Every January and July, the same conversation happens in every government office I've ever sat in. Someone says DA is going up 3%. Someone else swears it's 4. A third person forwards a WhatsApp screenshot that's usually wrong. And nobody can actually prove their number, because none of them have sat down and done the arithmetic.
The thing is, you can. Dearness Allowance isn't a decision somebody makes in a room. It's a calculation on published data — and once you build yourself a proper DA calculation sheet, you stop guessing and start knowing. I've kept one in a spreadsheet for years, and it has never once been surprised by the official figure.
Let me show you exactly how to build it, what goes in each column, and how to read it so you can work out the next hike months before the order comes.
What a DA calculation sheet actually is
Strip away the jargon and a DA calculation sheet is just a small worksheet with one row per month. Each row holds the AICPI-IW index for that month — the All-India Consumer Price Index for Industrial Workers, published by the Labour Bureau. From those monthly numbers you compute a 12-month average, run it through a fixed formula, and out pops the DA percentage.
That's the entire machine. Twelve numbers in, one percentage out. The whole reason it feels mysterious is that most people never see the twelve numbers lined up in front of them.
The formula, in plain terms
Here's the formula the Department of Expenditure uses on the current 2016 = 100 index base:
DA% = floor( ( 12-month average AICPI-IW × 2.88 − 261.42 ) ÷ 261.42 × 100 )
Two numbers in there need explaining, because they trip everyone up.
The 2.88 is a linking factor. The index was rebased from 2001 = 100 to 2016 = 100 back in 2020, and the 7th CPC formula was originally written on the old series. Multiplying by 2.88 bridges the new numbers back to the old scale so the formula still works.
The 261.42 is the AICPI-IW value as it stood on 1 January 2016, when the 7th CPC reset DA to zero. Everything is measured as a rise above that base.
And floor just means you round down to the nearest whole number. DA is always a whole percent, and the fraction is dropped, not rounded up. Miss that and your sheet will read one percent too high half the time.
If you'd rather skip the arithmetic entirely, our current DA rate page and Expected DA calculator run this exact formula on live data — but building the sheet yourself is the best way to actually understand what those tools are doing.
Building the sheet in Excel, step by step
You need four columns. That's it.
Column A — Month. One row per month, oldest at the bottom or top, whichever you prefer. Just be consistent.
Column B — AICPI-IW. The index for that month, taken from the Labour Bureau's monthly press release. This is the only data you have to source; everything else is formulas. You can pull the verified monthly numbers from our AICPIN archive, which carries every release.
Column C — 12-month average. In the row for the latest month, average the last twelve entries in column B. If your months run newest at the top in cells B2 to B13, the formula is =AVERAGE(B2:B13). Drag it down and each row shows the rolling average as of that month.
Column D — DA%. Apply the formula to column C. In Excel: =FLOOR((C2*2.88-261.42)/261.42*100, 1). The FLOOR(...,1) is what rounds down to the whole number. Copy it down alongside column C.
That's the working sheet. Add a fifth column if you like — your own basic pay times the DA% divided by 100 — and now every row also tells you the rupee value of DA in that month. I keep that column, because seeing the actual money makes the whole thing feel real.
A worked reference table
Here's what a stretch of the sheet looks like with real recent figures, so you can sanity-check your own. The average column is the rolling twelve months ending in that row's month.
| Month | AICPI-IW | 12-mo average | Implied DA |
|---|---|---|---|
| Jul 2025 | 146.5 | ~145.4 | ~55% |
| Sep 2025 | 147.3 | ~145.8 | ~56% |
| Nov 2025 | 148.2 | ~146.4 | ~57% |
| Jan 2026 | 148.6 | ~147.1 | ~58% |
| Mar 2026 | 149.1 | ~147.6 | ~59% |
| May 2026 | 150.8 | ~148.1 | ~60%+ |
Read down that "implied DA" column and you can watch the next revision building, month by month, long before it's announced. That's the whole point of keeping the sheet.
A full worked example, start to finish
Let me run one revision all the way through, so you can see there's no magic in it.
Say you're working out the DA effective 1 July 2026. The window is the twelve AICPI-IW readings from July 2025 to June 2026. Suppose those twelve months add up to a total of about 1,777. Divide by twelve and your average is roughly 148.1.
Now feed that into the formula. Multiply 148.1 by 2.88 and you get about 426.5. Subtract the base of 261.42, which leaves about 165.1. Divide that by 261.42 and multiply by 100, and you're at roughly 63.1. Floor it — drop the fraction — and your DA lands at 63%.
That's it. No committee, no judgement call, no rounding up. Twelve published numbers, one average, one formula, one floor. When the Department of Expenditure issues the order two or three months later, it will say 63% — because it did exactly the same sum you just did. The only thing you had that they didn't is a head start of a few months.
If your own total for the twelve months is a little different from mine, that's fine — you might be using slightly revised index figures, and the Labour Bureau does occasionally revise a month. That's precisely why keeping your own sheet, sourced from the verified releases in our AICPIN archive, beats trusting a forwarded number.
Which twelve months actually count
This is where people's sheets go wrong even when the formula is right.
DA is revised twice a year — effective 1 January and 1 July. Each revision uses the twelve-month average ending the month before it takes effect. So the 1 January revision is built on the average of the previous January to December, and the 1 July revision on the previous July to June.
If you're trying to predict the July 2026 DA, the window you care about is July 2025 through June 2026 — not the latest rolling twelve, and not the calendar year. Line up the wrong twelve months and your prediction drifts. Getting the window right is half the skill.
Why the hike always arrives late
Once you're keeping the sheet, one thing becomes obvious: you'll know the number before the government announces it. That's not a loophole — it's built into the calendar.
DA takes effect on 1 January or 1 July, but the final AICPI-IW reading for the window is only published at the end of the following month, and the Department of Expenditure then processes the order. So the January DA is typically notified around March, and the July DA around September. The gap gets paid as arrears in a single tranche.
Your sheet closes that gap. By the time ten or eleven of the twelve months are published, the answer is all but locked — a single remaining month can only nudge it by about one percentage point. That's why a well-kept DA calculation sheet is never really surprised.
The same sheet works for pensioners
One thing worth adding if you've retired, or you're helping a parent who has: the exact same sheet gives you Dearness Relief (DR).
DR is simply DA under a different name for pensioners. Same percentage, same revision dates, same formula — the only change is that you apply the percentage to basic pension instead of basic pay. So the DA% column in your sheet is your DR%. If your sheet says the July revision is 63%, then a pensioner's DR is also 63%, applied to their basic pension (on the un-commuted portion). You don't need a second sheet; the column you already built does both jobs.
The mistakes I see in people's sheets
Four errors account for almost every wrong DA prediction I've come across:
- Using the wrong index. It has to be AICPI-IW (Industrial Workers), not CPI-Urban, CPI-Rural or the combined CPI. Different index, different number, useless for DA.
- Forgetting to floor. Rounding to the nearest percent instead of rounding down inflates the answer. DA drops the fraction.
- Averaging the wrong window. Using the latest twelve months when the revision needs a specific January–December or July–June window.
- Old base confusion. Pulling an old 2001 = 100 figure into a sheet built on 2016 = 100, or forgetting the 2.88 linking factor entirely.
Fix those four and your sheet will match the official order every single time — because it's doing the identical calculation.
Do you even need the spreadsheet?
Honestly, for most people, no — the live tools do it for you. Our Expected DA calculator projects the next revision and shows the rupee impact on your basic pay, and the AICPIN archive keeps every monthly reading with the running average already computed.
But I still recommend building the sheet once, by hand. Not because you'll use it forever, but because after you've done it, you'll never again believe a wrong DA number someone forwards you. You'll just glance at the last few AICPI-IW readings, do the arithmetic in your head, and know.