## The Correct Way to Teach Economics

I have previously tried to emphasise that not everything taught on an economics course is worthless and needs to be abandoned. Here I’m going to take two examples of models/theories you are taught, one that is taught scientifically, and one that is taught, erm, economically. This may be slightly zzz for some but it is important, particularly for trying to get some mainstream economists on my side (assuming that’s possible).

The first, ‘good’ example is how income is related to consumption.

You are presented with a hypothesis: people do not consume all of their income (Y), which can be denoted the Marginal Propensity to Consume (MPC). You are also told that people have to spend some money to stay alive, which can be denoted autonomous consumption (Ca). So total consumption is equal to:

C = Ca + MPC*Y

You then look at the evidence, which suggests this is roughly accurate, but not entirely. You are told this must be explained somehow, and presented with the Permanent Income Hypothesis, which is incorporated into the equation thusly:

C = Ca +MPC*Yp

Where Yp = Permanent Income or Average Lifetime Income

This appears to explain the data better, but still not completely. So you must then incorporate inflation/money illusion, interest rates, credit availability and so forth. This continues as long as is practically feasible and until the models are satisfactorily accurate.

Contrast this with demand-supply, where the following methodology is used:

“Students, this is a market. When demand does x, y happens and when supply does z, w happens (P.S. Adam Smith).”

“There are some assumptions we have to use to get it, though. Here they are:

Perfect information

Perfect competition

Homogeneous products and preferences

Perfect mobility of goods/services

People are rational utility maximisers

(probably some more I’ve forgotten)”

This isn’t an exaggeration. Economists like to suggest that these assumptions are simplifying but students find it much easier to follow the first example than the second, for understandable reasons. I’m not exactly sure what would replace the demand supply diagram, and I do agree that it captures some obvious truths at its heart. But the above is simply not an acceptable way to teach science.

1. #1 by Rest Assured on February 6, 2012 - 3:52 pm

The Permanent Income Hypothesis is still a neoliberal theory, which explains why economists were so quick to take it up. Friedman’s idea that temporary changes in income have small effects on current consumption has been used by right wing politicians to justify not temporarily cutting the income taxes of the lowest paid people in the country, or temporarily raising unemployment benefits to stimulate consumption spending to get the economy going. Keynesian consumption theory was dangerous for economists because it could (and was) used to justify progressive policies, which alarmed the ruling elite whom economists served, which is why Friedman came to their rescue with this right wing theory to justify laissez faire.

• #2 by Unlearningecon on February 6, 2012 - 6:03 pm

Yes it does have RW echoes about it, though the data and introspection seem to square with a weaker form of it.

2. #3 by Ruskin on February 6, 2012 - 4:14 pm

You’re right that the demand-supply framework is not an acceptable way to teach science. But it’s been taught because economics is not, and has never been a science, but simply a particularly pernicious ideology developed by powerful people to provide an “intellectual” defence of the status quo. This is the purpose of demand and supply diagrams. That simple diagram describes a world where competing needs and wants are effortlessly coordinated by the magic of the “free market”, delivering an equilibrium where no one has an incentive to change his decision. This make-believe world of Demand and Supply has been used to show how unemployment shouldn’t exist, and the only reason it does is because of trade unions or minimum wages, and how food shortages are caused by price ceilings. Steve Keen demolishes the mathematics behind demand and supply curves, while psychologists have destroyed the assumption that individuals are rational utility maximisers. And yet economists keep playing around with these toy models, completely oblivious to the damage they’ve caused around the world.

• #4 by Unlearningecon on February 6, 2012 - 6:05 pm

Well, yes.

• #5 by Isaac "Izzy" Marmolejo on February 8, 2012 - 4:37 pm

Ruskin,
I agree that “economists keep playing around with these toy models, completely oblivious to the damage they’ve caused around the world”, but
simple models are in no way a sole defense for a free market. This is a poor argument to make at best. “The same theory that describes the efficient nature of perfect competition also suggests that in many, perhaps even most, instances perfect competition doesn’t apply. And if one takes a general equilibrium view of market economies, it is clear that full equilibrium could never be obtained by a free market. The reason is the widespread existence of market imperfections such as monopoly power, imperfect information, incomplete markets and externalities that all lead to market failure.”*

*From Karen Vaughn’s ‘Economic Policy for An Imperfect World’

• #6 by Unlearningecon on February 8, 2012 - 5:24 pm

The thing is I’m not sure why economists still use the model of perfect competition when they know it rarely applies in the real world. The two examples I can think of are a market like Koh San Road in Bangkok, where prices are negotiable, and eBay. But other than that I’m even beginning to question the existence of an entity called ‘the market’ in modern capitalist economies, which are characterised by massive top down corporations worth about half of the average economy.

I object to the ‘this is an ideal condition, but the real world deviates from it’ mentality.

3. #7 by Isaac "Izzy" Marmolejo on February 10, 2012 - 8:43 am

Unlearningecon,

“But other than that I’m even beginning to question the existence of an entity called ‘the market’ in modern capitalist economies…”

I am interested what you mean by this (I do not exactly see where are you going with this).

• #8 by Unlearningecon on February 10, 2012 - 12:20 pm

My undeveloped view in this is probably better outsourced to Bruce Wilder, who has two excellent comments:

• #9 by Guillermo on February 10, 2012 - 5:56 pm

I agree with you about the quality of Bruce Wilder’s crookedtimber comments – I’ve long been a fan (there are many other good ones in the archives). Actually, I think comment number 15 in the “Colin Crouch” discussion is even better than #4.

Someone should convince him to write a book :).

• #10 by Unlearningecon on February 10, 2012 - 6:02 pm

Wow, yes it is. In fact that was the one I had in mind – I read it a while ago and it has stuck in my mind. #15 is recommended.

• #11 by Isaac "Izzy" Marmolejo on February 10, 2012 - 6:45 pm

Unlearningecon,

Im still confused on what he is saying. I mean the way I interpret it is that he is making the ideal state that neoclassicals use equal to the actual market. And anything deviating from this ideal state, thus the ‘market’, shouldn’t be considered a market phenomenon.

I agree that ‘market’ explanations are somewhat inconsistent in a neoclassical sense. They talk about how a ‘market’ price is determined, how there is a ‘natural’ rate of X, what are the ‘market’ costs of X etc. and they can say such things because they are assuming perfect knowledge , perfect mobility, no uncertainty etc and in reality, this is just not the case. But it does follow that because this ideal market doesn’t exist that the market as a whole doesn’t exist. It can be the case that neoclassicals have a flawed definition of what the market is and does in reality though. But then this was Ludwig von Mises’ basic point when he claims the market as a process and not a place, thing, or entity.*

*Ludwig von Mises, Human Action p 258

• #12 by Isaac "Izzy" Marmolejo on February 10, 2012 - 7:11 pm

Sorry,
“But it does follow that because this ideal market doesn’t exist that the market as a whole doesn’t exist.”

should be

“But it does NOT follow…”

• #13 by Unlearningecon on February 11, 2012 - 4:49 pm

I don’t think he’s saying that markets don’t exist because they aren’t ideal; rather, that the way the real economy behaves isn’t analogous to millions of individuals expressing certain preferences but to various rules and large powers interacting to produce certain states.

I mean, look at the ‘market’ for phone contracts. Whilst there is choice, the choices are relatively limited and the way phone contracts are bought and sold does not seem to resemble a spontaneous, amorphous information processing system but one characterised by fairly rigid rules and conventions.

4. #14 by JazzBumpa on February 12, 2012 - 7:51 pm

Unlearningecon -

Interesting post. Maybe you can help me understand this:

Where Yp = Permanent Income or Average Lifetime Income

This appears to explain the data better,

What is this data, and where can I find it? Looking at what I can find regarding income and consumption (U.S. data from FRED), I find an absolute lock-step correlation, at both gross and fine detail levels.

http://www.angrybearblog.com/2012/02/income-and-consumption.html#more

I’d really like to get this right.

Cheers!
JzB

• #15 by Unlearningecon on February 13, 2012 - 5:21 pm

Hey,

Our results are not really conflicting – income alone is *very* close, but certainly for the UK, incorporating permanent income improves the fit somewhat. See here for standard absolute income:

http://img840.imageshack.us/img840/8632/consumption1.jpg

and once a variant of the permanent income hypothesis is incorporated:

http://img696.imageshack.us/img696/6349/consumption2.jpg

I’m afraid I don’t have the details of this data as it was simply given to us in this form. I appreciate that’s not particularly helpful!

• #16 by Min on March 8, 2012 - 5:05 pm

But isn’t this data population data? How do you differentiate between averaging across a lifetime from averaging across a population, with people at different stages of life?

• #17 by Unlearningecon on March 10, 2012 - 4:17 pm

I think it’s weighted according to age demographics.

• #18 by Min on March 10, 2012 - 5:55 pm

Well, weighting may make the data more representative of individuals, **if the group behavior mimics individual behavior**.

There was a simple psych rat running experiment where rats ran a T maze, and to get fed they had to turn in the same direction at the T every time they ran the maze. A graph of the data showed a nice, smooth curve starting at 50% correct turns, gradually increasing to 100%.

At first this was taken as evidence that rats did not have insight or jump to the conclusion that the food was always on the right (or left, depending on which group they were in), but gradually changed their behavior. However, taking a look at the data for the individual rats showed that, in fact, they made instantaneous or almost instantaneous changes in their behavior. Because different rats made these changes after a different number of trials, the population data made a smooth curve that did not reflect the behavior of individual rats.

Same kind of problem here, I think.

5. #19 by yorksranter on February 13, 2012 - 11:26 am

..and one in which the actual price of the goods and services involved, and even their ownership status, is pretty much completely opaque to the customer.

6. #20 by Min on March 7, 2012 - 9:56 pm

Scuse me, but you went a little quickly.

“You are presented with a hypothesis: people do not consume all of their income (Y), which can be denoted the Marginal Propensity to Consume (MPC). You are also told that people have to spend some money to stay alive, which can be denoted autonomous consumption (Ca). So total consumption is equal to:

C = Ca + MPC*Y ”

Err, where did the Ca come from? Is it not part of Y? From your description I would have expected this equation:

C = Ca + MPC*(Y – Ca)

“You then look at the evidence, which suggests this is roughly accurate, but not entirely.”

But isn’t the equation a definition? What is not to be inaccurate? Is it assumed that MPC is a constant?

Thanks.

• #21 by Unlearningecon on March 8, 2012 - 12:29 am

The idea is that people can spend more than their income by borrowing or eating into savings, though of course that depend son your definition of ‘income’.

• #22 by Min on March 8, 2012 - 1:50 am

Sure, people can and do spend more than their income. But you can’t assume that, right? And that spending is not necessarily what I would call money to stay alive. It sounds like you need more variables and a disjunctive relation. No?

• #23 by Min on March 8, 2012 - 5:07 pm

I know this is not what you mean, but it sounds something like

Ca = what people borrow to stay alive

• #24 by Unlearningecon on March 8, 2012 - 5:22 pm

Ca is by definition what people have to spend to stay alive. If they have no income they can borrow, eat into savings or be given money by relations/friends. This spending does not vary with income. Of course, it’s difficult to separate exactly what portion of income is autonomous, but it’s a useful approximation and appears to fit the data well.

This:

C = Ca + MPC*(Y – Ca)

Could be useful with the qualifier that Y>Ca

• #25 by Min on March 8, 2012 - 6:19 pm

“C = Ca + MPC*(Y – Ca)

Could be useful with the qualifier that Y>Ca”

A disjunctive relation.

Thanks.