Climate risk and fragility: Building Adaptive Social Protection in Sierra Leone

A reflection on designing climate-responsive safety nets in one of the world's most vulnerable economies

📄 Original World Bank Project Document:
Sierra Leone Productive Social Safety Nets and Youth Employment Project Additional Financing (P180035)

Sierra Leone sits at a dangerous intersection. The country ranks among the 15 worst climate-affected economies globally, while simultaneously grappling with persistent poverty, fragile institutions, and a youth bulge that threatens social stability. During my time at the World Bank, I worked on a project that tried to address both challenges at once—and the lessons remain relevant as climate shocks intensify across developing economies.

The Perfect Storm

The August 2022 protests in Freetown weren't just about rising food and fuel prices. They were a symptom of deeper fragility—a country where 77% of the population is under 35, where the economy needs to create 100,000 new jobs annually just to keep pace with demographic pressure, and where climate change is making everything worse.

The numbers tell a stark story. By 2050, climate impacts could reduce Sierra Leone's GDP by 9-10%. Temperature increases and erratic rainfall patterns are already threatening agriculture and infrastructure. In Freetown, unplanned development in disaster-prone areas means that the poorest communities are overexposed to floods, mudslides, and landslides. The areas with the highest concentration of poverty are precisely those most vulnerable to climate disasters.

This isn't just an environmental problem—it's an economic development crisis with a climate multiplier.

Beyond Traditional Safety Nets

The conventional social protection playbook focuses on consumption smoothing through cash transfers. Important, yes—but insufficient when your target population lives in flood-prone informal settlements and climate shocks are becoming the new normal rather than the exception.

Our project took a different approach: adaptive social protection. The idea is elegantly simple in theory but complex in execution—build social protection systems that can rapidly scale up in response to shocks, while simultaneously strengthening the systems that predict and manage those shocks.

The project had two key innovations:

1. Digital Public Works for Climate Data

We designed a pilot program that would provide short-term employment to 2,000 urban youth while simultaneously collecting critical climate risk data in Freetown's disaster-prone areas. Think of it as killing two birds with one stone—or more accurately, addressing two market failures simultaneously.

Young people, particularly women and persons with disabilities, would be trained and paid (at minimum wage rates, around $1.45/day) to:

Digitize historical records of climate disasters
Map evacuation routes and shelter locations
Document WASH facilities in vulnerable areas
Identify sources of flood risk
Assess the quality of dwelling construction near hazards

This wasn't make-work. Sierra Leone has a chronic shortage of digitized disaster data. Much of the historical information on floods, their duration, scale of damage, and impacts exists only in paper records scattered across government agencies. Without this data, you can't build effective early warning systems or target disaster response.

The economics were compelling: estimated per-beneficiary costs of $450, with projected net present value ranging from $129 to $4,541 per individual depending on impact assumptions. Even the conservative estimate suggested positive returns.

2. Operationalizing Shock-Responsive Systems

The second pillar focused on system-building:

Expanding the social registry to capture households in disaster-prone areas
Developing data-driven triggers for early action before weather-related hazards
Institutionalizing contingency plans to scale up cash transfers during disasters
Improving coordination between social protection agencies and disaster management authorities

Sierra Leone had already demonstrated capacity during COVID-19, providing emergency cash transfers to over 65,000 households through the Ep Fet Po program. The emergency response (ECTI and ECTII) showed that the delivery systems could work—71% of recipients were women, and there were measurable positive impacts on labor supply, albeit short-duration ones given the one-off nature of the transfer.

The challenge was to move from ad-hoc crisis response to systematic, predictable, and pre-financed shock response.

The Implementation Architecture

The institutional arrangements mattered as much as the program design. We built on the National Commission for Social Action (NaCSA), which had successfully implemented social safety nets for nearly a decade, but added new partnerships:

Freetown City Council provided local disaster-prone area identification and leveraged their "Transform Freetown" initiative
The National Disaster Management Agency supplied hazard mapping and disaster definitions
The Anti-Corruption Commission managed grievance redress mechanisms
Statistics Sierra Leone supported geographic targeting
Local universities partnered on beneficiary recruitment and skills training

This multi-agency coordination wasn't bureaucratic box-checking. In fragile contexts, implementation capacity is often the binding constraint. You need to leverage existing systems while carefully building new capabilities.

The Hard Tradeoffs

Every development project involves difficult tradeoffs, and this one was no exception.

The Skills vs. Poverty Dilemma: Digital tasks require some level of literacy and technical capability. But the social protection objective demands reaching the most vulnerable. We addressed this through tiered task design—grouping activities by skill requirements and matching them to youth profiles. More complex digitization tasks went to those with more education; simpler field mapping went to those with less.

The Self-Targeting Puzzle: Setting the wage at minimum wage ($1.45/day) meant only youth with limited alternatives would participate. But was this low enough to effectively target the vulnerable, or so low that it would discourage the semi-skilled youth we needed for quality data collection? This tension is inherent to public works programs.

Short-term Employment vs. Long-term Skills: With just 15 days of work per participant, could we provide marketable skills beyond the immediate task? We tried to address this by linking to the broader Youth Opportunity Platform under the project, but the binding constraint remained the brevity of the engagement.

Gender and Disability Inclusion: At least 50% of beneficiaries needed to be women, including women with disabilities. This required targeted outreach through civil society organizations, flexible working hours to accommodate household responsibilities, and careful attention to task selection and GBV risk management.

What the Economic Theory Says

The underlying economics here draws on multiple literatures:

Labor market theory suggests that public works programs can provide income support while building human capital, though the evidence on lasting labor market impacts is mixed. The key is whether the skills developed are transferable to other employment.

Information economics points to the value of reducing information asymmetries in disaster response. Better data on vulnerabilities and climate risks should, in theory, improve the allocation of scarce disaster response resources.

Behavioral economics reminds us that cash transfers can reduce negative coping mechanisms—pulling children out of school, selling productive assets, taking on predatory debt—that entrench poverty. The question is always: are the transfers large enough and timely enough to matter?

Political economy warns that in fragile contexts, elite capture and corruption are real risks. Hence the emphasis on third-party payment providers, anti-corruption commission involvement, and community monitoring.

The Financing Reality

Total project financing reached $140 million (parent project plus additional financing), with the digital public works pilot receiving $2 million from the Global Risk Financing Facility. The CERC component mobilized $6.1 million for emergency cash transfers when fuel and food prices spiked in 2022.

These numbers sound large, but context matters. With 35,000 households receiving emergency transfers, that's about $174 per household. With 2,000 youth in the digital works program, that's $1,000 per beneficiary including all overhead.

The economic question isn't whether these amounts are large enough to transform lives—they're clearly not. The question is whether they're sufficient to smooth consumption during shocks, prevent negative coping strategies, and build the data infrastructure needed for better future responses.

Lessons for Adaptive Social Protection

Three insights stand out from this work:

First, data is infrastructure. Developing countries systematically underinvest in the data systems needed for effective disaster response and climate adaptation. Finding ways to make data collection itself a social protection activity is one path forward, though it's not yet proven at scale.

Second, institutional coordination is hard but essential. Adaptive social protection requires disaster management agencies, meteorological services, social protection bureaucracies, and local governments to work together. Each has different incentives, capabilities, and political masters. The transaction costs are real.

Third, there are no silver bullets in fragile contexts. Sierra Leone's challenges—weak institutions, limited fiscal space, vulnerability to shocks, youth unemployment—are deeply interconnected. Social protection can help, particularly when designed to be shock-responsive, but it's one tool among many needed for resilient development.

The Road Ahead

The project faces substantial risks: political volatility (there was a general election in June 2023), macroeconomic instability, limited implementation capacity, and the inherent technical challenges of piloting new approaches in resource-constrained settings. The overall risk rating was "Substantial."

But the alternative—continuing with social protection systems that can't respond to climate shocks in countries where climate shocks are intensifying—is worse.

As climate change accelerates, more countries will face Sierra Leone's dilemma: how to protect vulnerable populations from increasingly frequent and severe climate disasters when fiscal space is limited and institutional capacity is weak. There are no easy answers, but experiments in adaptive social protection—linking safety nets to disaster response, investing in climate data collection, and building shock-responsive systems—are worth trying.

Because when floods meet fragility, the cost of inaction compounds over time. The children pulled out of school during emergencies, the productive assets sold in distress, the families pushed back into poverty—these aren't just humanitarian tragedies. They're economic losses that undermine long-term development.

The work in Sierra Leone is one small attempt to break that cycle. Whether it succeeds will depend on implementation quality, political commitment, and the continued support of development partners. But the approach—building climate-responsive social protection systems that simultaneously address youth employment and data gaps—offers a template worth refining.

This post reflects on work done as part of the World Bank's Sierra Leone Productive Social Safety Nets and Youth Employment Project Additional Financing (P180035). Views expressed are my own and do not represent those of the World Bank, the IMF, or any other institution.