In analysing urban growth and change, economists assume that it makes little difference whether land or property is rented or owner-occupied. This paper argues that this is a false assumption, particularly in the case of housing. When property is rented the division between landlord and tenant separates the owner from the occupier. So if the demand for property in an area increases then landlords increase rents forcing renters to pay up or move. But for owner occupiers the increased demand increases the value of their property so that, as owners, they make a capital profit which cancels out the increased opportunity cost of occupying the property and significantly reduces, indeed virtually eliminates, the pressure to move. Thus owner occupation lengthens the adjustment process, possibly by decades. In theory, of course, they should move to downsize in order to readjust the risk element in their portfolio of investments, since they are now overinvested in housing. But they do not do this for a number of economic and psychological reasons which are developed in the paper building on ideas put forward in Evans (2004). So adjustment in the market depends on retirement or death and requires decades. In the meantime relative prices are out of equilibrium on a virtually permanent basis. One result, as in the Tokyo metropolitan area, may be increased sprawl. Another result, as in south east England, may be prices significantly higher than they would be otherwise as a large proportion of the older population continues to occupy properties that are unaffordable by younger families with similar incomes.