Econ 106I (Galles) – Lecture 2 (January 19^th, 2005)

Hayek – about central planning

Is efficient central planning possible?

No. It is logically impossible.

You cannot possibly centralize decision making due to informational considerations

Wealth creation is a function of details

Soviet cement industry – very inefficient transportation. They shipped way more cement than they needed due to uncertain quality.

A bottle of water has no objective value. It gets its value by what someone is willing to pay.

These details are crucial to every aspect of wealth creation.

A firm manager must centrally plan some things

Need to figure out how to make the most of your information

Two main problems: change, equilibrium

If you are in equilibrium, there is no change.

In the competitive model, we assume everything is in equilibrium.

**Diagram**

We assume transitions between equilibria occur

It takes time

We assume the equilibrating factor is price.

What if I have a lower delivery cost?

I can offer free shipping.

I have a comparative advantage.

Cutting prices is not the same as free shipping which is not the same as offering 90 credit without interest.

All the hard problems in life are caused by change.

In an old, peaceful, agricultural society, couldn’t you just memorize what your parents did?

With change, you cannot memorize your way to success.

This is why our education focuses on teaching you to acquire skills. Teaches flexibility.

An equilibrium approach captures none of the problems.

Under Hayek’s argument, why do we see firms?

They don’t centralize everything – the problems are reduced.

There are offsetting advantages – superior interior markets, team production, market costs

Criticism of Schumpeter – father of “process of creative destruction”

Assumption that what you know is that which is true. This is not accurate

Focus on details (application)

Why do we have models of growth and not of shrinking?

If you are a manager, aren’t you only fully using your resources when it is time to grow? Don’t you know what constraints are binding and what to buy to get around

them?

With growth, these details are easy enough to acquire.

How hard is it to figure out how best to shrink?

Much harder decisions. “Tech wiz example” – who is doing stuff that is not part of his/her job description? Details

Additional costs – people get mad at you for laying them off

Why are there underground city tours? As long as the city is growing, shouldn’t we fix potholes? What if the city is falling apart? Maybe it is easier to just tear it down than to fix it up.

Pave over the whole damned thing.

Shrinking – The details are difficult to acquire. Hire “corporate fixers.”

Don’t we assume homogeneous capital and labor?

This assumption removes the role of managers (p. 77 of Alchian)

Superior interior labor and capital market in light of heterogeneity.

If everyone is identical, there is no advantage of acquiring information about people.

Heterogeneity – It is advantageous to know where to place people.

Is the firm a production function? (p.82)

Are economic decisions important only once per long interval of time? Definitely not

We assume away the ability for manager to be productive (Klein, Crawford, Alchian)

Buy printing press (specialized versus flexible)

Specialized press puts you at risk. Flexible press affects your production function.

Long-run cost curve:

**Diagram**

**Diagram**

The lower envelope of the average cost curves is the LR cost curve.

But how do you know that you will produce q* forever?

LR cost analysis tells us to pick specialized.

Why might we see people choose flexible?

How do we get to q*? In present value terms, it might be more valuable to grow slowly.

If I keep you at the same job for 30 years, wouldn’t my cost curve rise over time? Managers try to keep these costs down.

Future technological change – we assume no future technological progress in our models.

“Turbocharging plants” technology – probably worth purchasing at large scales. What if a new, better technology arises in the near future?

This is what managers decide.

Tin example (p. 85) – used to be the most famous example. Now it is not even mentioned

Suppose one person knows of a new valuable use for tin or that an old supplier disappeared

All people know is that tin has become more valuable. This is the crucial detail anyway.

The central planner would need to know everything.

A central planner probably couldn’t even get the tin market correct. What about the market for aluminum (substitute), copper (complement) and normal goods in the tin producing area?

The market system works all this stuff out. The government would be unlikely to figure out all these ripple effects.

The market rapidly communicates the truth through market behavior.

Alchian, Chapter 1:

Is there any uncertainty in our models?

A model that assumes away uncertainty is not very useful.

Alchian Chapter 2: Search costs

If you assume away heterogeneity, you assume away the role of management.

Alchian Chapter 1: Published in a major journal.

First paragraph: I am going to throw away profit maximization.

For profit maximization to make sense, don’t you need no uncertainty?

What do you have to replace it with?

There are two useful purposes of profit maximization

“What if…” games

Useful for historical study – the uncertainty has already been resolved.

Not useful for telling you what to do, however.

Call a coin: Can you maximize profits while it is still in the air?

We assume the coin has already landed before the choice needs to be made.

He says his new approach is more powerful (p. 34)

It is straightforward to start with complete uncertainty and no motivation and then add foresight where it is realistic to do so.

The opposite approach – maximize profits from your certainty assumptions – must throw these away once you introduce uncertainty.

You can now do only things that you can actually do. This is why it is a more powerful approach.

Every micro book cites this article and then quickly goes back to the math.

Certainty equivalents – You need to know everything about the uncertainty in order to use certainty equivalents.

(pp 18-19) – Admit uncertainty and use mean – disregards variance

A certainty equivalent assumes the answer

Certainty equivalents:

**Diagram**

Make unbelievable assumption that you know the entire distribution

Assume Von-Neumann Morganstern utility functions over distributions

Pick 2 unbelievable assumptions to solve one other unbelievable assumption (using the mean)

We cannot observe these distributions

What if we maximize market share, sales, sales growth instead?

Is the problem the word “profit” or the word “maximize?”

You cannot maximize in a world of uncertainty

What about trial and error? 2 responses

1. Need to know if trial is a failure or success

How do you separate the effects of your policy from the fact that the world is changing?

You cannot iterate this

2. Nearsighted grasshopper problem.

**Diagram**

You will end up at the top of hill if you know which direction to move.

What if profit hill has a local maximum and a different global maximum? You might end up at the local maximum instead.

Hopping towards the local maximum is not optimal.

You cannot sneak profit maximization in the back door.

Alchian is either a half century too early or a half century too late.

Biological evolution models.

Can’t randomness explain success?

What is survival? Realized positive profits.

(p 16) – Where foresight is uncertain, profit maximization is meaningless as a guide to optimal policy.

Adaptation and imitative behavior.

Why doesn’t profit maximization tell you what to do?

Each action translates into a distribution of possible results, some of which may overlap.

**Diagram**

Japanese green field example – build brand new, huge factories

Export their way to success – this worked extremely well.

Did anyone know this was a good thing to do? No!

Why was this successful?

Technology increase during war by leaps and bounds

New plants are far better than old ones.

Huge scale production was available

We did not go back into depression

Trade barriers were lowered

Why didn’t other countries do this? They didn’t make the bet.

Why were there so many conglomerates (a firm which is a combination of firms) in the Soviet economy?

What if everything had price ceilings? Does your money have any value if you can’t spend it?

Your firm as an advantage in bartering with other conglomerates.

What if we suddenly eliminated price controls?

Conglomerates may no longer be the best way of doing things.

Does profit maximization make sense in a world of certainty? No

There would be no profit in a world of certainty.

Economists are better at telling you what not to do than what to do.

i.e. Don’t take sunk costs into account.

Is diversification the way to maximize profit? No. Is it the standard advice? Yes. The standard solution doesn’t maximize profit. It reduces risk.

Oligopoly theory – Do you know how to maximize profits? You need to understand strategic interaction and you need more information

But this isn’t only the case for oligopolies. You always need more information.

Why doesn’t Saudi Arabia cheat in OPEC? They are too big for it to go unnoticed.

There are lots of things that are beyond your control.

Necessary condition for survival – positive profits.

There are only two types of investors: smart and unlucky.

Should you trust someone to be right on the eleventh time?

If you are surviving and you are doing behavior that does not follow my model, who is wrong? You or my model?

A firm in perfect competition does have room for error

What if you are competing with a lot of bad competitors?

Can four competitors be more competitive than 40 competitors? Of course. It depends on the quality of competition.

Separation of ownership and control. Managers manage for their own interest – disadvantage for corporations.

What if there are other benefits of corporations?

(p. 21) – Biology analogies. White butterfly in London analogy.

If you do something that was randomly smart, others will imitate.

Asymmetric bird – males are getting more asymmetrical (bigger right wings). Why?

They fly counterclockwise when fighting each other for mates.

Economists can still say useful stuff about who will survive even if no one is planning.

Chicago travel story.

We could discover the best route by taking risks.

What do you “keep” doing? Only those things you are successful at.

(p. 22) – Price discrimination – Look for differences in elasticities.

Nicholson assumes you do this correctly

You don’t need to do this perfectly for it to be successful.

Difference in prices between lunch and dinner – price discrimination.

They do this imperfectly – people can arbitrage (store it in the fridge and microwave it later)

(pp23-24) – Borel picker

Where there is uncertainty, people’s attitudes will differ.

Disagreement among managers is not consistent with profit maximization in a certain world.

Management is judgment – everyone will make mistakes.

Why is it possible to disagree? No one knows what the ultimately important variables are

Different time-preference rates

Only in a world of uncertainty does it make sense to disagree.

Inconsistent with standard model

(p. 26) – Representative firm. Biology analogies versus physics analogies.

Marshall looks at an average firm to think about general industry-wide problems, but no individual firm is like this.

Say I own 1000 acres of forest. Is any tree in equilibrium? Can’t I describe an average tree to describe the forest, though?

Physics analogies require that everything is the same as the average one. A typical firm is not any particular firm.

Physics can analyze individual behavior, but economics has to make far too many assumptions in order to do so as well.

Read all of Alchian 2 and Coase.