Reflections on the Saudi Wars
Meet the Doomers
It was early 2007, and I was riding high at The Oil Drum. I had written a number of articles on energy policy, and a consistent theme of mine was that biofuels weren’t going to replace our current level of fossil fuel usage. For the most part these essays were very well-received, until I turned my attention toward the topic of oil production in Saudi Arabia.
Realize that while there is a diverse readership at TOD, there are quite a few very vocal contributors who are ‘doomers.’ What exactly is a doomer? Doomers believe that peak oil will inevitably lead to a Malthusian collapse of society. Many cheer for stories that support their idea of doom (e.g., “biofuels will not save us”), but they can be downright vicious if what you are writing implies that things may not be exactly as bad as they think. The latter was the case with my Saudi essays.
Matt Simmons and Saudi
I have been highly interested in what is going on with Saudi oil production for a long time. Saudi has a tremendous amount of economic leverage because of their oil production, and if their production declined sharply, then a lot of doomer points would start to look more plausible. Thus, I am keenly interested in understanding the true situation in Saudi. This was one of my primary motivations for reading Twilight in the Desert.
Twilight was published in 2005, and argued for a near-term collapse in Saudi oil production, with an inevitable price shock to follow. Following publication of the book, Matt Simmons made a $10,000 bet with New York Times columnist John Tierney that oil prices in 2010 would average over $200/bbl (see the Simmons-Tierney bet). This bet is useful for understanding the time frame Simmons had in mind for a Saudi collapse; certainly by now we would be in the midst of a full-fledged Saudi production collapse.
Given his message, it should come as no surprise that Simmons has gained quite a following among the doomers. He is held in very high regard by many at TOD, and a number of people have used his work as a jumping off point for their own claims of a Saudi collapse. And in late 2005 when Saudi production began to fall, it seemed to many that Simmons’ analysis had been spot on and very timely. The bandwagon began to fill up; the decline had begun and Simmons’ star was on the rise.
The Saudis maintained that the declines were voluntary because the world oil markets were oversupplied. But they would say that, wouldn’t they? Or would they? I went back and forth on that point; I could see pros and cons either way. But the doomer contingent had decided: The Saudis were bald-faced liars. I lost count of how many times I saw the Saudi Oil Minister accused of lying when he maintained that the declines were voluntary, because the doomers “knew” good and well they weren’t.
I Had My Doubts
I was especially curious to get to the bottom of whether Saudi was on the brink of a production collapse. Saudi production fell from the end of 2005 through the end of 2006 by one million bpd even as oil prices were rising. But I started really trying to get my head around this issue, and the more I looked, the more I was convinced that the Saudis were not lying. The declines did appear to be voluntary.
I laid out much of my reasoning in When Will Saudi Arabian Oil Production Peak? My position had three major points. First, worldwide crude inventories were at record highs and rising when the Saudi cuts began. We had this information directly from the OECD, but I also found news accounts of this coming from important non-OECD consumers like China and India.
Second, I took a long hard look at one of the major tools being used to project that Saudi had peaked. The tool was called Hubbert Linearization (HL), and I tested it first by plugging in historical data to see if it would have predicted previous peaks. In the case that was being used as a proxy for Saudi – Texas – it would have predicted peak production 16 years too early (as shown in the previous link). It would have also had a large degree of uncertainty until about 5 years after the peak. So for 21 straight years, one could have made the argument that Texas had peaked in that particular year on the basis of the HL.
Worse, I found that it would always predict a peak even if I fed the model an infinite series of constant, or even mildly rising production rates. And as more data was fed to the model, it predicted higher and higher recoverable reserves. In the case of Texas, what was predicted to be recovered in 1960 was far lower than what has been produced to date.
HL was the mathematical version of a dowsing rod. There was so much wiggle room that you could predict peak based on very liberal criteria. For many doomers, 2005 was that year, and I received a great deal of verbal abuse and hate mail for pointing out that the technique didn’t really work. I documented some of that in Peak Oil and the Lunatic Fringe, and that led to me taking an extended leave from TOD. (I had to block two regular TOD contributors because they bombarded me with e-mails over this).
There was one final point that convinced me that Saudi production declines were probably voluntary. First, it is true that Saudi reserves are not an open book to outsiders. They have withheld detailed data on their reserves since 1982. They raised their reserve estimates by 90 billion barrels in 1990, once again leading to chants of “Liar, Liar” about their reserve numbers. Presently their reserves are estimated to be 267 billion barrels. Doomers will tell you that this is laughable. The HL technique was pointing at a remaining reserve number of only 70 billion barrels.
However, I did a little sanity check on this number (a more detailed analysis than what follows is here). It is true that Saudi stopped publishing detailed data in 1982, but prior to that their reserves were an open book. In 1982, their reserves were estimated to be 164.6 billion barrels. Even if I assumed no new discoveries and just subtracted subsequent production, I came up with 95 billion remaining barrels – already well above the HL prediction.
But of course they would have had new discoveries as well and technology has increased the amount of oil that can be recovered. Look at what happened in the U.S over that same period of time. In 1982, U.S. reserves were estimated at 27.9 billion barrels. Over the next 24 years U.S. production was 56.9 billion barrels. Yet in 2005, U.S. reserves were still 21.8 billion barrels. So over that 24 year-period the U.S. produced 57 billion barrels of oil and pulled reserves down by only 6 billion barrels. To me this was another piece of evidence that the HL technique had to be wrong about Saudi.
I also tried to put myself in the shoes of the Saudi Oil Minister. How would I manage their oil? Pretty much just as he was doing it. I wouldn’t manage oil just so American consumers could have cheap gas. I would try to maintain prices at the highest possible level that could be tolerated by the economy. That oil endowment would have to serve future generations, so I would want to maximize the value. That’s a fine line, and if you are too aggressive you can cause economic havoc. But if I saw that global inventories were rising, I would begin to cut production as well to avert a future price collapse.
So my conclusion – which I stated numerous times starting in 2006 was: The Saudi production decline was voluntary, and if global crude inventories starting dropping they would raise production.
The Critics Emerge
If you want to get a real flavor for the kind of trollish commentary I had to deal with over this issue, see the comments following Stuart Staniford’s TOD essay A Nosedive Toward the Desert (…Or, Why the Decline in Saudi Oil Production is Not Voluntary). (By the way, none of this is meant to pick on Stuart. Reasonable people can disagree about the data, and that’s how I would characterize my debate with Stuart. I think his analysis was data-based, unlike many of the others. He was not using the HL as the basis for his analysis, and he did come around to the view that the HL wasn’t useful for predicting a peak).
Stuart called me out in that essay, suggesting that my arguments for why the Saudi decline was voluntary were “completely implausible.” His argument was the polar opposite of mine. He wrote “Declines are rather unlikely to be arrested, and may well accelerate.”
But people really went after me in the comments section. I was dealing with one attack after another not only on TOD, but they even spilled over to other sites. There was this great thread as well at the Peak Oil message board. One commenter who belonged to the “I love HL” and “Saudi has peaked” fan clubs had this to say (among other snarky comments):
Robert Rapier was among the more optimistic (David Cohen being another) regulars at the Oildrum. Those two are in decendence as the very convincing argument for SA decline by Westexas, Stuart, Euran (and tons of others) continue to gain validity.
And then this one, by the same poster (responding to a comment from someone else):
“Robert is way too optimistic regarding SA. I always side with west texas on those debates.”
and so do I, and it appears the majority at the Oildrum agrees that Mr. Rapier is no longer a major player. The ball is definitely in Stuart’s court. I understand Stuart has submitted his analysis to Science magazine for publication.
I was no longer a “major player” because I took the view that Saudi was not on the verge of terminal decline, which a lot of doomers didn’t like. A major player can’t give them an opinion contrary to what they “know.” If they do, they are by definition not a major player. Well, I would just have to settle for the consolation prize of being correct.
Saudi Production Turns Around
Look at what has happened since Stuart’s post. When Stuart wrote Nosedive in March 2007, production (C+C) in Saudi was 8.6 million bpd (Data from the EIA). I predicted that the declines would stop by summer, and little did we know that when Stuart published that essay, declines had just stopped and would be stable until late summer before beginning to rise.
The Saudis had production back above 9 million bpd by December 2007, and by July 2008 they had production at 9.7 million bpd – the highest level in almost 30 years (and without the aid of some of the major new projects that were expected to bump production a little). Their production then pulled back after prices collapsed. Just the fact that production flat-lined for 7 months with no new major projects coming on says without a doubt they were sitting on spare production when I was arguing that they were. If they hadn’t been, they would have declined a bit each month and could have only reversed that by bringing new projects online.
One argument that many people made for a permanent decline was that if Saudi had spare production they would have brought it online in 2006-2007 as prices climbed. As I replied at the time “Not if inventories are full.” (Of course Saudi production rose with the price of oil in 2008, and hit 9.7 million bpd in the same month that oil prices hit $147). This argument (and I am not naming names, but many of you will know who I am talking about) goes like this: “If Saudi had just kept producing at their 2005 levels, they would have produced X billion more barrels and made XX billion more dollars. Thus, it is implausible that their declines are voluntary.”
Later, a friend sent me a paper explaining that Saudi often cuts production in the face of rising prices. That’s because they are looking at data besides prices. See Saudi Production Management.
So what’s the point of this post? Am I just gloating? Not really, but after some of the treatment I received as a result of my arguments, I think readers could forgive me for doing so. I have to admit that it wasn’t all bad; I always had supporters as well. It is just that the kind words of a supporter have less impact than a bitter diatribe and volley of e-mails from someone whose world view you are threatening.
Anyway, three things motivated me to write this post. First, a reader commented after the previous post that they had appreciated the critiques of the HL. That planted the idea for maybe taking a look back at how Saudi production played out following the predictions of imminent doom and my counter-predictions of a production rise.
Second, I have observed that the amnesia and selective memory have really gotten bad on this point. People who made dire predictions seem to have completely forgotten about them, or they rationalize them away by saying that the declines are right around the corner. Or, they say that they knew all along that the decline would really be the plateau we have seen instead of a steep drop – and that the financial crisis would be the real story.
The level of rationalizing has been impressive; I have seen none of the vocal predictors own up to being wrong about this issue. Some people have simply stopped talking or writing about it, but others are still out there making the same sorts of predictions (some even insisting that their predictions of steep declines were correct; that the Saudis are lying about their production).
Finally, today I saw a post over at The Oil Drum by Leanan, the Drumbeat editor who really captured the mass amnesia in a nutshell:
Back then, it was a topic of much debate here. Was Saudi heading for “a nosedive into the desert”? Or would they “turn on the taps” later in the year, proving they were not yet at peak oil?
In reality…neither happened. Production did not crash, nor did it sharply increase.
I did respond by saying I disagreed; that in fact Saudi had increased production by 1.1 million bpd in the 15 months following Stuart’s essay. If over that same time period production had fallen by that amount (which was the magnitude of many predictions), I think we would have agreed that this would have been a crash. So it is hard to argue that a 1.1 million bpd swing in the opposite direction was anything but a sharp increase. But I also thought to myself “I should go ahead and write up my historical perspective on this, which I have never done.”
Something that was repeatedly misrepresented was that this was a debate over the actual peak date of Saudi oil production. It was not. It was a debate over a faulty methodology used to come up with a date that was being heavily promoted.
One thing is clear now in hindsight: Saudi did not go into terminal decline in 2005. Proponents of that theory have now shifted their position to “I will give up the idea that 2005 was the peak when the January-December average production exceeds that of 2005.” That’s wrong on two counts. First, production in 2008 rose hand in hand with oil prices, and by July when prices hit record levels the production rate was at the highest level in almost 30 years. If 2005 was the peak, no way would that have been possible.
Second, they seem to forget their argument. Assume for a moment that Saudi produces at only 90% of the 2005 rate, but do it for the next 40 years. Will the 2005 peakists maintain that 2005 was the geological peak? As I pointed out recently to someone who made that argument (“I am correct that 2005 was the peak until production for a calendar year exceeds 2005 production”) – Saudi production in 1980 and 1981 were both higher than for 2005. By their logic, I must conclude that 1980 was the Saudi peak.
I think ridicule and loss of credibility is inevitable if you are out making predictions based on shoddy analysis – which I felt was the basis for many of the imminent Saudi decline predictions. I believe when you are wrong about something, you try to learn from it so that future projections are better. If you simply rationalize away wrong predictions, you will likely continue to make them. But I have also learned that people using shoddy analyses to make predictions are also unlikely to own up to failed predictions. There appears to be a strong correlation between them embracing shoddy analyses that gives them the “right” answers – and rationalizing when the “right” answers turn out to be wrong.
Finally, while I feel like we won’t see the sharp declines in Saudi production right away, I still don’t like being dependent upon Saudi (or Venezuela) for U.S. crude supplies. I would rather see us proceeding with a plan that discounts their future production. Even if production doesn’t decline sharply, I think Chinese demand will keep pressure on prices, and therefore it would be a good idea if we seriously try to wean ourselves away from oil.