understanding the importance and impact of anonymity and authentication in a networked society
navigation menu top border

.:home:.     .:project:.    .:people:.     .:research:.     .:blog:.     .:resources:.     .:media:.

navigation menu bottom border
main display area top border
« Some Thoughts on Camera Phones, Space and Gender | Main | Comment on U.K.'s Proposed 'Supercomputer' »


posted by:Ian Kerr // 11:59 PM // January 09, 2007 // ID TRAIL MIX


With the lights out its less dangerous
Here we are now
Entertain us
I feel stupid and contagious
Here we are now
Entertain us

Kurt Cobain

all is quiet on new years day. the pizza last night at colonnade was warm and comforting not unlike our nearby fireplace, where i now sit cross-legged with my laptop, well, on my lap. it occurs to me just lucky i am and how nice it was to spend a low key evening with family. (i have always thought of new years eve as a night that you have to kiss people you normally wouldn’t spit on…happy to have transcended that phase of life.)

rather than waking erin and newton, both of whom have an incredible ability to sleep easily and peacefully, i strangely find myself reflecting on the time that i spent last summer nosing around the caselaw on sniffer dogs.

at the time, i was preparing for an NJI conference, where my idtrail colleagues and i offered 150 judges a full day workshop on the reasonable expectation of privacy. we will be doing a funky-fied version of it again in montreal at CFP 2007, which is being organized this year by privacy guru and director of strategic policy and research at the office of the privacy commissioner of canada (and former idtrail research coordinator) stephanie perrin.

sniffer dogs look a lot like my dog — however, unlike newton, they are specially trained to “sniff-out” contraband such as illegal drugs or explosives.

i am not a criminal law specialist, nor do i claim any particular expertise in the law of search and seizure. but i do care a lot about it. it is the area of law most significant to the development of a crucial legal construct: the reasonable expectation of privacy.

the ‘reasonable expectation of privacy’ standard provides the benchmark for circumstances in which the state (and sometimes the private sector) is constitutionally permitted to interfere with an individual’s privacy interests. my own interest in the subject and part of the contribution that i had hoped to make at the NJI workshop stems from my deep concern that emerging surveillance technologies will be understood by courts to diminish our expectation of privacy; as i discuss in my NJI presentation, i think that this would be a dreadful, terrible mistake. an epic social disaster, really.

the crossroads for an exploration of the most recent sniffer dog cases, in my view, is a case that had nothing to do with sniffer dogs: a 2004 supreme court of canada decision called tessling.

at issue in tessling was the RCMP’s use of FLIR (forward looking infrared), a technology that captures the infrared portion of the electromagnetic spectrum. as the supreme court described it, it is a camera that takes pictures of heat instead of light. in tessling, the RCMP used an airborne FLIR system so that officers could measure from way-up-in-the-sky the heat emanating from a house occupied by a guy called walter tessling.

you see, the RCMP were suspicious that tessling had a grow-op in his house. but their sources were unreliable. the point of using airborne FLIR was to get better evidence — without trespassing on the property — so that an officer could appear before a judge and assert reasonable grounds to believe that tessling was growing pot in his house. (mere suspicion is not enough to get a search warrant.)

tessling’s lawyer argued that, without a warrant, the airborne use of FLIR to measure the heat coming off the walls of his house amounted to an illegal search and that the FLIR-evidence ought therefore be inadmissible in court.

the supreme court disagreed, stating at paragraph 63 of the decision that:

external patterns of heat distribution on the external surfaces of a house is not information in which the respondent had a reasonable expectation of privacy.

i could go on at some length about the decision and the legal framework used to determine that walter tessling did not have a reasonable expectation of privacy in the heat waves emanating from his house. but a discussion of the tessling decision or the legal framework it supports is not my current purpose. (if you are looking for that background, you should *most definitely* check out my friend and colleague jane bailey’s superb NJI presentation.)

since my focus is on sniffer dogs, tessling is relevant only insofar as its outcome seems to invite subsequent courts to consider adopting an extension of its logic — namely, in searches involving sniffer dogs, to ask:

whether external patterns of smell on the external surface of a knapsack is, or is not, information in which a person holds a reasonable expectation of privacy.

with the growing concern about various sorts of contraband, this is a burning question in canadian courts.

the answer to this question, i suggest, needs to be understood within the broader context of a recent and increasing trend in law enforcement — the adoption of ‘jetway’ programs (smell the irony in the ‘jetway’ nomenclature: this program is geared towards the surveillance of people who can ill-afford to travel by plane; it mostly takes place at bus terminals.)

jetway was developed in the US but has been used across canada for about 5 years. according to our courts, this program targets travelers said to look “out of the norm” in terms of their clothing, their behaviour, or their actions. once targeted, the abnormal-looking-individual is approached by a police officer along with a four-legged friend. the officer immediately shows police identification and engages the target in conversation, watching all the while for unusual behavior. the target is either discounted quickly and allowed to walk away, or is further engaged in a conversation that swiftly becomes more “personal and intrusive”. police describe these encounters as “strictly consensual”, claiming that the target is free at any time during the conversation to walk away. however, the police will often demand to see the target’s travel tickets and identification; sometimes this leads to a further ‘request’ that the target ‘consent’ to a baggage search. targeted persons usually capitulate. most don’t realize that they have any choice in the matter.

even if a target refuses to consent to a baggage search, the police dog does with its nose what the officer was not permitted to do with his or her hands and eyes: the pooch determines the contents in the bag and reacts in response to certain forms of contraband. permission or no permission, the sniffer dog sniffs.

funded to the tune of more than $500,000, canada’s federal jetway training course has been responsible for training hundreds of RCMP officers and enforcement officers from other agencies. similar provincial and municipal programs exist. approximately 5% of officers trained are reported to participate in it daily.

Q – do these sniffer dog programs constitute a ‘search’ of the sort that ought to invoke constitutional safeguards?

it turns out that canadian courts are all over the map on this…

(so much so that i am willing to supervise the PhD of anyone who can convince me that they could provide a theoretical account that reconciles the different decisions that the courts have rendered on the issue.)

consider the following sample of judicial pronouncements from across canada:

1. “The use of investigative tools and aides such as police dogs to detect contraband or explosives on public buses is not beyond the realms of reasonable expectations of the traveling public.
The dog sniff does not constitute a “search” within the purview of section 8 of the Charter. As there was no “search”, there could be no breach of [the target’s] right to be secure from unreasonable search or seizure.” [R. v. Gosse at para 28 and 40 (New Brunswick)]
2. “I find that the police conducted searches without the consent of [the targets] prior to their arrest by the use of the police dog. I reject the argument that [the dog] was simply used as an investigatory technique. It is clear from the evidence … that the dog was extremely reliable in detecting the odour of drugs emanating, as previously stated, either from drugs themselves, a recent presence of drugs, or items such as cash that have been in the presence of drugs or handled by persons who have themselves handled drugs. The sole purpose of the dog being at the bus depot that day was to assist the officers in locating drugs.” [R. v. Dinh at para 28 (Alberta)]
3. “In conclusion, I am of the opinion the [target] did not have a subjective expectation of privacy that could reasonably be supported. [The target] chose to travel by public transport which would provide no control or protection from others entering his immediate space. The use of dogs by police was known and he was aware of the effect of passing in close proximity of such a dog. The use of trained police dogs to detect the scent of contraband in public areas such as train, bus and airplane depots is a legitimate police investigatory tool and does not infringe on any legitimate privacy interest protected by section 8 of the Charter.” [R. v. McCarthy at para 36 (Nova Scotia)]
4. “I am not persuaded that the judgment of the Supreme Court of Canada in Tessling is supportive of the … position that a dog sniff is not a search. In Tessling, the house of the accused was specifically targeted as a result of information that the accused was involved in a marijuana grow operation. I see a significant difference between a plane flying over the exterior of a building (on the basis of information received) and the taking of pictures of heat patterns emanating from the building, and a trained police dog sniffing at the personal effects of [the targets] in a random police search.” [R. v. A. M. at para 47 (Ontario)]
5. “Justice Binnie in Tessling notes that FLIR imaging generates information about the home but section 8 protects people, not places. As I noted earlier, he emphasizes the fact that the information generated by FLIR imaging about the respondent does not touch on "a biographical core of personal information", nor does it "tend to reveal intimate details of his lifestyle".
Nor does the information that the dog’s actions supply.”
I conclude that [the target] did not have a reasonable expectation of privacy in the area surrounding his vehicle. The dog sniff did not constitute a search. [R. v. Davis at para 21-23 (British Columbia)]

if one were to start counting judicial noses in the dozen or so reported canadian decisions, almost half of them have held that the use of sniffer dogs without a warrant constitutes a search that infringes the section 8 Charter guarantee to be secure against unreasonable search and seizure. slightly more than half deny that the use of sniffer dogs constitutes a search — usually on the basis that people do not have a reasonable expectation of privacy in the smells that emanate from their personal effects.

part of the explanation for the apparent schizophrenia in the caselaw is that the reasonable expectation of privacy test requires a decision based on the ‘totality of the circumstances’. to be fair to the courts, the fact patterns for the above decisions range from dogs sniffing knapsacks at bus depots, to dogs sniffing rental cars on open roads, to dogs sniffing kids’ school lockers. it is not surprising that such different facts could lead to different judicial pronouncements regarding the reasonable expectation of privacy in at least some cases.

judicial inconsistencies aside, there are, in my view, several other troubling aspects to the jurisprudence.

for starters, i am uneasy about the pickwickian logic adopted by an overwhelming majority of our courts across canada. it goes roughly like this: if there is no reasonable expectation of privacy, then there was no search. in other words, if i do not have an expectation of privacy in the smells emanating from my backpack, then when three police officers show up at my law school and randomly comb the halls with their trusty german sheppard, sniffing for students with dope, this is not to be considered a police search.

to me, this smacks of humpty dumpty’s scornful response to alice in through the looking glass that, "When I use a word it means just what I choose it to mean – neither more nor less." if police and their dogs are on duty and doing their thing but are not ‘searching’, then what exactly are they doing?!

this approach to defining police searches is further problematized by the recent trend to understand and define privacy in terms of informational privacy. my concern, one shared by many of the participants at our NJI workshop, is that the informational privacy approach is excessively reductionist in nature. once police activities are understood as nothing more than ‘capturing heat emanating from the wall of a building’ or ‘intercepting chemical emissions oozing through a backpack’, it is no longer possible to appreciate the deep social significance of RCMP planes beaming infrared lights at our homes in the middle of the night or OPP police officers and their guard dogs randomly patrolling our high schools, city streets and bus stations. (thankfully, as indicated in quotation #4 above, this point was not lost on the Ontario Court of Appeal.)

i am also troubled by the fact that, practically speaking, it matters squat what the courts think about privacy or how they define a search. regardless of whether sniffer dogs are said to conduct searches or whether the court finds that a target has a reasonable expectation of privacy in escaping odours, the evidence gathered through jetway programs is, at the end of the day, almost always admitted in law courts. according to most courts, to exclude the evidence, would bring the administration of justice into disrepute.

by admitting evidence in spite of the fact that it was obtained through a privacy breach, and/or by failing to provide any alternative remedies in the case of such privacy breaches, the courts are relinquishing the strongest deterrent available to prevent police from orchestrating investigations that are designed to interfere with privacy. without deterrents or remedies, such investigatory techniques are sure to become standard practice. and once they are standard practice – you guessed it – it becomes unreasonable for us to expect the police to act otherwise. quotation #3 above demonstrates this point nicely: since the target must have known that police regularly use dogs to sniff out drugs, he could no longer reasonably expect privacy with regard to smells emanating from his personal effects.

here, the notion of ‘expectation’ is completely stripped of its previous normative commitments.

instead, we are forced, as herbert hart might have put it, to take an external perspective of our expectations of privacy. like holmes’ ‘bad man’, who has not internalized the law as a reason for behaving a certain way but only sees legal rules as mere predictions about what the courts will do in fact, the reasoning adopted in quotation #3 above and by many members of the canadian judiciary tends to reduce our privacy expectations to nothing more than predictions about how the police will in fact behave and what technologies the consider state of the art.

its no longer about how they ought to behave.

the discourse is no longer centered on democracy, rights, duties or even interests. it is about the state of the art and the current standards of practice. as such, the ‘reasonable expectations’ test becomes a strange kind of factual inquiry.

the reasoning in quotation #3 above perfectly illustrates the crucial problem with the ‘reasonable expectations’ standard stripped of its normative meaning. once an expectation is understood as nothing more than a prediction: if you want to change the standard, all you have to do is change the expectations. and if you want to change the expectations, all you have to do is change the standard.

it’s a circle that rolls round upon itself.

and once that ball is rolling, it doesn’t take long to snowball. although i am quite certain that this is not what justice binnie had intended, the post-tessling trend in courts across canada has been to reduce our understanding of police search activities to an impersonal, non-social, merely informational transaction in a way that tends to shrink our reasonable expectations of privacy.

i find this trend particularly disconcerting in light of concurrent surveillance programs in the private sector and in light of rapidly developing surveillance technologies.

in what has to be one of the most unnoticed ‘anti-piracy’ surveillance news stories in 2006, the motion picture association of america very recently sponsored a world tour of two sniffer dogs named ‘lucky’ and ‘flo’. this gorgeous pair of black labs can be seen in this video sniffing-out polycarbonates, a byproduct of CDs & DVDs. with demonstrations across north america, central america, europe and asia, the purpose of this tour is to convince customs agents and border authorities, worldwide, to use anti-piracy canine units at airports, seaports, and anywhere else that bootleg CDs & DVDs are being transported. (a special shout-out to JereMe for bringing this story to my attention)

clearly, this is absolute תעגושמ. but, I assure you, it is not fiction. far from it.

and in the same world where dogs are being trained to sniff-out DVDs in gym bags, technologists are perfecting new means of remote sniffing. from simple devices that detect, measure and analyze electricity consumption to gas chromatography and other advanced forms of machine olfaction that are used to detect, measure and analyze odours in the air that even dogs cannot.

to take one very primitive technological example, consider digital recording ammeters (DRA). DRA is a technology that is capable of measuring the flow of electricity and producing graphical representations of the cycle of electrical consumption that takes place within a residence. among other things, these graphs can be used to identify grow-ops which, as it turns out, produce a very particular pattern of electrical flow. [grow-ops tend to use 18 hours of light and 6 hours of darkness to grow the plants and then switch to 12 hours of light and 12 hours of darkness in order to simulate autumn, thus producing the buds, which are the saleable product from the marijuana plant.]

with DRA, the police no longer need the expensive infrared fly-overs used in tessling. they just need to hook-up one of these load profile devices to a nearby public utilities pole; DRA can determine with nearly 100% accuracy whether there is a grow-op (though the DRA cannot, with any degree certainty, determine what is being grown). of course, DRA could also be used to determine other activities going on inside a home.

canadian courts have considered whether the police’s use of DRA constitutes an unreasonable search. like the sniffer dog cases, these decisions are all over the map. in a slender majority of the cases that i have read (6 to 5), courts have applied a tessling type analysis to DRA. for example, this court, this court, this court and this court all held that use of the DRA device to monitor a home was not a search and it therefore did not interfere with the target’s reasonable expectation of privacy — even though the entire reason for using DRA was to surreptitiously determine the nature of a target’s activities inside of a dwelling without transgressing its physical boundaries.

in my view, DRA and other primitive technologies need to be understood in light of one of the most rapidly developing areas in the field of information and communications technology: sensor networks. through the use of wireless technologies consisting of spatially distributed autonomous sensor devices, we are developing an astonishing capability to monitor and meter personal, physical and environmental conditions at greater and greater distances.

and, practically anyone can use these devices for practically innumerable purposes!

it doesn’t require much imagination think beyond today’s prototypes. consider the ingenious feral robotic dogs. with a few clever hacks, commercially available toy dogs (such as the famous sony aibo) have been turned into robotic sniffer dogs — enabling citizens to ‘sniff-out’ corporate contaminants from remote distances. granted, this project (by the brilliant natalie jeremijenko) is a happy use of sensor networks. a form of counter-corporate sousveillance. but it is not hard to see that there will be other uses, less happy.

the stunning technological developments that are just around the corner should give us some pause when we think about the simplistic and reductionistic way that our courts are becoming more and more inclined to think about the sniffer dog cases in particular and the reasonable expectation of privacy in general.

okay. newton has come downstairs to sniff-out a second breakfast and i hear erin’s quiet footsteps, so i’ll end now with an allegory.

kurt cobain, the troubled soul behind the legendary grunge band, nirvana, set out in the early 1990s to write what he called, “the ultimate pop song”; a song that would bust-down the barricade between alternative and mainstream rock music and perhaps serve as an anthem for generation x; a song that he called smells like teen spirit.

according to pop folklore, the rocker subsequently described as the spokesman for (or, was it against?!) the coming generation of excessive consumption borrowed the song title from a line spray-painted on the wall of his bedroom by his pal kathleen hanna:


as the story goes, since he and kathleen (lead singer of the riot grrrl punk band bikini kill) had recently spent late evenings talking about the politics of anarchy, the future of alternative music and the plagues of humanity, cobain took her graffiti message as a slogan expressing how he had captured the spirit of a generation through his music.

apparently, kurt cobain had no idea what he smelled like!!

comically or tragically (depending on one’s point of view), the paint sprawl had a less inspired meaning. it turns out that hanna’s words were to be taken literally. she simply meant to say that kurt smelled like Teen Spirit,™ the deodorant worn by tobi vail (kathleen hanna's bandmate and kurt's then pelvic affiliate). cobain had not heard of this colgate product, nor had he realized that his friends thought of him as branded by his soon-to-be-causual-ex-partner’s scent.

i suspect that none of us really know what we smell of, or who is smellin’ us.


a denial, a denial, a denial, a denial, a denial,
a denial, a denial, a denial, a denial

ian kerr
new years day, 2007


An interesting article and thought-provoking observation on the perceived erosion of rights and privacy. One of the points that seems to be missing from this reader's standpoint is the issue is not one of "reasonable expectation of privacy" so much as a prevention of an "UNREASONABLE search". Do we expect in a public place to have an invisible plastic bubble around ourselves and not have any unwanted contact with third parties (especially law enforcement)? But when does privacy become a weapon to be used by wrongdoers to prevent reasonable protections for those around the wrongdoer? Do you feel safer in an airplane knowing that every passenger has gone through a security check? Do you wish to know if a grow-op with millions of dollars of drugs is down the street from you, and the potential for criminal activity increasing dramatically for you and your children? Seems as though the average law-abiding citizen is willing to accept the annoyance of a breathalyzer at Christmas time on the roads, a security scan at the airport or a dog smelling their luggage at a bus depot. Is this a slippery slope into Big Brother and eastern bloc midnight invasions into homes and the gulag? I don't see it that way anyway. And extending the logic to have my baggage checked by a DVD sniffing dog to prevent my illegal copyright infringement would protect the wrongdoer and not the legitimate interests of the owners.

Posted by: Rob at January 11, 2007 10:47 AM

As always if you get to frame the question you can influence the answer. The blogger poses the question as:

whether external patterns of smell on the external surface of a knapsack is, or is not, information in which a person holds a reasonable expectation of privacy.

This of course is not what a dog smells. Rather it smells the air surrounding the knapsack which contain molecules of the smelly substance which have escaped (bad luck or bad management by the bad guy) into that air. So the question is whether there is a reasonable (constitutional) expectation of privacy in that air as it wafts away from the knapsack. I know some smelly emanations from some people that should be kept private. Imagine that poor dog in a room full of such people.
Also the dog has not been specially trained to smell this stuff. He always could. What he has been trained to do is to tell his handler when he has come across that smell. I wish I could find one that could smell out winning lottery tickets.

Posted by: Doug at January 11, 2007 02:49 PM

Hi Ian,

Thanks for this fabulous post. I think you've hit the nail squarely on the head with your comment about the problematic logic of the inference from "there was no reasonable expecation of privacy" to "there was no [unreasonable] search". Even if one is inclined to agree with the Court in _Tessling_ that the respondent had no reasonable expectation of privacy with respect to the information gleaned by the FLIR cameras because it wasn't personal information about him, it simply doesn't follow that the manner in which the information was acquired must be okay. I think it's pretty clear that there can be many cases in which there is an unreasonable search even for nonpersonal information. (Many stalkers, I suspect, sadly get their kicks precisely by searching for information about their targets that is in some sense publicly available. That hardly makes their searching reasonable.)

Perhaps what we're dealing with in cases like _Tessling_ and the sniffer dogs is an overemphasis on the reasonable expectation of privacy. If the intuitive wrongness of the searches in such cases is not captured by our considered opinions about whether individuals had their reasonable expectation of privacy violated, maybe we should be looking elsewhere. I'm willing to bet that there's a lot more to the notion of an unreasonable search than that.

Posted by: David Matheson at January 12, 2007 04:23 PM

I understand the concern for privacy and the state's interest in "providing security" to its citizens. Both are very legitimate concerns, but as they say, one citizen's rights end just where the other citizen's rights begin. The writer's point seems extremely relevant to our daily lives; since as he exposes: reasonable expectations shifts as new standards are set. My concern is then, that as we gather more technological widgets, this reasonable expectation (or not) of privacy will disappear. So, in the end, the potential problem presented by this issue is that we might not have any expectations of privacy. Just like the DRA widget, you can no longer have reasonable expectation of privacy in terms of how you spend time at home (or your money on electricity for that matter).

Posted by: Miguel Nicolas Moreda at January 13, 2007 12:53 PM

Post a comment

Remember Me?

main display area bottom border

.:privacy:. | .:contact:.

This is a SSHRC funded project:
Social Sciences and Humanities Research Council of Canada