Examples of failures of the free market

[innonimatu]

Then who implements social welfare?
Not centrally planned government! I think you might be interpreting "social welfare planner" differently than we mean in economics. In the field, such a planner is just used to talke about a centrally planned economy where the goal is to maximize utility. In the perfect world given full information on all aspects of the economy, no shocks, and a static condition, the socially planned economy will equate to the free market economy. Note the bolded part, which violates reality in each condition. Here there is a mathematical property that shows that central-planning can never produce an optimal equilibrium. That's why the math is so useful, because well, you can't politicize calculus!

Oh, but you can! You just make it complex enough that other people will not look into the details, and then claim that it proves something.

Consumer utility curves make some sense for any given individual, but are useless if you try to aggregate them. Which basically means you can't do anything useful with a model based on utility curves.

Private firms generically seen to maximize profit, but the way the go about doing it is too complex to model, so most economists simply assume certain conditions for firms, in particular a minimum for marginal costs, which make no sense in the real world.

 

[Cutlass]

5. Markets where property rights have not been assigned (the Coase problem)

I wanted to ask about this little part. How common is this, and what would be examples?

 

[Hygro]

Well that makes more sense. I generally don't think of one system's success as the failure of another, but for the purposes of this discussion I suppose it doesn't matter.

Valid point.

 

[JerichoHill]

Oh, but you can! You just make it complex enough that other people will not look into the details, and then claim that it proves something.

Consumer utility curves make some sense for any given individual, but are useless if you try to aggregate them. Which basically means you can't do anything useful with a model based on utility curves.

Private firms generically seen to maximize profit, but the way the go about doing it is too complex to model, so most economists simply assume certain conditions for firms, in particular a minimum for marginal costs, which make no sense in the real world.

Err, actually, for private firms in perfect competition, that would be marginal cost equal to marginal revenue, which makes sense if you use your intuition.

As for politicizing calculus, you brought up how statistics can be manipulated by the end-user. But, the end-user cannot make 1+1 = 3, so somewhere, there is some truth.

And as for utility curves, well, if you agree that you cannot aggregate people's preferences to pareto optimal conditions, then point of central planning is?

 

[JerichoHill]

I wanted to ask about this little part. How common is this, and what would be examples?

Property rights not being assigned? Very common!

 

[Cutlass]

Property rights not being assigned? Very common!

How so? examples? I can't think of any.

 

[innonimatu]

Err, actually, for private firms in perfect competition, that would be marginal cost equal to marginal revenue, which makes sense if you use your intuition.

As for politicizing calculus, you brought up how statistics can be manipulated by the end-user. But, the end-user cannot make 1+1 = 3, so somewhere, there is some truth.

I'm not complaining here about statistics being manipulated. That, at least, tends to become obvious. I'm complaining about the representation of reality being manipulated. Theories that supposedly describe how things work, and prescribe actions, and which are absurd. But their absurdity is so well lidded that most people never see it.

Yes, marginal costs equals marginal revenue, perfect competition, etc, is the standard formula for the behavior of firms. And what lies behind that? An assumption of rising marginal costs, after some minimum. In the real world there is no such minimum, any given industry can produce a larger number of units cheaper. But that'd be troublesome, because then marginal cost and marginal revenue would never meet... in the real world it's falling prices (and therefore marginal revenue) which constrain production, not rising marginal costs. The "perfectly competitive" condition is there to ensure that demand and supply are independent. But either a firm can freely increase production (without meeting any rise in marginal costs) or, if it is constrained because it is using some finite resource, that means it is so important on that particular market (for the goods it produces) that perfect competition and the independence of supply and demand no longer hold.

And as for utility curves, well, if you agree that you cannot aggregate people's preferences to pareto optimal conditions, then point of central planning is?

Who talked about central planning? Governments intervene in markets according to circumstances, not according to any central plan or to some economic theory. That's what economists don't get, probably because it'd huts their self-esteem to admit that their most elaborate theories are next to useless in practice.

 

[LightFang]

There are models and models... an example: governments must model what they'll collect in taxes, well budgeting (well, they should... nowdays they just borrow, it seems).

We must always use models. What we must not do is enshrine models - we must be aware they're approximations, always, and only valid for the circumstances for which they arose - usually in a certain place and during a limited time frame. Many economists, however, keep pretending there is some holy grail of economics, some perfect and eternal economic model, as if society could be mathematically described and ordered, now and forever.

What's really dangerous is the presentation of any one economic model as the one good way of social organization. "Economics" is not separable from the whole of society, and whoever tries to impose a particular economic model is always furthering some political agenda. Enshrining a model as the one true model is just a strong form of propaganda.

The guy who wrote the book... does he have some problem with the use of normal distributions in statistical analysis of real world data, or what? The problems with that have been very well known for a long time, and still we must use it under many circumstances. Any good boon on numerical analysis discusses the subject.

Actually, I really do agree with this post a lot. I think you hit everything right on the money. Except for the part that we must use it. See, it doesn't matter if you find thousands of instances where you can use the Gaussian curve; all you need is one instance to disconfirm it.

The five largest movements in the markets have all been, like, one day events. The Gaussian does not allow for these incredibly rare events since the law of large numbers is supposed to cancel things out.

That's not to say that it's not useful in some circumstances, but it's rather misleading in a lot of areas where it's applied.

I disagree. Just because the macro economic models for stabilization policy are imperfect, doing noting is still the worse approach. Experience shows that not stabilizing the economy results in economic collapses.

I think that Taleb would say then that, as a skeptical empirist, if doing stuff proves to fix things, then stuff should be done. What he does not like absolutely is molding the facts to fit the theory. But if this really is true (and I don't doubt it), then yeah, do things. But also try to find better theories and avoid the trend in macroeconomics to just have grand theories and no data to support it.

 

[Bill3000]

How so? examples? I can't think of any.

China and its pirating problems?

 

[Ecofarm]

Examples of failures of the free market:

1. Presumes an informed consumer. Who would have thought that at some point in the future, we would consume so much crap that we don't bother to research/investigate/read about most of the crap we buy.

2. Externalities, aka The Commons being finite.

 

[aelf]

Egalitarianism does not equate to socialism. The "market" is a bunch of math about how firms and consumers make decisions. Whether it is centrally planned or decentralized into a "free market", consumers still have utility curves of consumption and leisure, and firms still produce items given technology, capital, and labor. You still have the requirement that whatever a firm produces is tat least worth as much as it cost to put together. You still have the requirement that the consumer will only work as much as they prefer.

Egalitarianism can exist within a free market framework. The two are not mutually exclusive, and its a mistake to assume such. Pick up a graduate level Macro book and look at a simple 1-period model of an economy and you'd see the mechanisms that would allow for such. If you've taken enough math classes, that is.

I won't really comment on the curves and whatnot, since you seem to have a discussion going on about those already. And, obviously, I do not have graduate level training at economics and would likely not be getting it. But before you say I'm not qualified enough, I do have some understanding of government (which I'm aware many Americans simply assume to be bad) and I can say that it is a necessary complement to economics in the real world.

Anyway, I would just like to point out that, from what I know, economic theory inevitably has to operate on a ceteris paribus basis. You can create models with many different factors, but the models become more cumbersome and less reliable. And the world is not a ceteris paribus place. That's a problem. But I think this has also been raised.

It's also ironic that critics of central planning who point out that it's impossible to accurately predict market behaviour are themselves claiming to be able to do so to some extent. So what's the deal? Can market behaviour be modeled at all or not? If you guys can make assessments and predictions, why can a central bureau not make them? What makes extensive government intervention necessarily bad, if it is able to do as competent a job as private consultants?

You further assume that the political economy would properly sort preferences and generate an optimal outcome. Sadly, this has been proven wrong, political governance never creates a pareto optimality (regardless of system)

That's a pointless argument unless you want to claim that pareto optimality can be achieved.

Not centrally planned government! I think you might be interpreting "social welfare planner" differently than we mean in economics. In the field, such a planner is just used to talke about a centrally planned economy where the goal is to maximize utility. In the perfect world given full information on all aspects of the economy, no shocks, and a static condition, the socially planned economy will equate to the free market economy. Note the bolded part, which violates reality in each condition. Here there is a mathematical property that shows that central-planning can never produce an optimal equilibrium. That's why the math is so useful, because well, you can't politicize calculus!

I don't know how much you know about politics and government, and I wouldn't presume that you don't know enough to do your job. However, whatever the math says, which I'm not totally ignorant about, there are some truths in this world that politics certainly acknowledge. First, human beings are not purely rational creatures. Second, there are human values other than profit maximisation that people actually do aspire to. Third, even given purely rational actors, there would not necessarily be ideal results, as demonstrated by market failure.

You might say that the irrationality is taken into account and rationality includes government intervention, where utility is weighed and decisions made based on rational assessments. But your argument would then become weak, in no way proving that the free market is superior. In that case, a more centralised economy might just be as good or better.

 

[Cutlass]

China and its pirating problems?

Not at all. See the property rights are clearly defined, just lack an enforcement mechanism. Now it's been many years since I read Coase, but I never did get how a property right is entirely undefined. Either someone owns it, or the government has control of it.

 

[JerichoHill]

@@=innonimatu

Yes, marginal costs equals marginal revenue, perfect competition, etc, is the standard formula for the behavior of firms. And what lies behind that? An assumption of rising marginal costs, after some minimum. In the real world there is no such minimum, any given industry can produce a larger number of units cheaper.
Inno, I am fairly certain that your conjecture here is false. You're assuming that there are economies to scale across every industry, which isn't simply true. A simple example is the concrete industry. Further, as an industry grows in relative size and scope it must bid labor away from other industry, thus increasing the cost of labor. A industry that consumes more power relative to others will face a higher cost of power at some point. These mechanisms work in non-perfectly competitive environments as well. I'm really struck here by how you're assuming that firms and industry face constant or decreasing returns to scale across the range of Q (0 to infinity). That doesn't pass a reality test!

Who talked about central planning? Governments intervene in markets according to circumstances, not according to any central plan or to some economic theory. That's what economists don't get, probably because it'd huts their self-esteem to admit that their most elaborate theories are next to useless in practice.
I'd argue the Fed has a semblance of a plan. Governments intervene in the economy for a variety a reasons, not just circumstances. If economics as a discpline is useless, then why would firms who want to make money, or governments for that matter, hire them? You can't have one scenario and the other at the same time.

 

[JerichoHill]

Actually, I really do agree with this post a lot. I think you hit everything right on the money. Except for the part that we must use it. See, it doesn't matter if you find thousands of instances where you can use the Gaussian curve; all you need is one instance to disconfirm it.

No, at that point you know that you have a USEFUL model, but it has to be applied in the right circumstance. A model that is accurate 99% is pretty darn good, and to expect perfection is ludicrious. There are always black swans

 

[Cutlass]

I think that Taleb would say then that, as a skeptical empirist, if doing stuff proves to fix things, then stuff should be done. What he does not like absolutely is molding the facts to fit the theory. But if this really is true (and I don't doubt it), then yeah, do things. But also try to find better theories and avoid the trend in macroeconomics to just have grand theories and no data to support it.

The Monetarist movement molded facts to fit their theories. The more mainstream overlooked a number of things at different times, but when it came down to a failure they tried to figure out why. Still there are somethings in macro economic policy that are pretty straight forward and used often. You have inflation, raise interest rates. You have recession, lower them. You have both, pick a priority.

 

[JerichoHill]

@@aelf

I won't really comment on the curves and whatnot, since you seem to have a discussion going on about those already. And, obviously, I do not have graduate level training at economics and would likely not be getting it.
If you REALLY want to get a leg up in understanding how this world operates, I MUST encourage you to get some higher-level mathematical training. It's worth it, for work and for managing your own life/finances.

But before you say I'm not qualified enough, I do have some understanding of government (which I'm aware many Americans simply assume to be bad) and I can say that it is a necessary complement to economics in the real world.
I never claimed government was bad. So I don't get why you're saying this.

Anyway, I would just like to point out that, from what I know, economic theory inevitably has to operate on a ceteris paribus basis. You can create models with many different factors, but the models become more cumbersome and less reliable. And the world is not a ceteris paribus place. That's a problem. But I think this has also been raised.
No, economics doesn't need to operate on ceteris paribus levels. Economics that is taught for undergraduates assumes ceteris paribus so that we can teach concepts. I can guarantee you that on day one of a PhD program, ceteris paribus is thrown out the window and into the garbage. Ergo, your arguement is invalid.

It's also ironic that critics of central planning who point out that it's impossible to accurately predict market behaviour are themselves claiming to be able to do so to some extent.
No such claim is being made. The claim is made that in a simple model case, the social planner would, in effect with perfect information, allocate exactly as a market would.

So what's the deal? Can market behaviour be modeled at all or not?
Yes, it can be modeled in an approximate fashion.

If you guys can make assessments and predictions, why can a central bureau not make them?
Does a central bureau have information on every single consumer, firm, their budget, utility function, production function? Is that even possible? No. Ergo, if we let individuals sort themselves, with the guesstimate that individuals will attempt to optimize their own situation given what the world is like, then they'll likely do a much better job than a few folks sitting in desks thousands of miles away would do. History shows us that extreme central planning always has led to economic maliase. For a US example, see the effects of Nixon's wage and price controls.

That's a pointless argument unless you want to claim that pareto optimality can be achieved.
Theoretically yes, realistically, probably not.

First, human beings are not purely rational creatures.
Economics can assume irrationality. Again, the practice of economics ISNT AN INTRODUCTORY ECONOMICS TEXTBOOK
Second, there are human values other than profit maximisation that people actually do aspire to.
Sure, and those can be considered part of our "utlity" function. That again, isn't a problem in economics.

Sigh. I just wish you folks wouldn't assume that your high school or basic college economics course actually represents economics as a practice. Just like any other discipline's intro courses, you get the 40,000 ft level. Practice operates at the 6 foot level.

 

[fishjie]

this whole subprime crisis. this is why there always needs to be some regulation. because otherwise, the irresponsible actions of predatory lenders (and incompetent greedy homebuyers) will ruin the ENTIRE economy. With companies this large able to bring the market to its knees, you must have protections in place. This is why the SEC was created after the great depression. lets hope something similar happens after this mess is cleared up.

 

[innonimatu]

@@=innonimatu
Yes, marginal costs equals marginal revenue, perfect competition, etc, is the standard formula for the behavior of firms. And what lies behind that? An assumption of rising marginal costs, after some minimum. In the real world there is no such minimum, any given industry can produce a larger number of units cheaper.
Inno, I am fairly certain that your conjecture here is false. You're assuming that there are economies to scale across every industry, which isn't simply true. A simple example is the concrete industry. Further, as an industry grows in relative size and scope it must bid labor away from other industry, thus increasing the cost of labor.

If you are talking about a huge industry (an whole industrial sector, for example the concrete industry as a whole), then that is true. But out goes "perfect competition" and out goes the independence of supply and demand - if they influence the market enough to change labor costs they'll also shift demand (and not even only for their product but for all products). And this means you can no longer calculate marginal returns and solve the MR = MI equation. You'd violate other fundamental conditions of the standard model of the firm.

By the way, the concrete industry (I assume you mean cement) is slightly limited by costs of transportation only (both of raw materials and products). And I stress slightly as it's common to ship cement across large distances (Turkey to France is common). And even clinker across the Atlantic. They don't really have increasing marginal costs, they have difficulties selling the product past a regional market of a few thousand kilometers. They can still monopolize that regional market with no problem, always with falling marginal costs.

A industry that consumes more power relative to others will face a higher cost of power at some point. These mechanisms work in non-perfectly competitive environments as well. I'm really struck here by how you're assuming that firms and industry face constant or decreasing returns to scale across the range of Q (0 to infinity). That doesn't pass a reality test!

And I'm struck how you assume they face increasing costs. We're not talking about infinity here, we're talking about control of the whole market targeted by the firm (monopoly). Or at least of a large enough fraction that this particular firm (for we're talking about firms, not industry segments) can influence prices. Once that happens perfect competition for this market is gone, and this (monopoly) can easily be achieved, profitably, by a firm.
That's another problem with standard economic theory: the mantra "monopolies are bad". Yes, they are, because they offer a single firm the power to charge higher prices, much over production costs. But they're efficient, as far as costs are concerned.

All this is not even new, the failure of the neoclassical theory of the firm has been known for 83 years, ever since Piero Sraffa published a demolishing critique against it. Yet it is still taught in schools. Pretty and simple mathematics are more important than reality...

I'd argue the Fed has a semblance of a plan. Governments intervene in the economy for a variety a reasons, not just circumstances. If economics as a discpline is useless, then why would firms who want to make money, or governments for that matter, hire them? You can't have one scenario and the other at the same time.

If the Fed had a plan the LIBOR for dollars wouldn't now be twice the "official" rate. The fed can't even lower rates any more, because that'd be ignored by banks, and expose the fact that monetary policy is now out of control.

In fact, if the managers of these firms had a plan, or the government and the regulators (are there any left?), the debt craziness wouldn't have happened, would have been prevented, and the largest banks, insurers of the world wouldn't be collapsing. The funny thing is that this situation was perfectly predictable for anyone with some common sense (debt can only grow until the weight of servicing it becomes unbearable), yet thousands of economists working for these collapsed banks kept on with their models and assumptions, right into bankruptcy.

 

[Hygro]

I'm enjoying this debate. Keep posting you two!

 

[JerichoHill]

Sraffa's work fits within general equilibrium models today.

As for monopolies, they aren't socially optimal in most situations. They also have a way of stagnating and losing their market advantage over time, unless they're a natural monopoly (say, an electrical utility)

Remember Inno, you have producer surplus, consumer surplus, and deadweight loss. Deadweight loss is the arguement against a monopoly.

And Inno, firms face decreasing, constant, and increasing cost to scale. You have to bid out other firms for capital, other firms for labor. That will increase the bottom line a firm pays.

Do you really think that we use models that only strictly assume the picture perfect world of perfect competition? No, of course we don't. We don't need to assume perfect competition. Robust GE models assume that market power varies and firms carry influence over wage rates, etc.