Housing prices and measuring the households' standard of living

The deflators used in the calculation of traditional measures of purchasing power and standard of living trace only part of the sharp increase in real estate prices observed since the late 1990s. Failure to include house prices in these measures would be meaningless if these prices had no effect on aggregate welfare. Three arguments are frequently put forward in this sense: neutrality between households (rising prices reduce the purchasing power of buyers but symmetrically increase the purchasing power of sellers), neutrality over the household life cycle (the additional savings effort due to rising prices is compensated by an increase in purchasing power later in the life cycle), and compensation by interest rates (lower rates compensate for the effect of rising prices). We take up the analytical framework of Bajari et al. (2005) and show that the validity of the hypothesis of neutrality between households depends on whether the price increase is temporary or permanent. Life-cycle neutrality is examined in the framework of a stylized model: it is verified only if the housing has no use value. When it offsets the effects of price increases during the homeownership period, lower interest rates lead to increased purchasing power later in the life cycle. The effects of a rise in property prices are heterogeneous according to income because of specific constraints (debt and minimum surface area). Thus, welfare losses have a stepped or inverted U-shaped profile, and are maximal for intermediate incomes. A constant-utility index imperfectly reflects this heterogeneity when the rise in prices induces an eviction effect from the real estate market.