|Multi-Stakeholder Public Policy Governance and its Application to the Internet Governance Forum|
The second set of issues that particularly confront international and transnational lawmakers, especially those engaged in public policy governance of the Internet, concerns the intersection of the international and domestic legal systems. These issues include the limited range of mechanisms available for the enforcement of international law, given that the international system lacks a conventional police force or judiciary. There are also difficulties in determining the appropriate domestic law that should govern a particular factual circumstance and in enforcing that law internationally, as well as difficulties in containing the extraterritorial effects of either domestic law or the conduct to which such law is directed. These problems will be dealt with in turn.
There are few formal mechanisms for the enforcement of international law. Whilst there is the International Court of Justice, the main limitation of its jurisdiction is that it requires the consent of each party in order to be enlivened. Furthermore whilst judgments of the ICJ against states may be enforced by the Security Council of the UN, the veto powers of each of the permanent members make it unlikely that such a judgment would ever be enforced against them.
Austin even concluded in 1832 that international law was not law at all because it lacked a judiciary or an executive to enforce it. Yet it has nevertheless been observed that “almost all nations observe almost all principles of international law and almost all of their obligations almost all of the time,” prompting research by international relations scholars (particularly institutionalists) as to why this should be so. It is now more widely accepted that the lack of a conventional mechanism of enforcement for international law is not fatal to its status as law, and that in any case there are a number of informal institutions by which it is enforced.
First, states can exercise self-help measures against other states believed to be in breach of their international law obligations. An obvious example is found in the United States military offensive against Iraq in the 2003 Gulf War. However as this action was taken in defiance of UN Security Council resolutions, it also illustrates the lack of legitimacy that such unilateral measures may carry as against actions authorised by the United Nations.
Second, particular international organizations have created frameworks within which international law can be enforced through methods other than reference to the ICJ or the Security Council. The best example is the WTO, the Dispute Settlement Body of which can impose a number of innovative penalties on non-compliant states including ordering the payment of compensation to other affected states. Of course, if the payment is not made then the WTO falls back on its ultimate self-enforcing remedy of trade sanctions.
Third and more fundamentally, there are forces which drive states towards voluntary compliance with international law, such as “concerns for reputation, reciprocity, perceptions of legitimacy, trust, and the like, which do not rest on coercion or sanctions.” Indeed, regime theory is built on the assumption that states will only establish an international regime (which includes a regime of law) if they consider it is in their long term best interests to do so.
Domestic law (including the rules of conflict of laws) tends to be premised on the assumption that events can be located territorially. Thus for example the postal acceptance rule provides that a contract is formed in the “place” where the acceptance is communicated to the offeror. But what if that place is the Internet? One of the earliest preoccupations of academics writing on cyberspace law was in tracing the implications of jurisdictional problems such as these where the location of a legally relevant event is obscure.
The difficulties in assigning online conduct to a jurisdiction have however sometimes been overstated, and most are soluble by the use of general principles of conflict of laws. Difficulty in finding a state to assume jurisdiction over online activity is unusual; more often there is more than one that might do so. But that is rarely a problem in a contractual context, as it is enough that one jurisdiction can be found to hear the case and generally irrelevant that more than one might.
Similarly in tort and criminal cases, the law to be applied is normally the lex loci delicti commissi (broadly that of the place where the act was committed), and even if there is more than one such place, then principles of forum non conveniens in civil matters and of double jeopardy in criminal matters will normally prevent the case from being tried twice before different tribunals.
True, the prospect of a multiplicity of states extending their jurisdiction over a single dispute is productive of some uncertainty, but in contractual cases that uncertainty can easily be resolved by the parties selecting their desired jurisdiction for the agreement pursuant to article 3(1) of the Rome Convention (and if they so wish having it enforced in some other jurisdiction of their choosing pursuant to the New York Convention). These choices can in general be elected online even more easily than they can offline, for example using so-called click-wrap agreements hosted on one party’s Web site (subject of course to any compulsory consumer protection provisions applicable in the jurisdiction of either party, such as article 12(2) of the EU Directive on Distance Contracts).
Where no such election has been or can be made contractually, there is still some provision through a patchwork of bilateral and multilateral instruments for the enforcement of court judgments between various jurisdictions, including the states of the EU, and between Australia and some 36 states and Canadian provinces pursuant to the Foreign Judgments Act 1991 (Cth).
As for criminal matters, the ability of a state to enforce its laws against a defendant in another jurisdiction will depend on the existence of an agreement between the two nations in question for the extradition of criminal defendants. The procedure by which Australia makes and receives extradition requests is set out in the Extradition Act 1988 (Cth), and the countries from which it receives requests are provided in regulations to that Act. This includes most of the same countries as included in the regulations to the Foreign Judgments Act, but with the notable addition of the United States.
A more serious issue arises in cases where the problem is not so much in locating an event in a particular jurisdiction, but in dealing with its effects that cross multiple jurisdictions. The effects may be those of wrongful conduct, such as tortious, criminal or intellectual property law infringements, or they may be those of the laws themselves, whereby an attempt by one jurisdiction to regulate conduct on the Internet impacts too widely against other sovereign jurisdictions. Such effects are known as “spillover effects.”
Spillover effects are seen in one of two forms. First, the least stringent regulations applicable to any Internet-connected jurisdiction may be those that prevail in practice; a “lowest common denominator” effect. This can be seen in the case of spam regulation, where the greatest amount of spam is sent from those countries with the weakest anti-spam laws, often on behalf of advertisers from elsewhere in the world. It is also a hallmark of content regulation that content providers will relocate their hosting operations to less highly-regulated jurisdictions offshore, as in the examples previously given of Australia’s Interactive Gambling Act and Broadcasting Services Act.
The second and converse effect of spillover is that online behaviour may be required to conform to the most stringent regulations in force worldwide, the potential for which is illustrated by the Gutnick defamation case and the Yahoo! Nazi memorabilia case both previously examined above, as well as by the effect of the EU Data Protection Directive which has, for example, led Microsoft Corporation to change the way it manages user data for its dot-NET Passport service, not only within Europe, but worldwide.
Without overstating the point, the protection of the First Amendment of the US Constitution, in particular, can result in very different regulatory regimes being applied inside and outside the United States in cases such as these. Taking the example of libel, it is for example necessary for American plaintiffs to demonstrate that a false statement published in the press was written with actual malice. Such differences as these may result in the more tightly regulated country’s laws “spilling over” upon those in the more loosely regulated jurisdiction.
In either case, whether it is conduct or law that spills over into less regulated jurisdictions, the outcome is that at least one sovereign nation’s law is effectively being trumped by another’s.
Goldsmith argues that the effects of spillover have again been overstated, in that it remains possible for indirect regulation to be exercised over much conduct that spills over jurisdictions, for example by criminalising the use of off-shore hosting facilities, or by controlling intermediaries within the local jurisdiction, such as banks, ISPs and domain registrars. A recent example of the latter is found in the withdrawal of credit card processing facilities by the Visa International network to AllofMP3.com, a Russian music download service that allegedly contravened United States (but arguably not Russian) law.
The spillover of law on the other hand, for Goldsmith, is unexceptional, as it is well within the competency of states to pass laws with extraterritorial effect in order to regulate effects within their own jurisdictions.
Post on the other hand, whilst not denying this as a general proposition, contends that there are sufficient differences between the way in which such laws impact upon online and off-line conduct that differential treatment of the former is warranted. One difference that he notes is scale, such as the way in which the automatic distribution of millions of Usenet messages through thousands of servers worldwide could lead to unreasonable liability in the event that a message so distributed infringed copyright. This, and the extent to which, in cyberspace, “virtually all events and transactions have border-crossing effects,” along with the fact that domestic regulation of same will almost always impact upon those who are not subject to that state’s sovereignty, point Post to the need for reforms to mitigate these effects.
And, indeed, there have been a number of efforts to do so, both technical and legal. Technical reforms aimed at overcoming the spillover effects of online conduct include filtering and geolocation. These work in approximately opposite ways. Filtering prevents content that is prohibited in a particular jurisdiction (such as unwanted speech or offensive images) from entering that jurisdiction via the Internet. In Australia, for example, both the spam and content regulation legislative regimes make use of co-regulatory codes drafted by industry groups and registered with the ACMA, pursuant to which ISPs are required to make appropriate filtering products or services available to their users.
Geolocation on the other hand, prevents content that is made accessible on the Internet, from being accessed from a particular jurisdiction. For this purpose the IETF has proposed a DNS-based method for determining the geographical location of Internet addresses, though more ad hoc geolocation services, both commercial and non-commercial, are currently more widely used.
Neither filtering nor geolocation technologies are foolproof, however, as both measures run counter to the Internet’s decentralised and cosmopolitan architecture. A 2002 study of six popular filtering products found that depending on how they were configured they would block between 89% and 91% of pornographic Web sites, but also between 33% and 91% of safe sex information sites. Similarly, the highest accuracy claimed for geolocation technology admits an approximate 10% error rate at the city level. These figures do not take into account users’ ability to deliberately attempt to evade the technologies in question by means such as encrypted tunnels or proxies.
As for the options to mitigate the effects of spillover through law reform, the first example to emerge was the development of a consistent doctrine in United States case law to determine whether a Web site, available internationally, should fall subject to local regulation, in the early case of Zippo Manufacturing v Zippo Dot Com, Inc.
The test established by that case described a sliding scale from “passive” Web sites, which were merely information-based, with minimal interaction with the user, through to “active” Web sites, of which the predominant purpose was to conclude business with the user. In between were the intermediate category of “interactive” Web sites. It was adjudged that the jurisdiction of the user in question could assert authority over active sites but not passive ones, whereas for interactive sites the level of interactivity and the commercial nature of the information exchanged would determine whether it would be treated as one or the other of the other categories.
Whether or not this test is an adequate one, it applies on a domestic level only. Since spillover is a transnational issue, it calls out for a uniform transnational solution. There are three ways in which this could be approached. The first and least radical would be to unifiy domestic choice of law rules, which was the intent of negotiations over a proposed Hague Convention on Jurisdiction and Foreign Judgments in Civil and Commercial Matters, which however fell through largely due to objections from the United States.
The next most radical is the imposition of a sui generis Internet law regime, which could still in appropriate cases be applied in domestic tribunals. A sui generis regime could take shape through the unification of Internet-related law through formal international agreements as has begun to happen with the UNCITRAL model law on Economic Commerce, or simply through the continued development of private transnational law, in what could become like the process by which the common law developed.
A final and even more radical solution would be the removal of Internet legal issues into a specialised Internet jurisdiction. This was the early approach of scholars like Johnson and Post, who reasoned that due to the inevitable jurisdictional problems associated with the application of domestic laws to the Internet, cyberspace should be left apart from territorial regulation to develop its own self-regulatory structures based on shared norms.
Whilst to date even the least radical of these alternatives has proved to be a dream, elements of all of them can be seen in development. Whilst we do not have the Hague Convention on Jurisdiction, we do have the Hague Convention on Choice of Law Agreements. We may not have a sui generis legal regime covering all Internet-related issues, but we do have the UNCITRAL model law. And whilst the world lacks a legal system exercising specialised Internet jurisdiction, there is such a system for the resolution of domain name disputes, the UDRP. These are the seeds from which bodies such as the IGF may be able to grow a more comprehensive transnational legal regime for the Internet in the future.
Statute of the International Court of Justice, 26 Jun 1945, 1975 ATS No 50 (entry into force for Australia 1 Nov 1945), article 36.
Slaughter observes that it is not all states, but more particularly liberal states who tend to comply with international law: Slaughter, Annie-Marie, International Law in a World of Liberal States (1995). But see contra Koh, Harold H, Transnational Legal Process (1996) and Koh, Harold H, Why Do Nations Obey International Law (1997).
Adams v Linsell (1818) 106 ER 250
Convention on the Recognition and Enforcement of Foreign Arbitral Awards, 10 Jun 1958, 1975 ATS No 25 (entry into force for Australia 24 Jun 1975) (New York Convention), an UNCITRAL treaty which provides for the enforcement of both agreements to arbitrate and the resulting arbitral awards in the state of each signatory to the Convention (presently numbering over 130). Compare also the Hague Convention on Choice of Law Agreements, concluded on 30 June 2005 (see http://www.hcch.net/index_en.php?act=conventions.pdf&cid=98), though this is yet to come into force and will not require signatories to enforce the awards of other states.
Burnstein, Matthew, A Global Network in a Compartmentalised Legal Environment (1998), 33. It provides, “Member States shall take the measures needed to ensure that the consumer does not lose the protection granted by this Directive by virtue of the choice of the law of a non-member country as the law applicable to the contract if the latter has close connection with the territory of one or more Member States.”
Pursuant to the Brussels Regime which comprises the EEX Convention, the Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters, 16 Sep 1988 (EVEX Convention) and European Commission, Directive on Privacy and Electronic Communications (2002).
See Section 220.127.116.11.
See Section 1.4.1.
See Section 18.104.22.168.
New York Times Company v Sullivan (1964) 376 US 254
Sandoval, Greg, Visa Halts its Service for Allofmp3.com (2006). The closure of the site was subsequently procured by the United States as a condition of Russia’s membership of the WTO, but it reopened shortly afterwards under a new brand at http://www.mp3sparks.com/: Halpin, Tony, Russia Shuts Down Allofmp3.com (2007) .
For a description of Usenet, see Section 4.1.
(1997) 952 F Supp 1119
See Geist, Michael, The Shift Towards “Targeting" for Internet Jurisdiction (2003) for criticism and an alternative proposal.