The African Development Bank responded to questions from MOVING AFRICA. Questions have been put to IATA and to Henley & Partners. Responses are pending.
We wrote about this once already. A contradiction we could not explain, and honestly still cannot fully explain, even after getting an answer.
Seychelles sits at the very bottom of the Africa Visa Openness Index, ranked 54th out of 54 countries. Its score is 0.000, which is the lowest a country can get on that index. The index is put together every year by the African Development Bank together with the African Union Commission.
Now look at a different ranking, the Henley Openness Index. On that one, Seychelles comes second in the entire world. Out of 198 possible nationalities, it lets in 197 without asking them to get a visa first.
Two rankings. Same country. Same border. One says it is the most closed place in Africa. The other says it is one of the two or three most open places on Earth.
That gap made no sense to us, so we went and asked the African Development Bank directly why they scored Seychelles the way they did.
They answered. And their answer pointed the whole story in a direction we did not expect.
What the Bank told us
We asked a simple question first. Seychelles lets travelers arrive without an approved travel authorization and finish the paperwork on the spot, for a 30 euro penalty. Why wasn't that classified under "visa on arrival," a category the index scores at 0.8, instead of under "visa required before travel," which scores zero?
Here is the Bank's answer, given to us word for word because it came from official correspondence and we are not going to trim it:
"In order to ensure a consistent methodological approach, the report assesses visa openness based on the requirements that are stipulated in the IATA database, and this informs the assessment of Seychelles' travel policies in the context of the AVOI methodology. During the data collection period (which ran from June-August 2025), the IATA database stipulated the requirement of an ETA ahead of travel – i.e. compulsory."
We also asked where the report got its description of what Seychelles actually requires from travelers, since we could not find that description on the main government summary page. The Bank pointed us to the FAQ page of the Seychelles Electronic Border System, which does spell it out. Turns out we had been looking at the wrong page. We corrected our earlier article to fix that.
Last question: South Africa is planning a 500 rand fee for visa-exempt travelers who use its new electronic travel authorization system. The Bank's own report recommends that these kinds of authorizations be free of charge for countries that already allow visa-free or visa-on-arrival entry. Isn't that a contradiction? The Bank told us the index supports easing movement across African borders in general, but that it could not comment on any one country's proposed fees.
Three answers. The first one is the one that matters.
The index does not look at Seychelles directly
Here is the part worth reading twice.
The Africa Visa Openness Index does not send someone to read the Seychelles government's website and decide, based on that, what the rules actually are. Instead, it pulls from a database run by the International Air Transport Association, known as IATA, and uses whatever that database says the requirement is.
The Bank explained why. Consistency. You cannot score 54 different countries fairly if 54 different researchers are each forming their own personal read of 54 different immigration systems. You need one source, applied the same way, every time. Using an industry database that airlines already rely on is a reasonable way to do that. Arguably it is the only workable way.
But this also means something important. The index's finding about Seychelles is not really a finding about Seychelles. It is a report of what one database said, between June and August of 2025.
Now compare that to Henley.
Henley pulls from the same source
The Henley Openness Index describes itself as basically a mirror image of the Henley Passport Index. According to Henley & Partners' own published methodology, the Passport Index is built using exclusive data from IATA, which Henley describes as the largest and most accurate travel information database in the world. Henley says its own research team then adds to and refines that data.
So both rankings, the one that puts Seychelles last and the one that puts it second, trace back to the same underlying data source.
That changes what kind of story this is.
This is no longer really a disagreement between a public development bank and a private advisory firm about what "openness" should mean. When you put their published rules side by side, they are actually pretty close to each other. Both essentially ask: does a traveler need government approval before they leave home, yes or no?
So this is not a disagreement about definitions. It is a disagreement about what a shared database actually says, or about how two different organizations are reading the exact same thing.
And that is not something either organization can resolve by itself.
Why the IATA database matters more than it looks like it should
Here is the part that turns this from a technical detail into something bigger.
The IATA database that both rankings lean on is not some reference book that sits on a shelf and gets consulted once a year by researchers. It is a live operational tool. Airline staff use it at check-in, right now, today, to decide whether a passenger has the paperwork needed to be let into their destination country.
That is the actual reason an airline can turn you away at the gate before you ever reach a border officer.
The person checking you in is not personally interpreting immigration law. They are looking at what a database says, and the database is telling them whether you are cleared to fly to wherever you are going.
So the exact same source that decides whether someone can board a plane to Seychelles is the same source sitting underneath both of these very different rankings of how open Seychelles is.
If the database says a travel authorization has to be approved before departure, that is not just a data point sitting in a spreadsheet somewhere. It is the actual instruction the airline follows when your boarding pass gets scanned.
The Bank told us plainly that during its data collection window, this database listed the Seychelles ETA as required ahead of travel, and described it as compulsory.
That makes Henley's decision to count Seychelles as open harder to explain, not easier. Henley's own methodology gives a score of zero to any system that needs government approval before departure, and that includes a visa on arrival that still requires advance clearance. According to Henley's published methodology page, a country only counts as open under its system when a traveler can get a visa-free entry, a visa on arrival, or an electronic travel authority "without needing pre-departure government approval." If pre-departure approval is required, the methodology assigns a score of zero for that pairing.
We genuinely do not know how Henley reaches its Seychelles classification given its own stated rule, and we are not going to guess. There are a few possible explanations, and none of them is confirmed.
One possibility: the two organizations checked the database at different moments, and the entry for Seychelles changed in between. That is not far fetched. Henley's Passport Index, and by extension its Openness Index, is described on Henley's own site as being based on data that updates monthly. The Africa Visa Openness Index, by contrast, is collected once a year, in a fixed window, which in 2025 ran from June to August. A source that updates monthly and a source that gets checked once a year, in a fixed window, will not always line up, even when they are pulling from the exact same underlying database.
Another possibility: Henley's research team, which by its own account "enhances" the raw IATA data, might handle the Seychelles ETA differently than the raw database entry does.
A third possibility: the database entry itself might contain more nuance than either summary captures, something that genuinely supports two different readings depending on how you interpret it.
We put these questions to both organizations. We are waiting to hear back.
Why this matters beyond one small island nation
Step back from Seychelles for a moment.
Conversations about mobility across Africa run on numbers like these. When the African Union talks about its free movement protocol, when officials debate the African Continental Free Trade Area, when a government wants to claim credit for opening its borders or a critic wants to argue the continent is closing up, visa openness scores are the currency people reach for.
The 2025 Africa Visa Openness Index recorded a continental average openness score of 0.445. According to the African Development Bank's own 2025 report, that is significantly lower than each of the three previous years and roughly on par with where the index stood back in 2021. The report attributes part of that drop specifically to Seychelles' shift toward an ETA system, along with similar shifts in countries like Guinea-Bissau, Mauritania, Nigeria and Somalia, all of which moved from broad visa-on-arrival policies toward requiring approval ahead of travel.
That 0.445 figure is going to get quoted in speeches, cited in policy papers, and used as evidence in arguments for the next year at least.
And underneath a meaningful part of it sits an entry, or several entries, in a commercial aviation database.
We want to be careful here, because this is not an accusation that the data is wrong. The IATA database exists because airlines genuinely need accurate, current entry requirement information, and the cost of getting it wrong, in fines, in stranded passengers, in reputational damage, is severe. It is likely one of the most carefully maintained sources of this kind of information anywhere in the world.
But three things follow from all this, and they deserve to be said plainly.
It is not public. If you are a citizen of an African country and you want to understand why your country got the score it did, you cannot go read the actual database entry yourself. The index gets published. What sits underneath it does not.
It was built to answer a different question. The database exists to help an airline decide whether to let someone board a plane. It was never designed to measure whether a border is "open" in the sense that African integration policy uses that word. Those two questions overlap a lot, but they are not the same question. Something can be simple for an airline gate agent to check and still be a real burden for the traveler who has to go through it.
It has ended up as the final word almost by accident. Nobody sat down and decided that a commercial airline database should be the reference point for measuring African mobility. It became that because it was the one source consistent enough to apply across 54 different countries at once. That might be a completely reasonable outcome. But it is worth naming out loud rather than quietly assuming.
What we have asked IATA
We sent IATA the following questions and are waiting on a response.
What does the database currently list as the entry requirement for Seychelles, and what did it list between June and August of 2025 specifically?
Does the database note that a traveler who shows up in Seychelles without an approved Travel Authorization can still complete the application on arrival, for a penalty fee?
How does the database classify electronic travel authorizations generally? Is there a difference in how it treats an authorization that gets granted automatically versus one that requires an actual human decision?
How often do entries in the database get updated, and what causes an entry to be reviewed?
Is this database used by outside organizations to build published rankings of visa openness, and does IATA have a view on whether it is well suited to that kind of use?
What we have asked Henley & Partners
We sent similar questions to Henley & Partners.
Their published methodology only counts an electronic travel authority as "open" when it does not require pre-departure government approval. Anything requiring that kind of approval gets a score of zero.
The government of Seychelles requires airlines to check that a passenger has an approved authorization before letting them board.
We asked Henley directly: on what basis does the Seychelles system meet the condition for an open score under their own stated rule?
We also asked about Guinea-Bissau, a country that holds a perfect score of 198 out of 198 on Henley's index. According to the African Development Bank's own 2025 report, Guinea-Bissau actually moved from visa on arrival to requiring a visa ahead of travel, and it launched a new e-Visa portal in the process. Under Henley's own methodology, a government-approved e-Visa required before departure should score zero, not a perfect 198.
We have not received a response from Henley yet, and we will publish it in full once we do.
What we actually know, and what we do not
Here is what we can say with confidence.
The African Development Bank has confirmed, in writing, that the Africa Visa Openness Index scores countries based on what the IATA database says, and that during its June to August 2025 data collection window, that database listed the Seychelles ETA as required ahead of travel and compulsory.
Henley & Partners states, in its own published methodology, that its Passport Index, and therefore its Openness Index, is built on IATA data that its research team then enhances further.
The two indices place Seychelles at opposite ends of their scales, one at the very bottom, one near the very top.
The government of Seychelles requires airlines to check for an approved travel authorization before allowing someone to board, while also allowing travelers who show up without one to apply on the spot for a 30 euro penalty, according to the official Seychelles Electronic Border System.
We cannot see what the IATA database currently says about Seychelles, because we do not have access to it.
We cannot see how Henley's research team processes or adjusts the raw IATA data before it becomes the Openness Index.
We cannot tell whether the two organizations pulled from the database at different times, or whether the entry itself changed in the gap between their two assessments.
Those are exactly the questions we have sent to the organizations that are actually in a position to answer them.
The bigger picture here
This whole series keeps circling back to one idea. Africa's borders are changing faster than the tools we use to describe them can keep up.
Permission used to happen at the border. Now it increasingly happens before you ever leave home. Authorizations that are technically not called visas are doing much of what visas used to do. Who gets easy access and who does not is increasingly sorted by nationality in ways that never show up if you just ask whether a visa is required, yes or no.
That "yes or no" question used to be enough. It no longer is, and to its credit, the African Development Bank's own report seems to know this. It explicitly warns against letting electronic travel authorizations quietly turn into visas by another name. It recommends that when a country already offering visa-free or visa-on-arrival access adopts one of these authorization systems, the system should be free of charge and should only ask for basic information. That is a genuinely thoughtful position, more careful than a single ranking number ever suggests on its own.
But even a careful methodology is still limited by whatever source it has to rely on.
If the foundation underneath Africa's visa openness measurement is a database built to answer a completely different question, whether to let someone board a plane, then the continent's picture of its own mobility is being shaped by a system that was never designed for this purpose, that is maintained by a private commercial company, and that the public whose movement it describes cannot actually go and read for themselves.
That arrangement might be entirely workable. It might genuinely be the best option available, better than the alternative of 54 separate national assessments done by 54 different research teams with 54 different blind spots.
But it should be something people have chosen deliberately, not something that happened by accident and now goes unmentioned. Right now, very few of the people quoting these numbers seem to know how they actually get made.
We did not know either, until we asked.
Part of MOVING AFRICA, a series on the cost and difficulty of travelling across the continent.
Sources referenced in this piece:
- African Development Bank & African Union Commission, Africa Visa Openness Report 2025 (visaopenness.org)
- Henley & Partners, Henley Openness Index Methodology (henleyglobal.com)
- Seychelles Electronic Border System, official government FAQ (seychelles.govtas.com)
- Club of Mozambique, coverage of the 2025 Henley Openness Index rankings