it doesn't make out a compelling interest on these facts, and that's the question presented. >> could you -- you know, we seem to be arguing rules. they say no beards. you say half inch is okay. and then the question begs itself how about 3/4 of an inch, how about an inch, or a full beard. what are we measuring this against? are you seeking to establish a rule that every prisoner has to be permitted to grow half inch beard and no more? or are you asking for a rule that applies just to your client and then articulate why for your client? >> no. we think a reversal here would establish a right to a half inch beard for all prisoners on this record unless some other state made a very different showing. all prisoners --

>> so what happens -- i know the magistrate judge or the judge below here said that it was preposterous to think that this prisoner could hide anything in his half inch beard. assuming that his half inch beard was not thickly grown, but some are, and some you can't see the skin. should that half inch issue be applied to that prisoner? wouldn't it be a different set of facts in that case to consider? >> well, a state might be able to show that's a different set of facts. you know, but the question is not just is it conceivably imaginable that some prisoner somewhere could hide something in a half inch beard. but could he hide something there that he couldn't much more easily and more securely hide in the hair on top of his head, in his shoe, in the lining of his clothes? >> for all half inch beards, right? >> yes. >> we have to assume all half inch beards are okay if god

tells you to grow them, right? >> well, you know, i think that's right, and again subject to somebody producing evidence that we're wrong about some half inch beards but i think, yes. >> well, i mean, whose burden is that? >> it is the state's burden, that is explicit in the statute. this is an affirmative defense we're talking about of compelling interest and least restrictive means. the only limit they impose on the hair on top of your head is it can't extend below the middle of the neck. you can have long hair, curly hair, afros on top of your head without a length limit. the hair on top of your head and the hair on the front of your head is not even rational. they could hide more and the prison warden testified to this, you could hide things on the hair on top of your head but that's not against the rules. they have singled out the beard

that is preserved -- that is more for religious reasons and not treated other things that are really indistinguishable. if there are no further questions, i'll reserve the remaining time. >> thank you, professor. >> mr. chief justice, may it please the court, this case involves the religious accommodation of a half inch long beard that bop and over 40 states would allow a prisoner to wear. the state failed in its burden of proving that denying a half inch religious beard would be the least restrictive means to further a compelling interest. >> is it your position that if, what, 90% of other institutions similar to the one at issue in the case permit the practice that is challenged, it cannot be a compelling interest? is that your position? >> no.

in fact, i think our position is that security interests in prisons are compelling, but the buen that is imposed p ed upon state is a burden to show the means selected, the denial of a half inch beard, is the least restrictive means to further that interest, and in this case that's a showing that has to be made in court. it is something that congress specifically recognized would be -- >> i understand that, but i'm asking what is the relevance of the fact that other -- >> well, the relevance -- >> are we going to say whenever an institution comes up here that has a restriction which other institutions of the same type do not have or at least a large majority of them, it's ipso facto bad? >> no, i don't think it's a dispositive factor, however it significantly undermines the state's showing that a similar restriction could not be in this context. and the state in the face of this example if raised in

litigation would have to provide a reasonable basis for explaining why there might be a distinction between what's going on in other states and what's going on in this state. and the showing in this case is exceptionally meager. >> why does it have to show a distinction? it can just say the other states are wrong. we think it's dangerous. i don't care what other states think. >> i think a little more will be required under the compelling interest test. now, this court in cutter -- >> suppose a state just simply says this, you know, actually there's nothing special about our prison. we can't show you some special circumstance, but what we can show you is that prisons are in the business of making trade-offs between security and other values. they do that every single day, and our state just thinks that the trade-off should be more security oriented. so we insist on a greater level of security than our peer institutions do. are you saying the statute 3re ven prevents a state from doing that?

>> no, i don't think the statute imposes a least common denominator amongst state prison systems. however, i think there is some bounds to the state's judgment that needs to be -- there are some limits, and the state needs to provide a reasoned explanation in order to get deference to its predictive judgment in this context. >> how -- >> go ahead. >> how do you reconcile deference with t strict scrutiny that the statute requires? >> well, this court in cutter, for instance, explained that the strict scrutiny -- when deciding what was required by strict scrutiny context -- >> there are two questions there. one is a good question. with does this idea of deference come from in this context but the question i was asking is assuming there is a role for deference, how does it fit together with what the statute expressly requires? >> i don't think there's any

dissidence between the idea of strict strut ni which is not a degree of proof required. it's whether you identified a compelling interest and shown that the burden is least restrictive means. that can be shown by at the pron der reince of the evidence. based on the fact they are charged with protecting the public and administering these prisons. when they provide a reasoned explanation based on experience and expertise, they don't have to point to a specific example of a half inch beard in the past resulting in something horrible -- >> i share justin alito's confusion on this point. all the things you're talking about are things we would never allow in the typical strict scrutiny context. you know, all this kind of, well, as long as they say something, they don't really have to prove it and it just has to sound kind of reasonable. that's the very opposite of

strict snut scrutiny generally. >> well, in this con tectextcon statute doesn't say strict scrutiny. >> it sounds like -- >> it's similar but remember this court in cutter has recognized that the application -- what you might broadly label strict scrutiny depends on context and in congress when they enacted laws understood these two concepts could be administered together. >> do you think it's the same standard in both of those statutes? >> for the large part, yes. i want to caveat that because -- >> the language is completely the same. >> yes, but in -- most of the language is completely the same. the standard is the same. but in one there is an additional profession that requires that the terms of the statute be broadly construed to the maximum extent possible to protect religious -- >> but i thought we actually

used that in the recent case to suggest something about riffra. >> that's because riffra definition of religious exercise incorporates the other definition. it's conceivable the court might have a broader -- >> mr. yang, before you sit down, your brief lists a whole series of cases on page 14 that were decided before rolupa. are all those practices which we approved up for grabs now under rolupa? there were restrictions on receipt of publications was one. >> i think the analysis is different now. it could be litigated. any of these claims could be litigated. the state would then have the burden of coming forward to show that the restriction would, in

fact, be a least restrictive -- >> so all those we approved, correspondence limitations, all those would have to be looked at anew under the standard? >> i think that's right because if you were to go back to pre-rolupa case law, no one would doubt that a state could, in fact, prohibit a half inch beard under the prior constitutional standards, but congress has set a higher bar and it imposes upon states the obligation to come forward to explain and justify it. now -- >> where are you on the full beard? >> on the full beard, i think there might well be a difference, but, again, rolupa depends on a showing in litigation by the state that the means selected is least restrictive means. a state may well be able to show that a full beard would run real risks that are just not present in the half inch beard that we have here. >> assuming that's the case, assuming they have some evidence of concealment or whatever in a

full beard, what do we do? just litigate a dozen cases until we settle on 1 3/4 inches or what? >> i think -- >> what i'm doing -- it's the same question i asked -- >> there's not -- >> what's the legal principle and if there is no direct legal principle, then isn't it a situation in which you would employ deference to the administrative judgment? >> i think that's exactly right, that there's going to be a bound, a range of reasonableness that courts will find appropriate to defer to predictive judgments by expert officials in various contexts. >> can i ask a similar question but, you know, lots of religions, including lots of religious of one, have dietary codes of various kinds. suppose a lot of prisoners say here is my dietary code personal to me and all of that costs money and let's stipulate as prisons have to spend money on that, they have less money to spend on things relating to

internal security. how is somebody supposed to think about those kinds of questions where it's just every time somebody makes a religious claim, the costs to the institution goes up and the ability of the institution to deal with security issues goes down? >> well, maybe not necessarily always the second. certainly you can have costs going up. it may not necessarily affect the -- >> but i'm -- they come out of some place. >> at some limit that's true. i mean, we all operate under a real world with limited costs and as the court recognized in cutter -- >> how do we do that? >> excuse me? >> so how do we do that? >> i think again it's going to depend. if the factor that is would be relevant is the context. does the increased cost prejudice other types of interests in operating the prison. that would have to be articulated by -- >> wait a minute. i mean, so vote more money. all you have to do is raise taxes. we're talking here about a compelling state interest. bear in mind, i would not have enacted this statute, but there

it is. it says there has to be a compelling state interest. and you're asking, let's balance things and be reasonable. compelling state interest is not a reasonable interest at all. >> it's not meyour honor. i think it's what the court actually recognized in cutter. the court in all quotes said the act needs to be applied in an appropriately balanced way with particular sensitive to security concerns and that accommodation must be measured so it does not override other significant interests. >> thank you, counsel. >> thank you, your honor. >> mr. kern. >> mr. chief justice, may it please the court, arkansas's security objectives are undermined by the petitioner's half inch beard because he could use it to alter his appearance, thwart identification, and conceal contraband in our

maximum security -- >> altering appearance, i thought it was conceded that at intake the prison could take a photograph clean shaven. >> your honor, that is not on the record. i concede that that was not sufficiently addressed to withstand the summary judgment posture in this case. the record testimony was just -- i agree with professor laycock it wasn't satisfactory. let me get to -- >> i don't understand what you just said. >> well, there's two points in altered appearance. one is identification within the prison itself and i want to get to that in one second. one is a post-escape scenario. where you're looking for the inmate. our testimony on that was not engaging in the question. there's no record testimony regarding that. so let me identification -- >> why do you need record testimony on a question such as that? if you're claiming if he escapes he can shave off his half inch

beard and thereby alter his appearance and the response is, well, just take a photograph of him before he grows his half inch beard, why do you need evidence on that point? it seems to me it's obvious. what prevents you from taking a photograph before he grows the half inch beard? which can then be distributed to police departments if he escapes? >> i agree, your honor. there are -- >> it's not an evidentiary matter at all. >> the point of identification within the prison though is an evidentiary matter on this record. and let me get to that because it's very important to understand this in our prison's unique environment. shaving a beard can enable an inmate to get into an area where he's not supposed to be in. that's a joint appendix page 104 and that a beard can enable an inmate to deviate from an inmate's appearance on an i.d.

batch, and, of course, a beard is one of the quickest and easiest ways to change one's appearance. that's on page 97 of the joint appendix. of course, the grooming policy itself speaks in terms of maintaining the standard appearance throughout the period of incarceration. let me explain a minute why this matters in our prison's unique environment and why we're different because it's a very important point here. the testimony on the record, page 101 of the joint appendix, was that we have a very different situation with barrac housing and inmates going outside the fence in large groups of 30 to 60 per barracks unit every day. it's a very high traffic maximum security facility where they come out of a large barracks holding 30 or 60 inmates, go out, and come back. there's a lot of traffic. in that environment rapid and accurate identification of the inmate by his face, his i.d. badge and the like, but also general familiarity with the

inmate is very important in that process. and if a mistake is made, an inmate could get into the barracks where he is not supposed to be in and an assault could occur. these are separated by enemies and the like and that is very serious in our environment. it was made on the record and it differentiates arkansas from every state mentioned by the petitioner and the united states. >> but you have no example of that ever happening. >> i have no example of a -- well, let me say this. in our brief on footnote 13 and on page 26 of the 18 states amicus briefs there are examples. there are no -- >> examples of what? of interprison identification problems. in the prison, a beard being used to thwart identification, and -- >> do you have that same concern with the pri er prisoners who h a very short beard for medical reasons?

jo n >> no, and let me explain why. there is confusion about what this so-called medical beard is. there is no exception of practice of a quantitative matter for medical beards. it is a means of shaving exception. our policy changed to reflect our actual practice about a year ago. what the practice is is that when a doctor's order says a person has a dermatological condition or some other scarring or skin condition that needs a shave they use barber style clippers, electric clippers, without a guard and they're used directly on the skin, and the result is a very clean shaven look. not quite as close as using a tamper resistant safety razor that other inmates use, but it is a very clean shaven face. the clippers are kept in the barber facility and a couple days -- >> how long are the whiskers when that is done? >> so they may take barber call maybe twice a week. so they will have a clean shaven face and then go a couple days, three days, and then go back to

the barber facility. >> are you saying they're completely clean shaven? >> i am saying they are -- what is clean shaven? some would say a razor -- >> clean shaven is somebody like you. >> i would say that's -- i have got a fairly dense hair but that's the appearance immediately after. >> and that would satisfy the medical problem? >> that's correct. >> to be shaved that closely? >> that's right. the doctors prescriptions invariably are getting a clipper shave and that brings a second point up, your honor, the policy's rationale is follow doctor's orders and we think that is fundamentally of a different nature than a religious reason because the eighth amendment law of deliberate indifference and the like admits of no countervailing security interest that come into play. our policy is we follow doctors orders and that's the end of the matter. undethe medical -- >> are you telling us that the quarter inch is wrong? i thought that that was in the

record as a given, that a quarter inch is allowed for medical reasons. >> the policy states that, your honor, and it's confusing. in practice there is no quantitative quarter inch rule for beards. there's clean shaven that's allowed some length to go to the next barber call. you can still see the skin the entire time in that scenario and if petitioner wanted to avail himself of that accommodation, we would let him do that. >> what about the argument that it's, never mind the least restrictive means, you have no comparable rule about hair on one's head where it seems more could be hidden than in a beard where you hide something in a beard, it might drop out. >> the material difference there isour professional difference is the disguise component of a

beard and shaving a beard is more than that on the head. yours is one to contraband. there's a length and gravity component to a head that's different than a beard for sure. the risk is still there. there is an interest in regulating the contraband element, but the head hair doesn't pose the same disguise related problem as a beard. >> why is that so? are you saying somebody with or without a half inch beard, that's a bigger difference than somebody who has longish hair versus the same person with a shaved head? >> in our professional judgment it is, yes, that's correct. because you're looking at the e sex features of a person face, the jaw lines and chin, and that's the mean by which we identify each other, so that is a significant difference in our vi and really the head hair policy complements the facial identification policy because it's not allowed to get to a length that could obscure the hair and that's the rationale

for that. >> mr. laycock characterizes your position as being all deference all the time. so i'll give you an opportunity to say when would deference be inappropriate? >> deference would be inappropriate when the explanation offered on the witness stand in the record of compelling interest and the least restrictive means is neither comports with logic or common sense. it sounds like we're on agreement on that. justice sotomayor asked which lower court decision would lend the most guidance. i think a straightforward but apt analysis is in the couch case in the fourth circuit where judge tracks ler joined by justice o'connor and judge shed went through the initial obligation is to explain the reason and common sense why that approach furthers a compelling interest and a least restrictive means, and once that happens, deference attaches but that doesn't mean either that you win

the case. it just means you have substantial weight. you're sort of a thumb on the scale so to speak and more evidence can come into play and you can still lose. >> can i go back to just so i'm clear in my head, is it two compelling interests, one in identification, one in contraband. >> that's right, justice sotomayor. >> is there a third or a fourth or are those the only two? >> those are the only two we're talking about. in this case the magistrate judge said it's preposterous to think you can hide something. so you don't have a security contraband -- >> let me take on preposterous if i might. if you look at the written findings, there is no such finding. in fact, the finding was to the contrary. i believe it's on page 167 of the joint appendix. the magistrate judge says the testimony about small bits of dangerous contraband is the most compelling in the entire case. then if you go back and look at the verbal musings from the bench, the judge sort of reflects a layman's view of,

well, the idea of contraband in a half inch beard seems almost preposterous and then the next par graph he says well then i heard the experience of highly experienced correctional professionals and they made me change my mind. he used the word impressed. the word impressed is 155 of the joint appendix. i think so let's assume what the

magistrate judge meant, which is what i assumed, you have a different read, that it's preposterous to think this prisoner could hide something in his beard but not preposterous to think that others might not be able to do so. assume my hypothetical. >> right. the question -- you pose a question i think, justice sotomayor, is whether the warden needs to do some sort of hair analysis of -- >> no, no, no, no. my question is one of whether you're obligated under this statute to look at the request of the individual and assume that the application of whatever rule you create can't have an exception as to that individual. >> on this testimony, your honor, you pointed out the testimony was a half inch beard, you can't see the skin. i think that's a functional difference, and we've got to think of how to administer a

rule where to the person that that level of granularity is just not functional. >> yes. but i don't know given the deference that was given here of the question, was it applied too broadly? what i'm getting at is does a court have to look at the individual request and figure out whether it can be accommodated in the least restrictive way. >> yes, i think it's fair to say that if the court actually did say it was preposterous, saying it defies common sense -- >> in this case. >> i think that's right, but i don't think that's an accurate finding of what the magistrate judge says and i think that's a problem whip that rule is not administratable. easier to say how about an eighth inch of beard where in all scenarioings ys you can't s skin. if he asks for that kind of accommodation, we would grant it. but he's not offered that. he's offered a half an inch, and

he's got a very complex lesser of evils type principle. >> on the change of looks, i'm still not sure. could you describe in more detail. i obviously missed it in the record. what is this barracks situation? where do they go when they leave the compound? >> so they're in a barracks situation. they have 30 to 50 or so in a room, and there are four barracks on each side of a common roof. they go out and get in a line and different shifts and they go out and they will go to chow and then they will do their business and they'll go out and work outside the prison fence in fields and they will come back again. it's a very high traffic environment. >> are these unprotected fields? >> there's guards there watching them. they're not just out working alone, but there's no prison fence there. it's up to the guard to keep up with them. what you have is an environment which on this record on page 101 of the joint i pen dix is we're not like california, we're not

like new york. they have cellblock housing, and there is no instance in which the government or the petitioner has said -- challenged that as to maximum security faciles. that's a big difference in the nature of how our institution runs. we think deference means anything -- it means you don't have to copy the prison policies of other states who don't even have the similar security concerns. >> did you establish that arkansas is unlike all these other states, that the other states don't have barracks, they don't have people going out to work in the field? i thought that that was not so, that there are other prisons that operate similarly with housing and having the prisoners work on a farm. >> two things in response to that, justice ginsburg. first is on this record there is only two states offered, california and new york, and the indisputed testimony in this record is they're different.

your question is what about all the footnotes and briefs to the court, if you look behind all the sources cited on those internet sites, which that's what petitioner mostly uses, and the government uses internet sites and also some case law examples, each one of those was referring to a minimum security institution. they have not offered any institutions line ours. as far as i can tell the only institutions that have something similar to ours have clean shaven rules. >> what about the federal prison system? i thought the rule was throughout the prison system. >> both the government and the petitioner cite a link to the regulation in their briefs and that is to a minimum security status inmates. >> when a prisoner goes out in the field and then wants to come back to the barracks, the prisoner is wearing an i.d., is that correct? >> that's correct. >> does it say which barracks that prisoner is supposed to go to? >> yes. and what happens is they trade i.d.s and they trade shirts. that happens even now. >> does the i.d. have a picture on it?

>> yes. so the person guarding the barracks, so to speak, the flow to and from the barracks relies on the i.d. and the face, but also general familiarity with who he's working with because that's -- that happens. >> i'm having difficult envisioning the scenario you're suggesting. so a prisoner who is supposed to be in barracks "a" has a half inch beard, has an i.d. that says barracks "a," has that person's picture, goes out in the field, brings a razor with him while he's out there, he shaves. he wants to come back and go into barracks "b." how is he going to get into baracked "a" if he has a barracks that says bararacks "b? he will trade with another prisoner? >> they would alter the i.d. -- what happens is you've got very fast recognition and if they favor each other at all -- this happens now, your honor.

and the shave would take place probably in the barracks in the morning. when they come back, the person monitoring the flow of 60 inmates through there gets beaten and that happens. >> he has to find somebody who also looks like him from barracks "b." >> i would think that's how that scenario would work but that happens. prisoners are capable of doing a lot of mischief in prison as you understand i think, and that kind of thing happens even now. we have assaults in the wrong bararacks because correctional officers get beat. >> one of the hazards is razors. you just said that they can shave themselves in the barracks. where do they get those razors and what happens to them? >> we have tamper resistant safety razors that are issued and they keep them in their personal possession and then when they're threw with them they can turn them back in on a one-on-one basis and get a new one if they exchange an old one. >> does your standard -- how if

at all does your standard differ from what it would be if we had no rolupa. any case that would come out differently in arkansas under rolupa than under the pre-existing law? >> well, i have to kind of get to the different elements. i think -- i'll talk first about compelling interest. alone credited prison officials' testimony that muslim inmates are sort of getting a good rehabilitative event by not having to go back into the prison for friday prayer because they might as well get used to an intolerant employer when they're out in the free world, that won't pass muster under compelling interest anymore. that was the old standard. and the interest has to be truly compelling. on least restrictive means, we think that interests grounded mostly in cost and hassle would

have survived under the old regime which had a lot of dietary cases and the like. those probably will fail a lot more often under rolupa than under the previous standard. if inkrumental like the yellow bear case, maybe an increment increase in more staffing ever so slight might say that is required to pass least restrictive means but it wasn't under the prior standard. i think even a deferential approach grounded in logic and common sense you will have more vigor under the rolupa standard and cases will go the other way more often. >> -- about this whole issue about cost and the statute. 2000 cc 3 c which says this chapter may require a government to incur expenses in its own operations to avoid imposing a substantial burden on religious exercise.

>> yes, your honor. >> isn't it anticipating -- >> it is. >> there might be expense. >> that's exactly right. even within the least restrictive means analysis there's a particular statutory command. now, i think courts are going to have to manage that with reason. we can't have each inmate has his own facility with ten guards around it. there's going to have to be some limit. cost. i don't know that this case implicates much of a cost issue. >> it doesn't implicate the cost issue. >> right. >> i thought earlier that you had pretty much abandoned the concealment justification for the policy. do you still -- and were relying upon the identification justification. >> where we think each justification stands on its own weight -- >> you think something can be concealed within a half inch beard. >> on this record, something as small as a sim card which the

court found compelling could. i think the identification within the prison is more weighty here. >> as far as concealment is concerned, what is the difference between half inch beard and hair on the head that's much longer? >> well, the testimony on the record was that a common sense view that longer hair is a better way to conceal contraband i think is the right one. the question really is, is a beard an unlikely place? the testimony here is not only can something fit in a beard, but correctional officers very likely will be extra reluctant to do a full search of the beard like they would with head hair. >> i take it there's no example, not a single example in any state that allows beard policies where somebody did hide something in his beard. >> i think that's mostly right, your honor. >> we have no example. >> there's no -- >> there's no such example. then do you think it might fit

within the language of that report which says that the fear of people hiding things in their beards is, to use their language, was it grossly exaggerated? i mean, 42 states, you know what i'm quoting from. i'm quoting from the report there. the exact words are what they're trying to get at is exaggerated fears. it doesn't even say grossly. would you say it's an exaggerated fear that people would hide something in their beards when in a country of a very high prison population not one example has ever been found of anybody hiding anything in his beard as far as you can tell and as far as i can tell. >> as far as -- >> do i have that right? >> as far as i can tell but let me make a caveat i think is important which is that just because we haven't found the example doesn't mean they aren't there and the cots -- >> no, there are a lot of things we've never found that might be there and i will refrain from

mentioning them but you see them on television on weird programs from time to time. >> the following -- this is briarwidth, these kind of things are buried in incident reports among thousands of other things and this court in the florence case, justice kennedy asked the assistant to the solicitor general, i thought the evidence of contraband was very skimpy. i was surprised to see there weren't more of this and the attorney's response was these things are buried in incident reports. we can't find all of these examples. it's the nature of prisons, but take my newspaper articles and the court actually, you know, took note of that as confirming its common sense intuition there. and i think that's just a problem in the prison environment. >> as far as searching a beard is concerned, why can't the prison just give the inmate a comb. you could develop whatever kind of comb you want and say, comb your beard, and if there's anything there, if there's a sim card or a revolver or anything

else you think can be hidden in a half inch beard, a tiny revolver, it will fall out. >> you know, i suppose that's a possible alternative. i think the concern there is there's no perfect way o searching and there's a lot of area there and you're going to have to really monitor to make sure they get all the spots, but -- >> do you really think that would be difficult? to say here is a comb, comb your beard? >> i don't think it would be that difficult. i'm not in the prison environment. it really wasn't raised on moo record. my clients might think it is. based on the information i have, i would think it sounds like something that could be done. and i do think it's important to distinguish sort of the rule that i would propose and that is -- what i think is very similar to what the government is offering here is really an effort to marry strict scrutiny with deference in a way that doesn't invite impeer simple.

this court's strict scrutiny jurisprudence hadn't always demanded examples especially in the prison context and i think that's important that we do be allowed to have profal lactic rules in some settings. justice ginsburg asked about what about literature? we have a rule that says racially inflammatory literature of a religious nature that incites violence isn't allowed in the prison. justice ginsburg in the footnote in the cutter opinion seemed to think that, of course, that's a concern that prisons ought to be worried about. that's not susceptible to any kind of empirical proof i don't think and as i understand my friends' understanding of the rule, we're in a land of strict scrutiny rereally that's akin to content speech religated. we ought to be using after the fact deterrence measures against maximum security inmates. they've already shown themselves not to sort of comport with that view of how to behave, and i

think it's particularly dangerous in the prison setting particularly in our prisons environment. thank you, your honor. >> thaur, counsel. professor laycock, you have five minutes left. >> on the issue of the written findings, magistrate said it's almost propos terous to believe you can hide anything this" this beard and then he madly said but there's a larger principle here which is i have to defer to these people and there were subsequent written findings based on that mistaken label of deference. he said three times i'm constrained by the case. the eighth circuit applied the rule and he gave that level of deference and he made written findsings apart from what he had actually seen. on the issue of identification inside the prison, prisoners can shave their heads, shave their

mustache, shave their medical beards. they don't claim that's a significant problem. the other 43 states do not appear to have found this to be a significant problem. it is a small and manageable problem. on the question of the quarter inch medical beard, the policy is in the appendix to our brief at page 11a. this morning is the first time we've heard what's really not a quarter inch rule, really some other kind of rule. first time we heard we let a religious claimant have a medical beard. they never said that before. and, you know, they had not been able to justify their policy. they do have to prove it, justice kagan, if the proof comes close they get deference. if they offer serious evidence, they get deference, but here they offered very limited conclusory testimony, no examples in a situation where there should be plenty of

examples. you can't administer a prison and maintain any kind of safety and security if you don't have some sense of where prisoners hide things. they don't have to dig out the data from the files. if prisoners were routinely hiding things in beards, these two witnesses would have known that, would have remembered it from the earlier rule in arkansas, and it would be easy to get examples of that from other states. there's simply no evidence in this record that it's a significant problem. >> what about the argument that there's no comparison, that arkansas is unique in the way it houses its prisoners and that the rules that were cited elsewhere have to do with minimum security facilities? >> arkansas may be somewhat different in how it -- in the number of maximum security prisoners working outside in the fields, but that does not make the half inch beard any more attractive of a hiding place. if i'm out in the fields i'm trying to smuggle something back in, i still have lots of better

places to smuggle it, including my shoes and my pockets and the lining of my clothes and as mr. quran just agreed, the hair on top of my head. again, there's just not a rational difference between where on the head the hair is located. >> you made a statement just then, mr. laycock, about how to think about deference in the context of this statute. and it's something that still troubles me so i'm going to ask you to expand on that. i mean -- and to say -- just seems like a contradiction in terms. i want to understand how it's not a contradiction in terms. >> there's obviously some tension here, but the legislative history says due deference. cutter opinion quotes that legislative history and says due deference to expertise. it doesn't say how much deference is due. that's the question to be decided here. we think the more informed and considered and well explained their decision, the more deference it naturally deserves,

the more deference is due, but they have to take some account of the prisoners' religious needs. they have to take some account of solutions that have been found to work in other states. if it's something so dangerous no one would ever try it, then, of course, you wouldn't expect examples, but here 43 states have tried it. arkansas tried it for years. in a situation like that where there ought to be plenty of examples if there's a problem, they ought to be able to produce some of the examples. the degree of deference depends on their consideration of the issue. here there's no indication they ever considered the religious needs of the prisoners in the adoption of this rule. and the testimony is very conclusory, devoid of examples, devoid of attention to other jurisdictions. the level of deference cannot be so great as to negate the statutory standard. you have to administer deference

within that standard, not substitute deference for the standard and the statutory standard is still compelling interest and least restrictive means. thank you. >> thank you, counsel. case is submitted. >> president obama says internet providers should not be allowed to cut deals with online services like netflix, amazon, or youtube to move their content faster. the president called for a ban on such deals. in response texas senator ted cruz tweeted net neutrality is obamacare for the internet. the internet should not operate at the speed of government. congresswoman anna eshoo tweeted, i strongly urge the fcc to adopt the president's approach. tonight on "the communicators qurtion, christopher yew professor at the university of pennsylvania law school and director of the center for technology,

innovation, and competition. >> people who oppose prioritization should take a look at the internet header, the ipv foreheader. it is the guts. that is the magic that makes the internet work. there's something in there called the type of service flag. that's different service classes really for high bandwidth services, low latency services, different forms of prioritization that was designed to the internet from the beginning, and people say, oh, that's just an old artifact. when we redesigned it for ipc 6 they included another field called a label field. if you look at the engineering design to suggest -- prioritization was never intended to be loud, i think a little engineering knowledge goes a long way. it's a design feature of the network from the beginning and if you talk to the way people are using the knellwork, they're using it today to deliver, for example, voice services.

we've all called on skype and been frustrated. the true completely ip-based voice service to your phone is called voice over lte. all uses voice prioritization. >> tonight at 8:00 eastern on "the communicators" on c-span2. here are just few of the comments we've recently received from our viewers. >> i just watched your show this morning on domestic violence and was very disappointed with what i saw and heard. i thought the guests were both weak and ineffectual, and it seemed that the bulk of callers were a bunch of whiny men. one woman is beaten every 15 seconds in this country by a husband or a partner. that is one woman every 15 seconds. this issue, alarmingly, is swept

under the rug in this country probably and most likely because most of the perpetrators are male. the only way this will ever change is if men are willing to look at their own bad behavior and address it head on. >> i'm listening to your commentator and the one from the bloomberg news, and they're talking about 2,000 and some bills being on the harry reid's desk to be presented for whatever. well, each and every one of those bills have a repeal of what they call obamacare or of the affordable care act. so, you know, whomever is your commentator, and that needs to bring out that point. i just heard the comments on a lady who called in and said i'm watching the show recorded, by the way, that it would be

good rather than having democrats leave a comment and then republicans leave a comment and then just airing them one at a time to let democrats and the republicans basically fight it out it sounds like verbally on the show. if you ever decide to do that, i'm up for that. >> continue to let us know what you think about the programs you're watching. call us at 202-626-3400. e-mail us at comments@cspan.org or send us a tweet at c-span #comments. like us on facebook, follow us on twitter. a look now at the future of self-driving cars. cato institute panelists discuss technology breakthroughs, the impact on urban transit, and privacy and security concerns. after their remarks, they took questions from the audience. it's about 90 minutes.

good afternoon, everyone. welcome to the cato institute. my name is matthew feeney, a policy analyst here. i'm very excited to welcome you here for what will i hope be an interesting discussion on self-driving cars. it's an interesting topic because we're at a stage in

history when where asking questions when it comes to the widespread use of these vehicles and we have three experts here to help talk about it. before we begin, i'd like all of you to please turn off your cell phones so we won't be interrupted i would also like to say we will do a q&a session after all speakers have had their turn. at which pointly be calling on you the three speakers we have today are randy o'toole, mark scribner and they'll be discussing a range of different issues such as transit and regional planning, the legal and regulatory issues as well as privacy. now the first speaker is randall oh o'toole. i'll introduce them before they speak. randall o'toole is a senior fellow at the kato institute where he works on urban growth, public land and transportation issues. he's the author of a number of books including "gridlock, why we're stuck in traffic and what to do about it" "the vanishing automobile and other myths"

"reforming the forest service" and "american nightmare, how government undermines the dream of homeownership." his writings have appeared in numerous national journals and newspapers and is the author of the most recent policy analyst that will be out later, the policy of implications of autonomous vehicles. he was educated in forestry at oregon state university and economics at the university of oregon. >> thank you. now we have to wait for the slide show to come up. i hope you're all familiar with driverless cars and the issues behind them what i want to do is get in-depth into how self-driving cars are going to impact our economy. and in particular urban areas. already four different companies have received licenses to

operate self-driving cars on an experimental bases any california, nevada, and a number of other companies have said that they're working on self-driving cars and have demonstrated them in various forms. google in particular has published a lot of its demonstrations of videos of self-driving cars dealing with things like getting around traffic detours, dealing with bicycles, dealing with obstctions in the road and so on and so forth. i think the important thing to understand if you aren't familiar with the technology is that self-driving cars have all the computing on board. they're not connected to any central computer telling them what to do. it's all on board which means that what is happening with the car is dealing solely with what the car sees and knows about the area. now the implications of

self-driving cars are, first of all, that we may see a major reduction inongestion because self-driving cars will have much faster reflexes than humans and most congestion is due to slow human reflexes. we're going see an expansion of mobility. right now only about two out of three americans have a driver's license. this will enabled non-licensed people to travel just as much as licensed people. i'm looking forward to the day when i'll be able to put my dogs in the car and send them to the vet. we're going to see soon the introduction of cars that twenty even have a human driven capability. so these cars will be specifically or especially for people who don't have driver's licenses so, for example, this gentleman is legally blind yet

he's happily driving the new google car without human-driven options. another major implications of self-drives cars is that it will change the way we look at transportation. right now about half of all americans say that their main constraint on travel is not cost but time it's not the monetary cost but time costs. and with self-drives caring that time cost largely goes away. you can surf the internet, play games with your children, train dogs on board the car while you're traveling if you have a self-driving car. that means that will change how we look at transportation and instead of trying to live a place that's near to where we work or do anything, we can have a fairly remote home and have a long commute. when we want to get groceries we send the car, we don't have to send ourselves. we're also probably going see a confluence of self-driving cars and car sharing. some people think that in the future all cars will be shared.

i'm not so sure. i think people will still want to own their own cars but people who don't will be able to use car sharing. and that will change the calculus of driving. right now most people if they own a car the cost of taking a flip a car is they're marginal or variable costs which is only about a third of the cost, the total cost of car ownership. so if you're car sharing, the cost of driving is going to be the average cost, a total of the fixed and variable cost. that means you probably won't drive as much or travel as much if you're car sharing than if you're not. that in itself may be one reason why some people won't want to car share, they'll want to own their own cars so they can reduce the marginal cost. what are going to be the implications of self-driving cars on urban transit? right now we have urban transit in almost every major city in the country.

in fact, lots of little teeny cities and minor cities have urban transit yet outside of new york urban transit plays a fairly minor role in transportation. in the new york urban area about 32% of all commuting is done by transit. . about 10% or 11% of all travel is done by transit. the next highest in san francisco at about 18%. these numbers are commuting, but when you talk about all travel the numbers are about one-third of commuting share which means in most urban areas transit carries about 1% or 2% or less of all travel. it's pretty insignificant. urban transit was mostly private before 1970. since 1970 it's been

nationalized or municipa palized and we've poured almost a trillion dollars of subsidies into urban transit and we've seen per capita transit trips for urbanites fall from 50 trips a year to 40 trips a year. so it's not been a high success. right now at one time urban transit was mainly for people who didn't have cars but right now only about 4.5% of workers in america live in households without cars and most of them don't take transit to work. only about 41% of them take transit to work. so transit isn't even important for people who don't have cars much less people do. if you think low income people are the main users of transit, well, it turns out you're most likely to use transit to get to work if you have earned $75,000 a year. re likely than if you earned less than $25,000 a year. so when you subsidize transit to some degree, you're subsidizing the rich rather than the poor. if you think transit is a good way of saving energy, it turns out transit saves hardly any

energy at all over driving. if you want to save energy, you encourage people to buy more fuel efficient cars such as a prius. if you think transit saves money, it turns out transit costs more than three times as much per passenger mile as driving so that's when you count subsidies of course. we subsidize transit to a far greater degree than we'd subsidize driving in order to make transit appear cost competitive with driving. so what happens when we take this heavily subsidized and largely failed transit industry and add driverless cars to the mix? well we look at nhattan where there's two million jobs in about seven square miles and three-fourths of them take transit. it's hard to imagine we could substitute self-driving cars for

transit. rail transit is always going to be important for manhattan as long as there's two million jobs in lower and midtown manhattan. but that's the densest job market in america. the second densest is the chicago loop where there's about 500,000 jobs, about half of them take transit to work. again, we probably can't see self-driving cars taking all those people to work but it might help for some. downtown washington has about 380,000 jobs, about half take transit to work boston has about 240,000 jobs, about half take transit to work. philadelphia 240,000 jobs, about half take transit to work. >> in these cases, i don't see self-driving cars being an ultimate replacement for transit however that's it. that's pretty much the line, those five or six cities are where transit really makes a big difference and where frying to get rid of transit and replace it with self-driving cars is going to cause too much congestion. for most of those cities outside of new york, bus transit

probably makes more sense than rail transit but that's another issue. then we go to atlanta. 173,000 jobs in downtown atlanta but only 14% take transit to work. so if we substitute self-driving car which is will be -- have faster reflexes and less congestion well i don't think you'll see any increase in congestion, you'll probably see a reduction in congestion. houston 170,000 jobs, only 13% take transit to work. denver, 10,000 jobs, 20% take transit to work but that's not going to be big enough of a market to support transit in the future. so basically outside of five or six cities i don't see transit as being a viable alternative to self-driving cars and car sharing in the future. i see car sharing and

self-driving cars as almost coletely replacing transit everywhere except for those few places. so we have to think about how are we going to wind down transit? how are we going to change transit in the future to adopt to the self-driving cars? we also have to think about long-range transportation planning. congress has for years mandated that urban areas have metropolitan planning organizations that engage in 20-year regional transportation plans and that rewrite these plans every five years. a few years ago i went through plans for the 70 largest urban areas in the country and i found that about half of them base their plans on what i call the fantasy model which is that we'll a imagine a world without cars and design for that world and hope for that people follow our imagination. for example, sacramento wrote this plan in 2006 and they specifically said they've engage 234d this fantasy that they can

live without cars for the last 25 years, which means the last five iterations of their regional transportation plan and for some reason it didn't work out. people are driving more and more even though they're not building any roads and there's more and more congestion, people are still driving even though they're spending lots of money on light rail, they're not luring people out of their cars? so what's the solution? their solution was to continue the policies of the previous plan. and we see this over and over again in cities across the country engaging in the fantasy rather than the reality building light rail despite fact that

construction costs are growing to be extremely high, my former hometown of portland in particular seems to be in a race with seattle for who can build the most expensive light rail in the universe. seattle is winning. but portland is coming back with a plan that will have a $2 billion tunnel. and i don't see any of these plans being viable in the future when we have self-driving cars. i don't see why we're going to need to have light rail or anything like that. this is not a surprise. what are self-driving cars going to do to the amount of travel? are people going to drive less, because they're car sharing, and they'll have a higher marginal cost of travel? or are they going to drive more, because more people are actually going to have access to self-driving cars? are they going to drive more, because their travel budget is different?

cost is not the issue, time is the issue, and now they can travel and be productive while they're traveling. nobody knows the answers to these questions. and urban planners, frankly, are ignoring them. not a single reg transportation plan that i've ever seen has even mentioned the possibility of self-driving cars in the future. most of them -- none of them are trying to model it. a few of them have asked these questions. and thrown up their hands and said, there are no answers to these questions. so about 60 american cities instead are planning for 19th century technology like speak cars and light rail rather than planning for 21st century technology. what should they be doing instead? i suggest they should focus on dumb infrastructure.

what do i mean by dumb infrastructure? well, the mini tell system is an example of smart infrastructure. they gave everybody in france a dumb terminal to access a smart system that would allow them to do things like make plane reservations, make restaurant reservations, buy theater tickets and things like that. but, the company that was managing it had to keep up the technology. they had to keep their smart infrastructure up, and they couldn't afford to do it, they couldn't keep up with the internet.

so instead, they abandoned it in about, what was it, 2003. and today france, just like everybody else, reliesn the internet which has the smart in your terminal. and the internet itself is a dumb communications infrastructure that doesn't contain the intelligence needed for you to do what you want to do on the internet. in the same way, highways can be smart or dumb. a dumb highway, basically, is pavement. and then the smarts come in your car that knows how to deal with that pavement. a smart highway has all kinds of communication systems in it that tells your car things like -- electronically tells you that there's an accident up ahead or red light up ahead or congestion or whatever. the problem is, maintaining that smart infrastructure is going to be vy expensive, and it's not going to work very well. so it's much better to have dumb infrastructure and let the smarts be in the vehicle.

another example of dumb infrastructure is rail transit -- or excuse me, smart infrastructure, is rail transit. it only goes to places we build the rail line. the trains will go there reliably as long as we maintain them. so we have a $60 billion maintenance backlog. trying to provide smart infrasucture, just provide basic dumb infrastructure, which means keep the streets paved, keep the pavement smooth, keep the stripes on the -- dividing the lines plain sight. and try to use a consistent form of signage across the country so that your smart car that works in california also will work in new york and virginia. in short, what we should do is try to solve today's problems today. don't try to foresee the distant future. instead, just try to leave the future with as many options as possible. so that they can solve their problems without being encumbered by a huge debt that we put out today, in order to build something that turns out not to bworthwhile at all. build and maintain dumb infrastructure, and -- i don't know why it didn't show my last point, which is, don't manage vehicle-to-infrastructure communications which i think our next speaker will talk about a little bit more. thank you.

>> thank you, randal. up next we have marc scribner, a research fellow at the competitive enterprise institute where he works on transportation, land use and telecommunicatns policy issues. he has written for "usa today," the "washington post," and the national review, and his work has been cited by the "wall street journal," "washington post," "boston globe," politico, he's widely cited. the bbc, c-span and more. he received his undergraduate degree in economics and philosophy from george washington university. marc? >> well, thank you, matthew. and thank you all for being here. i'm going to talk about some of the regulatory iues that we have coming up. and i'm not going to -- let's just go for the overview.

i'm going to begin by talking about the recent regulatory developments at the federal and state levels. i'll follow up with some discussion of the national highway traffic administration. and sort of our traditional safety philosophy at the federal level, spoiler alert i'm not a big fan. then i'm going to give some examples of how we're already potentially screwing up the regulation of automated vehicles. and i'll close with some principles for sound public policy. so recent automation policy development. these states in green here are states that enacted legislation that specifically recognizes the legality of automated vehicles. these states in yellow are considering similar legislation.

so in imme meanting the statutes, we have a few examples so far. nevada was first out of the gate in 2012. california has released the first part of its rules earlier this year that came into effect last month. these governed manufactured testing. the district of columbia here has proposed rules in april. they haven't gone anywhere yet. i think you'll see why. there are some problems with them. at the federal level, we haven't seen any specific regulations yet. nhtsa did issue a policy back in may 2013. among other things, what they did there was basically caution states about overregulating, erlegislating at this early stage. and they also laid out the definitions of automation. i'll show you those in a second.

while these aren't automation specific, there are two recent developments that have come out of nhtsa that will likely impact automated vehicles in the future. the first example was this tesla motor vehicle safety standard. in august of this year, nhtsa proposed rule making on vehicle-to-vehicle communications, what randal mentioned when he was closing. and i'll get to those later. here are the automation levels as defined by nhtsa. zero, simple, no automation. that should be pretty obvious. level four is full self-driving automation. this is where you can start talking about having murphy be and wet bars in your vehicles, giving new meani to the term fully loaded. but that is where you have the driver has no responsibility, and possibly no ability to retake manual control at any point. these levels in between, i'm not going to focus so much on them. i don't find them as interesting. although we're seeing level one. level two vehicles available to consumers. level three are coming, depending on how the regulatory and liability issues are sorted out. but i want to begin by talking about nhtsa. i think this letter evans quotes gives idea on where i stand on this.

i'm not saying anything about the people who work at nhtsa, i just think their philosophy has gotten highway safety priorities backwards. so what nhtsa has done for the last two decades, in fact, what its purpose was, was to focus on the effective and safe automobiles. e problem with this approach is that most crashes have always been and continue to be caused by driver behavior. so what we end up -- what happens is, here's the policy failure, so we focus on rare product defects blown way out of proption. we work to mandate low value, high cost safety technology, particularly technology that deals with post-crash, air bags and the like, and then we downplay everything else. i think the latest example is gm's effective ignitioch.n swit i just wanto say right now, i'm not defending gm or th crappy ignition switch, but it's garnered a lot of media coverage and gone to several congressional hearings. to put this in perspective, this is likely responsible for a few deaths per year since 2005. compare that to the 30,000 annual deaths annually that can

be attributed to driver behavior in some part. in fact, since 2005, there's about 130,000 gm deaths that are largely a tributed to behavior that we're essentially ignoring. if we were to -- we might get closer to my position, repeal the vehicle safety act. i don't think that's going to happen. what can we do to best promote automated vehicle innovation and update by consumers. the motivation here is pretty simple. if these vehicles are in fact safer, and i think we all think they will be, that any policy that results in unnecessary costs or delay will result in additional property damage, injury and death. so death by regulation is really something to be concerned about. and i know for a fact that there arpeople within the federal policy making world, are cognizant of this fact. so what i think nhtsa should do is focus on the federal motor

vehicle safety standards that may innovate techno lodge innovation in our automated world. an example was this petition from tesla motors to nhtsa regarding fmbss 11, and they were seeking to comply with the mirror rules with cameras, rather than using mirrors. even if tesla were to replace all mirrors' viewing functions with cameras, they would still be required to install mirrors. i think this is an early example. they just want to have the option to comply with mirrors. i think we'll run into a lot more of these going forward as automated vehicles get more and more advanced. so if congress is to do anything, and i don't think they should do anything at this point, perhaps not even in the future, they should have nhtsa report on these potential conflicts. i'm not sure what we'll get out of this. at least it's a starting point

and it's something that congress can sort of use as a jump-off point. w, like i said, there's already some examples. and this quote is from d.c. council member shay, her criticism of my bill that ultimately had passed. i'll get to why this is flawed in a second. she was under the belief that it was the technology that was currently then, that may have been true briefly. it's certainly not true now. i think you'll see why in a second. so i'm going to use washington, d.c., we're all here, as the

case study. 2012, this bill is introduced. the original bill from council member shane, all vehicles be powered by alternative fuels. impose a new mileage based tax on all autonomous vehicles. if you can see, there's a logic here. council member shay got a ride in one of google's toyota priuses and realized, if all of these automated vehicles are going to be powered by alternative fuels, they're not going to be paying the fuel tax, so we need to tax them somehow. there's a logic there, i just

don't think it's a good logic. and mandating a licensed driver be in the seat with autonomous operation. that driver seat rule remained. if you look at the april 2014 proposed rules, from these, you'll notice that they seem to require that an operator have a special d.c. driver's license endorsement. so the implication there is that the district is calling for -- or calling that all test drivers be d.c. residents and have a driver's license issued by the district of columbia. given that we live in a metropolitan area where most people live outside of the

district of columbia, even from a regional perspective, that seems to be restricting your potential test driver pool unnecessarily. so i think that's a sort of a ridiculo requirement. and hopefullthe final rules won't reflect that. but, you know, beating up on d.c. is easy. california and michigan also have these driver seat requirements. so california, i'll focus on them now. so, california, they're another place they rolled out their manufactured testing res earlier this year. they imposed this drive seat rule. as randal showed you that video of the latest prototypes, google developed this pod car, low-speed vehicle. and they want to take out the steering wheel and pedals and

basically be fully automated. the operator will have no ability to retake, or take manual control at any point. but state testing, and federal low-speed vehicle rules require the installation of the steering wheel, all this stuff they don't want. what we have here is a regulator promulgating a rule forcing the innovator to take a step back. we're already having negative impacts from regulation of autonomous vehicles. this is really unfortunate. i don't think it was intentional. it just goes to show, you know which roads good intentions can pave. so the vehicle-to-vehicle communications mandate that randall mentioned, nhtsa did

issue an advanced notice of rule making in august. what they wanted to do is they wanted to develop a final rule by the end of the decade that will require all new highway vehicles be enabled with dedicated short-range communications, vehicle-to-vehicle technology. and this mandate wouldn't be requiring that all cars have these little yellow circles around them. what the d.o.t. image is trying to illustrate is the cars will be talking to each other. they wouldn't be completely reliant on onboard sensors. for instance, in finding hazards. and keep in mind, this is not for automation. they're talking solely about warning drivers of hazards. imagine there's a car, a few car lengths ahead of you, slams on its brakes. this would send some data back to your car and tell you, hey, some guy slammed on his brakes ahead of you. take whatever action the car is going to tell you to take. it remains to be seen how effective that will be, or how drivers would actually respond to that in real world settings. so there's some problems with the v-to-v mandate. the proponents argue it's low cost. they're imagining that all nhat communicate with each other. others say it will cost a few hundred dollars. what's the big deal. we're looking at $25,000. so this isn't a big increase in the sticker price. the problem is, that the benefits are likely to be very low in the short run. we're going to have -- these things are -- at earliest, we're expecting a mandate by 2020. it's going to take at least ten years before you have significant market penetration for the auto fleet to turn over, that this is going to do any good for anyone. i mean, i think you need about 70% of vehicles enabled for you encountering one on the road to be greater than shant.

we expect it to take at least ten years, if we proceed as nhtsa appears to want to do right now. dsrc has had this 75 megahertz has been blocked off since 1999. the proponents obviously want to keep it blocked off for transportation purposes only. but there are forces out there, namely those who make consumer electronic -- or portable electronic devices that are wi-fi enabled that would love to have access to the spectrum. there's a battle going on at the ftc that hasn't been resolved yet over the spectrum. i think that reason alone, i think it makes premature for nhtsa to dive into this, but they thought otherwise. they ignore competing technologies. nokia earlier this year announced it has developed an advanced form ofte, that would be able to perform these same vehicle-to-vehicle communications' functions. but you would be able to use technology -- you would be able to use things like cell phones much more easily. and we already have an lte network infrastructure out there. when we get into sort of more precise vehicle-to-infrastructure kind of applications, that would require, if we're dealing with dsrc installing lot of roadside boxes, and we just frankly don't have the money for that. nhtsa has ignored some of these technologies. there is a chance it's already obsolete, too. if we're talking 2030, 2035, that, you know, what good is --

at least the way nhtsa is imagining this, and not that the context of automation is not bad. if you have a fully -- a self-driving vehicle, we have no responsibility to take control of that vehicle at any point. perhaps not even the ability, what good is a warning light, or audible alert going to do for you. the answer is not much. there's some unanswered cybersecurity and liability questions. this proceeding just opened an industry that's very interested in these. they're very concerned. as far as an automation specific cybersecurity potential problem, well, if you have automation based on sensors and onboard computers, how would a v-to-vk -- mandated v-to-v system, it's onboard the car, how will the

two systems interact, if at all. so that's why i think the best case for fully automated vehicles under this v-to-v mandate as currently conceived is automakers will be forced to install completely useless technology in autonomous vehicles. that might not be bad, but that would certainly increase the price. and that gets back to unnecessarily delaying the rollout to the consumers. i think the takeaway for v-to-v and dsrc, how nhtsa currently envisions it, we should be skeptical. v-to-v is so valuable automakers would not consider installing it voluntarily. there may be legitimate institutional problems with industry collaborating on standards.

but then perhaps those should be addressed directly, rather than forcing a mandate down everyone's throat. so i'm going to end with some sound automated vehicle public policy principles. that i think are sound principles. i think we should start with recognizing and promoting the huge potential benefits. i talked a lot about the safety. randal did mention these traditionally, mobility disadvantaged populations, the disabled, the elderly and the youth who don't have access to the mobility many of us take for granted. we should reject the precautionary principle. this is a new technology. we don't have data. it doesn't mean we should shift the burden to the innovators and say, you need to prove that this is safe. before we get these to consumers. i don't think any auto manufacturer is going to release to the consumer market until there's some demonstrated safety level. but i think going down a precautionary principle route is a good way to keep these out of the hands of normal people. we shouldn't presume to know how the technology and law will evolve. the technology right now, we're

talking about proprietary things. the only thing we know about this is generally the kind of the press releases we get from industry that's currently developing this. and the law -- there's no court cases yet. we don't know how this is going to -- do we need legislative intervention to update our liability laws, things like that? we don't know. there's certainly a ssibility that common law liability can evolve without any sort of intervention. and then, i think, you know, number four logically follows, we should always seek to minimize legislative and regulatory intervention. regulators are slow, however well intentioned. this is moving pretty quickly. we should let the innovators innovate, and keep the regulators as far away from these things as possible. doing maybe some very minor things around the edges, but

they shouldn't be involved in the sort of developing the technology. or deploying the technology. and then finally, once we get to the state where we're talking about, you know, updating motor vehicle codes to reflect this new automated world, and really doing some serious legislating, if it comes to that, we should focus on developing clear and simple rules that have neutrality i see a big risk of the first mover. so the first company that ends up with a consumer-ready vehicle coming to market, and then regulators saying, well, we're done, this is the technology we're going to mandate, i think it's a terrible idea to mandate technology in generation one, despite the fact that regulators may believe that this will enable to get generation one tech out to consumers more rapidly. so with that, i am finished. and i look forward to any questions. thank you very much. >> thank you, marc. our next speaker is adam thierer.

he's a senior research fellow with the policy program at george mason university. he specializes in technology, media, internet and free speech policies. his writings have appeared in the "wall street journal," the economist, "washington post." adam has authored or edited eight books ranging from media regulation and child safety issues and high technology market. his latest book is the continuing case for comprehensive technological freedom. adam was president of the progress and freedom foundation, director of telecommunications here at cato. he got his masters in business management at the university of maryland. and his bachelors in journalism at indiana university. adam?

>> thanks, matthew. and great to be back here at cato. it's always a pleasure. it's also a pleasure to follow randal and marc. i learned a lot from what they've both written on this, and many other issues. and i recommend it to all of you. my remarks will be focused primarily on the privacy and security technologies in driverless cars. you can download my paper from the cato website. in thinking about privacy and security in smart cars, or driverless cars, we need to begin first by understanding, acknowledging that security and privacy are relative concepts, with very amore fosed boundaries. not everyone affixes the same values on this. some people are hyper cautious, and hypersensitive about the privacy. others are risk takers that are just somewhat indifferent or

more pragmatic about their privacy. we all say we love our privacy, but sometimes we do things that actually in the real world act differently. this is called the so-called privacy paradox. we should understand the security and privacy norms tend to evolve over time. and do so very rapidly. with any new highly disruptive technology, such as intelligent vehicle technology, we often panic at first. especially about the privacy situations of the new technologies such as these. but we establish new ethical and legal base lines about new technologies fairly rapidly. i've written about this in recent law review articles and in my new book. there's essentially as i describe it in that work, a cycle of work of initial resistance ta technology, gradual adaptationthen, and eventual assimilion of that technology into our lives. not without some heartburn along the way, of course. this was just as true for the first cars that came along over a century ago and it will be true for the new smart car technologies that are evolving today. but it is especially true that these norms evolved as it

pertains to privacy and security, that they're subjective and relevant. third point, for almost every perceived privacy or security concern or harm, there is a corresponding consumer benefit that sometimes balances those out, or even outweighs those perceived harms or fears. we see this reality at work with the broader internet. and with digital technologies. and we'll see it at work with shall driverless vehicles. compare today's telemattics and intelligent technologies one board our cars with the tracking technologies onboard every smartphone that all of you are carrying with you right now. the reality is, these technologies, the ones we carry in our pockets, the ones we'll have in our cars and already have in our cars today, are capable of tracking us. that sounds sinister. tracking, of course, creates enormous benefits for us as well. we now know in realtime what traffic looks like when we're in our cars, not just because the technology in our cars, because, of course, we can see it on a map.

in realtime that happens because we're all connected and we're all being tracked. so again, we have to be clear that not every theoretical boogieman is just a big bad awful thing to be disposed of. it has a corresponding benefit we have to take into account before we address it. fourth point. as it pertains to intelligent vehicle technologies, today's privacy and security concerns are not the same as yesterday's and they're not going to be the same as tomorrow. today's intelligent vehicle technologies are likely to be more pressing i would argue than tomorrow's. that's things like event data records and telemattics are recording our realtime actions today as we are doing them with our hands on the wheel. this raises a variety of

interesting questions, which is, can that information be used for automobile insurance purposes. and it is today already on a voluntary purposes. or breaking the law by driving erratically or too fast. they raise obvious privacy considerations. that may even lead to considerations or concerns about discrimination associated with the data that's collected by our vehicle. but it's worth asking this question. what happens as we make this transition to fully autonomous vehicles. and what happens when our cars are less of a final good that we own and more like a service that we just use or rent on occasion. what happens when we combine the power of today's sharing economy with the power of self-driving technology. a car of the future may look like a robotic chauffeur, as randal described. they're just waiting out for us, we dial them up on our cell phones as we need them.

needless to say, in that environment the privacy considerations are very different than they are today. clearly we can still be tracked, our activities can be tracked, but not us personally manually operating that vehicle, and so it's a different type of consideration. at's something to take into account. fifth point. this is the most important point i'll make here today, that any security or privacy solution should take into account the source of considerations that i've just outlined, and must be able to accommodate the many different types of views and values that people have as it

pertains to privacy and security. i write about this in my recent book, there are no silver bullet solutions to concerns about safety, security and privacy. it will be very difficult for law to keep pace with not just rapid innovation in this space, but the rapid evolution of consumers' taste and values. this is why we need a flexible layered approach to addressing security and privacy concerns. we need to borrow a phrase from richard epstein of chicago, simple rules for a complex world. we need to rely on things like contracts, and enforce existing terms of service. we need to think about how common law and torts might evolve, or specifically products liability. for that, and i want to highly recommend a wonderful new paper from the brookings institution, where he talks about the law historically and say it's not expected to stop there. when confronted with new often complex questions involving produc liability, courts have generally gotten things right. products liability has been highly adaptive to new technologies in recent decades and it will be emerging as the need arises for autonomous hicles. i agree with that. it will be interesting to see how the liability and insurance standards evolve. again, as we move from our vehicles being final goods to services. because this will change what economists call the least common -- i'm sorry, the least cost avoider in these situations. right now, you as the driver of your vehicle, are the person who can take action to avoid potential types of harms, whether they be safety related

or otherwise. maybe privacy and security related. but what happens when you no longer own your vehicle. what happens when your vehicle anthe people who created it know more information about the vehicle than you do. the knowledge and therefore the responsibility and liability moves from you to them. they already know this. they're already thinking about this. this is the least cost avoider principle in economics. we shouldn't be surprised to see legal norms come to reflect that over time if we allow common law to evolve spontaneously. now, in the process, there are steps that manufacturers of intelligent vehicle technology could alleviate a lot of headaches for themselves. especially as they come to deal with more liability eventually. and this leads to a lot of proposals for so-called privacy by design, or security by design. the idea of baking in best practices regarding data security, or data collecti, or data availability, in systems as you manufacture them. this could include things like

data collection, minimization, or limitations, basically not allowing third parties to access certain types of data that are being collected about you and how you operate your vehicle today, or about how it's used in the future. it could include best practices regarding better transparency about how your data is being used, if shared with other parties. potentially, clear use practices about how you use these things apprriately. with better information or empowerment or education of the users. and clear consent for any types of new uses or iterations of technologies as they're evolving and more innovation is occurring. you might want to take a look at what the privacy forum is coming up with in this regard. this is an ongoing process. there is no end point in this process. these standards and these best practices evolve not just every couple of years, but every couple of months. because there's always some sort of new technology collecting more data, that raises still yet another privacy or security

concern. here's a query here we might have. we might say, this is all fine and well, but shouldn't there be a minimal legal standard? shouldn't there be a state or federal regulation regarding driver privacy or security? after all, these values are pretty important to a lot of people. but i think for all the reasons i've already sted, we now know especially in light of what randal and marc have already presented, that these technologies are a fast-moving target. i mean, this is very, very hard for me to imagine that any law we put in place today will be applicable to whatever is down even 18 months down the road. you need to have that flexible ongoing evolutionary approach of voluntary best practices, along with the evolution of common law, to probably better address the evolving cases and controversies that will develop in this field. as marc and randal already noted, we're probably better served by having a wait-and-see strategy, precautionary principle based approach, and instead we should continue to

have our default be innovation. this is the point i do the plug for the book. continuing case for -- i argue the benefit of the doubt should be given to those innovating, and innovation more generally. ongoing experimentation and trial and errowill permit us to find new and better ways of doing things. new and better ways of being safer, more secure, and even potentially more privacy. we can't be living in fear of short-term worst case scenario, or else long-term best case scenarios will never come about. so one final point i want to make about privacy and security. especially privacy. special consideration needs to be paid to the role that government plays in this regard. government actions can affect user privacy obviously in a very profound way. ma of the privacy and security concerns regarding private sector data collection with regards to intelligent vehicle technology are problematic, and

obviously deserve some discussion, governmental data collection and use is an entirely different level of concern and set of issues. they cannot tax you or imprison you, because they lack the powers the government has, although it is possible to ignore or refuse to be a part of certain types of technologies, including driverless technologies, you don't have to use them, the same is not true of governments who can obviously be abated. special protections are needed for law enforcement agencies and officials as it pertains to the technology. when government seeks access to data collected from these technologies, strong constitutional and statutory protections should apply. we need stronger fourth amendment constraints.

and the courts are going to need to consider revis tinge the third-party doctrine which holds an individual sacrifices his or her fourth amendment interest when they divulge it to a third party or share it for whatever reason even if that party promised to safeguard that data. what we want is a world where many different intelligent vehicle companies are competing on security and privacy, and maybe some are offering those who are more privacy sensitive better protection than industry norms or best practices. industry is already moving in this direction. we had a recent jao report that surveyed ten makers of these technologies and found they are all taking steps to protect these things to some degrees. but more could be done, and more is being done especially with the liability question looming large. when the government comes in and says we need back doors into all these systems, we're having this debate right now with our smartphones, right?

apple and google start using better encryption on our smartphones. the government is saying, no, you can't do that. a lot of people are clamoring for better security on their smartphone. that's the exact same debate that will unfold with the smart car and driverless vehicles. we need to allow innovation to go forward. thank you very much. >> thank you, adam. and thanks to all of our speakers. we are now going to enter the q&a part of the event. a few notes before we begin. it for the microphone. we have a few interns here to help out. and before your question, please announce your name and your affiliation. and can you please make your question, a question. this is the question-and-answer part of this statement. if you would like it directed at one of the speakers, please make it clear who you would like to ask. i'll begin with the gentleman at the back. is the microphone on? >> thank you. mark carr.

i have my own company which works in freight. i'm actually here thinking about a role that i play, transportation research. and for the last speaker, you talked about privacy and all. i'm involved in freight. and i'm wondering where in freight, and whether it comes to our homes, or whether that's commodity freight going back and forth between ports, where do trade secrets fall in with the privacy? because you can envision shippers who don't want other folks to know what their trade patterns are or their customers and such are. >> that's a wonderful question. i have not spent a lot of time researching it. i think i'll have to now.

i'm always careful to wade into the waters of intellectual property, patent and trade secrets, it's a complicated field. i would hope through ongoing experimentation with different business models, that companies find ways to protect thoserade secrets. although it may be difficult. having fleets of freight vehicles that are robotically opered raises a whole host of security questions about who has access to that data and information. that's part of the spectrum upon which companies are going to be competing, to offer eater security to make sure those trade secrets don't get out. it deserves more consideration than i'm giving it here. >> this gentleman at the front, please? >> i'm tom curry, a reporter with congressional quarterly role call. i talked to the head of the contra costa, california, transportation authority last week. they have the new test track for the mercedes autonomous vehicle in their county. he said one of the things in the future benefits that he sees is

that these vehicles could benefit to the county, is that these vehicles could go out and detect potholes, and immediately transmit the data back to the contra costa transit, or any trngs authority. the fleet of vehicles would be communicating information in realtime, so that they could maintain the transportation system better. i just wondered for mr. scribner and mr. o'toole, what's your view of that, and the cost also of transmitting all this data back to a county transportation authority? >> i don't see anything wrong, if he wants to get that information, for him to design an app, and upload it to apple and android, and people can volunteer to download it. and if they're driving on a road that's unusually bumpy, the app will detect the vibration and transmit the information to the road authority that there's a

serious problem with the road. there's no reason why it has to be mandated. or why it has to be applicable only to driverless vehicles. >> yeah, i agree with that. but sort of going from there, i think e -- it is really valuable, particularly in the context of automated vehicles to have our roads in very good condition, particularly if they're solely sensor based. these aren't things that can -- vehicles presumably, at least initially, won't be communicating a lot of data between each other, and certainly not with infrastructure. so that would be great. i don't think a mandate would be needed. but certainly, yeah, that's -- i think that's a very good use of

that technology. >> the gentleman who had his hand up earlier. >> thank you. lawrence gasman, small tech publishing. i'm interested in the time frames of this. randal o'toole put up a slide that showed the different types of intelgence the car could have. i think that came from some government agency. but i'm particularly interested in the advanced ones, because you have cars that can park themselves now. it's fascinating, but it doesn't seem like a huge regulatory issue. but you put your dogs in the car and they take themselves to the vet, that's a completely different deal. it will be quite a few years before you start seeing that in any, you know, outside california, i guess, is what i'm thinking. what sort of time frame do you think those will be? and i guess the question is, how long do we have to get this right? >> well, it depends on who you ask. certain automakers are saying that -- toyota is saying, we're not going to focus on this really at all. but then you have companies -- google is still saying 2017, presumably level three, level

four. that's still a goal they're sticking to. nissan is saying 2020. presumably level three, possibly level four. continental is also saying 2025. so it could be sooner rather than later. and, you know, it really does -- you know, we need -- we don't necessarily need special regulations for this, but it just goes to show that regulators are going to be well behind the technology developing. assuming these optimistic industry forecasts are true. >> about 200 experts in the field were surveyed on this question in a conference last summer, and the median answer was that google driverless cars, a car that has a steering wheel, but the car can take over pretty much completely in all conditions, will be available by 2020. the pod car that doesn't have a steering wheel, or any way that the driver can control it, other than start and stop, will be

available by 2030. i would go further and say b 2040, we'll start talking about closing roads to human-driven cars, because they're too dangerous. >> quickly, i think there's an important distinction to be made between highway and non-highway vehicles. certain things that are happening, namely city mobile 2 in the eu where they're focusing on the low-speed geographically restricted almost paratransit kind of vehicles. those you could perhaps see earlier. they would be deployed in college campuses, retirement communities, places like that, where you don't need to meet these stringent, say nhtsa guidelines. so there's the potential for that, too. happening before we see the actual highway vehicles. >> this gentleman in the front. >> thank you. i was wondering about insurance implications, especially when you have automated and manual vehicles. has anybody looked into that? are there any insights?

>> the interesting thing is insurance companies are some of the biggest backers of thing it insurance companies are the biggest backers of self-driving cars because they figure it will reduce problems and accidents. one concept i have heard is that instead of when you buy a driverless car like an experience policy for the car, the auto company will buy the insurance policy and that will be included in the price of your car. if there is any liability involved, then the auto company will have their own insurance to deal with. they don't have to deal with extended lawsuits.

>> a quick question. the issue that was the russian monitoring stations that they wanted to build for gps, that was luckily quashed last december. they have the companies that owned others that is putin's car company. theoretically and start thinking about this technology and security. we do not now have that monitoring station being the russians have the access. it's with foreign companies. any concerns or comments?

>> they lef in complete fear and find solutions tohe pblems and figure out how to make the systems as secure as possible so that that is packing for whatever reason that is avoided in whatever possible. i will say this. relative to the base line we are operating o a world where over 33,000 people are dying behind the wheel because of human error, almost 100 people a day. will there be hacking that results in someone resulting in an accident or even death? potentially yes. i'm not here to say no. in the world of technology, it may be that we have tens of thousands of people's lives being saved. we are going to have to try to make that balanced. i know it sounds crude or overly utilitari utilitarian, but they have to take it at the corporal base line. >> one of the reasons i opposed a mandatory v 2 vehicle to

vehicle communications is that a mandatory system will be a lot easier to hack. it's going to be a lot harder to defend against because the government that mandated the system will not be as motivated to defend where as if you have a competitive system out there with different companies each offering their own software and others offering apps and so on and so forth, they will all have a competitive reason to keep their systems from being hack and it will be harder for a hacker to attack and have a widespread attack. you can attack one app or software system, but you won't be able to attack everybody's at the same time.

that's called iro bot. a lot of them may troelt things you have an interest in that were promoted in this movie. i was wondering, are we or are the americans most likely to debut this type of driverless car or is it more likely that south koreans would be likely to debut. they seem to be ahead of us when it comes to mobile phone technology the way they are using that. that we are not necessarily using it just yet. are they ahead of us on this? >> they are uniquely positioned to be one of the first places to have these available in widespread testing which we are seeing early testing on public roads. one reason is we are not party to the vienna road convention

that may restrict the testing and operation they are working on updating the convention to allow for more availability in the future. it has proven to be a problem. i know a bunch of them are being led by a number of german parts and auto manufacturers right now. they are very well aware of this. as it stands right now, the u.s. is in a good position. >> there will be extra points for you if you can put a minority report into that. i want the gentlemen in the middle, red shirt -- >> it's a great movie, so there is that.

>> two of my colleagues right motorcycles and wonder how that would fit into our future. >> i thought about including a motorcycle video in our show and they have a driverless motorcycle. it's a little and the first entry to when they were running the challenges, it was an autonomous motorcycle and he messes something up at the last minute and it falls over. that was sort of -- you can look at it as the genesis of the self-driving motorcycle.

>> gentlemen on the right? >> i have two questions. this is an appropriate day to have this panel. this is the birthday of elwood haynes, one of the early pioneers of the automobile whose first gas-powered driving automobile is in the smithsonian. elwood haynes also was the first to get involved in an automobile accident. i would ask the panel to talk a little bit more about how they would see the common law personal injury evolving with that driving cars. my second question is, if in fact people could have time to read and do work and watch movies are going to be interested in commuting longer, why hasn't that happened?

people already have that constitution. >> and why we haven't seen an increase in commuting time already. >> briefly on the 50 one, i mentioned products liability and go alongside that. they have situations where someone in this short of vehicle that is fully autonomous and may be ensured by the service provider is in an accident with human driving like me, a car he renovated in the garage and is still that dumb bag of flesh and bones that plows into it. i have insurance myself. i we will work these things out. it will take time and we have a litigious society. there is a good question as to whether our overly litigious nature will ultimately lead to a lot of innovators being sued.

i don't know what we should do outside of changing the loser pay law. good luck with that. i am confident that the personal injury law will work out overtime and there will be fewer injuries. >> unlike an admission problem or a sticky accelerator pedal, the driverless cars record everything that is happening around them all the time. if they are involved in an accident that they will have excellent 360 degree recordings of what happened prior to the accident and it will be fairly easy to determine who was at fault in the accident. if the manufacturers of driverless car is at fault, they will pay up and update their software. that will take care of a lot of

litigation and liability problems. >> i will agree with both adam and randall. i think we should let the law evolve when they arise. one, taking adam's example a step further would be what happens if you are in a level three automated vehicle and you collide with a level four, how do you start then sort of breaking that out and a signing liability to various parties. it will be interesting. something i think the courts hopefully will be well-prepared to deal with. that's the best option we have as opposed to preemptive prescription delegation. you can do more harm doing that. i forget the second question.