The War On Spam Takes A Novel Turn

Professor Marshall Van Alstyne of Boston University School of Management has a plan to win the war on spam. He wants spammers to pay you for wasting your time.

"We're really trying to give you back what is a property right in your own attention," he says. "Since interruptions are costly, what you're basically doing is asking the sender to make these interruptions worth your time."

Technology companies impose a cost on spammers by blocking spam. The government imposes a cost on spammers by locking them up. But to date, the potential profit for spamming continues to exceed the likely cost. Spam continues because it pays.

Leave it to an economist to price spammers out of the market. Van Alstyne describes his anti-spam scheme in "An Economic Response to Unsolicited Communication," a research paper co-authored by graduate students Thede Loder and Rick Wash that should be accepted by a prominent economics journal in a month or so. He is proposing an "Attention Bond Mechanism," money put up by E-mail senders as a form of spam insurance. In the paper, he argues that an attention bond designed to promote valuable communication can outperform technical solutions designed to block low-value content. Economics, in this theory at least, trumps technology.

This might be a matter of solely academic interest if it weren't for the company Van Alstyne has been keeping recently -- last week he briefed industry experts at Google, Stanford University, Sun Microsystems, and Symantec. Last year, he met with Microsoft and the Federal Trade Commission. Next month, he's presenting the idea to the Federal Communications Commission.

Van Alstyne isn't alone in his approach -- economic solutions to spam are starting to enter the market. A startup company called Paritive Inc. hopes to commercialize the idea. Thede Loder, who studied with Van Alstyne, is the CTO. And a company called Vanquish Inc. late last year began selling to ISPs E-mail gateways that implement a financial bond system.

"Vanquish isn't the first to propose sender payment systems for E-mail," CEO Philip Raymond writes via E-mail. "But I believe that there have been no public references to the more subtle approach of sender liability prior to the end of 1997." The subtlety here is that there's no payment if the recipient isn't irritated with the message.

Raymond says he first heard of Van Alstyne and his co-authors in January during a spam conference. "While our theories are strikingly similar, they were arrived at from different perspectives," he says. With patents and a shipping product, Vanquish may well be the company that establishes the attention market if it plays its cards right.

Imposing a cost on spammers isn't exactly unheard of. Return Path Inc. uses financial bonds to improve message delivery and deter spamming. The difference is where the money goes. If a participant in Return Path's Bonded Sender program sends spam and generates enough complaints, the sender's bond gets paid to the Internet Education Foundation, a non-profit Internet advocacy group. And since participation in the program is voluntary, spammers can simply forego the greater rate of deliverability they'd get in the program and rely on volume to overwhelm filters.

Microsoft has been exploring related ideas. The Penny Black Project at Microsoft Research is one such example. It's a scheme to impose a computational cost, making spammers expend extra CPU cycles when sending messages, thereby slowing the number of messages that can be sent per second and reducing profits as a result. However, computation is a tax that no one collects, making it less beneficial than a cash distribution.

The way Loder, Raymond, and Van Alstyne see it, bond money should be available to the recipient rather than a third party, in effect creating a market for attention. It's not a tax because only unwanted messages incur a cost. It's a guarantee not to waste your time.

For E-mail recipients, the appeal is obvious -- it's hard to argue with the prospect of getting paid, even if the amount per message typically would be only a few cents. And it has the advantage of giving E-mail users control over their privacy. Those who don't want any E-mail solicitations could price access to their inbox so high that no marketer would risk that much. The privacy benefit accrues to senders as well.

"We can allow people who want to remain anonymous to do so," Loder says.

An attention market would even be useful in a non-commercial context. An executive like Bill Gates could price access to his inbox to reflect the value of his time. And those who had legitimate reasons to correspond with Microsoft's chairman could rest easy, knowing that he wouldn't cash in the substantial bond required to get his attention.

Marketers would benefit, too. Risking less than the cost of a direct mail campaign in the form of a bond -- an amount affordable to legitimate businesses but not for spammers sending millions of messages -- marketers could easily determine who wanted to hear from them. It would be a one-time cost, too, since under Van Alstyne's proposal, senders only risk their bond when initiating contact for the first time.

"Suppose L.L. Bean, Citibank, or Expedia, are trying to market camping gear, financial products, or travel," he says. "You happen to have an interest in travel. You could simply seize the bonds on these other marketing advertisements, indicating to the companies that you're not interested in that."

But in the case of the travel-related E-mail, he explains, you would choose not to collect the bond, which would demonstrate to the sender that you're interested in future communications of that sort. The result for marketers would be self-cleaning marketing lists and an efficient way to determine customer interests.

Loder says that attention bonds compare favorably with direct-mail costs. He estimates that a direct-mail campaign might cost around 50-cents per piece -- 15 cents to buy the name and address, 25 cents for postage, and 10 cents in production costs. He says most consumers would price their attention at a fraction of that.

But the question is whether the Internet community in general will buy it. The primary stumbling block is that the infrastructure needed to support micropayments for spam is non-trivial and implementing it so that established players don't feel threatened or ignored won't be easy.

Ray Everett-Church, chief privacy officer for ePrivacy Group and counsel for the Coalition Against Unsolicited Commercial E-Mail, observes that the Internet community has been bickering for years over far less sweeping architectural changes. And he says that any system that puts too much control in the hands of a single company would face resistance.

Bruce Schneier, noted security expert and CTO of Counterpane Internet Security Inc., is similarly skeptical. In reference to bond proposals such as Van Alstyne's, Schneier wrote in his blog last week, "These solutions generally involve re-engineering the Internet, something that isn't done lightly"

"There will be some technology adoption costs, and we would strongly recommend an open standard," concedes Van Alstyne. But there is precedence for this sort of infrastructure -- the Automatic Clearing House Network, which coordinates fund transfers nationwide between banks. A cryptographic system to secure the micropayments also would be necessary. Van Alstyne points to Peppercoin Inc. as a possible provider of that technology.

Despite the obstacles, Van Alstyne has faith in the curative power of the market. "If you can assign property rights in the problem, then you get efficient trading on it, then you get a better solution than almost any other possible alternative," he says. "That's why I think it will work."

Perhaps the most pressing question is whether anyone will get it. According to Raymond at Vanquish, it's not easy to explain how an E-mail warrant system benefits everyone. Says Raymond, "We don't really promote it because it confuses people."

This story was updated May 17.