Short's Law

Perhaps the economists among TerraNova’s readership can help me out here.

Courts routinely issue fines based on offenders’ ability to pay. Last week, the UK government narrowly avoided defeat over its flagship education bill by effectively conceding that the cost of a university education would be made variable depending on the ability of students to pay.

What if we translated this into virtual worlds? What if objects cost more money depending on a character’s ability to pay? Instead of pricing the sword at 500 units of currency (UOC) it would be priced at 500 UOC + 0.01%? For transfers between players, the percentage factor would be transformed accordingly, but only when the rich bought from the poor (to stop muling). Example: I have 1,000,000 UOC, you have 10,000 and I buy something from you for 500 UOC + 0.01%. I lose 500+100=600, you gain 500+1=501.

OK, so players would never accept this. Nevertheless, it’s an interesting thought experiment: what would happen if everyone paid their own individual price for everything they bought?

The trouble is, I don’t where to look for the answer as I don’t have a background in economics. I suspect I’m feeding Google with the wrong keywords (“ability to pay” when I should be using “Short’s Law” or something).

Anyone care to point me in the right direction?

Comments

[dan]

1.

Players will get around it. Rather than sell, they will switch to barter for large transactions, or create their own currencies, in effect.

But isn't this just another attempt at countering mudflation, a drain of virtual wealth? Player will, as you suggest, see it as unfair, and a sort of virtual flim-flam perpetrated by the game designers. And they will metagame it.

Rather than this artificial sort of sharing the wealth, focus more on drains that scale with wealth. By that I don't mean that something in a shop costs based on the wealth of the buyer, I mean that there are incremental benefits that the wealthy pay for, whether status-based or 'real.' Leave the poor with "it lets me get by," the middle class with "gets the job done just fine," but make sure there are lots of very slightly better things that cost significantly more... and that these drain cash from the economy, not just transfer it from one player to another (though transfers are fine, need something to draw down mudflation a bit).

That's one thing that SWG could have more of, cash drains. Maintenance and travel tickets are good, but they doesn't come close to offsetting the influx of wealth from the program's generation out of thin air (from a modeled economy standpoint) of the resources most everyone mines. I'm not suggesting they should balance, but they are terribly lopsided in their relationship.

The drains should be tied more to time actually played too, not just to real time. The present maintenance system in SWG penalizes casual players. They effectively pay more maintenance per hour played, meaning they have to spend more of their play time "working." Travel is better balanced in that you only pay travel costs if you're playing. I know maintenance has other issues, like keeping the database slimmer, but drains need to hit those that play the most proportionately to those that play the least.

Don't tax players for having more. Tax them for playing more, for spending more time in game (where they are taxing system resources with their presence). The current systems for charging players real money for entertainment doesn't do this, but that doesn't mean the ingame system can't. It might even help trim back, somewhat, the more outrageous "commercial" ventures.

[Peter]

2.

Dan, with a comment as well composed as that, we need to get you your own blog...

[Douglas Galbi]

3.

The terms you want are "price discrimination" and "revenue management." What you propose is called among economists "perfect price discrimination." Information problems and transaction costs, as well as customary practice and norms of fairness, limit the extent of price discrimination in practice.

Price discrimination is prevalent when there is a large difference between average cost and marginal cost. Thus price discrimination tends to be found in transportation pricing, hotel pricing, communication services pricing, and the pricing of information products. It's likely to occur to some extent for game world objects.

[Cory Ondrejka]

4.

Richard> What if objects cost more money depending on a character’s ability to pay?

IANAE, but doesn't eBay et al accomplish this? Right now, all items in virtual worlds are priced based on a mix of in-world currency and time to in-world. eBay, IGE, Gaming Open Market, &c, all allow users to change the equation to be a mix of in-world currency, time in-world, and real-world currency.

[Julian Dibbell]

5.

Yep, like Douglas said, "price discrimination." The network theorist Andrew Odlyzko has some interesting thoughts about the subject in his paper "Privacy, Economics, and Price Discrimination on the Internet":

http://www.dtc.umn.edu/~odlyzko/doc/privacy.economics.pdf

The gist, as I understood it, is that while virtual environments make price discrimination easier than ever (and while there are good reasons to want to encourage it), the privacy concerns it raises make it broadly unpalatable.

In other words, the players would hate it.

[P.D.]

6.

Julian Dibble writes: "The gist, as I understood it, is that while virtual environments make price discrimination easier than ever (and while there are good reasons to want to encourage it), the privacy concerns it raises make it broadly unpalatable."

You seem to be thinking of price discrimination based on players' real world wealth. I think the original post is considering discrimination based on characters' in game wealth.

Either one might have interesting effects-- the latter raises no real privacy issues.

[Julian Dibbell]

7.

Actually, I agree, PD. I was putting words in Odlyzko's mouth when I implied he was talking about the kind of virtual environments we're interested in here. He's talking about digital networks generally.

But more to the point, I'm not so sure that even in those broader areas privacy is really the sticking point. I suspect that to the extent privacy is a voiced concern, it's really a proxy for issues of fairness. In other words, people don't care so much whether the pricing system actually knows how much money they have -- they care whether it *acts* on that knowledge. When it does, they see it as violating egalitarian market-economic ground rules that reward people's efforts with wealth. Price discrimination undercuts the reward mechanisms that keep people on the treadmills that drive the economy.

Or so the objection goes, anyway. Needless to say, it definitely applies to players' in-game wealth accumulation.

[Richard Bartle]

8.

Douglas Galbi>The terms you want are "price discrimination" and "revenue management."

Thanks - just what I wanted!

I've come across the term "price discrimination" before (for example in http://www.gamestudies.org/0302/castronova/), but always read it as a regular English phrase rather than something with special connotations for economists.

Having now done some googling, I note from http://www.ftc.gov/bc/compguide/discrim.htm that in some cases price discrimination may be illegal. Whether that would include the internal economies of virtual worlds or not is another one for the lawyers...

Richard

[Richard Bartle]

9.

Dan>Players will get around it. Rather than sell, they will switch to barter for large transactions, or create their own currencies, in effect.

Not if they were buying from or selling to NPCs, they wouldn't.

P.D.>I think the original post is considering discrimination based on characters' in game wealth.

Yes, that's what I meant. The virtual world knows exactly how much cash an individual has, and can do whatever it liked to transactions involving that cash.

The price would have to have a non-scaled component as well as a scaled component, because otherwise people with hardly any money would have the same purchasing power as people with lots of it, but beyond a certain wealth the percentage component would come to dominate.

My guess is that if this kind of economy would lead to people keeping their wealth in goods, since there's little reason to save. In other words, as soon as they got money they'd spend it, down to some limit below which the absolute part of the price would become significant.

Richard

[Jeremy Neal Kelly]

10.

This pages gives a terrific explanation of the theory of price discrimination:

http://en.wikipedia.org/wiki/Price_discrimination

Here's my cursory economic analysis:

First, effective price discrimination requires knowledge of the demand curve, which developers don't necessarily have, and which this scheme ignores in any event. While this 'progressive pricing' method might offer a decent estimate of player demand, it will certainly be incorrect in many cases. This will lower the efficiency of the market, which will decrease the volume of trade and (from a strictly economic standpoint) injure all its participants, rich or poor.

Second, the goal of price discrimination and taxation (which this idea also suggests) is to transfer wealth between two parties, presumably with the understanding that both value it. This scheme, however, merely destroys that wealth. This may or may not cause a net loss of value, depending on one's ideas about the nature of game wealth, but it doesn't appear to benefit anyone. There might be some indirect effect which increases value by improving the game, but this, presumably, is a design issue rather than a strictly economic one.

Richard Bartle> My guess is that if this kind of economy would lead to people keeping their wealth in goods, since there's little reason to save.

When I first read your post, I assumed that by 'ability to pay', you meant the player's entire wealth (including the value of their goods) as opposed to their cash balance. Were you hoping to deprecate the use of currency for some reason?

[Fred Hapgood]

11.

Price discrimination is often hard to define, but you might get a taste of what you are looking for in games with a significant charitable component, since the more money
people have the larger their charitable contributions. (Something like 7% of all US
charitable spending is given by only ten people. Or around there.) In SL you are encouraged to
'rate' objects and avatars, which means to throw
them a few bucks out of general appreciation. I bet rating correlates with asset holdings.

Imagine a game economy in which there were no
fixed prices. You just gave the creator what
you wanted, with the idea that if you didn't
reward him for his efforts he would disappear
and you wouldn't be able to get any more of whatever he does. Bet those contributions, which would be like patronage, would go up with wealth.

[DivineShadow]

12.

Richard,

Players would not only find the system demotivating them from amassing wealth in cash, they would find plenty of workarounds.

The controls needed to enforce strict price discrimination on all transactions would be unpalatable. You might get away with doing it to some services or things that "charge" currency upon "use", but not your everyday game objects.
One possible workaround players will quickly come up with if you do this price discrimination is to have a poor player buy from an NPC, turn around, drop item on the floor and grab the pile of cash the rich guy next to him dropped. A few hours later the first escrow house will open staffed with apparently poor characters, the commisions will of course be given out to a "cashkeeper" character or account.

[Jeff Freeman]

13.

What Dan said.

I don't think this would be a very good idea.

I've always been fond of Traveller's model, where wealth was a question of cash-flow rather than cash accumulation. Where 'being rich', meant one had to make *and spend* a lot of money - 'being poor' meant that you neither made nor spent a lot of money.

But that's a different thing altogether.

[Staarkhand]

14.

Perhaps my general ignorance of economic theory is catching up with me here, but can I ask why? If high level/power gamers have too much money, give them something to spend it on, not some arbitrary taxation that will only, as pointed out, be worked around, NPC's or no. I can think of several ways to get around that right now with nothing more creative than multiple accounts and/or some friends. And historically I'm not particularly clever at working the system, at least compared to a player base of hundreds of thousands of people essentially working full time to come up with ways to get around pointless gamisms.

Example: Dialbo II: LoD's price scaling for gambling on unique/rare items based on character level. That worked for about 10 seconds. Answer: gambling bots (or mules, choose your dialect). Change the rules and the metagame will change.

[magicback]

15.

Richard,

As indicated by Jeremy, the scenario you put forth is a tax. Players will demand full accountability for the use of the tax revenue or scream “unfair money sink.” I’ll leave the dynamics of tax and wealth transfers to more able experts.

In the current debate over the cost of universities in UK, charging based on the ability to pay is a second tax on wealth: first when citizens pay their general tax (based on wealth) and second when citizens attends university (based on wealth again). This second-tax concept assumes that citizens will attend university at some point in their life.

Charging based on ability to pay is “price discrimination.” You’ll probably agree that most people would not apply this concept.

What companies and consumers do now is dissect the different values embedded in each price point. For example, airlines have different prices for the same seats because there are different benefits embedded into the price. You have full fares with no restrictions, restricted fares, group fares, last minute fairs, etc. You get to go from A to B, but you get different flexibilities.

So with the pursuit of finding embedded values, you get auctions for price discovery (finding the true value) and Priceline’s reverse auction (finding individual’s perceived value). However, value dissectors will say that the ability to set your own price is the embedded value. Examples of this are paying an extraordinary high tip, charity, random twinking, etc. Each action has a true value and perceived values.

The ‘money sink’ solution suggested by many in different forms is to allow players to speculate on their wealth assets: spending A will give you a probability of getting B. Diablo II’s gambling feature and SL’s volunteer-payment rating system are examples.

I recalled a MUD where I have the ability to pay more than what the NPC vendor requested. I didn’t know whether paying more will be of additional benefit, but I paid extra anyway just in case :)

Frank

[Edward Castronova]

16.

Man. I go away from here for five minutes and when I come back, everybody's an economist. And good ones too.

Yeah. Price discrimination, workarounds, legality/fairness issues, effectively just a progressive tax, debate on what equitable taxation is, problems of enforcement. Yeah, yeah, yeah. I feel rather unnecessary :(

But I'll say something anyway! This issue comes down to a debate about redistribution. I think in-world redistribution would enhance both the fun factor and would help reduce a number of practical problems, such as mudflation. I'm not sure price discrimination is the way to do it. Usually, PD operates not off of income differences but price sensitivity differences. Example: coupons. The store charges you more if you don't take time to cut coupons from the newspaper. Because you're richer? Nope, because you're time-starved. Time-starved people don't price-compare so much. People with lots of time do. So the coupon cuts the price to people with time to shop around. Same thing with 'senior and child discounts' - they could be renamed 'working parent surcharges.'

If I wanted to balance wealth in a virtual world, I would put in a wealth tax that's consistent with the lore. How about this: bank vaults require protection. The more things you have in there, and the higher quality, requires more guards. More guards cost more money. So everyone gets 8 bank slots free, but after that the fees go up steeply and depend on the quality of the items you're storing.

[Julian Dibbell]

17.

Ted, here's a question for you (or anybody else who feels up to it):

If I understand things correctly, the redistributive taxes you and Richard propose -- gold sinks both -- differ from real-world taxes in that they redistribute wealth not directly but indirectly. With real-world taxes, that is to say, revenues taken from one part of the economy generally result in an expenditure in another part of the economy, whereas with these taxes wealth is simply destroyed at the high end of the economy. In theory, this makes wealth scarcer overall and thus increases the value of assets held by the poorer folks -- effectively redistributing wealth without explicitly doing so.

My question is: What are the meaningful differences between the two approaches? Why choose one over the other? (And a bonus question: Does the gold-sink approach have any real-world analogues, or is it just another one of those game-world anomalies?)

[Jeremy Neal Kelly]

18.

Julian Dibbell> ...with these taxes wealth is simply destroyed at the high end of the economy. In theory, this makes wealth scarcer overall and thus increases the value of assets held by the poorer folks -- effectively redistributing wealth without explicitly doing so.

This is truly an excellent point. Most people forget these kinds of effects when discussing game economies.

Julian Dibbell> My question is: What are the meaningful differences between the two approaches? Why choose one over the other?

If the intent is to create a gold sink or implicitly redistribute wealth, then I think a simple wealth tax of the sort suggested by Dr. Castronova is by far superior. Aside from being easier to understand and politically more feasible, such a system introduces no distortion to game markets that might diminish trade between players. Whatever anyone happens to think about the effect of economic forces within games, I think we can all agree that in-game trade is desirable, if only for the socializing it entails.

Julian Dibbell> And a bonus question: Does the gold-sink approach have any real-world analogues, or is it just another one of those game-world anomalies?

This reminds me of the way the Fed regulates the money supply by 'purchasing' money with Treasury bills, but really the comparison is fairly superficial.

[Ernest Miller]

19.

One question, one comment:

Why not let players have some choice in how the wealth system works for them, by allowing them to choose "wealth classes"? For example, by playing a "poor" character, you might have less starting wealth and earn less through jobs (reflecting a lack of business acumen, contacts, what have you), and you might not be able to enter certain "wealthy" establishments, however maintenance costs would be lower. Conversely, you could play a "wealthy" character that earned money more easily, but also had higher maintenance costs. Or something similar.

Since I can't spend much time in virtual worlds playing, I really like the concept of tying maintenance costs more to time spent in the virtual world then real time.

[Richard Bartle]

20.

Jeremy NEal Kelly>This pages gives a terrific explanation of the theory of price discrimination:
http://en.wikipedia.org/wiki/Price_discrimination

You're right, it does - thanks. It looks like a much wider concept than I was describing, with people discriminating prices on all sorts of things (of which overall wealth/cash would be just one).

>When I first read your post, I assumed that by 'ability to pay', you meant the player's entire wealth (including the value of their goods) as opposed to their cash balance. Were you hoping to deprecate the use of currency for some reason?

I wasn't planning on using the idea at all; as I said, the players wouldn't accept it - end of story. This wasn't a serious suggestion of a way to do [whatever] in virtual worlds, it was research: I was interested to know what effects it might have were it to be implemented. It obviously has SOME uses; could any of these be co-opted into a virtual world? Would it be help or hinder that we have precise knowledge of wealth and inventory we have for individual characters?

It seems that price discrimination as a concept is ripe for exploits, except in the very particular case when what you buy is non-transferrable. Thus, you can make it that a character buying training from an NPC trainer has to pay more the better the training they have beforehand, and few people complain. It's not like you can pay someone else to undertake your training and expect you'll go up a rank.

So, from what people have been saying it looks like the answer to my question "what would happen if everyone paid their own individual price for everything they bought?" is "only bad things".

Just as well it was only a thought experiment and not a live one, then!

Richard

[Cory Ondrejka]

21.

Julian> Why choose one over the other?

Second Life used to tax in the redistribution sense. However, it was quite unpopular, as TN covered previously. With the 1.2 release, we changed the model significantly, including the removal of taxes.

Ted> I would put in a wealth tax that's consistent with the lore.

I was originally going to mention how easy this is to circumvent with external traders (Gaming Open Market, &c) but, while that is still true for individuals, the GOM rep will then be one of the richest users in the world. So, running with Ted's thought, the banker then becomes the biggest, most well defended, but most intriguing target :-)!

Of course, if the banker actually loses the money, we're back into previous TN discussions about stealing in virtual worlds.

[DivineShadow]

22.

Richard,

"Would it be help or hinder that we have precise knowledge of wealth and inventory we have for individual characters?"

Since the consequence of dodging the 'tracking' system is effectively nil (it's hard to audit the player's ingame tax return and fine them), the value of the tracking data gathered by the game is going to be directly proportional to the difficulty in excecuting the easiest dodging manuver, whose frequency will in turn be proportional to the benefit gained from it's execution.

I'd say better leave that data for other purposes...

[Richard Bartle]

23.

DivineShadow>the value of the tracking data gathered by the game is going to be directly proportional to the difficulty in excecuting the easiest dodging manuver

So if the easiest dodging maneuver were impossibly hard, that would mean the value of the tracking data is very good?

Virtual worlds can do things that the real world can't (yet). For example, in the real world price discrimination is optional, but in a virtual world you could make it mandatory (so anyone who bought things at lower prices in order to sell them to people who would be charged higher prices at source would themselves have to charge the higher prices).

Example: a character with a lot of cash wants to buy a sword from a character with very little cash. They agree on a price, but there's a sliding-scale sales tax imposed so that the cash-rich character loses a lot more than the cash-poor character gains. In order to try to dodge this tax, the cash-poor character drops the sword on the floor and the cash-rich character drops some coins. Except, when the character drops those coins they cease to be "500 gold pieces" and start to be "500 gold pieces or 0.01% of your total wealth, whichever is the lower". When the sword-seller picks them up, they convert into 2 gold pieces instead of the 500 they were expecting.

Again, I'd like to stress that I'm not proposing anyone actually does this in a live virtual world. VWs do have facilities at their disposal that the real world doesn't at present, however, so some of the standard arguments against price discrimination may apply differently.

Richard

[Bruce Boston]

24.

I think that its important to remember that the value of a good/item/currency in any secondary market determines its intrinsic value in the primary market.

So if you give players more money as they go up the ladder, but tax them for that reward, its the same as not giving them the reward in the first place.

It would be like me telling my 5-year old that his allowance went from $5/month to $205/month, but he now has to make the $200/month car payment. He'd pretty much figure that out the first time around that he was out the cost of the stamp.

As far as PD is concerned, I think the general rule is that members have access to more information than the developers, so if anything players are the ones using PD against the developers, not the other way around. As Raph has pointed out before, it's pretty clear that MMORPGs have a much higher value than the developers are able to charge for. Being a gamer, my guess is that most players won't pay more simply because they know they don't have to.

Besides, a progressive tax on senior players would be like a frequent flyer program that made it so you would earn downgrades the more you flew with an airline.

-Bruce

[Tom Hunter]

25.

Going back to Richard's original question: "What would happen if everyone paid their own individual price for everything they bought?"

In VWs its a thought experiment but in real life its taxation and we do pay very different prices for the services our governments provide to us.

People react to taxation in all sorts of ways. Most of us just pay. Many seek to avoid or reduce the burden legally, some find ways to expliot the system, some attempt to change the system to work in thier favor, some cheat.

Interestingly from time to time an enlightened ruler trys to reform (as opposed to modify) the system in a way that benefits the majority of taxpayer. When this works taxpayer become happier and economies more productive. The Kemp Roth tax reform in the US in the 1980s comes to mind as an example of this.

We can see that running the economy well has positive effects in the real world. Judging by the amount of discussion VW economic management generates people think its important that economies run well in VWs as well.

But is that really true? Why?

[Tom Hunter]

26.

Speaking as some one who spends too much time on airplanes:

"Besides, a progressive tax on senior players would be like a frequent flyer program that made it so you would earn downgrades the more you flew with an airline." -Bruce

That is pretty much what the airllines offer.

:(

[magicback]

27.

Tom Hunter - "We can see that running the economy well has positive effects in the real world. Judging by the amount of discussion VW economic management generates people think its important that economies run well in VWs as well.

But is that really true? Why?"

-

Well everything have to run well, no?

However, I think we just need to have a sound and robust economic framework. Whether the VW economies are run well or whether it is of positive benefit will depend on the setting.

For example, the setting may be post-apocalyptic earth where the economy is in complete shambles because for some mysterious reason wealth and resources randomly disappears. The economy is in shambles, but the world setting may be interesting.

Bruce - "Being a gamer, my guess is that most players won't pay more simply because they know they don't have to."

Bruce, perhaps this view could be rephrased as "simply because they don't preceive the fair value of paying more."

If the average level gain rate per month is 2 level, then if you offer a differentiated pricing plan of 1.5x the monthly subscription fee to automatically gain 1 level per month but a cap at average rate (gain no more than average), players will value this pricing model and perhaps not yell about spending RL$ to gain in-game advantage because the only advantage is to catch up.

It would be great to implement a tax structure that lore consistent, but the structure also need to be consistent of player's metaphysical ideas of fairness-a big bag of socio-anthro-political philosophy stuff.

I recall the concept of "rational optimizers" in modern finance theory. Guess, I should go back to my textbooks on this :)

Frank

[magicback]

28.

Oh, so many grammatical errors. Better type slower and proofread. My apologies.

Frank

[Edward Castronova]

29.

Too many things to comment on...

Bruce> As Raph has pointed out before, it's pretty clear that MMORPGs have a much higher value than the developers are able to charge for. Being a gamer, my guess is that most players won't pay more simply because they know they don't have to. Besides, a progressive tax on senior players would be like a frequent flyer program that made it so you would earn downgrades the more you flew with an airline.

It goes back to price sensitivity. Charging level 65s more makes economic sense - you're taking advantage of their high switching costs. If you live in the high-rent district, you have to pay more. People consider real estate taxes fair because they are a (roughly) equal percentage of wealth. But the point is, the tax is higher as the wealth is higher. In MMORPGs, the very rich pay the same monthly fee as the new player, and there's no in-world tax at all.

Disclaimer: I do not WANT games to charge more $US for top-end players. That would stink. I'm just saying, it's something they could do if they wanted to fight that fight. A lore-consistent in-world progressive tax would be much better.

[Edward Castronova]

30.

Several commenters> This tax woul be easy to work around, use external markets, etc.

Agreed. But RL governments face the same situation - anyone can ship their money overseas at tax time, and so on. They handle this by putting regulations and observation points on these money flows. MMOGs don't (I guess), or if they do, they don;t do anything about it. So I can have 10 million gold pieces appear in my bank account overnight and nobody asks why. Or, that same 10m GP can disappear overnight, and again, nobody asks why.

Imagine that in the real world:

Dear Internal Revenue Service,

Thank you for inquiring about the sudden drop in my wealth on December 31, that resulted in me having no tax burden. What happened was, I just gave the money away to a poor family in Barbados.

And as for the big increase on January 1, that I am living off now? That was a friend of mine from Switzerland, he had some extra and decided to give me some. Thanks for asking.

See you next year!

Sincerely,

Taxscammer101

****************

The point being, an effective economic control system needs to observe flows in and out, and have the in-system controls respond to those flows.

[Edward Castronova]

31.

Julian> What's up with taxing and throwing the money away as opposed to redistributing?

A question I often ask myself. OK, traditional design systems in these places let every single player run a low-cost mint. You just go up to a monster, kill it, and bam you have some gold pieces (or loot that you sell for gold pieces, same thing). If money is cheap to mint, there will be lots of it. Inflation.

Presumably this design decision comes from the need to get new players (all players really) some rewards, and some purchasing power. But why give them new money when the system already has money in it? You just have to arrange things so that players always get their money from other players. A progressive tax where the proceeds are used to support newbie/lowbie hunting grounds and delivery missions, for example. (Raph: fund the mission terminals through storage fees on Master players). Redistribution is better policy than letting every player be a self-standing mint.

Richard's connection between the idea of taxing and destroying the wealth and the Fed buying dollars with securities is 100 percent accurate. The whole point of money supply contraction is to reduce money supply, to reduce the amount of money that's spendable out in the system. It's like taking away gold pieces and throwing them in the river. The reason it is fair in RL is this: the central bank doesn't just take your money and throw it away. It induces you to give up some money now, in return for a promise that you will get your money back later with interest. COmpletely fair - it's a deal, not a tax. All it does is defer purchasing power to the future. And that is sufficient to reduce purchasing power now, which is all that matters.

[Jeremy Neal Kelly]

32.

Bruce Boston> As Raph has pointed out before, it's pretty clear that MMORPGs have a much higher value than the developers are able to charge for.

Can anyone tell me where he talked about this?

[Jeremy Neal Kelly]

33.

Edward Castronova> Richard's connection between the idea of taxing and destroying the wealth and the Fed buying dollars with securities is 100 percent accurate.

Are you talking about my post, or did Dr. Bartle post something that I missed? Either way, I don't think the analogy is necessarily that clear. The Fed's actions don't destroy value; deleting game money very well might, depending on one's ideas about the value of game goods.

[Hellinar]

34.

Why does the tax money have to disappear into thin air? Perhaps it could be used to fund the payments newbies got from early rat killing quests etc. As the general population got richer, the low level quests would pay better, and newbies would catch up faster.

Should developers be paying more attention to connecting the gold sinks to the gold sources? I could imagine a world in which there was only a fixed amount of gold to circulate. I can also see a bunch of problems with such a model though, such as the effects of hoarding by first movers.

One system that can work, in my experience, is player created currency. In A Tale in the Desert, the developers didn’t provide any coin, only a press for printing bank notes. A variety of currencies sprang into being. My character had a large hand in creating the most successful and long lived of these, the Trade Note (TN). The TN is backed by a “basket” of widely used commodities, canvas, iron etc. To obtain a 100TN note, you have to give the Bank one unit of each of these commodities, which is then stored in the vaults. TN prices of basic commodities have been very stable for the last six months. Some “luxury goods” have increased in price over time, others have dropped in price as better technology made them more obtainable. Not much sign of “mudflation” there.

I don’t know if such a system would be useful in the “loot based” economies of the typical MMORPG though. A Tale in the Desert is at heart a construction game, which models a real world economy better than the typical MMORPG.

[Hellinar]

35.

Hehe. I see Edward was making the same point on redistribution to newbies as I typing in my comment. Nice to be in tune with a real economist :)

[Jeremy Neal Kelly]

36.

Hellinar> Should developers be paying more attention to connecting the gold sinks to the gold sources?

Much attention has been paid to this idea. For a good introduction, See Zachary Booth Simpson's study of UO:

http://www.mine-control.com/zack/uoecon/uoecon.html

[Mithra]

37.

I'm no economist, but my take is that this system would be impracticable (aside from being undesirable from the player standpoint).

Richard Bartle talked earlier about a system whereby a sales tax could be imposed on trades where the low-level player ends up receiving less gold than the high-level player offered. He also suggested players would attempt to work around this by simply dropping items on the floor, and furthermore the system could be made to work by having the dropped currency convert to some relative value based on the takers wealth.

I don't believe this could work, (aside from the obvious arguments that would erupt when the seller determines he's been scammed) because the other component of the trade doesn't change in value, which is the SWORD.

Here's what I would do if I were a rich player :

I'd cache my wealth somewhere, either by handing it to a mule, or hiding it behind a tree or wall, so that the server regarded me as 'poor', go buy items for the cheap price, retrieve my wealth, then have a cash transaction with my customer. Unlike currency, the sword won't depreciate having been handed to the customer.

One other consideration:

If a poor player dropped 2 gold and a rich player picked it up, would it covert upwards to 500 gold? And if so, what then becomes the point of acquiring wealth or even having currency? Would PD be appropriate in the currency/item economy, when the level/skill/item-usage economy remains traditional?

I apologize if I'm not presenting my thoughts very eloquently, but it seems a PD system is technically a currency-less system. Will players bother to play your game if there is no opportunity here to excel? Wealth acquisition is the metagame for many, and quite frankly one of the reasons I haven't subscribed (plagued if you prefer) many of the newer MMOG's is because their economies were weak for one reason or another. Take away the economic treadmill and there isn't a great deal of game left, you might as well be writing Battlefield 1942.

[Nick Douglas]

38.

VW economies must run on non-real-world rules because of two properties of the physical world.

Location: As Edward Castranova earlier noted, stores offer coupons lest they lose customers who have time to shop around. Then they charge more to the time-poor.

As many discussers noted, high-income areas have high prices. When my uncle visits the family out in Rochester, NY, he buys groceries to cart back to New York City.

Both of these situations depend on the expense and time of travel. But the web has no space, as David Weinberger said in his book "Small Pieces Loosely Joined." The Internet and VWs are places, but not spaces. Thus we can comparison shop with ease, because every place is right next to every other.

By forcing players to traverse long distances, spending either time or money, game developers could induce a geographical scarcity that would induce a certain price discrimination.

Entropy: "Life is pain. Anyone who tells you otherwise is selling something." Most of life is maintenance. Maintenance isn't fun. When a VW requires gamers to mine their resources in a boring way, maintenance takes over and players find workarounds, using mules to support their fun interacting characters. Eventually, game wealth creation, like all maintenance, will be fun or will have a workaround. Thus, lazy players will not remain poor.

Even with the dullest maintenance methods, there is little disincentive. Mining gold in meatspace breaks your back; mining gold in UO wears out your finger. Everyone willing to sacrifice this small inconvenience eventually becomes rich.

I imagine more creative forms of wealth creation will (and already do) fight mudflation as well as forced travel will allow reasonable price discrimination.

[Hellinar]

39.

Hellinar> Should developers be paying more attention to connecting the gold sinks to the gold sources?

Jeremy> Much attention has been paid to this idea. For a good introduction, See Zachary Booth Simpson's study of UO:

Ah, yes. Been a while since I read that. My personal Online World experience started with the launch of EQ, so I tend to forget about the details of UO. That experiment was obviously very influential in turning people off closed economies and on to the faucets and sinks model. Seems to me, in a closed economy, some sort of anti-hoarding strategy is needed. But this is very unpopular with players.

In the online world I am most familiar with, the Egypt of A Tale in the Desert, there is little or no cost to hoarding. Players, who can make almost any desired goods, tend to build stockpiles of most everything. Taxes aren’t required to reduce these stockpiles though. The world is quite dynamic, with new technologies and buildings to house them arriving regularly. The developers simply look at what is being excessively stockpiled, and introduce a new building or technology that consumes that resource. The ultimate resource sink, or wealth tax, is due to arrive soon in that world. The Tale will end, and the civilization vanish, taking all (or most?) of its works with it.

[Edward Castronova]

40.

Edward Castronova> Richard's connection between the idea of taxing and destroying the wealth and the Fed buying dollars with securities is 100 percent accurate.

JNK>Are you talking about my post, or did Dr. Bartle post something?

Arf, sorry. Yes, you were the one who connected it to the fed. Mea culpa.

[Richard Bartle]

41.

Mithra>I'd cache my wealth somewhere, either by handing it to a mule, or hiding it behind a tree or wall, so that the server regarded me as 'poor', go buy items for the cheap price, retrieve my wealth

Ah, but at this point it wouldn't be wealth any more. If you have 1,000 UOC and drop 500 UOC behind a tree, that 500 UOC is "500 UOC or 50% of your wealth, whichever is the lower". So, you go off an buy something for 200 UOC, leaving you with 300 UOC, then you go back to the tree and pick up your 150 UOC.

Note that in this situation you could drop 500 UOC then immediately pick it up and you'd only have 750 UOC, so it's not going to please many players. Then again, this is never going to be implemented anyway.

>Unlike currency, the sword won't depreciate having been handed to the customer.

Well we could get into "objects wearing out", ie. yet another tax, but that's a bit off the subject.

>If a poor player dropped 2 gold and a rich player picked it up, would it covert upwards to 500 gold?

No. Although in a "pure" system it would, in the version I was suggesting it wouldn't. "2 gold or 100% of your wealth, whichever is the lower".

>And if so, what then becomes the point of acquiring wealth or even having currency?

Because the price of objects contains a fixed component. It's not that swords cost 0.01% of your wealth, because that way you could buy an infinite number of them (well, infinite as far as rounding errors will take you). Rather, it's that the sword has a base value and a tax. In real life, we have sales tax on the base value; what I was asking about was the implications of a sales tax on how much cash you had, which itself was variable (something like a loaf of bread might be +0.001%, whereas something like a crown might be +0.1%).

>Would PD be appropriate in the currency/item economy, when the level/skill/item-usage economy remains traditional?

No, it wouldn't be. Bits of it may, however, have parts to play, which is what I was investingating.

>Will players bother to play your game if there is no opportunity here to excel?

Well no, of course not. That's why, in my opening post, I said "players would never accept this".

Richard

[Ryan Kelln]

42.

Just wanted to point out a nice correlation that lead to some thoughts about virtual economies:

Richard Bartle points out that, "it seems that price discrimination as a concept is ripe for exploits, except in the very particular case when what you buy is non-transferrable." He gives the example of character paying for training.

Did you realize Richard this has equivalence to the real world price discrimination for British education?

Galbi mentions that "price discrimination tends to be found in transportation pricing, hotel pricing, communication services pricing, and the pricing of information products." With the exception of information products these are are examples of difficult or untransferrable products.

Reading the other sources here I didn't see much mention of the transferability of the product affecting the viability of price discrimination. Richard is right though - if you cannot transfer what you are paying for price discrimination is unavoidable.

In a virtual world it be possible to make every item only transferable once. What sort of effects would that have on an economy? Do you still have an economy? I'm having trouble conceiving of the full implications... especially for currency. If currency is generated out of thin air (mob loot) then it could be accumulated but it couldn't be spent... which makes it more like current experience systems.

Suppose you can accumulate currency and purchase good with it, but instead of the money trading hands it just disappears. This is basically what happens with NPC shopkeepers, but it could be the same for player created items. Obviously there would have to be some other benefit besides wealth for the player crafters - perhaps crafter skill is tied to how much their items are used (isn't that what SWG is doing?)

So that sort of system could work for faucet/drain economies, but since price discrimination is more of a progressive tax I think it's more interesting in a closed economy... but if currency can't change hands then you have to rely on barter. In addition you can only barter objects you have created or a service.

Bartering services might actually be a great design in an virtual world. Players can create new services at will - the rich players will be the most active (and thus wealth is represented as how much is "spent" not how much is accumulated).

An example:
Alice wants a new Stupendous Sword. Her friend Bob can make one but the sword "recipe" requires Bossmob to be killed and 10 Trinkets to be collected. Alice and Bob make a deal Alice must kill Bossmob, collect the 10 Trinkets, and kill Bob's enemy Carol. When she has done all of the above she can return to Bob and he will hand over the Stupendous Sword. How does price discrimination fit in? If Alice is rich (she already owns Amazing Armor, Super Shield, etc) then the "recipe" for the Stupendous Sword could be more invovled. Newbs and casual players would then have lesser challenges for similar rewards.

Perhaps the Stupendous Sword is a "charged" item that can be recharged by killing HardToFindMobs. Alice can have other players kill the HardToFindMob for her to recharge the sword. This time if Alice is poor each HardToFindMob that is killed recharges more than if she was rich.

Creating an interface for bartering services might be challenging but you could even go so far as having "witnessed" undefined services.

Is there a gameplay benefit to promoting barter and services over a currency based economy? A "single transfer" economy might solve eBay problems.

Anyways, I never would have thought about this without this disucssion. Thanks!

[Bruce Boston]

43.

Frank> Bruce, perhaps this view could be rephrased as "simply because they don't perceive the fair value of paying more."

Possibly. The other possibility is that there is an oversupply of MMORPG fun in the market. Market price has as much to do with supply as it does demand.

Frank> If the average level gain rate per month is 2 level, then if you offer a differentiated pricing plan of 1.5x the monthly subscription fee to automatically gain 1 level per month but a cap at average rate (gain no more than average), players will value this pricing model and perhaps not yell about spending RL$ to gain in-game advantage because the only advantage is to catch up.

UO is doing this with Advanced Character Creation Service, and EQ is also going after extra dollars in the market through their Legends servers. But, I think you are correct in saying that players have a different attitude to ‘catch-up’ services then they do towards ‘get ahead’ services, especially when RL$ are concerned.

Frank> It would be great to implement a tax structure that lore consistent, but the structure also need to be consistent of player's metaphysical ideas of fairness-a big bag of socio-anthro-political philosophy stuff.

Although I haven’t played Horizons yet, from what I have heard the ‘Dragons’ race allows some players to play the role of ‘tax man’. As they have the ability to take items/currency out of the world permanently. In fact, the whole point of the race seems to be who can destroy the most wealth. At the same time, it is a very lore consistent role and allows players to decide ‘who has too much’.

Ted> In MMORPGs, the very rich pay the same monthly fee as the new player, and there's no in-world tax at all.

I’m not sure if that is entirely accurate. High end armor/weapons/food/transportation/etc/etc normally costs much more than low-end items, either in currency or time or raw resources. Silver tipped arrows always cost more than bone tipped, and you have to have silver tipped to kill stuff at the upper levels. Not only is this very lore consistent it is also a very common design practice.

So, maybe I’m missing the point, what design problem are we trying to solve with PD and progressive taxes?

Ted> The reason it is fair in RL is this: the central bank doesn't just take your money and throw it away. It induces you to give up some money now, in return for a promise that you will get your money back later with interest. Completely fair - it's a deal, not a tax. All it does is defer purchasing power to the future. And that is sufficient to reduce purchasing power now, which is all that matters

Getting players to take money ‘effectively’ out of the live money supply doesn’t seem to be a problem that hasn’t been solved, in fact MMORPGs have a better solution than the Fed’s, it’s called hoarding. There are a number of ways to encourage hoarding and it seems to have the same end result as US bonds, except that game developers don’t have to pay interest. But, the effects of hording in RL or VW have their own set of issues.

Jeremy> Can anyone tell me where he talked about this?

In particular, I was referencing his statement in the recent Warcry interview:
http://www.warcry.com/scripts/columns/view_sectionalt.phtml?site=15&id=102&colid=1712
Raph: “Don't react with horror over this, but the fact is that back when UO got launched, we probably underestimated the price the market would bear.”

Hellinar> The TN is backed by a “basket” of widely used commodities, canvas, iron etc. To obtain a 100TN note, you have to give the Bank one unit of each of these commodities, which is then stored in the vaults.

Absolutely Brilliant!! I have often though what it would take to create the global hard currency, and I have come to a similar conclusion. Simply that it would be possible, if you used a sort of hedge fund model with the currency being backed by a global basket of commodities. In any case, the fact that you were able to do this in a MMORPG and that it worked is very exciting to read about!

Ted> How about this: bank vaults require protection. The more things you have in there, and the higher quality, requires more guards. More guards cost more money.

Worlds of Warcraft is planning to use this very model, with players starting out with 5 slot bank accounts and being able to purchase additional slots up to 14 additional slots based on a progressive scale, so that the 6th slot is cheap, but the 20th slot will be very expensive. The rumor is also that they will compound this design logic with increasingly expensive containers that can be used in each slot.

-Bruce

[Estariel]

44.

The underlying premise here seems to be that it is desirable in a VW to drain the "excess" wealth out of the system, targetting the richest characters.

I suspect that this premise may be flawed for two reasons.

Firstly, that striving to achieve the sense of being comparatively very wealthy is what drives many players to play the game, and do the work to get there. If the wealthiest players, were made less wealthy somehow, would that not decrease their personal feeling of achievement (and being more uber than others), and therefore have a negative feedback on motivation and thus subscriber retention?

Secondly, this inflation seems to be built in by design to the game worlds, which are still being extended via expansions. EQ is about to deliver its 7th(?) expansion, which one might expect, like the previous ones, to make available better stuff than ever before. Why? So that the current player base will have to buy it in order to keep up with their peers.

The corollory is that, as before, more older equipment will become junk status. Very few items from the first two releases now have any kind of market value.

The ingame game economy is being deliberately tweaked to deliver out-of-game subscription revenue. The most active and high level players ARE paying more for the game, because their ingame social context requires them to have all expansions.

[Bruce Boston]

45.

Just to have fun with Ted’s letter to the IRS, I think the only way we could make MMORPGs ‘fair’ through PD and progressive taxes would be to require players to declare RL wealth and MMORPG experience in an effort to create a sort of ‘MMORPG Handicap’ similar to golf.

*************************************
Dear Sony,

I recently purchased a copy of SWG and would like to apply for the ‘uber trader but can’t type worth beans’ handicap.

I am also forwarding my IRS, eBay and Paypal records to certify my RL disposable income. Please use my previous experience in EQ as a reference to my ability to twink characters, work with currency traders and still get killed because I can’t remember the shortcut to my evac potions.

I understand that the ‘uber trader but can’t type worth beans’ handicap comes with a 50% tax on all purchases and trades, but NPCs will kill me at a 20% reduced rate. I think this is only fair, as I expect to be able to out trade almost 95% of the players in the Galaxy and would very much enjoy dying less often.

Sincerely,

Mahji
Ahazi Server

*************************************************************

-Bruce

[Edward Castronova]

46.

Bruce> Worlds of Warcraft is planning to use this very model, with players starting out with 5 slot bank accounts and being able to purchase additional slots up to 14 additional slots based on a progressive scale, so that the 6th slot is cheap, but the 20th slot will be very expensive. The rumor is also that they will compound this design logic with increasingly expensive containers that can be used in each slot.

Gee, now I wonder where they got THAT idea??? {wink}

Hilarious letter btw.

[magicback]

47.

Quick thought: connecting Richard’s progressive cost of training and Bruce’s handicap cost, does SWG have a progressive cost of cloning (not on the items but on the character)? Meaning, the more experienced the character, the higher cost of cloning.

Frank

[Paul "Phinehas" Schwanz]

48.

Hellinar> "One system that can work, in my experience, is player created currency."

I've always thought player created currency could be a very interesting method for promoting various kinds of economics-related gameplay. I wrote a concept piece that incorporated the idea of player-created currencies a couple of years back. It can still be found in the MUD-dev archives at the following location:

http://www.kanga.nu/archives/MUD-Dev-L/2001Q2/msg00126.php

The write-up covers a lot of historical and cultural design ideas providing a foundation for the economy, but you can skip down to the section entitled "The Enurium Economy" to see a outline of what seems to be echoed rather closely by your description of Trade Notes. It is very gratifying to hear that ATITD has put a design structure into place that would allow for the sort of player created currency that I envisioned.

--Phin

[Richard Bartle]

49.

Ryan Kelln>Did you realize Richard this has equivalence to the real world price discrimination for British education?

Yes, that's exactly what I had in mind when I referred to working applications for price discrimination.

Richard

[magicback]

50.

Bruce,

RL PD can work for additional catch-up services with just in-game data.

---
Are you a hard working person, but don’t have time to keep up with the Jones in Everquest?

Don’t Fret! For $4.95 per month, you can be AVERAGE and never fall behind the Average Joe.

Here’s how this works…
---

It’s like people buying a combo cable-PVR-box to keep track of their favorite shows. I still don’t see it as PD, but as value pricing. Players will accept value pricing, but not anything close to “D”.

But going back to Richard thought experiment on PD…

Based on the thread so far PD based on wealth can be reasonably implemented on non-transferable goods where the “state” is the provider. Example: WOW implementing PD pricing for vault space.

This implies that PD for consumption is more acceptable than for transaction. So, you can have a % tax on the transfer of possession, but a PD tax on the use of the possession. This structure is similar to having min. and recommended levels for items.

Frank

[DivineShadow]

51.

Quick thought... General observation... On VWs that have combat, while higher-level players can tackle tougher opponents, they also expend (drain out) quite a bit more resources doing so. In an absolute scale they do get more value out of the higher-level encounter, but it's generally not a straight-line proportion. A low-level character might get a return of 150% from doing something 'at their level' and a higher level character would see a 125% return of something 'at their level'. In essence, that higher-level character is being 'charged' more, and the more a character tackles something out of their league the more they are 'charged' down past their break-even point, at which time they decide the creature is too difficult/not worth it.

Interesting...

[Bruce Boston]

52.

I guess it all depends on how we define Price Discrimination, and how that differs from Product Differentiation or Yield Management, both of which are sub-categories of the above mentioned Revenue Management.

Personally, I think there are some fundamental differences between the philosophical reasonings upon which Price Discrimination and Yield Management are based, to the degree that if the the two are defined accurately, they don't intersect at all. However, in reality, people often use PD to describe practices that are really YM.

On the surface these two terms may look alike, essentially creating different prices for different customers, however, the process that each uses to set price and how these practices are justified are very different.

YM assumes that the supply and demand of a product are always in fluctuation. For example, the demand for red roses is much higher on 2/14 then say 2/15, therefore the intersection point of the supply and demand curve, ie price, shifts considerably somewhere between 2/14 9PM and 2/15 9AM. In the same way, a fully loaded Tercel and Camry have very different prices, as would a MMORPG bank account with 5 slots and 15 slots.

Factors that cause these types of supply and demand shifts are innumerable, including differences in time, location, condition, quantity, transactional costs, secondary risks, etc, etc, etc. And, again YM is always there to justify that if there is even the slightest difference in any two items than the supply and demand for those two products are different, and so should the prices also be different.

On the other-hand, Price Discrimination works on a different level. It justifies pricing differences not by differences in the product, but differences in the customer. Personally, I have only heard one reason that I agree with for varying price based on the customer and that is when risk is involved. The classic example being insurance, where there is a very good reason to give one price to one person and another price to another person, but even in this case, it is important that insurance companies carefully document these differences in risk levels.

On the other-hand, a classic example of 'Perfect Price Discrimination' is that a bottle of aspirin 'should' cost Bill Gates $1,000, and someone who is sick but very poor $1.

Frankly, there are more than a dozen reasons why I think the costs of implementing PPD in a market far outweigh any marginal benefits that a market might gain by using PPD.

To start, what makes PPD a more efficient way to allocate limited resources, or determine optimal production levels in a market, or encourage production?

Even philosophically I don't agree with PPD, as it seems to contain the notion that access to wealth carries with it obligations back to society; a statement that would only be true of that same access to wealth had caused that same society some sort of debt that needed to be recovered.

To take this the next step, we would then have to assume, that the biggest producers in any market were also the producers of the largest debts in that same society. This would infer that society was actually better off without uber-producers of value, or that these producers are not actually producing any value, or, even worse, that the market was only tricked into purchasing products that had no value, leaving society as a whole poorer and a few richer. Even if we were to assume that this was true, we would then also have to assume that the market was no longer safe with its own money and no longer able to make rational decisions, a statement that I believe to be far from true.

This is not to say that I think all practices of all producers are, or have been, totally ethical. However, I do believe that when producers use unethical practices that those producers should be held responsible personally, instead of simply spreading societiy's loss across all producers because thats easier.

I am also not saying that governments don't use PD, as its a pretty common practice, but I think we would all agree that 'common government practice is not equal to economically efficient'.

For the most part, the productivity of the producers in any market is a major factor in determining any society's standard of living, and charging $1000 for a bottle of aspirin is a very strange way to show appreciation.

-Bruce

[Mithra]

53.

Bruce, I couldn't imagine a more eloquently worded argument against PPD.

[Julian Dibbell]

54.

But Bruce, isn't there a pro-PD argument based not on ethics but market efficiency? Here's what I'm remembering, probably not very well:

Say I'm selling a market report I've written up. This is a product some people will pay a lot for and others will pay not so much for. Let's say for simplicity's sake that there are 5 people willing to pay $100 for it and 5 who will pay no more than $50. If I set the price at the $75 average, then I make $500 and 5 people get to see the report. But in a world where I can know just what everybody's price is and set it accordingly, I make $750 and 10 people get to see the report. Greatest good for the greatest number -- slam dunk, no?

[Edward Castronova]

55.

Charging all consumers their willingness to pay (perfect price discrimination) does not affect total production in a partial equilibrium analysis (i.e. one in which larger systems, such as norms of fairness, are assumed to remain unchanged). Each consumer is still willing to buy the good, they just pay their WTP minus a penny for it. What PPD does is transfer the consumer's surplus (the difference between willingness to pay and price) to producers. Therefore, since efficiency is defined relative to production levels, PPD doesn't affect efficiency of markets. It does affect the distribution of the benefits of markets. Under PPD, all net value in the market (the difference between consumers' valuation of goods and what it costs to make them) ends up in the pockets of producers.

To riff on Julian's example for a minute, if you have 10 people, five WTP $50 and five more WTP $100, and you have a report whose marginal cost of production is zero, then profit maximization without PPD implies a price of $50, for a profit of $500. Five consumers pay exaqctly their WTP and thus make no surplus on the trade - they get what they pay for. But five others actually capture some surplus - they were WTP $100 but only had to pay $50. So consumers as a whole capture $250 in surplus, while producers get $500 profit. And this profit-maximizing strategy also results in efficient production - 10 reports. If you allow PPD, then profit maximization still involves 10 reports, the efficient number, but now producer profits are $750 and consumer surplus is $0. PPD doesn't change the number of reports traded - profit-maximization itself ensures that the optimal number, 10, will be sold - but it does affect how producers and consumers gain from the trade.

In a general equilibrium (i.e. one that, unlike the above partial equilibrium, allows everything in the universe to react to a situation), transferring well-being from consumers of goods to producers of goods will have some serious feedback effects. I.e., consumers will generate norms against PPD, or they will just get mad and pass laws to prevent it.

[Julian Dibbell]

56.

Clarifying, Ted. Thanks.

But one question: In the non-PPD world, why do I bother selling 10 widgets at $50 when it's less hassle to sell 5 at $100? Presumably I do so because there are other producers willing to undersell me if I don't. But doesn't this example introduce the possibility of local-optimum problems in non-PPD economies? In other words, if the market is split between a small niche of very-high-WTP buyers and a broad swathe of very-low-WTP buyers -- doesn't the competition have to get *very* fierce before sellers will come down off the high end? And doesn't that situation, at least, recommend PPD on efficiency grounds?

I'm imagining that the international market for AIDS drugs (modulo various complicating factors) might be an example of such a situation. For the sake of relevance, though, I'm hoping that someone can imagine a virtual-worldly situation that might just as well make the case.

[Cory Ondrejka]

57.

Julian> I'm imagining that the international market for AIDS drugs

Professor Lessig talks about this example in the most recent Wired.

[Edward Castronova]

58.

Right Cory: the issue there was, why don't drug companies cut their prices to sell to poorer nations. One answer, that Larry mentioned, is that a lot depends on whether the low-price people can sell the goods to the high-price people. If they can, then the company can't sell any goods to the high-price people; it only finds that it sells 10 to the low-price folks. PPD depends on the absence of a resale market (and in the VW context, to assume that is, well, not good, as Bruce keeps reminding us). With resale markets, the outcome is still efficient, but now the gains are split three ways: producers, high-price consumers, and resellers.

But assuming there's no resale possibility, what about the hassle factor? That is, what about the possibility that, if you have to pick one price, you might find it sensible to set a price so that only the high-WTP people buy, but not the low-WTP people. This would be because selling to the lowbies is a costly proposition in some sense. And this is, indirectly, the same thing as the idea that there are some competitors who drive the price down. See, if there ARE such competitors, it must mean that the core production costs are low at the margin; my competitors are pricing according to these low costs, and I am trying to do a mark-up. In that case, yes, the price comes down and the lowbies get to buy the goods. But the premise of the question here was, that it is a hassle to sell to the lowbies. I read that as 'there is a marginal cost of production such that making 5 units is fairly cheap, but making 5 million gets awfully expensive.' Under the circumstances, yes, the hassle factor *would* make you restrict your market to the high end.

In fact, in any normal supply-demand diagram, the first units made and sold - the quantities to the far left in the diagram - are ones that are highly valued (reading off the demand curve, the WTP) and also cheap to make (reading off the supply curve, the marginal cost). Basically, the presence of increasing marginal costs (upward sloping supply) would always cause the market to settle on a price that will exclude some potential buyers. It's just too much hassle to make and market the good to them at a price they are willing to pay.

[Bruce Boston]

59.

Hey Julian,

Again, I think for all practical purposes most economics books that I have seen treat Perfect Price Discrimination and Yield Management as essentially the same concept. And maybe its my business background that drives my desire to define them as two very different concepts.

However, if I were to label the example that you described, I would say that it is a YM practice not a PPD practice. Why? Because in most PPD examples, the arguments reads the same, having a number of 'shoulds' and 'becauses'. For example, Bill Gates 'should' pay $1000 for a bottle of aspirin 'because' he can afford it, and someone who can't afford aspirin 'should' only pay $1 'because' they need to get to work the next day.

YM arguments sound very different to me. In fact, as I mentioned above they are very closely linked to very precise calculations of supply and demand. And, contain words closely related to 'willing', which you used a number of times.

Just to take a step backward, I think the assumption is, yes, the standard demand line for any product is typically a curve. I have yet to find a single product of the several billion products on the market that every single person on the Planet is willing to pay the exact same price for.

With most YM techniques there is a basic pattern: start with the top bidder and work your way down the demand curve until the marginal cost outweighs the marginal benefit of producing the next unit of the product. This pattern is used by almost every industry from Automobiles to Airlines, from Taxis to Video Games.

Going back to your example, and using YM, you would then start by selling the first 5 copies at $100. At the very second that you finalized the sale of that 5th copy, both the supply and demand curve for that very product would fundamentally change. Where the high end of the demand curve was $100, it would now read $50, with 5 potential customers willing to pay just half for the next copy of that very same report you just finished selling at $100 each.

The pattern may actually repeat itself when you have sold the next 5 copies for $50 each, in this case resulting in a demand of zero at any price. And no, these are not simple calculations to make in real life, and yes, at some point you will have to worry, if you sell a copy of a report for $10 are you incurring some risk that you will no longer be able to sell your next report for the full $100 the first time around? In most cases, it is this perceived cost of risk, call it risk of brand degradation, not the actual cost of reproducing another copy of the report that is the final production stopper for most report/automobile/hamburger/etc factories.

Additionally, the most sophisticated uses of YM included the practice of product differentiation. Where if used correctly, you would never sell more than 5 copies of the same report.

If we are going to go ahead and assume perfect knowledge, then the most profitable perfect knowledge would be knowing exactly what each potential report buyer wanted.

This could be a small differentiation like time; this month($100) or in two months($50). Or format: PDF ($120) vs Paper($100) vs Fax($50). Or credentials, Special preview section by Dr. Castronova($150), vs. personal notes/signature by you($120), vs. typed name of your assistant($60), vs. rubber stamped by the company mail room clerk($30). Or content; includes information on 8 VWs($150) vs 5 VWs($100) vs 2 VWs($30). The more adept a company is at targeting and delivering the biggest bang for a customer's buck, the more likely they are to be very successful at using YM.

On the far other side of the spectrum you have governmental PPD, the classic examples being progressive taxes, like (I'm assuming) the above mentioned British education system. And, yes, it is so easy to recognize any argument in favor of any governmental PPD practice because it almost always starts out the same: Mr. or Ms. So and So 'should' pay more/less 'because' blah, blah, blah.

There is never the question of 'why', or 'willingness', or 'value', it is almost always simply assumed that Mr. and Ms. So and So, can/can't afford the real costs, therefore they 'should' pay more/less.

And, God forbid anyone ask, in an open forum, if the government could use a variable pricing system to openly offer different levels of service at different price points, as if to assume that the demand for government services might also be on a curve.

So again, maybe calling what businesses do with YM and Product Differentiation and what governments do with PPD isn't as simple as calling one 'Yield Management' and the other 'Perfect Price Discrimination', but to lump the two together under the same PPD umbrella doesn't seem very effective as they are such different practices based on some very different philosophical rationals.

-Bruce

[Bruce Boston]

60.

Dr. Castronova> But assuming there's no resale possibility, what about the hassle factor?

Yes, the hassle factor!

Actually this holds true in markets with resale as well. When I was in Japan working with Toyota, I sold some $1.4M in Japanese Pokémon and Japanese Beanie Babies on the side through my living room on eBay. I often noted that it was goods that were very easy to store, package and ship, like baseball cards or small collectors items that had the highest resale values because they had the lowest hassle costs.

The reverse was also true, in that no matter how valuable a collector's item is, if it has a high transactional cost, the resale value, and eventually the primary market value drops very quickly.

This principle was certainly true even in the resale markets of EQ, where before they had the automated bazaar the prices were at one set of levels, and as soon as they introduced a system with a lower hassle cost, the prices all jump to a different set of levels.

-Bruce