Revisiting the Radius of Trust Problem: Who Counts as “Most People”?

General trust in social surveys is usually measured with a Single-Item General Trust (SIGT) question that asks whether “most people” can be trusted. The concept of the radius of trust challenges this approach by highlighting that trust may vary with the social distance between those who express trust and those who are trusted. This has led to the addition of six specific trust questions to the World Values Survey (WVS) questionnaire since Wave 5, which ask respondents to rate their trust in their family, neighborhood, and people they know personally, as well as in people of another religion, people of another nationality, and people they meet for the first time. Analyzing respondents’ answers to the SIGT and six specific trust items in the Integrated Values Surveys (IVS) dataset (1981-2022) across 88 countries and territories and the three most recent survey waves, we find that trust is a complex construct whose invariant property across cultures is a fixed-length vector composed of the squares of its two components: ingroup and outgroup trust. The phrase “most people” in the SIGT question appears to refer to “familiar strangers”—people whom respondents do not know personally but are likely to encounter in everyday life. It is trust in these familiar strangers, rather than in more distant outgroups, which is most strongly associated with the Human Development Index. This association, however, becomes pronounced only at very high levels of human development.