It was one of the tightest counts the country has seen. Not until around three on Sunday morning did the No vote pull clear of the Yes. The final figures were 52.8% to 47.2% against reopening accession negotiations with the European Union, on a turnout of 82.6%, the highest in any Icelandic election since 2009. A shift of 6,500 votes would have reversed it.

Abroad the result is being read as a verdict on Europe, or as a signal of which way the geopolitical winds are blowing. What was decided, however, was smaller and closer to home: Icelanders were settling long-standing domestic arguments about the management of their economy and the control of their resources, and about what membership would do for either. Iceland's existing integration was never in question: inside the single market since 1994, inside Schengen, and applying a large share of EU law. Since the vote the firm cross-political consensus is that the authorities must now redouble their efforts to ensure the EEA's effective functioning and to keep cooperation with Brussels strong.

What follows is a first reading, three days after election night, with its inherent limitations. We focus on three questions. What the campaign was fought over. What the result leaves Iceland to work with. And what Brussels and Washington can take from it.

The campaign

The referendum was unusual in itself. No government had put the opening of EU accession negotiations to its voters before. Iceland had a good reason to. The country's 2009 application rested on a parliamentary vote alone and was contested for it; the 2015 withdrawal went to Brussels by letter, without a vote at all, leaving the question to simmer ever since. It was also fought across the summer holidays, when much of the electorate was abroad or in their summer hyttes. And it was a vote in which independent campaign organisations, rather than political parties, dominated the campaigning as never before, while no law effectively governed their operations, including what they could raise, or from whom.

The Yes case, and the case it did not make

The coalition had promised a referendum no later than 2027, and then brought it forward. Trump's brazen threats against Iceland's cousins to the west were part of the reason, though by no means all of it. There were the tariffs, and there was the sense that the wider order was coming apart. At Davos in January the American president repeated his claim to Greenland and referred to Iceland while doing so, and in the same week Mark Carney told that audience the rules-based order was over. Politico reported in February that the vote could come as early as August. Alþingi set the date on 28 May.

The government had every reason to be confident. Yes led in the first polls. It also had the mover's advantage, and the ask was modest. Voters were asked whether they wanted to find out more.

There was a case to be made, too. Iceland has one of Europe's highest rates of home ownership, and roughly half of all households own with a mortgage, a share at the top of the European range alongside the Netherlands and Norway. They are also among the most expensive mortgages in Europe. A small, thinly traded currency carries a risk premium, and the central bank has to hold rates higher to keep inflation anchored. Most mortgages are then indexed to the consumer price index, so inflation passes straight into the household's debt. An independent currency does buy flexibility, and it proved its worth after the 2008 crash. But when the adjustment comes it lands on the households, while a weaker króna rewards the export sectors. Add a small market with few firms in each sector, little foreign competition and foreign investment, and the Yes campaign had a set of grievances most households meet monthly.

It framed the referendum narrowly. Voters were asked to think procedurally: whether to look at the terms, with a second referendum on any agreement. Já til að sjá, yes to see, was the slogan. The Yes camp left the final destination out of the argument; that debate was to belong to the second referendum, which will not now be held. The framing was to focus instead on core bread and butter issues, such as interest rates, mortgages and inflation. Iceland's position in a less secure world was not pushed as a defining issue of the campaign.

That was a considered choice, and its contours were visible in the Prime Minister's early speeches. Since taking office she had argued that the question should be settled through an argument conducted in good temper about how Icelanders actually live, not on fear of what was happening outside Iceland. She had been criticised before the election that brought her to power for not pressing the European question, on the grounds that it was not a priority, and she entered this campaign perceived in some quarters as a reluctant traveller and not committing to it fully until the final stretch. Eiríkur Bergmann, of Bifröst University, put the failure more broadly to the New York Times: the Yes side made a measured, tactical argument about what talks might bring, without saying why Iceland would want to join at all. "What was the higher goal?" he asked. "That was never explained."

Miscalculations

The narrow framing was understandable, and, with the unfair benefit of hindsight, misguided. Loss of control mobilises voters. Procedure, not so much. The claim that a country is handing authority to a body it cannot hold to account has decided EU referendums before: Denmark on Maastricht in 1992, Ireland on Nice in 2001 and Lisbon in 2008, France and the Netherlands on the constitutional treaty in 2005, the Netherlands again on the Ukraine agreement in 2016, the United Kingdom that same year.

The second miscalculation concerned organisation. By the time the campaign opened, the institutional ground had already shifted. The Federation of Icelandic Industries, once the most pro-membership voice in the country, was campaigning against. The Confederation of Icelandic Enterprise, led by a former Independence Party leader, was hard against. The unions were now divided, several leaders having crossed over to the No side and the rest sitting mostly on the fence. Behind the No campaign stood organised interests with money, organisational capacity, and a plan. The Yes side improvised more. On election night the director of the Yes campaign said she had arrived on her first day with no job description, and that she was proud of how organically the work had developed. As the campaign heated up the Yes side lost control of the framing, and the polls began to slide.

The turn arguably came when the public debate was taken over by whether Iceland could secure an exemption from the Common Fisheries Policy. The answers from Brussels were inconclusive, and probably could not have been anything else. The Commission said it was ready to explore creative solutions, and had signalled earlier that tailored arrangements were possible; it did not come across as that permanent exemption was available. Into that space the No campaign hammered home that membership meant handing over control of the resource, and received support from outside the court. During a parliamentary session, Dutch PM Rob Jetten stated that all new member states would have to accept the CFP. Irish MEP Ciaran Mullooly cautioned that Irish fishermen would not accept an exception for Iceland. Both instances were covered in detail in Morgunblaðið.  From midsummer the vote was no longer about finding out what was on offer. It was increasingly about what Iceland would have to give up.

The argument for flexibility was being made in public at exactly that moment. On 22 June, Valérie Hayer, the French MEP who leads the Renew Europe group, wrote in the Guardian that the Union should be moving heaven and earth to bring Iceland in, and that an enlargement model demanding full alignment across every policy area from day one no longer serves it. In so doing, she was writing into a long running debate over whether the accession model itself needs rebuilding.

In these debets, Iceland's experience is a case to study, though extrapolation from it must be measured against how unusual the case is. A wealthy Nordic democracy with strong institutions and a settled human rights record, inside Schengen and the single market, with much of the Union's law long since on its books. To states that have queued for a decade and longer, expecting to be courted on those grounds looks like privilege, and it probably was. But Icelanders had spent three decades at the bar with the other wealthy EFTA countries, watching the EU states dance, and would probably not have minded being flirted with a good deal harder.

Where the campaign was won

Crucial to the No side's strategy was influencing and mobilising the youth vote, and it worked. Gallup had 18 to 29 year olds backing negotiations by 63% to 37% in June, the most pro-European age group in the country. In its final poll, two days before the vote, they opposed by 60% to 40%. The national margin moved seven points over those ten weeks. Among the young it moved forty-six. Turnout reached 82.6%, so this was not a matter of any one group staying at home.

Young voters are the least anchored group in any electorate, and the most movable for it. Both camps told ODIHR's needs assessment mission in July that they expected to campaign primarily on social media. There the No side had the stronger operation and the more emotive material. The same advantages have belonged to Eurosceptic and right-wing movements on the continent for much of the last years.

The No campaign also had the deeper bench: retired politicians with national standing, union leaders from the hard left, and business leaders they normally meet across a negotiating table, each speaking to a different constituency. While the older age groups moved less than the young, the bench offered familiar and experienced figures through whom they were used to having things explained, and by whom they were used to being swayed. In broad terms their case was that the EEA already does the job, that what membership might add is a maybe and what it costs is control of resources. Their closing argument was that all the effort and attention would be better spent on Iceland's economic situation and the reforms waiting at home.

The No campaign was better funded, too. The advertising showed it, and so did the space it occupied online. Nothing, however, requires either campaign organisation to say where its money came from. Iceland's political finance law covers parties and candidates; the organisations that ran this referendum fall outside it, and the National Audit Office, which audits the parties, has neither a mandate over them nor the power to impose a sanction of its own. Those the ODIHR mission met in July called this one of the principal shortcomings in the framework. 

Each camp spent the campaign accusing the other of outspending it. Opponents of Áfram Ísland put its budget at up to ISK 300 million; Áfram Ísland made the same charge about the money behind the Yes campaign. Neither claim was established, and nothing obliged either camp to publish anything. On a register of about 270,000 voters, ISK 200 million works out at about 730 krónur a head, roughly five euros. After the vote, the campaigns gave lower and almost equal figures, the No side about ISK 90 million and Yes about ISK 80 million. A spokesman for the No campaign declined to name its donors. It was a matter of confidence with the people who had supported the campaign, he said, and many of them did not want to be named.

The Yes campaign would not name its donors either. Gréta María Grétarsdóttir of the Yes campaign gave the split, 46 per cent of donations from individuals and 54 per cent from companies, but no names. When they were raising the money, she said, they told donors the names would not be published, and they would stand by that.

While it is clear that lessons need to be learned about transparency in campaign financing, it is harder to say whether the same is true of disinformation and outside influence. Iceland has institutions for this. The Ministry for Foreign Affairs co-ordinates work on hybrid threats and foreign information manipulation, CERT-IS was moved into the ministry's statutory security framework last year and watched for new referendum-related websites, and Netvís, the media literacy centre run by the Media Commission, opened a public reporting portal on 21 July. We could find no public communication from any of them while the campaign ran, and if there was any, the signal was weak.

That said, the anecdotal impression is that misinformation stayed low by continental standards, though higher than in previous Icelandic elections. Familiar imports circulated, the EU army and the billions to Brussels, but the online campaigning appears to have been locally run. With no public assessments, nothing concrete can be claimed. What can be argued for is the case for timely public reporting that rests on data and documented cases, so that this fast-emerging concern is argued on firmer ground than it has up until now.

The immediate fallout

The institutional reaction in Brussels was thin. The Commission's response came as a spokesperson's text message to selected correspondents on a Sunday: it takes note of the outcome, respects the choice of the Icelandic people, and regards Iceland as a close partner. Ursula von der Leyen said nothing under her own name. António Costa exchanged messages with the Prime Minister during and after the count, said the Union fully respects the democratic choice, and called Iceland one of its closest and most trusted partners, including through the EEA. One EU official, quoted anonymously in Politico, was less careful: it was like being turned down by someone you had not yet asked out.

Reykjavík was equally careful. Foreign minister Þorgerður Katrín Gunnarsdóttir took the question to Alþingi's foreign affairs committee before calling her counterparts in Brussels, and the chair, Pawel Bartoszek, said what the meeting was not: no decision on the application would be taken there, only consultation before the minister reported the result to the Union. Opposition leaders have already announced a parliamentary resolution to withdraw the 2009 application and end Iceland's standing as an applicant. The government can be expected to successfully resist it: the Foreign Minister has said withdrawal could make any future membership process longer and more difficult, and that she is ruling nothing in or out. Prime Minister Kristrún Frostadóttir has said only that the government will work out with parliament "how best to close this chapter", that it will take no firm decision yet, and she has not said when the matter reaches the floor. Whatever is decided will go through Alþingi, which is not how the file was closed in 2015.

Speaking on the state broadcaster on election night, before the results were in, Kristrún Frostadóttir did not rule out that a fundamental shift in the external environment could move Icelandic opinion, while making clear it would not be for her to say: the will to open the question again would have to come from below. Otherwise her line has held: a no puts membership off the table at least for the term of this government, which runs to 2028, and a referendum is not something a country calls every other year. She expects Europe to feature in the next general election even so. What such a shift would be is hard to see now, short of a major geopolitical event in or around Iceland, another economic collapse or a Norwegian decision to restart accession, which this referendum has made less likely rather than more.

Back in the EEA

With EU membership off the table for the foreseeable future, the status quo is EEA membership outside the customs union, the Common Agricultural Policy and the Common Fisheries Policy, with the exposure that comes with it: EU trade defence measures, a currency risk premium and thin domestic capital markets.

It is a settlement with broad support. Both camps campaigned on the importance of the EEA. The Eurosceptic People's Party has EEA sceptics in its ranks, but every party in Alþingi affirms the agreement, and there is no Icelandic constituency for leaving the single market. Guðrún Hafsteinsdóttir, leader of the conservative Independence Party, put the winning side's version: Iceland wants to stay in close relations with the European Union and with Europe, and the message of the vote is that Icelanders intend to steer that course themselves.

Which puts weight on an agreement Iceland has not always served well. The backlog of EU acts awaiting incorporation remains high, some files long outstanding, and the EU side has been pressing the EFTA states on timely transposition and on homogeneity in the internal market. Jacob Funk Kierkegaard of Bruegel puts the practical consequence this way: "One can well imagine that the Icelandic authorities now say: we are stuck in the EEA agreement and we have to make it work."

Doing that is mostly a question of staffing and political will. Iceland does not have a vote when EU law is adopted. What it has is the phase before adoption: the expert groups, the committees, the early consultations where drafts take shape. Influence there depends on how many competent Icelandic officials are in the room, or more often in the corridor outside it, and that depends on what a small mission in Brussels, and the ministries behind it, are funded to do.

Two live files may show how well Iceland defends its corner now that the Union has less reason to be accommodating. The Commission's safeguard measures on ferro-alloy imports, in force to November 2028, cover third countries including Iceland and Norway, and Icelandic producers and the authorities behind them now make their case with no accession prospect behind them. 

The second is the emissions trading system. Iceland's exemption for international aviation, which keeps transatlantic traffic through Keflavík on different terms from traffic inside the EEA, was due to expire this year; in July the Commission proposed extending it to 2030. Reykjavík had asked for a permanent arrangement built on geography, and for shipping to be covered too. It got neither. Asked at the time whether the referendum had influenced the decision, the Commission's climate chief, Wopke Hoekstra, called the extension "an investment in our cooperation and in our shared future. What that looks like is up to Iceland." Without the exemption the economics of the Keflavík hub change fundamentally, with consequences for how easily Icelanders travel abroad, for the cost of air-freighting high-value exports, and most drastically for tourism, on which the country has grown steadily more dependent.

What Brussels lost

For Brussels, Iceland was a tantalising prospect: a wealthy Nordic democracy with strong institutions, a long-standing member of the single market and Schengen, the first net contributor to join in more than thirty years, and on geopolitically prized ground.

An Icelandic Yes would also have served an argument Ursula von der Leyen is making for other reasons. At Munich in February she said Europe must become more independent, and that Article 42(7), "one for all and all for one," has to be brought to life, a commitment that "only carries weight if it is built on trust and capability." In the same speech she listed Iceland with the United Kingdom, Norway and Canada as close partners the Union should work with. Membership was a separate question, but a country already that integrated choosing membership as well would have supported her case that the Union is now worth joining for reasons of security and not only of trade.

The Union has not grown since Croatia joined in 2013, and with Brexit it has shrunk. Iceland was to be the easy case, and the wealthy partner to pair with Montenegro, and possibly Albania, at the leaders' summit in mid-October. Montenegro now carries it alone, with about half of its 33 chapters closed and a 2028 target date. The instruments being prepared for that summit, anti-backsliding clauses and time-limited vetoes for new members, were designed to reassure sceptical capitals in the west, though they are also the kind of conditionality Iceland would have resisted.

Nothing in the arrangements between Iceland and the Union changes, so what was lost is essentially a story: the Union's presumed attractiveness. Steven Blockmans of the International Centre for Defence and Security told the Wall Street Journal that the result reinforces a view that membership appeals only to poorer and conflict-ridden countries, not to affluent European democracies. James Batchik of the Atlantic Council put the counterfactual to the same paper: a yes "would have shown that countries, particularly rich countries in the north, are buying what the EU is selling as a source for stability."

The story the Union lost is one its opponents happily claimed. Marine Le Pen said within hours that Icelanders had "recalled an obvious truth": "membership in the European Union is not the natural horizon for every European democracy," and that "another Europe is possible, a Europe of free, sovereign nations cooperating voluntarily on matters of common interest." Nigel Farage congratulated Iceland on "voting to keep their independence." Neither reading really matches the campaign that produced the result. The No side argued for the EEA and Schengen rather than against them, and no party in Alþingi proposes leaving the single market. The claim abroad will nonetheless outrun the Icelandic evidence.

Set against all that, the assessment from inside the Brussels policy world is cooler. Heather Grabbe of Bruegel put it this way: "This is not a big deal, because the door is still open and Iceland remains a state that could apply and could carry on from where it left off. And there are of course many other candidate states forming a long queue. This would have been a rose in the buttonhole at this point, but it is mainly just a disappointment."

The security pitch

On paper, Iceland was the place where von der Leyen's security argument should have worked: wealthy, exposed, Arctic, and integrated. It did not. Iceland's security rests on the 1951 defence agreement with the United States, on NATO membership, and on partnerships such as the Joint Expeditionary Force and NORDEFCO, and what Article 42(7) would add to that was never made obvious. Kristinn Árni Hróbjartsson of Varða told the Wall Street Journal that part of the reason Icelanders rejected the talks is likely that they feel well protected as things stand.

So it was not the argument that decided the vote, and it was not really an argument that was had. Hulda Þórisdóttir of the University of Iceland told Euronews that security "was not a decisive issue during the campaign, as it was much less present than many anticipated," and that defence was top of mind for a political elite worried about international developments while most voters were concerned with day-to-day economics. Valur Ingimundarson of the same university put it from the other end: the government moved the referendum forward because of geopolitics, and the majority did not treat geopolitical uncertainty as a reason to change the status quo.

Chris Kremidas-Courtney of the European Policy Centre called the result "a cold splash of reality" on where the Union stands as a security provider. Membership offered no unique territorial guarantee, so the vote turned on the economic bargain, and "Icelanders still see the EU primarily as a market to join, or not." Jacob Funk Kierkegaard of Bruegel reached the same conclusion: the debate about defence, Trump and the Union as a counterweight "clearly gained no traction with the Icelandic public," which he calls a problem for a Union that wants to be seen as a security actor in the Arctic, "also as regards other nations there, such as Norway."

If the argument failed in Reykjavík, its purchase in Oslo is also doubtful. Some officials had hoped an Icelandic yes would prompt Norway to look again; that conversation seems to be over almost before it began. Espen Emil Thygesen of Agenda reasoned why that was always unlikely. Norway is already integrated in EU defence cooperation, and the average voter reckons membership would cost more than the security it adds. "If the Norwegian government did not believe the NATO security guarantee was credible any longer, it would have already started discussions on EU membership."

Washington

Washington said little, and what it said was fairly routine. The State Department took note of the referendum, said the United States respects "the sovereign decision of the people of Iceland," and expressed "confidence in Iceland's continuing trajectory as a strong and valuable NATO ally and close partner of the United States and Europe." Washington had little reason to regret the result. A Yes would have complicated the bilateral channel through years of accession negotiations, and eventual accession would have planted the EU's flag in the middle of the GIUK gap, giving Brussels greater standing in North Atlantic and Arctic questions. Saturday's No leaves things just as they were.

American coverage of the referendum led on Trump. The vote was framed as held under the shadow of his designs on Greenland (CNN, CNBC, Fox News), and many in Washington will read the result the same way, as a verdict on his pressure. That however confuses the timetable with the campaign. His threats were part of why the government brought the vote forward from 2027, but in the campaign itself he was mostly absent, and his name was rarely raised in the leaders' debates.

Icelandic apprehension about the United States is neither new nor confined to Trump. The American footprint here was wound down long before he took office. In 2003 Iceland supported the Iraq war and joined the coalition of the willing, partly in the hope of keeping American forces at Keflavík. The Air Force left in 2006 anyway, and abruptly as it was seen from Reykjavík, ending a presence that had run since 1951. It was received here as a defeat, and a personal one for the then prime minister who had insisted on his strong personal rapport with President Bush.

Two years later came the economic crash. The Central Bank asked the Federal Reserve for a dollar swap line in June 2008. In September the Fed opened lines to Sweden, Norway and Denmark and turned Iceland down; the memo explaining the decision arrived on 3 October, three days before the prime minister's "God Bless Iceland" address. When Iceland went to the IMF in November, the programme rested on Nordic financing, with Poland and the Faroe Islands lending directly. No assistance came from Washington. Back the war, lose the base, be turned away at the Fed, sit out the rescue.

The relationship has recovered to some extent. American funding for the Keflavík facilities resumed in the middle of the last decade and rotational anti-submarine deployments followed, with the pace increasing after Russia’s full scale invasion of Ukraine. Allied submarine surveillance is flown from Iceland, US Navy aircraft were here every day of last year, and the security area at Keflavík is being rebuilt with funding from NATO, American military authorities and the Icelandic state.

The temptation in Washington could be to read Saturday's result as a vote of confidence in the United States, and as evidence that the tariffs and the pressure over Greenland cost nothing here. That would be off the mark. American conduct was not on the ballot; the campaign turned on domestic issues, and a vote against Brussels was not a vote for anyone else. Iceland's westward orientation is set by geography, the North Atlantic Treaty and the 1951 agreement. What may vary is the discretionary margin, for example host-nation support and how far Reykjavík is willing to go in backing American positions. In the past, that margin has usually tracked how Icelanders judge American reliability.

One question settled, more ahead

Icelanders were asked whether to reopen negotiations, and that is really all they decided, whatever else the result is now said to possibly prove. Membership is closed for this electoral term, and probably well beyond it.

There will be calls for the government to resign, and their effect will mostly be performative. Nothing in the result obliges the coalition to go, and nothing suggests it will over this. But it spent political capital it could not easily spare, and enters the autumn session weakened, facing an opposition that will hold the result up for as long as its useful as evidence of the government’s poor judgement.The Independence Party, polling at historic lows a year ago, has recovered and now matches or leads the Prime Minister's Social Democratic Alliance, much of that ground taken back from the more nationalist Centre Party. In a Eurosceptic moment the established right gained at the expense of the nationalistic right, which is not the pattern seen elsewhere in Europe.

Two issues are set to test the new parliament. The first is Iceland's standing as an EU candidate: the opposition wants the 2009 application formally withdrawn, the government prefers to leave it dormant. The second is Protocol 35, which gives incorporated EEA rules precedence over conflicting Icelandic law. The Foreign Minister will table it this autumn and the hardest Eurosceptics in Alþingi will treat it as accession by other means, and as an opportunity to extend the glow of the win.

Behind both sits the larger institutional and mental adjustment. The EEA now has to be defended as an end in itself rather than managed as a waiting room. In practice that means influencing EU law while it is still being drafted, which Iceland can do only with officials in the room and on the corridors, and only as many as the mission and the ministries are funded to send.

The rest of the agenda is unglamorously domestic. Half the country had looked abroad for access to a remedy, and that route is closed for now. The interest burden on households. A currency that carries a risk premium and is not going anywhere. The competitiveness of Icelandic firms and the lack of foreign direct investment. An agricultural and rural policy that will be reformed at home or not at all. Food security, energy security, and the defence of the country in a changed world. The list is long.

There is a gain from the summer though, and it is not a small one. After a lively election campaign, the electorate knows much more than it did in May about how the EEA works, about how European rules become Icelandic law, and about where Iceland sits in a changing order. Whether that knowledge and interest translate into added capacity to legislate and reform is an open question.

The last burst of reform and institution-building in Iceland followed the 2008 crash. The IMF programme, the first accession bid, and parliament’s truth report into the crash, produced a decades’ worth of reforms in the space of a few years. This time there is no shock to force the work, and no outside institution with the standing to insist on it. Whether Icelandic politics can generate that impetus on its own, after a decade in which little of comparable ambition was attempted or thought needed, is the test the referendum has set. And so Áfram Ísland (Forward Iceland), the slogan the winning campaign ran under, becomes something else: a domestic obligation.

Reply with any questions or comments, and please forward to anyone interested.