Ending the zero-sum game: how to increase the productivity of the Fourth Amendment.

Author:Simmons, Ric
Position:II. Applying the Model B. Surveillance Techniques Which Are as Effective But Less Intrusive through Conclusion, p. 579-604 - Privacy, Security, and Human Dignity in the Digital Age
  1. Surveillance Techniques Which Are as Effective but Less Intrusive

    Another way to increase productivity is to develop surveillance techniques that provide the same level of output but cost less in terms of privacy costs. In general, law enforcement agencies will be indifferent as to whether to use these methods, or equally expensive but more intrusive methods because once again they are indifferent to the "savings" in terms of the cost to privacy.

    1. Binary Surveillance

      Binary surveillance refers to a surveillance method that only produces one of two results: positive (meaning that illegal activity has been detected) or negative (meaning that illegal activity has not been detected). (92) The surveillance provides no other information about the person or area being monitored, and so represents a relatively minor intrusion on the target's privacy. In fact, the Supreme Court has held that binary surveillance does not even count as a "search" under the Fourth Amendment because it does not infringe on an expectation of privacy that society is prepared to recognize as legitimate. (93)

      A simple example of a binary surveillance technique is a field test for narcotics. If a law enforcement officer reasonably believes that a certain substance may be narcotics, she can legally seize a very small amount of the substance and mix it with certain chemicals. (94) If the substance tests positive for narcotics, the law enforcement officer knows that the substance is in fact contraband and that a crime has occurred. If the substance tests negative, the officer knows nothing about the substance other than the fact that it is not an illegal drug. Therefore, because the suspect has no legitimate interest in possessing contraband, assuming other procedural prerequisites are satisfied, the surveillance does not implicate the Fourth Amendment (95)--the officer either learns nothing at all about the defendant or learns that the defendant is engaging in illegal activity.

      Under our analysis, binary surveillance would theoretically provide us with a very high level of surveillance productivity because the cost in terms of intrusiveness is almost nothing. In practice, however, there are a number of complications when applying our model to binary surveillance. First, simply because the surveillance does not implicate the Fourth Amendment does not mean it has no cost in terms of privacy intrusion. For example, a drug-sniffing dog that approaches an individual's belongings, home, or person represents some level of intrusion, (96) even if there is no chance that the surveillance will reveal any legitimate information about your possessions. Thus, the binary surveillance itself creates some privacy cost. Furthermore, many types of binary surveillance will lead to a more intrusive search. A drug-sniffing dog alerting to a suspect's car or suitcase will trigger a thorough physical search. If the drug dog has a very low false positive rate, then it would result only in a more intrusive search on those rare occasions when contraband actually was present. Thus, this extra privacy cost would be relatively low. But in practice this is not the case: According to some estimates, when drug dogs alert, they are correct only between 26% and 44% of the time. (97) Therefore, in calculating the total privacy costs of binary surveillance, the privacy cost of a physical search times the percentage chance that the dog has improperly alerted must be added to the equation.

      Finally, there is one more element to our total privacy cost calculation: a correct positive alert will also result in a physical search. Thus, we must include the privacy cost of a physical search multiplied by the percentage chance that contraband actually is present. Although the Supreme Court has stated that individuals who possess narcotics have no legitimate expectation of privacy in the contraband they carry, (98) a physical search that finds this contraband will involve a significant level of intrusion and will reveal other information about the suspect that is protected.

      Accordingly, the privacy cost of using a drug-sniffing dog is: [C.sub.P](drug dog sniff) + (false positive rate)*[C.sub.P](physical search) + (correct positive rate)*[C.sub.P](physical search)

      To measure the overall productivity of binary searches, we also need to determine the administrative cost of the surveillance and, on the other side of the equation, the expected benefits of the surveillance. In the case of drug-sniffing dogs, the administrative costs can be calculated relatively easily. An officer with a drug-sniffing dog costs more than an officer working alone--trained drug-sniffing dogs cost between $5,000 to $8,000 up-front and then approximately $600 per year to maintain. (99) Drug-sniffing dogs make up for this higher cost, however, with their ability to quickly scan through many suitcases or cars, allowing them to conduct their surveillance much more quickly than an officer would be able to working alone, searching by hand. Given this greater efficiency, a police officer with a drug-sniffing dog can conduct an individual search for far less money than a police officer without a drug-sniffing dog.

      Unfortunately, measuring the expected effectiveness of drug-sniffing dogs is challenging. Most of the existing studies focus on the chance of a false positive: that is, whether or not a drug-sniffing dog will alert when there is no contraband present. To measure the effectiveness of the drug-sniffing dogs, however, we need to know the rate of false negatives--how often the drug-sniffing dog does not alert when contraband is in fact present. Finally, we need to multiply the chance of a successful search by the severity of the crime--in this case, drug possession or drug trafficking, depending on the amount of contraband which is found.

      Of course, quantifying all of this data would be nearly impossible. But this analysis is still useful because it allows us to make rough comparisons between different types of surveillance. In the context of investigating narcotics possession, we can use our model to compare the use of drug-sniffing dogs to the more traditional physical search for drugs to find which method has a higher productivity. Suppose a car has been pulled over, and the police wish to determine whether the car has narcotics inside. A traditional physical search will take a substantial amount of time, which increases both the administrative and privacy cost of the surveillance. The physical search will also reveal a large amount of protected information to the law enforcement officer--essentially the entire contents of the car--which further increases the privacy cost. On the other side of the equation, a thorough search is likely to turn up narcotics if in fact any exist, though the most serious crimes are more likely to remain undetected because drug traffickers are likely to go to greater lengths to hide their contraband.

      A drug sniff, on the other hand, imposes a much lower administrative cost to the government and a much lower privacy cost to the suspect--a traditional search is carried out only if there is a positive alert by the drug-sniffing dog. Even if the false positive rate is over 75%, the use of the drug-sniffing dog still represents a lower privacy cost than a traditional search. And if a drug-sniffing dog is used, any narcotics that are present are more likely to be found: Frequently the dog will alert to a certain area, and law enforcement officials can concentrate their search on that area, conducting a more careful and comprehensive search than they would have done if a drug-sniffing dog had not been used.

      This conclusion does not mean that drug-sniffing dogs should be used indiscriminately, without probable cause, on every car that police stop on the road or find in a parking lot. It means only that searches with drug-sniffing dogs are preferable to searches without them because the former are more productive--they carry a lower cost with a higher rate of return. Thus, we should encourage their use over other, less productive searches, such as physical searches. Indeed, the current laws reflect this reality: Dog-sniff searches conducted during lawful Terry stops are permitted without a warrant, (100) whereas physical searches of a car without a warrant require probable cause. (101) This provides police with an incentive to use searches with drug dogs as opposed to relying on physical searches.

      Two other common examples of binary surveillance are hash value searches and gun detectors. "Hash value" searches are searches of a computer hard drive or other digital storage device in which a piece of software is used to examine every file stored on the drive. Every computer file has a unique identifier known as its "hash value," which is as individualized as a fingerprint. (102) The Federal Bureau of Investigation (FBI) maintains a database containing the hash value of every known computer file containing child pornography. Thus, a computer program could be installed into a suspect's hard drive and could sift through every digital file stored there, looking for a match with any of the thousands of known contraband hash values. If no match were found, the program would return no results, and the law enforcement officer monitoring the program would learn nothing about the suspect's computer other than the fact that it did not contain any known files of child pornography. If a match were found, the law enforcement officer would know with near-certainty that the suspect's computer did contain at least one such file.

      The software searching for these files could be secretly downloaded into a suspect's computer as an attachment to an e-mail, or it could be attached to an internet service provider (ISP) to monitor all internet traffic to and from the suspect's computer. Alternatively, this surveillance technique could be conducted on a massive scale, one that makes the potential scale of...

To continue reading