Toward a better understanding of poverty (wonkish)

Allow me to get my wonk on. It's becoming an increasingly mainstream idea that our traditional poverty measures just don't cut it. For years the UN Human Development Report has been playing with alternative measures. Well, this year we get the latest: the Multidimensional Poverty Index (MPI). In their own words, the new index "measures the combination of deprivations that each household experiences. The MPI uses microeconomic data to reflect the percentage of households that experience overlapping deprivations in three dimensions—education, health and living conditions." This month we get a lil taste of what's to come. The Oxford Poverty & Human Development Initiative (OPHI), with the support of the United Nations Development Programme, have released some preliminary results (104 developing countries) in advance of the upcoming 2010 Human Development Report to be released in the fall. 

First off, a little bit more about the MPI. The idea is that by looking beyond mere income poverty, one can ascertain a better understanding of the specific things (health, education, sanitation, electricity), that contribute to the overall picture of poverty. Governments, NGOs, or even random bloggers can then look at the data to figure out what the greatest needs are, and where it makes sense to intervene. The MPI also calculates the intensity of poverty by looking at how many different measures a certain population is lacking. As OPHI notes, "A person who is deprived in 70% of the indicators is clearly worse off than someone who is deprived in 40% of the indicators."

Sounds good in theory, but what about in practice? Well, some countries have already adopted a similar index, including Mexico. The index can be tailored to individual countries specific circumstances, different countries have different needs, different areas in need of improvement. In Mexico for instance the indicators used are: Current income per capita, Education, Access to healthcare, Access to social security, Housing quality and space, Basic services in homes, and Access to food. In December 2009 Mexico became the first country to implement a form of the MPI, with some astonishing results. As OPHI writes:

The results show that there is a striking contrast between deprivation in income-only and the multidimensional measure: only 1.2 per cent of indigenous people are vulnerable strictly in terms of income. Even across the entire population, only 4.4 per cent of Mexicans are income vulnerable only, whereas 44.2 per cent live in multidimensional poverty.

The measure distinguishes between a household which is poor in one dimension and one that is poor in several dimensions simultaneously. It is also decomposable by population group (indigenous/non-indigenous, over 65, under 17, etc) and by geographic regions. 



For example, comparison of Mexico City and Oaxaca shows that households in Oaxaca are more deprived in terms of basic services at home, but residents of the Mexico City are lacking in healthcare access. Nationally, the rate of extreme multidimensional poverty (defined as at least three deprivations plus insufficient income) is 10.5 per cent with an average of 3.9 deprivations, whereas among the indigenous people of Mexico the rate of extreme multidimensional poverty is 39.2 per cent with an average of 4.2 deprivations.

Some really interesting and positive findings. Positive in the sense that one can gain a much more accurate picture of what the problems are, and where they are. I would think that these sorts of measures would gain in popularity pretty quickly, maybe especially in Latin America. Venezuela, although not included in the preliminary findings, would stand to be an ideal candidate for this measure. Poverty measures in Venezuela largely rely solely on income, but over the last decade access to health care, education, food, etc. have, by and large, all increased dramatically. A new poverty measure may more accurately reflect that reality.

In any case, onto the data. One thing to remember though is that not all these numbers are from the same year. Bolivia for instance is calculated based on 2003 data, Argentina with data from 2005. Also, for each country measured there is a more detailed country breakdown, so if you got a hankerin' to see the complete breakdown of poverty in, say, Burkina Faso, look no further. Perhaps the coolest thing about the data is the interactive map that lets you scroll over countries, sort the data every which way, and is pleasing on the eyes as well. But, since it's not embeddable, we'll just give you less pretty, but just as useful charts. I suggest checking out the site though and playing with the data yourself.

Let's take a look at the three dimensions the MPI looks at, deprived in education, deprived in health and deprived in living standards. But first a look at the percent of people who are MPI poor, after being put through Maladjusted Charts™:

Pretty interesting stuff here. First thing that jumps out at me is the low levels seen in many countries, especially Ecuador for example, with just 2.2% of people MPI poor (2003 data). Compare this to the 51.2% poverty rate from 2004, based on CEPAL's numbers. My gut reaction is that this could be counter-productive by making it look as though poverty isn't as big an issue, although maybe it's just a reflection of some really outstanding social policies. On the other end of the spectrum, free market darling Peru has an outstanding 19.8% of her people MPI poor (the data is from 2004).  This puts Peru closer to the Central American countries than to most of South America.

Next, we'll take a look at the breakdown of those "poor and deprived in education":

So for instance Brazil, who is at the lower end of MPI poverty rates, still has a significant problem with access to education. This is the type of analysis these alternative measures allow. On the other hand, it seems like Peru's high MPI poverty rate is not being driven by a lack of access to education, but perhaps health? Let's take a look:

So, it doesn't look like access to health care is driving Peru's high MPI rate, must be the final measure, "poor and deprived in living standards". Colombia, where there is currently a pretty large debate over health care, ranks pretty poorly here with 17.5%. Argentina, middle of the road for education, has the top rank in health. Finally, a look at living standards, which includes things such as electricity, sanitation, having a floor, cooking fuel, etc.

Indeed, here we see that Peru's high MPI rate is largely driven by deprivation in living standards, even topping Bolivia and Honduras in this category. Although for the most part you can tell what is causing the MPI rates by looking at these three measures, the specific country breakdowns provided by OPHI are even more detailed and contain pretty lil pie charts like this one, for Peru:

You might need to click that for a larger image, but you can see the green shaded area (living standards) makes up the largest contribution to MPI. Although it's pretty evenly distributed, a lack of cooking fuel is a significant driver. Overall, a high level of child mortality is the leading contributor to Peru's MPI. Each country that the researchers look at are broken down like this, so I chose to look at Peru, but go to the site and pick your poison. The country breakdowns also include comparisons to the national poverty measure, and other measures like % living on $2 a day, $1.25 a day, etc.

There's a lot to chew on here, and plenty more you can find out over at the database, but definitely some food for thought. I'm not sure if this is the best measure, and as the example of Mexico shows, it may be more beneficial to tailor these indexes on a country level, although this would sure make comparisons harder. What is clear is that how one calculates poverty has a significant impact on the results, and it would certainly behoove us all to have a more accurate picture of what poverty is, and how to combat it. I'm sure there are flaws here, not the least of which is the outdated data, but I commend the researchers at OPHI and UNDP with a solid step in the right direction.

(cartoon from Polyp, check out the website for more)