The ‘Building on government systems for shock preparedness and response: the role of social assistance data and information systems’ webinar took place on 7 February 2019. It coincided with the launch of a publication and accompanying infographic on this topic (see full page here).
The webinar was led by the Australian Department of Foreign Affairs and Trade (DFAT), in coordination with the World Food Programme (WFP), and the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ). The event was moderated by Clare O’Brien (Senior Consultant, WFP), who was joined by speakers Valentina Barca (Senior Consultant, OPM), Rodolfo Beazley (Senior Consultant, OPM), and Tom Mtenje (Deputy Team Leader, Social Protection Programme, GIZ Malawi).
The contribution of social assistance data and information in preparedness and response to shocks
Valentina Barca and Rodolfo Beazley opened the webinar, referring to the Grand Bargain of 2016, which pledges practitioners to use existing resources and capabilities to shrink humanitarian needs over the long term. Routine social assistance data and underlying information systems can potentially make an important contribution to preparedness and response to shocks, as:
- They can support vertical and horizontal expansions, play a fundamental role for new programmes piggybacking on existing systems, and also provide important information for design tweaks and shadow alignment (see more on these concepts here).
- They differ from other government information systems, as they are a source of household (and not only individual level) data; comprehensive socio‑economic data; operational data (that is useful to identify, trace and deliver benefits); often geo‑referenced and always geographically‑disaggregated data; and (in an increasing number of countries) data that can help to capture shock vulnerability in advance of a shock.
- They also sometimes feature interoperability or data sharing arrangements with other government registries and are underpinned by established capacity to collect, store, and manage data.
- They could perform better that alternatives in terms of guaranteeing timeliness of response, coverage of population and needs, reducing costs and duplications and ensuring predictability and sustainability.
Yet, is it the case that using existing data is better than alternatives, e.g. collecting new data from scratch?
To start with, social assistance registries and information systems differ greatly depending on the country (see Figure above for some core variations), affecting how existing data can be used in an emergency. The panellists (see Section 3 in the report) proposed six complementary dimensions that can be used as a framework to assess their potential utility in response to shocks:
- Completeness (coverage): Referring to the number of records compared to a system with 100% coverage – the higher the coverage, the higher the potential. Important distinctions need to be made between data on beneficiaries (much lower coverage) and registered non‑beneficiaries, acknowledging that neither are likely to offer full coverage of populations affected (and that different countries have very different levels of coverage). However, given that shocks can affect the poorest disproportionately, social registries often have significant overlaps with affected populations post-shock as recent examples from Latin America have shown.
- Relevance: is suitable for the intended purpose (given that it was originally collected for a different purpose)? This can be analysed from two perspectives:
- Targeting: Is the available data capable of predicting households that may be in need of support? Often existing social assistance data focuses on chronic poverty and not resilience to shocks and some aspects of post-shock scenarios cannot be assessed in advance. Yet there are examples of countries that are moving towards strengthening data to include vulnerability to risks and shocks and geo-referencing, such as the Dominican Republic’s Índice de Vulnerabilidad ante Choques Climáticos (IVACC) and the National Socio-Economic Registry (NSER) in Pakistan.
- Timely delivery: Data on potential beneficiaries (e.g. in social registries) often does not include operationally relevant data, such as bank accounts, unless it is explicitly collected in advance. This is possible in contexts of recurrent and predictable shocks, as is the case of Kenya’s Hunger Safety Net Programme (HSNP), which pre-enrolled populations in areas prone to shocks.
- Currency: The degree to which data are current and representing households' real circumstances at the required point in time, ideally taking into account the disaster. It is, of course, impossible for standard social protection data to reflect the reality after a disaster, meaning some form of post‑disaster revalidation is always required. The relevant factor is how up to date existing data are overall – often an issue for concern in many countries reviewed.
- Accessibility: This refers to the ease with which potential users ‑ most likely national or local government agencies and departments, or their humanitarian partners ‑ can obtain the data. Accessibility can vary widely depending on who the users are and what processes and authorisation levels are in place for data sharing; the underlying policy and legislation; whether or not data are maintained and stored digitally; existing provisions for data security and privacy; what type of data interfaces are provided, etc
- Accuracy: Data is accurate and trustworthy if it is free from errors and omission, which affects usability.
- Protection: Data is defined as secure when it is protected against unauthorised access, misuse, or corruption. Data privacy is guaranteed when protecting an individual’s privacy preferences and their personally identifiable information. In emergency contexts, concerns regarding misusing or losing such information – potentially exposing households to further vulnerability – are heightened.
In practical terms, data and information systems can be used along the different stages of the shock cycle – as also discussed in Section 4 of the Report:
Focusing on the ‘Response Phase’, vertical expansions and new programmes piggybacking on beneficiary data are fairly common (all necessary data is already available). The main problem of this is that only beneficiaries of social assistance programmes will be reached (low coverage). As in the cases of Peru and Fiji, this challenge may be avoided by vertically expanding across multiple programmes, consequently increasing coverage. In these cases, the accuracy of the targeting response depends on the correlation between the targeting design and the effects of the shock.
Using routine non-beneficiary data for horizontal expansions and new programmes could also potentially allow for a timely response. However, there have been few experiences of this globally. This model has been considered several times and often rejected, for reasons such as: lack of protocols and plans, political concerns on errors of inclusion, inaccessibility of the data and lack of relevant (operational) data.
There is also a lot of potential in leveraging existing social assistance capacity and information systems for new data collection after a shock, as exemplified by Chile’s post-disaster needs assessment: Ficha Básica de Emergencia (FIBE).
Ultimately, routine social assistance data can play an important role in responding to shocks in countries with robust enough systems – if this is complemented by extensive preparedness measures and a broader focus on system preparedness (e.g. ensuring immediate financing) – see Section 6 of the Report. Moreover, key trade-offs between timeliness (see here) and targeting accuracy (see here) will need to be tackled in advance!
Using social assistance data and information for humanitarian response: The case for Malawi
Malawi’s social assistance database is known as the Unified Beneficiary Registry (UBR), which contains both beneficiaries and non-beneficiaries. It was initially designed to contain 50% of the poorest households, and data is collected by targeted registration every four years, collecting demographic and geo-location data as well as data on asset holding, housing condition, land ownership, and food consumption.
The UBR incorporates a proxy means test score, which allows poverty ranking at the sub-district level. The data is collected using extension workers as enumerators and is subjected to a community verification process, particularly for the poverty ranking. The UBR was primarily employed to target the Social Cash Transfer Programme and Public Works programme beneficiaries.
The humanitarian response is typically led by the Malawi Vulnerability Assessment Committee (MVAC) - a multi-stakeholder mechanism that informs and coordinates the emergency humanitarian response of the most prevalent and frequent shock in Malawi, which is food insecurity arising from climatic shocks. It uses a combination of weather data, crop production estimates, the national food balance sheet, and the household economy approach to estimate food insecurity, coupled with community-based targeting to identify the most vulnerable households post-shock. Then, a community administered tool with a final list is validated by the community, with beneficiaries who will receive food or cash.
In accordance with the six criteria to define the appropriateness of information systems for shock response, the Unified Beneficiary Registry stands as follows:
- Completeness (coverage): The registry originally covered only 50% of the population in the first 10 districts. Going forward, it will cover 100% of the forthcoming districts in rural areas. Details of the expansion coverage can be found in section 3 here.
- Relevance: The UBR does contain MVAC and other food consumption related indicators but some variables become obsolete almost immediately (e.g. meals eaten in the last 7 days), speaking to the issue of the need to validate the material prior to the shock-response.
- Currency: The UBR has updates every four years but MVAC requires annual (if not seasonal) updates.
- Accuracy: There are different perceptions between implementers (reduced elite capture) and communities (reduced community participation). Data consistency is facilitated across databases with the rollout of the National ID programme with the National ID now being collected by the UBR. Geo-location data is also collected.
- Accessibility: At the current moment, special access is granted for trials on a case by case basis. Data sharing protocols exist but are yet to be formalised, pointing to the need to formalise institutional ownership of the database and the data.
- Protection: Security is currently attained by restricting access to the UBR to the two government programmes. The access and privacy protocols for other non-government users are not yet formalised.