CPI, WPI and core

Someone says "inflation is high" and everyone nods. But high on which measure? India publishes more than one inflation number, they routinely disagree, and the one that makes the news is not always the one that moves policy. Before you can read what inflation does to a company — the next few modules — you have to know which thermometer is in whose mouth.

This module is the vocabulary. Three numbers do most of the work: CPI, the price of the shopping basket a household actually buys; WPI, the price of goods further up the chain, at the factory gate and the wholesale market; and core, which is CPI with its two wildest items — food and fuel — taken out. Learn what each one measures, and a lot of confused commentary suddenly sorts itself.

There is more than one inflation number because "the price of things" is not a single thing. The price a farmer gets for onions, the price a steel mill pays for iron ore, the price a wholesaler charges a shop, and the price you pay at the counter are all different prices, moving at different speeds. Each index is a deliberate choice about which of those prices to track.

CPI — the Consumer Price Index — asks a plain question: what does it cost a typical household to buy the same basket of goods and services this month as last? It is released monthly by the Ministry of Statistics (MoSPI, the government's statistics office) around the middle of the following month. Food and beverages carry a large share of India's CPI basket — close to half — which is why a bad vegetable season can swing the whole print. This is the number that matters most to you as a shopper, and, as we will see, the one the central bank is legally told to watch.

WPI — the Wholesale Price Index — tracks prices earlier in the chain, where goods change hands in bulk before reaching the shelf. It is dominated by manufactured goods and primary commodities, and it contains no services at all. That single design fact — heavy on commodities, no services — is why WPI is far more volatile than CPI and can even turn negative while CPI stays firmly positive. When global metal or crude prices tumble, WPI feels it hard and fast; your grocery bill barely notices.

Core inflation is not a separate survey. It is CPI with the two most jumpy categories — food and fuel — stripped out, to reveal the steadier trend underneath. Food swings with the monsoon; fuel swings with crude and taxes. Both can spike and reverse for reasons that have nothing to do with the broad economy overheating. Core is the attempt to hear the signal after turning down that noise: it captures whether price pressure is becoming general and sticky — spreading into rents, services, and everyday manufactured goods — rather than concentrated in a few volatile items.

The cleanest way to hold this is to picture where in the economy each index takes its temperature, and what sits inside its basket. They are measuring different points on the same journey from raw material to your kitchen.

Now the fact that ties the room together: the RBI targets headline CPI. Under the framework India adopted, the central bank is given a legal inflation target — a rule that its job is to keep consumer inflation at a set number, currently 4%, within a tolerance band of 2% to 6%. That target is written in terms of headline CPI, not WPI and not core. WPI is watched as a clue about input costs; core is watched to judge whether pressure is sticky; but the number the RBI is answerable for is the full CPI print.

Why does that distinction matter to a reader? Because it tells you what the central bank is likely to ignore and what it is likely to chase. A food spike from a bad monsoon lifts headline CPI, but if core stays calm the RBI may judge it temporary and sit still. A rise that shows up in core — in rents and services and everyday goods — is far harder to dismiss, because it signals the pressure has become general. So the split between the numbers is not trivia; it is how you anticipate whether a hot print will actually provoke a policy move. That is the whole of the next module on the RBI's reaction, and it starts here.

One more piece of the construction to hold on to. Every one of these indices is measured against a base — the same month a year earlier. If prices jumped sharply last year, then even a normal rise this year can show as a low inflation number, simply because the comparison point was already high. That is the "base effect", and it can flatter or worsen a print for reasons that have nothing to do with what is happening right now. We return to it in Part Ten, but flag it here: the number is a comparison, and the thing it is compared against shapes it.

Take one composite month and read the three numbers together. illustrative

Suppose a month prints headline CPI at 6.2%, core at 4.1%, and WPI at −1.0%. A beginner sees three numbers pointing three ways and concludes the data is a mess. It is not — read in order, it tells a clean story.

Start with the gap between headline and core: 6.2% versus 4.1%. That two-point wedge is food and fuel doing the damage. Something in the volatile basket — say a poor vegetable crop after a weak monsoon — has pushed the shopping bill up, while the steadier core is sitting near the RBI's 4% comfort zone. So the household is genuinely hurting, but the broad economy is not obviously overheating.

Now the negative WPI. Because WPI is stuffed with commodities and manufactured goods, a soft global metals and crude backdrop can drag it below zero even as your groceries climb. For a manufacturer that buys metal and turns it into products, that negative WPI is quietly good news — cheaper inputs. For the household staring at the vegetable stall, it means nothing at all. Same country, same month, two different price worlds.

Put together, this composite month says: the pain is concentrated and possibly temporary (food-led, with calm core), and input costs upstream are actually easing (soft WPI). A central bank reading this is more likely to wait than to slam rates higher — precisely because core, the part that tends to stick, is behaving. Notice what you did: you did not react to the scary headline alone. You read which basket moved and how sticky it is. That is the skill.

An inflation number is a rear-view measurement of an average basket. Hold on to what it cannot do, however precisely it is quoted.

It cannot tell you your inflation. The CPI basket is a national average; your household's spending — more rent, less food, or the reverse — can run quite differently. The famous number is a composite of millions of baskets, and no single family lives inside the average.

It cannot tell you what the RBI will do. The target is headline CPI, but the response is a judgement, not a formula — the bank weighs core, the source of the shock, and where it thinks inflation is heading. A high headline with calm core often produces no move. Reading the print straight into a rate call is the error the next modules exist to prevent. As Galbraith warned, reading next quarter's policy off this month's print is closer to astrology than analysis.

And it cannot tell you what it does not measure. The number is a construction — a chosen basket, fixed weights, a base year — and the choices shape it. That is not a reason to distrust it, but it is a reason to read it as a made object rather than a fact of nature. As Darrell Huff taught a generation, an average is a design, and you should ask what went into it before you lean on it.

The same handful of mix-ups catch readers month after month.

1. Treating WPI as consumer prices. WPI is wholesale and producer-level, commodity-heavy, with no services. A negative WPI does not mean shop prices are falling — it usually means commodities have softened. Read it as an input-cost gauge, not a shopping bill.

2. Reading a high headline as an automatic rate hike. The RBI looks through food and fuel spikes it judges temporary. The number that changes minds is core, because core is the part that stays.

3. Ignoring the base effect. A low print can simply mean last year's month was high. A number is a comparison; always ask what it is being compared with before calling it good or bad.

4. Forgetting food's giant weight. Because food carries such a large share of India's CPI, the whole headline can be hostage to the monsoon and a few vegetables. "Inflation is up" is often "onions and tomatoes are up."

5. Confusing your basket with the basket. The published number is a national average. Your own cost of living can move quite differently, and that is not an error in the data.