Attenuation bias, recall error and the housing wealth effect

The greater use of microeconomic and survey based data in addressing key financial stability related questions is a natural outcome of the recent financial crisis. Amongst other benefits, the use of such data enables a more precise understanding of the differing attitudes and responses of individual agents such as households to financial shocks. However, some difficulties can arise with the use, in particular, of survey data in this regard. In this paper we calculate measurement error in the house prices “recalled” by a representative sample of mortgaged Irish households and illustrate the degree of attenuation bias consequently introduced into estimates of housing wealth effects, when recall as opposed to actual house prices are used. The presence of “recall” or measurement error in house prices could be one reason for the relatively small housing wealth effects found in certain micro-based cross-country studies. JEL classification: E21, C81, D12.