People form long-term partnerships with others of similar educational attainment at rates well above chance, which concentrates advantage within households.
The regularity is robust across countries and has strengthened in several as higher education expanded and partners increasingly met through it. The consequence is arithmetical rather than moral: pairing two high earners and two low earners produces a wider distribution between households than the same four people paired at random, with no change in any individual's position.
In practice
- A country whose individual earnings gap is stable can have a widening household gap, and the change is in who married whom.
- A university is an efficient labour market and an efficient marriage market, and the second has the larger effect on the next generation.