Column: No Proof Required – CSO’s Believe It or Not!

By: | Updated: February 24, 2015 11:29 AM

The miraculous FY14 growth has essentially come from artificially lowering the numbers for FY13

The Indian statistical authorities (Central Statistical Organization, or CSO) recently released a revised set of GDP numbers that have set the statistical house on fire. The new salvos being fired from CSO’s shoulders suggest a radically different interpretation of the Indian growth story over the last few years, and indirectly, on what happened in the national elections in May 2014. Just to recall: in May 2014, the ruling Congress party was decimated and registered its lowest seat tally ever, 44 seats in a 543-seat Parliament. For those keeping records, this was the second-largest decline in parliamentary history in the world, possibly ever. The record holder for the worst thrashing was the Progressive Conservatives party of Canada. Under Martin Brian Mulroney, the Progressive Conservatives were re-elected in 1988, with 169 out of 295 seats. At the subsequent election, in 1993, the party collapsed to win just 2 seats, falling from first to fifth place.

Many of us thought that a major reason for the magnificent loss of the Congress party was the economy, stupid. We were told that GDP growth for two successive years was less than or equal to the 5% mark, 4.7 % in FY13 and 5% in FY14, a toxic decline from the heady years of near-double-digit growth. If GDP growth declines by half, voters are likely to be upset, and we all concluded that the voting reflected the pocketbook. Much ink was spilled, and many trees were cut, to document the close relationship between economic performance and voting behaviour.

Now, with the sensational results reported by the CSO, on the basis of revised and improved GDP, more trees are going to be felled. What the CSO now says is GDP growth in India was 5.1% in FY13 and a miracle-growth rate of 6.9% in FY14. Again, to recall and reiterate, just yesterday all of us thought that FY14 was the near-worst GDP year in recent times, and only marginally better than the retrospective tax year of FY13. With some people now saying, “Wow, Sirji, what a recovery!”, looks like the economy went from paralysis to a surge in activity faster than a BMW reaching 80 mph!

What the data do suggest is a smoke-and-mirrors interpretation of the CSO statistics. Before documenting, some caveats and disclaimers in defence of any revised numbers, put out by the CSO or anyone else. Data and methodological revisions are undertaken by all countries, and developing countries like India do undertake these revisions on a larger and more frequent scale.

New data for manufacturing; change in the base year from FY05 to FY12; segregation of crop and livestock production; estimation of value addition from extraction of sand through an indirect method, in accordance with its use in construction; inclusion of accounts of stock brokers, etc; movement to market prices from factor cost, etc—all these changes are necessary, and welcome. So, why question the surge in GDP growth in FY14?


Because the CSO revisions do not pass a basic smell-test. When data are revised, they are revised for all the years under consideration, in this case FY13 and FY14. New information, base year changes, etc, should affect the data for both the revised years and affect them approximately equally. Different economic conditions in different years will result in differences; however, these differences should be minor. And if not, smell-tests require that a more-than-adequate explanation be provided. And this has not been provided by the CSO.

Let us go down the list. In the two years, GDP growth was 9.7% in the old data, and 12% in the new data. That is a difference of 2.3 percentage points; thus, for each year, the GDP growth estimate should be raised by 1.15 percentage points. This is the gross effect of the new information, different base year, etc. If the new GDP growth rates were 5.9 % and 6.2 % (instead of 4.7 and 5), this would have been a dog-bites-man story and would not have gathered any newsprint, or dust. Instead, the gap in the two years is reported as 1.8 percentage points (5.1% versus. 6.9%).

Now that is a small dog biting a big man Some of the explanations offered to explain this gap (nobody is questioning the revisions) is that there was a surge in productivity in FY14, or new industries began to flourish in FY14, etc. Really? In a policy paralysis year, with bank balance sheets broken, and ease of doing business worse, there was a surge in investment activity? Old data: capital-formation growth declined from 0.8 % to minus 0.1%; new data: capital-formation growth BMW-accelerated from minus 0.3%to, hold your breath, 3%. Smell-test? Fail.

The bigger, egregious “error” is with regard to government consumption expenditures. These expenditures are for teachers’, hospital- and government-workers’ salaries, etc. When Finance Minister Chidambaram went on an expenditure-cutting spree in FY14, it was precisely these expenditures that he was reducing. There is no new information, or new survey, etc, with regard to government expenditures; if some local-body expenditures now enter the system, they should enter in both the years. So, what explains this wide discrepancy between reliable reality and unreliable statistics? Actually, all the explanations suggest that government consumption in FY14 grew at a lower rate than in FY13; instead, the new CSO data suggest that real government expenditures accelerated from an only-1.7% growth in FY13 to a 8.2 % growth rate in FY14. According to the reality-check old CSO data, real government consumption expenditures decelerated from 6.2% growth in FY13 to 3.8% growth in FY14. Smell test? Big fail!

How much difference do the corrections for investment and government spending make? About 1% less growth in FY14 than the 6.9% growth stated by CSO, with 0.4 % less contributed by real real government spending and 0.5% less contributed by real real investment spending.

Delhi election tailpiece: As you read this article, you will be going out to vote. Delhi election has seen much excitement, and our “poll of polls” calculations suggest that the excitement is yet to begin. The median vote-share, according to 8 polls undertaken in February 2015, suggest the following vote shares for each of the parties: the AAP–43%; the BJP–37% and INC–13%. The corresponding seats: the AAP–43; the BJP–25 and the INC–2 seats. It may be the case that the opinion polls are wrong, but a large selection of such polls being wrong is unlikely.

Bhalla is co-author, with Ankur Choudhary, of a recently-released book Criconomics–Everything you wanted to know about ODI cricket and More.

Twitter: @surjitbhalla

Get live Stock Prices from BSE and NSE and latest NAV, portfolio of Mutual Funds, calculate your tax by Income Tax Calculator, know market’s Top Gainers, Top Losers & Best Equity Funds. Like us on Facebook and follow us on Twitter.

Switch to Hindi Edition