[Dave]

Falx is certainly correct that mainstream economics relies on baseless axioms in order to build solveable mathematical models, but that doesn't necessarily refute Todd (who is not wedded to any isoquant fetish as far as I'm aware), nor can you "solve" an equation for a dynamic economy, although in many situations defining economic exchanges as Walrasian equations and finding the maxima and minima can be useful.

[Falx]

You used the analogy "Ship is sinking" and "Shit is hitting the fan" in the same post. That's what I was responding to.

I would call a recession that was due to market deregulation "a sinking ship" and the largest the largest bankruptcy in US history shit hitting the fan.

What are you basing this on?

Mathematical properties of differential equations, feel free to take issue with the stuff said here. Or buy a text book on them, I'm sure nets has suggestions more in tune with economics than the ones I use for physics, that however does not change the fundamental mathematics of the situation.

[Todd D.]

I would call a recession that was due to market deregulation "a sinking ship" and the largest the largest bankruptcy in US history shit hitting the fan.

So....you ARE saying that the sky is falling? If the ship is sinking, that means a collapse, not a mere recession, at least as far as I can interpret the analogy.

Mathematical properties of differential equations, feel free to take issue with the stuff said here. Or buy a text book on them, I'm sure nets has suggestions more in tune with economics than the ones I use for physics, that however does not change the fundamental mathematics of the situation.

What in the world are you talking about? Help me out here, seriously. What are you basing your claim that "the fact that economics almost exclusively uses paper and pencil models"? I don't particularly care that Differential Equations "exist". YOUR claim was that Economics fails to account for this. I want to know why you think this.

[Falx]

So....you ARE saying that the sky is falling? If the ship is sinking, that means a collapse, not a mere recession, at least as far as I can interpret the analogy.

The sky is falling is when several hundred million people are sure to die in the next few days/weeks/months. Do you think this is the case? If not then you've just answered your own question.

What are you basing your claim that "the fact that economics almost exclusively uses paper and pencil models"?

Compared to other disciplines who use computer models extensively economic models are extremely simple and primitive so much so that most of them can be solved by hand analytically.

I don't particularly care that Differential Equations "exist". YOUR claim was that Economics fails to account for this. I want to know why you think this.

Bloody hell, my point is that so that they can guarantee that their models aren't just parasitic solutions on a computer they need to keep the functions nice, that is no discontinuities etc as needed by the existence and uniqueness conditions. The assumptions needed for this make the models ridiculously restrictive and hence not able to fully describe reality. The alternative is that you can lump as much detail in as you want and you won't be sure that there is any solution. Not a problem for the sciences where you can just run the experiment again and again, a huge problem if you decide to run your economy based on a computer glitch.

I don't particularly care that Differential Equations "exist".

You've just disqualified yourself from conversations about anything more than qualitative ideas in any subject which uses mathematics.

[Dave]

Economics is not a natural science and strictly speaking doesn't require necessarily mathematics at all. Addressing the deficiencies in microeconomic modeling certainly doesn't refute other economic schools.

[Falx]

Economics is not a natural science and strictly speaking doesn't require necessarily mathematics at all.

Care to elaborate, I'm not sure I understand what you mean here.

Addressing the deficiencies in microeconomic modeling certainly doesn't refute other economic schools.

Every economic school as far as I know has, or can be extended to have, microeconomic predictions. Those predictions will enable models to be build given initial conditions. I guess you could build a theoretical economic model where everything is vague enough that you can't use a mathematical model but that would mean it has no predictive power and won't be at all useful.

[Dave]

Economics is not a natural science and strictly speaking doesn't require necessarily mathematics at all.

Care to elaborate, I'm not sure I understand what you mean here.

Economics is a social science. It is quite possible to approach it without the use of mathematics, much in the way other social sciences like sociology or philosophy are approached. Certain economic schools make almost no use of mathematics.

Addressing the deficiencies in microeconomic modeling certainly doesn't refute other economic schools.

Every economic school as far as I know has, or can be extended to have, microeconomic predictions. Those predictions will enable models to be build given initial conditions. I guess you could build a theoretical economic model where everything is vague enough that you can't use a mathematical model but that would mean it has no predictive power and won't be at all useful.

Some economic schools reject the micro/macro division entirely. Most heterodox schools do not use the kind of modeling you rightly criticize, although they may use mathematics in other contexts. Since economic forecasting requires analyzing human behavior, you can't solve an equation per se the way you can in physics. It can be much more beneficial even to rely on intuition.

Seasoned, skillful traders are usually the best economic forecasters around and more or less rely on gut feeling. The entire approach of "approximating" human behavior into a mathematic model relies on, as you say, unrealistic assumptions and the predictions are not useful. I could see math having great applications in economic planning (where planners control most of the variables), as within a socialist economy or within a corporation with predictable markets (rare since the early 1970s), but certainly not in dynamic economic forecasting.

I've been predicting this economic crisis since 2003 and have used no models. I had predicted it to arrive in 2007 rather than 2008, but whatever. This is due to my personal theory of business cycles and some math where I calculated debt/income ratios would become unsustainable.

[Falx]

Economics is a social science. It is quite possible to approach it without the use of mathematics, much in the way other social sciences like sociology or philosophy are approached. Certain economic schools make almost no use of mathematics.

That makes sense.

The rest is what I was talking about, it is possible to build a system in which you'll have a "recession" which will be "deep" and last "some time" with effects being "far reaching". Those statements are vague enough to be next to useless though if you want a quantitative answer to the question "how much money will I lose from this" that is unless you count in the traditional human of none, one, two, three, many.

[Dave]

I'm not sure how it would be possible to calculate that quantitative solution.

[Falx]

Me neither, that's why I'm a humble mathematical physicists in training.

[Dave]

To use terminology from biology, market economies operate on evolutionary principles. It is not possible to use mathematics to forecast outcomes in evolutionary systems. Command economies operate by intelligent design and can be analyzed mathematically with some accuracy.

[ThereBeDragons]

And yet people run biological simulations.

[Falx]

To use terminology from biology, market economies operate on evolutionary principles. It is not possible to use mathematics to forecast outcomes in evolutionary systems. Command economies operate by intelligent design and can be analyzed mathematically with some accuracy.

Intelligent design is jargon common to biology about as much as "Good day this is your captain, WE'RE ALL GONNA DIE, have a nice day." is common to airlines.

And yet people run biological simulations.

Those aren't economic models, I've seen and read papers about similar models used in economics, even chaos based ones, but they are mostly regarded as curiosities by academics and don't figure in policy making. Which is extremely unfortunate.

[Nets]

Or buy a text book on them, I'm sure nets has suggestions more in tune with economics than the ones I use for physics, that however does not change the fundamental mathematics of the situation.

I was taught Diff Eq from this textbook, which is for engineering, so I have a feeling its similar to what you learned Falx. Its a great text, IMO, but it really weak on PDEs. (Obviously, its not intended for those, but meh).



Let me just add that Diff Eq sucks balls.

Anyways, I agree with most of what Falx has said. The models are imperfect, but are not a lost cause. I believe a greater emphasis on incorporating greater historical empiricism can save them. You have to remember that in the 40s and 50s when people built these models, they didn't have access to the wealth of data sets we do today.

Also, I'd like to add that Mathematical Economics Econometrics, they are two different fields that have some overlap. Econometrics is fairly sound, IMO. I understand the criticisms against mathematical economics, on the other hand. They shouldn't be confused.

[Dave]

Alchemical formulae for the transmutation of base metals are imperfect, but not a lost cause.

In any case dynamic stochastic general equilibrium models address the faults Falx pointed out, but as they are microfounded they are of course still completely useless.

[Falx]

What's these fancy new editions? I learn from the good ol'e 2nd edition with its "new" Fortran code for machines that manage the astonishing speed of "several thousands calculations per second". God I hate my maths department, its like 1908, no calculating machines of any king are allowed in exams, even a vintage Curta (!), fucking hate it.

[Nets]

What's these fancy new editions? I learn from the good ol'e 2nd edition with its "new" Fortran code for machines that manage the astonishing speed of "several thousands calculations per second". God I hate my maths department, its like 1908, no calculating machines of any king are allowed in exams, even a vintage Curta (!), fucking hate it.

Heh heh FORTRAN. Lol. Only physicists use FORTRAN.

We also aren't allowed to use calculators. Although most math exams, they wouldn't help anyway.

I once brought a slide rule to a Linear Algebra exam (thinking I was being clever) and they almost kicked me out.