Ageing and Housing Demand: Dutch Cohort Evidence

In the Netherlands, and indeed elsewhere in Europe, homeownership (rates) and housing consumption has increased over time. On basis of repetitive cross-section analysis one can conclude that homeownership rates are lower for younger and older people, while people aged in-between are overrepresented in the homeownership sector (m.m. for housing consumption). Based on these findings, one could conclude that the ageing of Europe will lead to lower demand for owner-occupied dwellings and more generally to lower levels of housing consumption. Actually, the relation between age and homeownership/housing consumption is more complicated than that. Three issues are at stake here: age, birth cohort (generation-effect) and time (period). On forehand, one can expect housing consumption by age follows an inverted U-shape, while housing consumption and both the cohort-effect and time-effect are positively related. The way to disentangle these processes is by cohort analysis. Cohort analysis seeks to de-compose the observed trend into distinct effects associated with the life cycle, birth cohort and time. The analysis treats an outcome variable - i.e. homeownership, housing consumption - as a function of age, cohort and time. We modelled several distinctive cohort-models to assess the relation between housing consumption and age/cohort/time. Our models show that in the coming decadeís one can expect - under the assumption of no structural breaks - that the ageing of society does not necessarily leads to lower housing consumption; i.e. the negative age-effect is offset by changing attitudes over time towards housing (positive cohort and time-effects).