John Hicks, the great economist, wrote in 1975, “Though the Social Product (or National Income, or Gross National Product) has been ‘sold’ to the public, it has been ‘sold’ by economists. We have responsibility for it — the responsibility of the manufacturer for the quality of the good which he sells. We cannot escape the duty of justifying it by pretending that it has been justified already.”
Almost half a century later, Hicks’ statement needs an amendment. These days, national statistics are still ‘made’ by economists and statisticians, but the ‘selling’ is done under the strict supervision of politicians. Since politicians are not good at ensuring quality, the consequences have been undesirable, as the experience of our country shows.
Economics data on important entities, such as national income, gross domestic product, consumers’ expenditure, price rise, employment-unemployment trends, balance of payments and a host of similar indices, provide a quick, objective view of the current economic condition of a nation and the standard of living of its citizens. Data and statistical indices are indispensable for governance as well as for judging the effectiveness of governance. International media and agencies routinely rank nations on the basis of their national statistics. Data, thus, are of immense importance. But to serve their purpose, data must be trustworthy. And therein lies the rub!
Data are manufactured entities. They need to be ‘improved’ for good performance. Little wonder then that there is a great temptation to doctor data. This fudging has become a part of politics.
The Cold War provides a striking example of how politics influences the official use of statistics. Nikita Khrushchev famously boasted in 1957: “Growth of industrial and agricultural production is the battering ram with which we shall smash the capitalist system.” The United States of America responded by making the growth rate, rather than the level of GNP, the centrepiece of its official statements backed by appropriate statistics.
For statistics to be influenced by politics is one thing; but statistics to be manipulated for overtly political purposes is quite another — this belongs to the realm of malfeasance. Soviet progress under central planning was real enough, but the official statistics required some scaling down. Similarly, Chinese data used to be notoriously unreliable. The quality has improved now, but it still calls for close scrutiny by experts.
The worst offender, though, in recent years is Greece. To qualify for European Union membership, countries must satisfy certain criteria, one of which is a sufficiently low budget deficit to GDP ratio. The Greek government grossly manipulated both national income and deficit figures to meet the criteria. Eurostat officials expressed grave doubts, but Greece was granted membership nevertheless for geopolitical reasons. That was a big mistake. In 2014-15, in the midst of the country’s debt crisis, Greek politicians claimed that national income figures were being understated in order to qualify for a bail-out by European governments. The economy, in their opinion, was stronger than the data. Andreas Georgiou, the then national statistician, was forced to resign in 2016. The country’s international creditors insisted that his replacement must be approved by an international panel of distinguished statisticians.
What is happening in India at present is alarming too. It has moved from simple data manipulation to malpractice of almost Grecian proportions. Let’s briefly look at the history of India’s data edifice.
Although an official statistical system was already in place during the raj, serious efforts to strengthen it began only in the 1950s when India embarked on its unique experiment of comprehensive planning in a democracy. The statistical infrastructure was revamped under the leadership of P.C. Mahalanobis backed by the former prime minister, Jawaharlal Nehru. Systematic data collection on a very wide front was started by the National Statistical Organisation using sample survey methods developed and refined by Mahalanobis and his team. Within a few years, Indian official statistics came to enjoy global prestige because of their high quality signifying scientific rigour in the collection, processing and presentation of raw data obtained from samples of farms, factories, other production units and consumers from all corners of a vast country.
With Mahalanobis’s death in 1972, the system lost its skilled leader. The Planning Commission’s power and prestige also declined over time. In the early 2000s, a high-powered committee was appointed under the chairmanship of C. Rangarajan for a thorough review. It suggested numerous reforms, of which only a handful were implemented half-heartedly.
After coming to power in 2014, the Narendra Modi government replaced the Planning Commission with the NITI Aayog. This think tank and the ministry of statistics and programme implementation are in charge of looking after official statistics. Are they doing their job well? Unfortunately, most experts feel that these two entities have brought the system on the verge of a major crisis. Statistics have become a political instrument; consequently, objectivity and credibility have nosedived. An honest appraisal of government policy using official data has become nearly impossible.
A dense data fog has been generated to hide the true state of the economy, especially in regard to the plight of the poor. For measuring the depth and the extent of poverty in a country, data on per capita consumer expenditure is essential. An embarrassing National Sample Survey report came out in late 2019, which was suppressed by the government. Leaked figures revealed a drop of 3.7% in real per capita monthly consumption spending over the period 2012-2017, with a larger decrease in rural areas compared to urban areas. Calculations made by a member of the World Bank’s Commission on Global Poverty using the leaked data showed that 320 million rural Indians lived in acute poverty in 2017-18, 50 million more than just six years earlier. This is consistent with the sharp drop in rural per capita consumer spending revealed in the survey.
The official response was that the survey’s estimates were contradicted by consumption data of National Accounts Statistics and were, hence, unreliable. This is hardly a tenable defence because NAS procedures are often dubious in themselves. (The persistent — widening — discrepancy between NSS and NAS data is a disturbing feature of the official statistical system.)
On account of the pervasive fog or the unavailability of data on consumer spending or unemployment, researchers are now forced to turn to surveys by private organisations like the Centre for Monitoring Indian Economy. The quality of data is beyond reproach but the cost is very high. Most researchers or their organisations cannot afford that cost. As a result, work on consumption-based poverty or deprivation measures has virtually come to a standstill.
India’s once reliable statistical system built over the years has been reduced to a shambles. It’s a familiar situation with many other institutions. One must keep on hoping for the day when good sense and good statistics will prevail once more.
Soumyen Sikdar is a former professor, IIMC
