Part I of this paper examines the work of Cass Sunstein and others who are of the view that the Internet tends to create balkanization of opinion and fosters and supports extremist groups. Noah Zatz's proposal for an electronic sidewalk, (referred to in this paper as the "cydewalk") is then evaluated in terms of how it may counteract these problems. Part II of the paper categorizes the academic writings on Internet regulation as four schools of thought: 1) The Secessionist Libertarians, 2) The Anti-Anarchists, 3) The Open-Source Regulators and 4) The Telecosmic Futurists. This part of the paper then shows how each school would likely view the cydewalk proposal, and what modifications it would make to it. Part III of the paper examines the trends in the constitutional and labour relations jurisprudence and how they would affect the various models of the cydewalk posited by the four schools of thought. Part IV of the paper summarizes the potential problems with each variation of the cydewalk proposal. It indicates that the Telecosmic Futurists' preference for letting the marketplace govern the Internet would be the best paradigm, since employers and trade unions would mutually benefit from entering into cydewalk agreements.
A February 18, 2000, item in Labornet Newsline stated that Yahoo! cancelled an electronic leafleting campaign by the Service Employees International Union (SEIU) against Argenbright Security. Yahoo! claimed that negative, targeted banner ads were contrary to its policy. SEIU had paid Yahoo! for the campaign which involved banner ads pointing the viewer to the union's site when the name of Argenbright's parent company was entered into the search engine.
In 2000, Argenbright was retained by more than 40% of the airports in the United States to provide passenger and baggage screening services, including at Newark where two of the hijacked planes originated on September 11, 2001. Argenbright paid its employees an average salary of about $15,000 per year. Prior to the union's electronic leafleting, Argenbright was found guilty of numerous violations of the National Labor Relations Act in connection with anti-union activities during SEIU's organizing campaign. Argenbright subsequently had all of its contracts with the U.S. airports cancelled in the aftermath of the tragic and horrible events of September 11 when it became evident that the company's security services were woefully inadequate.
An article on the August 9, 2002 CNN.com site reveals that the terrorist group al-Qaeda, which brought down the World Trade Centre on September 11, has its own web site. The article also indicates that a patriotic U.S. operator of a string of pornographic web sites had briefly infiltrated the web site to gain information to hand over to the F.B.I., but was unmasked while trying to attract the F.B.I.'s attention. The web site then closed down but the al-Qaeda was able to re-open another one shortly afterwards using an off-shore server.
What lessons then can we learn from these two tragically related news items? Firstly, the Internet is not a friendly place for unions which wish to engage in leafleting or picketing of an electronic nature. Secondly, the Internet is a hospitable place for extremist groups who wish to use a web site for communicating with small numbers of their members scattered across the globe.
Most Canadians would agree that it is wrong for extremist, hate or terrorist groups to use a web site for purposes of coordinating activities of their members. The Canadian jurisprudence is consistent with this view, and the courts forced Ernst Zundel and his anti-Semitic web site out of the country and into the U.S. where he was permitted to operate it until his visa expired, since the freedom of expression protections in the First Amendment have generally been interpreted by the U.S. courts to permit hate sites, although not terrorist ones.
Canadians’ views are probably mixed on the issue of whether unions should be allowed to use the Internet for electronic leafleting or picketing, such as was attempted in the Argenbright situation. The only court case to date in Canada and the U.S. on electronic picketing in a labour relations context, the British Columbia Automobile Association decision (“BCCA”) supports the general notion that unions should have the right during a labour dispute to express their views to persons visiting the employer’s web site. However, this is only one trial level decision from a Canadian provincial court and the future judicial trends and the general public opinion on this topic remain to be seen. Although the facts of the Argenbright situation suggest a fairly compelling rationale for electronic picketing, critics of the concept may argue that the situation was a unique one and that, in any event, the union had alternatives of picketing and leafleting in physical space which it could have utilised instead of using targeted negative ads on the Internet.
Is there, though, a common set of conditions on the Internet that led to both the al-Qaeda web site, which the majority would vociferously condemn, and the cancellation of the electronic picketing of Argenbright, which would produce a more mixed response? If there is such a common set of conditions then would it not make sense to fix them so that the problems with both types of situations, i.e. the extremist site and the dysfunctional employer-union relationship could be remedied together?
The common set of conditions on the Internet creating both the extremist site and the limited opportunity for electronic freedom of expression in a labour context has been the subject of what is now referred to as "second generation" Internet scholarship, the leading proponent of which is Cass Sunstein. In his book, Republic.com, Sunstein argues that the present structure of the Internet provides few meaningful opportunities for public discourse since it contains no effective public forums. "Chat rooms" on the Internet, he postulates, tend to attract like-minded people of similar viewpoints who are merely interested in basking in an electronic echo chamber, rather than being exposed to differing ideas. He also points out that the Internet encourages fragmentation, polarization and extremism, due to filtering technologies which allow individuals to tailor the information that they receive so that they are not exposed to any new, different or opposing ideas. As an example, he refers to "The Daily Me" a hypothetical personalized electronic newspaper which subscribers can have e-mailed to them. He also notes the rise of hate and other extremist groups on the Internet, the fact that the Internet enables them to efficiently coordinate small numbers of members scattered over large geographical distances and that most of these numerous extremist sites do not have cross-links to sites expressing opposing viewpoints.
Sunstein recommends several possible solutions to these problems, including deliberative domains, disclosure requirements, voluntary self-regulation, economic subsidies and "must carry" rules requiring popular or controversial sites to provide links to other sites with opposing views.
His book has provoked a variety of responses ranging from an effusively positive review by Senator Edward Kennedy to a scathing, extremely critical review by Dan Hunter of the University of Pennsylvania. The truth probably exists somewhere between these two extremes of opinion.
The first major salvo in the academic literature in the way of a counter-response to Sunstein was by Mark Nadel, an attorney with the U.S. Federal Communications Commission. Although Nadel’s article is generally quite thoughtful and contains some interesting information about deliberative web sites, it is sometimes naïve, too U.S.-centric, and ignores the special situation of trade unions.
Nadel argues that the Internet already has a wide variety of serendipitous encounters and public forums which broaden an individual’s horizons, i.e. unsolicited mass e-mail or "spam," banner ads, chat rooms and browsing or "surfing." How spam broadens one’s horizons in any meaningful sense is beyond me. Trade unions are not likely to use this method since e-mail lists are generally proprietary and access to them must be with consent of the owner of the list. A manufacturer, retailer or an employer is obviously unlikely to sell a list of its customers’ e-mail addresses to a trade union. Similarly, targeted banner ads must either be purchased from the web site owner, the search engine owner, or the web site provider. As we have seen from the discussion of the Argenbright case at the outset of this paper, such banner ads are not likely to be sold to trade unions and other advocacy groups with controversial messages.
Nadel’s discussion of chat rooms is interesting, particularly with regard to discussion sites. The general wisdom to date has been that due to the size and diversity of the political community, chatrooms or discussion sites aiming for effective democratic discourse would require mediation to be truly effective. The resources required to provide such mediation would likely require advertising or sponsorship, which would tend to reduce the usefulness of chat rooms as a forum for free public discourse. Nadel counters this conventional wisdom by referring to new unmoderated sites such as www.quorum.org where the visitors collectively rate the various comments as to their value so that other visitors can ignore or read them accordingly. This type of system does address the problem of an excessively high “noise to signal” ratio on unmoderated discussion sites. However, although this type of site may create a good form of quasi-public forum which would assist deliberative democracy, the problem is that it would tend to attract only the genuinely open-minded.
Finally, there is browsing or "surfing". A person who enters "discount stores" into a search engine will end up with hundreds or thousands of links mentioning that phrase, many of which will be critical such as www.walmartsucks.org. Does the trade union then not have the same "right of overture" which it has in the real world, e.g. offering a leaflet on the street? The answer is no, due to a number of differences between the real and virtual worlds. Leafleters in the real world can 1) accompany the offering of the leaflet with a sales pitch, 2) target their efforts to persons visiting the site of the organization being criticized, 3) delay visitors’ access to the target site while determining whether they are interested in receiving the information, and 4) present their message without having to compete with hundreds or even thousands of other leafleters simultaneously trying to win the person’s attention. Where leafleting escalates into picketing, there is, of course, the well-known "signal effect" of the picket line in the real world, (which also entails increased opportunities for taking advantage of the grey area between impeding access and conveying information) which is not available in the current virtual world.
Although we will see later in the discussion of the BCCA case that the technology of meta-tag copying does offer some limited opportunities for unions to present their full message to Internet surfers seeking services of the type provided by the employer, this tends to be an unsatisfactory method since it is based on misleading the surfer. Furthermore, once these individuals have found the employer’s address, they are likely to book-mark it for future reference, and thereby avoid future encounters with the union’s point of view.
The next major salvo from the anti-Sunstein camp is an article by Dan Hunter, of the University of Pennsylvania. In this rather vitriolic 50-page long book review, Hunter makes a not so subtle reference in his critique to the fact that he has degrees in cognitive psychology and computer science, as well as law, and that Sunstein merely has "lawyerly qualifications". He claims that Sunstein’s entire thesis rests upon the idea that the Internet allows for perfect filtering of information. Then he proceeds to demonstrate the technical flaws in such a filtering process in the sense that it is not capable of producing a completely accurate picture of one’s entire tastes and interests. However, this is quite beside the point. Sunstein’s thesis does not depend upon completely perfect filtering in Hunter’s sense. Sunstein merely notes that using personalized news services on the Internet, you can restrict your news to items about certain very narrow topics or to items by commentators whose views you agree with, and that this creates the danger of polarization and balkanization. Hunter’s parading of his knowledge of "machine learning" and the fact that computers cannot yet be accurately taught how to anticipate a human’s every whim is interesting but really not germane to Sunstein’s arguments.
Hunter next attacks Sunstein’s arguments citing research studies showing that individuals deliberating in a group tend to move toward an extreme position. He argues that Sunstein neglected some research showing that in some cases, the group arrived at a consensus which was more cautious than that of its individual members. However, he buries in a footnote the admission that this cautious consensus tended to appear more where a greater number of novel arguments were introduced into the group. Obviously, in cases of groups which tend to be slanted one way initially, which is what Sunstein is concerned about, the novel argument pool will be small or non-existent and therefore Hunter’s cautious consensus research studies are of little relevance.
Hunter also criticizes Sunstein’s suggested technological remedies, i.e. "must-carry" links to opposing sites, by arguing that in a world where perfect filtering is possible, these remedies could be easily over-ridden by the close-minded through use of such filtering. However, this argument again erroneously pre-supposes that Sunstein’s thesis is based on an assumption of perfect filtering in Hunter’s sense.
Finally, Hunter attacks Sunstein’s postscript and sequel to Republic.com. Hunter objects to Sunstein’s pointing to the 2000 U.S. Presidential election as an example of balkanization produced by the Internet. However, Hunter does not offer any other explanations for the statistical anomaly of the outcome of tens of millions of votes hinging on the counting of a few hundred. Nor would Hunter be able to explain the unprecedented second place finish of Le Pen, the far right candidate in the 2002 presidential election in France, or the unusually close 2002 election in Germany.
The most recent parry against Sunstein is by Professor Anupam Chander of the University of California, Davis. Professor Chander focuses on the issue of minorities and other traditionally disadvantaged groups. He argues that these groups’ exclusion from mainstream media can be remedied by access to the Internet. There is some doubt as to whether his assertions about the mainstream media are correct, given the proliferation of cable television specialty channels, e.g. for African-americans, Asian-americans, gays etc. However, leaving aside this issue, the fact remains that even if the Internet is necessary for minorities to maintain and reinforce their vibrant identities, at the same time, it creates greater opportunities for organization and proselytizing by extremist groups which would deny those minorities their rights. The whole process is a vicious circle with the extremists threatening the minorities, who in response entrench themselves further into their unique cultures, increasingly agitating the extremists, and so on.
There has been an academic backlash against the "second generation" of Internet scholarship as represented by Sunstein and others. Although this paper’s review of the anti-Sunsteinites is necessarily brief, it is evident that their attacks seem to be not sufficiently well grounded to refute the notions that the Internet does produce balkanization, fosters and aids extremist groups and provides little in the way of meaningful public forums especially for trade unions wanting to bring their message to persons visiting employers’ web sites.
In the Argenbright example, had the electronic picketing by the union not been prematurely terminated by Yahoo!, the union might have won better working conditions and Argenbright may have been able to attract more dedicated employees (or at least ones who were not asleep at the switch due to having to work at 2 or 3 jobs to support their families) and maybe September 11 could have been prevented. The electronic picketing in question involved the SEIU local at the Los Angeles airport (LAX) and not one of the airports where the hijacked planes originated. However, if SEIU had been able to reach a collective agreement with Argenbright at LAX, would this have set a pattern for other airports where Argenbright provided screening services, such as Newark where two of the hijacked planes originated? In the minds of reasonably alert, properly trained and committed employees, would alarm bells have gone off when 5 or so young Arab men each equipped with a razor-sharp box cutter attempted to board a plane? Or was the tragedy primarily due to a failure of U.S. international intelligence? These questions are purely speculative and the answers would be based on the luxury of 20/20 hindsight, so they are best left as rhetorical ones, i.e. unanswered. However, the fact that they can even be seriously asked indicates that there may be a fundamental flaw with the present structure of the Internet.
The labour issue is just one piece of the larger puzzle. However, it is a good paradigm for purposes of examining the problems raised by the second generation of scholarship about the Internet, as well as suggesting solutions for those problems. The second generation of Internet scholarship shows that the metaphor of the present Internet as an "information highway" tends to be misleading. Instead, it should be analogised to a series of wormholes in cyberspace which enable persons to travel almost instantaneously from A to B without transiting the public spaces, e.g. the highways, streets and sidewalks which one must normally pass through to reach one’s destination. This lack of being forced to pass through public spaces in order to reach one’s destination renders the Internet unique and at the same time minimises its usefulness as a public forum.
This paper will discuss the notion of the electronic sidewalk or "cydewalk" as a countervailing measure to compensate for the shortcomings of the existing Internet as a public forum. It is somewhat ironic that the first person out of the gate (no pun consciously intended) with regards to the electronic sidewalk was Microsoft, which initiated sidewalk.com in 1996 in Seattle as a form of local electronic entertainment guide/yellow pages. This electronic sidewalk was expanded to over a dozen cities and in 1999 was sold by Microsoft to City Search/Ticketmaster, which now operates it as a series of web sites which are constituted by a city’s name followed by .com, e.g. www.toronto.com. Clearly, it is not a public forum in any meaningful sense. The Microsoft "sidewalk" is one cleansed of all picketers, protesters, leafleters, pan-handlers, and the homeless, and as such is a sidewalk in name only. In order to distinguish the Microsoft business from the proposal which is the subject of this paper, I have called the latter the "cydewalk".
The first reference to the cydewalk proposal in the academic literature is an article published in 1998 by Noah D. Zatz, Sidewalks in Cyberspace: Making Space for Public Forums in the Electronic Environment. Zatz’s proposal is to institute a legally-mandated and government-administered system of cydewalks on the Internet. Persons seeking permission to convey information on the cydewalk outside of a targeted web site would apply to the government agency responsible for the cydewalk and would pay a license fee to discourage frivolous applications. Only non-commercial speech would be allowed. Furthermore, messages that have some connection to the target site would have priority over others that had no such connection. Zatz’s proposal does not make any distinction between labour and non-labour protestors.
Under Zatz’s proposal, the cydewalk could be either server-based or browser-based. To explain the difference in technological layman’s terms, the server is owned by the ISP (Internet service provider, e.g. Bell Sympatico, AOL Canada etc.) and stores the information contained in the target’s web site, while the browser is the part of the information-seeker’s system which permits that individual to find the information for which he or she is searching. The primary advantage of the browser-based system is that it would permit the cydewalk agency (presumably a federal Canadian agency) to have jurisdiction over Canadian information-seekers whether they are visiting Canadian sites or ones outside of Canada (such as the Zundel or al-Qaeda site). The corresponding disadvantage of the browser-based system is that information-seekers located outside of Canada, but visiting sites in Canada, would not be subject to the cydewalk system. Another disadvantage of the browser-based system is that it could be more easily evaded by the information seeker than a server-based system.
For both the server and browser based models, Zatz’s proposal sets out several options for giving protesters access to the web sites of the organizations they are protesting against, and the corresponding obligations on the persons visiting those sites. Under these options, the visitor to the target web site can do the following:
a) Click on an icon to check for protestors.
b) Read a short "pop-up" window message from the protestor, and choose to click on it to visit the protestor’s web site, ignore it or close it.
c) Be automatically directed to the protestor’s web site with time, place and manner restrictions based on the nature of the opened page, (e.g. no audio clips so as to prevent unwanted intrusions of noise into home or office).
d) Be automatically directed to the protestor’s web site with no restrictions.
e) Any of (a) to (d), but allowing the visitor to limit the amount of protestors
f) Any of (a) to (e), determined according to the context of the nature of the site being visited and/or the nature of the protestor.
Zatz’s proposal has been mentioned in subsequent academic writing, for example, in Cass Sunstein’s Republic.com. However, the proposal has not yet been discussed in the Canadian context or, more particularly, from the viewpoint of labour relations. The purpose of this paper is to fill that gap and to use the labour relations issues as a paradigm for understanding the flaws of the Internet generally.
Part II of this paper will set out several different schools of thought of what the Internet is and should be vis-à-vis the role of the government in its regulation. This part will analyze how these schools of thought relate to the cydewalk proposal.
Part III will examine the Canadian constitutional and labour relations issues associated with the cydewalk proposal, in the context of electronic picketing, as measured against the various schools of thought on the proper role of the government in the regulation of the Internet.
Part IV will set out some recommendations and the rationale for them.
The Secessionist Libertarians represent the earliest school of thought on the role of the government vis-à-vis the Internet. This school of thought originated in the 1970’s and continued in the 1980’s, when the Internet was primarily used by academics and commercial uses were frowned upon. These academics generally trusted one another and believed that government regulation was not necessary. Their "hands-off" approach to government intervention is perhaps a bit ironic given that the Internet’s direct predecessor, the ARPANET, launched in 1968, was developed by a government agency to deal with the concerns voiced by the Defence Department in the 1960’s, namely, that the centralized design of the United States’ telecommunications system made it vulnerable to a Soviet nuclear first strike.
The Secessionist Libertarians rallied around a document called "A Declaration of Independence in Cyberspace" written by John Perry Barlow in 1996 soon after the Communications Decency Act was passed into law. Barlow, co-founder of the Electronic Frontier Foundation, was of the view that the Internet or "cyberspace" (the term coined by science fiction writer William Gibson in the novel Neuromancer) is actually a separate sovereign territory which is egalitarian, communal, open, self-regulating and based on enlightened self-interest according to the "Golden Rule". I have called this school of thought, represented by Barlow, the Secessionist Libertarians because they have declared cyberspace independent from all other governments and, at the same time, they espouse self-regulation with no formal government system, if possible.
Barlow’s declaration was subsequently taken up and refined by academics, most notably Professor David Post, a law professor at Temple University in Philadelphia, in his well-known 1996 article, Law and Borders – The Rise of Law in Cyberspace. Post argues that the development of cyberspace destroys the traditional link between geographical location and legal phenomena, since the net enables transactions between persons who in many cases do not know the other’s physical location. He notes that, for example, a domain name assigned to a computer, for example, .uk, may be physically moved with the computer to another country without changing the domain name. He further notes that it is often impossible to equate the address with an individual, who may be anonymous, or one of several using the same address.
It has been suggested by Professor Michael Geist of the University of Ottawa that the cyberspace self-government posited by Post in 1996 is already arguably out of date in terms of the technological framework upon which it is based. The most important technological difference between 1996 and current developments in 2002, in terms of Post’s thesis, is that now there are geographic identification technologies which enable the determination of the location of any Internet address. For example, a company called Infosplit claims, with proprietary algorithm technology, to now be able to accurately pinpoint the country of origin with 98.5% accuracy, the state or province with 95% accuracy and the city with 85% accuracy. This technology is primarily being used at present to focus advertising to a local audience and also to limit transmission of entertainment to areas for which it is licensed for distribution. However, the implications are much broader, in terms of undercutting Post’s thesis that cyberspace should be considered sovereign because it destroys the traditional link between geographic space and legal phenomena.
The Secessionist Libertarians are not only in favour of a separate set of rules governing cyberspace, but they also support self-government and self-regulation. How has this other notion fared in terms of the intervening six years (almost an eternity in Internet time) since the Post article and the Barlow declaration?
The issue of self-governance has mainly been put to the test in the context of the Internet Corporation for Assigned Names and Numbers (ICANN), a non-profit private sector corporation registered in California in 1998, to which the U.S. government has contracted out the co-ordination of the registration of Internet identifier systems, e.g. domain names and Internet protocol addresses.
In one of ICANN’s first attempts at Internet self-governance, it briefly experimented with on-line elections of directors, and discontinued this practice in June, 2002, alleging problems of fraud and vote-rigging, (e.g. individuals voting more than once through multiple e-mail addresses). It is granted that ICANN’s commitment to the concept of on-line democracy is certainly put into question by the August 5, 2002 Los Angeles Superior Court decision ordering ICANN to disclose information to Karl Auerbach, a director elected in the on-line vote in 2000. (This decision is discussed in more detail in Part III of this paper.) However, this lack of commitment is somewhat excusable in that true internet-wide on-line democracy is probably somewhat of a pipe dream, since there are currently several hundred million internet users world-wide. The problems associated with the small ICANN election (in which about 38,000 members voted) would be miniscule compared to the difficulties of a truly inclusive internet-wide vote in which at least 100 million members would have to participate in order to make the election representative in any true sense of the word. Clearly, this is impractical, given the evident problems with verifying registrations, even with a small group of voters such as in the ICANN on-line vote.
One of the obvious disharmonies between the Secessionist Libertarians and the Zatz cydewalk proposal is that Zatz is of the view that a government agency is probably required to administer access, whereas the Secessionists are of the view that the Internet in all respects should be self-governing based on ground-up decision-making. In the view of the problems with Internet-wide democracy outlined above, it seems impractical for the Zatz proposal to be modified to accommodate the concerns of this school of thought, for example, by having the cydewalk administered by an entity in the nature of a reformed ICANN which would conduct Internet-wide on-line votes for Directors.
The other and less obvious disconnect between the Secessionist Libertarians and the Zatz cydewalk proposal is a normative difference between the two. Zatz is of the view that there is convergence between cyberspace and the physical world, e.g. the remote surgery, and that the cydewalk would be a logical extension of this convergence. He states:
Interpreting the Internet as an environment that facilitates and structures action hardly necessitates positing it as a "separate" or "alternate" space, as is commonly done when drawing distinctions between "virtual" and "real" worlds. In this Section I show that one way cyberspace facilitates particular forms of activity and interaction is the construction of places quite analogous to the structuring of our material environs.
The Secessionist Libertarians are of the view that cyberspace is a fully separate place. As Zatz himself notes, from a functional point of view, this separateness is fundamentally inconsistent with the notion of convergence resulting in the extension of physical space into cyberspace. It would be more difficult to modify the Zatz proposal to reflect this particular point of view of the Secessionist Libertarians, but it would not be impossible. One could envision the Zatz proposal as being changed to provide other types of public forums in cyberspace that are not merely replications of the physical world. These alternative types of electronic public forums could include "must-carry" rules in the form of mandatory links to other web sites with opposing views.
This school of thought represents a reaction to the Secessionist Libertarians and expresses a fear that the establishment of cyberspace as a separate sovereign territory with its own rules, developed in a libertarian fashion, is a dangerous form of anarchy. This group generally thinks that the Internet is not qualitatively different from other communications technologies, and therefore should not be immune from national laws. Some of them have referred to the Internet as no more than "cable on speed." Perhaps a more apt analogy, though, would involve a more addictive substance, e.g. "cable on crack".
The views of this group are well expressed in a 1998 article by Professor Jack L. Goldsmith, a law professor at the University of Chicago Law School in his article, Against Cyberanarchy. In his article, Professor Goldsmith sets out several reasons why he thinks that it is misguided to say that the Internet cannot and should not be subject to its own separate set of rules. He notes that the Internet can be regulated by local laws through a consent-based system whereby an individual entering a particular web site would agree to be bound by the local laws, dispute resolution mechanisms and enforcement mechanisms where that site originated. He concedes that web-site consents do not protect third parties or persons incapable of giving consent, e.g. minors, but notes that most harmful local effects of off-shore Internet activity can be regulated through laws that focus on the local end-users, and on local intermediaries such as Internet service providers and credit card companies.
How would the Zatz cydewalk proposal fit in with the Anti-Anarchist school of thought? Although there is some element of anarchy inherent in open public forums generally, the Zatz proposal would limit the number of organizations picketing on the cydewalk and would provide for a government-run licensing system to discourage frivolous e-picketing. Furthermore, the premise of the Anti-Anarchists that cyberspace is not a separate place would seem to fit in with the Zatz concept of the convergence of the two through the replication of physical space in cyberspace. Also, the Anti-Anarchists may support the cydewalk proposal on the basis that it would assist in co-opting the so-called "Hactivists" who advocate electronic civil disobedience, e.g. hacking into web sites.
What position would the Anti-Anarchists take on the browser-side versus server-side question? Bearing in mind that the problem of extra-territoriality is one of the arguments that the Secessionist Libertarians use to support the need for a sovereign realm of cyberspace, would the Anti-Anarchists prefer one type to the other in terms of minimizing the potential for extra-territoriality?
To recap the previous discussion on browser-side versus server-side cydewalks, the former would permit the regulating jurisdiction to determine what information-seekers within that jurisdiction would see on their screens, regardless of whether they visited sites located on servers within that jurisdiction or outside it. The server-based system would allow the regulating jurisdiction to determine what information seekers see on their screens when they visit a web site located on a server within that jurisdiction, regardless of whether the information seeker is inside or outside the jurisdiction. In other words, the browser system would affect web sites outside of the jurisdiction, whereas the server system would affect information seekers outside of the jurisdiction.
Do either of these effects represent extra-territoriality in the true sense? As a fairly obvious example of extra-territoriality, for purposes of measuring these effects, we can look at the notorious May, 2000 decision of a French court which ordered Yahoo! Inc. in the United States to make it impossible for persons located in France to view sites on U.S. Yahoo! servers which contained auctions of Nazi memorabilia, which is illegal in France. How would this be different, for example, from the Canadian Parliament passing a browser-based cydewalk law which required Canadians visiting foreign web sites to view protests by e-picketers? It would be different because it would not direct certain actions to be taken in a foreign country. What then if, under a browser-based system, a Canadian visited the web site of a foreign government which prohibited picketing of government facilities and this Canadian encountered an e-picketer from Greenpeace on the cydewalk? Would this be equivalent to an extra-territorial attempt to extend Canadian freedom of speech laws to the foreign country? Probably not, since again it does not involve ordering any actions to be taken by foreign nationals. Furthermore, the law would not be extra-territorial since it would be analogous to a Canadian law governing Canadian decoding of foreign satellite signals.
What if the Canadian Parliament passed a server-based cydewalk law which required Canadians and foreigners visiting Canadian web sites to view protests by e-picketers? Although arguably this would be an attempt to regulate the behaviour of persons outside the jurisdiction, it is really no different from the accepted practice of blocking access to a Canadian web site for visitors located in foreign countries where the content is not licensed for distribution. In one case the change in behaviour is to see something less and in the other to see something more. In both cases, the change is only in the material that is offered for viewing and in neither is the foreign individual ordered to do something.
The third school of thought is championed by Lawrence Lessig of Stanford Law School. Lessig notes that behaviour can be regulated in each of four different ways: 1) the law, 2) social norms, 3) the market and 4) architecture. He notes that the law can also indirectly regulate behaviour by using the law to change any or all of social norms, the market and architecture. As a historical example of the law indirectly regulating through changing architecture, he cites Napoleon III’s replacement in 1853 of many of the small and winding streets in Paris with wide, straight boulevards, thereby making it impossible for revolutionaries to barricade key parts of the city. As a modern example, he cites the control of vehicular traffic through the placement of "speed bumps". He points out that the architecture of the Internet is primarily software based, and that this architecture regulates behaviour. He notes that in some cases it may be appropriate for the government to step in and indirectly regulate the Internet by requiring or encouraging programmers to include certain features in software code, provided that such regulation is transparent due to the accessibility of the details of the code to the public. I have called this school of thought the "Open-Source Regulators" to capture their beliefs that the Internet should be regulated through software code and that this regulation should be made transparent by providing the details of the coding to the general public.
This school of thought seems to combine some aspects of the Secessionist Libertarians (i.e. the belief that the software architecture of the internet should be made publicly available) with some beliefs of the Anti-Anarchists, (i.e. the belief that the Internet can be regulated by national governments). Presumably, since the Internet is actually a series of linked networks in different parts of the world, the Open-Source Regulators would be content with a pluralist approach in which different governments would require different architectures for various parts of the Internet.
How then would the Open-Source Regulators view the Zatz cydewalk proposal? They would likely be equally as positive as the Secessionist Libertarians about the need for the cydewalk in the first place, in terms of facilitating public forums on the Internet. However, they would probably disagree with both the Libertarians and the Anti-Anarchists in terms of how the goals of the cydewalk proposal would best be implemented. The Open-Source Regulators are of the view that there may have to be some government regulation of the Internet, but it should be indirect, through requiring or encouraging software coders to insert certain features in the code, and transparent, through requiring that code to be publicly available for examination. Lessig also points out that most members of the public cannot read software code and therefore in many cases it would be appropriate for the open-source code to be modular so that the part of the code embodying the regulation could be simply discarded by the user if she wishes. Lessig points out that this modular approach would have the advantage of discouraging or mitigating draconian attempts to regulate. He cites an example from Soviet Russia:
Say you are a Soviet propagandist and you want to get people to read lots of information about Papa Stalin. So you declare that every book published in the Soviet Union must have a chapter devoted to Stalin. How likely is it that such books will actually affect what people read?
Books are an open source software; they hide nothing; they reveal their source – they are their source! A user or adopter of a book always has the choice to read only the chapters she wants. If it is a book on electronics, then the reader can certainly choose not to read the chapter on Stalin. There is very little the state can do to modify the reader’s power in this respect.
The same idea liberates open source code.
Lessig also points out that, in the case of the Internet, there could be some pro-active measures taken by the government to encourage users to retain the regulatory portion of the code, which would allow the users to weigh any burdens of the regulation against corresponding benefits. This could take various forms, e.g. faster access to sites which have a cydewalk "in front" of them (architecture), government funding of programs to enhance awareness of the value of the cydewalk system (social norms), or the government using its purchasing power to flood the market and establish a de facto cydewalk software standard (the market).
Finally, there is a fourth school of thought that I refer to as the Telecosmic Futurists. The term "Telecosm" was coined by economist and Internet guru George Gilder to refer to the entire communications system – telephones, television, radio, and the Internet. Together with forward-looking lawyer and M.I.T. professor Peter Huber, Gilder is of the view that emergent technologies will remove the need for government regulation of the telecosm on the basis that these progressive technologies allow the broadcast spectrum to be shared by an infinite number of users. For example, ultra-wide band broadcasting (UWB) frees a broadcaster from using a specific frequency and allows it to share frequencies or use others already dedicated to different purposes, by layering its own signals on top using extremely short (billionths of a second) low-power precisely timed bursts to transmit information. This would eliminate the need for the bandwith-parcelling activities which occupy the time of Federal Communications Commission and Canadian Radio and Television Commission bureaucrats. Huber supplements the technological and economic expertise of Gilder by arguing that most of the remaining telecosm issues currently dealt with by government agencies, e.g. free speech, censorship, obscenity, piracy, copyright etc, could be managed without the FCC and the CRTC or any other government agency, simply by using the common law, the court system and technology such as the V-chip and the Net Nanny.
Members of this school of thought would obviously disapprove of a cydewalk created by legislation and administered by the government. However, their reaction would be different from that of the Secessionist Libertarians. The Telecosmic Futurists do not think of cyberspace as a unified world-wide sovereign territory, but are more inclined instead to have it governed by the common law, which of course is only present in common law jurisdictions and, even then, varies greatly according to the jurisdiction involved. The Telecosmic Futurists would probably leave it to market forces in conjunction with common law property rights to carve out a cydewalk.
In this part of the paper, I will take the various perspectives on the cydewalk suggested by the four schools of thought and test these perspectives against the recent trends in the Canadian and U.S. constitutional and labour relations jurisprudence. To lay the foundation for this discussion, I will examine the only case to date in Canada and the U.S. which deals with electronic picketing in a labour context. This case provides support for a legitimate role for electronic picketing in labour disputes.
To the author’s knowledge, British Columbia Automobile Assn. v. Office and Professional Employees’ International Union, Local 378 is the first and only case to date in Canada and the U.S. which involves electronic picketing in a labour dispute. Local 378 of the Office and Professional Employees’ International Union (OPEIU) was on strike against the British Columbia Automobile Association (BCAA). During the strike, the union produced a web site which initially incorporated the employer’s trademarks and the employer’s own web site design. The union also copied the meta-tags from the employer’s web site. Meta-tags are computer code on a part of a web site which is normally hidden from the user, containing word strings relevant to the user’s inquiry, which are used by Internet search engines, such as Netscape or Google. The employer’s web site included, for example, in its meta-tags not only "British Columbia Automobile Association" but also non-union related phrases such as "international driver’s permit". The original name of the union’s web site was bcaaonstrike.com. Subsequent to the commencement of the lawsuit by the employer, the union twice modified its web site. The second set of modifications substantially reduced the similarity between the appearance of the two sites, changed the name of the union’s site to bcaabacktowork.com and reduced the similarity in the meta-tags, although not in any substantial way.
The employer’s lawsuit against the union alleged that there was infringement of the employer’s copyright under the Copyright Act, depreciation of goodwill under s. 22 of the Trade-marks Act, and also the tort of passing-off. The court ruled that the first two versions of the union’s web site infringed the employer’s copyright in copying its web site design, but dismissed the trade-mark complaint. With regards to the passing-off claim, the court found that, although the initial web site constituted a passing-off, the second and final web sites did not, even though the meta-tags remained seriously misdescriptive.
The court’s rationale for finding that the second and final web sites did not constitute a passing-off, despite the misleading meta-tags, was stated by Mr. Justice Sigurdson as follows:
¶ 128 I think the plaintiff's complaints are unrealistic. I recognize that the use of BCAA in the domain name is intentional. However, the use of BCAA in the Union domain name and the references that are currently made in the meta tags are not a misrepresentation provided they do not represent the site as that of the plaintiff or associated with the plaintiff. Here, I think that here there were legitimate reasons for the Union to use the domain names and the meta tags it presently uses. The defendant's current use of meta tags is not a misrepresentation. I think that if a site wishes to operate as a lawful vehicle during a strike or as a consumer criticism site, it must be able to reach people who are attempting to find an employer's or a producer's site. Otherwise the Union's lawful activities or the activities of consumer groups would be significantly frustrated.
¶ 129 The language used in the current meta tags (which I note is different from the language used in the first two Union websites) would likely result in people using search engines to find B.C. Automobile Association finding the Union website high on the list of sites on the search results. While I think that the use of similar meta tags unconnected to a defendant's business or operation might indicate deception and might be a significant factor in determining if there is a passing-off, I think that the use of these particular meta tags is not objectionable because it is a reasonable way for the Union to bring its message to people wishing to do business with the employer. The Union must be free to identify its employer and communicate about its labour dispute. The fact of a labour dispute, which is apparent from the Union website is, I think, immediately clear to any reasonable or rushed observer. It is equally clear that the site is not endorsed by or the property of the B.C. Automobile Association.
¶ 130 I agree with the defendant's argument that the common law should be interpreted in a manner consistent with the Charter. When a website is used for expression in a labour relations dispute, as opposed to commercial competition, there is, I think, a reasonable balance that must be struck between the legitimate protection of a party's intellectual property and a citizen's or a Union's right of expression. I think that the principles in KMart require such a balance and the common law should not be interpreted in a way to unreasonably infringe a person's freedom of expression. In connection with this claim for passing-off, that balance favours the Union.
In paragraph 128, Mr. Justice Sigurdson states that whether an organization is a union involved in a labour dispute, or a consumer advocate group, it must be able to communicate its concerns to people who are attempting to find the employer’s or the producer’s or retailer’s web site. He states that if these organizations are deprived of this ability, then their lawful activities would be "significantly frustrated." Here, he entrenches electronic picketing into the law, without any background analysis, e.g. distinctions between businesses that conduct their operations only over the Internet (e.g. amazon.com) and those that operate both on the Internet and in the real world (so-called "clicks and mortar" businesses), or the free speech and labour relations principle of the availability of alternative forums.
In paragraph 129, Mr. Justice Sigurdson finds that there is no passing-off since the key test for that tort is not whether there has been an intent to deceive, but rather whether there was a misrepresentation which was "likely to deceive" the average Internet user. He notes that the case involves a labour relations and not a commercial matter, and that the union’s site did not purport in any way to be endorsed by the BCCA. He concludes, therefore, that there is little or no likelihood that the average Internet user, whether rushed or not, would be deceived or confused as to the origin of the union site. In paragraph 130, he goes on to rule on a Canadian Charter of Rights and Freedoms argument concerning passing off, perhaps to buttress his conclusion about the effect of the alleged deception, or possibly merely as an alternative ruling to shield himself from an appeal. He states that in the context of labour relations as opposed to commercial speech, the common law tort of passing off should be interpreted in a manner consistent with the union’s freedom of speech rights under the Charter.
A case comment on the decision by the plaintiff’s lawyer correctly notes that:
To the authors’ knowledge, this case represents the first time a Canadian court has employed a Charter right to limit the scope of an intellectual property right. As might be expected, the court’s analysis of freedom of expression raises more questions than it answers.
This part of the ruling may also now represent somewhat of an anomaly in the United States, where the trend seems to be that copyright is generally considered to be immune from first amendment claims, to the chagrin of scholars such as Lawrence Lessig.
Sigurdson’s analysis also distinguishes between labour and commercial competition speech in terms of interpreting the common law of passing off so that it is consistent with the Charter. Is this distinction justifiable? The courts have recognized that free expression is particularly critical in labour relations, but if the distinction between labour and non-labour speech is intended as a content-based distinction, it would likely not be a permissible one. Content-based distinctions, (as opposed to time, place and manner restrictions) are generally prohibited under the First Amendment in the U.S. Although in Canada, a purpose versus effect test is used instead of the content versus time, place and manner one, a content based distinction would generally be considered equivalent to a purposive one. Thus a bare distinction in the common law allowing labour but not commercial competition speech would probably be automatically prohibited under section 2(b) of the Charter, and not justified under section 1, on the basis that it was a distinction the purpose of which was to restrict freedom of speech rather than a distinction that merely had that effect.
Is Sigurdson’s labour versus commercial competition speech distinction then a content/ purpose-based one, or is it merely a modality/effect-based one? It is probably the latter, since it seems that the rationale for Sigurdson’s distinction between labour and commercial competition speech is the greater potential for customer/visitor confusion where the context is a commercial one and the two web-sites are in competition with each other. However, this point is already built into the common law tests for passing off, so one wonders why Sigurdson bothers. Perhaps he is just measuring the common law to ensure that it is in accordance with the Charter.
What lessons then can be learned from the BCCA case? We can glean three important ideas from the case: 1) electronic picketing has a legitimate information-providing purpose in labour relations, 2) copying the employer’s meta-tags for purposes of conveying information to the employer’s web site visitors is currently the only effective non-criminal and Charter-protected model of electronic picketing (i.e. non-criminal in that it does not involve hacking into the employer’s web-site) and 3) in justifying the copied meta-tag, and therefore legitimizing electronic picketing, the court took the unorthodox and controversial position that, in a labour dispute, intellectual property rights are watered down by freedom of speech rights under the Charter. It is not clear from the wording of the decision whether this Charter finding is part of the ratio of the case or where it is merely obiter dicta. If it is the former, then the decision arguably illustrates the ancient adage that "hard cases make bad law" and also may indicate that to avoid further entrenching this bad law into the jurisprudence, legislative intervention is needed to create places to picket in cyberspace, i.e. the cydewalk.
What about the background analysis which is missing from Sigurdson’s decision concerning the issues of the availability of alternative forums, and the lack of a distinction between a purely Internet-based business such as amazon.com and a combined Internet/physical world ("clicks and mortar") business such as the BCCA? Rather than protecting freedom of expression through the unorthodox device of using the Charter to water down intellectual property rights, did he merely reach the wrong result in that the union should have used the alternative of picketing in physical space? Of course, this alternative would not have been available if the employer were a purely Internet-based business such as amazon.com, but that is not the case here.
The relevance of the issue of the availability of alternative forums is recognized in both Canadian and U.S. law. Richard Moon, of the University of Windsor, states:
When judging the legitimacy of a time, place and manner restriction, the central issue is not the proper balance between the value of expression and the value of privacy or quiet but whether the individual seeking to express himself is left with adequate alternatives for his communication.
Professor Moon also notes that the focus on the available alternatives will be less where the means of expression are critical to the effectiveness of the message, in contrast to other situations, such as the Supreme Court of Canada’s example in IrwinToy where an individual parks his car in an illegal spot as a form of political protest against parking by-laws:
Because the Supreme court [sic] of Canada has defined expression broadly to include all acts intended to convey a message, any act is potentially an act of expression. This also means that any law is potentially a time, place and manner restriction on expression. Understandably, the courts are reluctant to require substantial justification for a law, such as a parking restriction, that would not ordinarily be seen as impeding expressive freedom. In the case of such restrictions, it will almost always be the case that the individual speaker has effective alternatives. In some cases, however, where the means of expression seem critical to the effectiveness of the message, the court may choose to exempt the expressive act from the rule’s application but not strike down the entire rule.
In the British Columbia Automobile Association case, Sigurdson found that the union’s lawful efforts would be "significantly frustrated" if it could not reach persons who were attempting to visit the employer’s web site during the work stoppage. This is essentially the same as saying that the electronic picketing is "critical to the effectiveness of the message" of the union as Moon puts it. However, Sigurdson does not explain his finding. His decision contains no references to the proportion of its business that the BCAA conducted over the Internet as opposed to physical locations. Neither does it contain any references to any evidence put forward by the union that would support the finding that the electronic picketing was essential to the communication of the union’s message. The court may have just assumed, and perhaps correctly so, that a significant portion of the employer’s business was conducted over the Internet (otherwise, it would not have contested the matter so vigorously) and therefore it was essential for the union to get its message out that it reach the sizeable portion of the market represented by persons who were attempting to find the employer’s web site. Furthermore, the court may have been influenced by the real possibility that, if the union could not picket electronically, the employer may, during the work stoppage, have transferred the bulk of its operations from the physical world to the Internet, thereby intensifying the freedom of speech problems, and perhaps precipitating an electronic "race to the bottom" by competing employers in terms of protecting themselves from union speech.
What about the possibility, then, of alternatives within the sphere of electronic picketing, i.e. less objectionable electronic alternatives to copying the employer’s meta-tags? The court did not delve into this issue at all and merely adopted the existing case law which supports the notion that meta-tag copying is essentially the only legal way to reach persons visiting the target web site. Was the court right in adopting the view that there are there no other less objectionable, viable electronic alternatives? Yes, the court was probably correct. Another potential method which has thus far been used in a commercial context only, but could conceivably be used in a labour dispute, is overlaying a banner or "pop-up" over the employer’s web-site through services offered by gator.com. Without the target site’s permission, the banners and pop-ups appear on the screens of subscribers to the free software distributed by gator when they visit the target site. Gator claims to have over 10,000,000 subscribers in the United States. However, in June, 2002, several large newspapers with web sites, including newyorktimes.com and washingtonpost.com launched an action in U.S. Federal Court in Virginia against gator.com seeking to prohibit it from delivering such banners and pop-ups, so the future of this innovative and arguably intrusive technology is uncertain. Also, the technology does not reach all persons visiting the target site, only gator subscribers. A further method of electronic picketing is a denial of service attack by tying up the employer’s server with "spam" messages, but these are to block access, not to convey information, so even a creative interpretation of the Charter would not protect them. There is also the method of the employees communicating directly with the employer, and vice-versa, during a labour dispute, but this does not allow third parties, e.g. customers, taxpayers, etc. to receive information. Finally, there is the method of picketing by hacking into the employer’s web site and posting information there but that of course is a criminal offence.
In summary, then, the British Columbia Automobile Assn. case lends support to the proposition that electronic picketing is essential for a union’s freedom of speech rights. This is certainly a valid viewpoint, given the importance of the Internet in disseminating information during a work stoppage. To the extent that electronic picketing may lessen the overall need for picketing in physical space, e.g. perhaps by shortening the strike through improving information flow, it makes sense from the point of view of efficiency. Electronic picketing does not entail the problems associated with picketing in physical space, i.e. violence, noise pollution, litter, interference with vehicular and pedestrian traffic, disruption of essential services, and the need for increased police involvement. Of course, from the union’s perspective, electronic picketing would be a supplement to and not a substitute for picketing in physical space, but that is a different question to be dealt with later in this paper.
What is disturbing, though, about the BCAA decision is the unorthodox and controversial part of the Charter analysis which states that intellectual property rights are watered down by freedom of expression rights. It is this unorthodoxy in the decision which may indicate that what is needed is a new solution (e.g. legislation) to deal with the issue of electronic picketing. This new solution may even have appeal for those such as Lawrence Lessig who think that the unorthodox view expressed by Sigurdson is correct. Lessig, as an advocate of open-source code, would probably be reluctant to base the edifice of free electronic expression by unions on the foundation of a less than transparent practice which diverts persons from their intended destination on the Internet.
By way of analogy, suppose that the only way that a union could convey information in the real world were to display in front of union headquarters a sign identifying the address as the employer’s business and then giving union leaflets to visitors as they mistakenly went inside? Would this not prompt an outcry against the union’s misleading or diversionary practices as well as a call for a legislative solution? It is granted that the practice of meta-tag copying does not always involve diversion, since in some cases, the individual will be seeking information about the employer from the union’s perspective, and the copied meta-tag may facilitate such a search. However, these instances will be far fewer in number than those in which the individual enters into the search engine, for example, the term "international driver’s license" and inadvertently ends up at the union’s site.
Meta-tag copying is not only based on a diversionary and less than transparent practice, but is also somewhat of an ad hoc and inefficient measure, since it only reaches first time visitors if they do not already know the employer’s web site address, and also cannot correctly distinguish between the true and the misleading links displayed in the search engine results. Therefore, it is flawed in its application, as well as its foundation. Lessig would probably consider it as an example of software code poorly regulating the Internet, with the need for legislative intervention to urge the code writers in the proper direction.
As mentioned in Part II of this paper, the Secessionist Libertarians would not only obviously object to government creation and regulation of the cydewalk, but would also probably object to the cydewalk proposal being essentially a replication of the physical world in cyberspace. The Secessionists would likely be of the view that a cydewalk in cyberspace that mimics the functions of its counterpart in the real world would demonstrate a functional convergence of the two worlds which is inconsistent with the separateness of cyberspace. The Secessionists, therefore, would favour alternatives to the cydewalk, such as cross-links between sites with opposing views. Even more fundamental to the platform of the Secessionist Libertarians, of course, is the notion that cyberspace is a sovereign territory which should be self-governed and immune from local laws. As we have seen in the discussion earlier in this part, self-government of the world’s Internet community through on-line elections or referenda is something of an elusive pipe-dream. The notion of cyberspace being sovereign and immune from local laws is probably a naïve, though noble, objective. It is being proven by recent trends in the jurisprudence not only to be impractical but also to create an unjust result in many cases, due to certain rulings by ICANN.
A very recent example of this phenomenon is found in the July 18, 2002 decision of Mr. Justice Blenus Wright of the Ontario Superior Court of Justice, Douglas Black v. Molson Canada. In this case, the court overturned a decision of an adjudicator appointed pursuant to ICANN’s Uniform Dispute Resolution Policy (UDRP) and Start-Up Trademark Opposition Policy (STOP). The adjudicator, a retired U.S. judge, had ruled that the domain name "canadian.biz" should be transferred from the applicant Douglas Black, Ph.D., a Canadian research consultant and web-site developer, to Molson Canada. The adjudicator based his ruling on a finding, pursuant to criteria set out in the STOP rules, that:
i) the domain name is identical to a trademark in which the Complainant (Molson Canada) has rights
ii) the respondent (Black) has no rights or legitimate interests in respect of the domain name; and
iii) the domain name has been registered or is being used in bad faith.
The Ontario Superior Court of Justice blasted the adjudicator’s ruling, finding that when Molson Canada registered the trademark "Canadian", it only intended to use it with respect to beer, and that it therefore did not have the exclusive right to
use the word. Mr. Justice Wright also noted that there are hundreds of Canadian businesses which use the word as a prefix or suffix for identification purposes.
He concluded the first criteria in ICANN’s STOP rules is flawed:
Simply because a domain name is identical or similar to a trademark name should not result in the transfer of the domain name to the trademark owner. In my view, unless there is some other evidence that the use of the domain name infringes on the use of the trademark name, a person other than the owner of the trademark should be able to continue to use the domain name.
Mr. Justice Wright then went on to criticize the adjudicator’s finding that Mr. Black had no legitimate interest in the name and acted in bad faith, noting that Black intended to use the name for bona fide business purposes.
On July 25, 2002, in his regular column "Cyberlaw" in the Globe & Mail, University of Ottawa Professor Michael Geist made the following perceptive comments on the adjudicator’s flawed ruling, noting that unfortunately it is fairly typical of ICANN rulings, (e.g. the transfer of "Ottawa.biz" from the capital of Canada to a Texas tractor manufacturer or the transfer of "Nissan.com" from an individual with that last name for 50 or so years to the Japanese car manufacturer):
It is ironic that while global e-commerce legislation has been focused on ensuring equality between on-line and off-line rules, domain name dispute resolution policy has granted trademark holders far more rights on-line than off-line and in the process created a new super-trademark. Perhaps this decision will begin to put a stop to policies that left traditional trademark norms behind long ago.
The Molson Canada case clearly demonstrates the problems inherent in constructing a sovereign system of law in cyberspace. The major difficulty is caused by using local laws as a reference point in cyberspace and then distorting them, with the result that the local law is applied in two different ways – one on-line and the other off-line, leading to unjust results and a perversion of the original intent of the local law. Although not a labour relations case, the Molson Canada decision suggests that similar problems could arise in the labour relations context. In creating a cydewalk or a variation of it in cyberspace, the Secessionist Libertarians would likely rely on an adjudicator similar to the ICANN appointed ones. This adjudicator would likely rely on principles gleaned from Labour Relations Board cases in determining the rights of union vis-à-vis the employer. The adjudicator would presumably consider labour relations legislation, but would not consider himself bound by it and, due to not being an expert in it, would likely distort it in the same manner as the adjudicator perverted the intention of local trademark law in the Molson Canada case. This would lead to a potentially unfair result, a misrepresentation of the purpose of the legislation, and a two-track labour relations system with one for the on-line world and the other for the off-line one. This would be unfortunate, especially given that e-commerce legislation, e.g. the Personal Information Protection and Electronic Documents Act, is designed to avoid disparate treatment of the real and virtual worlds.
The most recent court decision in the ICANN saga is the August 5, 2002 decision Auerbach v. ICANN by the Los Angeles Superior Court. Karl Auerbach, an ICANN director elected in the on-line election in 2000, had requested ICANN to allow him to inspect and make a copy of its general ledger and other records, so that he could properly carry out his duties. After over 10 months of stalling, ICANN agreed to permit Auberbach to inspect certain records but would not provide him with an electronic version or allow him to make copies, and required him to sign a statement of confidentiality and to agree to other conditions. Since ICANN is registered as a California non-profit public benefit corporation, Auerbach, who was sponsored by the Electronic Frontier Foundation, filed suit in California against ICANN, asking for a Writ of Mandate to compel inspection and copying of ICANN’s records. The court granted Auerbach’s request, noting that section 6334 of the California Corporations Code states that "Every director shall have an absolute right at any reasonable time to inspect and copy all books, records, and documents of every kind and to inspect the physical properties of the corporation of which such person is a director." The court noted that the burdensome and unreasonable restrictions imposed by ICANN violated section 6334 and also were not even authorized by the ICANN Board of Directors, but were promulgated by "an ad hoc group of functionaries".
Auerbach’s victory is a Pyrrhic one, since ICANN, likely influenced by the law- suit, has decided to move away from on-line elections for directors, and he will soon be replaced by a director appointed by corporate interests. Even before the lawsuit, ICANN was probably less than fully committed in the first place to the concept of a self-governing Internet based on ground-up decision-making, due in part to its reliance on low interest loans from corporations such as IBM.
However, as was shown previously in Part II of this paper, even if ICANN were more committed to the idea, true internet-wide democracy still remains something of a pipe dream due to difficulties of properly managing the registration and voting processes in a world-wide vote involving 100 million or more electronic ballots. As we have seen, the primary problem would be with ensuring the integrity of the registration process. Even sending out registration forms by "snail mail" to all eligible voters, a gargantuan task, would not be certain to eliminate most voter fraud in the form of individuals using multiple e-mail addresses to vote more than once. The technology (e.g. home-based biometrics) may one day be available to virtually eliminate fraud in such a vote, but it is not available at present.
Of all the four schools of thought on the Internet, the Anti-Anarchist school is perhaps the most consistent with the cydewalk proposal as formulated by Zatz. It is consistent with having national laws apply to the Internet, and having a government body regulate and license access to the cydewalk. The Anti-Anarchists although somewhat in favour of the status quo generally, could probably at some point be persuaded to support the cydewalk proposal on the basis that it would take the wind out of the sails of the Hactivists.
The same concerns that drive the Anti-Anarchists to fear disorder as a result of a self-governing sovereign cyberspace would also likely push them towards the direction of a cydewalk that is strictly regulated. Access would not only be denied, for example, to certain individuals on the basis that their views are frivolous, defamatory, fraudulent, or constitute hate crimes, but also because of sheer logistics so as to avoid overcrowding. After all, it takes a great deal of commitment both of time and money (i.e. the foregoing of earnings) as well as physical endurance to stand outside on a picket line during all types of weather. Even if a union has an extensive strike fund, it will have difficulty finding sufficient numbers of individuals so that the picket lines could be manned on a 24/7 basis. Picketing on the cydewalk, on the other hand, could take place on a 24/7 basis rain or shine with a minimal commitment of time, money and manpower, and with an equal if not a greater opportunity to convey information than in the real world. It is true that the signal effect of picketing on the cydewalk may not be as great as in the real world, especially with the less intrusive options such as the flashing icon, but this would probably be compensated for by the fact that the number of persons exposed to one e-picket line would usually be much greater than the number exposed to any one picket line in the real world. Furthermore, although the signal effect in the real world may depend to a large extent on public exposure and peer pressure, i.e. the fear of negative opinions of those who observe the individual crossing the line, this syndrome could still exist on the cydewalk if there is at least the perception that the individual’s privacy against cyber-snooping by the union is not absolute. Finally, it has been pointed out that the signal effect in the real world may be exaggerated, particularly in large cities where there is more diversity.
Since the cydewalk would be such a tempting tool for unions (as well as other protesters) to use on a 24/7 basis, during work stoppages as well as during labour "peace", some measures would have to be put into place to ensure that there would not be a "tragedy of the commons" as Lessig puts it where overuse by all destroys it for everyone. These measures could include the following types of restrictions:
i) a ban on messages that are fraudulent, defamatory, hate crimes, child pornography etc.;
ii) a small license fee to discourage frivolous applications and overstaying, with exemptions for the indigent;
iii) a ban on commercial speech;
iv) a preference given to protestors whose message content is relevant to the site in question;
v) a ban on protestors in front of personal or family web sites (the equivalent of a residence)
vi) a preference given to trade unions during a work stoppage; and
vii) restrictions on the number of protestors at a given time.
These restrictions would raise a whole host of possible freedom of speech issues, however, most of them are probably necessary to avoid exacerbating the very problem which the cydewalk was designed to curtail, i.e. the propagation and cultivation of extremist views through fragmentation and balkanisation. Most of these restrictions could be justified though the argument that, by passing legislation to create a cydewalk, the state is, in effect, expropriating a portion of the web site owner’s property for public purposes. The issue then becomes one of the access to state property for purposes of freedom of expression. The applicable test would be the one set out in the Supreme Court of Canada’s decision in the Dorval Airport case. In that decision, the court was unanimous in finding that the restrictions on political pamphleteering at the airport were unconstitutional, however, there were three different rationales put forward by the justices. Chief Justice Latimer found that a restriction that is based on the incompatibility of the particular form of expression with the state’s property use does not violate section 2(b) of the Charter and so does not require justification under section 1. He found that the restriction on the distribution of political pamphlets at the airport did violate section 2(b) on that basis.
Madame Justice McLachlin stated that a restriction will violate 2(b) even though it is incompatible with the use of the property, if it can be shown that the expression would advance the constitutional values surrounding freedom of expression, in particular whether there is a relationship between the forum and the particular expressive activity. McLachlin then states that if there is a violation of 2(b) on this basis, a section 1 analysis would require asking several types of questions including whether the location has special symbolic significance for the information being communicated, whether there are alternative forums, and what the claimant would lose by being denied the opportunity to communicate in that particular location. Finally, Madame Justice Heureux-Dubé stated that any time the state restricts expression on state property, there is a violation of section 2(b) which must be justified under section 1. However, she indicated that the Oakes section 1 test of rationality, minimum impairment and proportionality should be applied strictly in these types of cases. She suggested a balancing of private and public interests under section 1, in which issues such as compatibility of the message with the forum, alternative forums, traditional openness of the property etc. are examined.
All three types of analysis in the decision would seem to support restrictions i) to iv) above. More problematic would the last three types of restrictions, i.e. a complete ban on picketing in front of personal or family sites, a preference given to labour protests during a work stoppage, and restrictions on the number of picketers. The issue of picketing in front of residences was considered by the Supreme Court of Canada in the Pepsi decision, where the court struck down the illegal per se line of cases on secondary picketing, but upheld the portion of the injunction regarding residential picketing on the grounds that on the facts of the case it involved the torts of nuisance and intimidation. Therefore, the court left open the possibility that non-tortious residential picketing would be permissible.
Would this mean that a complete ban on access to the cydewalk in front of personal or family web sites would be unconstitutional? I would argue no, on the basis that the principle of forced speech indicates that a stricter standard should be applied where the "expropriation" of web site space by the state involves the property of an individual as opposed to a corporation or an organization. In Canada and the U.S., the constitutional prohibition against forced speech applies on the basis that "not only is it wrong for the state to prevent an individual from communicating with others, but it is also wrong to compel an individual to communicate against his or her will." The defence to a forced speech claim is that there is no real danger that the views expressed would be misattributed to the claimant. Although the danger of misattribution is low in a labour dispute due to the public’s knowledge of the traditional opposition of employers and unions, this would not necessarily apply in picketing of personal web sites, where a visitor may confuse the union’s views with the web site owner’s views. Furthermore, the prohibition against forced speech is stronger where an individual’s property is involved, since that involves the invasion of the person’s "sphere of intellect."
What about a restriction that would give a preference to union protesters during a work stoppage? Why would such a restriction be necessary at all? One reason may be that the extent of the signal effect of picketing varies in direct proportion to the number of picketers on the line at any given time. For example, a union with 50 members picketing against their employer at a location in the real world would thus argue that to give the same cydewalk access to the union picketers as to 50 other non-union protestors, each of whom had a separate non-union related complaint against the employer, (e.g. environmental, consumer complaint etc.) would be unfair in the sense that the signal effect of the union would be severely diluted. Furthermore, although these non-union protestors would not likely be prepared to invest the time and effort into picketing on a 24/7 basis in the real world, they would be more prepared to set up a cydewalk picket given the relatively small amount of time and effort involved (even with token licensing fees to discourage frivolous picketing). Also, what about the possibility that anti-union groups, which would normally not dare to demonstrate at a premises at the same time as a union was picketing for fear of provoking physical violence, would lay claim to cydewalk access during a work stoppage, knowing that in cyberspace, there is no violence? Should the union be given preference to better replicate the situation in physical space?
The answer is that, generally, preference should not be given to labour over non-labour speech, since it would be an impermissible content-based distinction of which the purpose was to restrict certain types of speech. The impermissibility of distinguishing between labour and non-labour speech was cemented into the law by the Supreme Court of Canada in the Pepsi decision where the court stated that one of the problems with the illegal per se prohibition on secondary picketing was that it distinguished between labour and non-labour speech. A time, place and manner restriction whereby only the trade union would be allowed access during a work stoppage would probably not be permissible in that it would distinguish between types of speech based on content.
As we have seen, in the BCCA case, the danger of forced speech due to misattribution is very low in a labour relations situation, because of the public’s general knowledge of the antipathy of employers and unions. Would this be a basis for allowing union access to the cydewalk to the exclusion of all other protesters, e.g. environmentalists, consumer advocates etc.? No, it would not. Firstly, there is some non-union speech which is speech that would not create a real risk of misattribution, e.g. consumer advocates criticizing a car manufacturer’s products. Secondly, there is some non-union speech with which the employer may well agree, e.g. anti-union protestors. Thirdly, it would be rather bizarre to attempt to assist employers to avoid a forced speech situation by creating a situation where their traditional opponent and sometimes their arch-nemesis, the union, has sole access to the cydewalk. Fourthly, in a system where the cydewalk would be mandated by legislation, the public, with the aid of suitable disclaimers, would probably come to understand that the views expressed on the cydewalk are not necessarily those of the site "behind" the cydewalk. As we will see later in this paper, this is in contrast to the optional cydewalk suggested by the Open Source Regulators school of thought.
What then about a possible solution of distinguishing between the employer’s intranet site (a secure internal network which only employees can visit) and the employer’s Internet site (which all the world can visit), for purposes of cydewalk access? Under such a distinction, the union would get preference or even sole access to the cydewalk in front of the intranet, but would not get such special treatment for the cydewalk in front of the Internet. This would allow the union to maximize its signal and informational effect vis-à-vis members of the employer’s workforce, e.g. management, members of other unions not on strike, replacement workers etc. Would this be permissible constitutionally? The answer is that the intranet cydewalk would probably be more analogous to the union bulletin board in the workplace rather than a public forum, since the intranet is not open to the public. As such, it would not be unconstitutional to restrict access to the cydewalk to the union, or at least to employees generally, whether union members or not. However, this would be an imperfect solution from the union’s point of view, since it would lose the opportunity to sway public opinion in its favour by picketing on the cydewalk outside the Internet.
Finally, what about the solution of simply limiting the number of protestors on the Internet cydewalk on a content neutral basis (e.g. first come first served) so that the union’s message is not unduly diluted by other protesters? The first problem with this solution is, of course, that the union might be one of the protesters which is excluded. Another problem with this solution is that the Ontario Court of Appeal confirmed in the Industrial Hardwood case that restrictions on the number of pickets would infringe upon freedom of expression unless it is demonstrated that the restriction is necessary to prevent blocking of or substantial interference with access to the premises. How would this rule translate into cyberspace? In the case of the cydewalk, the restrictions would not involve the number of union pickets, but rather the overall number of protestors, both union and non-union. Would such a restriction be reasonably necessary to prevent blocking of access? This depends on the type of cydewalk that is implemented. Clearly, if it is the type that automatically diverts the visitor to the protester’s site, then access would be effectively blocked, or at least unreasonably delayed, if there is a series of 50 protestors, all of whose sites the visitor must visit before he can get onto the employer’s site. If, on the other hand, it is an icon type of cydewalk, where there are a group of 50 protestor icons which can be turned off all at once with one click, then there is no real interference with access. In fact, this would resemble most user’s computer desktops, mine included, (with the exception of the protest part). The union’s message would definitely be drowned out in such a situation.
It might be of some assistance to the union if, as Zatz proposes, the number of protesters could be determined by the Internet user, i.e. the person visiting the employer’s site. However, where a person selects a small number of protesters’ icons to be displayed, there is the risk that the union may be excluded from the mix.
We have seen a few pages earlier that the issue of forced speech is not problematic from the point of view of union messages on the cydewalk, due to the low danger of misattribution. The other side of the coin of forced speech is the problem of forced listening and privacy. This tenet of freedom of expression provides that where the expression of a person’s views would require holding the listener captive, or intruding into the privacy of the listener’s home with no reasonable way of avoidance (e.g. loudspeakers on the sidewalk outside a residence), then it is reasonable to restrict the expression to avoid the forced listening that it entails. Zatz concludes that the cydewalk proposal would not run afoul of the principle of the prohibition against forced listening, since the Internet user is not captive, (he or she can terminate an unwanted speech encounter with the simple click of a mouse or a keystroke or a voice command). Also, the Internet user takes affirmative actions to enter cyberspace, (e.g. logging on, entering a web site address etc.) similar to the actions of going out the door and leaving one’s house, which involve a voluntary assumption of the risk of receiving exposure to unwanted information. Finally, the user is not in a "zone of privacy" (since Zatz’s view is that entry into cyberspace is associated with the same limited level of privacy which one would expect in the course of a walk down the street in the physical world).
Zatz’s notions that the Internet user is not a captive audience since he can click a mouse to close a window, and that the user voluntarily assumes the risk of unwanted information when he logs on, are fairly easy to accept. However, his notion that the Internet user has no more of a reasonable expectation of privacy than does an ordinary individual walking or driving down the street is somewhat disturbing. When an individual walks or drives down the street, he currently may be subject to video (police, corporate), or satellite surveillance (GPS, military etc.), either continuous or selective, recorded or unrecorded. The streets of Manhattan, for example, are continuously monitored by approximately 2400 video cameras, public and private, fixed to buildings. The Washington D.C. Police Department has, in the wake of September 11, announced plans to install over 1000 outdoor fixed cameras to monitor its streets. Face recognition software continually combs through the data in order to match file photos of suspects. Canada is not immune to this Orwellian trend and even the small town of Kelowna, British Columbia boasts several R.C.M.P. cameras continuously surveying and recording images of the people on its streets. This surveillance is the subject of a recent court challenge by the Canadian Information and Privacy Commissioner under section 8 of the Charter, which prohibits unreasonable searches and seizures.
Is it tenable then to state that the Internet user should have no more of an expectation of privacy than a character in the George Orwell novel, 1984? Most Internet users, and particularly the Secessionist Libertarians, would all give a resounding "No!" The only way that one could reconcile Zatz’s view with that of this majority would be to distinguish between the zone of privacy for purposes of not being exposed to unwanted information, and the zone for purposes of anonymity and confidentiality. Does the law recognize such a distinction? It appears that it may do so. For example, the draft Washington, D.C. by-law to implement the enormous expansion of its police video surveillance contains a prohibition against camera operators zooming in on or taking note of the content of leaflets distributed on the street. Presumably, this is because the D.C. lawyers have advised that the individual distributing the leaflet has a reasonable expectation of privacy as to the content of the leaflet, but the individual to whom the leaflet is targeted does not have a sufficient expectation of privacy to enable him to be shielded from the unwanted information. There is no reason to believe that the result would be any different in Canada, since the relevant U.S. and Canadian constitutional provisions are quite similar. Therefore, we see that Zatz’s countering of any forced listening objections to the cydewalk proposal does not jeopardize any arguments that would bolster confidentiality and anonymity on the Internet.
The issue of surveillance has some application in the labour relations context as well. Under the cydewalk proposal, unions have the potential ability to collect and track information about persons visiting the employer’s site through devices such as the "cookie". When an individual is diverted automatically from the employer’s web site to the union’s site, the unions’ server will track that individual’s browser responses, e.g. whether the individual clicked on various items in the union site and, if so, for how long, or whether he returned immediately to the employer’s site (the electronic equivalent of crossing the line), and the information will be deposited on the individual’s server in the form of a "cookie". The union’s server can retrieve the information in the "cookie" from the individual’s server at a later date when the individual visits the union server a subsequent time, (which would occur, for example, in the case of a telecommuter who is required to visit the employer’s site regularly to "report" to work.) However, although the information would include the individual’s Internet protocol address (IP), current technology would not enable the union to link the IP to the individual’s name, telephone number and physical address. Furthermore, "cookies" can be easily disabled by the individual.
The union may argue that it should have access to the information about the identities of the electronic line-crossers, in order for its members to properly exercise their right of free speech. However, such information would probably be used by the union to intimidate persons and deter them from visiting the employer’s web site during the strike, particularly if they were members of the union who wished to continue working from home during the strike and needed to visit the employer’s web site in order to do so. Although freedom of speech extends to protect even intimidating speech (as long as threats of violence or other criminal acts are not involved), it would not extend to requiring disclosure of the identities of the recipients of the speech for the purpose of facilitating intimidation of them.
There are options in the Zatz cydewalk proposal which would provide less intrusive methods than automatically diverting persons from the employer’s to the union’s site, e.g. placing a "pop-up" message on the employer’s web site which would allow a person to choose between entering the employer’s site or visiting the union’s site. Under this method, even with possible future technology that would allow linking of IP and physical addresses, the union would not be able to determine the identities of electronic line crossers, since they would never get onto the union site at all, but the employer, conversely would be able to determine the identity of individuals who chose to honour the line. Would this be as problematic as permitting the union to identify electronic line crossers? Probably not, since, amongst its employees, the employer can simply deduce who is honouring the line by subtracting the subset of employees who actually come in to do e-work during the strike. However, the union, on the other hand, would find it more difficult to identify electronic line crossers without this form of electronic spying, since it would have no quick way of identifying the identities of all the members who crossed the e-line. Therefore, the "pop-up" option would seem to involve a lesser temptation for misusing the system for any electronic spying that may be available in the future through linking of IP and physical addresses.
The privacy of the identity of electronic line-crossers would also be supported by the Canada Personal Information Protection and Electronic Documents Act (PIPEDA) known as Bill C-6, which took effect on January 1, 2001. PIPEDA was designed to address privacy problems stemming in part from the rapid flow of information through the Internet. The act applies to federal works or undertakings, e.g. airlines, broadcasting etc., and also provincially regulated entities if they transmit the data outside of the province. After a three-year period, it will apply to all organizations, public or private, federal or provincial, collecting personal information in a province unless the province has adopted provisions similar to the federal law and the federal government decides to exempt them.
PIPEDA incorporates the Canadian Standards Association (CSA) Model Code for the Protection of Personal Information as Schedule 1 of the act. This Model Code is based on the O.E.C.D.’s 1981 Model Code for Privacy. Section 4.3 of the CSA Code requires the individual’s knowledge and consent to collection, use and disclosure of personal information, except where inappropriate, e.g. law enforcement etc. In cases of a direct marketer seeking to purchase a mailing list from another organization that compiled the list, the consent must be obtained from the individuals by the compiling organization and not the direct marketer. Consent must be meaningful in the sense that the individual must be able to reasonably understand at the time of giving consent what are the intended purposes of collection of the information.
However, there are some potential loopholes and flaws in the consent provisions. For example, section 4.3.6 of the Code states: "An organization should generally seek express consent when the information is likely to be considered sensitive. Implied consent would generally be appropriate when the information is less sensitive." To compound the uncertainty introduced by section 4.3.6, section 5(2) of the act states that the word "should" in the Code indicates a recommendation and does not impose an obligation. Furthermore, section 4.3.7 of the Code provides that consent can be given by means of the so-called "negative option" method, whereby an individual who does not check off a box stating that he does not wish to consent to the collection, will be deemed to have consented. If this box is buried at the bottom of a web page, how could this constitute meaningful consent?
The CSA Code provides a lesser degree of protection than the E.U. Data Privacy Directive which, for example, in Article 8 gives a special status standard of protection to certain types of information, e.g. health, political opinions, religious or philosophical beliefs, and trade union membership unless there is "explicit" consent to collection (which presumably would rule out the negative option method). This type of extra protection would have certainly been beneficial in the CSA Code, which, due to its loopholes, may not prevent unions from attempting to negotiate picketing protocols with employers that would deem the employer to have obtained implied consent from line-crossers as a term or condition of employment during the strike. The union may even wish to extend this beyond the employees of the employer involved in the labour dispute. For example, the union may want to identify employees of the employer who are members of other unions, who do not honour the line, and to inform those other unions of the behaviour of their members for possible disciplinary behaviour. Of course, most employers would have sufficient negotiating strength to reject such demands by the union, but the occasional employer may submit to them, either through lack of bargaining power or a failure to properly appreciate the ramifications. In the United States, the privacy protection tends to be even weaker than that of the CSA Code, due to the fragmentation of privacy legislation amongst the states, and also due to concerns about terrorism in the wake of September 11. However, in the United States, unions tend to be less powerful and have less overall penetration than in Canada.
Therefore, we see that privacy concerns should be an important factor amongst the various technological options which Zatz sets out for the cydewalk proposal, e.g. pop-up window, flashing icon, automatic redirection to protester’s site etc., especially if future technological developments will allow linking of IP and physical addresses. Even if the least intrusive option is chosen, any legislation implementing the cydewalk should likely have special privacy provisions of its own to overcome the weaknesses in the CSA Model Code, as well as the fragmentation of U.S. privacy protections.
The Open Source Regulators would likely support the principles behind the cydewalk proposal, due to their emphasis on transparency and the free exchange of information, but they would probably suggest that the cydewalk’s objectives be accomplished in somewhat different ways. This group would favour open source modularity in the cydewalk code, which essentially means that owners of web sites could choose whether or not they wanted to have a cydewalk in front of their sites. In addition, this group would suggest that site owners be strongly encouraged to choose the cydewalk option, for example by means of government action affecting Internet architecture (e.g. faster access times to sites with a cydewalk), social norms (e.g. education campaigns fostering public awareness of the social benefits of the cydewalk) or the market (e.g. government using its purchasing power to create a de facto cydewalk software standard).
In terms of the trends in the constitutional and labour relations jurisprudence, there are a number of issues that are raised by the Open Source Regulators’ version of the cydewalk proposal. Dealing first with the fundamental idea that web site owners should have a choice as to whether or not they want a cydewalk, we see that this disposes of any arguments that could be made that the creation of the cydewalk involves an expropriation of private property. Instead, the web site owner would be voluntarily ceding an interest in a portion of his web site to the state for public purposes. What, however, would this do to the issue of forced expression? On first blush, it seems that the voluntary cydewalk would dispose of this concern also, on the basis that the web site owner was not forced to cede the cydewalk to the state in the first place. However, on looking at the issue more closely, it becomes evident that an argument could be made that there is a possibility of the forced speech issue succeeding due to the danger of misattribution of the views of the protesters to the web site owner. Suppose that a web site owner voluntarily agrees to have a cydewalk in front of his or her site. Everything is going fine, until one day when a protester appears on the cydewalk expressing a point of view that is particularly objectionable to the web site owner. The web site owner can revoke his agreement concerning the cydewalk at any time, since the agreement is not for a fixed period, but does not wish to do so because he would lose certain state provided incentives for having a cydewalk. He consults his lawyer and seizes upon the idea of launching a Charter challenge alleging forced expression.
Would this have a reasonable chance of succeeding? This really depends on the risk of misattribution of the protester’s views to the web site owner. If there is a serious risk of this, then the web site owner would probably be able to make a good argument that there is forced expression which violates his or her rights under section 2(b) of the Charter and which is not justified under section 1. The reality of the matter is that, in a world where the cydewalk is voluntary, there is the danger that the web site owner’s continued agreement to the presence of the cydewalk would, despite any suitably worded disclaimers, create the danger that the public may associate the protester’s views with the web site owner. Professor Richard Moon refers to this phenomenon in commenting on the U.S. Supreme Court decision Wooley which held that Jehovah’s witnesses had a constitutional right to cover up the "Live Free or Die" slogan on their New Hampshire license plate. He perceptively comments:
Only if car owners were given a right, which was sometimes exercised, to alter the message on their license plates, might we see the message on a particular plate as reflecting the owner’s personal views. Failure to cover up the message, then, might be seen as an acceptance or affirmation of it.
It is true that, in the license plate example, the choice is to cover up or not cover up one message, a state sponsored message at that, whereas in the cydewalk example, the right is not to cover up a particular message but rather the right to opt out of any messages at all. Furthermore, the license plate is on personal property, whereas the cydewalk is often in front of corporate property. Should this make a difference to the misattribution issue? It may, but only to a degree, since the inference would remain that if the web site owner can cancel the cydewalk at any time, then the owner must agree, or at least not strongly disagree, with the viewpoints which continue to be expressed upon it.
Of course, as we have noted earlier, in the context of employer-union disputes, the danger of misattribution is much lower, due to the public’s general knowledge of the traditional adversarial roles of the employer and the union. However, this does not take away from the fact that the danger of misattribution would be there for other protesters in some cases (e.g. the case of an extremely anti-union group speaking on the cydewalk in front of a moderate employer). As we have also noted earlier, it is not constitutionally permissible to distinguish between speech by unions and other groups in terms of allowing access to the cydewalk, so the danger of misattribution could not minimized by giving preference or exclusive access to union speech.
There would also be the problem of employers opting out of the cydewalk during a labour dispute, and then opting back in after a collective agreement is reached. This would, of course, probably create good grounds for an unfair labour practice complaint of anti-union animus, under labour relations legislation. It would probably also lead to a constitutional challenge that the government’s design of the cydewalk, in the sense of making it voluntary, facilitated the employer’s anti-union actions and the denial of the union’s speech opportunities. Here, the union might rely on the Dunmore decision of the Supreme Court of Canada, where it was held that the legislative exclusion of agricultural workers from the Labour Relations Act violated the claimants’ right of freedom of association under section 2(d) of the Charter, and was not justified under section 1, even though the exclusion did not itself deny the right of association but merely facilitated the denial of the right of association by private actors.
The Open Source Regulators’ modular view of the cydewalk would also include strongly encouraging web site owners to opt for the cydewalk, through various forms of government action, including faster internet access to sites with a cydewalk, public awareness campaigns, or using the government’s purchasing power to flood the market with a de facto cydewalk software standard. The first example of the faster access times incentive is an illustration of the use of architecture to modify behaviour. Lessig uses some real word analogies ranging from the use of speed bumps to deter urban speeders to Napoleon III’s widening and straightening of the streets in Paris so as to make it harder for revolutionaries to block off access to key parts of the city. What sort of constitutional and labour relations issues would be entailed by the architectural model of behaviour modification in the context of access?
One issue that would arise is due to the fact that, in the voluntary cydewalk system, opting out of the system would tend to be the norm amongst certain types of web sites, i.e. sites with extreme views of one form or another. These sites would then suffer the disadvantage that their visitors would have slower access than other more mainstream sites which did opt for the cydewalk. A site hosted by a political party with radical views, e.g. The Communist Party, could then conceivably launch a challenge under section 15 of the Charter, claiming discrimination on the basis of political beliefs. They would probably cite the recent decision of the Prince Edward Island Supreme Court which held that political beliefs should be considered as a prohibited ground of discrimination under section 15 of the Charter. However, an effective response would likely be that its political beliefs do not compel it to withdraw from the cydewalk. The Communist Party would not be discriminated against on the grounds of a particular set of political beliefs, but rather on the basis of its general inability to tolerate the views of others, which is common to many different kinds of organizations outside the mainstream.
What if the pattern of opting out of the cydewalk included not only extremist sites, but also unionised employers due to their fear that the cydewalk would be used to extend picketing opportunities during a labour dispute? We have discussed already the challenges that unions might launch to this type of scenario, but what about the challenges that an employer might launch in that its visitors had to endure slower access to its site as a result of the exercise of its choice to opt out of the cydewalk system? Even if the political beliefs ground of discrimination could somehow be stretched to include a general dislike of unions, such a challenge would be effectively met by replying that the employer’s decision to opt out was its own doing. The Dunmore principle that state action can be found where it sets the stage for private actors to adversely affect Charter rights would not include the complainant itself amongst those private actors.
The other aspects of the Open Source Regulators’ suggestions would include educational campaigns to encourage taking up of the cydewalk option, and attempt to influence the market through the government’s use of its large software purchasing power. The educational campaign is certainly a benign method and not controversial, but probably not very effective. Equally and perhaps even more ineffective, in Canada at least, would be the government as model purchaser option, since the purchasing power of the Canadian government is miniscule compared to that of the U.S. government, and it is in the U.S. primarily that the software is made.
The proponents of the Telecosmic Futurists school of thought adhere to the belief that the telecosm, which is the total communications universe made up of telephone, television, radio, and the Internet, should be free of regulation by government agencies and should instead be governed by the courts, the common law and the marketplace. They are emboldened by emerging ultra-wideband broadcasting technology (UWB), new intelligent digital radios, continuous improvements in the capacity of fibre optic cable, and mesh networks which will free government agencies from the need to parcel out bandwith allocations to deserving mega-corporations. This school of thought, represented by lawyer and former M.I.T. professor, Peter Huber, and by economist and Internet guru George Gilder, believes that technology and the market may create public spaces in the telecosm, including the Internet. Peter Huber states:
But the one certain thing is that true wisdom in matters this complex does not emerge from centralized commissions, nor even from visionary pundits. Wisdom emerges from markets. Markets find ways of reassembling private pieces into public spaces when that is the most profitable thing to do. They may take more time than an omniscient central authority, but finding omniscient central authority takes even longer. For now, the thing to do is to get the spectrum out of government hands, however it can be done, and leave it to the market to re-create the public commons. It will, if the economics are there.
Certainly, in the real world, an individual’s private property is worth a great deal more if it is adjacent to public property, e.g. streets, roads, highways etc., so that there is a means of access that individual’s property. Thus, the marketplace originally created public spaces that subsequently received the blessing and protection of the courts and governmental authorities which protected them from private encroachment. However, at the time when public spaces were first created by the marketplace, society was much less complex and diverse. In the digital world of the 21st century, can the marketplace really be trusted to create public spaces, particularly a cydewalk? Or is this new world a McLuhanesque "Global Village" which is just in the beginning stages of creating its own unique public spaces, as primitive villages did thousands of years ago?
There are several economic advantages to the parties that would accrue from a joint employer-union agreement to provide a cydewalk for purposes of labour expression. From the employer’s point of view, such an agreement could enable it to obtain concessions from the union concerning picketing in physical space, e.g. restrictions on numbers of pickets, limits on allowable delay, bans on pickets in front of certain locations. The employer would to the extent that picketing in physical space is regulated, save money on private investigators hired to gather evidence and lawyers retained to obtain court injunctions, as well as saving work time otherwise lost by non-striking employees whose access is interfered with by the union. It would also discourage electronic civil disobedience on the part of the union (hacking into the employer’s site) and would also eliminate the need for legal, but fundamentally misleading practices by the union such as meta-tag copying. A further benefit to the employer would be that it might be able to negotiate a reciprocal right for the employer to convey its message to employees visiting the union’s web site during the work stoppage, thereby correcting any misinformation that the union had been disseminating.
From the union’s point of view, picketing on the cydewalk would allow it to express its views to all persons visiting the employer’s web site, rain or shine, on a 24/7 basis with a minimal investment of money and manpower. Also, unions could, as part of such an agreement, negotiate provisions which would prevent the employer from transferring the bulk of its operations into cyberspace during the work stoppage so as to frustrate the union’s opportunities to convey its message to persons doing business with or working for the employer. Finally, a union could maximize the signal effect of the cydewalk if it could negotiate with the employer a provision that would allow the union to determine the identities of members who crossed the electronic picket line (through a provision requiring the employer to make the employees’ consent to disclosure a term or condition of employment). Obviously, the union would have to be in a very strong bargaining position to obtain such a provision, but it should be borne in mind that the more the cydewalk could be made to replicate its physical equivalent, the stronger incentive there would be for the union to agree to make concessions that relate to the real world in exchange for the right to picket on the cydewalk.
From the point of view of the mutual interests of both parties, an agreement for picketing on the cydewalk would be beneficial in that it may tend to tone down the rhetoric on both sides and mitigate the divergence of opinion through encouraging cross-fertilization of ideas. For example, a union may hesitate about putting an inflammatory personal attack against a member of the employer’s bargaining team on its web site if it knows that this will almost immediately produce an electronic response from the employer that would be displayed on the bottom one inch of the screen of all persons visiting the union’s site. Similarly, the employer, for example, may think twice about putting an ill-conceived "spin" on its latest offer when it knows that this will almost instantly evoke a response from the union which will be seen by all of the persons visiting the employer’s site. Not only could the cydewalk therefore tone down the rhetoric and cool down the passions that inflame the parties and interfere with meaningful collective bargaining, but from an efficiency perspective, the cydewalk could result in a faster and more effective way of communicating information to the general public about the work stoppage, than would periodic news conferences, thereby preventing "bargaining through the media", which takes the parties away from the table.
There are some constitutional and labour relations issues associated with letting the marketplace govern the cydewalk, where the employer is in the public sector, especially if the employer grants access to the cydewalk to the union, but denies it to other groups. As we have seen, this argument would not arise in the context of a cydewalk in front of the employer’s intranet, but may arise where the cydewalk is in front of the public sector employer’s Internet. This Charter argument would be based on the statement in the Pepsi decision that, for freedom of expression purposes, it is impermissible to distinguish between labour and non-labour speech.
There is also the issue of the enforceability of cydewalk agreements. These agreements would be similar to a picketing protocol that is worked out between the police, the employer and the union during a strike. A picketing protocol is not generally enforceable unless it forms part of a court order through a consent decree. This is because under labour relations legislation in Ontario and most jurisdictions in North America there can only be one collective agreement between the employer and the union. The collective agreement is defined as including all ancillary agreements including protocols, and the collective agreement must have expired before the right to strike arises. Therefore, during a strike, a protocol, such as a picketing protocol, is not recognised as a legally enforceable agreement. However, this lack of enforceability may operate to the advantage of "early adopters" of cydewalk agreements, since it would in effect make the agreements terminable at will by either party, and would encourage responsible use of the cydewalk.
After the "early adopters" stage, parties may be inclined to ask the court to incorporate the agreements into consent orders, so that the parties have the certainty associated with enforceability. This, in turn, may result in the courts starting to look at the cydewalk as an established alternative forum in terms of deciding what kinds of legislative limits on picketing in physical space may be permissible under the Charter, particularly concerning secondary picketing. In this regard, it is noteworthy that the Supreme Court of Canada stated in the Pepsi case:
The legislatures too may play a role. Clarification of the status of picketing at common law should not be viewed as a restriction on legislative intervention. Rather it should be seen merely as a tool to assist the courts where federal and provincial laws remain silent. As mentioned earlier, different circumstances in different parts of the country may call for specially tailored legislative regimes. Legislatures must respect the Charter value of free expression and be prepared to justify limiting it. But subject to this broad constraint, they remain free to develop their own policies governing secondary picketing and to substitute a different balance than the one struck in this case.
Would this possibility dissuade unions from entering into such agreements in the first place? It should not, since any modifications made by the legislatures following Pepsi would still very likely be limited to clarifying and codifying the common law torts, which should not be very controversial.
The Telecosmic Futurists also strongly believe in the role of the common law in regulating the Internet. What possibility is there that the cydewalk may evolve out of the development of the common law? Between 1920 and 1927, when radio stations first began to broadcast in the U.S., and there was no Federal Communications Commission, disputes between broadcasters were resolved by the common law, using as a reference point the common law of real property. Although this was a short-lived phenomenon due to the creation of the F.C.C. in 1927, it is possible that this could re-occur in the 21st century in the context of creating public spaces on the Internet. There is a whole body of common law jurisprudence relating to common law real property rights and public spaces, e.g. the cases on rights of way over private land to get to a public beach. Using the example of a union which hacked into an employer’s web site to display its message there, the union could argue that there is a common law property right on the Internet equivalent to the right in the physical world to have access through private property to get to public property, e.g. a beach. This would be a novel argument, since as noted earlier, although the doctrine of trespass to chattels has been found to apply in an Internet context, the doctrine of trespass to real property, or anything analogous to it, has not yet been found to apply in such a context. Also, there is the problem of defining what part of the space is public and what part private. If the union must penetrate the employer’s property line in order to reach the "cyber commons" where exactly is that commons located? In the case of an employer that operates a retail operation on the Internet, e.g. Amazon.com, it may be difficult to point to a part of the employer’s site as a cyber-commons. However, what about the on-line shopping mall? Does that contain a cyber-commons? Although there is the obiter dicta by Madame Justice McLachlin in the Dorval airport case that section 2(b) of the Charter does not grant the right to freedom of expression on private property, this may be offset by the finding in the more recent KMart decision that section 2(b) allows for peaceful leafleting by trade unions in a shopping mall while the stores are open on the basis that a shopping mall is the equivalent of the public market or main street of the past. Although the KMart case involved British Columbia labour legislation, it is possible that the Charter analysis in it could be applied to develop and interpret common law property rights on the Internet, (as Mr. Justice Sigurdson did in the BCCA decision for purposes of interpreting the common law doctrine of passing off).
However, it is likely that such a view of the Charter would be at least as controversial and unorthodox as Sigurdson’s use of the Charter to water down intellectual property rights in the BCCA case. Although the court in Pepsi indicated that the Charter should guide the development of the common law, it also indicated that the common law should be developed and changed in an incremental manner. Interpreting the common law to include common law property rights on the Internet would be a quantum leap and not incremental change at this point. However, given the pace of "Internet time", what the future may bring in even a few short years is really anyone’s guess.
If a representative of each of the four schools of thought had to view the cydewalk with an optical device, what kind of device would each person choose?
The Secessionist Libertarian would choose rose-coloured glasses, due to his misplaced idealism concerning the possibility of Internet self-government, perhaps with a pair of blinkers on the sides preventing him from seeing the convergence between the real and virtual worlds. Through these glasses he would see hyperlinks leading off to other strange sites like the signpost in the television series "M*A*S*H" pointing to numerous far-off destinations which no one ever visited.
The Anti-Anarchist would likely see the cydewalk through a pair of bi-focals. Initially, as she tilted her nose up in disgust at the novel idea, she would see only a chaotic blur. As she levels her gaze, perhaps enticed by the notion of co-opting the Hacktivists, the image would become more focussed and she would see the trade unions struggling to make themselves clearly seen in the "data smog" of other protesters.
The Open Source Regulator would view the cydewalk through a prism that would divide it up into the full spectrum. At the ends of the spectrum, he would see the cydewalk fading into nothingness where extremist sites opted out of it.
In the middle of the spectrum, he would clearly see that many sites had a cydewalk but there would be some blurry patches where the primary colours overlap, showing parts of the cydewalk where misattribution occurred.
The Telecosmic Futurist would view the cydewalk with a kaleidoscope. She would see an infinite number of ever changing patterns based on mutual agreements between employers and trade unions. These kaleidoscopic patterns would obtain some structure from labour relations and constitutional law and jurisprudence. Eventually, these patterns would include agreements between other types of adversaries such as oil companies and Greenpeace, and the World Trade Organization and anti-globalization advocates. Thus the labour paradigm of the negotiated agreement would become an excellent method of seeing the vast potential of the cydewalk.
This paper started with two news items which are both related to the tragic events of September 11, 2001 – the first item involved the failed attempt at electronic picketing of Argenbright Security and the other involved the al-Qaeda web site. I hope that I have shown how the mutual agreement paradigm of the cydewalk could remedy dysfunctional employer-union relationships such as in the first item. However, this paradigm would obviously not remedy problems of extremist and terrorist organizations with web sites as in the second item. The al-Qaeda is not likely to enter into an agreement with the U.S. to allow electronic picketing of each other’s sites, no matter how much the Internet becomes de-balkanized.
What, then, about the idea of supplementing the mutual agreement approach with a legislated "floor" in the same manner that employment standards legislation is used to provide basic requirements which can be built on during collective agreement negotiations? If this legislated floor were browser-based, it could cover extremist sites which are based offshore, e.g. al-Qaeda, or in the U.S., e.g. Zundel. However, such a solution would have several problems:
1. It would stifle the development of the common law property rights for Internet access. Once established, the common law tends to be more durable and flexible in the long run than statute law, so the former is preferable.
2. It would interfere with the development of mutual agreements. It is easier to negotiate a pay increase of $X over the minimum wage than it is to provide special electronic access by the union over and above that provided by statute given that, with some exceptions such as hate sites, the legislation would have to avoid content or purposive discrimination. Even if the employer could voluntarily give the union sole access to a special section of the cydewalk, there is the problem that dividing up the public’s attention is a distributive problem; the public only has so much tolerance for pickets and the visitor is unlikely to want to encounter two levels of picketing – one on the legislated cydewalk and the other on the mutually agreed one.
3. It would antagonise the party owning the picketed site and would tend to have a chilling effect on that party’s expressing criticism or views that could provoke a counter-response on the cydewalk.
4. Although extremist sites which involve hate or terrorist activities could be blocked from picketing on the cydewalk in Canada for the same reasons that the Zundel site itself was banned from Canadian servers, Nazi sites such as Zundel’s could have access to the cydewalk in the U.S., where they could counter-picket mainstream organizations, thereby multiplying their proselytising opportunities.
The problem of extremist sites is best dealt with by decreasing the level of balkanization of opinion on the Internet as a whole so as to create a greater
atmosphere of tolerance in society and thereby immunize it against the disease of hatred. The first step in this decrease in balkanization would be to explore mutual agreements concerning the cydewalk between trade unions and employers, and the next step would be to use the labour relations model as a paradigm for other types of adversaries.
