It is intuitive that poverty levels are likely to be significantly higher for those living in remote or isolated locations, but so far there has been relatively little quantitative evidence to substantiate this. Such evidence as there has been has often been only at a highly aggregated level (e.g. north versus south of a country): for policymakers seeking to tackle poverty in often hard-to-reach areas, it is important to have such information at a finer resolution. This paper uses available quantitative data to develop composite indices of isolation for Uganda based on a series of indicators of access to infrastructure and services, and uses these to examine the relationship between isolation and different aspects of poverty.
The paper begins (Section 2) by reviewing different concepts of isolation and remoteness considered in the existing literature, which is focused predominantly on rural areas. It highlights the complexity of the concept, which has many different dimensions: isolation can be understood in terms of distance from infrastructure, from services, from an economic core or from political decision making, and these different dimensions are often strongly interrelated. This section also reviews existing literature on why isolation is likely to be associated with poverty.
This leads into a review (Section 3) of the current evidence on poverty in Uganda, considering both monetary and non-monetary dimensions. Particular attention is paid to geographic disparities in poverty, including an initial discussion of their possible causes and historical origins. Existing analysis does not distinguish between more and less remote areas of Uganda, though, so this paper then discusses the construction of a summary index of isolation for districts and communities in Uganda (Section 4). This draws on extensive fieldwork, which collected information on a series of indicators of isolation at the district level (physical remoteness; access to infrastructure; availability of services, facilities and means of communication). Factor analysis is used to create composite indicators of isolation in Uganda, thereby reducing the many different dimensions into single indices based on correlations apparent in the data. An analysis of these indices at the district level reveals significant differentiation within the four regions of Uganda.
Data from the 1999/00 Uganda National Household Survey (UNHS) are then used to examine the relationship between these measures of isolation and different dimensions of poverty, including measures of vulnerability. There is a strong correlation between isolation and poverty, including chronic poverty. Further analysis considers some of the channels underlying this association. Households in more remote areas have lower levels of market participation (including commodity and financial markets), itself associated with poverty; they make less use of public services (which are often more remote); and household members (women and children) in remote areas have to devote more time to fetching wood and water.
The paper then summarises the implications of these findings for likely future evolution of poverty, in particular considering likely future spatial patterns of growth. It discusses policy implications for tackling poverty in remote areas of Uganda and integrating those living there into the process of growth. It also identifies future research priorities, both specific to Uganda and to the issue of isolation and poverty more widely.
