Why researchers ask about trust
Trust is one of the most widely tracked indicators in studies of community life. The idea is that when people expect others to behave fairly and reliably, everyday cooperation becomes easier, from lending a tool to a neighbour to taking part in local decisions. Researchers often treat trust as a clue to the overall climate of a place, though it is better thought of as one indicator among several than as a complete measure of community health.
Social trust, in the sense used by most surveys, means a general willingness to rely on other people. It is distinct from trust in institutions such as governments or the media, and from trust in a specific person, such as a close friend. Keeping these apart is the first step in reading any figure.
The classic general trust question
The most common survey item asks, in various wordings, whether most people can be trusted or whether you cannot be too careful in dealing with people. Respondents choose between these two positions or place themselves on a scale. Variations of this question have been asked in many countries for a long time, which makes it useful for broad comparison.
It also has well-known weaknesses. “Most people” is vague, and different respondents may picture different groups: neighbours, strangers, people in their city or people in general. The question also offers a rather stark either-or choice, so a person who trusts others in some settings and not in others has to pick one answer. Researchers sometimes supplement the item with questions on specific groups to get a more nuanced picture.
Asking about neighbours
Community surveys often add more local questions, such as how much respondents trust people in their neighbourhood, whether they would expect a lost wallet to be returned, or whether neighbours help one another. These tend to be more concrete than the general item, and they often show different levels. People frequently report higher trust in those they know than in strangers, and responses can differ between people who have lived in an area for years and recent arrivals.
Because “neighbourhood” is understood differently by different people, a question using that word is not geographically precise. One respondent may think of the street, another of a district. Some surveys provide a definition, such as the area within a short walk, while others leave it open, which affects comparability.
Why wording shifts answers
Small changes to a trust question can move results noticeably, which is a reason to be cautious about comparing figures from different surveys. Several features of question wording tend to matter.
- The target. Trust in “people in general”, “people in your neighbourhood” and “people you meet for the first time” are different questions.
- The response options. A two-option choice, a five-point scale and an eleven-point scale produce different patterns, even from the same attitudes.
- The order of questions. A question about crime or a recent news event placed just before a trust item can influence the answer.
- The framing of the extremes. Phrases like “you can’t be too careful” can nudge people towards caution because that option sounds sensible.
- The survey mode. People may answer more candidly in a self-completed form than in a conversation with an interviewer, in line with the effects of social desirability.
None of this means the questions are useless. It means that changes over time are most trustworthy when the same wording, options and mode are used throughout, and less so when they vary.
Trust, behaviour and context
Reported trust and actual behaviour do not always line up. Some studies pair surveys with small experiments, such as games involving sharing money or leaving items in public, to see whether stated attitudes predict conduct. The results are mixed and depend on the design, and such experiments have their own limits since a laboratory choice is not the same as everyday life.
Context also matters. Trust is influenced by personal experience, local conditions, recent events and people’s own circumstances, and a drop in a survey figure might reflect any of these. People who have experienced unfair treatment, or who live in places where they feel less secure, may report lower trust for reasons that are understandable and not a character trait. Averages for a country or city can also hide very different experiences within it, which is the logic behind the ecological fallacy.
Who answers matters
Community surveys depend on who responds. People who are at home more, who are more settled in an area, or who are more comfortable with surveys may be more likely to take part, and these groups may differ in trust from those who do not. Weighting can adjust for known differences such as age and sex, but it cannot fully correct for traits that are not measured. When a local survey has a low response rate, it is worth being cautious about the reported figure, since the people who did not respond may differ in ways that are hard to know.
Reading trust headlines
A claim that trust has fallen, or that one place is more trusting than another, is more credible when it states the exact question, the response options, the sample and the years compared. If a story gives only a percentage without these details, the number may be real but its meaning is uncertain. It is also reasonable to ask whether a change could come from the question or sample rather than from people. Trust figures are valuable as one input to understanding communities, and they are most useful when read alongside other evidence, such as participation, local records and the accounts of residents themselves.