
Trust & Verification, 21 October 2026, 11 min read
The Eighteen Percent: Why Intra-African Trade Stays Low
Tariffs within Africa average 6%. Non-tariff measures cost the equivalent of 18%. Much of that is information: finding, checking and trusting a counterparty across a border.
There is a number that explains more about Africa's economy than its GDP does.
Start with the GDP anyway. The IMF's World Economic Outlook data (October 2026 edition) put the combined 2025 output of Africa's 54 economies at about $3.2 trillion, produced by about 1.49 billion people. Set that beside three other economies of comparable or larger population:
| Economy | People (2025) | GDP (2025, current US$) |
|---|---|---|
| United States | ~342 million | ~$30.9 trillion |
| China | ~1.40 billion | ~$19.6 trillion |
| India | ~1.46 billion | ~$3.9 trillion |
| Africa (54 countries) | ~1.49 billion | ~$3.2 trillion |
Source: IMF World Economic Outlook database, October 2026. The Africa row is the sum of the 54 country estimates.
India and Africa have roughly the same population. India produces more, and India is one country. Africa is fifty-four.
The instinct is to reach for the usual explanations: not enough industry, not enough infrastructure, not enough capital, not enough institutions. All true. All real. And all of them take a generation and a great deal of money to fix.
Every one of those is a problem of missing inputs. They describe an Africa that is poor because it lacks things.
That framing misses the most interesting part. Africa is not only short of new capacity. It also underuses the capacity it already has. The continent has the people, the land, the minerals and, increasingly, the businesses. What it lacks is an easy way to connect them to each other across a border. That particular gap, unlike the others, does not require a generation or a sovereign balance sheet to close.
The conversion gap
Start with the fact that should be more shocking than it is. Afreximbank's African Trade Report 2026, launched in June 2026, puts trade between African countries in 2025 at about $213.8 billion, out of roughly $1.5 trillion in total African merchandise trade (Afreximbank press release). That is about 14%, or one dollar in seven. In Europe and Asia, an IMF working paper on the AfCFTA notes, intraregional trade is above 50%. An African business is more likely to sell to Rotterdam than to its neighbour two countries over.
For years the standard explanation has been tariffs and the fragmentation of 54 separate markets: different rules, customs regimes, currencies, standards. Cross a border in Africa and you re-paper everything. That diagnosis was correct, and it produced the right response: the African Continental Free Trade Area, which its Secretariat describes as the world's largest free trade area by number of participating countries. Trading under it began on 1 January 2021. If you want the detail of what it covers, read what the AfCFTA settles and what it leaves to firms and our guide to AfCFTA benefits for SMEs.
So here is the uncomfortable question. If the tariffs are being handled, why is the trade still not flowing?
What is left when the tariffs are gone
The answer is in one well-documented estimate that deserves far more attention than it gets. Writing in the IMF's Finance & Development in September 2023, Jihad Azour and Abebe Aemro Selassie put the average import tariff within Africa at 6%, and the non-tariff measures (everything that is not a customs duty) at the equivalent of an import tariff of 18%.
Eighteen percent, three times the tariff. That is the real wall.
The IMF's modelling points the same way. Its May 2023 departmental paper, Trade Integration in Africa, finds that cutting tariffs on intra-African trade by 90% and non-tariff measures by half would raise the median merchandise trade flow between African countries by about 15%. Pairing the AfCFTA with comprehensive reforms to the wider trade environment raises it by about 53% over the long term. Of those 53 points, 15 come from the tariff and non-tariff cuts and 38 from the environment around them. The headline cuts were never the whole constraint.
That gap is worth handling carefully. It is not a fee someone is skimming that could be refunded. It is commerce that never happens at all: the deal that is not struck because the buyer and seller never found each other, could not check each other, or could not risk a first transaction across the border. The eighteen percent is less a tax on African trade than a measure of the African trade that does not exist.
It also survives every tariff outcome. The tariff fights make the news: AGOA's lapse in 2025 and its extension through 2028, new duty schedules from partners in every direction, the AfCFTA's own phased cuts. Each of those moves operates on the tariff layer. Win them all and the 18% behind them still stands, because it was never written into a tariff schedule. (On what the latest US decision does and does not change, see what a trade deal can't do for US–Africa trade.)
So what is this 18%, concretely? Strip away the economist's language and much of it resolves into three plain failures:
- A buyer in Lagos and a supplier in Nairobi never find each other.
- When they do, neither can cheaply check that the other is real, solvent and competent.
- And even when both check out, there is no structure that lets them trust a first transaction across a border, in a different currency, under a different legal system, with no shared history and no one to call if it goes wrong.
Finding. Checking. Trusting. None of these is a tariff. None of these is a road. Every one of them is a problem of information, and information problems are the cheapest problems to fix, because fixing them requires no concrete, no port and no act of parliament.
One market, read as one
Look back at the table. The United States is one market. So is China. So is India. A business in any of them can reach hundreds of millions of customers under one readable set of rules: one framework for who counts as a real company, one way to check a counterparty, one body of law a stranger's contract can be enforced under. A firm in Chengdu can sell to Shanghai without re-papering its existence.
Africa's 1.49 billion people are not one readable market. They live in 54 countries, each with its own rules, registries, standards and ways of establishing who is real. The continent does not lack participants. It lacks a way for those participants to read each other across a border.
It would be easy to read that as a counsel of despair: too many countries, too many systems, nothing to be done short of redrawing the map. The European Union is the evidence against it. The EU did not start as one market. It started as a continent of sovereign states that did not share a language, a currency or a legal tradition, after centuries of war with each other. Nobody dissolved those countries. France is still France. What Europe built was the connective tissue: common standards, mutual recognition, shared ways of checking and enforcing. That let separate systems behave, for commercial purposes, like one. Europe did not remove its fragmentation. It made the fragmentation legible across the seams.
That is the most useful fact for Africa in the comparison, because it means the binding constraint is not the number of countries. It is whether a counterparty in one country can read, check and trust a counterparty in another. Europe solved that slowly and expensively, through treaty and institution: thirty-six years passed between the 1957 Treaty of Rome and the launch of the single market on 1 January 1993. The AfCFTA is Africa's treaty layer, and it is already doing the tariff work. The deepest layer of legibility (can I find this firm, confirm it is real, and risk a first deal with it) no longer has to be built only out of treaties.
The word for this is legibility
There is a useful way to name what is missing. It is not trust. Africa has enormous quantities of trust. The trader who extends goods on a handshake, the rotating savings group that has never needed a contract, the investor who wires money to a cousin's business on reputation alone: these are high-trust systems. They work. They have worked for generations.
The problem is that this trust is illegible to anyone outside the relationship. It does not travel. A Nairobi supplier's twenty-year reputation in its own market is invisible to a buyer in Accra and worthless to an underwriter in London. The reputation exists; it simply cannot be read by the people who would need to read it to do a deal. M-Pesa is the proof of concept, and it is worth sitting with: it did not create trust between Kenyans, who already trusted each other. It made existing trust legible to a system. Kenya did not get more trustworthy. Its trust got readable.
Africa's deficit is not a trust deficit. It is a legibility deficit: real, deep, relationship-based trust that the contract-based systems of cross-border commerce cannot see.
Proving the same firm five times
The cost of illegibility is easiest to see by following one firm. Take a hypothetical mid-sized agro-processor in Kumasi. Ask who it sells to and buys from, and the answer is rarely one country. It is a portfolio: a European buyer with its own traceability requirements, a Gulf importer with others, a Chinese supplier of inputs, an American distributor buying under AGOA, a regional customer across an AfCFTA border.
That portfolio is not indecision. It is the rational posture of a firm that has watched outside partners change terms. AGOA lapsed on 30 September 2025, was restored retroactively by a one-year extension signed on 3 February 2026, and was extended again in September 2026, through 31 December 2028. A firm that depends on one partner's terms learns to keep others.
But the portfolio carries a cost that is rarely named. The Kumasi firm proves its existence, ownership, capacity and standards to the European buyer. Then it proves all of it again, in a different format, to the Gulf importer. Then again to the American distributor's procurement team, to the bank financing the inventory, to the insurer underwriting the shipment, and, in the most painful version, to a regional buyer two borders away who has no way to read a Ghanaian firm's local reputation at all. Same firm, same facts, re-established from scratch for every counterparty that asks. None of those readers can see what the others already checked.
Non-tariff measures cover exactly this ground: the documentation, the standards, the checks, the cost of one party establishing to another that it can be trusted. My argument is that a large part of the 18% is spent re-proving the same firm to a new reader who cannot see the prior proof. And the more partners in the portfolio, the more times the firm pays.
What would help is not another preference scheme. It is portable credibility: a record established once, from one documented foundation, that the European buyer, the Gulf importer, the American distributor, the regional customer and the bank can each read in their own terms. That layer is partner-neutral. It serves whoever is asking, which is why it is worth more to a multi-partner firm than any single relationship in its portfolio.
Why this is the cheapest fix on the continent
A caution, because this is where arguments like this one usually overreach. The $3.2 trillion economy is not anyone's addressable market. No platform captures a slice of continental GDP or trade.
The honest claim is narrower and stronger. Of the things suppressing African output, most are capital-heavy and generational, and one is informational and immediate. The legibility layer is the lowest-cost response to the largest fixable barrier, not because it is worth trillions to whoever builds it, but because everything else on the list (the factories, the grids, the deep capital markets) needs counterparties who can find, check and trust each other first. Legibility is not one item on the development checklist. It is a precondition for most of the others.
The tariff wall is coming down. The 18% wall behind it is largely made of missing information, and that is a wall you can take down with documentation, checking and structure rather than with concrete and capital. The countries and firms that make their existing trust readable to the rest of the world will convert a far larger share of what they already have.
If you buy or sell across an African border, take your last cross-border deal and write down where its time and money went before it closed: finding the counterparty, checking them, and securing the first payment. Those three lines are the costs this essay is about. For the margin side of the same problem, read where the margin goes; for the checking side, our guide to finding and checking export suppliers in Nigeria, Kenya and South Africa.
AfroSynergy connects international buyers with African exporters. Listings open on 15 January 2027. To source a product or find a counterparty now, send us an enquiry.
Ready to Explore Opportunities?
Discover funding opportunities, partnerships, and projects across Africa. Founding-cohort access is by invitation during pre-launch.
Comments
Sign in to leave a comment. Comments are moderated.
Keep reading
Sourcing from Africa?
Listings open on 15 January 2027. Tell us what you need to source and we will come back to you.
Or get new guides by email.
By subscribing you agree to our Privacy Policy. Unsubscribe anytime.


