Forty-seven thousand compliances
India has spent five years deleting rules. The burden that remains is less about the number of rules than about who bears the risk of reading them wrong.
The decriminalisation push is real
The government's regulatory-burden programme has removed or decriminalised tens of thousands of compliance requirements since 2020, and the Jan Vishwas Act converted a large set of minor offences from criminal to civil. This is a genuine and unusually concrete reform. It deserves to be recorded as such.
It also has not changed the experience of a mid-sized manufacturer as much as the headline number suggests, and it is worth being precise about why.
Count is the wrong metric
A compliance burden is not the number of obligations. It is the expected cost of getting one wrong, multiplied by the probability of being found to have got one wrong, plus the fixed cost of the apparatus you maintain to avoid it.
Deleting ten low-risk filings and leaving one criminal-liability provision in place barely moves that product. The apparatus a firm maintains — the compliance officer, the retainer, the consultant — is sized by the worst-case obligation, not by the average one. That is why firms report the burden as unchanged while the official count falls, and both are telling the truth about different quantities.
The second issue is overlap. Central, state and municipal requirements are reformed on separate timetables by separate authorities. A firm operating in three states does not face one regime that got simpler; it faces three that changed independently.
The asymmetry that matters
Under most Indian compliance regimes, ambiguity resolves against the filer. If a provision can be read two ways, the firm bears the risk of the reading the assessing officer later prefers — and bears it with interest and penalty attached, often years later.
This produces a specific and expensive behaviour: firms over-comply, hire defensively, and structure transactions to minimise interpretive exposure rather than to maximise efficiency. The dead-weight loss is invisible in any compliance count, because it consists of the better structure that was never chosen.
Binding advance rulings, delivered quickly and honoured, would address this directly. The mechanism exists; the timelines do not yet make it useful for ordinary commercial decisions.
The data behind this
Drawn from the same series as the rankings pages. If the number moves, this chart moves with it.
Regulatory quality
World Bank (Worldwide Governance Indicators)
View as table
| Period | Regulatory quality |
|---|---|
| 2024 | -0.096 |
| 2023 | -0.127 |
| 2022 | -0.088 |
| 2021 | -0.17 |
| 2020 | -0.149 |
| 2019 | -0.098 |
| 2018 | -0.156 |
| 2017 | -0.216 |
| 2016 | -0.254 |
| 2015 | -0.33 |
| 2014 | -0.381 |
| 2013 | -0.372 |
| 2012 | -0.374 |
| 2011 | -0.34 |
| 2010 | -0.278 |
| 2009 | -0.244 |
| 2008 | -0.266 |
| 2007 | -0.209 |
| 2006 | -0.193 |
| 2005 | -0.177 |
| 2004 | -0.205 |
| 2003 | -0.441 |
| 2002 | -0.39 |
| 2000 | -0.375 |
| 1998 | -0.586 |
| 1996 | -0.639 |
Ease of Doing Business (discontinued)
World Bank
View as table
| Period | Rank |
|---|---|
| 2020 | 63 |
| 2019 | 77 |
| 2018 | 100 |
| 2017 | 130 |
| 2016 | 130 |
| 2015 | 142 |
How to read this
This is an argument, not a measurement. The charts above are sourced and checkable; the reasoning between them is editorial judgement and should be treated as such. Where a claim has no series behind it, it is stated as judgement rather than fact.