THE REAL MONEY TRADE - HOW THE REAL AND VIRTUAL ECONOMY INTERACT

There is a curious tension in the goals of every video game player. On the one hand, players enter a game in order to be challenged. They are apparently willing to pay money for developers to place obstacles in their way as they work toward some final goal - saving the princess, slaying the demon, building a city, and so forth.

RMT refers to any activity in which players exchange real world currency for virtual assets and services (Heeks, 2009). Players may also sell entire game accounts for money. Players typically agree on a price for the assets either by chatting inside the game or by meeting on a third-party site, such as eBay. After the money changes hands in the real world, the two players log in to the game and exchange the associated virtual assets. Depending on the game, the size of large transactions can be remarkable. One hundred gold pieces might sell for $20, a legendary sword for $200, and an entire game account - including currency, sword, and one or more high-level characters - could go for $2000 or more.⁵

The RMT can take place in a primary market or a secondary market. Primary RMT refers to any RMT in which a virtual world company sells virtual assets directly to players. Secondary RMT refers to any RMT that takes place between players (Lehtiniemi, 2007). Whatever the form, firms either sanction RMT or they do not. This is summarized in Table 24.1.

Table 24.1 Sanctions and virtual trade

24.3.1 The Problem with RMT

During the 2000s, RMT was an important issue for virtual worlds. There are several reasons for this. First, when large-scale virtual worlds were first released in the late 1990s, the end user licensing agreements (EULAs) - to which all consumers had to give assent before entering the world - usually stipulated that the game publisher was the sole owner of any virtual goods that players produced or used in the game.⁶ RMT was considered a violation of property rights by the game publisher because people who did not own virtual property were making money from its sale. In the case of services like speed­leveling, sharing account information was also against the EULA.

Second, because game publishers did not usually sanction RMT, exchange of goods required money to change hands first in the real world before it could change hands in the virtual one. Thus, buyers of virtual assets were at the mercy of sellers to follow through on their side of the bargain. Players would often hand out credit card information to pay for virtual assets or services, and then later find that they were victims of identity theft. In other cases, players would pay for a good or service and the seller would keep the money but never deliver. If the player handed over their account for character-improvement services like speed-leveling, they might never get the account back.

Players who were victims of fraud often took their complaints to game companies.

A third reason for the issue’s salience was that many players could not afford RMT, or were otherwise unhappy with its existence. RMT was thought by many players and developers to impinge upon the so-called ‘magic circle’, a barrier between the real and the virtual worlds in which players are supposedly ensconced when logged into the game. This violation of the wall between fantasy and reality reduced the value of the game for many players (Castronova, 2006a). Detractors also saw RMT buyers as cheaters because they paid outside contractors either for virtual goods or for other services that granted ‘unearned’ advantages. Allowed to run rampant, RMT gave many players an excuse to quit the game. At that time, the industry made most of its revenues through subscrip­tions, so early cancellation was a major concern.

Fourth, unchecked RMT made virtual economies difficult or impossible to manage using the faucet-drain system. RMT encourages sellers to specialize in gathering currency as quickly as possible. The influx of additional currency from these sellers can lead to an increase in the amount of money in the virtual economy relative to the number of goods and services. This leads to inflation, and only the incomes of RMT buyers can keep up with that inflation. Under normal circumstances, a game company facing inflation would simply widen the sinks so that more money would exit the virtual economy. However, in the case of RMT such a policy response adversely affects players who are not sup­plementing their virtual incomes. Thus, operators who combat inflation in the normal fashion face the ire of their players, while operators who do not act face the possibility of hyperinflation and economic collapse (Simpson, 2000).

Despite these issues, however, the total effect of RMT on a game and its profits

are unclear for several reasons. First, the presence of RMT may cause non-buyers to leave, but it may also cause buyers to stay longer. On the other hand, the accelerated advancement of RMT buyers might cause them to leave sooner. Second, RMT may cause the number of goods and services in a game economy to rise faster than they otherwise would; that is, it may cause players to gain virtual wealth at an undesirable rate. As we discussed before, this can decrease the player’s effective lifetime in a game because they can get through the content more quickly. Third, although players who feel that RMT is cheating may quit, other players may be attracted to a game because of the availability of RMT (Huhh, 2009). Fourth, to the extent that online games depend on social network externalities for success, the existence and extent of RMT will affect the value of the game for all players. Finally, if a game is so challenging that it turns players off or causes them to quit, RMT may reduce that challenge and attract more players, without requiring more expensive work by the game operator in an effort to rebalance the game.

Needless to say, these are highly complex and dynamic variables to consider and balance, and it is unclear how deeply game operators analyze the issue when forming policy responses to deal with RMT.⁸ One thing is certain: the difficulties and intrigue of fighting RMT (or not) gave many virtual world developers pause to consider other approaches to the phenomenon, and to their revenue model.

24.3.2 Gold Farming

Early RMT was supplied by a domestic, cottage industry: Western sellers serviced Western buyers in the EU and the USA, and Chinese and Korean sellers serviced their Eastern counterparts (Dibbell, 2007). The situation in the West changed radically with the release of Blizzard-Activision’s World of Warcraft (WoW) in 2004.

Predominantly based in China, gold farms are businesses that industrialized the process of producing virtual assets for sale on the RMT market. Employees work in shifts, producing assets, ferrying them to buyers, advertising wares, and so forth. Gold farms sell virtual assets directly to players or through intermediaries (ibid.; Heeks, 2009).

The rise of gold farming is notable for two reasons. First, it greatly exacerbated many of the problems RMT created for virtual world operators. By industrializing it, gold farmers drove down the price of virtual currency, making it affordable to far more players than the cottage industry could have achieved. This brought the problem of RMT to a head, and forced the industry to evolve to deal with it in new ways. Second, gold farming represents an important new source of employment for many developing nations. As outlined by Lehdonvirta and Ernkvist (2011), gold farming and similar services represent an important new industry whose cost structure is well suited for poor and developing nations. Gold farming operations require a small initial technol­ogy investment, but the main element of the business is cheap labor. By having low- wage workers conduct resource-extraction tasks in the game for eight-hour shifts, the gold farming business can rapidly acquire a large stock of in-game resources. Those resources can then be sold via the Internet at considerable profit to players from high- wage countries.

24.3.3 Modern Approaches to RMT

As early as 2000, there were some virtual world operators that tried to integrate RMT into their virtual economies. These attempts met with mixed success. Habbo, formerly Habbo Hotel, run by Finnish firm Sulake, is one of the few early, successful virtual worlds that integrated primary RMT.

Linden Lab’s Second Life allows RMT on both primary and secondary markets. The company sells the virtual world’s currency, the Linden, at a fixed rate of exchange to create a price floor, but residents and other third parties can trade with one another at prices above that floor. Linden Lab guarantees the transaction and takes a small cut. Linden Lab also sells and leases virtual land to residents. The remainder of the economy of Second Life is entirely player driven. Residents must create objects using the toolset provided by Linden Lab. Those items can then be bought and sold through Linden Lab or via another third-party broker. Since Linden dollars are only created when players buy them from the company, RMT does not create major problems for Second Life. Despite early hype, however, Second Life and similar worlds like MindArk’s Entropia Universe did not enjoy the success of contemporary, game-based virtual worlds.

More recently, game companies have begun to experiment with market-based solu­tions to the RMT problem. In CCP’s EVE Online, players may purchase a virtual item called a Pilot License Extension (PLEX) using real money. One PLEX grants its user a one-month extension to his or her subscription ($14.99 as of June 2014). The player who purchases the PLEX may, if he or she chooses, sell it to another player in exchange for EVE Online’s virtual currency, the ISK. This allows players who have extra cash a way to obtain additional ISK from players who have a surfeit of that currency, and who may wish to use it to purchase a month of access to the game.¹⁰

24.3.4 Free-to-play Games

One particularly important response to the RMT was the rise of the video game revenue model known as free-to-play, or F2P. F2P publishers give away some or all of the game content to players. The players may then purchase additional virtual goods and services from the publisher. Some publishers sell these goods and services directly in exchange for cash, while others prefer to sell so-called ‘hard currency’ for cash. Players then use this hard currency to buy items and services from the company (see Lehdonvirta, 2009; Hamari and Lehdonvirta, 2010).

There are a few early, game-based virtual worlds that sold virtual assets directly to players. Korean firm Nexon’s MapleStory was released in 2003 and remains quite suc­cessful. Korean and other Chinese game companies in general demonstrate much less hostility toward RMT than their Western counterparts, and embraced the phenom­enon early on. Though invented by Western game developers, the F2P business model for games exploded in countries like China and South Korea (Wang et al., 2012). This model makes sense in countries where a $40 upfront cost for a game client and a $10 to $15 monthly subscription - common in the EU and USA in the late 1990s and early 2000s - represent a substantial chunk of the average disposable personal income.

In the West, and with notable exceptions, F2P was a very small part of the video game industry until 2007, when Apple released the iPhone and Facebook released its Facebook platform. These two application development platforms aided the rise of the F2P indus­try in wealthier countries. Both platforms made it possible to market a game to a large number of people across a broad range of tastes, gaming experience, and skill levels. At the time of this writing, common F2P platforms like tablets and mobile phones are the only areas of the video game industry that are currently projected to experience sub­stantial growth in revenues.¹¹ While there are many factors contributing to this growth, including the age of the last console generation and the growth in the number of people owning mobile devices, it is clear that F2P is an important innovation.

What has this to do with virtual economies? As mentioned, F2P often depends for revenues on players purchasing a hard currency directly from the developer. This busi­ness model is a response to the RMT issue, and it has led to a proliferation in the number of virtual economies and associated virtual currencies. The real consequences of this substantial shift of real currency into virtual scrip - already a hallmark of more tradi­tional industries like airlines and their miles programs - are as yet poorly understood (Castronova, 2014). Especially in games where trade remains possible between players, the realization of gains from trade may mean that there is a large amount of important economic activity that is going unnoticed and unaccounted for in games.

24.3.5 Virtual Property

Although it is clear that goods and services in virtual economies have economic value, the question of ‘whose value?’ remains open. From the earliest legal writing on virtual economies, the question of property rights has remained central (Lastowka and Hunter, 2004). As we will see in a moment, courts have not been especially helpful in clarifying these issues. As a result, the legal status of virtual property remains fuzzy.

Virtual property is an intangible property, much like a URL or a brand name. Virtual property physically exists in the form of sequences of bits in databases on hard drives. Yet it is manifested to people, and used by them, in ways that are completely analogous to tangible property. As a result, virtual property evokes several legal metaphors whose implications conflict.

A game can be viewed as a document authored by the designer, in which case the virtual property within it is the intellectual property of the designer. Copyright issues are evoked. But a game can also be viewed as a forest. The designer plants the trees and allows ordinary people to go among them, for a fee. According to a labor theory of property, a man who gathers wood there owns the wood. It would follow that goods in games belong to the people who make them, that is, the people who use the software to create virtual items. Finally, a game can be viewed through the lens of its contract. Most virtual economies operate under an EULA between the game owner and the player. The EULA generally declares that everything in the game is merely licensed to the player; there is no ‘property’ in a legal sense. Thus under contract law, virtual property is not property; what looks, acts, and smells like a duck is declared not to be a duck (Fairfield, 2009).

Courts have not spoken uniformly on the issue of virtual property. In some countries (Korea), the legal system clearly recognizes the property status of virtual goods. In others (the USA), judgments tend to rely on contract law and intellectual property arguments (see Lastowka, 2011 for an overview).

Much of the discussion of virtual property treats that property as individual and poten­tially private, when in fact a great deal of property in games and social media is common. Thus an unfortunate side-effect of the legal respect now accorded to developer-to- individual contracts, is that player community legal agreements are impossible (Fairfield, 2008). If switching costs were low, this would not present a problem. Developers would make games and rule them, but sorting among the players would allow each player to find the ideal game. Unfortunately, switching costs in these environments can be high, meaning that generally there would exist a reallocation of policies and players that would raise welfare. A system of covenants among players (much like neighborhood asso­ciations) could relieve some of these inefficiencies. Virtual property certainly exists and certainly has value, but disputes over ownership of the value will continue.

24.4