Taking Global Poverty Stats for a Joyride

Dr Scott Wisor

Earlier this month, some very committed and talented people at one of my favourite think tanks, the Centre for Global Development, took global poverty estimates for a bit of joyride. (A similar effort occurred at the Brookings Institution, but I will focus on CGD here). The occasion was the release of the International Comparison Program’s latest round of price information. To compare economic activity across countries, it is necessary to compare the purchasing power of different currencies. This information cannot be gleaned from currency exchange markets, which fluctuate wildly and don’t tell you anything about how much you can buy once you change your pounds into rupees and arrive in India to spend them. To compare purchasing power across countries, the International Comparison Program tracks prices for a huge range of goods from all across the globe. By comparing like to like, you should be able to determine how many rupees are required to purchase the same quality items you purchase in the UK.

If you think the World Bank’s method of calculating absolute poverty at USD 1.25 per day is a reliable way of tracking the levels and trends of global poverty (I don’t), you might think a quick update on global poverty statistics based on the latest ICP is in order. Although the method used by researchers at CGD and Brookings cannot give any meaningful interpretation of the state of global poverty, they nonetheless gave it a try. By using the new ICP data, the authors estimated new poverty counts for all countries and found a considerable drop in the number of poor people. However, these efforts mistakenly assume that the new price information applies to national consumption and income figures but not to the setting of the poverty line itself. (Instead, the authors simply update the global poverty line using US inflation). As Martin Ravallion, former architect of the Bank’s poverty assessments notes, this is misguided and deceptive. The poverty line itself needs updating using the new price information, rather than simply updating the line using the US consumer price index. Kashuk Basu ,chief economist at the World Bank, also cautioned against using the new ICP data for international poverty comparison.

I don’t like the Bank’s method for measuring global poverty, and its flaws are inherited by other efforts undertaken by Brookings and CGD. Briefly, the method for setting the poverty line is implausible (an averaging of the poorest countries national poverty lines even though those poverty lines may not be plausibly anchored or even democratically discussed), and the method of making comparisons across contexts and over time takes account of the amount and prices of goods consumed across the whole economy, rather than focusing on those goods that are most consumed by poor people. This greatly risks understating the extent of global poverty. Why? Because if the prices of basic necessities rise (for example, the price of food) while the prices of goods consumed by middle and upper classes fall (for example, the price of big screen televisions), both will affect calculations of global poverty, even though only one is relevant to determining how well or badly poor people are faring.

One way to independently assess whether the global poverty figures are giving good information is to examine other indicators of human deprivation that are not related to figures on income and consumption. For example, how did it come to be that only 8.3% of Indians are poor after the new poverty calculations, but according a recent report on India’s children “approximately 60 million – are underweight, about 45% are stunted (too short for their age), 20% are wasted (too thin for their height, indicating acute malnutrition), 75% are anemic, and 57% are Vitamin A deficient.” Surely it cannot be that a child is at once free from poverty and unable to meet minimal nutritional requirements to grow.

Consider how a single country was treated in these revisions. The Philippines’ poverty rate plummeted dramatically according to CGD’s PPP revisions. Only 2.5 million people (about 2.6%) in the country are poor. On the Philippines’ own national poverty lines (one based on the cost of only simple foodstuffs, the other based on the cost of basic foodstuffs and other basic needs) the figures are much higher (10% and 24%, respectively). Using the multidimensional poverty index, the Philippines poverty rate is also higher than CGD estimated. In measurements using the Individual Deprivation Measure developed with several colleagues over the last four years (to be debuted later this summer), we found in a nationally representative survey that 48% of Filipinos count as deprived, very deprived, or severely deprived. Using the World Bank’s old method of calculating poverty at the higher 2.50 USD 2005 PPP line, 41% of Filipinos are poor. When CGD pronounces that almost no one in the Philippines is poor, as opposed to half the populations as other measures indicate, then one gets the impression that poverty is a small problem.

In my view, it is professionally irresponsible and morally objectionable to proceed with analysis of global poverty statistics in a way that implausibly suggests poverty is in fact much less of problem than previously supposed. (I equally think it is irresponsible to overstate the extent or trend of global poverty, as much commentary does). Why? Because how people perceive global poverty matters. If it is a problem that is not so big after all, and diminishing over time, then it is not very urgent that people take steps to end it.

Once this information is posted, it spreads. The Washington Examiner (as one example) ran an article headlined ‘Number of global poor falls in half’. Well, that is great news. Except that it probably isn’t true.

The CGD blog post ends on a puzzling note. They write: “Perhaps most importantly, it is worth repeating what didn’t change between Tuesday and Wednesday.  The people who have just been classified as ‘not absolutely poor’ don’t actually have any more money than they did yesterday, and will still struggle in terms of getting a decent job, and many still face grim daily tradeoffs between buying school supplies or ensuring their kids are well nourished.  In fact, if the new PPP numbers suggest anything it is that the quality of health or education or access to services associated with a given income has just gone down. The new numbers are good news for people who care about poverty, but they matter much less for people who are poor.”

But there aren’t two moral perspectives here, one the view of the poverty statistician and the other the view of the poor person. The only morally relevant consideration is the deprivation and hardship that people face. It is hard to see what good came from proclaiming that much of this deprivation and hardship is not evident in a rough calculation of new poverty numbers. Better to not make any pronouncement at all on the implications of the new price comparisons than to proclaim rapid poverty reduction as a result of calculations which are arguably mistaken and misleading.



Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: