Every day millions of people deal with others they know nothing or very little about. A Norwegian tourist buys a carpet in Casablanca. A woman in Mexico City hails a cab on the street. A person with a never-before-experienced eye pain asks an ophthalmologist for advice. In each case, individuals must form a belief about the reliability of a counterparty to decide whether to deal with this person at all.
How do individuals form beliefs about how others will behave in the absence of prior interaction? And do these initial beliefs persist in the face of evidence?
Even though most of our (economists’) models of — and hence predictions about — everything from two-person interactions all the way up to entire national economies depend crucially on such beliefs, little is known about the formation and evolution of beliefs.
In the absence of a true understanding of belief formation and evolution, to make our models work we economists typically make more-or-less well-founded assumptions about this process. For example, at the heart of many economic models is the assumption of “rational expectations” — i.e. that individuals’ beliefs turn out to be correct in equilibrium — or that initial beliefs are commonly held (common prior) or even that, initially, everybody believes any imaginable outcome is equally likely (flat prior).
Questioning the traditional assumptions: The false consensus approach
Economists have recently begun questioning the descriptive validity of our traditional convenience assumptions and investigating myriad mechanisms which may systematically colour beliefs.
- One such mechanism, which has a long tradition in social psychology, is a tendency for individuals to extrapolate from how they themselves would behave to form beliefs about how strangers will behave — a phenomenon labelled “false consensus” (Ross et al. 1977).
Many economists even argue that false consensus is perfectly rational. As Nobel Prize laureate Thomas Schelling once wrote “you can sit in your armchair and try to predict how people behave by asking yourself how you would behave if you had your wits about you. You get free of charge a lot of vicarious empirical behaviour" (1966). A handful of studies by economists document false consensus in situations with financial incentives for correct predictions, although its persistence and generalisability across contexts remains controversial (Selten and Ockenfels 1998; Engelmann and Strobel 2012; Massey and Thaler 2012).
Even if false consensus is not fully general across all contexts, it may still be a particularly relevant concept in specific situations. One such situation concerns the ubiquitous and economically fundamental decision about whether to trust a stranger about which one has little information on past behaviour — as in the examples given above.
Generalised trust has been shown to be:
- highly persistent across and within generations, and
- highly heterogeneous across individuals even within the same population.
This dual pattern of persistence and heterogeneity can be parsimoniously explained by false consensus.
- Parents teach values,
- values determine trustworthiness, and
- trustworthiness colours trust through false consensus.
This pattern yields simple testable implications. Trust should be persistently related to own-trustworthiness; and own-trustworthiness should be related to parentally instilled values.
In Butler et al. (2012a), we test exactly these hypotheses and find substantial evidence for both. Individual trust beliefs are related to individual trustworthiness, which is related to the values parents transmit to their children. Interestingly, anchoring of beliefs to inherited values continues to persist even when current beliefs are contradicted by experience. Obviously, if someone forms trust beliefs about an unknown person by attributing to others his own trustworthiness, he is bound to make mistakes, unless it happens that his own trustworthiness is close to the average trustworthiness of the unknown p