Here's how the British government describes the risk of a smallpox outbreak. "We are currently at alert level O. Smallpox remains eradicated. No credible threat of a smallpox release."

So, in response to this non-existent threat, it has published 122 pages of central plans. Each of the nine English regions maintains a Smallpox Diagnosis and Response Group, which in turn supports five Smallpox Management and Response Teams, one of which is on duty at all times. There are smallpox centres all over the country and lists of doctors, nurses and support staff prepared to run them, laboratories ready to multiply vaccines, and planning committees involving scores of different agencies.

The plans, in other words, must have cost millions. They use thousands of hours of specialist time every year. But step forward the man or woman who believes the government should abandon them.

The chances that this extinct disease might break out here are extremely remote - one in a million perhaps - but they cannot be dismissed while the US and Russia disgracefully refuse to destroy their stockpiles. Stealing, weaponising and distributing the virus would require capabilities beyond those of any known terrorist group. The government's plans are almost certainly a waste of time and money. But they are a waste of time and money that makes sense.

This is what government is for: to prepare for the worst, however unlikely it may be. The UK, like all rich nations, maintains an elaborate network of agencies to defend us from unlikely events: the ministerial sub-committee on protective security and resilience, the Civil Contingencies Secretariat, the domestic horizon scanning committee, the National Risk Register, the Research Capability Programme, the National Recovery Working Group, the Regional Resilience and Emergency Response Division, the Centre for Emergency Preparedness and Response, and endless departmental and regional bodies.

But this great state safety net is full of holes. The government has a strangely unbalanced approach to risk, overemphasising some contingencies - terrorism, anarchy, attacks by rogue states - while underplaying, even promoting, others. It was Gordon Brown, for example, who told the bankers of the City of London in his Mansion House speech of 2004 that "in budget after budget I want us to do even more to encourage the risk-takers".

There is one respect in which the government's approach seems utterly bonkers: a threat with a high likelihood of occurrence, for which it refuses to make any plans at all. I've been banging on about this for a while, with my usual absence of results. But now I've received a letter that makes its dismissive response look like outright lunacy.

There is nothing certain about the hypothesis that global supplies of conventional petroleum might soon stop growing and then go into decline. There is a large body of expert opinion, marshalling impressive statistics, which is convinced that peak oil is imminent. There is also a large body of expert opinion, marshalling impressive statistics, which insists that it's a long way off. I don't know who to believe. The key data - the true extent of reserves in the Opec nations - is a state secret. Anyone who tells you that oil supplies will definitely peak by a certain date or definitely won't peak ever is a fraud: the information required to make these assessments does not exist.

In February 2008 I sent a Freedom of Information request to the Department for Business, asking what contingency plans the government has made for the eventuality that global supplies of crude oil might peak between now and 2020. The answer I received astonished me. "The government does not feel the need to hold contingency plans specifically for the eventuality of crude oil supplies peaking between now and 2020."

As it revealed in a parliamentary answer, the government relies primarily on the International Energy Agency for its assessment. When I made my first request, its cavalier attitude chimed with the IEA's. But at the end of last year the agency suddenly changed tack. Its World Energy Outlook report upgraded the annual rate of decline in output from the world's existing oilfields from 3.7% to 6.7%. Previously it had relied on guesswork. This time it had conducted the world's first comprehensive study of decline rates, covering the 800 largest fields.

The report also contained a word the agency had hitherto avoided: peak. It proposed that "although global oil production in total is not expected to peak before 2030, production of conventional oil ... is projected to level off towards the end of the projection period." When I interviewed the IEA's chief economist for the Guardian, he tightened this up: "In terms of non-Opec, we are expecting that in three, four years' time the production of conventional oil will come to a plateau, and start to decline ... In terms of the global picture, assuming that Opec will invest in a timely manner, global conventional oil can still continue, but we still expect that it will come around 2020 to a plateau as well ... I think time is not on our side here." He told me that we would need a "global energy revolution" to avert this prospect. Nothing of the kind is happening.

So I sent the British government a new request: in the light of what the IEA has revealed, what contingency plans has the government made? The response has now arrived. "With sufficient investment, the government does not believe that global oil production will peak between now and 2020, and consequently we do not have any contingency plans specific to a peak in oil production."

I just don't get it. Let us assume that there is only a 10% chance of the IEA, and everybody else predicting that global oil supplies will soon peak or plateau, being right. That still makes peak oil about 100,000 times more likely than an outbreak of smallpox in the United Kingdom.

As the report by Robert Hirsch - commissioned by the US department of energy - shows, the consequences of peak oil taking governments by surprise are at least as devastating as a smallpox epidemic. "Without timely mitigation, the economic, social and political costs will be unprecedented." Hirsch estimated that to avoid global economic collapse, we would need to begin "a mitigation crash programme 20 years before peaking". If he's right and the IEA is right, we are already 10 years too late. But my conversations with government officials suggest to me that they wear the absence of plans almost as a badge of honour, like the Viking berserkers who went into battle without armour to show how mad they were.

The only explanation I can suggest is that the concept of insufficient oil cannot be accommodated within the government's worldview. Its response to a smallpox epidemic accords with its messianic tendencies: government as superman, defending us from crackpots carrying vampire pathogens. The idea that we might be undone by an issue as mundane and unresponsive as resource depletion just doesn't fit.

But at least we know where we stand: we'll have to make our own contingency plans. Does anyone have a spare AK-47?