Garvan Walshe is a former national and internationals security policy adviser to the Conservative Party.

The rewriters of history have got to work in the weeks since the American rout in Afghanistan. The mission failed because Afghanistan was a graveyard of empires. Or because humanitarian intervention never works, according to realist high priest Stephen Walt.

This is far from the truth, as Robert Kagan explains at length. We went into Afghanistan because of 9/11, not for purely humanitarian reasons. Even in more overtly humanitarian interventions, like NATOs in Kosovo or the removal of Gaddafi in 2011, strategic considerations mattered.

Success in Kosovo owed much to Milosevic, always an opportunist with a healthy sense of his own self-preservation, agreeing to give in after NATOs bombing campaign. The alternative, an invasion of Serbia and Montenegro through Albania and Hungary, while Serbs in Bosnia and Croatia rose up against the Dayton peace settlement would probably have led in Serbia to a destructive guerilla war not unlike that in Iraq or Afghanistan. The last minute paradrop into Pristina airport should remind us that even late Yeltsin Russia would not have been helpful.

Nonintervention is not the easy option frozen-blood realists like Walt would like it to be. They would have stood by as Kosovars were raped and murdered, or the inhabitants of Benghazi driven into the sea. They got what they wanted in Syria: hundreds of thousands of dead and millions of people displaced. Whether this sits more easily with their conscience or ideas of a globally influential West, I leave to readers to judge.

It is harder still where there is a strategic objective. How comfortable would we have been, for example, relying on Polish tanks to dash through Eastern Ukraine to protect Lviv? (Hint: it matters that the Poles call the city Lwów).

Or consider the French intervention in Mali, against central African jihadists. Would you be the President pulling troops out only to find your citizens attacked by a plot originating there? What of an allied government, with which you had important commercial or strategic ties, being toppled by hostile rebels?

Commercial ventures, military operations, people, religion and ideas now flow across borders more easily than they could in the past. Or rather, they have started to flow back the other way. Western business, armies, people and ideas have after all been flowing out of Europe to the Americas, Asia and Africa for five hundred years. Along with supply chains and investment flows, conflicts have become globalised.

It’s not viable any more for a single country to retreat, and there is no disputing the principle that if we’re to provide security at home, we need to get involved abroad. Disagreement is only over the manner of involvement.

In the last twenty years, the difficulty has never been to remove a hostile force from power, either through direct intervention, or Western air power supported by allies. Problems have set in afterwards, even though the need for long term post-conflict stabilisation is very much a “known known” (however much Rusmfeld himself was in denial about it). The question is why we’ve made the same mistakes over and over again.

Afghanistan suffered from intermittent attention and dispersed accountability. It only drew high level political focus at the beginning and when problems mounted. Different administrations tried varying strategies, with greater or lesser emphasis on state building, smaller footprints, or a “surge” of troops.

Meanwhile the mission was split, between the mission to capture bin laden and that to stabilise the country. The former a unilateral American operation. The latter a multilateral NATO one. Similar problems bedevilled the postwar reconstruction in Libya (with France and Italy backing rival governments), or Iraq, with the US reducing troops only to find it had to increase them to fight ISIS.

Without attention, disorder was allowed to fester, more civilians and troops got killed, and governments were unable to justify the intervention to their publics. Politicians picked up the public dissatisfaction, and rushed to leave as soon as they could.

Direct political control works best when there’s a single locus of accountability and continuous attention on the problem. In these multilateral interventions there’s neither, so public attention wanders, and the pressure on the different components of the alliance causes friction. This should not have been a surprise. These problems affect all complex and long-term international cooperation, which needs a certain amount of structure if it is not to become a sequence of ad-hoc adaptations to circumstance.

Towards the end of the Cold War, the CSCE (later OSCE) was set up to supervise disarmament, and continues to engage in security and democracy related aspects of the European international architecture. On climate change, the “Conference of Parties” has evolved into an organisation with an indefinite timescale (we are now on number 26).

A specific, but permanent organisation has a number of advantages: consensus on strategy is achieved through multilateral diplomacy. Participants allocate budgets that are spent by the organisation as a whole. A permanent secretariat maintains focus even when political attention is lacking. Membership can be limited to countries that agree with the organisation’s aims (to avoid the fate of the UN Human Rights Council).

Perhaps it is time to consider some sort of international stabilisation and counter-terrorist organisation.

Those establishing one will face a number of difficult questions about how it should work, not least over how to get security forces and human rights organisations to tolerate each others’ involvement, and over who should be included. For example, what roles should hostile powers like China, or highly relevant friendly countries with terrible human rights records, have?

But the last twenty years of unstructured unilateralism have hardly been an unqualified success. It’s surely time to give more structured alternatives a go.