The efficacy of ability proxies for estimating the returns to schooling: A factor model‐based evaluation
用因子模型框架评估了用认知和非认知能力代理变量校正能力偏差的常见做法，发现这些代理变量无法充分捕捉潜在能力，导致估计偏差。
Summary A common approach to addressing ability bias is to augment the earnings‐schooling regression with proxies for cognitive and non‐cognitive skills. We evaluate this approach using a factor model framework, which allows consistent estimation of the returns to schooling without relying on proxies. The factor model estimators may be viewed as implicitly estimating proxy measurement error and/or accounting for omitted dimensions of ability. A bias decomposition quantifies the contribution of the proxies while the estimated latent skills are used to construct direct tests for their viability. Both sets of results confirm the inadequacy of the proxies in capturing the latent skills.