I read your comment with great interest as it reflects a lot of my reactions when I took economics, particularly from the way Mankiw teaches it; much like this blog post, he starts with roughly 10 "assumptions" that the course will make comprising general consensus opinions that most economists use to make their theories. These are such broad assumptions like "people always make rational decisions", "people always think at the margin", "the market is comprised of many different roughly equal players that can freely enter and exit", etc. which I found hard to swallow and made me wonder if there was going to be anything valuable in the conclusions. Once I got over that hump, I did find that economists have come up with some interesting ideas in their dabblings with logical purity; sure they only work correctly within the artificial sandbox, but once in a while things line up well enough with the real world that you can make fair comparisons. I'm sure that the more you delve into economics, though, the less overlap you will find between any "simple" model and the behavior of societies, since economies in the real world almost always act on a societal if not global scale, incorporating causality in a way that definitely evokes "the butterfly effect" or chaos theory more than a few silly equations...
Among your many colorful examples, I'm going to call out your transistor and grain market examples, because they seem to be a misrepresentation of supply and demand. S&D doesn't say that "if we buy more transistors the price will go up", that's a misrepresentation of the theory. The whole point of S&D is that there is a relationship between the two and you can only predict price shifts (and quantities purchased) by seeing how both behave, not just one, and certainly not by just looking at quantity purchased. So demand for transistors has picked up over the decades, but the supply side has responded at an even greater magnitude, owing to improved technologies bringing the cost of supplying transistors down. This is the reason price has shifted down while quantities have risen. If anything is pretty unassailable in micro, it's S&D in a fair market with commodified goods. The numbers there actually match reality fairly well, once you have data on the shape of each curve (ah, econometrics). And it is totally legit for either curve to evolve over time; that's part of how you assess how people will react to taxes, supply shocks, etc. I would fix your examples for this point, the others seemed more legit.
Silly ASCII diagram for the transistor market:
P P
|\ /S -->+4 | \ S
| \ / | \ /
| X ==time=> | \ /
| / \ | X
|/ \D -->+2 | / \D
---------Q ---------Q
++demand, ++++supply
=> quantity purchased increases
=> price goes down
As for how economics treats technology, one of the more interesting conclusions of Mankiw's macro syllabus is that in the long run, only greater efficiency can increase the rate of growth of GDP. Simply increasing the growth of labor supply, growth of money supply, or capital investments, or decreasing rate of aggregate consumption has no long-term effect. Efficiency is essentially output per worker, so things like education and research are apparently the best effectors for accelerating GDP (and therefore the most important to preserve long-term). I can't argue with the sensibility of that result ;-)
What I said about supply and demand curves is well within common presentations by econ profs.
What you did was patch up such presentations. There are many ways to do that, and with enough such patches the theory can have fewer counterexamples but, then, look so narrow it looks useless.
One way to patch up the theory is to say that transistors in 1955 are NOT THE SAME good or product as transistors in 2010. Then my data is not a counterexample, but the theory looks much more narrow.
You went a little farther and tried to relate the supply and demand curves for the two different points in time. Okay.
Similarly for wheat from 1800 to 2000.
Then I will say, even at one point in time, if I buy a wheat contract on the futures market, then I pay much less per bushel than in a feed supply store. So, again, the patch up is that the two sources of wheat do not supply the same same product. Take one scoop of each wheat and try to tell the difference and don't see any, but academic econ says that they are not the same. What's not the same is that their theory is not the same as the way prices work for real wheat.
Similarly volume discounts are common all across the economy. So, we have to say that if we buy Grey Poupon mustard in the big bottle at A&P then that is not the same Grey Poupon mustard in the small bottle at A&P. Now we are getting to be absurd. It's the same GD mustard, from the same tube, from the same vat, from the same plant. No chemist can tell the difference, but academic econ can! And they do it looking only at the bottle and never the contents. SUCH a science!
And, if I operate a fleet of trucks and want to order 100 from Ford, then academic econ has to say that those are not the same trucks I get if I order just two. E.g., for 100 trucks, Ford may put on an an extra shift to supply them. Their 'marginal cost per truck' goes down, and they let me have some of the savings. Besides, if they don't want to deal, then maybe Dodge, Chevy, or Toyota will. And there's MUCH more to the price than just the 'marginal cost' -- MUCH more. Easy it's not; the academic econ supply and demand curves are easy, and that's part of why they don't work.
So, have to strain to find where supply and demand curves actually describe the real economy.
Net, supply and demand curves need so many patch ups they are basically just nonsense. That's just not how the real economy works.
For optimization, the idea is that each shopper in a grocery store solves some complicated problem in nonlinear, integer, stochastic optimization just to get the groceries. This nonsense is part of their rational assumption. Intellectual self-abuse.
They want to apply this optimization also to the micro economics of the firm assuming that, of course, with optimal material requirements planning from, say, SAP software, that of course any significant firm actually does such things. No they don't. Sorry 'bout that.
The nonsense goes on and on: They are talking about their imaginary, "sand box" economies and not any real economies.
I have an imaginary "sand box" too: I'm young, rich, athletic, handsome, and all the girls like me! That's not real, either.
Apparently by the time you got to the course, the field had been hit over the head often enough with objections such as mine that they started off the course with a defense, a long list of assumptions. Yup, if pigs could fly, then cast iron umbrellas would sell well.
Physics, chemistry, engineering, medical science, and technology actually DO work very well and do not work anything at all like academic econ, and academic econ essentially just doesn't work in any comparable sense. That dog won't hunt. It's a late parrot. They are like medicine back in the days of snake oil -- they haven't made any real progress in their subject and just don't know what they are doing. And they are killing people and don't care.
The first and last thing I saw in economics that made any sense was the Leontief input-output matrix. For a one point in time, first cut of the macro economy, it made some sense. It was used in US WWII production planning. Alas, the econ profs hated any such things, and it lost traction in academics.
One of the worst sins is that they take their simplistic assumptions and the corresponding conclusions and say that this is the way that a real economy should be FORCED to work. Now they are killing people.
As we saw in the 1930s, have seen in the US since the 1950s, and see again now with the Great Recession, academic economics doesn't have a weak little hollow hint of a tiny clue about how the US economy works. Bernanke, Geithner, etc. have been struggling and just hoping. They don't know what will happen to employment, taxes, the deficit, interest rates, house prices, the stock market, imports, exports, oil, the dollar, etc. They just don't know, even roughly. They don't know where we will be in five years, in a world war like we were in the 1930s or growing nicely. They don't know. Yup, it's an invisible hand that no one in economics can see.
In particular, academic econ and even the Federal Reserve Board staff are just not in touch with the real economy. They driving a big truck looking only at their belly buttons and never out the window. If anyone had had any reasonable empirical description of the US economy, and any credibility, then they could have given some advice to Clinton or W or even Barney Frank, and certainly to AIG, Lehman, Bear Stearns, BoA, Citi, and the FDIC, on the clear and present dangers. But, no one did. In this age of computers, data storage, and Internet communications, no one has even the first-cut, basic data, data that we now know is absolutely crucial. We don't even have a basic, competent dashboard.
When I looked at academic econ, the first thing I was looking for was the first-cut, empirical description of the US economy. So, we need to know the inventories and the flows, of money, goods, people, etc. Then for the 102 version, we need the same for the world economy. Nope: Not there. No meaningful data at all. At least Chairman Greenspan commonly was up to his neck in empirical data; he was also up to his ears in Ayn Rand, smoking funny stuff, total, dangerous nonsense.
At one point I was ready to chip in and try to make some progress. Sure, I would have started with what it is said that Bernanke did: Get a LOT of numerical data, get a description, and then try to build some decent theories. But, for what is in academic econ, dump at least 99 44/100% of it in the trash. But the field was so full of their arrogant incompetence that it was clear that there was no hope of doing anything constructive within their field.
Bernanke SHOULD be the man of the hour since he did his research on the details of the Great Depression. Still, he is clueless.
Net, for the US and the world, the situation is clear: The economy is REALLY important, and academic econ is from next to useless down to really dangerous.
So, instead of academic econ, proceed VERY carefully, almost entirely just empirically. For econometrics, be careful, be very, very careful. Yes, there may at times be some cases where someone from academic econ, in spite of their field, actually has some positive contribution to make. Maybe. But their hands are dripping with blood.
I enjoyed your original rant against economists and the damage some of their bold but ultimately baseless claims can do, but many of your examples seem ill-chosen. It's really no more complex to incoporate bundling vs budget constraints and the time value of money into neoclassical economic models than it is to incorporate friction into a physical model. Appending "given no corresponding increase in supply" to a claim that the price of a commodity will rise given an increase in demand is really no more of a "patch up" than appending "given the absence of air resistance and other frictional forces" to a statement that the earth's gravitational field will cause a falling object to accelerate at approx 9.81ms^2. "Demand" being the level of quantity which would be purchased at a given price at a given point in time is as fundamental to microeconomics as the concept of net force to Newtonian classical mechanics.
Sure, a number of economists attach undue weight to some underlying assumptions which are only partly true, such as economic actors to some degree exhibiting "economically rational" (perhaps akin to a disciple Newton fan asserting the wave nature of matter has no meaningful implications for their bodies, this kind of abstraction is harmless when applied to sufficiently large numbers). But whilst this economic model appears to break down at levels as simple as individuals being ineffective at computing stochastic optimisations in their heads, it works rather well when analysing the behaviour of crowds. It's not so good at identifying when bubbles will burst, but then meteorologists can't predict the weather next year with any reasonable degree of accuracy either, and that's arguably a simpler system to model.
Where economists overstep the mark is not so much the tendency to oversimplify as the fusion of normative arguments with their models, such as the sleight of hand that dresses Pareto "efficiency" as a static optimisation problem rather than a rights claim made on behalf of the status quo.
I'm pretty sure I don't agree. Many of the fundamental assumptions in economics aren't just first-order approximations; they're about tractability. You can leave out friction and get reliable ballpark figures, but when you assume monotonicity, you're liable to end up wildly off base sooner or later.
That's what I consider the main problem of economics. To use the gp's expression, they have physics envy when they need to be demanding statistics and psychology. The problem area is difficult and overpopulated, and what's worse, worth money. That means that the most apt people do business for themselves instead of suggesting on the business of others.
Much of economics is useful, but little is reliable. As much as it pains me to say this, I believe we could do with more (quiet) focus on economic policies - just not from the math people, but from the social studies people.
> For optimization, the idea is that each shopper in a grocery store solves some complicated problem in nonlinear, integer, stochastic optimization just to get the groceries. This nonsense is part of their rational assumption.
I do this, and it's a real problem. It takes me forever to buy groceries.
Among your many colorful examples, I'm going to call out your transistor and grain market examples, because they seem to be a misrepresentation of supply and demand. S&D doesn't say that "if we buy more transistors the price will go up", that's a misrepresentation of the theory. The whole point of S&D is that there is a relationship between the two and you can only predict price shifts (and quantities purchased) by seeing how both behave, not just one, and certainly not by just looking at quantity purchased. So demand for transistors has picked up over the decades, but the supply side has responded at an even greater magnitude, owing to improved technologies bringing the cost of supplying transistors down. This is the reason price has shifted down while quantities have risen. If anything is pretty unassailable in micro, it's S&D in a fair market with commodified goods. The numbers there actually match reality fairly well, once you have data on the shape of each curve (ah, econometrics). And it is totally legit for either curve to evolve over time; that's part of how you assess how people will react to taxes, supply shocks, etc. I would fix your examples for this point, the others seemed more legit.
Silly ASCII diagram for the transistor market:
As for how economics treats technology, one of the more interesting conclusions of Mankiw's macro syllabus is that in the long run, only greater efficiency can increase the rate of growth of GDP. Simply increasing the growth of labor supply, growth of money supply, or capital investments, or decreasing rate of aggregate consumption has no long-term effect. Efficiency is essentially output per worker, so things like education and research are apparently the best effectors for accelerating GDP (and therefore the most important to preserve long-term). I can't argue with the sensibility of that result ;-)