Measurement Error in Income and Schooling and the Bias of Linear Estimators

• 成果类型: Article
• 署名作者: Bingley, Paul; Martinello, Alessandro
• 署名单位: Lund University
• 刊物名称: JOURNAL OF LABOR ECONOMICS
• ISSN/ISSBN: 0734-306X
• DOI: 10.1086/692539
• 发表日期: 2017
• 页码: 1117-1148
• 关键词: Administrative data; earnings; extent

摘要

We propose a general framework for determining the extent of measurement error bias in ordinary least squares and instrumental variable (IV) estimators of linear models while allowing for measurement error in the validation source. We apply this method by validating Survey of Health, Ageing and Retirement in Europe data with Danish administrative registers. Contrary to most validation studies, we find that measurement error in income is classical once we account for imperfect validation data. We find nonclassical measurement error in schooling, causing a 38% amplification bias in IV estimators of the returns, with important implications for the program evaluation literature.