When Self-Assessment Meets System-Assessment: The Trouble with iTax Return Validation Regime

As the June income tax filing season reaches its peak, many taxpayers are encountering an unexpected shift in compliance. Filing a return is increasingly becoming less about declaring taxable income based on their own records and more about reconciling figures generated by the Kenya Revenue Authority’s (KRA) systems.

The introduction of prepopulated income tax returns was welcomed as a step towards modernising tax administration. The promise was straightforward: that taxpayers would benefit from returns pre-filled with information already available to the Commissioner, reducing errors, easing compliance burdens and improving efficiency.

However, the final implementation has taken a different form. Instead of the envisaged pre-population, the KRA has introduced a validation mechanism that compares taxpayer-declared incomes and expenses against data held within its systems, drawing from multiple sources, including eTIMS invoices, withholding tax filings and customs data captured through the Integrated Customs Management System (ICMS).

To support taxpayers, portions of this data are available for download via iTax, accompanied by user guides on how to align returns with KRA records. Certain items such as depreciation, airline passenger ticketing and staff costs are excluded from validation, while all others are subject to system checks at the point of filing.

On paper, this appears structured and logical. In practice, the experience has proven far more complex.

Taxpayers and practitioners continue to report material discrepancies between business records and the KRA’s data. Instances of duplicated sales transactions have resulted in inflated income amounts. Expense information is often incomplete, particularly where the underlying data sources do not capture the full range of deductible business costs. Yet these same figures are being used as the benchmark against which taxpayers' returns are validated. The result is a compliance process that places taxpayers in a frustrating position.

Businesses already invest heavily in maintaining accounting records, operating financial systems and engaging auditors to ensure that financial statements are accurate and independently verified. These processes are designed to produce reliable figures for tax reporting. Nonetheless, even where accounting records and audited financial statements are accurate and properly supported, returns may still be rejected because they do not align with system-generated figures. In the looming shadow of the return filing deadline, taxpayers are now having to incur additional costs to enhance their systems or otherwise outsource digital solutions in relation to tax technology in hopes of procuring timely tax reconciliations. These challenges undermine the compliance efforts and costs traditionally borne by taxpayers prior to the point of filing a return.

More concerning is the limited transparency within the validation process itself.

When a validation error occurs, taxpayers are typically presented with a single aggregated figure representing their expected income or expense position. However, the error message provides little information that would assist taxpayers in resolving the discrepancy. It does not show the corresponding amount declared in the return, the variance between the two figures, or any indication of transactions that may be driving the mismatch. Worse still, in some instances, the amount displayed as the expected total turnover or expense position varies from the data available for download on iTax. From the user guides shared, the KRA has indicated the variance in expenses could relate to entries in the download data such as duplicates and self-supply transactions that are excluded from the total downloaded figure before validation.

This creates an obvious problem. If the system excludes certain transactions when validating a return but continues to display the unadjusted downloadable totals in its validation message, taxpayers are effectively being shown figures that do not represent the amount actually being validated against the return, therefore rendering the taxpayer’s reconciliation process futile. Consequentially, the filing process then turns into an exercise of “trial and error”, resulting in a chaotic manual reconciliation exercise during the busiest compliance period.

In order to promote transparency in the reconciliation and filing process, the KRA should provide all information relied on by the system, including withholding tax records, and publish to taxpayers the actual details of validation errors preventing return filing.

While the KRA has introduced additional fields to the income tax return for declaration of non-eTIMS compliant expenses as well as adjustments to income and expenditure, many remain hesitant to utilise these fields fully. Their concern is understandable, as disclosures in these fields may effectively become a roadmap for future audits, with such amounts automatically attracting scrutiny and placing a retrospective burden of proof on taxpayers. Even where such concerns are only perceived rather than confirmed in practice, perception plays a critical role in tax compliance. Systems that rely on voluntary compliance must be seen to operate fairly, transparently and predictably.

Even more concerning is the fact that once adjustments to income and expenditure are registered on iTax, there is no further audit trail of the breakdown of the adjustments on iTax, only showing the total adjustment as one aggregated figure in the tax return. This would surely create a governance concern and present as a reconciliation and record-keeping nightmare in future periods, particularly for larger organisations where finance and tax personnel may change over time.

A further concern lies in sequencing. Modern tax administrations are increasingly embracing digital tools, third-party data and real-time reporting to improve compliance. However, these mechanisms are most effective where underlying data is reliable and complete. Where it is not, rigid validation risks shifting the burden of system inaccuracies onto compliant taxpayers rather than identifying non-compliance.

The KRA’s efforts to digitise tax administration are not in question. However, digitisation requires more than technology. It requires accurate data, clear processes and meaningful engagement with taxpayers who must operate within these systems.

The current validation framework risks shifting the nature of tax filing from compliance to reconciliation, with taxpayers bearing the cost of correcting inconsistencies in administrative data over which they have no control.

This creates a paradox. Financial statements prepared in accordance with accounting standards, reviewed by professionals and supported by audit evidence are effectively subordinated to administrative datasets that may not reflect the full economic reality of a taxpayer’s operations.

Beyond these operational challenges lies a more fundamental question about the structure of Kenya’s tax system.

Kenya operates a self-assessment regime across all tax obligations. Taxpayers are responsible for determining their taxable income, maintaining supporting records and filing returns based on audited financial statements. The Commissioner then retains the power to review returns, conduct audits and issue assessments where necessary.

Under this framework, the taxpayer’s return is always the starting point, and where any discrepancies arise after the filing process, each party can present his case.

The current validation mechanism appears to blur this distinction as it shifts the system towards what may be described as a dual assessment: one based on taxpayer records, and another imposed through system validation.

This raises a critical question: if a taxpayer’s records are accurate but the system’s data is not, which should prevail at the point of filing?

If indeed the intention is to shift toward a dual-assessment regime, sufficient thought ought to be given to its enabling frameworks, inclusive of the introduction of statutory provisions to support such a transition. Further still, considering the intention to roll out a new and improved tax compliance system to replace the current iTax, it is crucial that the challenges currently faced with the digitisation of tax filings are taken as a learning point and the process improved in the new system.

These are not merely technical concerns. They speak to the coherence of the tax system itself. Technology should support self-assessment by improving accuracy and efficiency, rather than becoming the authority that determines whether a taxpayer is permitted to file a return based on their own records.

As Kenya continues to modernise its tax administration, the challenge will be to ensure that digital systems strengthen the principles of self-assessment rather than gradually replacing them.

In the wake of the 30^th June income tax filing deadline, it remains to be seen how the compounded effect of these challenges will impact tax compliance; whether taxpayers will choose to meet the deadline by any means necessary, including incurring even further cost or choose instead to abandon the process and embrace the ensuing penalties and interest.

Netty Nyong’a is a Consultant, in the Tax & Legal team at Deloitte Kenya. The views presented are her own and do not necessarily represent those of Deloitte. She can be reached at nnyonga@deloitte.co.ke and here LinkedIn.