Monday, April 21, 2008

Thoughts on Demand Destruction

Oil is currently well over $100/barrel. Demand is effectively holding steady in the US despite this recent run-up in price. There are some measures that suggest a decrease in demand, and the press has seized upon these to “prove” that high oil prices are causing people to drive less. I think this is cherry-picking of statistics: one commonly watched demand indicator, the one-week domestic gasoline demand figure as published by the Energy Information Agency in This Week in Petroleum actually shows a 91,000 barrel per day increase in gasoline demand for the week ending April 11, 2008 over the week ending April 13, 2007 (9.338 mbpd in ’08 vs. 9.247 mbpd in ’07). That’s a 0.98% increase year on year—so where’s the demand destruction??

This flies flat in the face of repeated statements recently in the press and blogosphere that gasoline demand is going down, so let’s look at it a bit more carefully. Here are the EIA’s full historical tables for gasoline demand, both week ending and 4-week average. Using the smoother 4-week average, the 2008 demand has been consistently lower than in 2007, but not by much. However, using the finer-resolution one week data, 2008 demand was higher than 2007 for the weeks ending 4/11/08, 3/28/08, but lower the weeks ending 4/4/08 and 3/21/08. For two of the last four weeks, demand for gasoline has been higher in 2008 than in 2007. This is hardly conclusive evidence of demand destruction, and completely ignores that the most recent demand figure shows a year-on-year increase.

Will we see significant demand destruction in the future? There is no clear answer to that at this time, but I think one thing is clear: it’s time to take a deeper look at the mechanics behind how demand destruction will work, if and when we see it (or, if we already are).

Does a lack of demand destruction when oil is well over $100/barrel mean that prices must go even higher to destroy demand? How much higher? Or is it enough that prices hold at this level for long enough to cause people to gradually make long-term purchases with this price in mind, and thereby destroy demand? How long? Finally, how much of current US demand destruction (to whatever degree it exists—even if only as a decrease in growth of demand) is due to current economic conditions, and how much can be attributed to price alone?

Figure 1: No significant demand destruction based on EIA’s gasoline demand chart… the most that can be stated definitively is that the past year has not shown appreciable US gasoline demand GROWTH over 2007.

Time-Lag in Demand Destruction: Major Purchases Drive Energy Consumption

One way that demand destruction occurs is that, when making major energy-consuming purchases such as a car or a house, people make more energy efficient choices based on the price of energy. These choices happen over time—everyone won’t (and couldn’t) rush out tomorrow to buy a more fuel efficient car, even if gas suddenly hit $10/gallon. How long is the time lag in these choices? Moody’s says that the average time between car purchases is 4.33 years. Even if we could figure out a magic number at which every consumer will pick a new car based on improved fuel efficiency, it would take at least 4 years to affect this transition. In reality, however, no one knows what percent of people would change to a more efficient car, and how much more efficient that new car would be, based on a given price of gas.

What about houses? Americans move houses on average every 5 years. Well, at least they did when they were upwardly mobile in a growing economy and sub-prime credit was easy to come by. It is yet to be seen how the current economic situation will change this figure, but it seems likely that our rate of moving will slow. In theory, when we move homes, we could choose more energy-efficient homes (better insulated, better solar design), or, possibly more importantly, homes that require less driving to commute to work. However, the massive sunk-cost in suburbia must be taken into account. While these homes may go down in value because of the commuting difference, they will likely remain largely occupied because, while the cost of commuting may skyrocket, the cost of ownership in the suburbs may decline to even this out. After all, the average American home is about 30 years old, and despite the promise of “New Urbanism” or downtown condo living to reduce gas consumption via commuting, the turnover of America’s housing infrastructure will take time.

Return on Investment Driving Demand Destruction

Demand destruction happens in other ways than buying a more efficient car or moving to a house closer to work. It is also possible to reduce demand by choosing a less convenient, less pleasurable, or slower option over another that consumer more gasoline. Take carpooling, for example. The passenger-miles-per-gallon of any car immediately doubles when a single commuter adds another commuter as a passenger. Four adults in a Honda Civic hybrid would average about 200 passenger-miles-per-gallon. Even four adults in a Hummer would get respectable mileage per passenger! If this is so simple, then why don’t we all do this? Because carpooling costs time, both in the time required daily to pick-up and drop off the additional passenger, time required to set-up the carpool system, and time in the form of inconvenience of people unexpectedly needing to work late, not being ready for pick-up on time, etc. How do we value this? There are no statistics that I’m aware of that track % of people who commute with one or more commuting passenger, or that track something similar, nor do I have any statistics for average “inconvenience time” per additional carpool passenger. At some gasoline price level, it makes sense for any given person to arrange to carpool. At $4/gallon, however, my impression is that most Americans will still value the time saved more than cutting their gasoline bill in half. The calculations for riding the bus, light rail, walking, riding a bike, etc. are essentially the same—how do you balance the money saved on gas with value of added inconvenience and additional time? For some people the decision clearly makes sense—but those are the people most likely to already carpool, ride the bus, etc. New demand destruction doesn’t occur until the price of gasoline changes the calculus, where it didn’t make sense at $3/gallon, but does makes sense at $X/gallon. How high would gas prices have to be for it to “make sense” for 50% of suburban commuters to carpool or ride the bus?

Economic Cycles and Demand Destruction

Ultimately, the kind of calculus suggested above is inextricably linked to the health of the broader economy. Rich consumers with large and growing disposable incomes are likely to value their time and potential inconveniences at a much higher rate than those struggling to buy groceries (notably, those with high disposable income are also the most able to pay now to upgrade to more efficient homes or cars, but least incentivised to do so). Another point to consider in evaluating demand destruction is the cause of economic problems. If economic problems are caused by high energy prices, then it seems accurate to consider demand destruction attributable to these economic problems as demand destruction caused by high energy prices. However, to the extent that economic problems are the result of an economic cycle, and not due to high energy prices, then the energy demand destruction that results does not seem accurately attributable to high energy prices. Our current economic troubles seem to be a function of both issues, but in my opinion more a short-term cyclical issue (inaccurate pricing of risk and the resultant correction, as I argued a few weeks ago (LINK)). At least some of the decrease in US oil demand can be attributed to economic cycles, and not to high oil prices, but we probably cannot separate these causes and isolate the portion of demand destruction caused by economic cycles. Can we even say whether or not demand would actually continue increasing at $113/barrel IF the US was in an economic boom? Does a statistic like GDP/barrel of oil consumed allow us to see through this fog? It might if we had a very accurate measure of inflation, but the CPI certainly doesn’t qualify. For that reason, comparing the 2006 GDP/barrel consumed vs. the 2007 GDP/barrel consumed is also problematic. Furthermore, it does not necessarily follow that, in a cycle-driven recession, GDP will shift to more energy efficient paths.


With gasoline well over $3/gallon, and oil well over $100/barrel, there does not seem to be any significant demand destruction in the US. Reasonable people can argue that demand is up about 1% or down about 1% since this time last year, but I am defining this entire range as “minimally significant.” What is the boundary of “significant” demand destruction? By significant, I mean significant impact on the supply-demand equilibrium for oil. If a low-end estimate of the decline rate for oil production post-peak is something between 2% and 5% per year, then I think that is the boundary for “significant” demand destruction. Demand destruction of 1% per year on an ongoing basis, compared with oil production decline of 5% per year, won’t have a significant impact on the supply-demand equilibrium. Conversely, a year-on-year demand destruction of 5% compared with an oil production decline of 5% has a very significant impact on the supply-demand equilibrium because it negates the impact of the production decline rate—this is a form of what Heinberg suggests in his Oil Depletion Protocol.

If this analysis tells us anything, it is that there is no easy way to calculate exactly what price point will cause demand destruction of X%. I remember when many proclaimed that $3/gallon gasoline would cause huge demand destruction. Now many of these same people proclaim that demand destruction will explode at $4/gallon or $5/gallon gasoline. Europeans, though admittedly in a very different situation, don’t seem to be driving significantly less at $8/gallon. In the end, we simply cannot know how demand destruction will unfold, and I think that is highly significant for calculating the economic impacts of rising oil prices—we have no empirical basis to either prove or disprove propositions as opposite as 1) present prices, if maintained indefinitely, will cause sufficient demand destruction to keep prices from rising significantly higher, or 2) prices will be able to at least triple before demand destruction begins to keep pace with supply declines. I know that there are nearly endless opinions on this point, but the significance of this analysis is that we cannot prove either point of view to be right or wrong. We can only wait and see what happens…

It's also worth pointing out that this analysis only considers US gasoline demand. Even if there is an ongoing demand destruction of 1% per year in the US, two significant factors overwhelm this: global demand growth remains strong, and net exports are falling precipitously (by 150,000 barrels per day in March alone). More on these items in future posts...


Tom Konrad said...

Good Job, Jeff.
You've given this subject the analysis it deserves.

I agree with you on "significant" demand desturction. Although I would consider the basically flat gas deamd we are seeing as evidence of demand destruction in a growing economy, it certainly is not enough to put a dent in the declines needed to offset falling supply. Conclusion: when we do see falling supplies, we'll also see much higher gas prices (relative to income) to produce "significant" demand destruction.

Rice Farmer said...

Indeed, I think the evidence for demand destruction is at best inconclusive at this point.

There is one point that I have mentioned here and elsewhere which I believe deserves much more attention than it's getting, and that is our economy's structural dependence on high energy consumption. Sure, some demand can be reduced without appreciable negative side effects. One example you give is the transportation choices people make.

But overall, I think this structural dependence is going to give us far more trouble than we realize. Reduced energy consumption itself is good, but in our economy that is the death knell for many businesses. For example, if enough trucks stop rolling, what happens to all the "truck stops" which depend on truckers to fill their tanks and bellies? (Not to mention what happens to businesses which don't get their deliveries.) If enough people stop driving on vacation, what happens to all the motels, theme parks, tourist traps, and other facilities that depend on tourists to come in their cars? Etc., etc. The more you look, the more you realize that if demand destruction really picks up, it's going to devastate the economy.

The whole socioeconomic system will have to be restructured, but since it's too late to do that in an orderly fashion, I think we are in for some shock therapy.

Jeff Vail said...

Thanks, Tom. I wonder sometimes how far prices must rise to really trigger "significant" demand destruction. I think there's a large psychological barrier to what seem on the surface like relatively painless and obvious solutions such as carpooling. I wonder how much dollar value to put on the "social stigma" of carpooling (negative in some circles, but probably positive in others...). It also seems like this is an area where failure of information is a key culprit--I think many cities like Denver could readily implement some kind of online ride-sharing network that pieces together individual ride segments, matches geography, and even allows for rating of riders and drivers in some manner. I know many commuters don't readily know people they could share a ride with... I doubt this kind of thing will take off at $4/gallon, or even $5/gallon. I'll go out on a limb and pick $8/gallon as a real tipping point for the American commuter. Something tells me I'll be able to test that prediction within a few years.

Jeff Vail said...

Christopher comments (via email) that demand destruction will likely be a non-linear event. My response:

"I agree that the changes Peak Oil will bring about will be highly non-linear, but I don't think we're near one of these bifurcation points yet. Yet. I think we still have some very "low hanging fruit" that we can pick before we have to get down to the more "difficult" business of reducing our gasoline consumption (using, of course, the royal "our gas consumption" of suburban America). Two indicators I'm watching are the line at the nearest drive-through Starbucks (as long as ever) and the frequency of carpooling in the morning commute (nonexistent, for now). Either one of these can effectively postpone the impact of oil prices on the suburban commuter until oil reaches double the prior level. When the drive-through Starbucks boards up its windows and I see one in four or one in five commuter cars with two passengers, then I'll believe that the slack has been removed from the system enough to make me think that a period of highly non-linear response is probable in the near future..."

Jeff Vail said...

Rice Farmer-

I agree that we have a huge sunk cost in our economic infrastructure that presupposes cheap energy. The flip side is that many businesses will benefit from higher energy prices. For every truck stop that goes out of business, we'll see some local producer become a little more cost-competitive with imports from China. For every distant tourist trap or interstate motel that closes up, we'll see some uptick in attendance at the local theater, visits to the local attractions or fairs that we, until recently, bypassed in favor of something more exotic. The key question in my mind is how do we compare the relative loss to the relative gain, and how smoothly are we able to intentionally adapt, rather than be forced into a rougher adaptation due to very uneven and unanticipated changes in the cost of energy. I'm especially worried about the latter...

Rice Farmer said...

Jeff, thanks for the observation. In other words, relocalization, right? Of course, that's what it will come down to (I am personally active in my locale trying to facilitate this change ASAP). My concern is that relocalization is not proceeding nearly fast enough. Inertia (our socioeconomic structure) is making it very hard to change direction. So that is why I'm certain that we are in for some "shock therapy." But if we work hard to prepare for and facilitate relocalization, we can mitigate the shock somewhat.

Nic said...

To the carpooling service: A little googling revealed that some people are trying to implement online services for transportation (mycarpoolstation, GoLoco).

I hope they put up a good technical infrastructure and manage to stay online until the oil price is high enough...

Jeff Vail said...

Nice graphic for the non-linearity/bifurcation discussion above (click here to view image)

Jeff Vail said...

UPDATE: The EIA's April 23rd "This Week in Petroleum" shows a significant jump in gasoline demand for the most recent 4-week period: 9.292 million barrels per day in 2008 vs. 9.207 million barrels per day in 2007... that's an increase of 0.9% year on year. So much for demand destruction...

Rice Farmer said...

Indeed. Destroying demand is not so easy, is it?

And then there's the coal situation in China, where growing demand is clashing with struggling supply.

klara said...

good post. I uploaded it to (like digg but centered around sustainability). It’s a great resource and I am trying to spread the word on it. Please upload your posts there if you can! Also I wanted to give you a tip about a company called climate counts. They are about to launch on may 7th. Here’s the link….
If you end up posting on this please let me know. I’d love to read it!

Jeff Vail said...

The EIA data for the week ending April 25th says:

"Over the last four weeks, motor gasoline demand has averaged nearly 9.3 million barrels per day, up by 0.4 percent from the same period last year. Distillate fuel demand has averaged about 4.3 million barrels per day over the last four weeks, up 0.7 percent from the same period last year."

Still no demand destruction...

Don Dwiggins said...

Some observations: looking for demand destruction in short-term statistics is somewhat like looking for proof or disproof of global warming in weather changes. Also, it's a common observation in peak oil circles that we'll only know we've hit the peak when we see it in the rear view mirror.

You might try looking a bit deeper for indirect evidence. Anecdotal example: we have two cars in our family; mine gets much better mileage than my wife's, so she's taken to using mine when it's available, even though she prefers to drive hers.

Other thoughts: maybe you could tease apart discretionary driving stats from non-discretionary; any reduction should show up first in discretionary. Look for evidence of shorter vacation trips. Look for evidence of "indirect" destruction of demand, that is reduction of demand for products whose rising prices can be attributed largely to increased energy costs.

fdg said...

fdg said...

I like your blog. Thank you. They are really great . Ermunterung ++ .
Some new style Puma Speed is in fashion this year.
chaussure puma is Puma shoes in french . Many Franzose like seach “chaussure sport” by the internet when they need buy the Puma Shoes Or nike max shoes. The information age is really convenient .

By the way ,the nike max ltd is really good NIKE air shoes ,don’t forget buy the puma mens shoes and nike air max ltd by the internet when you need them . Do you know Nike Air Shoes is a best Air Shoes . another kinds of Nike shoes is better . For example , Nike Air Rift is good and Cheap Nike Shoes .the nike shox shoes is fitting to running.

Spring is coming, Do you think this season is not for Ugg Boots? maybe yes .but this season is best time that can buy the cheap ugg boots. Many sellers are selling discounted. Do not miss . Please view my fc2 blog and hair straighteners blog.
.thank you .

I like orange converse shoes ,I like to buy the cheap converse shoes by the internet shop . the puma shoes and the adidas shoes (or addidas shoes) are more on internet shop .i can buy the cheap nike shoes and cheap puma shoes online. It’s really convenient.
Many persons more like Puma basket shoes than nike air rift shoes . the Puma Cat shoes is a kind of Cheap Puma Shoes .
If you want to buy the Cheap Nike Air shoes ,you can buy them online. They are same as the Nike Air shoes authorized shop. Very high-caliber Air shoes and puma cat shoes . the cheap puma shoes as same as other.

polo shirts

ralph lauren polo shirts
chaussure puma

chaussure sport

chaussures puma

puma CAT

ed hardy clothing

ed hardy clothes

ed hardy womens

ed hardy sunglasses