Unveiling a novel approach to corruption measurement: Leveraging household survey data on income and expenditure through forensic analysis☆

Corruption exacerbates and perpetuates the global problems of poverty and inequality. However, accurately quantifying corruption, a prerequisite for any sensible analysis and policy formulation, is hindered by its clandestine nature and the reticence of involved parties. To overcome these obstacles, we conduct a forensic investigation to uncover the extent of bureaucratic corruption in Vietnam. Leveraging self-reported income and expenditure data from a nationally representative dataset and an instrumental variable approach, our analysis reveals that bureaucratic corruption accounted for 2.68% of the country's GDP in 2018, with variations observed across government sectors and with seniority levels. We identify inadequate remuneration, ambiguous regulations, and inconsistent enforcement as primary drivers of corruption, particularly in the education sector. Validating the findings with a unique dataset on extra-class expenditure, we demonstrate the empirical plausibility of our econometric estimates. This methodology can leverage a more reliable and comparable cross- countries corruption database.