Tech flaws in I-T system target honest taxpayers, miss real evaders

My friend Milind called in a panic after receiving a notice from the Income-Tax (I-T) Department. It asked him to explain the source of a fixed deposit worth a couple of crore, made in July 2021, as his declared income for 2021-22 was only in single-digit lakh. This implied the deposit came from unaccounted income.
 

Another friend Subhas received a notice for spending Rs 25 lakh on a foreign tour in 2023, when his declared income for the year was around Rs 10 lakh. 
 

A third friend Nitya was also asked to explain the source of her Rs 50 lakh investment in a startup when her declared income for the year was Rs 30 lakh.
 

All of them had valid explanations. Milind had sold shares in 2020-21 and deposited the proceeds into a capital gains account of a nationalised bank in 2021-22 to avail of capital gains tax exemption. Subhas, a retired executive, had used his hard-earned savings from his 35-year career to fund his well-deserved foreign trips. Nitya had cashed in her Employee Stock Ownership Plan (ESOP) from a startup a year ago and had invested a part of it in other startups.
 

Now, all three must face the faceless tax regime to prove that their investments and spending were from accounted sources. All this happened because the much-vaunted AI-driven Data Analytics Package (DAP) makes a simplistic assumption—that spending or investment for a year can happen only from that year’s income.
 

The DAP does not consider past I-T returns to match a taxpayer’s resources with their expenditures or investments. The term used is always “disproportionate to known sources of income” and never “disproportionate to known net worth”. It is what I call the stock and flow problem—the belief is that water can be drawn only from the rainfall for the year, ignoring the dam upstream. The system wrongly assumes that funds are drawn only from current income and ignores accumulated wealth.
 

This inability to distinguish between stock and flow is all-pervasive across the government system. For instance, the Financial Intelligence Unit–India (FIU-IND) requires regulated entities to report suspicious transactions. It has fined stockbrokers for not reporting suspicious transactions—clients whose transactions exceeded their declared income. Many traders, post-Covid, trade frequently and have transaction values far higher than their declared income. The database of suspicious transactions reported by stockbrokers must be overflowing by now.
 

One may argue that shortlisting taxpayers based on such simplistic criteria for scrutiny is harmless. They can provide the same explanation they gave me and eventually prove their innocence. However, being treated as a tax evader shakes honest taxpayers’ trust, even if they are cleared eventually. The actual evaders often escape unnoticed, while millions of man-days are wasted chasing the innocent.
 

Truth be told, the use of technology was meant to simplify matters for taxpayers. Certain aspects like refunds have indeed been hastened. But the DAP is not yet effective at identifying true tax evaders. The safest place for tax evaders to hide is among a crowd of honest taxpayers. DAP must find the needle (tax evader) in a large haystack. But it is limited by the attitudes of the people who provided the specifications for it. It is currently a blunt tool that shortlists lakhs of honest taxpayers in the same net as the few evaders. 
 

The effort required to isolate the tax evaders from these lakhs of honest taxpayers is immense. It’s about time DAP is designed without the limitations of past attitudes. It can then provide insights that will help find tax evaders and avoid the bad press it gets when it targets honest taxpayers like Milind, Subhas, and Nitya.

The writer heads Fee-Only Investment Advisors LLP, a Sebi-registered investment advisor; X (formerly Twitter): @harshroongta