Empowering the very poor: The Ultrapoor Programme in Bangladesh

This large scale project examines the fundamental question as to whether it is possible to lift the very poorest individuals in society out of poverty. It focuses on the poorest women in rural Bangladesh who typically do not have access to microfinance, have very limited economic opportunities, are often excluded from vital public services and are often both socially and psychologically disempowered.

In the programme, participatory rural appraisal (PRA), is used to sort households into different community defined wealth bins and identify the poorest households. After further cross-checks from BRAC are carried out a set of ultrapoor households are identified which become programme recipients. For 18 months these households receive training and weekly assistance with starting a small business (e.g. cow or goat rearing, rearing chickens for eggs, growing vegetables for sale). For this period they also receive monthly visits from health volunteer and have access to a lawyer if they need one. Between 18 and 24 months they take part in confidence building sessions about how to use microfinance and are enrolled in village level microfinance organizations which are self-sustaining.  The subsidized element of the programme is highly intensive but only lasts for two years.

The programme is therefore an exercise in economically, socially and psychologically empowering the poorest women in Bangladesh. A randomized evaluation which involved tracking 30000 households in treated and control villages will shine light on how the programme works. It is designed to identify the full impact, both on the targeted households and on the rest of the community. This is achieved through comparing the targeted and the non-targeted population in the areas where the programme takes place (treatment spots) to the population in areas where the programme does not take place (control spots). In order to identify the effects, the treatment and control households are chosen to be as similar to each other as possible before the programme. This is achieved through the following way:

Step 1: The ultrapoor programme decided which branch offices in the targeted districts will be included in the programme from 2007 onwards.
Step 2: Out of these branch offices, the evaluation team randomly chose one treatment and one control branch office in each sub-district (upazila).
Step 3: Using the participatory wealth ranking assessments (PRAs) carried out in both the treatment and control spots, a sample of targeted and non-targeted population within each treatment spot, and a corresponding sample of ultra poor and non-ultra-poor population in the control spots was formed.

The evaluation strategy involves a baseline survey and repeat surveys carried out in the treatment and control spots. A great attraction of the survey is that we can follow the same households over a full five years to see whether any improvements in living standards of ultrapoor households are maintained over time. The main principles behind the evaluation strategy are as follows:

(i)         By comparing the changes in the lives of the population in the control spots over time, we will be able to identify the dynamics of poverty in the absence of the programme.
(ii)         By comparing the targeted households in the treatment spots to the equivalent ultrapoor households in the control spots over time, we will be able to identify the effects of the programme on the households that are targeted by the ultrapoor programme.
(iii)        By comparing the non-targeted households in treatment spots to equivalent households in the control spots, we will be able to identify the full extent of the spillover effects of the programme.
(iv)        Furthermore, the evaluation strategy is designed to enable us to not only identify the extent of the spillover effects, but the mechanisms through which these effects take place (such as the social networks) and how these mechanisms themselves are affected by the programme over time.

BRAC and the LSE are collaborating an large scale randomized survey of 30000 households. The survey was designed by research staff at BRAC and the LSE. The survey is currently in the field picking up baseline information on ultrapoor and non ultrapoor household in both treatment and control spots. The baseline survey will be completed by the end of 2007 prior to asset transfer to ultrapoor households in treatment spots. To assess the impacts of the ultrapoor programme there will be follow up surveys in 2009 and 2011. Following households for five years will also enable us to examine what effect enrolment into BRAC microfinance groups will have on households exposed to the ultrapoor programme. In addition to the larger 30000 randomized trial we will also be carrying out a subsample study on 30000 households (2000 drawn from treatment spots and 1000 from control spots). For this subsample we will be surveying all households in a spot therefore enabling us to draw complete business and social networks in these spots. The subsample census will be fielded annually between 2007 and 2011 and will help us to better uncover the precise mechanisms via which exposure to the ultrapoor programme economically, socially and psychologically empowers both ultrapoor households and their neighbours (via spillovers) in the treatment spots.