Raw XBRL provides greater insights than commercial data set in tax case study
XBRL US has published an interesting case study, presenting student research on whether big companies are paying their “fair share” of tax. “Controversial questions require the right dataset,” it argues, with access to high quality, timely data enabling students (and others!) to carry out more meaningful analyses.
Students at the University of Mississippi’s Patterson School of Accountancy investigated 55 large companies that paid no tax in 2020, comparing their Effective Tax Rates (ETRs) with those of peers. Initially, they used a commercially available data set. As the case study explains, such data sets are usually ‘normalised’, or structured in accordance with a set of norms set by the data provider. Normalised data often aggregates companies’ reported facts, making it easier to compare multiple companies across a single line item, but reducing specificity and detail – and indeed, the students were unable to determine statistically significant associations between the companies and various economic factors, including ETRs.
On the other hand, the students were able to examine machine-readable XBRL data from individual companies, exactly as reported. Using an example, they were asked to determine how a profitable firm can pay no tax, reconciling its ETR with the US Statutory Tax rate of 21%, and delved deep into the individual digital tags to pinpoint factors such as tax breaks and the use of foreign subsidiaries to record earnings.
“The higher degree of specificity available in as-reported data allows students to answer more questions when researching the impact of policy and other economic factors on individual companies,” argue authors Christine Cheng of the University of Mississippi and Campbell Pryde of XBRL US, with the consistency of underlying XBRL tags enabling more accurate comparisons. The ability to visualize the metadata associated with each reported fact through Inline XBRL can enhance the learning experience for students, while they also build valuable skills for their future careers.
The case study materials, including a teaching case and Python data script, are now available for use by other academics. XBRL US encourages all those interested in accessing XBRL data for classroom and research use to get in touch.
And if you’re already using XBRL data to generate your own timely insights and improved resolution on important questions anywhere in the world, we at XBRL International would love to hear more about it, and share your case study with our readers.
Read more here.