Peak Oil Warning From an IMF Expert : Interview with Michael Kumhof

Michael Kumhof is the deputy chief of the modeling division at the research department of the International Monetary Fund (IMF). Together with other researchers from IMF, this 50 years old German economist has published several papers focused of “peak oil”, sending warnings regarding the possible imminence of a decline of world oil production and its impacts.

Echos of these concerns about peak oil are nowhere to be found in the official policy line of IMF, nor has Michael Kumhof ever been asked to present his research to the political leadership of IMF. That does not make his work pointless…

Ignoring the peak oil issue would be “highly unscientific, even irresponsible”, says  Michael Kumhof. (DR Anna Simonsson).

What made you focus on the peak oil issue ?

Michael Kumhof – I’ve have been personally concerned with this question for almost ten years. I am not certain of when a decline of world oil production might occur. But when I look at the world economy, the implications of such a possible decline would be so grave that I feel that, as a scientific researcher, I have to consider this issue very seriously.

 

What is the time frame between the most “pessimistic” and the most “optimistic” scenarios, before the decline starts ?

That is one of the things we are trying to figure out. One fact is that since late 2005, the growth rate of world crude oil output has been very close to zero. If you add in things like Natural Gas Liquids and unconventional forms of oil, then the recent trend becomes a little more positive. Nevertheless, if you just look at crude oil [which accounts for around 80 % of liquid fuel production], you could say, well maybe we are already at the historic maximum level of production.

If you take the whole liquid fuel output into account, it is too early to say when total peak production will be reached. Some credible researches have suggested that the time may come very soon. In our latest working paper at IMF, we just wanted to simulate what would happen at the moment when decline starts, whenever that happens.

 

Can you describe the level of gravity of the implications of a possible decline of world oil output ?

Let’s assume that at some point, world oil output will decline by 2% every year for a number of years. If that were to happen, our simulation says that could have very substantial effects.

According to our model, the reduction in the growth rate of GDP would be almost 1 percentage point per annum in the US and in the Euro area.

 

It does not sound that catastrophic, does it ?

Well, 1% of growth is a very large magnitude. If it happens over a number of years, it would mean that GDP ends up approximatly 20% below its previous growth trend after 20 years. That’s not a negligible quantity at all.

What you also have to look at is the price of oil: in our simulation, with a year-after-year decline of 2 % per annum of world oil output, the price of oil would rise to very high levels, in fact to almost 800% after 20 years ! I’m not sure that the world economy could easily take something like that. Now, what we are speculating at the end of our latest paper is that perhaps our model is missing something: once oil prices go above a certain level, the effect of GDP could be much larger.

 

What do you mean ?

The effect of oil prices on GDP can be non-linear. This means that beyond, let’s say, $200 per barrel, a lot of businesses might not be able to cope. You could think of transportation : trucking, airlines and the whole car industry might suffer quite gravely from such a high price. And that might have knock-on effects on other sectors of the economy.

Our model assumes that a relatively smooth adjustment takes place, where industries are just going to re-allocate to other energy sources. It is not clear to me that this is the right assumption at extremely high oil prices, but I can only speculate.

 

You acknowledge that your model has its limits when it comes to simulate the impact of a future decline of world oil production. Where does this model come from ?

It is a theoretical model that we have been using for many years, to run simulations of the world economy, for instance for past fiscal stimulus policies, and you know, for many of the things that have happened to the economy and have occupied the headlines during recent years. These were all economic scenarios : while being quite dramatic, none have produced extreme results. But an oil price of over $200 a barrel is a world we have never been in, especially if such high prices were to last not for a few months, but basically for ever. That is a world we just do not know.

 

You insist that we are not in a peak-oil world yet, with never ending high oil prices. Yet you underline the fact that crude oil output has been on a plateau for seven years now, staying flat despite the rise of oil prices.

At this moment, it is a little bit hard to disentangle whether this plateau is exclusively due to geological constraints, or whether it is also due to other important factors such as the financial crisis.

At the current point, the reason why oil output as been so flat may not be all about geology. Yet, our May 2012 working paper “The Future of Oil: Geology versus Technology” finds that, even if the world economy wanted to make oil output grow by about 0.8% per annum (which is what the International Energy Agency has been predicting), oil prices would have to rise very dramatically, by almost 100% over the coming decade !

There are several reasons why our model comes to this conclusion, including the fact that geology is definitely starting to play a role in our ability to grow output, while higher oil prices statistically only have a very small impact on growing the production of oil and on reducing the demand for oil.

As a caveat however, there is a lot of uncertainty surrounding our forecasts. For instance, you have the new US oil shale play going on. I think it’s going to help a little bit, but not that much.

 

So what can you say regarding the possible closeness of a decline of world oil production ?

In our latest paper, we refer to a very thorough study published in 2009 by the UK Energy Research Centre [pdf], which is by far the best study ever published on the “peak oil” issue, as far as we know. The authors of that study basically say that it is too early to be very certain of when the decline may happen. But they also say that it is almost impossible that it will happen after 2030. According to them, it will happen before, with a significant probability that it might happen quite a few years before 2030.

Another sound report directed by Robert Hirsch [interviewed on this blog in 2010], which was published by the US Department of Energy in 2005 [pdf], stated that to cope with peak oil would require a huge investment effort, without any equivalent in history, in order to realign the industrial infrastructure. That report showed it would take 10 years at the very least in order to achieve such a realignment without to much damage.

 

Do you find those studies convincing ?

I have no way of judging them with absolute certainty, but these results seem plausible to me.

 

You developed what you called an “entropy boundary scenario”. “Entropy” is not an economic concept, one you would expect to find in an IMF study…

No, unfortunately it is not. The concept of Entropy comes from physics and thermodynamics. Basically, it says that any system will always exhibit energy dissipation. Energy will always go from a state of low entropy to a state of high entropy, meaning energy will always tend to dissipate, from being concentrated to being diffuse and therefore useless. Many structures that we are using were built with the help of enormous amounts of concentrated energy, mostly in the form of fossil fuels. If you leave those structures alone, if you do not keep adding energy to the system, then after a few decades, those structures will just fall apart.

Economists have been neglecting entropy, and I think that is a mistake. There was a work published in 1971, by Nicholas Georgescu-Roegen. Even before that, the first person who ever tried to put the concept of entropy into economics was Frederick Soddy, who was the Nobel Price winner in chemistry in 1921. Their work got some attention at the time they were published, but afterward they got pretty much ignored. I think that in the future, entropy is something economists should look at a lot more.

 

What would you say of the degree of awareness within the IMF regarding the peak oil issue ?

Several people within the research branch of the IMF have contributed to research papers on the future of oil. But we in the research department are not producing the official policy line of the IMF.

 

Does the political leadership of the IMF ignore the problem ?

At this point the IMF actively encourages scientific, impartial research on this issue. But so far it has not taken an official position. We have presented this work internally at the IMF, but only to a Research Department audience.

As far as I am concerned, looking at the evidence, I cannot just dismiss the geological perspective anymore [i.e. the perspective of an absolute limit to future extraction capacities]. To just dismiss that perspective would be highly unscientific, even irresponsible.

 

Matthieu Auzanneau, The Oil Man blog, Le Monde blogs