InfoTech & Public Policy

A while ago, Cory Ondrejka, VP for Product Development from Linden Labs came to Princeton to talk about Second Life. He argued that the success of Second Life came from two things – freedom and ownership. These two factors have interesting implications for the network neutrality debate. If we see Linden Life as the gatekeeper to the Second-Life network and Verizon and Sprint as gatekeepers to the internet, it is amazing how different their philosophies are toward regulating access.

There are eerie parallels between Second-Life and the Metaverse of Neal Stephenson’s SnowCrash (I swear, I’m not a geek, that’s one of the few sci-fi books I’ve read). The similarities go beyond the way avatars interact with a 3D world to the transportation system for getting those avatars around and the complete customizability of the system (Second Life uses quaternions as datatypes, which is fairly ridiculous). It is the customizability that gives Second Life and the Metaverse their richness and makes them vulnerable. The premise of SnowCrash is that there is an eponymous virus inside the Metaverse (which is also a drug in the real world) that threatens to enslave humanity (yeah, kind of far-fetched). A significant amount of the plot revolves around how the Metaverse is open enough for hackers to plant this virus and also open enough for the protagonist to save the day by hacking. To a large extent, this is how Second-Life works – people are given the freedom to tinker, but that comes at a price. One of the more interesting elements of Cory’s talk was about how two people in Second Life went around as alien avatars in a B-movie spaceship abducting other users and giving them TV-shirts that read “I got abducted by aliens and all I have to show for it is this lousy T-Shirt.” Openess allows Linden Labs to tap the skill and labor of millions of users and enhance the richness of the game at no direct cost. Of course, this openness has great risks: the so-called “W-Hats WMD” (in which objects replicated rapidly enough to shut down servers) is to Second Life what the SnowCrash virus is to the Metaverse.

There are also obvious parallels to the internet, in that the creativity of users to do whatever they could over HTTP has made the internet the vibrant world that it is. This is in stark opposition to the Bell-Head model of networks – in which the default policy is to forbid uses outside of a narrowly defined set. As Susan Crawford noted in her IT policy talk, this has not led to much innovation in phone service over the last fifty years. This dichotomy has parallels to international economics: the telephone system is a typical caricature of European or Indian Bureaucracy; the internet is much more like the open capitalist system the US or Taiwan. Most economists argue that the freer economic systems allow for greater growth. If the telcos are too heavy handed in regulating permissible traffic they could destroy the vibrancy of the internet – the very product they are selling.

But neutrality is not enough – property ownership is another key ingredient. The idea is that to motivate entrepreneurs, one has to let them profit from their creations. In Second Life, the game creators grant property rights to in-game creations. This, Cory claims, has led to a whole slew of businesses: from fashion designers of virtual clothes and notaries (using standard cryptographic techniques). Again, millions of users are enhancing the richness of the game at no cost to Linden Labs. There is a parallel with theories on jump-starting economies in developing nations. Many economists argue that the lack of strong property rights in those countries is one of the key factors inhibiting economic growth. In the network neutrality context, this also argues for a more hands-off approach for the Bellheads. If they charge websites too much for traffic (as they are trying to do), they could destroy the entrepreneurial incentives that make people want to buy internet access in the first place.

Second Life makes a strong case for the Openist perspective toward network neutrality.

This isn’t my weekly post, but here’s an interesting article about Microsoft and Ads. They even claim there might be ads in Word, although I’m not sure how much of that is hype –probably most. (The only reasonable alternative might be a cheap version with ads.) Is advertising revenue really a good business proposition, especially with the prevalence of ad-blockers? It is optimal (from a game-theoretic perspective) to mimic your competition if you are ahead. Maybe that’s what MS is doing?

For the moment, I’ll be cynical and note that Google’s toolbar offers a pop-up blocking feature which has no effect on their own ads since they are not in the form of pop-ups. This raises certain network (or at least portal) neutrality issues. I wonder if Microsoft will try a similar strategy to filter out non-MS ads. Will they be safe from antitrust authorities if they discriminate against certain technologies, rather than from certain companies? How finely can we draw that line?

In this week’s discussion of broadcast regulation, the issue of localization and boarders came up. In particular it was noted that Canadian broadcasts had quotas for Candian and French Language content. The rules are actually quite specific: 35% of weekday radio broadcasts between 6AM to 6PM must be Canadian and 55% within the same timeframe must be in French. They even have specific rules on what qualifies as Canadian content.

Most students expressed the thought that protection of culture was silly. The fact that it was the French Language (of all languages) that was being protected made these regulations even sillier. After all, if there is one thing that is popular across America, it’s France Bashing (if you have any doubt, try “I’m Feeling Lucky” for “French Military Victories”).

The typical argument against cultural regulation (as expressed by some in class) is that people vote with their feet: if Canadians wanted to watch more Canadian Television, they would not need government subsidies for it. I have a few responses to that. First, by living in America, one has a distinctly asymmetrical view of broadcast culture. American airwaves are remarkably impermeable to international content, while international airwaves are filled with American content. Chances are, you have never heard of MC Solaar (a popular French Rap Artist), but many of my friends overseas have heard of Eminem. You have probably never heard of the Spring Festival TV Show, which (at least a few years ago) was the biggest TV event in mainland China. And yet, the students I have talked to (in East Asia and Europe) have all heard of American television shows. (Unfortunately, the most popular example, cited universally, is “Friends.” If you don’t believe my anecdotal evidence, check out the French, German, and Chinese language sites, which are always the first hit in the appropriate national language Google search). This asymmetry goes a long way in explaining the divide between countries concerning cultural protections. So then if American culture does better in the free market, does it mean it is somehow better – more marketable, larger popular appeal?

Not necessarily. The name of Frederick II’s palace in Potsdam is “Sans Souci” – it’s French for “without worries.” Why did this quintessentially Prussian ruler name his palace in French? And more importantly, why does it imitate the intricate Rocco style of Versailles (and for that matter, why do all the palaces of that era?) During the eighteenth century, France was the preeminent power and French culture and language were dominant. It has even been remarked that Russian Aristocrats spoke better French than they did their native tongue. Similar geopolitical overtones exist in the modern debate over culture and language preservation: people in foreign countries view American content as an invasion of something they cherish dearly.

A possible response might be, “well, I’m American; as a policy matter, I should oppose all cultural protections because they hamper the free market mechanisms I believe in and inhibit the spread of my culture.” Its pretty clear that Americans do not always think this way. One of the biggest applause lines in Avenue Q was from “Everyone’s a Little Bit Racist:”

Everyone makes judgments
Based on race.
…
Now not big judgments, like who to hire
or who to buy a newspaper from –
…
No, just little judgments like thinking that Mexican
busboys should learn to speak goddamn English!

On the policy front, a provision of the current immigration bill requires that newly “amnestied” immigrants learn English. My point is that Americans are just as culture and language conscious as everyone else.

While I believe that countries have a very legitimate right to support their own culture, I don’t think broadcast quotas are the right way of accomplishing that goal. For one thing, it costs a regulator the same amount of ink to place the quota at 1% for domestic content as 99%; the danger is that policy makers will pick unrealistic numbers without concern for their effects. Subsidizing cultural events is a much better method as increased support comes with increased costs. This is very much in tune with American policy: the current budget for the National Endowment for the Arts is around $140 million. Of course, one could argue that NEA money is squandered too: if the masses choose Britney spears over Itzhak Perlman then let them listen to teenage pop. This is a completely different debate, but I think that most Americans support some form of subsidization for the Arts.

Our discussion of viruses – no matter what the proximate cause – invariably led to the ultimate cause being blamed on Microsoft and its monopoly on the OS market. And not just with viruses, but for many of the perceived problems in the way technology works today, the typical response is “blame Microsoft.” So then is MS’s fortune (and everyone else’s presumed misfortune) a result of historical happenstance, business savvy or anti-competitive behavior. Well, probably a bit of everything. They were fortunate enough to have a contract with IBM for distributing an OS with one of the original PCs at the beginning of the PC boom and clever enough to capitalize on their position as the market grew. It’s also clear from our anti-competition debate that the company engaged in anti-competitive practices vis-à-vis Netscape in the Browser Wars, although the degree to which browsers were a threat to the Microsoft monopoly is still hotly debated.

But I believe there is another factor that gives Microsoft an advantage, and that is the desire to standardize. When I exchange documents with people, I would like them to be able to read my documents and vice versa. Having a myriad of different formats causes confusion, and translations between formats is often flawed (this comes from personal experience of trying to lug files home to a PC from my school’s Mac). If one format gains a dominant position, users will migrate towards the popular format to facilitate exchange of files; people will tend look at what their friends use and purchase accordingly. As it currently stands, files formats are closely tied with operating systems so if the overwhelming majority of users use a certain format, then there is little incentive for individual users to change. Controlling a standard allows a company to control the rules of the game. The more interconnected computers are becoming, the greater the need to agree on how computers talk to one another, i.e. the more homogenized the standards will become.

An industry driven solution to the corporate controlled standard is an open standard that companies negotiate. HTTP, is an excellent example of this. The openness of the standard has contributed to a wide variety of websites that greatly enhance the richness of the web. There are millions of websites and traffic is spread fairly widely amongst them. There is basically one WYSIWYG – MS Word – which rests a proprietary format controlled by MS.

Stifling competition means fewer useful features that consumers actually want (not fewer features, Word has plenty of those). Companies also suffer from a tendency towards group-think, which can be disastrous. Early on, AT&T contracted McKinsey about the potential market for cell phones. The consulting firm’s analysis was that the market would be 40,000 – 50,000 max, convincing AT&T not to invest in developing the technology. They ended having to buy their way into the market as a latecomer. In class, some have argued that a few decisions in the early stages of the Microsoft OS have led to many of its current security flaws. It’s hard to argue that the designers should be responsible for every conceivable hack, but if their products weren’t so dominant, their decisions wouldn’t be so critical.

The trend is for Microsoft to tighten control of their monopoly on standards. Richard Stallman warns that

There are also reports that Microsoft is planning to use patented extensions to XML as the basis for a future Word format; anyone who implements free software to read those files could be sued for patent infringement by Microsoft.

This would put an end to the whole open office project: if people cannot easily make the switch to open office by having their old documents migrated over, then it will be hard for the Open Office XML-standard to gain a foothold. I admit that Stallman is a bit of a zealot on these issues and closer post-Brower War scrutiny on Microsoft puts some restraint on the company. But Stallman is not alone in his skepticism

The proposed patents apparently seek to protect methods other applications could use to interpret the XML dialect, or schema, Office uses to describe and organize information in documents.

There was also much recent hype that in advance of Hu’s visit to the US, China would require all PCs to be pre-loaded with an OS. If this turns into how PC sales work in the US (where Windows on a new computer is the default) that would mean over a billion people subscribing to the proprietary Microsoft standard.

[The power went off in the dorms delaying the uploading of this blog post]

I want to shift the debate from Openists and Deregulationists to the distinction between “Netheads” and “Bellheads” given by Susan Crawford in her IT policy talk this Thursday. Netheads view the network for its potential to deliver applications and content in new and interesting ways. They see it as a neutral medium on which more interesting things can happen. They are generally Openists. Bellheads are those who view networks in terms of the old Bell telephone service model. There was a high degree of reliability, but everything, from the types of services that could be offered to the prices were heavily regulated. Bellheads see the network as a set of wires that they own. In this way, they are like deregulationists, but they would actually prefer regulation to the extent that they understand its ins and outs so that it presents a barrier of entry to competition.

This is an old debate between a regulated monopoly versus dynamic free market: essentially, the debate between Keynesians and Neoclassical economics. Keynesian economics grew out of the Great Depression and Second World War, a time when free-market excesses were thought to contain the seeds of their own destruction and government intervention and regulation of the economy was thought to be the savoir. In the 50s, the norm in America was towards large, often highly regulated, companies. GM President Charles Wilson claimed “what was good for the country was good for General Motors and vice versa” and Americans generally agreed. This started to change with economists like Milton Friedman who argued that regulated business were inefficient compared to a deregulated free market. Their policies were adopted around the 80s, for example, in the deregulation of airlines[1] and the breakup of AT&T[2]. It seems that the cost to consumers has gone down in each of these sectors and the deregulation has allowed for increased competition. Of course, many complain that the level of service has gone down with price.[3]

In recent years, America’s enthusiasm for the laissez-faire economics has faded. Part of it was pricked in the tech bubble and another part stemmed from the failings of deregulated energy as evidenced by Enron (for those of you keeping track, Jeff Skilling took the stand on Thursday). In a post 9/11 world (I feel like that date crops up in every policy discussion) Americans seem to yearn for the security and reliability that the old Bell system represented. The current political climate towards network policy is very much in favor of the Bellheads (if only because they have more lobbyists). I wonder if in pushing away from the absolute free market policies of the 90s, we are also pushing away from the intensely-competitive ideal of the early internet. There was a great synergy (to use another term popular of the 90s) between these two ideals: the Darwinian competition between websites was supported by a similar ideals about economics; those ideals were further confirmed by the high stock prices of Yahoo and Microsoft. In discrediting one, it’s very natural to discredit the other, and to think that a network regulated by a Bellhead authority is better than the wild, untamed internet of today.

Many students have posted to the courseblog concerning virtual worlds and the need to regulate them. On one side, are those who believe that regulation is ridiculous and unsustainable or that their real world irrelevance makes regulation unnecessary. Others argue that events in the virtual world have real world consequences on players and economics which necessitates real-world regulation

To a certain extent, both sides are talking past each other rather than engaging in a debate because the terms of debate have not yet been set. That is, with many other issues we have discussed the question was clear: online gambling is becoming a problem, find a solution that effectively curbs it given the technical possibilities and respects civil liberties. We knew the objective of intended regulation.

The rationale for regulation follows from societal perceptions of the (usually) economic and (sometimes) social consequences of an activity. For instance, we regulate financial market fraud because we believe having functioning capital market where investors can trade with confidence is necessary for a free market. That is, regulation is intended to foster behaviors deemed socially beneficial. On the other hand, we place many restrictions on gambling because we fear the ills it brings to society.

It is tempting to argue that people who spend all their time engrossed in MMORPGs are contributing nothing to the economy and therefore such activities should be discouraged. If one took this as a premise, then one might take the same attitude towards virtual world regulation as one does gambling[1]. However, simply because one is not churning out widgets in a factory does not mean one is not helping the economy. For instance, few would argue that watching a movie is a productive activity, but this form of entertainment is an important part of the American economy. In fact, we deem it to be so important that we have created a complex set of laws called copyrights to ‘incentivize’ the production of creative content. Without any consumers, products would have no purpose: all work and no play stifle the economy. If one is to justify regulations against virtual worlds, one would have to do so on the basis of negative externalities generated by online addiction, much as one might justify regulation of gambling based on gambling addiction. However, one would have to find significant negative externalities. As a society, we don’t believe in regulating behavior simply because it is “unsocial:” there are no rules barring you from being glued to your TV set or being a workaholic.

But there is another way to view virtual worlds, and that is as a platform for economic activity. The Second Life economy is booming: in January 2006, the Linden dollar equivalent of $5 million US was traded. If a sizable market springs up, then the government has a strong incentive to regulate n order to smooth out the rough edges of the free market. (Alternately, if one is of a libertarian persuasion, one could take the cynical view that the government will find a way to regulate any sizable market, especially when tax dollars are at risk.) Accepting the premise that virtual worlds are a perfectly legitimate (and socially desirable) medium in which to transact business, one would consider regulation (or the lack of it) to promote growth. This is not terribly unlike the stock market analogy: in both economically significant bits are being flipped in servers and these bits are redeemable for real and virtual services. However, it is difficult to say that the LindenX plays the same fundamental economic function that AMEX or NYSE does so any proposed regulation will have be tempered by this “irrelevance.”

The above is a discussion of two views of the purpose of government regulation. I don’t know if I subscribe to either view entirely; both have their strengths and shortcomings. Neither argument deals explicitly with the form and scope of regulation. On this point, I think Tim Wu made a good point. He gave a talk March 30^th on campus entitled “Who Controls the Internet?” in which he brought up the importance of physicality. Internet-based forms of punishment inflictable by game masters – lowering one’s score, or banishing someone from a site – do not seem to have an effect on certain individuals (consider, for instance the attacks by W-Hats on the Second-Life servers). However, even these people are afraid of real world sanctions that the government is capable of imposing – fines, prison time. While virtual punishments should be a first choice they sometimes need to be backed by real-world consequences.

There have been a lot of discussions (or in some cases, rants) concerning the RIAA and MPAA specifically and the clash between copyright and cheap convenient electronic distribution of content in general. In class, many suggested distribution of content along a subscription model would be better for consumers and movie studios. In this post, I will give an economic analysis of this hypothetical model for the movie industry, looking at Netflicks as an example.

Netflicks is a video rental service where a consumer may request DVD titles which are mailed to their house and are mailed back for new ones. Depending on the plan, consumers are limited to a different number of DVD’s that they can keep at any one time. This model already offers several conveniences to consumers. First, there is no marginal cost. DVDs might be $15 to purchase or $5 to rent, but add nothing to the monthly subscription fee. Second, unlike their competing brick-and-mortar movie rental operations, Netflicks charged no late fees. This was a major selling point in their advertising campaigns until the major brick-and-mortar operations were forced to eliminate late-fees altogether (a wonderful example of how new competition lowered costs for consumers). Third, a centralized content delivery system can hold more titles than your average Blockbuster or DVD store. This makes it a lot easier to find obscure or older titles that are not as popular. (I realize I sound like a salesperson but I swear I do not subscribe to this service myself, although I have several friends who are happy customers).

Often, such proposals are criticized as hopelessly utopian and unsustainable economically, so let’s look at the profit potential. According to the industry consumers spent roughly $24.3 billion renting and purchasing DVDs in 2005 and 82 million households owned DVD players. This works out to just under $25 a month per DVD-owning household. Keep in mind these households are what technologists call ‘early adopters:’ those who are relatively well off economically and enthusiastic about new technologies. As the market expands, the ‘average consumers’ will pay something less than this figure. (Interestingly enough, DVD sales were flat between 2004 and 2005 despite an extra 12 million homes with DVDs). Let’s say DVD players are ubiquitous and that the average monthly fee per household is around $18, which Netflicks claims is their most popular monthly plan. (Obviously, there would be price discrimination to get to the $18 average: at one end, a $35 per month deluxe membership that offers director’s cuts and behind the scenes footage while at the other, a $15 per month “value package” for only the most popular movies.) There are 170 million households in America, and some of them, either due to poverty or stubbornness, will never buy a DVD player. Let’s assume 150 million of those homes eventually get a player (this is comparable to the current 90% prevalence of VHS players in homes). That still means $32.4 billion annually for the movie industry, which is better than their current combined VHS, DVD rental and sales revenue.

By the time DVD players become commonplace, broadband will probably also become prevalent. Distributing content electronically eliminates two problems with NetFlick’s current model. One is the shortage of rare DVDs: if five people suddenly decide to watch an obscure movie, they have to wait in line as Netflicks will only have one copy to distribute. Many customers choose Netflicks specifically to watch obscure movies and such inconveniences have lost them customers in the past. The second problem is the cost of first class postage. Netflicks has been accused of “throttling” frequent rental customers – that is, purposely delaying requests to keep down shipping and processing costs. Online delivery would eliminate both these problems (the requisite DRM protections are a different issue that I won’t begin to touch here). I know I’m not the only person to have thought of this. MovieBeam broadcasts movies (“piggybacking on TV towers”) to a box in your home although they charge for the box and per movie. If only some company could combine subscription with electronic delivery … or maybe I’m too picky as a consumer.