Meet Cynthia Cheboi: The ATRN’s best research paper winner

In the world of academia, few accomplishments carry as much prestige as receiving the African Tax Research Network (ATRN)’s Best Research Paper Award. On 5 September, Ms Cynthia Cheboi, a data scientist, dedicated researcher and tax professional at Kenya Revenue Authority (KRA), was named the recipient of this distinguished award for her groundbreaking paper, 'Predicting Risky Taxpayers Using Machine Learning. We explore the significance of this award, the key insights from Ms Cheboi's award-winning work, and the impact it could have on tax policy and the broader academic community. 

Please tell us more about your research paper, what is it about?

The title of the research was 'Predicting Risky Taxpayers Using Machine Learning.' The study aimed to classify taxpayers based on specific attributes to identify those likely to evade payment of taxes. This would in turn facilitate targeted interventions, such as enforcement actions, tax audits, and taxpayer education, with the goal of minimizing revenue loss and optimizing resource allocation for effective tax administration.

What inspired you to explore the use of machine learning in tax administration especially on identifying risky taxpayers in Kenya?

I was inspired by the growing need to enhance tax compliance and resource allocation within the tax system. With tax evasion being a persistent challenge, I saw the potential of Machine Learning (ML) to transform how we identify risky taxpayers. ML offers a data-driven, objective approach to detecting non-compliance patterns that may not be immediately obvious through manual methods. By leveraging technology, we can optimize tax enforcement and ensure more efficient allocation of resources to critical areas.

What are some of the practical implications of your research for tax authorities in Kenya and across Africa?

The practical implications of my research for tax authorities in Kenya and across Africa include:

1. Targeted Interventions: By utilizing machine learning models, tax authorities can identify high-risk taxpayers more effectively, allowing for targeted audits and resource allocation to combat tax evasion.

2. Generalization of Findings: The methodology can be adapted to assess fraud risk across different tax types, providing a broader framework for tax compliance initiatives throughout the continent.

Following your ATRN best paper award, the Commissioner General of KRA awarded you a scholarship to study in Japan. How will the scholarship shape your future research? 

The scholarship to study in Japan is a remarkable opportunity that I deeply appreciate. Japan is one of the most developed countries, known for its advancements in ICT and the tax field, innovation and systems development. I aim to learn from these innovations and explore how they can be adapted and applied to improve systems in Africa, particularly in tax administration and public policy, which will serve to address contemporary issues. Additionally, I look forward to experiencing Japan's strong work ethic and effective leadership practices, which I believe will further enrich my professional growth. I believe it is also an opportunity to experience a different culture.

What was the most rewarding part of the research process for you?

The most rewarding part of the research process for me was receiving recognition at the ATRN Congress and understanding that my research could contribute to meaningful changes in tax policy and administration was incredibly fulfilling. Engaging with peers and experts in the field further enriched my perspective and fuelled my passion for continuous improvement in tax compliance practices.

How did you feel when they announced that you’d won the “Best Tax Research” award?

It was such a humbling, exciting, and fulfilling experience. All the effort that went into writing the paper had paid off, and I am very grateful to God for its success.

How has winning this award impacted your career?

Winning the ATRN Best Paper Award has significantly impacted my career and academic journey by enhancing my credibility as a researcher in tax administration. It has opened new opportunities for collaboration, mentorship, and networking. The recognition has also motivated me to further pursue innovative solutions in tax compliance and has positioned me as a thought leader in leveraging technology for improved tax policy. 

What advice would you give to students or researchers who are just starting out in tax research?

My advice to students or researchers starting out in tax research is to remain curious and open-minded. Embrace a multidisciplinary approach, as tax issues often intersect with economics, technology, and public policy. Collaborate with peers and mentors to enrich your understanding and gain diverse perspectives. Prioritize staying updated on the latest advancements in data analytics and technology, as these tools can significantly enhance your research. Finally, focus on real-world implications and seek to address practical challenges faced by tax authorities to make your research impactful and relevant.

How do you see your work contributing to Africa’s development goals?

My future research plans focus on leveraging Artificial Intelligence (AI) to improve the accuracy and reliability of predictive models in tax administration. This will enable better case selection for targeted interventions, especially in the context of limited resources. Additionally, I aim to explore ways to enhance revenue through tax base expansion, simplify tax processes for micro and small taxpayers, reduce the cost of compliance for both taxpayers and tax administrators, and how tax authorities can effectively address hard-to-tax areas, such as informal economies.