ADB Outlines Two-Year Roadmap for Data-Driven Tax Administration

Rather than attempting to audit every taxpayer, tax administrations should deploy analytics to identify compliance risks, prioritizing investigations where they yield the highest financial impact. By combining filing histories with third-party data, authorities can segment taxpayers into risk tiers. This allows for focused audits on high-risk cases while streamlining services for compliant businesses. The bank warns, however, that algorithms are not inherently objective; historical enforcement data often carries biases that require human oversight, rigorous fairness testing, and transparent audit trails to protect taxpayer rights.

The proposed two-year plan moves from basic data management to institutionalized analytics. In the first six months, administrations are tasked with establishing governance and data inventories, followed by a shift toward operational risk-based selection by the end of the first year. By the 24-month mark, the goal is an integrated model where analytics routinely supports debt management, revenue forecasting, and executive planning. This phased approach discourages heavy spending on sophisticated AI tools before fixing fragmented databases or addressing skill shortages.

Beyond enforcement, the report positions analytics as a cornerstone of fiscal planning. By integrating tax information with economic indicators such as inflation and employment, governments can improve revenue forecasting and use microsimulation to test the impact of policy changes before implementation. For international development partners, the focus is shifting from financing proprietary software to building institutional resilience through open-source technologies and workforce training. Ultimately, the objective is not a fully automated tax authority, but a more capable state that uses evidence to collect revenue efficiently and treat citizens with greater fairness.