Reimagining Sri Lanka’s Social Safety Nets Systems
Authored by Ms. Ayushka Nugaliyadda and Ms. Sarika Warusavitarana of Citra Social Innovation Lab, and Ms. Gayani Hurulle and Mr. Merl Chandana of LIRNEasia.
A joint article by the Citra Social Innovation Lab and LIRNEasia aimed at sharing an evidence-based perspective based on the engagement of both teams in the social protection space, to inform national policymaking. The article explores how the operation of social welfare benefits schemes might be revised to ensure that they are responsive to rapid contextual changes and are able to reach and meet the needs of newly vulnerable groups. In light of the challenges posed by existing systems, the article proposes a tri-part model focused on registration, verification and the provision of cash disbursements.
Introduction
The Government of Sri Lanka currently implements a range of schemes that focuses on providing safety nets (social assistance) to vulnerable groups — the Samurdhi Programme operates as its primary initiative, adopting a triple bottom-up approach by not only providing social assistance, but also microfinance and livelihood development. However, the legal and institutional environment is complex and siloed. As reflected in the UNDP’s Comprehensive Landscape Mapping of Current Social Protection Systems in Sri Lanka conducted in 2021, the programmes fall short of providing adequate coverage due to systemic inefficiencies. The Department of Census and Statistics estimated that in 2019, 13 social assistance programmes, including Samurdhi, and disability and senior citizens’ allowances, only led to the proportion of the population living in poverty reducing from 18% to 14.3%.
Emergency cash transfers were mobilized during the COVID-19 pandemic, and subsequently the economic crisis. These were not only extended to existing beneficiaries of select schemes, but also those on beneficiary waiting lists, and families who had lost livelihoods due to the pandemic who could self-register for the benefits. As highlighted by LIRNEasia’s 2021 research, this allowed for greater coverage, with 86% of the poorest households receiving benefits (vs. 45% for regular benefits). However, this database was not collated nor digitalized, and was therefore not available for use to support decision-making or any further rounds of cash transfers; this was a missed opportunity. Digitally enabled and data-driven processes have the potential not only to minimize such administrative inefficiencies but also improve coverage and targeting. This article examines the extent to which digital approaches and data have been included in existing and proposed processes, and highlights areas for further integration.
Source: LIRNEasia COVID-19 Impact Survey, 2021. Representative of all households in Sri Lanka with a +/- 2.8% margin of error at a 95% confidence interval (n = 2,501). SEC refers to the tiers of socio-economic classification in Sri Lanka, with ‘SEC A’ representing the highest tier and ‘SEC E’ representing the lowest tier.
Registration of New Beneficiaries:
The World Bank’s Social Safety Net Project, which was initiated in 2017, has laid the groundwork for using digital approaches and data for improved equity, efficiency, and transparency in Sri Lanka’s social safety net programmes. It proposes a Unified Social Registry Information System (SRIS) containing data of all current and former beneficiaries and new applicants, and developing an Integrated Welfare Management System (IWMS) for harmonised management of all programmes.
There has also been a re-activation of the Welfare Benefits Board (WBB), which has been mandated to develop a digitalized social registry that would identify beneficiaries and feed into the IWMS. Currently, the WBB is spearheading a fast-tracked registration process to capture three categories of beneficiaries: (1) those already registered in the four main social protection schemes in the country (Samurdhi, senior citizens allowances, and disability benefits and Chronic Kidney Disease benefits); (2) those on the waiting lists for these four schemes; and (3) those not captured in the existing databases or waiting lists, but would like to register — i.e., the ‘newly poor’ and those that have experienced exclusion errors thus far. Registration began in mid 2022 and over 3.7 million applications were received by March 2023. Noteworthy however, is that during LIRNEasia’s research fieldwork in late 2022, some individuals (mostly newly poor) were unaware of the reregistration programme despite mass media campaigns. In contrast, those who were already receiving benefits and had pre-existing relationships with local officials were more aware of the need to reregister. Therefore, while this initiative is a welcome and necessary change to the status quo, it is imperative that communication channels are continually improved to allow those who feel they need assistance to register for the programme.
Verification of Registrants:
The submitted documents are then verified against pre-defined criteria for eligibility. At present, the criteria being utilised by the WBB is primarily based on the selection criteria for welfare systems defined by the Extraordinary Gazette № 2128/24 of 2019. However, determining the eligibility of beneficiaries currently requires manual cross-checks against the criteria — a task that not only requires a great deal of training around data collection, sensitivity and coordination, but is also prone to considerable human error, delays, and reluctance and/or a lack of knowledge on how to use digital means of collecting information. It is therefore our view that this verification process could be further strengthened by utilising non-traditional sources of near-real-time data, if available and can be utilised for automated cross-checking, thereby reducing the errors and resource constraints of extensive primary data collection. Some potential examples include:
- Mobile phone usage data on consumption variables, mobility variables and strength of social networks
- Household electricity consumption
- Bank data indicating non-performing loans
- Supermarket data, where available and where on loyalty schemes, indicating purchasing patterns.
The optimal combination of these variables needs to be established for the Sri Lankan context based on the analysis of available data streams and verification with the reality on the ground. An iterative process that involves human-supervised verification and fine-tuning of the chosen criteria will be required at the beginning. Once this approach reaches desirable levels of targeting, human effort can be directed at interventions for reducing potential exclusion errors through grievance handling.
Given that Sri Lanka has passed the Personal Data Protection Act (No. 9 of 2022), it is vital to ensure compliance, both when the information is collected from beneficiaries at the point of registration, as well as the data provided on such individuals during the verification stage. This should be communicated to beneficiaries at the point of data collection. Further concerns around taking photographs and uploading documentation should be taken into consideration and duly accounted for.
Disbursement of Cash
There has been a lack of cohesion in the delivery mechanisms amongst the different social assistance programmes in Sri Lanka. Samurdhi ‘Banks’ have served as a collection point for the Samurdhi monthly cash transfers, while senior citizens’ allowances, disability benefits and kidney patients’ allowances are disbursed via post offices, state banks and divisional secretariats, respectively¹.
In the 2023 Budget Speech, the government indicated that these benefits would be delivered via bank transfers from April 2023. If is this is disbursed via all government and commercial banks, there would over 4000 points from which benefits could be collected. For those receiving Samurdhi, this is four times the number of collection points available to them at present. This would, in turn, lead to significantly less travel time, making the collection of benefits more convenient.
Over time, the government may also consider including other delivery methods, such as mobile money, and giving beneficiaries the option to choose the channel through which they wish to receive the funds.
Next Steps & Further Analysis:
The benefits of digitalization and datafication seem multi-fold. However, there are several elements that should be considered/undertaken:
- Understanding how many households we intend to target. It is crucial to accurately assess how many households are poor. UNDP and LIRNEasia are both conducting nationally representative surveys to understand poverty, including multidimensional poverty, in Sri Lanka. UNDP is collecting household-level data on household finances, coping strategies and access to basic necessities and social services across all 25 districts, with a sample size of 25,000 households. LIRNEasia’s survey across 25 districts, with a sample size of 10,000, is collecting data on income and expenditure, and indicators to measure multidimensional poverty.
- Assessing if the criteria being used by the government are good proxies to identify the poor. The Government is utilizing 22 indicators (as per the Extraordinary Gazette №. 2128/24 of 2019) for identifying those in need of benefits. However, the extent to which it can correctly classify those in poverty is unclear. LIRNEasia will use data from its nationally representative surveys to better understand this. Further, an analysis which examines the feasibility of using electricity consumption data as a proxy indicator of poverty is also underway, which if deemed sufficiently reliable, will allow for quick and easy verification.
- Evaluating the availability of sharable digitalized data streams to integrate into the verification process. If any criteria (those identified by the government at present, or any others) are found to be good proxies to identify those in need of benefits, it would be important to understand: (i) whether the government has access to the data; (ii) if the data is digitalized; and (iii) how often it is updated. LIRNEasia is undertaking an initial audit of this for the 22 criteria used by the government at present.
- Identifying if the poor are receiving means-tested social assistance. Sri Lanka has multiple means-tested social assistance schemes, but pre-crisis data indicates that there were targeting errors with many inclusion and exclusion errors. LIRNEasia will use its nationally representative survey of these errors for multiple schemes that are supposed to target the poor, including but not limited to Samurdhi cash transfers and senior citizens’ allowances.
- Classifying chokepoints and opportunities for digitalization in all social assistance programmes: There is a need for a comprehensive assessment of registration and delivery processes associated with all social assistance programmes– both de jure (on paper) and de facto (in practice), as highlighted in UNDP’s Comprehensive Landscape Mapping of Current Social Protection Systems in Sri Lanka. LIRNEasia is undertaking qualitative research in the form of focus group discussions, in-depth interviews and key informant interviews across 12 districts to understand these from a citizen’s point of view for several social protection programmes.
- Conducting a comprehensive advocacy campaign: To operationalize the above-mentioned mechanism in a rapid and comprehensive manner, it is vital to simultaneously roll out a comprehensive national awareness campaign, to ensure that the public is always aware of the basis for all measures taken going forward and the immediate and long-term impacts expected, leveraging the current public appetite for transparency and accountability and ensuring its continuation. Information should be presented in an easy-to-understand format, including eligibility criteria, simple steps to be followed to register, and who to approach for troubleshooting and grievance redressals. All content should be disseminated in both local languages, including sign language, to target beneficiary groups effectively and inclusively.
To create evidence-based, citizen-centric, effective, and efficient social protection systems, it is essential to conduct research not only on the issues outlined above, but also on the challenges involved in implementing such mechanisms, especially cash disbursements, and to identify the changing needs of newly vulnerable groups at the grassroots level.
[1] There was a short period of time in which senior citizens’ allowances, disability benefits and kidney patients’ allowances were also channelled through Samurdhi Banks, though this has changed.
