How many hours to become friends

Jeffrey Hall published in 2019 two surveys, one of people who had recently moved, the other of first-year students, in which participants estimated time spent with a new acquaintance and characterised the resulting relationship.

From these he drew thresholds: several dozen hours to move from acquaintance to casual friend, around a hundred for a friend, more than two hundred for a close friend. It is those three figures, and not the reasoning around them, that circulated.

The choice of the two populations is not incidental. A move or the start of a university year places people in a period when many relationships begin at the same time, which makes observation easier but makes these situations unrepresentative of ordinary adult life.

Time spent is estimated by the participants themselves, retrospectively. Duration estimates are known to be imprecise and sensitive to the importance attached to the relationship, a memorable evening weighing more heavily in memory than ten unremarkable lunches.

The characterisation of the relationship is also self-reported and made at the same moment, which creates a possible circularity: judging a relationship close may lead to estimating that more time was spent on it. The direction of the relationship between the two variables therefore remains undetermined.

The existence of an increasing relationship between shared time and reported closeness is found, which is not obvious in a domain where many long-standing relationships persist without frequent contact.

Hall also finds that the nature of the time matters: shared leisure time is associated with closeness, shared work time is not or barely, which is the article's most interesting result. It suggests that the setting in which people see each other weighs as much as the quantity of presence.

The same work and others find that the content of exchanges weighs more than their duration: conversations involving personal disclosure are associated with faster progression in reported closeness, a result consistent with the social penetration theory formulated by Irwin Altman and Dalmas Taylor in 1973.

That observation qualifies the hourly thresholds: two hundred hours of shared presence without such conversation does not produce the same result as fifty hours with it.

A precise threshold from a self-report study on a small sample should not be read as a constant. Its value is to indicate an order of magnitude, and above all to make visible that time is a scarce resource, one that relationships compete for.

Public coverage often retained the figure while forgetting the method, a configuration that has become usual for easily memorable results. A round number travels better than a methodological caveat, and that is precisely why it deserves to travel with its own.