OPINION

STEPHEN WM. SMITH, United States Magistrate Judge.

In three separate criminal investigations earlier this month, this court denied the Government's request under the Stored Communications Act (SCA) to compel cell phone service providers to produce cell site information for target cell phones. Among other things, each application sought "records or other information pertaining to subscriber(s) or customer(s), including historical cell site information and call detail records (including any two-way radio feature mode) for the sixty (60) days prior to the date the Order is signed by the Court (but not including the contents of communications)."1 Each application identically defined the requested information as "the antenna tower and sector to which the cell phone sends its signal," specifically including "the cellsite/sector(s) used by the mobile telephone to obtain service for a call or when in an idle state." In other words, the Government seeks continuous location data to track the target phone over a two month period, whether the phone was in active use or not.

This court has previously granted such requests.2 However, recent months have brought to light important developments in both technology and caselaw raising serious constitutional doubts about such rulings. Accordingly I denied these requests, but invited the Government if it disagreed to submit a brief to justify its position with appropriate legal and factual support, which it has now done.3

Five years ago the first reported decisions on government acquisition of cell site information from telephone companies appeared.4 The focus of those early decisions was the appropriate legal standard for obtaining prospective location information under the Electronic Communications Privacy Act (ECPA). Thereafter, a handful of decisions addressed the related problem of law enforcement access to historical cell site data collected and maintained by providers over time. A few courts have held that such requests triggered the Fourth Amendment warrant requirement,5 but most courts to date have granted government access to such information under the SCA, which imposes a less-than-probable cause standard.6

Several weeks ago U.S. Magistrate Judge James Orenstein, who authored the very first cell site opinion, suggested in a new opinion7 that courts re-examine the constitutionality of historical cell site requests in light of recent appellate court decisions, such as that of the District of Columbia Court of Appeals in United States v. Maynard.8 As if to underscore his point, two weeks later the Third Circuit became the first federal appellate court to issue an opinion dealing with government access to historical cell site data.9 Rather than definitively resolving the Fourth Amendment issue, the court remanded the case to the district court, concluding that the factual record was insufficient to resolve whether such records "could encroach upon ... citizens' reasonable expectations of privacy regarding their physical movements and locations."10

Though significant, the caselaw developments have been outstripped by advancing technology. Recently, committees in both the House and Senate have conducted hearings on proposals to update ECPA, the 1986 statute establishing the regulatory regime governing electronic communications. Expert testimony at those hearings reveals that regulatory and market forces have produced dramatic advances in location technology over the past half-decade. As will be shown, this new technology has altered the legal landscape even more profoundly than the new caselaw.

Mindful of the Third Circuit's admonition to base a Fourth Amendment adjudication on an adequate factual record, the court begins with the following findings of fact. These findings are based on judicially noticed facts derived from material contained in the record appendix, including publicly available industry studies, independent surveys, provider policies, and product specifications. The most significant findings are based on expert testimony recently given at a House Judiciary Subcommittee hearing entitled "ECPA Reform and the Revolution in Location Based Technologies and Services." The purpose of this oversight hearing was not to consider a particular bill, but to educate Congress on the current state of location technology in the telecommunications industry, so that needed reforms to the 1986 statute could be identified, drafted, and debated.11 Given that such testimony was not offered for partisan purposes or to advocate specific legislation, the court finds it particularly appropriate for judicial notice under Rule 201 of the Federal Rules of Evidence.12

Findings of Fact

Cell Phone Technology in General

1. Unlike conventional wireline telephones, cellular telephones use radio waves to communicate between the user's handset and the telephone network.13

2. Cellular service providers maintain networks of radio base stations ("cell sites") spread throughout their geographic coverage areas.14

3. A wireless antenna at each cell site detects the radio signal from the handset, and connects it to the local telephone network, the Internet, or another wireless network.15

4. Cell phones periodically identify themselves to a nearby base station as they move about the coverage area, a process called "registration." The registration process is automatic, and occurs whenever the phone is on, without the user's input or control. The registration signal is carried over a channel separate from the channel used to carry the call itself.16

5. During a call, if the phone moves nearer to another base station, the call is "handed off" between base stations without interruption.17

6. No longer just big three-sided radio towers, base station antennas can be mounted outdoors on roof-tops, building-sides, trees, flagpoles, and church steeples, or indoors in homes and offices. Many are no larger than a conventional stereo speaker.18

Wireless Location Technology

7. There are two distinct technological approaches for fixing the location of a cell phone: handset-based (GPS) and network-based (cell site).19

8. GPS is the acronym for Global Positioning System, which is comprised of at least 24 satellites constantly orbiting the earth in six low earth orbits.20

9. For GPS location, special hardware in a user's handset receives signals from at least four global position satellites, allowing the handset to calculate its latitude and longitude whenever it is in unobstructed satellite range.21

10. Current GPS technology can achieve spatial resolution typically within ten meters.22

11. Despite its relative precision, GPS has at least three fundamental drawbacks as a location tool: (a) it is not available for all handset models, especially older models; (b) it works reliably only outdoors, when the handset has an unobstructed view of several GPS satellites in the sky above; and (c) perhaps most significantly, it can be disabled by the user.23

12. For these reasons, GPS is neither the most pervasive nor the most generally applicable phone location system, especially for surveillance purposes.24

13. For network-based location, the position of the phone is calculated by the network based on data collected and analyzed at the cell site receiving the phone's signals, without explicit assistance from the user or his handset.25

14. A variety of techniques may be used for network-based location. The most basic technique is to identify the particular base station (or sector) with which the phone was communicating every time it makes or receives a call and when it moves from one sector to another.26

15. The relative precision of cell site location depends on the size of the cell sector. The smaller the sector, the more precise the location fix.27

16. In early cellular systems, base stations were placed as far apart as possible to provide maximum coverage. At that time, a sector might cover an area several miles or more in diameter. Today this is true only of sparsely populated, rural areas.28

17. Due to a combination of factors, the size of the typical cell sector has been steadily shrinking in recent years.29

18. As the density of cellular users grows in a given area, the only way for a carrier to accommodate more customers is to divide the coverage area into smaller and smaller sectors, each served by its own base station and antenna.30

19. New services such as 3G Internet create similar pressure on the available spectrum bandwidth, again requiring a reduction in the geographic size of sectors.31

20. Another factor contributing to smaller sector size is consumer demand for more reliable coverage in areas with unfavorable radio conditions (e.g., elevators), which again requires additional base stations to cover such "dead spots."32

21. The number of cellular base stations in the U.S. has tripled over the last decade, and the rate of growth is accelerating. By one industry estimate, there are now over 251,000 reported cell sites operating in the United States. There were only 913 the year before ECPA was passed.33

22. The trend toward smaller cell sectors has accelerated with the deployment of smaller-scale base stations designed to serve very small areas such as particular floors of buildings, or individual homes and offices.34

23. This new generation of cellular base station is generally known as a "microcell," and smaller versions are sometimes referred to as a "picocell" or "femtocell."35

24. Microcell technology is increasingly used by many carriers, including AT & T, Verizon, and Sprint. A microcell has a range of 40 feet (12 meters).36

25. The effect of this trend toward smaller sectors is that knowing the base station (or sector ID) handling a call is tantamount to knowing the user's location to within a relatively small geographic area. In urban areas and other environments that use microcells, this area can be small enough to identify individual floors and rooms within buildings.37

26. The decreasing size of cell sectors is not the only factor making network-based location more accurate. New technology allows providers to locate not just the sector in which the phone is located, but also its position within the sector.38

27. By correlating the precise time and angle at which a phone's signal arrives at multiple sector base stations, a provider can pinpoint the phone's latitude and longitude to an accuracy within 50 meters or less. Emerging versions of the technology are even more precise.39

28. Such enhanced location technologies are commercially available, and many carriers contract with specialized companies that provide "off the shelf" location-based products and system upgrades.40

29. Many of these companies were formed in response to directives from Congress and the FCC to develop wireless location technology in order to enhance the nation's emergency response (E-911) system.41

Data Collection and Retention

30. Cell location information is quietly and automatically calculated by the network, without unusual or overt intervention that might be detected by the target user.42

31. Carriers typically create "call detail records" that include the most accurate location information available to them.43

32. Historically, before more advanced location techniques were available, carrier call detail records typically included only the cell sector or base station identifier that handled the call. Today, the base station or sector identifier carries with it more locational precision than it once did.44

33. As even more precise location information becomes available, call detail records can now include the user's latitude and longitude along with the sector ID data. Some carriers also store frequently updated, highly precise, location information not just when calls are made or received, but as the device moves around the network.45

34. The cost of collecting and storing high resolution location data about every customer has become much cheaper in the last few years. Such information is valuable for network management, marketing, and developing new services. This trend toward greater and more extensive data archives is likely to continue.46

35. Some carriers effectively outsource the task of collecting, analyzing, and storing location information to companies offering specialized location technology.47

36. One such company installs multiple auxiliary receivers (called "Location Measurement Units", or LMUs) on existing cell towers and base stations to enhance location accuracy. These auxiliary receivers are very accurately time-synchronized to each other, and very sensitive; at any given moment, a single handset may be in communication with 30 or more LMUs.48

37. This same company has deployed over 100,000 LMU's.49

38. The company not only transmits this detailed location information to the carrier, it can also manage and analyze historic location and calling activity data. Such data can also be organized and aggregated to reflect current user activities, mobile events, and interaction with other devices.50

39. Most carrier systems use a variety of large and small sector configurations. A mobile user, in the course of her daily movements, will periodically move in and out of large and small sectors. The locational precision of cell sector data recording those movements will vary widely over the course of a given day, from relatively less to relatively very precise.51

40. Neither the user nor the carrier can predict how precise the next location data will be. For a typical user, over time, some of that data will likely have locational precision similar to that of GPS.52

41. Given these advances in technology, it is no longer valid to assume that network cell sector records will yield only an approximate user location.53

42. As cellular network technology evolves, the traditional distinction between "high accuracy" GPS tracking and "low accuracy" cell site tracking is increasingly obsolete, and will soon be effectively meaningless.54

Cell Phone Use Statistics

43. Today there are more than 285 million active wireless subscriber accounts in the United States. Many households no longer have traditional "landline" telephone service, opting instead for cellular phones carried by each family member.55

44. Cell phones are frequently used in the home or in other places not open to public view: one study shows that at least 52% of cell phone calls are made indoors;56 another study indicates that two out of three adults sleep with their cell phone nearby.57

45. In 1999, the number of reported wireless minutes of use was less than 200 billion. A decade later, the number has grown to more than 2.2 trillion minutes.58

46. Over the same decade, the annual number of text messages has jumped to 1.56 trillion.59

47. According to a 2008 Nielsen survey, the average U.S. cell phone user made or received 204 voice calls every month.60 A 2010 Pew Research study of adult cell phone use show that the median number of voice calls for a typical user is 5 per day, while the average (mean) is 13.1 calls/day. This study also shows that African American and Hispanic cell users make more calls (and texts) on average than their white counterparts.61

48. Similar patterns are reflected in cell phone texting. The 2008 Nielsen survey reported the average cell phone user made or received 357 text messages a month.62 According to the 2010 Pew Research study, adults send and receive a median of 10 texts daily; the average (mean) is 39.1 texts/day. Both figures are more than double the levels reported by Pew just 8 months earlier in September 2009. Teen use of text messaging is substantially heavier: the teen median level is 50 texts daily, and the mean is 112.4.63

49. Based on these numbers, even if limited to the beginning and end of actual phone calls and text messages, cell site data for a typical adult user will reveal between 20 and 55 location points a day. This data is sufficient to plot the target's movements hour by hour for the duration of the 60 day period covered by the government's request.64

50. If registration data were also collected by the provider and made available, as the Government has requested, such records would track the user on a minute by minute basis, compiling a continuous log of his life, awake and asleep, for a two month period.

Conclusions of Law

A. Under Current Location Technology, Cell Site Information Reveals Non-Public Information About Constitutionally Protected Spaces

The Fourth Amendment guarantees "[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures." At its core, the Fourth Amendment safeguards "the right of a man to retreat into his own home and there be free from unreasonable government intrusion." Silverman v. United States, 365 U.S. 505, 511, 81 S.Ct. 679, 5 L.Ed.2d 734 (1961). "With few exceptions, the question whether a warrantless search of a home is reasonable and hence constitutional must be answered no." Kyllo v. United States, 533 U.S. 27, 31, 121 S.Ct. 2038, 150 L.Ed.2d 94 (2001). A "search" occurs when an expectation of privacy that society is prepared to consider reasonable is infringed. United States v. Jacobsen, 466 U.S. 109, 113, 104 S.Ct. 1652, 80 L.Ed.2d 85 (1984); see also Katz v. United States, 389 U.S. 347, 353, 88 S.Ct. 507, 19 L.Ed.2d 576 (1967).

The refinements in location-based technology detailed in the findings of fact have decisive Fourth Amendment consequences. Perhaps most significantly, they bring cell site location data squarely within the protective ambit of United States v. Karo, 468 U.S. 705, 104 S.Ct. 3296, 82 L.Ed.2d 530 (1984).

Karo considered whether Fourth Amendment rights were violated by law enforcement monitoring of an electronic tracking device known as a "beeper."65 Law enforcement had placed the beeper inside a can of ether to be used to extract cocaine from clothing imported by drug dealers, and then monitored the beeper's movements over a period of several months. At one point the beeper was detected near a commercial storage warehouse; however, the beeper equipment was too imprecise to pinpoint a specific locker, and so such monitoring was held not to intrude upon the suspects' expectation of privacy in their own storage locker. Id. at 720 n. 6, 104 S.Ct. 3296. On another occasion, however, the beeper was monitored while inside a private residence, a place not open to visual surveillance. The Court held that such warrantless monitoring violated the Fourth Amendment, because it revealed "that the beeper was inside the house, a fact that could not have been visually verified." Id. at 715, 104 S.Ct. 3296.

Like the beeper signal from the residence in Karo, cell site information permits the government "to determine by means of an electronic device ... whether a particular article—or a person, for that matter—is in an individual's home at a particular time." Id. at 716, 104 S.Ct. 3296. Over the course of two months, it is inevitable that dozens if not hundreds of calls and text messages of a typical user will be sent from home, office, or other place out of public view. Each of these calls and messages will generate network-based location information, much of it as precise as GPS data.66 Even if no calls or texts were ever made, the phone's presence within the home at a given time would likely be revealed by the automatic registration process.67 This is precisely the type of information that Karo held subject to Fourth Amendment protection; it makes no difference that the electronic device employed is more sophisticated than a traditional beeper.68

Before the Third Circuit, the Government argued that cell site location data no more precise than 200 feet was insufficient to trigger Fourth Amendment protection.69 By the same token, when seeking to acquire GPS or similarly precise location data (often referred to as "E-911 data") from this court, the Government has regularly proceeded by way of a Rule 41 warrant, a tacit concession that such data would reveal facts about a constitutionally protected space.70 Now that cell site technology permits a location fix approaching the precision of GPS—e.g., microcells with a radius of 40 feet71—the technology undergirding the Government's policy has gone the way of VHS and Betamax.

Likewise, court decisions allowing the Government to compel cell site data without a probable cause warrant were based on yesteryear's assumption that cell site data (especially from a single tower) could locate users only imprecisely.72 Given that network-based technology is now capable of isolating a mobile phone user to a particular floor or room within a building, and that such increasingly precise "call detail records" are now kept by providers, the continuing vitality of those decisions must be doubted (with all due respect).

Even if an exact latitude and longitude is not yet ascertainable or recorded for every single mobile call, network technology is inevitably headed there.73 As the Supreme Court observed in Kyllo v. United States regarding the ongoing research and development of radar surveillance devices:

533 U.S. 27, 36, 121 S.Ct. 2038, 150 L.Ed.2d 94 (2001). Like the thermal imaging devices in Kyllo, the cellular location technology in use or development today crosses the "firm but also bright" Fourth Amendment line that the Supreme Court has drawn at the entrance to the house. Id. at 40, 121 S.Ct. 2038. Accordingly, the cell site records generated by that technology are subject to constitutional protection.

B. Historical Cell Site Records Are Subject to Fourth Amendment Protection under the Prolonged Surveillance Doctrine of United States v. Maynard

It is true that cell site records for a single day may not always reveal particularly intimate details about the user's private life but merely that the user's cell phone (like the Karo beeper) was present in the home at a particular time. Nevertheless, as Justice Scalia has observed, "[i]n the home, our cases show, all details are intimate details, because the entire area is held safe from prying government eyes." Kyllo, 533 U.S. at 37, 121 S.Ct. 2038 (emphasis in original). In this case, the records sought by the Government are likely far more intrusive—not a single snapshot at a point in time, but a continuous reality TV show, exposing two months' worth of a person's movements, activities, and associations in relentless detail.

In his decision denying warrantless access to historical cell site information, Judge Orenstein relied most heavily on the recent decision of the Court of Appeals for the District of Columbia in United States v. Maynard, 615 F.3d 544 (D.C.Cir.2010). In light of the technological developments noted above, reliance on the Maynard precedent is not essential to the ruling here. Even so, Maynard's treatment of month-long GPS surveillance is instructive, and provides additional support and alternative grounds for this decision.

Police in Maynard tracked the movements of a suspected drug dealer 24 hours a day for four weeks with a GPS device they had installed on his Jeep without a warrant. The Government used the resulting pattern of those movements—"not just the location of a particular `stash house' or Jones's movements on any one trip or even day"74—as evidence of drug trafficking conspiracy. Jones argued that the month-long warrantless tracking of his Jeep violated the Fourth Amendment prohibition of "unreasonable searches," and the D.C. Circuit agreed.

The Maynard court began by distinguishing United States v. Knotts, which had held that "[a] person traveling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements from one place to another." 460 U.S. 276, 281, 103 S.Ct. 1081, 75 L.Ed.2d 55 (1983). The police in Knotts had monitored a beeper placed in a five-gallon container while it was driven in a car 100 miles over public roads to a cabin in rural Wisconsin. Because the defendant by driving on public roads had "voluntarily conveyed to anyone who wanted to look" his progress and route, the Court held the beeper monitoring had violated no reasonable expectation of privacy, and hence was not a search under the Fourth Amendment. Id.

As Maynard correctly notes, the Knotts opinion expressly reserved the question whether a warrant would be required for prolonged or twenty-four hour surveillance.75 The D.C. Circuit held that Knotts was not controlling, because the tracking was not limited to a single trip from one place to another, but instead covered "Jones's movements 24 hours a day for 28 days as he moved among scores of places, thereby discovering the totality and pattern of his movements from place to place to place." 615 F.3d at 558. The court persuasively elaborates why prolonged location surveillance is different:

615 F.3d at 562 (footnote omitted).76 The court concluded that an individual has a legitimate expectation of privacy regarding the "intimate picture of his life" revealed by prolonged surveillance, citing various state privacy laws as well as the "considered judgments of every court to which the issue has been squarely presented."77

As Judge Orenstein observed, there are certain differences between the real-time GPS tracking in Maynard and the historical cell site records at issue here, but none support a different result. The temporal distinction between prospective and historical location tracking is not compelling, because the degree of invasiveness is the same, whether the tracking covers the previous 60 days or the next. In Judge Orenstein's words, "The picture of Tyshawn Augustus's life the government seeks to obtain is no less intimate simply because it has already been painted."78 Nor does it matter, for Fourth Amendment purposes, that the information sought was neither created nor maintained at the direction of law enforcement. See United States v. Miller, 425 U.S. 435, 443, 96 S.Ct. 1619, 48 L.Ed.2d 71 (1976) ("This [Fourth Amendment] analysis is not changed by the mandate of the Bank Secrecy Act that records of depositors' transactions be maintained by banks"). The fact that the records are presently in the hands of a third party might be dispositive if they had been "voluntarily conveyed" to the provider by the customer, but, as explained in the next section, that is not true of cell site tracking data.

In several respects, the historical cell site records sought here are more invasive than the GPS data revealed in Maynard. The duration and volume of information sought is more than doubled—60 days as opposed to 28 days of movement. As we have found, the level of detail provided by cell site technology now approaches that of GPS, and its reliability in obtaining a location fix actually exceeds that of GPS.79 Moreover, as Judge Orenstein points out, cell phone tracking is likely more revealing than a GPS device attached to a car, because the cell phone is carried on the person.80 It will also inevitably be more intrusive, because the phone can be monitored indoors where the expectation of privacy is greatest. By contrast, the GPS device in Maynard revealed no information about the interior of a home or other constitutionally protected space.

Finally, the Government's brief suggests that, as in Maynard, Fourth Amendment concerns are best addressed at a suppression hearing, after the search has taken place. But magistrate judges do not have the luxury of retrospective adjudication, waiting until a search occurs to decide whether a search warrant was required. If asked to issue an order that in our considered view violates the constitution, our sworn duty is to deny that application. Sometimes, the law is uncertain, because the Supreme Court has not definitively ruled. In such cases it is especially important for magistrate judges to explain their reasons on the record, giving affected parties (including the Government) the right to seek appellate review and correction, if necessary, by the Supreme Court. Murky areas of law like the ECPA remain murky decades after passage for two principal reasons—a dearth of reported district court decisions to generate appellate review, and a regime of sealing and gag orders to conceal court rulings from the general public and affected parties.81

For all these reasons, I join Judge Orenstein in holding that Maynard's prolonged surveillance doctrine precludes the Government from obtaining two months of cell phone tracking data without a warrant.

C. Because the Government Has Not Shown That the Location Data Sought Was Voluntarily Conveyed by the User, Smith v. Maryland Does Not Eliminate a Legitimate Expectation of Privacy

The Government urges that no Fourth Amendment interest is implicated here, because it is merely seeking the production of cell site data voluntarily conveyed by the target phone user to the provider. As the Supreme Court stated in Katz v. United States, "[w]hat a person knowingly exposes to the public ... is not a subject of Fourth Amendment protection." 389 U.S. at 351, 88 S.Ct. 507 (1967). In United States v. Miller, the Court found no legitimate expectation of privacy in bank checks, deposit slips, and financial statements, because they "contain only information voluntarily conveyed to the banks and exposed to their employees in the ordinary course of business." 425 U.S. 435, 442, 96 S.Ct. 1619, 48 L.Ed.2d 71 (1976). Perhaps the most directly relevant application of this doctrine is Smith v. Maryland, 442 U.S. 735, 99 S.Ct. 2577, 61 L.Ed.2d 220 (1979), where the Court found a telephone user had no legitimate privacy interest in phone numbers he dialed, because

Id. at 744, 99 S.Ct. 2577.

As with any Fourth Amendment claim involving records, a court "must examine the nature of the particular documents sought to be protected in order to determine whether there is a legitimate `expectation of privacy' concerning their contents." Miller, 425 U.S. at 442, 96 S.Ct. 1619; see also Smith, 442 U.S. at 741, 99 S.Ct. 2577("[I]t is important to begin by specifying precisely the nature of the state activity that is challenged."). The records at issue here are "historical cell site information and call detail information" for the target phone "for the sixty (60) days prior to the date the Order is signed by the Court."82 As defined in the Government's applications, this information includes "the cellsite/sectors used by the mobile telephone to obtain service for a call or when in an idle state."83 Clearly, these requests seek the phone's location not only at the beginning and end of calls, but also "registration" information as the phone moves about the network. In other words, the Government is asking for all available records tracking the phone's continuous location and movement during a two month period.

The first thing to note about this tracking data is that, although perhaps generated in the ordinary course of the provider's business, it is not a proprietary business record subject to unfettered corporate control, such as a marketing plan or an expense report or a soft drink formula. In 1999, Congress passed the Wireless Communication and Public Safety Act (WCPSA),84 which amended the Telecommunications Act to place limits on the carrier's use or disclosure of a cell phone user's location information. The existing statute obliged the telecommunications carrier to protect the confidentiality of "customer proprietary network information" (CPNI), that is, information about a customer's use of the service that was made available to the carrier by the customer solely by virtue of the carrier-customer relationship. 47 U.S.C. § 222(f)(1) (1996). In order to enhance privacy protection for wireless consumers, the new statute amended the definition of CPNI to include "location," and added the following section:

47 U.S.C. § 222(f)(1999). The privacy concerns animating this legislation were well articulated by one of the bill's sponsors on the House floor:

Wherever your cell phone goes becomes a monitor of all of your activities. 145 Cong. Rec. H9858-01, at H9860 (daily ed. Oct. 12, 1999) (statement by Rep. Edward Markey). Other members expressed similar privacy concerns.85

Of course, the statute did not place location information beyond the reach of law enforcement, and its privacy protections are not conclusive on whether a warrant is necessary to compel its production.86 Nevertheless, an act of Congress affecting proprietary interest in a thing is undeniably relevant to the legitimate-expectation-of-privacy inquiry.87 The WCPSA operates as a type of federal property law, balancing the interests of cell phone users and carriers in personal data generated as a result of providing telecommunication service. The fact that WCPSA imposes heightened privacy protection for a customer's call location information is highly relevant, especially given that the topic of wireless phone location is specifically mentioned in only one other federal statute, CALEA, which forbids its acquisition by law enforcement under the very relaxed standard for pen registers.88 At the very least, the WCPSA ties certain "strings" to call location data, pulling them out of the category of ordinary business records subject to a carrier's exclusive dominion and control.

Contrary to the Government's claim, the WCPSA does not "by its terms allow[] compelled disclosure pursuant to the SCA." Government brief, at 16. The statute does not mention the SCA. It merely recognizes an exception to its disclosure restrictions "as required by law." 47 U.S.C. § 222(c)(1). This language is perfectly consistent with a Fourth Amendment warrant requirement.

With this in mind, we return to the crux of the Government's argument—that cell site location information has been "voluntarily conveyed" by the cell phone user to the carrier, and thus Miller and Smith preclude Fourth Amendment protection. This contention has been directly addressed by two appellate courts to date, and both have rejected the claim. In United States v. Forest,89 the Sixth Circuit considered a drug dealer's claim to suppress the cell site data used to convict him. Law enforcement had dialed the defendant's phone without allowing it to ring, and the resulting cell site data was used to track his movements while driving on a public highway. The court was persuaded that Smith did not apply because the defendant had not voluntarily conveyed his cell site data to anyone; the agent, not the defendant, had dialed the number which caused the phone to send out signals.90 By necessary implication, in order to "voluntarily convey" location information under Smith, the user must do more than merely turn the phone on.

More recently, the Third Circuit went a step further, declaring that a cell phone user does not voluntarily convey location information by making or receiving a call. Rejecting the Government's analogy to the phone numbers dialed in Smith, the court explained:

620 F.3d at 317-18 (emphasis in original). This court reached a similar conclusion in its 2005 cell site decision.91 Although not all district courts have followed our lead on this point,92 nothing in those decisions or the Government's brief convincingly challenges that proposition.

Unlike the bank records in Miller or the phone numbers dialed in Smith, cell site data is neither tangible nor visible to a cell phone user. When a user turns on the phone and makes a call, she is not required to enter her own zip code, area code, or other location identifier. None of the digits pressed reveal her own location. Cell site data is generated automatically by the network, conveyed to the provider not by human hands, but by invisible radio signal. Thus, unlike in Miller or Smith, where the information at issue was unquestionably conveyed by the defendant to a third party, a cell phone user may well have no reason to suspect that her location was exposed to anyone. The assumption of risk theory espoused by Miller and Smith necessarily entails a knowing or voluntary act of disclosure; the Government has cited no case (and the court has found none) where unknowing, inadvertent disclosure of information by a defendant thereby precluded Fourth Amendment protection of that information.

One might argue that all cell phone users ought to know that their precise location will be conveyed to the provider because it is necessary to connect the call—otherwise the call could not be made. But that premise is simply untrue. As recent congressional testimony shows,93 before the advent of GPS and the current generation of network-based technology, it was not possible to locate cell phone users with any degree of precision.94 For that very reason, the Federal Communications Commission in 1997 issued final "Enhanced 911" (E-911) rules requiring providers to upgrade their systems to enable emergency responders to locate mobile units making emergency 911 calls for rescue or assistance.95 Thus, even a techsavvy cell phone user would not expect that anything more than an approximate location, such as his general neighborhood or area code, would be necessary for the network to complete a call.

The T-Mobile privacy policy tendered by the Government says no more than that: "Our network detects your device's approximate location whenever it is turned on (subject to coverage limitations)."96 Elsewhere the policy informs customers that call details and call location information are CPNI and reassures them that "Under federal law, you have a right, and we have a duty, to protect the confidentiality of CPNI and we have adopted policies and procedures designed to ensure compliance with those rules."97 Included in the record appendix are the terms of use for Metro PCS,98 the service provider for one of the other target phones, which describe arguably different policies and practices concerning the collection and retention of location information. Parsing the differences among the particular record-keeping practices of various providers would do little to advance the constitutional inquiry, however. As the Supreme Court wryly observed in Smith v. Maryland: "We are not inclined to make a crazy quilt of the Fourth Amendment, especially in circumstances where (as here) the pattern of protection would be dictated by billing practices of a private corporation."99

Of course, the tech-savvy user may now understand that there is a risk that the provider can calculate and record his location and movements very precisely. But the bare possibility of disclosure by a third party cannot by itself dispel all expectation of privacy. Otherwise, nothing would be left of Katz, because it was surely possible in 1967 for the phone company to wiretap and disclose a private conversation in a public phone booth. Similarly, it is possible that a carrier may open and inspect a letter or sealed package, but that risk alone does not eliminate the legitimate expectation of privacy in such effects. United States v. Jacobsen, 466 U.S. 109, 114, 104 S.Ct. 1652, 80 L.Ed.2d 85 (1984); Ex parte Jackson, 96 U.S. 727, 733, 24 L.Ed. 877 (1877).

In sum, Miller and Smith do not permit warrantless law enforcement access to all historical cell site data, because the user has not "knowingly exposed" or "voluntarily conveyed" that information to the provider, as those phrases are ordinarily understood. Historical cell site data are not ordinary business records of the providers. Congress has placed limits on the use and disclosure of call location information absent customer approval, and specifically forbade implying such approval based on mere use of the phone. Thus, consumers are not forced to sacrifice locational privacy as the price of using cell phones. This judgment of Congress may not be conclusive as to Fourth Amendment protection, but neither should it be ignored, especially when, as in the case of cell site data, it jibes comfortably with Fourth Amendment precedent.100

Conclusion

The "inexorable combination of market and regulatory stimuli ensures that cell phone tracking will become more precise with each passing year."101 In 1789 it was inconceivable that every peripatetic step of a citizen's life could be monitored, recorded, and revealed to the government. For a cell phone user born in 1984, however, it is conceivable that every movement of his adult life can be imperceptibly captured, compiled, and retrieved from a digital dossier somewhere in a computer cloud. Now as then, the Fourth Amendment remains our polestar.

It is true that the Government's warrantless requests are here limited to 60 days, but the logic of its position admits no temporal restraint. Two months' worth of hourly tracking data will inevitably reveal a rich slice of the user's life, activities, and associations; the D.C. Circuit has required a search warrant for half as much. If the telephone numbers dialed in Smith v. Maryland were notes on a musical scale, the location data sought here is a grand opera.

For these reasons, I arrive by a slightly different path at the same destination as my colleagues from Pennsylvania, New York, Massachusetts, Indiana, and Austin, Texas. Compelled warrantless disclosure of cell site data violates the Fourth Amendment under the separate authorities of Karo and Maynard. Accordingly, the Government's requests for that information under the SCA are denied.