Do it yourself

Recently, NITI Aayog published a working paper titled “Towards India’s Tax Transformation: Decriminalization and Trust-Based Governance”. The paper is a follow up to the recent enactment of the Income Tax At 2025. It is an extremely important step for making the tax governance structure trust based.

The paper seeks to provide a comprehensive and critical analysis of the criminal provisions within the Income-tax Act, 2025, mapping the present extent of criminalization, documenting omissions and modifications, and recommending a trust-based regulatory transition for India’s direct tax regime.

Recognizing the evolving policy landscape which stresses ease of business, citizen-centricity, and the need to move away from “fear-based” enforcement, the paper evaluates each criminal provision through a principled criminal law-making framework rooted in jurisprudence, comparative regulatory best practices, and provides recommendations. Its central premise is that decriminalization, rationalization of punishments, and emphasis on proportionate sanctions will collectively align India’s income-tax law with the vision of a fair, accessible, and modern compliance regime.

Summary of the working paper

·         The paper argues that India’s direct tax regime historically over-criminalized even minor procedural defaults. The Income Tax Act 2025 reduces criminalization but still criminalizes 35 acts across 13 provisions.

·         Applying a principle based criminal law creation framework (harm, necessity, proportionality, intent, clarity), it proposes further decriminalization, removal of mandatory jail terms, restoring burden of proof to prosecution, clearer drafting, and periodic review.

·         It recommends prosecution should be reserved only for willful / fraudulent / significant harm behaviors. All else must be a civil penalty domain.

Strong points

The paper correctly identifies the problem of reverse burden of proof in ITA 2025.

Appropriately flags the disproportionate penalty design relative to the BNS

Correct categorization: obstruction / evasion / payment failure / public servant breach

For the first time a government working paper explicitly shifts “normative anchor” from “enforcement” to “trust”. To this extent there is a paradigm shift.

Gaps

Economic modelling absent: The paper doesn’t quantify fiscal impact of decriminalization vs retaining criminal penalty for 6 core offences. This would be a critical input for developing a policy consensus amongst various organs of the policy ecosystem.

Operational tradeoffs not analyzed: CBDT has historically used criminal threat to solve compliance capacity problem — paper assumes that trust-based regime reduces admin cost but does not model enforcement substitution cost.

Digital rights section inadequate: The section on access to passwords / virtual digital space rightly flags constitutional risk, but it does not propose a constitutional safe harbour architecture. Merely stating risk is not sufficient policy, in my view.

No sunset architecture: The Paper proposes periodic legislative review but does not create an actually enforceable review mechanism (eg automatic expiry unless reviewed).

No tiering by taxpayer class: The compliance behavior of various entities - MSME, gig, self-employed, salaried - is structurally different. A single criminalization framework may not be optimal.

No distinction between pre-assessment vs post-assessment domain: Anecdotally, most criminal threats in India are misused in pre-assessment stage. That is where maximum fear is. Paper doesn’t solve this location of harm.

Suggestions

Policy Design Enhancements

·         Create a mandatory 3 tier compliance framework

Tier 1 → civil / monetary

Tier 2 → administrative sanctions (suspension of benefits, loss of fast track processing eligibility)

Tier 3 → criminal only for fraud / fabrication / concealment > threshold

·         Make fraud the only anchor for criminal prosecution — remove “willful” / “knowledge” from the criminal domain entirely.

·         Introduce statutory requirement of “loss to revenue proven / or reasonably demonstrable” as condition precedent for criminal prosecution.

·         Introduce decriminalization sunset — every 5 years offences auto lapse unless reconfirmed.

·         Create statutory economic impact statements for introduction or retention of each criminal offence (OECD style).

·         Shift burden of proof back to prosecution & codify it explicitly.

Digital Rights

·         Separate digital compelled access into national security statute not tax statute - create clear bright line: no compelled password disclosure for tax.

·         Any digital compelled access must require judicial pre-authorisation (not officer discretion).

Conclusion

This is one of the most important tax policy drafting direction papers in the last 20 years. However, there is a risk that this remains normative but NOT enforceable reform. If NITI fixes some missing parts and layers, especially in fiscal impact modelling and behavioral economics, this could become an illustrative reform for most emerging market governments.