Risk-based data analytics in the government sector: A case study for a U.S. county

This study explores the use of data analytics for improving the quality of government audits through the lens of processing fluency theory as the driver behind the need for data analytics. Little is known about the benefits of data analytics to government expenditure audits in a data-rich environment. Accordingly, this study proposes a risk-based prioritization framework and applies it to the real procurement dataset of a US county. The results indicate that the framework increases the efficiency and effectiveness of identifying true duplicate payments compared to either the scanning or sampling benchmarks. Specifically, it significantly reduces the number of potential duplicate candidates that require auditor review to approximately 12% of the duplicate records. As such, it enables the capture of true duplicates in a shorter period. These results suggest that the framework offers one way to mitigate the low processing fluency effect of information overload on auditor judgment.