trump’s criticisms of the FBI’s investigation as an example of overreach and political bias, the nature of Biden’s criticism was that the FBI investigation did not do enough.

Table 1: Treatment and Control Groups

Following the treatment, respondents were questioned about whether they believed the FBI is a biased institution that favors one party. I then employed an emotions grid for each of the following topics: Donald Trump, Joe Biden, and the FBI. The grid asks “How often would you say you have felt each of the following ways because of ____, or because of something they have done?” A set of ten emotions was listed, such as “happy,” “fearful,” “angry,” and “relieved.” Respondents were instructed to fill out the grid for each emotion, identifying how frequently they felt that way about the topic on a 5-point scale from “Always” to “Never.”

Responses to the emotions questions were used to generate scaled dependent variables. An answer of “Always _____” (angry, happy, etc.) was recoded to equal 1, or 100%, while an answer of “Never _____” equaled 0, or 0%. The answers in between then corresponded to 0.75 (75%), 0.5 (50%) and 0.25 (25%). Each of the emotions variables was then regressed on the four treatment groups in order to make a statement regarding how each treatment increased or decreased the frequency of feeling a given emotion.

A. Sample Statistics 

My sample consists of 542 respondents, 358 (66%) of whom identify as Democrats or Independent-leaning Democrats, 159 (29%) who identify as Republicans or Independent-leaning Republicans, and 25 (5%) who identify as true Independents. Forty-three percent identify as strong Democrats and 18% identify as strong Republicans. The low percentage of Independents included in my sample is comparable to recent nationally representative estimates, with a 2018 Pew Research survey suggesting that only about 7% of American voters identify as true Independents (Pew Research Center 2019). However, a disproportionate number of respondents identified as Democrats or Independent-leaning Democrats relative to the American population. This is important to note as I approach my data analysis because it reduces the external validity of my findings.  

Fifty-one percent of the sample is male, 48% female, and 1% nonbinary. As for education, 13% of the respondents have completed a post-graduate program, and 43% of respondents have a bachelor’s degree. While gender is relatively well-balanced in the sample, educational composition is not representative of the American population. Similar to the partisan composition of the sample, the disproportionate presence of highly educated respondents reinforces the limitations of this study’s application to the broader population.

B. Generating Variables and Scales

First, I developed a scale to measure partisanship, in which “0” signified Republican identification, “1” signified Democratic identification, and “0.5” signified Independent. 

Second, I created a political knowledge scale using the questions shown in Table 2. Respondents who answered all questions correctly received a score of 1, while those who answered them all incorrectly received a score of 0. Those who only answered some questions correctly scored between 0 and 1.

Table 2: Political Knowledge Assessment

These questions vary in degree of difficulty by design. It was important to include difficult questions because, as previously mentioned, the Amazon MTurk sample pool is disproportionately more educated than the average American. By including challenging questions, I was able to effectively test the sample for political knowledge. As Figure 4 illustrates, the majority of my respondents have high levels of political knowledge, defined as scoring 0.5 or higher on the scale. 

Prior research indicates that there may be a connection between political knowledge and partisanship. For example, Pew Research Center (2016) found that Democrats are more likely to be highly educated and thus demonstrate higher levels of political knowledge, on average, compared to Republicans. However, my survey demonstrated little variation in the levels of political knowledge. Democratic respondents scored an average of 0.66 in levels of political knowledge, while Republicans averaged 0.61. 

A scale was also required to test my alternate hypothesis, which I call “Regard for the Law.” It incorporates the results from two questions administered during the pre-treatment survey, which are outlined in Table 3. 

Table 3: Regard for the Law Assessment

Responses to these two questions have a correlation coefficient of 0.43, which is moderately high. I also ran a Cronbach’s alpha test to justify creation of the scale with additional rigor. Taber’s (2017) work outlines the characteristics of the test, noting that scores between 0.6 and 0.7 are acceptable values for the scale reliability coefficient (1). Testing these two questions yields a reliability coefficient of 0.60, demonstrating that this scale is appropriate to incorporate into my data analysis. 

Figure 5 displays the distribution of respondents on the “Regard for the Law” scale, with a majority holding the law in fairly high regard. 

These variables and scales were essential to my data analysis. They allowed me to test how other variables interacted with my treatments, namely whether partisanship, political knowledge, or regard for the law had moderating effects on my results.  In the sections that follow, I will discuss key results from various regression models that I ran to analyze the relationships between my independent, dependent, and control variables. 

V. Findings

A. Emotions toward the FBI

The first important dependent variables to analyze are the measures of emotions toward the FBI. I found statistically significant results for the emotions of happiness, relief, anger, and fear. 

I. Emotions of Happiness

With regard to partisanship, I hypothesized that Republicans would have the greatest shift in negative emotions toward the FBI. My findings provide support for this hypothesis. In treatments of Presidents Trump and Biden criticizing the FBI for misallocation of resources, Republicans express feelings of happiness toward the FBI 21% less often than comparable Republican participants in the control group. Additionally, Republicans express lower happiness with the Bureau compared to Democrats, revealing a partisan divide in responsiveness to partisan cues for this emotion. 

Table 4: Emotions of Happiness Toward the FBI

Table 4A: Emotions of Happiness Toward the FBI by Partisanship

As for political knowledge, I hypothesized that respondents with low levels of political knowledge would express more negative emotions towards the FBI, a prediction that is not borne out by my findings. Such respondents were unaffected by the treatments, whereas politically knowledgeable respondents exposed to Biden’s critiques of the FBI’s resource misallocation express happiness toward the Bureau 15% less frequently than the control group. This is significant because it suggests that political knowledge had a moderating effect on the treatments.

Table 4B: Emotions of Happiness Toward the FBI by Political Knowledge

Next, I tested whether my alternative hypothesis impacted these results. If the alternative hypothesis were correct, I would find that respondents with low regard for the law would express more negative emotions toward the FBI, regardless of treatment. However, I did not observe statistically significant results for these individuals. In contrast, I found that respondents with high levels of regard for the law exposed to Biden’s resource misallocation critique express feelings of happiness toward the FBI 15% less frequently than comparable respondents in the control group. These findings show that respondents with high regard for the law view the FBI more negatively after treatment exposure. Potentially, the topic of criticism – misallocation of resources – could signal disorder at the FBI. As a result, this could cause respondents to feel more negatively about the agency. Further analysis must be conducted to draw a broader conclusion from these findings, but the main implication is that they refute the alternative hypothesis.

Table 4C: Emotions of Happiness Toward the FBI by Regard for the Law

II. Emotions of Relief

This study also found a statistically significant relationship between exposure to partisan frames and feelings of relief toward the FBI. Respondents exposed to Biden criticizing the Bureau over the misallocation of resources felt emotions of relief 9% less frequently than the control group.

Table 5: Emotions of Relief Toward the FBI

In terms of partisanship, it is apparent that Republicans drive this result. For Democrats, exposure to the treatments did not generate statistically significant changes in the frequency with which respondents expressed feelings of relief toward the FBI relative to the control group. In contrast, Republican respondents exposed to Trump’s critique of the FBI’s misallocation of resources felt relief toward the FBI 23% less frequently than Republicans in the control group. 

Somewhat more surprisingly, Republicans’ feelings of relief toward the FBI also dropped, occurring 19% less often, when exposed to Biden’s criticism. I had hypothesized that Republicans would be less influenced by critiques issued by Democratic partisan elites, as I believed in-party cues would generate greater emotional responses. However, these findings contradict my expectations given the mere 4-percentage point variation in responses when exposed to different partisan elites. Generating a similar pattern to the previous analysis, this partisan division demonstrates that partisanship does play a part in generating feelings toward the FBI, even if the result is more mixed than expected. This finding also supports my hypothesis that Republicans would experience more negative emotional responses to the treatments than Democrats in their ratings of the FBI.

Table 5A: Emotions of Relief Toward the FBI by Partisanship

When analyzing the effects of political knowledge, politically knowledgeable respondents drive the results. Respondents with high levels of political knowledge who were exposed to the treatment of Biden criticizing the FBI over misallocation of resources expressed lower feelings of relief toward the FBI. Respondents within the treatment group with high political knowledge expressed feelings of relief toward the Bureau 12% less frequently than politically knowledgeable respondents in the control group. This could be attributable to the greater awareness of current events that such respondents possess. Since this treatment likely highlights a new criticism of the FBI, it may cause a greater change among respondents who are more attuned to the potential ramifications of such a transgression. 

Table 5B: Emotions of Relief Toward the FBI by Political Knowledge

Like the results for emotions of happiness, there are no statistically significant changes in emotions of relief among respondents with low regard for the law. Similarly, this study did not yield statistically significant results for those with high regard for the law either.

Table 5C: Emotions of Relief Toward the FBI by Regard for the Law

III. Emotions of Anger

This study also found that partisan framing had a statistically significant relationship with emotions of anger about the FBI. Respondents exposed to treatments of any criticism by President Trump or criticism of the Russia investigation by then-candidate Biden expressed feelings of anger toward the FBI more frequently than the control group. 

Table 6: Emotions of Anger Toward the FBI

In terms of partisanship, the magnitude of the relationship between treatment exposure and frequency of anger felt toward the FBI was greater for Republicans than Democrats. Republicans exposed to the treatment of President Trump criticizing the FBI over the misallocation of resources felt anger toward the FBI 16% more frequently than Republican respondents in the control group. Democrats exposed to the treatment of President Trump criticizing the FBI over the Russia investigation expressed anger toward the FBI 10% more frequently than Democratic respondents in the control group. 

Table 6A: Emotions of Anger Toward the FBI by Partisanship

There was statistical distinction among respondents with varying levels of political knowledge. When exposed to any criticism by Trump, or criticism of the Russia investigation by Biden, individuals with high levels of political knowledge expressed emotions of anger toward the FBI 10% more frequently, on average, than politically knowledgeable respondents in the control group. Among respondents with low levels of political knowledge, these treatments did not generate statistically significant change in emotions of anger toward the Bureau. 

Table 6B: Emotions of Anger Toward the FBI by Political Knowledge

This study also found statistically significant relationships between exposure to partisan framing and attitudes toward the Bureau among respondents with both high and low levels of regard for the law. Respondents with high regard for the law who were exposed to then-candidate Biden’s criticism of how the FBI handled the Russia investigation expressed feelings of anger toward the Bureau 9% more frequently than corresponding respondents in the control group. These individuals also displayed the same percent increase in how often they felt anger when exposed to President Trump’s criticisms of misallocation of resources. Among respondents with low regard for the law, exposure to President Trump’s criticism of the FBI’s decision to investigate his campaign’s ties to Russian officials was associated with a 16% increase in how often individuals expressed feelings of anger toward the Bureau.

Table 6C: Emotions of Anger Toward the FBI by Regard for the Law

IV. Emotions of Fear

The final emotion for which this study found statistically significant results was fear. Exposure to the treatment of President Trump criticizing the FBI over the misallocation of resources increased how often individuals expressed feelings of fear about the Bureau.

Table 7: Emotions of Fear Toward the FBI

Republicans exposed to the treatment of President Trump criticizing the FBI over misallocation of resources expressed feelings of fear 13% more often than Republicans in the control group. In contrast, I did not identify any statistically significant associations between exposure to partisan framing and feelings of fear toward the FBI among Democrats. These results align with previous findings that Republicans had stronger emotional responses to the FBI after treatment exposure than Democrats. 

Table 7A: Emotions of Fear Toward the FBI by Partisanship

Respondents with high levels of political knowledge exposed to treatment of President Trump criticizing the FBI over the misallocation of resources felt fear 11% more frequently than politically knowledgeable respondents within the control group.

Table 7B: Emotions of Fear Toward the FBI by Political Knowledge

Respondents with low regard for the law who were exposed to the treatment of President Trump criticizing the FBI over the Russia investigation expressed feelings of fear toward the Bureau 14% more often than comparable individuals in the control group. I speculate these results occurred because respondents with low regard for the law may have been more fearful of the FBI to begin with, perhaps already believing it could not effectively or fairly enforce the laws. As a result, criticism highlighting the FBI’s flaws may cause these respondents to express more fear than respondents in the control group. 

Table 7C: Emotions of Fear Toward the FBI by Regard for the Law

VI. Discussion: Trends in Analysis 

In viewing these results, some trends emerge. For expression of positive emotions, like happiness and relief, only the treatments criticizing the FBI for the misallocation of resources generated statistically significant negative results. This finding runs counter to my hypothesis. I predicted that treatments involving the Russia investigation would generate more significant findings than the misallocation of resources. This may be a result of the increased publicity of the Russia investigation. Respondents may have already had strong negative emotions associated with the Russia investigation, making control group members more reactive to the issue than they may have been without prior knowledge of the incident and past exposure to political elite framing.  Exposure to additional elite framing may therefore have had less of an impact on individuals in the treatment group, relative those in the control group, if respondents from both groups had already formed strong opinions on the issue. It is sensible that introducing a new topic of criticism – the FBI’s misallocation of resources – would produce more pronounced differences in respondent emotions across the treatment and control groups.

Contrasting with these findings, we see both topics of criticism generate statistically significant results for the negative emotions of anger and fear. I believe this is because the expression of such emotions toward the FBI aligned with the partisan criticism displayed in the treatments. Pannico’s (2020) research supports this speculation as he explains the impact of in-party cues sponsored by political elites. In my study, respondents were incentivized to express strong negative emotions in an attempt to best align with their partisan identities.

Overall, Republicans drove the majority of the findings, displaying a greater emotional response to the treatments. Additionally, respondents with high levels of political knowledge generated more statistically significant results. Finally, respondents with high regard for the law accounted for most of the statistically significant results except for the results involving emotions of fear and anger toward the FBI. 

B. Is the FBI a Partisan Entity?

In addition to evaluating emotional responses to the treatment groups, I also sought to examine perceptions of politicization of the FBI. The major finding of my analysis is that respondents view the FBI as a politicized entity favoring the Democratic Party. I ran regressions to see how the treatment groups interacted with the dependent variable, belief that the FBI is a partisan entity. The question, located in the post-test, reads: “Do you think the FBI tends to treat both major political parties equally, or does it tend to favor either the Democratic Party or the Republican Party?” Respondents selected one of 5 different options ranging from “Strongly favors the Democratic Party” to “Strongly favors the Republican Party.”

When running an initial general regression on my entire sample, I found statistically significant results. Respondents exposed to treatments of President Trump criticizing the FBI over misallocation of resources and the Russia investigation were more likely to view the FBI as expressing favoritism towards the Democratic Party. 

Table 8: Perception that the FBI Favors Democrats over Republicans

Next, I examined the association between exposure to partisan framing and beliefs that the FBI is a partisan entity among respondents who identify as Republicans. I hypothesized that Republicans would perceive the institution to be more susceptible to political bias following exposure to President Trump or then-candidate Biden’s criticism of the Russia investigation. My findings support this prediction. Republican respondents exposed to treatments of Trump or Biden criticizing the Russia investigation, as well as Trump’s criticism of resource misallocations, viewed the FBI as more strongly favoring the Democratic Party. When exposed to Trump’s criticism of the FBI for the Russia investigation, Republicans’ perception of favoritism increased by 20%. As for the misallocation of resources, perceived favoritism for the Democratic Party increased 16% relative to Republican respondents in the control group. Finally, when exposed to Biden’s criticism that the FBI’s investigation of the Russia investigation had been inadequate, Republicans’ perceptions that the FBI had become a politicized entity rose 18% relative to comparable individuals from the control group.  

In other words, Republicans exposed to criticism dictated by Trump or involving the Russia investigation viewed the FBI as more strongly politically biased toward Democrats, even when the latter message was delivered by Biden. This finding is interesting because it demonstrates how partisan individuals, when signaled by an in-party sponsor such as President Trump, can be influenced to view the FBI as a biased entity whose goals are aligned with those of the opposing party. 

Finally, I investigated the association between exposure to partisan framing and beliefs that the FBI is politicized among Democrats to facilitate comparisons across party lines. Democratic respondents exposed to treatments of Trump criticizing the FBI over misallocation of resources and the Russia investigation increased perceived favoritism for the Democratic Party by 9% and 7%, respectively, compared to Democrats in the control group. However, these results were not as strong as those found among Republicans. 

Table 8A: Perception that the FBI Favors Democrats over Republicans by Partisanship

Despite my expectation that Republicans and Democrats would diverge in their views of FBI favoritism, my results suggest that exposure to elite framing increases how strongly the FBI is perceived to be a partisan agency biased toward the Democratic Party. However, there are still substantial differences in the strength of each partisan response to this measure. 

One reason that the Republican response could be stronger is because many of my treatments involve close scrutiny of President Trump and his allies through the Russia investigation. Both items can potentially act as signals to prime Republican identity, which could generate a stronger response. Since the partisan cues were mainly in-party for Republican respondents, they may have had greater incentives to align in accordance with the criticism. In contrast, Democratic respondents may not have been primed as effectively by an out-of-party cue, thus making their responses less intense. This is a pertinent finding that contributes to my overall hypothesis that people conceptualize the FBI as a political entity. Additionally, partisanship is employed when expressing opinions toward the FBI.

These findings are also inconsistent with my hypothesis that Democratic respondents would believe the FBI favored the Republican Party. Democrats may respond to the out-of-party cue despite its messenger because the resulting sentiment – a Bureau favoring their party – was in alignment with their partisan goals. These results may reflect the mentality of a familiar proverb: the enemy of my enemy is my friend. Democrats may view the FBI as favoring their party because exposure to Republican sponsored criticism puts them on the same political “team.”

Next, I analyzed whether political knowledge levels interacted with these results. I hypothesized that respondents with low levels of political knowledge would view the FBI to be more strongly politicized in favor of the Democratic party after exposure to partisan framing. My results were inconsistent with this hypothesis. However, I did find statistically significant results for politically knowledgeable individuals; those exposed to treatments of President Trump criticizing the FBI over misallocation of resources and the Russia investigation increased the degree to which the FBI was perceived to favor the Democratic Party. Additionally, respondents with high levels of political knowledge exposed to then-candidate Biden’s criticism of the FBI’s handling of the Russia investigation viewed the FBI as more strongly favoring the Democratic Party relative to comparable individuals in the control group.

Table 8B: Perception that the FBI Favors Democrats over Republicans by Political Knowledge

My analysis suggests that politically knowledgeable respondents were influenced more by the treatments than respondents with low levels of political knowledge. One potential explanation for this finding is that politically knowledgeable respondents were more likely to have pre-existing knowledge and opinions about the political parties, then-candidate Joe Biden, or President Trump that were challenged by the treatments. Exposure to treatments may have therefore produced stronger feelings of cognitive dissonance for these individuals, prompting them to re-evaluate the validity of their original positions. Meanwhile, respondents with low levels of political knowledge had less background context to reconcile with the narratives presented during treatment. Thus, it may be less likely that the frames included in this study contradicted their previous conceptualizations of such figures or events.

Finally, I wanted to determine if my “Regard for the Law” scale had any moderating effects on my finding that partisan framing is associated with stronger perceptions of FBI favoritism toward the Democratic Party. When compared against those from my baseline model, my “Regard for the Law” results (Table 8C) hardly differ from the original values in Table 8. Furthermore, the coefficients for individuals with low and high regard for the law are similar in magnitude across treatments, further reinforcing my conclusion that levels of political knowledge did not have a moderating effect on beliefs that the FBI is biased in favor of Democrats.

Table 8C: Perception that the FBI Favors Democrats over Republicans by Regard for the Law

Overall, these findings support my hypothesis that the FBI has become a politicized entity and is conceptualized in terms of partisan identity. This has significant implications for the functioning of the FBI within American society. The FBI is seen as favoring the Democratic Party by both Republicans and Democrats when exposed to treatments containing criticism from President Trump. In the minds of many Americans, the FBI has taken a side within the political arena. Whether this perception will change under the current administration remains unclear at this time.

VI. Conclusion 

My findings regarding the influence of partisan identity and elite framing on attitudes toward the FBI represent a substantive contribution to the literature. Additionally, I found that both Democratic and Republican respondents viewed the FBI as favoring the Democratic Party. The concept that the Bureau is viewed within a partisan frame by many Americans was not traditionally the case in survey research. The growing partisan divide may explain this recent shift, which has important implications for the future of the FBI as a federal law enforcement agency. 

The FBI depends on the confidence of the American people to uphold its mission of maintaining national security. Without broad public trust in its integrity, the FBI cannot perform its job properly. The FBI is dependent on outside sources and collaboration with private companies to carry out its mission. If this cooperation is tainted by perceptions of partisan preferences, the FBI is placed at an operational disadvantage.

Of course, circumstances have not been static since I conducted the survey at the heart of this study while President Donald Trump was still in office. Looking to the future, President Joe Biden’s presidency will likely impact attitudes toward the FBI in new ways. It remains possible that President Trump’s rhetoric could have a lasting effect on some individuals’ views of the FBI. Recent events, like the storming of the Capitol Building on January 6, 2021, also present case studies for future analysis of how partisan framing can influence public opinions toward law enforcement and other government entities. With the FBI back in the news cycle amid investigations into the Capitol attack, will a new opportunity arise for further shifts in public attitudes?  

This study indicates there is much left to explore on this topic, but my approach provides a framework for further research. It is clear that the role of the FBI has changed in the minds of many Americans, with the type and source of criticism influencing the activation of partisan identity in perceptions of the Bureau. The future of the FBI’s relationship with the public remains an open question, dependent on whether the phenomena observed here continue and not only erode perceptions of the agency, but also its capacity to function effectively.