SOURCES

Office of the Economic Adviser, Department for Promotion of Industry and Internal Trade, Ministry of Commerce and Industry, wholesale price index, as published on DBIE: the 2011-12 series for every commodity code, monthly from April 2012 to April 2026, with each code’s weight, and the 2004-05, 1993-94 and 1981-82 series, monthly. The Office’s manual on the 2011-12 series (2017), its press release and answers to frequently asked questions on that series (May 2017), its pages of linking factors, its press release of 14 May 2026 for April 2026, and its press release of 15 June 2026 and note to users of 1 June 2026 on the 2022-23 series that replaced it. National Statistics Office, Ministry of Statistics and Programme Implementation, consumer price index (rural, urban, combined), 2012 = 100, all-India combined, by group, monthly from January 2013 to December 2025, as published on DBIE, with the weights from its press release of 12 January 2016. Labour Bureau, Ministry of Labour and Employment, wage rates in rural India, men, all India and by state, by occupation, monthly from November 2013 to April 2025, and its consumer price indices for rural labourers and for agricultural labourers (1986-87 = 100), as published on DBIE, and the Bureau’s brief on the series. Reserve Bank of India, monthly average prices of gold and silver in Mumbai, as published on DBIE. US Bureau of Labor Statistics, consumer price index for all urban consumers: services, commodities, hospital and related services and televisions, January 2000 to September 2025.

FURTHER READING

DOWNLOADS

METHOD

What the index prices. The wholesale price index measures the change in the prices of goods at their first sale in bulk: farm produce at the mandi, minerals at the mine, petroleum products at the refinery and manufactured goods at the factory gate, after discounts and before indirect taxes. It prices no services, and most of its quotes carry no shop’s margin, though milk and LPG are priced in retail markets (the Reserve Bank’s Patel committee, 2014). A good’s index compares the prices of its quotations, 8,331 of them across the 697 goods (the Office of the Economic Adviser’s manual, 2017), with their own prices in 2011-12, as a geometric mean; it is not a price, and the rupee prices are never published. Above the goods, the index is a weighted average, and a good’s weight is its share of the value of goods traded in India in 2011-12: output plus imports, less exports.

The series. The 2011-12 series ran from April 2012 to April 2026. On 15 June 2026 the Office replaced it with a series based on 2022-23, with 957 goods, and said the wholesale price index would be published for five more years beside new producer price indices and then stop (its press release of 15 June 2026). DBIE carries the 2011-12 series, and the page uses all of it. The last two months, March 2026 and April 2026, are the Office’s provisional figures, as DBIE carries them; its final figures for all commodities, published in its later releases, are a few tenths higher (161.2 and 167.6), and individual goods were revised by more, in both directions. On the final March figure one good, hydrogen peroxide, moves to ₹100.2, which would leave 32 cheaper. DBIE also prints the base year itself, April 2011 to March 2012, as 12 rows of exactly 100 for every code; they are placeholders, not prices, and are left out.

Goods, not groups. A good here is a leaf of DBIE’s commodity tree, a code no other code extends, with one correction: DBIE files the textile-machinery sub-group as a good and open-end spinning machinery as a sub-group, and they are swapped back here as the Office’s own file has them. That leaves 697 goods, the 117 primary articles, 16 fuel and power items and 564 manufactured products of the Office’s own count, whose weights add to ₹100. The 29 groups’ weights add to ₹100 too and, at those weights, their indices rebuild the whole basket within 0.061 of an index point in every month.

Years against years. Every comparison of a good with 2011-12 takes the average of 2025-26, April to March, against that base year, which is 100 by definition. The latest month, April 2026, is given where it is used. A single month against a single month would mislead for anything seasonal.

The chart. Each line on the chart is the index averaged over the twelve months to each month, so a good’s line ends at its average over the twelve months to April 2026, close to the 2025-26 average the sentences use, and a month of tomatoes at ten times its base price shows as a rise spread over the year that follows. The captions and the markers on the chart give single months. Every line leaves one origin, the 2011-12 average of ₹100, drawn at the middle of that year. The good’s emoji rides the drawn line, and the vertical scale is linear: a rise of 50 points takes the same height anywhere on the chart.

Seasonal goods. When a fruit or vegetable is out of season the Office compiles no index for it and spreads its weight over the rest of its sub-group; DBIE prints a 0 for the month. Six goods are priced in part of the year: green peas in 6, cauliflower in 8, mango in 5, orange in 8, grapes in 7 and litchi in 2 months of 2025-26. Their averages are over the months they were priced.

Flat series. A price quotation that does not arrive is carried forward unchanged, and some goods rest on one or two quotations, so a long run of identical months can be a price that did not move or a price nobody reported. Eleven goods have three or fewer distinct values since April 2012, or sixty or more months unchanged in a row: the personal computer index did not change for 101 months and the laptop index for 71. Three of the eleven are among the 33 goods that got cheaper: tanning materials, electrical ballast & chokes and modems. By a looser test, half or more of their months unchanged or two years at one value, 66 goods weighing ₹7.0 hardly move; they are 12 of the 33 that got cheaper, against about one in 12 of the rest. They are in the downloads and the receipts, and none of them is named in the sentences above.

No allowance for a better TV. The Office holds each quotation to a fixed specification and, when a product is replaced, splices the new price onto the old, so the gap between two models counts as quality and the index follows each model’s own price. It makes no other adjustment, and none for the goods that change fastest, which its own terms of reference asked for and did not get. Statistical offices that adjust for quality, as the United States does for computers, find larger falls for electronics than this index can show.

Since 1982. The Office links its series the same way every time: an older series is divided by its own average over the next base year. The 2004-05 series averaged 156.1 over 2011-12, the 1993-94 series 187.3 over 2004-05 and the 1981-82 series 247.8 over 1993-94, which are the Office’s published factors of 1.561, 1.873 and 2.478. The three major groups use their own factors; the Office publishes them for the two later joins, which these reproduce, and only the all-commodities factor for the 1981-82 join. Each base has its own basket, specifications and sources, so the long series compares baskets as well as prices, and the Office links nothing below the major groups. DBIE’s monthly table ends the 1981-82 series in December 1999 and begins the 1993-94 series in January 2000; the yearly figures from 1994-95 to 1999-00 come from the 1993-94 series in DBIE’s table of yearly averages, so the join is at 1993-94, where the Office makes it. Joined instead at another year in which both series ran, the basket’s multiple ranges from 10.4 to 11.4, grown goods’ from 15.1 to 17.1, fuel and power’s from 15.0 to 22.1 and factory goods’ from 7.8 to 8.5.

Days of work. A good’s rise between 2014-15 and 2024-25, the first and last full years of the Labour Bureau’s current list of occupations, is divided by the rise in the all-India day wage of a non-agricultural rural labourer, ₹238 to ₹396. DBIE carries men’s wages only; the all-India figure is the Bureau’s average over every village quotation, not weighted by state; and the wage is for a normalised eight-hour day, not earnings. A farm labourer’s day rose from ₹225 to ₹398 and a carpenter’s from ₹343 to ₹569, so the count moves little with the yardstick: 641 against the farm labourer’s day, 619 against the carpenter’s, 614 against the average of the 24 occupations. It moves more with the starting year: from 2015-16, after the oil price fell, it is 585. Averaging the 18 states’ rises equally, instead of every village quotation, gives 68% rather than 67%. The Bureau widened its sample from 600 villages in 20 states to 787 in 34 states and union territories from July 2025, after the last month on DBIE, and warns against comparing across that change. It deflates these wages by its consumer price index for rural labourers, which rose from 802.2 to 1310.9 (1986-87 = 100) over the same years, 63%, and the farm occupations by its index for agricultural labourers, which rose from 799.5 to 1299.1; both comparisons are given beside the goods-only one.

The consumer index. The consumer price index is compared over calendar years, 2013 and 2025, the first and last full years of its 2012 series in DBIE’s monthly table, and the wholesale index over the same years; over the financial years 2013-14 to 2024-25 the general index rose 72%. The two have different baskets, weights and base years, so the comparison is of rises, not levels. The consumer groups mix goods and services: transport and communication includes petrol and diesel, education includes books and stationery, and health includes medicines as well as doctors’ fees. A new consumer series based on 2024 began in January 2026 and joins the old one only at the general index.

Nothing typed by hand. Every number here is read from the DBIE series by a program that checks the tree, the weights, the placeholders and the linking factors every time it runs. The sentences above use those same numbers. The only figures typed in are those cited from other publications, which are named where they appear.