William Faulkner wrote, “all human behavior is unpredictable, and considering man’s frailty…and the ramshackle universe he functions in, it’s all irrational.” Despite this claim, scientists from disciplines ranging from economics and mathematics to anthropology and psychology have laboriously attempted to uncover patterns within the morass. Game theory is one such approach that starts with the assumption that humans are rational agents and then develop mathematical models to predict outcomes between these agents. Typically, these models have a set of players, payoffs, a set of strategies, and a framework that maps the strategies to a set of payoffs for the players. Broadly speaking, it is considered the science of logical decision-making in humans, animals, and computers. Scenarios such as the ultimatum game, prisoner’s dilemma, and dictator game are used to model human behavior.
In the above scenarios with finite players and finite options, the mathematician, John Nash, algebraically proved that there is at least one equilibrium state for each of these scenarios. This became known as the Nash equilibrium. In such an equilibrium state(s), there is no stakeholder that is necessarily behaving irrationally, but all stakeholders are strategically responding to their local incentives to improve their respective positions. It describes how rationally self-improving individuals can lead to self-harming groups. Nevertheless, if rationality is defined as utility maximization of the self, then humans repeatedly fail to be “rational.” In real life experimental models with humans, the system is consistently different than the predicted Nash equilibrium. For example, in experiments replicating the prisoner’s dilemma, approximately half of the people choose to confess. The Nash equilibrium for the ultimatum game would predict many low offers and few rejections. But in fact, in Western societies most people offer half and many reject offers below fifty percent. In contrast, chimpanzees never reject offers in the ultimatum game and their equilibria approximate closer to a theoretical Nash equilibrium.
Humans diverge from theoretical Nash equilibria because these models do not take into account social norms. Competition, cooperation, altruism, and selfishness might be proximal explanations for human behaviors, but the prime movers are the social norms created by our respective cultures. The disposition to form and learn norms have coevolved with our psychology and shaped brain development over a million years of cultural evolution. We are dispositioned to live in a world governed by social rules dictated by our cultures and subcultures. We effectively acquire these local norms as children, internalize them as intrinsically and inherently real, and rapidly develop cognitive abilities and motivations for discerning potential norm violations. Over time, this process of internalization makes our responses more automatic, so we often carry out these behaviors unreflectively. As policymakers in healthcare implement policies to steer behavior, primarily focusing on price signals and the cultures of market economics and industrialization, without regard to the culture of medicine, it will yield unpredictable, suboptimal results at best and a failed, unsustainable system at worst. However, herein lies the immensity and intractability of the challenge. It is not always easy to know where culture begins and where it ends. It is often not bound in place, nor is it fixed in time, and they all have their own set of rules. At all scales ranging from the individual to the system, there are multiple partially nested but also non-overlapping, multidimensional, and non-intersecting cultures – each with its associated explicit and implicit rules. An obvious example is an inherent tension between the societal culture of consumerism, industrialism, capitalism, and autonomy with the subculture in medicine of benevolent, patriarchal caregivers making decisions in service and for the sick and the vulnerable. How can you devise incentives that are specifically particular or sufficiently generalizable?
In my last essay, I discussed how healthcare is failing patients and physicians in the United States. The system is in an inferior Nash equilibrium partly because our policymakers have not considered the expanded definition of “rationality” that includes the milieu of cultural norms that have coevolved and with our psychological dispositions and adapted to local conditions. Incentives have to be structured at the very least to take into consideration the various cultural norms that are created by the complex interplay of the people in medicine, the institutions that train them, and the larger society within which they both function. In medicine, policymakers must consider if incentive schemes of price signals and economic utility maximization borrowed from the cultures of industrialization and capitalism mesh with the ideals of benevolence, care, compassion, vulnerability, and frailty that are historically part of the culture of medicine? However, in reality, culture is not static (in place nor time), so what might result from this single-dimensional culture clash is a new synthesis, transitioning towards a new Nash equilibrium.
Source(s): The Secret of Our Success
[…] machine learning and social science models. In prior essays, I discussed adjacent possibles and the Nash equilibria to describe how the healthcare system is moving towards an inferior optimum. In this essay, I add […]
[…] “is like” with the homologous “is”, which often puts artificial parameters on the range of adjacent possibles. For example, the pervasiveness and all-purpose use of industrial and commercial applications have […]