Finance & Development, September 2015, Vol. 52, No. 3

Alan Wheatley

PDF version

Sabina Alkire is passionate about measuring and eradicating poverty

There are many development economists, but Sabina Alkire is one of the few who is also an ordained priest. Alkire, the director of the Oxford Poverty and Human Development Initiative (OPHI), wears her religious beliefs lightly. In her small, functional office in Oxford University’s plant sciences building—no dreaming spires here—the only spiritual signifier is a mandala of the endless knot, one of the most auspicious symbols in Tibetan Buddhism.

“People who are deep friends think I’m daft to be a person of faith,” she says with an infectious, almost girlish, laugh. “I don’t see any distinction between myself and an atheist or humanist friends. We’re all coming out of a passion.”

Yet as she discusses the multidimensional poverty index with which her name is associated, it is clear that she is driven by more than a purely academic passion to observe and better measure poverty as a precondition for eradicating it. “We who work in development service, there’s some deep commitment to humanity and to justice—even if the faith is different and even if the doctrine is different,” she says.

A detailed picture

The aim of OPHI’s index is to supplement the traditional benchmark of income poverty, $1.25 a day, by painting a more nuanced picture of exactly how people are poor in different parts of their lives. “You need both of them to get a good read on poverty,” Alkire, a U.S. and U.K. dual citizen, says.

The global Multidimensional Poverty Index (MPI), based on household surveys, consists of 10 weighted indicators in three areas: health, measured by nutrition and child mortality; education, with years of schooling and school attendance as proxies; and living standards, assessed by access to electricity, sanitation, water, type of floor, cooking fuel, and ownership of basic assets. Anyone deprived in one-third or more of the indicators is deemed “multidimensionally poor.” The June 2015 global index covered 101 countries with a population of 5.2 billion people, about three-quarters of the world total, and found about 30 percent of them on average to be MPI poor. Alkire’s team—and she is at pains to stress that this is a team effort—has broken down the findings into 884 subnational regions, providing information that national averages would miss.

One of the attractions of OPHI’s MPI is that governments can tailor the methodology of the index to their own circumstances, for example by adjusting the weights and cut-off thresholds. Indeed, Alkire says her team spends most of its time now not on the global index but on national MPIs. Because the MPI can be broken down by indicator, policymakers can not only see the headcount poverty ratio but also zoom in on how different categories of the population, say by region or ethnicity, are deprived. In other words, the index captures both the incidence and the intensity of poverty at the household level in different dimensions, thereby helping governments to target policy.

“These two components allow people, especially policy analysts, to get better insight into the poverty of a country by making comparisons over time, better insight into the dynamics, and so on,” says Milorad Kovacevic, chief statistician at the Human Development Report Office of the United Nations Development Programme (UNDP) in New York. For its annual flagship Human Development Report in 2010, the UNDP replaced its human poverty index with an MPI constructed at OPHI by Alkire and Maria Emma Santos, now an assistant economics professor at the Universidad Nacional del Sur in Bahía Blanca, Argentina.

In 2014 the UNDP started calculating the index independently of OPHI because of some methodological differences. But they have since made up and have agreed to produce a single MPI again in 2016. “They are great colleagues and it will be nice to be working together again,” Alkire says.

Bhutan, Chile, Colombia, Mexico, and the Philippines have already adopted official national MPIs to help allocate resources and measure whether policies are being implemented appropriately. Several other countries, including Tunisia, are getting ready to follow their lead.

Colombia is using its MPI to guide the country’s 2014–18 national development plan, according to President Juan Manuel Santos. “The fight against multidimensional poverty is harder but much more effective,” Santos told the third annual meeting of the Multidimensional Poverty Peer Network, held in Cartagena, Colombia, in June. The group, a South-South initiative, counts officials from 40 countries.

Alkire says the index must not “sit and gather dust.” She wants it to be part of a data revolution to guide the fight against poverty. “What I really love is that we work with passionate and committed people who take the measurements and use them for policy. We’re in a very dynamic and creative phase right now,” she says.

If the multidimensional poverty index sounds as though it owes a lot to the work of Amartya Sen, that’s because it does. Poverty, Sen wrote in his 2000 book Development as Freedom, must be seen as “a deprivation of capabilities, rather than merely as low income.” Alkire says she regards the Indian economist very much as her mentor. Indeed, he had agreed to be her doctoral examiner but had to withdraw because of time pressure after winning the Nobel Memorial Prize in Economic Sciences in 1998.

Her bookshelves bulge with his works. A flier for a lecture he gave in Oxford in 2013 sits atop her filing cabinet, and a banner outside her office bearing a quotation from Sen’s “Possibility of Social Choice” lecture sums up OPHI’s mission: “How can it be possible to arrive at cogent aggregative judgments about poverty given the diversity of preferences, concerns and predicaments of the different individuals within the society?”

Sen is an adviser to OPHI and Alkire is in touch with him quite often. “But he’s not hands-on, in the sense that the MPI is our work,” she says. “He might be critical of it. He might take a different idea. So he gives a lot of freedom to people who try to develop his work. He’s not at all trying to direct it.”

A roundabout route

Alkire took a somewhat roundabout route before arriving at poverty measurement. She was born in 1969 in the German university town of Göttingen but left as a baby when her father took a job teaching chemical engineering at the University of Illinois at Urbana-Champaign. Alkire graduated in 1989 from the same university in sociology and premed studies and was accepted into medical school at Johns Hopkins. During her “gap year,” however, she decided not to take up the place, to avoid getting into debt.

That year included three months as a volunteer with a nongovernmental organization that conducted systematic immunization among Afghan refugees outside Peshawar in Pakistan. “I was living with a family. Their son was studying at my university and the sisters were my age, so I got completely into the culture and the language and loved Islam,” she recalls. In India she spent several weeks at Mother Teresa’s home for the dying in Kolkata, visited Tibetan settlements in Himachal Pradesh, and worked in Sri Lankan refugee camps in the south of the country. “It was the normal student gap year experience of simply trying to absorb as much as possible.” And then Alkire decided to branch out into theology. “I didn’t know what it was, so it was quite an odd choice,” she says, disarmingly. “I had a deep faith and still have—it’s a big part of my life. I wanted to learn about God and I realized that’s not what theology is. It seems to be about the study of texts that have to do with God.”

Alkire completed a Diploma in Theology at Magdalen (pronounced maudlin) College, Oxford, in 1992, but says it was a close thing: “To be honest, I nearly failed doctrine—I never grasped very clearly the difference between evil and suffering.” But she did earn a distinction mark for her Islam paper. “It may have been because I loved the course and also had been learning Koranic Arabic so could put bits in in Arabic.” Alkire went on to do an MPhil in Christian political ethics, seeking to explore ways to be useful in development while respecting other people’s cultural and spiritual values. It was not, she puts it diplomatically, “a match made in heaven” for Oxford’s theology department. Her 1994 master’s thesis, “The Concept of Poverty Alleviation in the World Bank since 1990: A Theological Analysis,” was too modern for the department’s liking and she was turned down for a PhD program. So she switched again, this time to economics, at the urging of Rosemary Thorp, then reader in the economics of Latin America, whom Alkire describes as a “wonderful woman.” Alkire says the progression was natural because she had been auditing classes from econometrics to development theory, finding them “very interesting and easy,” and was already delving into Sen’s work. By 1995, she had a master’s under her belt and by 1999 a doctorate followed, both on Sen’s capabilities approach to development.

After joining the World Bank, where she worked from 1999 to 2001, Alkire undertook ordination training, which is how she met Edmund Newell, then chaplain to the bishop of Oxford. Whereas Alkire had migrated from theology to economics, Newell studied economics and economic history at Oxford before becoming a priest in the Church of England. They went on to coauthor a book, What Can One Person Do? Faith to Heal a Broken World, which examined the United Nations Millennium Development Goals through a theological and practical lens.

“She treats her economics as vocationally as she treats her Christianity. She is totally devoted to it—to the people behind it. That was one of the things that came across really strongly working with her,” says Newell, who now heads an education charity and study center outside London.

“This was not about abstract economic theory. It was about practical ways of helping the world’s poor. That shines through in the way we worked and shines through in everything she does,” he adds.

Alkire had wanted for some time to apply Sen’s capabilities approach to gauging multidimensional poverty, and a breakthrough finally came in 2006, when she started to collaborate with James Foster, now a professor at The George Washington University in Washington, D.C., and a leading figure in measurement methods. While a graduate student at Cornell University, Foster in 1984 developed with Joel Greer and Erik Thorbecke (also at Cornell) the eponymous FGT set of poverty indices, which are still extensively used to measure single variables such as income, consumption, and calorie intake. Foster was at first “quite dubious about everything multidimensional,” according to Alkire, who was setting up OPHI at the time. “But we had a week of just head-on talking it through,” she recalls. “Then, I guess, the penny dropped and we saw what was possible in terms of the methodologies that built on his work,” she recalls.

The Alkire-Foster methodology underpinning the MPI stemmed from that marathon brainstorming. Alkire says she has learned a huge amount from Foster. “He has had more of an orientation towards theoretical work. When it comes to proofs, he can do those. It’s sort of a language I understand but can’t speak,” she says. Cue more laughter. “I try, and I’m trying to learn more, but he’s head and shoulders above me in that.”

A stern critic

Alkire says that learning to meditate soothed away what she describes as the “temper tantrums” of her childhood. Still, her equanimity must have been tested in 2010 when word reached her in Bhutan—during a meditation retreat, no less—that Martin Ravallion, then a leading World Bank researcher, had torn into the new enthusiasm for multidimensional poverty indices. “They had to call me to walk three miles to the hotel and find the Internet to respond,” Alkire remembers.

Taking aim at the UNDP’s new MPI, Ravallion argued that it was simply not credible to suppose that a single index could capture all the dimensions of poverty. “We can all agree that reducing child mortality is a hugely important development goal, but how can one contend . . . that avoiding the death of a child is equivalent to alleviating the combined deprivations of having a dirt floor, cooking with wood, and not having a radio, TV, telephone, bike, or car?” he wrote.

Five years later, Ravallion, now an economics professor at Georgetown University in Washington, D.C., is still a stern critic. For him, distilling various measures of poverty into a single index is akin to combining all the dials and gauges in a car into just one instrument. “The human development dimensions of welfare are just so important for assessing social progress, but I don’t want to add them up in some composite with material goods,” he says. “I want to look at them separately. I want to see where a country is doing well or not.” Aggregating different indicators into an MPI could be positively harmful if policymakers do not understand the trade-offs that are built in to the index, in Ravallion’s view. “I would give a strong warning to handle with care,” he says. “Governments look at these indices. They don’t know what went into the stew that made them. I don’t think that makes good policymaking.”

It’s an understatement to say that policy analysts like multidimensional indices such as OPHI’s MPI more than the economics community does. In addition to the subjectivity entailed in choosing the components and their weights, the index is open to criticism that it relies on international surveys and indicators that may not be applicable in every country. Is it appropriate, say, to measure poverty in Africa the same way as in countries of the former Soviet Union? Statisticians are also sniffy about the headline MPI, which uses ordinal not cardinal data. “The purpose of all these composite indices is communication. No one is pretending that they are very precise,” the UNDP’s Kovacevic says. “But if people are curious and need to know why things went in one direction and another country’s went in another, they can pull down the index and look at the components.”

Charles Kenny of the Center for Global Development, a Washington think tank, agrees with Alkire that it is misleading to view the battle against poverty exclusively through the lens of income. Countries such as Haiti and the Democratic Republic of the Congo have recorded modest improvements in child mortality and education even though incomes have stagnated, says Kenny, author of Getting Better: Why Global Development Is Succeeding—and How We Can Improve the World Even More. But he too has misgivings that the MPI might be substituting one politically convenient catchall figure—$1.25 in income a day—for another. “One thing I wonder about is that, after arguing that poverty is multidimensional, the MPI ends up as a single number.”

Another issue is that any index is only as good as its underlying data, and in emerging market economies that quality is often inadequate. “The knowledge problem in development is twofold: we know less about poorer countries and less about the poorer people in poor countries,” says Morten Jerven, author of Africa: Why Economists Get It Wrong. For example, on-the-ground surveys to check whether households really do have effective access to electricity—one of the MPI indicators—are simply not frequent enough. “The MPI does some disaggregating and does it a little better than some other measures,” Jerven, an associate professor at Simon Fraser University in Burnaby, British Columbia, says. “But if the numbers that go into this are not updated often or based on real observations, these trends may be meaningless.”

The quest for better poverty metrics coincides with growing doubts about the ability of conventional statistics, especially GDP, to gauge economic growth in the digital economy, let alone well-being, welfare, and environmental sustainability.

“In order to accurately measure our progress towards sustainable lifestyles, we feel we need to get beyond GDP measures,” Gudrun Kopp, Germany’s parliamentary state secretary for economic cooperation and development, said in explaining why Berlin had begun to support OPHI’s work on multidimensional poverty.

Diane Coyle, author of GDP: A Brief but Affectionate History, advocates tracking a dashboard of indicators that contribute to social welfare, such as the Organisation for Economic Co-operation and Development’s Better Life Index. “There is a counterargument that you need a single number because that’s what voters and politicians pay attention to,” Coyle says. “But that, though attractive, just submerges some of the trade-offs that policy has to be about.” Alkire acknowledges the usefulness of a dashboard approach but suggests it should include the MPI. “What we don’t agree with is that you should only look at deprivations one by one and never see who is deprived in many at once,” she says.

OPHI’s website commendably provides links to papers and blogs critical of the MPI. “We’re not out of the water because there’s still a lot of resistance to our work,” Alkire says. The OPHI team itself “fights vigorously and happily” over how to make the index better. “Nothing is beyond criticism, and certainly I have plenty of criticisms of my own of the index and the methodology,” she adds. Not all data are comparable or up to date, and there are holes in subnational regional surveys. But the index is pretty good and getting better fast as the quality of the survey improves. “There’s a whole set of criticisms I understand and respect. On the practical side, though, I disagree with them,” she says firmly.

To-do list

So what next for Alkire? She is determined to keep improving the MPI and to broaden its coverage to measure employment, empowerment (“my passion”), and violence. She would love to produce an index that sheds more light on the way women live and are treated, but such an initiative would need broad political backing. “You can’t get too far ahead of where countries are,” Alkire says. “This is a decision the international community has to take. If they want a gendered index, we know how to do it, but we don’t have the data.”

It was a misconception that measuring multidimensional poverty requires a lot more data gathering than tracking income or consumption poverty. For instance, the MPI draws on only 39 of 625 questions in the U.S. Agency for International Development’s Demographic and Health Surveys, one of OPHI’s main sources (UNICEF’s Multiple Indicator Cluster Surveys is the other). Little tweaks to the questions would suffice. “Just a few seconds and you have much better information,” she says.

In September 2016 Alkire will begin directing her program from the United States, where she has accepted a full-time professorship at The George Washington University—alongside James Foster. “GW has been kind to offer a gentle and wise transition,” she says. However, research into measuring poverty will continue at Oxford in one form or another. “How that will actually unfold we have no idea,” Alkire says. “I simply want to be able to continue working on this agenda.”

In the meantime, she has enough on her plate. The church is keeping her away from her hobbies of cooking, exercise, and meditation. In addition to her duties as honorary chaplain at Magdalen College, she has been heavily committed to her parish of Cowley St John in East Oxford, which has been without a vicar for far too long. “It’s been a year, but it’s certainly changed my life. I have to be home most Sundays. It’s weird,” she laughs. God’s work, it seems, is never done. ■