The Enshittification of Global Mobility

I grew up on Andrew Henderson’s content from Nomad Capitalist, around the same time I was getting into crypto. Back then his material was genuinely engaging and he was one of the first people to teach me what geo arbitrage was. These days there is less substance to it, and once you look closely you notice that most of what’s actually useful now sits behind a service, and that service is only open to people his own marketing describes as seven- and eight-figure entrepreneurs. I’ve come across people who clear those thresholds and still never dealt with him directly, because they got the team instead.

This is a trend I tend to see quite often. Smaller agencies spill real alpha to build an audience, and then as they grow they gate the exact same alpha that earned them the audience in the first place. It raises a question nobody selling this stuff particularly wants to answer, which is why the barrier to entry for global citizenship has to be so expensive at all, given the underlying facts are all published by governments for free.

Nobody is checking the numbers

I started building a dataset of citizenship and residence routes mostly for myself, expecting the hard part to be finding the rules, when actually a lot of what’s already published about them is simply wrong.

The Investment Migration Council is the industry’s own trade body. Its homepage carries a world map of citizenship and residence programmes, which turns out to be one base64 blob inlined in the page, so you can read the whole dataset in a single request. That map still lists Montenegro’s citizenship programme as open, four years after it closed in 2022. It prices the US EB-5 at $900,000, a figure that stopped being true when the Reform and Integrity Act passed in 2022 and moved it to $800,000 in a targeted area and $1,050,000 otherwise.

Governments are not much better about their own pages. Andorra raised its passive residence threshold from €600,000 to €1,000,000 in February 2026 under Llei 2/2026, and converted a refundable deposit into a non-refundable payment, yet its own e-tramits application page still quotes the old €600,000 as of July 2026.¹ The law is in the gazette; the government’s own front door hasn’t caught up.

Why the data rots

Rule changes land irregularly, across official channels that themselves keep moving, in a dozen languages, and the stale version sits there afterwards looking every bit as authoritative as the current one. Nobody is paid to notice when a number moves, so mostly nobody does.

Underneath that sits a problem I keep running into, which is that not everything legally true is true in practice. Paraguayan citizenship is technically available, but in practice it’s a presidential decision that essentially never gets invoked, so a dataset that only reads statutes will confidently tell you that door is open when it plainly isn’t. The same applies to the strangest route I found, where Argentina treats the Falkland Islands as Argentine soil, meaning islanders born there have been recognised as Argentine by birth on declaration, which would hand them Mercosur residence rights. The legislation is inconclusive, recognition is case by case, and take-up is essentially nil.²

No guide I’ve read draws that line between what the statute says and what the ministry will actually do with your file.

What I built

Flag Paths is an open atlas of citizenship, residence and mobility rules covering 236 countries and territories. Every route links back to the government source it came from, along with the date it was last checked.

What I didn’t expect, until it was all in one place, is that buying a passport outright is the rarest way in of the lot, somewhere around one route in thirty-five, and it’s the one the entire industry is built on. Almost everything else runs through where you were born, who your grandparents were, or how long you’re willing to stay.

Every route also carries a confidence label, so you can tell which is which instead of being handed a table where everything looks equally certain. Closed programmes stay in the dataset too, recorded with the instrument that closed them, so asking whether the UK investor visa is still open returns a sourced no rather than a lead form from someone still selling it.

It works because the atlas monitors itself. A pipeline sweeps official gazettes and ministry sources daily and refuses to publish anything it cannot match word for word against the original source text in its original language.³

There is also very little agreement about what residence ought to cost. The cheapest published minimum in the whole set is €4,500, which is the reduced threshold Americans get on the Dutch self-employment route under a friendship treaty from 1956, and the most expensive asks for $25m.

The windows move in both directions

Everyone selling this tells the same story, which is that the era is ending and you should move before the door shuts. The closures are real enough, since over four years the UK’s investor visa closed, Ireland’s IIP closed, Portugal stripped real estate out of its golden visa, Australia shut its programme entirely and Canada paused its Start-up Visa. What gets left out is that roughly the same number of programmes opened or loosened over exactly the same period.

Prices move in both directions too, and the moves that matter most are the ones nobody bothers to announce. Japan raised its business manager capital requirement sixfold in October 2025, from ¥5m to ¥30m. Cayman roughly doubled its investor fees in May 2026 under the renamed Caymanian Protection Act, which I only caught because I’d been reading Cayman regulation for other reasons.⁴ Bahrain went the other way entirely and dropped its golden residency property threshold from BHD 200,000 to BHD 130,000 without announcing it anywhere. I found that by diffing two captures of their own page, and as far as I can tell nobody has reported it.⁵

The churn is why I ended up building something that re-checks itself instead of writing a guide, since a guide written once would only keep its value if the windows moved in one direction, and they clearly don’t.

What I can’t tell you yet

Taxes aren’t in it, and adding them isn’t a matter of one more column. Two people can take the same route into the same country and end up nowhere near each other: an American in Dubai still files with the IRS on worldwide income, a Brit in the same apartment doesn’t. What you actually owe turns on your own nationality, how many days you were where, where the money arose, and which treaty was in force when it did. I can’t hang any of that off a country, and a country is what everything else in the atlas hangs off.

Bloc membership and route pricing are already joined, so I can ask for the cheapest way from a given passport into Mercosur or the Gulf and get something back. What comes back is investment routes, because those publish their numbers and they went in first. A job offer, a spouse, a university place, none of those are in there yet, so it’s answering a narrower question than the one you asked, and doing it confidently. It needs the ordinary routes before I’d let it recommend anything to anyone.

The theory-versus-practice problem doesn’t fully go away either. I can label confidence and record that a route is discretionary, but I can’t tell you whether a particular ministry will actually process your file this year. That gap is where the expensive intermediaries genuinely earn their fee, and I’d rather say so than pretend the dataset closes it.

The whole thing is open source, and I’d treat every medium-confidence route in there as a standing request for corrections. If you know a jurisdiction properly, particularly one where the published rules and the practice diverge, I’d like to hear from you.