Dr Gindo Tampubolon, Lecturer in Poverty, Global Development Institute
Developing countries are often marked by spatial disparity with the centre hoarding political influence, wealth and services. Increasingly they are also marred by neglect of disability with very little known about its distribution and consequences. Here new evidence on spatial disparity in disability and poverty is shown using the new Uganda Demographic & Health Survey 2016 released early this year.
The survey collected information on six domains of disability following the UN recommended instruments, the Washington Group measure of disability. The six domains are hearing, seeing, walking or climbing steps, remembering or concentrating, self-caring, and communicating. In each domain people reported whether they faced no difficulty, some difficulty, a lot of difficulty or completely unable to perform it. I summarise the report of 46,034 Ugandans aged 15 and older on a map. The map showed the South Central region (encircling Kampala) has a low percentage of people with some disability as rendered with a lighter shade of green whereas Kigezi in the far south west is home to the highest percentage of people reporting some disability. Disability like any other disadvantages has a spatial character. See map.
Like other disadvantages which often overlap, so for instance there are more sick people among those in poverty, it is likely that disability goes together with poverty. This can be for at least three reasons. Physically, people with disability may find themselves facing difficulty in securing a decent paying job compared to people without disability. Second, infrastructure in public and in workplaces may be inadequate to support people with disability despite their potential to contribute. Last, there may be stigma with certain forms of disability. In specific cases these three may all combine to reinforce the association between disability and poverty.
Given the evidence on spatial disparity in disability on the map, some maybe due to unobserved factors across districts, I fit a multilevel model of poverty on disability. One new twist to this story concerns mobile phone technology. It is widely known that the technology, a quintessential communication technology, has offered users myriad benefits all over the world. By making communication cheap and timely, all sorts of economic transactions and social interactions are facilitated. In Uganda, any MTN account holders can easily transfer funds across the span of the country from Kigezi to Karamoja without going through Kampala. So use of this technology may also affect poverty. Most interesting however is the question: does it matter more for people with disability?
To answer this question I examine the interaction between mobile phone use and disability or mobile phone use in the hands of people with disability when explaining poverty. Specifically the outcome of interest is multidimensional poverty, along the three axes of education, health and living standards as is recently done for the Global Multidimensional Poverty study. The indicators include, for education, no household member aged 10 years or older has completed five years of schooling, any school-aged child is not attending school up to the age at which he/she would complete class 8. Meanwhile for health they include any child has died in the family in the five-year period preceding the survey, or any adult under 70 years of age or any child for whom there is nutritional information is undernourished in terms of weight for age. For living standards, there is a long list of indicators including the household has no electricity, the household’s sanitation facility is not improved (according to MDG guidelines) or it is improved but shared with other households, the household does not have access to improved drinking water or safe drinking water is at 30-minute walk from home, the household has a dirt, sand, dung, or other type of floor, the household cooks with dung, wood, or charcoal, the household does not own more than one of these assets: radio, TV, fixed telephone, bicycle, motorbike, or refrigerator, and does not own a car or truck. Following the common practice when individual incomes are not collected, poverty is assessed for each household and assigned to all its members.
Though focusing on the interaction between disability and technology in explaining poverty, other covariates are accounted for including gender, age, education, residence (urban-rural) and random district effects. The results of marginal probabilities of multidimensional poverty is summarised in a table.
Disability, technology and multidimensional poverty
Random intercepts logit with 2002 districts as higher level units and adults as lower level units. Source: UDHS 2016. |
||||
Use mobile phone | ||||
1 | 2 | 3 | 4 | 5 |
No | Yes | Diff. | ||
Any disability | No | 28% | 20% | 8% |
Yes | 38% | 23% | 15% |
How much difference does the technology make? With the usual caveats of correlations in the table above, they nevertheless are striking. Those with any disability (bottom line) have higher probabilities of multidimensional poverty (38% versus 28% or 23% versus 20%). Those technology users (column 4) have lower probabilities of multidimensional poverty (20% versus 28% or 23% versus 38%). The last column i.e. the difference is most intriguing. Among able bodied adults, the difference that the technology makes is 8%, whereas among those with disability the technology associates with 15% lower level of poverty, i.e. about twice as large.
In Uganda there is something to be done about exploring further as to why and how disability and technology can interact to uncover such a substantial difference in lower poverty rates among the most disadvantaged.
Find out more about our ICTs for Development MSc