Living alone does not predict being isolated

The proportion of households made up of one person has risen sharply in Europe and North America since the 1960s, driven by ageing, the retreat of couple living and longer periods of independent life. In France, single-person households have become, census after census, one of the most common configurations.

That trend is robust and documented by censuses. It is regularly cited as proof of rising isolation, an inference relational data support only partly. Eric Klinenberg offered in 2012, in Going Solo, a sociological reading that stresses the economic and urban resources this way of life presupposes.

People living alone report on average as many or more outside contacts than people living as a couple, a result found in several European surveys. It runs against intuition, and yet it is regularly found.

The explanation is mechanical: living together supplies a substantial share of daily interaction, which reduces the need to seek it elsewhere. The absence of a cohabitant produces the opposite effect. Living alone does not mean being alone, a distinction demographers regularly point out.

The aggregate mixes opposite situations: young adults early in their path, separated people, widowed older people. Their exposure to isolation differs greatly. Some live alone at the start of a life course, others at the end of a shared life, and nothing justifies adding them together without care.

Analyses separating those groups find a marked association between living alone and isolation among older people, and virtually none among young urban adults.

Among the oldest, the association stems partly from factors that accompany living alone without being reducible to it: recent widowhood, reduced mobility, distance from relatives. Analyses that include them take them into account separately.

Studies distinguishing residential contexts find substantial gaps: living alone in a dense, well-served municipality does not have the same relational consequences as in an isolated area. Access to transport, shops and meeting places conditions the very possibility of seeing others.

That factor is rarely introduced in published analyses, which contributes to the apparent contradiction between studies on this subject. It also explains why the same proportion of people living alone may correspond, from one area to another, to unrelated realities.

Using the share of single-person households as an isolation indicator produces mistaken diagnoses, notably territorial ones, since young urban areas then appear the most affected.

Statistical agencies recommend using direct relational measures, available in several surveys, rather than a convenient demographic proxy. Living alone is, moreover, for some of those concerned, a deliberate choice, which demographic indicators do not distinguish from an imposed situation.