What the end of working life does to relationships

Relationships born at work make up a substantial share of employed people's networks. Maintaining them depends largely on daily contact, which stops abruptly. That context supplied both opportunities and a reason to meet.

Longitudinal work finds a fall in the number of active relationships in the years following the end of employment, more marked for people whose network was mainly professional. The meta-analysis by Cornelia Wrzus and colleagues, published in 2013, places this fall within a longer trend: the size of personal networks declines over adulthood, and retirement is one of its stages.

Some people compensate through voluntary engagement, collective activities or family investment. Surveys find a rise in associational participation in the first years after retirement.

That compensation is very unevenly distributed: it is more frequent among people who already had relationships outside work, which widens pre-existing gaps.

Two old readings of this moment stood opposed. Disengagement theory, formulated by Elaine Cumming and William Henry in 1961, saw it as a mutual and expected withdrawal; activity theory saw it as a loss to be offset. Neither has withstood scrutiny as a general law.

A chosen departure and an imposed one, for health reasons or redundancy, produce different trajectories in surveys, the second situation being associated with a sharper fall in participation.

Analyses distinguishing those situations are few, and most published series aggregate departures of very different natures. An average that blends these trajectories may mask two movements in opposite directions.

Differing retirement timetables within a couple alter the organisation of time, and surveys find effects on both partners' participation. When one retires while the other is still working, a period opens in which daily rhythms diverge.

Those analyses are recent and their results vary by country, depending notably on retirement rules and the usual age gaps within couples.

Schemes aimed at older people mainly measure participation in organised activities, an indicator of isolation rather than of loneliness.

The gap between the two measures, established elsewhere, explains why schemes effective on participation may produce no measurable effect on reported feeling.

The shrinking of the network does not, moreover, mean the same thing for everyone. Socioemotional selectivity theory, proposed by Laura Carstensen in the early 1990s, sees it partly as a choice: favouring the relationships judged most important. A tighter circle is therefore not automatically to be read as a loss.