Relationships across generations, outside the family

Age is among the dimensions where measured homophily is strongest, comparable to that of education. Personal networks consist overwhelmingly of people in the same age band. The review by Miller McPherson, Lynn Smith-Lovin and James Cook published in 2001 ranks it among the best-documented forms of homophily.

That concentration is largely induced by institutions: age classes at school, cohorts entering employment, activities segmented by age. It needs no preference to occur. Gunhild Hagestad and Peter Uhlenberg described in 2005 this institutional segregation of ages, produced by the very organisation of the life course.

Workplaces, voluntary activities and collective commitments are the main contexts where people of distant ages meet regularly outside the family.

Surveys find that these contexts produce durable relationships less often than age-homogeneous contexts, the difference stemming partly from diverging life timetables. A person of twenty-five and another of sixty do not go through the same stages at the same time.

Work on access to information and on entry into employment finds a role for relationships with older people, who occupy different positions and hold non-redundant information.

That reasoning is the weak ties argument, formulated by Mark Granovetter in 1973: usefulness stems not from the intensity of the relationship but from the distance between the positions occupied.

The benefit is not one-way. Older people also gain access to practices, knowledge and networks that would otherwise remain foreign to them, even if this direction of exchange is measured less often.

Mentoring, intergenerational cohabitation and volunteering schemes deliberately organise such encounters. Available evaluations mainly concern school mentoring programmes. In France, intergenerational home-sharing was given a legal framework in 2018 by the ELAN housing law.

These evaluations find modest effects on the measured indicators, clearer on participation and attendance than on school results, with strong dependence on the duration and stability of the relationship.

Personal network surveys rarely code the age of the people named, which makes measuring these relationships difficult at population scale. They ask for the respondent's age, rarely that of their relationships.

Existing work therefore concerns mainly organised schemes, a particular population, which says nothing about ordinary intergenerational relationships. It leaves in the shadows friendships born of a neighbourhood, a pastime or a former working relationship, whose frequency remains unknown.