Tuesday, February 27, 2007


Spam, or Unsolicited Bulk E-mail (UBE) to give it its proper name, is a huge problem. Spamhaus, an organisation that tracks the Internet’s worst spammers, estimated that as of February 2005 over 75% of all e-mail traffic was spam. Of course a large proportion of that never reaches people’s mailboxes because it is filtered out by ISPs and e-mail service providers. But enough of it still gets through to make it more than just a minor irritation for most Internet users. Much of it attempts to persuade recipients to buy pharmaceutical and other products that are often of dubious quality or even illegal. Some e-mails carry viruses, Trojans and other “malware” or attempt to extract personal information by directing users to websites masquerading as official bank websites (“phishing”).

Why is spam a problem? First, because there is so much of it networks can become congested. Time and resources are wasted in blocking and filtering it. It costs individuals and companies time and money to deal with it and productivity is reduced. Because of virus and phishing scares it may also have a deterrent effect on e-commerce. A number of studies have attempted to quantify the economic cost of spam. A study undertaken for the European Commission in 2003 estimated that the connection costs just to receive junk e-mails amounted to around 10 billion euros a year. Leung (2003) says that the annual cost of spam to US corporations is of the order of $8.9 billion. Costs relate not only to the time taken by consumers and businesses to delete unwanted e-mails but also the extra network capacity and file space that is needed to accommodate these unwanted messages, the cost of producing and using mail filters and blocking software and the more regular denial-of-service occurrences and network breakdowns that are caused by spam. In addition some important e-mail messages may be misclassified as spam (“false negatives”) and so never get delivered and read.

The two main approaches to dealing with the problem of spam that have been tried are legal and technological. Both can undoubtedly play a role in combating spam but in my opinion they do not get to the root of the problem which is essentially an economic one. The key fact is that it costs virtually nothing to send an e-mail message (in marginal cost terms) so a spammer can flood the system with unwanted e-mails without having to face the costs that he is imposing on the recipients of these messages. (This is an example of a negative externality – where private and social costs diverge. Spammers are either not aware of the costs that they are imposing or they don’t care about them.) Although there are some costs of maintaining an Internet presence and in creating or buying lists of mail addresses, bulk e-mails can be sent out for next to nothing. And even a very low response rate can then make it profitable. Leung estimates that the response rate to spam is as low as 0.005% - or 50 in every million people. But that is more than enough to generate sufficient revenue to make the operation profitable.

What can be done? A basic suggestion, originally strongly promoted by
Brad Templeton (the founder of the ClariNet company) is that e-mail messages should have to carry an attached electronic stamp (e-stamp). In fact the idea was originally patented by Todd Sunsted in 1999 and is supposed to deter spammers but not bona fide e-mailers. Only a small charge would be needed so it wouldn’t cost much for those who send even hundreds of e-mails a day – but it would start to add up to big money for those who send out millions.

The idea has been examined in a number of academic papers, including Kraut et al (2005) and Loder et al. (2006). As Kraut et al. note “…charging postage for mail causes senders to be more selective and to send fewer messages..” and “..charging a small price to send a message shifts the task of screening messages from recipients who don’t know the content of a message to senders who do.” Loder et al. also refer to this “information asymmetry” and propose an “attention bond” through which recipients can define a price that senders must risk to deliver a message. As many writers on this subject have noted the difficulty in designing a suitable system is in its detail. Much of the discussion has concerned who should collect the revenue and where it should go. For example some schemes suggest that recipients might wish to waive the charge for messages that they wish to receive and to donate charges levied for unwanted messages to charity.

A particularly ingenious version of an e-mail charging system is described by Scott Fahlman (2002). He talks about “interrupt rights” and argues that callers (the scheme can be applied to telephone calls as well as e-mails) who interrupt you in an effort to gain your attention should be forced to pay a price for doing so. Fahlman’s scheme allows each recipient to set their own price, so if an initial price appears not to be screening out enough calls or e-mails it can be raised. If interruptions are no longer free it should make advertisers and marketers think more carefully about targeting their messages. The beauty of Fahlman’s scheme is that it requires no new laws or special e-payment schemes as it can be attached to existing telephone and Internet accounts. And of course recipients can have a “white list” of callers for whom the charge is waived, which can be modified at any time.
Similarly there can be a “black list” which contains the addresses of anyone for whom no price would be high enough for an interruption to be acceptable. In between these extremes different prices could be set, according to how tolerable a message might be for its recipient.

As yet these schemes haven’t really taken off. Perhaps it needs more active promotion from ISPs and e-mail service providers. Or maybe just the backing of a big corporation like Microsoft. Bill Gates is certainly on record as supporting an economic approach to dealing with spam “In the long run, the monetary (method) will be dominant” he is
reported as saying one of the World Economic Forum meetings in Davos. However that was in January 2004 and at that time he also predicted that the spam problem would have been solved by 2006. It is high time that something more practical was put into place to deal with spam pollution!

FAHLMAN, S (2002) Selling Interrupt Rights: A Way to Control Unwanted E-Mail and Telephone Calls. IBM Systems Journal 41(4)759-766
LODER, T, VAN ALSTYNE, M and WALSH, R (2006) An Economic Response to Unsolicited Communication. In Advances in Economic Analysis and Policy, Vol 6 No 1 .
HERMALIN, B E and KATZ, M (2004) Sender or Receiver: Who Should Pay to Exchange an Electronic Message?,
RAND Journal of Economics, Vol. 35, No. 3 (Autumn). Earlier version available online
KRAUT, R E , SUNDER, S, TELANG, R and MORRIS, J (2005) Pricing Electronic Mail to Solve the Problem of Spam.
Human-Computer Interaction Vol 20 pp195-223. .

Labels: , ,


Post a Comment

<< Home