Big data can help the KRA boost revenue collection

The World Bank’s recent finding that the rate at which Kenya’s tax revenues are growing is lagging economic growth raises concern over the country’s tax buoyancy.

Tax buoyancy is an indication of the degree to which tax collected responds to changes in the overall economy. A tax-buoyant system needs to have an elasticity with respect to growth in nominal GDP of at least one.

The 2018 World Bank Kenya Economic Update, however, says the buoyancy of Kenya’s main tax heads such as income tax and VAT is much weaker than the ideal ratio.
The trend in revenue mobilisation suggests that the capacity to mobilise more revenue has plateaued and can only be rejuvenated with structural reforms to address the anomaly.
Such are the observations that have seen the Kenya Revenue Authority (KRA) announce a raft of new measures aimed at netting hitherto invisible taxpayers, as it aims to promptly plug the existing revenue collection gap vis-à-vis the ever-growing national budget.
For instance, the taxman wants to mine additional data by gaining access to key databases in both the private and public hands. Amongst these are records maintained by telecoms operators, statutory bodies like the National Construction Authority, Kenya Power, payrolls such as those of hospitals and registers maintained by regulatory and professional bodies.
As a developing economy, Kenya is still grappling with debilitating taxpayer apathy. A significant chunk of the income-generating population has an averse attitude to paying taxes and are either indifferent or evasive of their tax obligations.
The KRA’s initiatives such as tax amnesties, tax education and an enhanced customer care support are yet to gain traction amongst them. But these taxpayers continue to thrive, at the expense of the tiny fraction of the current ones, who are unfairly shouldering an ever-growing tax burden and are indeed getting tax-fatigued.
It’s intriguing to note that, against a backdrop of 19.6 million registered voters as at 2017 and 29 million active M-Pesa subscribers as at June 2018, only 3.2 million Kenyans filed their 2017 self-assessment tax returns by June 2018.
The reality is that a large number of those who did filed nil returns, even though they didn’t earn zero income in 2017. Some only declared a portion of their income.
For some entities, nil end-year returns were filed even though the businesses were actively trading during the year, with PAYE and VAT monthly returns processed.
The world over, tax authorities, the KRA included, are in the throes of utilising emerging technologies that develop new capabilities and establish a culture that is data-driven and intelligence led.
The simple premise is that if the taxpayer isn’t forthcoming with all declarations, the taxman’s systems shall mine these in an effort to assess the taxes due.
Thankfully, digitisation of the economy has led to unprecedented growth in data flows that are leading to a paradigm shift towards what is now known as Big Data, an emerging field in business and government and a topic deliberated upon in the just concluded Annual KRA Tax Summit.
Simply put, Big Data is an evolving term that describes large and complex datasets that have the potential to be mined for information. Big Data in itself is not a database. It is data collected from a large variety of sources, including the Internet, social media, sensors, text messages, video, images and audio files as well as other sources that are often unstructured. It is characterised by four V’s, namely Volume, Variety, Velocity and Value.
Volume means that large amounts of data are involved and Variety means that big data may be structured, semi structured or even unstructured. Velocity calls for big data to be constantly analysed quickly to derive the greatest value while Value means that big data analytics offers users the opportunity to extract better insights out of existing data.
Traditionally, data available to revenue bodies came from tax returns, or through third parties, often as part of withholding tax arrangements and supplemented by information collected during tax audits or assessments and research.
Typically, the bulk of data that taxpayers and intermediaries supply was and continues to be in a structured form specified by tax administration in accordance with the law. Unstructured tax data sets were not frequently accessed or supplied, often because taxmen lacked the tools to analyse and utilise them.
The widespread use of inexpensive and constantly connected digital devices has changed the model, from data capture, analysis and deducing customer behaviour, to direct measurement.
This availability of new sources of data is now allowing the monitoring and reaction to issues and opportunities happening in real-time based on live data, making tax assessments possible in real-time or near real-time.
Big Data tools can provide the ability to monitor and automatically send reports, whether it is tracking the performance of business assets, monitoring quality and risk, or providing performance information.
Real-time tax compliance monitoring, coupled with matching and analysis of information flows should lead to enhanced taxpayer transparency thereby reducing the propensity for tax evasion, avoidance and fraud.
The taxman can examine and understand historical activity and taxpayer behaviour through storage of original, unaltered information. Analysis can also be done across multiple periods, taxpayers and tax domains, allowing the tax administration to plan its compliance, control and risk management activities.
As expected, there are obvious ethical and privacy issues already raised that need to be addressed. The Kenya Data Protection Bill 2018 needs to take cognizance of these and offer appropriate solutions before it becomes law.

Ndegwa is a Tax & Business Advisor at Anchinga & Associates (CPA-K). [email protected].