OECD Taxation Working Paper No. 45

The potential of tax microdata for tax policy

Author: Seán Kennedy

Abstract:

This paper explores one distinctive form of the ‘big data’ of economics – individual tax record microdata – and its value and potential for tax policy analysis. The paper draws on OECD collaborations with Slovenia and Ireland in 2018 where tax microdata was used. Much of empirical economics is based on survey data. However, the current trend of low and falling response rates has placed a question mark over the future value of survey practice generally. By contrast, administrative microdata are increasingly used in some of the world’s most credible and influential economic research on public policy. Although tax microdata have limitations, they offer vastly greater scale and coverage, particularly among the highest earners, and have an inherent longitudinal structure with near perfect tracking rates. This paper argues that these characteristics produce analytical advantages in the depth and breadth of possible analysis. The paper also argues that the future of best-practice tax policy analysis will likely combine the unique advantages of tax microdata with survey and national account data. The complementary advantages of these combined data will be important for policymakers to address future policy challenges – including ensuring sustainable tax revenues in an era of population ageing, protecting citizens from rising income inequality and preserving fairness in a changing labour market. Despite its potential for tax policy analysis, access to and use of tax microdata remains limited and sporadic. The primary reason for limited access is to protect privacy. Furthermore, new skills and technology will be needed to manage the unstructured and multidimensional nature of these data, which are not designed for research. Despite these challenges, tax administrations are increasingly adopting new technologies, statistical software and trained analysts to capture and process large volumes of data securely. This development could support new data access solutions where data are used more readily for tax policy research while limiting privacy risks.

Download the paper

Comments are closed.