So where do liberal interventionists go from here? The spectre of the possibility of failure and defeat in Afghanistan and Iraq haunts our every move and emboldens the isolationists. There are four dimensions to the challenge liberal interventionists face – the political; the physical and financial capability to act; the issue of mission definition; and finally, the legal framework in which interventions should take place.

The political context and prospect look bleak. Neither the Labour party nor the wider electorate favours an interventionist foreign policy that takes the responsibility to protect seriously. The place of the responsibility to protect in global political discourse echoes this domestic isolationism. The general assembly of the United Nations adopted a ‘summit outcome’ at the end of the world summit in 2005. This restated in clear terms the responsibility of all states in the UN with respect to protecting populations from genocide, war crimes, ethnic cleansing and crimes against humanity. The same document also restated the commitment of the United Nations to ‘support all efforts to uphold the sovereign equality of all states, respect their territorial integrity and political independence’.

For liberal interventions, in cases of genocide, the first commitment supersedes the second. For a brief period from the 1990s it seemed that the international community might also be moving towards this position. The lessons of Rwanda and the example of Kosovo sounded loudly in our ears. But after the conduct of the US response to 9/11 there has been a backlash and we are back in the position in which it is national sovereignty that seems to matter more than the responsibility to protect. The situation in Darfur is the clearest current example of this. The paradox is that the operations in both Afghanistan and Iraq are sanctioned fully by the democratically elected governments of those states. But we are some way beyond being able to win or even have a reasoned argument about the justification for our continued military role in Iraq or Afghanistan. For many, there are no facts that will alter the conviction that it is time for British troops to cut and run from both these struggles.

This urge to withdraw, coming as it does from inside the military as well as from inside the Labour party and across the electorate, is based in part on a growing perception of overstretch. There are 7,200 operational troops in Iraq and 5,000 in Afghanistan, 300 on UN missions and 8,500 in Northern Ireland. It is hard to see where a British contingent of the proposed UN peace-keeping force in the Sudan would come from. This is not just a problem for the UK military. The US armed forces are in a similar position. But the issue of capability is not just logistical. Overlaying the current malaise of liberal interventionists is the sense that these tasks are actually not doable, that these conflicts are not winnable.

To an extent, this problem arises because of the weakness of mission definition in many of these interventions. Mission definition is not only about the ‘why we are there’ question, it is also about the ‘when we get out’ question. In Sierra Leone and Kosovo, the mission was clear in terms of the intervention and the long-term follow-through needed. We were good at making the case for the intervention, much weaker on explaining the need for a long-term commitment to post-conflict reconstruction. There was no ambiguity in the difficulty of the task in Afghanistan and Iraq because these were part of the global war on terrorism and not part of the developing policy of liberal interventionism. The case for intervention in Iraq could have been made under paragraph 138 of the world summit but it was not. If the mission in both these cases had been consistent with the responsibility to protect and thereby endorsed by a much more broadly based coalition, then things might have turned out differently.

Nevertheless, we have to deal with the situation we are in and not the one we would have liked. The civil war within the Islamic world between fascism and moderation will continue to produce humanitarian disasters like the situation in Iraq. It will also provide the cover needed by governments intent on ethnic cleansing, like the government in the Sudan. Politically this will make constructing the case for liberal intervention even harder than it is now. That case can, however, be strengthened if we are clear what our objectives are when we embark on an intervention.

It must also be a requirement that these operations take place in a legal framework, supported by the United Nations and covered by appropriate resolutions. Operations in Iraq and Afghanistan are both now within the international legal system but were launched without specific resolution coverage. In Afghanistan existing resolutions were broadly endorsed as justifying an extension of the concept of a right to self-defence. There was no second resolution on Iraq and there should have been. We need the legal framework to be implemented but we also need it to be modernised to reflect the contradiction between sovereignty and the responsibility to protect. If states are using violence against their own people there should be a generally recognised threshold which, if crossed, forfeits their place in the international community.

Liberal interventionism is in crisis. It is unpopular politically, it has been undermined in practice and seems weakly defined and defended in theory. But what are the alternatives on offer? Some on the left believe that all western intervention makes things worse so we should do nothing at all. Others that only operations that fully accord with the UN commitment to national sovereignty should ever take place. Both positions are riddled with nostalgia for the cold war. The super-power confrontation gave us a cover for our indifference. We tolerated and worked with genocidal dictators like Saddam Hussein because we were playing the great game. We left large sections of the world to rot because we could not intervene in the Soviet sphere of influence. We allowed the UN to be manipulated into inertia because those were the rules of the game. Today, the anti-war left and beyond are trying to go back to a world in which we let genocidal regimes kill with impunity. A world in which national sovereignty trumps in all circumstances the responsibility to protect, in which prevention of mass murder is not the responsibility of those states with the means to stop it: indifference as a system of international relations.

We should not abandon liberal interventionism. It is a new system for the relations between states and we are in the very early days of developing it. Of course there have been terrible mistakes and the cost in human life of those mistakes should haunt anyone who advocates interventions in the future. Moreover, the nature of the interventions we favour need to become much more long term, linking the right to development with the responsibility to protect. So we can do it better and we should do it better. There is a legal case for intervention and under the UN charter we are obliged to fulfil our responsibility to protect. Victory in the global war on terror will only be achieved by policies that link the use of force with economic intervention to promote development. But more than all of these things, democratic socialists should retain a sense of our moral commitment to global citizenship. Universalism is deeply unfashionable at present but the moral imperative of our responsibility for each other as human beings is more urgently required than ever.