When Numbers are Treated as Political Weapons Jayati Ghosh

India has a robust and admired statistical system. But the government is suppressing data to suit its narrative. It is perilous not to know the reality of the governed

No government in India has been as adept in shaping the political narrative as the Modi government since 2014. This has been accomplished through a domination of most of the mainstream media, significant control over social media, ruthless intimidation and repression of dissident voices, and reliance on a web of disseminators on the ground. A critical aspect of this has been the ability to control, manipulate and suppress official data in ways that are designed to conform to the desired narrative.

This means that an aggressive attitude towards economic statistics and indeed to all data that could force some accountability has become one of the defining features of the government. As a result, a national statistical system that was once the pride of India and the envy of most developing countries is now weakened, eroded and increasingly less reliable.

On the face of it, this may seem like a contradictory, even self-defeating, strategy for any government. After all, even minimal governance requires some basic knowledge about those who are to be governed—at the very least, how many people there are and where they are. Much more information is required to be effective in improving the conditions of the people: about how they live, work and earn; about disparities in income and wealth; about access to basic needs including food, health, housing, sanitation; about levels of education and skill; about ecological conditions and much more.

India’s statistical system was remarkable for a relatively poor country. Even before independence, the decennial Census exercise of the colonial government provided some basic information. From the early 1950s, India’s statisticians sought to develop methodologies for surveys that could capture consumption expenditure and employment changes and patterns, using imaginative techniques suited to an economy that was largely informal. These were periodically tweaked as gaps and anomalies were recognised. National income indicators like the gross domestic product (GDP) were developed in conformity with the UN’s System of National Accounts. As economic planning developed, other official data were generated, such as input-output tables and human development indicators that also recognised different social categories. Poverty estimates were based on data generated through the consumption surveys, and other backwardness criterions based on survey data also became important for policy formulation.

But since 2014, several moves have dramatically changed the availability and reliability of such essential data. This began with the revision of the base year for estimates of national income, an exercise which occurs every few years. The change to the new base year of 2011-12 brought changes in methodology. An important change was in how industrial output is measured, shifting away from mainly using the Index of Industrial Production, which is a physical measure of output. Now company reports are used, even though that data includes a significant proportion (around a third) of dummy companies, shows very wide fluctuations and can reflect incentives for companies to misstate their returns. This led to a dramatic increase in manufacturing output that was not confirmed by other economic indicators and appears to have been designed to present the Modi government in a better light than the previous one.

Next, though the results of the Periodic Labour Force Survey of 2017-18 were available in late 2018, they were withheld from the public until after the 2019 elections, presumably because they indicated stagnant or declining employment and high rates of open unemployment. Even worse, the results of the survey on consumption expenditure for 2017-18 were withdrawn and the government simply refused to release the data. A leaked report procured by journalists indicated alarming declines in rural consumption and significant increases in rural poverty between 2011-2012 and 2017-2018. Since then, no fresh consumption survey has even been conducted, even though that is essential for assessing the extent of absolute poverty, food consumption and inequality. So there is no information on how these crucial indicators have moved after 2011-12—more than a decade ago.

In the absence of essential and comparable data on poverty, the Niti Aayog has developed its own indicator—an index based on a collection of different estimates including mobile phone usage. All too predictably, this index shows a sharp decline in poverty in the country over the past decade, conforming to the preferred official narrative.

The Covid pandemic provided an excuse for the government to delay the Census that should have been conducted in 2021, even though other major infection-spreading events like assembly elections and associated rallies continued. Remarkably, there has been no attempt to conduct the Census since then and it has officially been delayed “until further notice”, presumably until after the next general elections. Since the last Census data relates to 2011, both the government and the public are now functioning in ignorance of the most basic data: how many people there are in the country, where, of what gender and age, and doing what.

Not only has the central government refused to conduct the national Census thus far, it is also resisting the efforts of state governments to try and fill the gap by attempting their own enumeration. The caste census, conducted by the Bihar government and now proposed in several other states, is being bitterly opposed by the central government, even though this should provide important information also on location, living and working conditions.

Other data that could shed light on the success or usefulness of government policies are also suppressed. The government’s own audit of the Ganga Action Plan, for example, has simply not been released. When one of the few remaining official surveys, the National Family Health Survey for 2019-2021, revealed that open defaecation still existed despite official claims to the contrary and that the incidence of anaemia had increased among women and girls, the government response was not to consider how this can be improved, but to punish the messenger. The head of the agency that conducts the survey was suspended on vague charges, presumably as punishment. This obviously sends a signal to the person in charge of the currently ongoing survey not to publish any results that could potentially embarrass the government.

None of this suppression of data would make sense for a government interested in developing and implementing economic policies that would benefit the people, because it is hard to implement policies in such ignorance of the people for whom they are intended. This instead suggests maximum government with minimum governance: a state focused on control and the management of public perception, rather than positively changing the reality.

(This article was originally published in The New Indian Express on October 19, 2023)