The Coming Second Life Industrial Revolution?
In previous articles (here and here), I explored the economy and monetary scheme of Second Life, an online three-dimensional virtual world where users can create virtual goods and services and exchange them with one another. I also discussed issues surrounding the increasingly interventionist stance that Linden Lab, Inc., the creator of Second Life, has taken toward its economy.
In the field of macroeconomic theory, the Austrian view is distinct from other schools of economic thought due to its emphasis on the role of capital. In attempting to extend the application of Austrian economics to virtual economies, it will be a worthwhile exercise to explore the nature of capital that exists within Second Life and observe how it is evolving.
What is capital in the context of a virtual world? The Austrian economist Eugen von Böhm-Bawerk, who elaborated the concept of capital as it relates to time, offered the following definition of (real world) capital: "a complex of goods that originate in a previous process of production, and are destined, not for immediate consumption, but to serve as means of acquiring further goods."
In the real world, one example of a capital good might be a bulldozer, which is not itself consumed, but is used to improve land by making it more suitable for human use. Analogously in Second Life, someone could use a scripted object to automate the shaping of virtual terrain. The object would only be capital, however, if reshaping the land were not an end in itself. Similarly, a real world bulldozer would not be considered capital if it were used strictly for entertainment purposes.
In addition to straightforward examples of capital like that above, the prototype of any item created in Second Life could also be considered capital. Unlike the real world, Second Life users possess the ability to replicate their creations at essentially zero marginal cost. A user may also restrict further replication of the replicas he produces and distributes to others, so that he alone retains the ability to make new copies. Accordingly, even if a user consumes his own creation by using it each day, he alone has the ability to sell its replicas at any time, thus potentially qualifying it as both a consumer good and a capital good, depending on its use.
The only items that cannot be considered capital in any sense of the term are nonreproducible replicas of goods that cannot be used to facilitate production of other items. One such example is the suit my Ludwig von Mises avatar is wearing in the picture accompanying my first article. The suit was purchased from a vendor and can either be worn (consumed) or sold to someone else, but it cannot be replicated.
Conceptualizing every prototype in Second Life as capital also has a real-world basis in the treatment of intangible goods (patents, copyrights, etc.) under accounting standards. Generally Accepted Accounting Principles (GAAP) allow firms, under certain circumstances, to classify intangible goods as capital assets on their balance sheets. Valuation of these assets is difficult, however, and GAAP generally allows firms to fully capitalize them only when they have a market price determined by an arms-length transaction, i.e., one in which the buyers and sellers of a product act independently and have no relationship to each other. Similarly, in Second Life, although I paid 500 Linden dollars (L$500 is about US$1.85) for a single replica of the Mises avatar's suit, it would be very difficult to assign an accurate value to the prototype without an arms-length transaction.
Having identified the characteristics of virtual capital, the next step in analyzing the role of capital within Second Life is to observe what entrepreneurs are producing and the processes that underlie production. Although there is no precise and comprehensive data on Second Life's economy readily available, anecdotal evidence collected from active users suggests that virtual real estate development is by far the largest industry in Second Life. Virtual real estate development firms purchase parcels of land that are periodically created by Linden, place structures and other objects on them, and then sell them to retail consumers. Although the process for this type of production might itself be highly intricate and complex, it can be accurately likened to a real world "cottage industry," where goods are manufactured from start to finish by individuals or small groups over a short period of time, and a minimal amount of capital is employed in production. This type of production stands in contrast to the real world's modern, industrialized society where the length of time taken to produce goods is much longer, large amounts of specialized capital are employed, and goods-in-progress often change hands from one business to another several times before finally reaching consumers.
Those familiar with the Austrian Business Cycle Theory (ABCT) will recognize the significance of the observation that Second Life's economy resembles a cottage industry. Without a significant capital structure wherein distortions may occur, the opportunity for malinvestment is much smaller, and the amplitude of fluctuations in investment and consumption may also consequently be smaller.
I claimed in my first article that boom and bust will necessarily occur under Second Life's current monetary scheme. I maintain that claim, but I suggest that the appropriate model for Second Life may differ from a pure application of real-world ABCT.
This may be seen more easily by likening Second Life's economy to an industry in the real world that is being subsidized by government. As long as government subsidies flow to the industry, lower consumer prices will result in excess consumption and greater profit, which will attract more investment. But once the subsidies stop, demand will return to a normal level and the industry will go bust, with declining consumption and declining investment.
In the case of Second Life, the subsidy comes in the form of Linden dollars, which are created out of thin air and are not explicitly backed by anything. As more and more Linden dollars are created, Second Life will eventually experience something like a "crack-up boom" when a sufficient number of people realize that their virtual money cannot be converted to real-world currency at a stable exchange rate. If this occurs, Second Life entrepreneurs will realize any money they invest in Second Life projects will be worth less when they attempt to convert it back to US dollars, and they will not invest in any new projects. Consequently, the economy will languish.
Returning to the concept of capital, we might wonder whether Second Life's economy will eventually make more intensive use of capital and evolve out of its current cottage-industry method of production. Given that the real-world economy did just that, it is certainly a reasonable expectation. To determine how that might occur, we can study some of the economic conditions that led to the Industrial Revolution and extrapolate them to a virtual context.
Industrialization occurred through the actions of profit-seeking individuals undertaking projects that had no guarantee of success; it was not achieved by the actions of monarchs or government agencies. Government's role in the Industrial Revolution was important, but only in the sense that it did not stand in the way of these entrepreneurs. The questions to be asked, then, are what motivation the entrepreneurs had, and what reassurances they had that their objectives could be achieved.
The goal of the industrial entrepreneurs was monetary profit. What conditions were necessary for profit seeking? First, they needed to believe that they could retain whatever profits they earned — that they would not be arbitrarily stolen by looters, employees, or government. Second, they had to believe that their enterprises could add economic value; that is, they could sell their finished products to consumers for more than the costs of the inputs used.
The first requirement basically boils down to the existence of property rights and contract enforcement. In the case of Second Life, property can take the form of virtual land purchased from Linden, reproducible prototypes, nonreproducible replicas, or Linden dollars. For each form of property, there are strong safeguards inherent in the computer programming of Second Life to protect one's rights over it. Landowners have the ability to remove other individuals and items from their property with the click of a mouse. Linden dollars cannot be stolen. Prototypes and replicas cannot even be replicated without the permission of the owner (barring hacking or other exploitations which Second Life has experienced briefly). Given these inherent protections, it appears that the first condition for large-scale investment and eventual industrialization — protection of property rights — is fairly well satisfied.
The second condition for the preservation of property is contract enforcement. Unlike property rights, which can be clearly delineated and programmatically enforced entirely within Second Life, it would be impractical, if not impossible, to provide a similar mechanism for contract enforcement. There are simply too many possible contractual provisions and conditions to be dealt with in an automated fashion. Still, entrepreneurs are able to rely on others' reputations to some extent, or they may seek a remedy from a virtual dispute resolution organization (e.g., Metaverse Republic or the e-Justice Centre), but it is necessary to resort to real world courts if one wishes to employ the threat of physical compulsion to enforce a contract. While today's available solutions still have room for improvement, the availability of real-world courts certainly provides adequate reassurance that legitimate contracts can be enforced in many circumstances.
The last aspect of protecting one's property and retaining profits is the ability to liquidate assets by converting them to money. While any large fixed investment, real or virtual, carries a risk that there will be no available buyer, an investment in a Second Life enterprise is especially risky because of the previously discussed fragility of the Linden dollar. If the Linden dollar becomes unstable, it will reduce the value of monetary holdings in Second Life and undermine entrepreneurs' ability to convert their Linden-dollar profits to US dollars. This circumstance certainly inhibits the prospect of industrialization in Second Life, but I do not think that it would necessarily prevent it altogether.
The other major condition for industrialization is the necessity of an industrial-type process being the most economic way to produce consumer goods. In the real world, factories and machinery have allowed people to produce large quantities of items quickly and efficiently, capturing economies of scale. In the case of Second Life or any other virtual world where goods can be replicated at negligible cost, there is no analogous opportunity to increase productivity. Instead, the only economic justification for industrialization would be to capture the benefits of specialization and the division of labor.
Given Second Life's digital, computer-programmed nature, a natural model for considering how specialization might occur is Object-Oriented Programming (OOP). OOP allows programmers to create basic "classes" with their own sets of attributes and abilities. Programmers can then create increasingly sophisticated classes by reusing and inheriting features from simpler classes that have already been created, avoiding the need to reinvent the wheel every time. Extending this concept to Second Life, distinct businesses could specialize in various tiers of classes and sell their creations to one another, just as goods in progress are sold from business to business in the real world.
An illustration of how this could occur in Second Life might be the creation of a virtual voice-controlled video camera. The designer of this item would select a voice-recognition module from among the many different versions offered by competing Second Life manufacturers. He would choose the module that he thought worked best, based on how well its highly specialized code converted sound waves into text. This module would presumably come with a manual that explains to the designer how to switch it on and off and access its output. The designer, having also selected a movement control module and a video camera module, would then create an interface between all of them and would tweak it as necessary to ensure proper functioning. Additional modules might be added to enable streaming the video to the Internet and whatever other functionalities the designer thinks consumers desire. This process of adding and combining modules could be continued indefinitely, creating ever-more complex goods.
To see precisely this kind of specialization in the real world, one need look no further than the personal computer software industry. Although antitrust cases might lead you to believe otherwise, there is not one single company that produces all the software that can be used on a personal computer. Many different operating systems (e.g., Microsoft Windows and Apple OS X) are produced by different companies. Applications (e.g., Microsoft Office, Adobe Acrobat, and McAfee VirusScan) are created by other companies and build upon the capabilities of the operating systems. On top of that, additional companies create modules that plug into these applications, allowing further specialization, all within the same framework.
Novel models are required for analyzing virtual economies and exploring such abstract economic concepts as capital. Although Second Life's capital structure is currently relatively simple by real-world standards, that may change one day. Entrepreneurs may move in the direction of creating products that consist of increasingly complex, specialized components that require lengthy periods of production.
However, such industrialization will ultimately occur only if it is economically feasible. If it does not, it will not necessarily be a failure on anyone's part, but will simply be because the most economic method for producing the goods that virtual world consumers desire would be for one group to produce them from start to finish. Profit-seeking individuals are constantly searching for the most economic ways to produce goods for consumers, and virtual goods are no exception.
Matthew Beller is a former employee of the Federal Reserve Board of Governors and the Securities and Exchange Commission. He currently works for a private investment firm in Los Angeles. Send him mail. Comment on the blog.
 The ability to reproduce objects at no marginal cost is specific to Second Life. Other virtual worlds do not necessarily grant this ability, which means that different economic models may be required to analyze different worlds.
 Linden has indicated that it has the ability to reduce the L$ supply by accepting "tier fees," which are similar in concept to property taxes, in the form of L$. Currently, tier fees are paid in US$. If Linden were to begin accepting tier fees in L$, it could potentially stabilize the L$/US$ exchange rate.
 I use the term "property" loosely. For a slightly more in-depth discussion of property rights in virtual worlds, see my previous article, particularly its discussion of the ethics of virtual interventionism.