How does rising political polarization impact the networks of LGBTQ+ individuals in the United States? Democrats and Republicans have become more polarized over the last 40 years with growing ideological distance (Pew Research Center 2017) and mutual dislike for the political outgroup (Abramowitz and Webster 2018). This polarization is also associated with both Democrats and Republicans having politically homogenous networks (Survey Center on American Life 2021). Homogeneity is fueled in part by individuals reducing ties with people who have different political views; one survey estimates that 22% of Americans’ friendships ended because of disagreements over Donald Trump (Survey Center on American Life 2021). LGBTQ+ individuals reported significant psychological distress after the 2016 Presidential Election (Veldhuis et al. 2018), but less is known about how these polarizing times impacted their social networks and social support.
LGBTQ+ people, in particular, are more likely to identify as Democrats (Pew Research 2016) and are more likely to be politically engaged compared to heterosexual adults, making politics a particularly salient issue for this demographic (Vandermaas-Peeler et al. 2018). LGBTQ+ people may have been especially motivated to reduce contact with others who supported policies targeting LGBTQ+ and other minority groups following the election of President Donald Trump. Within hours of taking office, the Trump Administration removed all inclusive mentions of LGBTQ+ people and rights from official public facing materials and, by February, announced that the administration would no longer uphold protections for transgender students in the United States (O'Hara 2017). The magnitude of large-scale, multisite protests of President Trump and his administration’s policies in 2016 and 2017 were unprecedented (Andrews et al. 2018) and brought together coalitions motivated by a range of issues, including LGBTQ+ rights, reproductive rights, racial equality, immigration, and climate change (Fisher et al. 2017).
While politics could be a factor in network loss, other factors, like closeness and family ties, may help to preserve ties with individuals who have different politics (Fischer and Offer 2020). However, this may work differently for LGBTQ+ people who, prior to the 2016 Presidential Election, already had differently structured networks compared with their cisgender and heterosexual counterparts. Although there are few systematic studies of LGBTQ+ adults’ networks, especially those of older LGBTQ+ adults, existing work suggests that LGBTQ+ adults have higher representation of other LGBTQ+ people in their networks and older LGBTQ+ adults have less diverse networks compared with younger LGBTQ+ adults (Erosheva et al. 2016). Other social and structural issues, such as homophobia, rejection, and policy barriers, have left many LGBTQ+ people with fewer spouses, children, and other kin ties (Carpenter and Gates 2008; Jones 2017; Reczek and Bosley-Smith 2021). Some LGBTQ+ individuals adjust for lower kin support by constructing their own “chosen families” that are strong ties without a direct familial bond (Heaphy 2009; Hull and Ortyl 2018; Knauer 2016).
These opposing processes thus suggest two countervailing outcomes related to polarization and network homogenization in the aftermath of the 2016 Presidential Election: 1) LGBTQ+ people may have been particularly motivated to reduce ties to individuals with different politics in the aftermath of the Trump election to reduce conflictual, nonsupportive relationships, or 2) LGBTQ+ people may have already had more politically homogenous networks and developed strategies for maintaining ties to individuals with different political views, like kin, to the extent that these individuals appear in their networks. To rigorously examine how the networks of LGBTQ+ individuals changed relative to their cisgender and heterosexual counterparts before and after the 2016 Presidential Election, we use a before-and-after design in a large panel sample of adults that includes a substantial oversample of older LGBTQ+ identified people. The 2016 Presidential Election was a particularly polarizing event (Gentzkow 2016) and serves as a useful moment to expand research on how network ties to individuals with different political views change in response to political events and among LGBTQ+ adults, and understudied and politically engaged population. In the next sections, we review key findings in the literature on political polarization and social networks and LGBTQ+ people’s networks. We then present our data, which is uniquely suited to address how networks changed among older LGBTQ+ adults in this period. LGBTQ+ adults and older adults in particular have differently structured networks from their heterosexual counterparts. Our study aims to evaluate how LGBTQ+ individual’s networks were influenced by a major political event.
Political polarization and social networks
The United States continues to become more polarized along partisan lines. Not only are Democrats and Republicans separating further on ideology (Pew Research 2017), but they are also growing in their dislike for each other (Abramowitz and Webster 2018; Finkel et al. 2020). Political polarization increases the strength and importance of political identities, which motivates partisans to process information that supports their own group while denigrating their outgroup (Van Bavel and Pereira 2018). Biased processing of political issues has been well-documented for Democrats and Republicans (Ditto et al. 2019), but such bias has grown to be quite broad. For example, Democrats and Republicans show a bias in hiring those who share their politics (Gift and Gift 2015), who they date (Easton and Holbein 2021), and often negatively stereotype each other (Deichert 2019). Studies have also shown how partisans will sometimes consider the political outgroup to be less than human (Martherus et al. 2021; Cassese 2021). A 2018 Pew Research Poll found that the majority of Democrats and Republicans believe that their differences go beyond policy as they cannot even agree on basic facts (Pew Research 2018). The broad bias of political polarization has been directly linked to the spread of misinformation (Van Bavel et al. 2020) and politization of health crises (Kushner Gadarian et al. 2020). Political scientists also express concern that such polarization threatens the foundation of our democracy (Levitsky and Ziblatt 2018).
Popular media reported on relationships ending during the 2016 Election (e.g., Filipovic 2016; Tavernise and Seelye 2016) and survey methods can help quantify the amount of network loss due to politics. A 2020 poll from the Survey Center on American Life found that 77% of Democrats and 75% of Republicans have politically homogenous social networks (Survey Center on American Life 2020). Social media also reveals significant political homogeneity as the average Facebook user was found to only have 23% of “friends” from the opposing political party (Bakshy et al. 2015). Not only do Democrats and Republicans avoid spending time with each other (Chen and Rohla 2018), they will even avoid conversing about nonpolitical topics such as sports or music (Settle and Carlson 2019). Lee (2021) found that close elections may even “close off” strong tied relationships because Americans are less likely to travel and have shorter Thanksgiving dinners following close elections. Thanksgiving dinners among politically diverse groups are also significantly shorter than those who agree on politics (Frimer and Skitka 2020). These studies are good indicators of a trend of political separation, but their designs do not directly test how political can change social networks overtime.
Social networks are an integral tool for evaluating the social outcomes of polarization as well as factors that contribute to further polarization. A 2020 Pew Research survey found both Trump and Biden supporters were less likely to have close friends with those who disagreed with their politics. This affects where people decide to live as well with 35% Democrats, and 50% of Republicans preferring to live in places with those who mostly share their politics (Pew Research 2014). Democrats are more than twice as likely to live in urban areas and Republicans are about twice as likely to live in rural areas across the United States (Mitchell 2018). This geographic and social distance further fuels political polarization. and could help explain why only 21% of Americans are in politically mixed marriages (Wang 2020).
Those with greater shared political agreement in their personal networks have stronger political beliefs (Bienenstock et al. 1990; Facciani and Brashears 2019) and the most extreme partisans are more likely to have politically homogenous networks on their social media (Boutyline and Willer 2017). Social influence does appear to impact beliefs directly as experiments show that having discussions with like-minded people increases belief strength (Keating et al. 2016) and respectful heterogenous discussions can decrease polarization (Levendusky and Stecula 2021a, b). Additionally, longitudinal network studies have found that people shift their ideological viewpoints to match their networks overtime (Lazer et al. 2010). Exposure different viewpoints does relate to polarization, but the type of relationships in the network matter. For example, exposure to opposing political information on social media does not reduce polarization (Bail et al. 2018) because following different Twitter accounts do not provide an opportunity for meaningful dialogue and reducing polarization through mutual understanding.
Facciani (2020) found that political heterogeneity within close networks was associated with reduced polarization. There was not an avoidance of discussion of politics with those who belonged to an opposing political group, but mutual respect was decreased in politically different network connections. Individuals are more likely to express political disagreement with those who they have strong ties with (Eveland et al. 2012). When people can learn more about the other side, they tend to become less polarized (Druckman et al. 2021; Levendusky and Stecula 2021a, b). Political disagreement can put stress on a relationship, but having a close connection appears to be critical for maintaining a politically different tie.
While researchers continue to uncover factors involved with political network composition, there is far less known about how political networks vary between different demographic groups. Race is a central political issue and could explain why White Biden supporters were twice as likely to have a Trump supporter in their network compared to Black Biden supporters (Pew Research 2020). Because marginalized groups are more likely to be impacted by current policies, political discussion may be especially relevant to them. The identities and human rights of LGBTQ+ individuals are often made into political issues, but there is a dearth of literature on how polarization shapes the networks of LGBTQ+ individuals.
LGBTQ+ and political networks
The average LGBTQ+ individual is more likely to be liberal and support the Democratic party (Edelman 1992; Egan and Edelman 2008; Lewis et al. 2011; Perrella et al. 2012; Smith and Haider-Markel 2002), although bisexual and transgender individuals are somewhat less liberal than lesbian women and gay men (Jones 2021). LGBTQ+ individuals are also significantly more politically active than their heterosexual counterparts (Egan and Edelman 2008; Swank 2018, 2019; Turnbull-Dugarte and Townsley 2020; Vandermaas-Peeler et al. 2018).
This increased political engagement could be caused by how often LGBTQ+ individuals have their rights directly challenged from policies. When Donald Trump became President in 2016 and Republicans regained the White House, there were significant concerns that previous protections of LGBTQ+ individuals would be removed. This may explain the increased stress and decreased mental well-being LGBTQ+ individuals experienced after the 2016 Election (Gonzalez et al. 2018; Price et al. 2021). Interpersonal connections are key to reducing this psychological distress. Greater social support and community involvement have been associated with positive health outcomes among LGBTQ+ individuals (Budge et al. 2013; Dyar and London 2018; Kertzner et al. 2009; Ramirez-Valles et al. 2005). LGBTQ+ identity and political identity also appear to be mutually reinforcing as a 2020 poll by the Williams Institute found that 72% of LGB Democrats agree that it is important to be politically active in the LGBTQ+ community. Thus, LGBTQ+ individuals may be especially motivated to reduce connections with those who add stress and conflict to their lives as a result of holding different, less liberal political views.
Importantly, prior to the 2016 Presidential Election, LGBTQ+ adults already had differently structured networks. LGBTQ+ individuals have a higher proportion of LGBTQ+ individuals in their networks while also reporting less kin (Frost et al. 2016; Grossman et al. 2000; Shippy et al. 2004). The 2013 Windsor and 2015 Obergefell Supreme Court decisions have increased marriage rates among gay men and lesbian women (Carpenter et al. 2021; see also US Census 2018); however, LGBTQ+ individuals are still less likely to have spouses and committed partners compared to their heterosexual counterparts (Badgett et al. 2021). LGBTQ+ individuals who are married are also more likely to be white and more highly educated (Badgett et al. 2021) compared to unpartnered LGBTQ+ individuals. Additionally, policies that limit adoption for LGBTQ+ couples as well as large costs associated with assisted reproduction results in LGBTQ+ individuals being significantly less likely to have children (Carpenter and Gates 2008). For a historically marginalized group like LGBTQ+ people, a lack of spouse, child, and other kin ties, especially in the household, is due to social and structural barriers, including homophobia, rejection, and policies that have historically limited opportunities for family formation, like same-sex marriage bans and laws limiting adoption and access to reproductive technologies for LGBTQ+ families.
Some LGBTQ+ individuals adjust for lower kin support by constructing their own “chosen families” that are strong ties without a direct familial bond (Heaphy 2009; Hull and Ortyl 2018). LGBTQ+ people also engage in “conflict work” to maintain ties with challenging members of their families of origin (Reczek and Bosley-Smith 2021). Thus, the ties that remain in their networks through adulthood and into midlife and older ages may be particularly resilient kin and nonkin ties. While studies of network change in the general population demonstrate less turnover among older versus younger adults (Bowling et al. 1995; Conway et al. 2013; van Tilburg 1998), we expect this may be especially pronounced for minority groups like LGBTQ+ people who whose networks may reflect the effects of personal, social, and structural effects in more pronounced ways.
The present study
The present study aims to describe the characteristics of network ties that hold different political views and investigates whether and how ties to individuals with different political views change after the 2016 Presidential Election. Drawing on network data collected before and after the 2016 Presidential Election, we also compare how network change before and after the election may differ for LGBTQ+ versus a random sample of cisgender and heterosexual adults. We first evaluate what types of relationships are most likely to contain political differences and predict politically network connections will be more likely to be kin or nonkin others because of political homogeneity within close networks (Facciani and Brashears 2019) and romantic partners (Easton and Holbein 2021). While individuals prefer political homogeneity, their kin may not share their political beliefs. We predict that politically different alters will be more likely to be kin compared to partners or friends.
H1
Alters with different political views are most likely to be kin or nonkin others (e.g., acquaintances, neighbors, know at work) rather than spouses/partners or friends.
Social support is vital as network losses can reduce one’s ability to cope with stress (Gerstorf et al. 2011). However, the 2016 Election caused significant stress for many Americans, and this could motivate individuals to decrease ties to others with different political views (Gonzalez et al. 2018; Price et al. 2021). While political disagreement can cause stress in any relationship, we predict kin will be preserved despite having different political views for both LGBTQ+ and non- LGBTQ+ adults.
H2
Kinship moderates the effect of different political views on alter loss for all adults.
-
H2a Kin alters with different political views are less likely to be dropped following the 2016 Presidential election.
-
H2b Nonkin alters with different political views are more likely to be dropped following the 2016 Presidential election.
LGBTQ+ people may have been especially motivated to reduce ties to individuals with different political views because of the salience of LGBTQ+ identities and other political issues during the 2016 election. However, they may also have already cut ties with politically different alters who reject them or developed coping strategies on how to maintain relationships with them (Reczek and Bosley-Smith 2021). Thus, we anticipate that there may be different patterns of loss and retention before and after the election for LGBTQ+ individuals compared to their heterosexual counterparts.
H3
LGBTQ+ adults have different patterns of alter loss following the 2016 Presidential election compared with cisgender and heterosexual adults.