While survey-based consumer confidence indices are useful and internationally available, it has a potential drawback such that survey measures may not be consistent intentionally. In this paper, we propose to measure consumer confidence by exploiting a discrepancy between observed productivity and information available to consumers using national accounts based on a simple consumption model. With a standard signal extraction with noisy information, we estimate model-based filtered consumer confidence indices for the U.S. and fifteen European countries. Since national accounts are internationally consistent and coherent for measuring activity, a structural method based on national accounts can provide a measure for consumer confidence that is internationally consistent.