Problems with H.R. 1748 (the "Smith Bill")

The reaction of many to SPAM (junk E-mail) has been "There oughta be a law" to stop it, and there are several laws in committee. One that's gotten a lot of attention is H.R. 1748, Congressman Smith's "Netizen's Protection Act," which is lobbied for by an ad-hoc group named CAUCE.

This essay was written in 1998, when this bill was pending. It died, as have all federal bills to this date, though many (unconstitutional) state laws have been passed. I leave this essay here because many of the elements of that bill have popped up in other bills, and the issues are interesting.

The bill attempts to amend the anti-junk-fax law to include unsolicited commercial E-mail. It proposes that if somebody sends you a single unsolicited E-mail "advertisement" (solicitation for business) you can sue for $500. If they do it in knowing violation of the law, you can sue for $1500. They must put on their real E-mail address and other contact info to make it easier to sue. The $500 price, while a million times higher than the cost of sending an E-mail, is small enough to be tried in small claims court. It is the same penalty as for an unsolicited Fax advertisement.

Laws are the last resort

The simplest reason this law is bad is that laws restricting freedom of speech are a last resort. Spam is a very young problem in social terms, having started seriously around 1995. The internet changes more in a long lunch hour than some industries change in a century, and it's far from time to call in the government yet. Even working at its best, the government is unable to deal with things that change this way. We saw that in the case of the Communications Decency Act. In the 1.5 years from the time the law passed to the time we had it struck down, an entire industry of tools to filter the net for children had sprung up.

While some methods have been tried against Spam with varying degrees of success, it is absurd to suggest that non-legal methods have had their turn and failed. The law calls for the government to regulate the content of E-mail, making single E-mail messages illegal, with a penalty of $500 or more, based on what they say, rather than how they are delivered.

Restricting Free Speech

"Congress shall make no law..." begins the U.S. bill of rights. While commercial speech has been ruled by the courts to have slightly less protection than other forms like political speech, it is still protected. Any law that proposes regulating a message based on its content rather than its form should be highly scrutinized.

The courts agree. Even laws that restrict commercial free expression must show that they protect an abiding national interest, that they do not prohibit any legitimate, protected speech, and that they are the minimum effort necessary in order to attain the goal. They must be as specific and understandable as possible. They must not make senders of legitimate, protected communications afraid or uncertain as to whether they can say their message, lest they create what is known as a "chilling effect" on legitimate expression.

That's a tall order, and it should be. Chances are none of the proposed laws pass this muster. If you can think of a simpler, less restrictive law (or non-governmental system) to adequately deal with the problem, then the law doesn't cut it.

This bill defines only a civil right of action -- which is to say it enables the receiver of an unsolicited commercial E-mail to sue and win $500 in small claims court, or $1500 if the sender flouted the law. You can bet that such a penalty will make people think twice about sending messages. But not just about junk mail, but also about borderline messages which are actually legitimate and turn out to be desired by the recipient.

Is a "signature" on the end that promotes a company or a web site an unsolicited ad? Probably not, one would hope, but people will be worried. If you send that message to somebody far away (you don't know where an E-mail recipient is) will you have to fly to Anchorage, Alaska to convince the judge not to fine you $500?

There are legitimate unsolicited commercial messages. Some areas of business thrive on them. Sometimes the recipient wants them, but doesn't know in advance to ask for them (and make them solicited.) The most common type of such message is the "referral." Perhaps you and I are friends. I have a meeting with a 3rd party, and learn she is starting a new business you might want to invest in. I tell her to contact you and drop my name. This happens very frequently, yet under this law it would be illegal for her to E-mail you based on that referral, unless you had given me an explicit declaration of permission to solicit such mail.

E-mail is becoming, and for many has already become, the medium of choice for both personal and commercial communications. As the #1 medium, it deserves the most protection of freedom of expression, not the least.

It's also important that it be clear that the online media are every bit as protected as the print media. If it's legal in print it has to be legal online, or we turn the online media into 2nd-class media, not protected by the 1st amendment the way others are. If you can do it in paper mail, it has to be legal to do it in E-mail. A law that says otherwise is a dangerous precedent. (And an unconstitutional one, based on the 9-0 supreme court decision in our case against the CDA.)

And even if it is legal to restrict commercial speech where political speech would be protected, it's inconsistent. The truth is that a Spam for a charity or religion or political party or personal cause is just as annoying as one trying to sell something. Many people may not recall that the first famous USENET mass-posting SPAM wasn't the Canter and Siegel immigration ad, but rather one saying "Jesus is coming."

You also can't select out commercial speech just because it is less protected, if non-commercial speech is just as annoying. The supreme court said that in Cincinnati v. Discovery Network.

A Fax is not an E-mail

The junk fax law, part of the Telecommunications Consumer Protection Act, applies these penalties to an unsolicited fax advertisement. Frankly, I think this law could have been better designed, too, but its case is a lot clearer than Smith's application of hte principle to E-mail.

Receiving a fax has incremental costs. When the law was passed, most faxes used thermal fax paper, which costs about 5 cents or more a page. Not enough to break you, but a real cost. Worst of all, however, a fax uses a regular telephone line. When a junk fax is coming in, your line is busy. A fax or phone call you want gets a busy signal, and that may mean you never get the message you wanted.

If the junk faxer is making a local call, they may not have incremental costs (unless they have measured service local calling, usually 1 cent/minute.)

An E-mail is cheap. Really cheap. Really, really, really cheap. We built it that way. In fact, the general decision was that E-mail was, unlike nuclear power, really too cheap to meter. There were metered E-mail systems before the internet, like MCI-Mail and mail on the online services. They lost out, killed quickly in competition by internet style shared-cost E-mail.

In sending an internet E-mail the sender and recipient share the cost. They each pay for their end, much like both ends of a local phone call. There are no "sender paid" calls and no "collect" calls.

Why? Well, almost everybody in the USA gets a flat-rate internet connection, and has no incremental cost to receive an E-mail. Estimates to come up with a partial cost for a typical user range from $.00002 to a maximum of $.003 for a 4kb E-mail. It turns out E-mail is the cheapest non-personal medium in which one can receive a message. Hardly one to have something banned for reasons of cost. However, the exact figures are not directly important. Whatever you might think the cost is, nobody denies it's very cheap, and getting cheaper.

This cost is less than it was years ago when E-mail systems were designed and it is going to continue to go down. Even in the past when it was an order of magnitude higher, E-mail system designers decided not to put metering on E-mail, because it wasn't worth trying to bill hundredths of pennies to mailers with all the infrastructure that would involve. They knew that there would be people who would send unwanted E-mail, but it just seemed a lot simpler to share the insignificant cost.

That lack of metering is in fact part of the Internet Cost Contract which says that you won't complain about the cost small volume misuses of your net connection.

They may have been wrong about that. If they were, the answer is not to fix their mistake with laws punishing particular types of E-mail, but to fix their mistake they way they would have fixed it.

The junk fax bill allowed the punishment of fax advertisers who did use some significant if small misappropriation of the receivers resources. As such it passed 1st amendment muster in federal district court. But the cost of a single E-mail is three orders of magnitude lower and will continue to decrease, and the court cited "significant" cost shifting as its prime reason for upholding the law.

Nobody is truly up in arms over hundredths of pennies -- or even pennies. The bill proposes damages over a million times larger than the appropriated cost of receiving an E-mail. And the truth is that all unwanted E-mail, not just business solicitations, appropriate that tiny cost from the recipient. That's the way E-mail was deliberately built.

The truth is we get Spam not because it transfers the cost to the recipient, but because the cost is simply very low. Even if there were a way for Spammers to pay the recipient's real costs of E-mail delivery I doubt it would stop them. That cost is just too small for either side to worry about. The cost of annoyance and invaded privacy, that's another matter.

But it isn't against the law to annoy people or waste moments of their time. Nor do we really want it to be. Can you imagine it being illegal to ask somebody a question on the street that annoys them, or to waste 2 seconds of their time in a line or driving down the road?

It is against the law to harass people -- to contact them repeatedly, especially when they have made it clear they don't want it, but this law punishes the first (and perhaps only) unwanted E-mail from a party.

Bulk is the problem

As discussed in my general essay on the problem of junk E-mail, it is cheap bulk that is the problem, not commerce. E-mail is cheap -- we (the builders of the network world) designed it to be very cheap. And that means bulk is cheap.

E-mail was designed as a "shared" cost system -- each side pays for their part of the costs. It was designed that way because it was felt the costs were so low it just wasn't worth the administrative hassles of trying to figure out billing. No surprise, this is how all internet traffic was set up. The phrase "too cheap to meter" never worked with nuclear power but people figured that it would with E-mail. Metering just seemed silly for quickly diminishing fractions of pennies.

It turns out almost all media are shared cost like E-mail. The sender pays to get the message to the recipient's "portal" and the recipient takes it from there. Only very rarely does the sender pick up the cost of whatever the recipient needs to receive and discard the message (TV set, newspaper delivery, mailbox, driveway, secretary, fax machine, computer, car etc.) E-mail turns out to be the cheapest medium, per message, in which to receive and discard a message, while fax is the most expensive. And E-mail keeps getting cheaper every year.

The mistake made was bulk. People built mail systems to welcome mail from all senders but the goal was to build a system that allowed one to one communication, and also allowed people to voluntarily subscribe to mailing lists. It was not expected that people would be put on mailing lists without their permission and against their will. Without cheap bulk there is no Spam problem, just as there is not one in the other media where bulk communications are not cheap.

Unfortunately HR 1748 attacks commerce rather than bulk. It punishes a single unsolicited commercial E-mail. But commerce is not the problem, it is just the most common motive for the abusers. The problem is unrequested bulk mail.

Put another way, it is advertising -- sending one message to a group of people using their equipment -- that is the problem, not person to person messages (soliciting or otherwise.)

Indeed, unsolicited bulk mail that says "Vote Republican" or "Save my child" or "Donate to the United Way" or "My god is the only true god" are just as annoying as advertising. And these forms come, though advertising dominates the Spam world.

Relay abuse is the problem

Another big cause of Spam today is what is known as "relay abuse." Some bulk junk mailers take advantage of the fact that most net mail servers were configured "open," and if handed mail for an address by anybody, would do their best to deliver it. This led some extremely unethical parties to abuse this to send bulk mail. They would find an innocent victim, and pass along a message to their computers with instructions to deliver it to 1,000 people. The computer, in a friendly mood, would do this, eating up its own resources and annoying those thousand people.

This should be illegal, and probably is, as theft of telecommunications. When it finally gets stopped by various means, it will cut down on a lot of the problem.

Yet HR 1748 insists on making a single E-mail unlawful, when the problem revolves around the abuse of bulk.

Less Restrictive Laws

In general, if one can design a workable law that is less restrictive on speech but still gets the job done, that makes the more restrictive law unconstitutional.

While we don't necessarily want to advocate more laws, the ability to provide less restrictive ones as examples is an argument that Smith's bill is the wrong way to go.

Because bulk is the problem, a law that limits bulk would still solve the Spam problem but be less restrictive than a law like H.R. 1748 which restricts a single piece of E-mail. It is possible to spot bulk mail and deal with it. Anybody can tell a piece of bulk mail when they get it with very good accuracy. It's almost always easy to tell the difference between a bulk message and one written just for you.

If you had a law which, like the Smith bill, allowed for civil action, one could apply it only to bulk mail. It's easy to prove bulk mail because in a civil case you just put the sender on the stand and ask them if they sent it as bulk mail. They can't "plead the 5th" -- they are compelled to answer or be in contempt, or commit perjury. Those are criminal offenses. And nobody is going to commit perjury when it would be so easy to get caught out. If you sent a mailing to 10,000 people, you had better be very sure 2 have them haven't gotten together before you testify under oath you didn't send it bulk.

And it's easy for 2 of them to get together. If they are both at a single site, like AOL, the mail system can immediately detect bulk mail. Otherwise recipients who suspect bulk mail need simply quickly bounce it on to a computer system that does those checks. Such systems are already in place to stop USENET spam.

Another less-restrictive law would be one requiring an "opt-out" policy for types of mail, rather than the "opt-in" that the Smith bill effectively mandates. In a free society, communications are never "opt-in" but it's fair to give individuals the choice to opt-out. Reliable Opt-out could be implemented in extended mailer protocols, for example. Establishing opt-out lists of e-mail addresses such as those used for telephone solicitation is more difficult, though not entirely impossible. (The difficulty comes because while there is only one telephone number for most lines, an E-mail address can often be expressed in many different ways.)

The best way perhaps to do "opt-out" would be with tagging on E-mail that required it to be identified ("tagged") as bulk or person to person, and indicated whether the sender was already known to the recipient or not. This would allow the development of E-mail tools that could allow individual users to control what sort of mail they get.

You can read more details about detecting bulk mail.

Other ways

Such less restrictive laws may not be necessary -- in fact, tagging may be possible without laws -- but the simple fact that such laws can be designed makes H.R. 1748 a bad law.

Other things being tried without laws are special filters which detect bulk, and "spam-bait" addresses which detect messages sent to addresses that only an "address harvester" would ever mail to.

Efforts are underway as well to punish people who abuse the net by sending Spam, by making sure ISP contracts bar their customers from such abuse and get them to lose their access when they do. Many blacklists exist to block mail from those abusers.

It's also possible to make "secretary" programs which challenge unknown mailers with a simple E-mail the first time they correspond with you, the way a secretary asks "who's calling?" to unrecognized phone callers. These stop Spam dead, with minimal impact on legitimate mailers.

These methods, and the proposed laws above, don't focus at all on the content of mail, but rather on things like how many people a message is sent to, and what the relationship of the sender is to the recipient. These things can be regulated without restricting freedom of expression. It is regulation based on the content and motive of messages that goes over the line.

E-stamps

Here's the trick. The E-stamp, in effect a digital cheque for a small amount of money (32 cents for example!), would be included with the mail. The recipient, not the mail company, could redeem it if they wanted to. Chances are they would only do so for anything but mail that annoyed them like Spam. In fact, it might be considered quite rude to redeem the E-stamp on mail from friends or any other one to one mail. If somebody you mail does redeem your E-stamp, you're hardly out much money, and you may decide not to correspond with them further, or figure it's worth 32 cents to annoy them.

But if somebody tries to send an ad to 100,000 people it's another matter. Many would redeem their stamps and the cost would be prohibitive. This would return the checks and balances that keep bulk messages under control in other media, like postal mail.

Of course, this is a lot of work to implement. Oddly enough, E-mail systems used to be sender-paid. MCI Mail and mail on the online services all cost money for the sender. And as such they didn't have bulk mail abuse problems. But shared-cost internet mail was so cheap, and too cheap to meter. It crushed the sender-paid E-mail systems dead. That was a choice we made, and we can un-make it, but we shouldn't use the government to solve the problems caused by the choice.

So why attack 1748? Are you defending the junk E-mailers?

Far from it. I hate them about as much as anybody, perhaps more. When I and the EFF sued to strike down the Communications Decency Act, we were accused of defending the pornographers as well. We weren't, except indirectly. We were defending the internet. Making a stand that what was legal in print should be legal online.

We're not defending spammers here, but the right of people to have, without government interference, open E-mail. But having free and open E-mail does mean the law doesn't directly forbid abuse -- it's a price one pays for that freedom.

Is it too much a price? Should we give up some of the freedom to avoid paying it?

A better way

There are lots of better ways. The EFF stands against government regulation of the content of E-mail and for the enabling of informed, individual choice. E-mail deserves to be protected, and subject to all the 1st-amendment safeguards that paper messages have. Let's fight junk bulk E-mail, but not by calling in the government to control its content. As much as you might hate Spam, this is not the way to fight it.