African urbanisation: Slower than you think

Reports by development agencies and economic consultancies have claimed in recent years that sub-Saharan Africa’s rapid urbanisation is one factor accelerating economic development in the region. But there are reasons to be cautious about these claims.

According to these analyses, sub-Saharan Africa is urbanising faster than any other part of the world. While populations in African towns do tend to grow faster than elsewhere, this is not how urbanisation is measured.  For a country to urbanise, towns must grow significantly faster than the national population. African countries still have high population growth rates and the speed at which their populations are becoming more urban has been slower than is generally assumed. Many Asian countries are urbanising much faster.

 The urban picture in sub-Saharan Africa is mixed and analysis is hindered by patchy and contested data. Nonetheless, censuses and remote sensing show that in the 1990s a few countries – Zambia, Côte d’Ivoire, Mali and the Central African Republic – were actually counter-urbanising; the share of the urban population was shrinking. The same happened in Zimbabwe between 2002 and 2012. More commonly the increase in the urban proportion of the population was under 2 percent, and in some cases 1 percent or less, over intercensal periods, often much longer than 10 years, stretching into the 2000s. 

Benin, Ethiopia, Malawi, Mauritania, Mozambique, Niger, Senegal, Sudan, Togo and Uganda are in this category. Analysis combining the limited available data from Nigeria’s 2006 census and from Africapolis – a French research project using aerial photographs to estimate town populations – suggests that Nigeria should also be included.

Indeed, Africapolis reports argue that urbanisation levels across much of West Africa have been overestimated by UN agencies and the World Bank. On the other hand, Ghana, Cameroon and Burkina Faso have been urbanising rapidly according to their censuses and conform more closely to the received wisdom. Both Ghana and Cameroon record that more than half their people now live in towns.

 If so many countries have been urbanising slowly, why is the opposite usually claimed? The main reason is the lack of regular censuses in the 1980s and 1990s which meant that governments and agencies, such as UN Habitat, often made growth projections based on earlier periods of rapid urbanisation in the 1960s and 1970s. These frequently proved to be incorrect but helped to create an idea of a rapidly urbanising region which has stuck. 

 There is a further complication about African urban statistics. In many sub-Saharan African countries very small settlements are counted as urban. The nature of economic activities is rarely accounted for. The assumption that urbanisation means people moving into more specialised occupations, generating more added value and higher incomes than typical rural occupations, is a natural one. From the perspective of assessing the role of urbanisation in economic development and sectoral changes, understanding what the workforce is doing is crucial.

In many small African towns the majority of adults are actually farmers. Were these settlements categorised according to criteria used in India, which factor in employment patterns, they would become villages and national urban growth rates and urbanisation levels would fall. Evaluating whether urban figures for Tanzanian and Kenyan censuses are recording economic change and development is very problematic for these reasons. Both are certainly less urbanised than reported. Kenya now includes peri-urban populations, which are often simply densely settled rural areas, as urban.  When excluded, its urbanisation level in 2009 falls from 30 percent to 23 percent. Tanzania’s 2002 census found that one third of those employed in towns were working in agricultural activities.

Most urban population growth in sub-Saharan Africa is derived from natural increase and reclassification of small settlements and not rural-urban migration. For example, a World Bank assessment of Tanzania found that only 17 percent of its urban growth in the year before the 2002 census was due to migration. The reasons lie in the drastic economic changes which occurred under structural adjustment programmes from the 1980s. These devastated urban incomes and urban economies informalised to a startling degree. Migration into towns continued but urban-rural migration increased, reducing the net contribution of migration to urbanisation. Circular migration had reduced in the early decades of independence when urban incomes were higher in real terms than they are now, but then revitalised as one of several strategies to cope with urban poverty and insecurity.

For poor, weak countries, a liberalised global economy means the laws of comparative advantage are a major determinant of what can be produced, what jobs are created, and what sort of urbanisation occurs. In Asia the outcome has been millions of urban-located, formal sector industrial jobs. Subsequent urbanisation has followed that economic development – not vice versa. In Africa economies have restructured back towards primary commodities, and formal urban job creation is limited. Recent rapid GDP growth has been driven by Asian demand for those commodities. This creates new wealth but it is usually exceedingly concentrated amongst a small elite.

Examination of African Development Bank data, which claimed a significant urban middle class is emerging, shows that in most sub-Saharan African countries the population with meaningful middle class incomes is still tiny. African regions with faster urban growth and higher urbanisation levels are often associated with mineral exploitation although this can create enclave economies with quite limited job creation. The significant new mineral finds will stimulate some local urbanisation but, so far, the broader-based urban employment creation experienced in Asia is missing.

Deborah Potts is reader in human geography in the Cities Research Group, Geography Department, King’s College London