Virtual World Economy: It's Namibia, Basically.

In my original study, EverQuest's productivity was measured at about $2,000 in terms of annual GDP per capita. This was misreported in the press as "Economist finds virtual economy bigger than Russia!" Absurd. No virtual economy could outproduce, in gross terms, Russia's hundred-million-worker economy. I wonder how many sensible people have rejected the whole line of research on the basis of this misreporting.

But let's address the question head on. Is it possible that virtual world economies might make a gross impact equivalent to that of a real country? Actually, yes. Virtual worlds are already that big, economically speaking. More accurately, some unfortunate countries remain that small.

Let's assume in the absence of other data that virtual worlds, on average, have a GDP per capita of about $2,000. As in the original study, this figure would be arrived at by taking the total production of valuable goods and services, whether marketed or not, and using real-dollar exchange rates from places like eBay to assign a value to that production.

Second, let's assume the current population of all virtual world users is about 20 million, and that the average user spends about 20 hours a week in the world. This would imply about (20)x(20)x(52) = 20.8 billion in-world hours annually, or (20.8b)/(365.25x24) = 2.37 million full-time equivalent people (people living there on a 24/7 basis).

I tend to think of these as conservative estimates, but of course I could be wrong. If they are not conservative today, they will be within five or six years.

Here is a list of real countries with populations between 1 and 3 million:
Panama 2940000
Mauritania 2629995
Jamaica 2617000
Oman 2538000
Mongolia 2449000
Latvia 2338000
Kuwait 2328000
Macedonia, FYR 2038000
Namibia 1985000
Slovenia 1964000
Lesotho 1777000
Botswana 1712000
Guinea-Bissau 1447000
Gambia, The 1389000
Estonia 1358000
Gabon 1315000
Trinidad and Tobago 1304000
Mauritius 1212000
Swaziland 1088000

And here is a list of real countries with GNI (Gross National Income, the Bank's preferred statistic) per capita between $1,000 and $3,000:
Dominica 3000
Belize 2970
Brazil 2830
St. Vincent and the Grenadines 2820
Jamaica 2690
South Africa 2500
Turkey 2490
Marshall Islands 2380
Maldives 2170
Europe & Central Asia 2160
Fiji 2130
Russian Federation 2130
El Salvador 2110
Peru 2020
Thailand 2000
Tunisia 1990
Micronesia, Fed. Sts. 1970
Suriname 1940
Romania 1870
Colombia 1820
Namibia 1790
Bulgaria 1770
Guatemala 1760
Jordan 1760
Algeria 1720
Iran, Islamic Rep. 1720
Macedonia, FYR 1710
Kazakhstan 1520
Ecuador 1490
Egypt, Arab Rep. 1470
Albania 1450
Tonga 1440
Samoa 1430
Serbia and Montenegro 1400
Belarus 1360
Bosnia and Herzegovina 1310
Cape Verde 1250
Swaziland 1240
Morocco 1170
Paraguay 1170
Syrian Arab Republic 1130
West Bank and Gaza 1110
Vanuatu 1070
Philippines 1030

(Source: 2002 World Bank World Development Indicators)
(Sorry about the formatting, I am too lazy to do more than cut and paste.)

The following countries appear on both lists:

Jamaica, population 2.6m, GNI/capita $2,690
Namibia, population 2.0m, GNI/capita $1,790
Macedonia, population 2.0m, GNI/capita $1,710
Swaziland, population 1.1m, GNI/capita $1,240

For comparison:
Virtual worlds, FTE population 2.4m, GNI/capita $2,000

In other words, the virtual economy would be roughly equivalent to the economy of Namibia or Macedonia, in terms of its overall role and position in the real global economy. It may be catching up to Jamaica. No, not Russia. Not even close. But not chopped liver, either.

Comments

[Nathan Combs]

1.

this figure would be arrived at by taking the and using real-dollar exchange rates from places like eBay to assign a value to that production.

Here is a question for the economists among us, are these numbers "double-booked?" E.g. Ebay transactions are underwritten in real money which is reflected in real economy statistics. So in Virtual Namibia's case, is its GNI is also reflected in the US (etc) GNI?

[Edward Castronova]

2.

The underlying concept here is GDP, which is a measure of production. As such, money doesn't enter the question at all - money is only a means to measure, assess, or assign value to production. It isn't production itself. Nobody eats money or wears it or drives it. You only use money to get things that you eat, wear, or drive. And GDP is about measuring the number of those things that are produced. Since it would be too cumbersome to list the actual numbers of things that get made - 10,000 cars, 2 million oranges, four billion silly hats - we use the current prices of those things to enable adding up. So, it becomes $4b worth of cars, $400m worth of oranges, $10b worth of hats. That's the only reason GDP appears to be monetary, to enable adding up disparate things.

In this case, the things in question are quite radically not subject to question as to locus. Things made in virtual worlds only exist there. Things made in the real world only exist here. So there's no double-counting; no case in which a virtual good gets counted once in the virtual world and once in the real world. It only exists there, and is only counted there.

See, what's been done here is to assign a real-world value ($2000) to the sum total of the virtual things that a player would make if she were to 'live' in the virtual world all the time for an entire year. This involves not the volume of eBay trade, but rather price of typical goods there. We use eBay prices to put values on things like currencies, swords, experience points (a form of capital). Then we add those disparate things up, to find the total value of virtual goods made inside all virtual worlds, regardless of whether or where they may be traded (eBay, in-world, or through an implicit transaction between the player as producer and the player as consumer).

(Speaking of the latter, economists will tell you that you do the same thing when you, as homeowner, rent housing space to you, as home occupier. Value is created but just not realized in an observable market transaction.)

(In the US, economists once proposed putting a sales tax on the homeowner rental transaction and an income tax on the homeowner's implicit income.)

(Because of this and other science-gone-mad proposals, typical of the scientist-gone-mad nature of many economists [present company excepted], economics is not a very well-respected field.)

I digress.

This all implies that the "GDP of cyberspace" by these calculations comes out to 2.4m x $2000, or $4.8 billion. There's plenty of countries with that level of GDP. And nope, the stuff on eBay doesn't appear twice.

[Dave]

3.

And nope, the stuff on eBay doesn't appear twice.

Unless someone figures out a dupe exploit.

[Bruce Boston]

4.

Reviewing these numbers I'm wondering if we are not forgetting to add the other side of the coin.

Doesn't 'government spending' account for approximately 1/3 of the US economy? I would think that the percentage is much higher in many smaller countries.

So, now I'm wondering what percentage of the "GDP of cyberspace" is government spending (developer spending). Does this get included? Is there any spending being done, or is everything being imported by the government from outside? As these economies grow, do we expect this internal and external government spending to increase or decrease? And, I wonder how this spending will effect the political systems inside and outside of these spaces?

In any case, it should be interesting to unfold these questions over the next few years.

-bruce

[Nathan Combs]

5.

Edward>

Things made in the real world only exist here. So there's no double-counting; no case in which a virtual good gets counted once in the virtual world and once in the real world.

Hmmm, interesting. In the spirit of pushing the envelope a bit here: is it entirely true, what about services? Consider where a RL product whose value is agumented via a VW "service" - e.g. brand advertising in SL?

[Tobold]

6.

Isn't there a major error in that calculation? You arrive at the $2000 GDP per capita by multiplying the total amount of platinum with the the value derived from EBay sales. But I think that EBay "exchange rate" is artificially high, because only such a small part of the platinum is actually traded.

Imagine a third world country growing a billion of bananas, mostly for local consumption. But 100 of these bananas are exported, and are sold for $1 each in a New York delicatessen. Would you then calculate that the third world countries GDP increased by a billion dollars? Or did it just increase by $100?

[CherryBomb]

7.

A billion bananas are a billion bananas, regardless of who eats them. If you can sell them for $1 each, they are worth a billion dollars. In developed countries, people produce essentially nothing for their own use; they get paid money to do things for other people, and they pay other people money to do things for them. In this way, MMOGs resemble subsistence agriculture economies, where most of the labor output is not monetized.

[ren]

8.

CherryBomb > A billion bananas are a billion bananas, regardless of who eats them. If you can sell them for $1 each, they are worth a billion dollars.

Riiight…
A glass of water is worth a lot more in the middle of the desert than a one stood next to a free water cooler.

Or to take an example that impacts us all right now – how much is a billion barrels of oil worth?

I think Tobold is probably right, the calculations would have to take some price curve stuff into account, but as an indicative set of numbers I think that Dr Cs figures are enlightening as always.

[CherryBomb]

9.

Ahh, but if you sell a glass of water in the desert, you're not really selling a glass of water, you are selling the transportation of it. Otherwise, I could make a fortune from my kitchen tap. If I grow bananas in my back yard and eat them myself (I have actually done this, by the way), the only rational way to value this production is what I could sell them for in my front yard (or on e-bay). This does break down if you try to value a large amount of production that is consumed directly by the producer, like in subsistence economies or on-line games. If everybody grew bananas in their back yard and tried to sell them, there would be no market. If everybody tried to sell their on-line stuff at the current prices, the market would get very elastic very fast, so it can be misleading to extrapolate sale prices and try to estimate the size of the entire game economy.

A billion here, a billion there; pretty soon, you're talking about real bananas.

[Bruce Boston]

10.

Tobold> "But I think that EBay "exchange rate" is artificially high, because only such a small part of the platinum is actually traded."

As is the case with any exchange rate. For example, not all Japanese Yen are exchanged for US$.

-bruce

[Staarkhand]

11.

Look to Dr. C's first response where he explains that the $ value is not indicative of actual sales, but just a common unit to which real production can be converted for comparison. Also the value of something is not predicated on the ability to instantly liquify said asset.

[Edward Castronova]

12.

Bruce> So, now I'm wondering what percentage of the "GDP of cyberspace" is government spending (developer spending).

I am having trouble thinking about what government spending is in this context. Within the virtual world, there is no government spending because there is no entity that taxes the users and then uses the gold pieces to purchase things or services (Second Life may be an exception).

If the developers spawn a new character with gear, that would be government production. And it would be valued the same as private production, and therefore it's already been included.

[Edward Castronova]

13.

Nathan> Consider where a RL product whose value is agumented via a VW "service" - e.g. brand advertising in SL?

I still think the service can be located, accurately, as 'within the virtual world.' GDP theory insists, by the way, that the location in which the good or service is produced is where its value is supposed to be located (for the purposes of this calculation). If Honda pays for a car to be made in America, and the car is then sold in Germany, the car's value enters America's GDP. It's a definitional thing.

[Edward Castronova]

14.

Tobold> Imagine a third world country growing a billion of bananas, mostly for local consumption. But 100 of these bananas are exported, and are sold for $1 each in a New York delicatessen. Would you then calculate that the third world countries GDP increased by a billion dollars? Or did it just increase by $100?

It didn't increase - the GDP was always $1 billion. 1 billion bananas were produced, right? And they had some value, right? Regardless of who consumes them, or what was paid, they are still yummy. But we just don't know what that value is, in dollar terms, until the bananas get sold in the deli. At that point, we have a valuation for the bananas - $1 each. In fact, even if only 1 banana were sold, we could/should/are forced to use that $1 price from that one transaction, to value all the bananas made. Economic theory argues for the accuracy of this step - called 'shadow pricing' - on the basis of a thing called the Law of One Price: if a good is available in two places, it cannot have a different price in them in the long run. Arbitrage will wipe out any price difference (buy at the cheaper place, sell at the more expensive one). This means that the guy who sold the bananas for $1 at the deli could have received $1 worth of goods or services for the banana in the little country; or, looked at another way, if he is willing to give it up for $1 in America, he would not be willing to part with it for less in Bananaland. And a banana is a banana is a banana: what was sold at the deli is no different from the bananas in the home country - they all have the same value, $1.

This abstracts from things like transportation costs and the possibility that bananas of different quality are selected for export. Fine. But those are quibbles. Basically, the price of gold on eBay is a fairly accurate estimate of the actual real-world value of gold to people in virtual worlds.

[Edward Castronova]

15.

Tobold> "But I think that EBay "exchange rate" is artificially high, because only such a small part of the platinum is actually traded."

This is another issue I encounter frequently - if everyone sold their characters on eBay, no way would they be worth $300 or whatever. By the same token, if everyone sold their Martha Stewart stock (and they should), its price would fall to zero. But we are asking about the value of the stock to each person who holds it today, and for each individual, they face a simple choice: keep it or sell it at the current price. Those that keep it must therefore value it at least as much as the current price. And therefore we can go ahead and use the current price to add up the total value of all Martha Stewar stock, to estimate the gross value of her company. That's how it is done in financial economics, and also in the area I studied in, called 'welfare economics,' which is about valuation of the impacts of government programs. Valuation theory is actually pretty standard in econ. The controversies are not about how to do it, but whether to do it at all. For example, I could calculate a cost of terrorism by adding all the physical damage, plus the cost in terms of lost human life. To get the latter, I could use some techniques (far funkier than anything I've ever applied to virtual worlds) that put a price on human life. One method: people expect more wages when they work at riskier jobs; it is as if they are trading the risk of death against cash, and from that trade-off you can infer a price that they would be willing to accept to have a 100 percent risk of a deadly accident - what they need to be paid to die. Economists quite casually accept that kind of number (I think it comes out to $2 million in most studies) as the value of a life, and fold it into estimates of the costs and benefits of pollution abatements, safety regulations, health care programs, and so on. There's even a price put on QALYs - "quality adjusted life years" so that if you get knocked in the head and wind up a little mumbly and slow, your value of life falls from $2m to $1.8m for the purposes of cost-benefit analysis.

Disclaimer: This price-of-life stuff is all barbaric to me.

But the point is, using eBay to price out virtual world production is well within the limits of standard practice. It's the pricing of life and art and the environment that gets iffy; pricing gold pieces is a pretty trivial application.

[Bruce Boston]

16.

Dr. Ted> "there is no government spending because there is no entity that taxes the users"

So, what's the argument against (for) considering subscription fees and upgrade fees a type of tax on the Users?

Today currency/goods are commonly sold by players, but I think in just a few years these same currencies and goods may be sold by developers, or be included in 'premium subscriptions'. Just looking ahead would this take these goods/services out of the GDP calculation because of whom they are sold by?

-bruce

[CherryBomb]

17.

You can do that kind of analysis eith MarthaStewartOmniLiving, but I don't think you can get away with it here. This is not a stock market. It is somrthing....else.

[Tobold]

18.

If 1000 Namibians went and exchanged their Namibian dollars for US dollars, it would not affect the exchange range very much. If 1000 Everquest players went and sold their platinum on EBay, the "exchange rate" would plummet significantly. Because the EQ PP is significantly less traded than the Namibian dollar.

I do not believe that the Law of One Price applies to the EQ platinum piece. Because there is a large majority of players that do not even consider it as a currency that could be exchanged for US dollars, they just consider it as play money. The EQ PP has some value to them, but that value is much lower than the EBay exchange rate would suggest. If there was not the hassle to actually exchange it, a "transaction cost" barrier, they would all gladly trade their PP for dollars at the current going rate.

[Richard Bartle]

19.

Edward Castronova>And a banana is a banana is a banana: what was sold at the deli is no different from the bananas in the home country

Well, except that the banana in the deli is in America and the banana in the home country is hanging off a tree in the yard.

In general, farmers get only a small proportion of the final price - 10% if they're lucky. They sell their produce to a buyer, who buys all the bananas in their area at rock-bottom prices because there's no-one else they can sell them to. The buyer sells the bananas to an export company, which ships them over to the USA and sells them to a distributor. The distributor drives them round in a van and sells them to the deli. The deli sells it for $1. So bananas are worth $1? You could get 10 bananas for $1 if you went to bananaland. Bananas are worth 10c, and getting them to you is worth 90c.

It may be standard Economics practice to regard a banana in bananaland as having the same value as a banana sold in a New York deli, but that says more about the way that Economics is studied than it does about the price of bananas.

That said, I agree with you that it's probably OK to use auction site prices as an estimate of how much people value their virtual goods. There are transaction costs, but they're not are all that great. Sure, the auction sites take a cut, but it's not 90%.

Richard

[Dan S]

20.

Richard,

The point isn't the "value" of the banana at the point of consumption, though, when calculating GDP. It's at the point of production. All of the transactions necessary to move the banana from the point of production to the eventual point of consumption are additional labor (and production in many cases) added to that initial production effort.

So a banana produced in bananaland and a banana produced in bananaland and shipped to NYC to be sold in a deli there are the same when in bananaland for purposes of calculating the GDP of bananaland. And the price used in arriving at that valuation is one calculated before as many external add-ons are tossed in as possible because all of those extras clearly make their own contributions to the GDP. To include them in the banana's valuation would inflate the GDP by duplicating, triplicating or more many elements.

It's not exact, but there is method to its madness.

[Richard Bartle]

21.

Dan S>the price used in arriving at that valuation is one calculated before as many external add-ons are tossed in as possible because all of those extras clearly make their own contributions to the GDP

So if a banana costs 10c in bananaland and $1 in New York, are you saying we calculate the contribution of bananas to bananaland's GDP at the rate of 10c per banana of $1 per banana? Ted seemed to be saying the former, but you seem to be saying the latter.

Richard

[Will Leverett]

22.

Could "government spending" in this sense be the costs of providing the service? Anywhere from 40% to 60% of a subscription cost goes straight to supporting the game (this includes hardware, bandwidth, and admin/GM support).

Shouldn't that be considered into a GDP of a virtual world?

[Edward Castronova]

23.

Another thing about GDP is that it measures the value of "final" goods and services. If I buy an egg for 10c at the store and eat, the 'egg' component of GDP goes up by 10c. If I buy the same egg for the same price and take it to my restaurant, cook it, and sell it to a customer for $2, the 'restaurent food' component of GDP goes up $2. Dan's right to say that counting the egg's 10c and also the $2 of restaurant food would be double-counting: the 10c price of the egg is included in the price I charge to the customer.

You measure the value of production at the point it reaches its final use. In the case of virtual worlds, we only observe one price point: when the goods are traded for US dollars. It's the only price we have. True, we could use in-world markets to get prices. To do that, we would have to assume that the in-world markets function properly and produce rational, reasonable estimates of the value of things. Personally, I believe that is true. But I knew that most of my readers would never accept that at face value. "Norrath produces a GDP of over 10 million platinum pieces per annum." Few readers of the Wall Street Journal would take notice, even if I were to show that a platinum piece trades for a penny or two. The analog is to the old days when the Communist countries used to announce their GDP, and nobody believed them because everyone knew that their internal markets never had prices that reflected the actual value of things. If you don't trust the internal markets, you have to go to markets you can trust. AND everybody HAS to trust eBay, it is by far the most successful market in human history. I would assert that this method of using external markets to value internally-produced goods is actually more valid in the case of virtual worlds that it is in its usual application, which is to 'broken' economies like Cuba, North Korea, and so on. The valuation markets for those countries - export markets - are also messed up. And there are notorious problems using real-world exchange rates to value real-world currencies, because of the prevalence of specualtion on macroeconomic indicators and such. Compared to all that, eBay is a wonderfully accurate market. I suspect we know much more about Norrath's market than Cuba's.

[Edward Castronova]

24.

Will> Shouldn't [game support, subscription fees, dev salaries, etc.] be considered into a GDP of a virtual world?

Chis would be analogous to the following situation: Fritz lives in Germany but works in France. When he works in France, he produces X mopeds per day. Mopeds are observed to sell at a price of P francs per moped. Therefore the value of Fritz's daily contribution to GDP is P*X. Now, since the mopeds were produced in France, Fritz's contribution is actually to France's GDP, not Germany's. Someone points out, however, that it costs Fritz Z German marks, each day, for a roundtrip ticket to France. The transportation is a valuable service, but note that it is produced in Germany, not France. It therefore belongs in Germany's GDP, not France's.

In other words, all the hard work that enables access to virtual worlds, and the fees that compensate for that work, reflect the production of entertainment
goods in the US, not the production of virtual goods in the virtual world. They belong in America's GDP, not cyberspace's.

[Bruce Boston]

25.

"They belong in America's GDP, not cyberspace's."

Sounds like a tricky point here.

Scenario 1: Dev makes 10k Holocrons, spreads 10k holocrons in 100k dragons, 1M player take 30mins each to kill dragons, Holocrons sell on eBay for $100 each. Result: Players in virtual world economy produced $1M in goods and services, or an average of $2 per player hr. [(10k*$100)/(1M*.5hr)]

Scenario 2: Dev makes 10k Holocrons, creates lottery to give away 10k Holocrons, Holocrons sell for $100 each on eBay? Developers in US Economy have created $1M in goods and services. [(10k*$100)]

So, first question: Why would the players get all the value creation credit in scenario 1, but not in 2?

This also leads to a second question, since these economies are mostly based on goods, and not services (despite what any eBay auction disclaimer may say: 'you are paying for the time it took me to pick the banana and not the banana itself'), I'm wondering if the value produced in VW isn't being overstated.

For example, Japan produces 1M cars. These cars are sold to US dealers for $20k each. Then the cars are marked up and sold for $25k each to consumers.

To then say that because 1M cars were sold in the US at $25k each, therefore '$25B of GDP was produced in the US economy' would be inaccurate. So, why is this same methodology OK in the case of goods in a VW?

-bruce

[Tobold]

26.

The value of something, as expressed in a market price, is found at the point where the supply curve and the demand curve cross. 1 platinum piece does not have the same value for everybody, so the people that value it lowest will be willing to sell it, while the people that value it highest will be willing to buy it. The market price is reached at the equilibrium, where the cheapest seller is demanding more than the most generous buyer is willing to pay.

In the case of the bananas, the equilibrium is shifted towards higher prices by the additional cost of transport and taxes. More people would be willing to buy bananas for 10c than there are people willing to pay $1. If banana buyers and sellers would be able to meet on a market without the transportation and tax cost, they would end up with a lower equilibrium price than $1, but probably slightly higher than the 10c.

In the case of the platinum piece, the equilibrium is also distorted, but not just by the EBay fees. To trade platinum pieces on EBay, you have to overcome some mental barriers. You have to consider platinum pieces as virtual property that can be bought and sold, you have to overcome peer pressure telling you that trading virtual property is dishonest, and you have to have a certain amount of trust in your trading partner (as it is well known that these contracts are not enforceable). All these barriers to PP trade shift the equilibrium towards higher prices. Many people value the PP much lower than the going EBay rate, but still wouldn't consider selling theirs (thereby moving the price down towards the real equilibrium), because of the barriers stated above. The PP market is not liquid enough to arrive at a true evaluation of the value of a PP, and thereby GDP. EBay exchange rate is the best value we have, but not good enough to calculate a proper GDP from it.

[Dan S]

27.

Yet as Ted points out, the valuation we get from eBay is arguably BETTER than those we use for real world economies.

I think we can and must consider those psychological elements you mention, Tobold, already factored into the price we derive from eBay. The already influence the price arrived at there by affecting the supply, demand, etc. That's the beauty of a "free market." It does the work for us.

[Edward Castronova]

28.

Bruce> Scenario 2: Dev makes 10k Holocrons, creates lottery to give away 10k Holocrons, Holocrons sell for $100 each on eBay? Developers in US Economy have created $1M in goods and services. [(10k*$100)]

But the developer's had to get the holocrons into the virtual world somehow. Even if the devs just created them in the database, they contribute $1m in virtual world production. The holocrons simply are not produced on Earth, so they don't count in Earth's GDP. And as for the dev's costs (labor, bandwidth, etc.) in making the holocrons available, that gets into an area called business investment - analog would be the expenses of Honda's Tokyo corporate office that contribute to the manufacture of a car in Ohio. I believe those head-office costs would count as Japan GDP. But the point here is, no part of the value of the car made in Ohio ends up in Japan's GDP. The car was made in Ohio. The holocrons were made on Bria. That's where they add to GDP.

[Richard Bartle]

29.

Edward Castronova>Even if the devs just created them in the database, they contribute $1m in virtual world production.

What if they're in the database but not accessible in the virtual world? They may be somewhere that people can't get to, and be released into a place where they can get them only under very specific circumstances.

This would make them like the diamonds in De Beers' vaults - worth millions, but only because they release them in a very controlled fashion. The difference is that whereas De Beers get their diamonds by digging holes in the ground, for a virtual world to get however many whatevers it wants could be as simple as incrementing a counter.

Suppose I had a never-ending bottle of Guinness. No matter how much I poured from it, it was always full. How much would that bottle be worth? Its existence wouldn't make the price of Guinness fall much, because even though I had the bottle I could only pour one drink at a time. However, it's an endless bottle, so whatever the price of Guinness, so long as it's non-zero my bottle should be worth an infinite amount of money.

Virtual world production is like that. The developers have an indefinite supply of everything. Players assign value to objects assuming that developers won't use their powers to flood the marketplace, but developers could nevertheless do that if they wanted.

There's something strange and interesting going on here (although since I'm offline all next week it'll be a while before I find out what they are!).

Richard

[Bruce Boston]

30.

Hi Dan, Hi Tobold,

Tobold> "EBay exchange rate is the best value we have, but not good enough to calculate a proper GDP from it."

There is a possibility that this is true. Take the following example;

Say an average of 100k people use my local supermarket. Of these people, 3% pay the supermarket $10 for delivery. Since this is the best data available of the value of delivery, I'm wondering if it's safe to assume that the people using the local supermarket are generating $1M in reportable economic value.

What happens when we vacuum our houses instead of pay someone $20 to do so, or drive to work instead of paying a taxi $20 to do so? Or comb our hair instead of paying someone to do so? Is it safe to assume that due to limited data points, that the value of anyone performing an activity is equal to the market value of the task? If so, I think that Earth governments are seriously understating the economic value created in their countries.

-bruce

[Bruce Boston]

31.

Richard> "However, it's an endless bottle, so whatever the price of Guinness, so long as it's non-zero my bottle should be worth an infinite amount of money."

Nope. For one, the amount of currency (paper, digital and otherwise) is finite. For two, this finite amount of currency gets regularly used on other things than Guinness. Therefore, at maximum the bottle is worth the total currency in the World minus all the currency that is being used for other purposes, which is a finite number not an infinite one.

But, actually it would be fairly easy to put a price on this endless bottle of Guinness. In the simplest of calculations it would equal the value of a trust fund that could consistently buy as much Guinness as the endless bottle produced. Even if this was 1 glass every 5 mins, there is a finite trust fund amount that could buy just as much.

-bruce

[Bruce Boston]

32.

"The holocrons simply are not produced on Earth, so they don't count in Earth's GDP."

Again, this seems tricky.

The basic argument seems to be that Holocrons don't exist on Earth, therefore they can't be produced on Earth, therefore if they exist at all (and have exchangeable value) they must exist in an economy outside that of Earth's.

By this argument I would expect the following to also apply.

When car is produced in Ohio, then sold in Ohio, then the car used in Ohio, and finally stolen in Ohio, the crime is not prosecuted in Japan, nor are damages awarded in Japanese yen, both of these events happen in Ohio because that is where the car 'exists'.

However, in the case of Holocrons, the courts seem to be prosecuting, awarding damages and defining criminal penalties in RL. Don't we have to somewhat assume that items exist where the law says they do, or is there now a new virtual property defense, 'these items don't exist on Earth, therefore Earth governments have no jurisdiction in prosecuting the crime'?

On the other hand, I tend to view VWs as subsets of the Earth economy, much like a city is a subset of the Earth economy. I don't disagree that value is being added to these items inside of the VWs themselves, and so the tricky part really seems to be deciding what value gets attached to which economy, and assigned to whom.

-bruce

[Cris Streetzel]

33.

The concept of devs adding to the value of the VWs GDP is interesting. Perhaps the point that needs to be enunciated is this: Economics is based on scarcity, but in VWs, scarcity of resources is completely artificial.

Perhaps the true difference between VW and the real world can be framed in terms of the age old economic debate: Should scarcity be measured in resources (bananas) or labour (time)? I think in the real world, we tend to think of it in resources, but you cannot import and export resources from the VW to the Real World. Perhaps VW economics need to be framed in terms of labour.

In that context, Dev created objects would impact GDP in the same way as the US Federal Reserve printing $1 Trillion dollars. It would not measured as part of GDP because it does not add to labour or increase labour's value.

Cris

[Bruce Boston]

34.

Hi Cris,

I often hear these types of arguments, i.e. 'the economic principles behind a VW economy are so different from RL economics, it must be a different economy all together'. However, I'm not sure I agree with either the premise or the conclusion. Let me give you some examples;

Cris> in VWs, scarcity of resources is completely artificial.

Yes, and no.

Clearly there are new objects being created all the time in VWs. Every time an upgrade package to EverQuest is created, I think that we can safely say that the number of objects in that VW have increased, and over the years they have increased substantially. Yes, it's true that once an object is produced in a VW, they can create allot of them, but as soon as the diminishing value of each item produced is equal to zero, the total value that can be produced in capped, and a finite number.

This actually isn't all the different from a number of industries. Most likely, a fair share of the assets in an industry like the Music industry are easily reproducible. And, yet this very large industry is often classified as part of the RL economy. There are a number of industries like this.

Cris> "but you cannot import and export resources from the VW to the Real World"

I grew up in San Diego, and went to school in Seattle. As much as Seattle would love to import San Diego's weather, coastline, and resultant tourist industry, they just can't. Are we going to then say that weather and coastlines have no economic value because they can't be exported, and the tourist industry doesn't really exist in the real world for that same reason?

Cris> 'Perhaps VW economics need to be framed in terms of labour'

If we are looking strictly at how much work an economy produces, then virtual worlds are second to only communism, (maybe tied for first in some cases). However, last time I checked Work was neither the goal nor the key valuable in an healthy economy.

Cris> 'It would not measured as part of GDP because it does not add to labour or increase labour's value.'

In the case of currency, sure. But when the government contracts Boeing to create a new airplane, this is very much part of the economy.

-bruce

[Michael Hartman]

35.

I am not sure why the ability of a developer to inject X amount of their game world currency into the game world is considered a sticking point.

Every government in the world has the ability to print money if they wish. The effect of this is to devalue the currency slightly (inflation).

Is that any different than when a developer creates and gives away game currency? I think not.

[Cris Streetzel]

36.

Bruce,
Many good points, so let me refine my argument.
The gov't is not producing anything in your Boeing analogy. They may be diverting resources thru taxes or borrowing or creating money from nothing, but it is Boeing (the PC, right?) that is contributing to GDP. Boeing may have built those airplanes for someone else anyway, so GDP might not be impacted at all - more likely just the inflation rate. There is also no creation of wealth.

We may need to put a dividing line between dev created currency, which may be used in payment of PC services ('quests') and dev creation of 'wealth' (items; resources). Your Boeing analogy would be along the lines of dev created currency being used to trigger a quest.

Now if the gov't had a magic nano-forge and created the airplanes and gave them away without any expense of time and labour, we might consider that an increase in wealth, but not an increase in GDP. I would argue that since no economic resources were required to create the airplanes, it should be classified as 'found' wealth, like a natural resource.

I like your comparison between VWs and the Music industry. I would argue that scarcity in the music industry is partially artificial - created by our laws regarding intellectual property. Certainly, the gov't has good reason to do that, but without someone forcing that scarcity, the economics of the music industry break down.

As for the Resources vs Labour debate, I'm not saying that resources don't effect the economic landscape - it certainly has an effect on labour productivity. My point is that if it cannot be exported, that it may not be the best way to frame the measurement. Even in your tourism analogy, you aren't exporting tourism... it is taking place in San Diego. If we define tourism as the production of 'fun', then the weather and beaches are resources that increase your production of 'fun', but no one in San Deigo is producing the weather (though they may be producing the beaches).

On the other hand, it makes sense to me to think of VW trade in terms of Labour specifically because the resource constraints are artificial and the resources cannot be exported. Labour can and is the only thing traded between the VW and the RW. The currency exchange can be framed as a labour exchange: I trade time creating wealth at my RW job for someone else's time creating wealth in the VW.

I'm not really proposing that the VW economy is completely different from the RW economy. As I said, Labour vs Resources is an old debate from the RW economy. The easy way out of the debate is to say that Labour *is* a resource. With VW's I think it is the primary resource consumed and the VW's productive output is 'fun'. While 'fun' is an intangible that cannot be exchanged, the labour portion can be exchanged. Any items or objects in the VW should be considered natural resources and should not be considered production. They may increase the production efficiency of converting labour into 'fun' and therefore have a intrinsic economic value.

VW Items are like discovering an oilfield under your house. Your house may now be worth millions, but the act of discovering the oil does not 'produce' the oil since there was little or no expenditure of resources. You may be an oil surveyer and make your living finding oilfields, but you are still not producing the oil. Your are simply finding wealth.

Hmm, I just realized I'm on the verge of arguing that VW's are not economies at all, but a capital tool for converting labour into fun. They may contain a marketplace for exchanging labour and other fun-creating resources, but it is a straight-forward exchange, not a productive act.

Cris

[Bruce Boston]

37.

“Your Boeing analogy would be along the lines of dev created currency being used to trigger a quest.”

Actually, I think I would classify this as ‘Government Spending’, and as you point out maybe we should be classifying Quests as VW Government Spending, as it seems to fit the pattern.

Example A: US Government asks Boeing to build an Airplane, Boeing does, US Government pays Boeing US$30M.

Example B: Star Wars NPC asks Kid Skywalker to gather 20 wombat skins, Skywalker does, NPC pays Skywalker 30k credits.

However, I would say that Earth governments often do create objects. For example, I’m sure there is a government or two that does not contract the work out, but has a government run airplane factory, in the case of VWs, I think it would be safe to say that they have a number of VW government owned ‘holocron factories’.

“I would argue that since no economic resources were required to create the airplanes, it should be classified as 'found' wealth, like a natural resource.”

You’d probably lose that argument.

“I would argue that scarcity in the music industry is partially artificial - created by our laws regarding intellectual property.”

I think there are a number of people that think that way, and it’s unfortunate. In reality, it’s these same laws that protect the investments being made in almost every industry, and it’s because of these investments that we have so much abundance in our economies today.

Music is just one example, but I think we could include everything from most forms of media, to software, to pharmaceuticals, to toys, to breakfast cereals, to footwear and handbags, to cars and almost any good that requires any amount of R&D. Believe me, if (for whatever reason) all laws protecting IP were discontinued next month, we would see the short-term effects by Christmas. Very cheap software, cheap pharmaceuticals, cheap footwear, toys, music, etc, etc.

The problem is that while the short-term effects look nice, the long-term effects are horrendous. Who would ever invest in creating an IP that could be used by anyone, and without new IP being created at the rate it is today, could we really expect the same rate of economic growth? This is a huge topic, and maybe I shouldn’t address it here, but it’s really to bad so many people think that IP laws somehow create an artificial scarcity, when in reality they do just the opposite by encouraging increased levels of investment.

“If we define tourism as the production of 'fun', then the weather and beaches are resources that increase your production of 'fun', but no one in San Diego is producing the weather (though they may be producing the beaches).”

So, I think I may tend to define the health of an economy in terms of Productivity, and the value of an economy in terms of Production. And, while I agree that labor is a resource, I still assert that Work is a poor measurement of economic health or value. Why? Because Work can be done in very un-productive ways. If two people can get the same job done using a computer that 10 people can do by hand, and if we only looked at work, one might say that more work is being done by the 10 people. In this same way, the economy in a VW with 20k members might very well be creating as much wealth as an economy in a VW with 100k members, and if we only measured Work, we might think that the second is healthier due to its size.

“Any items or objects in the VW should be considered natural resources and should not be considered production.”

As far as I know, anything that exists in a virtual world exists there because someone placed it there. These are hardly creations of Mother Nature.

“They may contain a marketplace for exchanging labour and other fun-creating resources, but it is a straight-forward exchange, not a productive act.”

So I think I’d disagree. Anytime one item is being converted into another item of higher value I would say that the difference in value is the value that is being created. If something is being converted into something of lesser value I think we also need to say that value is being destroyed.

But, I might agree with you in saying that these are more straight-forward transactions than one might think at first glance. I might also say that as such, the transactions in VW economies are not so different than the ones we find in other industries, and therefore these transactions can be included with other industries in Earth economies.

-bruce

[Estariel]

38.

The BBC has covered this story now. http://news.bbc.co.uk/1/hi/technology/3570224.stm

[Martin Im]

39.

I guess all these questions arise because we are searching for VW analogies with nation states and other social constructs, the worlds as we know it today. There are a number of implications when calculating GDP and counted one single time for a single country (or whatever) doesn't give us any kind of answer to anything. Calculated for Sweden in 2005, and only Sweden and only in 2005, says nothing really. How you choose to do it and what you decide to include impacts the result and the accuracy of the calculation. But more important is that it's done the same way between countries and between years. Otherwise it's a completely useless figure. I have my doubts whether it's possible to actually compare a VW with a nation state (for one they are much more volatile) but I have no doubts about the potential economical impact of VW, and there will certainly be a way how to calculate the value of the players actions in real currency. So more important is to establish a generally accepted way of calculating GDP (or what it might be called) for VW and then collect this data over a number of years while understanding how this information relates and can be compared to GDP of countries. If a straight comparison isn't possible there's always the option to adjust for this or that (like in all statistics). But then again GDP is such a rough figure and there a so many other fun things to look at in VW too.

And about scarcity, you'll find a lot of artficially sustained prices in RL too. We tend to forget that there is no such thing as a perfect market.

[Nikos P.]

40.

"As far as I know, anything that exists in a virtual world exists there because someone placed it there. These are hardly creations of Mother Nature.
-bruce"

Indeed, but in a VW the devs play the role of God or Mother Nature herself. The costs to produce an item that players call "vendor trash" (i.e. useful only for selling to NPCs) or an ultra-rare weapon that could fetch hundreds of dollars in an eBay auction are identical (and almost negligible). Thus, all items found in a VW should be considered as "natural resources" in this respect.

Another point to consider is that not all MMO economies work in the same manner. For example, Ultima Online presented a much more controlled economy than most games. Constantly selling skins to a vendor would bring diminished returns and, finally, nothing at all until such time as the goods could be "absorbed" (virtually of course). Compare this model to World of Warcraft today which, much like most games, utilizes a fixed price system for vendors. It doesn't matter if you sell 1 skin or 100, the price per piece remains the same.

[shakuria green]

41.

i love all african american but i hate white people they aint nothin but trash