[MikeT]

I have viewed the debate on oil peak over some time and the impression I get now is that there is a shift from 'whether' production will peak to 'when'. There seems also to be an increasing agreement that this will be sooner rather than later.

Richard Heinberg's book (amongst others) paints a pretty gloomy picture of the consequences. I have no reason to doubt this overall assessment.

However I would appreciate some guidance on whether there are any other analyses or models on the economic impact of oil peak - and the characteristics of businesses and country economies that would make them more or less susceptible to these changes.

Regards

Mike

[trespam]

$this->bbcode_second_pass_quote('MikeT', '
')Richard Heinberg's book (amongst others) paints a pretty gloomy picture of the consequences. I have no reason to doubt this overall assessment.

However I would appreciate some guidance on whether there are any other analyses or models on the economic impact of oil peak - and the characteristics of businesses and country economies that would make them more or less susceptible to these changes.

[WebHubbleTelescope]

Can't read your mind. Offer up a curve and see how it fits historical price changes. You probably won't get anywhere because humans use reverse psychology to defeat the effects of the best predictions.

On the other hand, the title of your post refers to "models of depletion". Price plays a second-order role in how that evolves. More precisely, price is a side effect of depletion.

For example, many people say that tar sands may become economical now because it costs $30 per barrel to extract. (Which is much below the current oil costs) But they forget that the $30 was the estimated costs before the current oil price hikes. And since the $30 included the energy used for extraction, that has to go up as well. And likely a lot. Basically, its hard to predict feed-forward (inflationary) systems.

[pup55]

$this->bbcode_second_pass_quote('', 'P')'=k(D-S) => P=k*int( t[0]->t , D-S) + C

[WebHubbleTelescope]

Elijah, keep talking and I will keep reading. I appreciate your approach of starting simple and working your way forward. If it doesn't match reality precisely, maybe it doesn't matter as long as people can understand mathematically how the market may respond or evolve.

By second order I mean as in a Taylor's series. You can get by most of the way in modeling depletion without introducing price. However, if oil prices were perfectly inelastic and we hit the inflection of a vertical supply curve, perhaps no one would ever buy oil again and we would stay in a perfectly stable nirvana. Nothing would deplete because it would cost too much to extract.

The two attempts I have made in modeling depletion happen to use a statistical, first-order Markovian models. These are the simplest models I can come up with:

Macro Depletion
http://mobjectivist.blogspot.com/2005/0 ... model.html

Micro Depletion
http://mobjectivist.blogspot.com/2005/0 ... model.html

[EnergySpin]

I think that Hotteling (the same guy who invented the T-square test for multivariate Student T data) worked out the economics of depletion of a finite resource in the 30s. If my memory is playing games with me he said that the price should go up by the same percentage as the rate of depletion (expressed at the same unit of time).
My introduction to economics has always been through math (this way one can nail the bastards when they are lying for political reasons). The tool that is most likely to be useful in modelling is not an ODE but a Stochastic Differential / Difference Equation , making the approach of WebHubbleTelescope especially appealing

[jaws]

$this->bbcode_second_pass_quote('EnergySpin', 'I') think that Hotteling (the same guy who invented the T-square test for multivariate Student T data) worked out the economics of depletion of a finite resource in the 30s. If my memory is playing games with me he said that the price should go up by the same percentage as the rate of depletion (expressed at the same unit of time).

[jaws]

$this->bbcode_second_pass_quote('ElijahJones', 'H')owever I have to tell you that I know the price of oil tomorrow will not be under $30 per barrel and not over $100. Now if you ask me how much I will bet on that that is the measure of my certainty. Yes, as we move forward in time the typical person is willing to bet less and less on a prediction about the price of oil. And yet since the resource is depleting and there is a current built in demand we feel good about saying that prices will continue to rise as long as these factors persist. You see people are both free and not free, this is true economically, sociologically and psychologically. EnergySpins idea of a stochastic dynamical system is right on the money.

[EnergySpin]

$this->bbcode_second_pass_quote('jaws', '')$this->bbcode_second_pass_quote('EnergySpin', 'I') think that Hotteling (the same guy who invented the T-square test for multivariate Student T data) worked out the economics of depletion of a finite resource in the 30s. If my memory is playing games with me he said that the price should go up by the same percentage as the rate of depletion (expressed at the same unit of time).

[EnergySpin]

$this->bbcode_second_pass_quote('jaws', '')$this->bbcode_second_pass_quote('ElijahJones', 'H')owever I have to tell you that I know the price of oil tomorrow will not be under $30 per barrel and not over $100. Now if you ask me how much I will bet on that that is the measure of my certainty. Yes, as we move forward in time the typical person is willing to bet less and less on a prediction about the price of oil. And yet since the resource is depleting and there is a current built in demand we feel good about saying that prices will continue to rise as long as these factors persist. You see people are both free and not free, this is true economically, sociologically and psychologically. EnergySpins idea of a stochastic dynamical system is right on the money.

[WebHubbleTelescope]

Elijah,

Using the Latex markup, are you going to convert the stuff to PDF's or is there a good XML type plugin for rendering the equations?

[CrudeAwakening]

$this->bbcode_second_pass_quote('ElijahJones', '
')Honestly, I really don't like your attitude, whatever your personal beliefs are and I think that to keep this thread from flamesville I will just ignore you until you start making some sense. Do you have a degree in something? Maybe we should comfort one another with an exposition of formal logic. Did your econ 101 class teach you axiomatics or prepositional calculus or how about graph theory? Do any of these have things to teach us about economic systems?

Peak Oil is a serious issue that is likely to cost the lives of millions of people and I am not interested in entertaining children or bullshit! So please, if you cannot respond here with logical argument kindly start your own thread. And do not flame the participants or I will report you to the moderators.

[kmann]

EJ,
I'd like to give you a word of encouragement. This is good stuff, it will be interesting to see where it goes. I wish I could contribute, but my math at this level has accumulated several layers of rust. I plan on following this thread, and if I'm not careful I might learn something.

[jaws]

$this->bbcode_second_pass_quote('EnergySpin', 'B')ut jaws ... most of the recent Nobel prizes in Economics have been given for mathematical breakthroughs (Black Scholes formula for example is an SDE solution). And the logical method as you call it was the foundation of Jaynes approach to Bayesian probabilistc reasoning. The subjective valuations as you call them can be codified in utility functions (zero-one if you are a Bayesian) and then examine expectations combining both uncertain events and personal utilities (de Finneti's approach).

[nero]

Elijah,

I see a problem with your initial model for P'= f(D,S) It perhaps is what jaws is also talking about. Your model is trying to model the inefficiency of the market. If the market was efficient then there would never be any lag between demand and supply. Prices would move smoothly to correct for any changes in the curves. While this doesn't happen in the real world the problem is we don't know what D (Demand at any instant in time) really is. It is extremely hard to quantify. Perhaps you could set up an experiment where D and S and quantifiable and sufficiently different so that the driving force you are after is significant. However in the real world, markets are much more efficient and the driving force is relatively small and hard to measure. Therefore economists are left talking theoretically about the ideal situation where D and S are always equal by definition.

[nero]

jaws,

Economists can perform experiments. As opposed to say astronomers or geologists or anthropologists. Economists have the same ethical and financial problems performing experiments that doctors or ecologists have. They also do not want to harm the patient or experimental subjects. They also are dealing with enormously complex situations where to perform useful experiments it takes alot of money, time and resources to isolate for the parameter of interest.

It's not impossible and some economists are doing it. On the other hand making a facile mathematical model from historic economic data isn't likely to be very useful. Before the math has to come the hypothesis or else it is probably only curve fitting. The data is just too dirty to hope to glean data from it otherwise.

[MacG]

Not "jaws", but anyhow...

$this->bbcode_second_pass_quote('EnergySpin', ' ')But jaws ... most of the recent Nobel prizes in Economics have been given for mathematical breakthroughs (Black Scholes formula for example is an SDE solution).

[nero]

$this->bbcode_second_pass_quote('EJ', 'E')fficiency by definition is "achieveing a goal as cheaply as possible" (Colander, G4) There is no reason to suppose that perfect efficiency is even possible or that the model makes this assumption. I agree the actual working models will be difference equations but ODE's are the continuous analogs of difference equations, I am making no logical err by starting here.

[EnergySpin]

The scientific method IS valid even in fields that cannot perform experiments (like geology or astronomy). The kind of Statistics that is applicable there is Bayesian statistics and models are used as proxies for a complex reality which cannot be known fully (hence models are uncertain variables themselves). Using traditional i.e. frequentist statistics in those fields is at best a grave mistake. One does not have repetitive experiemnts and in many cases symmetry assumptions cannot even be madel.

Many of the objections that Jaws raised about the application of mathematical models in economics are identical to objections raised against the mathematization of medicine and biology. Objections that were raised then had to do with the complexity of biological organisms/humans, the fact that no two humans are identical etc. The debate about quantitative reasoning in medicine actually dates back 2500 years ago with people debating whether a logic-based or data-based approach is suitable for the field. The fact that the argument seems to have been settled in favour of an approach combining theoretical mechanisms of disease AND empirical measurements in an increasingly Bayesian framework does show that there are NO limits to quantitative reasoning. Maintaining that a logic/language based approach is not amenable to mathematization is incorrect. First of all traditional (i.e. Aristotelian logic) can be represented by a Boolean Algebra notation; the extention to situations of uncertainty IS the Bayesian paradigm and that has been rigorously proved by both Jaynes and De Finetti. If one argues that it is not possible to reduce a linguistic/logic based approach to mathematics then one is not using logic in the first place , thus making his or her statement either vacuous or extremely suspect.

John Kenneth Galbraith has actually explored the question whether economics is science or not in his book "History of Economics"; he openly admitted that in the majority of historical cases economics was the "scientific" cloak that was used by the power structures to maintain control. Both the "invisible hand" and the "socialist state" are examples of such constructs IMHO.

Regarding the use of equilibrium models. It is fascinating to me to watch economics go through the same transformation that quantitiatve physiology or quantitative medicine went through. First there were qualitative observations (i.e. Louis Paster's germ theory), then quantitative data (i.e. the German physiologist of the late 19th early 20th century working on the cardiovascular system) and equilibrium as well as steady state models followed by differential equation modelling (i.e. Hodkin and Huxley models of nerve cells) and stochastic approaches in the last 20 years (i.e. ion channel kinetics).
Sorry to say that the future of all sciences is mathematics. Unless one adopts a dualistic philosophy towards the nature of man, all reality is based on matter and its properties. Laws on higher levels of organisation are grounded AND restricted by physical laws operating in the lower level . Even though I do not expect to see a quantum dynamical description of the economy or one of my patients, it is useful to remember this simple truth. Of course If one accepts the existence of a soul that is not subject to material laws then there is no point in debating the whole issue at all. For whenever a dualist is cornered he or she will always evoke a law based on the "properties of soul or human nature" thus destroying any opportunity of a real scientific debate.

[nero]

$this->bbcode_second_pass_quote('EJ', 'T')he equation P'=k(D-S) lacks many real world properties but as a starting point is has the following. Price goes up when D-S>0 and it goes down when D-S<0. Howmuch price changes under these conditions has to do with elasticity so k is not really constant, but that would be the next step in building the model. No k does not have to go to infinity for D-S to be zero. In the real world D-S=0 is the equilibrium the market is seeking but never finds.