Natural Disasters and Multidimensional Poverty: A Study of the Floods in Punjab

Zafar, Sameen (2016). 'Natural Disasters and Multidimensional Poverty: A Study of the Floods in Punjab' Paper presented at the annual conference of the HDCA, Tokyo 2016.

Natural disasters have plagued mankind since centuries. Climate change has caused the loss of millions of lives and livelihoods. Pakistan was affected by severe flooding in 2010 when nearly one-fifth of the country was inundated by floodwater. Almost one million people were affected by these floods leaving them in a state of destitute poverty and homelessness and without their livelihoods and possessions.
This study is an attempt to capture the state of multidimensional poverty before and after the floods in Punjab province of Pakistan. The methodology proposed by Alkire and Foster (2007) will be used to construct the multidimensional poverty index in 2008 and 2011 and a comparison will be drawn between the two years to gauge the extent of deprivation after the floods. A comparison will also be drawn between the determinants of monetary poverty and multidimensional poverty to analyse the important determinants of income and non-income poverty for targeted policy interventions. The multidimensional poverty index (MPI) is calculated for the two surveys using three dimensions of health, education and living standards using the same indicators as proposed by Alkire and Santos (2010).
MICS datasets for Punjab are used for this study. MICS is a provincially representative survey of households and it includes data on about 100,000 households with nearly half a million individuals living. The survey was conducted by the Punjab government with assistance from UNICEF.
The analysis is conducted at the household level and results are presented at the aggregate level for Punjab as well as for districts and towns within Punjab. Punjab is Pakistan's most populous province and contains nearly three fifths of the nation's people.
In order to analyse the determinants of poverty, a probit as well as a bivariate probit methodology is adopted to analyse whether the determinants of income poverty differ from those of non-income poverty. Furthermore, a spatial analysis is conducted to investigate the number of towns which were affected severely and not so severely by the floods and a difference-in-difference technique is used to comprehend the severity of the floods at the town level.
The results for districts show higher poverty levels in the South as compared to North Punjab, with the exception of a few southern districts. This result is supported by prior evidence since most of the resources of the Punjab government are heavily expended in the northern parts with the southern parts usually lagging behind. In addition, child school attendance and child nutrition are major variables which determine whether households are multidimensionally poor or not. The calculation and determinants of income and non-income poverty is currently being conducted.
It is essential to examine how multidimensional poverty changes when natural disasters occur. This study is the first one in tackling this issue using the multidimensional poverty index proposed by Alkire and Foster (2007). It can aid policy makers to understand the long term needs of victims of the flood affected areas and also identify the vulnerable households to prevent them suffering from floods in the future.

scroll to top