FBI Report on Wiretap Capacity 2/96
[Congressional Record: February 9, 1996 (Senate)] [Page S1166-S1171] From the Congressional Record Online via GPO Access [wais.access.gpo.gov]
Mr. LEAHY. Mr. President, we took an important step in the last Congress to preserve law enforcement's wiretapping tool and increase privacy protection for our telephone and computer communications by passing the Communications Assistance for Law Enforcement Act. This law expanded privacy protection to cordless telephones, restricted the ability of law enforcement to obtain transactional information from e- mail messages, and improved the privacy of mobile phones by expanding criminal penalties for stealing the service from legitimate users.
This new law also imposed new requirements to ensure that court orders for electronic surveillance can be carried out, rather than stymied by new technologies used on our telephone networks.
Significantly, these new requirements for our Nation's telephone networks were accompanied by guidelines designed to bring public oversight and accountability to the process of implementing them. For the first time, decisions on how new and existing telecommunications systems will remain accessible to government surveillance must be made in the sunshine of public scrutiny.
Thus, the new law requires for the first time that law enforcement's demands regarding the number of wiretap, pen register, and other surveillance orders that telephone companies must be able to service simultaneously, are published in the Federal Register and scrutinized in a public procedure.
In compliance with this new requirement, the FBI published in the Federal Register on October 16, 1995, a proposed notice setting forth its capacity demands. According to the proposed notice, these capacity demands were predicated upon a historical baseline of electronic surveillance activity and an analysis of that activity. Yet, the Federal Register notice did not include publication of this underlying information.
Shortly after the notice was issued, I wrote to the FBI Director requesting copies of this information, and urging him to release the information to the public to ensure the fullest dissemination of the information.
I am aware that the comments filed by the deadline on January 16, 1996, in response to the proposed notice on behalf of civil liberties groups, telephone companies, and cellular companies have criticized the proposed notice for failing to disclose the supporting data for the capacity demands. As one set of comments filed by the Center for Democracy and Technology and the Center for National Security Studies noted, ``it is impossible to say whether or not the capacity requirements proposed in the notice are justified'' without the supporting data.
The FBI has now provided me with a 13-page report explaining how they collected information about past law enforcement surveillance activity from Federal and State court offices, State attorneys general offices, and over a thousand telecommunications carriers. This report also describes the method the FBI used to analyze this information and figure out how much wiretap capacity law enforcement would need for the foreseeable future and up to 10 years from now in three broad categories of geographic areas.
I fully appreciate the amount of hard work that went into collecting this critical data and doing the analysis. Indeed, this is the first time that law enforcement has ever been required to estimate its future surveillance activity and the demands that will be made on telecommunications carriers. This is also the first time that this information has ever been required to be made public.
I am pleased to make this report available for public dissemination. It does not identify which geographic areas fall into which of the three different categories with differing capacity requirements. Thus, it does not tell us whether Vermont is in category I, where the greatest level of interception activity has occurred and is likely to occur in the future, or in category III, where the number of law enforcement wiretaps have been low or nonexistent. Telecommunications carriers doing business in Vermont and Vermonters will want to know which category we fall into.
The FBI has assured me that they are in the process of preparing two additional documents that will explain the proposed capacity notice in greater depth. I look forward to examining those two additional documents upon their release by the FBI.
The public process set up in the new law is working. I commend the FBI Director for his efforts to fulfill the public accountability provisions of the law by making available this report and future reports on the capacity requirements.
Since these reports were not available prior to the deadline for comments on the original proposed notice, however, I would urge the FBI to consider issuing a revised or supplemental proposed notice to give interested parties an opportunity to comment on the proposed capacity demands with the benefit of this new information.
Mr. President, I ask unanimous consent that the report be printed in the Record.
There being no objection, the report was ordered to be printed in the Record, as follows:
Information Concerning Implementation of the Communications Assistance for Law Enforcement Act [CALEA]
CALEA was enacted to preserve law enforcement's ability, pursuant to court order or other lawful authorization, to access communications (content) and associated call- identifying information in an ever-changing telecommunications environment. Because many interceptions \1\ in the future will be initiated through equipment controlled by carriers, CALEA obligates the Attorney General to provide telecommunications carriers with information they will need (a) to adequately size and design their networks to accommodate the maximum number of simultaneous interceptions that law enforcement potentially might conduct after October 25, 1998, and (b) to be capable of accommodating the actual number of simultaneous interceptions law enforcement potentially might conduct as of October 25, 1998. The CALEA specifically refers to two levels of capacity, maximum and actual.
Footnotes at end of article.
CALEA requires the Attorney General to provide a notice of the maximum capacity required to accommodate all of the communication interceptions, pen registers, and trap trace devices that the Attorney General estimates that government agencies authorized to conduct electronic surveillance may conduct and use simultaneously after the date that is 4 years after enactment (i.e., after October 25, 1998). The Attorney General must also provide a notice of the actual number of communication interceptions, pen registers, and trap trace devices, representing a portion of the maximum capacity, that the Attorney General estimates that government agencies authorized to conduct electronic surveillance may conduct and use simultaneously as of October 1998.\2\Maximum Capacity
Unlike actual capacity, CALEA indicates that the maximum capacity applies to a time, not specified, after the date that is 4 years after the date of CALEA's enactment. The maximum capacity has been interpreted as chiefly a design requirement for telecommunications carriers that will be utilized to size and define an upper bound capacity ceiling for the mid-term to intermediate future (i.e., 1998-2004), as discussed in more detail below. This ceiling is intended to provide carriers with certainty and stability, as well as with a framework for cost effectively designing and engineering future capacity requirements into their networks.
As CALEA makes clear, the maximum capacity is in no way synonymous with actual capacity (i.e., the interception capacity that law enforcement may actually use as of October 25, 1998). Rather, maximum capacity is viewed as relating to a longer term, more enduring design requirement that would serve as a defined technological bound to interception activity, but yet would provide room for expeditiously accommodating certain future interception-related ``worst case scenarios.'' For example, it would be impossible for law enforcement to predict unusual, anomalous, but nonetheless very serious or violent events such as those associated with certain acts of terrorism or extreme instances of drug-trafficking or organized crime activity. Maximum capacity provides a safety-net that would enable telecommunications carriers to expeditiously respond to serious, unpredictable, emergencies that require very unusual levels of interception capacity.Actual Capacity
Under CALEA, estimates of actual capacity are to apply to all simultaneous interception activity that may be conducted by the date that is 4 years after the date of enactment of CALEA. CALEA makes clear that actual capacity represents only a portion of the maximum capacity. Actual capacity thus pertains to the amount of interception activity that potentially may be needed when many of CALEA's requirements are scheduled to take effect in October 1998. Consequently, when viewing the maximum and actual capacity levels set forth in the Initial Capacity Notice, realistically only the actual capacity estimates should be construed as in any way reflecting the amount of simultaneous interception activity that potentially may be conducted by law enforcement in any geographic area as of October 1998.Methodology Overview
The CALEA mandate which obligates the Attorney General to estimate future capacity marks the first time (a) that information has been required to be provided to carriers in order for them to properly design and size future networks with reference to interception activity; and (b) that the entire law enforcement community has been required to project its collective potential future interception needs, thereby in effect placing possible technological limitations on its lawful authority to conduct electronic surveillance. This circumstance, as viewed by the law enforcement community, obviously generates great and legitimate concern, because historically telephone technology placed no constraints on law enforcement's court ordered electronic surveillance authority. If not implemented with care, CALEA could have the unintended effect of potentially placing restrictions on the lawful use of electronic surveillance authority. Thus, if law enforcement errs by underscoping its potential, legitimate, and lawful interception needs, effective law enforcement will be hampered and, more importantly, the public safety will be jeopardized.
The FBI, which is implementing many of the responsibilities conferred upon the Attorney General by CALEA, was required to issue the capacity notice. In order to meet this obligation, we proceeded by employing a rigorous, comprehensive methodology to acquire critically needed information and to establish analytic tools for determining the simultaneous interception activity of the recent past and for estimating future maximum and actual capacity.
The methodology used was intended to take into consideration the concerns of the parties principally affected by CALEA. On the one hand, CALEA provides law enforcement with an opportunity and means to ensure that its future electronic surveillance needs can be met. On the other hand, CALEA presents an opportunity and means for telecommunications carriers to understand the nature and extent of their obligations to accommodate law enforcement's electronic surveillance needs and to do so in a way that will not be unduly burdensome or excessive. Law enforcement's approach and perspective regarding its electronic surveillance needs relate to its mission to combat serious crime, acts of terrorism, and acts of violence. Traditionally, this mission has been tied to law enforcement agencies' geographically-based jurisdiction and associated jurisdictional legal authority. Telecommunications carriers' approach and perspective regarding law enforcement's electronic surveillance needs, however, chiefly relate to the effect such needs may have on particular components within their systems that are used to serve subscribers within a given market (i.e., switches and other network components).
As a goal, law enforcement sought to harmonize the different approaches and perspectives of these parties in its methodology. The methodology sought to produce a result that acknowledged and factored in the interrelationship between the geographic, jurisdictional contours of law enforcement interception activity and the geographic service areas covered by the carrier switching facilities that will be likely used to provide interception access. By identifying key pieces of data associated with these approaches and perspectives, we were able to formulate and then analyze past interception activity in terms that enabled us to establish one of the key components (one pertaining to past peak levels of interception activity in localities of various sizes) in an equation leading to an estimate of future interception capacity. The other key component (one pertaining to projected growth) was derived from a statistical model that included a number of variables reasonably believed to predict potential future capacity bounds and potential future interception needs. Once the projections were made to estimate future capacity needs, the results were reviewed and adjusted to ensure reasonableness before publishing the results in the Initial Capacity Notice.
By presenting capacity levels in the Initial Capacity Notice with reference to engineered capacity (discussed below), the FBI intended to express interception capacity in an understandable fashion as a percentage of a carrier's switch or other network component capacity. This approach was intended to enable carriers to tailor their technological responses to law enforcement's potential interception needs within specific types of geographic areas. The following discussion highlights the methodological steps used to arrive at the actual and maximum capacity figures published in the Initial Capacity Notice.Part One: Deriving Key Data from Past Interception Activity
As a first step, we sought to establish a baseline of past simultaneous interception activity. Information concerning the actual number of all types of simultaneous interceptions occurring throughout the United States in the recent past was collected. Such information, needed to establish the baseline, had never before been collected and did not exist in a single repository. Amassing this detailed and extremely sensitive information required an unprecedented and time- consuming effort on the part of law enforcement. It involved identifying sources from which accurate information could be retrieved in an efficient and effective manner. Specifically, we sought to identify sources that could provide the exact number of all types of interceptions (to include call content, pen register, and trap and trace) performed by all federal, State, and local law enforcement agencies, in terms of the actual number of telephone lines \3\ intercepted at each locality.
To obtain specific line-related information regarding past simultaneous electronic surveillance activity, records of interception activity were acquired from telecommunications carriers, law enforcement officials, and most importantly, from the Federal and State Clerks of Court's offices--the official repository for all interceptions conducted simultaneously between January 1, 1993 and March 1, 1995, for all geographic locations. Specifically, highly sensitive interception information pertaining to each interception start/end dates, and to area code and exchange was collected. The period January 1, 1993 to March 1, 1995 was chosen in order to obtain recent interception information that was reasonably retrievable given the time constraint of one year imposed by the CALEA with regard to publishing a capacity notice.
Approximately 1500 telecommunications carriers (those serving the majority of subscribers in the U.S., and representing the largest of the wireline and cellular telephone companies (as of March 1995)) were requested to provide information that would identify where and how many interceptions had occurred within their networks during the period of study. Acquisition and examination of sensitive electronic surveillance records maintained pursuant to statute under seal with the Clerks of Court offices was pursued through two separate efforts. All Federal court order information was collected pursuant to special court orders directing the unsealing of interception court orders for the limited purpose of enabling the Attorney General/FBI to comply with the capacity notice obligation. State and local law enforcement information was obtained through the State Attorneys General (AG) offices. Each State AG was requested to coordinate the collection of interception information within the AG's respective State.
By reviewing the data collected, the number of simultaneous interceptions that had occurred within switching facilities was identified throughout the country during the study period. One of the key pieces of relevant information was the highest number of interceptions that had ever occurred simultaneously within any telecommunications carrier's switch.\5\
For the years studied, the highest number of simultaneous interceptions occurring within any one switch in the United States was 220; that is, 220 pen register, trap and trace and/or call content interceptions were active at the same time within a particular switch. Further analysis revealed that the 220 interception number far exceeded the number of simultaneous interceptions conducted in other high activity switches and was due to a single unusual occurrence. The switch that had the second highest level of simultaneous activity supported 120 pen register, trap and trace and/or call content interceptions.
The distribution of baseline simultaneous interception activity by switch was varied. Among switches which had supported interceptions during the period studied, there was a broad and varied distribution of simultaneous interceptions ranging from 1 to 120. Switches with the greatest levels of interception activity frequently existed in urban areas. Switches with lower levels of interception activity existed across a variety of geographic areas, encompassing urban, suburban, and rural areas.
The review of the baseline of interception activity yielded certain key data on simultaneous interceptions for specific switching systems, but law enforcement usually approaches crime within the context of geographic or jurisdictional areas. The next analysis challenge was to associate the baseline simultaneous interception data on specific switches with law enforcement's need to express requirements relative to geographic areas. To do this, law enforcement sought to identify geographic boundaries that could provide common reference points between law enforcement and telecommunications carriers.
A number of geographic boundaries which define service areas of telecommunications carriers were considered (e.g., State lines, local access transport areas (LATA), metropolitan statistical areas (MSA), rural service areas (RSA), and major and basic trading areas (MTA, BTA)). However, in each case, these boundaries did not provide a good match with defined law enforcement areas. Law enforcement jurisdictional legal authority in a great number of cases is defined by county boundaries (i.e., for local law enforcement agencies). County boundaries rarely change and are not disputable. In most cases, a group of counties are encompassed by the boundaries that define a telecommunications carrier's service area. Consequently, county boundaries were used as the common reference tool in formulating an analysis for future capacity.
From the baseline of interception activity, interception data was aggregated for a group of specific switches by county. Switches were assigned to counties based on their geographic location. The data can be divided into their different levels (or categories) or interception activity: high, moderate, and low.
As part of the consultative process with the telecommunications carriers, law enforcement understood that one of the primary carrier concerns with regard to capacity was that each switch in a carrier's network should not be held to the same requirement for capacity levels. The carriers had indicated generally that although CALEA requires the Attorney General to estimate future capacity sizings and to estimate future potential interception activity that may occur in their network, every attempt should be made by law enforcement to express its needs in accordance with the variability of interceptions that had occurred in the past. Recognizing this, we chose to define law enforcement's potential future interception needs using categories and not just one absolute number that would apply without regard to the often varied nature of interceptions in a carrier's geographic service area.
As a result of considering the relative relationships between law enforcement and telecommunications carriers geographic areas, it was determined that three separate categories for stating capacity requirements should be created. The use of categories permitted capacity requirements to be stated in a way that reasonably reflected, and was responsive to, law enforcement interception needs without unduly burdening all carriers with the same level of capacity or requiring a particular carrier to provide the same capacity level everywhere in its network. As alluded to in the Initial Capacity Notice, Category I represents a small number of geographic areas where the greatest level of interception activity typically has occurred and is likely to occur. Category II also represents a relatively small number of geographic areas, which consist of some urban and suburban areas. Category III represents the vast majority of areas where electronic surveillance activity has historically been low or nonexistent, principally in rural and many suburban areas.
With regard to Category III, as the nation moves toward the future, law enforcement must have the capability to conduct a minimum level of electronic surveillance in any locality, regardless of previous levels of criminal activity or prior levels of electronic surveillance activity. Terrorism, drug trafficking, and violent crimes are constant and unpredictable threats to the public in all localities, as evidenced by the Oklahoma City bombing in April 1995. Consequently, unlike Categories I and II, which are defined principally with reference to past levels of electronic surveillance activity, Category III constitutes a minimum interception capacity for any location in which law enforcement may need an interception capability in order to protect the public and effectively enforce the law.
Once the geographic areas could be generally (but only tentatively) associated with each category, the historic baseline of interception activity for these areas could be used as a way of defining the distinct levels of past electronic surveillance activity. For the Category I level, as noted above, the highest number of simultaneous interceptions from a switch was 120. Our analysis also determined that within the Category II level the highest number of simultaneous interceptions from a switch was 42.The Use of Percentages and Engineered Capacity
Law enforcement considered it appropriate and prudent to express capacity requirements in the Initial Capacity Notice in terms of a percentage rather than as a fixed number for several reasons. First, percentages are very appropriate, if not essential, with regard to new service providers, new switches, new services, and new technologies. If absolute numbers were set forth, the introduction of new basic service, as well as more advanced services and features, could be impeded--a prospect that is unacceptable to either the Congress, law enforcement, or the telecommunications industry. Second, percentages allow capacity requirements to adjust slightly to a changing subscriber base as it increases or decreases over time.
To enable carriers to apply the percentages to the affected systems in their networks properly, capacity level percentages were tied to engineered capacity. Engineered capacity was referred to in the Initial Capacity Notice as the maximum number of subscribers that can be served by a carrier's equipment, facilities, or service. In the main, a carrier's switching facility was the key network component associated with this foregoing terminology. With regard to the highest level of simultaneous interception activity in the baseline data with regard to Category I (120), the switch associated with that interception activity was one that typically served an average of 35,000 subscribers. This led to the historical electronic surveillance activity being expressed as a percentage of engineered capacity of 0.34. This percentage was derived by dividing the 120 simultaneous interceptions by 35,000 subscribers. With regard to the highest level of simultaneous interception activity noted in the baseline data occurring with regard to Category II (42), the switch associated with that interception activity was one that typically served an average of 27,000 subscribers. This led to the historical electronic surveillance activity being expressed as a percentage of engineered capacity of 0.16. This percentage was derived by dividing the 42 simultaneous interceptions by 27,000 subscribers.
As discussed above, the interception activity which was now being associated with Category III reflected little or no activity in the past. Hence the capacity level associated with Category III was dervived in a different manner. To establish a percentage regarding areas where there had been little or no past interception activity, for purposes of future analyses, we examined the distribution of historical interception activity for those switches with very little or not interception activity. The majority of these switches had electronic surveillance activity of less than 0.1 of the capacity associated with such switches. This led law enforcement to assign 0.1 as the level of historical electronic surveillance activity for Category III, expressed as a percentage of engineered capacity. This figure, therefore, was selected essentially to ensure a bare minimum capacity to support law enforcement interceptions.Part Two: Estimating and Using Growth Rates to Project Future Capacity and Interception Activity
CALEA's mandate that law enforcement identify capacity requirements marked the first time future carrier capacity sizings and future potential interception activity were required to be estimated. Although CALEA provided no specific date as an outer bound for the projection for maximum capacity, the year 2004 was used for its projection of maximum capacity. We used the year 2004 because it reflects a ten year period from the last date for which historical data was available (1994). A ten year period is commonly used as a period of time within which to analyze and prepare projections. An analysis of a period of this length also provides the affected telecommunications carriers with the information that would produce the greatest level of stability for the mid-term to intermediate future. By comparison, CALEA specifies October 1998 as the date for projecting potential future interception activity (actual capacity).
Four steps were followed to determine the maximum and actual capacity percentages which were published in the Initial Capacity Notice. These four steps are as follows:
Step 1: Establishing a Statistical Model; Projecting Title III Court Order Trends as a Predicate for Projecting Future Design Capacity and Potential Future Interception Activity
Projections regarding future design capacity that may be required to accommodate unusual future interception activity, as well as the amount of potential future interception activity itself (call content interceptions, as well as the much more prevalent pen register and trap and trace interceptions), in terms of the number of lines that possibly could be implicated, are not readily and easily derived. For example, when one considers just one of the relevant groups of information (past Title III court order authorizations) it becomes apparent that a simple straight line trend does not exist. In reviewing the number of such authorizations on a yearly basis over the last 15 years (1980-1994), there is over a 100 increase (from 566 to 1154). However, the year to year increases are anything but consistent, with variations from -12 to +19 occurring by way of yearly changes.
Although it may be impossible to discern precisely why increases (or decreases) occur on a yearly basis with regard to all of the types of interception activity, a number of factors were considered (discussed further below) which over time presumably would logically influence such activity. In brief, these factors relate to population, telecommunications technology, law enforcement resources, and relevant crime statistics.
Notwithstanding the difficulty in making long range estimations about a design capacity level capable of accommodating reasonable growth over the intermediate future (1998-2004) and which, more importantly, could also address totally unexpected worst case scenarios (maximum capacity) and future potential, interception activity (actual capacity), in terms of the number of lines implicated, we were obligated to respond to CALEA's mandate to produce capacity estimates.
Law enforcement determined that it was imperative to construct and utilize an analytical statistical model in order to address the variability related to relevant historical data that could be associated with interception activity. The model selected was a widely-used model referred to as a multi-variable linear regression statistical model. With such a model, future trends central to predicting capacity could be projected.
In broad terms, this model sought to project potential future behavior based upon an analysis of the relationships between two data groups for which historical data was available over the last 15 years. One data group was the number of criminal Title III court orders authorized between 1980 and 1994. The other data group was composed of a suite of relevant factors (multi-variables) related to population, telecommunications technology, law enforcement resources, and relevant crime statistics. Once trend relationships were noted between the second group and the first, an equation was produced from which projections as to the future level of the first group (Title III court orders) could be established based on growth projections of the second group. The projection for the criminal Title III court order authorizations was an important analytic first step in projecting the design capacity and future potential interception activity in terms of the number of lines (interceptions) that may be involved.
The first data group consisted of the number of Title III court order authorizations recorded per year in the Wiretap Report, compiled each year by the Administrative Office of the United States Courts. In an attempt to establish a trend, we examined the criminal Title III court order authorizations granted per year over the last 15 years (1980-1994). As noted above, however, an examination of this data showed significant yearly fluctuations thereby precluding a simple, straight line trend for approximating future authorizations. Also, as noted above, the Wiretap Report only documents the number of criminal Title III court orders; it does not specify the actual number of call content interceptions associated with each order, nor does it address at all the vastly greater number of interceptions associated with pen registers and traps and traces. Nonetheless, because it was the only longstanding electronic surveillance data source in existence, it was concluded that it necessarily should play an important role in the model because it did document past, relevant interception-related information from which projections of design capacity and future potential interception activity could later be made.
The second data group consisted of factors or multi- variables deemed relevant with regard to the conduct of all types of interception activity. These factors, as noted above, were population, telecommunications technology, law enforcement resources, and relevant crime statistics. Historical and projected future data on population totals was collected from the United States Census Bureau. Historical and projected future data for wireline telecommunications subscribers was collected from the Federal Communications Commission. Historical data on wireless subscribers was acquired from reviewing publications issued by the Cellular Telecommunications Industry Association (CTIA). Projected future data for wireless subscribers was derived based upon an analysis of information found in trade journals concerning future projections of the number of wireless subscribers and attributed to the CTIA, the Personal Communications Industry Association, and other industry sources. Historical data on total number of law enforcement officers and reported incidents of violent crime was collected from the FBI's Uniform Crime Report. Projected future data on the total number of law enforcement officers and crime was derived based on an assumed linear growth rate of the historical data. For each of the above factors, projections for future data were made out to the year 2004.Step 2: Computing Growth Rates
Using the projected future data produced by the equation, the statistics indicated a growth rate of 130 in the first group between 1994 and 2004, a factor which was included in the subsequent analysis used to project maximum (design) capacity. By comparison, a projection for potential future interception activity (actual capacity) was arrived at by noting the statistics in the first group, that had applied the same trends projected by the statistical model, suggesting a growth rate between 1994 and 1998 of 54 percent. As noted above, when one reviews the number of criminal Title III court order authorizations on a yearly basis during the last 15 years (1980-1994), there is more than a 100 increase (from 566 to 1154). Also, when one reviews available official reports regarding the vastly more prevalent pen register and trap and trace court orders granted during the last 8 years, one observes an increase of 219. Finally, for the same period, with regard to such pen register and trap and trace court orders ``the number of persons whose telephone facilities were affected'' increased by 345.\6\ Consequently, the model's results are considered to be reasonable and consistent with past interception-related activity and appropriate for use as part of the analysis.Step 3: Projections for Design Capacity and Potential Future Interception Activity
As noted above, the historical interception activity expressed as a percentage of engineered capacity for Category I was 0.34 (120 simultaneous interceptions out of a switch serving a possible 35,000 subscribers); and the historical interception activity expressed as a percentage of engineered capacity for Category II was 0.16 (42 simultaneous interceptions out of a switch serving a possible 27,000 subscribers). For Category III, the minimum level of interception activity expressed as a percentage of engineered capacity was 0.1. The computed growth rates of 130 and 54 were converted into growth multipliers of 2.3 and 1.54, for maximum and actual capacity, respectively. By multiplying the historical interception activity figures for Category I, Category II, and Category III by the growth multipliers, we calculated numbers for a ``raw'' maximum and actual capacity, as illustrated in the table below.
RAW MAXIMUM AND ACTUAL CAPACITY PREDICTED BY THE GROWTH FACTORS ------------------------------------------------------------------------ Maximum Actual ------------------------------------------------------------------------ Category I...................... 0.34%2.3=0.78%.... 0.34%1.54=0.52% Category II..................... 0.16%2.3=0.37%.... 0.16%1.54=0.25% Category III.................... 0.10%2.3=0.23%.... 0.10%1.54=0.15% ------------------------------------------------------------------------Step 4: Applying Reasonableness Criteria to the Raw Calculations
In projecting future design capacity (maximum capacity) and capacity for potential future interception activity (actual capacity) that may be needed by all law enforcement agencies, for publication in the Federal Register, the FBI made downward adjustments to the numbers that were set forth with regard to the raw actual capacity in both Categories I and III. A substantial downward adjustment was made in Category III. These downward adjustments were made because law environment recognized that a majority of the affected telecommunications carriers fall in Category III, and that our historical interception activity was minimal in this category. Therefore, we chose to lessen the burden on the telecommunications industry and minimize the costs of installing solutions, consistent with assuring an essential minimum level of interception capacity.
The numbers set forth for the raw maximum capacity in Categories I, II, and III were adjusted upward for a number of reasons. First, as noted above, the interception activity associated with pen registers and traps and traces by far represents the most frequently used type of interception. The growth rate in the number of pen register and trap and trace court orders far exceeds that projected in the statistical analysis. As noted above, the percentage of increase in such court orders during the past 8 years was 219 and the number of persons whose telephone facilities were affected increased 345.
Second, although the peak number of simultaneous interceptions identified in the baseline data was 220, we elected not to use it in the statistical analysis because it was deemed to be an anomaly. However, this level of interception activity, although anomalous, is indicative of the type of worst-case scenarios that law enforcement should not easily ignore or completely discount. Hence some provision needs to be made for such situations.
Third, although every effort was made to capture as accurately as possible the actual levels of past interception activity in compiling our baseline of historical intercept information, there may be some instances where data was not fully collected or reported. Also, there is an amount of interception activity associate with national security that must be accounted for in the final capacity projections.
Fourth, during the study period (1980-1994), the number of States granting State and local law enforcement interception authority by statute increased and a number of States expanded interception authority to cover additional types of crimes and/or additional types of communications devices (e.g., cellular telephones and pagers). There is a reasonable likelihood that in the future other States may grant similar interception authority and/or increase the scope of existing interception authority.
Finally, law enforcement believed that judgment needed to be brought to bear on the numbers for raw actual and maximum capacity, in terms of making appropriate upward or downward adjustments. In particular, the FBI, acting on behalf of the entire law enforcement community, had a responsibility to be careful not to overstate or understate capacity needs. Importantly, however, if capacity needs were understated it could pose direct harm to the public safety and effective law enforcement.Initial Capacity Notice
On October 16, 1995, the FBI published in the Federal Register its Initial Capacity Notice. The estimates of actual and maximum capacity, expressed as a percentage of engineered capacity, were stated as follows in the Initial Capacity Notice:
ESTIMATES FOR PROJECTED MAXIMUM AND ACTUAL CAPACITY AS PUBLISHED IN THE FEDERAL REGISTER ------------------------------------------------------------------------ Maximum Actual ------------------------------------------------------------------------ Category I...................................... 1.00% 0.50% Category II..................................... 0.50% 0.25% Category III.................................... 0.25% 0.05% ------------------------------------------------------------------------
After the FBI's publication of the Initial Capacity Notice in the Federal Register, you requested two documents in your correspondence: (1) the historical baseline of electronic surveillance activity, and (2) an analysis of that activity. By way of response and clarification, although we examined past electronic surveillance activity and utilized certain key pieces of information derived therefrom as discussed in this letter, no ``document'', as such, was ever created. Similarly, the factors utilized in our analysis were never compiled into a document.
Nonetheless, because of the interest and the misunderstandings that have been associated with this matter, we currently are in the process of preparing two methodology documents which will explain our capacity notice efforts in greater depth. The first document will describe the process used to collect historical electronic surveillance information. The second document will describe the analysis used in developing the Initial Capacity Notice, as well as the Final Capacity Notice. Regarding the latter, it will take into account the written comments we have received and the input from our ongoing meetings with the telecommunications industry and other interest groups. The FBI will provide copies of these two documents to you upon their completion.Summary
As you are aware, in 1968 when Congress statutorily authorized court ordered electronic surveillance, telephone technology permitted law enforcement to execute, without impairment, essentially all court orders--a 100 percent capability/capacity to accommodate law enforcement's court ordered electronic surveillance needs. However, the onset of new and advanced technologies has begun to erode the capacity and capability of the telecommunications industry to support law enforcement's interception needs. For the first time, technological limitations will potentially be set on law enforcement's ability to lawfully conduct electronic surveillance. In view of these potential limitations, an attempt was made to estimate law enforcement's capacity needs in an accurate and reasonable manner with the goal of striking a balance of meeting law enforcement's interception needs without introducing unduly burdensome or excessive requirements on telecommunications carriers. This in no way changes law enforcement's fundamental statutory responsibility to obtain proper legal authority to conduct electronic surveillance. CALEA's capacity requirements simply ensure that, after law enforcement obtains proper legal authority, telecommunications carriers will have sufficient capacity to accommodate lawfully authorized electronic surveillance activity.
To project capacity needs, the previously described national methodology was employed in order to prepare an Initial Capacity Notice for publication in the Federal Register. Since release of the Notice, law enforcement has met with privacy advocates, the telecommunications industry, and other interested parties to clarify and further describe how best to apply capacity needs within a given carrier's network, and it is reviewing and considering the formal written comments provided in response to the Initial Capacity Notice.
In closing, CALEA is important to the public safety and national security of this nation and its full and timely implementation is critically important to the American public. On behalf of the law enforcement community, we would again like to express our appreciation for your support and leadership regarding this significant and complex issue. With the continued support of the Congress, we are confident that CALEA will be fully implemented in an effective and efficient manner.
\1\ For purposes of this discussion, the word ``interception'' refers to all types of interceptions: (1) interceptions of communication content (e.g., Title III); and (2) interceptions of dialing-related information (call identifying information) derived from pen registers and traps and traces.
\2\ 47 U.S.C. 1003 (1994).
\3\ For purposes of this discussion, the word ``lines'' refers to the transmission path from a subscriber's terminal to the network via a wireline or wireless medium.
\4\ Although a valuable source for historical information on criminal Title III (call content) court orders, The Wiretap Report, published annually by the Administrative Office of the United States Courts, does not identify the actual number of interception lines associated with each court order or, more importantly, with the vastly greater number of lines associated with pen register and trap and trace interceptions that have been performed by all law enforcement agencies. Even though the FBI used information pertaining to the number of court orders reported in The Wiretap Report as an essential element for forecasting purposes, the Report does not contain the line-related information that was needed to identify the actual level of past interception activity related to specific switches.
\5\ For purposes of this discussion, the term ``switch'' also represents a licensed cellular service area.
\6\ 18 U.S.C. 3126 (1986). By law, the Department of Justice is required to report to the Congress on a yearly basis information on the use of pen registers and trap and trace devices by law enforcement agencies within the Department of Justice. This report includes information concerning the number of original orders, extensions, the number of investigations, the number of persons whose telephone facilities were affected, and the number of overall dialings.
Share this page:
EPIC relies on support from individual donors to pursue our work.
Subscribe to the EPIC Alert
The EPIC Alert is a biweekly newsletter highlighting emerging privacy issues.
by Ryan Calo, A. Michael Froomkin,