India's fiscal year ends 31 March and new budgets open in April, so awards ramp up through the year and reset each spring. We used to headline this as the "March rush" - then we stress-tested it, and it does not earn red-flag status. Here is the pattern, and the test that demoted it.
Share of awards by calendar month. If procurement were even, every bar would sit near 8.3%.
March's share of awards, year by year - a consistent, predictable calendar effect.
A year-end "use it or lose it" scramble should show up as money, not just paperwork - big contracts dumped before the deadline. We tested that on the full deduplicated corpus, and it is not there:
March holds 10.9% of award count but only 10.0% of awarded value - barely above the 8.5% an even calendar would give, and lower than January (12.7%). The Jan–Mar quarter carries 30.2% of the year's value, inside the government's own 33% norm for year-end expenditure.
The median March award is Rs 8.6 lakh - identical to the median award in the rest of the year. What remains is the mechanical rhythm of a fiscal calendar: budgets close in March, new ones open in April, and tenders take weeks to process - a gradient, not a scandal.
One more data point on where the rush actually lives: tender publication timing is close to flat all year (see Buyers & Patterns) - it's award finalization that clusters toward March, not the decision to open a new tender. The scramble is in closing out approvals already in the pipeline, not launching new procurement.
We keep this page because the rhythm is real and worth knowing when reading the other figures - but we present it as a calendar effect, not an integrity signal. Retiring an overreach is part of the methodology: see how we audit ourselves.