Why media's data illiteracy should worry us more than the Election Commission's SIR goofups

Sometimes, it is worth stepping back from the controversies of the day to understand why democracies fail to grapple with real issues. In a democracy, information is more freely available than in autocracies. But information is only as good as its source, and how rigorously it has been researched, in order to come as close as possible to the truth. 

In the current-day cacophony of social media voices (where practically anyone with a smartphone can be a “journalist”), and adding the shrill attempts of TV anchors and even newspapers to create clickbait headlines in order to compete with instant social media virality, it is highly unlikely that the average citizen can come anywhere close to real “facts”. Consider the typical human tendency to listen more to people they like than those they dislike, add the social media algorithms that feed the same echo chamber, and voters are more likely to be influenced by shrillness and “feel” rather than the underlying reality, or the truth.

In this article, I am going to argue that data-illiterate journalists are more guilty of peddling flawed narratives than the Election Commission is of wrongly deleting voter names in the Special Intensive Revision (SIR) of electoral rolls, which is even now not without remedies. But the journalistic deficit in numeracy is a real crime against the truth.

The problem begins with the core advice budding journalists get when they start out on their careers: find an angle, find out what makes a story. An “angle” is, by definition, a limited, even slanted, view of reality. We get to define an elephant by the part we get hold of, not the whole of it. In the real world, truth comes in shades of grey, not black and white. And in today’s hyper social media world, the focus is on “narrative” and “virality”, not truth at all. Just look at the sheer number of young people at Jantar Mantar trying to generate viral video content on their smartphones.

Forget the amateurs. Today’s “journalists” themselves are not numerate by and large. Three or four incidents strung together make a trend, as we saw in 2014-15 when a few untoward events involving churches led to a narrative that Christians were under attack. It was only later, when the police said similar things were happening with much more regularity in temples and gurudwaras, that the narrative was quietly buried, and even so it is more than likely that it was not “facts”, but the rape of an aged nun in West Bengal - which the media hoped was done by a Sanghi - turned out to be an illegal Bangladeshi migrant. Media embarrassment over how the rape was reported was what buried the story, not facts. 

Given the essential gullibility of journalists, and their lack of data numeracy, you can grab the headlines if you make the wildest claims in terms of numbers. This is why when someone says “13 crore voters” were knocked off the electoral rolls during the SIR exercise, everyone takes notice without giving this headline number the scrutiny it requires. Once the number is repeated 10 times in the media, it becomes the accepted figure in terms of public perception.

Psychologists call this the anchoring effect of initial numbers. If you are in a shop bargaining with the salesman, the latter will deliberately indicate a high price initially so that even after bargaining you still end up paying more than you should have. Having created a mental anchor with a high number, the bargain hunter has lost the game from the word go.

Then there is the other psychological phenomenon of fear of losses. Given a choice between gaining Rs 5 and losing Rs 5, most people forgo the option to gain and focus on likely loss to make their decisions. This is why vested interests, with something to lose, tend to be more vocal and more influential in policy choices than much larger beneficiaries. In the current controversy surrounding voter deletions in the SIR, the one fearing a deletion is likely to be more vocal and active than the overwhelming majority who have not faced a deletion.

These two mind traps, the anchoring effect and the fear of loss, are key to understanding what happens in the policy sphere, since politicians who have to make those choices also have to worry about how voters will see an issue. Data-illiterate journos often end up amplifying a false or half-baked narrative.  

Consider how the anchoring effect influenced policies in the past. When Comptroller and Auditor General Vinod Rai gave us the figure of Rs 1,76,000 crore “presumptive loss” in the award of telecom licences in 2008 without an auction, he gave us three more (lower) estimates of presumptive losses to the exchequer, but only the biggest number became a political football. It didn’t matter to the opposition or TV debaters that a notional loss can never be real. Also, there can be good reasons for giving out spectrum at low prices, but the high loss figure made everything look suspicious. The fact that there was a real possibility of favouritism in the award of telecom licences got subsumed in the massive loss figure, which then made it impossible for any future government to consider selling spectrum without setting super high base prices. 

The same goes for other large numbers that are uninformed by other nuances. One is the massive public sympathy for farmers. Politicians with a vested interest in the farm vote always talk up farm suicides to prevent us from thinking about the underlying implications for extending large loan waivers or raising minimum support prices well above market prices. According to a study by Down To Earth, between 1995 and recent years, some 3,90,000 farmers committed suicide, and probably around 10,000 continue to die annually by taking their own lives even today. 

While this is certainly a human tragedy worth noting and doing something about, it hides the fact that this number was always bound to be huge given the fact that 45-46 percent of the Indian workforce is dependent on agriculture or related activities, which means over 600 million people. A cumulative suicide total of 3.9 lakh on a base of multiples of billions of population is not that significant (Note: it is not just the current population which is the denominator in calculating the suicide rate of farmers, but a much larger denominator that has to be arrived at by adding the total farming population in each year since 1995). Suddenly, the farmer suicide issue  will be seen in the right perspective - a human tragedy, but nothing as significant as to skew subsidy and loan waiver policies in unsustainable directions. 

Consider a reverse fact where the numbers are real, but no one cares. Data shows that the male suicide rate in India is two-and-a-half times that of women, but no one is rushing to declare a male mental health emergency as yet. This is because we are programmed to discount male suffering since patriarchy still rules and women visibly suffer more than men. We do not want to know why men kill themselves when it is obvious that women suffer more. Both facts can be true, but we want to believe only one.

Now let us come to the big number that is evoking such visions of a catastrophic democratic slide in India: the proposed deletion of 13 crore names from the voters’ list by the Election Commission in the ongoing SIR exercise. In my Business Standard column in September 2026, I offered two other contrasting figures to show why the deletion number may not actually be that large. India has a large domestic migrant population of around 45 crore; and upto one crore people die every year from natural and other causes. Between the last SIR exercise in 2002-04 and today, at least 10-12 crore citizens would have died, and most may still be on the voter lists because deletion is not something one worries about (why bother, when the dead anyway can’t vote). The 13 crore figure has got former election commissioners and even sitting and retired judges excited about the possibility of a large-scale attempt to disenfranchise voters. Justice Ujjal Bhuyan, speaking at Nalsar the other day, claimed “millions and millions” of Indians had been disenfranchised, and another former judicial worthy, Rohinton Nariman, went further and alleged that 90 lakh voters were knocked out in West Bengal so that the BJP could win. This is utter nonsense, and it is a travesty that a former Supreme Court judge is saying this. First, not all 90 lakh voters were necessarily citizens with a legal right to vote, as many of those deleted were either dead or had migrated permanently. 

In the current 13 crore figure of deletions (and these can be restored through an appeals process), some 2.8 crore were listed as dead, and 6.3 crore as possible permanent migrants, and one crore had duplicate registrations. Some three crore were absent from homes when the SIR surveys were done. The figures on the dead and migration are not inconsistent with the actual levels of people dying or migrating for work nationally. But the media narrative is about the figure that gets mentioned again and again: 13 crore deletions.

This is not, in any way, to suggest that many voters were not unfairly deleted from the rolls, and this is what needs fixing. Their numbers may be small (probably less than five million nationally, which would be one-half of one percent of voters, now estimated to be just under one billion). This is where loss aversion comes into play. These disenfranchised voices will be spreading negativity more than those 99 percent of registered voters who were not deleted in the SIR exercise.

What the Election Commission is up against is the negative framing of its failures in absolute terms, which is always the case in India. It has to come up with a better and more robust and continuous process to clean up the voter rolls, with stronger remedies for wrongful exclusions. 

The truth is not that the Election Commission somehow conspired to delete many voters; it is that it did not test its processes for robustness in a country where gross numbers are always large. 

As for journalists, it is time they brushed up a bit on understanding numbers. It is a scandal worse than the EC’s alleged mass disenfranchisement.