ADAPTING SOCIAL SAFETY NET OPERATIONS TO INSECURITY IN THE SAHEL

, and a repertoire of options for adapting social safety net projects to unprecedented levels of insecurity. It fills an operational knowledge gap regarding project design, implementation, and supervision under insecurity – as called for by the World Bank Group’s Strategy for Fragility, Conflict and Violence: 2020-2025 . Based on a mix of desk research and field insights, we map operational security risks and identify ingredients for an appropriate response in risk assessment, design, beneficiary targeting, and payment systems.

Uncertainty, fluid conditions, and higher risks of violence against vulnerable populations defy the sequential approach of the standard project cycle.This recognition lies at the heart of the World Bank Group's new Strategy for Fragility, Conflict, and Violence: 2020-2025.Given the protracted and complex nature of FCV, development actors must prioritize interventions that prevent the drivers of fragility from materializing into conflict and violence.But even if a crisis has erupted, development actors must now be on the ground.This implies an acceptance of higher risks, and a willingness to work closely with partners at the humanitariandevelopment-peace (HDP) nexus.Remaining engaged also requires proactive risk management, combined with targeted and rapid support when risks materialize.According to the Strategy, operations in FCV-affected states must therefore factor security issues into all phases of the project cycle: from design to implementation, supervision, and monitoring and evaluation (M&E). 10cial protection programs -and adaptive social protection (ASP) in particular -are often the first line of defense to protect the poor and vulnerable in emergencies.They are a primary means for governments to provide quick and direct support to people affected by crises. 11In response to shocks such as food insecurity, refugee inflows, and Ebola, the World Bank has supported cash transfers and cash-forwork programs in Burkina Faso, Niger, Cameroon, Central African Republic, and the Democratic Republic of Congo.Social assistance programs also make up about 60 percent of the 1,055 social protection programs implemented worldwide since the start of the COVID-19 pandemic. 12Demand for social safety net operations that must be launched and implemented in insecure environments is bound to rise.How can these projects be designed and implemented effectively when security risks are high and the World Bank's greater exposure to operating in insecure environments is new?Source: IOM, https://reliefweb.int/report/burkina-faso/central-sahel-liptako-gourma-crisis-monthly-dashboard-8-20-august-2020

NOTE 1 | NOVEMBER 2020
How can project teams navigate security risks?To answer this question, Diagram 1 depicts cascading levels of insecurity that often characterize the new development landscape in FCV locations.Whereas in the "green scenario" security risks are negligible, between "yellow" and "red" violence escalates and the presence of security actors in the project area becomes more constant -culminating in a "black" scenario in which projects must be suspended for long periods or cancelled altogether.The arrows going back and forth between each scenario are meant to emphasize an essential point: this process is fluid, unpredictable and areas switch along the "yellow-orange-red" continuum within a matter of days or weeks.The principles identified in the following section should help navigate this challenge.

Insecurity in Operations:
A Stylized Risk Assessment Framework 4 Photo by Ollivier Girard/CIFOR we provide a repertoire of adaptations in social safety net projects which can be customized to specific projects (Annex I).Third, we offer a typology of security actors and risks in the Sahel which practitioners can use for project preparation and supervision (Annex II).This paper is part of a series of operational and policy notes supported by the Sahel Adaptive Social Protection Program (SASPP).These notes will highlight how social protection teams on the ground are responding to operational challenges stemming from FCV.For this reason, the SASPP Operational and Policy Notes Series will reflect "work in progress" rather than present a finite set of conclusions.
In almost every crisis, development practitioners hear that "business as usual is no longer an option," and that "we cannot do more of the same."This paper aims to go a step further: it offers a framework and examples for task teams to think through operational implications of insecurity on project design, delivery, and supervision.Rather than providing specific "how-to" guidance that is too formulaic or highly context-dependent, we aim to distill a set of basic operating principles derived from a mix of ongoing desk research and insights from the field.This paper proceeds as follows: First, we sketch the contours of a security risk assessment framework to support project teams in understanding intensities of violence and identify a set of six principles of security risk management.Second,

Adapting Social Safety Net Projects to Insecurity:
What can be done?3

NOTE 1 | NOVEMBER 2020
For development specialists, the notion of becoming security stakeholders may seem like a stretch.But greater acceptance of residual risks implies a higher probability that social safety net operations, beneficiaries, and implementing agencies will all be disrupted by violence.This means that attacks against project beneficiaries and assets, interruptions of activities due to military operations, threats to contractors and implementing partners, or temporary suspensions of works are all likely to materialize and may even become more frequent.
What could task teams do?First, scenario planning about what might go wrong and why can help shed light on implicit assumptions of acceptable risk thresholds.Making these assumptions explicit could help uncover blind spots or unrealistic expectations from Project Implementation Units (PIUs), partners, and stakeholders.Second, teams and clients can explore ways to boost the project's response capacity.Options include: (i) establishing real-time, discreet, and reliable incident reporting mechanisms; (ii) allocating contingency budgets to make room for the unexpected; and (iii) clear duty of care obligations for contractors and implementing partners -including the need for insurance against security incidents.Development actors are becoming security stakeholders -so they must be prepared 5.1

NOTE 1 | NOVEMBER 2020
For crisis response to be effective, projects teams should try to gather granular data on the spatial distribution, frequency, intensity, and types of security incidents.Teams working in Africa have gone about obtaining this information in multiple ways (see Diagram 2). 13First, through low-cost tools such as geo-enabling for monitoring and supervision (GEMS) and iterative beneficiary monitoring (IBM).These solutions equip World Bank staff, PIUs, or beneficiaries with phone-based technology to report on local perceptions of insecurity or satisfaction with project activities. 14Second, through technical or social audits hired by PIUs with project funds or third-party monitors (TPM) contracted directly by the World Bank to fulfill supervision duties in areas where insecurity makes it impossible to visit project sites.
Internally, country management units, safeguards specialists and corporate security experts can set up contingency plans and escalation protocols based on predefined trigger events.For example, when a project area becomes less safe -going form "orange" to "red" in Diagram 1 -the World Bank could have identified thresholds for the temporary or permanent suspension of works, relocation to safer areas, or project cancellation.Furthermore, partnering with the United Nations (UN), NGOs or private firms, can help transfer some security risk to those agencies operating under third-party execution (TPE) arrangements.
In addition to undermining project development objectives (PDOs), security incidents can also turn into reputational risks.Even if the World Bank does not have a legal responsibility for managing responses to security incidents for beneficiaries and implementing partners, there is always the risk that attacks, abductions, or allegations of sexual and gender-based violence can undermine trust in the project.Effective citizen engagement and extensive community outreach can help minimize these risks.

DIAGRAM 2. HOW TO MONITOR WHAT'S GOING ON IN INSECURE ENVIRONMENTS?
Source: Authors' adaptation from World Bank (2019), unpublished.To anticipate what can go wrong, we need granular data and intelligence 5.2

NOTE 1 | NOVEMBER 2020
Data is important, but so are logistics and intelligence insights that may be harder to quantify.Partnerships with humanitarian actors and the UN enhance development access and staff safety in hard-to-reach places -for example, by setting up arrangements for UN airlift support during supervision or sharing threat assessments for remote high-risk areas where armed groups stage frequent attacks. 15Ensuring a direct line of communication with UN peacekeeping and/or political missions could provide access to country coordination mechanisms where crucial information about security challenges is shared by security officials who are not typical World Bank counterparts.
World Bank teams can also support PIUs in setting up informal contacts with local leaders that enable them to know when high security threats are about to materialize.As these arrangements are almost never codified, determining the PIU's ability to build and maintain a network of trusted stakeholders that can provide credible information in near real-time is critical to achieving PDOs.Faith-based groups who hold unparallel access to local communities could thus be helpful partners. 16r projects in insecure environments to work, task teams need to think about the ways and means needed to adapt to ever changing conditions."Flexible design" is thus a real necessity and some concrete examples are emerging from cash transfers and cash-for-work projects in Africa.These include: (i) updating the geographical regions of project activities in the Operations Manual based on criteria linked to security to avoid costly, time-consuming restructuring later on; (ii) allowing for a Contingency Emergency Response Component (CERC) to add flexibility to changing circumstances; (iii) ensuring that project beneficiaries can be switched between cash transfers or in-kind support depending on the levels of insecurity; and (iv) gradual rather than abrupt phasing in or phasing-out of activities.
Operational agility also bears concrete a meaning in highrisk areas.For example, the ability to switch activities and locations depending on the intensity of insecurity can be a powerful risk management strategy.Projects in Africa usually span geographies that change quickly from "yellow" to "orange" to "red" and back (as showin in Diagram 1).Yet, a project site going "red" or "black" due to attacks from armed insurgents or community militias does not necessarily imply temporary or full suspension of activities if the PIU can quickly redeploy to "orange" or "yellow" sites.

At the geographic level, the presence of forcibly displaced populations can be an additional indicator when selecting a region or community. At the individual level, the process of physically surveying individual households for a Proxy
Means Test (PMT) 17 or Household Economic Analysis (HEA) 18 will be riskier, as it requires enumerators to spend substantial amount of time in an unsafe environment.This kind of survey cannot be replaced by a phone-based survey, but it can be simplified to alleviate risks.For example, rather than enumerating a large sample of dwellings, a subgroup can be pre-selected according to criteria observable at a distance (or known to community representatives) such as the material of roofs or walls.The survey is then administered only to the smaller subgroup, and with an abbreviated questionnaire, substantially shortening the time spent by people traveling and administering the targeting process.If humanitarian agents have already enumerated the area for particularly vulnerable groups such as IDPs, their database may be sufficient to support or replace a PMT.For cash-for-work projects, self-targeting based on the announced (normally very low) wage is a viable alternative.
In high-risk areas, flexible design and agility are more than usual buzzwords 5.3 Strategies to target the right people in high-risk places must also change 5.4

NOTE 1 | NOVEMBER 2020
Beneficiary payment systems, no matter how versatile, remain at risk.For example, accompanying cash distribution with security escorts (private or national forces or UN peacekeepers) can be a double-edged sword: military presence can serve both to deter, as well as to attract attacks from extremist groups or rebels.Moreover, some national forces have a doubtful human rights record. 19witching to mobile money services is an option, but phone penetration in target areas often remains a challenge at the outset.However, the prospect of populations' increased purchasing power via cash transfers can alter mobile operators' business plans towards higher coverage.Allowing an alternative household member to collect cash transfers could help project beneficiaries who are unable to attend registration and payment events.But such arrangements need to be carefully arranged and monitored to avoid possible misconduct.

Even when PIUs work hard to alleviate shocks to the poor and vulnerable, other elements of the state -both civilian and military -can use violence to undermine this process.
As such, ensuring that clients adhere to the security and safety-related provisions of the World Bank's Environmental and Social Framework (ESF) becomes a must-do for effective risk management.More specifically, governments must now prepare Security Risk Assessments (SRA) and Security Management Plans (SMPs) to comply with the Environmental and Social Standard 4 (ESS4): Community Health and Safety.SRAs and SMPs identify threats to human security and specify how the government will act to minizme them -including how it will ensure that the possible deployment of armed forces in project areas does not cause harm to local populations.
Evaluating security risks is part of safeguards' requirements for World Bank due diligence.Security due diligence assessments help project teams and decision-makers think in a structured way about security threats, vulnerabilities, and risks -and their impact on project development objectives and sustainability.Apart from identifying security issues that clients should address, they can provide a playbook for implementation support and adaptative risk management. 20nother lesson from the MENA region is the importance of providing adequate resources for both governments and project teams to conduct these analyses, as well as to monitor and ensure compliance.

The purpose of this paper was to provide a framework, key principles, and a repertoire of options for adapting social safety net projects in the Sahel to unprecedented levels of insecurity. The findings are drawn from ongoing
Photo by Ollivier Girard/CIFOR Versatile payment systems surpass some, but not all the security-related access constraints 5.5 In FCV locations, remember that the state can be a source of insecurity 5.6

NOTE 1 | NOVEMBER 2020
projects, hence they reflect work in progress.The adaptations to insecurity for project design, delivery chain, and supervision are described in greater detail in Annex I.
For the benefit of busy task teams, we conclude with Diagram 3, which depicts operational adjustments according to the levels of insecurity identified in Diagram 1.In the "green scenario" security risks are negligible, and operations can proceed normally with field visits and on the ground supervision.Between "yellow" and "red," violence escalates and the presence of security actors in the project area becomes more constant.Adjustments to operations here are additive: adjustments in "yellow" can be carried over in other contexts which require additional tweaks and amendments to the traditional modus operandi.Between "yellow" and "red," operations can be characterized by increased reliance on technology, partners, and third parties, as well as the simplification of interventions, and a growing conflict sensitivity and social cohesion lens.This process culminates in a "black" scenario in which project activities must be cancelled altogether.But whether or not a "black" scenario leads to full project cancellation will in part depend on our ability to operate simultaneously in areas of differing security risks, and thus to add a new dimension to the meaning of Adaptive Social Protection.In countries with fast-changing security contexts, teams may want to consider even safer regions to be treated as "orange" to allow them to anticipate any potential security changes.

Conclusion:
From Principles to Practice 6 • Regular project preparation and identification, adaptation to country context based on lessons learned, best practices, etc.
• Emphasize inclusion, transparency and social accountability measures, as well as building national technical and institutional capacity.
• Regular project preparation and identification, adaptation to country context based on lessons learned, best practices, etc.
• Regular implementation support: World Bank staff can access project without any security-related constraints or safety risks.
Consider ICT add-ons to im-prove data collection.

ANNEX I: ADAPTATIONS OF SOCIAL PROTECTION OPERATIONS TO INSECURITY
This table identifies a set of adaptations in operational design, supervision, and implementation in response to intensifying violence.For each level of violence depicted in the Diagram 1 -ranging from green to yellow, orange, red, and black -task teams can explore a repertoire of options such as phased deployment, mobile payments, ex-ante operational security due diligence, and third-party monitoring or execution (TPM/E).As the intensity of violence changes quickly and unpredictably, these options must be considered on a spectrum and are not mutually exclusive.• LIPW: Self-selection with PMT or selfselection with lottery.Take into account potential ethnic sensitivities.Give priority to local workers as well as youth to be employed on the site.

REGISTRATION
• Rural areas: Provide beneficiaries with program/project ID, ensure photo on ID, and include alternative household member who would be allowed to collect transfer should main beneficiary be unable to.
• Urban areas: Consider registration points in the relevant targeted neighborhoods and undertake a staggered and scheduled registration process.Ensure photo on ID and include alternative household member who would be allowed to collect transfer should main beneficiary be unable to.
• ICT: Register biometric information as a priority, or QR codes on cards (or similar) as a second priority.

PAYMENTS
• • LIPW: (i) consider using a conflict sensitivi-ty index to determine geographic targeting (ii) consider the risk of completing PW in rural areas if high insecurity in neighboring regions/area vs. need to provide access to infrastructure to communities; (iii) coordi-nate on target areas with partners, ensur-ing no sensitive areas or disenfranchised ones are excluded, or that certain areas are overly targeted (South Sudan, Yemen, Niger); (iv) self-selection with lottery (CAR).

REGISTRATION
• Rural areas: Provide beneficiaries with program/project ID, ensure photo on ID, and include alternative household member allowed to collect transfer should main beneficiary be unable to (Niger, Somalia) • Urban areas: Consider registration points in the relevant targeted neighborhoods and undertake a staggered and schedule registration process.Ensure photo on ID and include alternative household member allowed to collect transfer should main beneficiary be unable to.

OPERATIONAL SECURITY DUE DILIGENCE
• FCV country analysis: Consult risk and resilience assessment (RRA).
• Operational security risk assessment: Perform analysis of spatial patterns of violence, relevant security actors and perpetrators of violence (state, non-state, foreign), and potential security threats to project beneficiaries: loss of life, physical injuries, trauma, extortion, SGBV, abuses by security forces.Increasingly performed as part of the Environmental and Social Framework (ESF) due diligence.
• Stakeholder mapping: Map all potential partners suitable for Third-Party Monitoring and Execution (TPM/E) in insecure areas: NGOs, UN agencies, faith-based groups.
• Determine partner capacity: Do central and field staff have standard operating procedures (explicit or implicit) for operating in insecure environments?Do they have collaboration protocols with the UN system or with bilateral aid agencies to share data?
• When engaging with UN or INGO, assess the compliance of their safeguards' framework to ESF requirements and identify gaps and additional resources needed to satisfy WB ESF requirements

ENVIRONMENTAL & SOCIAL FRAMEWORK
• Factor-in security risks in World Bank due diligence (see operational security risk assessments).
• Security Risk Assessment (SRA): Borrower to perform a SRA independently or as part of the Environmental and Social Impact Assessment (ESIA).See ESF/ESS4: Community Health and Safety.Critical to assess and decide if security forces will be used/how they will be trained.
• Incident reporting and contingency plans: Protocol for rapid incident reporting and emergency procedures -for both World Bank and Government.
• Ensure the Environment and Social Incident Response Toolkit (ESIRT) procedures are known to local implementers (identify specific safety focal points).
• Assess in advance whether project benefits would be reimbursed or relatives would be compensated in case of death (Mali: upon police report, CTs lost after an armed group would be reimbursed by government at the end of the transfer cycle).

RED: SECURITY RISKS ARE HIGH AND OMNIPRESENT.
Critical

CONTINGENCY PLANS
• Partial/temporary project suspension for short-term withdrawal with openness to reengage when conditions improve.
• Plan for how equipment will be temporarily secured or stored.
• Plan for how long PIU staff would be maintained.
• Agreement with the Government on criteria for re-engagement.
• Plan for crisis management/reputational risk/ relationship with the Government.
• Clear communication with implementers and contractors.Implement a clear plan and timeframe to settle contractors payments.
• TPE if the Government can no longer access areas.
• Updated community awareness strategy.

N/A N/A CONTINGECY PLANS
• Consider full and immediate project suspension followed by cancellation.
• Plan for crisis management/reputational risk/relationship with Government.
• Updated community awareness strategy.• Loss of life, physical injuries, trauma.
• Loss of productive assets and livelihoods due to kidnapping, extortion, and/or physical or mental harm.
• Recruitment into extremist organizations to compensate for livelihoods.
• Threats to staff safety, travel restrictions, security escorts.Interruptions of missions for long periods resulting in delays of activities.
• Insecurity results in inadequate beneficiary targeting or flawed design, undermining of PDO.
• Limited ability for supervision activities leading to dependence on third parties with greater UN access: local NGOs, private firms, United Nations agencies.Increased budgetary costs.
• Potential loss of life, kidnapping, extortion of government, partner NGO/INGO staff, or contracted consultants.
• Increase in costs due to reliance on Third-Party Execution (TPE).
• Reduction in the ability to implement the project (temporarily or permanently).
• Necessity to deploy security forces at project sites.
• Risk that project facilitators and implementing were conflict parties (current or former recruits in violent extremist groups).

ANNEX 2: SECURITY STAKEHOLDERS AND SECURITY RISKS IN THE SAHEL
The table below maps security risks for three types of stakeholders: project beneficiaries, the World Bank, and government counterparts.Risks such as potential loss of life, sexual and gender-based violence (SGBV) or violent community resistance are grouped according to the main perpetrators of violence in the Sahel.These actors include: (i) violent extremist armed groups or insurgencies; (ii) community self-defense militias formed in response to these groups; (iii) national or international security and defense forces -including United Nations (UN) peacekeepers; (iv) organized crime networks; and (v) host communities, refugees, and IDPs -to the extent that inter-communal tensions between potential project beneficiaries can flare up and undermine development objectives.• Loss of life, physical injuries, trauma due to inter-communal violence surrounding access to land, grazing rights, imposition of strict security measures by community militias to prevent banditry, etc.
• Forced displacement due to intercommunal violence which may or may not be related to militias.
• Threats to safety and severe travel restrictions • Community resistance to the project due to fears of perceptions of unfairness in beneficiary targeting or location selection (i.e.projects for IDPs and refugees can be perceived as favoring some vulnerable groups over others).
• Potential desire of community militias to be paid or play a role in project activities.
• Potential loss of life, kidnapping, extortion of government, partner NGO/INGO staff, or contracted consultants in case they belong to different ethnicities.
• Community resistance if association with the state becomes a security risk vis-à-vis violent extremist groups or secessionist movements fighting the state.

Security and Defense Forces
Military, police, gendarmerie, of Burkina Faso, Mali, Niger, Cameroon, Chad, Mauritania, Operation Barkhane (France), UN peackeepers.National forces prone to human rights abuses, weak discipline and combat effectiveness.
• Human rights abuses, including arrests and potential summary executions of beneficiaries from ethnicities overly represented in extremist groups (i.e.Fulani/Peulh).
• Association with security actors or abuses by security forces in project areas transforms World Bank into a potential target of violent extremist armed groups.
• Necessity to discuss with governments the rationale and risks of deploying security forces on project sites (Environmental and Social Framework).
• Pressure on the World Bank to be required to pay through project proceeds for security ar-rangements (i.e.payment for security forces).
• Potential security risks for government implementing agency and partners if accompanied by security forces.
• Community resistance due to abuses by security forces.
• Loss of life, physical injuries, trauma.
• Loss of productive assets and livelihoods due to kidnapping, extortion.
• Distortion of local markets.
• Recruitment into criminal networks.
• Insecurity related to organized crime results in inadequate beneficiary targeting, flawed design, reputational risks, undermining of PDO.
• Ineffectiveness of cash transfers due to distortions of local markets.
• Elite capture and participation in organized crime potentially enhances fiduciary risks and project integrity.
• Potential tensions among host communities, refugees, and IDPs due to pressures on scarce resources or the perception of differential treatment by authorities could affect project activities.
• Community resistance to the project due to fears or perceptions of unfairness in beneficiary targeting or location selection (i.e.projects for IDPs and refugees can be perceived as favoring some vulnerable groups over others).Unpredictable flow of displacement.
• Overall loss of trust form communities, disengagement with the project.

DIAGRAM 3 .
Implementation adjustments to levels of insecurity

Six Principles to Adapt Social Safety Net Operations to Insecurity 5 Green Yellow Orange Red Black SECURITY RISKS ARE NEGLIGIBLE.
Absence of violent conflict, operational environment is normal, project risks related to regular implementation issues across the delivery chain: possible complaints about targeting, selection of locations, timing of payments, etc.SECURITY RISKS ARE SIGNIFICANT AND COULD ESCALATE.Low-intensity conflict related to violent extremism, insurgency, communal militias, and other crime a ecting community safety in the broader project area.Possible recruitment to extremist groups.Limited attacks on local communities hosting internally displaced persons (IDPs) and refugees.Military present for reconnaissance and intelligence, but no active operations.

SECURITY RISKS ARE HIGH AND OMNIPRESENT. Critical
intensity of violent events in or close to project area.Sustained attacks on local communities.Violent demonstrations vs. the project or government.Military operations.