The economists focus on recoverability, arguing that the more wells you drill, the more oil you will recover; the more technology improves the more oil you will be able to recover; and, crucially, the more you can sell oil for the greater will be your incentive to go after difficult-to-extract oil. In other words oil fields that were uneconomical to exploit (because, say, they were in deep water) when oil was selling at $10 a barrel will become more attractive prospects when oil is selling for $30 a barrel. The amount of oil reserves depends, then, on economic and technological factors, not just on geological factors. You might say that if the price is right non-conventional oil will become conventional.

In effect, the amount of oil &ldquounder Hubbert's curve” grows over time and will continue to grow (there is a plenty of oil per se and only

2% of the world’s coal has been used), thus pushing the dreaded peak ever further into the future. Furthermore, the reserves in young oil fields tend to grow over time so the fact that oil discovery has been falling since the 1960s is not so worrying (since by definition these finds are of &ldquoyoung” fields). Adelman gives two dramatic examples of reserve growth in fields in The Genie out of the Bottle and also shows plenty of URR estimates that were quickly surpassed (15-17). I raised this point with Colin Campbell, who attributed much of this &ldquoreserve growth” to the caution of oil engineers:

&ldquoWhen you compile the individual fields you get the company reserves, as reported to the stock market – these are deemed to be financial data and quite rightly are presented extremely conservatively. Stock exchange rules are designed to stop fraudulent exaggeration but they don't have much trouble with under representing what you find. So as a result of all this the reserves of fields like most of the large North Sea fields initially were reported at 30 percent less than what they eventually delivered and this did not reflect any particular technological brilliance or anything special. It was simply the nature of the reporting procedures…”

&ldquoIt's easier to underreport a big field than a small one and so this underreporting phenomenon is now really historical because most of the large fields have been found … The underreporting of the past is really a historical phenomenon, not to be extrapolated into the future.”

Campbell finds some support here from Hallock (2004, 1694), who claims that most &ldquonew” oil comes results from the revision of conservative estimates of old fields. Geologists maintain that the world is by now thoroughly explored and it is unlikely that there will be major finds of oil. The new finds are smaller in size as, in the nature of exploration, the bigger fields tend to be found first. Colin Campbell admits there may be some possibilities in the polar regions and in deep water areas (no longer the frontier of exploration that they once were), but &ldquogenerally the world is now so thoroughly explored that it becomes more difficult and in fact if you look at the discovery record you'll find the peak of discovery was in 1964. It has been falling relentlessly since 1964 – that's more or less 40 years – despite the worldwide search, despite all the technology, despite the very favourable economic regime whereby most of the cost is offset against tax. So I can think of no good reason why this trend should change direction.”

I asked Michael Lynch if the argument that as oil prices rise the URR will increase was not academic: after all, if there is plenty of oil to be had but only at $200 a barrel will that not have the cataclysmic effect on the world economy predicted by his opponents in the debate anyway? He replied: &ldquoOne of the (many) arguments made is that much of the oil resource is expensive. This includes small fields, which contain a very large portion of the ultimate resource. Higher prices (or lower costs) mean that oil which was not previously recoverable (and thus not part of the URR) becomes economically viable, and thus increases the URR. If you study the curves of field size published by many, including Campbell and Laherrere, you will note that they are very flat at the end. This means that a small increase in price can move you far out the curve. For example, tar sands are now producing quite a lot of oil in Canada; 20 years ago, they were not economically viable. A significant portion of UK oil production is from small fields that weren't economically viable 20 years ago, but advances in technology and infrastructure lowered costs.” (Hallock considers Canada's reclassification of oil produced from tar sands as conventional &ldquounwarranted” (1677).)

I raised this issue with another economist, Daniel Crowley: &ldquoObviously oil is not just going to run out one day; it just becomes more and more expensive to extract. Once the price rises beyond that of competing power production technologies, everyone will switch to alternates, so there is a theoretical maximum oil price (though due to switching costs, the price would have to stay, or be predicted to stay, above a certain level for a long time). This maximum price also falls over time as competitors improve.”

&ldquoAssume the top limit is $200 per barrel. As the price rises, the profits of the oil companies stay roughly the same (all other things being equal), until they reach close to $200 and competing technologies make a big impact. Therefore the oil industry itself is fine, up to that point. The economic impact obviously falls elsewhere, i.e. on everyone else. Since the higher prices are a real cost, the economy overall suffers – it's not just a redistribution of wealth, but a loss of wealth due to more difficult or inefficient extraction techniques.”