Opting out of cooperation is individually feasible but not in the aggregate

Evidence

An interesting case arises where there is tension between the intensity of our interest in acting and the value of p. We may either not know whether to trust, or even know that we distrust somebody, but if we were to refrain from engaging (or attempting to engage) in cooperation, our losses could nevertheless be unacceptably high: opting out is a feasible option, but we would pay for it dearly. The option of avoiding or exiting from a relationship, for instance, may be present with respect to any one individual agent, but not when considered in aggregate terms: we may have a choice as to which restaurant we assign a high enough probability of not giving us food poisoning, but there are circumstances in which we can hardly afford to distrust all restaurants without perhaps suffering unpleasant consequences. — Diego Gambetta, Trust - Making and Breaking Cooperative…, 2763

Categorization

• XREF: Connects to the reader’s trust/observability synthesis with Luhmann and Hardin — this is exactly the ‘trust requires observable constraints’ axis: when exit is denied in aggregate, trust thresholds collapse and hope substitutes for verification. — connects to: Cooperative civic activities decline faster than individual ones — Robert D. Putnam, Bowling Alone The Collapse and Revival…, 127; Membership dilemmas persist even in border-sealing preference-maximizing states. — Douglas A. Kysar, Regulating from Nowhere: Environmental…, 461
• DEFINE: Sets the threshold-of-trust concept: what p must reach for cooperative action, and why it varies with stakes and alternatives.