Former CEA of India Arvind Subramanian is now on record that India’s data is fudged, and not just in recent years but since 2011, including the period when he was in office.
By Charan Singh
In the recent article, Arvind Subramanian, former Chief Economic Adviser (CEA), Government of India, has observed that India’s growth estimates are significantly overestimated since 2011-12. In fact, official estimates are around 7 per cent of annual average growth but according to Arvind Subramanian, actual growth rate, most probably, would have been 4.5 per cent, or even as low as 3.5 per cent. Arvind Subramanian uses various econometric techniques on data from India and cross-sectional panel, to justify his argument. Arvind Subramanian also argues that his criticism is not politically motivated but is technical in nature.
Now, this is an extremely unique situation, as the former CEA of India is now on record that India’s data is fudged, and not just in recent years but since 2011, including the period when he was in office. Hence, the article has an impact on the credibility and consequentially, ratings of the country. Earlier, there were rumblings of dissent and opposition to methodology and high growth rate figures by Arvind Subramanian which found a mention in the Economic Survey of 2016-17. The matter should have rested there, especially, when the statistical team from the IMF had visited India to examine the methodology and accepted the changes.
The timing of the article is very critical as the first Economic Survey by a recently appointed young CEA is being finalized and the Union Budget of the new Government by the new Finance Minister is being prepared. The country is grappling with real issues relating to slowing down of the economy in recent months, kick-starting growth after a strenuous and uncertain period of general elections, and the traditional problem of poverty and unemployment, especially amongst the youth. The article by Arvind Subramanian, sensationalizing the issue, therefore, is political rather than a genuine case of initiating an intellectual debate on the computation of revised series. This is a clear case of intellectual treason against the country.
Globally, computation of growth in any economy is complex and various methods and institutions are involved. Estimations are made, some based on surveys which are decades old because it is expensive to hold country-wide surveys annually, and checks and balances ensure that law of averages operate. The trend is most important rather than just the levels. Illustratively, computation of GDP in agriculture, with rice, spinach and lemons was never easy despite the availability of different technology. Similarly, production of lakhs of goods and services in the unorganized sector across the vast country like India. Statisticians and economists, under the aegis of the United Nations, work to standardize the procedures and methods of computing national income. As in a small African country totally dependent on coconut to the most modern country with advanced technology, computation of GDP is always a challenge, widely known to officials, despite standardization of methodology.
Interestingly, computation of growth of any country is rather very complex and Arvind Subramanian would know it best as he served in the IMF too. The IEO study of the IMF reveals how complex it is to forecast growth, globally, as well as individual countries. There are studies that establish that in many countries, forecasts perennially have an upward bias, especially in the advanced countries, and generally during election times. In contrast, there are countries where there is generally a downward bias in the estimation, like India. Similarly, there are numerous and regular cases of revision in methodology and base year, which in most countries result in an upward bias in estimations. Even in the USA, recognition, and inclusion of artwork happened recently resulting in back-revision of GDP. In any emerging country, with more availability of data and refinements in methodology, as well as increasing computerization, such revisions are expected and anticipated to have an upward bias.
Focusing on technical aspects of the paper by Arvind Subramanian, the sheen of analysis is lost when examined closely. Should India be compared with a set of many countries or should the focus be on similar characteristics of an emerging country? Arvind Subramanian attempts regressions in the article, arguing his case, based on data from high and middle-level countries, middle-income countries and with all countries. Illustratively, apart from mentioning poverty levels, and population size, India’s GDP per capita is less than $2,000 and that of the US is around $60,000.
In the IEO study on Research at the IMF, it became apparent that many IMF economists learnt econometric techniques in school, and applied those extensively in-country research, without realizing lack of data and understanding country context, and with the help of dummy variables, binary 0 and 1, tailored their findings. The paper by Arvind Subramanian is a perfect example of that shoddy plumbing. Should the comparison of India be done with select emerging countries or with any group of countries? Actually, it can be argued that India is a unique emerging country which is not resource rich, is democratic and did not manipulate its exchange rate to push exports and accumulate international reserves. Is there a parallel then?
The changes that the Indian economy has registered in the last decade are phenomenal. The crowded airports and modern shopping malls, expensive cars on the road, surge in luxury hotels, improvement in living standards, increasing banking, insurance and internet penetration, expanding cell phone market, the persistence of parallel economy, and a general rise in the monetized economy are all examples of advancing India. The Indian economy has diversified, and the unorganized sector has expanded exponentially. To many, given the smell test, India’s growth is understated. To any casual traveller in India, prosperity and development is apparent and palpitating
In such a dynamic and vibrant economy, depending on the correlation of some select indicators to negate the growth trends computed by highly sophisticated methods and manned by a technically qualified and globally recognized cadre of economists and statisticians, does not establish the credibility of technical aspects of the challenge. As an illustration, should foreign tourist arrival be taken as a proxy or number of Indian tourists travelling domestically and abroad? Similarly, consider the case of financial markets after Lehman, and its delayed impact on India.
In such a situation, domestically, would real credit to industry be a suitable variable, especially when many banks in India were in the grip of Prompt Corrective Action or alternative variable, like Sensex, could have been examined? With global growth suffering, is export from India a good predictive variable? Econometrically, high R-bar squares have long lost their admiration and has been replaced by economic logic in the equations. Technically, this paper is a classic illustration of what should not be done in good academic or policy-oriented research!
Finally, the services sector, construction and housing have been booming but Arvind Subramanian did not consider those variables. And, it is to the credit of computing agencies, CSO of India in this case, that despite the election year, and rural-dominated country, the reporting consistently revealed that growth is suffering in agriculture. How appropriate is it to cast aspersions on the loyal army of technocrats who faithfully reported to Arvind Subramanian, for nearly four years until June 2018.
India is on the cusp of massive growth trajectory and needs constructive suggestions to improve growth. The paper by Arvind Subramanian lacking rigour and thought has resulted in wasting much of national time of serious researchers and scholars, as well as media. It has certainly, given the background of Arvind Subramanian, cast doubt on India’s data, intentions, and credibility. Kennedy School as also other important global public policy institutions could do a favour to emerging and developing countries not to distract nation-building efforts by such slapdash research hypothesis which have not been tested appropriately.
As in biological sciences, vaccines and medicines undergo arduous tests before reaching the market, so should research output because investment and growth through rating agencies can get impacted, especially when the author has been a high official in the respective country. It is also for the Government of India to consider as to who all can be trusted with august offices of the country. In a complex socio-economic structure of India, verstehen approach would probably work better, rather than a short-term visiting assignment by an NRI. Resident Indians are not experimental mice, and India is not a banana republic, please.
- Charan Singh is Chief Executive and Director, EGROW Foundation, a Noida-based think-tank. Views are the author’s own.