What Should You Read to Get a Better Sense of Poverty Line Estimates?

Ali, an FYMSc student at GIPE reached out with a couple of questions about poverty line estimates – their nature, are they adjusted for inflation, etc. That was in response to my post about the tricky nature of poverty lines. Ali’s comment is towards the bottom of that post.

Here’s my list of things to read in order to understand the concept of poverty lines better:

First, the World Bank itself:

Q: What is the new poverty line, and based on this new measure, how many people are living in extreme poverty in the world?
A: The new global poverty line is set at $1.90 using 2011 prices. Just over 900 million people globally lived under this line in 2012 (based on the latest available data), and we project that in 2015, just over 700 million are living in extreme poverty.
Q: Why raise the poverty line? What was wrong with the $1.25 a day line that we are all used to?
A: As differences in the cost of living across the world evolve, the global poverty line has to be periodically updated to reflect these changes. The new global poverty line uses updated price data to paint a more accurate picture of the costs of basic food, clothing, and shelter needs around the world. In other words, the real value of $1.90 in today’s prices is the same as $1.25 was in 2005.


That answer’s at least one of Ali’s questions – the poverty line is indeed updated to account for inflation. But keep in mind that indices are tricky little devils at the best of times. When it comes to measuring inflation, particularly for a basket of goods that the poor are likely to consume across different countries over long time horizons, it is all but impossible.1 But still, to the extent possible, poverty lines are adjusted for inflation.

But Lant Pritchett has his reservations:

There is no line at dollar a day in income dynamics. One might think a poverty line exists that demarcates a “poverty trap” and that people “in poverty” have a hard time escaping poverty—except that it doesn’t. All of the available evidence that tracks households over time finds enormous fluidity across the dollar a day threshold–and no evidence that it is harder to increase incomes from just below than just above—there is no line.


… and if you read the entire essay, it is hard to disagree with his point. He has an entire page on his website dedicated to talking about poverty lines, and the articles/videos deserve a closer look by any student of development economics.

I cheated a little bit while writing what I wrote above, because I wanted to give a chronological view of developments in the last decade or so. What’s written above is from 2013-14, and then this happened in 2017:

Starting this month, the World Bank will report poverty rates for all countries using two new international poverty lines: a lower middle-income International Poverty Line, set at $3.20/day; and an upper middle-income International Poverty Line, set at $5.50/day. This will be in addition to the $1.90 International Poverty Line – which remains our headline poverty threshold, and continues to define the Bank’s goal of ending global extreme poverty by 2030.

Let us be completely clear: The World Bank’s headline threshold to define extreme global poverty is unchanged, at $1.90/day. The Bank’s goal of ending poverty by 2030, and the United Nations Sustainable Development Goal 1.1, are both set with respect to this line. However – as Amartya Sen noted early on, and the Atkinson Commission reminded us – poverty is not a uniquely defined concept. There is an inevitable element of arbitrariness in choosing any poverty line, no matter how carefully it is constructed.


Here’s more from the World Bank, if you’re interested. And if it is data you’re after, try this.

Speaking of Amartya Sen, this paper is a classic, and a must read. Excerpting from this paper is difficult, please, read the whole thing.2

Before we get to India specific literature, a couple of see-also’s: the Wikipedia page on poverty thresholds (of course) and this lovely website that accompanies Martin Ravallion’s book.3

Now, about India specific work on poverty lines.

Here is a short summary by PRS India on the work on poverty lines in India.

Here is a useful survey paper on measuring poverty in India:

Poverty can be measured relatively, but a measure of absolute poverty is more useful for making cross-cultural comparisons. Unfortunately, the measurement of absolute poverty is difficult, because of inter-individual and intra-individual variations in minimum needs over time. As a result, simplistic assessment methods and confusion have marked many of the estimates of absolute poverty in less-developed countries. Using Indian material as an example, this paper attempts to trace the progress of the methodology; to explain how widely varying poverty estimates have come about; and to draw some tentative conclusions about the extent and pattern of absolute poverty in India today

Cutler, P. (1984). The measurement of poverty: A review of attempts to quantify the poor, with special reference to India. World Development12(11-12), 1119-1130.

Any GIPE student should know about Dandekar-Rath (part-II here), of course, but this paper goes a bit beyond, and is therefore a better introduction (in my opinion). Speaking of Dandekar, the original paper is worth reading, of course, but this is also an enjoyable read.

Finally, here is an excellent speech by Abhijit Banerjee on different (potential) ways to measure the poverty line.

And finally finally, also look up the FGT formula, and here are past posts on EFE that mention poverty lines.

  1. and with PPP considerations to boot![]
  2. If I had to excerpt something, the bicycle passage would have been my choice, but without context it is quite confusing.[]
  3. The explanation about the cover to the book alone is worth a click-through[]

EC101: Links for 26th December, 2019

  1. On some articles about Baumol’s cost disease.
  2. A topic that is very, very dear to my heart: teaching economics better, and to younger folks.
  3. A topic on which I changed my mind this year, and therefore this year ought to count as a success. Props to Murali Neelakantan for helping me do so! On patents.
  4. Two sets of links about this year’s Nobel. One set is rather informative
  5. While the other is more critical.

Etc: Links for 15th November, 2019

  1. Bibek Debroy about Abhijit Banerjee’s father. This was fascinating!
    “There were people who didn’t have an exceptional publication record. They were simply superb teachers.Dipak Banerjee was one of them. Except for a paper on utility he wrote while he was at LSE (London School of Economics), he rarely published. He was an exceptional teacher who produced exceptional students. Bhaskar Dutta, Subhashis Gangopadhyay, Dilip Mukherjee and Debraj Ray should be familiar names. They (all Dipak Banerjee’s students) edited a collection of essays in his honour in 1990. Mihir Rakshit primarily taught us macroeconomics and Dipak Banerjee primarily taught us microeconomics. Mihir babu’s teaching was precise. He never deviated from the topic. Dipak babu’s teaching was also precise, but he deviated from the topic and told us “stories”, especially at tutorials. In the course of these stories, we learnt he had two sons. He wasn’t worried about his younger son, who was “street smart”. But he worried about his elder son, who wasn’t that street smart. We learnt this elder son was called Jhima and that he had a middle name of Vinayak because he was born in Mumbai and because his mother (Nirmala Banerjee) was a Maharashtrian.”
  2. A short article about the “perils” of Amazon Prime
    “”Because of multiple Prime orders, Amazon has had to think more about packaging. Recognizing some customers’ “wrap rage,” they are using more bubble envelopes. Aware that the excessive space occupied by smaller inexpensive items increases transport costs, they’ve been developing algorithms that match box size to contents to avoid “over-boxing.” And they want manufacturers to know that online packaging needs to be compact rather than attractive.”
  3. “We therefore predicted that reactivating previously unsolved problems could help people solve them. In the evening, we presented 57 participants with puzzles, each arbitrarily associated with a different sound. While participants slept overnight, half of the sounds associated with the puzzles they had not solved were surreptitiously presented. The next morning, participants solved 31.7% of cued puzzles, compared with 20.5% of uncued puzzles (a 55% improvement).”
    Fascinating is an understatement – Alex Tabarrok on being productive while sleeping.
  4. “For me, sleep is the #1 important factor for my cognitive productivity. I typically get between 6½–7¼ hours per night. Much less, and I feel my brain turning to goo when I try to do anything cognitively demanding. I track my sleep with a fitness tracker so I can anticipate when I should expect a “bad day” and plan accordingly.”
    On the importance of sleep, and holidays. Please look up Jensen’s inequality as well.
  5. “…music streaming subscriptions are typically far cheaper in emerging markets than they are in the US and Europe, but hardware built to play that music – often from the very same companies running the music services – is significantly more expensive.”
    I pay 179 INR per month for Spotify – for six family members. INR 189 per month for YouTube Premium – for six family members.

The Nobel Prize in Economics, 2019

Abhijit Banerjee, Esther Duflo and Michael Kremer have won the Nobel Prize in Economics for the year 2019.

Does Giving Aid Work?

When I teach classes in development economics, I often speak about the Easterly-Sachs spectrum*. Without getting into long, boring details, here is the point: William Easterly is of the opinion that giving aid does more harm than good. At the other end of the spectrum is Team Sachs: no way are countries ever going to develop without aid.

Now you, I and everybody else we know may have an opinion about this spectrum, and we could spend the rest of our lives arguing about our opinions. But if the issue is ever to have a chance of being settled, we need evidence, one way or the other.

Where to get that evidence from? How?

This year’s Nobel Prize has been awarded to the three people listed above for their attempt(s) at answering these questions.

The issue is impossibly difficult to deal with. Say you give aid to a country, and say the country does well next year. How much of the improvement was because of your aid? How much of the improvement would have taken place any way? Might it be the case that the improvement would have been (horror!) even more had you not given the aid? What if you gave aid to improve, say, primary healthcare for youngsters, and educational outcomes improved. Did kids learn better because of the aid given for healthcare? And on and on, making the issue all but impossible to resolve.

Nobel Prizes ought to be given for making all but impossible to resolve issues tractable, and from that viewpoint, this is an excellent choice.

The way the Nobel Prize winners sought to resolve this issue was by setting up experiments. Or, to use the jargon du jour, by setting up Randomized Control Trials. Or Impact Evaluations, if you will.

What are RCT’s?

Malaria is a frustrating disease to think about, because one the simplest ways to deal with it is to prevent it. And prevention is, for the most part, simply caused most effectively by using mosquito nets. And so an NGO called TAMTAM started distributing mosquito nets in Kenya.

Would people buy the nets if they weren’t subsidized? What if they were partially subsidized? What if they were free? This example is drawn from Poor Economics, but the idea is very simple (to describe, at any rate!): find out what works by experimenting.

Whichever method works best, well, deploy it.

Banerjee, Duflo and Kremer have all run a series of experiments on a variety of issues to figure out precisely this: which method is working best? Set up, as carefully as possible, an experiment to find out what works, and to what extent. Read Alex Tabarrok’s post over on Marginal Revolution to get a sense of some of the experiments that have been carried out.

There have been experiments carried out on gender, on mosquito nets, on savings behavior: the works. If you are curious about what kind of experiments have been carried out, this is a useful website.

So, awesome! Right? Well…

A long(ish) video, but here’s one of the dissenting voices.


But still…

All of the above being said, RCT’s have been fairly interesting in terms of helping us understand how a specific part of the world works for a specific period of time.

More than that is difficult to say, but that doesn’t mean at least that much shouldn’t be said!

Additional Links

  1. A book about randomized control trials.
  2. The latest book by Duflo and Banerjee (not out yet!)
  3. The Twitter thread announcing the news, from the official Nobel Prize Twitter handle.
  4. Via Niranjan Rajadhakshya’s Twitter feed, an old profile he had done of Banerjee and Duflo.
  5. Here is an old talk given by Michael Kremer on RCT’s.

Thursday’s links, needless to say, will be other posts written about this years prize winners.

The Bottom Line

Work done by Banerjee, Duflo and Kremer has helped us uncover some surprising truths, changed our guesses about what would happen to y if x was done, and we have come along much more in terms of the science of setting up credible, well designed, carefully constructed experiments. Whether or not you agree with their current popularity, they have aided our understanding of the world we live in, and that ought to be celebrated.








*Duflo and Banerjee speak about this at the very start of their excellent book, Poor Economics, which is where I got the idea from, of course.