Agent Based Model of Anti-Vaccination Movements: Simulations and Comparison with Empirical Data

"[T]he AV [anti-vaccination] movement is a perfect example of the role of the new communication landscape (technological and social) in determining the dynamics of opinion in modern societies, linking the virtual and real worlds."

Phenomena related to the anti-vaccination (AV) movement are extremely complex and present several challenges, including: the competition between rational and emotional decision-making, the role of highly visible individuals in shaping societal trends, the role of an active versus passive approach to information, and the interaction between different factors driving polarisation (e.g., between splits in political opinion and a pro-vaccination/anti-vaccination stance), to name but a few. This paper combines large-scale analysis of data covering AV activity on the internet with an agent-based model (ABM) that describes the behaviour of various relevant actors through use of big data observations. The purpose was to overcome some of the weaknesses of previous data analyses of the AV movement - specifically in the domains of understanding causation, uncovering hidden dynamics, and guiding further data collection.

The researchers divide AV activists into two groups, depending on their origin:

1. A relatively small group of anti-vaccination "professionals", which includes the few members of the medical profession or scientists who oppose vaccination, as well as certain celebrities. They see their activities as a global "mission" and, if challenged (for example, in legal processes, leading to expulsion from the medical profession), adopt a "martyrdom" stance that actually increases their influence rather than decreases it.
2. People who became afraid of vaccination and not only seek confirmation of their fears but also spread the "AV gospel". Their participation in internet discussion groups or on social media - especially through their personal contacts - creates a critical mass. Their personal stories serve as "proof" of the validity of the AV accusations and the universality of their fears. The "common people" strengthen the individual appeal of the AV position in ways official/government communication cannot.

As the researchers explain, ABMs are a way of investigating social phenomena often used in the context of studying "what-if" scenarios and uncovering causal mechanisms linking various aspects of a social system. The ABM framework typically consists of a combination of: agents (representing one or more types of social actors, which may be individuals, organisations, or even impersonal entities), the connections between the agents (and between agents and the external world), and agents' activities and interactions. By providing a simulation of the studied social system, an ABM model allows the theoretical hypotheses on which it is built to be tested. ABMs can also uncover unexpected behaviour resulting from interactions. "Unfortunately, most of the current generation of ABMs used to describe the dynamics of opinion changes is too simple to catch the complexity of the opposition to vaccines."

The datasets that were analysed in this study cover multiyear periods preceding the COVID-19 pandemic, documenting the behaviour of vaccine-related internet activity with high temporal resolution. The data covered the United States, or US (Reddit comments - treated as US representative, although participants may come from any geography) and Poland (contributions to a discussion forum on Interia, a news integrator and community portal.) The fact that the content in both sources was freely created by the general public (as opposed to traditional media) means that vaccination-related comments could serve as a proxy of the public interest that could be compared with activity modeled in the ABM.

To understand the empirical observations - in particular, the mechanism driving the peaks of AV activity, the researchers propose an ABM of the AV movement that includes the interplay between multiple driving factors: contacts with medical practitioners and public vaccination campaigns, interpersonal communication, and the influence of the infosphere (social networks, Web pages, user comments, etc.). There are four types of agent entities in the model:

• Doctors - representing official communications and actual doctors, in personal contact with agents. The number of doctors is relatively small (100).
• The studied population ("patients") - with the presented simulations using N=2,000.
• Initiators - representing the small number of special agents whose opinions are unchanged, and whose "task" is to create the initial stream of persuasive messages.
• The "messages" in the infosphere that are created by the initiators and by patients with sufficiently extreme opinions (such patients become "activists").

All four types of agents are characterised by an opinion on vaccination, ranging from +2 (strong support for vaccination) to –2 (strong opposition to vaccination).

The datasets studied show the presence of short-lived, high-intensity activity peaks - in particular, the bursts of active posting and re-posting - that are seemingly random in size and time separation. Such behaviour suggests a nonlinear nature for the social interactions driving the AV movement, instead of the slow, gradual growth typical of linear processes. The researchers are concerned about this "apparently stochastic, spiky nature" of infosphere-based AV communications because, "[i]n contrast to slow changes in opinions, which allow governments and other stakeholders to prepare, apply, and adjust strategic responses, a sharp peak in vaccination reluctance, driven by fear and other strong emotions leave no time for such preparations."

The paper demonstrates that the ABM simulations closely reproduce the observed temporal behaviour of the AV interest. For a range of parameters, the simulations result in a relatively small fraction of people refusing vaccination; however, a slight change in certain parameters (e.g., willingness to post AV information) may lead to a catastrophic breakdown of vaccination support in the model society, due to nonlinear feedback effects.

An increase in intensity of standard pro-vaccination communications by government agencies and medical personnel is found to have little effect. On the other hand, focused campaigns using the internet and social media and copying the highly emotional and narrative-focused format used by AV activists is found in the simulation to diminish AV influence. Similar effects result from censoring and taking down AV communications by social media platforms. The benefit of such tactics might, however, be offset by their social cost - for example, the increased polarisation and potential exploitation of the tactics for political goals, or increased "persecution" and "martyrdom" tropes on the part of AV activists.

Reflecting on the characteristics that are missing from the version of the model presented in this paper, the researchers note that it lacks: a social network structure (present in the AV movement and in most forms of social communication); strong homophily in interpersonal contacts (only partially modeled through rejection processes present in the model); and more realistic distributions (in that it uses the same fixed lifetimes for all agents and messages). To address these weaknesses, the researchers are gathering more detailed empirical information regarding the dynamics of AV communications. "Another direction of model development is related to modeling the situation created by the COVID-19 pandemic....The combination of fears related to health with the severe effects of social and economic lock-downs creates an emotional landscape in which distrust of the medical industry in general and anti-vaccination arguments in particular can easily take hold." This work is in progress as data accumulates, with the researchers hoping "to develop an extended model taking into account the lessons learned (so painfully) during the COVID-19 pandemic."