The missing middle: why economic models struggle with households

Standard economic models treat households as individual utility-maximisers or as a single collective unit ('unitary household model'). Neither captures what actually happens inside families. The unitary model predicts income pooling — it shouldn't matter which household member earns the money. Empirically it very much does: conditional cash transfers to mothers in developing-country programs produce different child outcomes than transfers to fathers, for the same income level. Bargaining models (Nash bargaining, separate spheres) do better but require you to specify fallback positions and outside options, which are hard to measure and vary a lot across cultural contexts. Why this matters for policy: education subsidies, pension design, property rights — all of these interact with intra-household dynamics in ways that aggregate models miss. A pension that vests in the husband's name alone has different effects than one that vests in both spouses, even for equal nominal amounts.