Few people are aware of the fact that a group of mutually-unfamiliar chimpanzees would tear each other to bits simply for lack of trust. Each wouldn't know enough about the others to feel safe not lashing out first. And yet, as strange as humans are to one another, we do a pretty good job of ignoring each other on the bus or in the market. This is common among humans—we often (too often) take it for granted. And yet humans manage to pull it off regularly enough that it has become our thing. I am talking, of course, about social trust—where mutual strangers are able to trust that, for the most part and with plenty of exceptions, people can be trusted not to slit your throat and rob your corpse. Or steal your identity or kidnap your children.

Social trust is essential to the quality of our lives, how society functions, and nearly everything else. This is where ignorance enters the picture. Ignorance is commonly treated as an epistemic failure. There are a lot of philosophers who hold that to be ignorant of something is to be intellectually or epistemically deficient.

This picture is incomplete, though, simply because knowledge is costly to acquire, maintain, and deploy, costing time, attention, cognitive effort, emotional bandwidth, and sometimes even social capital. Because those resources are scarce, agents face trade-offs about what to believe and how to pursue evidence for those beliefs. Once this point about scarcity and trade-offs is taken seriously, the normative landscape changes and the salient question here is whether there are forms of ignorance that are rational, and even socially and morally preferable.

One place where the value of ignorance becomes especially clear is in the role social trust plays in sustaining large-scale cooperation. After all, there is a real sense in which social trust dramatically reduces the amount of knowledge individuals must acquire about each other before they can cooperate. Where social trust is high, people can transact, cooperate, and rely on strangers without too much detailed information about their motives or competence. But where social trust is low, that same ignorance becomes dangerous, and agents react by investing heavily in verification and enforcement.

From an economic perspective, social trust lowers transaction costs, that is, the cost of trust, transfer, transacting business with each other. From an epistemic perspective, social trust substitutes for information in the sense that social trust allows agents to act under conditions of uncertainty without attempting to eliminate that uncertainty through costly knowledge acquisition. Here social trust enables what might be called productive ignorance, where the deliberate forgoing of information that would be expensive to obtain and of limited marginal value is avoided.

Consider the alternative where society with little social trust and where ordinary interactions become epistemically demanding with buyers vetting sellers, employers scrutinizing employees (more than they otherwise would), etc. In the absence of social trust, each transaction invites background checks, redundant documentation, contractual safeguards, surveillance mechanisms, and legal recourse. And these forms of verification are costly, placing an increasing epistemic burden on individuals.

By contrast, in a high-trust society, many of these epistemic demands are relaxed. This is because agents can reasonably assume norm compliance and reciprocity in most cases. They need not investigate every counterparty because they rely on stable social expectations backed by reputational incentives and institutional enforcement to do the verifying for them, in a sense, allowing them the freedom to forgo knowledge.

Seen in these terms, social trust enhances efficiency in a way that lightens the epistemic load on individuals and institutions allowing them to economize on knowledge by relying on socially-shared expectations rather than exhaustive scrutiny. And this point brings us to a deeper normative point, namely, that if ignorance is the inevitable byproduct of high social trust, and if there are strong reasons, independent of the alleged value or disvalue of ignorance, then attempts to eliminate that kind of ignorance are likely to be misguided and do more harm than good. For example, requiring verification where there is no need implicitly sends a signal that people are operating in a low-trust environment.