Poverty and Groundwater Salinity Survey, 2016

SAMPLING PROCEDURE
The Bangladesh Poverty and Groundwater Salinity Survey 2016 collected data from a total of 1,500 households in three sub-districts or upazilas in Bangladesh - 500 households in each upazila distributed across 50 primary sampling units (PSUs). The three upazilas selected for this study are the following: (i) Taltoli upazila in the Barguna district of the Barisal division; (ii) Morrelganj upazila in the Bagerhat district of the Khulna division; and (iii) Shyamnagar upazila in the Satkhira district in the Khulna division. Each upazila was allocated an equal size of households in order to get poverty estimates of similar precision. The sampling frame consists of a list of all rural villages developed by the Bangladesh Bureau of Statistics (BBS) based on the Census Enumeration Areas (CEAs) constructed for the 2011 Census of Population and Housing. PSUs are constructed by dividing rural villages into listing blocks or Enumeration Areas (EAs) of around 50 households each and then randomly selecting one block for listing.

The three upazilas included in this study where selected based on discussion with a water salinity expert in Bangladesh and practical considerations using a two-stage procedure. In the first stage, we combined upazila level poverty data from the official 2010 Bangladesh Poverty Maps with upazila level information on groundwater salinity collected by the Bangladesh Water Development Board (BWDB) with support from the Institute of Water Modelling (IWM). Using these combined dataset, we classified all 146 upazilas in coastal areas in four groups: (i) high water salinity and high poverty rate; (ii) high water salinity and low poverty rates; (iii) low water salinity and high poverty rate; (iv) low water salinity and low poverty rates. Figure 1 shows the spatial distribution of coastal area upazilas based on these four categories. In the second stage, we selected one upazila from each of the first three categories as focal areas for this study after discussion with a groundwater expert on availability of other water-supply options (e.g. managed aquifer recharge) and practical considerations. This categorization of upazilas also serve as our three sampling strata - high water salinity and high poverty rate, high water salinity and low poverty rates, and low water salinity and high poverty rate.