Rebuilding Sierra Leone: A Ten Year Plan, 2027 to 2036

Stability first, then services, then production. A ten year plan that says what it costs and who has to deliver it.

Sierra Leone did the hard part in 2025. Inflation fell to 4.4% by December, the leone stopped sliding and the budget stopped bleeding. Then a war thousands of miles away pushed up fuel prices, and by August 2026 inflation was back at 15.7%, according to the World Bank. If you buy rice, pay a bus fare or run a shop, you didn’t need the statistic to know.

Summary

The country doesn’t lack plans. The 2024 to 2030 development plan has sensible priorities. What’s missing is the chain that links a plan to the money that pays for it and to a result somebody can check. An academic review of audit reports found procurement breaches in every year from 2018 to 2024. Clinics run out of medicine. A power system that depends on rain can’t keep a hospital lit in March.

So this blueprint is built around ten moves, roughly in order of urgency.

1. Protect stability while it’s fragile, and shield poor households with targeted cash, not blanket subsidies.
2. Lift public revenue from about 11% of GDP towards 17% by 2036, mostly by closing leaks and widening the base, not by squeezing people who already pay.
3. Make public money traceable. Publish contracts, name who really owns the companies, and follow up every audit finding.
4. Pay for maintenance before new buildings, so clinics, schools, roads and water points keep working.
5. Back farmers with seed, storage, rural roads and credit, and cut rice imports without pretending the country can stop buying food abroad.
6. Reform mining so the state collects what it’s owed and some of it turns into schools, clinics and roads.
7. Get reliable power into hospitals, schools, water systems and farms first.
8. Teach children to read, and train young people for jobs employers have said they’ll fill.
9. Give district councils money by a published formula, then audit what they do with it.
10. Run all of it through a small delivery unit with a legal basis that publishes results every quarter.

None of this is quick. By 2036 it could mean prices that don’t outrun wages, a farmer who gets the harvest to market before it spoils, a health centre with power, medicine and a trained midwife, a Grade 3 child who can read a short passage and understand it, and a business owner who can register and pay tax without slipping something to someone on the side. It won’t mean a rich country, and I’d rather say that now.

Where the country stands

About 9 million people live in Sierra Leone, and the median age is around 20 (Worldometer). That’s a young country with a lot of people about to look for work. Here’s the baseline, with a source and a date beside each number.

Baseline, as of 23 October 2026

Why some numbers don’t match

Inflation first. Spring forecasts assumed a peak under 12%, and the August reading came in higher, so I’ve planned around 15.7%.

Revenue is trickier. The 10.8% figure is the Finance Ministry’s, after GDP was rebased. The IMF series shows 9.1%. Grants and the rebasing probably explain most of the gap, though I can’t prove it from the documents.

Electricity depends on who’s counting. The World Bank counts any access at all, even a small solar kit. The utility counts grid customers. A household survey found 21% on the grid or a mini grid, plus another 15% off grid. None of them tells you how many hours of power a home actually gets, which is the number people care about.

Countryside and city

In 2018, poverty stood at 74% in rural areas and 35% in towns. Freetown was at 23%. The North was poorest, with Tonkolili, Pujehun and Falaba hit hardest. Electricity, health staff and school quality thin out in the same direction, so district towns need attention from the first day, not the fifth year.

What can’t be measured yet

There’s been no national labour force survey since 2014 and no poverty survey since 2018. Other basics exist, but they’re old or partial. A 2022 survey found basic water access at 62.6% and basic sanitation at 13.6% (UNICEF). Only about 40% of primary roads are paved, and roughly a third of people live near a road that works all year (World Bank). In November 2025 the Human Rights Commission counted 5,476 prisoners in cells meant for 2,009 (Awoko). Nobody publishes a national court backlog, and nobody has a verified count of tourist arrivals. Fixing that is the cheapest useful reform in this whole plan.

Why progress keeps stalling

It’s tempting to pick one culprit. Corruption, aid cuts, the iron ore price. The evidence doesn’t back any single one. It points to several problems feeding each other.

The symptoms are easy to see: high prices after the 2026 shock, a hungry lean season, empty shelves in clinics, long blackouts. Underneath all that is an economy that earns too little. Iron ore is about 69% of mineral exports, farming is small and rain fed, and rural roads wash out every year. On top of that comes a procurement law that’s reasonable on paper and poorly obeyed. The procurement authority found contracts awarded on unapproved bidding documents (Awoko).

Shocks keep landing too. Ebola arrived in 2014 just as iron ore prices collapsed, and the 2025 aid cuts hit a health system that leaned on donors. And there’s politics. The U4 Anti Corruption Resource Centre records concern that enforcement can look selective, and notes the 2021 suspension of the then Auditor General. Those are concerns, not court findings. Still, a plan that runs on public trust has to deal with the perception as well as the fact.

Here’s how the pieces feed each other. Low revenue means no maintenance, so roads and power fail. Failed roads and power keep farms and firms unproductive, which keeps tax receipts low. Suspected corruption pushes up the price of every contract and makes people less willing to pay tax. Weak schools and clinics produce a workforce investors don’t want, so capital goes into mines, not factories. You can’t pull one lever and expect the rest to move. That’s why this plan starts with revenue, procurement, maintenance and data, not a showpiece project.

What to keep from existing plans

The 2024 to 2030 national development plan rests on five priorities: Feed Salone (food), human capital, youth jobs, public service reform, and technology and infrastructure (plan). It aims for middle income status by 2039. I’d keep most of it, with some changes.

The priorities are sound. The costs and the scorecard aren’t, so I’d re cost it and extend it to 2036. Feed Salone deserves a test, not a verdict. The government reports a 35% jump in 2023 paddy output, but nobody outside government has checked it, and the published import figures don’t fully reconcile. Free school education stays, with the focus moving from getting children in to getting them to learn. The education plan (2022 to 2026) and the child survival plan (2023 to 2025) have expired or are about to, so they need replacing.

The IMF programme stays, including about US$211.5 million approved in June 2026 (IMF). So do the Public Procurement Act 2016, the Public Financial Management Act 2016 and the Anti Corruption Act as amended in 2019. The weakness there is compliance, not the text, and a new law won’t fix a habit.

Where this should get the country by 2036

Inflation low enough that families can plan. Public money you can follow from budget to building. Rural homes with a road, power and a market. Children who read, mothers who survive childbirth, and growth that doesn’t rest on one mine.

Proposed targets

Every one of these has to be checked against baselines measured in 2027.

• Inflation: 15.7% now; below 9% by 2030, below 7% by 2033, 5% to 7% by 2036.
• Revenue, share of GDP: 9% to 11% now; 13.5%, then 15.5%, then 17%.
• Debt risk rating: high now; moderate by 2030; low to moderate by 2036.
• Reserves, months of imports: 1.5 to 2.1 now; 3.5, then 4, then 4.5.
• Poverty: 5, 9 and 12 points below the 2027 survey figure.
• People in acute hunger: about 19% in 2023; one third lower by 2030, 40% by 2033, half by 2036.
• Households with electricity: 21% to 36% now; 45%, 55%, 65%.
• Hospitals and health centres with reliable power: all hospitals and 70% of centres by 2030; 95% of centres by 2036.
• Grade 3 reading with understanding: double the 2027 baseline by 2030, 2.5 times by 2033, 3 times by 2036.
• Child deaths per 1,000: about 105 now; 85, 72, then 62.
• Maternal deaths per 100,000: 354 to 443 now; 280, 230, then 180.
• Audit recommendations acted on within a year: 50%, 65%, then 75%.
• Contracts competitively tendered, by value: 70%, 80%, then 85%.

The poverty targets assume incomes per person grow about 2.5% a year, because growth of 4.5% to 5% gets eaten up by population growth near 2%. I’ve left corruption perception scores out of the targets on purpose. A good public relations campaign can lift a perception score. Audit follow up and competitive tendering can’t be faked.

What comes first

1. Price stability and cash for poor households.
2. Revenue and procurement integrity.
3. Maintenance of what already exists.
4. Rural roads, storage and farm inputs.
5. Power for hospitals, schools, farms and small factories.
6. Skills and jobs, because about 200,000 young people reach working age each year (my estimate).
7. Mining reform, the largest single revenue lever.
8. Industry and exports, which need everything above first.

Government, public money and the courts

A quick word on costs before the detail. I give each programme as a rough share of GDP per year once it’s fully running. For scale, 1% of 2025 GDP is about NLe1.7 billion, worked out from the Finance Ministry’s own figures. Treat these as planning ranges. Each one needs a feasibility study before anyone commits money.

Parliament, civil service, elections and courts

The laws are mostly there. What’s missing is the habit of obeying them. Audit reports have found procurement breaches every year from 2018 to 2024, and nothing forces a ministry to answer. Many people also believe jobs depend on who you know, and trust in elections and courts splits along party lines. That’s a problem no statute fixes on its own.

Parliament’s Public Accounts and Finance Committees could start tomorrow. They should require a written reply to every audit finding within 90 days, hold hearings on the biggest ones and keep a public register of what happened next. It costs almost nothing. For hiring, the Public Service Commission should run published vacancies, structured tests, an outsider on each interview panel and public results, and permanent secretaries should sign annual performance contracts with published scorecards.

Payroll needs an audit too. Check teachers, health workers and security staff, then verify each person biometrically. Discipline individuals through a fair process only. I’d steer clear of anything that looks like a purge, because it’s unfair and it hands the next government a precedent.

Elections are a harder conversation. The next general election is due in 2028. The Tripartite Committee made 80 recommendations in 2024, many needing constitutional change, and by December 2025 none of the amendments under discussion had been adopted (ISS Africa). Whatever happens to those, the Electoral Commission can publish the voters’ register audit, results by polling station and the outcome of every complaint. The courts need a protected budget line, published judgments and case statistics, and judicial appointments made on published criteria.

All of this comes to about 0.1% of GDP. By 2030, every audit finding should be tracked in public and 80% of senior posts filled by open competition. If officials resist, start in five ministries and spread from there.

Corruption and public money

Sierra Leone scored 34 out of 100 on Transparency International’s 2025 index, up one point from a low base. Audit reports still show losses that come round every year. Independent observers have raised concerns that enforcement can look selective, so the first job is protecting the referees. A special anti corruption division of the High Court has existed since 2019, and the weakness is how consistently and impartially it’s used.

I’d start with the watchdogs. The Auditor General and the Anti Corruption Commissioner should be removable only for stated cause, through an open tribunal. They should serve single fixed terms and have budgets set by Parliament, so no minister can quietly starve them.

After that, publish by default. Contracts, procurement plans, awards, the real owners of contractors and licence holders, and audit follow up, all on one open portal. Move bidding online within three years for every ministry and council, and require written justification, published within seven days, for any single source award. Let data pick the high risk contracts for audit, then track whether the recommendations get carried out. Add verified asset declarations for senior officials, a law that protects whistleblowers, and disclosure of large party donations and campaign spending.

One warning belongs here. Suspects keep the right to a lawyer and to bail, charges get independent review, convictions can be appealed, and cases against any party follow the same rules. A tool used for revenge would wreck the genuine cases along with it.

The budget is about 0.15% of GDP. I won’t claim savings from competitive bidding until pilots measure them, because the claim is easy to make and hard to prove. By 2030, 70% of contract value should be competitively tendered and half of audit recommendations acted on within a year. If the independence amendment stalls, publish a binding protocol and have donors tie budget support to it. If online bidding stalls, start with the 20 biggest buyers.

Justice and security

People don’t trust slow courts and police, and they’ve got reasons. Pre trial detention is long, prisons are crowded and sexual violence is under reported. The Human Rights Commission counted 5,476 prisoners in space for 2,009 in November 2025, and no national court backlog is published, so 2027 has to establish one. The 2025 criminal session processed 1,659 inmates, and the Criminal Procedure Act of 2025 is meant to speed cases up (Awoko).

The practical fixes are plain enough. A case tracking system with published court statistics. Paralegals in every district. A police complaints body that can investigate. Strict rules on use of force, independent prison inspections, and specialised units for survivors under the Sexual Offences Act. Community safety forums help, and so does an honest national conversation about the civil war. Together they cost about 0.1% of GDP. By 2036, public confidence in police and courts should pass 50%, with no more than a 15 point gap between supporters of different parties.

Digital services and local councils

Too many payments and records are still on paper, which means delay and leakage. The government should build a payment system that works with mobile money, and a digital ID based on the civil registry, with a paper alternative for people without phones. It also needs a data protection law and an authority in place before large databases get linked. Cabinet approved a National Data Protection Policy in April 2026, and a combined data protection and access to information bill is being drafted (Ministry of Information). The Cyber Security and Crime Act 2021 already exists, though journalists have raised concerns about how it’s applied (SLAJ). Budget about 0.1% of GDP.

Councils run basic services, yet they often lack engineers, accountants and procurement staff, and the money they get doesn’t follow need. The first step is publishing which tasks belong to national government, districts and councils under the Local Government Act 2004 and its later amendments. Then pay grants by a published formula based on population, poverty and service gaps, with a bonus for clean audits. By 2033 every council should have audited accounts within a year of closing its books. This costs about 0.15% of GDP, partly replacing spending the centre does now. Local elites can capture the money, so scorecards and national audit have to travel with it.

Letting citizens in

Hold annual district hearings where ministers and council chairs answer questions on service data. Run a free complaints line with published response times, and fund independent civil society monitors through an open grant window. Authorities should reply in writing within 30 days and say what they did. The Right to Access Information Act 2013 only works if refusals can be appealed quickly. Cabinet agreed in April 2026 to replace the commission’s law with a combined data protection and access to information act, so the appeal route belongs in that bill. By 2030, 80% of complaints should be answered within 30 days, judged by an independent panel.

Keeping prices steady, and paying for it

Stability

The stabilisation was real, and it wasn’t cheap. Inflation fell from 47% in 2023 to about 30% in 2024, then to 4.4% by December 2025, thanks to tight money and a steadier leone (AfDB). Tightness hurt borrowers. Reserves are still thin, debt is rated at high risk of distress, and the energy shock pushed inflation back above 15%. I don’t think spending cuts alone can fix this. Cuts that land on clinics and schools cost growth and goodwill, so stability also needs more revenue and better spending.

Concretely, the government should stay with the IMF programme, which now includes about US$211.5 million under a resilience facility. Publish the documents and let Parliament debate the benchmarks. Budget for three years, with a target for the primary balance, a debt ceiling that matches the official debt analysis, and parliamentary approval before big supplementary spending.

Forecasts should be honest, which means basing revenue on last year’s actual collections, not hopes. Issue monthly cash plans, finish the Treasury Single Account and commit money only against cash. In the first 100 days, audit and publish all arrears and contingent liabilities, including state enterprises such as the power utility, then ban new arrears and set a schedule for the old ones.

On borrowing, prefer concessional loans. Anything else needs parliamentary approval and a published estimate of its effect on debt service. The IMF sees debt falling towards 37.5% of GDP in the early 2030s if the programme holds (IMF), but that depends on the revenue gains described below. Leave the central bank alone. No printing money for the budget, and the bank must be able to raise rates when it has to, with the government honest about what that costs borrowers.

For poor households, targeted cash beats broad fuel subsidies, which are expensive and mostly help people who own cars. Add public price monitoring for rice and fuel. The systems cost about 0.1% of GDP, and targeted shock transfers might run 0.3% to 0.6% for a year or two. The IMF’s downside case has inflation at 13% to 14.5% in 2026 and 2027 if the war drags on, so be ready to extend them. By 2030 the aim is inflation below 9%, no new arrears and reserves above 3.5 months.

Tax and revenue

Revenue is about 9% to 11% of GDP, far too low to run a decent state. Income tax made up about 52% of revenue in the second quarter of 2025 (Awoko), which tells you the base is narrow. A few people and firms carry most of the load.

My rules are simple. Widen the base before raising rates. Cut exemptions that mostly help the well connected. Make life easier for small traders. Leave basic food and household fuel alone. And never change rules retroactively for investors. The numbers below are planning ranges, as a share of GDP at full effect by 2033, and each needs a tax expenditure or customs study behind it.

• Review and end pointless exemptions: 0.6%. Mostly affects big firms and connected importers.
• Customs reform (valuation, risk based inspection, port targets): 0.8%. Needs safeguards against bribery by staff.
• Mining tax checks (royalty audits, transfer pricing): 0.5%.
• Sales tax compliance and electronic invoices: 0.4%.
• Excise on tobacco, alcohol, sugary drinks: 0.2%. Helps health too.
• Property tax run by councils, with land registration: 0.2%.
• Simple turnover tax for small traders, payable by mobile money: 0.5%.

That adds up to 3.2 points. The other 2.8 points of the climb from 11% to 17% must come from growth and from more firms joining the formal economy, and I wouldn’t promise that. Running costs are about 0.2% of GDP a year. In 2027, review exemptions, audit mining, start risk based customs checks and push digital payment. Add invoices, property tax pilots and the small trader regime in 2028 and 2029. The targets are 13.5% of GDP by 2030, 15.5% by 2033 and 17% by 2036. If politicians resist, phase the changes. If taxpayers push back, publish their rights alongside the new rules.

Finance, savings and the diaspora

High interest rates squeezed private credit, and many households rely on informal savings groups and moneylenders. Remittances help families, but they can’t stand in for public reform.

The aim is to make mobile money, bank and microfinance accounts work together, and to build a credit history system and a registry so vehicles and stock can back loans. Microfinance rules should centre on clear disclosure and fair debt collection. Partial credit guarantees for farm and small business loans can help, if they’re priced to cover expected losses. Otherwise they’re a hidden subsidy.

Diaspora bonds sound attractive, and I’d be cautious. They should go ahead only if proceeds go to named projects, repayment comes from a ring fenced source, the size fits the debt analysis, independent trustees publish audited reports, and the currency risk is spelled out. If any of those fail, steer diaspora money into co investment in projects that already pass a feasibility test. Cost is about 0.1% of GDP.

Farms, mines and the land itself

Farming and food

Rice is the staple, and for years the country has bought about a third of it abroad. Government figures put imports at about 430,000 tonnes in 2022 (Awoko). The government says paddy output rose 35% in 2023 and imports fell by over 60%, but its own numbers don’t quite agree. Going from 430,000 tonnes to under 250,000 is a fall of about 42%, and no independent source confirms either claim. I’d want independent data before planning around it. Meanwhile the World Food Programme warned in May 2026 that 78% of people were food insecure.

The causes are ordinary and fixable: poor seed, little water control, bad rural roads, losses between field and market, weak storage, insecure customary land rights and little affordable credit. I’d tackle them roughly in that order.

Seed and advice come first, from 2027. Scale up certified seed through the agricultural research institute and private growers, and put extension workers in the field with phone messages timed to planting. Next is water and land. Develop inland valley swamps and small irrigation where studies show a return, with farmer groups responsible for upkeep. Register customary land rights under the Customary Land Rights Act 2022, which bars refusing anyone land rights because of gender, age or marital status, and the National Land Commission Act 2022, so women and smallholders can invest without fearing eviction.

After the harvest, the losses begin. Drying floors, village stores and milling hubs help, but measure losses first so you know if you’ve cut them. Add input credit through cooperatives, guarantees, weather index insurance pilots and a price information service. Crop choice should follow markets: rice and cassava for local sale; cocoa and coffee for export, with the traceability buyers now demand; oil palm on smallholder and degraded land, never primary forest; vegetables for Freetown; licensed fisheries with a crackdown on illegal fishing; and poultry where feed and vets are within reach.

Timing matters as much as money. Seed and credit must arrive before planting, roughly April to June. Hunger peaks from June to August, so cash transfers and school feeding belong in that window. Stores and trucks must be ready for the September to November harvest, with roads repaired before the rains end.

The target should be a smaller import share, not zero: about 25% of rice consumption by 2030 and 15% to 20% by 2033, down from about a third. Full self sufficiency would push up food prices, and imported rice is sometimes cheaper. The programme costs about 0.7% of GDP a year, from the budget, World Bank, AfDB and IFAD loans, and donor grants. By 2033, post harvest losses should be down a third and acute hunger down 40%.

Mining

Minerals were over 76% of export earnings in 2024, and mineral exports hit US$1.3 billion in 2025 (Rio Times). Yet the state reported only about US$55.8 million from licences and royalties. Royalty rates reported in April 2025 were 3.5% for iron ore, 3% for bauxite, 5% for gold, 3% to 5% for diamonds and just 0.5% for rutile. The 2022 mining law also gives the state a 10% free carried interest in large mines and the option to buy up to 35% more (Mining Focus Africa). Whether those terms are fair can’t be judged without seeing the contracts, which is exactly the problem.

The country also knows what instability costs. Disputes over the Marampa and Tonkolili iron ore mines shut exports for long stretches (Mining.com; Miningmx). Investors don’t like that, and neither do the people whose wages depend on the mines.

Each mineral needs its own handling. For iron ore, a US$230 million processing plant at Tonkolili broke ground in April 2025, so the job is to audit shipments against payments and agree stable, fair terms. Rutile’s 0.5% royalty looks low beside 3.5% for iron ore and 5% for gold, so compare it with regional norms and explain the rate. Bauxite pays 3%; run a transfer pricing audit and plan for local suppliers. For gold, Baomahun’s developer closed US$330 million of senior debt in December 2025 and expects about 150,000 ounces a year over 12.5 years, though no first production date is set (FG Gold). Publish the agreement and a clean up bond before production starts. Diamonds suffer from smuggling and weak traceability, so track stones digitally from mine to export. Offshore gas and other prospects are early stage. Don’t spend against resources nobody has proved.

What would help most is a short list. Open licensing: publish a licence map, award new areas competitively, and make cancellations follow due process under the Mines and Minerals Development Act 2022. Open contracts and payments, with company and government figures reconciled every year. Joint revenue checks by the minerals agency and the revenue authority, with independent assayers and port scanning. Real local benefit, through supplier lists, community agreements with actual budgets and rehabilitation bonds held by the state. And support for small scale miners to form cooperatives and sell to licensed buyers.

Then there’s the question of what to do with the money. I’d pay mining income above a baseline into a Resource Revenue Window: a smaller reserve for price crashes and a larger share for health, education and maintenance, under parliamentary watch. Saving abroad is the textbook advice, but with high debt and urgent needs I’d rather earmark spending with strict project rules, and I accept that reasonable people will disagree.

Running cost is tiny, under 0.05% of GDP, and the revenue effect is about 0.5% by 2033. By 2030 all licences and contracts should be public. Investors worry about rule changes, so keep the rules predictable. If iron ore prices fall, the reserve and a wider economy are the defence.

Environment and climate

Freetown’s hillsides, coast, watersheds and fisheries are under heavy pressure, and the 2017 Regent mudslide showed what building on unstable slopes costs. A World Bank assessment cited in the IMF’s June 2026 country report puts the climate financing need at about US$2.97 billion through 2035 (Awoko).

The steps are practical. Enforce environmental impact assessments, with real consultation, before any major mine, dam, plantation or road. Protect the forests and watersheds that feed hydropower and Freetown’s water. Build erosion defences in the worst hit towns, and link early warning to the weather and disaster agencies. Enforce drainage and building rules, especially against building in waterways and on steep slopes. Fix waste collection and monitor fisheries. The bill is about 0.4% of GDP, with grants and climate funds covering much of it. By 2030 early warning should reach every flood prone district, and by 2033 no new housing should be approved in mapped hazard zones. Where people must move, offer fair compensation and secure alternative land.

Power, roads, water and housing

Electricity

Picture a hospital that loses power in the dry season. That’s the problem in its plainest form. Sierra Leone leans on the Bumbuna hydropower plant, and hydropower gives only about 50 MW in the rainy season, with output swinging with the weather (Connecting Africa). Access runs from 21% of households on the grid or a mini grid (ESMAP) to 35.5% on the World Bank’s wider measure. The utility, EDSA, loses much of what it generates and struggles to collect. A World Bank backed 40 MW solar and 35 MWh storage project, which cost US$65 million and was switched on in July 2026, is expected to lift access from about 23% to 36% (Connecting Africa). Earlier promises, including universal access by 2025, weren’t kept, so I’d treat any new target with some scepticism, including mine.

No single technology wins, and what follows is my own assessment, which needs least cost planning studies behind it. Grid expansion makes sense only after the utility’s finances improve; it’s high cost and takes three to seven years. Regional power imports through the CLSG line can cover dry season shortfalls, under capped contracts. Solar mini grids are the main route for district towns, clinics and markets the grid won’t reach for ten years or more. Rooftop solar with batteries is the first choice for hospitals, schools and water pumps, using certified products only. Hydropower, including a Bumbuna upgrade, gives cheap baseload, but only after independent checks on dry season output, resettlement and cost. Diesel should be backup only, and the 2026 fuel shock showed why.

What would I do first? Fix the utility from 2027, with a performance based management contract that sets published targets for collections, losses and outages, prepaid meters for businesses first, and open tariff setting by the regulator. Power the places that matter. Within three years, put solar and batteries into every district hospital, then as many health centres, schools and water stations as the budget stretches to. Set tariffs with care: cost reflecting prices for large users, a lifeline tariff for low users, and any subsidy paid openly from the budget. Don’t raise household prices before supply improves. And plan maintenance from day one, with an upkeep account for every project, local technician training and supplier guarantees.

The public cost is about 0.8% of GDP a year, with more from private and concessional money. Household access should reach 45% by 2030, 55% by 2033 and 65% by 2036, reported in hours of supply. If the grid disappoints, lean on mini grids and stand alone solar, which put less on the government’s books.

Roads, ports and transport

In the rains, many rural roads and bridges become impassable and farmers can’t reach markets. Port delays and checkpoint payments add cost, and projects sometimes start without money to finish or maintain them. Only about 40% of primary roads are paved, most secondary and feeder roads aren’t, and about a third of people have an all season road (World Bank). A current national road condition survey is a 2027 task.

Choose projects on merit. Score each one on economic return, access to markets, climate resilience, upkeep cost and social benefit, publish the scores and review big ones independently. Nothing should start without money to finish it and a maintenance line. Maintain first, by strengthening the Road Maintenance Fund so fuel levies pay for routine upkeep through performance based contracts. Then build feeder roads and bridges in the highest producing districts, using local labour where it suits, and fix drainage and climate proofing, especially in Freetown and coastal towns.

Ports need targets for how long containers wait, a customs single window and fewer informal charges at checkpoints. On safety and public transport, enforce helmet and speed rules, inspect vehicles, regulate motorcycle taxis and minibuses, and pilot a managed Freetown bus service with open fares. And stop the usual waste: publish contract values and unit costs, require parliamentary approval for loans, and compare prices with similar projects. Big prestige crossings or airports need independent appraisal. Don’t assume them.

The cost is about 1.2% of GDP, roughly half of it maintenance and rural roads. By 2033 maintenance should be fully funded from the fund, and by 2036 road deaths should be down a third from a measured baseline. If capacity is thin, use smaller packages and performance bonds.

Water, sanitation and housing

Freetown suffers water shortages and flooding. Waste is badly collected, sewage treatment is minimal, informal settlements are built in hazard zones, and land disputes put builders off. A 2022 survey found basic water access at 62.6%, basic sanitation at 13.6% and basic hygiene at 11.6% of households (UNICEF), so the 2027 survey should update them.

Protect water sources, cut leakage at the Guma Valley Water Company, extend SALWACO to district towns and put solar pumps on rural water points with community run upkeep. Use community led programmes to end open defecation, attended toilets in markets and sludge treatment. Councils should run waste collection on performance contracts. Digitise land records so tenure can be proved and used as security. For housing, offer serviced plots and small step by step building loans, not large expensive estates. Give Bo, Kenema, Makeni, Koidu and other towns their own water, drainage and waste plans.

The bill is about 0.6% of GDP. By 2030, water points in monitored districts should work at least 80% of the time. By 2033, basic water access should be up 15 points and sanitation up 12 from the 2027 baseline. Evictions are the main risk, so design with communities and resettle only with fair compensation.

Schools, clinics and the people who fall through

Education

Free school education, launched in 2018, got more children into classrooms, and a year of pre primary became compulsory in 2021 (Brookings). Getting children in was the easier half. Learning is weak. The World Bank reports that only 4.2% of ten year olds can read a simple text, while the education minister says most can read words but not understand them (Premier Media). Both versions describe a serious problem. Payroll problems are old too: about 1,000 ghost teachers were removed from the payroll in 2010 (News24), so the audit should be repeated.

I’d begin with one national reading and maths test in 2027, repeated every two years, so every target is measured the same way. Then teach Grades 1 to 3 well, with lesson guides, readable books and regular coaching, building on the 8,000 Primary 1 teachers already trained. Fix teacher pay and attendance by verifying teachers biometrically, linking payroll to the school census, paying on time by mobile money, and letting parents and school committees report attendance. Certify untrained teachers, offer hardship allowances in remote schools and publish the deployment formula.

Buildings matter, but only as part of a plan. One textbook per child, repairs, water and girls’ toilets, and no classroom block without a maintenance plan. Keep the policy that lets pregnant girls and young mothers stay enrolled, add scholarships at senior secondary level and screen for disability. Tie technical training to jobs through skills councils with employers on them. Fund courses only where employers confirm demand and offer placements, and publish what happens to graduates. Independent inspection, with published school reports, completes the picture.

The cost is about 0.7% of GDP a year, mostly recurrent, so it needs a permanent budget line. Treat donor money as temporary. Reading with understanding should double by 2030, reach 2.5 times by 2033 and triple by 2036. If payroll politics or rural shortages bite, pay community teaching assistants openly and protect classroom basics before new buildings.

Health

About 105 of every 1,000 children die before turning five. The health minister reports maternal deaths falling from 443 to 354 per 100,000 between 2020 and 2023, which is real progress if it holds, though the level is still among the world’s highest. Households pay over half of health costs themselves, according to UNICEF’s 2026 health budget brief. A SEND Sierra Leone assessment of 44 rural facilities found that over 80% of staff reported severe shortages of equipment and medicines, and only 28% had reliable piped water (Awoko).

The system also leans hard on donors. The Centre for Global Development estimated that US global health aid in 2022 and 2023 equalled 52% of the government’s own health spending, and noted that the 2026 budget gave health less than the 2025 budget. UNICEF puts health at 4.6% of the 2026 budget, down from about 5.0%, with a 7% fall in the allocation (CGD). The ministry told Human Rights Watch that US cuts meant a US$45 million reduction in maternal, child and adolescent health projects (Foreign Policy).

So the order of work matters. Plan the shift from donor to domestic money in 2027, and publish health budget releases every quarter so people can see whether approved money arrives. Pay community health workers on time, and build an emergency referral system with ambulances, phone links and fuel funds for obstetric emergencies. Put a skilled attendant at births in every chiefdom, emergency obstetric care in district hospitals, protect vaccination from funding gaps and screen for malnutrition.

Medicine supply needs electronic stock tracking, quarterly deliveries and published stock levels of 20 tracer medicines by facility. Staffing needs a costed workforce plan with rural postings and hardship pay, and no hiring beyond what the payroll can sustain. Stop informal charges by publishing what’s free, letting patients report illegal fees anonymously and reimbursing facilities promptly. And keep the lights on: fund maintenance teams, power and water in facilities.

On free care for everyone, I’ll be blunt. This plan doesn’t promise it. The maternal, newborn and child package should be guaranteed and fully funded. Wider benefits can follow as money and supplies come through, because a promise without drugs and staff damages trust more than an honest, smaller offer.

The cost is about 0.8% of GDP a year. Child deaths should fall to 85 per 1,000 by 2030, 72 by 2033 and 62 by 2036. Maternal deaths should reach 280, 230 and 180 per 100,000. If donors cut more, ring fence maternal and child care first.

Women, children and other vulnerable people

Poverty and hunger hit hardest among rural households, women who farm without secure land rights, children who miss school or work, and people with disabilities.

Start with one beneficiary registry and regular cash to the poorest, food insecure households, with a lean season top up, public targeting rules, mobile payment and a way to complain. Scale payments up fast when prices spike. Secure women’s customary land rights through registration, widen guaranteed credit, and apply the Gender Equality and Women’s Empowerment Act 2022, which sets 30% quotas for parliament, councils, the civil service and larger employers (IPS). Enforce child labour laws and the Prohibition of Child Marriage Act 2024, which makes marriage under 18 a crime, put child protection officers in every district and tie school feeding to attendance. Make schools and clinics accessible and add a disability grant. The review of the Persons with Disability Act 2011, which began with a consultation in 2024 (UNDP), should be finished.

Who pays and who gains? The costs fall mainly on mining companies, large importers and higher earners, and on donors. The gains go to the poorest households, women and children. Basic foods and household fuel stay off the list of new taxes. The programme costs about 0.5% of GDP a year. By 2030 every extremely poor household in the registry should get regular support, and by 2033 an independent audit should confirm it. Diversion is the risk, so use biometric checks and community validation.

Jobs, business and exports

Young people and work

About 37% of the population is under 15 (Population Pyramid). From that I’d estimate around 200,000 young people reach working age each year. That’s my own rough arithmetic, so Statistics Sierra Leone should confirm it. The official youth unemployment rate of 3.7% hides the real story: most people work in low paid farming and informal trade, and the last labour force survey was in 2014. So the first job measure is a new survey in 2027.

Several sectors could employ many people at decent pay, though each needs employer surveys behind it. Construction and maintenance, including solar installation, come first: pay contractors on time and use local labour on road upkeep. Food processing (rice, cassava, palm oil, cocoa, fruit) puts mills and drying hubs near farms, with many jobs for women. Fisheries need licence enforcement, ice and storage at landing sites. Tourism and hospitality need training tied to hotels and tour operators. Digital and business services should be piloted with employer commitments first, since bandwidth and power are still weak. Light manufacturing, such as building materials, packaging and furniture, comes after a competitiveness study.

What would I do about it? Teach skills employers will pay for, through apprenticeships with placement agreements signed by employers, recognition of skills informal workers already have, and published job outcomes. Create paid public work, such as road and drainage upkeep and lean season cash for work, paid on time by mobile money. Back firms that already have orders with guarantees and working capital. I don’t believe loan schemes can replace enough real jobs, and loans without a market only create debt.

Drug use and irregular migration are health and opportunity problems before they’re policing problems. The synthetic drug kush is a serious public health concern. Combine prevention, treatment at district level, and enforcement aimed at traffickers more than users, plus honest information on legal migration and action against recruitment agencies that break worker rights.

Together, jobs, business, industry and tourism cost about 0.5% of GDP a year. By 2030, 100,000 apprenticeship or training places with employer agreements is a reasonable aim. If employers don’t show up, scale only programmes with good placement rates.

Doing business

Registering a firm, getting a licence, paying tax and clearing goods are slow and often involve informal payments. Enforcing a contract or using property as collateral is hard. The fixes are known: one online registration and licensing window with fixed deadlines and published fees; simpler small business tax and customs; a commercial court division with time limits and compulsory mediation; a registry for loans secured on movable assets and a credit reference system (the Bank of Sierra Leone began looking at a private credit bureau with IFC advice in 2012, so the job is to finish it); insolvency rules that let viable firms restructure; and land registration so property can back a loan. Keep labour, safety and environmental protections, but write them clearly and enforce them consistently so they don’t become a source of bribes.

Investors respond to stable rules, reliable power, secure land and enforceable contracts far more than to tax holidays. By 2030, registering a business should take a few days online.

Industry and exports

Exports are mostly raw minerals. I’m wary of prestige factories, so I’d use a five part test. Can the sector compete without a permanent subsidy? How many jobs does it create per unit of investment? Can its energy and transport needs be met? Are the skills there? Is there a market? On present evidence, food processing, fish processing, building materials, some light manufacturing and regional logistics look most promising, and each needs a competitiveness study. Support should be time limited, tied to performance and published. Special economic zones make sense only where power, roads and water are guaranteed. This all depends on earlier progress in power, roads and skills, which is why I’ve put it in 2032 to 2033.

Tourism and culture

The country has long beaches near Freetown, the Banana Islands, Tiwai Island, forest parks, music and food, and heritage linked to the Atlantic slave trade and to Freetown’s founding by freed Africans. UN Tourism reported notable growth in international arrivals in 2025, though no verified count is published (Tourism Board), so I’d treat tourism as a modest business that grows in a few places, not the answer to the whole economy.

Commission a market study in 2027. Concentrate on a few destinations, such as the Freetown Peninsula, Tiwai and selected heritage sites, with safe access, clean sanitation and trained staff. Back community owned lodges and make sure conservation fees reach local people. Tell the history of slavery and the civil war with dignity and accuracy, through museums and memorials built with survivors and affected communities. The cost is about 0.05% of GDP. Judge growth against the study’s forecast, not a hopeful number, and go slowly on fragile coasts.

Who delivers it, and in what order

A delivery unit with a legal basis

In Sierra Leone the gap between an announcement and a result stays wide. A plan needs someone whose whole job is to chase it.

I’d create a Delivery and Results Unit by Act of Parliament, not by decree. It would track a short list of national priorities, unblock delays and report results. It could coordinate and recommend, but it couldn’t override a minister’s legal duties, the Auditor General or any independent body. It would hire analysts, engineers and finance staff competitively on fixed terms, have a head appointed openly and removable only for stated cause, and work with an advisory council that includes civil society, business, academics and opposition nominees. It would report quarterly to Parliament and the public. A review after six years would decide whether to keep, merge or end it.

Every reform gets a lead institution, a budget line, milestones and a named Senior Responsible Officer, such as a permanent secretary or council chief administrator, whose performance contract states the milestones.

Continuity is the part most plans skip. A general election is due in 2028 and another in 2033, and a plan that dies with a government isn’t a ten year plan. I’d pass a National Development Planning Act requiring a successor plan beyond 2030, independent yearly evaluation and a handover of the dashboard to any incoming government. I’d also seek a cross party National Recovery Compact on the priorities that shouldn’t be partisan: stable macro policy, procurement integrity, maintenance, child survival and learning. That doesn’t take away voters’ right to change the government. It protects what no party should want to reverse. Cost is about 0.1% of GDP.

Who does what

The ten year roadmap

The order matters. Stability, revenue, procurement and data come first because they pay for everything else and prove whether it worked. Basic services come next. Industry and exports depend on power, roads and skills, so they follow. The last phase is about protecting the gains.

2027 to 2028: stabilise and repair.

• Stay on the IMF programme, audit arrears, review exemptions.
• Pass the watchdog independence law, launch the contracts portal, audit payroll.
• Run the national test, survey baselines and a labour force survey.
• Fund road maintenance, put solar in hospitals, sign the utility contract.
• Start the cash transfer registry and the health financing plan.

2029 to 2031: expand services and productive investment.

• Revenue 13.5% of GDP; electronic invoices; electronic bidding everywhere.
• Feeder roads in the top farming districts; water control, stores, mini grids.
• Coaching at scale in schools; health workforce plan; regular cash transfers.
• Apprenticeships, public works and guarantees for small firms.

2032 to 2033: industry and exports.

• Revenue 15.5%; debt risk rated moderate.
• Processing hubs for export crops; local suppliers at 25% of mining purchasing.
• Power and water in all health facilities; skills councils placing graduates.

2034 to 2036: consolidate.

• Revenue 17%; maintenance fully funded; household electricity 65%.
• Child deaths 62 per 1,000, maternal deaths 180 per 100,000, reading three times the 2027 baseline.
• A successor plan agreed across parties, and the dashboard handed over to whoever governs.

Three things have to start in 2027, however much else slips: surveys and baselines; the arrears and contract inventory, with procurement and audit follow up; and maintenance of what already exists, along with targeted cash for the poorest. Other things depend on earlier progress. Big power and road projects need project scoring first. Processing hubs need power and roads. Health insurance needs a working supply chain. And if the 2027 surveys show a different bottleneck from the one I’ve assumed, the Delivery Unit should propose a new sequence to Parliament, in public.

Partners

The IMF anchors the macro programme and advises on revenue and debt, and its documents should be public. The World Bank and the African Development Bank bring concessional money for power, schools, health and roads, on budget where possible. UN agencies cover food aid, vaccines and surveys. The World Food Programme’s draft country plan for 2026 to 2030 has a stated cost of about US$148.7 million, with a six month shortfall of US$38.5 million (WFP). ECOWAS matters for trade, regional power, corridors and health cooperation.

Bilateral donors give grants and technical help. Treat aid cuts as a planning fact and agree multi year transition plans. Investors bring capital and skills for mining, energy, food processing and tourism, through open tendering and stable rules. Universities and think tanks can do research and independent evaluation. The diaspora brings skills, networks and co investment, through audited vehicles only.

Publish every loan and grant agreement, tell Parliament the debt terms, and fold donor reporting into one national dashboard so officials spend their time delivering, not writing parallel reports.

What it costs, how to track it, what could go wrong

Costs and funding

The plan is affordable only if revenue rises, spending is used better and borrowing stays cheap. In my central case it needs about 7% of GDP a year of extra spending once fully running, around 2031 to 2033, in a range of 4.6% to 10.2%. These are planning ranges, not feasibility study results, and they leave out temporary price shock transfers.

Annual cost at full roll out, share of GDP (low, central, high):

• Governance, audit, justice, digital, councils: 0.3, 0.5, 0.7. Needs a capacity review first.
• Revenue administration and financial inclusion: 0.1, 0.2, 0.3. Needs tax expenditure and customs studies.
• Social protection: 0.3, 0.5, 0.7. Needs registry design.
• Farming and food: 0.5, 0.7, 1.0. Needs irrigation and storage studies.
• Electricity (public share): 0.5, 0.8, 1.1. Needs a least cost power plan.
• Roads, ports, transport: 0.8, 1.2, 1.8. Needs a road condition survey.
• Water, sanitation, housing: 0.4, 0.6, 0.9. Needs an access baseline.
• Education: 0.5, 0.7, 1.0. Needs payroll and school census checks.
• Health: 0.5, 0.8, 1.1. Needs a financing strategy.
• Jobs, business, industry, tourism: 0.3, 0.5, 0.8. Needs a labour survey.
• Environment and climate: 0.3, 0.4, 0.6. Needs hazard maps.
• Delivery, statistics, evaluation: 0.1, 0.1, 0.2.
• Total: 4.6, 7.0, 10.2.

About 1% of 2025 GDP is NLe1.7 billion. Education and health are about 80% and 75% recurrent, which is the real trap, because donor money fades and salaries don’t.

Where the money could come from (low, central, high):

• New domestic revenue given to the plan: 2.2, 3.0, 3.6. Possible. About half of the six point revenue climb is needed to replace falling grants and cover interest.
• Reprioritised spending: 0.5, 0.8, 0.9. Possible. No procurement savings counted until pilots measure them.
• Grants: 0.5, 0.8, 1.0. Partly confirmed. Aid cuts make this uncertain.
• Concessional loans: 0.9, 1.2, 2.2. Partly confirmed. The IMF facility and World Bank and AfDB projects are under way.
• Private investment and partnerships, mostly off budget: 0.5, 1.2, 2.5. Possible. Needs bankable projects.
• Total: 4.6, 7.0, 10.2.

What’s actually confirmed? The IMF arrangement, the 2026 budget with NLe4.1 billion of borrowing, the World Bank backed solar project (now running) and the WFP plan with its funding gap. Feed Salone’s fundraising isn’t confirmed. Nearly everything else is possible but unconfirmed, including most of the US$2.6 billion the Planning Minister said is still needed.

A warning on debt. This is back of the envelope arithmetic, not a debt analysis. The central case means new concessional borrowing of about 7% of GDP over ten years. The IMF sees debt falling from about 49% of GDP towards 37.5% without that borrowing, so the plan could leave it in the mid 40s. That’s below today’s level but too close to the high risk zone to relax. A formal debt analysis should test the borrowing every year. Dollar priced power contracts, state enterprise arrears and guarantees can add to debt suddenly, so they need to be in the debt register before anyone signs.

If money falls short, protect things in this order: maternal and child health, early reading, medicines, power for clinics and schools, rural road upkeep, and cash for the poorest. Delay new buildings, big grids and showpiece projects first. In the low case, the plan stretches timelines by two to three years.

Tracking progress

The Delivery Unit should publish a quarterly dashboard, and an independent evaluation should be commissioned every year by the Auditor General, Parliament or a university group.

• Every month or quarter: inflation and food prices; tax and customs collection against target; share of contracts tendered competitively; teacher attendance; availability of 20 tracer medicines; hours of electricity supplied; complaints answered within 30 days.
• Every year: road condition and rural access; working water points; audit recommendations acted on; jobs and earnings from 2028; real household income; service satisfaction from an independent survey; acute hunger twice a year.
• Every two or three years: reading and maths tests; poverty survey; a demographic and health survey about every five years.

Any baseline marked “2027” in this plan has to be measured in that first year.

What could go wrong

Plenty, and pretending otherwise would be the first mistake. The biggest risk is resistance to reforms that cut discretion, which is highly likely and would hurt a lot. The answer is a cross party Compact, early wins and published results. If it fails, narrow the plan to five core reforms: revenue, procurement, payroll, maintenance and child survival. Insiders capturing new programmes is the next worry, and open contracting plus independent audit are the defence. If a line gets captured, pause it and re tender.

Election year spending in 2028 and 2033 is almost certain to test the plan. Limits on supplementary budgets and an arrears ceiling help; if they fail, Parliament can withhold release and the IMF safeguards kick in. A commodity price shock is less likely but severe, so build a stabilisation reserve, assume cautious prices and cut capital projects before services. Floods and mudslides will come, so invest in early warning and resilient design, and if they hit, redirect resilience funds and appeal for emergency help. Disease outbreaks need surveillance, stockpiles, community health workers and ring fenced emergency money.

Debt distress is a medium risk with a very high price. Borrow concessionally, run a yearly debt analysis and, if it comes to it, reprofile with the IMF while protecting core services. Weak delivery capacity is likely, so phase the rollout, get technical help and simplify designs. Regional insecurity is a lower risk, covered by ECOWAS cooperation and a contingency fund. A donor shortfall is likely, so put domestic revenue first, agree transition plans and use the low case with the protection order above.

The last risk is the one people forget: misuse of emergency powers. Emergency measures must be time limited, reported to Parliament and open to court challenge, and never used against opponents, media or civil society.

Starting well, and an honest verdict

The first 100 days

No government can finish structural reform in 100 days, and none should pretend otherwise. What it can do is establish the facts, stop leaks, protect the vulnerable and show the programme is real. Day 1 is the start of the programme, whichever government or ministry leads it.

1. Day 30: Cabinet directs that non concessional borrowing and guarantees need parliamentary approval. Lead: Ministry of Finance.
2. By day 45: a costed list of priority maintenance for roads, clinics, schools, water points and clinic power, paid from existing budgets, with first works by day 100.
3. By day 45: start the service baseline, counting clinics, schools, water points and rural roads, and verifying teacher and health worker payroll. First results by day 100. Lead: Statistics Sierra Leone with line ministries.
4. By day 60: a public expenditure review comparing what was approved, released and spent by ministry. Lead: Finance with the Audit Service.
5. By day 60: targeted lean season and price shock cash through the existing registry, with a complaints line. Lead: National Commission for Social Action.
6. By day 75: risk review of the 50 largest open contracts, with no award proceeding without a published reason. Lead: procurement authority and Audit Service.
7. By day 90: an audit of arrears and contingent liabilities, including state enterprises and power contracts, published as a public register. Lead: Accountant General and Auditor General.
8. By day 90: an inventory of all loans, guarantees, concession and mining contracts, public unless a legal test requires secrecy.
9. By day 100: conflict of interest guidance, an asset declaration compliance report, a whistleblower channel, and the watchdog independence bill sent to Parliament.
10. By day 100: district hearings in every district and a cross party meeting on the Recovery Compact.
11. By day 100: contracts signed for the labour force and poverty surveys, and one national learning test settled.
12. By day 100: the first public dashboard with about 10 indicators.

The first two years

Only a few reforms have to be funded and delivered first, because each one pays for the others or makes them possible.

1. Revenue and customs (2027). Exemptions review, mining royalty audits, risk based customs checks and digital payments, through the 2027 and 2028 Finance Acts.
2. Procurement and audit integrity (2027 to 2028). The watchdog independence law, an open contracts portal, electronic bidding for the 20 largest buyers and a public audit follow up register. Draft by mid 2027 and enact before the 2028 campaign hardens positions.
3. Payroll checks and service basics (2027 to 2028). Verify teachers and health workers, pay them reliably, and deliver the first year of coaching, books and medicines. No dismissals without a documented case and an appeal.
4. Maintenance first (2027 to 2028). A funded Road Maintenance Fund, a project scoring rule, power for clinics and schools, and the utility contract.
5. Cash for the poorest (2027 to 2028). Audit the registry, contract mobile money payers and set up complaints.
6. Data and delivery (2027 to 2028). Pass the Delivery Unit Act within the year and hire the team competitively.

All six need the same political groundwork. Seek a short cross party statement on them in the first half of 2027, make the legal changes before the 2028 election, name an accountable officer for each, and line up the 2028 budget to fund year one.

The verdict

What can change in ten years. With discipline, Sierra Leone could have lower and more predictable inflation, revenue near 17% of GDP instead of about 11%, procurement and audit systems that visibly work, and a debt risk rating moved down from high. It could have power for most public facilities and about two homes in three, clinics with medicines and trained midwives, children who read with understanding, fewer rice imports, and poverty about 12 points lower. Child deaths could fall from about 1 in 10 under fives to about 1 in 16. That would be a huge human gain, even though it’s still far from rich country levels.

What can’t be fixed quickly. Poverty won’t disappear. The World Bank still classes Sierra Leone as low income for 2026 to 2027, meaning gross national income per person of US$1,175 or less, and the lower middle income band starts at US$1,176 (AfriDigest). One review put the figure at about US$500 in 2021, before rebasing (Springer). Crossing the line would likely mean per person incomes more than doubling. With population growth near 2% and economic growth of 4% to 5%, I don’t think that’s credible by 2036, and the government’s 2039 aspiration would need faster, broader growth than anything in today’s numbers. Trust in institutions, teacher quality and the habit of maintenance take a generation.

Trade offs people should understand. Some exemptions will end, which will annoy powerful people. Large firms, mines, tobacco and alcohol may pay more tax while basic food is protected. Prices may stay high for a while, and the state will help the poorest with targeted cash, not across the board price controls. Transparency means bad news, including failed audits, gets published. Showpiece buildings will wait while maintenance comes first. Donors may give less, and the country will pay for more of its own services.

What happens if nobody acts. About 200,000 young people reach working age every year, by my estimate. Too few good jobs raises the risk of exclusion, drug use, irregular migration and unrest. Interest payments and arrears will crowd out services, and child and maternal deaths will stay at levels most countries would find unacceptable. Delay isn’t neutral, because every year without maintenance makes roads, clinics and power systems dearer to repair.

A last word. I’d ask every citizen, official and party to treat this as a ten year agreement that outlasts any election. Governments need to tell the truth about money and results. Opposition parties should back reforms that work even when others take the credit, and challenge the ones that don’t. Parliament, the courts, the Auditor General and the press need to be free and funded. If those commitments hold across governments, a child born in 2027 could start secondary school in a country where the school has books, the clinic has medicine, the road stays open all year and the budget is something people can read and trust.

Sources

Economy, budget and debt

• IMF, Country Report 26/168, June 2026
• IMF, Country Report 26/40, February 2026
• IMF, Press release 26/213 on the resilience facility and third review
• IMF, Sierra Leone country page
• World Bank, Sierra Leone overview
• World Bank, Macro Poverty Outlook, April 2024 and a later outlook
• AfDB, Economic Outlook and Country Focus Report 2025
• BudgIT Sierra Leone, 2026 budget analysis and citizens budget
• FRED, general government revenue
• Awoko, fiscal gap report and NRA revenue strategy

Plans, poverty, food and jobs

• Government of Sierra Leone, Medium Term National Development Plan 2024 to 2030
• Premier Media, financing for development conference, December 2025
• Child Health Task Force, Child Survival Action Plan 2023 to 2025
• Statistics Sierra Leone and World Bank, Poverty in Sierra Leone, 2018
• FAO, country brief
• Premier Media, WFP warning, May 2026; WFP Executive Board, draft country plan
• Awoko, rice output claims; Table Debates, rice and imports
• World Bank, labour force survey 2014; youth unemployment series
• Population estimates: Worldometer, Population Pyramid

Governance and mining

• Transparency International score via Trading Economics
• U4, Corruption and anti corruption in Sierra Leone
• Review of procurement compliance; Awoko on the NPPA report
• Rio Times on critical minerals and the mining economy
• Mining.com and Miningmx on the Marampa dispute

Energy, schools and health

• ESMAP, Beyond Connections, 2024; World Bank data via Statistics of the World
• Connecting Africa
• Brookings, early grade learning assessments
• Education Above All, literacy programme; Premier Media, education minister; World Bank, early grade literacy
• UNICEF, SDG country data
• Centre for Global Development, health aid; Foreign Policy, Can Sierra Leone keep its mothers alive?; Awoko, maternal health at the UN; ASLM, community monitoring report
• Springer, under five mortality study

Laws, elections, services and other data

• ISS Africa, election reform risks
• Ministry of Information and Civic Education, data protection policy and bill
• Mining Focus Africa, overview of mining legislation; IEA, Mines and Minerals Development Act 2022; FG Gold, Baomahun financing
• Customary Land Rights Act 2022, National Land Commission Act 2022, Gender Equality and Women’s Empowerment Act, Prohibition of Child Marriage Act 2024, Persons with Disability Act review
• UNICEF, WASH survey 2022 and 2026 health budget brief
• World Bank, roads and connectivity; AfriDigest, income classifications
• Awoko, prison numbers, 2025 criminal session, rural health facilities, climate financing
• Tourism Board, tourism growth