Global Policy Forum

 In the immediate aftermath of the fall of Muammar Gaddafi's regime in Libya last summer, Stewart Patrick, writing in Foreign Affairs, made a bold prediction. The fall of Tripoli, opined the former US State Department official, was "the first unambiguous military enforcement of the Responsibility to Protect norm, Gaddafi's utter defeat seemingly putting new wind in the sails of humanitarian intervention".

Even as Patrick wrote, his argument was apparently bolstered by apresidential study directive on mass atrocities that offered a menu of potential policy options in the face of large-scale human rights abuses. It was a document, he averred, which was a significant triumph for Samantha Power, the author of A Problem From Hell and the "intervention hawk" credited with persuading President Obama to back the anti-Gaddafi forces militarily.

That was then. Now, on the first anniversary of the uprising against the regime and with Libya in increasing turmoil, the certainties of last summer look less compelling. As recent reports by human rights groups and journalists have made clear, the country has descended into rival fiefdoms of competing militias, not least in Misrata, which, as the Guardian argued on Friday, has set itself up as a "city state" with its own prisons and justice system. Human rights abuses are rife. Corruption is endemic. The new post-Gaddafi state, far from coalescing into meaningful institutions, is becoming ever more fractured.

As Ian Martin, the UN's envoy to Libya, argued late last month: "The former regime may have been toppled, but the harsh reality is that the Libyan people continue to have to live with its deep-rooted legacy; weak, at times absent, state institutions, coupled with the long absence of political parties and civil society organisations, which render the country's transition more difficult." And the lessons of what has happened in Libya cannot be seen in isolation. Rather, they add impetus to the question of when and how humanitarian military intervention should be employed at the time when calls for a new intervention in Syria are mounting. For the reality is that far from being an unambiguous success, Libya has proved once again the limitations of military intervention for regime-change in its various guises.

In Iraq, Afghanistan – in Kosovo to a lesser degree — and now Libya, what has been left after intervention has been a series of weak and corrupt fragile states, where violence is often commonplace and anything resembling real democracy utterly absent.

Part of the problem stems from an overarching naivety in the terms of the doctrine of intervention – in particular "Responsibility to Protect", pushed by the likes of Power – which has operated on the assumption that removing a bad regime must lead inevitably to a happier outcome. That view, in turn, has its roots in a confused understanding of how the concepts of legitimacy and the use of force interact in times of war and how the recent history of conflict can create the permissive climate for further violence.

For while few would deny that states using violence against their own populations delegitimise themselves, when that abuse is then deployed to argue for the use of force to remove regimes, it creates a complex dynamic that risks normalising conflict in the new political space, as has occurred in Iraq and Libya. Perhaps even more worrying has been the starkly visible trend towards ever-more hands-off engagement in the post-conflict reconstruction that has mirrored an apparent desire for intervention to be ever cheaper in terms of blood and treasure.

After the fall of the Taliban and Saddam Hussein's regime, efforts were made, even if they were ultimately botched and deeply flawed, to remould the political space backed by huge resources. In Libya, "intervention lite" has been followed by an even lighter reconstruction, now unravelling with increasingly disastrous results.

Indeed, the failure of recent high-profile interventions – to a greater and lesser degree – far from putting new wind in the sails of humanitarian intervention, as Patrick claimed, has served to dramatise its ambiguities and shortcomings. These failures have raised once again the vexed questions of what should be the threshold for intervention, of proportionality and how far the organising notion of sovereignty should be undermined in international law.

All of which has led to a fundamental paradox. While it is difficult to counter the core moral principle of humanitarian intervention articulated by the likes of Power and US legal academic Fernando Tesón – the latter has argued that because the major purpose of states and governments is to guarantee human rights, then governments that violate human rights should not be protected by international law – it is far more difficult to commend the real and practical outcomes.

I have visited enough human rights-abusing regimes to understand the force of the moral argument in favour of preventing such abuses, especially when they are conducted on a grand scale. But having covered the interventions – and the bloodshed that has followed in Libya, Iraq and Afghanistan – I find it hard to be as complacent as the pro-intervention lobby often appears to be, finding as much to criticise in the abuse after intervention as before.

Perversely, the greatest danger for those pushing most forcefully for intervention is that the dubious consequences of recent interventions may ultimately discourage states from intervening in clear-cut and egregious cases of widespread atrocity and genocide of the kind that inspired the anger of the likes of Power in the first instance – in Bosnia and Rwanda.

What to do then? The answer is that if the notion of humanitarian intervention is not to be utterly discredited, there has to be a rigorous, realistic and practical understanding of what is required – not simply to remove abusive regimes, but to guarantee genuine freedoms, democracy and transparency in the post-conflict period.

For that to occur requires that the doctrine be married with a far higher threshold for intervention and a more profound understanding of both the actors involved and the potential consequences. Because the grim alternative – still visibly present in the relationship with countries such as Bahrain and Saudi Arabia – is a return to a kind of Kissinger-style policy realism that turns a blind eye to abuse.

If intervention is to be a tool, it must be a tool of last resort, backed by the promise of serious post-conflict engagement, costly and time-consuming as it is, with an explicit understanding that "Responsibility to Protect" should not simply mean the prevention of widespread atrocities in the first place, but responsibility for the prevention of civil war in the conflict's aftermath and for reconstruction.

Otherwise, the doctrine invoked to end one horror will become known as the doctrine that gave birth to so many others.