The simplest explanation I can think of is that an economy can produce goods and services having a combined value greater than the sum of the raw materials and raw labor used to create them.
Think of it this way. A DVD player has about $5 worth of raw materials in it. It sells for $300. That $250 value added increases the wealth of all those workers and employees who had a hand in building it and getting it to you.
I’ve been mulling these things over. Still no go, but I’m trying. Let me try to establish some of my difficulties. I find the concept of closed vs. open systems to be useful, even though it’s not clear how it translates to the games. Now, there seem to be three qualitatively different examples being bandied about here:
[ol]
[li]The blackjack: two people wager on a game (zero sum game).[/li][li]The commodity: one person buys a product from another (non-zero sum game).[/li][li]The table: labor is added to raw materials, making the end result worth more than the materials themselves (non-zero sum game).[/li][/ol]
Blackjack seems the clearest example of a zero sum game to me. The game starts when money is put aside in a pot. The game ends when the pot is redistributed. The system is closed, measured by the amount in the pot. OK, got it.
Commodity is less clear to me. The game starts when the buyer/seller begin their transaction. The game ends when money is exchanged for the item (or even barter occurs). Nothing is actually being produced (created); rather, the items themselves are simply transferred, but not transformed. The supposed wealth creation is due to subjective (and different) valuations of items being exchanged. As I’m conceptualizing it right now, this form of non-zero sum game is analogically similar to a Stirling engine (where different value assignments equate to temperature differences and created wealth equates to work produced by the engine). In a Stirling, there needs to be heating (or cooling) exerted from the outside to make it produce work. A closed system produces no work. Similarly, when buying/selling an item, the system is closed at the moment the game starts, assuming the utility function remains constant throughout the game. (If we assume a variable utility function during the course of the game, I’m not sure how assigning benefit has any meaning.) Again, a closed system produces no work (wealth). I have more thoughts about this that I’ll get to below.
Table seems the clearest possible example of a non-zero sum game. However, I still don’t see it. The game starts when the person gets the raw materials and ends when they trade their finished product for something. The system is closed from this viewpoint, as the person’s labor is worth exactly the difference between the cost of raw materials and the amount received upon sale.
Additional thoughts (pardon me if they’re a bit rambling or disjointed, it’s a product of trying to work this out):
I suppose another way to approach this is to set up stereotypical value tables as generally used in game theory, like those used to express the choices available in a prisoner’s dilemma. What might they look like in the table or commodity examples?
In the table example, the scope of the game can be adjusted (i.e., the game starts when the transaction starts), in an attempt to open the system. But then I’m not clear on what defines the system’s boundaries. Alternatively, we can exclude labor as a cost; but I’m not clear on how that would be justified. (I believe this is the basis for Marx’s analysis; a worker can never be compensated for more than their labor is worth, hence, the worker always loses in a capitalistic system.) As I said above, the only place I see for wealth creation is that there is no cost to grow trees (although putting them into usable form clearly has a cost).
Come to think of it, I think excluding labor is exactly the reason economics transactions can be considered non-zero sum. What I mean is, from the point of view of the owner, the profit is equivalent to however much they are stiffing the worker. It’s a capitalist system, so of course the system is defined by the capital.
Now, in the commodity example, yes, one party values an item more than the other. The measure of the system might be the break-even point? Does that mean that created wealth can always be attributed to the poor bargaining skills of one party? Wait – how about this: in the same way that the prisoners embroiled in their dilemma cooperate to gain benefit, the parties involved in trade each settle for less than their break-even price. In that sense, the amount of wealth created is equal to the amount each is willing to forego.
Odd, but that’s the only thing (thus far) that makes sense to me…I’m just gonna post these ramblings in hopes of getting feedback. Thanks all who made it this far…
This isn’t really very complicated. In a typical transaction, the buyer needs the good or service he’s buying more than he needs the money he’s giving for it, and the seller needs the money more than he needs the good or service. So, for example, if I go to a bakery and pay two dollars for a loaf of bread, I need the bread more than I need the two dollars (I’m hungry and can’t eat the money). The baker, on the other hand, has lots of bread, so he has more use for the two dollars than he does for one more loaf of bread. Each party in the transaction is better off than he was before the transaction took place. Thus, it’s a non-zero-sum game.
Yes, I think I’ve got it now. I don’t what my stumbling block was. It was making the comparison to the prisoner’s dilemma that finally did it for me; answering the question “What would it mean for two parties to cooperate in a transaction?” is what made it clear.
Is there a name for getting so wrapped up in posting that one loses the line of argument? I was thinking “over-Doping” – an overdose of the SDMB. I’ll attribute my last post to over-Doping and just be glad it helped me work it through…
Consider this example:
Let’s say we need to dig sand to make a living. We can make $1 for every pail full of sand we deliver.
Now consider a situation where one person has two shovels, and one has two pails. In each case, the marginal value to the owner of the second shovel or pail is very low. I can’t do a damned thing with two pails and no shovel, and you can’t do anything with two shovels and no pail. So we agree to an exchange: One shovel for one pail. Both of us have now improved our lives dramatically, and our own personal wealth has each increased. Because in a sand economy, someone with a pail and shovel is much wealthier than someone who lacks one or both.
This was not a zero-sum transaction. It was a huge benefit to both parties. Note that the guy with two shovels was more than happy to trade a shovel for a pail. But once he had a shovel and a pail, he wouldn’t trade his only shovel for two pails. Or three. The ‘value’ of a shovel to him is entirely dependent on circumstance. It’s not an intrinsic property of the shovel itself.
That’s why most trades in a capitalist system are not zero-sum. As people collect things, the value of each thing changes over time. Values of things can change even based on interest - the value of my daughter’s tricycle was high three years ago. But now that she’s too big, the thing has almost no value. That’s why you see tricycles at garage sales for $5. Of course, we bought that tricycle at a garage sale ourselves, because we valued a trike a lot more than $5.
Yes. For anyone else who’s trying to make sense of this (and to provide an opportunity to correct any misconceptions I might have), I’m going to spend a bit of time sketching this out. In table form (akin to the prisoner’s dilemma):
| shovel
| keep | give
----------------------
p | |
a keep | 0 | -1/+5
i | |
l give | +5/-1 | +5/+5
In the table above, the number pairs represent the value to the shovel/pail person, respectively. I made it such that giving an object away results in a minor loss, but receiving an item results in a larger relative benefit.
For items that are bought and sold, the table becomes a little more complex (which was throwing me for a loop). In it’s simplest form, I’d think there are three entries per side:
| seller
| loss | even | gain
---------------------------
b loss | -/- | 0/- | +/-
u even | -/0 | 0/0 | +/0
y gain | -/+ | 0/+ | +/+
The loss, even, gain is meant to indicate value judgment, which have slightly different meanings for the seller than for the buyer. For the buyer, even is the maximum amount he/she would pay for an item; loss is a cost above that value and gain is a cost below that value. For the seller, even is the cost of production/manufacture. The meaning of loss and gain to the seller should be clear. In this table, the +/- pairs represent the value to the seller/buyer, respectively. I left actual values out because that’s where the notion of utility value comes in. The greater the difference between even and gain, the more wealth is generated.
See, this was one of my stumbling blocks and I feel it’s a mis-statement. For any particular “game”, the value does not change. It is not reflected in any one example of a game, but can only be considered in iterated games. Unless one also supplies some form of utility function with a time frame involved.
Is that all good? I’d appreciate any corrections…
Well . . . it appears to me that you have defined “a game” in this sense to be a single, simple economic transaction. If that is the case, then no time passes during the “game”, so of course the values do not change. Note, however, that the “owned value” or “personal sense of prosperity” does change for each player during the transaction. To me, it appears you are taking a snapshot of a river and then wondering why people say that teh water moves.
The only way to get “a time frame involved” is to model more than one transaction. And, in fact, this is the heart of macroeconomics.
BTW – changes in value due to subjective valuation are only one way in which an economic system is not “closed”. Technological improvements, new resource availability, monetary policy changes, natural disasters, population fluctuations, exhaustion of resources, etc. also represent means by which the “total chips in the game” might change over time. Subjective value is the key to understanding a single transaction in a free market, but on the macro scale many more elements come into play.
Yes, that is the case. In the context of game theory, I’m not sure what else a single “game” could be. I’d think that, by definition, macroeconomics necessarily depends on an aggregate of individual transactions. While the characteristics of micro/macro under consideration may be different, without the micro, we got no macro. But this wanders off the subject and into complex systems theory; if you’d like to discuss emergent properties and the like, I’d think it belongs elsewhere.
No, I think this is exactly wrong. Any change is due to completion of the transaction. Right?
For multiple transactions, yes, we’re in complete agreement. However, one could adjust one’s utility function without ever entering into a transaction. That makes sense to me in terms of utility of a tricycle for a child – as the child grows older, the value of the tricycle will decrease, whether or not an economic transaction has occurred.
Yes, that’s right – thank you for pointing it out. I’d think that one could also incorporate such things into one’s utility function at a micro-scale.
Oh, and – I didn’t mean to sound curt regarding emergent properties and the like; I’m just trying to be careful about keeping to the OP, as this is in GQ.
Well, economic systems are a pretty classic example of a CAS, but I think the questions you are asking do not require such analysis. Sticking to game theory, a game need not be defined by a single transaction. Consider the Prisoner’s Dilemna that you turned to for initial insight. that game consists of multiple itterations of a choice. Winning strategies for the game depend specifically upon completing multiple itterations. While some games are defined by a single event, those are by no means the most interesting ones, nor do they make the most compelling model for complex real world situations.
You read my sentence a bit too literally. I was not proposing that we model individual transactions as having a duration (though in the real world they do.) As you say, the change in value is a result of the transaction having taken place.
Well, utility function is a poor descriptor. valuation is by no means driven exclusively by utility. You are correct, though, that an economic transaction need not take place for a value to change. However (and this is key) there is no way to describe or bound the change in value unless an economic trnsaction is offered. The only way to find out how much a child values the tricycle at any given moment is to determine what he would consider an acceptable exchange.
Upon further reflection, I stand corrected. I suppose what I was thinking of was a “round”, so long as iterations are specified. But this seems to be niggling details of a literal reading.
I was under the impression that a utility function was all-encompassing, including all factors that went into making a decision. Is that not correct?
It depends upon which economist you ask. 
Actually, I think we are definitely getting into the area of picking nits rather than fostering understanding. My apologies.
The point that I meant to communnicate (before getting sidetracked into pedanticism) is that a utility function describes an ordinal relationship. It doesn’t (rigorously) make sense to speak of the value of the tricycle decreasing over time unless there is a means for exchanging the tricycle from one time to another. Again, the important point is that the tricycle itself does not have a value. A person will value the tricycle, but the only way we have to describe that relationship mathematically is to order it as a preference in relation to other goods or services (or states of being, for the philosophically minded).
Or, to appraoch from another direction, the transaction is wholly irrelevant to both the valuation/utility function/preference relation. The mathematics describes what a rational person should do in a given context[sup]1[/sup]. Conversely, without the opportunity to make a transaction, there is no relationship from which to determine a preference. There are no economics in a vacuum.
[sup]1[/sup][sub]For the record, one common counterpoint for utility functions as a complete model for economic behavior is that they do not account for feasability. On the other hand, in game theory I believe that a utility function is considered to be all encompassing. Then there’s the nitpick that a utility function is really only a mathematical convenience to describe the preference relationship (or set of indifference curves) . . . yadda yadda yadda . . . lobster bisque [/sub]
I find it much easier to think in terms of destroying value rather than creating value. Say I go out and buy a hamburger for $3, I now have $3 less cash but I also have a hamburger valued at whatever. Now say I eat that hamburger, I still have $3 less cash but nothing of value left from that transaction. In effect, I have just destroyed $3 worth of wealth from the world. Everything we consume is destroying value so the economy is not a zero sum game.
Try it. I’ve told you what a game is. That is the proper definition from game theory. I, probably in error, combined two things together. Regardless, this is from Microeconomic Theory by Mas-Colell, et al., which is about as close to the bible of microeconomics as one would wish to get (pp 219-220):
There you have it. That is the definition of a game. So before you declare anything to be a game, answer those questions. If you can’t, it’s not a game. (Yeah, you won’t know individual preferences, but I’m sure you get the idea.) So, the traffic zipper, where two lanes merge into one. Do you merge early or late?
i. Players are the drivers ahead of the merge
ii. Merge early or merge late, such that one must merge at some point
iii. If everybody merges early, outcomes are such-and-such; if everybody merges late, outcomes are this-and-that; if some merge early and some late, outcomes are what-not
iv. How are people going to feel about the outcomes?
Okay, that is a crude and ugly example, but from looking at it you can see that the traffic zipper is indeed a game, no?
What’s a zero-sum game? (pp 221)
Okay? Open vs. closed has nothing to do with it. When friends get together to play poker, there is no money on the table before they show up and none on the table after they’ve left. All money lost is matched exactly by all money won. It is a zero-sum game.
Black jack is not a zero-sum game because the house, to the extent that it is not a player, takes some of the money. Most likely the players will have an overall net loss; thus it is not a zero-sum game. If the house is considered to be a player, however, then it becomes a zero-sum game. Make sense?
Making a table isn’t a game. I’m not strategically interacting with anybody when I make a table. I’m fashioning wood. An exchange of goods may be considered a game, but there is no reason to consider it a zero-sum game.
I think you may be making a common mistake that plagues economics: you are thinking in terms of money or stuff. Economics is not about money or stuff. It’s about welfare. The basic notion is that people have preferences—rats & pigeons, too—and they act to maximally satisfy their preferences.
Shoot, gotta run. I hope that helps some, at least.
Sorry about that.
People have preferences and they act to satisfy them. Economic well-being is measured in terms of per capita GDP and stuff like that because it is an easy proxy that is at least useful, if not perfect. For now, I’d like to hit on two sources of gain from exchange.
The first comes from specialization. We are all generally better at some things than others, or like some things more than others, or have different relative abilities over the things we can do. This is where specialization comes in. None of us depend on ourselves for producing and consuming for all our needs. We go to doctors, mechanics, teachers, &c. This form of advantage, comparative advantage, arises from the differences in relative prices for producing. Bob may be better at both cooking and sewing than I am at them, but because the relative trade-offs between the two activities are not the same for Bob and I, it pays for me to specialize and for him to specialize and for us to trade.
Another form of gain, that I alluded to above, is called consumer & producer surplus. I might be willing to pay $3 for a pint of Guinness, but if the market has the price set at $2.50, then I’m getting $0.50 bonus benefit from every pint I drink. The only person who doesn’t gain from buying a pint is the person who values the pint at exactly $2.50. Producer surplus works in the opposite direction, but I think you get the idea.
Now for another thing that you may not understand. There’s a book called The Arithmetic of Life and Death. I wouldn’t recommend reading it. In one chapter the author discusses how much one has to produce to be worthwhile as an employee. So he goes through overhead, equipment, &c., and comes up that a typical office worker, say, must produce $20,000 a year for the company to break even in emplying the person. It makes sense, right? And it certainly would be intimidating to your average slacker who wants to goof off.
But there’s a problem. It’s bullshit. The thesis of the book is that simple arithmetic can be applied to all sorts of real-life scenarios in order to gain insight and understanding. What the author fails to understand, is that some problems really aren’t best tackled with arithmetic, and calculus is one of them. This is bad for the author, because the problem he was trying to address with the office worker is a problem that needs to be tackled with calculus.
What the author was doing was getting the average costs and using that, but that’s not the relevant figure at all. Not even close, really. He needed to be dealing with the marginal cost which can paint quite a different picture. It is unfortunate, but this is the same sort of problem that Marx had. When you take the revenue function of a factory, you’re dealing with a curve. To find the additional value of the n[sup]th[/sup] worker, you need to evaluate the derivative at that point. What you find is that the marginal revenue is actually less than the average revenue; the worker should be paid the marginal revenue because that’s the value that she actually brings to the firm.
You may disagree, saying that it should be the average of the workers; however, then you are committing the crime against the owner: it is her capital that is at risk, and without it the workers wouldn’t be working at all. If you take the average revenue for each unit of capital, you have a situation where the capital can never be compensated for more than it’s worth, so the workers necessarily exploit the factory owner. But this dilemma is just an illusion; it’s a modern-day Zeno’s paradox!
I probably really botched the explanation, but I hope it helps.