Multidimensional Energy Poverty and its Dynamics in Rural and Small Towns of Ethiopia: A Fuzzy Set Analysis

No Thumbnail Available



Journal Title

Journal ISSN

Volume Title




Ethiopia, like many others in the sub-Saharan Africa, hosts its largest share of population in its rural areas and small towns. And among many factors, these areas of the country are characterized by having lower access to modern energy fuels. Thus, using panel data set constructed from the three rounds of Ethiopian Socio Economic Survey, this paper has attempted to investigate the level of multidimensional energy poverty and its dynamics in these areas. Unlike number of researches made on energy poverty, this paper has attempted to divert from the traditional crisp poor non poor dichotomization using a fuzzy set approach which represent each household by the degree of energy poverty it faced in values that range between zero and one through the use membership functions. Given selected dimensions which include type of cooking fuel, indoor air pollution, source of light and access to media and communication, and applying methods of average, intersection and union as techniques to aggregate degree of deprivation across each dimension selected, a fuzzy multidimensional energy poverty index for the study area was determined, and it was further decomposed to see which dimensions contribute the most for it. During the first survey, households in the study area were found to be faced with 79.98% average deprivation while experiencing a minimum degree of deprivation that reaches to 47.64% in each selected dimension. These numbers fall only slightly during the last survey where average deprivation declines to 72.41% while minimum deprivation in each selected dimension falls to 34.61%. Across each survey, households were found to be faced with a degree of deprivation that is close to 100% at least in one of the dimensions selected. These results are further supported by a fuzzy set longitudinal analysis made with the application of joint membership functions. And the rate of re-entry is found to be 98.11% and exit rate 42.24%. Further, given the fractional nature of the dependent variable, dynamic fractional regression was used to investigate the presence of state dependence, and a one percent rise in the propensity of experiencing deprivation in all dimensions in the previous period was found to cause a 3.8% rise in a given year’s deprivation overlap. And among other selected determinate variables of energy poverty, number of rooms available, proportion of children aged seven and above, log of real income per adult equivalent and year of survey are found to have significant and negative effect on households energy poverty. And proportion of labour aged female household members, living in rural areas, and compared to region Tigray living in other selected regions of the country are found to cause a significant and positive influence on households energy poverty.



Multidimensional energy poverty, fuzzy set analysis, dynamic fractional regression