Goodbye Neighbor
How shifts in close knit communities are damaging our nation and what data science can do to help.
How shifts in close knit communities are damaging our nation and what data science can do to help.
In his book The Vanishing Neighbor: The Transformation of the American Community, author Mark Dunkelman lamented the collapse of American communities and the unraveling of structures that had for centuries maintained the structure and cohesiveness in this country [1]. These tightly knit communities, connected by bonds of proximity but also affection, at one time formed the core of American exceptionalism. The erosion of these bonds has led to new kinds of communities, ones connected by shared beliefs. The impact of this seismic shift in community has been largely negative. Whereas incidental connections between doctors and lawyers, grocers and plumbers once fostered cross-socioeconomic debate, closely knit communities now encourage groupthink, cliquishness, and factionalism.
This disintegration of bonds of proximity in favor of bonds of likemindedness has contributed to the increasingly profound divide between Republican and Democrat, religious and secular, and economically challenged and well-to-do. It has changed the way that cities are planned, science is communicated and understood, how academic papers are peer reviewed, whether people enjoy success in business, and it has contributed to the collapse of balanced mainstream journalism in favor of unfiltered bias. The polarization of our nation is clear from the data with the gap between political parties growing in lock step with the expansion of biased media [2].
Human beings are social animals and we enjoy close bonds with others as a matter of course. These close knit connections can come from many directions in our lives. Traditional ones are through family, school, social clubs, churches, as well as shared hardship including perhaps the closest of bonds: between those who have fought in war together. Positive close-knit bonds are those that are cross cultural, cross-social economic, and cross-ideological. These bonds form based on proximity and interdependence as occurred in early American communities naturally. While clannishness has always been with us, anyone knows that not all members of a family think alike. Now, unfortunately, it is possible for people to abandon biological and proximal families for a new kind of family, one based on shared ideology.
While ideological groups are nothing new, technology has effectively removed any barriers to participating in them including any stigma associated with being part of a group advocating violence against women, minorities, those of other nationalities, or even the overthrow of the government. It has become easy to find brothers and sisters in denial of mainstream scientific views such as climate change or evolution, advocacy of extremist views and policies, and criminal behavior.
More mundane close-knit groups can form in other settings including at work, in religious communities, and in schools. These groups of likeminded individuals tend to reinforce one anothers’ sense of inclusion in the group by excluding others who are different and by reinforcing each others’ preconceived notions. While seeming benign, these groups are even more damaging than ideological ones. They stifle creativity by rejecting multiple perspectives to solve problems, exclude those who are different into “out-groups”, and create mistrust and suboptimal work and learning environments. They can even be hotbeds for bullying, harassment, and bystander inaction that allow both to continue.
Data science is key to solving the vanishing neighbor problem. While the past will never return, the future need not be one of xenophobia and ideological fracture. The rifts we see today between red and blue states and haves and have-nots are simply the result of a break down in communication and inter-group cross-pollination. As with plants, self-pollination of ideas weakens us. It damages our ability as a nation to be a leader on a global stage. It creates gridlock in our nation’s government and encourages wasteful and ineffective compromises on policy. Make no mistake, the current state of affairs will not go away on its own. To achieve a renewed conversation between left and right, rich and poor, minority and majority, religious and secular, police and policed, we need to analyze the data on where these groups exist in each kind of organization, what they believe, what their concerns are, and then develop ways of breaking down the barriers between them.
Social network analysis shows that close-knit communities behave differently than loose-knit but highly connected ones such as workplaces. While the workplace can contain close knit communities, project teams, for example, tend to be loose knit since they are artificially constructed. In-groups and out-groups naturally form over time in large organizations [3]. These groups can cause problems when people from different groups form project teams and create anti-social barriers to collaboration. In-group formation tends to cause excluding behavior as well as avoiding seeking multiple perspectives on problems. Big Data can help here in identifying such groups and determining barriers to inclusiveness. It can also suggest ways for in-groups and out-groups to have cross interaction which may create greater inclusiveness and foster more creativity. Looking at the communications between individuals in an organization, in-group membership can be determined. While privacy is an issue, social network analysis need not expose identifying information (although most organizations monitor internal communication anyway). The important point is to determine where the connections are tight and where they are loose and to create additional connections and chances of interaction between clusters of individuals in the workplace.
Close knit group behavior is a problem inside the scientific establishment as well. Social networks of scientific communities show polarization when two closely knit structure are loosely connected to one another [4]. Typically one group is more successful than the other and makes better decisions while the other group maintains its position despite worse outcomes and worse decisions because of its mistrust of the other group. What is fascinating about these polarizations, is that once these two groups form, like Republicans and Democrats or Conservative and Liberals, the polarization is stable. The graph maintains the two groups. As an example, in the 1990s, physicians who treated Lyme disease fell into two camps: those who believed that Lyme disease was a short term condition that required just a course of antibiotics, and those who believed it was a chronic condition that required long-term antibiotics. This disagreement led to what is now called the Lyme Wars in which mistrust existed on both sides and accusations between camps were hurled like flaming arrows. This polarization, unfortunately, has the effect of making both sides worse at making good decisions because they have to maintain their distance from the other side.
In the anti-science camp, groupthink behavior encourages members to tighten their membership when presented with evidence to the contrary of their beliefs. Thus, science causes these groups to become even more closed off. When presented with charts and graphs that contradict their opinions, people cling to their convictions even more [5]. They seek out false knowledge to the contrary, and it is easy to find. Facebook, Twitter, and news sources that are off the beaten path offer a cornucopia of alternative facts that soothe cognitive dissonance, ensuring people never have to change their ideas. Whether it is anti-vaccine arguments, climate change denial, conspiracy theories, or false information about coronavirus, the information exists because people are looking for it. They want it to be true.
Multinational corporations have used these groups as weapons at times to attack science that was unfriendly to their profits. For example, the Union of Concerned Scientists published an expose on how ExxonMobil funded a disinformation campaign to attack climate science using a playbook straight from Big Tobacco [6]. Part of this campaign was “funding an array of front organizations to create the appearance of a broad platform for a tight-knit group of vocal climate change contrarians.” Essentially, the mega corp funneled $16M to a number of organizations which all promoted the misinformation of a small number of science apostates, creating the impression that these contrarians had broad support in the community [7].
Oxytocin has been shown to help open up close-knit groups to inclusiveness. In a 2017 study, oxytocin combined with positive social cues was shown to reduce xenophobia in a double-blind controlled study [8]. The influx of refugees and immigration in general as well as mixed signals from leaders have encouraged xenophobia. The exacerbation of in-group and out-group behavior on the bases of nationality or culture can encourage violence as well as more subtle forms of discrimination. Therefore, it is important to mitigate. While the researchers relied on intranasal injection, hugging and other forms of social touch can also encourage the production of oxytocin naturally. Nevertheless, it is hard to see people all snorting oxytocin or randomly hugging one another.
Many ideological groups are not close-knit, and this is a critical piece of information that social network and textual analysis can provide. Multiple studies of anti-vaccine parents have shown that public information campaigns are ineffective. When presented with scientific evidence on the effectiveness and safety of vaccines, parents became more convinced that vaccines were harmful [9][10]. In a study published last year on mapping the anti-vaccination movement on Facebook, researchers found that anti-vaccine networks were mostly “small world” meaning clusters interconnected by longer links. Most of the members were women. Text analysis showed that the discourse largely centered around outrage over structural oppression by the government, typical of conspiracy theorists. One of the major discoveries, however, is that anti-vaccine communities are not close-knit [11]. They are not highly connected groups providing mutual support. Rather, they are loosely bound to one another through ideology and transient commentary. In other words, the majority of anti-vaxxers tend to like or share pages but do not necessarily receive much support. This suggests that an effective campaign may not be a global or national advertising blitz, but one that targets close-knit groups that women with children tend to be members of or connected to (or drawing them into such groups) by exploiting external social links to the larger community. Since many anti-vaccine parents are homeschoolers but homeschoolers are rarely all anti-vaccine, targeting homeschooling groups might be a good place to start.
While conspiracy theory groups tend to exist outside the mainstream, mainstream political opinion polarization is a much bigger problem, one that is currently tearing the nation to pieces. In 2012, on the eve of the Presidential election, the New York Times reported in “The Vanishing Battleground” that the country was dominated by solidly blue states on the coasts and solid red on the interior [12]. A common misconception among academics and the public, however, is that Twitter, blogs, Facebook, and other social media tend to be divided into left and right with little civil debate between them. The theory goes that social media has enabled users to filter out, algorithmically, any news source that is not likeminded enough. A number of studies in recent years have attempted to model social media “echo chambers” without analyzing whether that hypothesis is true. A highly cited study in 2014 and a follow up in 2015 showed, however, that increased social media use led to an increase in the likelihood of people being exposed to differing opinions [13][14]. This is good news because it suggests that, for all its faults, social media is beneficial to restoring some of the lost cross-ideological connection of yesteryear. While chance encounters are less common in real life, online we tend to encounter our friends, family, and coworkers often. Moreover, people are more likely to share their political opinions online than in person. Social media companies can do more, however, such as using something like this MIT recommender system that purposely presented to users opinions that may differ from their own [15]. Thus, data science can help to adjust algorithms to reduce conflict between polarized opinion spheres.
It seems that far from being lost, those bygone communities may be in the process of moving online. Meanwhile, moral panic over the loss of those communities may be overblown. As the authors of “Lost and saved… again: The moral panic about the loss of community takes hold of social media” argue “before we hated smartphones we hated cities” as 17th century philosopher Thomas Hobbes mourned in his class Leviathan where rapid urbanization and cultural changes in England were leading to a “war of all against all” [17]. Two thousand years earlier, Socrates and Plato complained that the invention of writing was corrupting people [18]. America today is undergoing similar upheaval.
That does not imply that none of the problems facing America today are real or that they do not stem from shifts in culture from urbanization to the growth of 24 hour media outlets to the increasing segregation of society afforded by gentrification. Nor does it mean that nothing needs to be done to stem the tide of vitriol plaguing our nation. This is where social network analysis and data science show both what is going on and a way forward.
Far from being culpable for the demise of American culture, social media may be its savior. While talk radio, cable news, and online news outlets from Breitbart to Salon are effective at maintaining their ideological bubbles, social media is far more porous. Most people maintain social relationships with people they know from school, church and other religious organizations, work, family, all those areas that once formed the backbone of Main Street America. They may not all agree with one another, and many may seek out more ideological communities, but, given their one-dimensionality and tendency for such communities to fracture at small differences, they are far from reliable as sources of emotional support. When it comes to mutual support, people turn to the people in their online neighborhoods of friends and family, exactly as they have always done. As more people move online and connect with others and as social media finds ways to encourage cross-ideological dialogue, America may see a return to a more peaceful and productive time.
[1] Dunkelman, M. J. (2014). The Vanishing Neighbor: The Transformation of American Community. United States: W. W. Norton.
[2] Doherty, Carroll. “7 things to know about polarization in America.” Pew Research Center (2014).
[3] Big Data at Work: The Data Science Revolution and Organizational Psychology. (2015). United Kingdom: Taylor & Francis.
[4] Groupthink in Science: Greed, Pathological Altruism, Ideology, Competition, and Culture. (2020). Germany: Springer International Publishing.
[5] Beck, J. (2017). This article won’t change your mind. The Atlantic, 13.
[6] Union of Concerned Scientists. (2007). Smoke, Mirrors & Hot Air: How ExxonMobil Uses Big Tobacco’s Tactics to Manufacture Uncertainty on Climate Science. Union of Concerned Scientists.
[7] Shaping the Message, Distorting the Science: Media Strategies to Influence Science Policy : Hearing Before the Subcommittee on Investigations and Oversight, Committee on Science and Technology, House of Representatives, One Hundred Tenth Congress, First Session, March 28, 2007. (2007). United States: U.S. Government Printing Office.
[8] Nina Marsh, Dirk Scheele, Justin S. Feinstein, Holger Gerhardt, Sabrina Strang, Wolfgang Maier, René Hurlemann. Oxytocin and social norms reduce xenophobia. Proceedings of the National Academy of Sciences Aug 2017, 201705853; DOI: 10.1073/pnas.1705853114
[9] Nyhan, Brendan, et al. “Effective messages in vaccine promotion: a randomized trial.” Pediatrics 133.4 (2014): e835-e842.
[10] Bedford, Helen. “Pro-vaccine messages may be counterproductive among vaccine-hesitant parents.” BMJ Evidence-Based Medicine 19.6 (2014): 219–219.
[11] Smith, Naomi, and Tim Graham. “Mapping the anti-vaccination movement on Facebook.” Information, Communication & Society 22.9 (2019): 1310–1327.
[12] Liptak, A.(2012, November). The vanishing battleground. The New York Times, SR1.
[13] Lee, Jae Kook, et al. “Social media, network heterogeneity, and opinion polarization.” Journal of communication 64.4 (2014): 702–722.
[14] Choi, Jihyang, and Jae Kook Lee. “Investigating the effects of news sharing and political interest on social media network heterogeneity.” Computers in Human Behavior 44 (2015): 258–266.
[15] Musco, Cameron, Christopher Musco, and Charalampos E. Tsourakakis. “Minimizing polarization and disagreement in social networks.” Proceedings of the 2018 World Wide Web Conference. 2018.
[16] Hampton, Keith N., and Barry Wellman. “Lost and saved… again: The moral panic about the loss of community takes hold of social media.” Contemporary Sociology 47.6 (2018): 643–651.
[17] Hobbes, Thomas. Leviathan. United Kingdom, n.p, 1676.
[18] Plato’s Phaedrus. (2009). (n.p.): Agora Publications, Incorporated.