The Worker Our Systems Cannot See
It is early evening in Adjamé market in Abidjan. A trader closes her stall, checks the payments that came in on her phone, pays a supplier and sends money to her mother in Bouaké. On a good day there is something left over. On a bad day the business barely covers the household.
She has income, customers, suppliers and dependants. She moves money every day, digitally, with a speed and frequency that would impress any treasury department. What she does not have is a payslip.
Now ask her to join a retirement arrangement whose forms begin with “employer”, whose contributions assume a fixed monthly salary and whose service centre opens during her busiest trading hours.
We call her informal. Her mobile money provider calls her a customer, and knows her better than any pension institution does.
The problem is not that she is hard to reach. It is that our retirement systems were never designed to look for her.
That is the challenge, and the opportunity, at the heart of PensionTech in Africa. Can we build retirement systems around people whose working lives do not fit the assumptions on which those systems were designed? And can we do it in a way that leaves the developed world studying us, rather than the other way round?
From the Boardroom to the Member’s Phone
In Conversation 6, we asked whether trustees could see the member experiences hidden behind reassuring board reports, and argued that governance should not wait for the next quarterly board pack.
That argument has a practical consequence. A board that wants earlier detection of errors, wider coverage, honest sight of funding adequacy and meaningful member engagement needs infrastructure capable of delivering it. Good intentions recorded in the minutes do not reconcile a single contribution.
PensionTech is that infrastructure: the technology used to enrol members, collect and account for contributions, enforce compliance, value assets and liabilities, administer benefits, support decisions and maintain trust across a retirement journey that may last sixty years.
Its real promise is not a faster version of existing administration. It is the chance to question the assumptions built into that administration, about who counts as a worker, where money should be collected, how often a fund’s health should be checked and what engagement actually means.
Six Conventions Africa Should Stop Borrowing
Much of Africa’s pension architecture was imported from economies where most people held salaried jobs. The models were sensible for their time and place. Applied unchanged to economies where most workers are self-employed, they have produced the coverage gap we now treat as inevitable.
Leapfrogging begins by questioning six of those inherited conventions.
“Coverage follows formal employment.” In much of Africa, waiting for formal jobs before extending pensions means waiting indefinitely. The ILO’s World Social Protection Report 2024–26 found that only 19.1% of Africa’s population was covered by at least one social protection benefit in 2023, against a global average of 52.4%. [1] The informal economy is not the edge of the market. It is the market.
“The payslip is the collection point.” For millions of Africans, the wallet already is. The GSMA reports that most of the world’s new registered and monthly active mobile money accounts in 2025 came from Sub-Saharan Africa. [2] A retirement system that can only collect through payroll is ignoring the continent’s most widely used financial infrastructure.
“Compliance is an enforcement problem.” It is first a data problem. Most defaults are not hidden; they are simply not being looked for. When the evidence of non-compliance already sits in tax, payroll and payment systems, inspection becomes the last resort rather than the first.
“Funding adequacy is a triennial event.” Many schemes formally learn their funding position once every few years, when the actuarial valuation arrives. Markets, inflation, membership and liabilities do not wait that long. There is no technical reason trustees cannot see an indicative funding position every month.
“Engagement means an annual statement.” A paper statement describing last year’s balance arrives long after the decisions that shaped it. Real engagement happens at the moment money moves, when a member is deciding what to do with today’s income.
“Leapfrogging means catching up.” It does not. The UK’s widely admired automatic enrolment reform was built around employers and payroll, and leaves the self-employed outside its default. Developed systems are now trying to retrofit gig and self-employed workers into structures designed without them. Africa has the chance to design for those workers from the start.
Africa’s advantage is not that it has better pension systems. It is that it has less legacy to dismantle.
Africa is not one pension market. Countries differ in infrastructure, institutional capacity, regulation and household resources, and mature systems still have much to teach us about governance, investment discipline and benefit delivery. The opportunity is to lead where our realities force innovation: irregular earnings, digital payment rails, low-cost small contributions and the data that connects them.
The Missing Millions Are the Growth Story
The industry usually discusses informal-sector coverage as a social obligation. It is that. It is also the single largest growth opportunity in African retirement savings.
Consider the scale. In Ghana, SSNIT reported in 2023 that only 0.85% of the country’s 6.7 million informal-sector workers were covered by its scheme. [3] In Nigeria, PenCom data for the third quarter of 2025 showed 215,914 informal-sector workers enrolled in the Personal Pension Plan, out of an informal workforce estimated at around 92 million. [4] Côte d’Ivoire counts roughly seven million independent workers. [5]
These are not marginal gaps to be closed by a marketing campaign. They represent tens of millions of potential members, and long-term domestic savings that could help finance the economies those members live in, a theme we will take up in Conversation 8.
But the opportunity will not be realised by offering informal workers a smaller, cheaper version of a salaried pension. It requires a different design: contributions that follow income rather than the calendar, collection through channels people already trust, costs that do not devour small balances, and a reason to keep contributing after the first enthusiasm fades.
Enrolment is a headline. Persistence is the business model.
What Côte d’Ivoire Is Teaching the Continent
Côte d’Ivoire offers one of the most instructive experiments on the continent, because it chose to extend mandatory social security to the self-employed rather than leave them to voluntary products.
Ordinance No. 2019-636 of 17 July 2019 created social protection regimes for independent workers, and the Caisse Nationale de Prévoyance Sociale (CNPS) launched the Régime Social des Travailleurs Indépendants (RSTI) under implementing decrees issued in 2020. [6] The RSTI covers retirement, with a lifetime pension from age 60 and survivor benefits, alongside sickness, work-accident and maternity protection.
Several design choices deserve attention across Africa.
Contributions built around real occupations. The RSTI charges 12% of declared income between 30,000 and 180,000 FCFA a month, with minimum contribution levels set by occupation, starting at 3,600 FCFA a month for traders. [6] Rather than asking a market trader to behave like a civil servant, the scheme begins from what different kinds of work actually earn.
Payment through everyday channels. Members can contribute through Wave, MTN and Orange Money, as well as through partner microfinance networks and CNPS offices. [6] In March 2023, the CNPS signed an agreement with the COFINA microfinance group to use its physical and digital network to enrol independent workers and collect their contributions. COFINA already served 37,000 independent workers. [5] That is distribution through institutions informal workers already trust.
Portability beyond the border. The CNPS has opened the RSTI to Ivorians living abroad, with portability of rights, recognising that working lives, and remittances, cross borders. [7]
Digital on the employer side too. The CNPS reports that more than 90% of employers now complete their procedures online, and that digitalisation has supported more than 294 billion FCFA in collections through 14 partner banks. [7]
The early growth was striking. Annual RSTI enrolments rose from 1,781 in 2020 to 206,277 in 2023, reaching 457,157 self-employed workers by April 2024. [6] By 2025, the CNPS reported coverage of close to one million non-salaried workers alongside about 1.2 million private-sector employees. [8]
The honest lessons are as valuable as the achievements. By April 2024, enrolment stood at 46% of the target of one million set for the end of that year. [6] The CNPS’s own analysis notes that about half of independent workers earn less than 55,000 FCFA a month and that 55% have no formal education. [6] Published enrolment figures also do not, on their own, tell us how many members contribute regularly.
That is precisely where PensionTech must now do its work. Côte d’Ivoire has built the legal mandate and the payment rails. The next frontier is persistence: knowing which members have stopped contributing and why, prompting them at moments when they can afford to save, and serving people who may prefer a voice message in Dioula to a form in French.
A mobile money integration that only collects payments is a pipe. One that understands members is an engine.
Kenya and Rwanda: Small, Frequent and Flexible
Côte d’Ivoire is not alone. In October 2026, Kenya’s NSSF announced the national rollout of an upgraded Haba Haba micro-pension product. Members can register and contribute from any mobile phone by dialling a USSD code, with a minimum contribution of KES 25 a day. Each contribution is split automatically: 30% to a short-term emergency wallet and 70% to a long-term pension reserve. The NSSF reported that a two-month pilot enrolled 489,197 members and mobilised KES 509 million. [9]
That split matters. It accepts a truth the industry has long resisted: a household with no emergency buffer will raid, abandon or never start its retirement savings. Designing liquidity in is not a concession. It is what makes long-term saving survivable.
Rwanda’s Ejo Heza offers a further lesson. In an October 2026 CGAP discussion, its leadership described a long-term savings arrangement accepting flexible contributions through mobile money, banks and other channels, supported by local institutions, while distinguishing enrolment from saving frequency and acknowledging the continuing challenge of adequacy. [10]
Across all three countries the pattern is the same. The rails work. The challenge has moved from access to adequacy and persistence.
Sealing the Compliance Gap: From Inspection to Detection
Coverage is not only about the informal sector. In most African countries, social security contributions are already mandatory for formal employers. Yet compliance remains stubbornly incomplete, and the gap is notoriously hard to close.
The scale is sobering. Kenya’s NSSF reported that unremitted deductions rose to Sh5.1 billion in the year to June 2025, from Sh3.14 billion a year earlier, with penalties on arrears of Sh11.6 billion. Across the Kenyan retirement benefits industry, around Sh72 billion in contributions was unremitted as of June 2025, 98% of it linked to county governments and quasi-government entities. [11]
We should name the most serious form of this problem plainly. When an employer deducts a contribution from a worker’s pay and does not remit it, that is not an administrative lapse. It is the worker’s wages held by someone else. The worker usually discovers it decades later, at the counter, on the day they try to retire.
A mandatory contribution that is not collected is a promise the system has quietly withdrawn.
Traditional enforcement relies on registration drives, inspections, demand notices and litigation. These remain necessary. But they are slow, expensive and reactive, and they depend on inspectors finding employers who have every incentive not to be found. PensionTech can change the economics of compliance in five ways.
Triangulate the data that already exists. An employer who files payroll taxes for 120 employees but remits social security for 40 has already told the system where to look. Matching social security records against tax filings, company registries, public procurement payments and bulk salary disbursements through banks and mobile money can identify under-declaration without a single site visit. Kenya’s Retirement Benefits Authority has indicated it is working with the Kenya Revenue Authority on collection efforts. [11] The aim should be routine, lawful and automated data-matching, not occasional cooperation.
Turn every member into an auditor. Imagine a monthly message to each worker: “Your employer deducted 2,160 from your pay for September. We have received 2,160.” Or: “We have not yet received your September contribution.” Millions of workers checking their own records is a compliance force no inspectorate can match, provided they have a safe, protected route to report a discrepancy without fear of losing their job.
Collect at source. Where salaries are paid digitally, contributions can be split at the moment of payment, through payroll software, bank or mobile money integration, so that the money never sits in the employer’s account long enough to be diverted. Contribution remittance then becomes a by-product of paying salaries, not a separate act of goodwill.
Make compliance status live and consequential. Several governments already require a social security compliance certificate before an employer can win a public tender or renew a licence. A paper certificate is a snapshot that can be months out of date. An API that confirms current compliance status at the moment of a tender award, licence renewal or major payment closes that window.
Detect distress early. Remittance patterns change before businesses fail: payments become late, then partial, then irregular. Analytics can flag those signals while there is still something to recover, and prioritise scarce inspection resources towards the employers most likely to default.
Technology will not solve every compliance failure. The largest Kenyan arrears sit with public bodies, and that is ultimately a fiscal and political question. What PensionTech does is remove the excuse of not knowing. It makes the gap visible, quantified and attributable, and it shifts the burden from the worker who has to prove a contribution was deducted to the institution that has to explain why it was not remitted.
The best compliance strategy is one in which defaulting becomes harder than paying.
Funding Adequacy Should Not Wait Three Years
Here is a question worth asking at any trustee meeting: what is our funding level today?
For many defined benefit and social security schemes, the honest answer is: “We will know at the next valuation.” Between valuations, which are often conducted every three years, boards govern with a figure that may describe a world of different interest rates, different inflation, different membership and different asset values.
In Conversation 6, we warned against the green dashboard. A funding level that is three years old is the ultimate green dashboard.
PensionTech makes a different model possible: continuous, indicative asset-liability monitoring built into the administration platform itself.
Assets, updated daily. Custodian and fund manager feeds can value the portfolio at market each day, by asset class, issuer, currency and liquidity.
Liabilities, rolled forward from live data. Because the administration system holds the membership record, liabilities can be re-estimated as members join, leave, retire, receive salary increases or die, using the actuarial basis the trustees have approved and updated discount rates and inflation assumptions.
A funding dashboard with agreed triggers. Trustees can see the indicative funding level, its trend and its drivers, with pre-agreed thresholds that prompt escalation, an interim review or a conversation with the sponsor long before the formal valuation.
Cash-flow and liquidity matching. For schemes paying monthly pensions, the system can project benefit outflows against contribution income and maturing assets, so that the board sees a liquidity squeeze months before it arrives. For social security institutions, the same approach can track the ratio of contributors to pensioners and the reserve ratio in near real time. When the CNPS raised pensions by 10% in 2025, as part of reforms that have lifted the average monthly pension from about 89,000 FCFA in 2012 to about 160,000 FCFA, the liability consequences of that decision could, in principle, be visible immediately. [7]
Stress testing that reflects African risks. African pension funds often hold large concentrations of domestic government securities. Boards should be able to see, on demand, what a sovereign repricing, a currency shock or an inflation spike would do to the funding position, and to member benefits.
For defined contribution schemes, the equivalent is an adequacy dashboard: the projected range of retirement incomes for each cohort, showing which groups of members are on track and which are drifting towards inadequate outcomes.
This does not replace the actuary. A monitoring estimate is not a statutory valuation, and its approximations must be understood. Nor should trustees react to every daily movement; long-term liabilities deserve long-term judgment. The governance discipline lies in agreeing in advance what level of deterioration requires action, who is alerted and what happens next.
A thermometer does not cure a fever. But no doctor would wait three years to take a patient’s temperature.
The Wallet Is the New Payslip
Return to the trader in Adjamé. Her mobile money history already reveals the rhythm of her working life: which days bring income, when suppliers are paid, how much goes home, and how much she spends on transport, airtime and stock.
Today, that data is used to sell her credit. Lenders across the continent already assess borrowers from their mobile money histories. If transaction data can tell a lender how much she can afford to borrow, it can also help her decide how much she can afford to save.
With her explicit consent, a PensionTech platform integrated with a mobile money provider could do four things no payroll-based system can.
Prompt at the moment of income, not the end of the month. When a large payment arrives after a strong trading day, she receives a prompt to set aside a small amount, rather than a reminder on the 25th when her wallet may be empty.
Offer saving rules that fit her life. Save 2% of every incoming payment. Round up each supplier payment to the nearest 500 FCFA. Save more in peak trading months and less when school fees fall due. Pause without penalty when circumstances change. The behavioural insight behind the West’s “Save More Tomorrow” could become Africa’s “Save More on Market Day”.
Show realistic projections, in her own terms. Instead of an abstract fund value, she sees what her current pattern could become, after fees and in today’s money, with a range rather than a promise.
Consider an illustration. A member who saves KES 25 a day, roughly KES 750 a month, for 30 years could accumulate about KES 370,000 in today’s money if her savings earn 2% a year above inflation after fees, or about KES 512,000 at 4%. Spread over 15 years of retirement, that is roughly KES 2,400 to KES 3,300 a month in today’s money. Doubling the daily amount doubles the result. These are illustrations, not forecasts: actual outcomes depend on returns, fees, inflation and how consistently she contributes.
Notice what this projection does honestly. It shows that small amounts add up to something real, and that the same amounts alone may not be enough. That is a far more useful message than either false comfort or discouragement. It invites her to save a little more on good days, and it tells policymakers where contributory saving must be complemented by public support.
Connect today’s spending to tomorrow’s income. “You spent the equivalent of KES 1,200 on mobile data this month. Saving a third of that each month could add roughly KES 1,300 to KES 1,800 a month to your retirement income in today’s money, on the same assumptions.” Framed with care, insight like this helps a member see a trade-off she has never been shown before.
This is where Africa could genuinely leapfrog. Western automatic enrolment uses the payroll as its default. Africa could pioneer consent-based “wallet enrolment”: an opt-out retirement saving feature offered at the moment a mobile money account is opened or upgraded, with clear disclosure, an easy exit and a liquidity bucket built in.
The guardrails must be as strong as the ambition.
Consent and purpose limitation. Spending data must only be used with the member’s informed and revocable consent, and only to help her save. It must not be sold, used to target her with credit or shared beyond what the service requires. Data protection laws in Kenya, Côte d’Ivoire, Nigeria, Ghana and many other countries set the floor, not the ceiling.
Affordability before persuasion. A prompt should never arrive when the wallet is nearly empty or a household is under strain. A nudge that ignores a member’s circumstances is not behavioural design. It is pressure.
No shame in silence. A missed contribution may reflect a poor harvest, an illness or a family obligation. The system should respond with options, not judgment.
If data can tell a lender how much someone can borrow, it can help that person decide how much they can afford to save.
A Retirement Record That Survives a Changing Working Life
Consider someone who moves from salaried employment to self-employment, takes seasonal work in another country and later returns to a formal job. Their retirement history should remain intelligible through all of it.
National digital identity can help establish who a member is and reduce duplicate records, as Ghana has sought to do by linking SSNIT membership to the Ghana Card. Connected systems can make it easier to locate accounts and support permitted transfers between mandatory and voluntary schemes. But identity verification alone does not create portability. Transfer rights, benefit rules and institutional agreements still matter.
There must also be a route for people whose documents are incomplete, whose names differ across records or whose biometric checks fail. A digital identity should help a member become visible. It should not turn a correctable discrepancy into permanent exclusion.
The Most Important Innovation Is Behind the Screen
A contribution has left the member’s mobile wallet. A message confirms the payment. Yet the pension account has not been credited. From the member’s perspective, the money is missing.
Every ambition in this article, from wallet-based collection and compliance matching to real-time funding monitoring, depends on one unglamorous foundation: an accurate, reconciled member ledger. Collections must connect to allocation. Reversals and duplicates need controlled treatment. Exceptions need an owner and a time limit. Acknowledgements should distinguish money received from money allocated.
An attractive app cannot repair an unreliable member ledger. A real-time funding dashboard built on unreconciled data is not real-time insight. It is real-time error.
The member’s trust depends on what happens after they press “send”.
Small Contributions Need Serious Economics
Technology reduces repetitive work, but digital services still carry costs: payment fees, verification, administration, support, security and investment management. When contributions are KES 25 or 3,600 FCFA, those costs decide whether inclusion is real.
A board approving an inclusion initiative should understand the full cost of serving a member, how charges affect different saving patterns and what happens if participation falls short of the business case. Mobile money operators, regulators and schemes should also discuss openly whether pension contributions deserve reduced or zero transaction fees, given their public value.
Shared infrastructure, automated reconciliation and aggregation through cooperatives, microfinance institutions and other trusted organisations can reduce costs, provided individual ownership and records remain clear.
The commercial model is part of the member outcome.
AI Should Make the Institution More Attentive
AI can help members obtain explanations in familiar languages and through voice, which matters when more than half of a target population has no formal education. It can help staff navigate scheme rules, help compliance teams spot under-declaration patterns and help administrators find contribution gaps for members approaching retirement while there is still time to fix them.
But an AI assistant explaining a scheme rule must draw from approved, current information, and it should say so when it cannot answer and connect the member to a person. An automated compliance or risk flag should prompt investigation, not punishment. An irregular contribution pattern may reflect seasonal work or caring responsibilities, not irresponsibility.
Consequential decisions need accountable human review and a meaningful route to challenge.
A Reminder Cannot Create Disposable Income
Behavioural design helps people act on intentions they already hold. It cannot create income that does not exist.
A trader choosing between restocking her business and saving for retirement is making a decision about present survival. For people who cannot contribute enough, contributory PensionTech must sit within a wider social protection response that includes social pensions and, where fiscally possible, matching contributions that reward low-income savers.
An inclusive system recognises the limits of what an individual can save, and is honest with members about them.
Digital Access Must Include a Human Way Through
A smartphone app may work beautifully for its intended user and remain unusable for someone with a basic handset, limited connectivity or difficulty reading the screen. Even confident digital users need help after losing a phone, changing a number or experiencing fraud. A bereaved dependant may be approaching the institution for the first time.
USSD and voice channels, plain language, assisted service through trained agents and trusted community organisations remain part of the infrastructure. Côte d’Ivoire’s use of microfinance networks and Kenya’s reliance on basic-phone access both recognise this.
We should judge accessibility by whether people can complete essential tasks, including when something goes wrong.
Build in Africa With Confidence and Discipline
As an African pension technology practitioner and co-founder of Systech, I believe deeply in building solutions that understand the continent’s realities. Our work with FundMaster is rooted in that conviction: varied scheme structures, changing rules, multiple currencies and languages, and the practical conditions under which administrators and members operate.
“Built in Africa for Africa” carries a responsibility to demonstrate quality. It asks African technology providers to meet demanding standards of accuracy, security, interoperability and service. It also asks buyers to assess capability through evidence. Can the system reconcile a difficult mobile money payment? Can it match employer data against external sources? Can it produce an indicative funding position the actuary recognises? Can a member correct a record? Can the institution keep operating when a component fails?
Regulators, mobile money operators, identity services, tax authorities, administrators and trustees each influence whether a promising idea can function safely at scale. Africa could export valuable lessons to retirement systems elsewhere by solving these problems well. Credibility will come from delivery.
A PensionTech Agenda for Africa’s Retirement Ecosystem
Leapfrogging is not something a single institution can do alone. It needs each part of the ecosystem to act.
For regulators and governments: establish lawful, secure data-sharing between social security institutions, tax authorities, company registries and payment providers; set open standards for contribution APIs; and require real-time verification of compliance status for public procurement and licensing.
For trustees: ask for an indicative funding or adequacy position between valuations, with agreed triggers; ask what share of enrolled members contributed in the last 90 days; and ask how much of the compliance gap has been identified, recovered and still outstanding.
For mobile money operators: treat retirement saving as a core product, not a payment category; explore reduced fees for pension contributions; and build consent-based saving features with the same energy invested in credit products.
For employers: integrate payroll directly with social security and pension systems so that remittance happens automatically when salaries are paid.
For technology providers: prove reconciliation accuracy, security and interoperability before promoting the interface, and make data exportable so institutions are never trapped.
Measure the Member Journey Before Celebrating the Launch
A launch proves that a platform exists. It does not prove that inclusion has improved.
The stronger questions follow the member journey. Who completes registration, and who drops out? What share of enrolled members contributes again after three months, six months, a year? How quickly is money allocated? What share of the employer compliance gap has been closed? How current is the board’s view of funding adequacy? Can members understand their projections and correct an error? Are benefits paid accurately and on time?
Results should be examined across groups. An improvement in the average can conceal continuing difficulty for women, rural members, people with intermittent incomes or those with limited digital access.
Through MOCA, the Member Outcomes Compliance Audit, I propose examining technology against the same seven pillars we apply to the institution: accuracy and benefit integrity; timeliness and predictability; clarity and member understanding; accessibility and dignity; fairness and equity; trust, voice and redress; and governance, learning and continuous improvement.
Begin with a clearly defined group and a real barrier. Establish the starting position, test the whole journey and assign responsibility for failures. Expand when the evidence supports expansion.
The Question for Africa’s Next PensionTech Investment
Return to the trader closing her stall in Adjamé.
Imagine that after a good day, her phone suggests setting aside an amount she can afford. She sees, in her own language, what her saving pattern could mean for her income at 60, and what a little more on good days would add. When she skips a week, the system asks whether she would like to pause rather than reminding her that she has failed.
Her brother, a salaried driver, receives a message each month confirming that his employer’s deduction has reached his account. When it does not, the institution already knows and is acting on it.
Their scheme’s trustees see the funding position each month, not every three years, and act on agreed triggers before a shortfall becomes a crisis.
Years later, she claims the benefit those contributions created, with a clear record and a dependable process.
That is a retirement journey worth designing for. It is also one the developed world has not yet built for its own self-employed workers.
For the next technology investment, I would ask one question:
“Whose retirement will this make possible, what gap will it close, and what evidence will show that it continues to work for them?”
Answering it requires ambition from technology providers, courage from regulators, judgment from trustees and practical collaboration across institutions that have rarely shared data before. It also requires honesty about affordability and the limits of technology.
The capital mobilised through wider participation brings its own responsibilities. In Conversation 8, we will ask whether Africa’s growing pension assets can deliver competitive risk-adjusted returns while helping to finance the economies in which members will retire.
Africa does not need to copy the West’s retirement systems. It needs to build the ones the West will wish it had designed.
When that trader reaches for her phone after a long day, will our retirement system have a place for her, and will it know when it has failed her?
This is Conversation 7 of my series, Rethinking Retirement in Africa: 10 Conversations the Pension Industry Must Have.
#RethinkingRetirementInAfrica #PensionTech #RetirementInclusion #InformalSector #SocialSecurity #MobileMoney #MemberOutcomes #MOCA #CNPS #Africa
Sources
[1] International Labour Organization. World Social Protection Report 2024–26, as reported in Forbes África Lusófona, 15 September 2024. https://forbesafricalusofona.com/?p=109733
[2] GSMA. The State of the Industry Report on Mobile Money 2026. https://www.gsma.com/sotir/wp-content/plugins/plugin_gsma_sotir/reports/The-State-of-the-Industry-Report-2026_English.pdf
[3] Ghana News Agency. “SSNIT Scheme covers 0.85 per cent of informal sector workers.” 31 October 2023. https://gna.org.gh/2023/10/ssnit-scheme-covers-0-85-per-cent-of-informal-sector-workers/
[4] Pension Policy International. “Nigeria: Only 215,914 informal workers enrolled in personal pensions as 92m remain uncovered.” January 2026, citing PenCom Q3 2025 data. https://www.pensionpolicyinternational.com/?p=79930
[5] Abidjan.net. “Sécurité sociale: la CNPS signe une convention avec le Groupe COFINA pour collecter les cotisations sociales des travailleurs indépendants.” March 2023. https://news.abidjan.net/articles/718455
[6] CNPS Côte d’Ivoire. Présentation des régimes de prévoyance sociale des travailleurs indépendants (RSTI/RCTI). IAPRP, April 2024. https://iaprp.org/wp-content/uploads/2024/04/rsti-cnps-bke.pdf
[7] Agence Ivoirienne de Presse. “La CNPS met en avant ses réformes pour renforcer les pensions et étendre la couverture sociale.” 4 June 2026. https://www.aip.ci/cote-divoire-aip-la-cnps-met-en-avant-ses-reformes-pour-renforcer-les-pensions-et-etendre-la-couverture-sociale/
[8] 225 Assurances. “CNPS: plus de deux millions de travailleurs couverts par la protection sociale en Côte d’Ivoire.” 1 May 2025. https://www.225assurances.com/2025/05/01/cnps-plus-de-deux-millions-de-travailleurs-couverts-par-la-protection-sociale-en-cote-divoire/
[9] Citizen Digital. “President Ruto unveils national rollout of upgraded ‘Haba Haba’ digital micro-pension scheme” (sponsored). 12 October 2026. https://citizen.digital/article/president-ruto-unveils-national-rollout-of-upgraded-haba-haba-digital-micro-pension-scheme-ad-n391731
[10] CGAP. Beyond the Payroll: Expanding Pension Coverage to Informal Workers. October 2026. https://www.cgap.org/research/podcast/beyond-payroll-expanding-pension-coverage-to-informal-workers
[11] Radio Generation Kenya. “Pension defaults rise as NSSF arrears top Sh5 billion.” 11 February 2026, citing the NSSF 2025 annual report and the Retirement Benefits Authority. https://radiogeneration.co.ke/news/71780/pension-defaults-rise-as-



