The head of the International Atomic Energy Agency, Mohamed El Baradei, has expressed concern that “nuclear weapons could fall into the hands of an extremist group in Pakistan or in Afghanistan” given recent events in Pakistan.

Pakistan has angrily rebuked the IAEA chief over these remarks and an Agency spokesperson has attempted to clarify matters by claiming that El Baradei was speaking about the need to enhance the security of nuclear weapons and fissile materials everywhere, not just in Pakistan. Two US analysts, Ashley Tellis and Lisa Curtis, have backed Pakistan claims in congressional testimony that its nuclear arsenal is safe and secure. Both, however, are associated with the Bush Administration. This does not necessarily mean that their stance on the security of Pakistan’s nuclear weapons is misguided but it does mean that we should not see their statements as necessarily being authoritative.

Whatever we may say about this latest diplomatic spat on Pakistan and the prospects for nuclear terrorism, it remains the case that El Baradei’s comments in the wake of the assassination of Benazir Bhutto serve to remind us of the stakes as Pakistan erupts into relative disorder. In essence, we are asking ourselves whether there exists the possibility of a catastrophic act of nuclear terrorism.

How real is that prospect?

Firstly, it should be pointed out that the confluence between failed states and nuclear danger is not a new one. The Soviet Union had both a vast arsenal of nuclear weapons, including small tactical nuclear weapons, and large quantities of fissile materials. During the collapse of the Soviet Union many were alarmed at the prospect of “loose nukes” getting into the hands of criminal groups or terrorists.

That it did not happen should not necessarily be viewed as a source of solace, but it does highlight the fact that there is no causal link between failed states and nuclear terrorism. In many alarmist reports since the assassination of Bhutto the existence of a link between the two has been presented as almost a law of nature.

It is quite clear that the first and most obvious danger would be if radical Islamist groups were to take power in Pakistan and thereby inherit Islamabad’s nuclear arsenal, which best estimates put at about 50-60 weapons. But it would be more likely that a newly installed Islamist regime would see its nuclear weapons capability as a deterrent against US or Indian efforts at regime change than as weapons of terror.

A distinct possibility would be an Islamist regime using nuclear weapons as a shield to support more conventional forms of terrorism. This is sometimes referred to as the “stability-instability paradox.” For instance, it is widely alleged that Pakistan has used its nuclear weapons as a shield to support insurgent groups in Indian-controlled Kashmir and even to send Pakistani troops across the Kashmiri line of control, safe in the knowledge that India would be prevented from escalating any conflict given the nuclear standoff between the two states.

Moreover, it is not obvious that a radical government would pass over either assembled nuclear warheads or fissile materials to independent terrorist groups, such as al-Qaeda. The fissile material for Pakistan’s nuclear weapons comes in the form of weapons grade uranium produced at the Khan Research Laboratories. It is estimated that Pakistan has produced 585-880 kg of weapons grade uranium.

However, the International Atomic Energy Agency has found traces of highly enriched uranium particles on Pakistani-origin nuclear technology exported by Pakistan to Iran. Checking up on the matter has increased our knowledge of Pakistani-enriched uranium. With a background database made possible by examples such as this it might well be possible through nuclear forensics for physicists to trace the origin of fissile material used in a nuclear terrorist attack.

If the terrorist explosion was relatively efficient then most certainly the weapon would have come from Pakistan’s military stockpile and nuclear forensics can be used to determine the efficiency of a nuclear weapon. Any Islamist regime would need to take this possibility into account before transferring either a fully functioning weapon or the necessary fissile material to a terrorist group. Even in the case of a radical Islamic state it is quite likely that the rationality criteria of nuclear deterrence theory is applicable.

One disturbing option that has beeen opened up by nuclear forensics and has been seriously considered in the White House is the promulgation of a “negligence doctrine” to deter nuclear terrorism. The idea here is that if a state were to lose control over fissile materials or nuclear weapons through “negligence,” and these materials were stolen and used in a nuclear explosive device by a terrorist group, then the United States would hold such a state “responsible” for the terrorist attack and strike back with nuclear weapons.

The possibility of such a nuclear strike, it’s argued, would deter “negligence.” But the concept makes “negligence” sound like a conscious choice made at the very highest policy levels, which it need not be. Sometimes at US nuclear weapons plants people have been caught sleeping on the job but surely the negligence doctrine would not apply if Bin Laden got his nuclear device because of a Homer Simpson.

In reality, a “negligence doctrine” would make an act of nuclear terrorism more likely. Jihadi groups like Al Qaeda are revolutionary - or, more accurately, counter-revolutionary - vanguards who see their main strategic task as mobilising a dissatisfied but apathetic population. In this sense they have been highly influenced by Lenin and the Bolsheviks. It is not hard to see how a “negligence doctrine,” rather than deterring nuclear terrorism, would actually encourage Jihadi groups to attempt to get their hands on the necessary fissile materials for a nuclear device because the prospect of a US nuclear counter-strike on such obviously immoral grounds would enrage, and hopefully radicalise, the entire Islamic world.

Could terrorist groups exploit the breakdown of law and order in Pakistan and actually steal a nuclear device? In so far as we are aware Pakistan does not store its nuclear weapons in fully assembled form. To get their hands on a nuclear explosive device a terrorist group would need to steal all the parts to assemble a weapon. That would be a very complex operation and would be a bridge too far. Indeed, even if they got one it would be a challenge to use the highly precise explosive lenses to focus a conventional shock wave and implode the fissile core. It is quite possible that the North Korean nuclear test may have been a fizzle for precisely this reason.

But they need not go to the trouble of getting a fully functioning device in this way. Pakistani nuclear weapons work by imploding a fissile core of metallic weapons grade uranium. To achieve this it takes about 15-20 kg of weapons grade uranium. If a terrorist group were able to steal more than three fissile cores it would theoretically be possible for them to develop their own crude nuclear explosive device.

This device would be a gun assembly device rather than an implosion device. It was a gun assembly device based on about 60kg of weapons grade uranium that destroyed Hiroshima. Firing a sub-critical mass bullet of weapons grade uranium into a slightly sub-critical metallic ring target will result in a nuclear explosion. It would be possible to do this with a mortar or a crude artillery piece; although the yield would not be the same as the Hiroshima bomb it would still pack a mighty punch.

A nuclear weapon based on this type of design is much easier to assemble and set off than an implosion device but would still be a tough call for a terrorist group, especially if the material came from a stockpiled weapon. Nuclear terrorism is therefore a theoretical possibility and the limiting factor involves assembly and delivery. The real issue in relation to Pakistan is the security of the fissile cores and the security of its stockpile of weapons-grade uranium.

The Khan Research Laboratories and the fissile cores are Islamabad’s “crown jewels” and thereby under very tight military security, but it might be possible to envisage security diminishing in the event that Pakistan becomes a failed state. Although it is reported that elite US Special Forces have assembled in Afghanistan ready to go into Pakistan and attempt to snatch the fissile cores if ordered to do so, this is risky without detailed actionable intelligence.

There is an inverse relationship between the safety of nuclear weapons and deterrence. The more threatened the regime in Islamabad feels the higher the salience it will place on nuclear deterrence; consequently, we will see a decrease in the safety of its arsenal. This makes most military based policy options rather unattractive.

Indeed, Pakistan has refused to place US Permissive Action Links on its weapons for fear that they may contain a “kill mechanism” preventing detonation. In this case deterrence trumped safety. Washington, after intense internal debate, has also refused to supply the mechanisms, fearing that Pakistan would gain too much insight into next-generation nuclear weapons design. Moreover, it is reported that Israel has recently tested the nuclear capable Jericho 3 ballistic missile, which is near the threshold between an intermediate and intercontinental range missile that would be able to reach Pakistan. This will undoubtedly increase the salience that Pakistan places on nuclear deterrence.

We now face a whole series of policy dilemmas in relation to Pakistan because of nuclear proliferation in South Asia. The big lesson here is that the proliferation of nuclear weapons should be the highest global security priority. But even here things are going backwards. The best scientific analysis demonstrates that the US-India nuclear trade and technology transfer agreement will enhance India’s capacity to produce fissile materials for nuclear weapons should it be implemented. Partly in response, Pakistan is expanding its capacity for producing weapons grade plutonium by starting construction on a new plutonium production reactor that would enable Pakistan to develop second generation nuclear weapons and to increase its nuclear arsenal.

So, where do we go from here? So long as there exists a theoretical possibility of nuclear terrorism, policy must be directed toward reducing the threat.

Clearly, the greatest danger would be if Pakistan were to become a failed state. Steps should be taken to help prevent that from occurring. What is not much appreciated by policy analysts is the influence that the Pakistani military has over the economy. This domination fosters many of the factors driving Pakistan toward becoming a failed state, such as endemic corruption and the existence of a kleptocratic elite.

Any process of democratisation that does not seriously tackle the military’s role in the economy will amount to nothing but a facade. Policy should be directed toward reforming the structure of Pakistan’s economy in order to prevent Pakistan from becoming a failed state. But this cannot be achieved so long as the military dominates politics.

Previous tentative attempts at democratisation have failed to buckle the economic power of the military. There exists little room now for further failure. The allure that we can offer liberal economic reformers is the prospect of integration into the regional and global economy. Global security would be well served by the liberal economic reform of the state-military sector in Pakistan.