Times of Better Bharat

The strengths first

India's statistical apparatus is, by the standards of countries at its income level, substantial. The National Sample Survey has run for seven decades. The Census, when conducted, is among the largest administrative exercises on earth. UPI, GST and the CPI produce high-frequency data that many richer countries lack.

The problems are specific rather than general, and naming them precisely is more useful than a blanket complaint.

Reported is not the same as occurred

Several of India's most important social series measure reporting rather than incidence, and are routinely read as if they measured incidence.

Crime against women is the clearest case. The recorded rate rose from 56.3 per lakh in 2014 to 66.4 in 2022. That rise is consistent with more crime; it is equally consistent with more reporting, following a decade of legal reform, awareness campaigns and improved registration. It is most likely both. No published series separates them, and any confident claim in either direction is going beyond the data.

This is not an argument for ignoring the number. It is an argument for stating what it is. A publication that reports the rise as 'crime against women increased 18%' is making a claim the source does not support, and one that ignores it entirely is worse.

The gaps that cost most

Three absences do identifiable damage.

There is no national crash-level road safety database. India records more road deaths than any country on earth, and manages that toll without the record-level data that would let anyone identify which stretches of road and which vehicle interactions produce it.

There is no clean, current national series for commercial dispute resolution time — the variable behind the enforcement problem described elsewhere on this site.

And the Census, whose decennial cycle underpins the denominator of nearly every rate published here, has been delayed well past its schedule. Every 'per lakh' figure on this site divides by a projection, and the longer the delay runs the more that projection carries.

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.