Missing price signals cloud GDP data

The gross domestic product (GDP) numbers in India have a problem and it pertains to the physical collection of prices. There are too few people employed by the government to collect data on prices, weekly and monthly. In addition, a United Nations-approved method to reconcile discrepancies that such shortage throws up is still making the rounds within the government.

At 3,249 statisticians, India's data crunchers number a fifth of what China puts on the ground to calculate its GDP. The US, despite deep computer penetration that makes data collection almost automatic, has nearly 4,500 federal employees to collect statistics.

The lack of foot soldiers is to be mainly blamed for India's recent trouble with data. It flared up after India aligned GDP measurement to the international best practice of gross value added and subsequently ran into episodes of deflation in the manufacturing sector. The first set of revised data for 2013-14 came out in January 2015.
 

The most common way to measure GDP is to multiply the production of any commodity with its price netting out the cost of goods that went into its production. In other words, it is the value added elements of all products and services. Since the result is expressed in rupees, it is easy to add up the entire range of commodities and services in a country to arrive at the GDP. The new method has made this process easy to calculate, which is why there is no dispute on the value of the nominal GDP arrived at this way. But when one wants to abstract the effect of price rise to measure real GDP, the lack of ground staff hurts. Especially, when prices fluctuate.

On a visit to a shopping mall this week, you could, say, find the price of shirts has risen by 3.8 per cent in a month, which is what the consumer price inflation chart shows. In the calculation of nominal GDP, this would not matter because the price printed on the shirt would be multiplied with the number of shirts sold.

To arrive at real GDP, one has to take into account the changes in prices of all items that went into manufacturing the shirt, in other words, create what is known as the GDP deflator. Statisticians do this by generally using the consumer price for stuff sold in shops to you and me and using the wholesale price for goods sold by manufacturers to each other.

But when prices fluctuate in the course of the year, as it did in 2014-15, it becomes a problem. Our shirt, for instance, includes at least two items, a polyester staple fibre whose price fell 10.2 per cent from April 2015 to March 2016 and a 1.5 per cent rise in the price of leather in the same period.

The task becomes complicated when the goods involve more parts than a shirt lined with leather. Estimating the price of a car when hot rolled coil prices have declined 27.5 per cent in the last financial year just when rubber prices moved up four per cent in the same timeframe is even more difficult.

How well these prices are recorded in the books of the enumerators determines how correct is the assessment of the real GDP. Price collection methods are still raw in India. One example is the request from the department of industrial policy - which measures the wholesale price index and is now running short of staff - to ask the ministry of statistics to monitor "the regular transmission of weekly price data… in respect of 3,813 units/factories in the organised manufacturing sector".

So, it boils down to how well the deflators have picked up all the price signals. If they do not, the statisticians have a problem on their hands.

For instance, the GDP can also be calculated by another method, known as the expenditure method, where statisticians sum up everyone's consumption expenditure and add to it investment, government expenditure plus net exports. In the absence of accurate measures of price, the GDP as estimated by this method would often under estimate the one derived via the income or production method.

The extent of this divergence is what is put down as discrepancies. In a year when prices have moved erratically making some prices shoot up faster while others move sluggishly, like those of steel and rubber, these discrepancies have become considerable.

India is not the first economy to face this problem. Others with mature statistical systems do too, and their way out it is to use a cheat sheet. Unlike its name, it functions like an input-output table, where if the price of an item is used on, say, the production side of the GDP equation, it is matched with the same item on the expenditure side. It means the discrepancies will be sorted out in the building of the data instead of being lumped at the end.

Work on this cheatsheet known as physical supply-use table has progressed quite a bit since last year when an interministerial group in March 2015 approved work on it subject to "availability of relevant statistics with ministries concerned and other data sources". The first such table has been presented to the advisory committee on national accounts already. It is now a judgement call if the statistics for 2016-17 will begin to use it.