Subjective well-being, disability and adaptation: A case study from rural Ethiopia

In many developing countries poor infrastructure – including sanitation and health facilities – exposes the population to high risks of disability. Low standards of health and safety at work and at home, coupled with political, ethnic, and domestic violence, also contribute to raising the risk of becoming physically disabled. The effect of physical disability on people’s lives is likely to be worse than in developed economies because of the reliance on physical labour for income generation – for example, in farming. Higher levels of national income and technological capability may also enable societies to make the investments required to enable disabled individuals to be productively employed. Finally, since formal social insurance is usually lacking in developing countries, the effect of disability on welfare is expected to be higher as disabled people must rely on social networks that have limited capacity to pool risks (Fafchamps and Lund, 2003). However there are also factors that tend to lower the proportion of disabled individuals in poor societies. The first one is lower life expectancy. In developed economies, the incidence of disability typically increases with age (e.g., loss of eyesight and hearing, paralysis due to stroke). This means that, other things being equal, populations with a larger proportion of elderly people have a larger proportion of disabled individuals. Put differently, many people in poor rural economies do not live long enough to become disabled. The second reason is that disability may have such dire consequences in terms of lost income and lack of support that disabled individuals have a much shorter life expectancy than they would have in a developed economy. If this is the case, the proportion of disabled individuals in the population may be low even though the risk of disability is high. In spite of the fact that disability is an important welfare concern, socio-economic studies on the effect of disability in developing countries are few in number. This chapter seeks to fill this gap by documenting the incidence of different forms of disabilities in rural Ethiopia. Using cross-sectional data from the Ethiopian Rural Household Survey collected in 2004, we examine the relationship between disability and welfare as captured by subjective well-being and self-reported wealth ranking questions. In particular, we test whether the negative effect of disability on welfare decreases over time. If it does, this would suggest that over time people adapt to disability. We also investigate whether the negative effect of disability on subjective well-being operates primarily through reduced material welfare. Empirical results indicate that, as expected, disability has a significant negative impact on welfare. This is true whether the person who answered the subjective well¬being question herself is disabled, or whether the disabled person is another member of the household. Even though we find some weak indication of adaptation with respect to specific forms of disabilities, the overall weight of empirical results suggests little adaptation: disability is associated with a lower subjective well-being irrespective of the time elapsed since the onset of the disability. We also find that, in the studied population, disability is associated with significantly lower material welfare. This lower material welfare is the main channel through which disability reduces subjective well-being. These findings stand in contrast with the psychology literature which has found that, in developed economies, the negative effect of well-being on subjective welfare becomes attenuated over time. But they are consistent with the local context: in an economy where there is no social protection against disability and where production depends on physical labour, disabled individuals are less able to contribute to household income, and this permanently reduces the household’s material welfare. This explains why the negative effect of disability on well-being is shared by all household members, whether or not they are themselves disabled. We should stress that, since we only have cross-section data, we cannot control for unobserved heterogeneity. In particular we cannot entirely eliminate the possibility that the relationship between material poverty and disability results from poor households having a higher risk of disability. We also cannot control for selectivity, that is, the possibility that a number disabled individuals are not observed in our data either because they left the household – e.g., to become beggars – or because they died prematurely as a result of abandonment or neglect (e.g., Miguel (2005). If this were the case, our results would underestimate the incidence of disability. To the extent that poor households are less able to care for disabled individuals, selectivity bias would affect poor households more, which means that the relationship between disability and material poverty may even be stronger than suggested by our results. To disentangle these issues longitudinal data on well-being and disability is necessary. The evidence presented here is nevertheless sufficiently strong to suggest that such longitudinal data should be collected. This chapter is structured in the following way. The next section briefly discusses the link between welfare, disability, and adaptation in general, and posits the adaptation hypothesis that is pursued in the empirical part of the chapter. Section 8.3 presents the survey from which the data are sourced and reports a number of descriptive results on the distribution of subjective well-being and disability as captured by the data. Section 8.4 presents the different tests for the existence of adaptation among households with disabled people in rural areas of Ethiopia. In Section 8.5 we investigate the relationship between disability and material welfare. Section 8.6 provides the conclusions.