DATE DOWNLOADED: Tue Apr 15 15:55:52 2025

SOURCE: Content Downloaded from HeinOnline

Citations:
Please note: citations are provided as a general guideline. Users should consult their preferred
citation format's style manual for proper citation formatting.

Bluebook 21st ed.

Paul Ohm, The Argument against Technology-Neutral Surveillance Laws, 88 TEX. L. REV.
1685 (June 2010).

ALWD 7th ed.

Paul Ohm, The Argument against Technology-Neutral Surveillance Laws, 88 Tex. L. Rev.
1685 (2010).

APA 7th ed.

Ohm, Paul. (2010). The argument against technology-neutral surveillance laws. Texas
Law Review, 88(7), 1685-1714.

Chicago 17th ed.

Paul Ohm, "The Argument against Technology-Neutral Surveillance Laws," Texas Law
Review 88, no. 7 (June 2010): 1685-1714

McGill Guide 10th ed.

Paul Ohm, "The Argument against Technology-Neutral Surveillance Laws" (2010) 88:7 Tex
L Rev 1685.

AGLC 4th ed.

Paul Ohm, 'The Argument against Technology-Neutral Surveillance Laws' (2010) 88(7)
Texas Law Review 1685

MLA 9th ed.

Ohm, Paul. "The Argument against Technology-Neutral Surveillance Laws." Texas Law
Review, vol. 88, no. 7, June 2010, pp. 1685-1714. HeinOnline.

OSCOLA 4th ed.

Paul Ohm, 'The Argument against Technology-Neutral Surveillance Laws' (2010) 88 Tex L
Rev 1685 Please note: citations are provided as a general
guideline. Users should consult their preferred citation format's style manual for
proper citation formatting.

-- Your use of this HeinOnline PDF indicates your acceptance of HeinOnline's Terms and
Conditions of the license agreement available at
https://heinonline.org/HOL/License
-- The search text of this PDF is generated from uncorrected OCR text.
-- To obtain permission to use this article beyond the scope of your license, please use:
Copyright Information
The Argument Against Technology-Neutral
Surveillance Laws

Paul Ohm*
Introduction
Should Congress write tech-specific or tech-neutral laws? Those who
have considered this question have almost always chosen neutrality: laws
should refer to the effects, functions, or general characteristics of technology,
but never to a particular type or class of technology.' Those who espouse
tech neutrality come from across the political and ideological spectrum and
embrace tech neutrality dogmatically, often referring to it as a "principle,"
one presumably violated only in exceptional circumstances for the most
compelling reasons.2
But a close examination of the arguments supporting tech neutrality
reveals many underappreciated flaws. At least three arguments support tech
neutrality-consistency, the need to avoid underinclusiveness, and the rec-
ognition of institutional shortcomings-but each is contingent and rebuttable,
and in many situations does not apply.
While other scholars have called Congress's blind adherence to the
principle of tech neutrality into question,3 none have explored the neutrality
of laws regulating government search and surveillance. This rich, important
context bears close scrutiny because the path of surveillance law so often
follows the twists and turns of evolving technology. Moreover, since 9/11,
Congress has more than once replaced tech-specific surveillance laws with
tech-neutral ones: for example, with the USA PATRIOT Act 5 Congress

* Associate Professor of Law, University of Colorado Law School. I thank Professor Bobby
Chesney and the editors of the Texas Law Review for the invitation to participate. I also thank the
participants at both the Symposium and the University of Colorado workshop series for their
comments. In particular, I would like to thank William Boyd, Joe Feller, Susan Freiwald, Jennifer
Granick, Lisa Graves, Marcia Hoffman, Clare Huntington, Orin Kerr, Derek Kiernan-Johnson,
Sarah Krakoff, Michael Kwun, Jon Michaels, Scott Moss, Helen Norton, John Radsan, Carolyn
Ramsey, Andrew Schwartz, Harry Surden, and Wendy Seltzer for their comments.
1. See, e.g., infra notes 29-30 and accompanying text.
2. See infra note 36 and accompanying text.
3. See Lyria Bennett Moses, Recurring Dilemmas: The Law's Race to Keep Up with
Technological Change, 2007 U. ILL. J.L. TECH. & POL'Y 239, 239 (complaining that tech-neutral
drafting may not be effective as technology changes); Chris Reed, Taking Sides on Technology
Neutrality, 4 SCRIPT-ED 263, 282-84 (2007) (advocating a three-step process lawmakers should
undergo when deciding between tech-neutral and tech-specific legislation).
4. See, e.g., John Schwartz, Debate over Full-Body Scans vs. Invasion of Privacy FlaresAnew
After Incident, N.Y. TIMES, Dec. 29, 2009, at A14 (discussing potential legislation to regulate the
use of newly developed full-body scanners in airports).
5. Pub. L. No. 107-56, 115 Stat. 272 (2001) (codified as amended in scattered titles of U.S.C.).
1686 Texas Law Review [Vol. 88:1685

brought neutrality to the Pen Register and Trap and Trace Devices Act6 (Pen
Register Act), and through the Protect America Act 7 (PAA) it did the same
for the Foreign Intelligence Surveillance Act' (FISA).
We should worry about this trend because the arguments in favor of
tech neutrality are especially misguided in the surveillance context. When it
comes to surveillance, every argument supporting tech neutrality can be met
with a powerful counterargument: Tech-neutral laws often force consistency,
even when inconsistency is preferable; they avoid underinclusiveness by
permitting overinclusiveness; and they address Congress's supposed institu-
tional shortcomings by cutting Legislative oversight over surveillance, even
though history has taught us to beware the surveillance of an unchecked
Executive. Given the deep flaws in the arguments for tech neutrality in the
surveillance context, we should stop treating tech neutrality as a principle
and instead treat it as a choice.
Finally, the blind adherence to the principle of tech neutrality pushes
Congress away from the many benefits of tech specificity. Most importantly,
a tech-specific surveillance law, even one imposing few constraints on the
agencies conducting surveillance, forces the Executive Branch to consult
with Congress whenever technology changes in significant ways, which
might help offset the troubling culture of secrecy in national security policy
by bringing broader, more participatory democratic oversight to the conduct
of national surveillance. Also, because technology evolves so rapidly and
constantly, tech-specific surveillance laws operate as a technology sunset,
expiring not on some arbitrarily defined timetable, but whenever the cir-
cumstances demand. Both of these benefits increase the Legislature's role in
national surveillance and national security debates and restore checks against
the Executive's power in ways that might have helped avoid some of the sur-
veillance abuses and excesses of the recent past.
This Article proceeds in three Parts, offering, in turn, the best arguments
for tech neutrality (Part I), the underappreciated counterarguments to those
arguments (Part II), and the case for tech specificity (Part III). Ultimately,
this Article tries to counter the pervasively held attitude that tech-specific
laws are indefensible mistakes to be avoided. Quite often, tech specificity is
the wiser course-the best way to balance the government's need to provide
security with the right to privacy.

6. 18 U.S.C. §§ 3121-3127 (2006).

7. Pub. L. No. 110-55, 121 Stat. 552 (2007) (to be codified at 50 U.S.C. §§ 1803, 1805a-
1805c).
8. Pub. L. No. 95-511, 92 Stat. 1783 (1978) (codified as amended in scattered titles of U.S.C.).
2010] The Argument Against Technology Neutrality 1687

I. Tech-Neutral Laws

A. Defining Tech Neutrality

Whenever Congress writes a law to address a problem caused by
technology, it must decide whether to draft tech-neutral or tech-specific
provisions. Tech-neutral provisions refer to technology in general, vague,
open-textured terms that specify purposes, effects, functions, and other gen-
eral characteristics. While Congress has used tech neutrality for surveillance
law inconsistently, for decades it has embraced neutral drafting in other tech-
heavy fields such as telecommunications 9 and copyright. Under the
Copyright Act, for example, copies are defined in part as "material
objects.., in which a work is fixed by any method now known or later
developed, and from which the work can be perceived, reproduced, or
otherwise communicated, either directly or with the aid of a machine or
device." '
In contrast, tech-specific provisions refer to specific types or classes of
technologies. For example, the Pen Register Act, a surveillance law, once
applied only to "a device which records or decodes electronic or other im-
pulses which identify the numbers dialed or otherwise transmitted on the
telephone line to which such device is attached."'" In the USA PATRIOT
Act, Congress replaced this tech-specific definition with a tech-neutral one
that broadly covers all "dialing, routing, addressing, or 3 signaling
information."' ' 2 We will revisit this example later in the Article.'
Most tech-centric laws lie along a spectrum from tech specificity to tech
neutrality with few as close to either endpoint as the laws just cited. Some-
times it can be tricky to tell near which end of the spectrum a statute falls. A
definition may seem tech specific on first blush because it lists specific types
of technologies, but sometimes the point of such a list is to exhaust
possibilities, covering the definitional waterfront, signaling that the list is
meant to cover everything neutrally. For example, the Computer Fraud and
Abuse Act-a Federal anti-computer-hacking law-defines a computer to
mean "an electronic, magnetic, optical, electrochemical, or other high speed
data processing device performing logical, arithmetic, or storage functions,
and includes any data storage facility or communications facility directly re-
lated to or operating in conjunction with such device."' 4 Despite providing a
long list of specific types of technology, Congress intended this definition to

9. See Reed, supra note 3, at 264 (recognizing that "technology neutrality has continued to be a
pervasive concept" in telecommunications policy).
10. 17 U.S.C. § 101 (2006).
11. Electronic Communications Privacy Act of 1986, Pub. L. No. 99-508, sec. 301, § 3126(3),
100 Stat. 1848, 1871.
12. USA PATRIOT Act of 2001, Pub. L. No. 107-56, sec. 216, 115 Stat. 272, 290 (codified at
18 U.S.C. § 3127(3) (2006)).
13. See infra notes 147-52 and accompanying text.
14. 18 U.S.C. § 1030(e)(1) (2006).
1688 Texas Law Review [Vol. 88:1685

have a broadly neutral meaning, and indeed the Seventh Circuit has inter-
preted it to cover not only laptop and desktop computers but "[e]very cell
phone and cell tower[,] ...every iPod, every5 wireless base station in the cor-
ner coffee shop, and many another gadget."1
Congress must often choose between tech neutrality and specificity
when it drafts surveillance laws because the great challenge of surveillance is
keeping up with the latest advances in technology. Over the decades, it has
written surveillance laws that fit at different points along the spectrum. Con-
sider one law in particular, the Wiretap Act, 16 and take a single, complex
subsection of this Act, 18 U.S.C. § 251 l(2)(g), which lists exceptions to the
general prohibition on wiretapping, and this subsection provides a menagerie
of examples from across the specificity spectrum. Under § 2511 (2)(g), it is
not an illegal wiretap to intercept electronic communications that are "readily
accessible to the general public" 17-- a classically tech-neutral rule-radio
communications "transmitted by a station operating on an authorized fre-
quency within the bands allocated to the amateur, citizens band, or general
mobile radio services" 18-a mostly tech-specific rule-or communications,
"the transmission of which [are] causing harmful interference" to another
radio "to the extent necessary to identify the source of such interference ' 9-
which seems to fall somewhere in between.
Through the first few technological epochs of electronic surveillance- 22
2° 2
from the earliest telephone wiretaps, to the spike mikes and room bugs
of the mid-twentieth century, up until the early days of computer-network
surveillance-Congress wrote many tech-specific surveillance laws. My
strong sense is that in the past decade or so, it has switched to writing only
tech-neutral ones. As one example, the precursors to the tech-specific
Wiretap Act provisions listed above were included in the original 1968

15. United States v. Mitra, 405 F.3d 492, 495 (7th Cir. 2005).
16. Omnibus Crime Control and Safe Streets Act of 1968, Pub. L. No. 90-351, tit. III, 82 Stat.
197, 211-23 (codified as amended at 18 U.S.C. §§ 2510-2520, 47 U.S.C. § 605 (2006)).
17. 18 U.S.C. § 2511(2)(g)(i) (2006).
18. Id. § 2 511(2)(g)(ii)(III).
19. Id. § 2511(2)(g)(iv).
20. See Olmstead v. United States, 277 U.S. 438, 456-57 (1928) (describing the wiretap as
small wires inserted into the telephone lines coming from the petitioners' houses).
21. See Silverman v. United States, 365 U.S. 505, 506-07 (1961) (describing the instrument
used as a microphone with a spike attached to it that was inserted into the house to become a
"conductor of sound"); Goldman v. United States, 316 U.S. 129, 131 (1940) ("They had with them
another device.., having a receiver so delicate as, when placed against the partition wall, to pick
up sound waves ...").
22. See Katz v. United States, 389 U.S. 347, 348 (1967) (describing the device used as capable
of intercepting communications while being place outside of a structure).
2010] The Argument Against Technology Neutrality 1689

Wiretap Act, 23 while the tech-neutral "readily accessible" provision was

added much more recently.24

B. Tech Neutrality in NationalSecurity Surveillance Law

Those who urge Congress to expand surveillance authorities to protect
national security often argue for tech-neutral surveillance laws. For example,
John Yoo and Eric Posner applauded the USA PATRIOT Act's amendments
to FISA for embracing tech neutrality.25 Thanks to the USA PATRIOT Act,
"FISA warrants.., are now technology-neutral... [and] allow continuing
surveillance of a terrorist target even if he switches communication devices
and methods. 2 6
While Yoo and Posner lauded the shift to a tech-neutral FISA warrant
standard, others remained dissatisfied about lingering tech specificity in the
law, even after the USA PATRIOT Act. In particular, in the middle part of
the first decade of the twenty-first century, Executive Branch officials
pressed Congress to fix one form of lingering specificity in FISA-the way it
treated communications bouncing through satellites differently than commu-
nications carried on fiber-optic cables.2 7 Under the widely accepted
interpretation of the statute's definitions, if the NSA wanted to monitor the
communications of a foreigner (or, to use the statute's term, a non-"United
States person") located outside the United States, it faced significant proce-
dural hurdles if the communications happened to travel over a fiber-optic
cable and almost no hurdles if they traveled via satellite.28

23. Omnibus Crime Control and Safe Streets Act of 1968, Pub. L. No. 90-351, sec. 802,
§ 2511(2)(a)-(b), 82 Stat. 197, 214.
24. Electronic Communications Privacy Act of 1986, Pub. L. No. 99-508, sec. 101(b)(4), 100
Stat. 1848, 1850.
25. See John Yoo & Eric Posner, The PatriotAct Under Fire, AEI: ON THE ISSUES, Dec. 1,
2003, http://www.aei.org/issue/19661 (describing tech neutrality as a "common-sense adjustment[]"
of necessity).
26. Id.
27. See, e.g., Modernization of the Foreign Intelligence SurveillanceAct: HearingBefore the S.
Select Comm. on Intelligence, 110th Cong. 27-28 (2007) [hereinafter Modernization of FISA]
(statement of Kenneth L. Wainstein, Assistant Att'y Gen. of National Security, United States
Department of Justice) (describing the distinction running throughout FISA between wire
communications and radio communications).
28. To state the complicated argument concisely, under the definition of "electronic
surveillance," FISA treats surveillance of "wire communications" differently than it treats
surveillance of "radio communications." Compare 18 U.S.C. § 1801(f)(2) (2006) (defining
"electronic surveillance" to include wire communications acquired without regard to intent or a
reasonable expectation of privacy), with id. § 1801(f)(3) (requiring intentional acquisition of the
transmission and a reasonable expectation of privacy for acquisitions of radio transmissions to
constitute electronic surveillance). Surveillance of radio is not regulated by FISA unless, among
other things, "both the sender and all intended recipients are located within the United States." Id.
§ 1801(f)(3). Thus, for radio surveillance, when the NSA knows at least one party is outside the
United States, FISA does not apply. In contrast, surveillance of wire communications falls within
FISA if only one party is "in the United States" and if the surveillance itself "occurs within the
United States." Id. § 1801(f)(2). This summary omits a few details.
1690 Texas Law Review [Vol. 88:1685

Executive Branch officials found this distinction untenable. Beginning

at least in 2006, officials from the Intelligence Community and Justice
Department pressed Congress repeatedly for a fix to FISA. Ken Wainstein,
the Department of Justice's first Assistant Attorney General in charge of the
National Security Division, suggested,
Rather than focusing, as FISA does today, on how a communication
travels or where it is intercepted, we should define FISA's scope by
reference to who is the subject of the surveillance. If the surveillance
is directed at a person in the United States, FISA generally should 29
apply; if the surveillance is directed at persons overseas, it shouldn't.
Former Director of National Intelligence, Admiral Michael McConnell,
agreed, testifying that "[o]ur job is to make the country as safe as possible by
providing the highest quality intelligence available. There is no reason to tie
30
the Nation's security to a snapshot of outdated technology."
Congress eventually gave the Executive Branch what it wanted. First,
in 2007, it enacted the PAA, which erased the wire and radio distinction for
some cases, but out of concern for a rushed legislative process, 31 it set a six-
month sunset on the law. 32 After the PAA expired, Congress enacted the
FISA Amendments Act of 2008,33 which took a different textual approach
than the PAA, albeit to the same ends. As amended by the FISA
Amendments Act, FISA no longer draws a distinction between
communications carried by satellite and those carried by fiber-optic cables
when non-U.S. persons are the target of the surveillance.34 Now, intelligence
analysts can listen to those communications no matter how they are carried-
whether by copper wire, fiber-optic cable, microwave radio, satellite radio, or
something else-under the same low standard. And because this part of
FISA is now tech neutral, the same rules will apply to any communications
technology developed in the future, regardless of how it operates, where it is
deployed, or if it implicates privacy in new ways.

C. The Arguments in Favor of Tech Neutrality

This story of how and why Congress made FISA more neutral is typical.
In many legislative debates over surveillance law, one side or another will

29. Modernization of FISA, supra note 27, at 30 (statement of Kenneth L. Wainstein, Assistant
Att'y Gen. of National Security, United States Department of Justice).
30. Id. at 19 (statement of J. Michael McConnell, Director of National Intelligence).
31. James Risen, Bush Signs Law to Widen Reachfor Wiretapping,N.Y. TIMES, Aug. 6, 2007,
at Al.
32. Protect America Act of 2007, Pub. L. No. 110-55, sec. 6(c), 121 Stat. 552, 557 (to be
codified at 50 U.S.C. § 1803).
33. Pub. L. No. 110-261, 122 Stat. 2436 (to be codified in scattered sections of 50 U.S.C.).
34. Id. sec. 101, § 702. As amended, FISA now allows the Intelligence Community to monitor
communications of non-U.S. persons not known to be in the United States, whether carried over
wire or radio, without prior judicial approval, subject to some safeguards and checks, including
mandatory notice to the FISA Court. Id.
2010] The Argument Against Technology Neutrality 1691

urge Congress to reject tech specificity in favor of tech neutrality. 35 Those

who argue for tech neutrality too rarely explain in detail the reasoning behind
their arguments. Quite often, tech neutrality is a principle or rule, and it al-
36
most seems to go without saying. Even when proponents of neutrality
explain their reasoning, they often do so cursorily. As a result, we lack satis-
fying theoretical explanations for tech neutrality. 37 Before I offer
counterarguments, I must first present the best arguments I can for tech neu-
trality in the surveillance context in order to try to avoid taking on straw men.
The arguments for neutrality are not inherently flawed, and sometimes
tech neutrality may be a good idea. Still, these arguments are not
unassailable, and they certainly do not support elevating the idea of tech
neutrality to the level of a principle. Instead, they have gaps and logical
flaws that suggest the shortcomings of the approach, which I will explore in
Part II. These arguments number three.

1. Consistency.-The most often recited argument in favor of tech

neutrality is the need for consistency-the need to avoid arbitrary
distinctions between technologies that should be treated alike.38 When
Congress enacts a tech-specific rule, it regulates a specific technology while
leaving unregulated similar technologies. 39 It makes no sense to treat these
similar technologies differently because the policy rationale justifying the
rule usually focuses on the40 effects of the technology, not on the function or
features of the technology.

35. See supra notes 25-30 and accompanying text.

36. See, e.g., Nathan Alexander Sales, Run for the Border: Laptop Searches and the Fourth
Amendment, 43 U. RICH. L. REV. 1091, 1127-28 (2009) (postulating that use limits are a better way
to regulate border searches of laptops because special-collection limits would "violate the principle
of technological neutrality").
37. Some scholars have developed lists of explanations for tech neutrality. Chris Reed cites
three aims: "futureproofing, online and offline equivalence, and encouraging the development and
uptake of the regulated technology." Reed, supra note 3, at 275. Similarly, Ilse van der Haar
argues that tech neutrality leads to "non-discrimination, durability, efficiency, and certainty."
Corine Schouten, EU Failed to Apply Technology Neutrality in Regulating Communication
Services, INNOVATIONS REP., Nov. 12, 2008, http://www.innovations-
report.com/html/reports/communication-media/eu-failed-apply-technotogy-neutrality-regulating-
124187.html. Almost all of these laudable qualities appear in the three arguments I present below.
38. See Stephanie Cooper Blum, What Really Is at Stake with the FISA Amendments Act of
2008 and Ideas for Future Surveillance Reform, 18 B.U. PUB. INT. L.J. 269, 299 (2009) (relating
that the FISA Amendments Act "proceed[s] in a technology-neutral and less arbitrary fashion" than
FISA).
39. See Modernization of FISA, supra note 27, at 10 (statement of J. Michael McConnell,
Director of National Intelligence) ("FISA was written to distinguish between collection [of
communications] on a wire and collection out of the air.").
40. See id. at 28 (statement of Kenneth L. Wainstein, Assistant Att'y Gen. of National Security,
United States Department of Justice) (explaining that by embedding tech-specific language in FISA,
Congress "use[d] the manner in which communications [were] transmitted as a proxy for the types
of targets and communications that the statute intended to reach"); id. at 30 (arguing that instead of
focusing upon "how a communication travels or where it is intercepted, [Congress] should define
FISA's scope by reference to who is the subject of the surveillance").
1692 Texas Law Review [Vol. 88:1685

Consistency sits at the heart of the Executive Branch arguments in favor

of the PAA and FISA Amendments Act. Admiral McConnell, Assistant
Attorney General Wainstein, and other Executive Branch officials repeatedly
argued against treating satellite and fiber-optic communications differently. 4'
Consistency arguments often invoke happenstance and chance. Should the
fact that a terrorist's communications happen to be carried over fiber-optic
cable rather than via satellite have any bearing on whether the 2NSA can listen
to them? Of course not, the tech-neutrality proponents argue.

2. Keeping Up with Technological Change.-Many argue that laws

should be written neutrally because technology changes too quickly for the
legislative process to keep up. 4 3 According to this argument, specificity
leads inevitably and rapidly to anachronism because by the time a bill be-
comes a law, the technology will have evolved. 44 To support Admiral
McConnell's call for tech neutrality in FISA, Andrew McCarthy of the
National Review argued, "Any statute that focuses on technology will
become obsolete (or worse, counterproductive) when technology
changes... ,A5 Those making particularly strong forms of this argument
seem to hold tech neutrality up as a form of enlightened modernity; a recog-
nition by Congress that something in society-technology--moves too
quickly for the legislative process.4 Outside the surveillance context, the
Seventh Circuit explained, "[L]egislators ... know that complexity is
endemic in the modern world and that each passing year sees new

41. See id. at 13 (statement of J. Michael McConnell, Director of National Intelligence)

("FISA's definitions of 'electronic surveillance' should be amended so that it no longer matters how
collection occurs (whether off a wire or from the air)."); id. at 34 (statement of Kenneth L.
Wainstein, Assistant Att'y Gen. of National Security, United States Department of Justice) ("In
keeping with the preference for technological neutrality, we would eliminate the distinction between
'wire' and 'radio' communications that appears throughout [FISA].").
42. See id. at 11 (statement of J. Michael McConnell, Director of National Intelligence) ("Our
job is to make the country as safe as possible by providing the highest quality intelligence available.
There is no reason to tie the nation's security to a snapshot of outdated technology.").
43. See id. at 15 (advocating amendments that would "make FISA technology-neutral, so that as
communications technology develops-which it absolutely will-the language of the statute does
not become obsolete").
44. See, e.g., id. at 33 (statement of Kenneth L. Wainstein, Assistant Att'y Gen. of National
Security, United States Department of Justice) ("As a result of revolutions in communications
technology since 1978, .. . the current definition of 'electronic surveillance' sweeps in surveillance
activities that Congress actually intended to exclude from FISA's scope.").
45. Andrew C. McCarthy, FISA Reform: The Bad Bill That Beats No Bill, NAT'L REV. ONLINE,
Feb. 14, 2008, http://article.nationalreview.com/348094/fisa-reform-the-bad-bill-that-beats-no-
bill/andrew-c-mccarthy?page=l.
46. See Modernization of FISA, supra note 27, at 33 (statement of Kenneth L. Wainstein,
Assistant Att'y Gen. of National Security, United States Department of Justice) ("Legislators in
1978 should not have been expected to predict the future of global telecommunications, and neither
should this Congress.... We should not have to overhaul FISA each generation simply because
technology has changed.").
2010] The Argument Against Technology Neutrality 1693

developments. That's why they write general statutes rather than enacting a
list of particular forbidden acts."47
Furthermore, tech-specific laws do not simply become unacceptably
anachronistic, but rather, they tend to become underinclusive with time.
Once the specific type or class of technology targeted by a tech-specific law
evolves into a new successor form, the law no longer applies. For those who
support the policy underlying the law, this makes the law underinclusive, as
they would prefer a law that expands to cover new versions of old
technology.
Proponents of the PAA and FISA Amendments Act complained that the
evolution of technology from satellite to fiber-optic cable communications
had narrowed FISA.48 According to their version of history, in 1978, when
Congress enacted FISA, almost all transoceanic communications bounced
49
through satellites using radio waves.
Times and technologies had changed. Thousands of miles of new fiber-
optic cable had been laid since 1978, and the telecommunications industry
had moved much of its operations from satellites to the new, cheaper,
plentiful fiber-optic alternative. 50 By the time of the debates over the PAA,
telephone companies were carrying most long-haul-phone calls over cables
including, of course, the calls of terrorists and agents of foreign powers,
creating an underinclusive, technological anachronism in the law. 51 What
Congress had chosen not to regulate in 1978, evolving technology had re-
regulated.
The narrower FISA severely burdened the Intelligence Community.
Admiral Michael McConnell argued, "Because technology has changed but
the law has not, this statute-meant to protect against domestic abuses-
instead protects potential foreign terrorists. We are significantly burdened in

47. United States v. Mitra, 405 F.3d 492, 495 (7th Cir. 2005).
48. See infra note 51 and accompanying text.
49. See Modernization of FISA, supra note 27, at 10 (statement of J. Michael McConnell,
Director of National Intelligence) ("When the law was passed in 1978, almost all local calls were on
a wire and almost all long-haul communications were in the air .... ").
50. See Declan McCullagh & Anne Broache, NSA Eavesdropping:How It Might Work, CNET
NEWS, Feb. 7, 2006, http://news.cnet.com/NSA-eavesdropping-How-it-might-work/2100-1028_3-
6035910.html?tag=rrmcol (explaining that today "an undersea web of fiber-optic cables spans the
globe-and those carry the vast majority of voice and data that leave the United States" so that "99
percent of the world's long-distance communications travel through fiber links (and t]he remaining
1 percent... are satellite-based").
51. See Modernization of FISA, supra note 27, at 10-11 (statement of J. Michael McConnell,
Director of National Intelligence) (explaining that, in 1978, because most local calls were wire
communications and most international calls were wireless communications, FISA's scope included
wire communications; today, "the situation is completely reversed").
52. See id. at 19 (statement of Kenneth L. Wainstein, Assistant Att'y Gen. of National Security,
United States Department of Justice) (explaining that technological "advances have largely upended
FISA's intended carve-out of intelligence activities directed at persons overseas" so that
"considerable resources of the Executive Branch and the FISA Court are now expended on
obtaining court orders to monitor the communications of terrorist suspects overseas").
1694 Texas Law Review [Vol. 88:1685

capturing overseas communications of foreign terrorists planning to conduct

attacks inside the United States.",53 Similarly, Assistant Attorney General
Wainstein testified that "sweeping changes since 1978-both in the nature of
the threat that we face and in telecommunications technologies-have upset
the delicate balance that Congress sought to achieve when it enacted
FISA. 5 4

3. Institutional Competence.-Finally, tech-neutral provisions respond

to institutional concerns, helping Congress do what it does well and avoid
doing what it does poorly. Those who argue against tech-specific statutes
often intimate or assert that Congress is not equipped to understand compli-
cated new technologies.55 These arguments echo themes from each of the
prior arguments-about consistency and the rate of technological change-
tying them specifically to Congress's perceived institutional shortcomings.
As Bruce Berkowitz of the Hoover Institution puts it, "Intelligence officials
know what they really require to do their mission, and legislators know how
to write authorizing legislation., 56 General Michael Hayden, then-Director
of Central Intelligence, testifying about FISA, suggested that legislators were
not equipped to keep up with changing technology: "Legislators in 1978
should not have been expected to predict the future of global
telecommunications, and neither should you.... [T]he statute we develop
should be technology neutral. 5 7

II. The Problems with Tech Neutrality

On the surface, these arguments have undeniable persuasive force, but
they fare poorly under closer scrutiny. Every purported benefit of tech
neutrality-consistency, avoidance of underinclusiveness, and institutional
competence-can be recast as a shortcoming instead. These shortcomings
are best illustrated through laws other than FISA, allowing us to draw lessons
from older debates about the laws governing criminal surveillance. Consider
the significant downsides of tech neutrality.

53. Mike McConnell, Letter to the Editor, ProtectingAmericans and Their Rights, N.Y. TIMES,
May 5, 2007, at A12.
54. Modernization of FISA, supra note 27, at 24 (statement of Kenneth L. Wainstein, Assistant
Att'y Gen. of National Security, United States Department of Justice).
55. Often arguments like these carry a hint of superiority and maybe even a sense of ridicule.
Perhaps other societal institutions can keep up with technology, but not Congress, which is stodgy
and out of touch, full of elderly members who are the same. See, e.g., Jim Puzzanghera, Weighing
High-Tech Bills in Analog: Political Issues Pile Up in the Fast-Evolving Sector, but Congress'
Expertise Isn't Up To Date, L.A. TIMES, Aug. 7, 2006, at CI (cataloging the frustration of business
leaders in educating Congress on technology and noting the substantial ridicule heaped on former
Senator Ted Stevens for describing the Internet as "a series of tubes").
56. Bruce Berkowitz, The Wiretap Flap Continues,WALL ST. J., Sept. 18, 2007, at Al5.
57. FISA for the 21st Century: Hearing Before the S. Comm. on the Judiciary, 109th Cong. 8
(2006) (statement of Michael V. Hayden, Director, Central Intelligence Agency).
20101 The Argument Against Technology Neutrality 1695

A. TreatingDifferences Alike
While the law should not treat different technologies differently when
doing so would reward happenstance and chance, it is also true that some
differences deserve to be treated differently. If instead Congress, trying not
to violate the "principle" of tech neutrality, treats such differences alike, it
will produce ineffective laws with unpredictable or pernicious effects.
As many have written, in our modem, information-driven world,
technology acts like architecture, constraining and enabling certain human
behavior.58 But because different technologies constrain to different degrees
and in different ways, 59 we should not regulate any specific technology until
we take the time to study it to allow us to tailor our laws and regulations to
the idiosyncrasies of the specific context. Policy makers fail to do this when
they enact tech-neutral laws.
Many information-privacy scholars have recognized this point, arguing
that policy makers should respond to the diversity of technology by tailoring
and differentiating regulation to the specific context. Helen Nissenbaum has
argued that expectations of privacy turn entirely on deeply contextualized
differences between situations. 60 Dan Solove has written extensively about
how changing technology brings new challenges to privacy. 6 1 In part be-
cause of the diversity of privacy-impacting technologies, he concludes that
privacy cannot be described monolithically but instead should be considered
as a complex of different values that relate to one another only through
Wittgensteinian "family resemblances. 62
Scholars writing about national security and criminal law have drawn
similar conclusions. Orin Kerr has written extensively about how specific
new forms of technology enable both new forms of surveillance and new
methods for committing crime. 63 He argues that these differences matter to
criminal procedure and suggests rules that take these subtle differences into
account. 64 Similarly, Jack Balkin and Sandy Levinson write persuasively
about how "new technologies of surveillance, data storage, and computation"

58. See, e.g., LAWRENCE LESSIG, CODE: VERSION 2.0, at 77-79 (2006) (arguing that the way in
which any given technology is implemented-and selected from among the many potential
architectures-is an exercise of power with political and social consequences).
59. See id. at 203-07 (cataloging the privacy consequences inherent in the specific architecture
of several modem technologies).
60. HELEN NISSENBAUM, PRIVACY IN CONTEXT: TECHNOLOGY, POLICY, AND THE INTEGRITY
OF SOCIAL LIFE 2-3 (2009).
61. See, e.g., DANIEL J. SOLOVE, THE DIGITAL PERSON: TECHNOLOGY AND PRIVACY IN THE
INFORMATION AGE 9 (2004) (decrying the inadequacy of existing law protecting information
privacy in response to the emergence of digital dossiers).
62. Daniel J. Solove, ConceptualizingPrivacy, 90 CAL. L. REV. 1087, 1090-91 (2002).
63. E.g., Orin S. Kerr, The Fourth Amendment and New Technologies: Constitutional Myths
and the Casefor Caution, 102 MICH. L. REV. 801, 864-67 (2004) (charting Fourth Amendment
treatment of various technological developments).
64. See id. at 868-75 (contrasting the institutional competence of the Legislature and Judiciary
in addressing the implications of new technology on these areas of law).
1696 Texas Law Review [Vol. 88:1685

have contributed to the rise of what they call the "National Surveillance
State," characterized by a significant increase in the amount of intelligence
and surveillance the government conducts in the name of protecting national
security. 65
The message from the overwhelming weight of legal scholarship about
technology, privacy, and national security recommends subtlety and nuance,
yet when Congress embraces uncritically the principle of tech neutrality, it
abandons subtlety and nuance in the name of consistency.
Consider the ongoing Fourth Amendment debate over the search of
laptops at international borders. The Supreme Court has held that govern-
ment agents at international borders can conduct a wide range of
suspicionless searches without violating the Fourth Amendment because of
the need to protect American sovereignty and because people crossing bor-
ders should and usually do expect less privacy. 66 Scholars have debated
whether this rule should extend to files stored on laptops being carried across
the border.6 7
Civil liberties groups argue that laptops are special technologies that
merit special treatment under the Fourth Amendment at the border.68
Because laptops store vast amounts of information and because the
information can be of a highly personal nature, laptops become extensions of
the self, more akin to a home than a pad of paper in a traveler's backpack.6 9
Former Bush Justice Department official, now law professor, Nathan
Sales disagrees, arguing that the "principle of technological neutrality" de-
mands a rule that treats pads of paper and laptops consistently. 70 But
Professor Sales errs if he means to invoke a freestanding principle of
neutrality, one that must be "violated" only with good justification. The only
principle Congress should invoke is this one: Treat similar technologies alike
and differing technologies differently. Arguing that a technology is not suf-
ficiently different to outweigh a principle of neutrality is to double count.
To be fair, Professor Sales relies not only on the principle of neutrality;
he also compares the privacy risks from searches of laptops to searches of

65. Jack M. Balkin & Sanford Levinson, The Processes of Constitutional Change: From
Partisan Entrenchment to the National Surveillance State, 75 FORDHAM L. REV. 489, 520-22
(2006).
66. See United States v. Montoya de Hernandez, 473 U.S. 531, 538 (1985) ("Routine searches
of the persons and effects of entrants are not subject to any requirement of reasonable suspicion,
probable cause, or warrant .... Automotive travelers may be stopped at fixed checkpoints near the
border without individualized suspicion even if the stop is based largely on ethnicity .... (citations
omitted)).
67. E.g., Symposium, The Fourth Amendment at the International Border, 78 MiSS. L.J. 241
(2008).
68. Brief for Amici Curiae Ass'n of Corporate Travel Executives & Electronic Frontier
Foundation in Support of Defendant-Appellee at 4, United States v. Arnold, 523 F.3d 941 (9th Cir.
2008) (No. 06-50581).
69. Id. at 11-12.
70. Sales, supra note 36, at 1115.
2010] The Argument Against Technology Neutrality 1697

"letters, address books, photo albums, and similar items." 71 The comparison,
however, should be the entire analysis; the invocation of a principle should
add nothing.

B. Technological Change

Tech-neutral laws too often avoid the problem of underinclusiveness by

permitting overinclusiveness. They expand to cover new technologies and
new circumstances. Consider the Communications Assistance for Law
Enforcement Act 72 (CALEA), a law that requires telecommunications
providers to design their systems to be readily wiretappable to accommodate
lawful government requests for access to customer communications.73
CALEA is a tech-neutral law, one directed at "telecommunications
carrier[s]" that governs what they must do with "equipment, facilities, or
services" that can be used by a customer to "originate, terminate, or direct
74
communications.
When CALEA was enacted in 1994, both the Justice Department, which
pressed for the law, and Congress focused mostly on problems associated
with digital telephone networks. 75 Although the Internet was growing in im-
portance at the time, almost all of the attention in hearings and committee
reports centered on how digital telephone switches were foiling lawfully au-
thorized wiretaps. 76 Motivated by such a tech-specific fear, Congress could
have written a tech-specific law, one focused on digital telephony or perhaps
even one that cited particular protocols or products by name. Instead,
Congress wrote a tech-neutral law.
As we should have anticipated, tech-neutral CALEA has expanded over
time. In 2005, the Federal Communications Commission (FCC), using
power delegated to it in CALEA, granted the Justice Department's petition to
apply CALEA to providers of broadband-Internet and interconnected-Voice-
over-IP (VoIP) services. 77 The FCC came to this conclusion over the objec-
tions of privacy groups and affected service providers, most vocally groups
representing libraries and universities that worried they would be required to

71. Id.
72. 47 U.S.C. §§ 1001-1021 (2006).
73. Id. § 1002(a).
74. Id. Under the law's definitions, a "telecommunications carrier" is "a person or entity
engaged in the transmission or switching of wire or electronic communications as a common carrier
for hire," id. § 1001(8)(A), but excludes "information services," id. § 1001(8)(C), those that
.,offer[]... a capability for generating, acquiring, storing, transforming, processing, retrieving,
utilizing, or making available information via telecommunications," id. § 1001 (6)(A).
75. See, e.g., H.R. REP. No. 103-827, at 14 (1994), reprintedin 1994 U.S.C.C.A.N. 3489, 3492
(calling on Congress to respond to the "'digital telephony' revolution").
76. See id. at 10-16, reprinted in 1994 U.S.C.C.A.N. 3490-96 (summarizing the hearings on
CALEA).
77. In re Commc'ns Assistance for Law Enforcement Act and Broadband Access to Servs., 20
F.C.C.R. 14,989, 14,989 (2005).
1698 Texas Law Review [Vol. 88:1685

include surveillance backdoors in their networks.7 8 Ultimately, the D.C.

Circuit rejected the challenges to the rulemaking.7 9
Even one who agrees with this interpretation of the language of CALEA
should concede that Congress did not say much about VolP and broadband
Internet when it considered whether to enact CALEA. When a tech-neutral
law like CALEA expands over time, it loses its tether to the evidence
Congress considered, the experts consulted in hearings, and the pages of
research compiled into committee reports.

C. Imprudent Delegation
Of all of the arguments that support tech neutrality, the most important
and the most flawed is the argument about institutional competence.
Although Congress may sometimes have difficulty understanding the subtle
nuances of technology or national security, a tech-neutral surveillance law
rarely delegates Congressional power to an expert agency better equipped to
understand such complexities. Instead, such a law almost always delegates
power solely to the Executive Branch, which is often no better situated than
Congress to understand such complexities. 80 When Congress switched from
regulating "numbers dialed" to "dialing, routing, addressing, and signaling
information,' ' 81 it surrendered its role in future discussions about evolving
technology because a tech-neutral law always expands with changing
technology, placing the power entirely in the White House, NSA, and Justice
Department.82

D. How to Decide Between Neutrality and Specificity

Thus, every argument that supports the principle of tech neutrality can
be met with a strong counterargument. We should never again treat legisla-
tive tech neutrality as a principle, default choice, or presumption; it is merely
one of two paths we might take, and whether it is the right path depends on
many circumstances. For example, to choose between tech neutrality and
tech specificity, legislators need to understand how the technologies work,
have been deployed, and have been used. Only by gathering accurate and

78. See, e.g., Final Brief for Petitioners at 43-44, Am. Council on Educ. v. FCC, 451 F.3d 226
(D.C. Cir. 2006) (No. 05-1404) (arguing, with libraries and universities among the petitioners, that
the FCC's interpretation would force private broadband providers to comply with surveillance-
capability requirements).
79. See Am. Council on Educ., 451 F.3d at 232-36 (rejecting petitoners' claims that Internet
broadband and VolP services classify as "information services" under CALEA).
80. See infra subpart Ill(A).
81. See supra notes 11-12 and accompanying text.
82. See Kenneth A. Bamberger & Deirdre K. Mulligan, Privacy Decisionmaking in
Administrative Agencies, 75 U. CHI. L. REV. 75, 76, 89 (2008) (discussing privacy-impact-
assessment requirements that apply to new technology and privacy risks and also highlighting
increased involvement by the White House in administrative action).
20101 The Argument Against Technology Neutrality 1699

complete information about these topics, can legislators decide, for example,
whether to treat two technologies alike or differently.
Unfortunately, in the debates surrounding the PAA and FISA
Amendments Act, Congress might not have had accurate information about
these critical circumstances, because according to some nongovernmental
observers, Executive Branch officials had painted a misleading picture about
the critical factual claim that "almost all transoceanic communications were
[satellite] radio communications., 83 This factual statement supported every
single Executive Branch argument for making FISA tech neutral, thus serv-
ing as the foundation for Congress's decision to expand the surveillance
power under FISA.
At the time the Executive Branch was making this factual claim, Kate
Martin and Lisa Graves of the Center for National Security Studies were re-
butting it in congressional testimony:
[E]ven a general examination of telecommunications history
reveals that the scenario [administration officials] posit claiming that
virtually all international calls of Americans were via satellite radio
and therefore intended to be obtained by the government is not
accurate. While satellites were increasingly used in the 1970s for
television broadcasting and some telecommunications, American
telephone companies were continuing to rely on trans-oceanic cables
for international calls, with newer transatlantic cables sunk even the
84
year after FISA passed ....
The pair concluded, "A more accurate statement than the
administration's description would be that for [the] past 29 years, US
telecommunications has relied on both wire and radio technology for domes-
tic and international calls. 85
These conclusions were corroborated by David Kris, writing at the time
as a private citizen but now the Assistant Attorney General (AAG) for
National Security in the Obama Administration. Mr. Kris rebutted the
Administration's claims about the evolution from satellite to wireline
communications, finding them "exaggerated," because "inand around 1978,
transoceanic communications were made in relatively large quantities by
both satellites (radio) and coaxial cables (wire); both kinds of systems were
expected to continue in service for many years; and the use of fiber optics
86
was already anticipated for undersea cables.

83. Modernization of FISA, supra note 27, at 29 (statement of Kenneth L. Wainstein, Assistant
Att'y Gen. of National Security, United States Department of Justice).
84. Modernization of FISA, supra note 27, at 195 (statement of Kate Martin, Director, and Lisa
Graves, Deputy Director, Center for National Security Studies).
85. Id.
86. David S. Kris, Modernizing the Foreign Intelligence Surveillance Act 9 (Counterterrorism
& Am. Statutory Law Series, Working Paper No. 1, 2007), available at
http://www.brookings.edul-/media/Files/rc/papers/2007/1115_nationalsecurity.kris/ 11i5_nationals
ecurity-kris.pdf.
1700 Texas Law Review [Vol. 88:1685

In other words, according to AAG Kris, the core factual premise

underlying the Executive Branch's argument for tech neutrality might have
been exaggerated. This might mean, as Ms. Martin and Ms. Graves argued
and contrary to the Executive Branch's claims, that when Congress enacted
FISA in 1978, it had good reason to treat radio and wireline communications
differently.87 Perhaps those reasons still merited inconsistent treatment in
2007 and 2008. My research has not yet confirmed which of the two ver-
sions of history should be believed. My point here is merely to suggest that
one reason Congress might have failed to do a better job untangling these
inconsistent histories is because it might have failed to see the importance of
the inquiry, once it placed too much stock in the principle of tech neutrality.

III. The Argument for Tech-Specific Surveillance Laws

To this point, I have offered only arguments that challenge
unchallenged claims for tech neutrality. There is no freestanding principle of
tech neutrality, and arguments to shift from a specific to a neutral rule should
be weighed on their own merits. But rejecting tech neutrality is not the same
thing as defending tech specificity. Policy makers should take care not to
make the same type of mistake in favor of tech specificity I have argued the
proponents of tech neutrality have made; treating tech specificity as a
freestanding principle is as bad as doing so with tech neutrality.
In this final Part, however, I will try to make that argument without
making that mistake, giving reasons to often favor tech-specific laws over
tech-neutral ones for surveillance. The most important reason was intro-
duced in Part II: Tech-specific rules check the Executive Branch by
authorizing narrow and circumscribed new forms of surveillance, permitting
the Executive Branch the freedom to act with the Legislature's blessing, but
only for a particular type of technology. We should prefer the active partici-
pation in surveillance decision making of two branches of government rather
than one.
In order to embrace tech specificity, however, we need to deal with two
practical difficulties, neither insurmountable. First, tech-specific laws expire
as people switch from using the specified technology to using a replacement
technology, leaving us adrift without legislative guidance. This would be
unacceptable if it permitted either unchecked surveillance or untraceable
crime or terrorism, but neither extreme is likely thanks to what I call the
"background rules of surveillance." 88 Second, once law makers decide to
create a tech-specific rule, they must decide how specific to make the rule,
requiring a difficult textual balancing act.

87. See Modernization of FISA, supra note 27, at 195 (statement of Kate Martin, Director, and
Lisa Graves, Deputy Director, Center for National Security Studies) ("[F]or [the] past 29 years, US
telecommunications has relied on both wire and radio technology for domestic and international
calls. From the beginning, FISA was written to accommodate that reality.").
88. See infra section IIH(B)(1).
2010] The Argument Against Technology Neutrality 1701

Finally, tech-specific rules serve one unappreciated benefit: they sunset

when new technologies are introduced. A law that governs only the use of a
telephone, for example, will not govern the use of the Internet. Technology
sunsets should be viewed as significant improvements over the traditional
time-based sunsets that Congress seems to favor for surveillance laws lately.
Technology sunsets enjoy many of the benefits and few of the downsides of
their traditional counterparts. For all of these reasons, Congress should con-
sider drafting tech-specific surveillance laws much more often than they
have.

A. Why We Should PreferSpecificity

Sometimes Congress should delegate its authority to experts-to those
with relative institutional advantages-but history has taught us to doubt that
surveillance is a proper situation for delegation. The Executive Branch sees
only one side to debates between security and privacy, and it tends to expand
its authority and decrease oversight at every step. History has proven this
repeatedly, from the well-documented wiretapping abuses at the FBI under
J. Edgar Hoover, 89 to the intelligence abuses at the CIA that led to the Church
Committee9" and the enactment of FISA, 91 to the NSA's Terrorist
Surveillance Program,92 and to abuses of the national security letter process
at the FBI. 93 The modem surveillance state needs information, and left with-
out proper oversight, the analysts and agents in the field always seem to
choose the path to more information and fewer administrative hurdles. 94 The
Executive Branch, especially one bent on finding hidden terrorists, has
shown that it cannot be trusted to act unchecked.95
The Legislative Branch also brings another institutional advantage over
the Executive Branch. The Executive Branch, especially the NSA, shrouds

89. See WHITFIELD DIFFIE & SUSAN LANDAU, PRIVACY ON THE LINE 163-64 (updated &
expanded ed. 2007) (detailing the wiretapping of seventeen people for political purposes during the
Nixon administration); Robert Bloom & William J. Dunn, The Constitutional Infirmity of
Warrantless NSA Surveillance: The Abuse of PresidentialPower and the Injury to the Fourth
Amendment, 15 WM. & MARY BILL RTS. J. 147, 148-52 (2006) (comparing President Bush's
warrantless wiretapping to Nixon's extensive wiretapping).
90. S. REP. No. 94-755, at 24 (1976).
91. See S. REP. NO. 95-701, at 5 (1978), reprinted in 1978 U.S.C.C.A.N. 3973, 3973-74
(providing a history of FISA and attempting to "make more explicit the statutory intent, to provide
further safeguards for individuals subjected to electronic surveillance").
92. James Risen & Eric Lichtblau, Bush Lets U.S. Spy on Callers Without Courts, N.Y. TIMES,
Dec. 16, 2005, at Al.
93. See Dan Eggen, FBI Found to Misuse Security Letters, WASH. POST, Mar. 14, 2008, at A3
(describing the FBI's use of national security letters to obtain personal data from U.S. citizens rather
than foreigners).
94. See Kevin Poulsen, FBISeeks Internet Telephony Surveillance, SECURITY Focus, Mar. 27,
2003, http://www.securityfocus.com/news/3466 (detailing a request by the FBI and the Justice
Department to require companies to make technical changes making eavesdropping easier).
95. See, e.g., Risen & Lichtblau, supra note 92 (detailing how President Bush allowed the
United States to monitor phone calls without court intervention).
1702 Texas Law Review [Vol. 88:1685

its entire operations in secrecy. 96 Although the Legislative Branch deals with
national security matters through classified hearings, select committees, and
security clearances, its members are all quintessentially public figures who
probably think more about the public's interest than a typical, nameless
Executive Branch analyst.
Thus, the Legislative Branch should not delegate away its checking
power. But that is precisely what it does when it writes a tech-neutral
surveillance law.

B. Implementation
Before we can embrace tech specificity wholeheartedly though, we need
to address two important implementation challenges. First, tech-neutral laws
have one clear advantage over tech-specific laws-longevity. A tech-
specific law applies only so long as people use the specific technology, and
when people shift to using other, newer technology, we are left with
uncertainty. The good news is that surveillance tends to be governed by
good enough background rules. Second, legislators drafting a tech-specific
law will struggle to set the proper level of specificity, and below I set out
some rules of thumb.

1. Background Rules.-Tech-specific laws, by definition, do not

expand or shift with every advance in technology; instead they expire as
technology progresses, sometimes quickly and sometimes gradually. The
expiry of an important surveillance law may seem like catastrophe, deregu-
lating both surveillance and privacy protection, permitting either
undetectable crime and terrorism, unchecked surveillance, or worse, both.
These worst-case scenarios should not worry us, however, once we recognize
that surveillance and privacy tend to be protected by important background
rules that step in to fill the void when statutes do not.
At the outset, note a seeming irony: background rules tend to be tech-
neutral rules. 97 Background rules apply when tech-specific rules expire pre-
cisely because they are not tied narrowly to a particular technology. Thus,
although this Article argues against tech-neutral statutes, it cannot dismiss
tech neutrality entirely. Without tech-neutral background rules, we would
not be able to enact tech-specific laws.98
The most important source for background surveillance rules is the
Fourth Amendment to the U.S. Constitution.9 9 The Fourth Amendment sits

96. See JAMES BAMFORD, THE PUZZLE PALACE 357 (1983) (describing NSA's informal
nickname, "No Such Agency").
97. See infra note 103 and accompanying text.
98. I thank Joe Feller for suggesting this point.
99. The Fourth Amendment provides,
The right of the people to be secure in their persons, houses, papers, and effects,
against unreasonable searches and seizures, shall not be violated, and no Warrants shall
2010] The Argument Against Technology Neutrality 1703

in the background because the Supreme Court interprets it in a generally

tech-neutral manner, but this has not always been the case. Before Katz v.
United States,100 the Supreme Court construed the Fourth Amendment with
tech specificity, for example, distinguishing between "spike mike" recording
devices that intruded physically into the offices of the people being moni-
tored and those that did not. 10 1 The seeming hyperspecificity of this rule
prompted the D.C. Circuit to note that it was "unwilling to believe that the
respective rights are to be measured in fractions of inches."' 10 2 Beginning
with Katz, however, the Court has construed the Amendment more neutrally,
asking whether new forms 10 3
of surveillance invade a person's "reasonable ex-
pectation of privacy."'
A neutral Fourth Amendment is necessary but not sufficient to serve as
an appropriate tech-neutral background for tech-specific surveillance
statutes. The Fourth Amendment must also avoid extreme conclusions-
absolute prohibitions or permissions for new surveillance techniques. If the
Fourth Amendment's default background rule for surveillance were an ab-
solute prohibition on the use of new surveillance technologies, then the
Intelligence Community would lose access to information, and in the worst
case, it would lose track of those trying to harm us. On the other hand, if the
Fourth Amendment's rule were absolute permission, meaning any unregu-
lated surveillance technology could be used to its fullest extent with no
possibility of review, then we would end up with far too many invasions of
privacy than we are willing to tolerate. Either result would be unacceptable.
The good news is that the Fourth Amendment's background rules for
surveillance almost never sit at either extreme. Instead, the Fourth
Amendment tends to operate somewhere in the middle, thanks to a feature of
its jurisprudence that is never celebrated by scholars-its lack of clarity.
To quote the first line of Anthony Amsterdam's seminal article, "For
clarity and consistency, the law of the fourth amendment is not the Supreme

issue, but upon probable cause, supported by Oath or affirmation, and particularly
describing the place to be searched, and the persons or things to be seized.
U.S. CONST. amend. IV.
100. 389 U.S. 347 (1967).
101. Compare Goldman v. United States, 316 U.S. 129, 133-34 (1942) (holding that the use of
a spike mike that did not enter the apartment was not a search), with Silverman v. United States, 365
U.S. 505, 509 (1961) (holding that the use of a spike mike that made contact with an apartment
baseboard was a search).
102. Silverman v. United States, 275 F.2d 173, 178 (D.C. Cir. 1960), rev'd, 365 U.S. 505
(1961).
103. Katz, 389 U.S. at 361 (Harlan, J., concurring). Despite the apparent neutrality of the
reasonable-expectation-of-privacy test, the Court still seems to treat different technologies
differently. See Orin S. Kerr, Digital Evidence and the New Criminal Procedure, 105 COLUM. L.
REV. 279, 307 (arguing that although Katz "was designed ...to achieve some kind of technology
neutrality within search and seizure law,... its impact on the law has been surprisingly narrow").
1704 Texas Law Review [Vol. 88:1685

Court's most successful product."' 0 4 Other scholars have complained that

"the Court has produced a series of inconsistent and bizarre results that it has
left entirely undefended.' '0 5 But for a background rule, inconsistency has a
silver lining.
The muddiness of the Supreme Court's rule causes intelligence agents
(and even more so their lawyers) to hesitate before charging ahead. As Carol
Rose has said in praising muddy rules in property law, "When a court intro-
duces ambiguity into the fixed rules that the parties initially adopted, it in
effect reinstates the kind of weighing, balancing, and reconsidering that the
parties might have undertaken if they had been in some longer term relation-
ship with each other."' 1 6 Because of the Fourth Amendment's muddiness,
rarely should a government lawyer, pressed to analyze some new surveil-
lance technology, tell an agent that he or she should proceed without
worrying about the law.
Specifically, the Supreme Court and the federal courts of appeals have
left unanswered two Fourth Amendment questions that arise in many con-
temporary surveillance situations: How does the Fourth Amendment apply to
the Internet, and how does the Fourth Amendment apply to national security
investigations involving foreign persons? We have only partial answers to
these questions. Smith v. Maryland'0 7 stands for the proposition that govern-
ment surveillance of some of the noncontent characteristics of electronic
communication (specifically, the numbers dialed on a telephone) are not
protected by the Fourth Amendment. 10 8 United States v. District Court
(Keith)'0 9 stands for the proposition that the Fourth Amendment applies to
national security investigations of domestic persons." 0 These cases leave
many important questions unanswered: Are the websites visited in a Web
browser like the numbers dialed on a telephone and thus unprotected under
Smith? Can Keith be extended to cover investigations of foreign persons?
These are important questions that the Court should answer.
But recognize how the confusion over the Fourth Amendment plays a
salutary role in the face of technological uncertainty. Smith provides a cau-
tious green light to some aggressive new forms of surveillance, and Keith
puts up at least a yellow light about national security investigations. The
cases give government lawyers hope that they might be able to permit what
their agents want to do without legislation, especially when the facts are
important enough, but prevent them from charging forward without imposing

104. Anthony G. Amsterdam, Perspectives on the Fourth Amendment, 58 MINN. L. REV. 349,
349 (1973).
105. Silas J. Wasserstrom & Louis Michael Seidman, The Fourth Amendment as Constitutional
Theory, 77 GEO. L.J. 19, 29 (1988).
106. Carol M. Rose, Crystals and Mud in Property Law, 40 STAN. L. REV. 577, 608-09 (1988).
107. 442 U.S. 735 (1979).
108. Id. at 742.
109. 407 U.S. 297 (1972).
110. Id. at 316-17.
20101 The Argument Against Technology Neutrality 1705

some limits and controls on agents, as a hedge against future, adverse inter-
pretations of the Fourth Amendment.
I have made two descriptive claims about the Fourth Amendment: After
Katz, Fourth Amendment rules tend to be tech neutral, and the neutrality of
these rules acts as a safety net, giving Congress the freedom to pass tech-
specific statutes without worrying too much about what happens when the
technology changes. But, turning to the normative, should the Fourth
Amendment's rules be tech neutral, in light of the arguments against tech-
neutral statutes in Part II? If so, then why might we value tech neutrality in
our Constitution but reject it for statutes?
This normative question allows me to wade a bit into an illuminating
debate that occurred between Professors Orin Kerr and Daniel Solove. 111
Although the pair disagreed about many things, they started from a point of
fundamental agreement: both Congress and the courts play important roles in
developing the rules of criminal procedure-Congress by passing the kind of
surveillance statutes discussed throughout the instant Article, and the courts
112 Solove referred to this as a
as interpreters of the Fourth Amendment. ' 113
"dualist system of criminal procedure."
The pair disagreed, however, about which branch we should trust more
to come up with good rules for criminal procedure, especially those designed
to respond to new technology. Kerr argued that the Legislature has compar-
ative institutional advantages over the courts for this task, 1 4 while Solove
wanted courts to play a more aggressive role than they had in the past. 115
Rather than take a side in this debate, I argue that it is good to have both
branches creating rules of criminal procedure. If nothing else, given institu-
tional differences between the branches, they are likely to come to different
conclusions about some surveillance practices, giving us more than one take
on a subject, allowing us to use the different branches as laboratories to play
out different ideas. Best of all, these approaches can support one another,
each doing what the other does not. While the Constitution might serve as
the wellspring of principle and baseline values, the statutes can fill in the
details, policing the specifics of privacy and security. As Professor Kerr

111. The back-and-forth took place in three law review articles. Kerr, supra note 63; Daniel J.
Solove, Fourth Amendment Codification and Professor Kerr's Misguided Call for Judicial
Deference, 74 FORDHAM L. REv. 747 (2005); Orin S. Kerr, Congress, the Courts, and New
Technologies: A Response to ProfessorSolove, 74 FORDHAM L. REv. 779 (2005) [hereinafter Kerr,
Response].
112. See Kerr, supra note 63, at 855 ("A broader look at the legal standards that govern
criminal investigations involving new technologies suggests that Congress has often taken the lead,
and... decisions interpreting the Fourth Amendment generally have played a secondary role. In
some instances, congressional action has followed Supreme Court decisions interpreting the Fourth
Amendment."); Solove, supra note 111, at 753 ("The rules regulating government investigations
have increasingly been those of federal statutes, not Fourth Amendment law.").
113. Solove, supranote 111, at 747.
114. Kerr, supra note 63, at 858.
115. Solove, supra note 111, at 777.
1706 Texas Law Review [Vol. 88:1685

noted, "[W]e should not expect the Fourth Amendment alone to provide ade-
quate protections against invasions of privacy made possible by law
enforcement use of new technologies .... Congress will likely remain the
primary source of privacy protections in new technologies thanks to institu-
tional advantages of legislatures."' 1 6 At the same time, when tech-specific
statutes, with their focus on detail and specifics, fail to apply because of
changes in technology, the Constitution's principles will provide the
bulwark.
But even the advantages of interbranch diversity fail to explain fully
why tech neutrality is so often a bad thing for Congress but not for the courts.
This answer lies in one important institutional difference between the
branches: courts adjudicate on a case-by-case basis, while legislatures design
rules of general applicability." 7 Given this difference, the amount of harm
caused by a bad rule is much higher for legislative rules than judicial rules.
When a legislature misreads the effect on privacy or security of a new tech-
nology or makes a bad prediction about the evolution of a future technology,
the flawed general rule it creates as a result will apply broadly and will be
hard to reverse. After enacting the rule, Congress will likely pay less atten-
tion to the question, making it hard for it to detect the effects of the bad rule.
Further, in order to reverse the bad rule, Congress will need to muster the
political will it needs to pass an amendment or repeal.
In contrast, when a judge crafts a rule based on a misreading of
technology, it directly impacts only the parties in one case. In subsequent
cases, judges applying the bad rule will have an opportunity to see how it
applies to a new set of facts, which might expose the rule's flaw. Law en-
forcement agencies or criminal defendants who disagree with the rule will
have both the incentive and the opportunity in later cases to point out prob-
lems and to argue why the rule should be narrowed or reversed.
In addition to the Fourth Amendment, a second set of "rules" similarly
sits somewhere between prohibition and permission, although it might seem
odd to call these rules. They flow from the increasing intermediation of
communications networks. In the early twentieth century, telephone and
telegraph networks carried communications in the form of simple, easily
captured analog signals, and surveillance targets tended to communicate
from fixed locations like stationary landline telephones." 18 On such simple
analog networks, the government could conduct surveillance often without
the help of an intermediary, for example attaching alligator clips to a wire

116. Kerr, supranote 63, at 838.

117. Id. at 884.
118. See K. A. Taipale, Whispering Wires and Warrantless Wiretaps: DataMining and Foreign
Intelligence Surveillance, N.Y.U. REV. L. & SECURITY, 7 Supp. BULL. ON L. & SECURITY, Spring
2006, at 3, available at http://ssrn.com/abstract=-889120 ("When FISA was being drafted it made
sense to speak exclusively about the interception of a targeted communication-one in which there
were usually two known ends and a dedicated ('circuit-based') communication that could be
'tapped. '").
2010] The Argument Against Technology Neutrality 1707

19
atop a telephone pole or in an office building's basement. Things are
much more complicated today. Digital packets have replaced analog signals,
surveillance targets can access their e-mail accounts or use their cell phones
from any place, and intermediaries
0
can track communications that would
have been untrackable before.12
Now that the government needs help from private parties to conduct
new forms of surveillance,' 21 a second background rule operates. Large
corporate telecommunications providers worry about being sued by their
customers for assisting the government. They worry especially about re-
quests for novel forms of surveillance that may be inconsistent with specific
122
congressional authority or at least unaccompanied by judicial order.
Sometimes, providers overcome this reluctance, as when telephone and
Internet providers complied with Bush Administration requests for assistance
following 9/11.123 Despite the pressure to cooperate with such requests,
however, some providers have resisted government requests that they have
felt might contradict the law.124 Like the Fourth Amendment's muddy rules,
intermediary risk aversion and exposure to liability leads to moderation.
Nervous intermediaries will resist overly aggressive, broadly worded, or in-
completely authorized new forms of surveillance, but they will also bend to
the will of law enforcement and the Intelligence Community when a case
seems important or urgent enough, as in the days following 9/11.
Both of these sets of background rules, Fourth Amendment rules and
intermediary conservatism, help prevent the worst scenarios after a tech-

119. See Olmstead v. United States, 277 U.S. 438, 456-57 (1928) (describing the government's
means of wiretapping as inserting small wires along ordinary telephone wires).
120. See Taipale, supra note 118, at 1 (describing FISA's inadequacy in addressing new
technological developments).
121. Kenneth R. Logsdon, Note, Who Knows You Are Reading This? United States' Domestic
Electronic Surveillance in a Post-9/ll World, 2008 U. ILL. J.L. TECH. & POL'Y 409,419 (discussing
the government's use of the private telecommunications industry in a new surveillance program).
122. See H.R. REP. No. 99-690, at 15-16 (1986), reprinted in 1986 U.S.C.C.A.N. 5327, 5341-
42 (noting that private parties are concerned with issues of liability when cooperating with FBI
investigations); Albert Gidari, Jr., Keynote Address at the University of San Francisco Law Review
Symposium: Companies Caught in the Middle (Oct. 28, 2006), in 41 U.S.F. L. REv. 535, 546-47
(2007) (describing cell-phone providers resisting requests for location-tracking information).
123. See Gidari, supra note 122, at 541 ("September 11 ... changed a lot of things for service
providers."); Leslie Cauley, NSA Has Massive Databaseof Americans' Phone Calls, USA TODAY,
May 11, 2006, available at http://www.usatoday.com/news/washington/2006-05-10-nsax.htm
(asserting that AT&T, Verizon, and Bellsouth all furnished the NSA with customer records after
9/11); Risen & Lichtblau, supra note 92 (describing a massive, warrantless monitoring effort made
on thousands of international phone calls and e-mails from people inside the United States).
124. See Cauley, supra note 123 ("Among the big telecommunications companies, only Qwest
has refused to help the NSA .... Qwest declined to participate because it was uneasy about the
legal implications of handing over customer information to the government without warrants.");
Katie Hafnier & Matt Richtel, Google Resists US. Subpoena of Search Data, N.Y. TIMES, Jan. 20,
2006, at AI (describing Google's refusal to comply with a broad subpoena for copies of its search-
query records); Declan McCullagh, DOJ Abandons Warrantless Attempt to Read Yahoo E-mail,
CNET NEWS, Apr. 16, 2010, http://news.cnet.com/8301-13578_3-20002722-38.html (describing
Yahoo's refusal to comply with a court order for evidence in a criminal investigation).
1708 Texas Law Review [Vol. 88:1685

specific rule lapses-unchecked permission or absolute prohibition. The

background rules, therefore, should give Congress the reassurances it needs
to build narrowly crafted tech-specific rules without worrying about chaos
after the new law expires. At the same time, because the Fourth Amendment
and intermediary cautiousness lead inherently to doubt and conservatism,
these rules must usually stay in the background only, and Congress should
eventually regulate to replace laws that expire.

2. How Specific?-After identifying and weighing background rules, if

Congress chooses to enact a tech-specific law, it next needs to describe the
technology at the proper level of specificity. Congress should strive to write
statutes that talk about technology specifically enough to allow for the bene-
fits of tech specificity but generally enough to prevent the need to revisit the
statute every six months.
Striking the balance between breadth and specificity can be difficult.
To start, Congress should look at the specific technology or technologies that
motivated it to act. Perhaps a news story or anecdote about a specific type of
surveillance technology brought the issue to Congress's attention. For
example, consider the barrage of media attention paid in late summer, 2000,
to Carnivore. 125 Carnivore was the name given by the FBI to a packet- 126
sniffing-and-filtering device that could be used to track Internet behavior.
Although the tool was originally vilified in the press and by privacy
groups,127 with the benefit of time, this criticism seems a bit mistargeted.
According to several scholars, the tool was used only with a court order and
only when an Internet Service1 Provider
28
(ISP) lacked the expertise to conduct
the ordered surveillance itself
Nevertheless, in 2000, Congress expressed concern and outrage over
Carnivore. Within weeks of the first news reports, Congress held hearings in
which members criticized Justice Department and FBI officials for having

125. See, e.g., Neil King Jr. & Ted Bridis, FBI's Wiretaps to Scan E-mail Spark Concern,
WALL ST. J., July 11, 2000, at A3 (describing how Carnivore is "[e]ssentially a personal computer
stuffed with specialized software [and] represents a new twist in the federal government's fight to
sustain its snooping powers in the Internet Age").
126. See Trenton C. Haas, Note, Carnivoreand the Fourth Amendment, 34 CONN. L. REV. 261,
271-73 (2001) (providing a detailed description of Carnivore).
127. See, e.g., Ted Bridis & Neil King Jr., CarnivoreE-mail Tool Won't Eat Up Privacy,Says
FBI, WALL ST. J., July 20, 2000, at A28 (discussing concerns about invading the privacy of
Americans not under investigation for crimes).
128. See, e.g., Patricia L. Bellia, Surveillance Law Through Cyberlaw's Lens, 72 GEO. WASH.
L. REv. 1375, 1445 (2004) (describing Carnivore as a "tool the FBI developed to overcome
difficulties service providers had in isolating and delivering the contents of electronic
communications or addressing or routing information in response to court orders"); Orin S.Kerr,
Internet Surveillance Law After the USA PATRIOT Act: The Big Brother That Isn't, 97 Nw. U. L.
REv. 607, 653 (2003) (noting that, at times, "ISPs lack the expertise or the willingness to implement
court orders on law enforcement's behalf').
2010] The Argument Against Technology Neutrality 1709

developed the device. 129 Congress turned this criticism and concern into
legislation, including in the USA PATRIOT Act a little-discussed provision
specifically targeting Carnivore and similar tools. Section 216 of the Act
obligates law enforcement to file a sealed report with a court when it uses
tools like Carnivore. 130 Congress did not refer to Carnivore by name, choos-
ing instead to refer to any
1 31
"pen register or trap and trace device on a packet-
switched data network."
This law seems appropriately tech specific, but consider other paths
Congress might have taken. One year after the USA PATRIOT Act, with
lingering fears about Carnivore on its mind, Congress passed another new
reporting law, one which required much more detailed reporting while at the
same time being much more narrowly defined. 32 In this new law, Congress
referred specifically to the name and model number given to Carnivore after
the publicity fiasco, DCS-1000. 133 This law required the Attorney General to
provide detailed reports about "the use of the DCS1341000 program (or any
subsequent version of such program)" for two years.
Congress made a mistake drafting such a specific provision.
Surveillance laws should not refer to specific tools by model and version
number, even with the caveat applying the law to "any subsequent version."
While this type of hyperspecificity might make sense for the expert pro-
nouncements of an administrative agency, Congress itself should rarely, if
ever, refer to technology by a specific model number.
But this lesson in overspecificity provides a road map for finding the
right level of generality. For any technology, one can recite a series of de-
scriptions of increasing generality. 35 In the case of Carnivore, we progress
from the most specific-DCS-1000-all the way to the most general-
surveillance software or, even more generally, software. 136 Congress should
avoid both extremes, the former being too specific, the latter too neutral by
its generality. One possible target is to describe the technology at one or two
steps above the most specific level. In this case, perhaps the ideal level of

129. See, e.g., Fourth Amendment Issues Raised by the FBI's "Carnivore"Program: Hearing
Before the Subcomm. on the Constitution of the H. Comm. on the Judiciary, 106th Cong. 107 (2000)
(statement of Rep. Bob Barr, Member, H. Comm. on the Judiciary) (noting that the impact of
Carnivore on the privacy rights of U.S. citizens is "immense").
130. USA PATRIOT Act of 2001, Pub. L. No. 107-56, sec. 216, 115 Stat. 272, 289 (codified at
18 U.S.C. § 3123(a)(3)(A) (2006)).
131. Id.
132. 21st Century Department of Justice Appropriations Authorization Act, Pub. L. No. 107-
273, 116 Stat. 1758 (2002) (to be codified in scattered titles of U.S.C.).
133. Id. sec. 305.
134. Id.
135. Copyright law embraces a similar "abstractions test," first recited by Judge Learned Hand.
Nichols v. Universal Pictures Corp., 45 F.2d 119, 121 (2d Cir. 1930).
136. See Kerr, supra note 128, at 653-54 (discussing Carnivore and "its progeny" the "DCS-
1000" in comparison to other software).
1710 Texas Law Review [Vol. 88:1685

abstraction would be "packet-capture device" or maybe "filtered-packet-

capture device." -
Although the standard outlined in this section is necessarily vague, it
may prove easy to apply. Consider a few other surveillance technologies that
in recent years have sparked the public's imagination: In order to regulate
these technologies, Congress should target "keystroke logging software" but
never "Magic Lantern" (too specific) nor "spyware" (too general); 137 regulate
"heat sensing cameras" rather than the "Agema Thermovision 210" or
cameras;138 and "whole-body scanners" instead of "L-3 Provision" or
"radiation scanners."1 39

C. Technological Sunsets
Because tech-specific laws expire when technology changes, we can
think of them as alternatives to traditional sunset provisions-legislative
enactments that expire after a set period of time. In the surveillance context,
140
Congress has enacted a number of sunset provisions in the past decade.
Tech-specific laws and laws with sunsets have much in common. Jacob
Gersen, who has written frequently about sunset provisions, 14 1 gives three
reasons legislators enact sunset provisions: to offset information
asymmetries, reduce error costs in the face of uncertainty, and correct limits
of cognitive bias.1 42 Tech-specific provisions can also satisfy these three
roles, by helping offset the doubt and uncertainty legislators have about the
evolution of technology.
For example, imagine that a legislative proposal authorizing a new form
of surveillance has a little less than a majority of Congress in support and a
vocal contingent fiercely opposed. To help muster the few more votes they
need, proponents of the bill might offer a traditional time-limited sunset
provision, expiring say in four years. This serves two purposes: it helps

137. See Ted Bridis, FBIDevelops Eavesdropping Tools, WASH. POST, Nov. 22, 2001, at A15
(describing the FBI's "Magic Lantern" technology that "would allow investigators to secretly install
over the Internet powerful eavesdropping software that records every keystroke on a person's
computer").
138. See Kyllo v. United States, 533 U.S. 27, 29 (200i) (discussing whether the use of an
"Agema Thermovision 210" thermal imager to detect infrared radiation emitting from Kyllo's home
constituted a Fourth Amendment search).
139. See Schwartz, supra note 4 (debating the use of screening technologies that can show the
contours of the body and reveal foreign objects in reference to risks of privacy invasion).
140. See, e.g., Protect America Act of 2007, Pub. L. No. 110-55, sec. 6(c), 121 Stat. 552, 557
(to be codified at 50 U.S.C. § 1803) (setting a 180-day sunset on select provisions of the Act).
141. See, e.g., Jacob E. Gersen, Temporary Legislation, 74 U. CHI. L. REV. 247 (2007)
(analyzing the "historical, legal, and political implications of temporary legislation"); Jacob E.
Gersen & Anne Joseph O'Connell, Deadlines in Administrative Law, 156 U. PA. L. REV. 923 (2008)
(discussing the use of deadlines to control the timing of administrative agency actions); Jacob E.
Gersen & Eric A. Posner, Timing Rules and Legal Institutions, 121 HARV. L. REV. 543 (2007)
(investigating constitutional, statutory, and internal congressional rules that affect the timing of
legislative and executive actions).
142. Gersen, supra note 141, at 248.
2010] The Argument Against Technology Neutrality 1711

convince undecided members to support the bill, by guaranteeing them a

second vote in the near future, and it dampens the intensity of the opposition,
who might fight less forcefully if they are guaranteed a future opportunity to
kill the law. But the bill's proponents should recognize another way they
might save the bill, by changing tech-neutral provisions into tech-specific
provisions. If undecided and opposition law makers recognize that a tech-
specific provision also expires at some point in the future, they may treat it
the way they treat a traditional sunset.
More importantly, tech-specific laws overcome a significant limitation
of ordinary sunsets. By "expiring" not according to an arbitrary timetable
but instead precisely when changes in technology give reason to reopen pol-
icy debates, tech-specific laws offer the benefits of sunset without some of
the downsides. To understand the relative advantages of technology sunsets,
we need to understand some of the more technical details of Gersen's model
as well as some of the model's shortcomings.
Gersen uses a transactions costs-public choice model to compare sunset
legislation to permanent legislation. 43 Legislators must expend "enactment
costs" when they enact or, in the case of a "sunsetted" law, reenact44
legislation, and they must expend "maintenance costs" at all other times.
For example, finding enough votes for passage is an enactment cost, while
beating back an 45 effort to repeal a law after it has been enacted is a
maintenance cost. 1
As Gersen himself concedes, this model, although clarifying, proves
difficult to apply because so much depends on unpredictable circumstances.
How high are enactment costs versus maintenance costs? How much do
legislators discount future enactment costs? Doubts about the answers to
questions like these prevent Gersen from coming to many categorical conclu- 146
sions about the differences between temporary and permanent legislation,
and they probably leave legislators making crude guesses about the effect of
using a sunset or the amount of time to give to a sunset period.
Think of these difficulties as the products of a simple calibration
problem. If a sunset period is set too far in the future, then the law may
persist after the time when legislators would have otherwise wanted to revisit
or even repeal the law. Even worse, if the sunset period is set to expire too
soon, legislators will be forced to expend the costs of reenactment, even
when there is no need for further review or debate. For any piece of

143. Id. at 261-66.

144. Id. at 263-65.
145. Id.
146. See id. at 266 ("While the analysis does not demonstrate that temporary legislation is
clearly less costly than permanent legislation, it does show that temporary legislation is not clearly
inferior-at least along the transaction-cost dimension."). Gersen comes to some tentative
conclusions, for example arguing that "[iut is almost certainly easier to block the repeal of
legislation than to pass new legislation. As a result, continuing permanent legislation is less costly
in the sunset year than reauthorizing temporary legislation." Id. at 264-65.
1712 Texas Law Review [Vol. 88:1685

traditionally sunsetting legislation, there is an ideal but unknowable term of

expiration. The reason the ideal term cannot be known is because of the dif-
ficulty predicting the rate of change of important facts, particularly when
those facts involve evolving technology.
Thinking of this as a calibration problem illuminates why tech-specific
laws are better. A well-written tech-specific law is calibrated to expire pre-
cisely when the most important facts have changed enough to justify a
reevaluation. As an example, consider how the technological shift from the
telephone to the Internet expired an old version of the Pen Register Act at an
optimal time.
The Pen Register Act regulates the government's ability to monitor the 47
so-called envelope information associated with electronic communications. 1
For example, pen-register orders are needed to observe the numbers dialed
by a telephone user) 48 Before the USA PATRIOT Act amended the Pen
Register Act, it referred only to "numbers dialed,"'' 49 which meant it could
expand without congressional reauthorization, but only to a point. As the
telephony state of the art shifted from landline phones to cordless phones to
mobile phones, the Pen Register Act expanded to cover each change, without
wasteful congressional intervention. 50 This seems appropriate: although the
surveillance of a mobile phone raises some issues not raised by the surveil-
lance of a landline telephone, the two technologies seem similar enough to
obviate the need for new congressional deliberation. The tech-specific law
avoids the problem of laws tuned to expire too soon.
But then, people began to communicate over the Internet. Surely
"numbers dialed" did not cover Internet-envelope surveillance, meaning
Congress had to reconsider envelope surveillance as more people began to
embrace this revolutionary new technology. The old technological sunset
had expired. This seems well calibrated. Seen through both the privacy and
law enforcement lenses, monitoring envelope information on the Internet
seems a difference in kind not merely in degree from telephone surveillance.
Precisely when the promise and peril of the Internet came into view,
Congress was thrust back into the conversation. To be sure, great transaction
costs were incurred-the first few times the Justice Department asked for
changes to the Pen Register Act, Congress refused, partly because privacy
advocates pushed back-but after it had time to deliberate fully, and once

147. 18 U.S.C. §§ 3121-3127 (2006).

148. Id. § 3121(a).
149. Electronic Communications Privacy Act of 1986, Pub. L. No. 99-508, sec. 301, § 3126(3),
100 Stat. 1848, 1871.
150. Cf In re Application for Pen Register and Trap/Trace Device with Cell Site Location
Auth., 396 F. Supp. 2d 747, 749, 752-53 (S.D. Tex. 2005) (permitting the use of a pen register to
obtain information from a mobile phone).
2010] The Argument Against Technology Neutrality 1713

spurred by 9/11, Congress granted the new authority as part of the USA
PATRIOT Act. 1 '
Imagine how much less efficient it would have been for Congress to
reevaluate the Pen Register Act on a fixed timetable, no matter what length
of time it chose. If Congress had set the Pen Register Act to expire after four
years, then at the end of the first term in 1990, there would have been very
little to discuss. Communications did not change much in that time period.
Congress would have been forced to expend resources to reenact the bill,
perhaps placing it back under another four-year term, and it probably would
have faced pressure after the first term to switch to a permanent term instead.
The opposite problem might have occurred had the original sunset been set
too far in the future, say ten years. In 1996, at the end of the first term, the
Internet explosion would have been still in its infancy, and it might have
been too soon to discuss an amendment. Then, if Congress had reenacted the
Act with a second ten-year term, it is doubtful that it could have waited until
the second date of expiry, in 2006, to finally get around to extending the Act
to the Internet. Instead, the technological sunset forced a reevaluation at
what seems to have been a near-optimal time: 15 2
five years after Americans be-
gan to adopt the Internet in large numbers.

Conclusion
Conventional wisdom suggests that Congress should write tech-neutral
surveillance laws most of the time. The conventional wisdom has it
backwards. Congress should narrowly target surveillance laws at specific
technologies most of the time. By doing so, it can assert its oversight role
over the Executive Branch, which too often abuses its surveillance power
when it acts unchecked, and shine a light on surveillance abuses.

151. See Beryl A. Howell, Seven Weeks: The Making of the USA PATRIOT Act, 72 GEO.
WASH. L. REV. 1145, 1194-95 (2004) (explaining how the amendments to the Pen Register Act
mirrored those "the Justice Department had suggested for several years" before 9/11).
152. See U.S. CENSUS BUREAU, COMPUTER AND INTERNET USE iN THE UNITED STATES: 2003
(2005), http://www.census.gov/prod/2005pubs/p23-208.pdf (reporting that between the years 1997
and 2000, the percentage of American households with Internet use at home rose from 18% to
41.5%).