The good Pragmatic Desi asked me to write about the release of the latest poverty numbers from NSSO. I do want to write a follow up post to my Pragati article. But the bad news is that I have accumulated a lot of heavy reading material in preparation for that follow up post which I need to read. The good news is that a long weekend is coming up, and I hope to get the reading and the post out by then.

But in the mean time, here is a rant.

Every so often, a poverty number comes out, and we go through essentially the same set of arguments . It is almost as if the dispute over poverty statistics in India is a battle over which headlines and sound bytes show up in the news rather than a serious discussion over which policies are best suited to reduce poverty.

The same errors are committed over and over again. For example, Dilip D’Souza tweeted: “Know anyone who lives in a city and earns Rs 860/mth? Understand this about them: they are not poor. “

 As many people (including me, 4 months back) have pointed out, a person earning less than  Rs860  a  month would probably be classified as poor, if she has a family to support, because the family is being forced to get by on less than Rs860 per day per person in a city. The poverty definition is based on consumption per person, not on the income of the breadwinner. Dilip made the same error in 2006, and hasn’t had occasion to update his knowledge in 6 years.  That, however, doesn’t stop him from accusing the Planning commission of bad faith in his tweets  (1, 2, 3, 4, 5 )

 (Dilip, if you are reading this, the explanation for the discrepancy in numbers is here.  In short, they did what you want them to do – they updated their definition of poverty.  The fact that you got confused explains why they shouldn’t do it very often.)

 When we do point out the error, the angry rejoinder is that the poverty line is still too low – which family can get by comfortably at Rs6,000 a month? we are asked. This question too I have answered in my Pragati piece, but it is worth addressing in detail, which I will do here.

 Let me use an analogy from the other end of the spectrum here – just as we disagree over what is poverty, we also disagree over what affluence is. If you are an Indian reading this post, I would estimate that an overwhelming majority of you are from the top 10% of the population, and I will bet that if you were asked to classify yourself, you would put yourself in the middle-class, reserving the epithet “rich” for people like Mukesh Ambani.

 We can argue till kingdom come over whether you are rich or middle-class, or we can recognize that some things are objectively determined, while others are a matter of classification. The objective fact is that you travel to work by car while Ambani does by helicopter. Whether your commuting by car makes you rich is a matter of terminology and classification, which may differ from person to person. The criteria for classification shouldn’t be entirely arbitrary – what criteria you use should depend on what you want to measure.

 But one thing is sure – if you use one set of criteria to classify people as poor or rich, you should take care to use the same criteria when you want to compare across time. If you use “commutes by helicopter” as your condition for being rich in 2004, but use “commutes by car or helicopter” in 2009, you will get a large increase in the number of rich people. But your data will be garbage.

  Likewise, you can get a higher poverty headcount ratio by raising the poverty line. But what will you achieve by doing so? You will get a scarier headline, but what will you learn from it about the impact of our policies? To find out what impact government policies has had on poverty, we need to use the same criteria across time to see if poverty is reducing or increasing, and by how much.

 So let’s say that we update our poverty line and find out that instead of dropping from 37% to 29%, our poverty rate has dropped from 50% to 45%. A drop of 5% instead of 8% as we originally thought.

 What will the headlines say? “Drop in poverty less than originally assumed”.  What is the real story? It will be “Growth has benefited the poorest the most”.

 Now, for the record, I don’t believe we are going to see such numbers if we update our poverty line. I do agree that while growth has benefited the poor, the rich (and the less poor) have benefited more – so if you are a lower middle-class person struggling to send kids to school, you have probably gotten more out of growth than the destitute construction labourer. So the headline will probably be the other way round  – and the headlines will be as misleading as ever. (“Poverty has fallen even more than believed!”)

 But the point is to illustrate the silliness of what is going on.  We have a system of measuring poverty. Yes it measures destitution rather than poverty, but in terms of methodology, it is one of the better functioning systems in the country (though it is annoying that they take 2 years to process and release data). And there is nothing wrong with focusing on destitution – wasn’t it Gandhiji’s dictum that  we should keep the face of the poorest person in the country in mind whenever we decide policies? In what way will including the moderately poor in addition to  the utterly destitute in the calculation improve our policy making?

 Now, here’s the thing. I believe that the best way to achieve reduction in poverty is economic growth. I believe that growth does not leave the poorest behind. Given these beliefs, I don’t believe that raising the poverty line is such a big deal. After all, if growth happens, the rising tide will lift the boats of the moderately poor just as it will raise those of the destitute – so I would expect to see a reduction in poverty no matter where we draw the poverty line.

 But suppose you disagree with me. Suppose you think that growth is not enough, and that there ought to be special focus on redistributing resources to the poorest. In that case, where would you want the poverty line to be? Shouldn’t you favour a poverty line that is drawn such that there is a sharp focus on the poorest and an accurate measurement of their number? Why would you want to dilute your metric by including the less poor in your measurement?

 At this point, I might hear an objection (stated with a supercilious smile on your face) “OK. In that case, why not lower the poverty line all the way to zero? Problem solved. No poverty!”.

 That objection would miss the point. The point is that any metric we use should give us useful information about the impact of our policies. Yes, our current poverty line is the destitution line. But it is important to note that nearly 3 out of 10 Indians continue to be below that line.  In our list of national priorities, improving their lot should come near the top – which is why we should measure the number of our destitute to see how well we are doing in terms of achieving our priorities.

 This doesn’t mean that this line is the one we should use for all time.  If we used Indian criteria to classify Americans as poor, we would get a ridiculously low percentage. Should the Americans use our criteria? No – in the US, destitution has more to do with things like drug abuse, broken families, mental illness and cultural issues than with their economic growth – while the Americans should measure destitution and should figure out how to reduce it, they shouldn’t use it as the most important metric to measure the impact of their government’s policies.  When we get richer, we should update our poverty line – we should do it transparently so that we can maintain comparability.

 The point is that the metrics we use should help us focus on our most important priorities. In my view, the current poverty measurement does so. If you disagree, you need to explain why. But an explanation that amounts to: “This poverty line gives us a lower poverty percentage, and a less scary headline” is frivolous. That this argument has dominated our national debate tells us a lot about the quality of our national debate.