The Relationship Between Social Media and Political Polarization Among Members of Generation Z

The Relationship Between Social Media and Political Polarization Among

Members of Generation Z

INTRODUCTION

There is no doubt among scholarly communities that social media is impactful in the way it

promotes images, ideas, and communication between groups of people. The debate about

social media has then become whether it is predominately harmful or helpful, which is a topic

that is broadly debated—from body image issues or increased violence to international

communication or business promotion. Social media has a profound ability to change the

methods in which we can articulate, change, and share discussions, and is considered as

separate in impact from other media sources. While studies have shown that media usage—

through television, radio, or podcasts—has been a factor in the formation of political

identities (Statista 2019; Choudhary 2020; Jurkowitz et al. 2020; Prior 2013), research on

social media’s impact on this field is sparse given its relatively recent development. Studying

social media usage and the formation of political identities in a generation that has grown up

with the existence of social media for its entire duration provides an interesting outlook into

the possibility of an even more polarizing medium than what had existed in the youth of

generations before. Because eighteen percent of adults in the United States reported social

media platforms as their preferred source for political and governmental issues in November,

2019, it is important to understand what correlation that may have with the formation of

political ideologies through social media (Jurkowitz et al. 2020). Social media is unique in

the way that it can “produce alternative forms of knowledge, create community, empower

users and revolutionize communications” which has been accredited in part for leading to

actual revolution as can be drawn from the Tunisian and Egyptian uprisings of 2011 (Nealon

and Giroux 2012:78). Knowing the power that social media holds in depicting stories and

feeding movements, I was left to consider its influence on the perception and practice of

politics by its users as well. 6 Through this study, I seek to understand whether this is true by

evaluating the political ideologies of ninety-four social media users within Generation Z

(born between 1997 and 2010) and how their social media experiences line up with and

encourage their political identities (Wood 2013). LITERATURE REVIEW Individuals

seeking out groups of like-minded individuals to share discussion with is no new concept.

Social Identity Theory developed by sociologist Henri Tajfel, and then later built on by

psychologist John Turner, is relevant to understanding the identification of oneself to a social

group, and what it means to be a part of that group. As part of this theory, an individual seeks

membership in what is defined as an “in-group”—which is a group of like-minded peers— as

opposed to being a member of the binary “out-group” (Tajfel and Turner 2004). To be

considered a member of an in-group one must follow a set of criteria including: self-identify

with said group, comparison of themselves to other members of the group using relevant

attributes, and comparison to a relevant out-group (Tajfel and Turner 2004). In the context of

political ideologies, my study follows the last criterion by developing two contrasting

groups—liberal and conservative—relying on the binary contrast between these two groups

to evaluate the presence of political polarization. It is also worth noting how important the

self-identity criterion is, as “social action can only be intelligible if we recognize that people

are guided to act by the structural and cultural relationships in which they are embedded and

by the stories through which they constitute their identities - and less because of the interests

we impute to them” (Somers 1994:624). It is still possible to police ideology and practice

within the in-group. Due to the stark contrasts the two groups provide for each other, there

exists a set definition of what ideology members should follow, and thus anyone whose

ideology remains slightly outside of the group is 7 encouraged to shape their own to match

the group’s ideology perfectly to avoid losing the support of the group (Tajfel and Turner

2004; Moon 2012). Within the context of my study, this ideological preset can result in

people not thinking they are “liberal enough” or “conservative enough,” thus pushing

themselves to fit more of the prescribed ideology to avoid criticism from the in-group—

which are the other members of their political alignment. This concept that an individual will

develop stronger political ideology by being a part of that social group follows the principles

set forth in Albert Bandura’s Social Learning Theory. Bandura 1969 makes the argument that

when an individual feels there is a reward for doing a certain action—whether that is voting

for a political candidate, participating in a protest, or sharing a political post on social

media—they will be encouraged to take that action again to feel supported by their peers.

They will then feel more rewarded in their practice of that political philosophy (Mutz 2002;

Bandura 1969). Most often the way an individual meets what is the “expected” practice of

their political ideology is through social models such as teachers, parents, and peers, but most

importantly those who hold prestige, power, and higher socioeconomic status such as

celebrities, public figures, activists, and political candidates (Bandura 1969). Social media

provides an outlet in which an individual is provided contact with those figures, as well as

their family and friends, so I found it essential to consider social media as a socialization

measure in political ideologies for individuals in Generation Z. The media sources one

chooses to follow typically align with what they already believe. It is through Limited Effects

Theory that this process is understood in its entirety. There has been evidence in favor of

media influencing political ideologies, yet it is highly probable that the ideology already

existed within the individual and had been simply reinforced by commentators from their

media sources (Choudhary 2020; Bode 2016). Through social media, individuals may 8

curate networks of users who think like themselves, post relevant content to their political

ideology, enable discussions within the group and in opposition to another group, and cycle

information they deem relevant. This constant reinforcement of the same content and

ideology leads users to stay within their own bubble of knowledge and become increasingly

involved in the political ideology they already follow (Bode 2016; Velasquez and Queenette

2018). It is given this context that my study seeks to understand what impact this careful

curation of social media has had on polarizing political practices. Limited Effects Theory has

shown up in practice of many different studies on social media, however, most focus on

networks like family, friends, and peers—as was mentioned through Dr. Bandura’s work. In a

2014 study done about Facebook and social mobilization, evidence shows that by seeing

friends and family post political opinions, news links, and news stories, an individual is more

likely to engage with what they see in front of them and become more interested in political

information and discussion (Theocharis and Quintelier 2014). As widely suggested, it is

through social media feeds—like those on Facebook—that individuals are able to see how

those inside their social networks are practicing politics, and in most cases use that model of

political practice to shape their own behaviors (Velasquez and Queenette 2018; Somers 1994;

Tajfel and Turner 2004). One Twitter study even suggests that by seeing political content that

opposes one’s own, individuals are more likely to become engrossed in their own political

ideology even further. The study required Democrats and Republicans in the United States to

follow accounts on social media that posted frequent political content that was opposite to

how the participants were identified (Bail et al. 2018). The outcome of this was Republicans

that felt even more conservative after being introduced to liberal-leaning content, and

Democrats 9 that felt slightly more liberal after being introduced to conservative-leaning

content (Bail et al. 2018). Given the ease of platforms like Facebook, Instagram, Twitter, or

Snapchat, and the audience most likely to use them, the question has often been asked

whether social media has effects on political participation offline as well. The current study

does not focus on this aspect of social media’s influence—I chose instead to analyze through

the lens of developing political ideologies and opinions, regardless of whether one acts on

them offline or not. However, political participation is still worth noting in hopes of better

understanding the interactions between social media and the political sphere. Social media

offers a different role than typical radio, television, and newspaper interaction, as will be

developed through the course of this paper. This new form of “digital participation” has been

thought to enable users to connect with politics and political movements online, and then

carry that participation into offline movements like voting and protests—especially for

younger generations (Nealon and Giroux 2012; Shah et el. 2005; Gil de Zúñiga et al. 2014;

Hsaio 2018). When users are engaged in political spheres online, and take to expressing

political ideologies of their own, it creates a center of communication between like-minded

people, thus motivating individuals to participate in social movements offline (Hsaio 2018;

Nealon and Giroux 2012; Mutz 2002). Without these platforms, individuals are “less likely to

express grievances; feel more socially isolated; feel emotionally detached to protest groups;

and, most importantly, think that they cannot do anything to redress their concerns,” thus

leading to political inactivity both on and offline (Hsaio 2018:3474). The best example of this

occurs when studying political spheres surrounding the 2008 and 2016 United States

presidential elections, where social media participation had proven to be beneficial to

political engagement. In 2008, the Obama administration integrated social media 10 and

Web-based communication into their campaign as a way to mobilize the younger generation

and create grassroots movements—which was not only one of the first campaigns to do so,

but was also successful in redefining the nation’s political culture at the time (Nealon and

Giroux 2012). In 2016, the behaviors of social media users online were starting to reflect

offline as well as seen through those users' political engagement, suggesting that how users

were interacting with politics on their social media platforms was influential to their practice

of politics outside of those platforms (Velasquez and Queenette 2018). There are many

reasons why these findings are consistent in suggesting social media as an outlet of political

circulation—first and foremost: social media is convenient. Numerous studies have credited

social media’s ease of circulation as a factor in spreading political ideologies. For example,

through social media, users can forward along messages through sharing, re-posting, linking,

that reduces the time an individual spends composing the message themselves (Gil de Zúñiga

et al. 2014). Thus, individuals use social media platforms as methods of communicating with

others and informing themselves through the networking, sharing, and messaging that takes

place at a rapid pace given social media’s easy user interface (Loader et al. 2015). Social

media is distinctive in its manner of sharing that also gives rise to voices not typically heard

through regular media platforms. While social media allows for self-expression and identity

formation—whether that be political or otherwise—it also holds the power of introducing its

users to other social groups that an individual may not have ever encountered if not for its use

(Gil de Zúñiga et al. 2014). This provides a space for political discussion of all kinds,

including the voices of marginalized communities who would like for their issues to be heard

and spread in a manner that tend to be excluded from traditional conversation in civic,

political, cultural and social spheres without intervention (Keating and Melis 2017). This is

11 arguably successful, as the example of the Black Lives Matter movement shows people

getting more upset about racism and police brutality after the mistreatment of people of color

in America was publicized over social media (Duggan and Smith 2016). It is in this manner

that social media’s convenience proves to be an essential factor of its widespread use, and

thus impact on social and political circles. Arguably, the generations most impacted by this

appeal in political engagement are those in younger age demographics who may use social

media more than generations above them. The current study follows the work of Boxell,

Gentzkow, and Shapiro (2017) who looked at which age demographics were the most

polarized, and if social media was a factor. Their findings argued “against the hypothesis that

the internet in general or social media in particular are the main drivers of increasing

polarization,” but this was largely relevant to the older age demographic they identified,

which consisted of individuals ages 65 and older (Boxell et al. 2017). While this study found

the largest amount of polarization in the oldest age demographic, the researchers did not

discredit the impact of social media among younger generations. Instead, they concluded that

social media may very well be a factor that increases polarization in younger generations, but

that whatever factor was increasing polarization in older generations was not social media

and was stronger in influence than social media (Boxell et al. 2017). Thus, the research that

follows does not directly contradict these findings, but rather offers an alternative lens to

identifying the correlation between social media and political polarization in younger

generations that was not considered above. Many other social media studies neglect to isolate

younger generations as a population to study individually, which is where the current study

seeks to expand knowledge. The differences in age groups are important, as once again, the

younger generations have grown up with social 12 media and the internet woven into their

daily lives as a method of communication and socialization that generations before may not

have fully experienced. A political participation study found this to be true as “the Internet

has an impact on the information seeking behavior of the younger age groups, who are the

heaviest users of the technology and who were also the most likely to use the Internet to look

for political information” (Gennaro and Dutton 2006:312). Another study from 2012—which

analyzed the same age group that I do, although earlier in their youth—reinforced the concept

that internet usage has a huge impact on how individuals are able to express and share

opinions and concerns, and even suggested huge potential in the future impact of social

media in mobilizing younger generations to be more active in civil and political engagement

(Lee et al. 2012). Given the context of previous research, as well as what can be expected

from Social Identity, Social Learning, and Limited Effects Theories, I designed the current

study with the assumption that political polarization would be apparent through the use of

social media platforms. The proposition that guided the following study can be stated as such:

given that Generation Z has grown up with the presence of social media, social media can be

expected to foster polarization in the political practices of Generation Z through the content

and ideologies that it presents and encourages. METHODS To conduct this research, I gave

an anonymous survey to currently-enrolled Otterbein students who were born between the

years 1997 and 2010 to constitute them as a member of Generation Z (Wood 2013). Once

approved by the Otterbein University Institutional Review Board, the survey was given

online via Microsoft Forms, and students received no compensation for their participation

(See Appendix A). Given the population and the topic studied at hand, an 13 internet survey

was the best method of reaching this demographic as those who use social media would

already have access to the internet, thus ensuring accessibility for the study population.

Overall, the advantage to this method is the creation of an accessible and anonymous

atmosphere for respondents to answer about political ideologies and topics without the

presence of an inperson interviewer to give social clues and inflict bias on respondents’

answers (Babbie 2017). The survey contained twenty questions and the option to leave the

survey at the end to take the Political Compass test (www.politicalcompass.org) and report

back findings. Participants were first asked demographic information such as their age,

gender, and ethnicity (For survey text see Appendix A). The rest of the survey then asked

questions relevant to measuring polarization on social media by first asking what social

media platforms they frequented the most. From there I measured polarization using three

categories: hostility towards others because of politics, reinforcement of political ideology,

and political identification. Hostility towards others was addressed in five questions

throughout the survey that considered whether participants would block someone on the basis

of the political content they posted,1 if they felt the political party opposite to theirs was

destroying the country or evil, if that message was supported with the political content they

saw, if they engaged in political discourse on social media, and whether that left them feeling

isolated as a result. Six questions throughout the survey dealt with the reinforcement of

political ideology, asking if the political posts they saw on social media aligned with their

own beliefs,2 if the users they followed aligned with their political beliefs, if their political

beliefs were idolized by other users, if they felt validated by engaging 1 They were given the

options Yes, No, and Other with the option to explain themselves. From the Other responses,

answers were separated into categories based on if their responses would lean towards no,

lean towards yes, or if it would be conditional on the content or subject at hand. 2 For this

question, Never was measured as 0 times a week, Sometimes as 1-4 times a week, Frequently

as once a day, and Very Frequently as more than once a day. 14 with other users, if they

purposefully sought out political content on their social media platforms, and if the public

figures they followed tended to align with their political beliefs. Political identification

covered the last questions of the survey which asked participants to self-identify on a scale of

zero (liberal) and ten (conservative), and then afterwards were given the option to take the

Political Compass test and report back their coordinates for comparison of where they

depicted themselves to fall with their political ideology, and then where a test placed their

political ideology. The Political Compass Test The Political Compass Test for measuring

political ideologies was chosen because of the similar Left/Right positioning it offers to the

zero to ten liberal/conservative rating scale given in the survey. The Left/Right spectrum on

the Political Compass Test seeks to describe economic positioning, coupled with the

Authoritarian/Libertarian scale that is used to measure positions on social issues vertically

(see Figure 1). Given the nature of how this test identifies political ideologies, one would

expect to see individuals that ranked themselves rather conservatively on the self-depicted

political ideologies scale to have coordinates in the first quadrant of the Political Compass

Test, whereas scores on the liberal side of the scale would be more likely to end up in the

third quadrant. Whether or not self-depictions of political ideologies matched where they

were anticipated to end up in the Political Compass Test was a measure of indicating how

one’s perception of their own identity could differ from their evaluated identity using this test

as a way of ensuring criterion-related validity. This helps in determining whether it was

possible for participants to have unknowingly been more polarized in their political ideology

than they had anticipated. 15 The Political Compass Test has also been critiqued for its

evaluation of the political spectrum. For example, Kastelein (2018) has pointed out the

Political Compass’s single-axis coordinate system is not as effective in measuring multi-

faceted political ideologies and suggests a more triangular or circular approach to

coordinating identities (Kastelein 2018). However, the research done in the current study is

focused on binary political ideologies as is necessary to analyze political polarization. For

this reason, the nature of the Political Compass Test suits the simplicity of the liberal to

conservative scale used towards the end of the survey. RESULTS Descriptive Statistics The

survey came back with 96 responses at the end of the survey period, 94 of which were used in

the following data analysis (Table 1).3 The ages of respondents considered ranged from 18 to

22, with an average of 19.55 years old. Participants were asked to select one gender category

that best applied to them which resulted in 71 Female-identified respondents (75.5%), 20

Male-identified respondents (21.3%), two Non-binary respondents (2.1%), and one

respondent who answered Genderqueer when prompted in the “Other” category (1.1%). With

regards to ethnicity, 79 participants self-identified as White (84%), nine as African American

(9.6%), nine as Asian (9.6%), three as Hispanic/Latinx (3.2%), two as Native

American/Alaskan Native (2.1%), and two as Other (2.1%).4 3 Two surveys had to be

excluded on account of the respondent’s age. The survey was interested in respondents with

ages falling within the range of Generation Z, and thus a 34-year-old was not considered part

of the target population. On the other hand, one 17-year-old respondent was not included on

the basis of age as well as requested by the Institutional Review Board to exclude this minor.

4 Although Pacific Islander was given as an ethnicity option, there were no responses for this

category. This question allowed multiple selections, and thus 8.5% selected two or more

ethnicities are considered “mixed race” for the purposes of this study. 16 Survey Questions I

then asked participants to answer a series of questions about political ideology and social

media which started with what social media platforms they used most frequently (Table 2).

Eighty-three respondents used Instagram (88.3%), 82 used Snapchat (87.2%), 71 used TikTok

(75.5%), 53 used YouTube (56.4%), 32 used Twitter (34%), 29 used Facebook (30.9%), 18

used Pinterest (19.1%), and two used an “Other” social media platform (2.1%).5 On these

social media platforms, six respondents encountered a post containing political content on

their feed Never (6.4%), 38 Sometimes (40.4%), 26 Frequently (27.7%), and 24 Very

Frequently (25.5%). Fifty-seven respondents (60.6%) noted getting their news from social

media first as opposed to a different news medium, and 32 respondents (34%) noted they

purposefully followed political topics on social media through news accounts, activists, or

law makers. To evaluate political content, participants were asked how often political posts

on their feeds aligned with their own political beliefs; eight said Never (8.5%), half said

Sometimes, 36 said Usually (36%), and three said Always (3.2%). Seventy-three respondents

felt like they had been exposed to a new political ideology through social media (77.7%).

Four respondents noted None of the users they followed on social media shared their political

beliefs (4.3%), whereas 25 noted Some (26.6%), 29 noted At Least Half (30.9%), 34 noted

Most (36.2%), and two noted All (2.1%). Out of the celebrities, public figures, and/or

influencers they followed, three felt like they posted political content that Never aligned with

their beliefs (3.2%), 37 Sometimes aligned with their beliefs (39.4%), 39 Mostly aligned with

their beliefs (41.5%), five Always aligned with their beliefs (5.3%), and 10 did not follow

celebrities, public figures or influencers (10.6%). 5 Participants who answered Other were

asked to fill in their other source. One said Discord, the other said tumblr. 17 Fifty-four

respondents felt like their political beliefs were often reinforced or idolized by other users on

social media (58.1%). Similarly, when participants were asked if they felt validated by

engaging with users who shared their political beliefs on social media, the responses were

split 50/50. When asked about how many times they engaged in political discourse on social

media, half of participants responded Never (51.1%), 37 Rarely (39.4%), eight Sometimes

(8.5%), and one Always (1.1%). If participants answered Rarely, Sometimes or Always to the

question above, they were then asked how many times the interaction left them feeling

further away from those with opposing political beliefs to which four answered Never

(8.7%), 27 Sometimes (58.7%), nine Mostly (19.6%), and six Always (13%). Sixty-six

percent of participants felt the political ideology that countered or opposed their own was

destroying the country or evil when asked, and when asked whether that message was

supported in the political posts they saw on social media, and the results were the same with

62 Yes (66%). Participants were also asked if they have or would ever unfollow, block, or

unfriend someone who posted about political content that did not align with their beliefs.

Forty-eight participants fell into the Yes or Leans Towards Yes category (51.6%), 36 No or

Leans Towards No (38.7%), and nine were Conditional on Content/Subject (9.7%). Finally,

participants were asked to evaluate their own political ideology on a scale from 0 (liberal) to

10 (conservative). Fifty percent of the participants fell between the 0 to 3 liberal range with

17 answering 0 (18.1%), seven answering 1 (7.4%), ten answering 2 (10.6%), and 13

answering 3 (13.8%). Thirty-four percent of the participants fell in the middle range from 4 to

6 with ten answering 4 (10.6%), 16 answering 5 (17%), and six answering 6 (6.4%). Sixteen

percent of the participants fell between the 7 to 10 conservative range with five answering 7

18 (5.3%), three answering 8 (3.2%), two answering 9 (2.1%), and five answering 10 (5.3%).

The mean political ideology for this sample was 3.71, leaning towards liberal. The mode for

this sample was 0, the most liberal value, followed by 5, the center value. Thirty-five

participants elected to take the Political Compass Test at the end of the survey. They were

asked to report the coordinates the test had given them for the Economic Left/Right (X-

Coordinate) and Social Libertarian/Authoritarian (Y-Coordinate) measures, both of which

were evaluated on a scale from 10 to –10. The mean coordinate set from these responses was

-3.51, -3.92, which puts the average in the third quadrant where we would expect to see

predominately liberal ideologies.6 The average of political ideologies from the 10-point scale

mentioned above for only these 35 participants was 2.54, leaning liberal. Bivariate Analysis

Firstly, I looked at age in comparison to how likely an individual was to engage in discourse

over social media about political content. Participants aged 18 were the most likely to engage

in discourse with 4.3% saying they Always did, 17.4% Sometimes did, 26.1% Rarely did, and

52.2% Never did (Table 3). This was followed by participants aged 19, 14.3% Sometimes

did, 42.9% Rarely did, and 42.9% Never did. For participants aged 20, 52.9% Rarely did, and

47.1% Never did. For participants aged 21, 43.5% Rarely did, and 56.5% Never did. Lastly,

participants aged 22 had a 100% response rate for Never engaging in political discourse on

social media. Female-identified participants were much more likely to engage in discourse

comparative to other genders (Table 4). Female respondents noted 1.4% would Always

engage in discourse 6 There were thirty-five respondents, however one elected not to share

their Social Libertarian/Authoritarian coordinate. 19 on social media over political content,

9.9% Sometimes, 40.8% Rarely, and 47.9% Never. This is followed by Non-binary and

“Other” respondents who 100% fell in the Rarely category. Lastly, 5% of Male respondents

noted Sometimes, 25% Rarely, and 70% Never, leading Males to be the gender least likely to

engage in political discourse on social media. Male participants were also the least likely

demographic to block, unfollow, or unfriend a user on social media over political content

(Table 5). Twenty-five percent of respondents said they would block, 70% said they would

not or lean towards not blocking, and five percent were conditional on the subject presented

in the political content. Out of female respondents, 57.2% said they would block or lean

towards blocking, 31.5% would not block or leaned towards not blocking, and 11.4% were

conditional on the subject matter. One hundred percent of participants who responded Non-

binary or “Other” said they would block. When considering gender in comparison to whether

a respondent thought the other political ideology was destroying the country or evil, 100% of

Non-binary or “Other” respondents said Yes, they did feel like the other side was evil (Table

6). This was followed by 66.2% of Female respondents, and 60% of Male. I also considered

whether a participant who thought the other side was destroying the country or evil would be

more likely to block a user on their social media over political ideology (Table 7). Out of

those who said they thought the other side was destroying the country or evil, 62.3%

participants said they would block, 8.2% said it would be conditional on content or subject,

and 29.5% said they would not. For participants who did not feel like the other side was

destroying the country or evil, 31.3% said they would block, 12.5% said it would be

conditional on content or subject, and 56.3% said they would not. 20 Social media platforms

had varying rates of polarization, explored using the number of users who felt the other side

was destroying the country or evil (Table 8). YouTube had the highest rate, with 75.5% of

users feeling like the other side was destroying the country or evil. This was followed by

72.2% on Pinterest, 67.6% on TikTok, 65.6% on Twitter, 65.1% on Instagram, 63.4% on

Snapchat, 50% on Other social media platforms, and 44.8% on Facebook. When considering

gender as a factor, the outcome of the political ideologies scale is also interesting to note

(Table 9). Using the categories as 0 to 3 being liberal, 4 to 6 being moderate, and 7 to 10

being conservative, 25% of Male respondents fell in the liberal categories, half fell in the

moderate categories, and 25% fell in the conservative categories. Female-identified

respondents, however, fell 54.9% in the liberal categories—with 22.5% of those responses in

the 0 liberal value alone—followed by 31% in the moderate categories, and 14% in the

conservative categories. Respondents who identified as Non-binary or Other fell 100% in the

liberal category.7 Lastly, I analyzed the relationship between where individuals self-identified

on a political scale and their likeliness to block an individual on social media based on

political ideology (Table 10). Starting off with the most liberal value, 70.6% of respondents

who answered 0 would block someone based on political content, 71.4% of respondents who

answered 1 would, half of those who answered 2 would, 53.8% of those who answered 3

would, 40% of those who answered 4 would, 26.7% of those who answered 5 would, 16.7%

of those who answered 6 would, 40% of those who answered 7 would, 33.3% of those who

answered 8 would, half of those who answered 9 would, and 60% of those who answered the

most conservative value 10 7 There was one non-binary respondent, and one respondent who

noted they were genderqueer when prompted in the “other” category. 21 would. This shows

those who identify towards the poles—either liberal or conservative—are more likely to

block those they disagree with about political content on social as illustrated in Figure 2.

DISCUSSION AND CONCLUSION The purpose of this research is to analyze the

relationship between social media and political polarization in members of Generation Z, in

which the results favor a correlation between more polarized political identities and practices

over social media. Political polarization is measured in this study by three categories:

hostility towards others because of politics, reinforcement of political ideology, and political

identification. Firstly, hostility towards others because of politics could be seen best in the

question that asked whether participants thought the political party opposing theirs was evil

or destroying the country, in which 66% noted they did feel that way, and that that message

was reinforced by their social media platforms. When participants were asked whether they

would block or unfriend someone on a social media platform based on political ideology,

51.6% noted they would. For 9.7% it was conditional on the subjects they posted about.

Willingness to block was also recorded in higher percentages for those who fell closer to

liberal or conservative poles, with those who fell in the middle of the self-identified political

ideology scale the least likely to block someone based on political content. The only

questions in this category where the majority of responses did not seem to support the general

hostility principle was in whether the participants would engage in discourse on social media

over politics—where the majority of respondents noted they Never would. As mentioned in

Bandura’s work with Social Learning Theory, individuals are more likely to engage with

content when they feel there will be a reward for doing so (Bandura 1969). By engaging with

members of an opposing political party who is telling an individual that their 22 ideology is

wrong, that individual is less likely to feel rewarded by engaging with them on their social

media feed, and likely resort to blocking the other individual, feeling as if they are evil

because they fall out of the group of peers they are used to rewarding them, or even in rare

cases engage in discourse surrounding the political ideology they disagree on. In my study, I

found this to be the case with the 51.6% of participants who would block those they disagree

with, increasing to 61.3% if including those responses where blocking is conditional on

subject. In Bail et al. (2018), this concept was reinforced by showing Twitter users who found

themselves feeling even more dedicated to their own political ideology after being introduced

to content that countered their own, instead of coming to a more middle-ground approach

(Bail et al. 2018). As in the current study, Bail et al. (2018) suggests that the hostility

principle in play has a hand in political polarization by creating users who are more likely to

feel negatively towards those who hold the political ideology opposing theirs, and more

positively towards those who hold similar political ideology, as predicted by Social Learning

Theory. Thus, the second category in evaluating political polarization becomes relevant:

reinforcement of political ideology. Many participants noted following users who aligned

with their beliefs; only 4.3% accounting for following no users that aligned with their beliefs,

and 26.6% who noted only some, the rest fell into At Least Half, Most, or All categories. For

those who followed celebrities, influencers, public figures and the like, this observation was

supported again with only 3.2% of respondents who thought none of the celebrities they

followed were aligned with their own political ideology, whereas 46.8% felt like they Mostly

or Always aligned. The only category that I saw to be slightly less reinforcing of the users

own political ideology was when looking at whether the posts seen on social media aligned

with the participants political ideology where half of respondents felt it Sometimes did, and

38.3% it 23 Usually did, suggesting a more even distribution of political posts that fell

outside of the polarizing Never or Always categories. Regardless, 58.1% of participants felt

their political ideology was idolized by other users on their social media. Interestingly, when

asked whether participants felt validated by engaging with users on social media who shared

their political ideology, the results were split down the middle, suggesting a different

conclusion than what would have been expected using Bandura’s work to evaluate

reinforcement of political ideology on social media, which might have leaned more towards a

majority of participants feeling validated by engaging with those like themselves. Given the

previous research primarily relating to Limited Effects Theory, I expected to see the majority

of participants engaging with political content that was similar to their own political

ideology—thus ensuring reinforcement—and for the majority of questions relating to the

reinforcement principle in this study, this seems to be the case (Choudhary 2020; Bode 2016;

Velasquez and Queenette 2018). That is not, however, to say that the responses

overwhelmingly support the reinforcement principle, but more that the majority of cases

appear to sway in that direction and suggest a correlation between political polarization

through reinforcement and how participants chose to curate their social media platforms. This

is consistent with previous findings such as Choudhary (2020), Bode (2016), and Velasquez

and Queenette (2018) that suggested a fair amount of political reinforcement through network

formation, thus leading to groups of individuals who all think alike and tend to stick with

those who will support their preexisting beliefs. Social Identity Theory in Tajfel and Turner

(2004) laid the basis for evaluating the third polarization component in this study: political

identification. Social Identity Theory outlines the tendency for an “in-group” to contrast

themselves with a comparative “out-group” 24 (Tajfel and Turner 2004). For the purposes of

this study, this is understood as strongly liberal ideologies contrasting with strongly

conservative ideologies, thus creating two binary groups with little room left for identities in

the middle. What this study found however, was 65.8% of respondents who fell at one of the

“poles” of either liberal or conservative—with 49.9% of respondents that fell into the liberal

category and 15.9% that fell into the conservative category—and 34.2% that fell into the

middle range. The mean to the political compass test fell at coordinates -3.51, -3.92, with the

majority of responses falling into what I identified for the purposes of this study as the liberal

quadrant, suggesting a fair amount of respondents who tended to fall in more distinctly liberal

categories than middle ground. This remains consistent with the liberal average 2.54 for the

self-depicted political ideology scale for those same respondents, but since taking the

Political Compass Test remained optional for participants, it is hard to truly identify if this

pattern would have existed for the entire sample, and not just the 35 who responded to these

questions on the survey. This comparison of perceived political ideology through self-

identification compared to identification of political ideology given a universal standard

could be an area for further research. Using the three categories of political polarization

outlined above, I seek to identify the relationship between the practices of users on social

media surrounding politics and how polarized those users are because of their engagement

with other users both in and outside of their own political ideology. While the findings above

present evidence in favor of witnessing political polarization on social media, it is not to say

that the relationship between these two are causal—as would need to be explored in further

research. This study instead suggests that political polarization measures have been observed

on social media, but cannot safely say that social media is the sole contributor of polarized

political interactions in the sample studied. 25 Notably, the generation studied is different in

its experience with social media, as it has been a generation to grow up with social media,

and therefore has been influenced more by its content than those of generations before. Social

media is not the primary source of political polarization among older generations as seen in

Boxell et al. (2017), but further research might seek to explain just how different political

polarization trends are between Generation Z and those before it to identify if social media

truly has had such a unique impact on Generation Z given their exposure to social media from

such a young age. In this study, I recorded the experiences of social media users in

Generation Z in hopes to observe patterns of political polarization, but these findings are not

compared to patterns of any older generations, and thus cannot be ruled conclusive of

identifying Generation Z as engaging with politics online any differently than those in

different age groups, more suggesting this as an area for further research as well.