In recent years SARS has been investing in an automated and data driven compliance programme whereby it is increasing the use of data, data analytics and artificial intelligence to understand taxpayer behaviour and to inform case selection for audit and verification purposes. The idea is for risk detection, assessment and profiling to be largely automated for standard matters, whilst manual risk profiling and case selection are reserved for complex matters.

Automated risk-profiling

SARS’ latest Strategic Plan describes “compliance” as the degree to which taxpayers fulfil their tax obligations, such as registration, filing/declaration and payment, as and when required by law. Voluntary compliance therefore requires taxpayers to accurately and timeously attend to their tax obligations without being prompted by SARS to do so.

In terms of SARS’ fifth strategic objective, its expanded and increased use of data is to substantively support its strategic intent of voluntary compliance. Therefore, SARS’ unwavering message has been to make non-compliance with tax laws hard and costly and to create an environment where taxpayer compliance is consistently monitored.

But risk-profiling (albeit automated) is a sensitive topic and categorising a particular taxpayer as high(er)-risk can ruffle feathers no matter how well-intended the approach may be. It is therefore not surprising that SARS’ reliance on its risk-engine has been met with some apprehension by taxpayers. The main issue is a perception that the methodology and parameters for automated risk detection are over-engineered and highly sensitive, thus giving rise to repeat verifications of the same VAT vendors for months on end.

Therefore, unless a taxpayer is able to build up a good compliance history to avoid constant scrutiny, the use of risk-engines may unintentionally lead to non-compliance.

Retrospective detection

VAT is a self-assessment tax, which means the taxpayer (vendor) bears the obligation to calculate its VAT liability or refund for each tax period and to declare the results in its VAT returns which it files with SARS. The only way in which to detect and enforce compliance is to do so on a retrospective basis. VAT is also transaction-based and the VAT treatment of a transaction is usually determined only once, at the inception of an agreement. Get it wrong and the error could affect multiple VAT returns before it is detected and remedied.

When risk detection and profiling are automated and based on historic data, an errant taxpayer can easily be labelled as a repeat offender and may therefore be placed in a higher risk category even though there was in fact only one instance of non-compliance. Based on SARS’ compliance philosophy as set out in its most recent Annual Performance Plan, a habitual non-compliant taxpayer can expect increased enforcement action to nudge it in the right direction.

Why trust matters

In terms of SARS’ Annual Performance Plan it believes that most taxpayers fall between two extremes where the majority of taxpayers will always strive to do the right thing whilst a certain number will always fall short. SARS therefore seeks to move as many taxpayers as possible up the compliance continuum.

What is required is for taxpayers to buy into SARS’ risk-based approach to compliance and data-driven enforcement. For tax practitioners, it is to facilitate robust debates about interpretational issues and to leverage off strong dispute resolution mechanisms. For taxpayers, it is to have certainty around their tax positions.