Economists have devoted increasing attention to the magnitude and consequences of measurement error in their data. Most discussions of measurement error are based on the "classical[equal, rising dots] assumption that errors in measuring a particular variable are uncorrelated with the true value...