Thursday, March 25, 2010

Good News For Africa: The Poverty Rate has Fallen

Good News For Africa: The Poverty Rate has Fallen
“African poverty is falling”, so says Xavier Sala-i-Martin and Maxim Pinkovskiy in a recent NBER paper (National Bureau of Economic Research, working paper 15775, February 2010).
The study of poverty in general and African poverty in particular goes in cycle. During the Johnson and Nixon administrations the economics profession devoted a great deal of efforts to measure US poverty rates and to come up with solutions to poverty. These solutions ranged from a negative income tax (Milton Friedman) to the family assistance plans.
Measurements of poverty however, have remained an engaging subject for many in the economic profession and especially for those at the University of Wisconsin--Madison Institute for Research on Poverty, the Urban Institute and, the Brooking Institution.
The economics profession’s interest in the study of poverty although was dominated by the study of US poverty, development economists notably Amartya Sen have exerted great efforts in measuring poverty in developing counties, identifying its causes as well as advancing solutions to address it. Such efforts notwithstanding, eradicating or substantially reducing poverty here and elsewhere has remained beyond the grasp of politicians and their economic advisors.
In the past couple of years, measurement of poverty has taken central stage. For the most part, the focus was on US poverty (see AEI, “The Poverty of the Official Poverty” by Nicholas Eberstadt, November 2008, and Brookings, “Improving the Measurement of Poverty”, December 2008). Poverty rates and “poverty line” in the developing world and especially in Sub-Saharan Africa did not escape the attention of economists motivated in part by the goal set in the 2008 Millennium Development Goals Report (MDG) (UN, 2008). The report has set a target for reducing developing countries’ poverty rates by the year 2015. The MDG report was not very optimistic about reducing Sub-Saharan Africa poverty rate by that date citing many factors that hinder growth.
In his presidential address to the American Economic Association (January 17th, 2010), Professor Angus Deaton of Princeton University provided an exhaustive analysis of world poverty and inequality. Most of the discussion however was devoted to the role played by the purchasing power parity (PPP) price indexes in the measurement of global poverty and global inequality.1 The Deaton address (paper) is 60 pages in length and for the non professional economist it is a bit “boring” and cumbersome. The critique aside, the general tenet of the paper and hence comparisons of poverty rates hinges on the revision of the PPP indexes reported in the International Comparison Project (ICP). If one were to believe in the soundness of the ICP revision of the purchasing power parity.
Thus according to the ICP, “Global inequality has increased and this reduced the global poverty line relative to the US dollar”.
Before getting into the “nitty-gritty” of measurement, few numbers may be helpful. Deaton provides in Table 1 of the paper, poverty head count ratios and the global poverty line expressed in year X (say 2005) PPP international dollars. The information reported are note worthy in that it contrasts changes in poverty ratios since 1981 (¬¬¬¬Three data points are given 1981, 1993, 2005). Measurement of the percentage of people in poverty in each of these years has to be calculated on the basis of the poverty line derived from the PPP at a given date. Thus, using the year 2005, the poverty line is defined by a number, in this case $1.25. This means that if a person in country Y for example Ethiopia or Nepal received $1.25 a day in income (whatever the source), then he is not counted as poor. At $1.24 he is counted as a poor person. The poverty line hence is used to count the ratio of the population that is labeled as “poor”. Given the significance of the poverty line to the count of people who are in poverty, Deaton’s 60-page-paper engages the reader into an explanation of what this number means for the count of people in poverty and hence changes in global inequality over time.2
Not to get tangled in the “details”, the paper provides a “powerful” message. Global inequality has fallen over the past two and a half decades. The total percentage of people in poverty (using the $1.25 poverty line) has fallen from 51.9 percent in 1981 to 25.2 percent in 2005. This is good news to be sure. How do people in one country or a continent fare against people in another or others depend on where they live. For example, contrast the poverty profile of a person in East Asia Pacific with that of a person in Sub-Saharan Africa. In the year 1981, 77.7 percent of the people in East Asia Pacific were classified as poor at the $1.25 poverty line. This percentage has fallen to 16.8 percent, a decline of almost 80 percent.
Compare this rate of decline with that for Sub-Saran Africa. Again, at the $1.25 poverty line measure, the percent of population that would have been classified as poor in 1981 was 53.4 percent. By 2005 this percent has fallen to 50.9 percent, a decline of almost 5 percent—Great news for Sub-Saharan Africa.
Looking at the count rather than the percentage, the global poverty (measured at $1.25 poverty line), has fallen from a count of 1,900 millions in 1981 to 1,374 million in 2005, a fall of more than 27 percent in a two and one half decade. That is not bad or is it?
Deaton‘s paper goes beyond the statistics I have cited. It is worth the time and effort to expend to understand the magnitude of the problem. One issue that is worth investigating is the sample entities, the size of the population in each member of the sample and the economic progress attained there.
Back to Sub-Saharan Africa. As mentioned earlier, Deaton tells us that the poverty rate there has fallen from 53.4 percent in 1981 to 50.9 percent in 2005. Xavier-sala-i-Martin and Maxim Pinskovskiy (NBER, 2010) are much more upbeat about the prospects of poverty reduction in Sub-Saharan Africa. The authors focus exclusively on African poverty using three poverty lines; daily income of $1, $2 and $3. Translated into yearly income in US dollars, they get a value of $365 in 1985 dollars for a poverty line of $1/day. The authors provide extensive data analysis of the African distributions of income for 4 data points: 1970, 1990, 2000 and 2006. Because of this selection and the choice of the poverty lines, values, their findings are not totally comparable to those of Deaton. Nonetheless, their results on changes in the African poverty rate between 1970 and 2006 at $1/day poverty line augment what one learns from Deaton’s.
Taking the range of $1/day and $2/day poverty lines one may be able to contrast their findings with those reported by Deaton. African poverty rate for 1981 at $1/day is given as 39.4 percent. At $2/day, the rate is 64.8 percent. The number reported by Deaton is somewhere in between, at 53.4 percent for $1.25/day. The difference is not unreasonable. In 1993, according to Sala-i-Martin and Pinkovskiy the poverty rate (surprisingly) rose to 42.2 percent at $1/day and to 67.1 percent at $2/day. For 2005, the respective rates are 33.1 percent and 60.9 percent. Over the 1981 to 2005 period, the decline in the poverty rate measured at $1/day is 15.9 percent but at $2/day it is only 6 percent, which is a bit closer to Deaton’s rate of decline of 5 percent estimated at $1.25/day.
How great is the decline? Measured at $2/say, it is 6 percent over two and half decades. Good news but nothing to write Home about.
Fortunately, the story does not end there. Sala-i-Martin and co-author Pinkovskiy predict that by 2015 the $1/day poverty rate will be 22.8 percent, a decline of 10.3 percent over 10 year period (from 33.1 percent in 2005), a one percent reduction per year. That is indeed remarkable if true. What makes it so is that it is within the range reported for the previous 14 years where the rate has fallen 1.4 percent per year from 1981 to 2005. With such progress the authors predict that the poverty rate of 21.0 percent by 2015 set by the 2008 Millennium Development Goals (UN, 2008), is attainable but perhaps in the year 2017 rather than 2015.
What accounts for the poverty rate decline? Whether one accepts the numbers given in Deaton or Sala-i-Martin and Pinkovskiy the answer is simple and expected: a “decent” growth rate of per capita Gross Domestic Product (GDPP). The correlation between the two holds for all countries, whether the country is developed or developing in Africa or in the EU and the US. The question that has and continued to be asked is what accounts for economic growth and whether observed growth in any country over any period is sustainable. This is a topic for another essay. For now it suffices to point out that the driving force in the evolution of poverty “is an almost exact mirror image of the evolution of GDP per capita” (Sala-i-Martin and Pinkovskiy, p.10). If growth is sustained then the authors’ prediction for 2017 will be achieved, if not the MDG target will neither be met in 2015 or 2017.
Authors’ prediction clearly hinges on the state of affairs in the world following the global financial crisis of 2008. But there is room for hope as the world economy seems to be moving, albeit slowly towards a decent recovery.
A final note which I shall revisit in the future. The notion that $1.25/day, $2/day or $3/day as a benchmark for counting people in poverty worldwide is a “bare” and “unfeeling” statistic. Think about what $1.25 buy an individual in Timbuktu, Ethiopia or Nepal. The bold hard fact does not convey what that individual be a male, female, adult, child or an elderly consumes in terms of goods and services. The number is said to be obtained from household surveys. Given that household survey data are available what would be most enlightening is to use it to show the composition and the size of the basket of goods which the $1.25 buy in each of these countries compared to the “standard” for a decent living. Undoubtedly, one can set the rate at $5, $6, $7 or what a decent living implies.
Deacon gives us a glimpse of what a decent living means derived from data on well being measures from the Gallop World Poll (Table 8, p. 55). Using the year 2006 as the base, 37.9 percent of the world population reported “poor living standards” with the percentage rising to 38.6 in 2009. For Sub-Saharan Africa population the corresponding number is 61.4 percent in 2006 rising to 62.6 percent in 2009. These numbers tell a different story, perhaps closer to what poverty means than the sheer count of the numbers in poverty.
One can easily define what a decent living standard requires (for example BEA basket of goods and services used to determine the cost of living index in the US). What is not easy is how to get there. If indeed the world organizations and we as individuals in affluent societies care about global poverty then efforts should be exerted to enable Sub-Saharan African countries (the region with the highest poverty rate) to make substantial advances in reducing poverty. We should not be satisfied with 5, 6, 10 or even 15 percent reduction in the poverty rate over a decade with a standard of living set at $1 or even $2 a day.
---------
1 PPP are needed to translate purchasing power in one country’s currency into US purchasing power. In other words, if a good costs 16 rupees in India, with PPP exchange rate, the cost is expressed in terms of dollars.

2 To get a feel of the numbers in a comparative setting, the poverty rate in the US in 1981 was 14.0 percent at $12.65/day. In 2005 the corresponding figures are 12.6 percent at $27.32/day.

No comments: