How to make a PATRIOT SAFE for Americans

October 20th, 2003

At the heart of the insights that animate the lasting genius of the United States Constitution is the principle that no individual (or branch of government) can be trusted with too much power.


The Constitution’s architecture is one of checks-and-balances, meaning that the different branches of government are independent of each other, are chosen via different mechanisms, are populated by people with different geographic and institutional bases of power and for different terms of office, exercise authority through different channels, and pursue different (presumably self-interested) objectives.  Though independent, the branches are not permitted to exercise power entirely alone.  At the broadest level, any major shift in federal governmental policy requires the cooperation (or, at least, the non-objection) of the executive, legislative, and judicial branches.  Closer to the ground, individual executive branch actions are subject to judicial review against the relevant statutes and regulations;  individual judicial determinations can be corrected by subsequent legislation;  any new law requires not only the agreement of both houses of Congress, but of the President (and, ultimately, if challenged as unconstitutional, the federal courts as well).  Within the judicial branch, district courts are reviewed by the circuit courts of appeals, which are reviewed by the nine-member Supreme Court.  Supreme Court decisions can be overcome by legislation or constitutional amendment, as required.  A constitutional amendment typically requires the concurrence of supermajorities in the Senate and House of Representatives, and then of the legislatures of three-fourths of the states.  Etc., etc.  Basically, the federal constitution is loaded with provisions that effectively prevent one individual or one branch of government from exercising plenary, unreviewable authority all by him/her/itself.


[Surely the most eloquent (and moving, if you get buzzed by this kind of thing) statement of these ideas was The Federalist #51, written by James Madison:



But the great security against a gradual concentration of the several powers in the same department, consists in giving to those who administer each department the necessary constitutional means and personal motives to resist encroachments of the others. … Ambition must be made to counteract ambition. … It may be a reflection on human nature, that such devices should be necessary to control the abuses of government. But what is government itself, but the greatest of all reflections on human nature? If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself. A dependence on the people is, no doubt, the primary control on the government; but experience has taught mankind the necessity of auxiliary precautions.


The longevity of the Constitution as a stable but evolving charter of rights and freedoms owes a great debt to the obsessive attention paid by the Founders to the institutional architecture of government.]


It is no surprise, then, that Americans tend to be hostile to claims that any single institution of government should be trusted to do the right thing, without the need for scrunity or review by other, independent institutions of government. It is precisely that sort of “Trust us!” claim that was codified in a number of the provisions of the USA-PATRIOT Act of 2001.  (Amazing how the bill’s title, “Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism,” just happened to generate that catchy USA-PATRIOT acronym, no?)


Last week, a surprisingly diverse cluster of Senators introduced a bill (S. 1709, or, for the brand-conscious MTV generation, the Security and Freedom Ensured (SAFE) Act of 2003) that would re-establish some of the traditional mechanisms for judicial review and oversight of the federal government’s now-heightened surveillance and investigative powers. Led by Senators Larry Craig (R-ID) and Richard Durbin (D-IL), the bill’s cosponsors include Senators Michael Crapo (R-ID), Russell Feingold (D-WI), John Sununu (R-NH), Ron Wyden (D-OR) and Jeff Bingaman (D-NM). 


What has brought this set of Senators — left, right, and center — together?  The answer is to restore checks-and-balances to the law of surveillance.  (In the audio vision of this blog, that last line would be read in a Yoda voice.)  The SAFE bill would essentially re-assert the notion that the FBI’s exercise of surveillance powers should be uniformly subject to independent judicial review. 


What does this mean, exactly?  Back to the insights of the Founders, for a moment.  A terrific example of checks-and-balances in action is the Fourth Amendment, which, as interpreted by the courts, means that executive branch officers (e.g., the FBI) must ordinarily obtain a warrant from a neutral magistrate (who must find “probable cause”) before executing a search or seizing evidence.  Over the past two hundred years, a number of exceptions to the warrant requirement has been recognized by the courts as reasonable, such as searches incident to a lawful arrest, automobile searches, border searches, and items in plain view.  But the basic thrust of the Fourth Amendment has consistently been upheld, namely, that officers of the executive branch should ordinarily not conduct searches or seize evidence without a warrant issued on probable cause by a neutral official of the judicial branch.  Put another way, the Fourth Amendment interposes the judicial branch as a screen between the executive branch and the citizens it wishes to snoop on.


In simplest form, the USA-PATRIOT Act granted to the FBI expanded authority to monitor the activities of Americans without meaningful judicial oversight.  The SAFE Act would add judicial review to a few of the more eggregious instances of that authority.  The SAFE Act would, for example, restore the legal status quo ante for so-called “sneak-and-peak” or “delayed-notice” warrants, meaning they would be allowed only when notice to the subject of the search would (1) endanger someone’s life or physical safety; (2) result in flight from prosecution; or (3) result in destruction or tampering with evidence.  SAFE would also require notice within seven days.  The SAFE Act would essentially terminate the availability of “John Doe roving wiretaps” (which do not have to specify the person or place to be surveilled), by restoring the traditional requirement that law enforcement must identify either the person or place to be wiretapped. Likewise, in the case of “place” wiretaps, the SAFE Act would restore the requirement that law enforcement be sure that the suspect is present at the place to be wiretapped before recording.  In the case of library searches, the SAFE Act would retain the expanded definition of “business records” to include “library records”, but would reinstate the pre-USA-PATRIOT standard for seizing those records, namely, that the FBI must demonstrate it has reason to believe that the person to whom the records relate is a suspected terrorist or spy. The SAFE Act would also exempt libraries from the section of the USA-PATRIOT Act that allows access to electronic communications on the strength of an administrative subpoena rather than a court order.


All of these changes would re-assert the role of the judiciary in protecting against abuse by law enforcement.  I’m surprised that Attorney General Ashcroft is fighting so hard to defeat the SAFE Act (and the numerous companion bills that would do similar things — see CDT’s excellent website for a list of what’s been introduced in Congress this year), given his post-Waco Clinton-era record of opposition to all things Reno.  He, as much as anyone, should share the conviction that no one — not the FBI, nor the Attorney General, nor even the elders of the Republican Party — can be trusted not to abuse unreviewable, unaccountable powers.  It’s not like the SAFE Act would interpose the ACLU between law enforcement and their surveillance targets:  it’s the federal courts we’re talking about.  The federal judiciary has consistently proven its trustworthiness in handling matters of national security.


Personally, I don’t share a lot of the civil libertarians’ conventional reservations about the USA-PATRIOT Act’s expansion of federal surveillance powers.  I take the threat of destructive terrorism seriously, and I want the government to have the authority to stop the next 9/11 before it happens.  A free and open society requires security from those who would abuse our freedoms to kill and destroy. 


What I do not like, however, is the USA-PATRIOT Act’s insistence on making its surveillance powers plenary and unreviewable.  That runs directly contrary to the constitutional traditions of checks-and-balances that have served this country so well over the last 215+ years.  The SAFE Act is a reasonable corrective, and should be enacted into law. 


Judge Davis Gets It: The Internet is Not a Telephone

October 17th, 2003

It’s a wonderful thing to see a federal judge actually understand the Internet.  In his recent decision permanently enjoining the Minnesota Public Utility Commission from applying state telecom regulations to Voice-over-Internet-Protocol (VOIP) services, Judge Michael Davis of the Minnesota federal district court wrote what is one of the most well-informed — and most important — Internet-related decisions in recent years.  Judge Davis’s memorandum merits a close read;  I’ve posted it here in .pdf format.


At the heart of the case is the difference between circuit-switching and packet-switching, and whether the laws developed decades ago for the circuit-switched traditional phone network should be applied to the packet-switched Internet.  It’s like asking whether the laws developed for the railroad should be applied to the automobile. (Answer: Hell, no!)  What most non-technical people (including federal judges) don’t readily understand is how fundamentally different the packet-switched Internet is from the circuit-switched telephone.  Happily for the future of the Internet, Judge Davis figured it out.


Here’s a quick summary of the issues in the case, and Judge Davis’s analysis.  Minnesota law requires telephone companies to obtain certification from the Public Utility Commission.  The certificate specifies certain obligations (such as 911 service) and tariffs.


Vonage is a VOIP service provider that allows broadband customers to use a special Cisco box to translate telephone calls into IP packets, and vice versa.  A Vonage customer uses an ordinary telephone to make calls through her home broadband connection.  If a Vonage customer calls a traditional telephone customer, the telephone call is (a) converted to IP packets right there in the home via the special Cisco box, (b) carried over the Internet to a Vonage computer in the recipient’s area code, (c) converted back into a traditional telephone signal, and (d) sent into the traditional telephone network as a local call.  If a Vonage customer calls another Vonage customer, the call is (a) converted to IP packets right there in the home via the special Cisco box, (b) carried over the Internet to the recipient’s home, and (c) converted into a traditional phone call on the telephone attached to the recipient’s Cisco box.  Vonage call quality is good, but not as clear or reliable as a traditional circuit-switched telephone call.


Judge Davis’s decision revolves around two key legal questions:  (1) whether, under federal law, VOIP services are “information services” (not subject to regulation) or “telecommunications services” (subject to regulation), and (2) whether the federal Telecommunications Act pre-empts Minnesota’s state law.


On the first question, Judge Davis has little trouble concluding that VOIP is an “information service”.  He first looks at the language of the Communications Act of 1996, and concludes that Vonage service meets the plain statutory definition of “information service.”  Judge Davis goes further, however, examining the nature of Vonage’s service and comparing it to the 4-part test tentatively used by the FCC to determine whether a given service is phone-to-phone IP telephony.  (In its 1998 report to Congress on universal service issues, the FCC tentatively concluded that phone-to-phone IP telephony “lacks the characteristics that would render them ‘information services’ within the meaning of the statute, and instead bear the characteristics of ‘telecommunications services.’”  Whether the FCC’s 1998 thinking is still valid is very much subject to doubt.)  The FCC’s tentative 4-part test:



In using the term ‘phone-to-phone’ IP telephony, we tentatively intend to refer to services in which the provider meets the following conditions: (1) it holds itself out as providing voice telephony or facsimile transmission service; (2) it does not require the customer to use CPE [customer premises equipment] different from that CPE necessary to place an ordinary touch-tone call (or facsimile transmission) over the public switched telephone network; (3) it allows the customer to call telephone numbers assigned in accordance with the North American Numbering Plan, and associated international agreements; and (4) it transmits customer information without net change in form or content.


Judge Davis concludes that Vonage’s service does not meet requirements (2) and (4):  Vonage’s service offering requires customers to use special VOIP equipment, and entails the conversion of calls from IP packets into traditional telephone signals.  Therefore, Vonage service is not regulatable phone-to-phone IP telephony.


Fortifying his reasoning, Judge Davis noted that when Vonage customers call each other, they are effectively making computer-to-computer calls (where the computers are the special Cisco boxes into which the home telephones get plugged) that do not come into contact with the traditional (regulated) telephone system.  From the point at which it leaves one home to the point at which it reaches the other, the calls consist entirely of packets. 


In the thorniest (and perhaps murkiest) bit of analysis in the decision, Judge Davis delves deep into the FCC’s generic 1998 discussion of IP telephony, in which it examined three classes of providers involved with IP telephony:  first, those that only provide hardware and software at customer premises;  second, ISPs;  and third, IP telephony providers that deploy a gateway within the network to enable phone-to-phone service.  According to the FCC’s tentative conclusions, the first two categories are not regulated “telecommunications services”, while the third category is.  In his decision, Judge Davis determines that Vonage does not qualify for categories one (it provides more than hardware and software) and two (it is not an ISP).  But Vonage does, however, maintain gateways that transform IP packets into traditional telephone signals, and vice versa. 


So how can Judge Davis conclude that Vonage is not an phone-to-phone IP telephony provider?  His opinion is less than crystal clear, but the answer appears to center on the fact that the Vonage system provides computer-to-computer or computer-to-phone service (keeping in mind that the Cisco box that sits between the customer’s broadband connection and her telephone is a computer), but never phone-to-phone.  Judge Davis goes on to quote a range of FCC policy statements that argue that the legal line between telecommunications and information services is one that should be maintained, even though the engineers might have a hard time distinguishing them on the basis of their technical infrastructure.  (Judge Davis is here running up against an increasingly pressing conceptual headache.  He’s not alone.  Many commentators, including me, have argued in recent years that the massive convergence of telephony and IP networks is going to make irrational the FCC’s infrastructure-based regulatory distinctions.)  Stressing that information services can legitimately be deployed on top of telecommunications infrastructure without being pulled into the orbit of telecommunications regulations, Judge Davis finds that Vonage uses telecommunications services, but does not provide them.  In passing, Judge Davis trashes the Minnesota PUC’s “quacks like a duck” analysis as a distracting oversimplification. 


[By the way: if Judge Davis is right that Vonage’s service doesn’t count as the kind of phone-to-phone IP telephony that is subject to regulation, what does?  What was the FCC talking about?  The answer is that the FCC wanted to make clear that telecom companies couldn’t remove themselves from regulation by migrating their backhaul networks to the Internet protocol.  Unlike Vonage (which packetizes phone calls right there in the customer’s home), telecom companies typically have IP-based networks at the core, with traditional switched telephone signals covering at least the last mile. From an engineering standpoint, this is a suspect distinction.  But it seems to be what motivated the FCC’s 1998 definitions.]


Whether federal law pre-empts state law typically depends on the clarity of Congress’s intent.  One thing Judge Davis gets right is Congress’s unusually specific statement of intent that the Internet remain as unregulated as possible.  Said Congress in the Telecommunications Act of 1996:



“It is the policy of the United States . . . to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.”


In short, Judge Davis concludes that Vonage provides an “information service”, that “information services” are expressly unregulated under the federal Telecommunications Act of 1996, and that Minnesota’s effort to regulate Vonage’s “information service” conflicts with federal law and must therefore be pre-empted.  He orders that Minnesota be permanently enjoined from enforcing its attempt to regulate Vonage.


In my judgment, this a fantastic victory for the Internet, and a hugely significant precedent.  For one thing, a federal court has declared that, for purposes of federal law, packets are packets are packets, even if some of them happen to be carrying voice traffic.  More importantly, VOIP promises to bring low-cost, highly-reliable, and feature-rich integrated communications devices within the reach of many many more people in developing countries for much less money than traditional telephony could ever realistically hope to achieve.  A strong stand against regulation by the US would be a powerful signal to other nations.


Finally, the decision bolsters the notion that Internet services should not be forcibly shoehorned into decades-old telecommunications regulations.  If VOIP is to be regulated, it should be approached in ways that respect the comprehensively different nature of packet-switched networks. 


For example: If 911 service is the issue, then the regulators need to make the mandates fit the capabilities of the technology.  Vonage already provides a limited 911 service similar to that for mobile phones (after all, your Vonage box can connect to the Internet anywhere in the world, so Vonage can’t tell with certainty where a 911 caller is, geographically, at a given moment).  The fact that mobile phones can’t convey geographic location to 911 operators was not considered to be a reason to slow or restrict mobile phone technology;  likewise, VOIP providers should be allowed to offer services with modified 911 capabilities, so long as customers know exactly what they’re getting.


Ineluctable modality of the risible

October 16th, 2003

I know it’s bad form to laugh at those who have tumbled headlong and humiliated into the rank dankness of the gutter.  But when the high-and-mighty has tripped over his own sanctimony, hypocrisy, and bloated ego, well, all bets are off.   In “Rush to Recovery,” Harry Shearer administers a therapeutic dose of cruelty.