Abstract
We oer a structural interpretation of survey measures of consumer confidence. Our approach is based on a simple forward-looking model of consumption. The model decomposes observed consumption fluctuations in changes due to fundamentals, and changes due to temporary errors caused by noisy information. Our model-based measure, estimated using national accounts, closely mimics out-of-sample survey data in the U.S. and a majority of European countries. The results provide a theoretical foundation for the use of survey-based consumer confidence indices. In addition, since national accounts are an internationally consistent measure of activity, our structural method provides an alternative and internationally consistent measure of consumer confidence.