Of all the developing regions in the world, the severity of poverty and malnutrition is greatest in sub-Saharan Africa, where it is also increasing at the fastest rate. While it can almost be taken for granted that faster and more stable economic growth will eventually contribute to an alleviation of poverty, immediate complementary actions are necessary to meet the needs of the poor.

The question is, what kind of actions would be most effective. The answers to that question require a thorough understanding of the characteristics and determinants of poverty. AERC is seeking to contribute to that understanding and find some of those answers.

The quest stems from a major recommendation flowing from the 1995 evaluation of the AERC research programme: To add a new theme and corresponding research group within a nexus of interrelated components consisting of poverty, employment, labour markets, human capital and the fiscal role of government with reference to human resources. The review also recommended the initiation of a collaborative project on poverty with suitable academic and development institutions.
In response, AERC launched its collaborative research project, Poverty, Income Distribution and Labour Market Issues in Sub-Saharan Africa (SSA), in 1997. During its first three years the project was run by and under the guidance of two coordinators, Professors Erik Thorbecke (Cornell University) and A.A.G. Ali (UN Economic Commission for Africa). In 2000, Professor Germano Mwabu (University of Nairobi) replaced Professor Ali as co-coordinator with Erik Thorbecke.

Laying the foundation
The first phase of the project was completed at the end of 2000 and the project is now embarking on a second phase. The most important contribution of the project during Phase I was the creation and institutionalization of new poverty research networks in a number of SSA countries and the strengthening of existing networks covering about a dozen countries in the region. The capacity to measure and analyse poverty is a necessary precondition to the formulation of poverty alleviation strategies. In this sense, the existence of national poverty networks can also be considered a requirement for the design of poverty reduction policies.

An essential instrument in building capacity was the organization of three training workshops (Kampala, Cape Town and Abidjan) in which prominent economists presented the state of the art relating to poverty analysis for participants from many African countries. The training workshops covered conceptual and analytical issues as well as hands-on computer work using data from African household surveys. Approximately 140 African economists attended these workshops and obtained first-rate knowledge of the methodology underlying poverty analysis.

The project research results have come in two forms: framework papers and country case studies. First, a set of 12 framework papers was commissioned and has been published as AERC Working Papers. They are currently being edited into a substantive volume, which should serve as a major treatise on poverty in Africa. The conceptual frameworks and methodologies developed in these papers provided useful guidelines to the African researchers preparing the country case study papers. Second, 12 country case studies were completed by teams of African researchers over a time horizon of two years. A most valuable capacity building mechanism was twinning visits to the collaborating institutions (Cornell University and the universities of Laval, Gothenburg and Copenhagen). Twinning visits gave the opportunity to African teams to improve their skills during the implementation of their research through closer contacts with renowned scholars and institutions in the field of poverty analysis.

For many of the research teams, poverty analysis requires new types of data, many of which are not especially user-friendly. During visits to the twinning institutions, the research teams were able to work through the details of their analyses from theoretical bases to data management and analysis, benefiting from the experience of twinning institution researchers, and the superior computer, software and library resources available at those institutions.

Applying the results
The research results are actively used by policy makers and the Poverty Reduction Strategy Paper (PRSP) teams to formulate a long-term strategy for poverty reduction in various African countries. The capacity built by AERC is at the fore of the preparation of PRSPs and has encouraged the network to deepen and expand its coverage.

African countries are already tapping into capacity built by AERC in several ways. In Madagascar, the AERC poverty research team is led by a statistician who took advantage of the network to grasp the concepts and analytical tools in the areas of poverty and income distribution. In Cameroon, the team of researchers (predominantly at the University of Yaounde) has used the skills acquired in the AERC training and research workshops to prepare the country’s 1999 Human Development Report for the UNDP. The researchers have also provided inputs into countries’ negotiations for debt relief using poverty analyses.

Likewise, an AERC researcher is a prominent member of the team preparing the Chad poverty reduction strategy. Similar achievements can be cited for research teams in Tanzania, Benin, Kenya and South Africa. The South African team included one member from the Labour Ministry and was jointly funded by the Poverty Project and the Labour Ministry. Their study has come out in a published volume and their recommendations are already being considered by the government. Finally, the Poverty Project has worked closely with different units of the World Bank dealing with poverty. The research output of the Poverty Project has been extensively used by some of these units.

In general the AERC case studies on poverty in Africa confirm high poverty rates in the continent, as shown in previous literature, while also revealing surprising new information about poverty profiles and causes of poverty. For example, although poverty is believed to be concentrated in rural areas, the case studies show exceedingly high poverty rates in urban areas of some countries. Moreover, even though education is associated with lower rates of poverty, some case studies show that living standards between people without education and those with primary schooling are about the same. An overview of case studies shows that some countries with high rates of literacy (e.g., Tanzania) also have high poverty rates, which suggests that the link between more schooling and poverty reduction is not a simple one. Female-headed households tend to be poorer than male-headed households, but this finding cannot be generalized across the continent or across regions in the same country. In Nigeria, and in a notable number of other countries, female-headed households have lower poverty rates than those headed by men, contrary to common beliefs.

The AERC case studies also show that the poverty rate reported for a country depends critically on the method used to derive a poverty line, a situation that complicates comparison of poverty profiles across the continent. However, despite the problem of comparability, the case studies have produced, for the first time in many countries, solid evidence on poverty that is being used by policy makers to discuss and decide on strategies for poverty reduction. There has been a surprisingly high interaction between policy makers and AERC poverty researchers, and this may be taken as some evidence of the project having aided a transfer of scientific information on poverty to policy makers in Africa.

Moving to the next phase
The two broad objectives in Phase I will be continued in Phase II: building capacity for poverty analysis and using the project to generate information to assist in the design and implementation of poverty reduction policies. However, given that new data sets have become available in a number of African countries (e.g., panel data in Ethiopia and South Africa; additional welfare monitoring, and demographic and health surveys in 19 SSA countries), and given new methodological contributions to poverty analysis, it will be possible to examine aspects of poverty in Phase II that could not have been studied in Phase I. In particular, the intent is to encourage as much as possible research that provides results relevant to policy makers.

The proposed duration for Phase II is four years and the specific objectives of the phase are:

• To enhance the understanding of poverty trends and poverty dynamics in specific countries in sub-Saharan Africa.
  
• To diversify poverty measurement to include non-money metric measures such as nutrition- and asset-based measures.
  
• To study linkages among poverty, social capital, agricultural development, rural institutions and globalization.
  
• To inform policy makers, donors, the corporate private sector and civil society of the status of poverty in Africa.

Reaching beyond money metrics
Given the success of the current collaborative research project, coupled with the continuing and increasing importance of poverty reduction as a policy target, AERC believes that it is valuable to expand and extend the project along the directions suggested above. There are several reasons for this consideration. First, the use of nutrition as a measure of poverty is consistent with the notion of defining poverty based on capabilities and functionings first formulated by Amartya Sen. The concepts of capabilities and functionings reflect deprivation better than money-metric measures based on household income or expenditure.

Second, an income or expenditure measure of welfare is a health input variable while a nutritional indicator is a health output variable, that is, an outcome. Third, nutritional indicators measure the status of individuals whereas money metric measures are derived from household level observations that have to be converted into individual measures by way of equivalence scales that tend to be arbitrary.

There is a further, albeit abstract, reason to extend poverty measures outside the class of money metrics. Poverty is a multidimensional phenomenon, reflecting also deprivation in non-economic aspects of life such as spiritual or immaterial assets, and lack of voice and empowerment in society. Despite measurement difficulties, there is a need to begin to bring to the fore non-economic factors in the study of poverty and in the formulation of poverty reduction policies. n

Erik Thorbecke is Professor in the Graduate School and the H.E. Babcock Professor of Economics and Food Economics Emeritus at Cornell University, USA, and a long-standing member of the AERC research network.