Perhaps no statement by economists causes physical scientists more aggravation than the claim that the economic importance of energy can be measured by its dollar share in total GDP. Here I explore some of the arguments for both sides of that claim.
Econbrowser readers will be very familiar with Stuart Staniford, whose thoughtful observations have added greatly to the quality of the comments sections on many previous posts. He’s now writing regularly for the Oil Drum, and made some very interesting points there last week:
An economy
consisting entirely of bodies in a state of rest or uniform motion would not be much of an economy. Instead, almost
all economically significant actions involve some forces being applied to some bodies….From an economic perspective, if you want action faster (more acceleration) you will need more force, and
if you want more stuff moved (greater mass), you will need proportionately more force also. So there
should already be some sense that the total amount of forces being thrown around in the economy must
be related somehow to the total amount of economic activity.
Although I agree with his physics, I don’t agree that the level of economic activity has any necessary connection to the amount of force being thrown around. Consider two of the areas in which we’ve seen the greatest economic growth over the last twenty years, computing and medical care. Today you can perform something like 50 billion floating point operations per second for 5 watts of electricity. Stuart could do a much better job than I could of coming up with a graph of FLOPS per watt as a function of time, but I have no doubt that it would exhibit a spectacular rising trend. The reason is that the nature of technical progress here has been to make the physical dimensions of what is being moved around smaller and smaller– today’s computers are better precisely because they’re trying to move electrons over shorter and shorter distances.
Or consider medical advances. Antibiotics, immunization, pharmaceuticals, and improved surgical methods have produced profound economic advances. We stand on the verge of almost inconceivable progress over the next century as we put into practical use our ability to read the genetic code of ourselves and our diseases. Here again what we actually need physically to induce are tiny events at the molecular level. Economic progress consists entirely of getting smarter at knowing exactly how to do that.
Nor are computers and health care all that exceptional. The graph at the right plots the ratio of the number of barrels of oil that the U.S. uses each year to real GDP (measured in year 2000 dollars). Year after year, economic activity has consistently grown at a faster rate than our physical consumption of oil.
None of this is to deny Stuart’s observation that if we had no energy, there would be no economic activity. But it would be equally accurate to claim that, if we had no labor, there would be no economic activity. That doesn’t mean we should subscribe to an energy theory of value any more than we should believe in a labor theory of value, or, for that matter, an aquatic theory of value, since, after all, you and I would perform most unsatisfactorily if we were to be completely denied access to H20. Energy, labor, water, and a great many other things are valuable, even indispensable, for economic activity.
But the question is not, how well would we function if we used no energy, but rather, how well would we function if we used a little less energy? If you had to reduce your gasoline consumption a little, what could you do? Maybe you would figure out a way to combine errands, make more use of public transportation, share rides, or use a more fuel-efficient vehicle. You’d rather not do those things, and you won’t do them voluntarily, unless the price of gasoline rises. If it does rise and you do cut your energy use, how would we measure how much worse off you are as a result of the changes you were forced to make? One approach would be to take the amount by which you’ve reduced your gasoline consumption (call it Q) and multiply it by the price (P). The reason that’s a logical measure is that you could have considered giving up any of a number of other things rather than reducing your gasoline consumption– fewer movies, fewer new clothes, or whatever. The dollar value of clothes, etc. you would have had to give up to keep your gasoline consumption constant is given by P x Q. You decided those things were worth more to you than the extra gasoline (else you would have given them up, not the gas). Thus the economic value to you of what you’ve lost by being forced to reduce your gas consumption is less than P x Q. This reasoning suggests that the economic importance to the overall economy of losing a given quantity of oil could be measured by the dollar share of that lost oil in total GDP.
Back to Stuart’s analysis:
I have a college level Macroeconomics textbook by my side. It devotes about two out of 519 pages to consideration of energy, wherein it says nonsense like: “Because energy constitutes a small proportion of the nation’s total expenditure on inputs, most statistical studies suggest that higher energy prices did not contribute much to the slowdown [in the 1970s].”
I made a related statement about standard macroeconomic models here, which may have been part of what encouraged Stuart to give up on my blog and start writing one of his own.
Despite the disagreements noted above, I expect that Stuart and I would agree on many of the substantive issues. For example, as I argued in the post just referenced, my own view is that energy prices did contribute to the economic slowdowns in the 1970’s, not because of the lower energy use per se (whose effect, I agree with the nonsensical macroeconomic textbook, could be measured from the dollar value of the lost energy), but rather because these oil shocks caused adjustments in a number of other inputs besides energy. What we’re all watching for in the present situation is whether similar adjustments again occur over the next several months.
Another key issue in which both Stuart and I share an interest is what will happen when global petroleum production begins to decline. As ever larger curtailments in energy use are sought, the economic cost of further reductions will become greater, essentially for the reasons that Stuart is giving. This means that energy’s dollar share will start to rise as oil production begins to decline. As Stuart has correctly pointed out in a different post, if depletion rates turn out to be very rapid, we could move very quickly into that regime in which energy has a large dollar share, meaning that further reductions in its supply have tremendous economic consequences. Precisely the same economic reasoning that might lead you to conclude that energy by itself should not matter that much in 2005 could lead you to conclude that energy by itself might matter a great deal in 2015.
Technorati Tags: peak oil
For a good discussion of the tremendous flops per watt and other currently exploding exponential trends, and extensive discussion of how fundamentally everything is information – and how we are progressing towards an evententual 100% computing/information society, check out the forthcoming book The Singularity Is Near by Ray Kurzweil.
The statistics I found say that the U.S. uses about 20 million barrels of oil per day. This is about 7.3 billion barrels per year. At $63/barrel this is $460 billion of oil per year. U.S. GDP is $11.75 trillion, so as a fraction of that, oil is 4%.
But obviously if we eliminated oil usage it would not cut GDP by only 4%.
I think what would happen is that in a severe oil shortage, the price would be bid up. The result would be that, paradoxically, oil would take a larger share of GDP, the less there was of it. If we had 10% less oil, as a wild guess maybe oil prices would double. Then the share of GDP would be greater by a factor of 1.8, not 10% less.
So the problem with this 4% formula is that with inelastic demand, even modest changes in availability can have a large impact on price, making the 4% figure highly volatile. I think the bottom line is that the impact of oil shortages is complex and hard to predict. It’s not going to have a 1 to 1 impact on the economy, because people can substitute and conserve. But it’s going to be more than just its before-shortage fraction of the GDP.
I haven’t given up on your blog at all JDH – I think there’s great hope for you, and I learn a tremendous amount from reading with you. I will save most of my response for later in my series of posts over at the Oil Drum. But for now, I point out the following:
We will need to distinguish two alternative
A) where energy has low price (and hence contribution to GDP) because there are plenty of economic goods where you don’t really need much if any energy, or could use something else instead of energy.
B) where energy has low value because, even though you need it for everything you do, you have a huge pool of it out the backdoor where you can always take a little of what you have and use it to go get a bunch more out of the pool.
At least some economists seem to believe that A) is the dominant situation, and some of your comments above have that flavor. I will argue that B) has been the dominant situation, and the difference matters greatly. In essence the laws of thermodynamics make it clear that energy is completely unsubstitutable, for deep physical reasons, though individual forms of energy may substitute for each other modulo some practical difficulties. Or at least, the laws of thermodynamics prove that, and I will try to explain it in a way that makes it clear.
In terms of your P*Q paragraph, let me ask this – why wouldn’t the elasticity of energy be the relevant measure, rather than the price? If something has extremely elastic demand, that says it’s non-essential – if the price goes up, we’re just going to switch to something else, or use less. On the other hand, if the demand is extremely inelastic, that says that if the price goes up, we will continue to use almost as much, suggesting we see few viable substitutes and are very resistant to using less. Energy is very inelastic (suggesting to me that the market is well aware of the laws of thermodynamics).
It is clear that economy cannot be reduced to energy physics and mechanics. But it is also clear that economy is not only money, prices, supply and demand, labour and capital. It has a physical, real fundament where the law of physics do work: goods are made and transported.
I see the economy as a complex, multi-layered structure. The layers are coupled together but with some flexibility. The upper layers (eg. financial) are where classical economics work. But under them are the physical layers where classical economics have no effective modelling tools. These layers have also somewhat different time scales. Stock market or currency rate fluctuations are very fast compared to bringing a new oil field online.
The energy economics has something to say. It can point out that the historically unprecedented rapid world economic growth (from about 0.1% in 18th century to 2 – 10% in modern times) coincides with fossile energy production and consumption. It might appear that energy production is only a consequence of this growth. No, it is the other way round, it is the driving force of the growth. It is possible to pinpoint the beginning of the British industrial revolution to the coal mines of Northern England. Energy economics can explain the origins of the modern world. This is not trivial.
The energy economist can also say that the correlation of energy use and GDP has changed in some countries but it is still positive. This means that diminishing energy supply will bring negative growth. If energy use is already very effective small decrease in it will bring greater adverse effects! And the only realistic way is to look at the whole world: the smokestack industries have not vanished, they have moved to China and India. The Chinese coal miners are keeping the world economy going.
One misconception in this kind of discussion is mixing oil price and quantity. Energy economics see that oil prices are just prices. As such they behave like any prices and can be modelled with the tools of the classical economics. Prices do not correlate with the physical supply level, they belong to the sphere of supply and demand.
What really matters is the physical supply. We have high oil and energy prices and relatively high economic growth at the same time precisely because the oil supply has been able to grow and the overall world energy supply too. Here wee can see the physical component and energy economics at work. More energy use means growth, no matter what. No global imbalances can prevent this. This is my viewpoint in the discussion between the “Goldilock economists” and the “doomers”.
It is no coincidence that China accounts for the bulk of total world energy consumption increase. China is 94% energy independent. That is more than in any big industrial country. Look at the new industrial countries (China, India, Indonesia, Brazil). They have much higher energy independence and domestic energy production growth than the old ones. The all have significant coal production (Brazil has a lot of biomass).
The energy economist looks the world much more at macro level and historically. The underlying energy layer of the economy is difficult see otherwise. The deepest impact of the energy layer is in the investments. No new energy-intensive investments are made unless enough affordable energy is available for 20 – 30 years. That is why the coming peak oil affects already the economy. Look at oil refinery investments. Look at India where power blackouts are routine. No amount of cheap labor helps if you don’t have electricity and transporting goods is difficult. So it is in Africa.
I’m having a problem understanding the PxQ formula. Let’s take a concrete example.
Suppose I use 1,000 gallons of gas per year (because I drive 15,000 miles in a 15 MPH vehicle). Suppose gas goes up from $2 to $3. How much worse off am I?
Most people would probably say that I was paying $2,000 a year for gas before, but now I’m paying $3,000 a year, so I’m a thousand dollars poorer. However JDH’s formula looks at how much I cut back. Suppose that my demand is fairly inelastic so I only reduce my driving by 10%, from 1,000 down to 900 gallons, saving 100 gallons. By the PxQ formula this costs me $3/gallon x 100 gallons or $300. This is less than 1/3 the simple answer above.
So which is it? In a year, am I $1,000 poorer or $300 poorer? Or maybe I’m $700 poorer because I’m using 900 gallons at $3/gallon for $2,700 where before it was 1,000 gallons at $2/gallon for $2,000.
It does seem that I have $700 less in my pocket at the end of the year, plus there is the inconvenience of driving less, so I should be at least $700 poorer. I don’t see how the $300 figure makes sense. Guess I need to think about it some more.
One problem that Andrew Oswald of Warwick University has pointed out:
the US also uses less *copper* per unit of GDP than it did 20 years ago
ie less oil/ GDP is simply a measure of economic growth (more productive use of inputs to get the same output). This is also true of just about any exhaustible resource that is consumed (like copper).
What makes oil unique (perhaps) is that the substitutes are just not as good (whereas aluminium wiring displaced copper, and is for practical purposes just as good and with a vastly greater supply).
A better metric might be oil/capita. So in the US this is approximately 20 barrels per person per year. Canada is close behind (about 18 I think) and then we get to Western Europe (12-14).
On this basis, the US is pretty vulnerable to oil depletion. In a world where US citizens can only afford to consume say 10 bl/capita pa, that is a big shift in energy consumption. My gut says its probably possible for most of us to cut our consumption by 10%, even 20%, without huge changes in how we live, but when you factor in the supply chain (all those big box stores burn a lot of energy to keep stocked), 50% is a very big shift in lifestyles.
Of course, the US has a big GDP so it can ‘afford’ to make that shift. But the implicit shift in resource allocation is huge. And since much of that reallocation takes the form of embedded capital (cannot think of the economic term) ie buying new, more fuel efficient cars, restructuring cities along denser, more fuel efficient lines, etc. it is a very big task for any economy.
Hal, you have right. The question doesn’t make sense. This is not the energy viewpoint in economics. If we call energy economics that part of economics that deal with energy usage and effectiveness etc. that might be somewhat relevant question but then it would link to the demand elasticity, effectiveness etc.
Otherwise we could ask the same question about beer. What happens if the price of beer goes up? How much poorer are you?
It is very difficult to handle energy at microeconomic level because there are so much embedded energy in every product. If gas price goes up it affects everything that is transported, persons or goods, everything that uses gas as a fuel in producing something (tractors etc.). If the underlying cause is crude price rise then this will affect a lot of raw materials (plastics etc.) and in the end, through substitution, price of other forms of energy.
But if this a price change only it all means only changing allocation. You pay more but somebody gets more. Changing allocation would mean costs and disruptions but if the energy (net) supply gets growing we have still growth. The Peak Oil is not necessarily the Oil Price Peak.
Oil is not water. Fuels are not just raw materials. No amount of water using will cause economic growth (only hydropower does but this not really water consumption). Water can be a constraint but cannot generate growth. World growth did not begin in the banks of Amazon river.
Hal, as I pointed above oil is not just a raw material, it is not like copper. But it is also that. Oil gives plastics and a lot more. The point here is that it can be handled like a raw material, a mere commodity, but it still is more than that.
Look, if you cut the US oil consumption by 10% and cannot be compensated by other fuels it means that total US energy consumption will go down about 3%. That is a lot and causes negative economic growth. It would do that even if the share of oil in GDP would be 0.01% (there are oil producing countries where gas prices are practically nil). Now add 1% population growth. What this means in real economy? Sharply increasing unemployment. Collapsing investment. For instance. You definitely commute less if you have nowhere to commute.
I agree, the society will not collapse and people will survive and still live quite comfortably. But if the oil crunch comes it will mean much more than driving some less.
Believe me, I have seen countries in sharp energy crisis.
I really dont think we have to get too heavy into thermodynamics. Though the deep physical reasoning associated with thermodynamics is interesting, I think it sufficient from an economic perspective to consider energy as an essential resource (non-substitutable), with energy available from depletable and renewable sources. E.g. the depletable forms are stored in stocks such as uranium and fossil fuels and the renewable in the form of solar radiation.
Now depletable energy is available to the economy from different sources, e.g. oil or coal, and these have different levels of valuein some measure of energyto the economy. One can say that the EROEI of a particular source of energy is lower, but that just tells me that the energy available in that stock is lower than in some form that has a higher EROEI. And that will affect the price.
With that, we can forget thermodynamics. Im trained as a physicist, but I see no reason to complicate the discussion. I think it detracts. Im not with the Julian Simon cornucopians, but I also think it possible to discuss this issue without getting too hung up on the thermodynamics. It seems more a red herring.
Summarizing: Weve got energy flowing into the economy from depletable and renewable sources. And it is essential to the economy. There is no economy without energy. Though there is also no economy without matter, and if were going to start E=MC**2-ing ourselves to say matter is just energy so we only need energy, were way way off base. Energy is essential. Matter is essential. (Yes, with enough energy, I can create whatever matter I want, but like E=MC**2 mentioned above, its just not necessary to drag the discussion to this level).
So now weve got an economy that is fed with a flow of energy that is mined from depletable stocks and renewable sources (e.g. solar).
Energy is valued in terms of its flow into the economysupply and demand is largely in terms of the supply and demand of that flow. Higher flow, lower value. In addition though, the flow of energy can also be valued in terms of the perception of the future price of energye.g. a perception that the depletable energy stocks arewell, depleting. So we expect the flow in the future to decrease.
Then, economically, its just a question of how the economy uses that energy. And of course, how efficiently. And there is a non-simple relationship between the cost of energy and its supply and demand.
Over time, we expect the stocks of depletable energy to deplete, the flow of energy into the economy to decrease and the price to increase.
The fact that energy use is inelastic need not be described as a thermodynamic issue. The discussion need not be complicated in that way. It is essential, non-substitutable, and it isnt necessary to say why. No energy, no economy. In a hunter gathering society, no solar flow of energy, no life. No life, no GDP.
So it really just comes down to the fact that energy is essential (non-substitutable), and it is available from depletable sources and renewable sources. Then determine the dependency of the economythe GDPon this essential resource. It wont be a linear relationship, as someone mentioned earlier. But thats probably true of many inputs into the economy. And then determine the ability of the economy to restructure as the price of energy increases. True, that restructuring will most likely require some of that essential inputenergy–in order to be accomplished.
Energy is essential. But so is matter. A universe of photons floating aroundsay background radiation–with no matter would be a pretty darn boring place.
As far as the depletion curves over at OilDrum, the two issues Ive brought up over there are (1) all these depletions curves to date are local, not global, and one cannot assume that the global curve will look anything like the local curvesthey could be worse but the higher prices may draw them out, only to drop much more quickly later. And (2) depletion curves are just that, based upon our experience with these local depletions, not taking into account that other sources of energy were available from non-depleting resources. What we are really interested in is the consumption of energy by the economy in response to global depletion. True, the consumption must be constrained by certain geophysical limits. But we also have to consider (1) the behavior of the producers after global depletion occurs and (2) demand destruction, which may undershoot the depletion curves.
If oil is increasing in value while sitting in the ground, there seems to be less incentive to pull it out of the ground as quickly as possible. In addition, demand may drop in response to higher prices in such a way that the depletion curves are drawn out to the right. Demand may drop because of efficiencies, or reduced GDP.
Energy use is strongly inelastic, it is true. E.g. investments have been made in automobiles. But behaviors are changing. I drive a Toyota Prius and a Honda Insight. You would not believe the number of people who are now approaching me. One woman with an SUV approached me in a parking lot and then, after talking, wondered aloud whether she could even sell her SUV now.
Now I agree. It takes energy to restructure the economy (as mentioned above). So the question is how the economy adaptsor doesntto the increasing price of energy (i.e. the decreasing supply or flow into the economy).
It seems to me that economics can handle this issue fine without getting too excited about thermodynamics. Its when economists dont admit that demand destruction of a rather insidious nature is a possibility, and even a strong possibility, of depletionthats when I get excited. I think economics can provide us the answers. Its just that we might not like them. Its then the job of physics (and engineering) to look for efficiencies–where possible–sources of energyif possible. And politics to distribute the gains or pains.
At least thats how I see it.
Well, I propose to describe the thermodynamics as intuitively as I can, and people can read or not as they choose 🙂
But for example, why should anyone believe you that energy must “flow” through the economy? Why can’t we just recycle it and use it over and over? Then we’d never run out, right?
T.R.: I agree with you. Thermodynamics and economics is the same than biology and social sciences. Humans are biological creatures but you cannot succesfully derive social laws from biology.
Classical economics and energy-based economics are like Newtonian physics and modern physics. Newtonian mechanics works fine in the given environment. But it doesn’t tell the whole story.
But everything doesn’t boil down to the fact that energy is essential. Energy has a different role in economy than water and air which are also essential. The latter are a prerequisite and constraning factor, energy enables the economy (economic activity is work) and can make it grow.
It is also necessary to emphasize the role of fossile fuels (non-renewable energy sources) in the modern economy. The present economy would not exist without that kind of external energy input.
Energy use has been quite elastic – mostly upwards. And the higher oil prices during the first oil crisis did curb consumption and slowed markedly the world oil consumption growth. And it must be remembered that the availability of energy (or the expected growth of energy supply) does affect directly investment decisions. There is more than just the price mechanism. Even in the most market-oriented economy.
Look at China and the US. The kind of industrial capacity growth that is happening in China would be impossible in the US. The building and running of that capacity is extremely energy-intensive and it is easy to see that the US could not meet the energy demand. China produces every year nearly 200 million tons of coal MORE. Could the US do the same? The Chinese still produce about half of the oil they consume. What about the US? China is 94% energy independent, the US is not.
I see that many who discuss here are quite fixated to their cars. But from the macroeconomic viewpoint private driving is just secondary activity, consumption. You commute because you have a job – because someone has made investments – and you yourself have invested in your house somewhere. You drive to shop because Wal-Mart has invested and built their logistics so and so. You drive for leisure but only if you have invested in a car and have money (wage) to pay the fuel. These things change. Oil supply changes have also other effects than bying a Prius.
hal is right. Period.
The confusion is between inputs and outputs. Oil is really an input, we don’t use oil for oil’s sake, we use it in order to produce output.
Even with declining energy use, it would be possible to increase utility if efficiency gains were enough to compensate. That is why you mention the higher gas mileage, and the hybrid and whatever. These are efficiency gains.
Again. Hal’s first post was right. If you are an economist and you can’t come to terms with that, I would recommend you take the time to figure it out because it is a fundamental concept.
The object at rest analogy you cited is lame too.
One more thing. If oil production plateaus, then the number of cars that can be on the road with be equal to the amount of gas produced per day dvivided by the amount of gas consumed per car per day.
N = (Production)/(Consumption per Unit)
That means that the N wealthiest people/economic entities will get to drive automobiles and everybody else will have to use some other means of transportation. With China and India coming online and becoming major producers, they will be more and more of this chunk of privileged people and the US will get less and less. In order for this to happen, the prices in the US will get as high as needed to achieve this number.
Again, if the cars are more efficient (on average) N will be a larger number and more people will be able to drive.
Enjoie!
Stuart: I agree that thermodynamics motivates why energy cannot be recycled. But once motivated, in the economic models, it can simply be described as nonrecycleable. It is, in a real sense, pollution that has been distributed through the environment. Not that different from highly toxic trace metals that have been distributed through the environment. My only point is that it is not necessary to describe why something is depletable or non-recyclable. The economic models don’t care and shouldn’t.
Now the economic models must describe the dependence of GDP on energy prices. And as some pointed out above, it can be huge.
Reading Material from Elsewhere…
Some interesting reading from my newest neighbors at the rankings of The Truth Laid Bear:
Econbrowser
T.R.: “Now the economic models must describe the dependence of GDP on energy prices. And as some pointed out above, it can be huge.” This is just classical economics. The effect of Peak Oil is not only and mainly through oil price. The is no law that says that the prices should rise to a peak simultaneously.
It is the physical supply. I have tried to say that even in such market economy as the US there are other, important mechanisms through which oil supply affects the economy.
See, if there were no more oil, oil would had no price because there would be no oil in the market. At this point classical economics wouldn’t see nothing. No supply, no price, no effective demand. Nobody would invest in oil production (drilling dry holes). But this situation would still have a huge effect on the economy.
What’s true in the aggregate doesn’t reveal the impact at more discrete levels. When CNBC’s Jim Cramer concluded that oil was too cheap compared to milk, he likely unknowingly picked a very energy intensive foodstuff. The cost of energy as a proportion of the cost of a gallon of milk has to be considerably more than 4% of average GDP input costs. Many of the businesses at the various farm-to-glass steps are low margin businesses in capital intensive, highly competitive commodity markets. Woe to the farmer who due to location has made the capital investment in diesel rather than electricity for irrigation of alfalfa. Something on the order of 50% of the input costs for conventional dairy producers is the purchase of feed for cows. The bulk of that is usually either corn silage (requiring lots of nitrogen from natural gas) or alfalfa, some 40% of the cost of which is energy of one form or another before the recent rise in crude oil. Milk from the cow has to be rapidly cooled from her body temperature down to refrigeration temperatures (electricity running large refrigeration compressors), held there except for rapid heating to pasteurization temperatures (natural gas) and cooled back down and held at refrigeration temperatures through the distribution chain (refrigerated trucks), the grocery store dairy case (electricity) and the consumer’s refrigerator. The rapid rise in natural gas and crude oil prices has to be causing a lot of pain to many of the businesses along the production chain in this sector. Substitutions of energy forms and technology will occur but the old joke is that once a farmer has poured the concrete (picked his technology), because the investment is so high relative to the return only other farmers can learn from his mistakes.
If total energy inputs plateaued today, the economy could still grow at about 1.5% per year. That’s because, according to the EIA, global energy intensity (Btu per GDP) declines about 1.6% per year, and it’s also about 1.6% for all OECD and for all Asia. When oil prices rose sharply in the oil shocks, energy intensity declined even faster, suggesting that more aggressive conservation measures (or at least ending taxpayer subsidies for wasteful energy usages) might reap additional long term gains.
So with a little extra effort, we might easily see global growth of 2% in an energetically ‘steady state’.
Alternate view: global energy availability could contract by 1.6%-2% per year without any reduction in global living standards.
And, of course, this is all before any increased reliance on renewable or nuclear energy sources…
Hal, here’s how to make sense of your example.
You are not just $700 poorer, you’re poorer by $700 PLUS you’re poorer by 100 gallons of gas. But the person who sells you gas is now $700 richer. The extra $700 you pay is not a net loss to society, but merely a transfer to those who have ownership claims on the oil from those who do not.
The net cost to society is measured by the fact that you now get to enjoy 100 fewer gallons of gas. PxQ is a way of measuring the value of that lost enjoyment.
TI: Heres where Im at with this whole price, stock, flow issue. There are several ways to think of this. (1) The flow of energy into the economy largely defines the price in terms of supply/demand right now. (2) The quantity and quality of the depletable energy stock (e.g. oil or coal or tar sands) largely defines the volume of the flow into the economy, though demand is also controlling that flow (producer behavior, investment, etc). (3) The price of energy may also be affected by perceptions about the future volume of flow into the economy, e.g. the perception of the size of stocks. (4) Sources of energy, e.g. oil, also have other uses in the economy other than energy. The nature of these uses is different and should be considered or handled differently. E.g. plastics come from oil. Plastics can be recycled. Fertilizers come from natural gas. Fertilizers, by and large, cant be recycled. (5) For the time being, and into the foreseeable future, I think price is a reasonable means to determine the value of oil coming into the economy. That oil has a energy components to it and a material component to it. And the economy has to price it. Im not saying that peak oil means peak price. It all depends on demand. As the flow of fossil fuels into the economy decreases, they will be priced higher. That pricing will take into account its energy value to the economy and its material value (e.g. plastics, fertilizers, etc). At a certain price, GPD will stagnate and in fact can and probably will start decreasing (recession, depression). (6) Im not an economist, but I would think classical economics can handle this problem. See, oil will never completely deplete. Its just that demand will drop. There can be some very unpleasant ways in which demand can drop, including lower standards of living (per capita oil consumption) or less heads (capitas). Im not sure that classical economics says everything has to be pleasant. And through all the unpleasantness, oil will still have a price.
Now I agree that one could instead come up with a emergy or Nicholas Georgescu-Roegen type description of energy economics, but Im not sure that is necessaryand would be extremely difficult. The economy prices oil, and as it depletes the market will price it higher, and it will become a significantly higher percentage of the economy because of that price increase and because the size of other parts of the economy may shrink. Just as food would become a larger percentage of the economy as, say, flat screen TVsor any TV–became an impossible luxury for most.
We all agree that if energy usage were reduced 100%, all economic activity would halt. Let’s also all agree that that’s not the question here. The question here is, if energy usage were reduced by 1%, by how much would economic activity be reduced? The fact that the answer to the first question is 100% does not mean that the answer to the second question is 1%.
To answer the second question, basically we want to take a derivative. Hal, the derivative I was taking was with respect to quantity, not with respect to price. (I see now that the words I chose may have obscured that). I was asking the question, if you had to give up a certain amount Q of the energy you consume, what would be the economic value of losing that quantity Q? The answer to that question is given by P x Q, and does not depend on the elasticity.
Where the elasticity comes in is when you talk about reducing energy use not by an arbitrarily small amount Q (the concept of a derivative), but by a large amount. Then Hal is exactly right, the elasticity does matter, because you’re going to start spending more and more of your income on energy as the uses you try to eliminate become more and more essential. This in fact was precisely the point I was making in the last paragraph.
Thanks to Stuart and James for a great discussion.
Just the same, I think the economic piece of the oil puzzle is smaller than these discussions imply. Economic analysis works within a stable, rational, market framework.
We’ve seen, with legislative responses to Katrina, that human societies are quite ready and willing to “bend” the market in response to a perceived emergency. Tax repeals. Gouging investigations.
I mean, look at something as small as CAFE standards. They are a clear market intervention. They didn’t have too much backing a couple months ago, but one hurricane later they are back on the agenda.
For what it’s worth, I think that will be the pattern for the future. I think we will be looking at irrationally high consumption of oil up until it gets scarce, and then irrational responses to scarcity.
If you see a X% fall in oil supply, do you really think you are going to be looking at a purely rational economic response?
It seems kind of pointless to calculate the economic impact of a large shortfall, when we will doubtless respond to the shortfall by changing all the rules.
oh, another example would be how release of oil from strategic reserves changed the whole psychology of the situation.
Odograph: What you say is true, but what to do with it? I’ll use an extreme example. Economist: “Observe modern industrial society. Observe my economic theories regarding modern industrial society.” Observer: “Great. What happens if I drop an atomic bomb on each of your economic centers.”
My point?: Certainly the rules can be changed. The rug can be pulled out from any theory. Hence my point above about politics playing a role in distributing gain or pain. And no theory is going to describe all these different possibilities and contingencies: politically motivated market interventions, wars, aliens from another planet, discovery of perpetual motion machines (and overturning of the kinddom of thermodynamics). All throw monkey wrenches into the machinery.
It appears that industrial society must make a transition into a realm in which (a) the energy flow into the economy decreases and (b) we face increasing costs due to environmental damage, pollution, etc (e.g. global warming). I’ve consider these issues from energy perspectives and economic perspectives, and I have to say that economics seems to be the only way to rationally discuss the tradeoffs and alternatives. Some of the “what ifs” can be handled. Market interventions, regulations, etc. Others “what ifs” are more appropriate for “futurists” who can verbally describe the issues without the need for analytical analysis or modeling.
As far as whether we will irrationally consume to the point of irrational responses, well, I’m not sure what to do with that. Where does one go with it? How does one define rational in the above sentence? I’m betting that we will consume, with consumption increases decreasing as prices increase, until finally an energy shock (supply disruption) of one sort of another knows the wind out of growth completely. Then we see demand drop significantly. Even if there are irrational or market interventions at this point, one can’t push a string. The Govt can’t totally prop up deflating economy–if that is what is necessary to destroy demand. I’m just not convinced that will happen.
Let me try to make clearer my concerns with your approach JDH. Since oil is a fairly inelastic good, the variation in the proportion of GDP spent on it comes far more from variations in its price than from variations in the amount we use (the latter has long term slow trends, but the former can and does change by factors of several over a few short years). So the main explanation of the amount of GDP consumed by oil is price.
What is the price? It is an expression of the relative bargaining power of buyers and sellers. I think the argument that “oil can’t matter than much, it’s only 4% of GDP”, is equivalent to saying, “look, if people really cared about this stuff, they’d be paying a lot more than 4% of their income on it”. This comes because we generally know that if somebody *really* *really* wants something, then they are in a weak bargaining position and will end up having to pay a lot for it.
However, I would argue that the historical situation with oil is that people do really, really want it, and they are in a potentially weak bargaining position, but, this is offset by the fact that the sellers have been in an even weaker bargaining position. The latter is due to the fact that there was a relatively large amount of oil in the ground relative to human decision-making horizons (a couple of centuries worth anyway), there were a large number of suppliers, and anytime society wanted more, there was always someone willing to go drill more wells with enough probability of success that society as a whole faced few barriers to using as much oil as it felt like.
So the fact that oil has been cheap says nothing about how much we *would* be willing to pay. It’s like this. Suppose I come across an old lady with an eighteenth century Chippendale chair in her attic. She, not knowing it’s value, offers to sell it to me for $10. I am willing to pay $10,000 but I’m about to get away with it for $10. Unfortunately, just then a bolt of lightning darts out of the sky and incinerates the chair. Did I lose $10, or $10,000? You’re saying I lost $10. It looks to me more like I lost $9990.
What *does* say something about how much we would be willing to pay is the elasticity. What happens when the price increases? Answer, in the developed countries: price can double and usage barely changes. Beyond that, do people start to roll over and say, “ok, I guess I will take the train”. By and large, no. They keep paying more, but they get really upset and start demanding that politicians cut taxes, stealing gas from gas stations, blocking the roads in fuel protests, etc.
I largely agree with TR about the need to depend on economics in a limited resource environment. The problem is that most current economic models are largely linear models with a structural framework that has been developed and refined over the last several decades when energy has been relatively abundant.
As energy becomes a larger part of the economy and the physical resource limits become clear, the structure of the model has to change. This requires thinking from first principles, deriving new concepts and building new models. It can and will be done; Hotelling and [Reynolds http://oilcrisis.org/reynolds/MineralEconomy.htm%5D have already made a start. However, it is not easy for existing practitioners to do so, which is why so many economists continue to cling to the tried and true ?rules of thumb?. Note that this attitude is not limited to economists only; as a practicing engineer, I seldom think from first principles unless I encounter a situation where I am forced so to do.
The question is whether mainstream economics will respond in time to create the models and tools that can be used to mitigate the effects of Peak Oil. As energy rises from 4% of GDP to 10%, 20% and even 30%, the modeling of the impact on growth, inflation, interest rates, etc. will be interesting, to say the least.
Stuart writes: “What *does* say something about how much we would be willing to pay is the elasticity. What happens when the price increases? Answer, in the developed countries: price can double and usage barely changes. Beyond that, do people start to roll over and say, “ok, I guess I will take the train”. By and large, no. They keep paying more, but they get really upset and start demanding that politicians cut taxes, stealing gas from gas stations, blocking the roads in fuel protests, etc.”
But what is the counterproposal to the JDH approach? Discussing thermodynamics has nothing to add. It won’t model the stealing of gas, the cutting of gas taxes, or otherwise.
The energy flow into the economy will decrease. There will be a variety of responses, market, political, sociological (increase crime, protests, etc). If you are saying that energy is so fundamental that we have to consider it differently, I claim that (a) it is fundamental and (b) why treat it differently. I see no effective counterproposals by which it can be considered.
Again Stuart, distinguish the question “do you really, really want to use 500 gallons of gas this year instead of 0” from “do you really, really want to use 500 gallons of gas this year instead of 499”. I’m saying that in order to answer the second question, the relevant magnitude to look at is the current price.
I have no doubt that there are some drivers who really, really do want that 500 instead of 499. But if we nationally reduce gasoline consumption by 1 gallon, they’re not the ones who will give up that gallon. The person who’ll give up the gallon is the person who would really, really rather have $3.00 than that gallon of gasoline.
Ray: I’ve been looking at Reynolds lately and in fact just ordered his book the other day “Scarcity and Growth Considering Oil and Energy: An Alternative Neo-Classical View”. Also read his papers. Interesting stuff. Now talk about inflation and high-costs, look at the price of his book at Amazon. Geez.
What to do with it? Maybe just recognize that economic analysis is going to work best in “somewhat similar” conditions. To the degree that events are “rule changing” we will need new analysis.
Maybe I’m saying “wait and see” is more appropraite at this point than over-massaging the data.
To continue though – One of my peeves is that out there in the world I hear the current gasoline price system described as a free market. I know it’s not, but I don’t know how to describe how free it is:
Government builds roads (and regulates their use).
Private (or semi-nationalized?) companies build cars.
Private (or semi-nationalized!) companies provide oil.
It’s a complex system. I can certainly see it staying within the same “economic framework” for certain levels of oil depletion – but it ain’t “free” and for other levels of depletion the rules are certainly going to change.
JDH writes:”But the question is not, how well would we function if we used no energy, but rather, how well would we function if we used a little less energy?”
It seems to me that JDH and Stuart will keep talking past each other as long as they each view the others position is at the extreme end of the scale; ‘we have no energy’ or ‘we have to reduce our energy use by a little amount”.
Maybe we could agree on a single question such as:
How will the economy function if energy supply drops by 5% a year over several years?
For the moment, let’s now argue about whether or not this is going to happen or even possible. I would love to see an answer to that question based on classical economic theory.
FWIW, I suspect 5% a year over several years is rule-changing. That might even be enough to trigger rationing.
T.R: I didn’t want to claim that oil would end altogether. That was just a thought experiment. But I only wanted to say that the price is not the only means that the effects of changing energy supply is carried to the economy.
This is my point. I would dare to say that the pricing of oil doesn’t take very much account for its value as an energy source. This is clearly quite obvious: oil price has generally been mainly determined by production costs, not its “usefulness”. Just compare it with the cost of muscle work. The energy content has had meaning in pricing only in comparison to other fuels.
We may ask if this will change when oil becames more scarce. It is not self-evident. Its price will still have a relation to other fuels and substitutes. Even when all energy becomes more scarce the energy prices will not rise to the sky. In a sense energy has no absolute price, it only has a relative price.
We know that energy is not the biggest item of household budget in those countries where energy use is low. Food is the biggest item (it is energy too, of course). In order to use fossile fuels you must afford to be a fuel user: have a car, have roads to drive on…
Using energy prices to model what happens when energy supply decreases is therefore not very useful.
What Silent E said about the decreasing energy intensity is important but the question is not so simple here either. Ayres has made some empirical studies using the concept “exergy”, which means something like effective net energy. These studies have guite close correlation between real economic growth and energy use.
Besides population growth brings some economic growth by itself (yes, manual labor is also energy, so is increased use of biomass etc.). And the fact that energy efficiency is mainly realized by new investments is also important. So the lower energy intensity might be partly a result of increased energy use. If it would be so, the curve might not be symmetrical. Or its form might change: improved energy efficiency might be gained by removing the least effective capacity, but this means receding economy, of course.
There are studies showing that main source of productivity is not R&D but investments. The technology must be realized and that means new investment. But investments need energy. So here we are. So I would like to see more empirical studies: what has really happened in economies where energy supply has decreased.
RayJ, I discuss the “small change” case because that’s the one I’m sure I have the tools to understand– I know how to take this derivative.
The casual answer to your question– what would happen with 5% reduction each year for several years– would be to integrate those derivatives (that is, add the effect of giving up the 500th gallon to the effect of giving up the 499th gallon to the effect …) with the energy price gradually rising (that’s where the elasticity of demand would come in). I say that’s a casual answer, because my personal belief is that macroeconomic frictions would start to be very important in such a scenario, that is, we would see a significant number of people become temporarily unemployed, which would greatly magnify the economic cost. So, I think that Stuart and I would agree that 5% per year for several years could be a very significant economic event, though we may reach that conclusion from rather different ways of thinking about it.
But, the next question is whether 5% per year for several years is something that’s about to happen. Remember it’s possible to smooth that flow by changing the current rate of production and storage– instead of using 100 – 95 – 90 mbd three years in a row, it’s possible instead to use 97 – 95 – 93. Regardless of what the geologists may think, that’s physically an option for the economy. And the economic incentives are to produce something like 97 – 95 – 93 rather than 100 – 95 – 90. But whether you agree with that last conclusion of course gets us back into the big discussion that began here:
https://econbrowser.com/archives/2005/07/how_to_talk_to.html
In any case, my main purpose in the present post was to explain why economists don’t subscribe to an energy theory of value.
JDH: I’m confused by your reply to Hal. The derivative measures the effect of a small increment or decrement, say in Q, as Q approaches zero. But you seem to imply that elasticity suddenly jumps in or out of the picture when Q is somewhere near zero.
I would think PxQ would measure a direct monetized loss from losing Q. That would reflect a partial derivative. However, the net loss to my ‘utility’ might be partially mitigated because, all else equal, PxQ (money) would become available to spend in a presumably less satisfying way on something else. That complicates my ‘calculation’. In the real world, though, P would probably be going up as Q went down because it’s more than just somebody swiping Q from me and me alone. That further complicates my calculation. So my experience would be better understood by adjusting PxQ according to an appropriate total derivative, and that involves elasticities but has no particular discontinuities as a function of Q?
What, then, is known about elasticity for energy or oil-product consumption at various time scales? Some of heat generated by considering oil or energy seems to emanate from a tacit assumption that energy or oil elasticity is essentially zero on all time scales, so that middle-school children will be trundled hundreds of miles merely to play one soccer game no matter the price. Or perhaps it is a nearly equivalent assumption that non-agricultural energy must eventually dwindle to zero when hydrocarbon fossil fuels eventually dwindle, causing elasticity to approach zero as total available per-capita energy dwindles toward human basal metabolism.
After all, at zero elasticity, civilization might indeed collapse at the slightest feather touch.
“the US also uses less *copper* per unit of GDP than it did 20 years ago”
Developed economies use less of *everything* material as they advance. Less in absolute terms per capita.
E.g., NYC today produces only 45% as much garbage per capita by weight as it did in 1940, even through consumption in real dollar terms is 4x larger today. That 55% reduction per capita with over 4x income exceeds an 80% reduction per real dollar of GDP. That’s nontrivial.
The profit motive drives reduced consumption of physical inputs to save their cost, the cost of shipping and storage, etc. So everything from beer containers to office buildings end up using a fraction of the material they did previously.
Greenspan weighs in on this, so to speak:
http://anasazi.umsl.edu/FIN455/NonLinear/GreenspanWeighs.htm
This is why Simon could be so sure of winning his bet with Ehrlich, of course, and did so very one-sidedly. If you’re green and want to prevent the planet from being stripped, promote rapid economic development!
“What makes oil unique (perhaps) is that the substitutes are just not as good (whereas aluminium wiring displaced copper,..)”
When you use less of *everything* as the economy grows it is not a matter of substituting — except, of course, that you are substituting increasing knowledge for material inputs.
And where the ‘there is no substitute for oil’ idea comes from I still don’t know. Oil products don’t come out of the ground, they are manufactured in plants, and any hydrocarbon is a substitute for any other as an input for this purpose, going back to whale oil.
Heck, one can even make motor fuel out of dead cats…
http://www.ananova.com/news/story/sm_1534821.html
… giving new meaning to “put a tiger in your tank” (watch that Soylent Super Supreme).
Germany used oil from coal during WW II and did fine with it (until its plants were bombed) South Africa used the same during the embargo years and still does, and you’re putting gas from Athabasca and Venezuelan oil sands in your car right now. So clearly there are substitutes for Saudi light and Texas tea, and they are already being used at an accelerating rate. (Not to mention substitutes from other sources of energy for hydrocarbons).
The question is the available amount of substitutes, and their price. As to that, here’s an interesting lecture by Prof Stanford Penner, who has a rather long list of credits in academia and with the DOE…
http://maeweb.ucsd.edu/SSPenner_Lecture/SSPENNER/
He estimates how long an economy of 10 billion people with current US-level GDP could be supported by various existing energy sources: nuclear, fossil fuels, solar, renewable, etc.
For “fossil fuels” he gives more than 1,000 years — that is using fossil fuels alone, nothing else. (The total depletion of all fossil fuels to date being 0.015% — 99.985% of all fossil fuels are still in the ground.)
As to price, he gives alternatives to conventional oil, such as shale oil, oil sands, etc., as being economic on a huge scale at around $35/b. (He was with the synthetic fuel program during the Carter years). As noted above, it’s already started. And with normal technological development this price/b will steadily fall over time.
The Prof’s conclusion: “we will not run out of fossil fuels until we decide not to use them” — which we may very well decide to do, due to CO2 and other factors, though not due to physical shortage.
BTW, this is only a small part of the lecture, he’s got a number of interesting observations on other long-run issues.
The price elasticity of oil (gasoline) is not a simple question. But it is not unique in this. Elasticity changes with time, the curve may not be smooth but change its form at higher price levels. We have evidence that people are not willing or able to pay just anything to get gas (they may shoot for it, because bullets are cheap).
But the reduction of oil use will not come primarily as a result of rising prices. It will come from geological constraints – not so much more oil in the ground. The price hike will be just a consequence of this and will only help to allocate the remaining oil. Somebody gets more money for selling oil, somebody pays more for that. Somebody uses relatively less, somebody more. That is why the price approach has only limited use.
The share of energy in GDP will not necessarily rise to 20 – 30% when energy supply starts decreasing. It was not that high when energy consumption volume was half of what it is today. We know that a depression can reduce demand so much that the prices fall back – but the usage will not rise again much.
By the way: I think it is not possible to have an energy teory of value as an economic concept. Energy itself is free, it has no price. Nobody made oil but somebody dug it up. Oil and other fuels are carriers of energy and we pay for them as such.
This may sound odd but as I said before it is not diffcult to see that oil price is not really related to its usefulness as energy source. We cannot give energy value of something (a product or service) in dollars but we can give it in kJ or kWh, in energy units. But conventional economic theory is not very good in handling kWhs. That is why this is so difficult.
Reynolds has shown that prices do not signal correctly scarcity. Here is another reason for thinking that the price approach is not useful in here.
The link between economic theory and energy is mainly through the real economy. Macroeconomics, economic history and policy has traditionally also included looking at production statistics, geographical factors, investment activity, population etc. We cannot derive economic laws from thermodynamics but we cannot get macroecomic forecasts merely from microeconomic considerations of US gasoline elasticity curves.
Strictly on the question of measure of value, and whether units of energy or prices set in the market better serve as such …
One thing that Penner (who’s an energy scientist, not an economist) notes is research showing that photovoltaic has negative return on energy investment. It provides less energy than you have to put into making the equipment.
OK, so granting this, if units of energy are our measure of value, then photovoltaic has negative value (until this situation changes). Forget it, it only makes us poorer.
But does it really make us poorer? Suppose you are in Arizona without a power line near you … or are operating a commercial space satellite?
Value is in *people’s minds* not things. Since people are in different situations, the value of the same thing X differs for different people — as well as for the same people at different times and places.
Market prices recognize this fact, since they reflect opportunity cost. Here in NYC to buy photovoltaic I’d have to give up the opportunity to buy a whole lot of other things that are more valuable to me personally for the same price. So I won’t do it. But if I’m in Arizona or operating a space sattelite I’ll happily give up the other things to buy photovoltaic energy that is more important to me personally than them.
So market prices allocate items to where they actually have the most value to people — such as photovoltaic to where it produces a net welfare gain for people, instead of a net loss.
Value systems that place value in units of things — labor, food, energy, whatever — just can’t get a grip on this and always fail.
I noted a while back that the NY Times once ran an article on green economics that seriously stated “most people agree” that without a physical universe there couldn’t be an economy at all.
Granting that most people are right about this, it doesn’t mean that we should use some fraction of the physical universe as our measure of economic value.
INAE but doesn’t the graph above fail to account for all the oil we import indirectly though the import of products (made with and from oil)? It seems to me that the graph says nothing more than that we have moved many manufacturing jobs out of the country.
The way I see it, there is a hierarchy of frameworks. Without an understanding of the all the interlocking conceptual “frames”, you’ll spin your wheels.
At the base is Stuart’s physics and thermodynamics. You have to understand irreversibility, for example, to understand that energy is not “recyclable” as in a perpetual motion machine. Entropy is another cornerstone as is building order and resisting disorder (it explains Paris Hilton, for example).
The energy flows and ordering of the universe are fine and dandy, but we want to understand what it means to life, that is, biology. For that we should study ecology, where one can trace and model a biological community in terms of the energy flows within it and in the environment in which is embedded. For example, we used to thing that all life came from the sun and photosynthesis was to driver for all life. The discovery of living communities around the “black smokers” deep in the darkest seas changed that view. Life had adapted to the chemical energy spewed from the volcanic vents there – no sulight involved.
I treat economics as human ecology. Humans are still motivated by the primal urges of all life forms – survive, reproduce, spread. Until the industrial age, economics was about trade and land. Land was the main way to capture energy for human use (fishing is another). I think it was Ricardo who wrote extensively as land as the source of all wealth. Malthus also captured pre-industrial human ecology but his theories were nullified by fossil fuels. (OK, I read “The Worldly Philosophers” about 5 times as a kid!)
Prior to fossil fuels, as TI and I have discussed previously, human economic growth was a function of the slow accumulation of wisdom, order, knowledge, and techniques over the ages, that accumulation the result of skimming what little excess energy existed into writing, libraries, wise men, security, and good government.
Today, we are sitting at the peak of 200 years of excess energy. The rate of growth accelerated greatly with the knowledge of finding, extracting, and using petroleum as a fuel source to supplement photosynthesis. It seems that the growth rate was limited by how wisely we could grow, and operated something like compound interest. The excess energy could be reinvested into energy consuming (or conserving) capital goods that in turn allowed more energy to be consumed and more productively (electrification, for example.)
Today, we foreseeing the end of a period of easy, low investment, energy from petroleum. Of course, petroleum is by no means the only energy source we have but it has been one of the easiest to exploit.
We have built a complex society, highly productive that we would dearly like to keep running. To do so, I think we will see a massive shift in capital expenditures into new energy developments. Tar sands, heavy oil, GTL, nuclear, deep water oil – all are more capital intensive than Ghawar or Spindletop. Even efficiency investments (public transit, for example) will become more competitive for investors.
Please note that “price” is different from cost. Many petroleum producers capture the rents from (price – cost) and reinvest. Remember “Petro Dollars” in the 70’s? Only ownership changed.
As an aside, computers and information technology has come as a result of hundreds of years of energy investment in science and math. I laugh when our venture capital community thinks it can make another killing on energy like they did on IT! The difference is that information technologies are based on mathmatics while energy is, at heart, physics. In IT, the closer one can approach pure math, and the least one must be encumbered with physics, the better and more efficiently it works.
Good question, Professor and Stuart!
JDH:
Let’s take the oil shock case, where suddenly, for some reason, we have to use X% less crude oil. X varies from 0 to 1. You are saying, since the price of gasoline right now is $3, the economic value lost is $X*3/GDP.
I agree with you that the meaning of the price being $3 is that someone is willing to give up a gallon of gasoline if it goes to $3.01. You agree with me that if X=1 there is no economy (though that has never been my main point). So then we need to look at what happens at intermediate values of X.
The meaning of the elasticity being very high is that there aren’t very many people willing to give up very much gas usage at $3. At least on the evidence to date, gas going from $1.50 to $3 has not reduced gasoline usage much if at all. If it goes to $6, we’ll find some more people, but most people are still going to drive more or less as much as they do.
So as X goes from 0 towards 1, the price is going to go up. It’s some function P(X), and the elasticity tells us that the derivative DP/DX is a large number near X=0. So if X is more than infinitessimal, then the integral of P(X)(1-X) from 0 to X will get big fast. In GDP terms, the impact of X being non-zero will quickly rise to being a great deal more than 4%. I think we agree on this also, right?
So the remaing questions are almost semantic in nature. What exactly do we mean by ‘economic importance’ in the following?
“Perhaps no statement by economists causes physical scientists more aggravation than the claim that the economic importance of energy can be measured by its dollar share in total GDP.”
To me, the “economic importance” must mean “how well can the economy do without this”? As you increase X, the withdrawn percentage of the quantity in question, what happens to GDP? Roughly speaking, we need to compute d(GDP)/dX’, and then integrate X’ from 0 to X. But to first order, we could assume there’s a linear regime around the current situation and just say that the ‘economic importance’ of a quantity is d(GDP)/dX
Would you agree with that?
Now, how does GDP change with X. So, in the case of inelastic quantities like gas, the main thing that happens for small to moderate X is the price increases a lot. So suddenly, as a nation, instead of spending Y*P for Y gallons of gas, we are spending approximately Y*(P+dP/dX*X). We didn’t get any additional benefit for that extra Y*dP/dX*X (in fact we got slightly less gas). But that money now can’t be spent on other things. So, if we don’t change our savings/borrowing ratio, economic activity for other stuff must go down by Y*dP/dX*X. That’s essentially a loss in real GDP, right? (If we hold imports/exports constant) So d(GDP)/dX = Y*dP/dX.
So, in summary, I argue that
d(GDP)/dX = Quantity*Elasticity
and that’s what we should mean by “economic importance” (at least for modest X – we’d have to integrate for larger X). No?
Brian, you are right. Energy is imported and exported also as embedded in products and services. Only the global approach will do here. I think it is a common misconception that the “importance” of energy is diminished. No, as I have pointed out earlier here, the energy-intensive smokestack economy has not vanished but only moved to China, India, Indonesia and elsewhere. Their coal and products keep the Western “service and information economy” going. Not so much is changed as many people might think.
Jim, the photovoltaic cell is just an energy storage: you put energy in it producing it and it gives some of it back by enabling you to use the current generated by solar energy. So it is like a battery. We can calculate an objective energy value by looking how much energy is embedded in it. Subjective value is another matter. Market value an another.
I should add a disclaimer here lest anyone be confused. I am not a working “physical scientist”. I have a PhD in theoretical physics, so my understanding of basic physics is solid, but have spent the last decade being part computer scientist, part entrepreneur.
Just want to throw in a link: “Oil: Geologists vs. Economists”
http://www.prudentbear.com/internationalperspective.asp
I would like to note that the deeper issue of an “energy theory of value” is another in a long line of theories that seek to reduce the economy to a single input. The most famous such theory was the labor theory of value, put forward as far back as the 1600s by William Petty, advocated somewhat by David Ricardo, and very famously so by Karl Marx. We have also seen land theories of value, with the physiocrats led by Francois Quesnay in the 1700s being such advocates. Arguably some of the early Austrian economists such as Bohm-Bawerk had a capital theory of value as well. All of these, including various versions of an energy theory of value, face the problem that no matter how important the input in question is (would there be an economy without labor or land?), there are other inputs that are also crucial.
Furthermore, there are various versions of the energy theory of value, and it has been around in one form or another for a long time. An entropy theory of value has been identified with Nicholas Georgescu-Roegen (1971, The Entropy Law and the Economic Process, Harvard U. Press), but he did not really believe in it. Modern ecological economists, such as Robert Costanza, have focused more on solar energy and the entropy process of its dissipation through the ecosystem and the issue of the “ecological footprint” of the human economy rather than the much more narrow issue of the role of oil in the economy that has been the focus of discussion here (yes, oil is important, but not, therefore, the most important).
Probably the earliest to make an energy theory of value argument were G. Helm, _Die Lehre von der Energie_, 1887, Leipzig; L. Winiarski, “Essai sur la Mecanique Sociale: L’Energie Sociale et ses Mensurations, II” Revue Philosophique, 1900, vol. 49, pp. 265-287, and W. Ostwald, _Die Energie, 1908, Leipzig. All of these guys were physicists, with the latter in particular making arguments quite similar to many made on this list in this discussion.
However, from my perspective one of the more interesting of these earlier arguments was by the economist, Julius Davidson (“One of the Physical Foundations of Economics,” Quarterly Journal of Economics, 1919, vol. 33, pp. 717-724) who argued that the law of entropy was the ultimate reason for the essentially universal truth of the law of diminishing returns.
Oh, man, where to start?
Yes, it takes energy to move things around, but the things we choose to move are getting lighter (as someone above points out that Greenspan pointed out a while back), the things we use to move them around are more efficient at doing it, and we have lower-cost substitutes for moving things.
We have substituted plastic and aluminum for wood and steel, so our furniture and other products are lighter. Modern engines are hundreds of times more efficient than engines used 100 years ago, and getting more efficient all the time, so we get more useful energy out of them for the same amount input. And, as JDH pointed out to begin with, we don’t have to drive or ship packages across the country when a phone call or e-mail (with attachments) will do.
Also, we still have smokestack industries. We make more steel today than we did in 1950, but with many fewer inputs (about a hundred million tons vs. 97 million tons, and a many-fold reduction in the number of people and amount of energy required). The reason people don’t think we have those industries is that they have diminished in proportion to the rest of the economy because of (1) productivity increases, and (2) new industries have come along. Y’all know the saying about how generals fight the last war? I think we may have the same problem with respect to the economy. We know it’s changing, but we keep trying to think about it as if it hasn’t. That’s why so many New Deal policies were agriculture-oriented when the real problem was the demographic shift *away* from farming. We’ve been going through a demographic shift away from heavy manufacture, with the high water mark of ignorance of the fact being Mondale’s 1984 “nation of burger flippers” campaign (people are still blissfully unaware that the services industry includes doctors, lawyers, and engineers). A demographic shift is the movement of people, but we grow more food and manufacture more goods than ever before.
I read recently that the total energy usage is up about 200 fold since 1800, and population is up 50 fold in that time, so our per capita energy usage has only gone up 4 fold. Would you say that we are 4x better off than someone in 1800? I believe that I am much better off than that, so I agree that the energy theory of value is tenuous.
From data on the EIA site (which I will try to post tonight), I see that Canada uses almost as much energy per capita as we do, but doesn’t get the same per capita income from it. You could argue that it’s colder, but that wouldn’t explain how the Japanese get almost the same output but at much lower per capita energy. Going by the Japanese model, I’d say that we could forsake a lot of energy before we started to see a decline in living standard (surely, many caveats apply). But the danger is that we move toward the Canadian model and see our living standards fall without a significant change in energy intensity.
TI, I disagree. A PV cell does not store energy. It stores some, embedded in its structure, but it does not give that up to make electricity, whereas a chemical battery does (the energy is embedded in the internal chemical structure which changes as it discharges). Furthermore, a PV cell will never yield energy in the dark. A PV cell is an energy converter, albeit a very inefficient converter of very low density energy.
Eric:
I don’t think it’s controversial that the energy efficiency of the economy can, has, and will improve gradually over time, or that efficiencies vary from economy to economy. I don’t think anyone would disagree with that.
Regarding Canada, US, and Japan.
I remember reading recently that our midwestern states have the higher miles driven per capita. Makes sense, the population is more distributed geographically, and every product and service has to move farther to serve them. Given that, I think it would be hard to improve Canada’s efficiency to that of Japan … unless everyone moves to BC.
One of the problems with using an energy metric is energy quality. Another is the nature of the economy.
Canada produces lots of primary metals. Doing so is very energy intensive but needs little labor (relatively). Examples are nickel and aluminum.
Japan produces highly labor intensive products and imports their raw materials.
A BTU of warm water is not the same as a BTU-equivalent of clean 120 volt alternating current.
Quick note – I see my dP/dX is actually upside down of the usual economics definition of elasticity which is generally defined as relative change in quantity demanded over relative change in price. I don’t think it changes my basic point though.
However, I’m making an assumption that GDP is a differentiable function of current inputs. I’ve seen this in numerous papers I’ve looked at recently – Cobbs Douglas function etc – but it doesn’t seem to be a really well justified assumption – there’s a lot of hysteresis in the economy (ie the macro-state of the economy depends in an important way on the history as well as the current values for the various inputs). Has this been explored in the economics literature?
As I read the preceding discourse, I am reminded of a study in the 1890’s that concluded that the city of New York would collapse by 1920 because there was no way to deal with all of the horse manure that was generated in the city.
Projecting tomorrow is difficult. Projecting society and the economy ten years hence is impossible.
The level of current oil consumption is simply a reflection that it has been a low cost resource. The reason that we have not all bought more efficient vehicles is that oil is a low cost resource. The reason that we have built hydroelectric facilities in the past fifty years is that oil is a low cost resource. The reason that we have not developed more efficient systems for utilizing solar power is that oil is a low cost resource. The reason that we need air conditioning is because energy is cheap.
The problems you cite are simply symptoms of our current economic environment. When you change that environment, humans will adapt. The horror of course is that not all adapt equally. But then, there has never been any security, only ignorance of the threats to our well being.
Bill
Stuart Staniford
You defined X as the %change in quantity of crude oil used. The price elasticity of demand is defined as the (% change in quantity demanded) / (% change in price). If Y is the quantity of gas then the price elasticity of demand for gas is (dY/Y)/(dP/P) = (dY/dP)*(P/Y), note that X = (dY/Y). In other words, dP/dX is not an elasticity, neither is dX/dP, rate of change is not elasticity.
“We didn’t get any additional benefit for that extra Y*dP/dX*X (in fact we got slightly less gas). But that money now can’t be spent on other things.”
Only if the people who got the money take it out of the economy. The people who received the extra Y*dP/dX*X have increased their income by that amount and that money goes right back into the economy when they spend it, assuming as you say that imports and exports are constant and the excess money isn’t going to Saudi Arabia and staying there.
Stuart, actually the math is a little simpler than you’re making it. Let Y denote nominal GDP and X the total physical quantity of oil used (so Y is measured in dollars and X is in barrels of oil). Then dY/dX is simply equal to P. You can deduce this using production functions (just assumes differentiability, don’t need the special case of Cobb-Douglas) the same way I made the argument in the original post, and maybe it would have been less confusing to Hal and PaulS if I’d conducted the analysis in terms of production rather than consumer’s utility. This equation implies d ln(Y)/d ln(X) = P x X / Y, in other words, the percentage change in real output for a 1% change in energy use is given by the dollar share of oil in total GDP.
JDH: Another “hit” post with the “peak oil” topic 😉
Mathematical economics deals best with small (or “marginal”) changes in variables. The JDH/Stuart debate seems to hinge on how to cope with expected changes which will eventually go far beyond marginal (in this case oil depletion).
AFAIK, there simply does not exist any mathematical theory ready to deal with this in a satisfactory way. Not physics, not economics.
There’s a “small delta” regime where economics does a semi-respectable job, and a “huge delta” regime where we’re all dead (eg. TR Elliott’s nuclear bombings). I think there is probably a mid-range delta regime where there is enough pain (however defined) that meaningful readjustments are made. Economics doesn’t do a good job of modeling how an economy retools to changing inputs . But history does suggest that human beings are remarkably adept at adjustments provided the changes aren’t huge.
Tough to say what threshold level of change is too hard for us to handle.
The problem reminds me most of phase transitions in thermodynamics, but molecules don’t “innovate”. Put another way, economic actors don’t seem simple or symmetrical enough to aggregate the way thermodynamics aggregates molecules.
JDH:
I’m afraid I’m not following your math at all. You’re going to have to spell it out for me if I’m to get it. Possibly I don’t have enough background to follow you and you’ll have to educate me.
What it looks to me like you’re doing is making the assumption that the main thing happening economically when there’s a small reduction in supply is that small reduction in supply, which you can approximate as happening at fixed price.
What it looks to me like the main thing happening economically when there’s a small reduction in supply is a much larger relative change in price (since the elasticity of oil seems to have a magnitude less than 0.1 (given that the recent doubling of price has produced at most a few percent reduction in usage even spaced out over several years).
To put it in the terms of your original post, you’re looking only at the effects of the marginal gallon on the family that decided not to use it, and ignoring the much larger aggregate effect of the .001c rise in price (or whatever it is that it took to get that marginal reduction) on every other family in the economy.
To me, this inelasticity is the way economics represents what seems physically intuitive – since energy (strictly exergy) is pervasively used for all economic activities, any reduction in the supply is going to have an effect slowing something down somewhere, and the economy’s only short term choice is what it picks to do the slowing (which obviously will tend to be something that was already economically marginal).[1]
George:
Clarification on definition of elasticity agreed, but I don’t think it changes my point.
Also, true the oil suppliers get the money, but if they try to spend the extra on anything that involves use of oil, then since supply is constrained, that’s going to force a further price increase which will make someone else have to conserve somewhere else. Since there isn’t a whole lot that happens without some use of oil….
[1] Before someone feels obliged to point it out, yes I know over the long haul we can make gradual efficiency improvements and that will be part of our long term response. That’s not in contention.
I agree with STS. Microeconomics is different from macroeconomics or economic history. Energy economics doesn’t help much in predicting daily oil price fluctuations but can help explain economic phenomena of large scale and long time frame. The energy viewpoint has become interesting exactly for this reason: if the oil peaks it will be just that kind of large scale phenomenon.
But I dare to say that I doubt a little those optimistic assumptions of incessantly growing energy efficiency. Of course a car driving on a rough road will consume more fuel than one on a paved highway. But constructing the highway has demanded a lot of energy. Here the investments are crucial. So energy begets energy efficiency. That means that increasing energy efficiency depends on increasing energy consumption.
We see this in societies that have experienced severe energy crunches. If the crisis continues for some time the energy efficiency starts to deteriorate because the infrastructure and production capacity starts to crumble in lack maintenance and new investments. People drive in older cars on rougher roads (examples are Cuba, Ukraine etc.)
Japan is an example of a basically stagnant economy. But its energy consumption grew on average 1.6% in 1990-2001. Japan is very energy efficient but this number shows that the possibilities to increase efficiency significantly are pretty much exhausted.
China is a different case. There the energy efficiency is quite low but the economy is growing very fast. But sizable part of this growth is infrastructure investments. They are just now building those energy efficiency improving highways and modernizing their production and energy sector. But as we see this requires much energy. (Note: only if there are enough cars will the paved highways be built. So this is really connected to higher energy usage).
What happens
Stuart, as far as non-linear models for impact of input changes on GDP, you might take a look at “What is an oil shock?” by one James D. Hamilton, ftp://weber.ucsd.edu/pub/jhamilto/oil.pdf
The abstract: “This paper uses a flexible approach to characterize the nonlinear relation between oil
price changes and GDP growth. The paper reports clear evidence of nonlinearity, consistent
with earlier claims in the literature – oil price increases are much more important than oil
price decreases, and increases have significantly less predictive content if they simply correct
earlier decreases. An alternative interpretation is suggested based on estimation of a linear
functional form using exogenous disruptions in petroleum supplies as instruments.”
This looks at non-smooth functions such as ones where an increase in oil price decreases GDP but a decrease in oil prices have no effect. The models considered also include ones with lagged terms to model hysteresis. One of the conclusions is that oil price increases have more effect if they occur after a period of stable oil prices than after a period of volatility.
I looked at the “What is an oil shock?” by James D. Hamilton. It confirms that the relation of the oil prices and GDP is asymmetric. It is natural that sharp price hikes are associated with supply disruptions. The economy must response to these almost immediately. This forces the decrease in GDP.
When prices go down there is ample supply but nothing that could force the economy to react rapidly to that and increase consumption in short time span. Increasing consumption takes time and maybe new investments.
So this can very well support the idea that physical oil (energy) supply is the real mechanism that controls economic growth rate. The price movements are secondary in this sense.
Time should be factored in. The economic impact of a macroeconomical loss of an energy input varies with the time over which such loss is diffused. I.e.: a SUDDEN “peak oil” (aka “shock”) versus a gradual substitution of inputs (hopefully not dead cats). We could adjust to another Ice Age, for example – an Ice Age comes on slowly. (Sidelight: why rebuild New Orleans if global warming is real?)
As to energy economics, energy is necessary but not sufficient to life and economy; and oil does not equal all energy!
If peak oil is gradual (meaning people are growing more and more aware of the possibility as well as there is less and less economic supply), the economy will shift inputs-driven by local knowledge-with scarcely a bump. But not if the spigot runs dry SUDDENLY. This is the time factor; how to represent it mathematically? It seems to me time is the most crucial factor as well as the hardest to conceptualize.
Short time to adjust = doom. Long time = no problem. 🙂
My New Orleans comment shows what contributes mightily to inelasticity and “stickiness” in economic change generally: malinvestment. And the greatest cause of malinvestment, IMHO, is statism i.e. a command-and-control economy.
Without govt. (political) misincentives, no one in his right mind would rebuild a tourist trap in what must be almost the worst flood plain in the world! Yet Bush et al. will probably direct X billions in rebuilding efforts in defiance of nature and all common sense. Yes, some may transfer money to themselves as a result (such as Mr. Hertz; google him) but it’s pretty clearly malinvestment IMHO.
So, another factor in estimating the impact of an input’s deterioration: how much political stupidity in terms of govt. redistribution can we survive in the face of the need to respond flexibly to declining supply etc?
From Policy PeterPetroleum Policy and Geopolitics http://policypete.com/
“Leigh Yaxley, a Petroleum Engineering Manager based in Indonesia, has sent in an analysis of the issue. http://policypete.com/Misc/USA%20net%20oil%20import%20costs.htm He suggests that the key ratio is the percentage of GDP spent on net oil imports. Based on the economic downturn following events in the late 1970s / early 80s, he believes that a sustained period where the ratio exceeds 1.5% will bring on stagflation.*
*It is worth noting that Saudi Arabia may have a similar view, as it has taken unilateral steps to reduce the selling price of its marker crude. It also should be noted that net oil imports currently account for a smaller fraction of total imports than they did in 1980.”
As DR points out, time needs to be factored in. A large increase in the price of oil will only affect us if it occurs rapidly.
From 1978 to 1981, when OPEC tripled the price of oil, we went from spending 2.99% of GDP on oil going into refineries to 5.24%. That 2.25% increase in only 3 years was a shock to the economy.
Since that time, the economy has grown a lot more than our usage of oil. In 2004 we spent 2.01% of GDP on oil going into refineries. Even with the price of oil increasing this year, well only spend about 2.5% of GDP on oil. We can probably handle that kind of in incremental increase. (Thats about $80 billion. The effect from Katirna will be twice that.)
A note on numbers. I tend to follow oil going into refineries rather than total oil usage. Thats because thats where most of it goes. The total usage number includes about 2 million bpd in refinery gain. Thats the increase in volume due to the refining process. The refineries dont pay for that oil twice. They already paid for it when it went into their crude units.
Also, be careful about using the oil price quoted in the press. It is generally the price for WTI or another light sweet crude. Heavy sour crudes are trading at significant discounts to WTI up to $15 less. Lots of US refineries are now equipped to handle heavy sour crude.
According to the work of R. Ayres and Warr (IS THE US ECONOMY DEMATERIALIZING?
MAIN INDICATORS AND DRIVERS) the use of mass/GDP or exergy/GDP is not relevant as an indicator of the economy dematerialization as long as the masss/capita is increasing. quote:
In terms of mass per capita, the US economy is not dematerializing at all.
There are two countervailing trends that appear to explain this result. One is the
phenomenon known as the rebound effect, which is the fundamental growth mechanism that operates in the economic system. In brief, increased efficiency of production or usage leads to reduced costs and greater demand. The latter more than
compensates for the efficiency gains. The other countervailing phenomenon is increased complexity. Simply stated, as functionality increases (e.g. computer chips become smaller and more powerful) the manufacturing process becomes more and more complex. The ratio of indirect material consumption to material actually embodied in the product is extremely large, which is, in itself, a new phenomenon.
I can’t contribute to the economics much, but for adjustements, look up: http://www.thenaturalhome.com, http://www.annualizedgeosolar.com, and http://www.sunnyjohn.com. It is technically possible to adapt to local production in many things, if we start soon enough. If you have any land, it wouldn’t hurt to learn the newest high efficiency organic gardening methods-if we can’t eat it is hard to do anything else!
Gud morning, can i ask a question? about theory of energy storage?