The Changing Landscape of the Relationship between US Professional Sports and Politics
Introduction
Following the nation-wide reckoning with racism and White supremacy in the US in 2020, many professional athletes, coaches, and the like poured their energy into advocating for racial justice, police reform/abolition, and voting rights. As many leagues had to shift their arena policies to have spectator-less games in accordance with CDC guidelines during the COVID-19 pandemic – the National Basketball Association (NBA) going as far as creating a ‘bubble’ in Orlando, Florida – their athletes amplified their messaging on social justice and combatting racism in the wake of the revived Black Lives Matter (BLM) movement (Shelburne & Andrews, 2020).
At its core, professional sports are rooted in ideals of racial capitalism that focus on the pleasures of consumerism (Montez de Oca, 2021) – professional sports leagues are a form of entertainment, with individual teams being valued at over several million dollars each (Ulas, 2021). In sports especially, this focus on consumerism works to empower Black athletes who have lived experiences of racism and often come from marginalized communities themselves. Black athletes, particularly in the NBA and National Football League (NFL), become celebrities in their own right, often providing commentary on social and political issues. However, when those athletes are empowered to dismantle systems of oppression, such as the institutions of policing in the US, the leagues themselves must strategically shape the manufactured illusion of support for their athletes that the viewers and fans see (Montez de Oca, 2021). These shifts have been seen in the last decade with the fierce condemning and blacklisting of Colin Kaepernick from the NFL in 2016 to the painting of “End Racism” on their fields four years later. A 2020 Nielsen study found that 83% of NBA fans and 81% of NFL fans supported athletes in their racial justice efforts, and 76% of NBA fans and 72% of NFL fans supported the Black Lives Matter movement (Lee, 2020); there is a clear business imperative in supporting athletes in their efforts to combat US racism. These examples of performative activism, such as preapproved social justice slogans or campaigns featuring other Black celebrities, are indicative of a shift in the public imagination of what a Black athlete can accomplish with their platforms. Historically, there has been a clear disconnect between the justice activism of athletes and the sociopolitical messaging by the leagues they play in, contributing to a sort of “shut up and play” mentality. However, following the resurgence of BLM and the massive steps that leagues took towards supporting these athletes, this disconnect seems to be fading – at least from a fan’s perspective.
But how does the actual business of sports contribute to a league’s political action? Are team owners – financial elites with the means to make large contributions to campaigns – aligned with their players’ and leagues’ social and political values? Overall, American professional sports team owners have spent upwards of $47 million in federal elections since 2015 (ESPN & FiveThirtyEight, 2020). The monetary contributions of those in front-office positions within these leagues is hard to ignore – sports ownership and high-level American politics have a long, connected history. A major example can be seen in Larry O’Brien, a leading political operative who held a key role in both John F Kennedy and Lyndon Johnson’s presidencies who later became the Chair of the Democratic National Committee. O’Brien acted as the NBA commissioner from 1975 to 1984 and is memorialized through the Larry O’Brien Trophy, the award given to the NBA Finals champion (Arnovitz, 2020). Additionally, NBA team owners and commissioners, such as David Stern, Herb Kohl (a U.S. Senator), and Larry Weinberg, have been heavily involved in American politics. Collectively, NBA owners donate more than $4 million per federal election cycle (Arnovitz, 2020).
There is a clear connection between professional sports ownership and politics. Sports ownership has become a role for the ultra-elites, and their political aims are not lost once they enter the professional sports arena. However, it seems that many of the donations made by these owners tend to reflect their personal gains, based on whatever industries they might have stakes in. For example, according to ESPN (2020), NBA Miami Heat owner Micky Arison is heavily involved in the cruise ship industry – when looking at his political donations in the past several years, his contribution patterns are not very uniform by party or political ideology, with donations going to both conservative Republicans and progressive Democrats. The constant across his donation recipients is that they all have influence over the cruise ship industry. Similarly, because team owners tend to also have stakes in real estate, they often donate to local officials as well. While owners’ contribution patterns seem random by party, in actuality, they are strategic based on their personal gains. How do these patterns impact the dynamic between athletes and front-office management, if at all? What do these patterns say about the larger (racial) capitalist habit of pushing surface-level, performative justice?
This project aims to explore patterns of political donations by team owners in the NBA and NFL and contribute a network perspective of an already investigated dataset. In 2020, FiveThirtyEight and ESPN worked together to publish a series of articles to investigate the political spending habits of professional sports owners in the US. Using a network perspective, this project will explore the question of homophily by political party in the donation habits of team owners in the NBA and NFL.
Methods
The NBA and NFL have very different connotations attached to their leagues. The NBA is often seen as an urban sport with significant ties to contemporary Black culture. The NFL, on the other hand, is seen more as an “all-American” league, representative of White, middle-America. These connotations, along with the specific sociopolitical climate of the 2016 and 2020 elections, provide the backdrop of this project in comparing the NBA and the NFL in team owner spending patterns. Because of these differing connotations and the actions by both leagues during the summer of 2020, it is expected that the NBA would be more Democrat leaning and the NFL be more Republican leaning. For example, in 2020, the NBA board of governors and the NBPA (National Basketball Players Association) agreed upon a deal to contribute $300 million to the Black community over the next decade, T-shirts and other merchandise with BLM messaging were worn by players, public statements supporting BLM and raising awareness were posted, and players of entire teams even agreed to boycott some of the first games in the NBA bubble (amongst many other actions – most of them being Black athlete-led) (Lee, 2020). The NFL on the other hand, was urged by its players to put out an official statement condemning racism and supporting Black Lives Matter. Shortly after, the NFL announced plans to paint phrases like “It Takes All of Us” and “End Racism” in field end zones, and play “Lift Every Voice and Sing” (i.e., the Black national anthem) before games (Montez de Oca, 2021). Additionally, the Washington Football Team got rid of its racist name and teams were encouraged to sign free-agent Colin Kaepernick – although a team has yet to sign a deal with him.
However, that is not the case. According to FiveThirtyEight and ESPN (2020), both leagues are heavily Republican leaning, with most contributions going to Republican campaigns and PACs. It should be noted that this is not a comprehensive dataset, as only clear cases of contributions from owners to campaigns were included and many wealthy elites have ways of hiding their political spending. Overall, owners in the NBA are reported to spend much more than owners in the NFL, each league spending $11 million and $5 million respectively (ESPN & FiveThirtyEight, 2020).
The data utilized for this project comes from the popular politics, sports, and science website, FiveThirtyEight.com. The data describes political contributions made by owners of American professional sports team owners in all major leagues (NBA, NFL, MLB, NHL, WNBA and NASCAR) from 2016, 2018, and 2020. It includes data such as team name and league, election year, political action committee (PAC) recipient, the monetary value of the donation, and the party of the PAC recipient. The original data was not configured as network data, so it was transformed into a two-mode network object using iGraph in R. For the purposes of this project, the data was filtered to only show NBA and NFL team owners and their donations in 2016 and 2020. In total, four networks were examined: the 2020 NBA network, the 2016 NBA network, the 2020 NFL network, and the 2016 NFL network.
These four networks are directed (i.e., ties are sent from one node to another and not necessarily reciprocated), two-mode networks; these kinds of networks, also called affiliation or bipartite networks, are those with two different kinds of nodes (e.g., actors and events). Only nodes of the same kind can be connected. For example, team owners cannot be connected to team owners, and PACs cannot be connected to PACs. In this network, nodes represent team owners and PACs. Edges indicate a donation made from a team owner to a PAC. Figure 1 illustrates the 2016 NBA two-mode network as a sociogram. The orange nodes on top represent team owners, while the purple nodes on the bottom represent PACs. As represented in the sociogram, edges only go from team owners (orange nodes) to PACs (purple nodes).
figure 1: Two-Mode Projection Example
Two-mode networks are interesting for an overall view of the relationships being examined, but for analysis purposes, two-mode networks are not very helpful. Therefore, two-mode networks are often transformed into one-mode projections to run common network analysis techniques like centrality measures or clustering routines. However, there are important considerations that should be made when analyzing two-mode networks and their one-mode projections. An examination of relationships between nodes in a one-mode projection should be done with caution, as ties between two nodes in a one-mode projection only indicate an opportunity for interaction or connection. In other words, two team owners who donate to the same PAC may only have that in common – just because they contribute to the same PAC does not necessarily mean they have a meaningful relationship. Similarly, computational and statistical analysis of one-mode networks should be done with the same assumptions in mind. If a node in a one-mode projection representing team owners has high degree centrality, it does not necessarily mean they are directly tied to many other owners; it means they donated to many PACs, several of them being the same as other team owners.
For the analysis portion of this project, the two-mode networks were converted into one-mode projections. Each two-mode network was converted into two one-mode networks, one illustrating relationships between PACs and the other illustrating relations between team owners. For this project, the PAC projection networks were analyzed. Figure 2 is a sociogram of the 2016 NBA PACs one-mode projection. The nodes of this one-mode projection are PACs and edges indicate that at least one of the same team owners donated to both linked PACs. For example, in 2016, Jerry Reinsdorf, the owner of the NBA Chicago Bulls, donated to both the Friends of John McCain Inc. and the Tim Scott for Senate campaign. This means that in the one-mode projection of the 2016 NBA PACs network, the Friends of John McCain Inc. and the Tim Scott for Senate campaign are connected.
figure 2: One-Mode Projection Example
This project will be utilizing the methodological technique of modularity calculation to test for homophily by political party of PACs. Modularity is the measure of the how well a network is divided into subgroups; in other words, it is is the measure of whether ties are concentrated within subgroups or between subgroups. A modularity score can range from -1 to 1, with values closer to 1 indicating more ties within groups and less ties between groups. These scores should be analyzed comparatively, meaning there is not one value of modularity that is standard and scores should be compared across different networks to derive any meaning.
Modularity is typically used to decipher whether a community detection algorithm is accurate in dividing subgroups within a network. Community detection algorithms are used to identify subgroups within a network – modularity is the measure of whether there are more ties within those subgroups or between those subgroups, verifying the results of the community detection algorithm. However, homophily is another network aspect that can be tested by using modularity. Homophily is “the principle that a contact between similar people occurs at a higher rate than among dissimilar people” (McPherson et. al., 2001), i.e., similar people tend to be connected more than dissimilar people. Exploring homophily with modularity is done by calculating modularity by a certain node-level attribute, in this case, political party. The result indicates whether the clustering of a network relates to the node-level attribute in question. For this network, having a high modularity score would indicate that PACs representing the same political party would be clustered together, meaning the same set of team owners tends to donate to the same group of PACs.
Results
As stated above, the modularity of these networks was calculated to explore homophily by political party of the recipient PACs. Four networks were analyzed in total, each being one-mode projections describing relationships between PACs. Overall, the modularity scores of all four networks are low, but a comparative lens in conjunction with an examination of each network’s sociogram can still reveal interesting stories about the ways in which professional sports team owners contribute to US politics. Each one-mode projection is illustrated by the sociograms below:
figure 3: 2016 NBA PACs network
figure 5: 2020 NBA PACs network
figure 4: 2016 NFL PACs network
figure 6: 2020 NFL PACs network
The modularity scores for each network are described in the table below:
There are several observations that can be made based on the sociograms and modularity scores of each network. The network with the highest modularity score is the 2020 NBA network with 0.0447, which show the one-mode projection of the PACs NBA owners donated to during the 2020 election. This means that the 2020 NBA network has the most evidence for homophily by political party compared to the other three networks. On the other hand, the 2016 NFL network has the lowest modularity score, with -0.0004. Additionally, when examining the network sociograms, owners in the NFL tend to donate more to bipartisan PACs than owners in the NBA. In the NBA, there are few donations from team owners to bipartisan PACs, with most donations going to partisan campaigns. Overall, there is more evidence for homophily in both NBA networks compared to the NFL networks. This means that, in the NFL, owners tend to donate to a more diverse set of PACs – in terms of their political affiliation – compared to owners in the NBA.
There are several preliminary conclusions that can be taken away from these observations. Firstly, NBA owners seem to be more consistent in the political party they donate to, whether that be Republican or Democrat, whereas NFL owners seem to donate to a wide range of campaigns across the political spectrum. This could be indicative of a tendency for NFL owners to donate in their own self-interest rather than by political interest. The example of Micky Arison, although an owner of an NBA team, is illustrative of this idea. As discussed above, Arison has historically donated to a wide variety of PACs because of his underlying motive of donating to those who are involved in the cruise ship industry. Similarly, when interpreting the modularity scores and sociograms of the NFL networks, it seems that other owners in the league may follow a similar pattern in that the PACs they donate to vary on the political spectrum. However, this is a suggestive conclusion without definitive evidence; it cannot be said for certain that the diversity in PACs is a result of NFL owners acting in their own self-interest. Secondly, in the 2020 NFL PACs network, there is more evidence for homophily by political party compared to the 2016 NFL PACs network, illustrated in both the modularity scores and sociograms. The modularity score of the 2016 NFL PACs network is -0.0004, compared to the 2020 NFL PACs network, with a score of 0.0094. In the NFL sociograms, there are visible differences in the amount of clustering within the network. The 2016 NFL PACs sociogram shows more nodes in the periphery that are more loosely connected, while the 2020 NFL PACs sociogram shows more distinct subgroups. This is potentially pointing to evidence that NFL owners are becoming more politically minded in their contributions and moving away from self-interest. Because they seem to be donating to campaigns of a similar party, it could be said that they are contributing more on political ideology than personal stakes. However, this could also mean that their personal stakes are now being supported by politicians of a similar party. The same could be said about the NBA networks, which also have a slight increase in modularity score from 2016 to 2020. The shift in the NFL networks is uniquely interesting because of the lack of homophily by political party in 2016, while the 2016 NBA network still has some evidence for homophily by political party. Again, this is a suggestive conclusion; as stated above, when analyzing one-mode projections, it is difficult to infer conclusive takeaways about the relationships between actors.
One potential reasoning for this shift in the NFL PACs network could be a sort of counteraction by team owners because of the NFLs shift to a more open dialogue about racism and bias in their league. Following Colin Kaepernick’s kneeling protests in 2016, the quarterback has essentially been blacklisted in the league. Kaepernick became a free agent after his departure from the San Francisco 49ers in 2017, with no other teams signing him since. This seemed to be the acceptable norm in the NFL – players who visibly protest on the field will get penalized. However, following the events of 2020, it was clear the NFL had to make drastic changes to its external image to save face during the resurgence of BLM. It could be that team owners – ultra-elites who are already quite republican leaning – personally disagreed with the NFLs decision to enter the racial justice dialogue and became more politically leaning in their donations. Other possible explanations include heightened political-mindedness due to the climate surrounding elections in 2020. Again, additional qualitative research must be done to support these suggestive conclusions.
Discussion and Limitations
This project explored the function of political party of recipient PACs on the homophily of one-mode projection networks, comparing across the NBA and NFL in both 2016 and 2020. Questions of political activism in professional sports and the disconnects between athlete, league, and owners were addressed by investigating whether NBA and NFL team owners are prone to donating to PACs based on party affiliation. Overall, owners of American sports team owners donate more to Republican campaigns – the only league that donates more to Democratic campaigns is the WNBA. It seems that NFL team owners tend to donate to a wider variety of PACs on the political spectrum, which is potentially due to their aims of protecting their personal stakes in certain industries. On the other hand, there was more evidence for homophily by party of PACs in the NBA networks. There is an overall shift in greater homophily by political parties in both leagues from 2016 to 2020. There are several factors that could have contributed to this shift, including the tumultuous nature of the 2020 election and the backdrop of the racial injustices that were being publicly debated during this time.
A limitation of this project was the lack of qualitative data to inform a more conclusive takeaway on the reasons why homophily by political party increased from 2016 to 2020. While there are certainly several factors that could have contributed to this change, additional investigations into motives for political donations are necessary to paint a fuller picture of the mechanisms at play. Other network analysis methodologies could be utilized to answer these questions as well. For example, using exponential random graph models (ERGMs) to test how political party of PACs shapes the odds of certain relational ties existing. Using a homophily parameter in this context could provide more insight into the relevance of political party in these networks.
Additionally, an investigation into homophily by geography could also reveal interesting patterns in donation patterns by team owners. While the PAC networks explored in this project did not have strong evidence for homophily by political party, there could be evidence for homophily by geography. Since it is known that many team owners donate based on personal interest (i.e., industries they are involved in), there may be stronger homophily by geography.
References:
Arnovitz, K. (2020, October 28). Election 2020: What motivates billionaire sports owners to donate to political campaigns? ESPN.com. https://www.espn.com/nba/story/_/id/30198394/election-2020-motivates-billionaire- sports-owners-donate-political-campaigns
ESPN & FiveThirtyEight. (2020, October 28) Inside The Political Donation History of Wealthy Sports Owners. FiveThirtyEight.com. https://fivethirtyeight.com/features/inside-the-political-donation-history-of-wealthy-sports-owners/
Lee, J. (2020, October 30). Election 2020: How does political perception differ from reality in American sports? ESPN.com. https://www.espn.com/mlb/story/_/id/30213498/election-2020-how-does-political-perception-differ-reality-american-sports
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a Feather: Homophily in Social Networks. Annual Review of Sociology, 27, 415-444.
Montez de Oca, J. (2021). Marketing Politics and Resistance: Mobilizing Black Pain in National Football League Publicity. Sociology of Sport Journal, 38, 101-110. https://doi.org/10.1123/ssj.2021-0005.
Shelburne R., & Andrews, M. (2020, October 30). What the work behind political change looks like for LeBron James and star athletes. ESPN.com. https://www.espn.com/nba/story/_/id/30213500/what-work-political-change-looks-lebron-james-star-athletes
Towler, C. C., Crawford, N. N., & Bennett, R. A. (2020). Shut Up and Play: Black Athletes, Protest Politics, and Black Political Action. Perspectives on Politics, 18(1), 111-127. https://doi.org/10.1017/S1537592719002597
Ulas, E. (2021). Examination of National Basketball Association (NBA) team values based on dynamic linear mixed models. PLoS ONE, 16(6). https://doi.org/10.1371/journal.pone.0253179
This project was completed as for the final assignment in Social Network Analysis (INSH 5304) taught by Professor Cassie McMillan at Northeastern University in Fall 2022.