Multidimensional Poverty: Evidence from Rural-Urban Migrants in Indonesia
Gultom, Sarah Elyzabeth (2016). 'Multidimensional Poverty: Evidence from Rural-Urban Migrants in Indonesia' Paper presented at the annual conference of the HDCA, Tokyo 2016.
This study presents the application of the Multidimensional Poverty Index (MPI) in a case of rural-urban migration in Indonesia. Rural-urban migration has been a well-known phenomenon in a developing country, and there is currently limited literature to discuss how the rural-urban migrants perform and progress in the cities especially when they attempt to escape from poverty. The notion of self-selection among migrants suggest that migrants are positively selected from the population of origin since they usually have better observable characteristics (e.g. educational attainment) and non-observable characteristics (e.g. abilities and drive) compared to those who stay in their native locations. Therefore, migrants tend to have the ability to adapt and assimilate faster in their new locations. As they might have favourable living conditions due to their positive characteristics and strong social network between each other, migrants can potentially outperform the non-migrants in terms of their standards of living.
Particularly for rural-urban migrants in Indonesia, past studies have shown that they have a lower probability to be expenditure-poor, and those who migrated during their childhood have a higher probability to be healthier and higher educational attainments when they are grown up. This suggests a possibility that rural-urban migrants in Indonesia have better standards of living and therefore, a lower level of multidimensional poverty compared to similar non-migrants. It is possible that, based on the notion of positive self-selection, these rural-urban migrants might possess certain skills or abilities that could drive them to improve their standards of living and, if they are poor, to increase their probability to escape from poverty. To analyse this issue further, this study aims to provide empirical evidence on how well rural-urban migrants have performed in their new locations and how different their standards of living are compared to the natives over time.
The MPI provides an informative measure that has a useful framework to analyse multidimensional poverty where it does not only view poverty as lacking of monetary possessions, but also as capability deprivation. This method allows poverty to be defined in a range of different dimensions, aggregately measures the joint distribution of the deprivations, as well as to allow flexibility in selecting dimensions of poverty and other relevant components. Its decomposability property allows a more focused and thorough poverty analysis of a certain group and to evaluate the contribution of each poverty dimension to the overall poverty level. This study tries to illustrate the advantages of MPI by comparing the multidimensional poverty level of rural-urban migrants and non-migrants, offering more insights regarding performance and progress of rural-urban migrants in cities. There is still literature yet to assess the poverty level of rural-urban migrants with a joint distribution measure as MPI.
The dataset used in this study is the survey of Rural-Urban Migration in Indonesia (RUMiI). The survey was specifically designed to gather information on various socioeconomic characteristics of rural-urban migrants in Indonesia. It oversamples the rural-urban migrants and was conducted in four major cities, which have a large number of rural-urban migrants to represent four of the largest islands in the country: Tangerang (Java), Medan (Sumatera), Makassar (Sulawesi) and Samarinda (Borneo). The data is yet to be nationally representative, but these cities can capture the diversity of the rural-urban migrant experience in Indonesia. The data were collected annually by following the same individuals from year 2008 to 2011. The nature and the richness of information that the dataset provides allows for a year-by-year analysis of multidimensional poverty level for both groups (i.e. rural-urban migrants and non-migrants).
Results suggest that rural-urban migrants are able to perform better than non-migrants in terms of their multidimensional poverty level. The multidimensional poverty level reduced for both groups over the chosen period of time, but rural-urban migrants tend to have a higher change in the reduction compared to the non-migrants. A higher standard of living achieved by the rural-urban migrants in the cities might suggest the notion of positive self-selection among them.