Of the various disaster risk mitigation Social Safety Net Programmes (SSNPs) run by the Government of Bangladesh, this article discusses three major ones. These are Vulnerable Group Feeding (VGF), Gratuitous Relief (GR) and Test Relief (TR). These SSNPs are designed to protect the poor and vulnerable people of society, especially those who are vulnerable to natural disasters, seasonal employment fluctuations, and food price shocks. The total amount of money allocated to these three programmes is about Tk 38.24 billion (3824 crore), which is about 8.5 per cnet of total allocation for social protection in 2018-2019 budget. Instead of focusing on what proportion of poor people are receiving SSNPs, the analysis presented below shows the profile of SSNP beneficiaries, whether or not they are from the poorest group and the selection criteria have been properly fulfilled. The main goal is to examine the quality of targeting of the three important SSNPs of the Bangladesh government.
Among many other criteria, the most important criterion for the inclusion in VGF, GR, and TR beneficiaries' lists are: whether a household or a person is poor (e.g., below poverty line or eats only two meals per day), is landless or has land less than 0.5 acres, is affected by natural disaster/food price shocks, and is a day labourer. The single most general criterion is whether a household or a person is poor/vulnerable. Thus, the analysis starts with the question: are all the beneficiaries included in the VGF, GR, and TR programmes poor? This criterion is measured in relation to the national poverty line estimated using 2016 HIES (Household Income and Expenditure Survey) data. Chart 1 shows that only 20 and 19 per cent VGF recipients are extreme and moderate poor, respectively. This indicates the majority (61 per cent) VGF beneficiaries are from the non-poor group. This is also evident for both GR and TR programmes. Among the three programmes, TR programmes perform worse in terms of including poor people as beneficiaries. Only 22 per cent of TR recipients are poor. This finding has further been confirmed using the per capita expenditure quintile as an alternative criterion. Chart 2 shows that the majority of the SSNP recipients are from the top three expenditure quintile. Of the total, 56 per cent VGF, 61 per cent GR, and 72 per cent TR beneficiaries are from the top three expenditure quintile, again indicating TR includes more non-poor people than the other two programmes. The average per capita expenditure of the third quintile group is Tk 2655, which is well above the 2016 national poverty line income of Tk 2268.
Whether a household is affected by natural disasters is our next criterion to see the quality of targeting of SSNPs. Chart 3 shows that of the total VGF recipients, only 8.0 per cent experience shocks from natural disasters and therefore were vulnerable. For GR and TR, these percentages were 23 and 12 per cent respectively. The corresponding figures for all kinds of shocks (natural disasters plus driven by human activities) are 12, 26, and 16 per cent for VGF, GR, and TR beneficiaries respectively. The results indicate the inclusion of a significant number of non-eligible beneficiaries.
We also see the distribution of household earning members for the recipients of SSNPs. A household will be included in the SSNP beneficiaries list if it has no adult earning member. The results in Chart 4 do not go in line with this inclusion condition. Chart 4 shows that only 3-5 per cent of SSNP beneficiaries have no earning member. About 63-88 per cent has at least one earning member, followed by 22-25 per cent of households with two earning members. This criterion also indicates the inclusion of a significant number of non-eligible households as beneficiaries.
In terms of household head employment status and education, the targeting of SSNP performs better. Households with day labourer heads are included more than other professions. About 48-62 per cent of household beneficiaries have day labourer heads while 24-30 per cent are self-employed followed by 6-9 per cent employees. About 6-9 per cent of households with unemployed heads are included in the SSNPs.
Households whose heads are less educated are most likely the recipient of the programmes (Chart 6). The results show that about 48-53 per cent of household heads of the SSNP recipients are illiterate, followed by 31-34 per cent with primary education only. The results indicate the effective targeting of SSNP beneficiaries in terms of household head education and employment status. The effectiveness of targeting of SSNP in terms of land owned by a household is even better. Most recipients of VGF, GR, and TR programmes are either landless or marginal landowners. About 8-25 per cent of beneficiary households have no land and about 71-78 per cent of households own less than 0.5 acres (marginal landowner), indicating the effective targeting of beneficiaries. This leaves us a few percentages (14-23 per cent) of the beneficiary households have land more than 0.5 acres.
The analysis presented above indicates that though the recipients of VGF, GR, and TR programmes are the largely marginal landowner (less than 0.5 acres), day labourer, and illiterate or less educated, they are largely from the non-poor group of society. The analysis suggests that the implementing agencies of SSNP put more weight on land owned by a household in targeting the beneficiaries of three SSNPs as discussed above rather than focusing on the poor and vulnerable condition of households. This also means owning a lower amount of land does not necessarily imply a particular household is poor and vulnerable to shocks. According to HIES 2016, about 70 per cent of the marginal landowner are non-poor. Thus, targeting SSNPs beneficiaries based on land ownership is ineffective. One could argue that the SSNPs lifted out people from poverty and therefore most of the recipients are non-poor. But, given the average amount of money received by per beneficiary of above-mentioned three SSNPs is Tk 50 per month (HIES 2016), it is less likely that the SSNPs help households to get rid of poverty/vulnerable situations to a large extent.
Finally, what can be done to improve the targeting effectiveness of SSNPs to reach out to the poorest and most vulnerable households of society? The implementing agencies of SSNPs may create a national database of the households based on the key selection criteria, particularly on income or expenditure, and should update it every six months or one year. Then, based on household income, the implementing agencies can distribute ''below the poverty line'' (BPL) and ''above the poverty line'' (APL) cards to its people. This will help to identify vulnerable households when an intervention is required due to shocks arise from natural disasters, higher food prices, and seasonal employment fluctuations.
(Opinions expressed in this article are those of the author; they do not necessarily reflect BIDS and/or UBC policies)
Dr. Zabid Iqbal is a Research Fellow at Bangladesh Institute of Development Studies (BIDS), Bangladesh and is a Pos-doctoral Research Fellow at the University of British Columbia (UBC), Canada. zabidecon@gmail.com