There have been a lot of interesting responses to our latest white paper on spam. We appreciate the feedback, and in response, we've complied a list of issues you've raised and provided our replies. We hope this helps clarify EFF's position on some of the more tricky questions in this debate.
The Argument Against "Pay to Play"
Some groups and individuals have argued that the best way to prevent spam is essentially to "tax" it through proposals like e-stamps. We believe it's unlikely that this kind of scheme would succeed in preventing any significant amount of spam. But more importantly, we believe that introducing artificial costs into Internet communications would do a lot more damage to the owners of noncommercial email lists than it would to spammers.
Spam is big business. It not only pays, it pays well, with tremendous profit margins for spammers. Placing an artificial cost-burden on sending email might reduce those margins somewhat, but we've yet to see any compelling evidence that it would actually drive any successful spammer out of business, much less affect the wildly successful ones that are the bulk of the problem.
Second, proposals like this miss the central point we're trying to make here -- that noncommerical, grassroots mailing lists, often administered by a single person in his or her spare time, are one of the great things about the Internet.
There has been a lot of heat produced by our discussion of Moveon.org's problems with its mailing lists, but very little light has been shed on the problems smaller, less well-funded groups have been experiencing. Mailing lists like Crypto-Gram, IP, Politech, and TidBits likely would not exist if the list owner was required to pay for each mail sent and received. It should never be the case that if a list-serve is popular and growing, or a list-serve topic sparks an especially lively discussion, the list owner (or any speaker) has to pay more money. More speech should be rewarded, not penalized.
The Argument for Global Standards
Some readers of our white paper have argued that we are imposing a US-centric notion of freedom of expression on an unwilling global community when we say that the problem with spam filtering is that it endangers free speech. But freedom of expression is also part of international law. It's included in the Universal Declaration on Human Rights and the Covenant on Civil and Political Rights, which have been adopted worldwide. There should be no question that freedom of expression is recognized as a basic human right all around the world.
Where the US differs from other countries is at the "margins," not the center, of the free speech issue; for example, we have different standards with regard to hate speech, libel, and pornography. But, again, this does not mean that the international community does not value free expression. It therefore makes no sense for EFF to pretend that protecting legitimate email is important only to US citizens.
The Argument for Freedom of Expression on Private Property
Some critics have noted that we suggest that free speech doctrines ought to inform the discussion about the problems caused by overzealous or poorly implemented anti-spam efforts. These critics claim that EFF doesn't understand the difference between government and private actors when it comes to speech.
As an organization that has devoted 15 years to fighting for freedom of speech, we obviously understand the difference between private and public actors in the law, and that is why you see no legal claims in the paper. But that shouldn't end the matter.
There is no real question that if the government started using current anti-spam mechanisms to decide which of your email messages you should receive and which you should not, with no due process, no method of redress, and no accountability, it would face serious First Amendment challenges. The effect on speech is the same when private parties engage in this kind of behavior. Wanted speech is being blocked from reaching a willing audience, often without that audience's knowledge or consent.
We argue that private actors running mail servers or providing anti-spam tools should care about silencing wanted noncommercial speech as a matter of principle, even if their acts do not create the basis for a First Amendment lawsuit.
Our goal is to have an honest discussion about these collateral effects and to ask the Internet community as a whole if the kind of censorship that we would not tolerate from a government is what we want our ISPs and anti-spam community to engage in on their own. We think that's a fair question.
The Argument for User-Controlled Filtering
Our critics often ask us what kinds of antispam techniques we recommend, since we seem to have negative things to say about the most common systems available.
We recommend user-side filtering using technologies such as SpamAssassin. One of our concerns, however, is that SpamAssassin can be misused (and often is). Despite their usefulness when deployed under ideal conditions, SpamAssassin and other filtering technologies can be used to block email based on the language it contains or the country of origin.
It's important to note that blocking by language or country is not a default setting in SpamAssassin. But that doesn't mean this tool isn't being used that way. Many systems administrators describe routinely blocking email from China and/or Korea using the SpamAssassin point system. This kind of blocking becomes an issue when it's done server-side and users have no control over it. If a user wants to block all mail from China, that's her prerogative. But when a sysadmin decides the question unilaterally, that's a problem.
Ultimately, EFF would like to see users given the right to determine for themselves what constitutes "wanted" and "unwanted" email. Spam is in the eye of the beholder, and the beauty of user-side filtering is that it allows end users control over the process of creating filters so that they can get rid of email they don't want while still receiving what they do.