Predicting Customs Fraud Using Machine Learning and Mirror Analysis in Togo

Predicting Customs Fraud Using Machine Learning and Mirror Analysis in Togo

Predicting Customs Fraud Using Machine Learning and Mirror Analysis in Togo

Authors: Pouwemdéou Tchila, Komlan Kawa Agbanho and Abalo Bouwe

ISSN: 2709-8575
Affiliations: Docteur en sciences économiques, Data scientiste, Chef division analyse risques et suiviévaluation, Office Togolais des Recettes & Chercheur associé au CREAMO (Université de Lomé); Docteur en sciences économiques, Inspecteur des Douanes, Chef section brigade à la Division des Opérations Douanières de Kwadjoviakopé de l’Office Togolais des Recettes; Master en Statistiques & Mathématiques, Data scientiste, Chargé de l’analyse des risques et de la programmation fiscale, Office Togolais des Recettes
Source: African Multidisciplinary Tax Journal, Volume 5, Issue 1 (2025), p. 1–26

Share

Cite this article

Tchila P; Agbanho KK; Bouwe A
Predicting Customs Fraud Using Machine Learning and Mirror Analysis in Togo
African Multidisciplinary Tax Journal Volume 5, Issue 1 (2025) p. 1–26
https://doi.org/10.47348/AMTJ/V5/i1a1

 

Abstract

Customs fraud is an inherent phenomenon of customs administrations and is most often responsible for undermining customs revenue collection. In an attempt to combat this phenomenon, customs administrations, particularly in developing countries, often conduct extensive and unstructured audits. This is not conducive to the fluidity of international trade. The objective of this study is to analyse the extent to which the use of machine learning and mirror analysis improves the identification of customs fraud, while preserving the objective of revenue mobilisation. Using data from the Togolese Revenue Authority and COMTRADE, the findings indicate that mirror analysis and machine learning can better enhance customs fraud detection. To this end, the study recommends the use of these tools in fraud detection.

Predicting Customs Fraud Using Machine Learning and Mirror Analysis in Togo

Understanding the Tax Payment Compliance of Companies: Evidence from Eswatini

Understanding the Tax Payment Compliance of Companies: Evidence from Eswatini

Authors: Phindile T Masuku, Dr Fabrizio Santoro and Ziyanda T Dlamini

ISSN: 2709-8575
Affiliations: Manager, Research, Strategy and Statistics Division at the Eswatini Revenue Service; Research Fellow, Institute of Development Studies; Economic Analyst, Research, Strategy and Statistics Division at the Eswatini Revenue Service
Source: African Multidisciplinary Tax Journal, Volume 5, Issue 1 (2025), p. 27–50

Share

Cite this article

Masuku PT; Dr Santoro F, Dlamini ZT
Understanding the Tax Payment Compliance of Companies: Evidence from Eswatini
African Multidisciplinary Tax Journal Volume 5, Issue 1 (2025) p. 27–50
https://doi.org/10.47348/AMTJ/V5/i1a2

 

Abstract

This paper investigates corporate income tax (CIT) payment compliance among corporations in Eswatini, an underexplored area critical for revenue-constrained low and middle-income countries. Using a unique administrative dataset (2017–2022), we analyse factors driving timely and full tax payments. We combine descriptive analysis with a more robust Heckman selection model to address sample selection bias. Results show that, while 82% of filed returns included payments, only 55% were fully compliant, and 42% were late. Compliance was higher among larger firms and those in urban tertiary sectors, while smaller and rural firms frequently overpaid, potentially due to penalties. Electronic payments exhibited the highest compliance, whereas mobile and cash payments lagged. Regression analysis highlights company size, provisional tax filings and electronic payments as key compliance predictors. This study contributes to the tax compliance literature with actionable insights for revenue authorities, from the simplification of tax processes for smaller firms to the larger implementation of electronic tax payments.