The National Social Protection Strategy (NSPS) puts together a comprehensive strategy document for social protection that seeks to meet the social protection needs of a rapidly changing Bangladeshi economy as it moves away from a primarily agro-based rural economy towards a manufacturing and formal services based economy, along with growing urbanisation.
Correct identification of the poor is a must to successfully implement any social safety net system. It anticipates Bangladesh achieving a middle income status by 2021. The social protection needs of this emerging society are vastly different from the present food-transfer dominated safety net system that also suffers from a multiplicity of programmes, substantial leakage, low coverage of the poor and low average benefits.
Accordingly, the NSPS proposes a combination of cash-based safety net transfers along with social insurance and labour protection policies for the formal sector. In this system, the financing of safety nets will happen through the budget, while social insurance and labour protection schemes will be funded by the private sector through beneficiary and employer participatory contributions.
Understanding social protection plans
To prevent leaks and ensure proper beneficiary selection for the tax-funded safety nets, the NSPS proposes a major overhaul of the administrative system, involving the establishment of a computerised single registry of beneficiaries, establishing government to people (G2P) payments through the financial institutions, instituting a sound selection criteria and a grievance mechanism, and significant administrative reform of the delivery of safety nets.
Since the safety nets will be financed though the national budget and resources are limited, it is imperative that the selection criteria ensure that resources are channelled to the poor and the vulnerable. This group is the natural target for any safety net system.
A key question is how to identify the poor and vulnerable. The NSPS proposes a combination of income and other criteria reflecting risks faced by the poor and vulnerable populations at different stages of the life cycle. The vulnerable population is identified as small children, school-age children, the vulnerable women (widows, single mothers, unmarried single women, women socially excluded owing to profession), the unemployed, the elderly, and the physically challenged.
The application of the income criteria presents a major challenge since reliable income data is not available. As a substitute, a proxy means test (PMT) is proposed. This is fine but it is well known that PMT suffers from significant exclusion errors that need to be managed.
The selection challenge for the income criteria is illustrated in the graph that shows the cumulative per capita consumption distribution in Bangladesh based on the Household Income and Expenditure Survey (HIES) 2010.
What is striking about this figure is the fact that the consumption distribution curve is heavily concentrated around the national poverty line. In 2010, 80 percent of the population consumed below 2 times the poverty line, close to 70 percent consumed less than the 1.5 times the poverty line and 50 percent of the population consumed less than 1.25 times the poverty line.
These suggest that the size of population that is both poor and vulnerable on income grounds is large. Thus, a shock, such as a major illness or other social crisis could easily push a large population below the poverty line, even if transitorily.
A related important question is how successfully a PMT will identify the poor. Since it is based on projecting income based on measurable proxy variables, there are understandably large prediction errors resulting from sampling errors, measurement errors and the like. These errors include both exclusion errors (excluding the poor) and inclusion errors (including the non-poor). Empirical work shows that these errors are large for small coverage but they fall with higher coverage. Thus, using a somewhat higher cut-off point to determine income eligibility will considerably reduce the exclusion risk, thereby, including many more poor.
Accordingly, the NSPS proposes an income cut-off point of 1.25 times the poverty line. This is a small increase in the income cut-off aimed at reducing exclusion risk and also bringing in the highly vulnerable population under safety nets. However, the income criteria is only one variable. The other selection criteria deal with gender, age, physical disability, and marital status.
The application of all these criteria yields a target population of poor and vulnerable of about 36 million in fiscal 2017-18 when the NSPS is expected to be fully functional. This amounts to only 22 percent of the population, which is a far cry from the 50 percent highlighted in The Daily Star report.
The 50 percent refers to the population that is consuming below 1.25 times poverty and is used to define income eligibility. The NSPS does not suggest providing income transfers to 50 percent of the population. Other criteria are used to prioritise and narrow down the actual recipients of transfer payments, which as noted is only 22 percent of the population.
The total cost of the life cycle based safety net transfers from the budget is estimated at 1.4 percent of GDP, excluding the government service pensions. Some 58 percent of the benefits go to children, 24 percent to the elderly and 13 percent to vulnerable women; the remaining 5 percent go to disabled people and freedom fighters.
These beneficiaries are either poor or concentrated around 1.25 times the poverty line. The entire group is highly vulnerable to all kinds of shocks. This small investment will likely yield substantial benefits in terms of social capital formation. Simulation results show that the coverage of the poor and average benefits will be much higher in the NSPS than the present safety nets. As such, the impact of the NSPS in terms of poverty reduction and lower depth of poverty will be larger than the present safety net system that has significantly higher budget spending (1.7 percent of GDP excluding the government service pensions).