Let us begin with the pressing policy issue of HIV risk taking behavior among homelessness teens to take home this point of social bond (de)construction. In the work of Rice et al.’s (2012) study of HIV risk behavior, homeless adolescents were located within the core (that dense ball of spaghetti in the middle of the network graph), were more likely to be female and were more likely to have been homeless for at least 2 years. The longer the teen, particularly for the young woman, is outside of the family unit, the teens form strong, compact ties with a new “family”. Surprisingly, being on the outside of this tight “family” that is found in the periphery of the network was protective against HIV risk taking. Highly connected, dense core are great for galvanizing information within that group. But a dense group may be more difficult to infiltrate. If the dense ball of teens are passing misinformation and reinforcing risky HIV behaviors, it is best to go your own way. But where can a young person go with such marginalized circumstances?
If the public policy being developed pertains directly to HIV risk taking reduction, perhaps targeting the core network to diminish risk could be a first step. But in the work of being connected to other people, the low risk teens may help each other or could transfer into the high risk group. But policymakers must remain mindful of what systemic changes can flow from targeting that portion of the network. People come and go into each other’s lives. Policy must be mindful that the longer the teen is outside of a traditional household, human connections will be made with the people that they have the most contact with. Could the teens in periphery have formed cliques that supported less risk taking? This may help these teens. Keep the periphery teens supported in their low-risk behavior. In the world of networks, there is something called homophily (birds of a feather). This means more than living in the same place. Above that shared space, the teens are ties together by something stronger: love, support, shared values, shared behaviors (see Feld & Carter, 1998; Kadushin, 2012). In other world, people live by forming bonds wherever they land.
If the policy lumps the new cliques (core and periphery) together, network membership can change over time. Teens that tie together two completely separate networks are called bridges. By theory, the networks would not have connected if not for this new bridge. Often the bridge has enough prestige and power to convince two divergent groups to join forces (Granovetter, 1973). Will the new members from the outside possess adequate social currency to offset the peer influence of the core members? Thus we have complexity. Can one policy that is meant to affect teens as if they share the same life chances and social embeddedness work? Most likely answer is no. While there may be an overarching goal set up the policy, parse out how different attributes of the teens may affect how the proposed policy works.
The longer a teen is away, it becomes more likely that their family will be in the same dire social straits and may not be protective in navigating good social choices and decisions. But the longer a teen is away, human nature requires connection and closeness, a family broadly defined. Being on the outside (periphery) of the homeless core protects against HIV risk-taking. Let us not forget a social purgatory between the instability of homelessness and the perceived caustic environment that the teen desperately calls to escape. The peripherals may be at risk in other ways that may lead to a greater risk of HIV risk taking once the teen is in the core. But if that teen has the ability to persuade those at risk, there is a possibility that the low risk taking of a strong teen could start to cascade low risk attitudes and values. But there is also a possibility that the teen will become enveloped and become high risk himself. It may be too much to ask of that teen to work to overhaul the collectively held value of higher risk sexual practices (Long et. al., 2013).
So in using the research on social networks, I propose systemic factors that should be accounted when attacking HIV among homeless teens:
1. Every homeless teen is not the same and each with present a different set of connections.
2. Being deeply connected in the homeless culture may place these teens at higher risk for unsafe sexual behaviors.
3. Targeting low risk teens on the periphery will require a different intervention to support the low risk behavior.
4. While there may be opportunities to use low-risk teens as “bridges” to the high-risk teens, this should only be done with extreme care and oversight. The bridge is more susceptible to falling into the activities of the core and may suffer from burn-out for the heightened sense that change is on that teen’s shoulders.
5. Watch the movement of teens from the core to periphery (and back again). This movement brings a whole new set of structural realities both for the teen as well as the network.
Social networks are powerful and are often underutilized in uncovering the underlying structure of health policies. But the policy work that we should hold dear must account for the power to combatting ecological gaps and failures, such as the personal and societal failing of just one homeless teen.
Feld, S. & Carter, W. (1998). “Foci of Activities as Changing Contexts for Friendship.” In
Placing Friendship in Context, eds. Rebecca G. Adams and Graham Allan. Cambridge,
UK: Cambridge University Press.
Granovetter, M. (1973). The strength of weak ties. American Journal of Psychology. 78 (9),
Kadushin, C. (2012). Understanding Social Networks. New York: Oxford.
Long, J., Cunningham, F. & Braithwaite, J. (2013). Bridges, brokers and boundary spanners in collaborative networks: a systematic review BMC Health Services Research 2013, 13:158
Rice, E., Barman-Adhikari, A., Milburn, N. & Monro, W. (2012) Position-Specific HIV Risk in a Large Network of Homeless Youths. American Journal Of Public Health. 102(1), 141-147.
NOTE: This white paper is a revision of a blog written by the author. An early version was originally posted on the Orgcomplexity Blog (Orgcomplexity.wordpress.com) on February 28, 2013.