PSA2016: The 25th Biennial Meeting of the Philosophy of Science Association

Full Program »

Some Philosophical Challenges Concerning the Inclusion of Social Norms in Epidemiological Models

Social norms play an important role in infectious disease: the norm to use bare hands during burial preparation exacerbated the Ebola epidemic; Marin county’s highly affluent population has norms that encourage healthy living, but they have arguably been overextended into an anti-vaccination movement. While epidemiologists recognize the importance of social norms, a lack of clear definitions of what norms are and how they can be measured has prevented their inclusion in epidemiological models. This work will make progress towards the development of coupled models of social norms and infectious disease and analyzes some challenges in validating them.

Unlike some areas of social science where social norms are treated as exoge- nous variables, many philosophers model social norms as emerging endogenously from structured interactions of individuals (Lewis, 1969; Skyrms, 2010; Alexander, 2007). Even though these models are highly abstract, couching them in certain game-theoretic frameworks makes them amenable to validation. For example, experiments can be done using the prisoner’s dilemma, the ultima- tum game, the stag hunt game, or other coordination games. In turn, such experiments provide data for estimating model parameters.

There are numerous ways to incorporate these types of social norm games in epidemiological models. Traditional epidemiological models, such as the SIR model, include various disease states that individuals could be in, such as suscep- tible, infectious, and recovered, and rates or probabilities (depending on whether the level of abstraction is at the population or individual level) for transitioning between these states. The transitions from susceptible to infectious and infectious to recovered are interpreted biologically. More contemporary epidemiological models include states that represent prophylactic behavior, such as vaccination or hand washing. The transitions between non-prophylactic and prophylactic states are interpreted cognitively – they involve a decision making procedure. It is in these decision making procedures that social norm games can come into contact with epidemiological models. But there are many possible points of contact. For example, one approach is to incorporate the utilities of different disease states into the payoffs of the social norm games, letting the decision making procedure for engaging in prophylactic behavior then be determined by the outcome of the social norm game. Another approach is to have two payoff matrices, one for the social norm game and one for disease states, and then weight these in the decision making procedure.

The problem is not to figure out which one of the numerous ways of coupling the two types of models is the right way. Rather, the problem is figuring out what assumptions are being made for the different ways that they can be coupled, and consequently, how they limit or facilitate possible avenues of validating the models. This work outlines several approaches and the challenges they face for validation.

Alexander, J. M. (2007). The structural evolution of morality. Cambridge Uni- versity Press.
Lewis, D. (1969). Convention. Cambridge University Press.
Skyrms, B. (2010). Signals: Evolution, learning, and information. Oxford University Press.

Author Information:

Bert Baumgaertner    
University of Idaho


Powered by OpenConf®
Copyright©2002-2015 Zakon Group LLC