Expanding the Frontier of Economic Statistics Using Big Data: A Case Study of Regional Employment
提出一个框架，量化大数据相对于官方统计在特定应用中的有用性，并应用于改进区域就业估计。使用薪资处理数据使州级就业估计误差降低11%，并生成更及时、细化的县级估计，其准确性不亚于官方统计。
ABSTRACT Designing effective public policies depends on accurate and timely economic measurement. Big data promises substantial benefits for improving economic measurement but also presents significant challenges. This paper introduces a framework for quantifying the usefulness of big data for specific applications, relative to official statistics. We weigh the potential benefits of additional granularity and timeliness while examining the accuracy associated with any new or improved estimates, relative to comparable accuracy produced in existing official statistics. We apply the methodology to improving timely regional employment estimates. We find that using data from a payroll processor reduces out‐of‐sample error in state employment estimates by 11%. Additionally, we produce new county‐level estimates offering more timely, granular insights than previously available. Applying a novel test, we cannot reject the hypothesis that the new county estimates have an accuracy in line with official measures. An application to COVID highlights the practical benefits of these estimates.