Economic growth, inequality and poverty: why has poverty reduction been slow in Africa?

At the Millennium Summit in September 2000, where the Millennium Development Goals (MDGs) were adopted, the international community set the goal of eradicating extreme poverty and hunger, with two specific targets. First, over the period 1990‑2015, halve the proportion of people whose income is less than USD 1 a day. Second, halve the number of people suffering from hunger. Measuring poverty is not only important for establishing the progress that has been achieved in terms of reaching the set targets, but also for knowing the extent to which economic growth and socio-economic policies affect the process of poverty reduction.

When the MDG targets were set, the indicator chosen to determine progress in poverty reduction was the incidence of extreme poverty, measured by calculating the proportion of people living on less than USD 1 per day using 1993 purchasing power parities (PPPs). After revisions of the estimates for purchasing power parities exchange rates were introduced in late 2007, the extreme poverty line was revised upwards to USD 1.25 in 2005 PPPs. This measure of poverty is called headcount ratio.

The complexity of the concept of poverty and its measurement has generated passionate debates about the methodologies used to measure poverty, the poverty lines that should be used to compare poverty incidence across countries, and the different results that different poverty lines and different methodologies have produced. Some have also argued that relying on household data to measure poverty makes the analysis of poverty dynamics in Africa difficult given the time separating household surveys. Attempts to rectify this problem by using other methods and different data sources remain contested.¹

Using the poverty line of USD 1.25 a day in 2005 PPPs, the proportion of sub‑Saharan Africa’s population living in extreme poverty increased from 54% in 1981 to 59% in 1996. By 2005, the most recent year for which comparable data are available, the proportion had declined to 51% (Chen and Ravallion, 2008). Hence, from 1996 to 2005, the proportion of poor people in Africa had been reduced by 8%. However, as a result of the crisis that has been affecting world economies since 2007, it has been estimated that the number of poor people in Africa might have increased by 50 million in 2009 and by another 39 million in 2010 (Ravallion, 2009a) relative to a baseline scenario with the absence of the crisis. This may have slowed the progress that the continent had been making until the onset of the crisis.

The poverty dynamic appears to mirror the evolution of economic growth rates. In the 1980s, the average per capita income of the African countries in the sample was USD 1 955 in 2005 PPP exchange rates. In 1996 the average income per capita had declined to USD 1 887, and it was USD 2 163 in 2005. As a result of the economic crisis, African rates of economic growth declined. From an average growth rate of 6% a year in 2006‑08, the rate of growth declined to 2.5% in 2009. The African Economic Outlook 2010 projected that growth in 2010 would be around 4.5% and in 2011 around 5.2% (AfDB et al., 2010). Taking into account the average rate of population growth of about 2.5% per year, these rates of economic growth indicate that relative to the period from 1996 to 2008, income per capita in the period 2009-10 was much lower. In fact, the rate of economic growth in 2009 suggests that there was no increase in income per capita that year.

These statistics suggest that growth of income per capita reduces poverty.²  As high economic growth increases per capita income, more people are pulled out of poverty, other things being equal. This reduces the proportion of poor people in the population. But to what extent will growth contribute to poverty reduction in Africa, and why has poverty reduction in Africa been slower than in other regions?

Owing to the paucity of reliable cross-country data on poverty incidence and its determinants in Africa, very few studies have attempted to deal with these questions at the regional level. Quantitative evidence shows that three key factors explain the weak response of poverty to economic growth in Africa. The first is that despite the rise in growth rates over the period 1996-2008, Africa’s average growth rates have not been high enough to have as strong an impact on poverty reduction as in other regions. The second factor is that the growth process in Africa has been more weakly linked with poverty reduction than in other regions. The third factor is that relatively high rates of inequality, as well as high levels of poverty, have hampered poverty reduction in Africa.

In an early attempt to analyse poverty at the continental level using cross-country quantitative analysis, the Economic Commission for Africa (ECA) (1999) found that sub‑Saharan Africa would need to grow by 7% per year to meet the first MDG of halving poverty by 2015.³ This finding suggests that if African countries are reducing poverty more slowly than other regions, the reason could be that the growth rates in Africa have not been high enough. Indeed, over the period 2001‑09, the continent’s average rate of economic growth was 5.3% per year. Only 9 out of 53 countries – Angola, Chad, Equatorial Guinea, Ethiopia, Mozambique, Nigeria, Sierra Leone, Sudan and Uganda – recorded average GDP growth rates that were higher or equal to 7% per year (AfDB et al., 2010). The implication is that to have a more substantial effect on poverty reduction, economic growth in more African countries will need to be higher than it has been so far.

A deeper analysis of the relationship between economic growth and poverty reduction in Africa shows that in the majority of the fastest growing countries, defined as the countries with growth rates higher than the African average of 5.3% of GDP over the period 2001‑09, economic growth had the weakest effect on poverty reduction.⁴ Out of 44 countries for which there is information on growth elasticities of poverty, 14 recorded high economic growth rates as defined here.⁵ These countries are Angola, Burkina Faso, Cape Verde, Chad, Ethiopia, Ghana, Mali, Mozambique, Nigeria, Rwanda, Sierra Leone, Tanzania, Uganda and Zambia. However, only three countries (Cape Verde, Ethiopia and Ghana), or about one‑fifth of the fastest growing countries, have high growth elasticities of poverty, defined as elasticities higher than the African average of ‑1.717 (see Table 4.1). Among the fastest growing countries, no oil or mineral exporter has a high growth elasticity of poverty. On the other hand, oil producers (although some are small oil producers) with high growth elasticities of poverty have low growth rates.

The weak growth elasticity of poverty suggests that Africa has been unable to reduce poverty as needed, even in some countries where the growth rate was relatively high. This illustrates that growth alone is not a sufficient condition for poverty to decline, raising the question of why growth in Africa is weakly related to poverty reduction relative to other regions. As Table 4.1 shows, sub‑Saharan Africa has the lowest average growth elasticity of poverty.⁶ It is half of the elasticity in Latin America and the Caribbean and only two‑fifths of the elasticity in eastern Europe and Central Asia.

The weak response of poverty reduction to economic growth may be due to the growth process not being inclusive enough, partially because it was not strongly linked to activities and economic sectors where the poor are. This could have been because growth did not take place in the sectors where the poor work (e.g. in agriculture), or in areas where the poor live (in rural areas), or simply if growth did not involve the poor by using their labour. It is known that high fuel and mineral commodity prices have strongly influenced the growth process in many of the fastest growing African economies in the period from 1996 to 2008, given that fuels and minerals account for the largest share of Africa’s exports (AfDB et al., 2010).⁷ These commodities, particularly fuel, are produced in “enclave” industries using highly capital-intensive technologies that tend to exclude the poor. Therefore, even rapid growth can have limited impact on poverty reduction when growth is generated from sectors weakly linked to the rest of the economy, particularly the sectors where large numbers of poor people are active. The limited impact of economic growth in reducing poverty creates a vicious cycle. Higher poverty diminishes even further the capacity of economic growth to reduce poverty (Ravallion, 2009b).

Countries with high levels of poverty can enhance the poverty impact of growth if their growth strategies focus investments on sectors where the poor are more active. Growth needs to be associated with employment creation, especially in rural areas where most of the poor live, because employment is the main channel through which economic growth affects poverty (Nkurunziza, 2007). This has not been the case in most African countries in the past, even though the situation may have changed to some extent in some countries. In this regard, the recent trend to refocus economic development in Africa on agriculture and rural sector development is expected to have a stronger impact on poverty reduction, as it will make agriculture and the rural economy important sources of economic growth.

Economic inequality is the third factor that inhibits the responsiveness of poverty to economic growth. The growth effect is stronger in more equal societies where different segments of the population more equally share the benefits of growth. In less equal societies, additional income created by economic growth is appropriated by a small group that accounts for a disproportionately large share of a country’s income. In a particular country, the extent to which economic growth reduces poverty depends on the level of inequality because economic growth and inequality have an opposite effect on poverty. Even though at the aggregate level Africa has the lowest inequality elasticity of poverty, on average, relative to other regions, the last column of Table 4.1 shows that the negative effect of inequality on poverty outweighs the positive effect of economic growth. Elasticity estimates show that 61% of countries in Africa have inequality elasticities higher than growth elasticities of poverty, four out of five of the highest inequality elasticities of poverty being for North African countries (Fosu, 2011).

On the basis of these factors, poverty reduction policies solely focusing on economic growth would have a more limited impact on poverty in cases where initial levels of inequality are high and persistent. In fact, if additional income arising from economic growth were equally distributed, reducing the negative effect of inequality on poverty reduction, poverty rates would fall faster than at the current rate (Bigsten and Shimeles, 2003). Therefore, as inequality in sub‑Saharan Africa over the period from 1981 to 2005 has tended to be persistent, the slow pace of the decline in inequality helps to explain why the impact of growth on poverty has been so unrelentingly low. To accelerate the rate of poverty reduction, the progress made in terms of increasing economic growth will need to be scaled-up, while, at the same time, finding ways to reduce inequality.