Methodological note: Corporate tax abuse
Summary
The document outlines a methodology for detecting corporate tax abuse by comparing reported profits of large MNCs against theoretical profits based on economic activity. It highlights that CBCR data provides unprecedented country coverage, particularly for low- and middle-income countries, though it requires machine learning to correct for data gaps and aggregation biases.
Key insights
- Country-by-country reporting (CBCR) data is superior to other sources because it offers more substantial country coverage, especially for low- and middle-income countries. For instance, US CBCR data includes tax and profit information for 25 African countries, whereas Bureau of Economic Analysis data only covers three.
- CBCR data is based on tax accounting rather than financial accounting, which means it reflects actual tax payments and is less likely to underestimate profit shifting compared to datasets like Orbis.
- The methodology employs a profit misalignment model where theoretical profit is calculated using a weighted combination of employees (25%), wages (25%), and unrelated party revenues (50%). Profit shifting is then identified as the difference between reported and theoretical profits.
- To prevent outliers, the model assumes that MNCs do not shift profits to countries with tax rates higher than 25%, setting the profit misalignment for such foreign observations to zero.
- The researchers used a Histogram-based Gradient Boosting Regression Tree (via scikit-learn) to estimate missing data for employees and sales in aggregated country categories, which typically increases total shifted profits by approximately 30%.
- Data completeness varies significantly by reporting country; for example, while the US 2017 data is nearly complete, China's reported data to the OECD only included 82 companies compared to 583 expected according to Orbis.
Cite the original document
- APA
- CCFD-Terre solidaire (2020). Methodological note: Corporate tax abuse. https://ccfd-terresolidaire.org/wp-content/uploads/2020/11/sotj_methodology_cbcr.pdf
- Chicago
- CCFD-Terre solidaire. Methodological note: Corporate tax abuse. 2020. https://ccfd-terresolidaire.org/wp-content/uploads/2020/11/sotj_methodology_cbcr.pdf.
- Wikipedia
- {{cite report |author=CCFD-Terre solidaire |title=Methodological note: Corporate tax abuse |date=11 November 2020 |url=https://ccfd-terresolidaire.org/wp-content/uploads/2020/11/sotj_methodology_cbcr.pdf |access-date=5 November 2026 |via=Climate Insights Directory (document ccfdts-001351)}}
- BibTeX
- @techreport{ccfdterresolidaire2020methodological, author = {{CCFD-Terre solidaire}}, title = {{Methodological note: Corporate tax abuse}}, institution = {CCFD-Terre solidaire}, year = {2020}, month = nov, url = {https://ccfd-terresolidaire.org/wp-content/uploads/2020/11/sotj_methodology_cbcr.pdf}, urldate = {2026-11-05}, note = {Indexed by Climate Insights Directory, document ccfdts-001351} }
Full text
Collected · Record updated