The Best Case for Yes — and Why I Still Vote No

I have spent this series arguing one side of Iceland’s referendum. That earns a reader’s suspicion, and it should. So before the vote on 29 August, I want to do the opposite of what an advocate usually does: make the strongest case I can for the side I am voting against — not a caricature of it, but the real thing, at full strength, the way its most serious proponents make it. Then I will tell you, honestly, why I have weighed it and still land on No. If the Yes case cannot survive being stated fairly, it was never worth much. I think it can survive being stated fairly. I just think, on balance, it loses.

Let me make the case for Yes properly.

The real advantages of joining

The euro removes a genuine, permanent cost. Set aside the disputed promise of dramatically lower interest rates. A small open economy running its own micro-currency pays a standing tax that never goes away: exchange-rate risk on every transaction, hedging costs, and the thin-market volatility that the government’s own report documented as real. Every importer, exporter, and traveller pays a spread a eurozone member simply does not. For a country where trade is a large share of the economy, eliminating that friction is a real, recurring efficiency gain. The report puts the cost-of-capital benefit at 25 to 100 basis points. That is modest, but it is not nothing, and it compounds every year forever.

The customs union closes a gap the EEA leaves open. This is the Prime Minister’s strongest point, and it is a fair one. The EEA gives Iceland the single market but not the customs union — which means Icelandic exporters still face rules-of-origin paperwork, some tariffs on processed marine products, and the standing risk of becoming a bargaining chip in a trade dispute. Membership would remove that friction. Brexit is the evidence that sitting outside the customs union carries real, measurable costs in delay and expense. This is a concrete gain, not a theoretical one.

Deeper capital markets — for exactly the companies I care about. This one I feel personally, because it cuts against my own life’s work. A foreign investor putting money into an Icelandic startup today takes króna risk on top of company risk. Inside the euro, and inside the EU’s Capital Markets Union, that currency layer disappears, and Iceland plugs into a continental pool of capital far deeper than anything it can build alone. The report itself notes that the capital-markets channel would especially help start-ups. My whole “build it here” argument assumes Iceland can build that depth domestically. A Yes advocate can fairly say: the euro gives you that depth overnight, instead of over fifteen years.

A seat at the table, instead of the fax machine. This is real, and I will not minimise it. Through the EEA, Iceland already adopts a large share of EU rules with no vote on them — they arrive from Brussels and Iceland implements them. Critics call it “fax democracy,” and they are right. Membership would convert Iceland from a rule-taker into a rule-shaper: six seats in the Parliament, a Commissioner, a vote in the Council. On the rules Iceland already lives under — including the ones that will govern artificial intelligence — having a small voice is more than having none.

The credibility argument — and this is the strongest one. Here is the Yes case at its most serious, and it uses my own series against me. I have spent weeks arguing that Iceland must summon the political will to fix indexation, the banking oligopoly, and fiscal procyclicality. A sharp Yes advocate turns it around: Iceland has had forty-seven years to fix indexation and has not. If the domestic political system genuinely cannot discipline itself, then an external anchor that forces discipline — the Maastricht criteria, the fiscal rules, the surrender of the devaluation escape valve — is not a cost. It is the entire point. Countries have done exactly this: Italy joined the euro in part to import a monetary discipline it could not generate at home. “Tie yourself to the mast” is a real and respectable theory of political economy, and it is the mirror image of my own “Are We Willing.” I say we must find the will. They say we have proven we will not, so we should bind ourselves. I do not find that stupid. I find it the hardest thing I have to answer.

And the security and AI dimensions. The EU adds a second layer of mutual defence beyond NATO, and deep integration on sanctions, cyber, and resilience. And on artificial intelligence — the defining technology of the age — the EU is the world’s most consequential regulator, the bloc that made Google, Apple, and Meta change their behaviour, the author of the AI Act. For a country of 380,000 with no leverage over the technology giants, sheltering inside the one democratic bloc large enough to set terms is not nothing.

That is the case for Yes, made as well as I know how to make it. Every point on that list is real. None of it is a strawman. If I could not feel the weight of it, I would not trust my own conclusion.

Why, weighing it, I still vote No

Now the weighing — and the honest form of this is not to deny the advantages, but to set each against what it costs.

The friction cost is real but small; the thing traded for it is permanent. The euro’s efficiency gain is a modest, recurring saving. What it buys is the irreversible loss of the one tool that, in 2008, let Iceland’s economy swing from a current-account deficit of a quarter of GDP to balance within two years — a faster adjustment, the government’s own report says, than the euro-periphery countries achieved through internal devaluation. You are trading a small perpetual saving for the elimination of your only shock absorber in the rare event that matters most. That is a bad trade for a small, volatile, commodity-shaped economy, however attractive the saving looks in calm weather.

The capital depth is real but buildable at home. Yes, the euro would deepen the capital available to Icelandic companies. But the specific bottleneck — the thin domestic market for hedging currency risk — has a domestic fix that the IMF itself recommended: deepen Iceland’s FX and derivatives market and loosen the limits on it. That keeps the currency and gets much of the depth. Fifteen years of building it yourself, with the levers in your hands, beats acquiring it by abolishing your central bank.

The customs-union gain is real but priced against the fishery — and the fishery is too dear. This is the crux. The customs union would help. But the price of admission is the Common Fisheries Policy, and France has already declined to guarantee Iceland permanent control over its own waters — a control that no acceding state has ever obtained in the Union’s history. A concrete gain in trade friction, weighed against surrendering authority over the resource the modern Icelandic state was built on, is not a close call. The gain is real and the price is your foundation.

The seat is real but tiny. Converting Iceland from a rule-taker to a rule-shaper sounds like a gain in sovereignty, and in one narrow sense it is. But the arithmetic is unforgiving: it is roughly 0.08% of the vote, in exchange for 100% of the national control Iceland now holds over the domains the Union decides collectively. Trading complete control of your own affairs for a fractional voice in someone else’s is not a gain in sovereignty. It is a change in its form, and a reduction in its substance.

The credibility argument — the strongest — has the deepest flaw. Yes, an external anchor imposes discipline. But look at what it disciplines. The euro binds Iceland’s hands by removing the currency — it amputates the tool rather than curing the weakness. And here is the part that undoes the argument: it leaves the actual domestic machine entirely untouched. Indexation stays. The banking oligopoly stays. The 3.5% pension reference rate stays in Icelandic law, because Frankfurt has no power over it. So the credibility import binds the wrong thing: it takes away the shock absorber while leaving in place every structural distortion this series has been about. You would arrive in the euro with your hands tied and your house still broken. Discipline imported by amputation is not the same as discipline built — and only one of them fixes anything.

And AI — the argument dissolves on inspection. The claim is: join to shelter from the AI threat. But Iceland already receives the EU’s AI rulebook through the EEA, by the same mechanism it receives GDPR and every other digital regulation — the AI Act is marked EEA-relevant and is being incorporated, and Iceland already sits as an observer on the EU’s AI Board. So the protection is already here without membership. What membership would add is a fractional vote on writing those rules. And on the thing that actually matters — capability, compute, the ability to build and defend at the frontier — the EU cannot help, by its own account. The Draghi report, the EU’s own competitiveness diagnosis, records that in 2023 the EU invested $8 billion in AI against America’s $68 billion; that last year the US produced 40 large foundation models, China 15, and the EU three; and that Europe’s own capital markets are failing even its flagship AI company. Sheltering inside the EU for AI means sheltering inside the bloc that leads the world in regulation you already receive for free, and trails it badly in the capability you would actually need. Iceland’s real AI advantage is the opposite of absorption: abundant green energy for compute, a tiny high-trust society that can adopt and pilot faster than any bloc, and a proven ability to build globally competitive technology from almost nothing. The AI age rewards the agile and the sovereign, not the absorbed.

What tips the scale

Add all that up and the ledger is close enough that an honest person could linger over it. The advantages are real; the answers are real. If the decision were only a sum of quantifiable costs and benefits, reasonable people would land on both sides, and I would not fault them.

But the decision is not only that sum, and here is where I stop being an accountant and tell you what I actually believe. Some things do not sit on the cost-benefit line at all. Independence — the standing to make your own choices. Strategic latitude — the freedom, having committed to the alliances a small nation genuinely needs, to stay unbound on everything else: to trade with whom you choose, take your own diplomatic line, and price your own strategic value case by case rather than dissolving it into a common position. Sovereignty — full authority over the resource, the currency, and the law. These are not line items to be netted against 50 basis points. They are the frame inside which every future choice — including the ones we cannot foresee today — will get made.

And this is the logic of my whole series returning at the end. In a power-law world, the value of keeping your own levers is precisely that you do not know which one you will need, or when. 2008 was not forecastable. The next one will not be either. What you are being asked to do is trade a modest, quantifiable, calm-weather gain for an open-ended, permanent, unquantifiable surrender of the ability to choose when the weather turns. A small nation that keeps its levers keeps its options. A small nation that spends them for a discount has made the one trade it can never reverse.

So I have weighed it, with the ledger open, and this is my judgment — not a proof, a judgment. The material advantages of joining are real, and I have tried to give them their full due. But independence, latitude, and sovereignty are worth more to a small nation than the union on offer, because they are worth more than any price that can be written down. On 29 August, weighing the whole of it, I am voting No.