At Yahoo, I write on the WPI with India’s Inflation Kool-Aid:

India’s economic growth is measured in terms of Gross Domestic Product (GDP) as reduced by Inflation. So we grew — according to the recent CSO release — a whopping 20% on GDP last year (2010-11), and because inflation was, well, pretty high, we grew a only a “real” 8.6% in 2011. How can you grow 20%  but only really grow 8.6%? Inflation eats up the rest — and in a way “deflates” the GDP.

This year, the GDP deflator is at 11%+. We have had only six higher figures in the past since 1961, and this year is the highest since 1991 (13.73%). But can you really rely on it?

The GDP deflator is calculated in a complex way, but according to Deepak Mohanty, Executive Director at RBI, much of that is based on inflation of the WPI — or the Wholesale Price Index. This is the measure of Wholesale prices of commodities and manufactured goods across the country, released monthly.

There are multiple price indexes in India: The WPI and four different indexes of Consumer Price Inflation (CPI). This is confusing, but we’re sure it all has a purpose — just one of them, the CPI (Industrial Workers) is used to calculate “dearness” allowance, a cost-of-living increase in wages.

But let me explore the WPI. Remember, the WPI is what is mostly used to determine our GDP growth in real terms; our politicians and economists refer only to 8% or 9%, so real growth is important. Therefore, it should be just as important to understand how the deflator, and therefore the WPI, works. And more interestingly, how it doesn’t work.

How it works: First, they have a bunch of items they categorize into a hierarchical list and give a weight to each item depending on how important it is to the overall basket. You have the outer broad categories of “primary articles” (mostly food, but also raw materials like iron and copper), Fuel and Manufacturing goods. These are weighted around 20%, 15% and 65% respectively. Within each of them you have subcategories, all the way till your actual item. For instance, Rice is found like so:

Primary Articles > Food Articles > Food Grains > Cereals > Rice (Wt. 1.79%)

Presumably, prices are collected every month and updated and instead of mentioning a price, they relate it to a base year. Current WPI statistics are linked to prices of 2004-05, which is given a price of 100, and a today’s index price of 200 means the item has doubled in price since 04-05.

About 700 item prices are published monthly by the Office of the Economic Adviser. With this granularity some defects become apparent; much of the data seems to not be collected. Take the example of Crude Petroleum. It’s 0.9% of the index, which is huge, for a single item — if you add the weights of Onions, Potatoes and Chicken, you get only 0.8%.

Crude petroleum prices haven’t changed, according to the WPI release, since the last four months. One doesn’t need to be an expert to know something’s off here — international prices went up nearly 20% in the same time.

Prices of Coke — not the cold drink, the fuel — haven’t been updated since April 2008. 36 months of no-change. Even coking coal, which was updated in March 2011, saw 16 months of being at the same level. Raw wool prices show the same since October 2009, and coffee powder since May 2010 (Coffee’s been making new highs worldwide, just for the record). This doesn’t make sense – either data wasn’t being updated or prices really didn’t change. I wouldn’t bet on the latter.

More than 80 items haven’t been updated for over four months, and these add up to 10% of the index. Some of these items don’t change regularly — like prices of electricity for instance. But, like we saw above, many items have just been overlooked. If the idea is that they don’t update prices for a very long time, and then do a single “bunched” update that will perhaps add up all the intermediate unrecorded price increases, we’ll see the inflation index take sudden large jumps. The index is supposed to be for transparency, and we lose all such benefits from “bunched” updates. But the problem goes deeper.

Some items don’t even belong in an index. Litchi prices for instance, although a very small weight (0.03% of the index) are available just one month of the year. I’m not sure a product with such a skewed supply should even figure in an index that determines, in the end, economic policy. (For the record, I love litchis and ice-cream.)

Yet others aren’t even represented. For a country that’s going ga-ga on mobile telephony, the total weight of mobile telephones in the index is… zero. (There is a category called TV Sets and VCD players — but not one for mobiles.) One could say mobile prices have come down and that’s a good thing, but all we know is that they’re not even in the index.

And then there are revisions. Every few months, the past indexes are revised based on data that is supposedly better. Such revisions are supposed to be minor, small variations from the first number. But not anymore. The January 2011 inflation number was revised from the first reported 8.23% up to 9.35% – a whole percentage point higher. This could just mean our March 2011 number — 8.98% – might actually be revised to double digits.

Finally, the Wholesale Price Index is not really a consumer basket — it contains no references to rent or housing, use of services like internet access or cable TV, or wages, all of which are substantial parts of the consumer’s spending. Yet, the WPI doesn’t even recognize any of that. In effect, we “guess” the inflation on the service sector by estimating the price increases through WPI — that is like guessing the price of a house through the increase in steel and concrete prices. To anyone who has not been living under a rock, it is obvious real estate prices have no correlation with anything; the equation is like this:

Steel prices go UP + Concrete prices go UP = House prices Go UP.

Steel goes DOWN + Concrete goes UP = House prices Go UP.

Steel goes DOWN + Concrete goes DOWN = House prices Go UP.

Someone dumps tons of steel and concrete for FREE = House prices Go UP.

The inflation in housing prices, captured by the simple formula above, does not reflect in our WPI.

(For the record, house prices can only go up so far. The fact that many of us haven’t seen a sustained price fall doesn’t negate its possibility)

For services, there is a new index called the CPI (Urban) which will, next year, replace the other CPIs and perhaps the WPI as the indicator of choice. But in its preliminary releases, it has demonstrated that it has similar degrees of blindness by declaring that the last nine months of 2010 saw no increase in housing prices in India. I admire how they can say this with a straight face.

WPI and perhaps even the CPI (Urban) when it comes, will therefore misrepresent inflation, and probably do so quite substantially — how much, we’ll never know unless they actually update real data. But one thing is certain — you can pretend only so long by not revealing the real data, but a price rise will eventually flood the entire economy. Even if we don’t update crude prices in the WPI, it is only a matter of time before plastics and textiles become expensive, and that spread to everything else. This is not something that I predict will happen — it is already happening today.

Underreporting inflation is a useful thing when you have only that much GDP growth to work with — a lower inflation number creates a higher “real” number to be proud of. The US does it all the time. But to drink our own kool-aid and therefore turn a blind eye to growing inflation is altogether more dangerous — we create a monster that we can’t control.

This is the time for the central banker to act — to recognize that inflation, as reported, is probably below actual numbers. It’s also time for the good people collecting WPI data to be more aggressive in keeping prices up-to-date so we get a more accurate picture. It would be a shame if we tried to “pretend-and-extend” our overheating economy.

More Yahoo Columns:

Now, tell them about it: