Up until I began doing my reading for this fall’s Criminal Procedure: Investigation course, I largely bought the heroic Warren Court story of privacy and the Fourth Amendment.

The story is simple: The Supreme Court, concerned only with helping businesses through decisions like Lochner, had left people unprotected from warrantless searches and seizures. In decisions like Olmstead v. United States (holding that a warrantless wiretap did not violate the Fourth Amendment), the Court threw privacy under the bus. But, as with the First Amendment, Brandeis and Holmes dissented, presaging the arrival of the glorious Warren Court, which overturned Olmstead in Katz v. United States.

Though, unlike many FedSocers, I love the Warren Court and its expansion and constitutionalization of personal liberties both procedural and substantive, the heroic story just isn’t quite right.

The tale has two problems. First, it lumps together the entire pre-Warren Court years as one big Bad Old Days of privacy. Second, it ignores the theories behind the decisions in pursuit of the results. The story’s troubles are interrelated.

In fact, back in 1886 – in a case called Boyd v. United States – the Supreme Court explained a theory of the Fourth Amendment that would be extremely protective of privacy. Haven’t heard of it? Neither had I. But its underlying message about the Fourth Amendment, which was hardly novel (indeed, the opinion traced it back to 18th century English cases), sounds like something a modern-day libertarian would write about privacy.

The case involved a statute requiring the owner of goods imported without paying the tariff to produce his records regarding the goods. This, the Court held, was an unreasonable search or seizure, in violation of the Fourth Amendment (and it compelled the defendant to give testimony againt himself, in violation of the Fifth Amendment). The reason was simple: the government had seized the defendant’s papers, which were his property. This made the case unlike the government seizing stolen goods to return to their rightful owners; that is not a “seizure” that deprives anyone of property rights. (And, the Court held, it did not matter that the law in question was not technically criminal, because the penalty for refusing to produce the papers was a default judgment, by which the goods would be forfeited. Yet another way the old Court was more “liberal” than the modern Court, which blindly accepts the “civil” designation legislatures stick on their penal statutes in order to evade the Constitution.) 

As subsequent cases such as Gouled v. United States made clearer, this holding did not apply just to papers, but to all property that was seized as “mere evidence” of the defendant’s guilt (not because the defendant was not entitled to the property).

But, concerned with the mere evidence rule’s effects on criminal investigations, the Court later weakened it and riddled it with exceptions. That trend explains the ruling in Olmstead, which had to explicitly limit Gouled to its facts. Olmstead faltered not because it refused to adopt a new conception of privacy but because it refused to adhere to the old. It held that, unlike a person’s mail, “the wires beyond his house and messages while passing over them are not within the protection of the Fourth Amendment.” Had the Court been true to Boyd, it would have noted that somebody owned those wires – perhaps the homeowner or perhaps the phone company, but certainly not the government – and the government searched and seized by trespass those wires by tapping them.

Katz came to the right outcome, but it should have arrived at restored Boyd to its full potential. Imagine what the Fourth Amendment would do today if it had. The Fourth Amendment would be no mere procedural protection, no simple rule of evidence, but a strong substantive limit on government. It may forbid, for example, net neutrality regulations that seize a company’s wires in order to achieve some supposedly socially desirable end. It may constitutionalize the principle of intervening action enshrined in Sec. 230, since the government cannot search or seize the papers of someone who has herself done no wrong and thereby forfeited her right to them. Best of all, it would go a long way to establishing in the minds of judges, politicans, and voters the best legal conception of privacy – privacy through property rights.

I’m not hopeful for such a big change. (The last time the Court really changed the fundamental theory behind a constitutional provision was in 2004, with the Confrontation Clause. And that revolution did not have the sweeping change in outcomes I’d expect from adopting a property approach to privacy.) But it’s interesting and useful to read cases from an era that got so much wrong to find one that came closer to getting it right than most legal scholars today.

As if we needed another. Over at Overcoming Bias, Robin Hanson points out that mandating balance leads to worse reporting.

We’ve spent a lot of time here at TLF talking about the sound economic arguments against net neutrality regulation. We argue that net neutrality regulation will result in worse consequences than leaving the internet relatively unregulated. But to me, the more important point is that net neutrality regulation is itself unjust.

Why do I make such a strong claim? Simply put, people own their stuff. People can decide what to do with their stuff. People can enter into mutually-consensual agreements about what to do with their stuff. As long as both parties agree on the terms, both parties are deciding what to do with the property they each bring to the table. All that is just. It is unjust, on the other hand, to take someone’s property. It is similarly unjust to use force upon someone (e.g. by taking their money, which is other property they own) as a punishment for doing something just. So, it would be unjust for me to reach into your wallet and take a “fine” from you because I don’t like that you sold your copy of Anarchy State & Utopia to Berin for what I think is far too low of a price. I could argue to you that Berin is giving you a bad deal or tell Berin to stop exploiting you, but it is unjust for me to steal from you to enforce my personal desires about the terms of your agreement.

People can certainly disagree with what I’ve just said. I have defended those views, which are the central core of the libertarian conception of justice, elsewhere. But if you accept what I said in the last paragraph, then net neutrality regulation is unjust. Why? Well, the ISPs own their pipes (or own the rights to use pipes in the way they’re using them). They can do whatever they want with them. They could just use their wires for their own purposes, e.g. hanging Christmas decorations or sending communications between their employees like an internal telephone system, or leave them unused. Instead, they reach mutually-consensual agreements with others (“customers”) whereby the customers get to use the wires in the way and for the purposes the ISP specifies in exchange for money. The ISP alienates some of its rights to use its pipes and transfers them to the customer. But the only rights the ISP alienates are the ones it consents to alienate. So, if the ISP says “You can use my pipes, but only on Tuesdays and Thursdays,” or “You can use my pipes, but they’re not connected to anything,” or “You can use my pipes, but only to access the parts of the Internet I like,” all of that is just. And it would be unjust to fine the ISP for making such an agreement.

So, regardless of how great the consequences of having a neutral network or even using regulation to mandate a neutral network are, using force to punish someone for making her network non-neutral is unjust.

Now, of course, one major assumption I have been making is that the ISP and the customer have actually both consented to the terms and both parties follow them. This would not occur if, for example, the ISP promised access to the full Internet but then blocked parts of it. That is unjust. It is not providing the right the customer now has, which the ISP alienated by its consent. Such scenarios may be part of the appeal of net neutrality regulation. And, indeed, they may have already happened. But we already have systems of law – torts and contracts – for punishing those sorts of offenses. Net neutrality regulation, on the other hand, is an attempt to impose, by force, the terms of an agreement mutually consenting parties come to about what to do with their property. That is unjust.

Alternate title – Sec. 230: Not just good for the consequences any more!

Since Sec. 230 has been a hot topic around here recently, I figured this would be a good time to fire up some controversy and cross-link to an old OpenMarket post. In it, I discuss the Principle of Intervening Action, a principle postulated by Alan Gewirth that states that we are responsible soley for our own actions. I argue that this principle is correct.

I mention its application to Sec. 230 (and to safe harbor under the DMCA), but do not lay out the argument explicity. Essentially, it’s this: if a user does something unjust, it is the user that should be held responsible. The website is not the one that performed the unjust action and thus cannot justly be punished for it. But perhaps PIA points in the opposite direction, in at least some cases. If a user orders a website to do something (e.g. posting an infringing video, though we can argue about whether that’s really unjust), PIA eliminates the “following orders” defense. If you do it, you’re responsible. So, is the user the one “doing” the action or is the user directing the site to “do” it?

Discuss (or Disqus)!

One of my favorite recurring themes here on TLF is the definitional dispute/clarification. We point out where a term has been used in many different ways and explain the positives and negatives of the various behaviors described by that term. I just did this with privacy.

Of course, it is somewhat pointless to argue about the “true” meaning of a term, but that’s not exactly what’s involved here. Yes, we libertarians can lament when terms that used to describe things we believe in, like “liberal,” “freedom,” “rights,” “choice,” etc., get appropriated by others and terms that used to describe things we don’t believe in, like “coercion,” get ascribed to us. There may be some battles we can win, some terms we can hold onto, but these disputes often end up with two ships passing in the night.

But I’m talking about something a little different. Lots of terms that have, or get, normative connotations – that sound like they describe something good (think “democracy”) or bad (think “terrorism”) – get way overbroadened. Speakers use such terms to describe nearly anything (as long as it’s vaguely related to the original meaning) to which the speaker wants to ascribe the good/bad connotation. We here on TLF catalog those various ways such terms have been used – break the term down – and describe which ways are really good and really bad. As I said, I just did this with privacy. If this were a more lawy, as opposed to techy, blog I’d do it with “activism,” one of my pet peeve words. (Maybe I’ll do it anyway; after all, I posted on the best and worst Supreme Court decisions even though they weren’t especially tech-focused.)

But today, it’s “regulatory capture.” We have discussed it a bit recently, including just tonight. Tim Lee did some great posts on it back in the day. It’s definitely a recurring theme here. We seem to have something fairly specific in mind when we use the term. As Tim put it, it is when “established businesses argue in favor of regulations that they perceive as hurting their competitors (often smaller competitors) more than themselves.” Indeed, I argued with a commenter on one of Wayne’s posts that this definition that makes the most sense given the meanings of the words:

Regulatory capture is when businesses capture regulatory actions and use them as tools, backed by the force of government, for imposing burdens on their competitors. Businesses banding together to oppose government intrusion is not “capture.” Fighting an enemy is not the same as capturing him and using him to do your bidding…

Call Tim’s and my definition the “appropriation” definition.

As I pointed out, however, others (in particular, fans of regulation) seem to think of “regulatory capture” as being – or at least including – when businesses use their connections with regulators to try to get the regulators to stop regulating. Call this the “abdication” definition.

There’s another sense, too, that’s an even more direct form of capture than appropriation – where businesses try to win government contracts or other favors for themselves (as opposed to trying to get new regulations to hurt their competitors). The defense industry is infamous for this, but it’s of course all over the place. Call this the “bribery” definition, since it basically describes bribery, broadly defined.

What makes these forms of regulatory capture happen? What are their effects? Should any be celebrated or all condemned?

Of course, Iron Triangles are a big part of what makes all forms of regulatory capture possible. Iron triangles are complex webs, of course, but I’m most familiar with the relatively simple example of lawyers going back and forth between government agencies (or the DoJ) and private firms. For example, a tax firm would sensibly love to hire someone who has worked at the IRS or Treasury and knows its internal procedures and thought processes. Relations develop from these revolving doors, of course. But what incentives do the regulators have to listen to their former colleagues, now working for the businesses they regulate? Well, this depends on what the businesses want.

In the case of bribery, they want contracts or other favors and can offer regulators bribes in exchange. That’s pretty simple. But it’s also pretty illegal, and so this is hardly the worst problem in terms of scope that comes from regulatory capture.

In the case of appropriation, it’s a little more subtle. Say AMD wants to disadvantage its big competitor, Intel, but without having to, you know, make better products for less money. Instead, they can go to antitrust regulators and get them to sue the pants off Intel. It costs AMD little. But what do the regulators get? They get to make themselves look great! It’s similar to journalists taking stories from activists – they get their jobs done from them, and in a way that makes them look good. They don’t even have to think of the choice sound bites. (I think we saw this with NebuAd and the regulatory fall-out as companies like Google scrambled to sic the FTC on their competitors.) The regulators can justify their jobs and even expanding their authority as being necessary and can look tough, fighting business, not beholden to it! Of course, they are beholden to one business by fighting another. Not only is this legal, it makes the regulators seem super-diligent. So, we see a lot of this kind of regulatory capture. This is, needless to say, bad. It causes massive deadweight loss. Companies should be fighting in the marketplace, where their wars result in better products at lower prices.

In the case of abdication, however, there is little the business can offer the regulator. The company wants the regulator to do less – to go home and put herself out of a job. Good luck with that sales pitch. If the regulator goes along, she looks like she’s in the pocket of big business and can only justify her actions by reciting true arguments (ugh) about the bad effects of regulation. And those are the same arguments those lobbyists made, so she must just be their puppets (regardless of whether the arguments are right). That’s why we don’t see a lot of successful “regulatory capture” of this type. In fact, it’s a rare cause for celebration when companies even try to get regulations lifted. (There’s a coordination problem from the companies’ perspective too. It’s much less costly to get the DoJ to go after a competitor than to take a stand that benefits the industry as a whole, since there’s a big free rider problem.)

Can anyone think of one big example of this kind of regulatory capture – a situation where a law passed by Congress has been left unenforced because of successful lobbying by businesses? To say the least, this seems much less common than laws being used way beyond their purposes, or even regulatory actions taken without any statutory basis whatsoever, but I’d love to hear counter-examples.

So, maybe – since it doesn’t happen and doesn’t really mesh with the meaning of “capture” – we should not call regulatory abdication “regulatory capture” at all. And we already have a word for “bribery.” (It’s “bribery.”) So, perhaps we could save the term “regulatory capture” for the specific pattern of companies convincing regulators to use their coercion wands to whack competitors.

In my last post, I touted an often-ignored benefit of targeted ads: that they directly enhance the browsing experience, compared to seeing dumb ads. This post argues that no one has a “right to her data” that ad-targeting takes away.

“Privacy” is a word of many meanings. The best explanation of the myriad ways the word has been used I’ve seen came from TLF’s own Jim Harper five years ago. People have a right to privacy in some senses, but not in others. They have a justice-based right (a political moral right) to freedom from government intrusion – the broad sense of privacy espoused in cases like Lawrence v. Texas. And they have a justice-based right to not have to give up information about themselves to the government, the sort of right violated by NSA wiretapping and that should be protected by the Fourth Amendment – the kind of right espoused in cases like Katz v. United States.

We do not have any justice-based right, however, to control what is done with those things we voluntarily give away to private individuals. If I sell you or give you my laptop, without any restrictions on your use, I have no rights-based complaint if you use it to do something I find objectionable, such as reading Perez Hilton. Nor do I have a rights-based complaint if you take the information I left on it and use it for your purposes. Even if that information is about me.

Of course, depending on the information and how you use it, you may be a bad person for using it; it may even in a sense be immoral for you to use it in that way. But it’s not immoral in the justice-based way, in the political-moral sense, as government rights-violations are. You do not deserve to be fined (have your property, to which you have rights, taken from you) or thrown in jail (have your liberty, to which you have rights, taken from you) for such actions. Perhaps you deserve to be denounced, boycotted, ostracized, etc. But, unless you have violated my rights, you do not deserve to have your rights violated in return.

Voluntarily selling or giving away a laptop with information on it is the same thing as voluntarily sending that information through data pipes that someone else owns, but you’ve contracted to use, to a third party such as a website. If you give your data away to Google, Google can use them as it sees fit. Perhaps not without being evil (though, if the company – with your consent – anonomizes the information and uses it to enhance your browsing experience, that doesn’t really seem evil to me), but definitely without violating your rights. Sending your data to a website is giving it away. If you do not want to give it away, don’t. Some websites may not let you access them, or not let you access certain features on them, unless you give them your data. (Indeed, the Internet is structured so that if you want to access any website you have to transfer some data, unless you’re using one of the clever tricks we discuss here.) Too bad. You make the call. If you decide that the benefits of Gmail outweigh the “privacy” costs, you don’t have any complaint that Google has violated your rights.

When government takes your data, they do it without your consent. You don’t “give it away”; it is taken from you. (More often, the government takes data about you from another source that actually posseses the data, either directly by theft or by a coerced agreement. Whether your rights or the possessors’ rights were the ones violated is another matter, and irrelevant for these purposes.)

But what websites do with “your” data (the data that you voluntarily gave them in exchange for accessing their sites) is a matter you may be concerned about, but it is not a matter for government intervention. Talk to your content providers, not to your Congressperson.

We here at TLF have long been pointing out the benefits of targeted ads. But recently, we have focused on what I call the “supply-side” benefits – that targeted ads make free content possible by increasing the price advertisers are willing to pay for each pageview and therefore the amount of revenue content providers collect. That is a crucial point, and one that has yet to be absorbed by Congress, the FTC, and even other experts in Internet policy.

But we haven’t talked a lot about what I call the “demand-side” benefits – that targeted advertising is better for the viewer, directly, than non-targeted advertising. We have been too quick, I think, to legitimize the other side’s concerns, which they label under the heading “privacy,” by discussing the situation as one of trade-offs and TINSTAFL.

Our arguments have sounded like those supporting free trade agreements because of the trade barriers the other countries are lifting, implying that we should not remove our own trade barriers unless other countries agree to remove theirs. This is the wrong argument to make, for the simple reason that trade barriers don’t just hurt the economies of other countries; they damage our own. As my economics professor, Jeff Miron, put it, when everyone is shooting themselves in the foot, you don’t wait until you can get everyone else to agree to stop with you; you just let go of the gun.

Similarly, we should be discussing how much better it is to see targeted ads than non-targeted ones. Who feels nostalgic for the dumb, annoying scrolling banner ads of the mid-90s?

Do we really want this again?

Not me! I love targeted ads. I have found them ridiculously useful. I would never have found several of my new favorite bands without those ads on the right side of facebook. Somehow, facebook’s advertisers – using my list of bands I like from my publicly-accessible social networking profile – figured out better than Pandora (into which I’ve inputed even more data, but which hasn’t been giving me many songs I haven’t already heard) what music I should try. I’ve discovered sales on airline tickets, new events, useful law firm research tools, free web tools, and a lot more.

Today’s ads take up less space, chew up fewer system resources (proportionally, at least), and are more relevant – all while generating more money for the content providers. That ads that sell for more per view are also better for the viewer is not an accident. As Berin points out in his neat summary of the arguments against ad regulation, advertisers pay more for ads that viewers are more likely to find useful. Annoying and irrelevant ads generate less in clickthrough and conversion, and therefore revenue, for the advertisers. How do you know if Ad A is better than Ad B? If you’re more likely to use it. If you’re more likely to use it, the advertiser will pay more for it. More revenue per view = better for you.

Does anyone remember what it was like trying to support a website when ads were cheaper? Much more webspace had to be devoted to the ads. They were more prolific and annoying. Websites can minimize the amount of space they devote to ads if each ad pays them more. Conversely, if you think more valuable ads are not necessary to allow content creators – from small-time bloggers to major web innovators – to support themselves, do a little math. If each ad is worth half as much because of privacy regulation, how many more ads would the site need to display to get the same revenue? (It’s probably more than twice, because more ads – especially more irrelevant ads – reduces the value of the site to users and thus the number of views.)

This post has been about the too-often-ignored, direct, demand-side benefits to targeted ads. My next post will discuss why we should not view those ads as having “privacy” costs in the same way as real concerns of privacy from government onlookers.

I’ve spent the past couple of months interning for a large Silicon Valley technology company doing export compliance work. The company I’m working for does an enormous amount of its business overseas. And it exports, well, technology products, many of which are controlled. Laws ostensibly designed to prevent terrorism and proliferation in fact control way more than weapons and chemicals – indeed, they regulate even extremely mundane goods like servers, software with encryption, and the technical data used to design and build such products.

As a result, it has to employ large numbers of people to comply with the US’s, the EU’s, and other countries’ export control regimes. The US’s is particularly complicated, with a long list of prohibitions, some which can be circumvented if an exporter gets a license and exceptions to the licensing requirement (based on the classification of the goods, the destination country, the end-user, and the end-use). In addition, there are lists of parties – companies, universities, and individuals – with whom no company can do business. Companies that provide lots of goods and services: hardware, software, courseware for training on the products, etc., have to screen those lists many times – when a customer buys a product, when she signs up for training, when a part is shipped from a manufacturer, etc. They also have to spend lots on classifying their products and devising schemes to ensure compliance at every step in their complicated supply and distribution chains. And, because of the US “deemed export” rule, they often cannot share information with their US-based engineers who are citizens of other countries (who were hard enough to obtain visas for in the first place!).

And, yet, the US system – with all its complexity – still requires less effort than some other countries’, which require a license to export every controlled good. That entails significant delay and processing costs. Unfortunately, too little attention has been paid to these costs on doing business internationally when passing feel-good “anti-terrorism” and “anti-proliferation” laws and regulations.

As Tim Lee points out, some of the more ridiculous encryption controls have finally gone away, but as technology advances, more and more products will fall into a category (which are often based on technical performance) that requires a license. So, as American products improve, the costs of sending them overseas increases! One would think politicians supposedly worried about the trade deficit would see this as counterproductive to their goals of increasing US exports and reducing imports… but that’s politics!

Better not be offering incentives!

As I previously reported, the DC Circuit recently upheld a decision by the FCC to forbid customer retention practices used by Verizon to incentivize its customers to stay with the carrier rather than leaving for a VOIP provider. In the earlier post, I analyzed the bad economics of the FCC’s ban. In this post, as promised, I go into greater detail on the court’s decision affirming the FCC.

The latest issue of the Center for Internet and Society’s publication, Packets, has arrived and with it my summary of the case. The Packets piece provides a more neutral (but detailed) summary of the DC Circuit’s decision, without much analysis.

The big question before the court was whether what the FCC did was really pursuant to the Telecommunications Act, which forbids a telco “that receives or obtains proprietary information from another carrier for purposes of providing any telecommunications service” from using the information for a marketing purpose. If not, then essentially the FCC just went AWOL; instead of enforcing the law, as it is supposed to, it simply made its own law.

Indeed, that is exactly what happened here. The natural reading of the language, as the court admits, is contrary to the FCC’s ruling. To use an example employed by the court, when one says “Joe received information from Mary for purposes of drafting a brief,” the court reasoned, “it is overwhelmingly likely that the speaker expects Joe to do the drafting.” But Verizon is getting the information from other telcos not in order to provide their customers with phone service, but to cut off service. It is the competitors who are using the information to provide phone service. Mary is drafting the brief, so the statute doesn’t apply! The court never fully explains why it refuses to limit the statutory language to its natural meaning – saying only that one could grammatically read it the other way.

The explanation for this double-backflip to try to read into the statute some power for the FCC to issue its ruling is Chevron deference. I explained a little more about Chevron v. NRDC in a recent post, listing it as one of the worst Supreme Court decisions ever. In Chevron, the Court said that in reviewing administrative agency actions courts should not figure out what the law is and then ascertain whether the agency applied the law. Rather, courts should give great deference to the agency’s interpretation of the law. This is absurd for many reasons. For one, the job of courts is to interpret the law. And higher courts always review lower courts’ findings of law de novo, giving no weight to what the lower court found! In essence, the Court said, administrative agencies are better at figuring out the law than courts!

But, of course, agencies want to expand their power, not mekely enforce the law! The FCC views its job as controlling TV, radio, phones, the internet, and the entire electromagnetic spectrum. It doesn’t care what the law says; it just cares what it thinks necessary. This was vividly demonstrated this summer, when the FCC ruled that Comcast violated some policy guidelines it set forth. But the policy guidelines weren’t based on any law, just on what the FCC thought ISPs should do with their networks!

In the Verizon case, the DC Circuit next considered Verizon’s point that the FCC was treating the VOIP providers as telcos for the purpose of this case, but leaving open the possibility that they aren’t telcos for the puposes of other cases. The FCC wanted to have its cake and eat it too – to be able to use one definition of “telecommunications carrier” when it helps them regulate Verizon and apply a different defition, excluding the very companies they previously included, when it helps them do something else. Under Chevron, though, the court is to reverse the FCC only if this decision was “arbitrary and capricious,” and the court didn’t think it was.

Finally, the court considered the First Amendment argument. Verizon wants to contact its own customers about potential savings. Why can’t it do so? Well, “commercial” speech is treated differently from “real” speech (even if that speech is designed to make money) and gets essentially no protection. Of course, this is not a distinction the First Amendment makes, when it speaks of “freedom of speech.”

All in all, the decision is a good example of extreme deference gone awry. Courts have stopped doing their jobs of enforcing the law and making sure that the lawmakers follow it. The rule of law says that one law should be applicable to all and that the government does not get to make up the rules as it goes. A criminal court couldn’t punish someone for wearing an ugly tie if there were no law against it. Why does the FCC get to punish Verizon and Comcast when there is no law against what they do?

CNN reports:

An Illinois sheriff filed a federal lawsuit Thursday against the owners of craigslist, accusing the popular national classified-ad Web site of knowingly promoting prostitution.

The sheriff is upset that the site maintains a bulletin board system which is very lightly policed by its creators. It is little more than a forum for people to place their own advertisements. Thus, principles of caveat emptor abound, as anyone who has tried to find an apartment through the service knows.

Without craigslist, back to street walking

More importantly, Craig’s List is perhaps the best example of a site that should be immune from prosecution for the actions of its users under Section 230 of the Communications Decency Act. It exercises little control over what its users do, and that’s what makes the service both valuable and free. If the company had to hire thousands more people to examine every post that comes before it, its service would become more like Apple’s iPhone/iPod Touch App Store.

Section 230 allows websites like Craig’s List, Google, YouTube, Blogger, and pretty much every other user-driven Web 2.0 site the security to know they can operate free of lawsuits about what someone else, their users, did. Adam Thierer goes so far as to argue that it makes possible a real world analog for Nozick’s meta-utopia. Moreover, it is philosophically required by the tenet of justice known as the “principle of intervening action.”

Yet attorneys general and other politicians have been seizing on high-profile internet-related misfortunes like the MySpace suicide to push against Section 230’s safe harbor promise. Adam Thierer recently gave an excellent summary of where the section may be heading in the US. Other countries are even worse.

Perhaps even more dangerous than overt legal erosion of Section 230 through bad precedents (there are still some judicial defenders of the section out there, after all) is its covert destruction through coerced “agreements” forced upon ISPs and websites by AGs. They started popping up all over the place this summer and there is no end in sight. Indeed, CNN pointed out:

Craigslist entered into an agreement with 43 states’ attorneys general in November to enact measures that impose restrictions on its Erotic Services section. The agreement called for the Web site to implement a phone verification system for listings that required ad posters to provide a real telephone number that would be called before the ad went public.

Let’s hope the new administration stops the trend and puts life back into Section 230.

In December, the Fourth Circuit upheld the conviction and 20-year sentence of a man who downloaded pictures, drawings, and text emails depicting minors engaged in sexual acts. Receiving obscene depictions of “a minor engaging in sexually explicit conduct” is prohibited by 18 U.S.C. § 1466A(a). The court held the statute constitutional on its face, and as applied to downloading materials from the internet.

Receiving via the internet, the court said, is unlike mere possession in one’s home, as is protected by the First Amendment and Stanley v. Georgia, but is rather trafficking in commerce and so can be constitutionally prohibited. Of course, it is very easy to inadvertently “receive” obscene materials through the internet, whether in one’s Spam folder or on a pop-up, but the court simply hoped that inadvertent access would not be targeted for prosecution, because the statute requires knowing access.

Moreover, the court held that Congress can prohibit accessing obscene text and drawings, not just pictures and videos of actual children. Though stopping child pornographers is a noble goal, the Fourth Circuit’s ruling goes well beyond targeting production and even consumption of child porn and into banning express of deviant thoughts.

Judge Gregory, writing in dissent, pointed to great works of literature that have wrestled with the taboo subjects of incest and attraction to children. He pointedly wondered if, under the majority’s ruling, Nabokov could have been convicted for emailing excerpts of Lolita. Judge Gregory worried that the majority gave Congress “the power to roll back our previously inviolable right to use our imaginations to create fantasies. It is precisely this unencumbered ability to fantasize that has allowed this nation to reap the benefits of great literary insight and scientific invention.”

He concluded:

The Constitution’s inviolable promise to us is its guarantee to defend thought, imagination and fantasy from unlawful governmental interference regardless of whether such thoughts, imaginings, or fantasies are popular with the masses. It is in these moments that our grip on the rule of law and our fidelity to constitutional values is tested.

We must not let the government’s zeal in combating child porn override our most important constitutional limitations, as it has been recently.

You can find more information on the case in the Center for Internet and Society’s latest version of their Packets publication.

During my summer internship at CEI, a couple of us interns discussed the book Cato’s Robert Levy published last May, The Dirty Dozen: How Twelve Supreme Court Cases Radically Expanded Government and Eroded Freedom. We looked at Levy’s list of the worst decisions and sent each other lists of our own. Now that I’m taking ConLaw, I feel as though the time has come to post my lists of the twelve worst and the twelve best Supreme Court decisions of all time. It is by no means an exhaustive list. My inclusion of different cases than Levy does not indicate that I disagree with his assessment that those decisions are terrible – just maybe not as bad as the ones I select.

The Worst:

1. The Slaughter-House Cases (1873). The very worst decision ever made by the US Supreme Court. Eviscerated the 14th Amendment only five years after its adoption. It is best known for reading the Privileges or Immunities Clause, which was supposed to be (and could have been) a vehicle for both incorporation and unenumerated rights, out of the Constitution. But it also wrote out the Due Process Clause and the Equal Protection Clause, though those two clauses eventually crawled back into existence, to a degree.
2. Katzenbach v. McClung (1964). It was tough to decide which of the various cases reading the Commerce Clause expansively enough to permit Congress to pass any law it desires, thus destroying the basis of the federal government as one of defined and limited powers to include. But McClung seems to be the most expansive in both its result and its holding.

3. West Coast Hotel v. Parrish (1937). Abdicated the Supreme Court’s responsibility to prevent states from violating “economic” rights. Unlike the Commerce Clause cases, which were weakly limited by Lopez and Morrison, the freedom of contract cases have never recovered from Justice Roberts’ despicable “switch in time.”
4. Dred Scott v. Sandford (1857). The Court almost always does a good thing when it invalidates a law as being unconstitutional. Not here, where Chief Justice Taney overturned the Missouri Compromise with its promise of some limitations on slavery, relying on a ridiculous, convoluted, and racist reading of the Constitution as mandating that all African-Americans forever be treated as property. Chief Justice Hughes was right to call it a “self-inflicted wound,” as it resulted in the stigmatization of the Court’s role in deciding the constitutionality of laws, in particular on substantive due process grounds.
5. Korematsu v. United States (1944). Held that the government could lock thousands of U.S. citizens in concentration camps because of military necessity. Introduced “strict scrutiny,” while demonstrating its patent inadequacy. On the plus side, Jackson’s and Murphy’s dissents are some of the best ever written.
6. United States v. Carolene Products (1938). Typical case of the post-’36 era ignoring the Court’s responsibility to decide the constitutionality of laws. But Footnote Four made it even worse. Justice Stone introduced the idea of levels of scrutiny, claiming that it was fine in some cases for the Court to abdicate its role as the arbiter of the Constitution but less so in others. Made “economic” rights count less than other rights – and ensured that no right would ever be upheld just because the Constitution guaranteed it. Rather, the Court should uphold unconstitutional laws as long as the legislature had a really good reason (only a not-totally-insane reason in other cases) for passing it.
7. Plessy v. Ferguson (1896). The Court decide that separate “but equal” was fine, and so refused to overturn blatantly unconstitutional Jim Crow laws for generations.
8. Shelley v. Kraemer (1948). The 14th Amendment – and the rest of the Constitution – impose limitations on what the government may do; that’s both in the text and the nature of what a constitution is. Shelley v. Kramer eviscerated the “No State shall…” language of the 14th Amendment, by invalidating private contracts as in violation of the Amendment.
9. City of Boerne v. Flores (1997). Bad for two reasons. First, it overturned the “ratchet” interpretation of Section 5 of the 14th Amendment given by Katzenbach v. Morgan, thus holding that Congress cannot prevent states from violating any rights the Court hasn’t already said the states can’t violate. Second, the particular law it overturned was the Religious Freedom Restoration Act, designed to restore Sherbert v. Verner‘s interpretation of the Free Exercise Clause over the Court’s interprtation of it in Employment Division v. Smith, where the Court held that laws of general applicability can constitutionally prohibit the free exercise of religion.
10. Home Building & Loan Association v. Blaisdell (1934). Abandoned the Contract Clause, upholding a state law impairing the obligation of creditors to pay their debts.
11. Chevron v. NRDC (1984). Introduced “Chevron deference,” allowing administrative agencies to decide what the law is for themselves. Though higher courts review lower courts’ legal determinations de novo, administrative agencies get a much broader say in what the law is.
12. Buck v. Bell (1927). Justice Holmes upheld forced sterilization of the mentally ill, saying “three generations of imbeciles are enough.”

The Best:

1. Marbury v. Madison (1803). Got the Court in the business of overturning unconstitutional laws, thus making the Court the best of the three branches (for it can pass no laws of its own, merely decline to enforce bad ones made by the others).
2. Lochner v. New York (1905). Lochner was not the first, the last, or the strongest case enforcing the Due Process Clause of the 14th Amendment to prevent states from passing laws restricting freedom of contract. But it has served as the icon of the era, probably because of Holmes’ dissent. The Court held that a person who wants to work more than a certain number of hours can do so, in spite of state protectionist legislation to the contrary.
3. United States v. E.C. Knight (1895). Pointed out that the Commerce Clause only allows Congress to “regulate commerce… among the several states,” not manufacturing of products which then go out of state.
4. Griswold v. Connecticut (1965). Overturned anti-contraception laws. The justices varied in their reasoning. Douglas’s majority spoke of “penumbras” around enumerated rights. Harlan spoke of substantive due process (an interpretation later ratified in Lawrence v. Texas). But the best was Goldberg’s concurrence, invoking the totally-neglected Ninth Amendment and arguing that it serves as an interpretative guide for the Constitution, saying that rights should be read expansively.
5. Brown v. Board (1954). Overturned Plessy and got the Court involved in overturning state laws violating civil rights.
6. Reynolds v. Sims (1964). Following in the wake of Baker v. Carr, the Court held that states could not deprive their citizens of equal votes by screwy districting.
7. Mapp v. Ohio (1961). Introduced the exclusionary rule to the states, putting some teeth into the Fourth Amendment.
8. United States v. Butler (1936). Limited the Taxing and Spending Clause by holding that Congress couldn’t use it as “a means to an unconstitutional end,” namely the regulation of intrastate activities.
9. West Virginia v. Barnette (1943). Justice Jackson’s well-written majority opinion overturned state laws forcing saluting the flag, thus allowing religious freedom for Jehovah’s Witnesses and expanding speech rights.
10. Schechter Poultry v. United States (1935). Applied EC Knight to police the Commerce Clause, but also held that Congress could not delegate lawmaking authority to administrative agencies.
11. Reno v. ACLU (1997). Applied strict scrutiny to laws infringing freedom of speech on the Internet and overturned the bad parts of the Communications Decency Act (leaving the good Section 230 safe harbor provision).
12. Engel v. Vitale (1962). Overturned state official-led school prayer and got the Court into policing the Establishment Clause in a meaningful way.

Bureaucrash has just posted a new round of libertarian lolcats. Many involve tech policy. Check them out if you’re in the mood for some feline-and-political-commentary-based hilarity!

One of the biggest problems with the present copyright system is transaction costs, inhibiting Coasian bargaining. If I want to make a movie and have to get permission from dozens of different copyright owners, I may just give up – especially if I can’t locate some of them. (For more on the specific problem of orphan works, see Tim Lee’s techknowledge article at Cato and some of the many discussions on TLF.)

What copyright regime would best deal with the problem of transaction costs, while ensuring sufficient incentives to create? Robert Merges argues that the fair use doctrine may hamper the formation of copyright clearing-houses (or “collective rights organizations”) and thus increase transaction costs because fair use results in somewhat uncertain rights. See Robert Merges, Contracting into Liability Rules, 84 Cal. L. Rev. 1293 (1996).

Would compulsory licensing, as is required of song covers, radio, and cable retransmission, solve this problem? But, as I have argued elsewhere, compulsory licensing is price-fixing… and makes particularly little sense in industries where the players are all well-known to each other (like cable rebroadcasting network TV).

I don’t know what the solution is, but I’d like to hear everyone’s proposals for a more efficient (and decently liberty-friendly) system. Registration? Some stringent form of equitable estoppel?

Over the summer, I blogged about an FCC decision to ban Verizon’s practice of offering incentives to departing customers to get them to stay. Yesterday, the DC Circuit upheld that bad decision. When a customer of Verizon’s phone service decides to leave for a VOIP company, Verizon gets a notice that the number is being ported. When Verizon got notified that the customer was trying to leave, the company would offer her incentives such as “discounts and American Express reward cards” to stay.

This worked well for the customers, who got discounts if they stayed. It also worked well for Verizon, for whom it costs much more to find a replacement customer than to keep the current one. And it was really the best way to do so. If Verizon had given the incentives any time a customer threatened to leave, but didn’t start the process of doing so, then customers would just bluff to get the incentives. Verizon instead looked for a costly signal from the customer. And if Verizon had waited until after the port was already completed, it would cost the customer, Verizon, and the new carrier a lot of effort to switch back.

But the FCC banned Verizon’s efforts and yesterday the DC Circuit affirmed the Commission. I will follow with more details, once my summary of the case comes out in the March issue of Packets, the Center for Internet and Society’s publication summarizing important new internet cases. But for now, I should just note that the court hinted that the FCC’s reading of the statute it relied upon was a bit counterintuitive, but was compelled by Chevron v. NRDC to give the administrative agency great deference in its bad reading of the law. The court even noted that Verizon offered uncontroverted evidence “that continuation of its marketing program would generate $75–79 million in benefits for telephone customers over a five-year period.” Further, the court rejected Verizon’s First Amendment challenge, because the lower standard for commercial speech compelled the conclusion that Verizon’s sound marketing efforts didn’t deserve protection.

These precedents need to be revoked, or the growing administrative state will keep swallowing up more and more of our most important freedoms while preventing sensible and beneficial policies.

Seeing Adam’s recent post on the stimulus and its advocates, I had to toss in my two cents.

2008 was the year of Schumpeter. Creative destruction was doing its thing, getting rid of many unproductive old-economy companies that were simply creating economic waste by keeping inputs from going to their highest-value use. But this scared a lot of people who had grown used to the benefits capitalism had given them and who were therefore quite risk-averse. Even the entrepreneurs, upon whose ingenuity growth rested, had grown risk-averse and were demanding bail-outs of their own. As the government gave into demands for stability, the risk-taking class upon which prosperity rested began withering away.

If 2008 was the year of Schumpeter, 2009 may be the year of Hazlitt. In Economics in One Lesson, Hazlitt describes a mode of argument all too common in politics: the broken window fallacy. The notion is that by taking money from some and spending it, the government is “creating jobs” and enhancing productivity because money is circulating. Of course, this ignores what the people whose money was taken would have done with it. In other words, it is not beneficial to just spend money, no matter how badly. That is precisely the point that Eugene Robinson and other stimulus proponents seem to have missed.

At Stanford Law School, I am a member of the Stanford Law and Technology Association and the Center for Internet and Society. I write for CIS’s publication, Packets. I just published a piece summarizing the recent Third Circuit case once again holding the Child Online Protection Act unconstitutional. When the decision was released back in July, Adam Thierer wrote a wonderful post here on it. Adam’s and my pieces are complementary. Though Adam gave a nice assessment of COPA’s future, my summary goes into a bit greater detail on the court’s legal reasoning. If you’re interested in the law or in the constitutional principles involved, you may want to check out this interview with my brother, who was a counsel on the case.

As Adam Thierer has previously commented on this very blog, Mythbusters is “the best science show on TV in years.” Since the show tackles ridiculous beliefs that have entered the popular culture, it would make sense that at some point, they’d expose some dumb government policy. But, generally, the Mythbusters stay away from terribly controversial topics. So, unlike Bill Maher, they don’t debunk religious beliefs. And, unlike Adam, they haven’t shown that concerns over airplane terrorism are overblown.

But maybe Adam and Jamie’s policy is changing. I just watched an episode (which originally premiered in 2006) where they test whether cell phone signals can interfere with airplane avionics. Shockingly, even when they hauled into a real plane a radio transmitter broadcasting all kinds of cell phone signals at hundreds of times their normal power, there was no interference at all! This makes sense; after all, there are lots of radio signals travelling through the air everywhere anyway. Airplanes are built to ensure that these signals don’t affect their navigation equipment. And the EU has allowed cell phone use on planes for years, without incident. Plus, we all know that people have used their phones on planes in the US, just more covertly.

I reported on the US’s absurd ban before over at OpenMarket. And it looks like, in spite of liberalizing moves on the part of the FAA and FCC, the ban isn’t going anywhere, thanks to Congress. I guess you can bust myths with science, but the government won’t listen. If only we could recruit Mythbusters to show that the FDA does more harm than good or that Social Security creates fiscal insecurity.

Yesterday, after my Criminal Law class, I went to a lunch talk sponsored by the Stanford Biolaw and Health Policy Society about “abandoned” DNA – that is, DNA traces that people leave all over the place. It was given by Prof. Elizabeth Joh, visiting Stanford Law this year from UC Davis Law. She focused on her recent law review article on the subject.

Joh’s basic argument was that DNA is fundamentally different than the other detritus we abandon on a regular basis. She contended that, though we might not have an expectation that the soda bottle we tossed into the public trash can won’t be seen by anyone, we have an expectation that it won’t be mined for our saliva and the genetic information it contains. Joh even argued that DNA traces are fundamentally different from fingerprints, since fingerprints can only identify us, but cannot give investigators a view into fundamentals about who we are (including our health risks).

Joh contrasted her view, which focuses on privacy, from what she called the “old” trespass view. Under that perspective, what was wrong about an FBI agent slipping into your house to implant bugs was not that the government could now listen into everything you say in your home, but rather the property violation involved in breaking in. Similarly, under the trespass view, a cop could not run a cotton swab on the inside of your mouth to collect DNA (without a warrant) because it would violate your property in yourself, not because it would reveal your genetic information to the government. But the trespass view would have no problem with the government picking up that soda bottle out of the trash and collecting your DNA from it, to match you to a crime.

Governments have been wary of actively collecting this abandoned DNA, however, and so have gone to great lengths to get suspects to voluntarily send the government their DNA. For example, they’ll ask suspects to mail them information, then get their saliva from the envelope or stamp they licked. Joh is wary of these practices as well.

Joh fears that if we let governments analyze abandoned or given DNA, they might create a national database of DNA, like the present national fingerprint database. Governments may not use this just to identify suspected criminals; they may also mine it for racial data or to study genetic predispositions for criminality.

Though I would strongly oppose the creation of such a database, I would do so on fundamentally different grounds than I would oppose FBI agents breaking into my apartment and installing hidden cameras. I guess I’m old-fashioned, but the trespass argument seems to me much stronger – and of a fundamentally different character – than the privacy argument. As Jim Harper has discussed, we don’t really have a fundamental right to privacy, in the same way as we have a fundamental right to property. Privacy is a set of relations we have with others, by which we share information with them. Property is a basic right to own ourselves and our stuff.

I would have no problem with the national DNA database if I were certain that the government was totally benevolent and could never do anything wrong with the information at its disposal. But, of course, that is not the world in which we live. We find the national DNA database scary not because it is inherently wrong, but because a lot of bad can be done with it. This is fundamentally unlike breaking into someone’s house. Even if I knew that the government would only do good things with the data it collected from my apartment and keep it entirely secure, I would still object to breaking into my apartment and vandalizing it. That action violates my rights, my rights to my property. The DNA database is “just” bad policy – a waste of money, and a dangerous new tool for governments to do evil.

Presumably, everyone reading this post has purchased software at some point in the past 15 years. If you have, you may have agreed to a contract unwittingly. Breaking the seal of the shrinkwrap around the box might bind you to the terms and conditions contained inside. This is but one of many new ways you can be determined to agree to contractual terms you may have never seen.

In the last decade, Gateway came under fire for its means of doing business with consumers. A customer would order her computer over the phone, but when it would come, it would contain a list of terms including things like a mandatory abritration clause – and always stating that the customer was deemed to have accepted the terms by not returning the computer (at her expense) by some period of time. A number of court cases raised the question whether this practice really created a binding contract. 

As libertarians, we are generally in favor of contracts. But a contract is a mutually consensual agreement. The critical question for shrinkwrap contracts and the like is whether both parties have really assented. In the Gateway cases, there are three main interpretations of what is going on:

1. The phone conversation ordering the computer (or, analogously, buying the software from the store) was just preliminary negotiations. When Gateway shipped the computer, it offered it to the customer with the terms that came with it. And since under traditional contract law “the offeror is the master of his offer,” Gateway could specify that keeping the computer without returning it would be deemed to be acceptance of the offer. So, the customer has accepted the terms and is bound by them. This was Gateway’s interpretation, and the one a New York appellate court adopted in Brower v. Gateway 2000.
2. The phone order was the customer’s offer to buy a computer from Gateway. Gateway’s shipping the computer accepted it, but its acceptance provided deviating terms. Under the Uniform Commercial Code (UCC), which governs the transaction, the deviating terms must be expressly assented to by the customer. Keeping the computer did not accept the deviating terms, so the contract has only the terms agreed to by both parties and the default terms provided by the UCC. This was the result a Kansas court came to in Klocek v. Gateway.
3. The phone conversation (or, analogously, the sale of the software) was actually a complete contract itself, which agreed on the item to be purchased and the price the customer would pay. The terms that came with the computer were Gateway’s proposal to modify the agreement. But since the customer didn’t agree to the terms explicitly, but rather just accepted the computer, the new terms were never accepted.

Libertarians might legitimately disagree about which of these interpretations is most plausible, but I think that #3 is the winner. We should be in favor of contract rules that allow people to form contracts in a number of ways, including telephone sales calls, but we should make sure that both parties really agree to the contract. This doesn’t mean that we have to look at whether the parties had “equal bargaining power” or thoroughly read the contracts before signing.

But what Gateway was doing in those cases, and what software companies do all the time, seem an awful lot like the homeless guy washing your car window and then demanding payment. This is not a contract. As the great libertarian philosopher Robert Nozick said, “You may not decide to give me something, for example a book, and then grab money from me to pay for it, even if I have nothing better to spend the money on” (Anarchy, State, and Utopia, p. 95.). Though the offeror must be allowed to require the offeree to do a lot to accept the contract (e.g. providing a car dealer with your credit history), the offeror cannot bind the offeree even though she hasn’t done anything. I should not be required to ship back a product, at my expense, to refuse assent.

But perhaps shrinkwrap contracts will become less of a problem as more sales – and even content acquisition – occurs online. Before I buy a song on iTunes, I have to agree to the terms Apple gives me. If I don’t like the terms, I can refuse to buy from iTunes and instead use Amazon Music Downloads or go buy a CD. And few people buy things over the phone any more. When you buy a physical product online, the seller can require you to agree to its terms before it accepts your order, charges you, and ships. These cases are explicit agreements, and libertarians should support protecting these. But, as Tim Lee has pointed out here, we should be careful before finding contracts where none really exist.