As I write this on my laptop, there’s a headline elsewhere on my screen saying the Sensex rose by 228 points. In fact, it goes like this: “Market snaps 9-day losing streak; Sensex gains 228 points."

Question: is this 228 points good news? Are investors and stockbrokers and other folks with an interest in financial matters dancing the bhangra over this increase? Over snapping the losing streak?

I trust you know the answer to those questions. Of course, a rise in the Sensex is good news, in the sense that it means most investors’ wealth has increased and that’s generally seen as a good thing. Similarly, of course the end of a losing streak is good news, because who wants their wealth to steadily decrease?

But let’s say the Sensex rose from 100,000 to 100,228. And let’s say over the previous 9 days, it had slumped from 200,000 all the way down to 100,000. Are we still talking of bhangra-level good news? Alternatively, suppose the Sensex rose from 100 to 328, and this after 9 days in which it slid inexorably from 110 to 100. Now what about that bhangra?

The point, and surely it is obvious, is that changes in the Sensex are meaningful not by themselves, but only in comparison to the value of the Sensex itself. In the first example above, the rise from 100,000 to 100,228 is all of 0.2%, barely a blip. What’s more, it’s hardly likely to raise the spirits of someone licking her wounds after the market lost a full 50% of its value— 200,000 to 100,000—over the previous 9 days. But the second example? There’s serious reason for joy. That 228 point rise more than triples the value of her portfolio, handsomely compensating for a 9-day loss of less than 10%.

You know just how a financial professional would react to these two cases. But only, in both cases, after she matches the changes in the Sensex against its value.

In some mathematical circles, headlines like the one above are called “numerator-only data". Which is to say, it’s as if I only told you about the numerators of fractions, not their denominators. For example, suppose you and I have to share a cake. I suggest two possibilities and ask which you want. I could cut it into two equal pieces and offer you one—meaning half, or 1/2, of the cake. Alternatively, I could cut it into eight equal pieces and offer you four—also meaning half, but now 4/8, of the cake. If I urge you to take the second option because you’ll get four pieces instead of one, I hope you will look at me as if I’ve lost my marbles. After all, you know the denominators—2 and 8—and therefore know that both options give you exactly the same quantity of cake. You would opt for four pieces only if you knew its denominator was less than 8: 4/5, for example, is considerably more than 1/2.

Not just that. If you didn’t know the denominator, I could try fooling you by suggesting that the number of pieces alone—the numerator alone—means you get more of the cake, 4 being more than 1. It’s true you probably won’t get fooled, because you’ll want to know the denominator, and thus the actual fraction. But still, I could try.

Again, obvious stuff, right? Yet it’s startling how often we are fed numerator-only data and expected to believe it. Sometimes, it’s innocuous and probably innocent, like with the Sensex gain above. But sometimes, it’s anything but innocent. Especially if we let “denominator" take a slightly wider meaning here—call it “data deliberately omitted".

This being election season, with its attendant doses of hatred and bigotry, there are plenty of examples of less than innocuous use of data. Take just one.

In Bhopal, a certain Mr Chouksey has printed and circulated a leaflet, a copy of which I have. It is titled “If you want to save your culture, understand the ‘Green Plan’". (My translation from Hindi). Maybe that already suggests what this leaflet amounts to: a two-page-long screed against Muslims. After all, writes Chouksey, Hindus believe in values like love, truth and non-violence, and they respect all other religions. In contrast, he writes, Islam considers those who follow other religions sinners, Muslims think it is sinful to pay respect to our Bharat Mata, and they are against “Vande Mataram". Also, they believe it is their religious duty to eliminate other religions, and that doing so will give them entry into heaven.

Etcetera. Yes, there’s more.

This baring of Mr Chouksey’s soul is hardly unfamiliar. In this country, we’ve heard this stuff over and over again from all kinds of people, some of whom have built political careers on it. This column is hardly the place to challenge these precepts. But Chouksey uses all this as context for a table of figures he offers, and this column is certainly a good place to challenge those figures.

Chouksey’s thesis is that Muslim numbers in India are steadily increasing, and this increase outstrips that of Hindus. Thus the Muslim “Green Plan": to spread Islam’s values across this land, specifically by deliberately increasing Muslim numbers till they are the majority here. This is the threat, Chouksey suggests, that confronts the 15,000 year-old Hindu civilization.

Chouksey’s table is supposed to support this spectre. It shows Hindu and Muslim population figures for every census since 1951, both actual numbers and as percentages, all taken from the Census of India website.

Briefly, the table says that between 1951 and 2011, the number of Hindus in India rose from 303.5 million to 966.2 million, while the number of Muslims rose from 35.4 million to 172.2 million. In the same six decades, the percentage of India that is Hindu dropped from 89.55 to 84.87, whereas the Muslim percentage rose from 10.44 to 15.13. There you have it: while both faiths have seen an increase in numbers, their respective shares of India’s population have gone in opposite directions: Muslims rising, Hindus falling.

The first thing to note here is that for every census, Chouksey’s Hindu and Muslim percentage numbers add to 100. This is clearly impossible, because this country also has plenty of Sikhs and Christians, Jains and Parsis, Jews and atheists and agnostics.

With that in mind, let’s look more closely at the population figures. Hindu and Muslim numbers have both risen dramatically. No doubt about that. But let’s start by putting denominators—in this case, the population in 1951. We find the Hindu population has increased by a factor of just over 3, the Muslim population by a factor of just under 5. No doubt there either.

If Chouksey thinks this difference is frightening, fair enough. But for one thing, he has ignored other religions. Putting their numbers into the denominators, as we should do, will change those factors somewhat. Perhaps not substantially, but change them nevertheless. For another thing, census figures also show that growth rates are slowing across India. And while the Muslim growth rate still outstrips the Hindu growth rate, it is declining faster than the Hindu rate is. Using the right denominators—for each census, the numbers from the previous census—will show this unmistakably.

But Chouksey commits one final obfuscation. If Hindus do nothing, he says, in 350 years they will become a minority in India. This is, he says, “straight arithmetic".

True, it is pretty straight arithmetic. (I get 265 years, but never mind). But is it reasonable to get people’s bile up about a prospect 350 years from now? That’s for you to answer for yourself. While you do, consider this: Chouksey got 350 by projecting today’s growth rates into the future and looking for the point when Muslim numbers will overtake Hindu numbers.

But what Chouksey does not tell us—call it the denominator, the data he has deliberately omitted—are those Muslim and Hindu numbers, 350 years from now. At today’s growth rates, there will then be something like 200 billion Hindus and 200 billion Muslims in India. (Many billion Jains and Sikhs and so on, but never mind them). Think of it: about 400 fellow-Indians occupying the space you occupy, as you read this.

Absurd, right? Which is why Chouksey hid those numbers. Which is why he didn’t tell you the truth about population growth: that it is indeed slowing across the board in India. In fact, our population will plateau sometime later this century. At that time and forever after, Hindus will outnumber Muslims as they do today, about four or five to one.

Put in the denominator, you see, and suddenly, even in an election-time leaflet, things don’t look quite as menacing as the Choukseys would like.

Once a computer scientist, Dilip D’Souza now lives in Mumbai and writes for his dinners. His Twitter handle is @DeathEndsFun

Close

Wait for it…

Log in to our website to save your bookmarks. It'll just take a moment.

Yes, Continue

Your session has expired, please login again.

Yes, Continue