What Would Actually Raise India’s CSR Ticket Size

Six weeks of filings, one stubborn number: about ₹35 lakh a project, for a decade.

Part 1: spend grew ~3×; tickets did not. Part 2: concentrators vs sprayers. Part 3: sector league. Part 4: high-spend states ≠ high-ticket states. Part 5: the typical year is ₹23 lakh total. Part 6: education and health are the volume machines.

Part 1 → Why India’s CSR Cheque Size Won’t Budge · Part 2 → Concentrators vs Sprayers · Part 3 → The CSR Sector Ticket League · Part 4 → The Geography of the Cheque · Part 5 → The ₹12 Lakh Company · Part 6 → Education and Health Are CSR’s Volume Machines

This week is not another diagnosis.

It is the physics.

If you wanted the national average to rise — as a board metric, a policy curiosity, or a provocation — what would actually move it? And what would look busy while leaving ₹35 lakh exactly where it is?

There is only one equation

Average ticket = total amount spent ÷ number of project rows.

So there are only three ways to raise it:

1. Same rows, more spend (intensive growth)
2. Same spend, fewer rows (consolidation)
3. Shift mix toward headings and maps that already run fatter tickets

India did the opposite of (1) and (2) at once: spend up, rows up in lockstep. (3) went the wrong way: education and health gained share while thinning.

Everything else is commentary.

First: who cannot lift a national mean

The ₹12 lakh company cannot. Part 5: median year is ~₹23 lakh out the door. You cannot write a ₹35 lakh average project from a ₹23 lakh budget. Lecturing 20,000 such filers is a category error.

The ₹25 crore+ giants, alone, mostly cannot either — not the way people think. Company-years that spend ₹25 crore+ hold about half of all CSR rupees — and only about 7% of all project rows. Their average row is already about ₹2.7 crore.

They are the weighted mean. They are not where the denominator lives.

The national average is a fight between a money tail and a row factory. If you only squeeze the tail, the factory keeps printing lines.

Five things that would move the number

These hold spend roughly constant unless noted. They are counterfactuals, not moral advice.

1. Stop fragmenting (the biggest lever in the file)

In FY 2014-15, companies filed about 1.7 project rows each. In FY 2023-24, about 3.7.

Same companies-count order of magnitude. More slices.

If FY 2023-24 had kept 1.7 rows per company with the same total spend, the average project would have been about ₹75 lakh — not ₹35 lakh.

Even a milder freeze at 2.5 rows per company → about ₹51 lakh.

That is not a new CSR law. That is not adding a line every time the budget ticks up.

Caveat, honestly: early years also had messier zeros and more one-line filings. 1.7 is not a golden age. It is still the cleanest proof in the series that the average died of addition.

2. Consolidate the movable middle (₹1–25 crore years)

This band has enough money to choose geometry and enough rows to matter.

If those company-years had capped at five project rows (same spend, rest of India unchanged), the all-years national average would rise from ~₹35 lakh to ~₹42 lakh.

That is the Part 2 sprayer, in arithmetic. Not the SME. Not necessarily the PSU giant. The filer who can afford a programme — and files a portfolio of small lines instead.

3. Fatten unit size inside education and health

Part 3 / Part 6: give only those two headings the average ticket of vocational skills (~₹58 lakh instead of ~₹38 lakh), keeping their row counts.

National average: ~₹35 lakh → ~₹45 lakh.

No other pair of headings has that leverage. They hold half the pie. If the volume machines keep minting ₹8–25 lakh school and camp rows, the country cannot average ₹50 lakh.

4. Change tickets on the volume map (not only the volume)

Part 4: if Maharashtra’s enormous row count had Odisha’s average ticket, the national mean would move ~₹35 lakh → ~₹42 lakh.

Localisation without larger units does the reverse. If Pan-India rows had been filed at Maharashtra’s ticket, the national average would fall toward ~₹29 lakh. Pan-India is currently propping the mean up.

5. Stop the mix from thinning

Hold recent years’ row mix by sector, but restore early-year ticket sizes inside each heading.

The late-period average would sit near ₹41 lakh instead of the actual ~₹34 lakh.

The decade did not only add education rows. The same labels started writing smaller cheques. Rural development is the exhibit: once a fat-ticket engine, now closer to the mean. Reversing that slide is a real lever. Adding more education share is not.

Five things that would not (or would barely)

Spend more at national scale. We already did. Rows came with the rupees. Totals are not tickets.

Bring more companies under Section 135. New filers look like Part 5. They add headcount and modest rows. They dilute, they do not fatten.

“We work in education / health.” That is how you become the average (Part 6), not how you lift it — unless the row size changes.

Send every rupee to a poorer district, without changing unit size. Need and ticket size are different maps (the equity debate is real; it is not this equation). A thousand ₹8 lakh Bihar rows will not raise a national mean.

Tell only the top 0.5% to “think bigger.” Many already do. Flooring ₹25 crore+ years at a ₹2 crore average ticket barely moves the all-India number — they are not the denominator.

Declare a mega-programme without killing the micro-rows. A ₹100 crore flagship next to eighty ₹5 lakh lines is a barbell (Part 2). The mean flatters you. The operating model is still a spray.

Multi-year is a measurement trick — and a real one

The portal’s unit is a filing row, often by year, state, partner.

A ₹6 crore, three-year school can appear as three ₹2 crore rows — or as thirty district-year slices. Ground spend unchanged. Measured ticket crushed.

We cannot perfectly re-stitch programmes in this file. Treat this as interpretation with teeth:

If boards reported programmes (multi-year, multi-district, one object) instead of annual fragments, the published average would rise even if not a single extra rupee reached a student.

That would be a better thermometer. It might also be a better operating habit: fewer contracts, longer commitments, one owner.

It would not, by itself, fix Part 5’s ₹23 lakh year. It would stop pretending that fragmentation is “more impact.”

Zeros are part of the ₹35 lakh story

About 15% of project rows sit in company-years with zero recorded spend. They inflate the denominator.

Count only positive-spend rows and the all-years average is about ₹41 lakh, not ₹35 lakh.

That does not make the series vanish — the median is still ~₹11–12 lakh, education/health still pin the middle, fragmentation still rose. It does mean some of “₹35 lakh” is file hygiene, not philosophy.

If you want a cleaner public metric: average among rows with amount spent > 0, plus median company-year ticket. Publish both. The mean without the median is how giants hypnotise a country.

So what is the shortest path, if the goal were the mean?

In order of leverage, from this file:

1. Do not add a project row when you add budget — especially in the ₹1–25 crore band.
2. Write larger units inside education and health, where the rows already are.
3. Keep (or grow) a few genuinely fat national / plant-district programmes — the Pan-India and Odisha-shaped lines that currently hold the mean up.
4. File programmes, not confetti — multi-year, fewer geographies per object.
5. Leave the median company alone on ticket size; ask them for one decent line, not a portfolio.

None of this requires a new Schedule VII. It requires treating row count as a design variable, not as a trophy.

What this is not saying

• Not that the purpose of CSR is to maximise average ticket.
• Not that small projects are waste.
• Not that Odisha-style or TCS-style portfolios fit every firm.

It is saying:

If your strategy for “deeper CSR” is more spend, more states, more education lines, and more partners — the filings say you have described the last decade. The average did not move. The equation did not forgive you.

Series note

Part 7 (this): What would actually raise the average.

Next — Part 8: The CSR Ticket Size Playbook — what CSR heads, boards, and implementing organisations can do on Monday without pretending they are the national mean.

If you run a ₹5–25 crore CSR year: you are sitting on the movable middle. The country’s ₹35 lakh is not a mystery. It is your row count, multiplied by everyone like you.

Data: National CSR Portal company-wise filings, FY 2014-15 to FY 2024-25. Counterfactuals keep total spent constant unless stated. Averages = amount spent ÷ project rows. Latest year may be incomplete due to filing lag. Also on LinkedIn.