The 2014 Wales Pledge Explained, Where NATO's Two Percent Came From

The most famous number in the North Atlantic alliance is not in the North Atlantic Treaty. That surprises people, including many who have shouted it at a television. The two percent figure is quoted as if it were a legal requirement and a membership fee and a moral test, all at once. It is none of those things, and the real story is more interesting than the slogan.

Let us define it, read the text, and then argue about what it means. Precision matters here, because a debate about alliance burden-sharing conducted on bad definitions produces bad policy on both sides.

Defining the terms

Three terms need pinning down before anything else.

The two percent guideline. It is a benchmark under which a member country spends at least two percent of its gross domestic product on defense. It is a political commitment among governments, not a clause of the treaty.

The Wales pledge. At the NATO summit held in Wales on September 4 and 5, 2014, the heads of state and government endorsed a Defence Investment Pledge. That is the formal origin of the commitment as most people know it, though the two percent benchmark itself was an older informal yardstick.

The treaty obligation. The legal duties are in the 1949 treaty itself: Article 3 on self-help, Article 5 on collective defense. Those are legally binding. The summit pledge is not a treaty and was never submitted to any legislature for ratification.

Keep those three apart and most of the confusion disappears.

What Wales actually said

The pledge came after Russia annexed Crimea in March 2014 and while it supported a war in eastern Ukraine. The allies were looking at a decade in which many had cut defense budgets after the global financial crisis, and at a security environment that had gotten worse.

The text, which is worth reading in full at NATO's website, makes three commitments, paraphrased here. First, allies whose defense spending was already at or above two percent of GDP would aim to stay there. Second, allies below that level would aim to halt any decline in defense spending, to raise it in real terms as their economies grew, and to move toward two percent within a decade. Third, allies would aim to spend at least twenty percent of their defense budgets on major new equipment, including the research and development tied to it.

Notice the verbs. "Aim to." "Move toward." Nothing in the pledge says "shall pay" and nothing in it sets a penalty. It was a statement of intent among governments, made in public so that the public could watch whether it was kept. That is the actual enforcement mechanism: embarrassment. It works about as well as embarrassment usually works, which is to say unevenly.

The twenty percent that nobody remembers

The second number deserves more attention than it gets. The twenty percent equipment goal was meant to answer a fair objection to the two percent figure: that a government could hit the target by spending nearly all of it on salaries and pensions while buying almost no modern gear.

It is a smart supplement. A country that spends two percent of its economy on a large force with old equipment has a different capability from one that spends the same share on a smaller force with new equipment. The equipment target forced a question about what the money buys.

Yet the headlines almost never report it. A reporter can fit "two percent" in a headline. "Two percent, of which at least a fifth on major equipment and research" does not fit in a headline. That is the reason the simple number has traveled so far and the sensible companion has stayed home.

The ten-year horizon

The pledge's timeline ran to 2024. A decade was written into the text, and when the deadline approached, the question became how many allies had in fact gotten there. The answer, as a matter of the alliance's own published reporting, was that the number of allies at or above two percent grew substantially over the decade, from a small handful in 2014 to a majority by 2024. I would not quote the exact counts from memory, and you should check NATO's annual defense expenditure report for them, which is public.

The direction of change is not in doubt, and it is worth noticing why it changed. The Wales pledge did not do it alone. The 2014 annexation made the threat visible, and the 2022 invasion of Ukraine made it undeniable. Pledges tend to be kept when an adversary reminds the signatories why they were made.

Later summits raised the ambition. Allies agreed in 2023, at Vilnius, to treat two percent as a minimum rather than a ceiling. At the June 2025 summit at The Hague, allies set a new goal of five percent of GDP by 2035, composed of a core defense share and a smaller share for related security spending. Those later commitments belong to a different story. The point here is that Wales is the hinge, and it is the one most people misquote.

How NATO counts defense spending

Next, the arithmetic, because the figure is a ratio, and ratios have a numerator and a denominator.

The numerator is what NATO calls defense expenditure: spending by a government to meet the needs of its armed forces, those of its allies, or of the alliance. NATO's definition includes personnel, including pensions in many cases, equipment, infrastructure, operations, and research and development. It also counts some items that surprise people: military pensions, and expenditure on certain forces that are paramilitary but can be deployed with the armed forces. It generally excludes civil defense and some things governments might wish to count.

The denominator is gross domestic product, the market value of everything an economy produces in a year.

A worked example, with invented numbers and labelled as such. Picture a country with a GDP of 500 billion in its currency. Two percent of that is 10 billion. If its defense ministry spends 10 billion on the NATO definition, it is at the target. If its economy grows by ten percent while defense spending stays flat at 10 billion, the ratio falls to about 1.8 percent, with no cut at all. That is the first oddity of the metric: a country can fall short because its economy grew.

Why GDP is a crude yardstick

That example points to the central weakness of the measure, and honest defenders of the pledge admit it. A share of GDP tells you what a country spends relative to what it earns. It does not tell you what a country can do.

Five problems follow.

One. It rewards a poor country for spending a lot on a little. A small economy hitting two percent buys a small amount of capability. A large economy at the same ratio buys a great deal.

Two. It ignores the quality of the spending. Money on gold-plated programs that never field and money on ammunition and readiness count the same. The twenty percent equipment goal partly addresses this, and only partly.

Three. It penalizes a growing economy and flatters a shrinking one. In our example the ratio falls as the economy grows. In a recession the ratio can rise while the force shrinks.

Four. It ignores geography and need. A country on the front line has different requirements from one with a buffer of allies in between. The measure treats Poland and Portugal the same.

Five. It ignores what a country contributes in other ways. Troops deployed on missions, basing for allied forces, and industrial capacity do not show up in the ratio.

That is the fair criticism of the metric. The fair defense of the metric is that it has the virtue of simplicity. A number that every voter can understand is a number that can be held up in a parliament and a press conference. A more sophisticated measure would be more accurate and much easier to ignore. The two percent figure's crudeness is also the secret of its political force.

The pledge versus the treaty

Now the point where the loudest arguments go wrong. The two percent target is not the treaty, and failing it is not a violation of Article 5.

The treaty was signed on April 4, 1949, in Washington. Its obligations are short and plain. Article 3 says the parties, "by means of continuous and effective self-help and mutual aid, will maintain and develop their individual and collective capacity to resist armed attack." That is a legal duty to maintain defensive capacity, but it sets no number. Article 5 says that an armed attack against one is an attack against all, and each party will assist by taking "such action as it deems necessary, including the use of armed force." That, too, is a duty with a discretion built in. The allies decide for themselves what action is necessary.

So the two percent figure gives content to Article 3's self-help requirement without being part of it. A country can be in compliance with the treaty while missing the target, and the treaty does not say that an ally who fails the target loses the protection of Article 5. Anyone who claims otherwise is describing a deal that nobody signed.

That cuts in two directions, and both are worth saying.

To the alliance's critics who say America should defend only members who pay: that is a policy position you can argue for, and the burden-sharing complaint is a real one. But it is a proposal to change the bargain, not a description of the current one. The full text of the treaty is covered in Read the North Atlantic Treaty, All Fourteen Articles Take Ten Minutes.

To the alliance's defenders who treat the guideline as a trivial technicality: the treaty's self-help article is meaningless if members can ignore it for decades. The guideline exists because the legal duty is vague. Treating the pledge as optional invites the free-rider problem that the pledge was meant to address.

A second worked example, with the dials turned

Two invented countries, labelled as such, make the problem concrete. Call them Northland and Southland. Each has a GDP of 1,000 billion in its own currency, so two percent is 20 billion for each. Both report exactly 20 billion in defense expenditure on the NATO definition, and on the chart they sit side by side, two bars of the same height, two allies in good standing.

Now open the books. Northland spends 14 billion on personnel and pensions and 6 billion on equipment and related research. Six is thirty percent of 20, so Northland clears the equipment goal of twenty percent. Southland spends 18.5 billion on personnel and pensions and 1.5 billion on equipment. One and a half is 7.5 percent of 20, so Southland misses the equipment goal by a mile. Same headline number, very different militaries.

That is why the second figure in the pledge matters. A summit communique that says "two percent" and stops is like a restaurant review that says "the bill was reasonable" and does not mention whether the kitchen was open. The equipment share tells you whether there is anything on the plate.

It is also why a reader should be a little skeptical of both of the loud camps. The camp that says "they all hit two percent, so the problem is solved" has not asked what the money bought. The camp that says "they all miss two percent, so scrap the alliance" has not asked what the treaty requires. Each is dealing in the number as a slogan, and the number was never meant to carry that weight.

The Founding and the free rider

There is an old principle at work, and the Founders would have recognized it. In a union of states, the temptation for each member is to let others carry the common load. The Articles of Confederation failed in part because Congress could request money from the states but could not collect it. The Constitution fixed the problem by giving the national government its own power to tax, in Article I, Section 8, Clause 1.

The alliance has no power to tax. It has a pledge, a table of published numbers and the pressure of fellow members. That is the Articles of Confederation model, and it explains both why the pledge is weak and why moral pressure is the only tool available. If you want to understand why burden-sharing is a permanent argument in the alliance, you need not look further than the structure of the arrangement itself.

What it means for a family

Why should a family in Ohio care about a figure agreed in Newport? Because defense spending by allies affects how many American service members and how many American tax dollars are needed to deter the same threats. If wealthy allies underinvest, the gap is filled by American forces, American stockpiles and American money. If they invest in their own readiness, the burden falls where it should.

For related reading on how America counts its own spending, see Defense Spending as a Share of the Economy, From Korea to the 2010s, and on the debate over whether the alliance should back Ukraine, Should Conservatives Back Ukraine? Two Honest Answers, Weighed Fairly. The Weinberger tests for committing troops, laid out in The Weinberger Doctrine Set Six Tests Before Sending Troops to Fight, bear on the same question from a different angle.

The case in brief

1. The two percent of GDP target is a political guideline endorsed at the Wales summit of September 2014, not a clause in the 1949 treaty.
2. Wales also set a goal of spending a fifth of defense budgets on major equipment, and a ten-year horizon to 2024. The equipment goal is the half of the pledge that people forget.
3. NATO counts defense expenditure by its own definition, and GDP is a crude yardstick: it can reward the wrong things, penalize growth and ignore geography and quality.
4. The treaty's real obligations are Article 3, self-help, and Article 5, assistance by "such action as it deems necessary," and neither sets a number.
5. The pledge matters because the alliance has no power to tax, so peer pressure and published figures are its only enforcement.

What you can do: open NATO's most recent public defense expenditure report, find the figures for three allies and for the United States, and check each against both parts of the pledge, the two percent share and the twenty percent for equipment.