Administrative data is incredibly valuable for research on taxation for several reasons. It is much more disaggregated than macroeconomic data on tax payments and income, it often represents the longest available time series of data, and tax records capture income more accurately than any other source of data. With surveys, respondents often do not provide honest or accurate answers about their income, so data is often underreported, and researchers have to impute likely tax payments based on a proxy for income, assuming full compliance with tax laws. Administrative data solves these problems because it includes information on declared incomes and actual tax payments, with no need for proxies.
Administrative data has a few drawbacks: only data declared to the revenue authority is captured, so information about informal income is missing; it is not always possible to match different sources of income to the same taxpayer because they may have multiple taxpayer identification numbers; and only the data recorded for tax assessment is available. For example, Rwanda does not record gender information in tax declarations. Still, there is much to be gained by using administrative data for tax research.
This has been demonstrated recently by research done using the Rwanda Revenue Authority (RRA)’s administrative data. Researchers from RRA, the international Centre for Tax and Development (ICTD), and the African to provide a comprehensive analysis of the Rwandan tax system, including quantifying audit probabilities, compliance gaps, and tax burdens, for which there was previously little evidence.
Authors
